a simpler origin for life - scientific american …researchers use the term "rna world" to...

© MEDIOIMAGES/CORBIS

February 12, 2007 | By Robert Shapiro |

Extraordinary discoveries inspire extraordinary claims. Thus James Watson reported that,immediately after they had uncovered the structure of DNA, Francis Crick "winged into theEagle (pub) to tell everyone within hearing that we had discovered the secret of life." Theirstructure--an elegant double helix--almost merited such enthusiasm. Its proportions permittedinformation storage in a language in which four chemicals, called bases, played the same roleas twenty six letters do in the English language.

Further, the information was stored in two long chains, each of which specified the

contents of its partner. This arrangement suggested a mechanism for reproduction,

that was subsequently illustrated in many biochemistry texts, as well as on a tie that

my wife bought for me at a crafts fair: The two strands of the DNA double helix

parted company. As they did so, new DNA building blocks, called nucleotides, lined

up along the separated strands and linked up. Two double helices now existed in

place of one, each a replica of the original.

The Watson-Crick structure triggered an avalanche of discoveries about the way in

which living cells function today. These insights also stimulated speculations about life's origins. Nobel Laureate H. J. Muller wrote that

the gene material was "living material, the present-day representative of the first life," which Carl Sagan visualized as "a primitive

free-living naked gene situated in a dilute solution of organic matter." In this context, "organic" specifies material containing bound

carbon atoms. Organic chemistry, a subject sometimes feared by pre-medical students, is the chemistry of carbon compounds, both

those present in life and those playing no part in life. Many different definitions of life have been proposed. Muller's remark would be

in accord with what has been called the NASA definition of life: Life is a self-sustained chemical system capable of undergoing

Darwinian evolution.

Richard Dawkins elaborated on this image of the earliest living entity in his book The Selfish Gene: "At some point a particularly

remarkable molecule was formed by accident. We will call it the Replicator. It may not have been the biggest or the most complex

molecule around, but it had the extraordinary property of being able to create copies of itself." When Dawkins wrote these words 30

years ago, DNA was the most likely candidate for this role. As we shall see, several other replicators have now been suggested.

When RNA Ruled the World

Unfortunately, complications soon set in. DNA replication cannot proceed without the assistance of a number of proteins--members of

a family of large molecules that are chemically very different from DNA. Proteins, like DNA, are constructed by linking subunits, amino

acids in this case, together to form a long chain. Cells employ twenty of these building blocks in the proteins that they make, affording a

variety of products capable of performing many different tasks--proteins are the handymen of the living cell. Their most famous

subclass, the enzymes, act as expeditors, speeding up chemical processes that would otherwise take place too slowly to be of use to life.

The above account brings to mind the old riddle: Which came first, the chicken or the egg? DNA holds the recipe for protein

construction. Yet that information cannot be retrieved or copied without the assistance of proteins. Which large molecule, then,

appeared first in getting life started--proteins (the chicken) or DNA (the egg)?

A Simpler Origin for Life - Scientific American http://www.scientificamerican.com/article/a-simpler-origin-for-life/?print...

1 of 7 12/31/2014 3:39 PM

A possible solution appeared when attention shifted to a new champion--RNA. This versatile class of molecule is, like DNA, assembled

of nucleotide building blocks, but plays many roles in our cells. Certain RNAs ferry information from DNA to structures (which

themselves are largely built of other kinds of RNA) that construct proteins. In carrying out its various duties, RNA can take on the form

of a double helix that resembles DNA, or of a folded single strand, much like a protein. In 2006 the Nobel prizes in both chemistry and

medicine were awarded for discoveries concerning the role of RNA in editing and censoring DNA instructions. Warren E. Leary could

write in the New York Times that RNA "is swiftly emerging from the shadows of its better-known cousin DNA."

For many scientists in the origin-of-life field, those shadows had lifted two decades earlier with the discovery of ribozymes, enzyme-like

substances made of RNA. A simple solution to the chicken-and-egg riddle now appeared to fall into place: Life began with the

appearance of the first RNA molecule. In a germinal 1986 article, Nobel Laureate Walter Gilbert of Harvard University wrote in the

journal Nature: "One can contemplate an RNA world, containing only RNA molecules that serve to catalyze the synthesis of themselves.

& The first step of evolution proceeds then by RNA molecules performing the catalytic activities necessary to assemble themselves from

a nucleotide soup." In this vision, the first self-replicating RNA that emerged from non-living matter carried out the functions now

executed by RNA, DNA and proteins.

A number of additional clues seemed to support the idea that RNA appeared before proteins and DNA in the evolution of life. Many

small molecules, called cofactors, play a necessary role in enzyme-catalyzed reactions. These cofactors often carry an attached RNA

nucleotide with no obvious function. These structures have been considered "molecular fossils," relics descended from the time when

RNA alone, without DNA or proteins, ruled the biochemical world. In addition, chemists have been able to synthesize new ribozymes

that display a variety of enzyme-like activities. Many scientists found the idea of an organism that relied on ribozymes, rather than

protein enzymes, very attractive.

The hypothesis that life began with RNA was presented as a likely reality, rather than a speculation, in journals, textbooks and the

media. Yet the clues I have cited only support the weaker conclusion that RNA preceded DNA and proteins; they provide no

information about the origin of life, which may have involved stages prior to the RNA world in which other living entities ruled

supreme. Just the same, and despite the difficulties that I will discuss in the next section, perhaps two-thirds of scientists publishing in

the origin-of life field (as judged by a count of papers published in 2006 in the journal Origins of Life and Evolution of the Biosphere)

still support the idea that life began with the spontaneous formation of RNA or a related self-copying molecule. Confusingly,

researchers use the term "RNA World" to refer to both the strong and the weak claims about RNA's role prior to DNA and proteins.

Here, I will use the term "RNA first" for the strong claim that RNA was involved in the origin of life.

The Soup Kettle is Empty

The attractive features of RNA World prompted Gerald Joyce of the Scripps Research Institute and Leslie Orgel of the Salk Institute to

picture it as "the molecular biologist's dream" within a volume devoted to that topic. They also used the term "the prebiotic chemist's

nightmare" to describe another part of the picture: How did that first self-replicating RNA arise? Enormous obstacles block Gilbert's

picture of the origin of life, sufficient to provoke another Nobelist, Christian De Duve of Rockefeller University, to ask rhetorically, "Did

God make RNA?"

RNA's building blocks, nucleotides, are complex substances as organic molecules go. They each contain a sugar, a phosphate and one of

four nitrogen-containing bases as sub-subunits. Thus, each RNA nucleotide contains 9 or 10 carbon atoms, numerous nitrogen and

oxygen atoms and the phosphate group, all connected in a precise three-dimensional pattern. Many alternative ways exist for making

those connections, yielding thousands of plausible nucleotides that could readily join in place of the standard ones but that are not

represented in RNA. That number is itself dwarfed by the hundreds of thousands to millions of stable organic molecules of similar size

that are not nucleotides.

The RNA nucleotides are familiar to chemists because of their abundance in life and their resulting commercial availability. In a form of

molecular vitalism, some scientists have presumed that nature has an innate tendency to produce life's building blocks preferentially,

rather than the hordes of other molecules that can also be derived from the rules of organic chemistry. This idea drew inspiration from

a well known experiment published in 1953 by Stanley Miller. He applied a spark discharge to a mixture of simple gases that were then

thought to represent the atmosphere of the early Earth. Two amino acids of the set of 20 used to construct proteins were formed in

significant quantities, with others from that set present in small amounts. (A description of the Miller experiment and the chemical

structures of an amino acid and a nucleotide can be found in "The Origin of Life on the Earth," by L. E. Orgel; Scientific American,


2 of 7 12/31/2014 3:39 PM

October 1994.) In addition, more than 80 different amino acids, some present and others absent from living systems, have been

identified as components of the Murchison meteorite, which fell in Australia in 1969. Nature has apparently been generous in providing

a supply of these particular building blocks. By extrapolation of these results, some writers have presumed that all of life's building

could be formed with ease in Miller-type experiments and were present in meteorites and other extraterrestrial bodies. This is not the

case.

A careful examination of the results of the analysis of several meteorites led the scientists who conducted the work to a different

conclusion: inanimate nature has a bias toward the formation of molecules made of fewer rather than greater numbers of carbon atoms,

and thus shows no partiality in favor of creating the building blocks of our kind of life. (When larger carbon-containing molecules are

produced, they tend to be insoluble, hydrogen-poor substances that organic chemists call tars.) I have observed a similar pattern in the

results of many spark discharge experiments.

Amino acids, such as those produced or found in these experiments, are far less complex than nucleotides. Their defining features are

an amino group (a nitrogen and two hydrogens) and a carboxylic acid group (a carbon, two oxygens and a hydrogen) both attached to

the same carbon. The simplest of the 20 used to build natural proteins contains only two carbon atoms. Seventeen of the set contain six

or fewer carbons. The amino acids and other substances that were prominent in the Miller experiment contained two and three carbon

atoms. By contrast, no nucleotides of any kind have been reported as products of spark discharge experiments or in studies of

meteorites, nor have the smaller units (nucleosides) that contain a sugar and base but lack the phosphate.

To rescue the RNA-first concept from this otherwise lethal defect, its advocates have created a discipline called prebiotic synthesis.

They have attempted to show that RNA and its components can be prepared in their laboratories in a sequence of carefully controlled

reactions, normally carried out in water at temperatures observed on Earth. Such a sequence would start usually with compounds of

carbon that had been produced in spark discharge experiments or found in meteorites. The observation of a specific organic chemical

in any quantity (even as part of a complex mixture) in one of the above sources would justify its classification as "prebiotic," a substance

that supposedly had been proved to be present on the early Earth. Once awarded this distinction, the chemical could then be used in

pure form, in any quantity, in another prebiotic reaction. The products of such a reaction would also be considered "prebiotic" and

employed in the next step in the sequence.

The use of reaction sequences of this type (without any reference to the origin of life) has long been an honored practice in the

traditional field of synthetic organic chemistry. My own PhD thesis advisor, Robert B. Woodward, was awarded the Nobel Prize for his

brilliant syntheses of quinine, cholesterol, chlorophyll and many other substances. It mattered little if kilograms of starting material

were required to produce milligrams of product. The point was the demonstration that humans could produce, however inefficiently,

substances found in nature. Unfortunately, neither chemists nor laboratories were present on the early Earth to produce RNA.

I will cite one example of prebiotic synthesis, published in 1995 by Nature and featured in the New York Times. The RNA base cytosine

was prepared in high yield by heating two purified chemicals in a sealed glass tube at 100 degrees Celsius for about a day. One of the

reagents, cyanoacetaldehyde, is a reactive substance capable of combining with a number of common chemicals that may have been

present on the early Earth. These competitors were excluded. An extremely high concentration was needed to coax the other

participant, urea, to react at a sufficient rate for the reaction to succeed. The product, cytosine, can self-destruct by simple reaction with

water. When the urea concentration was lowered, or the reaction allowed to continue too long, any cytosine that was produced was

subsequently destroyed. This destructive reaction had been discovered in my laboratory, as part of my continuing research on

environmental damage to DNA. Our own cells deal with it by maintaining a suite of enzymes that specialize in DNA repair.

The exceptionally high urea concentration was rationalized in the Nature paper by invoking a vision of drying lagoons on the early

Earth. In a published rebuttal, I calculated that a large lagoon would have to be evaporated to the size of a puddle, without loss of its

contents, to achieve that concentration. No such feature exists on Earth today.

The drying lagoon claim is not unique. In a similar spirit, other prebiotic chemists have invoked freezing glacial lakes, mountainside

freshwater ponds, flowing streams, beaches, dry deserts, volcanic aquifers and the entire global ocean (frozen or warm as needed) to

support their requirement that the "nucleotide soup" necessary for RNA synthesis would somehow have come into existence on the

early Earth.

The analogy that comes to mind is that of a golfer, who having played a golf ball through an 18-hole course, then assumed that the ball


3 of 7 12/31/2014 3:39 PM

could also play itself around the course in his absence. He had demonstrated the possibility of the event; it was only necessary to

presume that some combination of natural forces (earthquakes, winds, tornadoes and floods, for example) could produce the same

result, given enough time. No physical law need be broken for spontaneous RNA formation to happen, but the chances against it are so

immense, that the suggestion implies that the non-living world had an innate desire to generate RNA. The majority of origin-of-life

scientists who still support the RNA-first theory either accept this concept (implicitly, if not explicitly) or feel that the immensely

unfavorable odds were simply overcome by good luck.

A Simpler Replicator?

Many chemists, confronted with these difficulties, have fled the RNA-first hypothesis as if it were a building on fire. One group,

however, still captured by the vision of the self-copying molecule, has opted for an exit that leads to similar hazards. In these revised

theories, a simpler replicator arose first and governed life in a "pre-RNA world." Variations have been proposed in which the bases, the

sugar or the entire backbone of RNA have been replaced by simpler substances, more accessible to prebiotic syntheses. Presumably, this

first replicator would also have the catalytic capabilities of RNA. Because no trace of this hypothetical primal replicator and catalyst has

been recognized so far in modern biology, RNA must have completely taken over all of its functions at some point following its

emergence.

Further, the spontaneous appearance of any such replicator without the assistance of a chemist faces implausibilities that dwarf those

involved in the preparation of a mere nucleotide soup. Let us presume that a soup enriched in the building blocks of all of these

proposed replicators has somehow been assembled, under conditions that favor their connection into chains. They would be

accompanied by hordes of defective building blocks, the inclusion of which would ruin the ability of the chain to act as a replicator. The

simplest flawed unit would be a terminator, a component that had only one "arm" available for connection, rather than the two needed

to support further growth of the chain.

There is no reason to presume than an indifferent nature would not combine units at random, producing an immense variety of hybrid

short, terminated chains, rather than the much longer one of uniform backbone geometry needed to support replicator and catalytic

functions. Probability calculations could be made, but I prefer a variation on a much-used analogy. Picture a gorilla (very long arms are

needed) at an immense keyboard connected to a word processor. The keyboard contains not only the symbols used in English and

European languages but also a huge excess drawn from every other known language and all of the symbol sets stored in a typical

computer. The chances for the spontaneous assembly of a replicator in the pool I described above can be compared to those of the

gorilla composing, in English, a coherent recipe for the preparation of chili con carne. With similar considerations in mind Gerald F.

Joyce of the Scripps Research Institute and Leslie Orgel of the Salk Institute concluded that the spontaneous appearance of RNA chains

on the lifeless Earth "would have been a near miracle." I would extend this conclusion to all of the proposed RNA substitutes that I

mentioned above.

Life With Small Molecules

Nobel Laureate Christian de Duve has called for "a rejection of improbabilities so incommensurably high that they can only be called

miracles, phenomena that fall outside the scope of scientific inquiry." DNA, RNA, proteins and other elaborate large molecules must

then be set aside as participants in the origin of life. Inanimate nature provides us with a variety of mixtures of small molecules, whose

behavior is governed by scientific laws, rather than by human intervention.

Fortunately, an alternative group of theories that can employ these materials has existed for decades. The theories employ a

thermodynamic rather than a genetic definition of life, under a scheme put forth by Carl Sagan in the Encyclopedia Britannica: A

localized region which increases in order (decreases in entropy) through cycles driven by an energy flow would be considered alive.

This small-molecule approach is rooted in the ideas of the Soviet biologist Alexander Oparin, and current notable spokesmen include

de Duve, Freeman Dyson of the Institute for Advanced Study, Stuart Kauffman of the Santa Fe Institute, Doron Lancet of the Weizmann

Institute, Harold Morowitz of George Mason University and the independent researcher G¿nter W¿chtersh¿user. I estimate that about a

third of the chemists involved in the study of the origin of life subscribe to theories based on this idea. Origin-of-life proposals of this

type differ in specific details; here I will try to list five common requirements (and add some ideas of my own).

(1) A boundary is needed to separate life from non-life. Life is distinguished by its great degree of organization, yet the second

law of thermodynamics requires that the universe move in a direction in which disorder, or entropy, increases. A loophole, however,


4 of 7 12/31/2014 3:39 PM

allows entropy to decrease in a limited area, provided that a greater increase occurs outside the area. When living cells grow and

multiply, they convert chemical energy or radiation to heat at the same time. The released heat increases the entropy of the

environment, compensating for the decrease in living systems. The boundary maintains this division of the world into pockets of life

and the nonliving environment in which they must sustain themselves.

Today, sophisticated double-layered cell membranes, made of chemicals classified as lipids, separate living cells from their

environment. When life began, some natural feature probably served the same purpose. David W. Deamer of the University of

California, Santa Cruz, has observed membrane-like structures in meteorites. Other proposals have suggested natural boundaries not

used by life today, such as iron sulfide membranes, mineral surfaces (in which electrostatic interactions segregate selected molecules

from their environment), small ponds and aerosols.

(2) An energy source is needed to drive the organization process. We consume carbohydrates and fats, and combine them

with oxygen that we inhale, to keep ourselves alive. Microorganisms are more versatile, and can use minerals in place of the food or the

oxygen. In either case, the transformations that are involved are called redox reactions. They involve the transfer of electrons from an

electron rich (or reduced) substance to an electron poor (or oxidized) one. Plants can capture solar energy directly, and adapt it for the

functions of life. Other forms of energy are used by cells in specialized circumstances--for example, differences in acidity on opposite

sides of a membrane. Yet others, such as radioactivity and abrupt temperature differences, might be used by life elsewhere in the

universe. Here I will consider redox reactions as the energy source.

(3) A coupling mechanism must link the release of energy to the organization process that produces and sustains life.

The release of energy does not necessarily produce a useful result. Chemical energy is released when gasoline is burned within the

cylinders of my automobile, but the vehicle will not move unless that energy is used to turn the wheels. A mechanical connection, or

coupling, is required. Each day, in our own cells, each of us degrades pounds of a nucleotide called ATP. The energy released by this

favorable reaction serves to drive processes that are less favorable but necessary for our biochemistry. Linkage is achieved when the

reactions share a common intermediate, and the process is speeded up by the intervention of an enzyme. One assumption of the small-

molecule approach is that coupled reactions and primitive catalysts sufficient to get life started exist in nature.

(4) A chemical network must be formed, to permit adaptation and evolution. We come now to the heart of the matter.

Imagine for example that an energetically favorable redox reaction of a naturally-occurring mineral is linked to the conversion of an

organic chemical A to another one B within a compartment. The favorable, energy releasing, entropy-increasing reaction of the mineral

drives the A-to-B transformation. I call this key transformation a driver reaction, for it serves as the engine that mobilizes the

organization process. If B simply reconverts back to A or escapes from the compartment, we would not be on a path that leads to

increased organization. By contrast, if a multi-step chemical pathway--say, B to C to D to A--reconverts B to A, then the steps in that

circular process (or cycle) would be favored because they replenish the supply of A, allowing the continuing discharge of energy by the

mineral reaction.

If we visualize the cycle as a circular railway line, the energy source keeps the trains traveling around it one way. Each station may also

be the hub for a number of branch lines, such as one connecting station D to another station, E. Trains could travel in either direction

along that branch, depleting or augmenting the cycle's traffic. Thanks to the continual depletion of A, however, material is drawn from

D to A. The resulting depletion of D in turn tends to draw material from E to D. In this way, material is "pulled" along the branch lines

into the central cycle, maximizing the energy release that accompanies the driver reaction.

The cycle could also adapt to changing circumstances. As a child, I was fascinated by the way in which water, released from a leaky

hydrant, would find a path downhill to the nearest sewer. If falling leaves or dropped refuse blocked that path, the water would back up

until another route was found around the obstacle. In the same way, if a change in the acidity or in some other environmental

circumstance should hinder a step in the pathway from B to A, material would back up until another route was found. Additional

changes of this type would convert the original cycle into a network. This trial-and-error exploration of the chemical "landscape" might

also turn up compounds that could catalyze important steps in the cycle, increasing the efficiency with which the network utilized the

energy source.

(5) The network must grow and reproduce. To survive and grow, the network must gain material at a rate that compensates for

the paths that remove it. Diffusion of network materials out of the compartment into the external world is favored by entropy and will

occur to some extent, especially at the start of life when the boundary is a crude one established by the environment rather than one of


5 of 7 12/31/2014 3:39 PM

the highly effective cell membranes available today after billions of years of evolution. Some side reactions may produce gases, which

escape, or form tars, which will drop out of solution. If these processes together should exceed the rate at which the network gains

material, then it would be extinguished. Exhaustion of the external fuel would have the same effect. We can imagine, on the early Earth,

a situation where many startups of this type occur, involving many alternative driver reactions and external energy sources. Finally, a

particularly hardy one would take root and sustain itself.

A system of reproduction must eventually develop. If our network is housed in a lipid membrane, then physical forces may split it, after

it has grown enough. (Freeman Dyson has described such a system as a "garbage-bag world" in contrast to the "neat and beautiful

scene" of the RNA world.) A system that functions in a compartment within a mineral may overflow into adjacent compartments.

Whatever the mechanism may be, this dispersal into separated units protects the system from total extinction by a localized destructive

event. Once independent units were established, they could evolve in different ways and compete with one another for raw materials;

we would have made the transition from life that emerges from nonliving matter through the action of an available energy source to life

that adapts to its environment by Darwinian evolution.

Changing the Paradigm

Systems of the type I have described usually have been classified under the heading "metabolism first," which implies that they do not

contain a mechanism for heredity. In other words, they contain no obvious molecule or structure that allows the information stored in

them (their heredity) to be duplicated and passed on to their descendants. However a collection of small items holds the same

information as a list that describes the items. For example, my wife gives me a shopping list for the supermarket; the collection of

grocery items that I return with contains the same information as the list. Doron Lancet has given the name "compositional genome" to

heredity stored in small molecules, rather than a list such as DNA or RNA.

The small molecule approach to the origin of life makes several demands upon nature (a compartment, an external energy supply, a

driver reaction coupled to that supply, and the existence of a chemical network that contains that reaction). These requirements are

general in nature, however, and are immensely more probable than the elaborate multi-step pathways needed to form a molecule that

can function as a replicator.

Over the years, many theoretical papers have advanced particular metabolism first schemes, but relatively little experimental work has

been presented in support of them. In those cases where experiments have been published, they have usually served to demonstrate the

plausibility of individual steps in a proposed cycle. The greatest amount of new data has perhaps come from G¿nter W¿chtersh¿user

and his colleagues at the Technische Universit¿t M¿nchen. They have demonstrated portions of a cycle involving the combination and

separation of amino acids, in the presence of metal sulfide catalysts. The energetic driving force for the transformations is supplied by

the oxidation of carbon monoxide to carbon dioxide. They have not yet demonstrated the operation of a complete cycle or its ability to

sustain itself and undergo further evolution. A "smoking gun" experiment displaying those three features is needed to establish the

validity of the small molecule approach.

The principal initial task is the identification of candidate driver reactions--small molecule transformations (A to B in the example

before) that are coupled to an abundant external energy source (such as the oxidation of carbon monoxide or a mineral). Once a

plausible driver reaction has been identified, there should be no need to specify the rest of the system in advance. The selected

components (including the energy source) plus a mixture of other small molecules normally produced by natural processes (and likely

to have been abundant on the early Earth) could be combined in a suitable reaction vessel. If an evolving network were established, we

would expect the concentration of the participants in the network to increase and alter with time. New catalysts that increased the rate

of key reactions might appear, while irrelevant materials would decrease in quantity. The reactor would need an input device to allow

replenishment of the energy supply and raw materials, and an outlet to permit the removal of waste products and chemicals that were

not part of the network.

In such experiments, failures would be easily identified. The energy might be dissipated without producing any significant changes in

the concentrations of the other chemicals or the chemicals might simply be converted to a tar, which would clog the apparatus. A

success might demonstrate the initial steps on the road to life. These steps need not duplicate those that took place on the early Earth.

It is more important that the general principle be demonstrated and made available for further investigation. Many potential paths to

life may exist, with the choice dictated by the local environment.


6 of 7 12/31/2014 3:39 PM

An understanding of the initial steps leading to life would not reveal the specific events that led to the familiar DNA-RNA-protein-based

organisms of today. However, because we know that evolution does not anticipate future events, we can presume that nucleotides first

appeared in metabolism to serve some other purpose, perhaps as catalysts or as containers for the storage of chemical energy (the

nucleotide ATP still serves this function today). Some chance event or circumstance may have led to the connection of nucleotides to

form RNA. The most obvious function of RNA today is to serve as a structural element that assists in the formation of bonds between

amino acids in the synthesis of proteins. The first RNAs may have served the same purpose, but without any preference for specific

amino acids. Many further steps in evolution would be needed to "invent" the elaborate mechanisms for replication and specific protein

synthesis that we observe in life today.

If the general small-molecule paradigm were confirmed, then our expectations of the place of life in the universe would change. A

highly implausible start for life, as in the RNA-first scenario, implies a universe in which we are alone. In the words of the late Jacques

Monod, "The universe was not pregnant with life nor the biosphere with man. Our number came up in the Monte Carlo game." The

small-molecule alternative, however, is in harmony with the views of biologist Stuart Kauffman: "If this is all true, life is vastly more

probable than we have supposed. Not only are we at home in the universe, but we are far more likely to share it with unknown

companions."

ROBERT SHAPIRO is professor emeritus of chemistry and senior research scientist at New York University. He is author or co-author

of over 125 publications, primarily in the area of DNA chemistry. In particular, he and his co-workers have studied the ways in which

environmental chemicals can damage our hereditary material, causing changes that can lead to mutations and cancer. In 2004, he was

awarded the Trotter Prize in Information, Complexity and Inference. Shapiro has written four books for the general public: Life Beyond

Earth (with Gerald Feinberg); Origins, a Skeptic's Guide to the Creation of Life on Earth; The Human Blueprint (on the effort to read

the human genome); and Planetary Dreams (on the search for life in our Solar System). When he is not involved in research, lecturing

or writing, he enjoys running, hiking, wine-tastings, theater and travel. He is married and has a 35-year-old son.


7 of 7 12/31/2014 3:39 PM

Insights & Perspectives

The RNA dreamtime

Modern cells feature proteins that might have supported a prebiotic polypeptide world but

nothing indicates that RNA world ever was

Charles G. Kurland

Modern cells present no signs of a putative prebiotic RNA world. However,

RNA coding is not a sine qua non for the accumulation of catalytic polypep-

tides. Thus, cellular proteins spontaneously fold into active structures that

are resistant to proteolysis. The law of mass action suggests that binding

domains are stabilized by specific interactions with their substrates. Random

polypeptide synthesis in a prebiotic world has the potential to initially pro-

duce only a very small fraction of polypeptides that can fold spontaneously

into catalytic domains. However, that fraction can be enriched by proteolytic

activities that destroy the unfolded polypeptides and regenerate amino acids

that can be recycled into polypeptides. In this open system scenario the

stable domains that accumulate and the chemical environment in which they

are accumulated are linked through self coding of polypeptide structure.

Such open polypeptide systems may have been the precursors to the cellu-

lar ribonucleoprotein (RNP) world that evolved subsequently.

Keywords:.domain selection; non-ribosomal peptidyl transferase; polypeptides;

proteolysis; ribozymes

Introduction

The standard model for the origin of lifeimagines that the first replication, trans-lation, and transcription systems weresupported by RNA without the interven-tion of proteins [1]. The impulses for thisconjecture were first, the discoveries ofintrons and exons in eukaryote mRNAsand second, the self splicing of someintron sequences in ribosomal RNA [2].

Here, a core assertion is that primordialmini-RNAs corresponding to the originalexons encoded the first polypeptides [1].In this scenario ‘‘RNA molecules beganto synthesize proteins, first by develop-ing RNA adapter molecules that can bindactivated amino acids and then byarranging them according to an RNAtemplate using other molecules such asthe RNA core of the ribosome’’. Twenty-four years later this dazzling speculation

has been reduced by ritual abuse tosomething like a creationist mantra.Hence, the title, borrowed from Collinset al. [3], alludes to an oral tradition oforigins passed on by the first Australians.Finally, the support for a prebiotic RNAworld consists solely of ingenious piece-meal chemical simulations in vitro thatwere obtained by chemists at great costand effort over a 20-year period [4]. Suchchemical simulations are accepted ashard evidence by RNA worlders, but intruth, they do not constitute proper evol-utionary evidence.

It might have been useful earlier onto address questions such as: Does theroutine identification of ribonucleopro-teins (RNPs) as remnants of a prebioticRNA world [4] support Gilbert’s conjec-ture [1] or do they beg the question? Ispartial simulation in vitro by a ribozymerunning at one-millionth the rate nor-mally catalyzed by a protein [5, 6] to betaken as evidence for the prebiotic pre-cedence of the ribozyme activity?Though there are impressive syntheticribozymes with convincing performancecharacteristics [7], why are thereno examples of naturally occurring,protein-free ribozymes to link the postu-lated protein-free RNA world to themodern cellular world [8, 3]?

The secret of the cage

RNase P is an RNP that mediates thematuration of transfer RNAs [9].Normally, RNase P has both RNA as wellas protein components. But recently, anRNA-free variant was discovered inhuman mitochondria and shown to

DOI 10.1002/bies.201000058

Department of Microbial Ecology, University ofLund, Solvegatan, Lund, Sweden

Corresponding author:Charles G. KurlandE-mail: [email protected]

Abbreviations:FSF, fold superfamily; NPTase, non-ribosomalpeptidyl transferase; PTC, peptidyl transferasecenter; RNP, ribonucleoprotein.

866 www.bioessays-journal.com Bioessays 32: 866–871,� 2010 WILEY Periodicals, Inc.

Hyp

oth

eses

mature tRNA precursors normally [10].If RNase P function can be supported byprotein alone, it is conceivable that suchprotein functions have participated inprebiotic systems as well. So, whichcame first, a protein or an RNP versionof the enzyme? In fact, a recent phylo-genetic survey suggests that an ances-tral protein version may have precededthe RNP version of RNase P [11].

Another instructive RNP is the bac-terial ribosome. Studies of these RNPsreveal the tension between routine bio-chemical observations and the expec-tations of RNA worlders, who refuse toaccept the simplest interpretations oftheir experiments. For example, gentledisruption of 50S subunits yields a com-pacted 23S RNA associated with a verysmall core of ribosomal protein and thisRNP supports peptidyl transferaseactivity with model substrates [12, 13].Though these are straightforwardresults, the authors are reluctant to con-cede the potential involvement of ribo-somal protein in the peptidyl transferaseactivity. Instead, they speculate that thecore of proteins in their active 23S RNApreparations is trapped in an imaginaryRNA cage; that is to say, the proteins arepassive prisoners and not active partici-pants [12, 13]. Noller [13] further com-ments that ‘‘true ‘ribocentrics’ willsimply view the latest aspect of the ribo-somal puzzle as a worthy challenge,whose solution promises to reveal oneof nature’s most ancient biologicalsecrets’’.

An attempt to simulate the peptidyltransferase center (PTC) with the aid of asynthetic ribozyme was equally obtuse.Variants of tailored polynucleotides thatcould simulate peptidyl transferaseactivity were synthesized and selectedin vitro [6]. The fastest variant of thetailored ribozymes had a kcat corre-sponding to 0.05 minute�1 per peptide,which was compared with an elongationrate of 15–20 seconds�1 for Escherichiacoli ribosomes [6]. In this comparisonthe kcat was at least four orders of mag-nitude smaller for the ribozyme than forthe ribosome. However, the kcat of E. coliribosomes in the peptidyl transferasereaction is closer to 100 seconds�1

[14], which indicates that the ribozymeis five orders of magnitude slower thanthe bacterial ribosome. Nevertheless,Zhang and Cech [6] suggest that the‘‘peptidyl transferase reaction of the

selected ribozymes is ‘fundamentally’similar to that carried out by the ribo-some’’. Fundamentally similar?

The first crystallographic reports ofthe structure of archaeal 50S ribosomalsubunits described a roughly 20 A-diameter protein-free RNA domain thatwas identified as the PTC [15, 16]. Theseobservations contrasted decades of bio-chemistry that had identified proteincontributions to the bacterial PTC [17].

So, a fatwa was issued to clarify theX-ray revelations: ‘‘From this structurethey deduced . . . that RNA componentsof the large subunit accomplish the keypeptidyl transferase reaction . . . Thus,ribosomal RNA (rRNA) does not exist asa framework to organize catalyticproteins. Instead, the proteins are thestructural units and they help toorganize key ribozyme (catalytic RNA). . . elements, an idea long championedby Harry Noller, Carl Woese . . ., andothers’’ [18].

In unambiguous contrast to thisclaim, Zimmerman and collaborators[17] showed that deletion of E. coli ribo-somal protein L27, long thought to bepart of the bacterial PTC, severelydepresses bacterial growth rates, whichare restored when the protein isexpressed from a plasmid. Likewise,deletion of one to three N-terminal aminoacids of L27 had a similar debilitatingeffect on growth rates as well as on pep-tidyl transferase reaction rates in vitro.Most informative was their finding thatdeletion of the three N-terminal aminoacids strongly inhibits the labeling of L27by photo-activated tRNA at the P site.The experiments of Maguire et al. [17]rather clearly confirm a substantial chainof biochemical experiments initiated in1973 that consistently implicated L27 as aclose neighbor and probable participantin the peptidyl transferase reaction ofbacterial ribosomes.

The identification by Maguire et al.[17] of L27 as an essential part of the PTCin E. coli ribosomes was confirmed by ahigh-resolution structure of 70S activeribosomes with tRNA-filled A site and Psite that revealed two proteins interact-ing directly with tRNA at the putativePTC [19]. One of these is L27 and theother is L16, which was also previouslyimplicated in peptidyl transferaseactivity [17]. Thus, the universalprotein-free ribozyme at the heart ofthe ribosome is history.

Does RNA replace protein?

Comparisons with the findings fromarchaeal subunits [15, 16] are sugges-tively complicated by the apparentabsence of an L27 homolog from thearchaeal 50S ribosomal subunit.According to Voorhees et al. [19] theprojected orientation of L16 in the arch-aeal 50S subunits suggests that, like itshomolog in the E. coli ribosome, it maybe interacting with the elbow of thetRNA in the archaeal A site. In fact itwould be highly instructive if, aftermore stringent structural determi-nations are made, archaeal ribosomeswere indeed found to exploit aprotein-free RNA peptidyl transferasein a high-resolution structure for active70S particles. In that case, the PTCwould be just another example of anRNP featuring an interchangeability ofRNA and protein functions as in theRNA-free RNase P [10]. Indeed a crypticevolutionary trade-off between RNA andprotein may account for the observeddifferences between some archaealand bacterial PTCs.

The growth efficiency of a cellularprocess such as translation or transcrip-tion can be measured by its rate normal-ized to the molecular mass that isrequired to carry out that process [20].No biosynthetic cycle is as expensive to aprokaryote cell as translation becausenothing else involves as large an invest-ment in macromolecular equipment.This means that the biosynthetic costof RNA in terms of growth efficiency isroughly one tenth that for a polypeptidechain. In this context of growth effi-ciency, the observed differences betweensome bacterial and archaeal 50S subu-nits are consistent with the notion thatthe evolution of the archaeal translationapparatus involved a trade-off betweenthe costs of making that equipment andthe efficiency with which it works.

For cells such as the bacteriumE. coli, the growth optimization favorshigh rates of function in relatively richmedia. That is to say, a fast, protein-richPTC is an acceptable ribosomal designfor optimal growth under relatively gen-erous conditions [20, 21]. However, forcells such as some archaea the optionsmay be different because they areadapted to growth under conditions ofenergy stress [22]. Here, ‘‘cheap’’ struc-tural solutions for ribosomes are at a

..... Insights & Perspectives C. G. Kurland

Bioessays 32: 866–871,� 2010 WILEY Periodicals, Inc. 867

Hypotheses

premium even if they might come at thecost of a somewhat slower rate of func-tion, as expected for ribozymes. Thepoint is that archaea growing underconditions of energy stress are con-strained to produce amino acids rela-tively slowly with the consequencethat optimal translation rates might becorrespondingly slow. Here, substi-tution of costly protein by less costlyRNAwould be a favored design strategy.Indeed, comparative data suggest thatthe evolution of the archaea involves aselective loss of proteins from ribo-somes, a loss that is most striking atthe crown of the archaeal ribosometree [23].

Valentine’s [22] insight into theadaptive specializations of the archaeamay account for the putative protein-free PTC of some archaeal ribosomes,and more generally, it may explainwhy there are two prokaryote domainsthat are descendents of the eukaryoteancestor: one for rich environmentsand one for more metabolically chal-lenging circumstances. Since the eukar-yotes are the ancestral lineage fromwhich the divergence of archaea andbacteria are thought to have been drivenby reductive pressure [24–27], relaxedreductive pressure would allow eukar-yote ribosomes to remain more protein-aceous than prokaryote ribosomes.

Accordingly, one prediction of thisscenario is that the PTC of eukaryoteribosomes will turn out to be moreprotein-rich than those of prokaryotes.Another is that the rates of protein-depleted archaeal ribosomes undercomparable conditions may prove tobe slower than those of bacterial ribo-somes. Astonishingly, it is currentlyimpossible to locate in the literatureribosomal rates of translation alongwith growth rates with which to makemeaningful comparisons.

Ribosomes areribonucleoproteinparticles, period!

An evolutionary trade-off between RNAand protein for the adaptations of ribo-somes to the realities of growth con-straints is consistent with an earlierview of ribosome structure. Voorheeset al. [19] note that even in E. coli ribo-somes the PTC seems to be rich in RNA.

This was not too surprising since ‘‘every-where’’ in the bacterial ribosome is rich inRNA. Prior to the revelation of RNAworld, an emergent idea was that riboso-mal RNA and proteins are not segregatedbut are intermixed with cooperative clus-ters of proteins organized around specificRNA domains [28]. Thiswas a data-driveninterpretation based on biochemicalstudies of ribosomes modified bymutations, antibiotics, cross-linkingagents, site-specific chemical agents,and partial assembly in vitro. TheE. coli ribosome along with all other ribo-somes was seen as a RNP particle, not asan RNA particle in protein drag.

Here, the self assembly of ribosomesgenerates mixed neighborhoods of RNAand proteins that provide the workingsurfaces for the translating ribosome[28]. The fact that a protein and anRNA domain cooperate in the assemblyprocess would not preclude either oneas a potential ligand for intermediates intranslation. In effect, the strict divisionof labor for ribosomal RNA andprotein proclaimed by Cech [18] is notnecessarily respected by ribosomes.Amino acyl-tRNAs, protein factors,and mRNAs are all relatively huge sub-strates. Accordingly, the binding ofthese substrates to ribosomal sites andtheir movements during translationspan correspondingly large RNPdomains [28]. This data-driven, low-resolution model has been well substan-tiated by high-resolution structural datafor the components of bacterial ribo-somes. For example, all proteins withthe exception of the oligomeric proteinL7/L12 are directly bound to ribosomalRNA [29]. Likewise, participation of RNAand protein at functional sites is con-firmed by high-resolution structures for70S ribosomes complexed with A-siteand P-site tRNAs [19].

Polypeptide world

Gilbert’s ‘‘big bang’’ scenario [1] can bereplaced by a data-driven, gradualistscheme in which a prebiotic polypeptideworld evolved into amodern RNPworld.Intelligent discussions of a putative pol-ypeptide world are found in Cairns-Smith [30], Kaufman [31], and Egel[32]. Here, the principle novelty is achemical scheme in which randomlygenerated, catalytic polypeptides may

have been selected through a proteolyticmechanism that enriches the popu-lation of polypeptides with biologicallyrelevant activities without the interven-tion of coding by RNA.

Highly relevant to this enrichmentscheme are the workings of a ubiquitouscellular catabolic pathway that pre-serves the stability of high-density cel-lular proteomes by destroying proteinsthat present aberrant sequences [33–35].Modern proteins are made up of one toseveral compact or folded domains (e.g.fold superfamilies or FSFs) along withterminal as well as interspersed linkersequences. The provision of robust fold-ing pathways for polypeptides synthes-ized on ribosomes is part of normalsequence selection in evolution [36].The result is that linkers are bound atsurfaces of domains or to other macro-molecules so that they, together withthe self-organizing domains, are pro-tected from the depredations of ubiqui-tous proteolytic ‘‘machines’’ suchas proteasomes [33–35]. In general,degrading enzymes require a sequencethat is less than 10 A in diameter to beaccommodated within the proteolyticsite [37]. Clearly, an amino acidsequence that is organized into a com-pact domain or one that is stably boundto ligands such as other domains, lipidmembranes, or chemical substrates isnot likely to pass through the 10 A gate-way leading to proteolysis.

The presentation of proteolyticallyaccessible amino acid sequences eitheras unbound linkers or as unfoldeddomains can result frommutation, trans-lation errors, and chemical modification,induced conformational rearrange-ments, or failure to bind a ligand [33–36]. Systematic destruction of aberrantproteins by proteasomes and their homo-logs rids cellular proteomes of potentialseeds for aggregation and precipitation,which otherwise would be lethal to cells.Indeed, the maintenance of the highprotein densities characteristic of all cel-lular proteomes has had a profoundinfluence on the evolution of cells[38, 26]. Human degenerative diseasesthat arise from mutations affectingprotein folding or from defective protea-some function underscore the impact ofproteolytic surveillance systems [39–42].The metabolic consequence of suchproteolytic surveillance is that proteindegradation is a significant catabolic

C. G. Kurland Insights & Perspectives.....

868 Bioessays 32: 866–871,� 2010 WILEY Periodicals, Inc.

Hypotheses

flow in healthy cells. As much as thirtypercent of newly synthesized proteinsare destroyed by proteolysis [43], whilea steady-state background of proteinturnover of about two percent per houris observed in growing cells [44].

This ubiquitous catabolic flow wasintroduced to illustrate a particularlyinformative dimension of protein evol-ution: this is that mutant variantspresenting unfolded domains or unpro-tected strings would tend to be culledfrom populations as lethal alleles [45].For example, even an allele that isexpressed with a fully active catalyticsite in an unstable fold might be a lethalallele if its substrate does not protect thedomain from destruction. Accordingly,protein sequence evolution is not a ran-dom walk through amino acid sequencespace. It is canalized through modularsequences that are expressed as self-organized compacted domains and pro-tected linker sequences, which are thesole survivors of selection by proteolyticmachines [45]. Indeed, such modulardomains may correspond to the prod-ucts of exons though it must be said thatprotein chemists insist that the bound-aries of exons and domains are not iden-tical [46]. Perhaps a measure ofsequence drift over time may accountfor these boundary discrepancies.

Numerous enzymes that synthesizeoligopeptides without the assistance ofan mRNA or the rest of the moderntranslation apparatus have beendescribed in all three superkingdoms[47–49]. Their products are short pepti-des with activities as diverse as anti-biotics as well as neural transmittersand that ubiquity itself speaks for theirancient origins. Though these non-ribo-somal peptidyl transferases (NPTases)produce oligopeptides with definedamino acid sequences, more primitiveancestral enzymatic activities lackingsuch amino acid specificity can beexpected to have arisen in stochasticpopulations of polypeptides assembledby geochemical mechanisms [30–32].So, the second tier of prebiotic polypep-tides may have been composed of ran-dom sequences produced by NPTases.In this second phase, NPTases would beable to autocatalytically increase therates with which random polypeptidesequences could accumulate.

Among sufficiently large popu-lations of random amino acid

sequences, some small fraction wouldbe expected to fold spontaneously intostable active domains with functionssuch as the NPTases and proteolyticenzymes as well as any number of‘‘metabolic’’ activities. This expectationis supported by the observations ofKeefe and Szostak [50], who recoveredbiological activities from relativelysmall populations of random amino acidsequences polymerized in vitro. Thus,stochastic populations of polypeptidescontaining NPTases and proteolyticactivities etc. could generate a dynamicsituation in which random polypeptidesare continuously synthesized, but mostof these would be recycled by proteasesthat regenerate the amino acids. Thesmall core of polypeptides that is resist-ant to proteolysis would be, by hypoth-esis, that which could spontaneouslyfold into stable compacted domains,which are resistant to enzymatic attack.These diverse domains, stable toproteolysis, would be enriched for acorresponding diversity of catalyticactivities. This proposal can be testedin systems such as those described byKeefe and Szostak [50] to determinewhether exposure of randomly synthes-ized polypeptides to proteolysisincreases the specific activities of thepolypeptides for assayable functions.

Finally, it is reasonable to expect thatfolding into active, proteolysis-resistantdomains is facilitated by the binding ofcofactors and/or substrates specificto individual catalytic polypeptides.Here, the specific binding of small mol-ecules from the geochemical system totheir cognate polypeptides would tendthrough the law of mass action to trapthe polypeptides in a folded state that isresistant to proteolysis. In this way, pre-biotic geochemistry may have selectedpolypeptides that mediate cycles ofmetabolic intermediates. The predictedeffect of substrates on the proteolyticstabilities of random polypeptides canalso be studied in vitro as above.

Three steps to cells

In fact, a prebiotic polypeptide scenariomight account for the origins of theribonucleotides that accumulated priorto the debut of RNA. There has been atortured history of attempts to generateribonucleotides in the laboratory as

precursors required for the RNAs ofa putative prebiotic world [51, 52].Recently, an ingenious approach hassucceeded to synthesize pyrimidineribonucleotides in a reaction that mightconceivably have been supported byvolcanic outgassing [52]. However, thereare still questions about the feasibility ofthis reaction scheme in a prebioticenvironment and it still remains toaccount for the purine ribonucleotides[51, 52]. Finally, the scheme of Powneret al. [52] is not the textbook pathway forcellular biosynthesis of ribonucleotidesfrom three amino acids. For thesereasons it is still worth considering thealternative in which the ribonucleotideswere accumulated prebiotically accord-ing to the textbook schemes that mod-ern polypeptides follow.

Several innovations are required toenable a transition from the prebioticworld of polypeptides to biological sys-tems. One is the introduction of anmRNA analog to encode proteins, andassociated with that an analog of theaminoacyl-tRNA adapter to translatemRNAs, as in Gilbert’s proposal [1].However, the adapter initially neednot have been a tRNA because the abil-ity to recognize and to bind nucleotidetriplets in mRNA is not unique to RNA.Proteins in the form of release factorscan bind and recognize triplet codonswith a discriminatory capacity muchgreater than that of aminoacyl-tRNAs[53]. Likewise, proteins in the form ofaminoacyl-tRNA synthases recognizesubtle side chain differences of aminoacids and match these with cognatetRNA structures. Without this polypep-tide function the adapter hypothesiswould be just another RNA worldfantasy.

So, whatever the disadvantages maybe, a primitive translation machine mayhave exploited proteins for all functionsexcept that of mRNA. That is to say, theevolution and tuning of the translationmachinery may have been driven by theintroduction and progressive expansionof RNA functions at the expense ofprotein. Such a tendency is consistentwith the notion that the large protein-rich eukaryote ribosome is ancestral tothe more economic archaeal ribosomesthat have evolved with reduced proteincomplements under stringent reductivepressures ([24–27], Wang et al., inpreparation).



Hypotheses

A second elementary innovation wouldentail the ability to copy mRNA fromcomplementary polynucleotides. Aconservative guess is that this inno-vation arose when polypeptide-depen-dent random synthesis of polyribo-nucleotides was transformed into anenzymatic copy mechanism. Initially,it is simplest to imagine that makingmRNA copies as transcripts and repli-cating them as genomes were one andthe same function. At some point thesetwo functions were separated as inmodern cells.

The third innovation, which isessential for the evolution of geneticallydetermined sequences, is the creation ofa link between individual genetic deter-minants and their products to enableselection of competitive characters.Evolution has chosen cellular bound-aries to provide this link. As a result,the composite features of a cell’s pro-teome can be selected by their inte-grated influence on, for example, acell’s growth phenotype. Without sucha link, selection is impossible.Obviously, selection was for this veryreason impossible in the original formu-lation of RNA world [1].

There are at least two ways to thinkabout the origins of cellular bound-aries. One is to imagine that short poly-peptides with amphiphilic characterthat could mimic lipid moleculesformed the first membranous bound-aries as for example in nanotubulesmade in laboratories [32, 54, 55]. Theother is to employ lipids synthesized bypolypeptides to spontaneously formvesicles. In appropriate solutions lip-ids associate and behave astonishinglylike cellular membranes in modelexperiments [56–58]. Of course somecombination of both lipids and amphi-philic peptides may be more relevant.Since the lipids as well as lipid-likepeptides are potentially products ofpolypeptide enzymes, their emergencein the prebiotic world would not beexceptional.

Since each of these three innovationsis in the present view a spontaneousexpression of protein chemistry, it isassumed that the order of their appear-ances was random. However, all threeinnovations would have to cometogether to enable the transition fromprebiotic to biological systems.

Conclusions

In the 1970s I attended an EMBOWorkshop at which a French philoso-pher of science asserted thatMolecular Biology was merely ‘‘roman-tic idealism’’. I of course indignantlyrejected that comment out of hand.But 30 years later while trying to under-stand the explanatory power of Gilbert’s‘‘big bang’’ theory [1] for the origin of lifeI drifted back to that comment aboutromantic idealism. And, I began tounderstand why RNA world did notneed to explain anything in order tobe attractive to nearly all molecularbiologists.

RNA world is an expression of theinfatuation of molecular biologists withbase pairing in nucleic acids played outin a one-dimensional space with noreference to time or energy: ‘‘DNAmakes RNA makes protein’’ [59]. Thisis not chemistry. It is genetics. And,when true believers apply their geneticdogma to studies of chemical mechan-ism, the result is ‘‘the secret of the cage’’and a five order of magnitude kineticdiscrepancy described as a ‘‘fundamen-tal’’ similarity [12, 13, 6].

The positive side of this infatuationhas been the development of robustgenomics, especially bioinformatics.But RNA worlders are not likely to findmuch comfort in genome sequences. Avery recent phylogenomic study byCaetano-Anolles et al. [60] based onhundreds of fully sequenced genomeshas revealed a time line for cellular evol-ution in which protein domains (FSFs)that support conventional metabolicpathways precede the debut of thenucleic acids as well as the proteindomains associated with nucleic acidsin gene expression. This cellular timeline for the gradual elaboration ofRNA functions [60] is not inconsistentwith the thesis that the prebiotic worldwas a polypeptide world.

Protein folding into compactdomains is a kind of self coding.However, the protein code is not asimple iterative code that lends itselfto a copy mechanism. That is to say, itwould be difficult to build geneticsaround polypeptide interactions alone.On the other hand, self folding andproteolytic editing might be just goodenough to create a prebiotic chemical

platform from which a cellular geneticsystem might take off.

If my own initial reactions are any-thing to go by, RNA world or ‘‘RNAmakes RNA makes protein’’ has imme-diacy for molecular geneticists that islacking in the present scheme for a pre-biotic polypeptide world. Thus, theappeal of a polypeptide fold editingscheme rests on some familiarity withthe contours of protein structure as wellas with translational editing in moderncells. Regardless of its spontaneousappeal, I suggest that RNAworld shouldnow take its place on the shelf of ‘‘niceideas’’ along with Aristotle’s identifi-cations of whales as fish and the workerbee as a male.

AcknowledgmentsFor stimulation, critique, unpublishedinformation, and guidance in the liter-ature I am greatly indebted to O. Berg,G. Caetano-Anolles, L. Collins, M.Ehrenberg, R. Garrett, A. Harish, A.Liljas, D. Penny, D. Valentine, and M.Wang. My gratitude also goes to IrmgardWinkler for help with the manuscriptand to the Royal PhysiographicSociety, Lund, as well as the NobelCommittee for Chemistry at the RoyalSwedish Academy of Sciences,Stockholm, for generous support.

References

1. Gilbert W. 1986. The RNA world. Nature 319:618.

2. Zaug AJ, Cech TR. 1986. The interveningsequence RNA of Tetrahymena is an enzyme.Science 231: 470–5.

3. Collins LJ, Kurland CG, Biggs P, Penny D.2009. The modern RNP world of eukaryotes.J Hered 100: 597–604.

4. Gesteland RF, Cech TR, Atkins JF. ed.2006. The RNA World. Woodbury, NY: ColdSpring Harbor Laboratory Press.

5. Jeffares DC, Poole AM, Penny D. 1998.Relics from the RNA World. J Mol Evol 46:18–36.

6. Zhang B, Cech TR. 1997. Peptide formationby in vitro selected ribozymes. Nature 390:96–100.

7. Shechner DM, Grant RA, Bagby SC,Koldobskaya Y, et al. 2009. Crystal structureof the catalytic core of an RNA-polymeraseribozyme. Science 326: 1271–5.

8. Collins L, Penny D. 2009. The RNA infra-structure: dark matter of the eukaryotic cell?Trends Genet 25: 120–8.

9. Guerrier-Takada C, Lumelsky N, Altman S.1989. Specific interactions in RNA enzyme-substrate complexes. Science 246: 1578–84.

C. G. Kurland Insights & Perspectives.....

870 Bioessays 32: 866–871,� 2010 WILEY Periodicals, Inc.

Hypotheses

10. Holzmann J, Frank P, Loffler E, Bennett KL,et al. 2008. RNaseP without RNA: identifi-cation and functional reconstitution of thehuman mitochondrial tRNA processingenzyme. Cell 135: 462–4.

11. Sun FJ, Caetano-Anolles G. 2010. Theancient history of the structure of ribonu-clease P and the early origins of Archaea.BMC Bioinformatics 11: 153.

12. Noller HF, Hoffarth V, Zimniak L. 1992.Unusual resistance of peptidyl transferaseto protein extraction procedures. Science256: 1416–9.

13. Noller HF. 1993. Peptidyl transferase:protein, ribonucleoprotein, or RNA?J Bacteriol 175: 5297–300.

14. Johansson M, Lovmar M, Ehrenberg M.2008. Rate and accuracy of bacterial proteinsynthesis revisited. Curr Opin Microbiol 11: 1–7.

15. Ban N, Nissen P, Hansen J, Moore PB, et al.2000. The complete atomic structure of thelarge ribosomal subunit at 2.4 A resolution.Science 289: 905–20.

16. Nissen P, Hansen J, Ban N, Moore PB, et al.2000. The structural basis of ribosome activityin peptide bond synthesis. Science 289: 920–30.

17. Maguire BA, Beniaminov AD, Ramu H,Mankin AS, et al. 2005. A protein componentat the heart of an RNA machine: the import-ance of protein L27 for the function of thebacterial ribosome. Mol Cell 20: 427–35.

18. Cech TR. 2000. The ribosome is a ribozyme.Science 289: 878–9.

19. Voorhees RM, Weixlbaumer A, Loakes D,Kelley AC, et al. 2009. Insights into substratestabilization from snapshots of the peptidyltransferase center of the intact 70S ribosome.Nat Struct Mol Biol 16: 528–33.

20. Ehrenberg M, Kurland CG. 1984. Costs ofaccuracy determined by a maximal growthrate constraint. Q Rev Biophys 17: 45–82.

21. Kurland CG. 1992. Translational accuracyand the fitness of bacteria. Annu Rev Genet26: 29–50.

22. Valentine DL. 2007. Adaptations to energystress dictate the ecology and evolutionof the Archaea. Nat Rev Microbiol 5:316–23.

23. Lecompte O, Ripp R, Thierry J-C, Moras D,et al. 2002. Comparative analysis of ribosomalproteins in complete genomes: an example ofreductive evolution at the domain scale.Nucleic Acid Res 30: 5382–90.

24. Poole A, Jeffares D, Penny D. 1999. Earlyevolution: prokaryotes, the new kids on theblock. BioEssays 21: 880–9.

25. Brinkman H, Phillipe H. 1999. Archaea sis-ter-group of bacteria? Indications from tree

reconstruction artifacts in ancient phyloge-nies. Mol Biol Evol 16: 817–25.

26. Kurland CG, Collins LJ, Penny D. 2006.Genomics and the irreducible nature ofeukaryotic cells. Science 312: 1011–14.

27. Kurland CG, Canback B, Berg OG. 2007.The origins of modern proteins. Biochimie89: 1454–63.

28. Kurland CG. 1974. Functional organization ofthe 30S ribosome subunit. In Nomura M,Tissiere A, Lengyel P. ed; Ribosomes.Woodbury, NY: Cold Spring HarborLaboratory Press. p 309–31.

29. Liljas A. 2004. Structural Aspects of ProteinSynthesis. Singapore: World ScientificPublishing Co.Pte. Ltd.

30. Cairns-Smith AG. 1985. Seven Clues to theOrigin of Life. Cambridge: University Press.

31. Kauffman SA. 1993. The Origins of Order.New York: Oxford University Press.

32. Egel R. 2009. Peptide-dominated mem-branes preceding the genetic takeover byRNA: latest thinking on a classic controversy.BioEssays 31: 1100–9.

33. Goldberg AL. 1972. Degradation of abnormalproteins in Escherichia coli. Proc Natl AcadSci 69: 422–36.

34. Goldberg AL. 2003. Protein degradation andprotection against unfolded or damagedproteins. Nature 426: 895–9.

35. Prakash S, Matouschek A. 2004. Proteinunfolding in the cell. Trends Biochem Sci29: 593–600.

36. Oliveberg M, Wolynes PG. 2006. The exper-imental survey of protein-folding energy land-scapes. Q Rev Biophys 38: 275–88.

37. Tyndall JD, Nall T, Fairlie DP. 2005.Proteases universally recognize beta strandsin their active sites. Chem Rev 105: 973–99.

38. Ellis RJ. 2001. Macromolecular crowding:obvious but underappreciated. TrendsBiochem Sci 26: 597–604.

39. Cookson MR. 2005. The biochemistry ofParkinson’s disease. Annu Rev Biochem 4:29–52.

40. Chitil F, Dobson CM. 2006. Protein misfold-ing, functional amyloid and human disease.Annu Rev Biochem 75: 333–66.

41. Gidalevitz T, Ben-Zvi A, Ho KH, Brignull HR,et al. 2006. Progressive disruption of cellularprotein folding in models of polyglutaminediseases. Science 246: 1471–84.

42. Powers ET, Morimoto RI, Dillin A, Kelly JW,et al. 2009. Biological and chemicalapproaches to diseases of proteostasisdeficiency. Annu Rev Biochem 78: 959–91.

43. Schubert U, Anton LC, Gibbs J, NorburgCC, et al. 2000. Rapid degradation of a largefraction of newly synthesized proteins by pro-teasomes. Nature 404: 770–4.

44. Pratt JM, Petty J, Riba-Garcia I, RobertsonDHL, et al. 2002. Dynamics of protein turn-over, a missing dimension in proteomics. MolCell Proteomics 1: 579–91.

45. Kurland CG, Berg OG. 2010. A hitchhiker’sguide to evolving networks. In Caetano-Anolles G. ed; Evolutionary Genomics andSystems Biology. Hoboken, New Jersey:Wiley and Sons. p 363-96.

46. Doolittle RF. 1995. The multiplicity ofdomains in proteins. Annu Rev Biochem 64:287–314.

47. Hancock REW, Chapple DS. 1999. Peptideantibiotics. Antimicrob Agents Chemother 43:1317–23.

48. Finking R, Marahiel MA. 2004. Biosynthesisof non-ribosomal peptide. Annu RevMicrobiol58: 453–88.

49. Marahiel MA. 2009. Working outside theprotein-synthesis rules: insights into non-ribo-somal peptide synthesis. J Pept Sci 15: 607–799.

50. Keefe AD, Szostak JW. 2001. Functionalproteins from a random-sequence library.Nature 410: 715–8.

51. Szostak JW. 2009. Systems chemistry onearly earth. Nature 459: 171–2.

52. Powner MW, Gerland B, Sutherland JD.2009. Synthesis of activated pyrimidine ribo-nucleotides in prebiotically plausible con-ditions. Nature 459: 239–242.

53. Freistroffer DV, Kwiatkowski M,Buckingham RH, Ehrenberg M. 2000. Theaccuracy of codon recognition by polypeptiderelease factors. Proc Natl Acad Sci 97: 2046–51.

54. Zhang S, Marini DM, Hwang W, Santoso S.2002. Design of nanostructured biologicalmaterials through self-assembly of peptidesand proteins. Curr Opin Chem Biol 6: 865–71.

55. Yokoi H, Kinoshita T, Zhang S. 2005.Dynamic reassembly of peptide RADA16nanofiber scaffold. Proc Natl Acad Sci 102:8414–9.

56. Szostak JW, Bartel DP, Luisi PL. 2001.Synthesizing life. Nature 409: 387–90.

57. Chen IA, Roberts RW, Szostak JW. 2004.The emergence of competition betweenmodel protocells. Science 305: 1474–6.

58. Hanczye MM, Szostak JW. 2004.Replicating vesicles as models of primitivecell growth and division. Curr Opin ChemBiol 8: 660–4.

59. Tooze J. 1983. DNA Makes RNA MakesProtein. Amsterdam: Elsevier.

60. Caetano-Anolles D, Kim KM, Mittenthal J,Caetano-Anolles G. 2010. Proteome evol-ution and the metabolic origins of translationand cellular life. J Mol Evol in press.



Hypotheses

Bernhardt Biology Direct 2012, 7:23http://www.biology-direct.com/content/7/1/23

COMMENT Open Access

The RNA world hypothesis: the worst theory of theearly evolution of life (except for all the others)a

Harold S Bernhardt

Abstract

The problems associated with the RNA world hypothesis are well known. In the following I discuss some of thesedifficulties, some of the alternative hypotheses that have been proposed, and some of the problems with thesealternative models. From a biosynthetic – as well as, arguably, evolutionary – perspective, DNA is a modified RNA,and so the chicken-and-egg dilemma of “which came first?” boils down to a choice between RNA and protein. Thisis not just a question of cause and effect, but also one of statistical likelihood, as the chance of two such differenttypes of macromolecule arising simultaneously would appear unlikely. The RNA world hypothesis is an example ofa ‘top down’ (or should it be ‘present back’?) approach to early evolution: how can we simplify modern biologicalsystems to give a plausible evolutionary pathway that preserves continuity of function? The discovery that RNApossesses catalytic ability provides a potential solution: a single macromolecule could have originally carried outboth replication and catalysis. RNA – which constitutes the genome of RNA viruses, and catalyzes peptide synthesison the ribosome – could have been both the chicken and the egg! However, the following objections have beenraised to the RNA world hypothesis: (i) RNA is too complex a molecule to have arisen prebiotically; (ii) RNA isinherently unstable; (iii) catalysis is a relatively rare property of long RNA sequences only; and (iv) the catalyticrepertoire of RNA is too limited. I will offer some possible responses to these objections in the light of work by ourand other labs. Finally, I will critically discuss an alternative theory to the RNA world hypothesis known as ‘proteinsfirst’, which holds that proteins either preceded RNA in evolution, or – at the very least – that proteins and RNAcoevolved. I will argue that, while theoretically possible, such a hypothesis is probably unprovable, and that theRNA world hypothesis, although far from perfect or complete, is the best we currently have to help understand thebackstory to contemporary biology.

Reviewers: This article was reviewed by Eugene Koonin, Anthony Poole and Michael Yarus (nominated byLaura Landweber).

Keywords: RNA world hypothesis, Proteins first, Acidic pH, tRNA introns, Small ribozymes

BackgroundThe problems associated with the RNA world hypothesisare well known, not least to its proponents [1,2]. In thefollowing, I discuss some of these difficulties, some ofthe alternative hypotheses that have been proposed (in-cluding the ‘proteins first’ hypothesis), and some of theproblems with these alternative models. As part of thediscussion, I highlight the support provided to the RNAworld concept by the discovery of some extremely smallribozymes. The activities of these provide support for

Correspondence: [email protected] of Biochemistry, University of Otago, P.O. Box 56, Dunedin, NewZealand

© 2012 Bernhardt; licensee BioMed Central LtdCommons Attribution License (http://creativecreproduction in any medium, provided the or

proposals we have made previously for the identity ofthe first tRNA [3], for the origin of coded ribosomal pro-tein synthesis [4], and for the evolution of an RNA worldat acidic pH [5] (see also [6]). I also revisit the proposalfor a replicase origin of the ribosome, and what has be-come the most commonly held model for the origin oftRNA.In modern biological systems, the components of

DNA are synthesized from RNA components [7], and ittherefore makes sense to view DNA as a modified RNA.Similarly, the ribosome – the universal cellular machinethat makes proteins – is composed mainly of RNA, andRNA is its active component, although there are indica-tions that proteins may be playing an increasing role in

. This is an Open Access article distributed under the terms of the Creativeommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andiginal work is properly cited.

mailto:[email protected]

http://creativecommons.org/licenses/by/2.0

Bernhardt Biology Direct 2012, 7:23 Page 2 of 10http://www.biology-direct.com/content/7/1/23

some instances e.g. [8,9] (even in the case of nonriboso-mal peptide synthesis [10,11], the protein enzymecomplexes that synthesize other proteins are of coursethemselves synthesized on the ribosome). RNA func-tions as both catalyst (e.g. in peptide synthesis andtRNA maturation) and genome (in RNA viruses such asHIV and influenza viruses). In contrast to nucleic acids,which associate according to the rules of base paircomplementarity, the intricacies of protein structure donot – normally – allow for an easy mechanism of repli-cation, which presumably explains the evolution of acoded system for their synthesis (for an interesting dis-cussion of the contrasting molecular requirements forreplication and catalysis, see [12]). Parsimony at leastwould seem to favour a scenario in which functionscarried out by two classes of macromolecules in themodern system were, at an earlier stage, carried out byonly one (for an alternative view however, see [13]). Sowhich came first, the chicken or the egg? Protein orRNA? This is an underlying current in the debate sur-rounding the RNA world hypothesis, which I addresswhen I discuss the ‘proteins first’ hypothesis.Before beginning, it is important to clear up a com-

mon source of confusion. The RNA world hypothesisdoes not necessarily imply that RNA was the first repli-cating molecule to appear on the Earth (although a newpaper by Benner and colleagues argues that this was, infact, the case [14]). The more general claim is that theRNA world comprised a stage of evolution preceding –perhaps immediately – the RNA/protein/DNA world wenow inhabit. In this way, the hypothesis is not incompat-ible with models such as the ‘crystals-as-genes’ conceptof Cairns-Smith [15], which proposes that the first repli-cators were imperfection-containing layers of clay thatwere able to pass on these imperfections to proceedinglayers (unfortunately, one experimental test of Cairns-Smith’s model suggests that replicated defects arequickly overrun by random defects or noise [16]). Simi-larly, it has been hypothesized that RNA was precededin evolution by a nucleic acid analogue – for example,one in which glycerol replaces ribose in the phospho-diester backbone – though pathways for the prebioticsynthesis of many such analogues are even less plausiblethan for RNA itself [17].

DiscussionThe following objections to the RNA world hypothesishave been raised:

RNA is too complex a molecule to have arisenprebioticallyRNA is an extremely complex molecule, with four differ-ent nitrogen-containing heterocycles hanging off a back-bone of alternating phosphate and D-ribose groups

joined by 3′,5′ linkages. Although there are a number ofproblems with its prebiotic synthesis, there are a fewindications that these may not be insurmountable.Following on from the earlier work of Sanchez andOrgel [18], Powner, Sutherland and colleagues [19] havepublished a pathway for the synthesis of pyrimidinenucleotides utilizing plausibly prebiotic precursor mole-cules, albeit with the necessity of their timed delivery(this requirement for timed delivery has been criticizedby Benner and colleagues [14], although most origin oflife models invoke a succession of changing conditions,dealing as they do with the evolution of chemical sys-tems over time; what is critical is the plausibility of thechanges). A particularly interesting aspect of the path-way is the use of UV light as a method of isolating thenaturally occurring nucleotides [18,19], suggesting apossible means of nucleotide selection (see also [20]).Although RNA is constructed with uniform 3′,5-linked

backbones, recent work by Szostak and colleagues hasdemonstrated that ribozymes and RNA aptamers retainpartial function when the standard 3′,5′-linkages arereplaced with a mixture of 3′,5′- and 2′,5′- linkages,suggesting that a degree of heterogeneity may be com-patible with (or even beneficial to) RNA function andsynthesis (J. Szostak, pers. commun.; [21]). This comple-ments an earlier study by Ertem and Ferris [22] thatshowed that poly C oligonucleotides with mixed 3′,5′-and 2′,5′-linkages are able to serve as templates for thesynthesis of poly G oligonucleotides by nonenzymaticreplication. Such work suggests that ancestral systemsmay not have been as tightly constrained as theyare today.Due perhaps to the molecular complexity of nucleic

acids, metabolism-first models (as opposed toreplication-first models such as the RNA world hypoth-esis) highlight the importance of the initial generation ofsmall molecules through chemical or metabolic cycles.Establishment of a plausible energy source is a criticalaspect of these models, some of which propose that lifearose in the vicinity of hot alkaline (pH 9–11) under-seahydrothermal vents, with energy provided by pH andtemperature gradients between the vent and the cooler,more acidic ocean [23-26]. In some ways, metabolism-first models appear not to conflict with the RNA worldhypothesis, as they potentially offer a solution to the dif-ficulty of ribonucleotide and RNA synthesis. A largepoint of difference, however, comes with the claim thatsuch nucleic acid-free systems are capable of Darwinianevolution. Addressing this claim, Vasas et al. [27] havereported a lack of evolvability in such systems, whileBenner and colleagues have noted the lack of experi-mental support from specific chemical models [14]. Amore recent paper by Vasas et al. [28], while seeminglycontradicting their earlier paper, uses a computational


modeling approach without reference to a real-worldchemical system (something noted by two of thereviewers in their published reviews).

RNA is inherently unstableRNA is often considered too unstable to have accumu-lated in the prebiotic environment. RNA is particularlylabile at moderate to high temperatures, and thus anumber of groups have proposed the RNA world mayhave evolved on ice, possibly in the eutectic phase (a li-quid phase within the ice solid) [29-33]. Two of thesestudies [31,32] demonstrated maximal ribozymic activityat −7 to −8°C, possibly due to the combined effects ofincreased RNA concentration and lowered water activity.A possible difficulty with this scenario is that RNAsequences have an increased tendency to base pair atsuch temperatures, leading in some cases to the forma-tion of intermolecular complexes [34] that potentiallycould reduce catalytic activity.A further problem is the susceptibility of RNA to

base-catalyzed hydrolysis at pH >6 [35]. The phospho-diester bonds of the RNA backbone and the ester bondbetween tRNAs and amino acids – something similar towhich would have been critical for the evolution ofribosomal protein synthesis – are both more stable atpH 4–5 [5,6]. With our proposal for RNA world evolu-tion at acidic pH [5], we have suggested that the primor-dial ‘soup’ may have been more like vinaigrette, whileHanczyc [36] has drawn a comparison with mayonnaise,with its emulsified mixture of oil in water (in light ofthese, could there be potential for food science to pro-vide insights for origin of life studies?) While Mg2+ isimportant for stabilizing RNA secondary and tertiarystructure, high Mg2+ concentrations also catalyze RNAdegradation, which has been identified as a particularproblem in the case of RNA template copying [21]. Heretoo, acidic pH offers a possible solution, as the positivecharge on protonated cytosine and adenosine residues inacidic conditions may reduce the requirement for diva-lent cations. For example, a self-cleaving ribozyme withmaximum activity at pH 4 isolated by in vitro selection,is active in the absence of divalent ions (including Mg2+)[37]. RNA secondary (and tertiary) structure would ap-pear to be compatible with the presence of protonatednucleotides, as we have found an increased number ofpotentially protonated A-C base pair ‘mismatches’ in thetRNAs from acidophilic archaeal species with reportedcytoplasmic pHs of 4.6-6.2 [5].

Catalysis is a relatively rare property of long RNAsequences onlyThe RNA world hypothesis has been criticized becauseof the belief that long RNA sequences are needed forcatalytic activity, and for the enormous numbers of

randomized sequences required to isolate catalytic andbinding functions using in vitro selection. For example,the best ribozyme replicase created so far – able to repli-cate an impressive 95-nucleotide stretch of RNA – is~190 nucleotides in length [38], far too long a sequenceto have arisen through any conceivable process ofrandom assembly. And typically 10,000,000,000,000-1,000,000,000,000,000 randomized RNA molecules arerequired as a starting point for the isolation of ribozy-mic and/or binding activity in in vitro selection experi-ments, completely divorced from the probable prebioticsituation. As Charles Carter, in a published review ofour recent paper in Biology Direct [5], puts it:

“I, for one, have never subscribed to this view of theorigin of life, and I am by no means alone. The RNAworld hypothesis is driven almost entirely by the flowof data from very high technology combinatoriallibraries, whose relationship to the prebiotic world isanything but worthy of “unanimous support”. Thereare several serious problems associated with it, and Iview it as little more than a popular fantasy”(reviewer's report in [5]).

1014-1016 is an awful lot of RNA molecules. However,the discovery of a number of extremely short ribozymessuggests that long sequences – and hence the hugenumbers of RNA molecules required to sample the ne-cessary sequence space – might not have been necessary.In a section titled ‘Miniribozymes: small is beautiful,Landweber and colleagues [31] discuss a number of suchsmall ribozymes, including a minimal size active duplexof only 7 nucleotides that self-cleaves. Regarding therelatively modest rate enhancement of this miniribozyme– three orders of magnitude less than the parent ribo-zyme from which it is derived – the authors conclude:“the smallest molecules are likely to arise first, and anyrate enhancement would have been beneficial in a pre-biotic setting” [31]. Another, closely related, miniribozymecan ligate a small RNA to its 5′ end, requiring only a sin-gle(!) bulged nucleotide in the context of a larger base-paired structure containing a strand break. Interestingly,the self-cleaving 7-nucleotide sequence forms a part of theligase ribozyme, demonstrating the closeness in sequencespace of the two, albeit related, functions [31]. Equally asinteresting from an RNA world perspective, Yarus and col-leagues have recently isolated by in vitro selection a ribo-zyme that is able to be truncated to just 5 nucleotides,while retaining its ability to catalyze the aminoacylation intrans of a 4-nucleotide RNA substrate [39]. Remarkably,only 3 nucleotides are responsible for this activity: 2 in theribozyme and 1 in the substrate. In fact, even this much isnot required: a variant of the parent ribozyme with a mu-tation of 1 of the 3 conserved nucleotides is able to


aminoacylate a substrate variant with the sequence GCCA(similar to the universal aminoacylated 3′ terminus oftRNA), albeit at a reduced rate [40] (we have previouslyproposed a possible sequence for an aminoacylating ribo-zyme based on this variant that could have base-pairedwith the universal 3′ CCA termini of tRNAs (and pro-posed RNA hairpin precursors [41,3] through a doublehelix interaction, while also forming specific triple helixinteractions – at acidic pH – with other nucleotides in thetRNA [5]). As with the small ribozymes discussed byLandweber and colleagues, the rates of aminoacylation ofYarus' ribozymes are somewhat underwhelming: that ofthe original 5-nucleotide ribozyme is only 25-fold higherthan the uncatalyzed rate [39], while that of the variant isonly 6-fold higher than the uncatalyzed rate [40] (for fur-ther discussion of the implications of such tiny ribozymessee [42], and [31] and references therein).Although not quite as small as the ribozymes dis-

cussed above, Gross and colleagues have demonstratedthat 12-nucleotide and 20-nucleotide nuclear tRNATyr

introns from Arabidopsis thaliana and Homo sapiens –understood to be cleaved by protein enzymes in vivo –are able to self-cleave in the presence of 10 mM Mg2+,0.5 mM spermine and 0.4% Triton X-100 [43-45]. Al-though the introns form part of a larger pre-tRNA se-quence, the nucleotides responsible for self-excision arepossibly confined to a 3- or 4-nucleotide bulge region.The discovery of this intrinsic activity (which admittedlyrequires the presence of a low concentration of surfac-tant) supports previous proposals for the origin of tRNA[41,3,4]. Although there exist a number of other modelsfor the origin of tRNA (one of which is discussed in detailin the following section), a hairpin duplication-ligation ori-gin stands as a credible hypothesis [41,3] that has receivedsupport from a number of sources [46-48]. Briefly, the idea- first proposed by Di Giulio [41] - is that two (either

Figure 1 A proposal for the origin of tRNA through the ligation of a hintron based on proposals by Di Giulio [41], and Dick and Schamel [4extension of one of the precursor hairpins formed by a transcriptional runoimply that an amino acid was necessarily attached here during the intron l

identical or very similar) hairpins, approximately half thesize of contemporary tRNA, formed a ligated duplex dueto the symmetry of base-pairing interactions, possibly byan intron-mediated mechanism [49] (Figure 1). It has beenproposed previously that contemporary protein-splicednuclear tRNA introns are descended from an ancestralself-splicing group I-type intron that catalyzed the ancestralligation [49] (as depicted in Figure 1, the ancestral tRNA in-tron may have derived from a 3′ extension of one of theprecursor hairpins by a transcriptional runoff error). Thefindings of Gross and colleagues [43-45] indicate that somenormally protein-cleaved nuclear tRNA introns have par-tially retained the ability to self-cleave. This ability to self-cleave implies the reverse reaction – self-ligation – is alsopossible, which could have produced the ligated intron-containing hairpin intermediate; subsequent intron self-cleavage could have produced the first proto-tRNA [49](Figure 1).

The catalytic repertoire of RNA is too limitedIt has been suggested that the probable metabolicrequirements of an RNA world [50] would haveexceeded the catalytic capacity of RNA. The majority ofnaturally occurring ribozymes catalyze phosphoryl trans-fer reactions – the making and breaking of RNAphosphodiester bonds [51]. Although the most efficientof these ribozymes catalyze the reaction at a comparablerate to protein enzymes – and in vitro selection has iso-lated ribozymes with a far wider range of catalytic abil-ities [9,51] – the estimate of proteins being one milliontimes fitter than RNA as catalysts seems reasonable, pre-sumably due to proteins being composed of 22 chem-ically rather different amino acids as opposed to the 4very similar nucleotides of RNA [12].It is frequently forgotten however that proteins too

have their catalytic limitations: after all, many enzyme

airpin duplex catalyzed by an ancestral self-splicing group I-type9]. In this depiction, the intron is shown as originating from a 3′ff error. aa indicates the amino acid binding site, but is not meant toigation events.


active sites contain cofactors and/or coordinated metalions, suggesting that some reactions are ‘too hard’ forproteins as well (it is estimated that ~50% of proteinsare metalloproteins [52], although of course not all thesemetal ions are found at the active site). RNA ribos-witches bind a range of protein cofactors, such as flavinmononucleotide, thiamine pyrophosphate, tetrahydrofo-late, S-adenosylmethionine and adenosylcobalamin (a formof vitamin B12) [53]. In the case of the glmS riboswitch/ribozyme, the metabolite glucosamine-6-phosphate bindsin the active site and appears to participate in catalysis[54]. Because of the ability of these naturally occurringRNA riboswitches to bind protein enzyme cofactors, andbecause many of these cofactors possess non-functionalfragments of RNA – one of the earliest pointers to apossible ancestral RNA world [55] – it is likely that atleast some of the cofactors now used by proteins werehanded down directly from the RNA world, where theyplayed a similar if not identical role in assisting catalyticfunction [53].One of the arguments for the RNA world hypothesis

comes from the observation that RNAs are, in most cases,worse catalysts than proteins. This implies that their pres-ence in modern biological systems can best be explainedby their being remnants of an earlier stage of evolution,which were too embedded in biological systems to allowreplacement easily. An alternative explanation is that theywere co-opted by a protein world due to their superiorproperties for the particular functions they perform. Whilesuch an explanation seems intuitively less likely, surpris-ingly it is held by some proponents of the ‘proteins first’model [56-60] (discussed in more detail below).

Proteins firstAn increasingly strident view is that protein either pre-ceded RNA in evolution or, at the very least, that RNAand protein coevolved, in what is known as the ‘proteins(or peptides) first’ hypothesis [56-60]. Take, for example,Charles Kurland in his 2010 piece in Bioessays [57],which is utterly scathing of the RNA world hypothesisand its fellow travelers:

“[The RNA world hypothesis] has been reduced by ritualabuse to something like a creationist mantra”, and

“[The] RNA world is an expression of the infatuationof molecular biologists with base pairing in nucleicacids played out in a one-dimensional space with noreference to time or energy” [57].

On a less emotional note, Harish and Caetano-Anollés[60] earlier this year published a phylogenetic analysis ofribosomal RNA and ribosomal proteins, concluding thatthe oldest region of the ribosome is a helical stem of the

small ribosomal subunit RNA and the ribosomal proteinthat binds to it. As this helical stem has the importantroles in the modern ribosome of decoding the mRNAmessage and in the movement of the two subunits rela-tive to each other (including translocation of the mRNAmessage and tRNAs), Harish and Caetano-Anollés con-clude that the original function of the ribosome was asan RNA replicase (this idea, which has been suggestedpreviously, is discussed in detail in the following sec-tion). In addition, because RNA and protein componentsof the ribosome apparently have similar ages, Harish andCaetano-Anollés surmise that peptide synthesis has al-ways been carried out by RNA in association with pro-teins, as is the case with the modern ribosome.Without debating the merits or otherwise of their

phylogenetic techniques, the most serious objection tothese conclusions is that phylogenetic analysis has thelimitation that it can only analyze the protein sequencerecord as it has been captured in DNA (this is true evenfor a phylogenetic analysis based on protein fold struc-tures, as the only record we possess of these folds is theirprimary amino acid sequence as captured in the DNA).Therefore, any information we can recover can only datefrom the advent of coded protein synthesis, as that is thepoint at which protein sequence became coded in nucleicacid. In an online report [61] on Harish and Caetano-Anollés’ paper, Russell Doolittle makes this same point:

“This is a very engaging and provocative article by oneof the most innovative and productive researchers in thefield of protein evolution,” said University of Californiaat San Diego research professor Russell Doolittle, whowas not involved in the study. Doolittle remains puzzled,however, by “the notion that some early proteins weremade before the evolution of the ribosome as a protein-manufacturing system.” He wondered how – if proteinswere more ancient than the ribosomal machinery thattoday produces most of them –“the amino acidsequences of those early proteins were ‘remembered’ andincorporated into the new system.” [61].

To which, Caetano-Anollés’ reported response isslightly puzzling:

“It requires understanding the boundaries of emergentbiological functions during the very early stages ofprotein evolution. However, the proteins that catalyzenon-ribosomal protein synthesis – a complex andapparently universal assembly-line process of the cellthat does not involve RNA molecules and can stillretain high levels of specificity – are more ancientthan ribosomal proteins. It is therefore likely that theribosomes were not the first biological machines tosynthesize proteins.” ([61]; italics in original).


It is certainly possible that there were functional noncodedpeptides prior to the advent of coded protein synthesis.These could have been formed either through random pro-cesses, by noncoded ribosomal synthesis prior to the adventof coding [4], by non-ribosomal peptide synthesis catalyzedby specific ribozymes (analogous to non-ribosomal peptidesynthesis catalyzed by protein enzymes in modern systems[62]), or by some combination of the above. It seems highlyunlikely, however, that proteins synthesized proteins prior tothe advent of the ribosome, as this would appear to suggestan infinite regression series. As Doolittle [61] suggests, thecritical point is that once coding evolved, the sequences ofthese noncoded proteins would have needed to be recapitu-lated by coded proteins; therefore the phylogenetic signalwould only go back to the point of recapitulation. Put an-other way, the earliest proteins phylogenetically speaking willbe the first proteins that were coded for. Presumably, if thesesequences can still be detected in modern genomes, theywould tend to be relatively short and somewhat indistincttraces only, as one might expect for the first proteins pro-duced by a rudimentary ribosome. In a sense then, one cansay that the advent of coded protein synthesis has drawn aveil over the previous life of proteins. Although it seems un-likely, complex proteins may have existed prior to this, but –as all record of them has been erased by the advent of coding– that is as much as we can say (for an in-depth discussionof the implications of non-ribosomal peptide synthesis forthe RNA world hypothesis, see [62]).

RNA replicase origin of the ribosomeAs mentioned above, Harish and Caetano-Anollés arenot the first to suggest an RNA replicase origin of theribosome (or small ribosomal subunit). The idea, whichwas possibly first proposed by Weiss and Cherry [63], isthat “the ancestor of small subunit RNA was an RNAreplicase that used oligonucleotides as a substrate” [63].The hypothesis has grown in scope to include the use ofexcised tRNA anticodons as the source of oligonucleo-tides, with the energy required for ligation provided byconcomitant peptide bond formation [64-66]. However,as pointed out by Wolf and Koonin [67], such a ligasewould have required a molecular machinery at least ascomplex as the modern ribosome, which would make itan unlikely evolutionary forerunner. This notwithstand-ing, Weiss and Cherry’s original, simpler, model mayhave some merit. If, as has been recently suggested, earlyRNA replication was performed by the ligation of shortoligonucleotides [68,69], or by a combination of nucleo-tide polymerization and oligonucleotide ligation [21], a‘decoding’ RNA able to proofread triplet base pair inter-actions for accuracy – similar to its role in the modernribosome of maintaining the fidelity of the triplet codon-anticodon interaction – might have played an importantrole. Interestingly, a 49-nucleotide hairpin comprising

part of the decoding site of the small ribosomal subunitRNA has been found to bind both poly U oligonucleo-tide and the tRNAPhe anticodon stem-loop in a similarfashion to the entire small subunit [70]. This hairpincontains the two mobile nucleotides A1492 and A1493

(numbered according to the Escherichia coli small ribo-somal subunit RNA sequence) that proofread theanticodon-codon helix in the modern ribosome [71]. Itwould be interesting to test whether this hairpin is ableto enhance the rate and/or accuracy of non-enzymaticligation using a single-stranded RNA ‘template’ and shortcomplementary oligonucleotides. If an enhancementwere indeed demonstrated, such a mechanism would beanalogous to that utilized by the large ribosomal subunit,for which substrate positioning of the two tRNAs mayconstitute one of its main roles in catalyzing peptidesynthesis [72].As part of their model of early RNA replication by

oligonucleotide ligation, Manrubia and colleaguespropose that an increase in the catalytic rate of the rep-licase/ligase would have occurred with an increase in se-quence length through a process of bootstrapping[68,69]. Furthermore, they suggest that the first RNAreplication possibly had a high error-rate:

“Highly mutagenic replication processes could haveproduced relatively large repertoires of short,genetically different molecules, some of them foldinginto secondary/tertiary structures able to performselectable functions” [68].

Similarly, we have proposed that, in an RNA worldevolving at acidic pH, non-standard base pairing interac-tions due to base protonation could have provided ameans of increasing RNA sequence variation throughnon-enzymatic replication [5].

The origin of tRNAWiener and Maizels’ genomic tag hypothesis proposesthat the 3′ (or ‘top’) half of tRNA originally functionedas a tag demarking the 3′-end of genomic RNAs for rep-lication, and thus was the first part of tRNA to evolve[73]. Sun and Caetano-Anollés [74,75] have publishedphylogenetic evidence that they believe supports thegenomic tag hypothesis by confirming, “that the ‘tophalf ’ of tRNA is more ancient than the ‘bottom half ’”[75]. Noller [76] has observed that the tRNA top half(comprising the T arm and the acceptor stem – includ-ing the amino acid binding site) interacts almost exclu-sively with the large ribosomal subunit, while thebottom half (comprising the D and anticodon arms)interacts almost exclusively with the small subunit. Be-cause peptide synthesis (a function of the large subunit)is usually viewed as more ancestral than decoding (a


function of the small subunit) – a view which has sup-port from a structural analysis by Bokov and Steinberg[77] – the top half of tRNA (which interacts with thelarge subunit) has been viewed as being more ancestralthan the bottom half [73,78]. However, this ‘standardmodel’ for the origin of tRNA, and the results of Sunand Caetano-Anollés that support this model [74,75],are apparently both in conflict with Harish and Caetano-Anollés’ [60] more recent findings on the relative ages ofthe ribosomal subunits. As described above, these find-ings suggest that the small ribosomal subunit was thefirst to evolve, which is difficult to reconcile with the factthat the bottom half of tRNA (with which the small sub-unit mainly interacts), is, by theirs [74,75] and others[73,78] estimation, the newer half of tRNA. Equally, theirfinding that the large ribosomal subunit evolved morerecently [60] is difficult to reconcile with the fact thatthe top half of tRNA (with which the large subunitmainly interacts), is, by theirs and others estimation, theolder half of tRNA. Incidentally, Caetano-Anollés andcolleagues’ finding [75,79,80] that the most ancienttRNAs coded for selenocysteine, tyrosine, serine and leu-cine not only runs counter to other work in the area(see e.g. [81]), but – as these tRNAs all possess long vari-able arms – appears to contradict their own finding thatthe “variable region was the last structural addition tothe molecular repertoire of evolving tRNA substruc-tures” [74].

As discussed above, a plausible scenario for the origin oftRNA is the duplication and subsequent ligation of an RNAhairpin approximately half the length of modern tRNA (or al-ternatively the ligation of two very similar hairpins) [41,3],with ligation possibly catalyzed by an ancestral self-cleavingintron [49] (see Figure 1). An important implication of suchan origin is that both tRNA halves are of equal antiquity, asboth would have to be present for ligation to occur! However,due to the symmetry of the tRNA molecule, the top half,which is considered to be the more ancient, is in fact moreancient-like, as it retains the base-paired 3′ and 5′ ends ofthe original hairpin from which it derives. In contrast, thebottom half, considered to be the more recently acquired,contains the ‘join’ between the two hairpins, which hasaltered the conformation of the original hairpin, giving thisbottom half a new structure. If one accepts a hairpinduplication-ligation origin of tRNA, this explains why the tophalf of tRNA interacts with the peptidyl transferase region ofthe large ribosomal subunit: it is because this half retains thesame structure (and possibly nucleotide sequence) as thehairpin from which it derives, which originally interactedwith the peptidyl transferase region of the large subunit.In fact - and this point has been made by others [49] – thisretention of structure probably favoured (or even enabled)the duplication event, as it meant the resultant tRNA wasable to be aminoacylated by the same ribozyme synthetase

that aminoacylated the hairpin precursor, and thereforethe tRNA was able to participate in ribosomal protein syn-thesis. At the same time, the appearance of a novel struc-ture at the ligation point – the anticodon loop – allowedfor the subsequent evolution of genetic coding [4,3].One of the strongest arguments in favour of the hair-

pin ligation being catalyzed by an ancestral self-cleavingintron [49] (as depicted in Figure 1) is the presence ofthe highly conserved ‘canonical intron insertion position’between nucleotides 37 and 38 in the anticodon loop[41], where almost all eukaryotic nuclear (and the major-ity of archaeal) tRNA introns are found, even thoughintrons are only found in a subset of tRNA isoacceptors[82]. It has been proposed previously that this conservedposition constitutes a 'molecular memory’ of the positionof the ancestral intron that was responsible for theligation that created the first tRNA [83]. If the canonicalintron insertion position is ancestral, it implies thateukaryotic nuclear tRNAs (and possibly archaeal tRNAs)have a more ancestral structure than eubacterial tRNAs,which usually lack tRNA introns altogether or possessself-splicing introns at a variety of different positions inthe molecule. Such a finding is consistent with theintrons-early hypothesis, and the proposal that eubac-teria have undergone a process of intron loss [84,85].

ConclusionsI have argued that the RNA world hypothesis, whilecertainly imperfect, is the best model we currently havefor the early evolution of life. While the hypothesisdoes not exclude a number of possibilities for what – ifanything – preceded RNA, unfortunately the evolutionof coded protein synthesis has drawn a veil over theprevious history of proteins. The situation is differentin the case of non-coding RNAs such as ribosomalRNA and tRNA, as these were able to replicate prior tothe evolution of ribosomal protein synthesis.As we have noted previously [5], the proposal that the

RNA world evolved in acidic conditions [5,6] offers aplausible solution to Charles Kurland's criticism [57]that the RNA world hypothesis makes no reference to apossible energy source. As de Duve [87] has noted, "thewidespread use of proton-motive force for energy trans-duction throughout the living world today is explainedas a legacy of a highly acidic prebiotic environment andmay be viewed as a clue to the existence of such an en-vironment" [87]. Although Russell, Martin and others[23-26] have argued that proton and thermal gradientsbetween the outflow from hot alkaline (pH 9-11) under-sea hydrothermal vents and the surrounding coolermore acidic ocean may have constituted the first sourcesof energy at the origin of life, the lack of RNA stabilityat alkaline pH ([5] and references within) would appear


to make such vents an unlikely location for RNA worldevolution.Although possible, it seems unlikely that the A-C base

pair 'mismatches' found in the tRNA genes of Ferro-plasma acidarmanus and Picrophilus torridus (two spe-cies of archaebacteria with a reportedly acidic internalpH) [5] are corrected by C to U RNA editing thatoccurs, for example, with some - but not other - plantchloroplast tRNAs [88,89]. Such editing of secondarystructure A-C base pair mismatches has so far not beenfound to occur in archaebacteria; however, in a singlearchaeal species (Methanopyrus kandleri) a tertiarystructure A-C base pair found in 30 of its 34 tRNAsundergoes C to U editing catalyzed by a cytidine deami-nase CDAT8 [90]. M. kandleri is a unique organism thatcontains many 'orphan' proteins. CDAT8, which con-tains a cytidine deaminase domain and putative RNA-binding domain, has no homologues in other arachaealspecies, including F. acidarmanus and P. torridus (LRandau, pers. commun.; [90]). Definitive proof, however,that the A-C base pairs in these two species are notmodified would of course require e.g. cDNA sequencingof the tRNAs.

AbbreviationsmRNA: messenger RNA; tRNA: transfer RNA.

Competing interestsThe author declares that he has no competing interests.

AcknowledgementsThis paper is dedicated to my mentor and colleague Professor Warren Tate,who was instrumental in my setting off on this life of adventure anddiscovery and who encouraged me to write this paper. Many thanks to HansGross, George Fox and Steven Benner for critical reading of an early draft ofthis manuscript and for their helpful suggestions. Thanks to Lennart Randaufor helpful information regarding his work on CDAT8 from M. kandleri.Thanks to Diana Yates from the University of Illinois News Service and RussellDoolittle for permission to use material which first appeared there. Theresearch was conducted during tenure of a Health Sciences CareerDevelopment Award at the University of Otago.The title is an adaptation of Sir Winston Churchill’s famous comment ondemocracy made in a speech to the House of Commons on 11 November1947: No one pretends that democracy is perfect or all-wise. Indeed, it has beensaid that democracy is the worst form of government except all those otherforms that have been tried from time to time.

Reviewers’ commentsReferee 1: Eugene KooninI basically agree with Bernhardt. The RNA World scenario is bad as ascientific hypothesis: it is hardly falsifiable and is extremely difficult to verifydue to a great number of holes in the most important parts. To wit, no onehas achieved bona fide self-replication of RNA which is the cornerstone ofthe RNA World. Nevertheless, there is a lot going for the RNA World(Bernhardt summarizes much of the evidence, and I add more below)whereas the other hypotheses on the origin of life are outright helpless.Moreover, as argued in some detail elsewhere [91], the RNA World appearsto be an outright logical inevitability. ‘Something’ had to start efficientlyreplicating to kick off evolution, and proteins do not have this ability. AsBernhardt rightly points out, it is not certain that RNA was the first replicatorbut it does seem certain that it was the first ‘good’ replicator. To clarify, thisdoes not imply that the primordial RNA World did not have peptides; on thecontrary, it is plausible that peptides played important roles but they werenot initially encoded in RNA.

Moreover, straightforward observations on modern proteins indicate that therole of RNA in the ancient translation system was much greater that it is in themodern system. Indeed, Class I aminoacyl-tRNA synthetases (aaRS) representonly a small branch on the complex evolutionary tree of Rossmann-likedomains, so the common ancestor of all 10 Class I aaRS emerged afterextensive diversification of this particular class of protein domains had alreadytaken place. Accordingly, one is compelled to conclude that a high-fidelitytranslation system that alone would enable extensive protein evolution existedalready at the late stages of the hypothetical RNA World [92].

All this discussion is not pointless play with hypotheses. Realization of theunique status of the RNA World among the origin of life scenarios is criticalfor maintaining the focus of research on truly important directions such asexperimental and theoretical study of the evolution of ribozymes rather thanfutile attempts to debunk the RNA World.

Referee 2: Anthony Poole

Harold Bernhardt’s review of the RNA world hypothesis is readable and timely.He presents a very open-minded review of recent results and how they impacton old ideas, and distills a large amount of material. Aside from the admirableattempt to synthesize a vast array of ideas, a valuable contribution hiddenwithin is the critical assessment of the view that the RNA world hypothesisneeds to be abandoned in favour of a peptides-first model.

Author’s response: I have revised the abstract and introduction to includereference to my critique of the ‘proteins (or peptides) first’ hypothesis.While I doubt that anyone seriously excluded peptides as part of a prebiotic milieu,the primacy of peptides does need careful consideration. In this regard, the explicitexplanation of why a pre-genetic code origin of proteins will not be detectablefrom comparative genomic analyses is an important contribution. Perhaps this isobvious to some, but in light of a growing view that non-ribosomal peptidesynthesis preceded ribosomal peptide synthesis, it would seem that the communityneeds a reminder, and Bernhardt spells it out in a very informative manner.Another issue with arguing for non-ribosomal peptide synthesis preceding theribosome is that there is an enormous difference in information input versusoutput. As discussed in [62], megaenzymes like cyclosporin are ~15000 aminoacids in length and produce products of 11 amino acids in length – a factor of104 is not trivial. While non-ribosomal peptide synthetases are modular andcould in principle be engineered into minimal entities, the challenge ofequalizing information input and output is significant regardless of one’sfavoured prebiotic starting point. It is clear from reading Bernhardt’s review thatthe RNA community is much closer to this than those who seek to replaceprimordial RNA-based replication with peptide-based replication.

Referee 3: Michael Yarus (nominated by Laura Landweber)

Almost always, progress to new understanding is sporadic, with insightscoming in separated locales. Difficulties temporarily immobilize discussion,but then are surmounted by a successful theory. This sometimes inchoatestagger toward a broader, more self-consistent argument is all that can beexpected, even of an ultimately successful idea. Discussions of the RNAworld sometimes forget this, and demand e.g., the ultimate replicase today!But this essay by Harold Bernhardt remembers what has happened for othersuccessful evolutionary ideas, like the big tree. For all its successes, the tree isstill being questioned under extreme prejudice in certain quarters, as is theRNA world.

Contrariwise, here we have here a sympathetic review of the support for theRNA world, which specifically makes the point that it fits our descent betterthan other ideas (You look like the son of a montmorillonite to me, yamangy mutant!). It will be useful to those who want an entry to the RNAworld literature, and could easily serve as the crux of a university course.However, this is also its weakness; the text is polite and respectful, even tothose whose ‘contribution’ has been otherwise. It treats even loony ideas(‘we need proteins to evolve translation!’) with deference. Or to put it inother words, it is edgeless – some attitude would be welcome. Some choicebetween hypotheses should go with the territory; some consequentmake-or-break predictions are the responsibilities of a guide. But as a gentleintroduction, you will not find better.


Author’s response: In revising the manuscript, I have – to some degreeinadvertently – added a bit more bite!

Received: 9 May 2012 Accepted: 11 July 2012Published: 13 July 2012

References1. Benner SA, Kim HJ, Yang Z: Setting the stage: the history, chemistry, and

geobiology behind RNA. Cold Spring Harb Perspect Biol 2012, 4:a003541.doi:10.1101/cshperspect.a003541.

2. Robertson MP, Joyce GF: The origins of the RNA world. Cold Spring HarbPerspect Biol 2012, 4:a003608. doi:10.1101/cshperspect.a003608.

3. Bernhardt HS, Tate WP: Evidence from glycine transfer RNA of a frozenaccident at the dawn of the genetic code. Biol Direct 2008, 3:53.

4. Bernhardt HS, Tate WP: The transition from noncoded to coded proteinsynthesis: did coding mRNAs arise form stability-enhancing bindingpartners to tRNAs? Biol Direct 2010, 5:16.

5. Bernhardt HS, Tate WP: Primordial soup or vinaigrette: did the RNA worldevolve at acidic pH? Biol Direct 2012, 7:4.

6. Kua J, Bada JL: Primordial ocean chemistry and its compatibility with theRNA world. Orig Life Evol Biosph 2011, 41:553–558.

7. Forterre P, Grosjean H: The interplay between RNA and DNAmodifications: back to the RNA world. In DNA and RNA Modificationenzymes: Structure, Mechanism, Function and Evolution. Edited by Grosjean H.Austin: Landes Bioscience; 2009:259–274.

8. O'Brien TW: Properties of human mitochondrial ribosomes. IUBMB Life2003, 55:505–513.

9. Strobel SA, Cochrane JC: RNA catalysis: ribozymes, ribosomes, andriboswitches. Curr Opin Chem Biol 2007, 11:636–643.

10. Koglin A, Walsh CT: Structural insights into nonribosomal peptideenzymatic assembly lines. Nat Prod Rep 2009, 26:987–1000.

11. Strieker M, Tanović A, Marahiel MA: Nonribosomal peptide synthetases:structures and dynamics. Curr Opin Struct Biol 2010, 20:234–240.

12. Benner SA, Burgstaller P, Battersby TR, Jurczyk S: Did the RNA world exploitan expanded genetic alphabet? In The RNA World. 2nd edition. Edited byGesteland RF, Cech TR, Atkins JF. Cold Spring Harbor, NY: Cold SpringHarbour Press; 1999:163–181.

13. Kunin V: A system of two polymerases–a model for the origin of life. OrigLife Evol Biosph 2000, 30:459–466.

14. Benner SA, Kim HJ, Carrigan MA: Asphalt, water, and the prebioticsynthesis of ribose, ribonucleosides, and RNA. Acc Chem Res 2012, Epubahead of print.

15. Cairns-Smith AG: The origin of life and the nature of the primitive gene.J Theor Biol 1966, 10:53–88.

16. Bullard T, Freudenthal J, Avagyan S, Kahr B: Test of Cairns-Smith's'crystals-as-genes' hypothesis. Faraday Discuss 2007, 136:231–245.discussion 309–28.

17. Anastasi C, Buchet FF, Crowe MA, Parkes AL, Powner MW, Smith JM,Sutherland JD: RNA: prebiotic product, or biotic invention? Chem Biodivers2007, 4:721–739.

18. Sanchez RA, Orgel LE: Studies in prebiotic synthesis: V. Synthesis andphotoanomerization of pyrimidine nucleosides. J Mol Biol 1970,47:531–543.

19. Powner MW, Gerland B, Sutherland JD: Synthesis of activated pyrimidineribonucleotides in prebiotically plausible conditions. Nature 2009,459:239–242.

20. Sobolewski AL, Domcke W: Molecular mechanisms of the photostabilityof life. Phys Chem Chem Phys 2010, 12:4897–4898.

21. Szostak JW: The eightfold path to non-enzymatic RNA replication. J SystChem 2012, 3:2.

22. Ertem G, Ferris JP: Synthesis of RNA oligomers on heterogeneoustemplates. Nature 1996, 379:238–240.

23. Russell MJ, Hall AJ: The emergence of life from iron monosulphidebubbles at a submarine hydrothermal redox and pH front. J Geol SocLondon 1997, 154:377–402.

24. Martin W, Russell MJ: On the origin of biochemistry at an alkalinehydrothermal vent. Philos Trans R Soc Lond B Biol Sci 2007, 362:1887–1925.

25. Martin W, Baross J, Kelley D, Russell MJ: Hydrothermal vents and the originof life. Nat Rev Microbiol 2008, 6:805–814.

26. Sleep NH, Bird DK, Pope EC: Serpentinite and the dawn of life. Philos TransR Soc Lond B Biol Sci 2011, 366:2857–2869.

27. Vasas V, Szathmáry E, Santos M: Lack of evolvability in self-sustainingautocatalytic networks constraints metabolism-first scenarios for theorigin of life. Proc Natl Acad Sci USA 2010, 107:1470–1475.

28. Vasas V, Fernando C, Santos M, Kauffman S, Szathmáry E: Evolution beforegenes. Biol Direct 2012, 7:1.

29. Bada JL, Bigham C, Miller SL: Impact melting of frozen oceans on theearly Earth: implications for the origin of life. Proc Natl Acad Sci USA 1994,91:1248–1250.

30. Kanavarioti A, Monnard PA, Deamer DW: Eutectic phases in ice facilitatenonenzymatic nucleic acid synthesis. Astrobiology 2001, 1:271–281.

31. Vlassov AV, Kazakov SA, Johnston BH, Landweber LF: The RNA world onice: a new scenario for the emergence of RNA information. J Mol Evol2005, 61:264–273.

32. Kazakov SA, Balatskaya SV, Johnston BH: Ligation of the hairpin ribozymein cis induced by freezing and dehydration. RNA 2006, 12:446–456.

33. Attwater J, Wochner A, Pinheiro VB, Coulson A, Holliger P: Ice as aprotocellular medium for RNA replication. Nat Commun 2010, 1:76.

34. Sun X, Li JM, Wartell RM: Conversion of stable RNA hairpin to ametastable dimer in frozen solution. RNA 2007, 13:2277–2286.

35. Oivanen M, Kuusela S, Lönnberg H: Kinetics and mechanisms for thecleavage and isomerization of the phosphodiester bonds of RNA byBrønsted acids and bases. Chem Rev 1998, 98:961–990.

36. Hanczyc MM: Metabolism and motility in prebiotic structures. Philos TransR Soc Lond B Biol Sci 2011, 366:2885–2893.

37. Jayasena VK, Gold L: In vitro selection of self-cleaving RNAs with a low pHoptimum. Proc Natl Acad Sci USA 1997, 94:10612–10617.

38. Wochner A, Attwater J, Coulson A, Holliger P: Ribozyme-catalyzedtranscription of an active ribozyme. Science 2011, 332:209–212.

39. Turk RM, Chumachenko NV, Yarus M: Multiple translational products froma five-nucleotide ribozyme. Proc Natl Acad Sci USA 2010, 107:4585–4589.

40. Chumachenko NV, Novikov Y, Yarus M: Rapid and simple ribozymicaminoacylation using three conserved nucleotides. J Am Chem Soc 2009,131:5257–5263.

41. Di Giulio M: On the origin of the transfer RNA molecule. J Theor Biol 1992,159:199–214.

42. Yarus M: The meaning of a minuscule ribozyme. Philos Trans R Soc Lond BBiol Sci 2011, 366:2902–2909.

43. van Tol H, Gross HJ, Beier H: Non-enzymatic excision of pre-tRNA introns?EMBO J 1989, 8:293–300.

44. Weber U, Beier H, Gross HJ: Another heritage from the RNA world: self-excision of intron sequence from nuclear pre-tRNAs. Nucleic Acids Res1996, 24:2212–2219.

45. Riepe A, Beier H, Gross HJ: Enhancement of RNA self-cleavage by micellarcatalysis. FEBS Lett 1999, 457:193–199.

46. Nagaswamy U, Fox GE: RNA ligation and the origin of tRNA. Orig Life EvolBiosph 2003, 33:199–209.

47. Widmann J, Di Giulio M, Yarus M, Knight R: tRNA creation by hairpinduplication. J Mol Evol 2005, 61:524–530.

48. Fujishima K, Sugahara J, Tomita M, Kanai A: Sequence evidence in thearchaeal genomes that tRNAs emerged through the combination ofancestral genes as 5' and 3' tRNA halves. PLoS One 2008, 3:e1622.

49. Dick TP, Schamel WWA: Molecular evolution of transfer RNA from twoprecursor hairpins: implications for the origin of protein synthesis. J MolEvol 1995, 41:1–9.

50. Benner SA, Ellington AD, Tauer A: Modern metabolism as a palimpsest ofthe RNA world. Proc Natl Acad Sci USA 1989, 86:7054–7058.

51. Hiller DA, Strobel SA: The chemical versatility of RNA. Philos Trans R SocLond B Biol Sci 2011, 366:2929–2935.

52. Thomson AJ, Gray HB: Bio-inorganic chemistry. Curr Opin Chem Biol 1998,2:155–158.

53. Cochrane JC, Strobel SA: Riboswitch effectors as protein enzymecofactors. RNA 2008, 14:993–1002.

54. Cochrane JC, Lipchock SV, Smith KD, Strobel SA: Structural and chemicalbasis for glucosamine 6-phosphate binding and activation of the glmSribozyme. Biochemistry 2009, 48:3239–3246.

55. White HB III: Coenzymes as fossils of an earlier metabolic state. J Mol Evol1976, 7:101–104.

56. Egel R: Peptide-dominated membranes preceding the genetic takeoverby RNA: latest thinking on a classic controversy. Bioessays 2009,31:1100–1109.

http://dx.doi.org/10.1101/cshperspect.a003541



57. Kurland CG: The RNA dreamtime: modern cells feature proteins thatmight have supported a prebiotic polypeptide world but nothingindicates that RNA world ever was. Bioessays 2010, 32:866–871.

58. Caetano-Anollés D, Kim KM, Mittenthal JE, Caetano-Anollés G: Proteomeevolution and the metabolic origins of translation and cellular life. J MolEvol 2011, 72:14–33.

59. Caetano-Anollés G, Kim KM, Caetano-Anollés D: The phylogenomic rootsof modern biochemistry: origins of proteins, cofactors and proteinbiosynthesis. J Mol Evol 2012, 74:1–34.

60. Harish A, Caetano-Anollés G: Ribosomal history reveals origins of modernprotein synthesis. PLoS One 2012, 7:e32776.

61. Study of ribosome evolution challenges RNA world hypothesis. University ofIllinois News Bureau. http://news.illinois.edu/news/12/0312ribosome_GustavoCaetano-Anolles.html.

62. Poole AM: On alternative biological scenarios for the evolutionarytransitions to DNA and biological protein synthesis. In Origins of Life: ThePrimal Self-Organization. Edited by Egel R, et al. Berlin Heidelberg: Springer;2011:209–223. (Part 4) doi: 10.1007/978-3-642-21625-1_10.

63. Weiss R, Cherry J: Speculations on the origin of ribosomal translocation.In The RNA World. Edited by Gesteland RF, Atkins JF. Cold Spring Harbor, NY:Cold Spring Harbour Press; 1993:71–89.

64. Gordon KHJ: Were RNA replication and translation directly coupled in theRNA (+ protein?) world? J Theor Biol 1995, 173:179–193.

65. Poole AM, Jeffares DC, Penny D: The path from the RNA world. J Mol Evol1998, 46:1–17.

66. Penny D: An interpretive review of the origin of life research. Biol Philos2005, 20:633–671.

67. Wolf YI, Koonin EV: On the origin of the translation system and thegenetic code in the RNA world by means of natural selection,exaptation, and subfunctionalization. Biol Direct 2007, 2:14.

68. Manrubia SC, Briones C: Modular evolution and increase of functionalcomplexity in replicating RNA molecules. RNA 2007, 13:97–107.

69. Briones C, Stich M, Manrubia SC: The dawn of the RNA World: towardfunctional complexity through ligation of random RNA oligomers. RNA2009, 15:743–749.

70. Purohit P, Stern S: Interactions of a small RNA with antibiotic and RNAligands of the 30 S subunit. Nature 1994, 370:659–662.

71. Ogle JM, Carter AP, Ramakrishnan V: Insights into the decoding mechanismfrom recent ribosome structures. Trends Biochem Sci 2003, 28:259–266.

72. Hiller DA, Singh V, Zhong M, Strobel SA: A two-step chemicalmechanism for ribosome-catalysed peptide bond formation. Nature2011, 476:236–239.

73. Weiner AM, Maizels N: tRNA-like structures tag the 3´ ends of genomicRNA molecules for replication: Implications for the origin of proteinsynthesis. Proc Natl Acad Sci USA 1987, 84:7383–7387.

74. Sun FJ, Caetano-Anollés G: The origin and evolution of tRNA inferred romphylogenetic analysis of structure. J Mol Evol 2008, 66:21–35.

75. Sun FJ, Caetano-Anollés G: Transfer RNA and the origins of diversified life.Sci Prog 2008, 91:265–284.

76. Noller HF: On the origin of the ribosome: coevolution of subdomains oftRNA and rRNA. In The RNA World. Edited by Gesteland RF, Atkins JF. ColdSpring Harbor, NY: Cold Spring Harbour Press; 1993:137–156.

77. Bokov K, Steinberg SV: A hierarchical model for evolution of 23 Sribosomal RNA. Nature 2009, 457:977–980.

78. Maizels N, Weiner AM: The genomic tag hypothesis: modern viruses asmolecular fossils of ancient strategies for genomic replication. In The RNAWorld. Edited by Gesteland RF, Atkins JF. Cold Spring Harbor, NY: ColdSpring Harbour Press; 1993:577–602.

79. Sun FJ, Caetano-Anollés G: Evolutionary patterns in the sequence andstructure of transfer RNA: a window into early translation and thegenetic code. PLoS One 2008, 3:e2799.

80. Sun FJ, Caetano-Anollés G: Evolutionary patterns in the sequence andstructure of transfer RNA: early origins of archaea and viruses. PLoSComput Biol 2008, 4:e1000018.

81. Trifonov EN: The triplet code from first principles. J Biomol Struct Dyn2004, 22:1–11.

82. Marck C, Grosjean H: tRNomics: analysis of tRNA genes from 50 genomesof Eukarya, Archaea, and Bacteria reveals anticodon-sparing strategiesand domain-specific features. RNA 2002, 8:1189–1232.

83. Di Giulio M: The non-monophyletic origin of the tRNA molecule and theorigin of genes only after the evolutionary stage of the last universalcommon ancestor (LUCA). J Theor Biol 2006, 240:343-352.

84. Poole A, Jeffares D, Penny D: Early evolution: prokaryotes, the new kidson the block. Bioessays 1999, 21:880–889.

85. Sinclair R: A quantitative approach to investigating the hypothesis ofprokaryotic intron loss. Available from Nature Proceedings. 2011. http://hdl.handle.net/10101/npre.2011.5770.1.

86. Yarus M: Getting past the RNA world: the initial Darwinian ancestor. ColdSpring Harb Perspect Biol 2011, 3:4. doi:10.1101/cshperspect.a003590. pii:a003590.

87. de Duve C: Blueprint for a Cell: The Nature and Origin of Life. Burlington,North Carolina: Neil Patterson Publishers, Carolina Biological SupplyCompany; 1991:179.

88. Su AAH, Randau L: A-to-I and C-to-U editing within transfer RNAs.Biochemistry (Moscow) 2011, 76:1142–1148.

89. Paris Z, Fleming IMC, Alfonzo JD: Determinants of tRNA editing andmodification: avoiding conundrums, affecting function. Sem Cell DevelBiol 2012, 23:269–274.

90. Randau L, Stanley BJ, Kohlway A, Mechta S, Xiong Y, Soll D: A cytidinedeaminase edits C to U in transfer RNAs in archaea. Science 2009,324:657–659.

91. Koonin EV: The Logic of Chance: The Nature and Origin of Biological Evolution.Upper Saddle River, NJ: FT press; 2011.

92. Aravind L, Mazumder R, Vasudevan S, Koonin EV: Trends in proteinevolution inferred from sequence and structure analysis. Curr Opin StructBiol 2002, 12:392–399.

doi:10.1186/1745-6150-7-23Cite this article as: Bernhardt: The RNA world hypothesis: the worsttheory of the early evolution of life (except for all the others)a. BiologyDirect 2012 7:23.

Submit your next manuscript to BioMed Centraland take full advantage of:

• Convenient online submission

• Thorough peer review

• No space constraints or color figure charges

• Immediate publication on acceptance

• Inclusion in PubMed, CAS, Scopus and Google Scholar

• Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit

http://news.illinois.edu/news/12/0312ribosome_GustavoCaetano-Anolles.html

http://news.illinois.edu/news/12/0312ribosome_GustavoCaetano-Anolles.html

http://dx.doi.org/10.1007/978-3-642-21625-1_10

http://hdl.handle.net/10101/npre.2011.5770.1

http://hdl.handle.net/10101/npre.2011.5770.1


IUBMB Life, 49: 173–176, 2000Copyright c° 2000 IUBMB1521-6543/00 $12.00 + .00

Hypothesis Paper

A Replicator Was Not Involved in the Origin of Life

Robert ShapiroDepartment of Chemistry, New York University, New York, NY 10003

Summary

Many scienti� c theories of the origin of life suggest that life be-gan with the spontaneous formation of a replicator (a self-copyingorganic polymer) within an unorganized chemical mixture, or“soup.” A profound dif� culty exists, however, with the idea of RNA,or any other replicator, at the start of life. Existing replicators canserve as templates for the synthesis of additional copies of them-selves, but this device cannot be used for the preparation of the very� rst such molecule, which must arise spontaneously from an un-organized mixture. The formation of an information-bearing ho-mopolymer through undirected chemical synthesis appears veryimprobable. The dif� culties involved in such a synthesis are illus-trated by considering the prospects for the assembly of a polypep-tide of L-®-amino acids, based on the contents of the Murchisonmeteorite as an example of a mixture of abiotic origin. In that mix-ture, potential replicator components would be accompanied bya host of interfering substances, which include chain terminators(simple carboxylic acids and amines), branch-formers, D-aminoacids, and many classes of substances for which incorporationwould disrupt the necessary structural regularity of the replicator.Laboratory experiments dealing with the nonenzymatic synthesisof biopolymers have not addressed the speci� city problem. Thepossibility that formation of the � rst replicator took place througha very improbable event cannot be excluded, but greater atten-tion should be given to metabolism-� rst theories, which avoid thisdif� culty.

IUBMB Life, 49: 173–176, 2000

Keywords Minerals; Murchison meteorite; origin-of-life; polypep-tide; replicator; RNA world; speci� city.

INTRODUCTIONMany scienti� c theories of the origin of life suggest that life

began with the spontaneous formation of a replicator within anunorganized chemical mixture, or “soup.” The term “replica-tor” will be used here to represent a self-copying organic poly-mer that contains a repetitious backbone and attached, variable,

Received and accepted 24 January 2000.Address correspondence to Prof. Robert Shapiro, Department of

Chemistry, New York University, 100 Washington Sq. East, New York,NY 10003. Fax: (212) 260-7905. E-mail: [email protected]

information-bearing subunits. DNA and RNA are examples ofsuch molecules that function within life today. Two-dimensionalinformation storage within minerals as proposed by Cairns-Smith (1) is suf� ciently distinct to be considered an alternativeproposal for life’s origins.

The roots of the replicator idea run deep (see Muller [2] andthe references cited therein); the following eloquent statementof this concept has been supplied by Dawkins (3):

Processes analogous to thesemust havegiven rise to the “primevalsoup” that biologists and chemists believe constituted the seas somethree to four thousand million years ago. The organic substances be-came locally concentrated , perhaps in drying scum round the shoresor in tiny suspended droplets. Under the further in� uence of energy,such as ultraviolet light from the sun, they combined into largermolecules . . . in those days large organic molecules could drift un-molested through the thickening broth. At some point a particularlyremarkable molecule was formed by accident. We will call it theReplicator. It may not necessarily have been the biggest or mostcomplex molecule around, but it had the extraordinary property ofbeing able to create copies of itself.

The idea of a replicator at the start of life has embedded itselfsuf� ciently in scienti� c consciousness that it has entered manyde� nitions of life, for example, that of Muller (2):

The “stripped down” de� nition of a living thing offered heremay be paraphrased : that which possesses the potential of evolvingby natural selection . . .: The gene material alone, of natural materials,possesses these faculties, and it is therefore legitimate to call it theliving material, the present-day representative of � rst life.

Such thinking has received wide endorsement, and a recent state-ment of this type has been termed the “NASA de� nition oflife” (4).

The discovery of the catalytic capability of RNA gave freshimpetus to these ideas. In particular, the “RNA world” theory(5, 6) postulated that RNA served as both genetic material andcatalyst in the � rst biosphere. In the past years, this hypothesishas been widely distributed by the media (7 ) and by textbooks.However, there is a profound dif� culty with the idea of RNA,or more generally, any organic polymer replicator, at the startof life. The argument of the present discussion will be that thedif� culty is severe enough to warrant the tentative acceptance

173

174 SHAPIRO

of the opposite hypothesis: that life began without such a repli-cator, and that RNA and DNA were introduced after a period ofevolution.

THE PROBLEM OF THE FIRST REPLICATORExisting replicators can serve as templates for the synthesis

of additional copies of themselves, but this device cannot beused for preparation of the very � rst such molecule, which mustarise spontaneously from an unorganized mixture. In the spe-ci� c case of RNA, the unavailability and instability of its com-ponents (8–13) have led some scientists to conclude that “thede novo appearance of oligonucleotides on the primitive earthwould have been a near miracle” (14). They have suggested thatRNA was preceded by a simpler or more accessible replica-tor (15), which functioned as the original hereditary material.Generally, a length of 30–60 residues has been assumed nec-essary to get a self-replicating system started (16). Among thecandidates suggested have been proteins, peptide nucleic acid(PNA), polynucleotides based on the pyranosyl analog of ribose(p-RNA), and polymers based on pyrophosphonates , hydroxyacids, glycol and glycerol phosphates, aminoaldehydes, dithiols,and others (14). The monomers components of some of thesereplacement replicators are more plausible than nucleotides ascomponents of prebiotic mixtures, but a severe problem remainsin connecting them.

A uniform secondary structure, such as a double helix, ap-pears necessary if an organic polymer is to carry out its templat-ing function according to a consistent scheme. Such a structurerequires a backbone of a repetitive nature, with no or few in-terruptions, and the replicators proposed above have met thisrequirement. Little attention has been paid to the circumstancesin which such a polymer could be obtained, however. The im-plicit assumption has been that monomers of a single chemicaltype would seek each other out in a prebiotic mixture and com-bine exclusively with one another. No theoretical or experimen-tal basis has been put forward to support such an assumption,however, and considerations of entropy would lead in the op-posite direction: The components of a mixture should combinehaphazardly, producing chaotic polymers.

To illustrate the dif� culties involved in constructing a replica-tor, let us examine the contents of a mixture produced by abioticorganic synthesis. Laboratory simulations of such mixtures mayvary according to the assumptions of the investigators, but na-ture has provided us with authentic and ancient examples of abi-otic organic synthesis: the meteorites. The Murchison meteorite,which has been subjected to extensive analysis (for reviews, see17 and 18), will serve as the basis of this discussion. Other mete-orite results are similar in their general outlook, although minordetails may differ (see, for example, 19).

Overall, the contents of the Murchison are quite complex,with many chemical classes represented. Within a class, all iso-mers of a given carbon number can be detected, if only a fewcarbons are present. The quantity of material of a given carbonnumber declines logarithmically with the number of carbons,however.

COULD A REPLICATOR FORM FROM AMURCHISON MIXTURE?

Orgel has written (20):

The subunits of an informational polymer are likely to be trifunc-tional—two functionalities are needed to hold the backbone together,and a third is needed to provide for speci� c interchain interaction.The prebiotic production of a particular trifunctional molecule maywell be accompanied by the production of numerous isomers andclosely related molecules.

The Murchison meteorite mixture contains three monomergroups that could self-combine according to this prescription:A suite of 80 amino acids can be found there, as well as 51 hy-droxy acids and > 50 hydroxydicarboxylic acids. Selected mem-bers of any of these groups could be linked together, in theory,to form an information-bearing homopolymer. (Other combina-tions involving alternating subunits are possible; dicarboxylicacids could be linked together by using a diol such as dihy-droxyacetone at every other position, for example. Such combi-nations appear less plausible than simple homopolymerization,however.) Here I will analyze only one possibility, that of theL- a -amino acids, because these occur naturally and polymers ofthis type have been suggested as possible replicators (21, 22).

Because the fraction of all peptide sequences that could func-tion as a replicator cannot be estimated readily, I will focus onlyon the likelihood that any linear combination of 30 to 60 suchresidues could arise spontaneously by dehydration of a mixtureresembling the components of the Murchison meteorite. Thefollowing considerations arise:

1. Availability of suitable components. The total concentra-tion of amino acids of all kinds in the Murchison meteoriteapproaches 700 nmol/g, with glycine (98 nmol/g) andracemic a -aminoisobutyric acid (117 nmol/g) as the mostprominent members. The latter is not utilized in proteins inlife today, but there is no reason to believe that it could notfunction in a hypothetical protein replicator. More seriousis the fact that a -amino acids, with the notable exceptionof glycine, will be present as racemates (I set aside the un-resolved question of possible enantiomeric excess). Thechance of obtaining a polymer of 30 L-residues by randomcombination of a -amino acids will be 1/230, or about 1 in109. However, other substances will also be present in themixture, which would further interfere with polypeptideformation.

2. The termination problem. The concentration of mono-functional carboxylic acids in Murchison is 10 to 20 timesthat of the amino acids (17 ); such carboxylic acids wouldbe expected to terminate most chains at the amino end un-der polymerization conditions. Acetyl amino acids havein fact been isolated from Murchison (23). Alkylsulfonicacids, which rival amino acids in concentration (18), couldalso terminate chains. Termination at both ends would beneeded, of course, to end the polymerization, but abun-dant C-terminators are also present, for example, ammo-nia (1100 nmol/g), methylamine (71 nmol/g), and otheramines.

NO REPLICATOR IN THE ORIGIN OF LIFE 175

3. Interruption of backbone regularity. Many substances ex-ist in the Murchison mixture that, although they couldallow polymerization to continue, might disrupt a helix,sheet, or other regular polypeptide structure. The list of in-terfering agents includes b - and c -amino acids, hydroxy-acids, dicarboxylic acids, and hydroxydicarboxylic acids.Many of these substances are present in concentrations ri-valing those of the amino acids. Dicarboxylic acids couldintroduce another complication by linking two peptidechains through their N-terminals.

4. Branching. Aspartic and glutamic acids, which are amongthe set of 20 used in proteins, have been detected in theMurchison meteorite in concentrations of 5 and 18 nmol/g,respectively. Their abiotic inclusion in a polymer chain,however, could involve either carboxyl group or both ofthem. Incorporation of the more distant (from the amino)carboxyl group into the polymer structure would disruptthe regularity of the backbone, whereas incorporation ofboth would afford a branched structure. Cross-linked poly-peptides apparently are produced upon the heating ofamino acid mixtures containing aspartic and glutamicacids. Such structures are incompatible with our currentconcepts of replicator function as we know it. Alternativeroutes to prebiotic polypeptide formation have been pro-posed that do not involve dehydration (25), but it is notclear why these routes should not run into the same dif-� culties of heterochirality, chain termination, backboneinterruption, and branching already discussed.

Similar arguments could be made concerning the obstacles tothe spontaneous synthesis of other polymers that have been sug-gested for replicator status. The additional dif� culties that comeup in the case of RNA have been discussed elsewhere (8–13). In-terestingly, alkyl phosphonates have been the only organophos -phorus compounds isolated thus far from the Murchison mete-orite (18).

SOLUTIONS TO SPECIFICITY PROBLEMSIN LIVING ORGANISMS

Biopolymers of uniform backbone are produced routinely inliving cells, of course. Their production is mediated by enzymes,which ensure that only the desired monomers are incorporated.The polymerases involved in DNA replication, for example,utilize an intricate structure involving “thumb,” “� ngers,” and“palm” regions to recognize an incoming triphosphate (26). Ad-ditional proofreading functions help to correct incorporation er-rors, and additional repair systemsact to eliminate surviving mis-takes. Less recognized is the contribution of cellular metabolismto � delity, in offering only a limited number of activated sub-strates for incorporation. When this safeguard is bypassed by theintroduction of arti� cial substrates, striking errors can be made,for example, the incorporation of 2,4-di� uorotoluene oppositeadenine (27 ), or the placement of a pyrene residue opposite anabasic site (28). A plausible replicator-based theory of the ori-gin of life must provide a mechanism by which speci� city in

polymer assembly could be attained without the support of pre-existing biocatalysts and metabolism. No such mechanism forspeci� city has yet been demonstrated.

EXPERIMENTAL NONENZYMATIC PREPARATIONOF BIOPOLYMERS

Numerous experiments have been directed toward the temp-late-free “prebiotic” preparation of a potential replicator, someof which have drawn considerable attention. In a review of onestriking accomplishment (29), von Kiedrowski commented (30):

Ferris et al. provide evidence that longer oligonucleotides andpeptides can be obtained if the polycondensatio n takes place on min-eral surface. . . This result is remarkable because it indicates that theclay surface may promote the formation of oligonucleotides up tothe length of small ribozymes. The latter are being discussed in thecontext of the “RNA world” . . .

The experiments described above, and others (see 16, 29,31, and the references cited therein), have demonstrated that inthe presence of minerals such as montmorillonite (for oligori-bonucleotides) and illite or hydroxylapatite (for polypeptides),activated monomers could be condensed to oligomers as longas 55-mers. Those studies were addressed to the question ofwhether oligomerization in aqueous solution could compete withhydrolysis; they answered it in an elegant manner. They did not,however, deal with the speci� city problem discussed here, po-tentially interfering substances having been excluded by the in-vestigators. In the absence of any competition studies, there is noreason to believe that a given mineral might preferentially adsorband combine the monomers that would be useful in constructinga particular replicator, while excluding the much greater numberthat would disrupt such a function.

Apart from this, the experiments were conducted in ways thatappear unlikely to take place outside a laboratory. The monomerswere speci� cally activated by procedures involving either themultistep preparation of nucleoside 5 0 -phosphorimidazolides(for nucleotide oligomerization) or the prior treatment of mono-mers with synthetic reagents such as carbodiimides or carbonyldiimidazole (for amino acids). Fresh batches of activated mono-mer were added to the given mineral according to a well-de� nedfeeding schedule. The relevence of these results to events on theprebiotic Earth is therefore questionable. The � nding that cer-tain minerals have a tendency to accumulate polymers is striking,however, and it would be interesting to learn whether they ac-tually do so outside the laboratory. Montmorillonite and illiteare found in a diverse variety of natural circumstances today. Insome cases, they may have been exposed repeatedly to chemi-cals that are not readily degraded. It would be interesting to learnwhether these minerals have assembled polymers, particularlyhomopolymers, of any type.

THE ROLE OF CHANCEArguments of the type presented here cannot disprove the

replicator hypothesis, but only render it improbable. Despite thedif� culties discussed above, circumstances in particular loca-tions may have improved the chances of replicator formation.

176 SHAPIRO

A more restricted set of components than that present in theMurchison meteorite may have been generated at certain sites,and by chance, a few of them may have had enriched concen-trations of replicator subunits. Also by chance, such a site mayhave contained a mineral or other natural catalyst that favoredpolymerization of those same components. After a large numberof trials, a replicator may have arisen in such an environment.This appeal to chance has been made, implicitly or explicitly,many times in origin-of-life literature and done so eloquently byWald (32). On the early Earth, the production of an information-bearing homopolymer within a complex mixture by chance can-not be excluded, but if such an event was required to start life,then the origin would have been an extremely improbable acci-dent, and prospects for life elsewhere would be diminished. Aproposal of this type violates no natural laws but is less to bepreferred in scienti� c methodology than one that hypothesizesa more probable origin for life.

ALTERNATIVES TO THE REPLICATOR THEORYSeveral scientists have put forth theories that do not require

an ordered polymeric replicator at the start of life. They pro-pose, instead, that life began with a mutually sustaining set ofcatalytic reactions involving smaller molecules (see, for exam-ple, 33–35). Such theories provide a robust alternative to ideasbased on a replicator. The details can differ; for example, thereaction set might be carried out on a mineral surface (36), orwithin a membrane-bound compartment (37–39). Insuf� cientexperimental attention has been given to such ideas, but if thehypothesis presented here is accepted, perhaps they will moveto the forefront of origin-of-life research.

REFERENCES1. Cairns-Smith, A. G. (1982) Genetic Takeover and the Mineral Origins of

Life. Cambridge University Press, New York.2. Muller, H. J. (1966) The gene material as the initiator and the organizing

basis of life. Am. Nat. 100, 493–517.3. Dawkins, R. (1976) The Sel� sh Gene. p. 16, Oxford University Press, New

York.4. Luisi, P. L. (1998) About various de� nitions of life. Origins Life Evol.

Biosphere 28, 613–622.5. Gilbert, W. (1986) The RNA world. Nature 319, 618.6. Yarus, M. (1999) Boundaries for an RNA world. Curr. Opin. Chem. Biol.

3, 260–267.7. Wade, N. (1999) Inside the cell, experts see life’s origin. The New York

Times April 6, F1.8. Shapiro, R. (1984) The improbability of prebiotic nucleic acid synthesis.

Origins Life 14, 565–570.9. Shapiro, R. (1988) Prebiotic ribose synthesis: A critical analysis. Origins

Life Evol. Biosphere 18, 71–85.10. Shapiro, R. (1995) The prebiotic role of adenine: A critical analysis. Origins

Life Evol. Biosphere 25, 83–98.11. Larralde, R., Robertson, M. P., and Miller, S. L. (1995) Rates of decompo-

sition of ribose and other sugars; implications for chemical evolution. Proc.Natl. Acad. Sci. U.S.A. 92, 8158–8160.

12. Levy, M., and Miller, S. L. (1998) The stability of the RNA bases: Implica-tions for the origin of life. Proc. Natl. Acad. Sci. U.S.A. 95, 7933–7938.

13. Shapiro, R. (1999) Prebiotic cytosine synthesis: A critical analysis. Impli-cations for the origin of life. Proc. Natl. Acad. Sci. U.S.A. 96, 4396–4401.

14. Joyce, G. F., and Orgel, L. E. (1999) Prospects for understanding the originof the RNA world. In The RNA World, 2nd ed. (Gesteland, R. F., Cech,T. R., and Atkins, J. F., eds.). pp. 49–77, Cold Spring Harbor LaboratoryPress, Cold Spring Harbor, NY.

15. Joyce, G. F., Schwartz, A. W., Miller, S. L., and Orgel, L. E. (1987) Thecase for an ancestral genetic system involving simple analogues of the nu-cleotides. Proc. Natl. Acad. Sci. U.S.A. 84, 4398–4402.

16. Orgel, L. E. (1998) Polymerization on the rocks: Theoretical introduction.Origins Life Evol. Biosphere 28, 227–234.

17. Cronin, J. R., Pizzarello, S., and Cruikshank, D. P. (1988) Organic matterin carbonaceou s chondrites, planetary satellites, asteroids and comets. InMeteorites and the Early Solar System (Kerridge, J. S., and Matthews,M. S., eds.). pp. 819–857, University of Arizona Press, Tucson.

18. Cronin, J. R., and Chang, S. (1993) Organic matter in meteorites: Molecularand isotopic analyses of the Murchison meteorite. In The Chemistry of Life’sOrigins (Greenberg, J. M., Mendoza-Gomez, C. X., and Piranello, V., eds.).pp. 209–258, Kluwer Academic Publishers, Dordrecht, The Netherlands.

19. Naraoaka, H., Shimoyama, A., and Harada, K. (1999) Molecular distri-bution of monocarboxyli c acids in Asuka carbonaceou s chondrites fromAntarctica. Origins Life Evol. Biosphere 29, 187–201.

20. Orgel, L. E. (1992) Molecular replication. Nature 358, 203–209.21. Lee, D. H., Granja, J. R., Martinez, K. S., and Ghadiri, M. R. (1996) A

self-replicating peptide. Nature 382, 525–527.22. Yao, S., Ghosh, I., Zutshi, R., and Chmielewski, J. (1998). Selective ampli-

� cation by auto- and cross-catalysis in a replicating peptide system. Nature396, 447–450.

23. Pizzarello, S., Feng, F., Epstein, F., and Cronin, J. R. (1994) Isotopic anal-yses of nitrogenous compounds from the Murchison meteorite: Ammonia,amines, amino acids and polar hydrocarbons . Geochim. Cosmochim. Acta58, 5579–5587.

24. Lyons, J. R., and Vasavada, A. R. (1999) Flash heating on the early Earth.Origins Life Evol. Biosphere 29, 123–138.

25. Minard, R. D., Hatcher, P. G., Gourley, R. C., and Matthews, C. N. (1998)Structural investigations of hydrogen cyanide polymers: New insights usingTMAH thermochemolysis /GC-MS. Origins Life Evol. Biosphere 28, 461–473.

26. Steitz, T. A. (1999) DNA polymerases: Structural diversity and commonmechanisms. J. Biol. Chem. 274, 17395–17398.

27. Moran, S., Ren, R. X.-F., and Kool, E. (1997) A thymidine triphosphateshape analog lacking Watson–Crick pairing ability is replicated with highsequence selectivity. Proc. Natl. Acad. Sci. U.S.A. 94, 10506–10511.

28. Murray, T. J., and Kool, E. T. (1998) Selective and stable DNA base pairingwithout hydrogen bonds. J. Am. Chem. Soc. 120, 6191–6192.

29. Ferris, J. P., Hill, A. R. Jr., Liu, R., and Orgel, L. E. (1996) Synthesis oflong prebiotic oligomers on mineral surfaces. Nature 381, 59–61.

30. von Kiedrowski , G. (1996) Primordial soup or crepes? Nature 381, 20–

21.31. Hill, A. R. Jr., Bohler, C., and Orgel, L. E., (1998) Polymerization on the

Rocks: Negatively-charge d amino acids. Origins Life Evol. Biosphere 28,225–243.

32. Wald, G. (1954) The origin of life. Scienti� c American 191, 144–153.33. Kauffman, S. (1993) The Origins of Order. Self-Organization and Selection

in Evolution. Oxford University Press, New York.34. Dyson, F. (1985) Origins of Life (Cambridge Univ. Press, Cambridge).35. De Duve, C. (1992) The thioester world. In Frontiers of Life (TranThan

Van, J., TranThan Van, K., Mounolou, J. C., Schneider, J., and McKay, C.,eds.). pp. 1–20, Editions Frontieres, Gif-sur-Yvette, France.

36. Wachtershauser, G. (1992) Groundworks for an evolutionary biochemistry:The iron–sulfur world. Prog. Biophys. Mol. Biol. 58, 85–201.

37. Morowitz, H. J. (1992) Beginnings of Cellular Life. Yale University Press,New Haven.

38. Deamer, D. (1997) The � rst living systems: A bioenergetic perspective.Microbiol. Mol. Biol. Rev. 61, 239–261.

39. Segre, D., Pilpel, Y., and Lancet, D. (1998) Mutual catalysis in sets ofprebiotic organicmolecules: Evolution through simulated chemical kinetics.Physica A (Amsterdam) 249, 558–564.

Review ArticleOrigins & Design 17:1

Gordon C. MillsDepartment of Human Biological Chemistry and GeneticsUniversity of Texas Medical BranchGalveston, TX 77555

Dean KenyonDepartment of BiologySan Francisco State University1600 Holloway AvenueSan Francisco, CA 94132

Introduction

One of the earliest published suggestions that RNA-catalyzed RNA replication preceded and gave rise to thefirst DNA-based living cells was made by Carl Woese in 1967, in his book The Genetic Code1. Similarsuggestions were made by Crick and Orgel2, for reasons that are not difficult to grasp. Prior to the discoveryof catalytic RNAs, proteins were considered by many to be the only organic molecules in living matter thatcould function as catalysts. DNA carries the genetic information required for the synthesis of proteins. Thereplication and transcription of DNA require a complex set of enzymes and other proteins. How then couldthe first living cells with DNA-based molecular biology have originated by spontaneous chemical processeson the prebiotic Earth? Primordial DNA synthesis would have required the presence of specific enzymes, buthow could these enzymes be synthesized without the genetic information in DNA and without RNA fortranslating that information into the amino acid sequence of the protein enzymes? In other words, proteins arerequired for DNA synthesis and DNA is required for protein synthesis.

This classic "chicken-and-egg" problem made it immensely difficult to conceive of any plausible prebioticchemical pathway to the molecular biological system. Certainly no such chemical pathway had beendemonstrated experimentally by the early 1960s. So the suggestion that RNA molecules might have formedthe first self-replicating chemical systems on the primitive Earth seemed a natural one, given the uniqueproperties of these substances.

They carry genetic information and (unlike DNA) occur primarily as single-stranded molecules that canassume a great variety of tertiary structures, and might therefore be capable of catalysis, in a manner similarto that of proteins. The problem of which came first, DNA or proteins, would then be resolved.

Self-replicating RNA-based systems would have arisen first, and DNA and proteins would have been addedlater. But in the absence of any direct demonstration of RNA catalysis, this suggestion remained only aninteresting possibility.

Then, in the early 1980s3, the discovery of self-splicing, catalytic RNA molecules (in the ciliated protozoanTetrahymena thermophila), put molecular flesh on the speculative bones of the idea of an early evolutionarystage dominated by RNA. These catalytic RNA molecules have subsequently been termed "ribozymes." "Onecan contemplate an RNA World," wrote Walter Gilbert in 1986, "containing only RNA molecules that serve

The RNA World: A Critique - Origins & Design 17:1. Mills, Gordon and... http://www.arn.org/docs/odesign/od171/rnaworld171.htm

1 of 8 12/31/2014 3:38 PM

to catalyze the synthesis of themselves."4

The phrase "RNA World" stuck to the general hypothesis, and has since come to denote the RNA-first,DNA-and-proteins-later scenario depicted in Figure 1. The long-standing "chicken-and egg" puzzle at theorigin of life indeed appeared amenable to a solution:

The primordial...conundrum -- which came first, informational polynucleotides or functionalpolypeptides -- was obviated by the simple but elegant compaction of both genetic informationand catalytic function into the same molecule.5

A second impetus to the RNA world hypothesis came from the cluster of technical innovations now knowngenerally as ribozyme engineering. Naturally occuring RNA catalytic activities are actually restricted to asmall set of highly specialized reactions, e.g., the processing of RNA transcripts primarily in eukaryotic cells.However, ribozyme engineering, made possible by techniques such as DNA sequencing, in vitro transcriptionand the polymerase chain reaction [PCR]6, allow molecular biologists to manipulate RNA to whatever extentthe molecule will allow. Thus, the catalytic repertoire of RNA can be expanded beyond the naturallyoccurring activities -- in the main, by two broad strategies of ribozyme engineering.

One strategy involves the direct modification of existing species of ribozymes, to produce better or evennovel catalysts. This has been called the "rational design" approach. The other strategy employs pools of short(often 50-100 nucleotide units) randomized RNA molecules, which are subjected repeatedly to a selectionprocess designed to enhance the concentration of RNA molecules with the desired functional activity. Thefew selected molecules are then multiplied a million-fold or more by using the polymerase chain reaction,which uses activated nucleotide precursors and enzymes. This has been termed the "irrational design"method.

Judging from the progress in ribozyme engineering in recent years, it seems likely that new and improvedtypes of RNA catalysts will be produced in years ahead. Moreover, molecular biologists may discoveradditional catalytic roles of RNA in living cells, although the variety of such roles is not expected to rival thatof the protein enzymes. Thus, one might expect that the RNA World hypothesis will continue to havesupporters.

Yet beyond the immediate foreground of RNA World excitement lies a disquieting landscape of chemicalproblems, largely ignored in the recent literature on ribozyme engineering. As researchers broaden their focusto include the chemical plausibility of the RNA World itself, however7, these difficulties cannot be avoided.

Furthermore, the relevance of ribozyme engineering to naturalistic theories of the origin of life is doubtful atbest, primarily because of the necessity for intelligent intervention in the synthesis of the randomized RNA;then again in the selection of a few functional RNA molecules out of that mixture; then, finally, in theamplification of those few functional RNA molecules [see box, "What Do Ribozyme EngineeringExperiments Really Tell Us About the Origin of Life?"].

Hubert Yockey, borrowing a metaphor from Jonathan Swift, suggests that current origin-of-life research,including the RNA World hypothesis, floats improbably in mid-air like the roof of a house built by anarchitect of the Grand Academy of Lagado. This savant had contrived a method of building houses bybeginning at the roof and working downwards. "The architect pointed out that among the advantages of thisprocedure," Yockey notes8, "was that once the roof was in place [before the walls or foundation] the rest ofthe construction could proceed quickly and without interruption by weather." That "roof" -- consisting in thisinstance of tiles which represent the catalytic activities of RNA -- may look solid to those believers in the


2 of 8 12/31/2014 3:38 PM

existence of a prebiotic RNA World. But is the roof really solid? Is it supported by walls and a foundation?

Once one peers over the edge of the roof to look beneath, we shall argue, the implausibility of the theoreticalstructure as a whole is inescapable. In what follows, we present the key postulates or presuppositions onwhich the RNA World hypothesis must rest (see Figure 2). Each represents an unsolved chemical problem, inevery case well-known to origin-of-life researchers. Unfortunately, in many articles on the RNA World, theseproblems are often collapsed into the "prebiotic soup" and "self-assembly" phases of the scenario, and receiveno discussion. We suggest that new discoveries about the catalytic activities of RNA should be seen for whatthey really are: not elucidating prebiotic processes on the early Earth, but rather as extending our knowledgeof the molecular biology of the cell in important ways (see below).

The relevance of catalytic RNA to the problem of the naturalistic origin of life is, however, a different matterentirely.

We take heart in noting that, despite the frequent neglect in much of the popular literature of the chemicaldifficulties of the RNA World scenario, many of the scientists involved with that hypothesis are quite candidin their assessment of the problems associated with it. These are represented for instance by the numerouscontributors to The RNA World7. Since the RNA World hypotheses are so broad, we will attempt to breakthem down into somewhat narrower postulates. In this way one may see more clearly some of thepresuppositions that are involved.

Problematic Chemical Postulates of the RNA World Scenario

Postulate 1: There was a prebiotic pool of beta-D-ribonucleotides.

Beta-D-ribonucleotides (see Figure 2) are compounds made up of a purine (adenine or guanine) or apyrimidine (uracil or cytosine) linked to the 1'-position of ribose in the beta-configuration.

There is, in addition, a phosphate group attached to the 5'-position of the ribose. For the four differentribonucleotides in this prebiotic scenario, there would be hundreds of other possible isomers.

But each of these four ribonucleotides is built up of three components: a purine or pyrimidine, a sugar(ribose), and phosphate. It is highly unlikely that any of the necessary subunits would have accumulated inany more than trace amounts on the primitive Earth. Consider ribose. The proposed prebiotic pathway leadingto this sugar, the formose reaction, is especially problematic9. If various nitrogenous substances thought tohave been present in the primitive ocean are included in the reaction mixture, the reaction would not proceed.The nitrogenous substances react with formaldehyde, the intermediates in the pathways to sugars, and withsugars themselves to form non-biological materials10. Furthermore, as Stanley Miller and his colleaguesrecently reported, "ribose and other sugars have suprisingly short half-lives for decomposition at neutral pH,making it very unlikely that sugars were available as prebiotic reagents."11

Or consider adenine. Reaction pathways proposed for the prebiotic synthesis of this building block start withHCN in alkaline (pH 9.2) solutions of NH4OH.12 These reactions give small yields of adenine (e.g., 0.04%)and other nitrogenous bases provided the HCN concentration is greater than 0.01 M. However, the reactionmixtures contain a great variety of nitrogenous substances that would interfere with the formose reaction.Therefore, the conditions proposed for the prebiotic synthesis of purines and pyrimidines are clearlyincompatible with those proposed for the synthesis of ribose. Moreover, adenine is susceptible to deaminationand ring-opening reactions (with half-lives of about 80 years and 200 years respectively at 37º C and neutral


3 of 8 12/31/2014 3:38 PM

pH), making its prebiotic accumulation highly improbable13. This makes it difficult to see how anyappreciable quantities of nucleosides and nucleotides could have accumulated on the primitive Earth. If thekey components of nucleotides (the correct purines and pyrimidines, ribose, and phosphate) were not present,the possibility of obtaining a pool of the four beta-D-ribonucleotides with correct linkages would be remoteindeed.

If this postulate, the first and most crucial assumption, is not valid, however, then the entire hypothesis of anRNA World formed by natural processes becomes meaningless.

Postulate 2: Beta-D ribonucleotides spontaneously form polymers linked together by 3',5'-phosphodiester linkages (i.e., they link to form molecules of RNA; see figure 2).

Joyce and Orgel discuss candidly the problems with this postulate14. They note that nucleotides do not linkunless there is some type of activation of the phosphate group. The only effective activating groups for thenucleotide phosphate group (imidazolides, etc.), however, are those that are totally implausible in anyprebiotic scenario. In living organisms today, adenosine-5'-triphosphate (ATP) is used for activation ofnucleoside phosphate groups, but ATP would not be available for prebiotic syntheses. Joyce and Orgel notethe possible use of minerals for polymerization reactions, but then express their doubts about thispossibility15:

Whenever a problem in prebiotic synthesis seems intractable, it is possible to postulate theexistence of a mineral that catalyzes the reaction...such claims cannot easily be refuted.

In other words, if one postulates an unknown mineral catalyst that is not readily testable, it is difficult torefute the hypothesis.

Joyce and Orgel then note that if there were activation of the phosphate group, the primary polymer productwould have 5', 5'-pyrophosphate linkages; secondarily 2', 5'-phosphodiester linkages -- while the desired 3',5'-phosphodiester linkages would be much less abundant. However, all RNA known today has only 3',5'-phosphodiester linkages, and any other linkages would alter the three-dimensional structure and possibilitiesfor function as a template or a catalyst.

Even waiving these obstacles, and allowing for minute amounts of oligoribonucleotides, these moleculeswould have been rendered ineffective at various stages in their growth by adding incorrect nucleotides, or byreacting with the myriads of other substances likely to have been present. Moreover, the RNA moleculeswould have been continuously degraded by spontaneous hydrolysis and other destructive processes operatingon the primitive Earth16.

In brief, any movement in the direction of an RNA World on a realistically-modeled early Earth would havebeen continuously suppressed by destructive cross-reactions.

Postulate 3: A polyribonucleotide (i.e. RNA molecule), once formed, would have the catalytic activity toreplicate itself, and a population of such self-replicating molecules could arise.

The difficulty with this postulate is evident in the following quotation from Joyce and Orgel:

...it is assumed...that a magic catalyst existed to convert the activated nucleotides to a randomensemble of polynucleotide sequences, a subset of which had the ability to replicate. It seems tobe implicit that such sequences replicate themselves but, for whatever reason, do not replicate


4 of 8 12/31/2014 3:38 PM

unrelated neighbors.17

They refer to this as a component of "The Molecular Biologists Dream," and discuss the difficulties inherentin such a view. In order for a stable population of self-replicating RNA molecules to arise -- a prerequisite forfurther evolution -- the RNA molecules must be able to replicate themselves with high fidelity, or thesequence specificity which makes self-replication possible at all will be lost. While "it is difficult to state withcertainty the minimum possible size of an RNA replicase ribozyme," Joyce and Orgel note, it seems unlikelythat a structure with fewer than 40 nucleotides would be sufficient. Suppose, then, that "there is some 50-mer[RNA molecule of 50 nucleotides length]," Joyce and Orgel speculate, that "replicates with 90% fidelity. ...Would such a molecule be expected to occur within a population of random RNAs?"

Perhaps: but one such self-replicating molecule will not suffice.

"Unless the molecule can literally copy itself," Joyce and Orgel note, "that is, act simultaneously as bothtemplate and catalyst, it must encounter another copy of itself that it can use as a template." Copying anygiven RNA in its vicinity will lead to an error catastrophe, as the population of RNAs will decay into acollection of random sequences. But to find another copy of itself, the self-replicating RNA would need(Joyce and Orgel calculate) a library of RNA that "far exceeds the mass of the earth."18

In the face of these difficulties, they advise, one must reject

the myth of a self-replicating RNA molecule that arose de novo from a soup of randompolynucleotides. Not only is such a notion unrealistic in light of our current understanding ofprebiotic chemistry, but it should strain the credulity of even an optimist's view of RNA'scatalytic potential. If you doubt this, ask yourself whether you believe that a replicase ribozymewould arise in a solution containing nucleoside 5'-diphosphates and polynucleotidephosphorylase!19

Postulate 4: Self-replicating RNA molecules wouild have all of the catalytic activities necessary tosustain a ribo-organism.

S.A. Benner et al. note20:

...one is forced to conclude that the last ribo-organism had a relatively complex metabolism thatincluded oxidation and reduction reactions, aldol and Claison condensations, transmethylations,porphyrin biosynthesis, and an energy metabolism based on nucleoside phosphates, all catalyzedby riboenzymes...It should be noted that this reconstruction cannot be weakened without losingmuch of the logical and explanatory force of the RNA World model.

Although Benner et al. speak of the last "ribo-organism," surely the first ribo-organism would have requirednearly all of the same metabolic capabilities in order to survive. It is also apparent that the scenario of Benneret al. would surely include enclosing the ribozymes within a membrane with the ability to transport ions andorganic molecules across that membrane.

Anyone who is familiar with biochemistry would recognize that it would take hundreds of differentribozymes, each with a particular catalytic activity, to carry out the metabolic processes described above. Itshould also be apparent that most of these metabolic capabilities would have to be functional within a shortperiod of time (certainly not hundreds of years), in the same microscopic region, or the ribo-organism wouldnever survive.


5 of 8 12/31/2014 3:38 PM

When one recognizes that catalytic activities of RNA are just as dependent upon specific sequences ofnucleotides in RNA21 as protein enzymes are of amino acid sequences, then the probability of postulate 4being valid is seen to be vanishingly small.

Benner et al. note that the diverse catalytic properties of enzymes often require coenzymes or prostheticgroups. They mention particularly the iron-porphyrin, heme, and pyridoxal, but have no suggestion how these(and other co-enzymes) could have functioned in the catalytic activities of early RNA molecules.

The other unproven assumption of postulate 4 is that RNA molecules initially had all of these suggestedcatalytic activities, but nearly all of these activities have been subsequently lost. RNA molecules withcatalytic activity that are known today predominantly have nuclease or nucleotidyl transferase activity withsome minimal esterase actitivy22. There is no solid evidence that RNA molecules ever had the broad range ofcatalytic activities suggested by Benner et al., even though a number of the authors of The RNA World speakof present-day RNA molecules as being vestiges of that early RNA World.

Conclusion

We have more to learn about RNA, both in vivo (as used by organisms) and in vitro, in terms of its chemistrygenerally and functional properties in particular. RNA is a remarkable molecule.

The RNA World hypothesis is another matter. We see no grounds for considering it established, or evenpromising, except perhaps on the objectionable philosophical grounds of philosophical naturalism (and itsoperational offspring, methodological naturalism), according to which the best naturalistic hypothesis isperforce the hypothesis to be accepted. We consider that historical biology should be open to all empiricalpossibilities, including design -- and see the molecular biological system of organisms, of which RNA is sostunning a part, as exemplars of design.

We find ourselves, however, distinctly in the minority of biologists. If design exists at all, it is a matter ofsubjective intuition, the majority of our colleagues would claim, asserting with science writer George Johnsonthat "the point of science is...to explain the world through natural law."23

We would put the point rather differently. The point of science is to explain the world, through natural lawsor whatever other causes best account for the phenomena at hand.

Philosopher of science Stephen Meyer captures the point well:

The (historical) question that must be asked about biological origins is not "Which materialisticscenario will prove adequate?" but "How did life as we know it actually arise on earth?" Sinceone of the logically appropriate answers to this latter question is that "Life was designed by anintelligent agent that existed before the advent of humans," I believe it is anti-intellectual toexclude the "design hypothesis" without consideration of all the evidence, including the mostcurrent evidence, that would support it.24

Detecting design is not a matter of subjective intuition.25 To see design as a real causal possibility, however,one must break free of the constraints of naturalism.

What do Ribozome Engineering Experiments Tell Us About the Origin of Life?


6 of 8 12/31/2014 3:38 PM

Notes

1. Carl Woese, The Genetic Code (New York: Harper and Row, 1967).2. F.H.C. Crick, "The origin of the genetic code," J. Mol. Biol. 38 (1968): 367-379; L.E. Orgel,"Evolution of the genetic apparatus," J. Mol. Biol. 38 (1968): 381-393.3. K. Kruger, P.J. Grabowski, A.J. Zaug, J. Sands, D.E. Gottschling, and T.R. Cech, "Self-Splicing RNA:Autoexcision and Autocyclization of the Ribosomal RNA Intervening Sequence of Tetrahymena," Cell31 (1982): 147-157.4. Walter Gilbert, "The RNA World," Nature 319 (1986): 618.5. I. Hirao and A.D. Ellington, "Re-creating the RNA World," Current Biology 5 (1995): 1017-1022; p.1017.6. Mullis, K.B. and Faloona, "Specific synthesis of DNA in vitro via a polymerase catalyzed chainreaction," Methods Enzymol 155 (1987): 335-350.7. G. Joyce, "RNA evolution and the origins of life," Nature 338 (1989): 217-224; T.J. Gibson and A.I.Lamond, "Metabolic complexity in the RNA World and implications for the origin of proteinsynthesis," J. Mol. Evol. 30 (1990): 7-15; G.F. Joyce and L.E. Orgel, "Prospects for understanding theorigin of the RNA World," in The RNA World, eds. R.F. Gesteland and J.F. Atkins (Cold Spring Harbor,NY: Cold Spring Harbor Laboratory Press, 1993), pp. 1-25.8. H.P. Yockey, "Information in bits and bytes: reply to Lifson's Review of Information Theory andMolecular Biology," BioEssays 17 (1995): 85-88; p. 87.9. R. Shapiro, "The improbability of prebiotic nucleic acid synthesis," Origins of Life 14 (1984):565-570; R. Shapiro, "Prebiotic ribose synthesis: a critical analysis," Origins of Life 18 (1988): 71-85.10. Recently it has been shown that reaction mixtures containing dilute glycoaldehyde phosphate andformaldehyde or glyceraldehyde-2-phophate will generate reasonably high yields of ribose2,4-diphosphate and a few other sugar phosphates in less amounts. See D. Muller, S. Pitsch, A. Kittaka,E. Wagner, C.E. Wintner, and A. Eschenmoser, "Chemie von alpha-aminonitrilen. Aldomerisierung vonglykoaldehydphosphat zu racemischen hexose- 2,4,6-triphosphaten und (in gegenwart vonformaldehyd) racemischen pentose 2,4-diphophaten: rac.allose-2,4,6-triphosphat und rac.-ribose-2,4,-diphosphat sind die reaktionshauptproduckte. Helv. Chim. Acta 73 (1990): 1410-1468; Joyce andOrgel, ibid. However, if these reactions are not also run in the presence of amines and other nitrogenouscompounds (i.e., in chemical mixtures of the complexity proposed for the "prebiotic soup"), theirrelevancy to the origin of life is problematical.11. Rosa Larralde, Michael P. Robertson, and Stanley L. Miller, "Rates of decomposition of ribose andother sugars: Implications for chemical evolution," Proc. Natl. Acad. Sci. USA 92 (1995): 8158-8160.The ribose half-lives are very short, Larralde et al. report: 73 minutes at pH 7.0 and 100º C and 44years at pH 7.0 and Oº C.12. J.P. Ferris, P.C. Joshi, E.H. Edelson, and J.G. Lawless, "HCN: a plausible source of purines,pyrimidines and amino acids on the primitive Earth," J. Mol. Evol. 11 (1978): 293-311.13. R. Shapiro, "The prebiotic role of adenine: a critical analysis," Origins of Life and the Evolution ofthe Biosphere 25 (1995): 83-98.14. Joyce and Orgel, ibid.15. Ibid., p.416. C. Thaxton, W. Bradley, and R. Olsen, The Mystery of Life's Origin (New York: PhilosophicalLibrary, 1984).17. Joyce and Orgel, ibid., p.7.18. Ibid., p.11.19. Ibid, p.13.


7 of 8 12/31/2014 3:38 PM

20. S.A. Benner, M.A. Cohen, G.H. Gonnet, D.B. Berkowitz, and K.P. Johnsson, "Reading thePalimpest: Contemporary Biochemical Data and the RNA World," in The RNA World, eds. R.F.Gesteland and J.F. Atkins (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press, 1993), pp.27-70; p. 57.21. T.R. Cech, "Mechanism and Structure of a Catalytic RNA Molecule," in 40 Years of the DoubleHelix, The Robert A. Welch Foundation 37th Conference on Chemical Research, 1993, pp. 91-110; seealso T.R. Cech, "Structure and Mechanism of the Large Catalytic RNAs: Group I and Group II Intronsand Ribonuclease P," in The RNA World, eds. R.F. Gesteland and J.F. Atkins (Cold Spring Harbor, NY:Cold Spring Harbor Laboratory Press, 1993), pp. 239-269.22. Ibid.23. George Johnson, Fire in the Mind: Science, Faith, and the Search for Order (New York: Alfred A.Knopf, 1995), p. 314.24. Stephen C. Meyer, "Laws, Causes, and Facts," in Darwinism: Science or Philosophy, eds. J. Buelland V. Hearn (Richardson, Texas: Foundation for Thought and Ethics, 1994), p.34.25. See William A. Dembski, "The Design Inference: Eliminating Chance Through Small Probabilities,"unpublished Ph.D. dissertation, 1995, Department of Philosophy, University of Illinois-Chicago Circle.

Copyright © 1996 Gordon C. Mills and Dean Kenyon. All rights reserved. International copyright secured.File Date: 6.22.96


8 of 8 12/31/2014 3:38 PM

14 The RNA World

By the mid-1980s many researchers concluded that both DNA-first and

protein-first origin-of-life models were beset with many difficulties. As

a result, they sought a third way to explain the mystery of life's origin.

Instead of proposing that the first informational molecules were pro

teins or DNA, these scientists argued that the earliest stages of abiogen

esis unfolded in a chemical environment dominated by RNA molecules.,

The first scientist to propose this idea was Carl Woese, a microbiologist

at the University of Illinois. Walter Gilbert, a Harvard biophysicist, later

developed the proposal and coined the term by which it is now popu

larly known, the "RNA world."1

The RNA world is now probably the most popular theory of how

life began. Scientists in some of the most prestigious labs around the

world have performed experiments on RNA molecules in an attempt to

demonstrate its plausibility, and in the opinion of many scientists, the

RNA-world hypothesis establishes a promising framework for explain

ing how life on earth might have originated.

I had an encounter with one such scientist in the spring of 2000. I

had just written an article about DNA and the origin of life in the April

issue of a prominent New York journal of opinion.2 When the letters

to the editor came in, I initially blanched when I saw one from a fierce

critic named Kenneth R. Miller, a biology professor at Brown Uni

versity and a skilled debater. Had I made a mistake in reporting some

biological detail in my argument? When I saw his objection, however, I

was relieved. Miller claimed that my critique of attempts to explain the

The RNA World

Figure 14.1. Walter Gilbert, photographed in front of a chalkboard in his office at Harvard. Courtesy of Peter Menzel/Science Photo Library.

297

origin of biological information had failed to address the "RNA first"

hypothesis. Miller asserted that I had ignored "nearly two decades of

research on this very subject" and failed to tell my "readers of experi

ments showing that very simple RNA sequences can serve as biological

catalysts and even self-replicate."3

Miller was half right. I hadn't told my readers about these experi

ments. But I knew that two decades of research on this topic had not

solved the problem of the origin of biological information. Because of

space constraints and the format of the journal, I had decided not to address this issue in my original article. But now Miller's letter gave me

a chance to do so.

At the time I had been studying research articles from origin-of-life

specialists who were highly critical of the RNA-world hypothesis, and

in my response to Miller I cited and summarized many of their argu

ments. I heard nothing more from Miller on the matter, but as I at

tended various conferences over the next several years, I discovered that

he was far from alone. Despite the pervasive skepticism about the RNA

298 SIGNATURE IN THE CELL

world among leading origin-of-life researchers, many practicing molec

ular biologists, including some very prominent scientists at famous labs,

continued to share Miller's enthusiasm. Moreover, I discovered that

many of these molecular biologists had recently initiated new experi

mental work inspired by their confidence in the viability of the RNA

world approach. Had they solved the information problem?

Second Things First

The RNA world is a world in which the chicken and egg no longer con

found each other. At least that has been the hope. Building proteins re

quires genetic information in DNA, but information in DNA cannot be

processed without many specific proteins and protein complexes. This

problem has dogged origin-of-life research for decades. The discovery

that certain molecules of RNA possess some of the catalytic properties

seen in proteins suggested a way to solve the problem. RNA-first advo

cates proposed an early stage in the development of life in which RNA

performed both the enzymatic functions of modern proteins and the

information-storage function of modern DNA, thus sidestepping the

need for an interdependent system of DNA and proteins in the earliest

living system.

Typically RNA-first models have combined chance events and a law

like process of necessity, in particular, the process of natural selection.

As Gilbert and others envision it, a molecule of RNA capable of copying

itself (or copying a copy of itself) first arose by the chance association of nucleotide bases, sugars, and phosphates in a pre biotic soup (see Fig.

14.2). Then because that RNA enzyme could self-replicate, natural se

lection ensued, making possible a gradual increase in the complexity of

the primitive self-replicating RNA system, eventually resulting in a cell

with the features we observe today. Along the way, a simple membrane,

itself capable of self-reproduction, enclosed the initial RNA enzymes

along with some amino acids from the prebiotic soup.4

According to this model, these RNA enzymes eventually were re

placed by the more efficient proteins that perform enzymatic functions

in modern cells. For that to occur, the RNA-replicating system first

had to begin producing a set of RNA enzymes that could synthesize

The RNA World 299

Figure 14.2. The RNA World Scenario in Seven Steps. Step 1: The building

blocks of RNA arise on the early earth. Step 2: RNA building blocks link

up to form RNA oligonucleotide chains. Step 3: An RNA replicase arises

by chance and selective pressures ensue favoring more complex forms of

molecular organization. Step 4: RNA enzymes begin to synthesize proteins

from RNA templates. Step 5: Protein-based protein synthesis replaces

RNA-based protein synthesis. Step 6: Reverse transcriptase transfers genetic

information from RNA molecules into DNA molecules. Step 7: The modern

gene expression system arises within a proto-membrane.


proteins. As Gilbert has explained, in this step RNA molecules began

"to synthesize proteins, first by developing RNA adapter molecules that

can bind activated amino acids and then by arranging them according

to an RNA template using other RNA molecules such as the RNA core

of the ribosome."5 Finally, DNA emerged for the first time by a process

called reverse transcription. In this process, DNA received the informa

tion stored in the original RNA molecules, and eventually these more

stable DNA molecules took over the information-storage role that RNA

had performed in the RNA world. At that point, RNA was, as Gilbert

put it, "relegated to the intermediate role it has today-no longer the

center of the stage, displaced by DNA and the more effective protein

enzymes."6

I knew that origin-of-life theories that sound plausible when stated

in a few sentences often conceal a host of practical problems. And so

it was with the RNA world. As I investigated this hypothesis, both

before and after my exchange with Professor Miller, I found that many

crucial problems lurked in the shadows, including the one I had seen

before: the theory did not solve the problem of biological informa

tion-it merely displaced it.

Because so many scientists assume that the RNA world has solved

the problem of the origin of life, this chapter will provide a detailed

and, in some places, technical critique of this hypothesis. My critique

details five crucial problems with the RNA world, culminating in a

discussion of the information problem. To assist nontechnical readers,

I have placed some of this critique in notes for the scientifically trained.

I would ask technically minded readers to read these notes in full, be

cause in some cases they provide important additional support for, or

qualifications to, my arguments.

Each element of this critique stands mostly on its own. So if you find

that the technical material under one subheading presupposes unfamil

iar scientific concepts or terminology, take note of the heading, which

summarizes the take-home message of the section, and skip ahead to

the next one, or even the final two, which address the theory's greatest

weakness: its inability to explain the origin of biological information.

The RNA World 301

Problem I: RNA Building Blocks Are Hard to Synthesize and Easy to Destroy

Before the first RNA molecule could have come together, smaller con

stituent molecules needed to arise on the primitive earth. These in

clude a sugar known as ribose, phosphate molecules, and the four RNA

nucleotide bases (adenine, cytosine, guanine, and uracil). It turns out,

however, that both synthesizing and maintaining these essential RNA

building blocks, particularly ribose (the sugar incorporated into nucleo

tides) and the nucleotide bases, has proven either extremely difficult or

impossible to do under realistic prebiotic conditions.7 (See Fig. 14.3.) Consider first the problems with synthesizing the nucleotide bases. In

the years since the RNA world was proposed, chemist Robert Shapiro

has made a careful study of the chemical properties of the four nucleotide

bases to assess whether they could have arisen on the early earth under

Figure 14.3. The chemical structure and constituents of RNA.


realistic conditions. He notes first that "no nucleotides of any kind have been reported as products of spark-discharge experiments or in studies of meteorites." Stanley Miller, who performed the original prebiotic simulation experiment, published a similar study in 1998. 8 Moreover, even if they did somehow form on the early earth, nucleotide bases are too chemically fragile to have allowed life enough time to evolve in the manner Gilbert and other RNA-first theorists envision. Shapiro and Miller have noted that the bases of RNA are unstable at temperatures required by currently popular high-temperature origin-of-life scenarios. The bases are subject to a chemical process known as "deamination," in which they lose their essential amine groups (NH2). At 100 degrees C, adenine and guanine have chemical half-lives of only about one year; uracil has a half-life of twelve years; and cytosine a half-life of just nineteen days. Because these half-lives are so short, and because the evolutionary process envisioned by Gilbert would take so long-especially for natural selection to find functional ribozymes (RNA molecules with catalytic activity) by trial and error-Stanley Miller concluded in 1998 that "a high temperature origin of life involving these compounds [the RNA bases] therefore is unlikely."9 Miller further noted that, of the four required bases, cytosine has a short half-life even at low temperatures, thus raising the possibility that "the GC pair" (and thus RNA) "may not have been used in the first genetic material." Shapiro concurred. He showed that it would have been especially difficult to synthesize adenine and cytosine at high temperatures and cytosine even at low temperatures. Thus he concluded that the presumption that "the bases, adenine, cytosine, guanine and uracil were readily available on the early earth" is "not supported by existing knowledge of the basic chemistry of these substances."10

Producing ribose under realistic conditions has proven even more problematic. Prebiotic chemists have proposed that ribose could have arisen on the early earth as the by-product of a chemical reaction called the formose reaction. The formose reaction is a multistep chemical reaction that begins as molecules of formaldehyde in water react with one another. Along the way, the formose reaction produces a host of different sugars, including ribose, as intermediate by-products in the sequence of reactions. But, as Shapiro has pointed out, the formose reaction · vill not produce sugars in the presence of nitrogenous sub-

The RNA World 303

stances.11 These include peptides, amino acids, and an;iines, a category of molecules that includes the nucleotide bases.

This obviously poses a couple of difficulties. First, it creates a dilemma for scenarios that envision proteins and nucleic acids arising out of a prebiotic soup rich in amino acids. Either the prebiotic environment contained amino acids, which would have prevented sugars (and thus DNA and RNA) from forming, or the prebiotic soup contained no amino acids, making protein synthesis impossible. Of course, RNAfirst advocates might try to circumvent this difficulty by proposing that proteins arose well after RNA. Yet since the RNA-world hypothesis envisions RNA molecules coming into contact with amino acids early on within the first protocellular membranes (see above), choreographing the origin of RNA and amino acids to ensure that the two events occur separately becomes a considerable problem.

The RNA-world hypothesis faces an even more acute, but related, obstacle-a kind of catch-22. The presence of the nitrogen-rich chemicals necessary for the production of nucleotide bases prevents the production of ribose sugars. Yet both ribose and the nucleotide bases are needed to build RNA. (See note for details).12 As Dean Kenyon explains, "The chemical conditions proposed for the prebiotic synthesis of purines and pyrimidines [the bases] are sharply incompatible with those proposed for the synthesis of ribose."13 Or as Shapiro concludes: "The evidence that is currently available does not support the availability of ribose on the pre biotic earth, except perhaps for brief periods of time, in low concentration as part of a complex mixture, and under conditions unsuitable for nucleoside synthesis."14

Beyond that, both the constituent building blocks of RNA and whole RNA molecules would have reacted readily with the other chemicals present in the prebiotic ocean or environment. These "interfering cross-reactions" would have inhibited the assembly of RNA from its constituent monomers and inhibited any movement from RNA molecules toward more complex biochemistry, since the products of these reactions typically produce biologically inert (or irrelevant) substances.

Furthermore, in many cases, reactions (such as the formose reaction) that produce desirable by-products such as ribose also produce many undesirable chemical by-products. Unless chemists actively intervene, undesirable and desirable chemical by-products of the same reaction


react with each other to alter the composition of the desired chemicals

in ways that would inhibit the origin of life. In sum, synthesizing the

building blocks of the RNA molecule under realistic prebiotic condi

tions has proven formidably difficult.

Problem 2: Ribozymes Are Poor Substitutes for Proteins

Another major problem with the RNA world is that naturally occur

ring RNA molecules possess very few of the specific enzymatic proper

ties of proteins. To date, scientists have shown that RNA catalysts or

"ribozymes" can perform a small handful of the thousands of func

tions performed by modern proteins. Scientists have shown that some

RNA molecules can cleave other RNA molecules (at the phosphodi

ester bond) in a process known as hydrolysis. Biochemists also have

found RNAs that can link (ligate) separate strands of RNA (by catalyz

ing the formation of phosphodiester bonds). Other studies have shown

that the RNA in ribosomes (rRNA) promotes peptide-bond formation

within the ribosome15 and can promote peptide bonding outside the

ribosome, though only in association with an additional chemical cata

lyst.16 Beyond that, RNA can perform only a few minor functional roles

and then usually as the result of scientists intentionally "engineering" or "directing" the RNA catalyst (or ribozyme) in question.17

For this reason, claiming that catalytic RNA could replace proteins

in the earliest stages of chemical evolution is extremely problematic. To say otherwise would be like asserting that a carpenter wouldn't need

any tools besides a hammer to build a house, because the hammer per

formed two or three carpentry functions. True, a hammer does per

form some carpentry functions, but building a house requires many

specialized tools that can perform a great variety of specific carpentry

functions. In the same way, RNA molecules can perform a few of the

thousands of different functions proteins perform in "simple" single

cells (e.g., in the E. coli bacterium), but that does not mean that RNA

molecules can perform all necessary cellular functions.

The RNA World

Problem 3: An RNA-Based Translation and Coding System Is Implausible

305

The inability of RNA molecules to perform many of the functions of

protein enzymes raises a third and related concern about the plausibility

of the RNA world. RNA-world advocates offer no plausible explanation

for how primitive self-re'plicating RNA molecules might have evolved

into modern cells that rely on a variety of proteins to process genetic

information and regulate metabolism.18

To evolve beyond the RNA world, an RNA-based replication system

eventually would have to begin to produce proteins, and not just any

proteins, but proteins capable of template-directed protein manufac

ture. But for that to occur, the RNA replicator first would need to

produce machinery for building proteins. In modern cells it takes many

proteins to build proteins. So, as a first step toward building proteins,

the primitive replicator would need to produce RNA molecules capable

of performing the functions of the modern proteins involved in trans

lation. (Recall from Chapter 5 that translation is the process of build

ing proteins from the instructions encoded on an mRNA transcript.)

Presumably, these RNA molecules would need to perform the func

tions of the twenty specific tRNA synthetases and the fifty ribosomal

proteins, among the many others involved in translation. At the same

time, the RNA replicator would need to produce tRNAs and the many

mRNAs carrying the information for building the first proteins. These

mRNAs would need to be able to direct protein synthesis using, at first,

the transitional ribozyme-based protein-synthesis machinery and then,

later, the permanent and predominantly protein-based protein-synthesis

machinery. In short, the evolving RNA world would need to develop a

coding and translation system based entirely on RNA and also gener

ate the information necessary to build the proteins that later would be

needed to replace it.

This is a tall order. The cell builds proteins from the information

stored on the mRNA transcript (i.e., the copy) of the original DNA

molecule. To do this, a bacterial cell depends upon a translation and

coding system consisting of 106 distinct but functionally integrated

proteins as well as several distinct types of RNA molecules ( tRNAs,

mRNAs, and rRNAs).19 This system includes the ribosome (consisting


of fifty distinct protein parts), the twenty distinct tRNA synthetases,

twenty distinct tRNA molecules with their specific anticodons (all of

which jointly embody the genetic code), various other proteins, free

floating amino acids, ATP molecules (for energy), and-last, but not

least-information-rich mRNA transcripts for directing protein syn

thesis. Furthermore, many of the proteins in the translation system per

form multiple functions and catalyze coordinated multistep chemical

transformations (see Fig. 14.4).

Is it possible that a similar translation and coding system capable of

producing genetically encoded proteins might first have arisen using

only RNA catalysts (ribozymes)? Advocates of the RNA-world hy

pothesis have defended the possibility because of the demonstrated

catalytic properties of some RNA molecules. Eugene Koonin and

Yuri Wolf, two prominent scientists at the National Center for Bio

technology Information, recently reviewed the results of research on

the capacities of RNA catalysts in an important article assessing the

plausibility of an RNA-based translation system.20 They note that in

the last twenty years, molecular biologists have documented, or engi

neered, ribozymes that can catalyze "all three elementary reactions"21

required for translation, including aminoacylation (the formation of

a bond between an amino acid and an RNA), the peptidyl-transferase

reaction (which forms the peptide bond between amino acids), and

amino-acid activation (in which adenosine monophosphate is attached

to an amino acid).

At first glance, these results may seem to support the feasibility of

an RNA-based translation system. Nevertheless, significant reasons

to doubt this aspect of the RNA-world hypothesis remain, as Koonin

and Wolf note. First, though ribozymes have demonstrated the ca

pacity to catalyze representative examples of the three main types of

chemical reactions involved in translation, they have not demonstrated

the ability to catalyze anywhere near all the necessary reactions that

fall within these general classifications. Moreover, the gap between

"some" and "all" necessary reactions of a given type remains signifi

cant. For example, ribozyme engineers have successfully designed an

RNA molecule that will catalyze the formation of an aminoacyl bond

between itself and the amino acids leucine and phenylalanine. 22 But

no one has yet demonstrated that RNA can catalyze aminoacyl bonds

oT#et 'P�TEI�

(�[r.. IIJITrATIC>N, el.OIJ[rATIC>N,

& TEtMIIJATIC>N 'FACTo��)

The RNA World

AkitJO AlI�

307

AkitJOAlYL-tttJA

�YtJT#ETA�e�

Figure 14.4. The main molecular components of the translation system:

twenty specific transfer-RNA molecules, twenty specific aminoacyl tRNA

synthetases, the ribosome with its two main subunits composed of fifty

proteins and ribosomal RNA, the messenger-RNA transcript, and a supply

of amino acids.


with the other eighteen protein-forming amino acids, still less with

the specificity required to make the resulting molecules useful for

translation. Yet establishing a genetic code requires molecules that can

catalyze highly specific aminoacylation for each of the twenty protein

forming amino acids. To say that RNA can catalyze "aminoacylation"

is true, but it obscures the distinction between part of a group and the

whole group, where having the whole group of molecules is necessary

to the function in question. Again, it takes more than a hammer to

build a house.

Second, unlike RNA catalysts (ribozymes), the protein-based en

zymes involved in translation perform multiple functions, often in

closely integrated or choreographed ways. Ribozymes, however, are

the one-trick ponies of the molecular world. Typically, they can per

form one subfunction of the several coordinated functions that a cor

responding enzyme can perform. But they cannot perform the entire

range of necessary functions, nor can they do so with the specificity

needed to execute the many sequentially coordinated reactions that

occur during translation.

Consider what ribozymes must do to rival the capacities of the syn

thetases that catalyze aminoacylation, which occurs between tRNA

molecules and their "conjugate" amino acids during translation in

actual cells. Researchers have demonstrated that certain RNA molecules

can bind a protein-forming amino acid, phenylalanine, to itself, thus

performing the function of aminoacylation. They have even isolated

a version of the RNA catalyst that binds only phenylalanine, achiev

ing a specificity of sorts. But the synthetase enzymes responsible for

aminoacylation in life must catalyze a complex two-stage chemical reac

tion involving three kinds of molecules: amino acids, ATP (adenosine

triphosphate ), and tRNAs.

In the first stage of this reaction, synthetases couple ATP to a spe

cific amino acid, giving it the stored energy (in the form of adenosine

monophosphate, AMP) needed to establish a bond with a tRNA

molecule. Next, synthetases couple specific tRNA molecules to spe

cific activated (AMP-charged) amino acids. These tRNAs have spe

cific shapes and anticodon sites that enable them to bond to mRNA

at the ribosome. Thus, synthetases help form molecular complexes

with a specificity of fit and with specific binding sites that enable

The RNA World 309

translation to occur in the context of a whole system of associated

molecules.

The RNA catalyst proposed as a precursor to the synthetase cannot

do this. It does not couple ATP to amino acids as a precursor to catalyz

ing aminoacylation. Instead, the ribozyme engineer provides "preade

nylated" amino acids (amino acids already linked to AMP molecules).

Nor does the RNA catalyst couple an amino acid to a specific tRNA

with a specific anticodon. The more limited specificity it achieves only

ensures that the RNA catalyst will bind a particular amino acid to

itself, a molecule that does not possess the specific cloverleaf shape or

structure of a tRNA. Moreover, this RNA does not carry an anticodon

binding site corresponding to a specific codon on a separate mRNA

transcript. Thus, it has no functional significance within a system of mol

ecules for performing translation. Indeed, no other system of molecules

has even been proposed that could confer functional significance or

specificity on the amino acid-RNA complexes catalyzed by the amino

acyl ribozyme.

Thus, even in the one case where ribozyme engineers have produced

an RNA-aminoacyl catalyst, the ribozyme in question will not produce

a molecule with a functional specificity, or capacity to perform coordi

nated reactions, equivalent to that of the synthetases used in modern

cells. Yet without this specificity and capacity to coordinate reactions,

translation-the construction of a sequence-specific arrangement of

amino acids from the specific RNA transcript-will not occur. 23

Similar limitations affect the RNA catalysts that have been shown

to be capable of peptidyl-transferase activity (i.e., catalyzing peptide

bonds between amino acids). These ribozymes (made of free-standing

ribosomal RNA) compare quite unfavorably with the capacities of the

protein-dominated ribosomes that perform this function in extant cells.

For example, researchers have found that free-standing ribosomal RNA

can only catalyze peptide-bond formation in the presence of another

catalyst. More important, apart from the proteins of the ribosome, free

standing ribosomal RNA does not force amino acids to link together

into linear chains, which is essential to protein function. (For more

details, see the note.)24


Why RNA Catalysts Can)t Do What True Enzymes Can There is a fundamental chemical reason for the limited functionality of

RNA catalysts-one that casts still further doubt on the RNA-world hy

pothesis and specifically on its account of the origin of the translation

system. Because of the inherent limitations of RNA chemistry, 25 single

RNA molecules do not catalyze the coordinated multistep reactions that

enzymes, such as synthetases, catalyze. Even if separate RNA catalysts

can be found that catalyze each of the specific reactions involved in trans

lation (which is by no means certain), that would leave us very far short

of a translation system. Each pony of the RNA world does only its one

trick. And even if all the ponies were present together, each one would

do only its particular trick separately, decoupled from the others. That's a

problem, because producing the molecular complexes necessary for trans

lation requires coupling multiple tricks-multiple crucial reactions-in a

closely integrated (and virtually simultaneous) way. True enzyme cata

lysts do this. RNA and small-molecule catalysts do not.

Here's the chemical backstory. Enzymes couple energetically fa

vorable and unfavorable reactions together into a series of reactions

that are energetically favorable overall. As a result , they can drive

forward two reactions where ordinarily only one would occur with

any appreciable frequency. Water runs downhill because of favor

able energetics provided by gravitational force. Water does not run

uphill, however, unless there is so much of it that it accumulates

and slowly rises up the ban k. Whether chemical reactions will occur

readily depends upon whether there is enough energy to make them

occur. Molecules with enough stored energy to establish new chemi

cal bonds will react readily with one another. Molecules with insuf

ficient stored energy will not react readily with each other unless vast

amounts of the reactants are provided (the equivalent of the rising

water flooding the banks).

Enzymes use a reaction that liberates energy to drive forward a re

action that requires energy, coupling energetically favorable and unfa

vorable reactions together. Enzymes can do this because they have a

complex three-dimensional geometry that enables them to hold all the

molecules involved in each step of the reaction together and to coor

dinate their interactions. But two independent catalysts cannot accom

plish what a compound catalyst (i.e., an enzyme) can. And so far RNA

The RNA World 311

ribozymes have demonstrated the capacity to act only as independent

catalysts, not true enzyme catalysts. RNA catalysts might catalyze some

energetically favorable reactions, but without the sophisticated active

sites of enzymes, they can't couple those favorable reactions to energeti

cally unfavorable reactions.26 (See Fig. 14.5.)

€"1ZYMATrlALLY loU'PL€1> TWo-S'r€'P �€AlTrOIJ (AMIIJO AlYLATr0"1)

AA + ATP -+ Al-\WOACYL-Ak'P + z.'P,:. Akii.ioACYL-Ak'P + t�i.iA - Al-\WOAlYL-t�i.iA + Ak'P

Figure 14.5. Enzymes couple energetically favorable and unfavorable

reactions together into a series of reactions that are energetically favorable

overall. Enzymes can accomplish this because they have a three-dimensional

specificity that allows them to sequester and correctly position all the

molecules involved in a series of such reactions. RNA catalysts cannot do

this. The figure above shows an enzymatically mediated reaction called

aminoacylation. The diagram shows the specificity of fit between a tRNA

synthetase and a tRNA molecule during this two-stage chemical reaction.

The synthetase links the tRNA to a specific amino acid (AA) using energy

from ATP, thus coupling energetically favorable and unfavorable reactions.

Amino acids and ATP molecules are not pictured. They would be enveloped

by the synthetase during the reactions represented by the chemical equations.


Thus, the demonstration that RNA can catalyze "all the elementary

reactions" of translation, but neither the suite of functions nor the co

ordinated functions performed by the necessary enzyme catalysts of

the extant translation syste1'1, does little to establish the plausibility of

ribozyme-based protein synthesis, let alone the transition to enzyme

based protein synthesis, that the RNA-world scenario requires. The

inability to account for the origin of the translation system and ge

netic code, therefore, remains a formidable barrier to the success of the

RNA-world hypothesis.

Problem 4: The RNA World Doesn't Explain the Origin of Genetic Information

As I sifted through the primary scientific literature on the RNA-world

hypothesis, it did not take me long to realize that the hypothesis faced

significant problems quite apart from the central sequencing problem

that most interested me. Yet I also realized that it did not resolve the

mystery of the origin of biological information-which I had, here

tofore, called the DNA enigma. Indeed, I now realized that I might

just as easily have called that mystery the "RNA enigma," because the

information problem looms just as large in a hypothetical RNA world

as it does in a DNA world. This is not actually surprising. The RNA

world was proposed not as an explanation for the origin of biological

information, but as an explanation for the origin of the interdepen

dence of nucleic acids and proteins in the cell's information-processing

system. And as I studied the hypothesis more carefully, I realized that it

presupposed or ignored, rather than explained, the origin of sequence

specificity-information-in various RNA molecules.

Consider the step in the RNA-world scenario that I just examined

getting from a primitive replicator to a system for building the first

proteins. Even if a system of ribozymes for building proteins had arisen

from an RNA replicator, that system of molecules would still need

information-rich templates for building specific proteins. RNA-world

advocates give no account of the origin of that information beyond

vague appeals to chance. But as I argued in Chapters 8-10, chance is

not a plausible explanation for the information necessary for building

The RNA World 313

even one protein of modest length, let alone a set of RNA templates for

building the proteins needed to establish a protein-based translation

system and genetic code.

The need to account for these templates of information stands as a

formidable challenge to the RNA world. Nevertheless, the hypothesis

faces an even more basic information problem: the first self-replicating

RNA molecules themselves would have needed to be sequence-specific

in order to perform the function of replication, which is a prerequi

site of both natural selection and any further evolution toward cellular

complexity.

Though the RNA world was originally proposed as an explanation

for the "chicken and egg" functional interdependence problem, not the

information problem, some RNA-world advocates nevertheless appear

to think that it can somehow leapfrog the sequence-specificity require

ment. They imagine short chains ( oligomers) of RNA arising by chance

on the prebiotic earth. Then, after a sufficiently large pool of these mol

ecules had arisen, some would have acquired the ability to self-replicate.

In such a scenario the capacity to self-replicate would then favor the

survival of those RNA molecules that could do so and thus would favor

the specific sequences that the first self-replicating molecules happened

to have. Thus, self-replication arose again as a kind of "accidental choice

remembered."27

But like Quastler's DNA-first model discussed in the last chapter,

this scenario merely shifts the specificity problem out of view. First,

for strands of RNA to perform catalytic functions (including self

replication), they, like proteins, must display specific arrangements of

their constituent building blocks (nucleotides in the RNA case). In

other words, not just any sequence of RNA bases will be capable of self

replication. Indeed, experimental studies indicate that RNA molecules

with the capacity to replicate themselves, if they exist at all, are ex

tremely rare among possible RNA base sequences. Although no one has

yet produced a fully self-replicating RNA molecule,28 some research

ers have engineered a molecule that can copy a part of itself-though

only about 10 percent of itself and then only if a complementary primer

strand is provided to the ribozyme by the investigator. Significantly,

the scientists selected this partial self-replicator out of an engineered

pool of 1,000 trillion (1015) other RNA molecules, almost all of which


lack even this limited capacity for self-replication.29 This suggests that

sequences with this capacity are extremely rare and would be especially

so within a random (nonengineered) sample.

Further, for an RNA molecule to self-replicate, the RNA strand

must be long enough to form a complex structure. Gerald Joyce and

the late Leslie Orgel are two prominent origin-of-life researchers who

have evaluated the RNA-world scenario in detail. They consider, for

the sake of argument, that a replicase could form in a SO-base RNA

strand, though they are clearly skeptical that an RNA sequence of this

length would really do the job. 30 Experimental results have confirmed

their skepticism. Jack Szostak, a prominent ribozyme engineer, and his

colleagues have found that it typically takes at least 100 bases to form

structures capable of catalyzing simple ligation (linking) reactions. He

estimates that getting a ligase capable of performing the other functions

that polymerases must perform-"proper template binding, fidelity and

strand separation"-may require between 200 and 300 nucleotides.31

The ribozyme mentioned above-the one that can partially copy itself

required 189 nucleotide bases.32 It is presently unclear how many bases

would be needed to generate enough structural complexity to allow

true polymerase function, since no molecule capable of both complete

and unassisted self-replication has yet been engineered. It may be as low

as 189 bases, but it may be much higher, or it may simply be impos

sible. 33 Moreover, the problem may be more basic than length. RNA,

with its limited alphabet of four bases, may not even have the capacity to

form the complex three-dimensional shapes and distributions of charge

necessary to perform polymerase or replicase function.

In any case, even if we suppose that RNA-based RNA polymerases

( replicases) are possible, experimental evidence indicates that they would

have to be information-rich-both complex and specified-just like

modern DNA and proteins. Yet explaining how the building blocks of

RNA might have arranged themselves into information-rich sequences

has proven no easier than explaining how the parts of DNA might have

done so, given the requisite length and specificity of these molecules. As

Christian de Duve has noted in critique of the RNA-world hypothesis,

"Hitching the components together in the right manner raises addi

tional problems of such magnitude that no one has yet attempted to do

so in a prebiotic context."34

Ihe RNA World 315

Certainly, appeals to chance alone have not solved the RNA information problem. A 100-base RNA molecule corresponds to a space of possibilities equal to 4100 (or 1060). A 200-base RNA molecule corresponds to 4200 (or 10120) possibilities. Given this and the experiments mentioned above showing the rarity of functional ribozymes (to say nothing of polymerases) within RNA sequence space, the odds of a functional, self-replicating RNA sequence arising by chance are exceedingly small. Moreover, the odds against such an event occurring are only compounded by the likely presence of destructive cross-reactions between desirable and undesirable molecules within any realistic prebiotic environment.

To make matters worse, as Gerald Joyce and Leslie Orgel note, for a single-stranded RNA catalyst to produce an RNA identical to itself (i.e., to "self-replicate"), it must find an appropriate RNA molecule nearby to function as a template, since a single-stranded RNA cannot function as both replicase and template. Moreover, as they observe, this RNA template would have to be the precise complement of the replicase. Once this chance encounter occurred, the replicase molecule could make a copy of itself by making a complement of its complement (i.e., by transcribing the template), using the physics of nucleotide base pairing. 35

This requirement, of course, compounds the informational problem facing this crucial step in the RNA-world scenario. Even if an RNA sequence could acquire the replicase function by chance, it could perform that function only if another RNA molecule-one with a highly specific sequence relative to the original-arose close by. (See Fig. 14.6.) Thus, in addition to the specificity required to give the first RNA molecule self-. replicating capability, a second RNA molecule with an extremely specific sequence-one with essentially the same specificity as the original-would also have to arise. RNA-world theorists do not explain the origin of the requisite specificity in either the original molecule or its complement. Orgel and Joyce have calculated that to have a reasonable chance of finding two such complementary RNA molecules of a length sufficient to perform catalytic functions would require an RNA library of some 1048 RNA molecules. 36 The mass of such a library vastly exceeds the mass of the earth, suggesting the extreme implausibility of the chance origin of a primitive replicator system. They no doubt vastly underestimate the necessary size of this library and the actual improbability of a self-replicating couplet of


RNAs arising, because, as noted, they assume that a 50-base RNA might be capable of self-replication. (See note for qualifying details. )37

Given these odds, the chance origin of even a primitive selfreplicating system-one involving a pair of sequence-specific (i.e., information-rich) replicases-seems extremely implausible. And, yet, invoking natural selection doesn't reduce the odds or help explain the origin of the necessary replicators since natural selection ensues only after self-replication has arisen. As Orgel and Joyce explain, "Without evolution [i.e., prebiotic natural selection] it appears unlikely that a self-replicating ribozyme could arise, but without some form of selfreplication there is no way to conduct an evolutionary search for the first primitive self-replicating ribozyme."38

Robert Shapiro has resorted to one of my old standbys-Scrabble letters-to illustrate why neither chance, nor chance and natural selection combined, can solve the sequencing problem in the RNA world. While speaking in 2007 at a private conference on the origin of life, he asked an elite scientific audience to imagine an enormous pile of Scrabble letters. Then he said, "If you scooped into that heap [of letters], and you flung them on the lawn there, and the letters fell into a line which contained the words, 'To be or not to be, that is the question,' that is roughly the odds of an RNA molecule, given no feedback [natural selection]-and there would be no feedback, because it [the RNA molecule] wouldn't be functional until it attained a certain length and could copy itselfappearing on earth."39

If neither chance, nor chance and selection, can solve the RNA sequencing problem, can self-organization do the trick? It can't. RNA bases, like DNA bases, do not manifest bonding affinities that can explain their specific arrangements. Thus, no one has even attempted to solve the RNA sequencing problem by proposing a "self-organizational RNA world scenario." Instead, the same kind of evidentiary and theoretical problems emerge whether one proposes that genetic information arose first in RNA or DNA molecules. And every attempt to leapfrog the sequencing problem by starting with supposedly "informationgenerating" RNA replicators has only shifted the problem to the specific sequences that would be needed to make such replicators functional.

In addition, not only does the origin of RNA self-replication depend upon sequence specificity (information), but the transition from the

The RNA World 317

RNA-based translation system to the current protein-based translation

system would have required at some point the production of more than

100 different proteins, each of which would have in turn required an

information-rich nucleic acid to guide its construction.

Once again, the pink stuff was spreading.

L

2..

��' #YPoiUETTlAL �NA

�EJ>UlA�E.

l�tJOT �"PLUATC (tJO �E"PUlA�E)

lAtJtJOT �e"PUlATe (NO Tek"PLATC)

+ f ,.

� '""'l111 ,.

' I IJ ' g.1 I u l u '�"�' _w ,....y

( loM.PLEM.Ef\ITA�Y

�Nt:>) E'IACrloPY

O'F Ol?IqitJAL

Figure 14.6. The minimal requirements for template-directed RNA self

replication as envisioned by Joyce and Orgel. They insist that any RNA replicase would need to come into close proximity to an exact complementary

strand, thus increasing the needed sequence specificity associated with

getting such self-replication (and natural selection) started.


Problem 5: Ribozyme Engineering Does Not Simulate Undirected Chemical Evolution

Because of the difficulties with the RNA-world hypothesis and the

limited number of enzymatic functions that naturally occurring

ribozymes can perform, a new cottage industry has sprung up in

molecular biology. Scientists sympathetic to the RNA world have

sought to design new RNA molecules with heretofore unobserved

functions. In doing so, these scientists have hoped not only to learn

more about RNA chemistry, but also to demonstrate the plausibil

ity of the RNA-world hypothesis and possibly even to synthesize an

artificial form of life.40

These ribozyme-engineering experiments typically deploy one of two

approaches: the "rational design" approach or the "directed evolution"

approach. In both approaches, biologists try to generate either more

efficient versions of existing ribozymes or altogether new ribozymes

capable of performing some of the other functions of proteins. In the

rational-design approach, the chemists do this by directly modifying

the sequences of naturally occurring RNA catalysts. In the directed

evolution (or "irrational design") approach, scientists seek to simu

late a form of prebiotic natural selection in the process of producing

ribozymes with enhanced functional capacities. To manage this they

screen pools of RNA molecules using chemical traps to isolate mol

ecules that perform particular functions. After they have selected these

molecules out of the pool, they generate variant versions of these mol

ecules by randomly altering (mutating) some part of the sequence of

the original molecule. Then they select the most functional molecules

in this new crop and repeat the process several times until a discernible

increase in the desired function has been produced.

Most ribozyme-engineering procedures have been performed on li

gases, ribozymes that can link together two RNA chains ( oligomers)

by forming a single (phosphodiester) bond between them. Ribozyme

engineers want to demonstrate that these ligases can be transformed

into true polymerases or "replicases." These polymerases would not

only link nucleotide bases together (by phosphodiester bonds), but also

would stabilize the exposed template strands, and use the exposed bases

as a template to make sequence-specific copies.

The RNA World 319

Polymerases are the holy grail of ribozyme engineering. According

to the RNA-world hypothesis, once a polymerase capable of template

directed self-replication arose, then natural selection could have become

a factor in the subsequent chemical evolution of life. Since ligases can

perform one, though only one, of the several functions performed by

true polymerases, RNA-world theorists have postulated ligases as the

ancestral molecular species from which the first self-replicating poly

merase arose. They have tried to demonstrate the plausibility of this

conjecture by using ribozyme engineering to build polymerases (or rep

licases) from simpler ligase ribozymes.

To date, no one has succeeded in engineering a fully functional

RNA-based RNA polymerase, from either a ligase or anything else.41

Ribozyme engineers have, however, used directed evolution to enhance

the function of some common types of ligases. As noted, they also have

produced a molecule that can copy a small portion of itself. Leading ri

bozyme engineers such as Jack Szostak and David Bartel have presented

these results as support for an undirected process of chemical evolu

tion starting in an RNA world.42 Popular scientific publications and

textbooks have often heralded these experiments as models for under

standing the origin of life on earth and as the leading edge of research

establishing the possibility of evolving an artificial form of life in a test

tube.

Yet these claims have an obvious flaw. Ribozyme engineers tend to

overlook the role that their own intelligence has played in enhancing

the functional capacities of their RNA catalysts. The way the engineers

use their intelligence to assist the process of directed evolution would

have no parallel in a prebiotic setting, at least one in which only undi

rected processes drove chemical evolution forward. Yet this is the very

setting that ribozyme experiments are supposed to simulate.

RNA-world advocates envision ligases evolving via undirected pro

cesses into RNA polymerases that can replicate themselves from free

standing bases, thereby establishing the conditions for the beginning of

natural selection. In other words, these experiments attempt to simulate

a transition that, according to the RNA-world hypothesis, would have

taken place before natural selection had begun to operate. Yet in order to

improve the function of the ligase molecules, the experiments actually

simulate what natural selection does. Starting from a pool of random


sequences, the investigators create a chemical trap to isolate only those sequences that evince ligase function. Then they select those sequences for further evolution. Next they use a mutagenesis technique to generate a set of variant versions of these original ligases. Then they isolate and select the best sequences-those manifesting evidence of enhanced ligase function or indications of future polymerase function-and repeat the process until some improvement in the desired function has been realized.

But what could have accomplished these tasks before the first replicator molecule had evolved? Szostak and his colleagues do not say. They certainly cannot say that natural selection played this role, since the origin of natural selection as a process depends on the prior origin of the self-replicating molecule that Szostak and his colleagues are working so hard to design. Instead, in their experiment, Szostak and his colleagues play a role that nature cannot play until a self-replicating system, or at least a self-replicating molecule, has arisen. Szostak and his colleagues function as the replicators. They generate the crop of variant sequences. They make the choices about which of these sequences will survive to undergo another round of directed evolution. Moreover, they make these choices with the benefit of a foresight that neither natural selection nor any other undirected or unintelligent process can-by definition-possess.43 Indeed, the features of the RNA molecules that Szostak and his colleagues isolate and select are not features that would, by themselves in a precellular context, confer any functional advantage.

Of course, ligase enzymes perform functions in the context of modern cells and in that setting might confer a selectable advantage on the cells that possess them. But prior to the origin of the first selfreplicating protocell, ligase ribozymes would not have any functional advantage over any other RNAs. At that stage in chemical evolution, no self-reproducing system yet existed upon which any advantage could be conferred.

The ability to link (ligate) nucleotide chains is, at best, a necessary but not a sufficient condition of polymerase or replicase function. Absent a molecule or, what is more likely, a system of molecules possessing all of the features required for self-replication, nature would not favor any RNA molecule over any other. Natural selection as a process selects only

The RNA World 321

functionally advantageous features and only in self-replicating systems.

It passes its blind eye over molecules possessing merely necessary con

ditions or possible indicators of future function. Moreover, "it" does

nothing at all when mechanisms for replication and selection do not yet

even exist. In ribozyme-engineering experiments, engineers perform a

role in simulating natural selection that undirected natural processes

cannot play prior to the commencement of natural selection. Thus, even

if ribozyme experiments succeed in significantly enhancing the capaci

ties of RNA catalysts, it does not follow they will have demonstrated

the plausibility of an undirected process of chemical evolution. Insofar

as ribozyme-engineering experiments using a rational-design approach

(as opposed to a directed-evolution approach) involve an even more

overt role for intelligence, they exemplify the same problem (for ex

ample, see note 28).

Conclusion

As I have investigated various models that combined chance and neces

sity, I have noted an increasing sense of futility and frustration arising

among the scientists who work on the origin of life. As I surveyed the

literature, it became clear that this frustration had been building for

many years. In 1980, Francis Crick lamented, "An honest man, armed

with all the knowledge available to us now, could only state that in some

sense, the origin of life appears at the moment to be almost a miracle,

so many are the conditions which would have had to have been satisfied

to get it going."44 In 1988, the German biochemist and origin-of-life

researcher Klaus Dose followed suit with an equally critical assessment

of the state of the field. Dose explained that research efforts to date

had "led to a better perception of the immensity of the problem of the

origin of life on earth rather than to its solution. At present, all discus

sions on principal theories and experiments in the field either end in a

stalemate or a confession of ignorance."45 After attending a scientific

conference on the origin of life in 1989, one of my Cambridge supervi

sors returned to report, "The field is becoming increasingly populated

with cranks. Everyone knows everybody else's theory doesn't work, but

no one is willing to admit it about his own."

The Origins of the RNA World

Michael P. Robertson and Gerald F. Joyce

Departments of Chemistry and Molecular Biology and The Skaggs Institute for Chemical Biology,The Scripps Research Institute, La Jolla, California 92037

Correspondence: [email protected]

SUMMARY

The general notion of an “RNA World” is that, in the early development of life on the Earth,genetic continuity was assured by the replication of RNA and genetically encoded proteinswere not involved as catalysts. There is now strong evidence indicating that an RNA Worlddid indeed exist before DNA- and protein-based life. However, arguments regarding whetherlife on Earth began with RNA are more tenuous. It might be imagined that all of the compo-nents of RNA were available in some prebiotic pool, and that these components assembledinto replicating, evolving polynucleotides without the prior existence of any evolved macro-molecules. A thorough consideration of this “RNA-first” view of the origin of life must recon-cile concerns regarding the intractable mixtures that are obtained in experiments designed tosimulate the chemistry of the primitive Earth. Perhaps these concerns will eventually beresolved, and recent experimental findings provide some reason for optimism. However, theproblem of the origin of the RNA World is far from being solved, and it is fruitful to considerthe alternative possibility that RNA was preceded by some other replicating, evolving mole-cule, just as DNA and proteins were preceded by RNA.

Outline

1 Introduction

2 An “RNA-first” view of the origin of life

3 An “RNA-later” view of the origin of life

4 Concluding remarks

References

Editors: John F. Atkins, Raymond F. Gesteland, and Thomas R. Cech

Additional Perspectives on RNA Worlds available at www.cshperspectives.org

Copyright # 2012 Cold Spring Harbor Laboratory Press; all rights reserved; doi: 10.1101/cshperspect.a003608

Cite as Cold Spring Harb Perspect Biol 2012;4:a003608

1

1 INTRODUCTION

The general idea that, in the development of life on theEarth, evolution based on RNA replication preceded theappearance of protein synthesis was first proposed over40 yr ago (Woese 1967; Crick 1968; Orgel 1968). It was sug-gested that catalysts made entirely of RNA are likely to havebeen important at this early stage in the evolution of life,but the possibility that RNA catalysts might still be presentin contemporary organisms was overlooked. The unantici-pated discovery of ribozymes (Kruger et al. 1982; Guerrier-Takada et al. 1983) initiated extensive discussion of the roleof RNA in the origins of life (Sharp 1985; Pace and Marsh1985; Lewin 1986) and led to the coining of the phrase “theRNAWorld” (Gilbert 1986).

“The RNA World” means different things to differentinvestigators, so it would be futile to attempt a restrictivedefinition. All RNA World hypotheses include three basicassumptions: (1) At some time in the evolution of life,genetic continuity was assured by the replication of RNA;(2) Watson-Crick base-pairing was the key to replication;(3) genetically encoded proteins were not involved as cata-lysts. RNA World hypotheses differ in what they assumeabout life that may have preceded the RNA World, aboutthe metabolic complexity of the RNA World, and aboutthe role of small-molecule cofactors, possibly includingpeptides, in the chemistry of the RNAWorld.

There is now strong evidence indicating that an RNAWorld did indeed exist on the early Earth. The smokinggun is seen in the structure of the contemporary ribosome(Ban et al. 2000; Wimberly et al. 2000; Yusupov et al. 2001).The active site for peptide-bond formation lies deep withina central core of RNA, whereas proteins decorate the out-side of this RNA core and insert narrow fingers into it.No amino acid side chain comes within 18 A of the activesite (Nissen et al. 2000). Clearly, the ribosome is a ribozyme(Steitz and Moore 2003), and it is hard to avoid the conclu-sion that, as suggested by Crick, “the primitive ribosomecould have been made entirely of RNA” (1968).

A more tenuous argument can be made regardingwhether life on Earth began with RNA. In what has been re-ferred to as “The Molecular Biologist’s Dream” (Joyce andOrgel 1993), one might imagine that all of the componentsof RNAwere available in some prebiotic pool, and that thesecomponents could have assembled into replicating, evolvingpolynucleotides without the prior existence of any evolvedmacromolecules. However, a thorough consideration ofthis “RNA-first” view of the origin of life inevitably triggers“The Prebiotic Chemist’s Nightmare”, with visions of the in-tractable mixtures that are obtained in experiments designedto simulate the chemistry of the primitive Earth. Perhaps thiscontinuing nightmare will eventually have a happy ending,

and recent experimental findings provide some reason foroptimism. However, the problem of the origin of the RNAWorld is far from being solved, and it is fruitful to considerthe alternative possibility that RNA was preceded by someother replicating, evolving molecule, just as DNA and pro-teins were preceded by RNA.

2 AN “RNA-FIRST” VIEW OF THE ORIGIN OF LIFE

2.1 Abiotic Synthesis of Polynucleotides

This section considers the synthesis of oligonucleotidesfrom ß-D-nucleoside 50-phosphates, leaving aside for nowthe question of how the nucleotides became available onthe primitive Earth. Two fundamentally different chemicalreactions are involved. First, the nucleotide must be con-verted to an activated derivative, for example, a nucleoside50-polyphosphate. Next the 30-hydroxyl group of a nucleo-tide or oligonucleotide molecule must be made to reactwith the activated phosphate group of a monomer. Synthe-sis of oligonucleotides from nucleoside 30-phosphates willnot be discussed because activated nucleoside 20- or30-phosphates in general react readily to form 20,30-cyclicphosphates. These cyclic phosphates are unlikely to oligo-merize efficiently because the equilibrium constant fordimer formation is only of the order of 1.0 L/mol (Ermanand Hammes 1966; Mohr and Thach 1969). In the presenceof a complementary template somewhat larger oligomersmight be formed because the free energy of hybridizationwould help to drive forward the chain extension reaction.

In enzymatic RNA and DNA synthesis, the nucleoside50-triphosphates (NTPs) are the substrates of polymeriza-tion. Polynucleotide phosphorylase, although it is a deg-radative enzyme in nature, can be used to synthesizeoligonucleotides from nucleoside 50-diphosphates. Nucleo-side 50-polyphosphates are, therefore, obvious candidatesfor the activated forms of nucleotides. Although nucleoside50-triphosphates are not formed readily, the synthesis of nu-cleoside 50-tetraphosphates from nucleotides and inorganictrimetaphosphate provides a reasonably plausible prebioticroute to activated nucleotides (Lohrmann 1975). Othermore or less plausible prebiotic syntheses of nucleoside50-polyphosphates from nucleotides have also been re-ported (Handschuh et al. 1973; Osterberg et al. 1973; Reim-ann and Zubay 1999). Less clear, however, is how the firstphosphate would have been mobilized to convert the nu-cleosides to 50-nucleotides. Nucleoside 50-polyphosphatesare high-energy phosphate esters, but are relatively unreac-tive in aqueous solution. This may be advantageous forenzyme-catalyzed polymerization, but is a severe obstaclefor the nonenzymatic polymerization of nucleoside 50-pol-yphosphates, which would occur far more slowly than thehydrolysis of the resulting polynucleotide.

M.P. Robertson and G.F. Joyce

2 Cite as Cold Spring Harb Perspect Biol 2012;4:a003608

In a different approach to the activation of nucleotides,the isolation of an activated intermediate is avoided byusing a condensing agent such as a carbodiimide (Khorana1961). This is a popular method in organic synthesis,but its application to prebiotic chemistry is problematic.Potentially prebiotic molecules such as cyanamide andcyanoacetylene activate nucleotides in aqueous solution,but the subsequent condensation reactions are inefficient(Lohrmann and Orgel 1973).

Most attempts to study nonenzymatic polymerizationof nucleotides in the context of prebiotic chemistry haveused nucleoside 50-phosphoramidates, particularly nucleo-side 50-phosphorimidazolides. Although phosphorimida-zolides can be formed from imidazoles and nucleoside50-polyphosphates (Lohrmann 1977), they are only mar-ginally plausible as prebiotic molecules. They were chosenbecause they are prepared easily and react at a convenientrate in aqueous solution.

Nucleotides contain three principal nucleophilic groups:the 50-phosphate, the 20-hydroxyl, and the 30-hydroxyl group,in order of decreasing reactivity. The reaction of a nucleotideor oligonucleotide with an activated nucleotide, therefore,normally yields 50,50-pyrophosphate-, 20,50-phosphodiester-,

and30,50-phosphodiester-linkedadducts(Fig.1A), inorderofdecreasing abundance (Sulston et al. 1968). Thusthe conden-sation of several monomers would likely yield an oligomercontaining one pyrophosphate and a preponderance of20,50-phosphodiester linkages (Fig. 1B). There is little chanceof producing entirely 30,50-linked oligomers from activatednucleotides unless a catalyst can be found that increases theproportion of 30,50-phosphodiester linkages. Several metalions, particularly Pb2+ and UO2

2+, catalyze the formation ofoligomers fromnucleoside50-phosphorimidazolides(Sleeperand Orgel 1979; Sawai et al. 1988). The Pb2+-catalyzed reac-tion is especially efficient when performed in eutectic solu-tions of the activated monomers (in concentrated solutionsobtained by partial freezing of more dilute solutions). Sub-stantial amounts of long oligomers are formed under eutec-tic conditions, but the product oligomers always contain alarge proportion of 20,50-linkages (Kanavarioti et al. 2001;Monnard et al. 2003).

What kinds of prebiotically plausible catalysts mightlead to the production of 30,50-linked oligonucleotides di-rectly from nucleoside 50-phosphorimidazolides or otheractivated nucleotides? It is unlikely, but not impossible,that a metal ion or simple acid-base catalyst would provide

O

OH

OH

OH OH

OH

OPO

O–

O–

O–

O–

O– O–

O–

O–

O–

O– O– O–O

O

Ni Ni

OPO

O

O

OH

O

OH

OPO

O

O

OH

OPO

O

OH

OOPO

O

OHOHOHOH

OH

OPO

O

O

OH

ONi+1

N3

N2

N1

N4

N5

N6

Ni+1Ni+1

OP

O

OP

O

O

OH OH

Ni

O OO

OOPO

P

O

OP

O

O

O

OOPOO

A

B

Figure 1. Phosphodiester linkages resulting from chemical condensation of nucleotides. (A) Reaction of an activatedmononucleotide (Ni+1) with an oligonucleotide (N1 –Ni ) to form a 30,50-phosphodiester (left), 20,50-phospho-diester (middle), or 50,50-pyrophosphate linkage (right). (B) Typical oligomeric product resulting from chemicalcondensation of activated mononucleotides.

Origins of the RNA World

Cite as Cold Spring Harb Perspect Biol 2012;4:a003608 3

sufficient regiospecificity. The most attractive of the otherhypotheses is that adsorption to a specific surface of amineral might orient activated nucleotides rigidly andthus catalyze a highly regiospecific reaction.

The work of Ferris and coworkers provides supportfor this hypothesis (Ferris et al. 2004; Ferris 2006). Theyhave studied the oligomerization of nucleoside 50-phos-phorimidazolides and related activated nucleotides onthe clay mineral montmorillonite (Ferris and Ertem1993; Kawamura and Ferris 1994; Miyakawa and Ferris2003). Some samples of the mineral are effective catalysts,promoting the formation of oligomers even from dilute sol-utions of activated nucleotide substrates. Furthermore, themineral profoundlyaffectsthe regiospecificityof the reaction.The oligomerization of adenosine 50-phosphorimidazolide,for example, gives predominantly 30,50-linked products inthe presence of montmorillonite, but predominantly 20,50-linked products in aqueous solution (Ding et al. 1996;Kawamura and Ferris 1999). Once short oligomers havebeen synthesized, they can be further extended by adsorbingthem on either montmorillonite or hydroxylapatite andrepeatedly adding activated monomers, resulting in theaccumulation of mainly 30,50-linked oligoadenylates up to40–50 subunits in length (Ferris et al. 1996; Ferris 2002).However, even when adsorbed on montmorillonite, thephosphorimidazolides of the pyrimidine nucleosides yieldoligomers that are predominantly 20,50-linked.

Long oligomers have also been obtained from mono-mers in a single step using a different activated nucleotidein which imidazole is replaced by 1-methyl-adenine (Praba-har and Ferris 1997; Huang and Ferris 2003). Using the1-methyl-adenine derivative of adenylate or uridylate, oligo-mers containing up to 40 subunits were produced, consist-ing of �75% 30,50-linkages for oligoadenylate and �60%30,50-linkages for oligouridylate (Huang and Ferris 2006).Oligomerization of the 1-methyl-adenine derivative of gua-nylate or cytidylate was less efficient, but all four activatedmonomers could be co-incorporated, to at least a modestextent, within abiotically synthesized oligonucleotides.

Detailed analysis of this work on catalysis by montmor-illonite suggests that oligomerization occurs at a limitednumber of structurally specific active sites within the inter-layers of the clay (Wang and Ferris 2001). These sites mustnot be saturated with sodium ions, which appear to blockaccess of the activated nucleotides (Joshi et al. 2009). Sev-eral different samples of montmorillonite have proven to begood catalysts, in part depending on their proton versus so-dium ion content. It will be important to determine if thereare other types of minerals that are comparably efficientcatalysts of oligonucleotide synthesis, and if so, to studythe regiospecificity and sequence generality of the reactionsthey catalyze.

2.2 Nonenzymatic Replication of RNA

If a mechanism existed on the primitive Earth for the poly-merization of activated nucleotides, it would have gener-ated a complex mixture of product oligonucleotides thatdiffered in both length and sequence. The next stage inthe emergence of an RNAWorld would have been the rep-lication of some of these molecules, so that a process equiv-alent to natural selection could begin. The reaction centralto replication of nucleic acids is template-directed synthe-sis, that is, the synthesis of a complementary oligonucleo-tide under the direction of a preexisting oligonucleotide.A good deal of work has already been performed on thisaspect of nonenzymatic replication. This work has beenreviewed elsewhere (Joyce 1987; Orgel 2004a), so only asummary of the results will be given here.

The first major conclusion is that most activated nu-cleotides do not undergo efficient, regiospecific, template-directed reactions in the presence of an RNA or DNAtemplate. In general, only a small proportion of templatemolecules succeed in directing the synthesis of a completecomplement, and the complement usually contains a mix-ture of 20,50- and 30,50-phosphodiester linkages. After a con-siderable search, a set of activated nucleotides was foundthat undergo efficient and highly regiospecific template-directed reactions. Working with guanosine 50-phospho-2-methylimidazolide (2-MeImpG), it was shown thatpoly(C) can direct the synthesis of long oligo(G)s in a reac-tion that is highly efficient and highly regiospecific (Inoueand Orgel 1981). If poly(C) is incubated with an equimolarmixture of the four 2-MeImpNs (N ¼ G, A, C, or U), lessthan 1% of the product consists of noncomplementary nu-cleotides (Inoue and Orgel 1982). Subsequent experimentssuggested that this and the related reactions discussed lateroccur preferentially within the context of double helicesthat have a structure resembling the A form of RNA(Kurz et al. 1997, 1998; Kozlov et al. 1999, 2000).

Random copolymers containing an excess of C residuescan be used to direct the synthesis of products containingG and the complements of the other bases present in thetemplate (Inoue and Orgel 1983). The reaction with apoly(C,G) template is especially interesting because theproducts, like the template, are composed entirely of Cand G residues. If these products in turn could be used astemplates, it might allow the emergence of a self-replicatingsequence. Self-replication, however, is unlikely, mainly be-cause poly(C,G) molecules that do not contain an excess ofC residues tend to form stable self-structures that preventthem from acting as templates (Joyce and Orgel 1986).The self-structures are of two types: (1) the standardWatson-Crick variety based on C†G pairs, and (2) a quad-rahelix structure that results from the association of four



G-rich sequences. As a consequence, any C-rich oligonu-cleotide that can serve as a good template will give rise toG-rich complementary products that tend to be locked inself-structure and so cannot act as templates. Overcomingthe self-structure problem using the standard C and G nu-cleotides is very difficult because it requires the discoveryof conditions that favor the binding of mononucleotidesto allow template-directed synthesis to occur, but suppressthe formation of long duplex regions that would excludeactivated monomers from the template.

Some progress has been made in discovering defined-sequence templates that are copied faithfully to yield com-plementary products (Inoue et al. 1984; Acevedo and Orgel1987; Wu and Orgel 1992a). Successful templates typicallycontain an excess of C residues, with A and U residuesisolated from each other by at least three C residues. Runsof G residues are copied into runs of C residues, so longas the formation of self-structures by G residues can beavoided (Wu and Orgel 1992b). In light of the availableevidence, it seems unlikely that a pair of complementarysequences can be found, each of which facilitates thesynthesis of the other using nucleoside 50-phospho-2-methylimidazolides as substrates. Some of the obstaclesto self-replication may be attributable to the choice ofreagents and reaction conditions, but others seem to beintrinsic to the template-directed condensation of acti-vated mononucleotides.

A related nonenzymatic replication scheme involvessynthesis by the ligation of short 30,50-linked oligomers(James and Ellington 1999). This is certainly an attractivepossibility, made more plausible by the discovery of analo-gous ribozyme-catalyzed reactions (Bartel and Szostak1993), but it faces two major obstacles. The first is thedifficulty of obtaining the substrates in the first place.The second is concerned with fidelity. Pairs of oligonucleo-tides containing a single base mismatch, particularly ifthe mismatch forms a G†U wobble pair, still hybridize asefficiently as fully complementary oligomers, except in atemperature range very close to the melting point of theperfectly paired structure. Maintaining fidelity wouldtherefore be difficult under any plausible temperatureregime.

Despite these problems, template-directed ligation ofshort oligonucleotides may be a viable alternative to the oli-gomerization of activated monomers. Ferris’ work dis-cussed above suggests that predominantly 30,50-linkedoligonucleotides might form spontaneously from activatednucleotides on some variety of montmorillonite (Ferriset al. 1996) or on some other mineral. Oligonucleotide50-triphosphates undergo slow but remarkably 30,50-regio-specific ligation in the presence of a complementary tem-plate (Rohatgi et al. 1996a,b). The combination of some

such pair of reactions might provide a replication schemefor polynucleotides starting with an input of activatedmonomers.

There also are efforts in what is sometimes termed “syn-thetic biology” to achieve nonenzymatic replication withmolecules that resemble biological nucleic acids, but arenot constrained by considerations of plausible prebioticchemistry. For example, the 20- and 30-hydroxyl groups ofactivated mononucleotides can be replaced by an aminogroup at either position, providing enhanced nucleophilic-ity and resulting in more rapid template-dependent (andtemplate-independent) oligomerization (Lohrmann andOrgel 1976; Zielinski and Orgel 1985). Dinucleotide build-ing blocks, consisting of 30-amino, 30-deoxynucleotide ana-logs can also be oligomerized in the presence of a suitablecondensing agent (Zielinski and Orgel 1987). With addi-tional modification of the nucleotide bases, it has been pos-sible to carry out the template-directed copying of nucleicacid sequences that contain of all four bases (Schrum et al.2009). These efforts, although not explaining the origin ofthe RNAWorld, contribute to understanding the chemicalchallenges that must be overcome in achieving the non-enzymatic replication of RNA.

2.3 The First RNA Replicase

The notion of the RNAWorld places emphasis on an RNAmolecule that catalyzes its own replication. Such a moleculemust function as an RNA-dependent RNA polymerase,acting on itself (or copies of itself ) to produce complemen-tary RNAs, and acting on the complementary RNAs to pro-duce additional copies of itself. The efficiency and fidelityof this process must be sufficient to produce viable “prog-eny” RNA molecules at a rate that exceeds the rate ofdecomposition of the “parents.” Beyond these require-ments, the details of the replication process are not highlyconstrained.

The RNA-first view of the origin of life assumes that asupply of activated ß-D-nucleotides was available by someas yet unrecognized abiotic process. Furthermore, it as-sumes that a means existed to convert the activated nucleo-tides to an ensemble of random-sequence polynucleotides,a subset of which had the ability to replicate. It seems to beimplicit in the model that such polynucleotides replicatethemselves but, for whatever reason, do not replicate unre-lated neighbors. It is not clear whether replication involvesone molecule copying itself (and its complement) or a fam-ily of molecules that together copy each other. These ques-tions are set aside for the moment in order to first considerthe question of whether an RNA molecule of reasonablyshort length can catalyze its own replication with suffi-ciently high fidelity.



Accuracy and Survival. The concept of an error threshold,that is, an upper limit to the frequency of copying errorsthat can be tolerated by a replicating macromolecule, wasfirst introduced by Eigen (1971). This important idea hasbeen extended in a series of mathematically sophisticatedpapers by McCaskill, Schuster, and others (McCaskill1984a; Eigen et al. 1988; Schuster and Swetina 1988).Here only a brief summary of the subject is provided.

Eigen’s model (1971) envisages a population of repli-cating polynucleotides that draw on a limited supply ofactivated mononucleotides to produce additional copiesof themselves. In this model, the rate of synthesis of newcopies of a particular replicating RNA is proportionalto its concentration, resulting in autocatalytic growth.The net rate of production is the difference between therate of formation of error-free copies and the rate ofdecomposition of existing copies of the RNA. For an ad-vantageous RNA to outgrow its competitors, its net rateof production must exceed the mean rate of productionof all other RNAs in the population. Only the error-freecopies of the advantageous RNA contribute to its netrate of production, but all the copies of the other RNAscontribute to their collective production. Thus the relativeadvantage enjoyed by the advantageous individual com-pared with the rest of the population (often referred to asthe “superiority” of the advantageous individual) mustexceed the probability of producing an error copy of thatadvantageous individual.

The proportion of copies of an RNA that are error freeis determined by the fidelity of the component condensa-tion reactions that are required to produce a completecopy. For simplicity, consider a self-replicating RNA thatis formed by n condensation reactions, each having meanfidelity q. The probability of obtaining a completely error-free copy is given by qn, which is the product of the fidelityof the component condensation reactions. If an advanta-geous individual is to outgrow its competitors, qn mustexceed the superiority, s, of that individual. Expressed interms of the number of reactions required to produce theadvantageous individual,

n , j ln sj=j ln qj:

For s . 1 and q . 0.9, this equation simplifies to

n , ln s=(1� q):

This is the “error threshold,” which describes the inverserelationship between the fidelity of replication, q, and themaximum allowable number of component condensa-tion reactions, n. The maximum number of component

reactions is highly sensitive to the fidelity of replication,but depends only weakly on the superiority of the advanta-geous individual. For a self-replicating RNA that is formedby the template-directed condensation of activated mono-nucleotides, a total of 2n – 2 condensation reactions arerequired to produce a complete copy. This takes into ac-count the synthesis of both a complementary strand anda complement of the complement.

It should be recognized that a marked superiority of onesequence over all other sequences could not be maintainedover evolutionary time because novel variants would soonarise to challenge the dominant species. However, a markedinitial superiority may be important in allowing an efficientself-replicating RNA to emerge from a pool of less efficientreplicators. In the absence of other efficient replicators, aprimitive self-replicating RNA that operates with low fidel-ity may gain a foothold by taking advantage of a somewhatless stringent error threshold. Whether or not this canoccur depends on its superiority. For example, an RNAthat replicates 10-fold more efficiently than its competitorsand does so with 90% fidelity could be no longer than 12nucleotides, and a similarly advantageous RNA that repli-cates with 70% fidelity could be no longer than fournucleotides. It seems highly unlikely than any of the 17 mil-lion possible RNA dodecamers is able to catalyze its ownreplication with 90% fidelity, and even less likely that a tet-ranucleotide could catalyze its own replication with 70% fi-delity. However, an RNA that replicates 106-fold moreefficiently than its competitors and does so with 90% fidel-ity could be as long as 67 nucleotides, and one that repli-cates with 70% fidelity could be as long as 20 nucleotides.

When self-replication is first established, fidelity islikely to be poor and there is strong selection pressure favor-ing improvement of the fidelity. As fidelity improves, alarger genome can be maintained. This allows explorationof a larger number of possible sequences, some of whichmay lead to further improvement in fidelity, which inturn allows a still larger genome size, and so on. Once theevolving population has achieved a fidelity of about 99%,a genome length of about 100 nucleotides can be main-tained, even for modest superiority values. This wouldallow RNA-based life to become firmly established. Untilthat time, it is a race between evolutionary improvementin the context of a sloppy self-replicating system and therisk of delocalization of the genetic information becauseof overstepping the error threshold. If the time requiredto bootstrap to high fidelity and large genomes is toolong, there is a risk that the population will succumb toan environmental catastrophe before it has had the chanceto develop appropriate countermeasures.

It is difficult to state with certainty the minimum pos-sible size of an RNA replicase ribozyme. An RNA consisting



of a single secondary structural element, that is, a smallstem-loop containing 12–17 nucleotides, would not be ex-pected to have replicase activity, whereas a double stem-loop, perhaps forming a “dumbbell” structure or a pseudo-knot, might just be capable of a low level of activity. A triplestem-loop structure, containing 40–60 nucleotides, offersa reasonable hope of functioning as a replicase ribozyme.One could, for example, imagine a molecule consisting ofa pseudoknot and a pendant stem-loop that forms a cleftfor template-dependent replication.

Suppose there is some 40-mer that enjoys a superiorityof 103-fold and replicates with 90% fidelity. This should beregarded as a highly optimistic but not outrageous view ofwhat is possible for a minimum replicase ribozyme. Wouldsuch a molecule be expected to occur within a populationof random-sequence RNAs? A complete library consistingof one copy each of all 1024 possible 40-mers would weighabout 1 kg. There may be many such 40-mers, encompass-ing both distinct structural motifs and, more importantly, alarge number of equivalent representations of each motif.As a result, even a small fraction of the total library, consist-ing of perhaps 1020 sequences and weighing about 1 g,might be expected to contain at least one self-replicatingRNAwith the requisite properties. It is not sufficient, how-ever, that there be just one copy of a self-replicating RNA.The above calculations assume that a self-replicating RNAcan copy itself (or that a fully complementary sequence isautomatically available, as will be discussed later). If twoor more copies of the same 40mer RNA are needed, thena much larger library, consisting of 1048 RNAs and weigh-ing 1028 g would be required. This amount is comparable tothe mass of the Earth.

At first sight, it might seem that one way to ease the er-ror threshold would be for the replicase ribozyme to acceptdinucleotide or trinucleotide substrates, so that copies ofthe RNA could be formed by fewer condensation reactions.Calculations show that, over a broad range of superiorityvalues, RNAs that are required to replicate with 90% fidelitywhen using mononucleotide substrates would be requiredto replicate with roughly 80% fidelity when using dinucleo-tide substrates or roughly 70% fidelity when using trinu-cleotide substrates. Thus the use of short oligomers offersonly a modest advantage because of lessening of theerror threshold, which likely would be outweighed by thegreater difficulty of achieving high fidelity when dis-criminating among the 16 possible dinucleotide or 64possible trinucleotide substrates, rather than among thefour mononucleotides.

If one accepts the RNA-first view that there was a prebi-otic pool of random-sequence RNAs, and if one assumesthat the pool included a replicase ribozyme containing,say, 40 nucleotides and replicating itself with about 90%

fidelity, then it is not difficult to imagine how RNA-basedevolution might have started. During the initial perioda successful clone would have expanded in the absenceof competition. As competition for substrates intensifiedthere would have followed a succession of increasinglymore advantageous individuals, each replicating withinits error threshold. After a period of intensifying com-petition, the single most advantageous species wouldhave been replaced by a “quasispecies,” that is, a mixtureof the most advantageous individual and substantialamounts of closely related individuals that replicate almostas fast and almost as faithfully as the most advantageousone (Eigen and Schuster 1977; Eigen et al. 1988). Underthese conditions the persistence of a particular advanta-geous individual is no longer the problem, but one mustunderstand the evolution of the composition of the quasis-pecies and the conditions for its persistence. This difficultproblem has been partially solved by McCaskill (1984b).The general form of the solution is very similar to the errorthreshold described by Eigen (1971), but with differentvalues for the constant in the inequality. Thus concernsabout the error threshold apply to the quasispecies aswell as to the succession of individuals. Practically speak-ing, however, once a quasispecies distribution of sophisti-cated replicators had emerged, the RNA World wouldhave been on solid footing and, barring an environmentalcatastrophe, unlikely to lose the ability to maintain geneticinformation over time.

Another Chicken-and-Egg Paradox. The previous dis-cussion has tried mightily to present the most optimisticview possible for the emergence of an RNA replicaseribozyme from a soup of random-sequence polynucleo-tides. It must be admitted, however, that this model doesnot appear to be very plausible. The discussion has focusedon a straw man: The myth of a small RNA molecule thatarises de novo and can replicate efficiently and with highfidelity under plausible prebiotic conditions. Not only issuch a notion unrealistic in light of current understandingof prebiotic chemistry (Joyce 2002), but it should strain thecredulity of even an optimist’s view of RNA’s catalyticpotential. If you doubt this, ask yourself whether youbelieve that a replicase ribozyme would arise in a solutioncontaining nucleoside 50-diphosphates and polynucleotidephosphorylase!

If one accepts the notion of an RNAWorld, one is facedwith the dilemma of how such a genetic system came intoexistence. To say that the RNAWorld hypothesis “solves theparadox of the chicken-and-the-egg” is correct if onemeans that RNA can function both as a genetic moleculeand as a catalyst that promotes its own replication. RNA-catalyzed RNA replication provides a chemical basis for



Darwinian evolution based on natural selection. Darwin-ian evolution is a powerful way to search among vast num-bers of potential solutions for those that best address aparticular problem. Selection based on inefficient RNAreplication, for example, could be used to search among apopulation of RNA molecules for those individuals thatpromote improved RNA replication. But here one encoun-ters another chicken-and-egg paradox: Without evolutionit appears unlikely that a self-replicating ribozyme couldarise, but without some form of self-replication there isno way to conduct an evolutionary search for the first,primitive self-replicating ribozyme.

One way that RNA evolution may have gotten startedwithout the aid of an evolved catalyst might be by usingnonenzymatic template-directed synthesis to permit somecopying of RNA before the appearance of the first replicaseribozyme. Suppose that the initial ensemble of monomerswas not produced by random copolymerization, but ratherby a sequence of untemplated and templated reactions(Fig. 2), and further suppose that members of the initialensemble of multiple stem-loop structures could be repli-cated, albeit inefficiently, by the template-directed process.This would have two important consequences. First, anymolecule with replicase function that appeared in themixture would likely find in its neighborhood similar

molecules and their complements, related by descent, thuseliminating the requirement for two unrelated replicasesto meet. Second, a majority of molecules in the mixturewould contain stem-loop structures. If it is true that ribo-zyme function is favored by stable self-structure, and if thebase-sequences of the stems in stem-loop structures are rel-atively unimportant for function, this model might providean economical way of generating a relatively small ensembleof sequences that is enriched with catalytic sequences.

How plausible is the assumption that replicases couldact on sequences similar to themselves, while ignoringunrelated sequences? This selectivity could be ensured bysegregating individual molecules (or clonal lines) on thesurface of mineral grains, on the surface of micelles, orwithin membranes. Closely related molecules might besegregated as a group through specific hydrogen-bondinginteractions (the family that sticks together, replicates to-gether). For any segregation mechanism, weak selectionwould result if the replicating molecules are sufficientlydispersed that diffusion over their intermolecular distanceis slow compared with replication. Computer simulationshave shown that under such conditions of segregation,evolutionary bootstrapping can occur, resulting in pro-gressively larger genomes that are copied with progres-sively greater fidelity (Szabo et al. 2002). Alternatively, the

(etc.)

Figure 2. Nonenzymatic synthesis of multi-stem-loop structures as a result of untemplated (open arrowhead) andtemplated (filled arrowhead) reactions. Template-directed synthesis is assumed to occur rapidly whenever atemplate, activated monomers, and a suitable primer are available. Once the complementary strand is completed,additional residues are added slowly in a random-sequence manner.



requirement for replication of related, but not unrelated,sequences might be met through the use of “genomictags” (Weiner and Maizels 1987). Among self-replicatingsequences, it is plausible that some are restricted to copyingmolecules with a particular 30-terminal subsequence. Areplicator that happened by chance to carry a terminalsequence that matched the preference of its active sitewould replicate itself while ignoring its neighbors.

Another resolution of the paradox of how RNA evolu-tion was initiated without the aid of an evolved ribozymeis to abandon the RNA-first view of the origin of life andsuppose that RNA was not the first genetic molecule(Cairns-Smith 1982; Shapiro 1984; Joyce et al. 1987; Joyce1989, 2002; Orgel 1989, 2004a). Perhaps RNA replicationarose in the context of an evolving system based on some-thing other than RNA (see the section “Alternative GeneticSystems”). Even if this is true, all of the arguments concern-ing the relationship between the fidelity of replication andthe maximum allowable genome length would still apply tothis earlier genetic system. Of course, the challenge to thosewho advocate the RNA-later approach is to show that thereis an informational entity that is prebiotically plausible andis capable of initiating its own replication without the aid ofa sophisticated catalyst.

2.4 Replicase Function in the Evolved RNA World

Although it is difficult to say how the first RNA replicaseribozyme arose, it is not difficult to imagine how such amolecule, once developed, would function. The chemistryof RNA replication would involve the template-directedpolymerization of mono- or short oligonucleotides, usingchemistry in many ways similar to that used by contempo-rary group I ribozymes (Cech 1986; Been and Cech 1988;Doudna and Szostak 1989). One important difference isthat, unlike group I ribozymes, which rely on a nucleosideor oligonucleotide leaving group, an RNA replicase wouldmore likely make use of a different leaving group that pro-vides a substantial driving force for polymerization andthat, after its release, does not become involved in somecompeting phosphoester transfer reaction.

From Ligases to Polymerases. The polymerization of ac-tivated nucleotides proceeds via nucleophilic attack by the30-hydroxyl of a template-bound oligonucleotide at thea-phosphorus of an adjacent template-bound nucleotidederivative (Fig. 3). The nucleotide is “activated” for attackby the presence of a phosphoryl substituent, for example aphosphate, polyphosphate, alkoxide, or imidazole group.As discussed previously, polyphosphates, such as inorganicpyrophosphate, are the most obvious candidates for theleaving group. The condensation reaction could be assisted

by favorable orientation of the reactive groups, deprotona-tion of the nucleophilic 30-hydroxyl, stabilization of thetrigonal-bipyramidal transition state, and charge neutrali-zation of the leaving group. All of these tasks might be per-formed by RNA (Narlikar and Herschlag 1997; Emilssonet al. 2003), acting either alone (Ortoleva-Donnelly andStrobel 1999) or with the help of a suitably positionedmetal cation or other cofactor (Shan et al. 1999; Shanet al. 2001).

The possibility that an RNA replicase ribozyme couldhave existed has been made abundantly clear by work in-volving ribozymes that have been developed in the labora-tory through in vitro evolution (Bartel and Szostak 1993;Ekland et al. 1995; Ekland and Bartel 1996; Robertsonand Ellington 1999; Jaeger et al. 1999; Rogers and Joyce2001; Johnston et al. 2001; McGinness and Joyce 2002;Ikawa et al. 2004; Fujita et al. 2009). Bartel and Szostak(1993), for example, began with a large population ofrandom-sequence RNAs and evolved the “class I” RNAligase ribozyme, an optimized version of which is about100 nucleotides in length and catalyzes the joining of twotemplate-bound oligonucleotides. Condensation occursbetween the 30-hydroxyl of one oligonucleotide and the50-triphosphate of another, forming a 30,50-phosphodiesterlinkage and releasing inorganic pyrophosphate. This reac-tion is classified as ligation because of the nature of theoligonucleotide substrates, but involves the same chemicaltransformation as is catalyzed by modern RNA polymeraseenzymes.

X-ray crystal structures of two RNA ligase ribozymes,the L1 and above-mentioned class I ligases, have been deter-mined, providing a glimpse into the mechanistic strategiesthat these two structurally and evolutionarily distinct ribo-zymes use to catalyze the same reaction (Robertson and

O

OH

OH

OH OH

OH

O P O

O–

O–

O

O Ni

Ni–1

O P O

O

O Ni+1 O P

R O

O –

Figure 3. Nucleophilic attack by the 30-hydroxyl of a template-boundoligonucleotide (N1 –Ni) on the a-phosphorus of an adjacenttemplate-bound mononucleotide (Ni+1). Dotted lines indicate basepairing to a complementary template. R is the leaving group.



Scott 2007; Shechner et al. 2009) (Fig. 4). Both crystalstructures capture the product of the ligation reaction,and consequently offer an incomplete view of the reactionpathway. For example, the pyrophosphate leaving group isabsent from the structures, so no conclusions can be drawnregarding potential ribozyme-assisted orientation of thereactive triphosphate or charge neutralization of the pyro-phosphate leaving group. There is, however, informationregarding other aspects of the reaction mechanism thatcan be inferred from the product structures.

Both the L1 and class I ligases are dependent on thepresence of magnesium ions for their activity. A prominentfeature of the L1 structure (Fig. 4A) involves a bound metalion in the active site, coordinated by three nonbridgingphosphate oxygens, one of which belongs to the newlyformed phosphodiester linking what originally were thetwo substrates. This magnesium ion is favorably positionedto help neutralize the increased negative charge of the tran-sition state and, potentially, to activate the 30-hydroxyl nu-cleophile and to help orient the a-phosphate for a moreoptimal in-line alignment. In the case of the class I struc-ture (Fig. 4B), no catalytic metal ions appear to have beenretained in the vicinity of the active site, although two mag-nesium ions are observed to participate in crucial structuralinteractions that help shape the active site architecture.Despite the lack of direct observation of a catalytic metalat the active site, there is what appears to be an empty metalbinding site formed by two nonbridging phosphate oxy-gens, positioned directly opposite the ligation junction ina manner similar to that observed for the magnesium bind-ing site of the L1 ligase and reminiscent of the arrangementseen in protein polymerases. The lack of a metal in the

crystal structure may simply be an artifact of the crystalliza-tion process or may imply a local conformational change inthe product that disfavors retention of the bound metal.These structures show that, despite some remaining gapsin the detailed understanding of how these ribozymesfunction, the available information points to a universalcatalytic strategy, very similar to that used by modernprotein-based RNA polymerases.

Subsequent to its isolation as a ligase, the class I ribo-zyme was shown to catalyze a polymerization reaction inwhich the 50-triphosphate-bearing oligonucleotide is re-placed by one or more NTPs (Ekland and Bartel 1996).This reaction proceeds with high fidelity (q ¼ 0.92), butthe reaction rate drops sharply with successive nucleotideadditions.

Bartel and colleagues performed further in vitro evolu-tion experiments to convert the class I ligase to a bona fideRNA polymerase that operates on a separate RNA template(Johnston et al. 2001). To the 30 end of the class I ligase theyadded 76 random-sequence nucleotides that were evolvedto form an accessory domain that assists in the polymeriza-tion of template-bound NTPs. The polymerization reac-tion is applicable to a variety of template sequences, andfor well-behaved sequences proceeds with an average fidel-ity of 0.967. This would be sufficient to support a genomelength of about 30 nucleotides, although the ribozymeitself contains about 190 nucleotides. The ribozyme has acatalytic rate for NTP addition of at least 1.5 min21, butits Km is so high that, even in the presence of micromolarconcentrations of oligonucleotides and millimolar concen-trations of NTPs, it requires about 2 h to complete eachNTP addition (Lawrence and Bartel 2003). The ribozyme

A B

Figure 4. X-ray crystal structure of the (A) L1 ligase and (B) class I ligase ribozymes. Insets show the putative mag-nesium ion binding sites at the respective ligation junctions. The structures are rendered in rainbow continuum,with the 50-triphosphate-bearing end of the ribozyme colored violet and the 30-hydroxyl-bearing end of the substratecolored red. The phosphate at the ligation junction is shown in white, and the proximate magnesium ion (modeledfor the class I ligase) is shown as a yellow sphere, with dashed lines indicating coordination contacts.



operates best under conditions of high Mg2+ concentra-tion, but becomes degraded under those conditions over24 h, by which time it has added no more than 14 NTPs(Johnston et al. 2001).

Further optimization of the polymerase ribozyme usinghighly sophisticated in vitro evolution techniques has led toadditional improvements in its biochemical properties. Bydirectly selecting for extension of an external primer on aseparate template, Zaher and Unrau (2007) were able toimprove the maximum length of template-dependentpolymerization to .20 nucleotides, with a rate that is�threefold faster than that of the parent for the first ninemonomer additions and up to 75-fold faster for additionsbeyond 10 nucleotides. In addition, although not rigor-ously quantitated, the new ribozyme displays significantlyimproved fidelity, particularly with respect to discrimina-tion against G†U wobble pairs. It is this improved fidelitythat appears to be the underlying source for the observedimprovements in the maximum length of extension andthe rate of polymerization.

A different RNA ligase ribozyme can operate on a sep-arate RNA template in a largely sequence-general manner,and does so with a Km that is at least 100-fold lower thanthat of the class-I-derived polymerase (McGinness andJoyce 2002). However, its catalytic rate is much lower aswell, and it is unable to add more than a single NTP. Yetanother RNA ligase ribozyme can operate on a separatetemplate with the help of designed tertiary interactionsthat “clamp” the template–substrate complex to the ribo-zyme (Ikawa et al. 2004). But it too is a relatively slow cata-lyst that cannot add more than a single NTP.

A highly pessimistic view is that because there is noknown polymerase ribozyme that combines all of the prop-erties necessary to sustain its own replication, no such ribo-zyme is possible. A more balanced view is that RNA clearlyis capable of greatly accelerating the template-dependentpolymerization of nucleoside 50-polyphosphates. Suchcatalytic RNAs can operate in a sequence-general mannerand with reasonable fidelity. It seems only a matter oftime (and likely considerable effort) before more robustpolymerase ribozymes will be obtained. Nature did nothave the opportunity to conduct carefully arranged evolu-tion experiments using highly-purified reagents, but didhave the luxury of much greater reaction volumes andmuch more time.

RNA Replication. Despite falling short of the ultimategoal of a general-purpose RNA polymerase ribozyme, arobust reaction system for RNA-catalyzed RNA repli-cation has recently been shown. The system uses a pair ofcross-replicating ligase ribozymes that each catalyze theformation of the other, using a mixture of four different

substrate oligonucleotides (Lincoln and Joyce 2009). In re-action mixtures containing only these RNA substrates,MgCl2, and buffer, a small starting amount of ribozymesgives rise to many additional ribozymes through a processof RNA-catalyzed exponential amplification. Whenever thesubstrates become depleted, the replication process can berestarted and sustained indefinitely by replenishing thesupply of substrates.

Because the substrates are recognized by the ribozymesthrough specific Watson-Crick pairing interactions, evolu-tion experiments can be performed by providing a varietyof substrates that have different sequences in these recogni-tion regions and different corresponding sequences in thecatalytic domain of the ribozyme. RNA replication wasperformed with a library of 144 possible substrate combi-nations, resulting in the emergence of a set of highly advan-tageous replicators that included recombinants which werenot present at the start of the experiment. Until the adventof a general-purpose RNA polymerase ribozyme, the sys-tem of cross-replicating ligases offers the best platform tostudy the biochemical properties and evolutionary behav-ior of an all-RNA replicative system.

Nucleotide Biosynthesis. RNA replicase activity isprobably not the only catalytic behavior that was essentialfor the existence of the RNA World. Maintainingan adequate supply of the four activated nucleotideswould have been a top priority. Even if the prebiotic envi-ronment contained a large reservoir of these compounds,the reservoir would eventually become depleted, andsome capacity for nucleotide biosynthesis would havebeen required.

Ribozymes have been obtained, through in vitro evolu-tion, that catalyze some of the steps of nucleotide biosyn-thesis. Unrau and Bartel (1998), for example, developed aribozyme that catalyzes a reaction between 4-thiouraciland 5-phosphoribosyl-1-pyrophosphate (PRPP) to form4-thiouridylate (Fig. 5A). The 4-thiouracil is provided freein solution and the PRPP is tethered to the 30 end of theribozyme. An optimized form of this ribozyme, containing124 nucleotides, has an observed rate of 0.2 min21 in thepresence of 4 mM 4-thiouracil (Chapple et al. 2003). Thisis at least 107-fold faster than the uncatalyzed rate ofreaction, which is too slow to measure. Unrau and col-leagues used a similar approach to develop two differentribozymes that catalyze the formation of 6-thioguanylatefrom 6-thioguanine and tethered PRPP (Lau et al. 2004),as well as a third guanylate synthase ribozyme that aroseas an unanticipated consequence of a related in vitroevolution experiment (Lau and Unrau 2009). The firsttwo guanylate synthase ribozymes are slightly larger andhave about twofold higher catalytic efficiency compared



with the uridylate synthase ribozyme, although guanylatesynthesis is expected to have a much higher uncatalyzedrate of reaction.

RNA-catalyzed synthesis of PRPP has not been shown,but a ribozyme has been obtained that catalyzes the50-phosphorylation of oligonucleotides using g-thio-ATPasthe phosphate donor (Lorsch and Szostak 1994) (Fig. 5B).The ribozyme shows a rate enhancement of about 109-fold compared with the uncatalyzed rate of reaction.Once a nucleoside 50-phosphate has been formed, it can-phosphoryl be activated by another ribozyme that cataly-zes the condensation of a nucleoside 50-phosphate and aribozyme-tethered nucleoside 50-triphosphate (Huangand Yarus 1997) (Fig. 5C). This results in the formationof a 50,50-pyrophosphate linkage, which provides an acti-vated nucleotide leaving group that can drive subsequentRNA-catalyzed, template-directed ligation of RNA (Hagerand Szostak 1997) (Fig. 5D).

None of these four RNA-catalyzed reactions has pre-cisely the right format for the corresponding reaction in ahypothetical nucleotide biosynthesis pathway in the RNAWorld. However, they show that RNA is capable of per-forming the relevant chemistry with substantial catalyticrate enhancement. It remains to be seen whether ribozymescan be developed that catalyze the formation of the funda-mental building blocks of RNA, D-ribose and the four nu-cleotide bases, using starting materials that would havebeen abundant on the primitive Earth.

3 AN “RNA-LATER” VIEW OF THE ORIGINOF LIFE

3.1 Abiotic Synthesis of Nucleotides

The RNA-first view of the origin of life proceeds from theassumption that pure ß-D-nucleotides were available insome prebiotic pool. How close to such a pool could one

NH

NH

NH

O

S

O

OH OH OH OH

OP

POPOO O

–O –OO–

O– O–O–

O

OO N O

S

PO

O

+ + HP2O73–

A

OH OHOH OHO–

OHO BaseOOPPO OP

O– O– O–

–SO O O

N

N

N

N

NH2

OO Base–SOP+ + ADP

B

OH OH OH

OH

OH

OH

O–

O– O– O– O– O– O–

O O OOPP OO O

BaseBaseOOPP O

O OP O–OO

Base+ + HP2O73–O

O

O PO

Base

C

OH

OH

OH

OHOH

OH

O–

O– O–

O–

O–

O O OOPP OO O

BaseA

+ + AMPOO BaseP

OO OO BaseP

OO

OO BasePO

OH

O

D

Figure 5. Known RNA-catalyzed reactions that are relevant to nucleotide biosynthesis. (A) Formation of 4-thiouridylate from free 4-thiouracil and ribozyme-tethered 5-phosphoribosyl-1-pyrophosphate. (B) 50-phosphor-ylation of an oligonucleotide using g-thio-ATP as the phosphate donor. (C) Activation of a nucleoside 50-phosphateby formation of a 50,50-pyrophosphate linkage. (D) Template-directed ligation of RNA driven by release of a50,50-pyrophosphate-linked adenylate.



hope to get without magic (or evolved enzymes) on theprimitive Earth? Could one hope to achieve replication ina pool containing a more realistic mixture of organic mol-ecules, including, of course, ß-D-ribonucleotides? The syn-thesis of a nucleotide could occur in a number of ways. Thesimplest, conceptually, would be to synthesize a nucleosidebase, couple it to ribose, and finally to phosphorylate theresulting nucleoside. However, a number of other routesare feasible, for example the assembly of the base on a pre-formed ribose or ribose phosphate, or the coassembly ofthe base and sugar-phosphate.

The classical prebiotic synthesis of sugars is by the pol-ymerization of formaldehyde (the “formose” reaction). Ityields a very complex mixture of products including onlya small proportion of ribose (Mizuno and Weiss 1974).This reaction does not provide a reasonable route to theribonucleotides. However, a number of more recent exper-imental findings, to some extent, address this deficiency.

The base-catalyzed aldomerization of glycoaldehydephosphate in the presence of a half-equivalent of formalde-hyde under strongly alkaline conditions gives a relativelysimple mixture of tetrose- and pentose-diphosphates andhexose-triphosphates, of which ribose 2,4-diphosphate isthe major component (Muller et al. 1990). Reactions ofthis kind proceed efficiently when 2 mM solutions of sub-strates are incubated at room temperature and pH 9.5 in thepresence of layered hydroxides such as hydrotalcite (mag-nesium aluminum hydroxide) (Pitsch 1992; Pitsch et al.1995a). The phosphates are absorbed between the posi-tively charged layers of the mineral. The reaction proceedsunder these milder conditions presumably because of thehigh concentration of substrates in the interlayer andbecause the positive charge on the metal hydroxide layersfavors enolization of glycoaldehyde phosphate. A reactionbetween glycoaldehyde or glyceraldehyde and the amido-triphosphate ion provides an ingenious and prebioticallyplausible route to glycoaldehyde phosphate and glyceralde-hyde-2-phosphate, respectively, the two substrates in theabove reactions (Krishnamurthy et al. 2000).

A number of other studies have addressed the prob-lems presented by the lack of specificity of the formosereaction and by the instability of ribose. The Pb2+ ionis an excellent catalyst for the formose reaction and en-ables yields of the pentose sugars as high as 30% to beachieved (Zubay 1998). Furthermore, it seems likelythat ribose is almost exclusively the first pentose productof the reaction and that the other pentoses are formedfrom it by isomerization. Other recent studies have ad-dressed the problem presented by the instability of ribose.The four pentose sugars, including ribose, are all stronglystabilized in the presence of borate ions or calcium borateminerals (Ricardo et al. 2004). However, the effect of

borate on the progress of the formose reaction has notbeen reported.

Many sugars, including the four pentoses, react readilywith cyanamide to form stable bicyclic amino-oxazolines(Sanchez and Orgel 1970) (Fig. 6A). Strikingly, the ribosederivative crystallizes readily from aqueous solution evenwhen complex mixtures of related molecules, including amixture of the amino-oxazoline derivatives of the otherthree pentose sugars, are present (Springsteen and Joyce2004). The crystals are multiply twinned, each crystal con-taining many small domains of each of the two enantio-morphs. Thus the reaction of a mixture of racemic sugarswith cyanamide followed by crystallization might stabilizeribose, segregate it from other sugars, and present it inenantiospecific microdomains. Much remains to be shown,but the reactions described above suggest that ribose syn-thesis, although still problematic, may not be the intract-able problem it once seemed.

The synthesis of the nucleoside bases is one of thesuccess stories of prebiotic chemistry. Adenine is for-med with remarkable ease from ammonia and hydrogencyanide (Oro 1961). This synthesis has been describedas “the rock of the faith” by Stanley Miller. Reasonablyplausible syntheses of the other purine bases and of thepyrimidines have also been described (Sanchez et al.1967; Ferris et al. 1968; Robertson and Miller 1995; Orgel2004b; Saladino et al. 2004). The coupling of the purinebases with ribose or ribose-phosphate has been achievedunder mild conditions, but in relatively low yield (Fulleret al. 1972). The corresponding reaction with pyrimidinesdoes not occur.

There is a different potential route for the prebiotic syn-thesis of pyrimidine nucleotides via arabinose amino-oxazoline that first was explored nearly 40 yr ago (Tapieroand Nagyvary 1971) and in recent years has begun to lookvery persuasive (Ingar et al. 2003; Anastasi et al. 2007;Powner et al. 2009). The earlier studies began with arabi-nose 3-phosphate, which, like arabinose and other sugars,reacts with cyanamide to give the corresponding amino-oxazoline (Fig. 6B). This in turn reacts with cyanoacetyleneto form a tricyclic intermediate that hydrolyzes to producea mixture of cytosine arabinoside-30-phosphate and cyto-sine 20,30-cyclic phosphate.

Sutherland and colleagues (Powner et al. 2009) havetaken this approach further by starting simply with glyco-aldehyde and cyanamide, which in the presence of 1 Mphosphate at neutral pH gives 2-amino-oxazole in excellentyield (Fig. 6C). The phosphate both buffers and catalyzesthe reaction, directing glycoaldehyde toward 2-amino-oxazole, rather than a complex mixture of aldomerizationproducts. Glyceraldehyde is then added, resulting in forma-tion of the various pentose amino-oxazolines, including



the arabinose compound. Arabinose amino-oxazoline, inturn, can react with cyanoacetylene, also in phosphatebuffer, to form cytosine 20,30-cyclic phosphate as the majorproduct. Perhaps equally intriguing, although given lessemphasis in these studies, is that reaction of arabinoseamino-oxazoline with cyanoacetylene also gives a substan-tial yield of cytosine 20,30-cyclic-50-bisphosphate, which ismore amenable to being converted to an activated mono-mer that would be suitable for polymerization.

There has been significant progress, especially recently,concerning the synthesis of the nucleosides and nucleotidesfrom prebiotic precursors in reasonable yield. However, thestory remains incomplete because these syntheses still re-quire temporally separated reactions using high concentra-tions of just the right reactants, and would be disruptedby the presence of other closely related compounds. Thereactions channel material toward the desired products,but other fractionation processes must be discovered that

A

B

O

H

OH

OHOHO

O NO

O

P

O–

O–O–

O–

O–

O– O–

O– O–

O–

O

OHHO

HO O

O

P O–

O–O

OHO

O OOH

HH

OH

PO

OHO NO

N

O

PO

O N

NO

OP

O

O

H

O

OO N

NOOH

PO

C CN CH+

+

CN

HO

HO

HO

HO

HO

HO

HO

HO

HOHO

C

C NH2

NH2

NH2

NH2

NH2

NH2

NH2

NH2

NH2

NH

N

O NO

OH

OHO

NOH

HO OH

O

O

NH2

O NCN

H2PO4–(pH 7.0) H2PO4

–(pH 7.0)

H2PO4–(pH 6.5)H2PO4

–(pH 6.5)

NH2

O NO

OH

NH2

O NO

NNH

OH

O N

NO

OP

O

O

C CN CH

O

NOH O

HO

HOHO

O

H2O

–H2O

Figure 6. Potential prebiotic synthesis of pyrimidine nucleosides. (A) Reaction of ribose with cyanamide to form abicyclic product, with cyanamide joined at both the anomeric carbon and 2-hydroxyl. (B) Analogous reaction ofarabinose-3-phosphate to form a bicyclic product, which then reacts with cyanoacetylene to form a tricyclic inter-mediate that hydrolyzes to give a mixture of cytosine arabinoside-30-phosphate and cytosine 20,30-cyclic phosphate.(C) Reaction of glycoaldehyde with cyanamide in neutral phosphate buffer, followed by addition of glyceraldehyde,to form ribose and arabinose amino-oxazoline (and lesser amounts of the xylose and lyxose compounds). Arabinoseamino-oxazoline then reacts with cyanoacetylene to give cytosine 20,30-cyclic phosphate as the major product.



provide the correct starting materials at the requisite timeand place. This “preprebiotic” chemistry likely would in-volve a series of reactions catalyzed by minerals or metalions, coupled with a series of subtle fractionations ofnucleotide-like materials based on adsorption on minerals,selective complex formation, crystallization, etc.

Even minerals could not achieve on a macroscopic scaleone desirable separation, the resolution of D-ribonucleo-tides from their L-enantiomers. This is a serious problembecause experiments on template-directed synthesis usingpoly(C) and the imidazolides of G suggest that the poly-merization of the D-enantiomer is strongly inhibited bythe L-enantiomer (Joyce et al. 1984). This difficulty maynot be insuperable; perhaps with a different mode of phos-phate activation, the inhibition would be less severe. How-ever, enantiomeric cross-inhibition is certainly a seriousproblem if life arose in a racemic environment.

It is possible that the locale for life’s origins was notracemic, even though the global chemical environmentcontained nearly equal amounts of each pair of stereoiso-mers. There likely were biases in the inventory of com-pounds delivered to the Earth by comets and meteorites.For example, some carbonaceous chondrite meteoritescontain a significant enantiomeric excess of L-amino acidsthat are known to be indigenous to the meteorite (Engeland Macko 1997; Cronin and Pizzarello 1997; Pizzarelloet al. 2003; Glavin and Dworkin 2009). These in turn couldbias terrestrial syntheses, although the level of enantiomer-ic enrichment generally declines with successive chemicalreactions. A special exception are a remarkable set of reac-tions and fractionation processes that amplify a slightchiral imbalance, even to the level of local homochirality(Kondepudi et al. 1990; Soai et al. 1995; Viedma 2005;Klussmann et al. 2006; Noorduin et al. 2008; Viedmaet al. 2008). These systems have in common both a catalyticprocess for amplification of same-handed molecules andan inhibition process for suppression of opposite-handedmolecules.

Some of the most appealing examples of chiral symme-try-breaking reactions involve saturating solutions of vari-ous amino acids that form an equilibrium between theliquid phase and solid phase. The solid phase consists of ei-ther racemic or enantiopure crystals, and the liquid phasereflects whatever enantiomeric excess exists at the eutecticpoint for the mixture. For some amino acids, such as serineand histidine, the enantiomeric excess at the eutectic is.90% (Klussmann et al. 2006). This means that, startingfrom a small concentration imbalance of D- and L-isomers,the imbalance is amplified as both isomers enter the solidphase and the solution phase approaches the eutectic equi-librium. This and related near-equilibrium mechanisms(Noorduin et al. 2008; Viedma et al. 2008) could provide

a means to achieve high enantiomeric enrichment in a localenvironment. This in turn could bias the production ofribose and the derived nucleotides.

Scientists interested in the origins of life seem to divideneatly into two classes. The first, usually but not alwaysmolecular biologists, believe that RNA must have beenthe first replicating molecule and that chemists are exagger-ating the difficulties of nucleotide synthesis. They believethat a few more striking chemical “surprises” will establishthat a pool of racemic mononucleotides could have formedon the primitive Earth, and that further experiments withdifferent activating groups, minerals, and chiral amplifica-tion processes will solve the enantiomeric cross-inhibitionproblem. The second group of scientists are much morepessimistic. They believe that the de novo appearance ofoligonucleotides on the abiotic Earth would have been anear miracle. Time will tell which is correct.

3.2 Alternative Genetic Systems

The problems that arise when one tries to understand howan RNAWorld could have arisen de novo on the primitiveEarth are sufficiently severe that one must explore otherpossibilities. What kind of alternative genetic systemsmight have preceded the RNA World? How could theyhave “invented” the RNAWorld? These topics have gener-ated a good deal of speculative interest and some relevantexperimental data.

Eschenmoser and colleagues have undertaken a system-atic study of the properties of analogs of nucleic acids inwhich ribose is replaced by some other sugar, or in whichthe furanose form of ribose is replaced by the pyranoseform (Eschenmoser 1999) (Fig. 7B). Strikingly, polynu-cleotides based on the pyranosyl analog of ribose (p-RNA)form Watson-Crick paired double helices that are morestable than RNA, and p-RNAs are less likely than the corre-sponding RNAs to form multiple-strand competing struc-tures (Pitsch et al. 1993, 1995b, 2003). Furthermore, thehelices twist much more gradually than those of standardnucleic acids, which should make it easier to separatestrands of p-RNA during replication. Pyranosyl RNA ap-pears to be an excellent choice as a genetic system; insome ways it seems an improvement compared with thestandard nucleic acids. However, p-RNA does not interactwith normal RNA to form base-paired double helices.

Most double-helical structures reported in the litera-ture are characterized by a backbone with a six-atom repeat.Eschenmoser and colleagues made the surprising discoverythat an RNA-like structure based on threose nucleotideanalogs (TNA) (Fig. 7C), although it involves a five-atomrepeat, can still form a stable double-helical structurewith standard RNA (Schoning et al. 2000). This provides



an example of a pairing system based on a sugar that couldbe formed more readily than ribose: Tetroses are the uniqueproducts of the dimerization of glycoaldehyde, whereaspentoses are formed along with tetroses and hexosesfrom glycoaldehyde and glyceraldehyde. A structural sim-plification of Eschenmoser’s threose nucleic acid has beenachieved by Meggers and colleagues (Zhang et al. 2005).They replaced threose by its open chain analogue, glycol,in the backbone of TNA, resulting in glycol nucleic acid(GNA) (Fig. 7D). Complementary oligomers of GNAform antiparallel, double-helices with surprisingly highduplex stabilities.

Peptide nucleic acid (PNA) is another nucleic acid ana-log that has been studied extensively (Fig. 7E). It was disco-vered by Nielsen and colleagues in the context of researchon antisense oligonucleotides (Egholm et al. 1992, 1993;Wittung et al. 1994). PNA is an uncharged, achiral analogof RNA or DNA in which the ribose-phosphate back-bone of the nucleic acid is replaced by a backbone heldtogether by amide bonds. PNA forms very stable doublehelices with complementary RNA or DNA. Work in theOrgel laboratory has shown that information can betransferred from PNA to RNA, or from RNA to PNA, in

template-directed reactions, and that PNA/DNA chimerasare readily formed on either DNA or PNA templates(Schmidt et al. 1997a,b; Koppitz et al. 1998). Thus it seemsthat a transition from a PNA World to an RNA World ispossible.

The alanyl nucleic acids (ANA) are interesting for a dif-ferent reason. They are polypeptides formed from nucleoamino acids (Fig. 7F), but pairing structures can be formedonly if the two enantiomers of their constituent a-aminoacids occur in a regular alternating sequence (Diederichsen1996, 1997). Because abiotic syntheses of potentially chiralmolecules would under almost all circumstances yield ra-cemic products, pairing structures that can be formedfrom racemic mixtures are particularly attractive. TheANA-type backbone of alternating D- and L-amino acidscould, in principle, support paired, double-strandedstructures based on a variety of side-chain interactions.

Eschenmoser and colleagues have examined repeatinghomochiral dipeptide backbones that have either triazinesor aminopyrimidines attached at alternating positions(Mittapalli et al. 2007a,b) (Fig. 7G). In this case, not onlyis the backbone a potential precursor to that of RNA, butalso the bases have been replaced by potential precursors,

O

OO

P

O

O

Base

O

HO

HO

HO

OO

P

OBase

OO

P

O

BaseO

A B D

OPOO

Base

OPOO

Base

OPOO

Base

O

C

OOPOO

Base

OOPOO

Base

OOPOO

Base

O

O

O

O

BaseO

O

PO

O

P OO

Base

O

OH

OH

OH

OPO

O–

O–

–O

–O O–

O– O–

O–

O–O–

–O

O–

O Base

E

N

Base

O

O

N

Base

O

O

N

Base

O

O

FNHNH

NH

NH

NH

NH

Base

Base

O

O

NH

NH

OBase

Base

GNH

NH2

NH2

NH2

O

NH

N

NN

O

NH

O

NH

NH2

N

NN

HNH

O

NH

O–O2C–O2C

NH

O

NH

O

O

O

NH

HN

NH

O

O

O

NH

HN

NH OS

BaseO

S

Figure 7. The structures of (A) RNA; (B) p-RNA; (C) TNA; (D) GNA; (E) PNA; (F) ANA; (G) diaminotriazine-tagged (left) and dioxo-5-aminopyrimidine-tagged (right) oligodipeptides; and (H ) tPNA. ANA contains a back-bone of alternating D- and L-alanine subunits. The diaminotriazine tags are shown linked to a backbone of alternat-ing L-aspartate and L-glutamate subunits; the dioxo-5-aminopyrimidine tags (shown unattached) can be linkedsimilarly. tPNA is shown with a backbone of alternating L-cysteine and L-glutamate subunits.



such as 2,4-diaminotriazine (TNN), 2,4-dioxotriazine(TOO), 2,4-diamino-5-aminopyrimidine (APNN), and2,4-dioxo-5-aminopyrimidine (APOO). Oligomers con-taining either TNN or APOO subunits were found to pairstrongly with complementary RNA, whereas oligomerscontaining either TOO or APNN subunits did not. Notsurprisingly, therefore, pairing between complementarysubstituted oligodipeptides of the same type (eitheroligo[TNN]†oligo[TOO] or oligo[APNN]†oligo[APOO])also was weak. However, cross-pairing between oligo(TNN)and oligo(APOO) was robust (Mittapalli et al. 2007a). Thisraises the intriguing possibility that an informational poly-mer could have a mixed composition of TNN and APOOsubunits, which would direct the synthesis of an opposingstrand that has a complementary sequence as well as a“complementary” backbone composition.

Even more radical are thioester peptide nucleic acids(tPNA) (Fig. 7H), containing a repeating dipeptide back-bone with cysteine residues at alternating positions, whichare transiently linked via a thioester to a nucleic acid base(Ura et al. 2009). The bases are in dynamic equilibriumbetween the solution and cysteine positions along the back-bone. Occupancy of a base at a particular position is en-hanced by the presence of the complementary base onthe opposing strand. In this way the informational polymercan self-assemble in a template-directed manner, with mis-matched bases exchanging rapidly and matched basesremaining thioesterified for an extended period of time.Perhaps genetic information could be propagated in sucha system, although the fidelity of replication, and thereforethe maximum number of informational subunits, is likelyto be modest.

The studies described previously suggest that there aremany ways of linking together the nucleotide bases intochains that are capable of forming base-paired double heli-ces. It is not clear that it is much easier to synthesize themonomers of p-RNA, TNA, GNA, PNA, ANA, or tPNAthan to synthesize the standard nucleotides. However, itis possible that a base-paired structure of this kind will bediscovered that can be synthesized readily under prebioticconditions. The properties of the TNN- and APOO-taggedoligodipeptides suggest that it may be fruitful to explore abroader range of potential precursors to RNA, changing therecognition elements as well as the backbone. A strongcandidate for the first genetic material would be any infor-mational macromolecule that is replicable in a sequence-general manner and derives from compounds that wouldhave been abundant on the primitive Earth, and preferablyhas the ability to cross-pair with RNA.

The transition from an RNA-like World to the RNAWorld could take place in two ways. The transition mightbe continuous if the pre-RNA template could direct the

synthesis of an RNA product with a complementary se-quence. Such a transition, for example, from PNA to RNA,would preserve information. RNAcould then act as a geneticmaterial in aformerly PNAWorld.However, even if chimeraswere involved in the transition, it is unlikely that the originalfunction of a PNA catalyst could be retained throughoutthe transition because PNA and RNA have such differentbackbone structures. A direct and continuous transitionfrom p-RNA to RNA would not be possible becausep-RNA does not form complementary double helices withRNA, but this limitation does not apply to TNA and GNA.

The second type of transition can be described as agenetic takeover. A pre-existing self-replicating systemevolves, for its own selective advantage, a mechanism forsynthesizing and polymerizing the components of a com-pletely different genetic system, and is taken over by it.Cairns-Smith (1982) has proposed that the first genetic sys-tem was inorganic, perhaps a clay, and that it “invented” aself-replicating system based on organic monomers. How-ever, he clearly recognized the possibility of one organicgenetic material replacing another (Cairns-Smith andDavies 1977). Genetic takeover does not require any struc-tural relationship between the polymers of the two geneticsystems. It suggests the possibility that the original geneticsystem may have been unrelated to nucleic acids.

The hypothesis of a genetic material completely differ-ent from nucleic acids has one enormous advantage—itopens up the possibility of using very simple, easily synthe-sized prebiotic monomers in place of nucleotides. How-ever, it also raises two new and difficult questions. Whichprebiotic monomers are plausible candidates as the com-ponents of a replicating system? Why would an initialgenetic system invent nucleic acids once it had evolvedsufficient synthetic know-how to generate molecules ascomplex as nucleotides?

A number of prebiotic monomers that might have madeup a simple genetic material have already been suggested.They include hydroxy acids (Weber 1987), amino acids(Orgel 1968; Zhang et al. 1994), phosphomonoesters ofpolyhydric alcohols (Weber 1989), aminoaldehydes (Nel-sestuen 1980), and molecules containing two sulfhydrylgroups (Schwartz and Orgel 1985). The list could be ex-panded almost indefinitely. The discussion here concernsa small class of these monomers that appear to be particu-larly attractive in the light of recent work on enzymemechanisms.

There is accumulating evidence that several enzymesthat make or break phosphodiester bonds have two or threemetal ions at their active sites (Cooperman et al. 1992;Strater et al. 1996). In the case of the editing site for phos-phodiester hydrolysis in the Klenow fragment of Escheri-chia coli DNA polymerase I, no other functional groups



of the enzyme come close to the phosphodiester bond thatis cleaved. This has led to the suggestion that the major roleof the enzyme is to act as scaffolding on which to hangmetal ions in precisely determined positions (Beese andSteitz 1991; Steitz 1998). A similar suggestion has beenmade for ribozymes on the basis of both indirect and directevidence (Freemont et al. 1988; Yarus 1993; Steitz and Steitz1993; Shan et al. 1999, 2001; Stahley and Strobel 2005).

Perhaps these observations can be extended to suggestthat, if informational polymers preceded RNA, they mayalso have been dependent on metal ions for their catalyticactivity. If so, the range of prebiotic monomers that needsto be considered is greatly reduced. In addition to the func-tional groups that react to form the backbone, the mono-mers must have carried metal-binding functional groups.If the metal ions involved were divalent ions such asMg2+ and Ca2+, the side groups are likely to have been car-boxylate or phosphate groups. If transition metal ions wereinvolved, sulfhydryl groups and possibly imidazole deriva-tives are likely to have been important.

Prebiotic monomers suitable for building polymersthat bind Mg2+ or Ca2+ include aspartic acid, glutamicacid, and serine phosphate among biologically importantamino acids. ß-amino acids, such as isoglutamic acid, hy-droxydicarboxylic acids, such as a-hydroxysuccinic acid,and hydroxytricarboxylic acids, such as citric acid, areother possible candidates. A polymer containing D-asparticacid, L-aspartic acid, and glycine as its subunits is typicalof potentially informational co-polymers that might, inthe presence of divalent metal ions, both replicate andfunction as a catalyst. Transition-metal ions might play acorresponding role for polymers containing cysteine orhomocysteine. The present challenge is to show replication,or at least information transfer in template-directed syn-thesis, in some such system.

What selective advantage could a simpler, metabolicallycompetent system derive from the synthesis of oligonucleo-tides? This is a baffling question. Most arguments thatcome to mind do not stand up to detailed analysis. If, forexample, one postulates that nucleotides were first synthe-sized as parts of cofactors such as DPN, one must explainwhy the particular heterocyclic bases and sugars werechosen. Even if one supposes that among the many “experi-ments” in secondary metabolism performed by earlyorganisms one happened by accident on a pair of comple-mentary nucleotides that could form a replicating polymer,one must still explain how polymerization subsequentlycontributed to the success of the “inventor.” Could oligo-nucleotides, by hybridization, have functioned at first as se-lective “glues” for tying pairs of macromolecules together?Could RNA have been invented by one organism as “anti-sense” against the genome of another?

The discussion so far, even though highly speculative,is still conservative in overall outlook. It supposes thatthe original information-accumulating system that led tothe evolution of life on Earth was either RNA or some linearcopolymer that replicated in an aqueous environment inmuch the same way as RNA. There remains a lingeringdoubt that the discussion is not on the right track at all;maybe the original system was not an organic copolymer(Cairns-Smith 1982), or maybe it replicated in a nonaqueousenvironment and RNA is an adaptation that permittedinvasion of the oceans. Perhaps systems of high complexitycan develop without any need for a genome in the usualsense (Dyson 1982; Kauffman 1986; Wachtershauser 1988;De Duve 1991; Eschenmoser 2007). Perhaps . . .

Laboratory simulations of prebiotic chemistry are de-pendent on organic chemistry and can only explore thekinds of reactions understood by organic chemists. Agood deal is known about reactions in aqueous solution,but less about reactions at the interface between waterand inorganic solids. Very little is known about reactionsin systems in which inorganic solids are depositing fromaqueous solutions containing organic material. It is hardto see how speculative schemes involving heterogeneousaqueous systems can be tested until much more is knownabout the underlying branches of chemistry.

4 CONCLUDING REMARKS

After contemplating the possibility of self-replicating ribo-zymes emerging from pools of random polynucleotidesand recognizing the difficulties that must have been over-come for RNA replication to occur in a realistic prebioticsoup, the challenge must now be faced of constructing arealistic picture of the origin of the RNAWorld. The con-straints that must have been met in order to originate a self-sustained evolving system are reasonably well understood.One can sketch out a logical order of events, beginningwith prebiotic chemistry and ending with DNA/protein-based life. However, it must be said that the details of thisprocess remain obscure and are not likely to be known inthe near future.

The presumed RNAWorld should be viewed as a mile-stone, a plateau in the early history of life on Earth. So too,the concept of an RNAWorld has been a milestone in thescientific study of life’s origins. While this concept doesnot explain how life originated, it has helped to guidescientific thinking and has served to focus experimentalefforts. Further progress will depend primarily on new ex-perimental results, as chemists, biochemists, and molecularbiologists work together to address problems concerningmolecular replication, ribozyme enzymology, and RNA-based cellular processes.



ACKNOWLEDGMENTS

This work was supported by research grant NNX07AJ23Gfrom the National Aeronautics and Space Administration.Previous versions of this article, which were published inthe First (1993), Second (1999), and Third (2006) Editionsof The RNA World, were coauthored by Leslie Orgel, whodied on October 27, 2007. Many portions of the text havenot been changed in the current edition because they re-main an accurate reflection of current scientific under-standing. The contributions of Leslie Orgel to this workand to the scientific literature of the origins of life are grate-fully acknowledged.

REFERENCES

Acevedo OL, Orgel LE. 1987. Non-enzymatic transcription of an oligo-deoxynucleotide 14 residues long. J Mol Biol 197: 187–193.

Anastasi C, Crowe MA, Sutherland JD. 2007. Two-step potentially prebi-otic synthesis of a-D-cytidine-50-phosphate from D-glyceraldehyde-3-phosphate. J Am Chem Soc 129: 24–25.

Ban N, Nissen P, Hansen J, Moore PB, Steitz TA. 2000. The completeatomic structure of the large ribosomal subunit at 2.4 A resolution.Science 289: 905–920.

Bartel DP, Szostak JW. 1993. Isolation of new ribozymes from a large poolof random sequences. Science 261: 1411–1418.

Been MD, Cech TR. 1988. RNA as an RNA polymerase: Net elongation ofan RNA primer catalyzed by the Tetrahymena ribozyme. Science 239:1412–1416.

Beese LS, Steitz TA. 1991. Structural basis for the 30-50 exonuclease activ-ity of Escherichia coli DNA polymerase I: A two metal ion mechanism.EMBO J 10: 25–33.

Cairns-Smith AG. 1982. Genetic takeover and the mineral origins of life.Cambridge University Press, Cambridge.

Cairns-Smith AG, Davies CJ. 1977. The design of novel replicating poly-mers. In Encyclopaedia of ignorance (eds. Duncan R., Weston-SmithM.), pp. 391–403. Pergamon Press, Oxford.

Cech TR. 1986. A model for the RNA-catalyzed replication of RNA. ProcNatl Acad Sci 83: 4360–4363.

Chapple KE, Bartel DP, Unrau PJ. 2003. Combinatorial minimizationand secondary structure determination of a nucleotide synthase ribo-zyme. RNA 9: 1208–1220.

Cooperman BS, Baykov AA, Lahti R. 1992. Evolutionary conservation ofthe active site of soluble inorganic pyrophosphatase. Trends BiochemSci 17: 262–266.

Crick FHC. 1968. The origin of the genetic code. J Mol Biol 38: 367–379.Cronin JR, Pizzarello S. 1997. Enantiomeric excesses in meteoritic amino

acids. Science 275: 951–955.De Duve C. 1991. Blueprint for a cell: The nature and origin of life. Neil

Patterson Publishers, Burlington, North Carolina.Diederichsen U. 1996. Pairing properties of alanyl peptide nucleic acids

containing an amino acid backbone with alternating configuration.Angew Chemie 35: 445–448.

Diederichsen U. 1997. Alanyl PNA: Evidence for linear band structuresbased on guanine–cytosine base pairs. Angew Chemie 36: 1886–1889.

Ding ZP, Kawamura K, Ferris JP. 1996. Oligomerization of uridine phos-phorimidazolides on montmorillonite: A model for the prebioticsynthesis of RNA on minerals. Orig Life Evol Biosph 26: 151–171.

Doudna JA, Szostak JW. 1989. RNA-catalysed synthesis of complementa-ry-strand RNA. Nature 339: 519–522.

Dyson FJ. 1982. A model for the origin of life. J Mol Evol 18: 344–350.Egholm M, Buchardt O, Christensen L, Behrens C, Freier SM, Driver DA,

Berg RH, Kim SK, Norden B, Nielson PE. 1993. PNA hybridizes

to complementary oligonucleotides obeying the Watson-Crickhydrogen-bonding rules. Nature 365: 566–568.

Egholm M, Buchardt O, Nielson PE, Berg RH. 1992. Peptide nucleic acids(PNA). Oligonucleotide analogues with an achiral peptide backbone.J Am Chem Soc 114: 1895–1897.

Eigen M. 1971. Selforganization of matter and the evolution of biologicalmacromolecules. Naturwiss 58: 465–523.

Eigen M, Schuster P. 1977. The hypercycle: A principle of natural self-organization. Part A: Emergence of the hypercycle. Naturwiss 64:541–565.

Eigen M, McCaskill J, Schuster P. 1988. Molecular quasi-species. J PhysChem 92: 6881–6891.

Ekland EH, Bartel DP. 1996. RNA-catalysed RNA polymerization usingnucleoside triphosphates. Nature 382: 373–376.

Ekland EH, Szostak JW, Bartel DP. 1995. Structurally complex and highlyactive RNA ligases derived from random RNA sequences. Science 269:364–370.

Emilsson GM, Nakamura S, Roth A, Breaker RR. 2003. Ribozyme speedlimits. RNA 9: 907–918.

Engel M, Macko S. 1997. Isotopic evidence for extraterrestrial non-racemic amino acids in the Murchison meteorite. Nature 389:265–268.

Erman JE, Hammes GG. 1966. Relaxation spectra of ribonuclease. V. Theinteraction of ribonuclease with cytidine 20:30-cyclic phosphate. J AmChem Soc 88: 5614–5617.

Eschenmoser A. 1999. Chemical etiology of nucleic acid structure.Science 284: 2118–2124.

Eschenmoser A. 2007. On a hypothetical generational relationship be-tween HCN and constituents of the reductive citric acid cycle. ChemBiodivers 4: 554–573.

Ferris JP. 2002. Montmorillonite catalysis of 30–50 mer oligonucleotides:Laboratory demonstration of potential steps in the origin of the RNAworld. Orig Life Evol Biosph 32: 311–332.

Ferris JP. 2006. Montmorillonite-catalyzed formation of RNA oligomers:The possible role of catalysis in the origins of life. Phil Trans R Soc B361: 1777–1786.

Ferris JP, Ertem G. 1993. Montmorillonite catalysis of RNA oligomerformation in aqueous solution. A model for the prebiotic formationof RNA. J Am Chem Soc 115: 12270–12275.

Ferris JP, Sanchez RA, Orgel LE. 1968. Studies in prebiotic synthesis. III.Synthesis of pyrimidines from cyanoacetylene and cyanate. J Mol Biol33: 693–704.

Ferris JP, Hill AR, Liu R, Orgel LE. 1996. Synthesis of long prebioticoligomers on mineral surfaces. Nature 381: 59–61.

Ferris JP, Joshi PC, Wang KJ, Miyakawa S, Huang W. 2004. Catalysis inprebiotic chemistry: Application to the synthesis of RNA oligomers.Adv Space Res 33: 100–105.

Freemont PS, Friedman JM, Beese LS, Sanderson MR, Steitz TA. 1988.Cocrystal structure of an editing complex of Klenow fragment withDNA. Proc Natl Acad Sci 85: 8924–8928.

Fujita Y, Furuta H, Ikawa Y. 2009. Tailoring RNA modular units on a com-mon scaffold: A modular ribozyme with a catalytic unit forb-nicotinamide mononucleotide-activated RNA ligation. RNA 15:877–888.

Fuller WD, Sanchez RA, Orgel LE. 1972. Studies in prebiotic synthesis.VI. Synthesis of purine nucleosides. J Mol Biol 67: 25–33.

Gilbert W. 1986. The RNA world. Nature 319: 618.Glavin DP, Dworkin JP. 2009. Enrichment of the amino acid L-isovaline

by aqueous alteration on CI and CM meteorite parent bodies. ProcNatl Acad Sci 106: 5487–5492.

Guerrier-Takada C, Gardiner K, Marsh T, Pace N, Altman S. 1983. TheRNA moiety of ribonuclease P is the catalytic subunit of the enzyme.Cell 35: 849–857.

Hager AJ, Szostak JW. 1997. Isolation of novel ribozymes that ligateAMP-activated RNA substrates. Chem Biol 4: 607–617.



Handschuh GJ, Lohrmann R, Orgel LE. 1973. The effect of Mg2+ andCa2+ on urea-catalyzed phosphorylation reactions. J Mol Evol 2:251–262.

Huang F, Yarus M. 1997. Versatile 50 phosphoryl coupling of small andlarge molecules to an RNA. Proc Natl Acad Sci 94: 8965–8969.

Huang W, Ferris JP. 2003. Synthesis of 35-40 mers of RNA oligomers fromunblocked monomers. A simple approach to the RNA world. ChemCommun 12: 1458–1459.

Huang W, Ferris JP. 2006. One-step, regioselective synthesis of up to50-mers of RNA oligomers by montmorillonite catalysis. J AmChem Soc 128: 8914–8919.

Ikawa Y, Tsuda K, Matsumura S, Inoue T. 2004. De novo synthesis and de-velopment of an RNA enzyme. Proc Natl Acad Sci 101: 13750–13755.

Ingar AA, Luke RWA, Hayter BR, Sutherland JD. 2003. Synthesis of acytidine ribonucleotide by stepwise assembly of the heterocycle on asugar phosphate. Chembiochem 4: 504–507.

Inoue T, Orgel LE. 1981. Substituent control of the poly(C)-directed oli-gomerization of guanosine 50-phosphoroimidazolide. J Am Chem Soc103: 7666–7667.

Inoue T, Orgel LE. 1982. Oligomerization of (guanosine 50-phosphor)-2-methylimidazolide on poly(C). J Mol Biol 162: 204–217.

Inoue T, Orgel LE. 1983. A nonenzymatic RNA polymerase model.Science 219: 859–862.

Inoue T, Joyce GF, Grzeskowiak K, Orgel LE, Brown JM, Reese CB. 1984.Template-directed synthesis on the pentanucleotide CpCpGpCpC.J Mol Biol 178: 669–676.

Jaeger L, Wright MC, Joyce GF. 1999. A complex ligase ribozyme evolvedin vitro from a group I ribozyme domain. Proc Natl Acad Sci 96:14712–14717.

James KD, Ellington AD. 1999. The fidelity of template-directed oligonu-cleotide ligation and the inevitability of polymerase function. Orig LifeEvol Biosph 29: 375–390.

Johnston WK, Unrau PJ, Lawrence MS, Glasner ME, Bartel DP. 2001.RNA-catalyzed RNA polymerization: Accurate and general RNA-templated primer extension. Science 292: 1319–1325.

Joshi PC, Aldersley MF, Delano JW, Ferris JP. 2009. Mechanism of mont-morillonite catalysis in the formation of RNA oligomers. J Am ChemSoc 131: 13369–13374.

Joyce GF. 1987. Non-enzymatic template-directed synthesis of informa-tional macromolecules. Cold Spring Harbor Symp Quant Biol 52:41–51.

Joyce GF. 1989. RNA evolution and the origins of life. Nature 338:217–224.

Joyce GF. 2002. The antiquity of RNA-based evolution. Nature 418:214–221.

Joyce GF, Orgel LE. 1986. Non-enzymic template-directed synthesis onRNA random copolymers: Poly(C,G) templates. J Mol Biol 188:433–441.

Joyce GF, Orgel LE. 1993. Prospects for understanding the origin of theRNA world. In The RNA world (eds. Gesteland R.F., Atkins J.F.), pp.1–25. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

Joyce GF, Schwartz AW, Miller SL, Orgel LE. 1987. The case for an ances-tral genetic system involving simple analogues of the nucleotides. ProcNatl Acad Sci 84: 4398–4402.

Joyce GF, Visser GM, van Boeckel CAA, van Boom JH, Orgel LE, vanWestrenen J. 1984. Chiral selection in poly(C)-directed synthesis ofoligo(G). Nature 310: 602–604.

Kanavarioti A, Monnard PA, Deamer DW. 2001. Eutectic phases in ice fa-cilitate nonenzymatic nucleic acid synthesis. Astrobiology 1: 271–281.

Kauffman SA. 1986. Autocatalytic sets of proteins. J Theor Biol 119: 1–24.Kawamura K, Ferris JP. 1994. Kinetic and mechanistic analysis of dinu-

cleotide and oligonucleotide formation from the 50-phosphorimida-zolide of adenosine on Na+-montmorillonite. J Am Chem Soc 116:7564–7572.

Kawamura K, Ferris JP. 1999. Clay catalysis of oligonucleotide formation:Kinetics of the reaction of the 50-phosphorimidazolides of nucleotides

with the non-basic heterocycles uracil and hypoxanthine. Orig LifeEvol Biosph 29: 563–591.

Khorana HG. 1961. Some recent developments in the chemistry ofphosphate esters of biological interest, pp. 126–141. Wiley & Sons,New York.

Klussmann M, Iwamura H, Mathew SP, Wells Jr DH, Pandya U,Armstrong A, Blackmond DG. 2006. Thermodynamic control ofasymmetric amplification in amino acid catalysis. Nature 441:621–623.

Kondepudi DK, Kaufman RJ, Singh N. 1990. Chiral symmetry breakingin sodium chlorate crystallization. Science 250: 975–976.

Koppitz M, Nielsen PE, Orgel LE. 1998. Formation of oligonucleotide-PNA-chimeras by template-directed ligation. J Am Chem Soc 120:4563–4569.

Kozlov IA, Politis PK, Van Aerschot A, Busson R, Herdewijn P, Orgel LE.1999. Nonenzymatic synthesis of RNA and DNA oligomers on hexitolnucleic acid templates: the importance of the A structure. J Am ChemSoc 121: 2653–2656.

Kozlov IA, Zielinski M, Allart B, Kerremans L, Van Aerschot A, Busson R,Herdewijn P, Orgel LE. 2000. Nonenzymatic template-directed reac-tions on altritol oligomers, preorganized analogues of oligonucleoti-des. Chem Eur J 6: 151–155.

Krishnamurthy R, Guntha S, Eschenmoser A. 2000. Regioselectivea-phosphorylation of aldoses in aqueous solution. Angew Chemie39: 2281–2285.

Kruger K, Grabowski PJ, Zaug AJ, Sands J, Gottschling DE, Cech TR.1982. Self-splicing RNA: Autoexcision and autocyclization of the ribo-somal RNA intervening sequence of Tetrahymena. Cell 31: 147–157.

Kurz M, Gobel K, Hartel C, Gobel MW. 1997. Nonenzymatic oligomeri-zation of ribonucleotides on guanosine-rich templates: Suppression ofthe self-pairing of guanosine. Angew Chemie 36: 842–845.

Kurz M, Gobel K, Hartel C, Gobel MW. 1998. Acridine-labeled primers astools for the study of nonenzymatic RNA oligomerization. Helv ChimActa 81: 1156–1180.

Lau MWL, Cadieux KEC, Unrau PJ. 2004. Isolation of fast purine nucleo-tide synthase ribozymes. J Am Chem Soc 126: 15686–15693.

Lau MWL, Unrau PJ. 2009. A promiscuous ribozyme promotes nucleo-tide synthesis in addition to ribose chemistry. Chem Biol 16: 815–825.

Lawrence MS, Bartel DP. 2003. Processivity of ribozyme-catalyzed RNApolymerization. Biochemistry 42: 8748–8755.

Lewin R. 1986. RNA catalysis gives fresh perspective on the origin of life.Science 231: 545–546.

Lincoln TA, Joyce GF. 2009. Self-sustained replication of an RNA enzyme.Science 323: 1229–1232.

Lohrmann R. 1975. Formation of nucleoside 50-polyphosphates fromnucleotides and trimetaphosphate. J Mol Evol 6: 237–252.

Lohrmann R. 1977. Formation of nucleoside 50-phosphoramidatesunder potentially prebiological conditions. J Mol Evol 10: 137–154.

Lohrmann R, Orgel LE. 1973. Prebiotic activation processes. Nature 244:418–420.

Lohrmann R, Orgel LE. 1976. Template-directed synthesis of high molec-ular weight polynucleotide analogues. Nature 261: 342–344.

Lorsch J, Szostak JW. 1994. In vitro evolution of new ribozymes withpolynucleotide kinase activity. Nature 371: 31–36.

McCaskill JS. 1984a. A stochastic theory of molecular evolution. BiolCybernetics 50: 63–73.

McCaskill JS. 1984b. A localization threshold for macromolecular quasis-pecies from continuously distributed replication rates. J Chem Phys80: 5194–5202.

McGinness KE, Joyce GF. 2002. RNA-catalyzed RNA ligation on an exter-nal RNA template. Chem Biol 9: 297–307.

Mittapalli GK, Osornio YM, Guerrero MA, Reddy KR, Krishnamurthy R,Eschenmoser A. 2007a. Mapping the landscape of potentially primor-dial informational oligomers: Oligodipeptides tagged with 2,4-disubstituted 5-aminopyrimidines as recognition elements. AngewChemie 46: 2478–2484.



Mittapalli GK, Reddy KR, Xiong H, Munoz O, Han B, De Riccardis F,Krishnamurthy R, Eschenmoser A. 2007b. Mapping the landscapeof potentially primordial informational oligomers: Oligodipeptidesand oligodipeptoids tagged with triazines as recognition elements.Angew Chemie 46: 2470–2477.

Miyakawa S, Ferris JP. 2003. Sequence- and regioselectivity in themontmorillonite-catalyzed synthesis of RNA. J Am Chem Soc 125:8202–8208.

Mizuno T, Weiss AH. 1974. Synthesis and utilization of formose sugars.Adv Carbohyd Chem Biochem 29: 173–227.

Mohr SC, Thach RE. 1969. Application of ribonuclease T1 to the synthe-sis of oligoribonucleotides of defined base sequence. J Biol Chem 244:6566–6576.

Monnard PA, Kanavarioti A, Deamer DW. 2003. Eutectic phase polymer-ization of activated ribonucleotide mixtures yields quasi-equimolarincorporation of purine and pyrimidine nucleobases. J Am ChemSoc 125: 13734–13740.

Muller D, Pitsch S, Kittaka A, Wagner E, Wintner CE, Eschenmoser A.1990. Chemie von a-aminonitrilen. Aldomerisierung von Glykolalde-hydphosphat zu racemischen hexose-2,4,6-triphosphaten und (in ge-genwart von formaldehyd) racemischen pentose-2,4-diphosphaten:rac.-allose-2,4,6-triphosphat und rac.-ribose-2,4-diphosphat sind diereaktionshauptprodukte. Helv Chim Acta 73: 1410–1468.

Narlikar GJ, Herschlag D. 1997. Mechanistic aspects of enzymatic catal-ysis: Lessons from comparison of RNA and protein enzymes. Annu RevBiochem 66: 19–59.

Nelsestuen GL. 1980. Origin of life: Consideration of alternatives to pro-teins and nucleic acids. J Mol Evol 15: 59–72.

Nissen P, Hansen J, Ban N, Moore PB, Steitz TA. 2000. The structural ba-sis of ribosome activity in peptide bond synthesis. Science 289:920–930.

Noorduin WL, Izumi T, Millemaggi A, Leeman M, Meekes H, Van Enck-evort WJP, Kellogg RM, Kaptein B, Vlieg E, Blackmond DG. 2008.Emergence of a single solid chiral state from a nearly racemic aminoacid derivative. J Am Chem Soc 130: 1158–1159.

Orgel LE. 1968. Evolution of the genetic apparatus. J Mol Biol 38:381–393.

Orgel LE. 1989. Was RNA the first genetic polymer? In Evolutionary tin-kering in gene expression (eds. Grunberg-Manago M., Clark B.F.C.,Zachau H.G.), pp. 215–224. Plenum, London.

Orgel LE. 2004a. Prebiotic chemistry and the origin of the RNA world.Crit Rev Biochem Mol Biol 39: 99–123.

Orgel LE. 2004b. Prebiotic adenine revisited: Eutectics and photochem-istry. Orig Life Evol Biosph 34: 361–369.

Oro J. 1961. Mechanism of synthesis of adenine from hydrogen cyanideunder plausible primitive earth conditions. Nature 191: 1193–1194.

Osterberg R, Orgel LE, Lohrmann R. 1973. Further studies of urea-catalyzed phosphorylation reactions. J Mol Evol 2: 231–234.

Pace NR, Marsh TL. 1985. RNA catalysis and the origin of life. Orig LifeEvol Biosph 16: 97–116.

Pitsch S. 1992. “Zur chemie von glykolaldehyd-phosphat: Seine bildungaus oxirancarbonitril und seine aldomerisierumg zu den (racemi-schen) pentose-2,4-diphosphaten und hexose-2,4,6-triphosphaten.”Ph.D. Thesis, ETH, Zurich.

Pitsch S, Eschenmoser A, Gedulin B, Hui S, Arrhenius G. 1995a. Mineralinduced formation of sugar phosphates. Orig Life Evol Biosph 25:297–334.

Pitsch S, Krishnamurthy R, Bolli M, Wendeborn S, Holzner A, MintonM, Lesueur C, Schlonvogt I, Jaun B, Eschenmoser A. 1995b.Pyranosyl-RNA (‘p-RNA’): Base-pairing selectivity and potential toreplicate. Helv Chim Acta 78: 1621–1635.

Pitsch S, Wendeborn S, Jaun B, Eschenmoser A. 1993. Why pentose- andnot hexose-nucleic acids? Pyranosyl-RNA (‘p-RNA’). Helv Chim Acta76: 2161–2183.

Pitsch S, Wendeborn S, Krishnamurthy R, Holzner A, Minton M, BolliM, Miculka C, Windhab N, Micura R, Stanek M, et al. 2003. Pentopyr-anosyl oligonucleotide systems. 9th communication. The ß-D-

ribopyranosyl-(40!20)-oligonucleotide system (‘pyranosyl-RNA’):Synthesis and resume of base-pairing properties. Helv Chim Acta 86:4270–4363.

Pizzarello S, Zolensky M, Turk KA. 2003. Nonracemic isovaline in theMurchison meteorite: Chiral distribution and mineral association.Geochim Cosmochim Acta 67: 1589–1595.

Prabahar KJ, Ferris JP. 1997. Adenine derivatives as phosphate-activatinggroups for the regioselective formation of 30,50-linked oligoadenylateson montmorillonite: Possible phosphate-activating groups for theprebiotic synthesis of RNA. J Am Chem Soc 119: 4330–4337.

Powner MW, Gerland B, Sutherland JD. 2009. Synthesis of activatedpyrimidine ribonucleotides in prebiotically plausible conditions.Nature 459: 239–242.

Reimann R, Zubay G. 1999. Nucleoside phosphorylation: A feasible stepin the prebiotic pathway to RNA. Orig Life Evol Biosph 29: 229–247.

Ricardo A, Carrigan MA, Olcott AN, Benner SA. 2004. Borate mineralsstabilize ribose. Science 303: 196.

Robertson MP, Ellington AD. 1999. In vitro selection of an allosteric ribo-zyme that transduces analytes into amplicons. Nature Biotechnol 17:62–66.

Robertson MP, Miller SL. 1995. An efficient prebiotic synthesis ofcytosine and uracil. Nature 375: 772–774.

Robertson MP, Scott WG. 2007. The structural basis of ribozyme-catalyzed RNA assembly. Science 315: 1549–1553.

Rogers J, Joyce GF. 2001. The effect of cytidine on the structure andfunction of an RNA ligase ribozyme. RNA 7: 395–404.

Rohatgi R, Bartel DP, Szostak JW. 1996a. Kinetic and mechanistic analysisof nonenzymatic, template-directed oligoribonucleotide ligation. JAm Chem Soc 118: 3332–3339.

Rohatgi R, Bartel DP, Szostak JW. 1996b. Nonenzymatic, template-directed ligation of oligoribonucleotides is highly regioselective forthe formation of 30 –50 phosphodiester bonds. J Am Chem Soc 118:3340–3344.

Saladino R, Crestini C, Costanzo G, DiMauro E. 2004. Advances in theprebiotic synthesis of nucleic acids bases: Implications for the originof life. Curr Org Chem 8: 1425–1443.

Sanchez RA, Orgel LE. 1970. Studies in prebiotic synthesis. V. Synthesisand photoanomerization of pyrimidine nucleosides. J Mol Biol 47:531–543.

Sanchez RA, Ferris JP, Orgel LE. 1967. Studies in prebiotic synthesis. II.Synthesis of purine precursors and amino acids from cyanoacetyleneand cyanate. J Mol Biol 30: 223–253.

Sawai H, Kuroda K, Hojo T. 1988. Efficient oligoadenylate synthesiscatalyzed by uranyl ion complex in aqueous solution. In Nucleic acidsresearch symposium series (ed. Hayatsu H.), vol. 19, pp. 5–7. IRL PressLimited, Oxford.

Shechner DM, Grant RA, Bagby SC, Koldobskaya Y, Piccirilli JA, BartelDP. 2009. Crystal structure of the catalytic core of an RNA polymeraseribozyme. Science 326: 1271–1275.

Schmidt JG, Christensen L, Nielsen PE, Orgel LE. 1997a. Informationtransfer from DNA to peptide nucleic acids by template-directedsyntheses. Nucleic Acids Res 25: 4792–4796.

Schmidt JG, Nielsen PE, Orgel LE. 1997b. Information transfer from pep-tide nucleic acids to RNA by template-directed syntheses. Nucleic AcidsRes 25: 4797–4802.

Schoning K, Scholz P, Guntha S, Wu X, Krishnamurthy R, EschenmoserA. 2000. Chemical etiology of nucleic acid structure: The a-threofur-anosyl-(30!20) oligonucleotide system. Science 290: 1347–1351.

Schrum JP, Ricardo A, Krishnamurthy M, Blain JC, Szostak JW. 2009.Efficient and rapid template-directed nucleic acid copying using20-amino-20,30-dideoxyribonucleoside-50-phosphorimidazolidemonomers. J Am Chem Soc 131: 14560–14570.

Schuster P, Swetina J. 1988. Stationary mutant distributions and evolu-tionary optimization. Bull Math Biol 50: 635–660.

Schwartz AW, Orgel LE. 1985. Template-directed synthesis of novel,nucleic acid-like structures. Science 228: 585–587.



Shan S, Kravchuk AV, Piccirilli JA, Herschlag D. 2001. Defining thecatalytic metal ion interactions in the Tetrahymena ribozyme reaction.Biochemistry 40: 5161–5171.

Shan S, Yoshida A, Sun S, Piccirilli JA, Herschlag D. 1999. Three metalions at the active site of the Tetrahymena group I ribozyme. ProcNatl Acad Sci 96: 12299–12304.

Shapiro R. 1984. The improbability of prebiotic nucleic acid synthesis.Orig Life Evol Biosph 14: 565–570.

Sharp PA. 1985. On the origin of RNA splicing and introns. Cell 42:397–400.

Sleeper HL, Orgel LE. 1979. The catalysis of nucleotide polymerizationby compounds of divalent lead. J Mol Evol 12: 357–364.

Soai K, Shibata T, Morioka H, Choji K. 1995. Asymmetric autocatalysisand amplification of enantiomeric excess of a chiral molecule. Nature378: 767–768.

Springsteen G, Joyce GF. 2004. Selective derivatization and seques-tration of ribose from a prebiotic mix. J Am Chem Soc 126:9578–9583.

Stahley MR, Strobel SA. 2005. Structural evidence for a two-metal-ionmechanism of group I intron splicing. Science 309: 1587–1590.

Steitz TA. 1998. A mechanism for all polymerases. Nature 391:231–232.

Steitz TA, Moore PB. 2003. RNA, the first macromolecular catalyst: Theribosome is a ribozyme. Trends Biochem Sci 28: 411–418.

Steitz TA, Steitz JA. 1993. A general two-metal-ion mechanism for cata-lytic RNA. Proc Natl Acad Sci 90: 6498–6502.

Strater N, Lipscomb WN, Klabunde T, Krebs B. 1996. Two-metal ion cat-alysis in enzymatic acyl- and phosphoryl-transfer reactions. AngewChemie 35: 2024–2055.

Strobel SA, Ortoleva-Donnelly L. 1999. A hydrogen-bonding triad stabil-izes the chemical transition state of a group I ribozyme. Chem Biol 6:153–165.

Szabo P, Scheuring I, Czaran T, Szathmary E. 2002. In silico simulationsreveal that replicators with limited dispersal evolve towards higher ef-ficiency and fidelity. Nature 420: 340–343.

Sulston J, Lohrmann R, Orgel LE, Todd MH. 1968. Nonenzymatic syn-thesis of oligoadenylates on a polyuridylic acid template. Proc NatlAcad Sci 59: 726–733.

Tapiero CM, Nagyvary J. 1971. Prebiotic formation of cytidine nucleoti-des. Nature 231: 42–43.

Unrau PJ, Bartel DP. 1998. RNA-catalysed nucleotide synthesis. Nature395: 260–263.

Ura Y, Beierle JM, Leman LJ, Orgel LE, Ghadiri MR. 2009. Self-assembling sequence-adaptive peptide nucleic acids. Science 325:73–77.

Viedma C. 2005. Chiral symmetry breaking during crystallization: Com-plete chiral purity induced by nonlinear autocatalysis and recycling.Phys Rev Lett 94: 065504.

Viedma C, Ortiz JE, de Torres T, Izumi T, Blackmond DG. 2008. Evolu-tion of solid phase homochirality for a proteinogenic amino acid. J AmChem Soc 130: 15274–15275.

Wachtershauser G. 1988. Before enzymes and templates: Theory of sur-face metabolism. Microbiol Rev 52: 452–484.

Wang KJ, Ferris JP. 2001. Effect of inhibitors on the montmorillonite clay-catalyzed formation of RNA: Studies on the reaction pathway. Orig LifeEvol Biosph 31: 381–402.

Weber AL. 1987. The triose model: Glyceraldehyde as a source of energyand monomers for prebiotic condensation reactions. Orig Life Evol Bi-osph 17: 107–119.

Weber AL. 1989. Model of early self-replication based on covalent com-plementarity for a copolymer of glycerate-3-phosphate andglycerol-3-phosphate. Orig Life Evol Biosph 19: 179–186.

Weiner AM, Maizels N. 1987. tRNA-like structures tag the 30 ends of ge-nomic RNA molecules for replication: Implications for the origin ofprotein synthesis. Proc Natl Acad Sci 84: 7383–7387.

Wimberly BT, Brodersen DE, Clemons WMJr, Morgan-Warren RJ, Car-ter AP, Vonrhein C, Hartsch T, Ramakrishnan V. 2000. Structure of the30S ribosomal subunit. Nature 407: 327–338.

Wittung P, Nielsen PE, Buchardt O, Egholm M, Norden B. 1994. DNA-like double helix formed by peptide nucleic acid. Nature 368:561–563.

Woese C. 1967. The genetic code, pp. 179–195. Harper and Row,New York.

Wu T, Orgel LE. 1992a. Nonenzymatic template-directed synthesis onoligodeoxycytidylate sequences in hairpin oligonucleotides. J AmChem Soc 114: 317–322.

Wu T, Orgel LE. 1992b. Nonenzymatic template-directed synthesis onhairpin oligonucleotides. II. Templates containing cytidine andguanosine residues. J Am Chem Soc 114: 5496–5501.

Yarus M. 1993. How many catalytic RNAs? Ions and the Cheshire Catconjecture. FASEB J 7: 31–39.

Yusupov M, Yusupova G, Baucom A, Lieberman K, Earnest TN, Cate JH,Noller HF. 2001. Crystal structure of the ribosome at 5.5 A resolution.Science 292: 883–896.

Zhang S, Lockshin C, Cook R, Rich A. 1994. Unusually stable beta-sheetformation in an ionic self-complementary oligopeptide. Biopolymers34: 663–672.

Zhang L, Peritz A, Meggers E. 2005. A simple glycol nucleic acid. J AmChem Soc 127: 4174–4175.

Zielinski WS, Orgel LE. 1985. Oligomerization of activated derivatives of30-amino-30-deoxyguanosine on poly(C) and poly(dC) templates.Nucleic Acids Res 13: 2469–2484.

Zielinski WS, Orgel LE. 1987. Oligomerization of dimers of30-amino-30-deoxy-nucleotides (GC and CG) in aqueous solution.Nucleic Acids Res 13: 1699–1715.

Zubay G. 1998. Studies on the lead-catalyzed synthesis of aldopentoses.Orig Life Evol Biosph 28: 13–26.



synthesis of prebiotic organic molecules.It has been generally accepted that at about 1.5 Ga

[Giga annum = billion years ago] the oxygen content ofthe air rose at least 15-fold. (Note that evolutionary/uniformitarian ‘ages’ are only used here for argument’ssake.) Before this, the oxygen had been reduced by Fe(II)in sea water and deposited in enormous bands as oxides orhydroxides on the shallow sea floors. The source of theferrous iron was hydrothermal vents in the company ofreducing gases such as hydrogen sulphide (H2S).

In 1993 Widdel and his team cultured non-sulphurbacteria from marine and freshwater muds. Theseanoxygenic, photosynthetic bacteria use ferrous iron asthe electron donor to drive CO2 fixation. It was a signaldiscovery that oxygen-independent biological ironoxidation was possible before the evolution of oxygen-releasing photosynthesis. Quantitative calculations supportthe possibility of generating such massive iron oxidedeposits dating from Archaean and Early Proterozoic times,3.5–1.8 Ga.4

AW SWEE-ENG

ABSTRACT

Profound advances in the fields of molecular biology in recent yearshave enabled the elucidation of cell structure and function in detailpreviously unimaginable. The unexpected levels of complexity revealed atthe molecular level have further strained the concept of the random assemblyof a self-replicating system. At the same time, the recent discovery of fossilalgae and stromatolites (primitive colonies of cyanobacteria) from as earlyas the Precambrian, have reduced the time for development of the first cellas much as tenfold. Together with implications of this for the oxidativestate of the primitive atmosphere, these developments will force researchersto rethink many fundamental ideas pertaining to current models of the originof life on Earth. The evidence for the nature of the primitive atmosphere isexamined and the possibility of ribonucleic acid (RNA) as the first self-replicating molecule is evaluated. The focus is then on DNA, proteins andthe first cells.

The Origin of Life:A Critique of

Current Scientific Models

THE EARLY ATMOSPHERE

The nature of the atmosphere under which life aroseis of great interest. The high oxygen content of the Earth’satmosphere is unique among the planets of the Solar Systemand could have been tied up with the composition of thecore and its crust. It has to be said that none of thehypotheses of core formation of the Earth survivesquantitative scrutiny. The gross features of mantlegeochemistry, such as its redox state (FeO) and its iron–sulphur systems, apparently do not agree with experimentaldata.1,2 There are outstanding questions relating to theformation and recycling of the Archaean crust.3

Interesting organic molecules such as sugars and aminoacids can be formed from laboratory ‘atmospheres’ ofdifferent proportions of CO2, H2O, N2, NH3, H2, CH4, H2Sand CO. This happens only in the absence of free O2.Oxygen is highly reactive, breaking chemical bonds byremoving electrons from them. A reducing gas (H2, CH4

or CO) is therefore thought to be essential for the successful

300 CEN Tech. J., vol. 10, no. 3, 1996

In-Depth ReviewsIn-Depth ReviewsIn-Depth ReviewsIn-Depth ReviewsIn-Depth Reviews

The Origin of Life — Aw

CEN Tech. J., vol. 10, no. 3, 1996 301


less hydrogen, methane and ammonia. Still, it seemsprudent to consider other mechanisms for theaccumulation of the constituents of proteins and nucleicacids in the prebiotic soup.For instance, the amino acids and nitrogen-containingbases needed for life on the earth might have beendelivered by interstellar dust, meteorites and comets.’15

In his essay on the origin of life on Earth, Orgel quotesthe experiments of Miller, and of Juan Oro' who used theMiller model to produce adenine with hydrogen cyanideand ammonia.16 His conclusions overall are:

‘Since then, workers have subjected many differentmixtures of simple gases to various energy sources.The results of these experiments can be summarizedneatly. Under sufficiently reducing conditions, aminoacids form easily. Conversely, under oxidizingconditions, they do not arise at all or do so only insmall amounts.’Saturn’s giant moon, Titan, has an atmosphere

composed mainly of molecular nitrogen and up to 10 percent methane. Carl Sagan and Bishun Khare of CornellUniversity simulated the pressure and composition ofTitan’s atmosphere and irradiated the gases with chargedparticles. A dark solid was formed, which on dissolvingin water yielded amino acids and traces of nucleotide bases,polycyclic hydrocarbons and many other compounds. Itwas then assumed that from this ‘wonderful brew’ lifewould have originated.17 In the text Molecular Biologyof the Cell the authors note that experimentalists are

In 1992 Han and Runnegar made a discovery whichimpinged on discussions of oxygen evolution during thePrecambrian. To everyone’s surprise they reported thespiral algal fossil Grypania within banded iron formations(BIFs) in Michigan, USA. Algae require oxygen, so theirexistence at this juncture shows banded iron formationsdo not necessarily indicate global anoxic conditions.5

Indeed, as early as 1980 two reports appeared on thediscovery of stromatolites in the 3.4–3.5 Ga WarrawoonaGroup sediments from the Pilbara Block, Australia.6,7

Similar remains were also discovered in Zimbabwe8 andSouth Africa.9

It is fair to conclude that the Earth’s early atmospherebefore 3.5 Ga could have significant quantities of oxygen.This should discourage the sort of hypothesising on abioticmonomer and polymer syntheses so often assumed to haveoccurred in Archaean times. Robert Riding says that theGrypania discovery

‘could spell the end of BIF-dominated models ofoxygen build-up in the early atmosphere . . . The catreally will be put among the pigeons, however, if[further] fossil discoveries extend the eukaryote recordback much beyond 2200 million years ago, into whatis still widely perceived to have been an essentiallyanaerobic world.’10

SCENARIOS FOR PREBIOLOGY

A number of revised textbooks on molecular biologycame out in 1994–1995 which, while conveying thestandard arguments for origin-of-life hypotheses, arecautious in their affirmation. Rightly so, because advancesin the field have uncovered exquisite details of intracellularprocesses. These challenge superficial explanations thattheir origin and subsequent refinement were fed byrandomness. After mentioning the famous simulation byMiller and Urey of prebiotic synthesis of organiccompounds (Figure 1), Voet and Voet handle the riddle ofthe formation of biological monomers with a caveat. Theywrite:

‘Keep in mind, however, that there are valid scientificobjections to this scenario as well as to the severalothers that have been seriously entertained so that weare far from certain as to how life arose.’11

The text of Molecular Cell Biology in its secondedition was well indexed on the evolution of cells,describing the Miller experiment in detail.12 The thirdedition has dropped the chapter on evolution of cells foundin the second edition.13 Similarly, Stryer’s fourth editionof his textbook on biochemistry makes no mention of theabiotic synthesis of organic molecules.14

‘Doubt has arisen because recent investigationsindicate the earth’s atmosphere was never as reducingas Urey and Miller presumed. I suspect that manyorganic compounds generated in past studies wouldhave been produced even in an atmosphere containing

Figure 1. Simplified apparatus for abiotic synthesis of organiccompounds as performed originally by Miller and Urey. Byvarying the mixture of gases, including using volcanic gasesof today, experimenters have been able to produce manytypes of organic compounds.

302 CEN Tech. J., vol. 10, no. 3, 1996

The Origin of Life — AwIn-Depth ReviewsIn-Depth ReviewsIn-Depth ReviewsIn-Depth ReviewsIn-Depth Reviews

gases issuing from the vents, withhydrothermal mixing there would emergepeptides, nucleotides and even protocellsof some sort. Miller and Bada, however,dispute the plausibility.‘This proposal, however, is based on anumber of misunderstandings concerningthe organic chemistry involved. Anexample is the suggestion that organiccompounds were destroyed on the surfaceof the early Earth by the impact ofasteroids and comets, but at the same timeassuming that organic syntheses canoccur in hydrothermal vents. The hightemperatures in the vents would not allowsynthesis of organic compounds, butwould decompose them, unless theexposure time at vent temperatures wasshort. Even if the essential organicmolecules were available in the hothydrothermal waters, the subsequentsteps of polymerization and theconversion of these polymers into the firstorganisms would not occur as the ventwaters were quenched to the coldertemperatures of the primitive oceans.’20

TIME-SPAN FOR PREBIOLOGY

A pillar of ‘prebiological evolution’has been the long period of timesupposedly available for the emergenceof ‘protocells’ whose development in turnprofoundly altered the climate of the

planet and its geology. For an estimated age of the Earthof 4.6 Ga this seemed initially to pose no problem.However, the discovery of stromatolites in WesternAustralia21,22 and in South Africa23,24 upset the timetableseverely. The finding of algal filaments dated at onlyslightly more than 1 Ga younger than the Earth itselfrestricted the time required for the evolution of the livingcell. Pari passu the list of processes thought to occurabiotically has been shrinking.25,26 Even the origin of thehuge banded iron formations of the Archaean can now beattributed to microorganisms,27 and Raup and Valentinehave suggested that bolide impacts have, at intervals of105 to 107 years, periodically erased more than one originof life.28 According to this scenario, ten or more extinctbioclades could have preceded the Cambrian. A biocladeis a group of life forms descended from a single event oflife origin. 4.2 Ga has been given as the date of the oldestrocks, which is ostensibly consistent with the cooling anddegassing of an active molten Earth that is said to be 4.6 Gaold.29 According to the isotopic carbon record insedimentary rocks, 3.8 Ga would date the origin of life.30

Fred Hoyle, the Cambridge astronomer and physicist,

beguiled by the ‘surprisingly easy’ manner in which organicmolecules form.18 Little store is laid for such crucial pointsas the lability of the organic products, or their reactivityamong themselves to form mixed polymers. Indeed, theproblem of spontaneously producing a simple homochiralcompound, say, L-alanine, from racemic reaction systemshas not been solved (see Figure 2).

Classical mechanisms generally rely on chance for theselection of L-amino and D-sugars by self-replicatingsystems. Mason has put forward the tantalising speculationthat a weak nuclear interaction will stabilise the L-aminoacids and their polypeptides over their D-forms. Thiselectroweak advantage is considered too weak to affectthe outcome of biochemical evolution. An imaginary flowreactor of a kilometre in diameter and four metres deepwould be needed to autocatalyse a change of 10-2 to 10-3

moles of one isomer over 10,000 years if the temperatureis kept at ambient. Admittedly a good thought experiment‘but it will find no popular primitive Earth scenarios.’19

The discovery of hydrothermal vents at oceanic ridgecrests has spawned several origin-of-life hypotheses. Itseemed an attractive suggestion that, given the dissolved

Figure 2. Optical activity and chirality. Ordinary light consists of waves vibrating in all possibledirections perpendicular to its path. Certain substances will selectively transmitlight waves vibrating only in a specific plane — plane polarised light. Mostcompounds isolated from natural sources are able to rotate the plane of polarisedlight a characteristic number of degrees for any specific substance. The significanceof this phenomenon to molecular biology and the origin of life is that stereoisomers,molecules of identical but mirror image structure, possess such ‘optical activity’.For example, in the case of the stereoisomers of the amino acid alanine shownabove, L-alanine will rotate the plane of polarised light in the opposite direction to D-alanine. Why biological systems utilise exclusively levorotatory (left-handed) aminoacids and dextrorotatory (right-handed) sugars remains unfathomable. Mixturesof organic compounds synthesised in Urey-Miller type experiments always consistof racemic (equal amounts of left- and right-handed) mixtures.


CEN Tech. J., vol. 10, no. 3, 1996 303


strands of RNA that reproducedthemselves, perhaps on clay surfaces.This conjecture is strengthened by the factthat in cells today there are segments ofsome eukaryotic pre-rRNAs which cancleave themselves off and join the two cutends together to reform the mature rRNA.In 1982 Thomas Cech and his colleaguesat the University of Colorado discoveredthis can take place in the absence ofprotein in the ciliated protozoanTetrahymena thermophila.35 Just asremarkable are the small nuclear RNAs(snRNAs), which complex with proteinto form small nuclear ribonucleoproteins(snRNPs; pronounced ‘snurps’).Particles called spliceosomes convert pre-mRNA to mRNA.36 Other ribozymesinclude the hammerhead variety andRNAse P, which generates the 5' ends oftRNAs. The former are found in certainplant viruses. Origin-of-life theories seeprebiotic significance in these ‘vestigial’post-translational mechanisms.

Though attractive, there are severalserious objections to the notion that lifebegan with RNA:–(1)Pentose sugars, constituents of RNAand DNA, can be synthesised in theformose reaction, given the presence offormaldehyde (HCHO). The products area melange of sugars of various carbonlengths which are optically left- and right-handed (D and L). With few exceptionssugars found in biological systems are ofthe D type; for instance, β-D-ribose ofRNA, which is always produced in smallquantities abiotically.(2)Hydrocyanic acid (HCN) undergoespolymerisation to form diamino-maleonitrile which is on the pathway toproducing adenine, hypoxanthine,guanine, xanthine and diaminopurine.These are purines: there is difficulty inproducing pyrimidines (cytosine, thymineand uracil) in comparable quantities37,38

(see Figure 3).(3)Neither preformed purines norpyrimidines have been successfully linkedto ribose by organic chemists. An attemptto make purine nucleosides resulted in a‘dizzying array of related compounds’.39

This is expected if sugars and bases wererandomly coupled. The prebioticproduction of numerous isomers andclosely related molecules hinders the

made some sobering calculations on theorigin of the cell.31 The probability offorming the 2,000 or so enzymes neededby a cell lies in the realm of 1 in 1040,000.This makes the conceptual leap from eventhe most complex ‘soup’ to the simplestcell in the time available (that is, about500 Ma) so dramatic that it requires somesuspension of rationality in order toaccept it. Small wonder that latterly it isbeing touted that life may have taken farless time to appear.

Carl Sagan has opined:‘If 100 million years is enough for theorigin of life on the earth, could 1,000years be enough for it (to appear) onTitan?’32

A RIBONUCLEIC ACID (RNA)WORLD

RNA is a linear polymer of ribo-nucleotides, usually single stranded. Eachribonucleotide monomer contains thesugar ribose linked with a phosphategroup and one of four bases: adenine,guanine, cytosine or uracil. RNA appearsin both prokaryotic and eukaryotic cellsas messenger RNA (mRNA), transferRNA (tRNA) and ribosomal RNA(rRNA) which are involved in proteinsynthesis with DNA the source ofinformation. Some viruses howevercontain genomes of RNA. The nuclei ofeukaryotic cells carry two other types ofRNA; heterogeneous nuclear RNA(hnRNA or pre-mRNA) and small nuclearRNA (snRNA).

In recent literature there is muchexcitement over the discovery that thereare RNAs that can catalyse specif icbiochemical reactions. These are theribozymes, that is, RNA with enzymaticfunctions.33 RNA can do this surprisingfeat by folding its linear chains toappropriate secondary and tertiarystructures thereby conferring ‘domain’type catalytic structures as seen in proteinenzymes.

That RNA can act as a template andalso now exhibits catalytic activity fuelledhypotheses for the evolution of an ‘RNAworld’.34 In this scenario RNA is theprimary polymer of life that replicatesitself. DNA and proteins were laterrefinements. So the first genes were short

Figure 3. The molecular structures ofdeoxyribonucleic acid (DNA)and ribonucleic acid (RNA)are built using thenitrogenous bases adenineand guanine (purines), andthymine, cytosine and uracil(pyrimidines), which are the‘letters’ of the genetic code.

304 CEN Tech. J., vol. 10, no. 3, 1996


likelihood of forming desirable mononucleosides.Furthermore, unless ribose and the purine bases formnucleosides rapidly they would be degraded quite quickly.

Purine and PyrimidineNucleotide Biosynthesis

Purine ribonucleotides (for example, AMP, GMP) aresynthesised from scratch by living systems in ways notremotely connected with the laboratory models. The purinering system is built up stepwise from an intermediate 5'-phosphoribosyl-1-pyrophosphate (PRPP) to a largermolecule inosine monophosphate (IMP). This involves apathway comprising 11 reactions.

The biosynthesis of pyrimidines is less complex, butagain the process is elegantly dissimilar to the in vitrochemistry, with some of the enzymes on the pathwayexercising regulatory functions.

The purine and pyrimidine biosynthetic pathways arefinely tuned, and defects such as enzyme deficiencies, theirmutant forms or loss of feedback inhibition, cause diseasesin man.

Suppose that we already have mononucleosides —purines (or pyrimidines) linked to ribose. Heating thesein a mixture of urea, ammonium chloride and hydratedcalcium phosphate has been shown to produce mono-, di-and cyclic phosphates of the mononucleoside. Thesubsequent chemistry would yield a rich (or untidy,depending on how it is viewed) racemic mixture of D andL-oligonucleotides in all sorts of combinations andpermutations. Internal cyclisation reactions would destroymuch of these oligonucleotides.40

Suppose further that we have a parent strand of RNAin a chirally-mixed pool of activated monoribonucleotides.By base-pairing the strand correctly aligns on itself theincoming monomeric units in matching sequence.Phosphodiester bonds are spontaneously forged.

The chief obstacles to efficient and faithful copyingappear to be threefold.41

(a) D-mononucleotides and L-mononucleotides hindereach others’ polymerisation on an RNA template.

(b) Short chains of nucleotides tend to fold back onthemselves to form double helical Watson-Cricksegments.

(c) Newly formed strands separate with difficulty fromtheir parent RNA strands. The process grinds to a halt.

Using activated monomers — both nucleotides and aminoacids — Ferris and his co-workers could form oligomersup to 55 monomers long on mineral surfaces. Suchsurfaces bind monomers of one charge (negative in theseexperiments) and strength of binding increases with chainlength. Desorption then becomes impossible.42

Joyce sums up the difficulties of conjuring up ahypothetical RNA world in these words.

‘The most reasonable interpretation is that life did notstart with RNA . . . . The transition to an RNA world,like the origins of life in general, is fraught with

uncertainty and is plagued by a lack of relevantexperimental data. Researchers into the origins of lifehave grown accustomed to the level of frustration inthese problems . . . . It is time to go beyond talkingabout an RNA world and begin to put the evolution ofRNA in the context of the chemistry that came beforeit and the biology that followed.’43

These sentiments are shared by Orgel, a long-time,well-known prebiotic chemist. In 1994 he wrote:

‘The precise events giving rise to the RNA world remainunclear. As we have seen, investigators have proposedmany hypotheses, but evidence in favour of each ofthem is fragmentary at best. The full details of howthe RNA world, and life, emerged may not be revealedin the near future.’44

As we have seen, the intuition that an RNA worldpreceded DNA and protein is based on some features foundin modern cells. But it appears to be contradicted by theavailable experimental evidence. In fact, the extra hydroxylof ribose renders it more reactive than deoxyribose and, inprinciple, makes the more stable DNA a more likelyprogenitor.

Other OptionsAttention switched to other molecules that can carry

information and replicate themselves. In 1991 a team ofDanish chemists led by Egholm strung the four familiarbases of nucleic acids along a peptide (polyamide)backbone forming a peptide nucleic acid (PNA).45,46

Unfortunately, PNAs bind natural DNA and RNA tightly(about 50 to 100 times stronger than the natural polymers

KEY POINTS

• The presumed rise of oxygen levels in a primitivereducing atmosphere formerly attributed to theevolution of photosynthesis can be explained byoxygen-independent biological iron oxidation.

• Recent investigations indicate that the Earth’satmosphere was never as reducing as previouslythought.

• Recent discovery of fossil stromatolites and algaefrom the Precambrian has reduced the time forevolution of the first cell ten-fold.

• The atmosphere of 3.5 billion years ago could havecontained significant quantities of oxygen.

• Under oxidising conditions, the formation of organiccompounds and their polymerisation do not occur.

• Biological homochirality of sugars and amino acidsremains an enigma.

• Hypotheses of ribonucleic acids (RNAs) as the initialself-replicating molecule have serious unresolveddifficulties.

• Extrapolating results of in vitro synthesis of purinesand pyrimidines should take into account thatbiosynthesis utilises different reaction pathways.


CEN Tech. J., vol. 10, no. 3, 1996 305


classes are presumed to have evolved. It has been proposedthat the pristine reductase enzyme, similar to present-dayclass III enzymes, arose before the advent ofphotosynthesis and therefore before the appearance ofoxygen.

Now the E. coli class III enzyme mentioned abovecan be induced by culturing the bacteria under anaerobicconditions. This enzyme is an Fe-S protein that in its activeform contains an oxygen-sensitive glycyl free radical.51

This poses a conundrum: the survival and continualevolution of an oxygen-sensitive enzyme when oxygenappeared. On the other hand, the class I reductases requireoxygen for free radical generation. Surely they could nothave evolved and operated in the anaerobic first cell in anoxygen-free environment.52 Moreover, one of the mostremarkable aspects of this E. coli ribonucleotide class Ireductase is its ability to maintain its highly reactive freeradical state for a long period. Interestingly, this is achievedin vivo by internally generated oxygen. Four proteinshave to be in place:–• Flavin oxidoreductase, which releases superoxide ion (O2

- � ),• Superoxide dismutase, to rapidly convert this destructive

radical to H2O2 and O2,• A catalase, to disproportionate H2O2 to H2O and O2, and• A fourth protein, thioredoxin, that functions as a

reductant.The oxygen oxidises Fe II and a deeply buried tyrosylresidue (Tyr122). Herein lies a difficulty. The reductasesare complex protein reaction centres acting in tandem oneach other and on the 2'-OH group of ribose. These mustall have co-evolved before DNA and along with RNA.Could this be seriously contemplated for a metabolicallynaive RNA ‘progenote’?

The origins of deoxyribose and of DNA thereforeremain unsolved mysteries.

Even if the DNA molecule were assembled abiotically,there is the instability and decay of the polymer byhydrolysis of the glycosyl bonds and the hydrolyticdeamination of the bases.53 Each human cell turns over2,000–10,000 DNA purine bases every day owing tohydrolytic depurination and subsequent repair. Geneticinformation can be stored stably only because a battery ofDNA repair enzymes scan the DNA and replace thedamaged nucleotides. Without these enzymes it would beinconceivable how primitive cells kept abreast of theconstant high-level damage by the environment and byendogenous reactions. If unrepaired, cell death wouldresult. Indeed, the spontaneous errors resulting fromintrinsic DNA instability are usually many times moredangerous than chance injuries from environmentalcauses.54 The enzymes of the DNA repair system are amarvel in themselves and have been rightfully recognisedas such.55

Reports of the culture of Bacillus sphaericus fromspores preserved in amber for over ‘25 million years’ does

bind among themselves) so that it is difficult to envisagetheir being a prebiotic replicating system. So strong istheir affinity for DNA that they would disrupt nucleotideduplexes unless they were removed from an evolving RNAmilieu. Their base-specificity for natural nucleic acids ofoligomers of 10 units or more, and consequently theirfidelity in copying RNA or DNA, is uncertain. Thismilitates against the co-evolution of multiple geneticsystems, a suggestion raised by Böhler and his co-workers.47 Using an unusual activated monomer, guanosine5'-phosphoro (2-methyl) imidazolide, they formed 3'-5'-linked oligomers with PNA as template. In fact, becauseof problems of cyclisation the activated dimer rather thanthe monomer was used. No oligomers of more than 10were formed, and there was present in the complex mixtureshort oligomers with unnatural 2'-5'-phosphodiester bonds,pyrophosphate linked oligomers and possibly cyclicoligomers.

THE DNA STORY

Like RNA, deoxyribonucleic acid (DNA) is a linearpolymer of nucleotides. Each nucleotide consists of apentose sugar, a nitrogenous base and a phosphate group.The sugar–phosphate linkages form an external backbonewith the bases sticking in and hydrogen-bonding withcomplementary bases of the opposite sugar–phosphatebackbone, zipper-fashion, producing the famous doublehelix structure of DNA. The helix can take on alternateforms in which it twists to alter the compactness of itsspiral and bends to change its overall shape. The packingof DNA in a microscopically visible chromosomerepresents a 10,000-fold shortening of its actual length.Little is known of the structure of DNA in the natural statewithin the cell. Clearly it is dynamic, and by assumingdifferent forms DNA controls various biological processessuch as replication, transcription and recombination. Thisis a fruitful area for research.

The Synthesis of βββββ-D-RiboseThe abiotic origin of DNA is beset with problems

similar to those seen with RNA.48 The synthesis ofdeoxyribose forms the nub. We have already mentionedthe difficult synthesis of even small amounts of β-D-ribosefor the in vitro production of RNA. Furthermore, we mighthave expected deoxyribonucleotides to be biosynthesisedde novo from deoxyribose precursors. In real life, however,DNA components (the deoxyribonucleotides dADP, dCDP,dGDP and dUDP) are synthesised from their correspondingribonucleotides by the reduction of the C2' position. Theenzymes that do this are named ribonucleotide reductases.There are three main classes of reductases. All replacethe 2'-OH group of ribose via some elegant free radicalmechanisms.49,50 The class III anaerobic Escherichia colireductase is thought to be the most closely related to thecommon reductase ancestor from which the three main

306 CEN Tech. J., vol. 10, no. 3, 1996


not tally with what is known of the physico-chemicalproperties of DNA.56

Several DNA ParadoxesThe total amount of DNA in the haploid genome is its

C-value. Intuitively we would expect that there should bea relationship between the complexity of an organism andthe amount of its DNA. The failure to consistently correlatethe total amount of DNA in a genome with the genetic andmorphological complexity of the organism is called theC-value paradox.57 This paradox manifests itself in threeways.(1) Many plant species have from two to ten times more

DNA per cell than the human cell. Among thevertebrates with the greatest amount of DNA are theamphibians. Salamander cells contain 10–100 timesmore DNA than mammalian cells. It is hard to makesense of the existence of such major redundancies inorganisms evolutionarily less complex than man.

(2) There is also considerable intragroup variation in DNAcontent where morphology does not vary much. Forexample, the broad bean contains about three to fourtimes as much DNA per cell as the kidney bean.Variations of up to 100 times are found among insectsand among amphibians. In other words, cellular DNAcontent does not correlate with phylogeny.

(3) Large stretches of DNA in the genome, say, of humans,appear to have no demonstrable function. This will bediscussed later.

INTRONS AND EXONS

Once the genes of unrelated cells were studied itbecame clear that the molecular genetics of higherorganisms are different from those of bacteria. Theprinciples uncovered in prokaryotes cannot simply beapplied to eukaryotes. For one thing, the precursor RNAfound in the nucleus, called heterogeneous nuclear RNA(hnRNA), was far greater in amount than the mRNA thatemerged from the nucleus into the cytoplasm. It wasdiscovered that the linear hnRNA molecule containedexcess RNA which was cut out, and the mRNA was thenconstructed from splicing together the in-between pieces.An editing process had taken place.58 The logical inferencefrom this finding was that the genomic DNA from whichthe hnRNA was transcribed must be similarly constructed.The notion of the co-linear relationship between a segmentof DNA and the protein for which it codes is not true, atleast for higher organisms.

The word ‘intron’ was used to describe such a non-coding region of a structural gene. They separate the‘exons’, which encode the amino acids of the protein.59

For instance, the human β-globin gene comprises, in linearsequence, three exons separated by two introns within atotal length of 1,600 nucleotides. Introns are abundant inhigher eukaryotes, uncommon in lower eukaryotes, and

rare in prokaryotic structural genes. Variations in the lengthof the genes are primarily determined by the lengths ofthe introns. Since the discovery of introns/exons theintricate processes of nuclear mRNA splicing have beenelegantly elucidated. Among these are the remarkable self-splicing introns60 and the equally revolutionary finding thatindividual nucleotides can be inserted into RNA aftertranscription altering them remarkably.61

The inevitable questions emerged. What role doeshaving genes in pieces serve? How have such interruptedgenes ‘evolved’ over time?

One hypothesis points out that exons usually encodefor a part of the protein that folds to form a domain. Whatconstitutes a domain has been a matter of controversy. Bydispersing individual exons of a protein among introns itis reasoned that breaking DNA and rejoining andrecombining different exons is that much easier. Thisprocess of shuffling exons/domains is presumed to havecreated new proteins with multi-domain structures. Thisis thought to be a more efficient way for a cell to createproteins rather than through random DNA mutations. Hereis a means of duplicating, modifying, assembling andreassembling units with modular functions into largerstructures. According to this hypothesis this is the reasonwhy introns have survived through time. Several queriesmay be raised. First, exon shuffling as a device to speedup evolution is logically tied up with a subsidiaryassumption that possessing similar domains qualif iesproteins for biochemical kinship, which is to say, theseproteins are alleged to bear the marks of descent from acommon ancestral protein.62 But the construction ofphylogenetic trees relies on unstable molecular clocks andother genetic mechanisms largely unknown63 and, asdiscussed below, should be approached with caution.

Biochemical kinship aside, would not domainsexercising similar function be structurally alike such aswe see between, say, the catalytic domains of the two serineproteases chymotrypsin and tissue plasminogen activator?

Second, RNA splicing is an accurate and complexprocedure comparable in complexity to protein synthesisand initiation of transcription. It is carried out by a 50S to60S ribonucleoprotein made up of small nuclearribonucleoproteins (snRNPs) as well as other proteins. Justas the ribosome is built up in the process of translation,the spliceosome components assemble in an orderlymanner on the intron to be spliced before the initialcleavage of the 5' splice site. The splicing must be carriedout precisely, joining the 5' end of the preceding exon tothe 3' end of that following. A frameshift of even onenucleotide would change the resulting mRNA message.The inescapable conclusion is that these interlockingcomponents must have ‘evolved’ together, as an imperfectsplicing mechanism is worse than none.

Third, were the original protein-coding units seamless,that is, uninterrupted by introns? And were the intronsbits of ‘selfish DNA’ that later insinuated themselves into


CEN Tech. J., vol. 10, no. 3, 1996 307


the hosts’ structural genes? What purpose then thesubsequent evolution of a multi-step complicated splicingmachinery to remove the introns?64-69 Would not simplyeliminating the introns make better sense for selectiveadvantage?

Fourth, and most importantly, transport of mRNA fromthe nucleus to the cytoplasm is coupled to splicing anddoes not occur until all the splicing is complete. Howdoes the RNA enter the cytoplasm for translation duringthe evolution of the splicing mechanism? This would havedisrupted protein synthesis and would be powerfullyselected against.70-72 Why is splicing in all its variants sorampant today?

The problem would arise too were introns abundant incells without nuclear membranes — the prokaryotes.Mattick wrote:

‘If introns were introduced into a procaryotic cell’sgenes, there would be no opportunity to remove thembefore protein is made, and the result would be“nonsense” non-functional proteins.’73

This is essentially correct because spliceosomes would beneeded for their removal, but again begs the question onthe viability of the transitional phases.

The relationships between exons and protein domainsremain to be worked out. Where introns came from andhow they were integrated into the genome is a mystery toevolutionists.74

THOSE OVERLAPPING CODES

Messenger RNAs generally contain only one readingframe which is dictated by the position of the initiationcodon. This correct reading frame translates the nucleotidecode into a functional protein. Starting at an AUG codon,translation continues in triplets to a termination codon. Thestarting point can be altered by a mutation, usually resultingfrom insertion or deletion of a single nucleotide to give analternate reading frame. A frameshift error results in thesynthesis of a polypeptide that does not resemble thenormal product. Typically, it will be inactive and, becausestop codons are abundant in the alternative frames, shorterthan the native protein.

Some organisms store information in their DNA in theform of overlapping codes. The overlapping codes arestill triplet but have different initiation points. In otherwords, the same stretch of DNA carries the informationfor producing two proteins of entirely different amino acidsequence. This discovery is truly startling, because thepossibility that genes might overlap in different readingframes imposes severe evolutionary constraints. Afavourable mutation in one frame must be favourable inthe other. A termination codon in the second frame wouldbe fatal to the organism as a whole. So the two overlappinggenes have to evolve in parallel. Yockey considered theproblem from the point of view of information theoryapplied to biology, itself a venture fraught with caveats.75

In his opinion information theory shows that transcriptionfrom two or even three reading frames in a DNA or RNAsequence is possible, provided the total informationalcontent to be transcribed does not exceed the fullinformational capacity of the DNA or RNA sequence. Thisinteresting bit of information is a necessary but not asufficient explanation for the origin of overlapping codes.The packing of information for synthesising additionalessential proteins through weaving such information intoa pre-existing nucleotide sequence is little short ofmiraculous, assuming that chance is the author.

Most of the known examples of such programmedframeshifts occur in viral genes.76,77 The notorious hepatitisB virus has four open reading frames on the long strand ofits DNA to produce four different proteins. In a strikingdemonstration of sheer economy it turns out that eachreading frame overlaps at least one other frame. Andthe code for the polymerase enzyme overlaps the otherthree.78 It is true that programmed frameshifts are notcommon, but they have been found across a wide spectrumof organisms. Yeast and E. coli also practiseframeshifting.79,80 The mechanisms by which they workseem to involve ‘shifty’ messages in the mRNA, wherethe ribosomes may read four nucleotides as one amino acidand then continue reading triplets. Or it may back up onebase before reading triplets in the new frame. ‘ShiftytRNAs’ are also implicated.81-83

THE NON-UNIVERSAL CODE

Even the code’s universality — a strong argument forthe hypothesis that life on Earth evolved only once — hasa large number of ‘exceptions’. These are usually creditedto later evolutionary developments, as the following quotefrom a paper by Jukes and his colleagues shows.Commenting on the dearth of molecular studies on

‘the more than 10 million species of organisms nowliving on Earth, all of which are derived from a singlepool of the ancestor’,

they continue:‘. . . nonuniversal codes have been discovered at arelatively high incidence. Codon UGA Trp has beenfound in seven Mycoplasma species and relatedbacteria; at least two kinds of nonuniversal code areindependently used in ciliated protozoans; the samecode change was found in two different organismiclines, ciliated protozoans, and unicellular green algae;a yeast line uses a still different code. All nonplantmitochondria that have been examined usenonuniversal codes, which are more or lesscharacteristic for each line. It is remarkable thatmitochondria from one species use more than twononuniversal codons; six in yeasts, four or five in manyinvertebrates, and four in vertebrates. Thus,nonuniversal codes are widely distributed in variousgroups of organisms and organelles. . . . . The

308 CEN Tech. J., vol. 10, no. 3, 1996


nonuniversal codes are not randomly produced, butare derived from the universal code as a result of aseries of nondisruptive changes.’84

All this just means that hypotheses of the origin of thegenetic code based on our understanding of the nature ofthe DNA, its transcription and translation have to besubstantially revised.

THE SILENT MAJORITY

It is now agreed that any theory on the origin of DNAmust take into account that the genomes of multicellularorganisms are characterised by high intron content. Mattickhas proposed that introns having a high sequencecomplexity be regarded as informational RNA (iRNA).85

Each chromosome is increasingly being viewed as acomplex ‘informational organelle’. At least some nowregard the idea that there is ‘junk’ or ‘useless’ DNA asuntenable,86 but the logical extensions are not usuallyfollowed through.

An unanswered question concerns the enormousamount of DNA in most eukaryotic genomes which appearsto serve no useful purpose. Introns contribute to thisexcess. The highly conserved nature of the sequences inintrons points to the possibility that they have servedimportant function(s) from the time of their f irstappearance in their hosts’ genomes. For instance, mouseand human T-cell receptor genes show 71 per centhomology over their entire 100 kb length even though lessthan six per cent of that length encodes the receptorprotein.87 Recent studies describe finding a RNA regulatorof gene expression originating from the introns of anothermRNA.88,89 This small RNA binds to the so-called 3'untranslated region (3'UTR) which lies at the end of eachgene’s mRNA, once again confounding the notion of‘functionless’ RNA.

Intron-containing genes have yet another intriguingproperty, uncovered in 1992 by Peng and his co-workersin Boston. They introduced a new quantitative method todisplay correlations in the sequence of nucleotides. Totheir surprise they discovered that the nucleotide sequencein intron-containing genes is correlated over remarkableranges of thousands of base pairs apart. Their results arebased on a statistical assessment of 24 viral, bacterial, yeastand mammalian sequences. This means that a particularnucleotide at one site would somehow influence whichnucleotide would locate at a remotely distant site. Thislong-range dependence indicates an intricate self-similaritythat is reminiscent of fractal dynamics.90 In addition thereare hints of a language structure, akin to that seen withordinary languages, in the lengths of non-protein codingDNA. Their findings support the possibility that non-coding regions of DNA may carry biological information.The two standard linguistic tests applied were those ofGeorge Zipf and Claude Shannon. The coding regions ofthe genes returned negative results for both tests.91

Distinctive and previously unsuspected features ofgenomic DNA are beginning to be revealed. What issurprising is the tiers of immense complexity which areburied in its structure. An analogy will not be out of place.Viewing from a great height a road traversing the lengthof a continent, a being from outer space might at firstwonder what purpose such a structure could serve.Unfamiliar with the ways of man, the alien realises thatthe ribbon-like structure actually links areas that areintensely bright at night, which are, of course, our citiesand towns.

Further study by the alien is even more revealing. Thenight-bright entities seem to correlate with the lie of theland, its mountain ranges, rivers and underground mineralresources. The alien may even be momentarily distractedby the question of whether the link or the entities camefirst! What he can conclude, however, is that the structurehe had examined is neither random in design nor intentionover its whole length, but serves to link entities whichthemselves evince design and purpose.

What is increasingly seen as the DNA story unfolds isprima facie evidence of intelligent design extending overthe whole molecule. What used to be thought of as aprodigious 95 per cent excess of repetitive and useless DNAturns out to be an interactive regulatory network controllinggene expression in the remaining five per cent. Even thehumble trinucleotide repeat sequence CAG has beenimplicated in the pathogenesis of a number of seriousneurological diseases.92 This illustrates the complicity ofthe simplest codes in the intricate regulatory network, andputs further strain on ideas of the code’s abiotic origin. Insumming up, let me quote the editor of Nature, who wrotein 1994:

‘The problem of the genetic code has several facets, ofwhich the most compelling is that of why it is why it is. . . . it was natural that people should look for anexplanation, both for its own sake and because anunderstanding of how the code evolved must certainlybe a pointer to the origin of life itself . . . . It wasalready clear that the genetic code is not merely anabstraction but the embodiment of life’s mechanisms;the consecutive triplets of nucleotides in DNA (calledcodons) are inherited but they also guide theconstruction of proteins.So it is disappointing, but not surprising, that the originof the genetic code is still as obscure as the origin oflife itself.’93

THE ORIGIN OF PROTEINS

As with the D-sugars of carbohydrates, so with theamino acids from which proteins are made. They aretypically L (left-rotating) in optical activity. D-amino acidsare found in bacterial products and peptide antibiotics, butthey are not incorporated into proteins via the ribosomalprotein synthesising system.


CEN Tech. J., vol. 10, no. 3, 1996 309


The almost total dominance of one chiral form in presentlife forms is an enigma. Vital processes such as proteinbiosynthesis, ligand-receptor activity, substrate binding,enzymatic catalysis and antigen–antibody interaction dependon the present chemical-handedness. Fisun and Savin haveprovided another example of monochiral utility by examiningproton transfer along the hydrogen-bonded chain formed byamino acids.94 After all, membrane proteins are structuredto enable such transfers to take place as a means of regulatingproton concentrations. The amino acids they examined wereL-tyrosine, L-serine and L-threonine. What would happen,they asked, if a long sequence of such OH-bearing acidswere interrupted by an unnatural D isomer? Their analysesrevealed that it suppressed transfer through the hydrogen-bonded network. The authors point out the generallydisruptive effects that deforming natural polymers with D-amino acids would have on diverse biological phenomena,such as information, charge, energy and mass exchanges.

The evolutionary explanation for left-handed aminoacids is simply that a common ancestor, by sheer coincidence,happened to have this mirror image. Well-wornexplanations, such as the anisotropic effects of refractedlight, are convincing only to those who propose them. ‘Chiralfields’ that could effect a critical prebiotic transition to onechiral species have been worked out on paper.95 The troubleis that, so far, there has been no success for the apparentlysimple problem of tipping the experimental scales to favourone of two isomers.

The problem of chirality is crucial to the origin of life.For Darwinian evolution involves selection, a winnowingprocess that separates the ‘fit’ from the ‘unfit’. The ‘fit’ arethen amplified to ensure a progeny. The ‘fit’ are those ableto do one of two things, depending on the school of thought.The ‘genes first’ school envisages primitive replicons thatlater surrounded themselves with metabolic cycles.96,97 The‘cells first’ school pictures primitive cells covered withprimitive membranes engaged in a sort of metabolicexchange with the environment. These propagatedthemselves by simple expansion followed by division.Genetic mechanisms of inheritance developed gradually.98,99

Both schools founder on the unsettling and unsettledquestion — which came first, homochirality or life?100 Ifone holds that homochirality came first, it is an admissionthat without ‘left-handed’ amino acids and ‘right-handed’sugars life’s structures and processes would have beenimpossible. One then has to account for the origin ofhomochirality. If one assumes that life came first, then oneis saying that chirality was not important to the origin oflife’s structures and processes as we now know them. Onehas to enter a special pleading for a vastly differentmetabolism in the ‘protobiont’, ignoring, for instance, thepivotal role of polypeptide homopolymers in hydrogen-bonded networks for proton and electron transport.101 Onehas also to account for the successful transition tohomochirality as we have it today.

The logical conclusion from these considerations is a

simple and parsimonious one, that homochirality and lifecame together. But evolutionary lore forbids such a notion.It claims to explain how life began, but on the profoundissue of life’s ‘handedness’ there is no selective mechanismthat it can plausibly endorse.

FOLDING PROTEINS

Much thought has been given to suggesting pathwaysas to how a polypeptide chain, freshly made, folds into itsunique shape.102 But biological systems are inherentlycomplicated and so are their components. Today the conceptthat proteins can self-assemble has been modified toincorporate the astonishing part played by accessoryproteins called chaperones, first identified inE. coli.103-107 Chaperones are found in all types of cells andin every cellular compartment. They bind to target proteinsto facilitate proper folding, prevent or reverse improperassociations, and protect their accidental degradation. Ofspecial interest are a subset of chaperones calledchaperonins. They are large, barrel-shaped, polymericproteins present in bacteria, mitochondria, chloroplasts andeukaryotes. They enfold protein chains in a cavity, aprotected micro-environment to allow their guest moleculesopportunity to fold correctly. Chaperones utilise the energyof ATP hydrolysis to bind and release their charges. Theyare also involved in many macromolecular assemblyprocesses, including the assembly of nucleosomes, proteintransport in bacteria, assembly of bacterial pili, binding oftranscription factors, and ribosome assembly in eukaryotes.A subset of molecular chaperones has even been implicatedin signal transduction. This follows upon the discovery thatsteroid hormone receptors, which are cytoplasmic proteins,combine not only with their respective hormones, but alsorequire chaperones in order to form functioning recyclingcomplexes.108 Such structural arrangements must be highlyconserved, seeing that these chaperones are found in similarmacromolecular complexes in organisms as diverse asmammals and yeasts.109 This is supposed to attest to theirgreat antiquity (if evolution is true), because properly foldedproteins are absolutely essential for a cell’s viability.

Lodish and his co-authors express their opinion:–‘Folding of proteins in vitro is inefficient; only aminority undergo complete folding within a fewminutes. Clearly, proteins must fold correctly andefficiently in vivo, otherwise cells would waste muchenergy in the synthesis of non-functional proteins andin the degradation of misfolded and unfoldedproteins.’110

How did cellular proteins avoid being tied up into kinksindividually and aggregates corporately before chaperonescame on the scene? If chaperones help other proteins fold,what mechanism helps chaperones to fold? And chaperonesare themselves complex proteins. A well-studiedchaperonin, Cpn60, has a unique structure, consisting offourteen identical subunits of a 60 kDa protein arranged

310 CEN Tech. J., vol. 10, no. 3, 1996


in two stacked rings of seven.111,112 It interacts with anotherconserved protein chaperonin Cpn10, itself a complex ofseven subunits.113 The answers to these questions wouldindeed be illuminating.

THE ANCIENT CELLS

Prokaryotes and EukaryotesThe existence of chaperones influences the

endosymbiont hypothesis of the origin of eukaryotes. Thishypothesis proposes that chloroplasts and mitochondriabegan as free-living aerobic prokaryote ancestors whichwere engulfed by, and formed, a mutually advantageousrelationship with an ancient large anaerobic prokaryotewith a nucleus.114,115 These endosymbionts became theorganelles mentioned, which then apparently lost many oftheir own genes to the nuclei of their hosts. Now, the time-frame of oxygen levels in the primitive Earth is extremelycontroversial in the face of conflicting palaeobiologicalevidence.116 Nevertheless, how a stable relationshipbetween ingested aerobic invaders and an anaerobic, oraerotolerant, host was possible, and why some genes andnot others should be transferred to the host’s nucleus isnot clear.

An idea of how many genes were ‘lost’ to the hostnucleus may be gleaned from the fact that the cytosolsynthesises for the mitochondria the following proteins:ribosomal proteins, DNA replication enzymes, aminoacyl-tRNA synthases, RNA polymerase, soluble enzymes of thecitric acid cycle and so on.117 It is clear that, since proteinsare made at two separate sites, nuclear-coded proteins mustbe imported into mitochondria and chloroplasts. This isnot made easy by the fact that imported proteins have tocross subcompartments to get into both organelles as theorganelles possess double membranes: twosubcompartments in the case of mitochondria, three forchloropasts because of the thylakoid membrane.

Here is where chaperones are needed to bind thepolypeptide chains just as they emerge through specialpores into the mitochondrial matrix. Assistance withprotein folding is given by yet other chaperones near athand.118 A similar process operates in the importing ofproteins into the chloroplast. As plant cells have bothchloroplasts and mitochondria, two different kinds of signalpeptides are also required to send proteins to the correctaddresses.119 The very complicated transport arrangementsdescribed force us to query how they arose and whatselective advantages there could be for originalendosymbionts to share genomes with the nucleus of thehost cell. As if this is not difficult enough, a further logicaland logistical problem is created by the fact that all of thehost cell’s fatty acids and a number of amino acids aremade by enzymes in the chloroplast stroma. We have nowa transfer in reverse.120

THE MOST ANCIENT CELL

We are running ahead somewhat becauseendosymbiosis could only take place when cells with well-developed metabolism were in existence. These were thethree prokaryotic lines — the Archaebacteria, theEubacteria and those nuclei-bearing prokaryotes destinedto initiate the eukaryotic line by acquiring organelles.121,122

Antedating these three in time was their hypotheticaluniversal ancestor, at the very root of the phylogenetictree — an anaerobic prokaryote shrouded in mystery,barely surviving on the simplest molecules diffusing infrom the surroundings. How simple was its metabolism?A recent textbook suggests that it must be glycolysis.

‘If metabolic pathways evolved by the sequentialaddition of new enzymatic reactions to existing ones,the most ancient reactions should, like the oldest ringsin a tree trunk, be closest to the center of the “metabolictree”, where the most fundamental of the basicmolecular building blocks are synthesized. Thisposition in metabolism is firmly occupied by thechemical processes that involve sugar phosphates,among which the most central of all is probably thesequence of reactions known as glycolysis, by whichglucose can be degraded in the absence of oxygen (thatis, anaerobically). The oldest metabolic pathwayswould have had to be anaerobic because there was nofree oxygen in the atmosphere of the primitive earth.’123

It is extremely unlikely that the earliest cell was sucha heterotroph ‘feeding’ on organic compounds such asacids and sugars. Many strictly anaerobic bacteria todaybreak down glucose through the Entner-Doudoroffpathway. This pathway comprises more than six enzymesacting in sequence and is therefore rather advanced forthe rudimentary first cell.

If the specific qualities of the ancestor are to reflectthe geothermal environment it occupied it should be athermophilic autotroph, that is, a heat-tolerant cellsubsisting on the simplest compounds. It happens that theArchaebacteria of today inhabit environments of extremeheat or salinity or acidity. They can utilise (fix) CO2,although not by the Calvin cycle, as in most photosyntheticorganisms. Indeed, current belief is that the closest to aprototype of the earliest cell are those Archaebacteria thatare completely anaerobic, with inorganic electronacceptors, and which use H2 and CO2 as sole reductantand carbon source, respectively.124 These cells calledchemolithotrophs are (were) able to extract energy andsynthesise their cellular constituents from simple moleculessuch as SO4

2-, S2, H2 and CO2. For most anaerobicArchaebacteria, CO2 can be used as the sole carbon sourcefor growth, and acetyl-CoA is the central biosyntheticintermediate or ‘building block’ for other molecules. Theformation of acetyl-CoA requires two molecules of CO2,a nickel enzyme complex and other cofactors.Furthermore, pyruvate obtained from the breakdown of


CEN Tech. J., vol. 10, no. 3, 1996 311


glucose is converted to acetyl-CoA by a thiamine-pyrophosphate (TPP) enzyme called pyruvateoxidoreductase.125

The recruitment of coenzymes such as TPP so early inevolution is puzzling. Recently, Keefe and his colleaguesattempted the successful synthesis of pantetheine, aprecursor to coenzyme A, presuming the abundance of theprecursor molecules on the primitive Earth. Heatingpantetheine with ATP or ADP failed to produce thedephosphocoenzyme A.126,127 All things considered, achemolithotroph, whether ancient or modern, is anythingbut simple for the kinds of enzymes and metabolicpathways it possesses.

REPRISE

Evolution is biology as a historical science.128

Evolutionists seek to unravel the tangled strands ofhypothetical ancient life forms assumed to have developedover billions of years. In so doing they hope to learn thesecret of that most profound of scientific enigmas, namely,the origin of life.

The driving forces for the enterprise are two: the fossilrecord of cellular structures, and the reasonable inferencethat nucleotide and protein molecular changes over timeshould enable their ancestral lineages to be traced.

Of the first, there is the hard evidence for the presenceof Precambrian stromatolites. This indicates that cellsidentical to modern cyanobacteria were thriving at3.5 Ga.129-132 This and the discovery of the algal fossilGrypania133 support the most ancient dates for the originof fully-developed cells and have skewed the currentopinion on the oxygen content of the primitive atmospheretowards higher values.134 Strong support also comes fromthe studies of Schidlowski on the fractionation of the carbonisotopes in the waxy carbon polymers (kerogens) ofArchaean sediments. In photosynthesis, somewhat moreof the lighter 12CO2 is fixed in slight preference to theheavier 13CO2. Enrichment of 12C with respect to 13C inkerogens extracted from 3.8 Ga rocks is evidence thatphotosynthetic life must have been around for almost4 Ga.135

The time available for the origin of the cell has shrunkto one-tenth or less than has been assumed.136,137 Therenow seems to be little or no time for the genesis of theanaerobic first cell — the progenote of the RNA world.138

Turning now to rooting the phylogenetic tree of life,investigators in the field have voiced concern over attemptsto do this and plead for greater understanding ofphylogenetic methods. Only recently, Hillis andHuelsenbeck caution that

‘current phylogenetic implementations of maximumlikelihood are limited to relatively simple and thereforeunrealistic models of evolution.’139

At the same time workers in Canada and Switzerland havecommented on uncertainties of trying to work outphylogenies using both parsimony and maximum-likelihood methods.140,141

The current belief that life’s ancestral lineage is throughthe Archaebacteria also faces major unsolved problemswith rooting the tree, as witness the following opinions:

‘However, using protein phylogeny to root the tree oflife is not safe; besides the possibility of lateral genetransfer, one cannot be sure that proteins compared inan individual tree descend from a single gene in thecommon ancestor, or from already duplicated genes.’142

Doolittle laments the fact that there is‘still profound disagreement among different kinds ofbiologists about what a phylogenetic taxonomy is.’143

In conclusion, molecular biology in recent years has

KEY POINTS

• How deoxyribonucleic acid (DNA) sequence integritycould have been maintained in the absence of themany enzymes which continually scan and replacemissing, incorrect and damaged nucleotides has notbeen satisfactorily explained.

• The amount of DNA in species does not correlateconsistently with organism complexity.

• Exon shuffling creates problems in molecularphylogeny.

• The numerous components involved in RNA splicingmust have all appeared simultaneously to beadvantageous because a partially completemechanism would function detrimentally.

• Introns introduced into a prokaryotic cell’s geneswould have no opportunity to be removed beforeprotein is made, resulting in ‘nonsense’ non-functional proteins.

• The weaving of information coding for onepolypeptide into an existing nucleotide sequencecoding for another imposes severe evolutionaryconstraints.

• The universality of the genetic code — a strongargument that all organisms are derived from asingle ancestor — in fact has many exceptions.

• Intron sequences correlate over remarkable rangesof thousands of base pairs, strongly suggesting theyare functional.

• It has not been explained how proteins could havemanaged to fold correctly in the absence ofchaperones — themselves complex proteins.

• In hypotheses involving the incorporation of aprokaryote to account for organelles such asmitochondria, it is not clear how a stable relationshipbetween anaerobic invaders and an aerobic oraerotolerant host was possible or why some genesand not others should be transferred to the host’snucleus.

• Current attempts to root the phylogenetic tree of lifeare based on relatively simple and thereforeunrealistic models of evolution.

• Accidental assembly of a self-replicating moleculenow has so many qualifications that its scientificintegrity is questionable.

312 CEN Tech. J., vol. 10, no. 3, 1996


revealed previously unimagined levels of sophistication inthe details of subcellular organisation and function.144-149

The available evidence from the field and the laboratoryis not amicable to the theory that life began with theaccidental assembly of a self-replicating molecule. It isnow accepted with so many qualifications that its scientificintegrity, even as a heuristic device, is questionable.

REFERENCES

1. Jones, J. H. and Drake, M. J., 1986. Geochemical constraints on core formationin the Earth. Nature, 322:221–228.

2. Towe, K. M., 1990. Aerobic respiration in the Archaean? Nature, 348:54–56.3. Bowring, S. A. and Housh, T., 1995. The Earth’s early evolution. Science,

269:1535–1540.4. Widdel, F., 1993. Ferrous iron oxidation of anoxygenic phototrophic bacteria.

Nature, 362:834–836.5. Han, T.-M. and Runnegar, B., 1992. Megascopic eukaryotic algae from the 2.1-

billion-year-old Negaunee Iron-Formation, Michigan. Science, 257:232–235.6. Lowe, D. R., 1980. Stromatolites 3,400-Myr old from the Archaean of Western

Australia. Nature, 284:441–443.7. Walter, M. R., Buick, R. and Dunlop, J. S. R., 1980. Stromatolites 3,400-3,500

Myr old from the North Pole area, Western Australia. Nature, 284:443-445.8. Orpen, J. L. and Wilson, J. F., 1981. Stromatolites at 3,500 Myr and a greenstone-

granite unconformity in the Zimbabwean Archaean. Nature, 291:218–220.9. Byerly, G. R., Lowe, D. R. and Walsh, M. M., 1986. Stromatolites from the

3,300-3,500 Myr Swaziland Group, Barberton Mountain Land, South Africa.Nature, 319:489–491.

10. Riding, R., 1992. The algal breath of life. Nature, 359:13–14.11. Voet, D. and Voet, J. G., 1995. Biochemistry, John Wiley and Sons, Inc., New

York, p. 21.12. Darnell, J., Lodish, H. and Baltimore, D., 1990. Molecular Cell Biology, second

edition, Scientific American Books, distributed by W. H. Freeman and Co., NewYork, pp. 1049–1071.

13. Lodish, H., Baltimore, D., Berk, A., Zipursky, S. L., Matsudaira, P. and Darnell,J., 1995. Molecular Cell Biology, third edition, Scientific American Books,distributed by W. H. Freeman and Co., New York, p. 9.

14. Stryer, L., 1995. Biochemistry, fourth edition, W. H. Freeman and Co., NewYork.

15. Orgel, L. E., 1994. The origin of life on the Earth. In: Life in the Universe,Scientific American, 271(4):53–61.

16. Orgel, Ref. 15, p. 56.17. Sagan, C., 1994. The search for extraterrestrial life. In: Life in the Universe,

Scientific American, 271(4):71–77.18. Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K. and Watson, J. D., 1994.

Molecular Biology of the Cell, third edition, Garland Publishing Inc., NewYork and London, p. 4.

19. Mason, S., 1985. Origin of biomolecular chirality. Nature, 314:400–401.20. Miller, S. L. and Bada, J. L., 1988. Submarine hot springs and the origin of life.

Nature, 334:609–611.21. Lowe, Ref. 6.22. Walter et al., Ref. 7.23. Orpen and Wilson, Ref. 8.24. Byerly et al., Ref. 9.25. Han and Runnegar, Ref. 5.26. Riding, Ref. 10.27. Widdel, Ref. 4.28. Raup, D. M. and Valentine, J. W., 1983. Multiple origins of life. Proceedings of

the National Academy of Science USA, 80:2981–2984.29. Compton, W. and Pidgeon, R. T., 1986. Jack Hills, evidence of more very old

detrital zircons in Western Australia. Nature, 321:766–769.30. Schidlowski, M., 1988. A 3,800-million-year isotopic record of life from carbon

in sedimentary rocks. Nature, 333:313–318.31. Hoyle, F., 1983. The Intelligent Universe, Michael Joseph, London, p. 16.32. Sagan, Ref. 17, p. 75.33. Sharp, P. A., 1994. Ribozymes. Current Opinions in Structural Biology,

4:322–330.

34. Lewin, R., 1986. RNA catalysis gives fresh perspective on the origin of life.Science, 231:545–546.

35. Cech, T. R., 1990. Self-splicing of group I introns. Annual Review ofBiochemistry, 59:543–568.

36. Steitz, J. A., 1988. ‘Snurps’. Scientific American, 258(6):56–63.37. Joyce, G. F., 1989. RNA evolution and the origins of life. Nature, 338:217–

224.38. Robertson, M. P. and Miller, S. L., 1995. An efficient prebiotic synthesis of cytosine

and uracil. Nature, 375:772–774.39. Joyce, Ref. 37, p. 221.40. Lohrmann, R. and Orgel, L. E., 1971. Urea-inorganic phosphate mixtures as

prebiotic phosphorylating agents. Science, 171:490–494.41. Orgel, L. E., 1992. Molecular replication. Nature, 358:203–209.42. Ferris, J. P., Hill Jr., A. R., Liu, R. and Orgel, L. E., 1996. Synthesis of long

prebiotic oligomers on mineral surfaces. Nature, 381:59–61.43. Joyce, Ref. 37, pp. 222, 223.44. Orgel, Ref. 15, p. 61.45. Nielson, P. E., Egholm, M., Berg, R. H. et al., 1991. Sequence-selective recognition

of DNA by strand displacement with a thymine-substituted polyamide. Science,254:1497–1500.

46. Egholm, M., Buchardt, O., Christensen, L. et al., 1993. PNA hybridizes tocomplementary oligonucleotides obeying the Watson-Crick hydrogen bondingrules. Nature, 365:566–568.

47. Böhler, C., Nielsen, P. E. and Orgel, L. E., 1995. Template switching betweenPNA and RNA oligonucleotides. Nature, 376:578–581.

48. Joyce, Ref. 37.49. Nordlund, P. and Eklund, H., 1993. Structure and function of the Escherichia

coli ribonucleotide reductase protein R2. Journal of Molecular Biology,232:123–164.

50. Sjöberg, B. M., 1994. The ribonucleotide reductase jigsaw puzzle: a large piecefalls into place. Structure, 2:793–796.

51. Mao, S.S., Holler, T. P., Yu, G. X. et al., 1992. A model for the role of multiplecysteine residues involved in ribonucleotide reduction: amazing and still confusing.Biochemistry, 31:9733–9743.

52. Reichard, P., 1993. From RNA to DNA, why so many ribonucleotide reductases?Science, 260:1773–1777.

53. Lindahl, T. and Karlström, O., 1973. Heat-induced depyrimidation ofdeoxyribonucleic acid in neutral solution. Biochemistry, 12:5151–5154.

54. Lindahl, T., 1993. Instability and decay of the primary structure of DNA. Nature,362:709–715.

55. Koshland, D. E., Jr., 1994. Molecule of the year: the DNA repair enzyme. Science,266:1925–1926.

56. Fischman, J., 1995. Have 25-million-year-old bacteria returned to life? Science,268:977.

57. Lodish et al., Ref. 13, p. 312.58. Dreyfuss, G., Matunis, K. J., Piñol-Roma, S. and Burd, C. G., 1993. hnRNP

proteins and the biogenesis of mRNA. Annual Review of Biochemistry, 62:289–321.

59. Gilbert, W., 1985. Genes-in-pieces revisited. Science, 228:823–824.60. Cech, T. R., 1986. RNA as an enzyme. Scientific American, 255:64–75.61. Benne, R., 1996. RNA editing: how a message is changed. Current

Opinion in Genetics and Development, 6:221–231.62. Patterson, C., 1987. Molecules and Morphology in Evolution: Conflict or

Compromise?, Cambridge University Press, Cambridge, UK.63. Doolittle, R. F., 1986. Of Urfs and Orfs, A primer on how to analyze derived

amino acid sequences, University Science Books, California, p. 38.64. Green, M. R., 1991. Biochemical mechanisms of constitutive and regulated pre-

mRNA splicing. Annual Review of Cell Biology, 7:559–599.65. Moore, M. J. and Sharp, P. A., 1993. Evidence for two active sites in spliceosome

provided by stereochemistry of pre-mRNA splicing. Nature, 365:364–368.66. McKeown, M., 1993. The role of small nuclear RNAs in RNA splicing. Current

Opinions in Cell Biology, 54:448–454.67. Nilsen, T. W., 1994. RNA-RNA interactions in the spliceosome: unravelling the

ties that bind. Cell, 78:1–4.68. Horowitz, D. S. and Krainer, A. R., 1994. Mechanisms for selecting 5' splice

sites in mammalian pre-mRNA splicing. Trends in Genetics, 10:100–105.69. Sharp, P. A., 1994. Nobel lecture: split genes and RNA splicing. Cell, 77:805–

815.70. Green, M. R., 1989. Pre-mRNA processing and mRNA nuclear export.

Current Opinions in Cell Biology, 1:519–525.


CEN Tech. J., vol. 10, no. 3, 1996 313


102. Weissman, J. S., 1995. All roads lead to Rome? The multiple pathways ofprotein folding. Chemistry and Biology, 2:255–260.

103. Ellis, R. J. and Hemmingsen, S. M., 1989. Molecular chaperones: proteinsessential for the biogenesis of some macromolecular structures. Trends inBiochemical Science, 14:339–342.

104. Ellis, R. J., 1990. Molecular chaperones: the plant connection. Science,250:954–959.

105. Matthews, C. R., 1993. Pathways of protein folding. Annual Review ofBiochemistry, 62:653–684.

106. Rassow, J. and Pfanner, N., 1996. Protein biogenesis: chaperones for nascentpolypeptides. Current Biology, 6:115–118.

107. Hartl, F-U., Hlodan, R. and Langer, T., 1994. Molecular chaperones in proteinfolding: the art of avoiding sticky situations. Trends in Biochemical Science,19:20–25.

108. Kimura, Y., Yahara, I. and Lindquist, S., 1995. Role of the protein chaperoneYDJ1 in establishing Hsp90-mediated signal transduction pathways. Science,268:1362–1365.

109. Smith, D. F. and Toft, D. D., 1993. Steroid receptors and their associated proteins(review). Molecular Endocrinology, 7:4–11.

110. Lodish et al., Ref. 13, pp. 74–75.111. Martin, J., Langer, T., Boteva, R., Schramel, A., Horwich, A. L. and Hartl, F-U.,

1991. Chaperonin-mediated protein folding at the surface of GroEL througha ‘molten-globule’-like intermediate. Nature, 352:36–42.

112. Brag, K., Hainfield, J., Simon, M., Furuya, F. and Harowich, A. L., 1993.Polypeptide bound to the chaperonin GroEL within a central cavity.Proceedings of the National Academy of Science, USA, 90:3978–3982.

113. Gray, T. E. and Fersht, A. R., 1991. Co-operativity in ATP hydrolysis by GroELincreased by GroES. FEBS Letters, 292:254–258.

114. Margulis, L., 1981. Symbiosis in Cell Evolution, W. H. Freeman and Co.115. Perna, N. T. and Kocher, T. D., 1996. Mitochondiral DNA molecular fossils in

the nucleus. Current Biology, 6:128–129.116. Knoll, A. H., 1992. The early evolution of eukaryotes: a geological perspective.

Science, 256:622–627.117. Alberts et al., Ref. 18, p. 716.118. Pfanner, N. and Meijer, M., 1995. Pulling in the proteins. Current Biology,

5:132–135.119. Watson, M. D. and Murphy, D. J., 1995. Genome organisation, protein synthesis

and processing in plants. In: Plant Biochemistry and Molecular Biology,J. L. Peter and R. C. Leegood (eds), John Wiley and Sons, pp. 197–219.

120. Alberts et al., Ref. 18, p. 696.121. Woese, C. R., 1987. Bacterial evolution. Microbiology Reviews, 51:221–

271.122. Woese, C. R., Kandler, O. and Wheelis, M. L., 1990. Towards a natural system

of organisms: proposal for the domains Archaea, Bacteria and Eucarya.Proceedings of the National Academy of Sciences, USA, 87:4576–4579.

123. Alberts et al., Ref. 18, pp. 13, 14.124. Fuchs, G., Ecker, A. and Strauss, G., 1992. Bioenergetics and autotrophic

carbon metabolism of chemolithotropic archaebacteria. In: TheArchaebacteria: Biochemistry and Biotechnology, M. J. Danson, D. W.Hough and G. G. Lunt (eds), Biochemical Society Symposium No. 58, PortlandPress, London and Chapel Hill, pp. 23–39.

125. Danson, M. J. and Hough, D. W., 1992. The enzymology of archaebacterialpathways of central metabolism. In: The Archaebacteria: Biochemistryand Biotechnology, M. J. Danson, D. W. Hough and G. G. Lunt (eds),Biochemical Society Symposium No. 58, Portland Press, London and ChapelHill, pp. 7–21.

126. Keefe, A. D., Newton, G. L. and Miller, S. L., 1995. A possible prebioticsynthesis of pantetheine, a precursor to coenzyme A. Nature, 373:683–685.

127. Ferris, J. P., 1995. Life at the margins. Nature, 373:659.128. Lodish et al., Ref. 13, p. 4.129. Lowe, Ref. 6.130. Walter et al., Ref. 7.131. Orpen and Wilson, Ref. 8.132. Byerly et al., Ref. 9.133. Han and Runnegar, Ref. 5.134. Riding, Ref. 10.135. Schidlowski, Ref. 30.136. Quastler, H., 1964. The Emergence of Biological Organization, Yale

University Press, New Haven, London.137. Shklovskii, I. S. and Sagan, C., 1966. Intelligent Life in the Universe, Dell

71. Spector, D. L., 1993. Nuclear organization of pre-mRNA processing. CurrentOpinions in Cell Biology, 5:442–447.

72. Jimenez-Garcia, L. F. and Spector, D. L., 1993. In vivo evidence thattranscription and splicing are co-ordinated by a recruiting mechanism. Cell,73:47–59.

73. Mattick, J. S., 1994. Introns: evolution and function. Current Opinions inGenetic Developments, 4:823–831.

74. Hurst, L. D., 1994. The uncertain origin of introns. Nature, 371:381–382.75. Yockey, H. P., 1979. Do overlapping genes violate molecular biology and the

theory of evolution? Journal of Theoretical Biology, 80:21–26.76. Moore, R., Dixon, M., Smith, R., Peters, G. and Dickson, C., 1987. Complete

nucleotide sequence of a milk-transmitted mouse mammary tumour virus:two frameshift suppression events are required for translation of gag and pol.Journal of Virology, 61:480–490.

77. Jacks, T., Madhani, H. D., Masiarz, F. R. and Varmus, H. E., 1988. Signals forribosomal frameshifting in the Rous sarcoma virus gag-pol region. Cell,55:447–458.

78. Tiollais, P., Purcell, D. and Dejean, A., 1985. The hepatitis B virus. Nature,317:489–495.

79. Wilson, W., Malim, M. H., Mellor, J., Kingsman, A. J. and Kingsman, S. M.,1986. Expression strategies of the yeast retrotransposon Ty: a short sequencedirects ribosomal frameshifting. Nucleic Acid Research, 14:7001–7016.

80. Craigen, W. J. and Caskey, C. T., 1986. Expression of peptide chain releasefactor 2 requires high-efficiency frameshift. Nature, 322:273–275.

81. Hatfield, D. and Oroszlan, S., 1990. The where, what and how of ribosomalframeshifting in retroviral protein synthesis. Trends in Biochemical Science,15:186–190.

82. Weiss, R. B., 1991. Ribosomal frameshifting, jumping and read through.Current Opinions in Cell Biology, 3:1051–1055.

83. Gesteland, R. F., Weiss, R. B. and Atkins, J. F., 1992. Recoding: reprogrammedgenetic coding. Science, 257:1640–1641.

84. Osawa, S., Jukes, T. H., Watanabe, K. and Muto, A., 1992. Recent evidence forthe evolution of the genetic code. Microbiology Review, 56:229–264.

85. Mattick, Ref. 73, p. 827.86. Nowak, R., 1994. Mining treasures from ‘junk DNA’. Science, 263:608–610.87. Koop, B. F. and Hood, L., 1994. Striking sequence similarity over almost 100

kilobases of human and mouse T-cell receptor DNA. Nature Genetics, 7:48–53.

88. Lee, R. C., Feinbaum, R. L. and Ambros, V., 1993. The C. elegans heterochronicgene lin-4 encodes small RNAs with antisense complementarity to lin-14. Cell,75:843–854.

89. Wightman, B., Ha, I. and Ruvkin, G., 1993. Post-transcriptional regulation ofthe heterochronic gene lin-14 by lin-4 mediates temporal formation in C.elegans. Cell, 75:855–862.

90. Peng, C-K., Buldyrev, S. V., Goldberger, A. L., Havlin, S., Sciortino, F., Simons,M. and Stanley, H. E., 1992. Long-range correlations in nucleotide sequences.Nature, 356:168–170.

91. Mantegna, R. N., Buldyrev, S. V., Goldberger, A. L., Havlin, S., Peng, C-K.,Simons, M. and Stanley, H. E., 1994. Physical Review Letters, 73:3169–3172.

92. Warren, S. T., 1996. The expanding world of trinucleotide repeats. Science,271:1374–1375.

93. Maddox, J., 1994. The genetic code by numbers. Nature, 367:111.94. Fisun, O. I. and Savin, A. V., 1992. Homochirality and long-range transfer in

biological systems. BioSystems, 27:129–135.95. Avetisov, V. A. and Goldanskii, V. I., 1993. Chirality and the equation of the

‘biological big bang’. Physical Letters, A172:407–410.96. Schuster, P., 1991. Molecular evolution as a complex optimization problem.

In: Self-Organization, Emerging Properties, and Learning, A. Babloyantz(ed.), NATO ASI Series B260, Plenum, New York, pp. 241–254.

97. Eigen, M., 1992. Steps Towards Life: A Perspective on Evolution (with R.Winckler-Oswatitch), Oxford University Press, Oxford.

98. Fox, S. W., 1986. Molecular selection and natural selection. Quarterly Reviewsin Biology, 61:375–386.

99. Fleischaker, G. R., 1990. Origins of life: an operational definition. Origins ofLife, Evolution and the Biosphere, 20:127–137.

100. Cohen, J., 1995. Getting all turned around over the origins of life on earth.Report on ‘The origin of homochirality in life’, February 15–17, 1995. Science,267:1265–1266.

101. Fisun and Savin, Ref. 94.

314 CEN Tech. J., vol. 10, no. 3, 1996


Publications, New York.138. Woese, Ref. 121.139. Hillis, D. M. and Huelsenbeck, J. P., 1995. Assessing molecular phylogenies.

Science, 267:255–256.140. Schluter, D., 1995. Uncertainty in ancient phylogenies. Nature, 377:108–

109.141. Benner, S. A., Jermann, T. M., Opitz, J. G., Stackhouse, J., Knecht, L. J. and

Gonnet, G. H., 1995. Uncertainty in ancient phylogenies. Nature, 377:109–110.

142. Forterre, P., Charbonnier, F., Marguet, E., Harper, F. and Henckes, G., 1992.Chromosome structure and DNA topology in extremely thermophilicarchaebacteria. In: The Archaebacteria: Biochemistry and Biotechnology,M. J. Danson, D. W. Hough and G. G. Lunts (eds), Biochemical SocietySymposium No. 58, Portland Press, London and Chapel Hill, pp. 99–112.

143. Doolittle, W. F., 1992. What are the archaebacteria and why are they important?In: The Archaebacteria: Biochemistry and Biotechnology, M. J. Danson,D. W. Hough and G. G. Lunt (eds), Biochemical Society Symposium No. 58,Portland Press, London and Chapel Hill, pp. 1–6.

144. Riding, Ref. 10.

145. Darnell et al., Ref. 12.146. Lodish et al., Ref. 13.147. Stryer, Ref. 14.148. Orgel, Ref. 15.149. Watson and Murphy, Ref. 119.

QUOTABLE QUOTE:The Origin of the Universe

‘What is a big deal — the biggest deal of all — is how you getsomething out of nothing.Don’t let the cosmologists try to kid you on this one. They havenot got a clue either — despite the fact that they are doing a prettygood job of convincing themselves and others that this is reallynot a problem. “In the beginning,” they will say, “there wasnothing — no time, space, matter or energy. Then there was aquantum fuctuation from which . . .” Whoa! Stop right there.You see what I mean? First there is nothing, then there is something.And the cosmologists try to bridge the two with a quantum flutter,a tremor of uncertainty that sparks it all off. Then they are awayand before you know it, they have pulled a hundred billion galaxiesout of their quantum hats.’

Darling, David, 1996. On creating something fromnothing. New Scientist, 151(2047):49.

Dr Aw Swee-Eng, M.B., B.S., Ph.D. (London), FRC Path.,MIBiol. (London), was Associate Professor of Biochemistryat the University of Singapore until 1978, and is now headof the Department of Nuclear Medicine and Director ofClinical Research at Singapore General Hospital. Authorof around 30 technical papers in his field of biochemistryand nuclear medicine, he has authored a critique of origin-of-life theories entitled Chemical Evolution — AnExamination of Current Ideas.

533

b1567 Biological Information — New Perspectives b1567_Sec4.2 8 May 2013 2:56 PM

Towards a General Biology: Emergence of Life and Information from the Perspective of Complex

Systems Dynamics1

Bruce H. Weber

Department of Chemistry & Biochemistry, California State University Fullerton and Division of Science & Natural Philosophy, Bennington College

Abstract

I argue that Darwinism is best described as a research tradition in which specific theories of how natural selection acts to produce common descent and evolutionary change are instantiated by spe-cific dynamical assumptions. The current Darwinian research program is the genetical theory of natural selection, or the Modern Evolutionary Synthesis. Presently, however, there is ferment in the Darwinian Research Tradition as new knowledge from molecular and developmental biology, together with the deployment of complex systems dynamics, suggests that an expanded and extended evolutionary synthesis is possible, one that could be particularly robust in explaining the emergence of evolutionary novelties and even of life itself. Critics of Darwinism need to address such theoretical advances and not just respond to earlier versions of the research tradition.

Key words: complex systems dynamics; Darwinian Research Tradition; emergence; expanded/extended evolutionary synthesis; genetical theory of natural selection; Modern Evolutionary Synthesis; origin of life; self-organization

My thesis is that the Darwinian Research Tradition, defined below, is being enriched, extended and expanded by new information and concepts and that a Darwinian evo-lutionary synthesis deploying background assumptions of complex systems dynam-ics can robustly guide further research into biological phenomena and lead to the

1 The Wistar Institute held a conference in 1966 to explore the adequacy of the neo-Darwinian inter-pretation of evolution, the proceedings of which were subsequently published by the Wistar Institute Press as Mathematical Challenges to the Neo-Darwinian Interpretation of Evolution [1]. In addition to mathematical critiques of the version of population genetics upon which the neo-Darwinian Synthesis, or more accurately the Modern Evolutionary Synthesis, was based, there were presenta-tions, particularly by Conrad Waddington, that pointed out that the synthesis had not adequately included developmental biological phenomena and was by implication incomplete. Two of the key figures in the development and deployment of the second phase of the neo-Darwinian synthesis, Richard Lewontin and Ernst Mayr, were participants, defending the Modern Evolutionary Synthesis even as they provided some criticism of the limitations of one version of the neo-Darwinian program that reduced all biological phenomena to population genetics.

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

534 B. H. Weber


development of a theory of general biology. Such a general theory could and should address issues of the emergence of life, topics properly previously screened off in the Darwinian discourse. After reviewing the history of neo-Darwinism and the Modern Evolutionary Synthesis in the Darwinian Research Tradition,2 and making the case for shifting background dynamical assumptions to those of complex systems, I will focus specifically on the current status of “ origin of life” research and how such work may contribute to a theory of general biology. Finally, I will argue that intel-ligent design theory does not provide a suitable scientific alternative in that it does not provide a conceptual framework for empirical and theoretical research on the phenomena of emergent complexity.3 However, criticisms from intelligent design theorists, among others, of on-going efforts to develop a new Darwinian evolution-ary synthesis can help sharpen the deployment of such a research program.

The Modern Evoluti onary Synthesis and the Darwinian Research Traditi on

In Darwinism Evolving and subsequent publications, David Depew and I have argued that there is not a single Darwinism synonymous with evolutionary theory,

This paper had its origins in a 2007 conference in Boston organized by Bruce Gordon under the auspices of the Center for Science and Culture at Discovery Institute, which funded the event, In the style and spirit of the Wistar Conference, it was meant to explore, some forty years later, the robustness of the earlier neo-Darwinian mathematical population-genetics theory of evolution in light of the pro-gress in molecular and developmental biology, as well as in ecology, in the intervening time. A number of the critics of Darwinism present at the conference articulated an alternative explanation of func-tional biological complexity known as ‘intelligent design’ or more succinctly ID. Others present, like myself, while moving beyond the specific program based upon population genetics, defended the more general concept of a Darwinian evolutionary synthesis under a ‘self-organizational’ rubric.2 Since there was a research program known as neo-Darwinism in the late nineteenth century based upon Weismannian inheritance that was taken to preclude any Lamarckian mechanisms of heredity, many historians of biology prefer to use the term ‘Modern Evolutionary Synthesis’ rather than neo-Darwinism, or neo-Darwinian synthesis, to characterize the genetical theory of evolution based upon population genetics (see discussion in [2]). I will use neo-Darwinism to mean the specific program based upon early Mendelian genetics and Modern Evolutionary Synthesis for a more broadly con-ceived synthesis that includes the version based upon population genetics. I will use the term ‘Darwinian Research Tradition’ to refer to an interlinked set of research programs over time that share a commitment to natural selection as a major, though not sole, source of biological adaptation, order, and innovation, even as the concept of natural selection is articulated against different sets of background assumptions about systems dynamics.3 This is not to say that there cannot be a productive research program based upon assumptions of intelligent design, particularly in areas studying cultural artifacts and social and cultural phenomena more generally. Also, I can imagine productive programs so based for studying atemporal aspects of biological phenomena.

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

Towards a General Biology…Complex Systems Dynamics 535


nor is the Modern Evolutionary Synthesis (often called neo-Darwinism, but see footnote 2) a monolithic research program [2–6]. Rather, we see a Darwinian Research Tradition, which has itself changed over time in light of new empirical data and conceptual advances, and which has assimilated new information and resolved entailing theoretical problems through a process of modifying underlying assumptions about the nature of biological systems and the dynamics of their changes over time. For example, we see “Darwin’s Darwinism” as being informed by Newtonian systems dynamics that emphasized differential survival of individ-ual organisms in populations and saw natural selection as analogous to a Newtonian force that acted gradually, instantaneously equilibrating with other forces (such as variation), to produce adaptation. For the two to three decades fol-lowing the rediscovery of Mendelian genetics in 1900 the discrete nature of muta-tions seemed to contradict the notion of small, continuous variation that was assumed by Darwin in his Newtonian conceptual framework. Indeed, many critics saw and/or hoped for the demise of Darwinism.

After all, Darwinism was not the only research tradition that addressed the phenomena of evolutionary biology. There were many evolutionary biologists in the nineteenth and early twentieth centuries who worked within a Lamarckian, a Geoffroyean, or a Spencerian conceptual framework and research program, in which internal factors, developmental processes, or natural laws of complexifi-cation, respectively, were taken as the driving force of evolution rather than natural selection as a Newtonian-type of force. All three of these alternatives seemed to be gaining adherents in the early twentieth century, even when such scientists called themselves Darwinians, which was for some just a label for accepting descent by modification. As Depew and I recount, the great concep-tual advance brought about by Sergei Chetverikov, J.B.S. Haldane, Ronald Fisher, and Sewall Wright that produced the basis of the “genetical theory of evolution.” This move, which formed the basis of the “Modern Evolutionary Synthesis,” involved shifting the underlying concepts of systems and systems dynamics from Newtonian to Boltzmannian. This shift took advantage of statis-tical insights used by Boltzmann in his development of statistical mechanics in which macroscopic, thermodynamic properties of matter and physical processes were re-described in terms of the aggregate behavior of the microscopic atomic and molecular constituents. The analogy of the action of selection on the fre-quencies of genes in populations with statistical mechanics was explicitly for-mulated by Fisher in his seminal The Genetical Theory of Natural Selection [7]. What mattered in this view was that the gradual shifting of the frequencies of a number of genes within an interbreeding population of a species due to the action of adaptive natural selection, by which change the fitness of the overall population was increased.

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

536 B. H. Weber


Not only did this first phase of the Modern Evolutionary Synthesis resolve the apparent conflict of discontinuous Mendelian genetical variation and gradualistic Darwinian natural selection by changing the background systems dynamics, it was attractive since it provided biologists with a mathematical theory of population genetics that could be rigorously tested. Further it placed biology within the broader “statistical revolution” that had already occurred in the physical sciences. Finally, during the 1930s and 1940s it provided the basis for a second phase and a broader synthesis of a number of areas of biology within the rubric of population genetics. The creative work of Theodosius Dobzhansky, Julian Huxley, Ernst Mayr, George Gaylord Simpson, and G. Ledyard Stebbins produced a more gen-eral synthesis of evolutionary biology, based upon population genetics, that incor-porated much of biology including botany, paleontology, systematics and population ecology [8,9]. This version of the Modern Evolutionary Synthesis, as noted above, is sometimes called neo-Darwinism or the Synthetic Theory of Evolution and continues to provide a basis for a robust program of empirical and theoretical biology [10].

Despite any misgivings about the completeness of the Modern Evolutionary Synthesis, its advocates assumed that the action of natural selection on gene frequencies over generational time (“microevolution” see [11]) could account for the phenomena of common descent over geological time (“macroevolution”). But this synthesis was not complete, as Conrad Waddington repeatedly argued, since it bracketed off developmental biological phenomena, which were assumed to be merely the readout of the genes in the conceptual framework of neo-Dar-winism [12–14]. Similarly bracketed off were aspects of ecology, such as energy flow and community interactions that went beyond population ecology [15–20]. Despite expectations that knowledge of the molecular sequence structures of biological macromolecules (DNA, RNA, proteins) would fit neatly into the neo-Darwinian framework, such knowledge has raised interesting puzzles and identi-fied new evolutionary phenomena that need to be either incorporated into an expanded version of the Modern Evolutionary Synthesis or serve as the basis for a new, yet Darwinian, Expanded and Extended Modern Evolutionary Synthesis [2, 21–23]. Paleontologists Stephen Gould and Niles Eldredge have argued that the Synthesis is unfinished and needs a hierarchical expansion with selection acting in different ways at different levels of the biological hierarchy [24–26]. Scott Gilbert has continued Waddington’s efforts to call for taking developmental phenomena seriously in an expanded and extended evolutionary synthesis, espe-cially in light of the advances in “evo-devo” [27–32]. Gilbert sees development as a complementary process working with natural selection, producing variation and novelty, rather than replacing population genetics [28]. Mary Jane West-Eberhard has shown how developmental plasticity can provide variation even

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.



when there are no changes in the genome and how such phenomena impact evo-lutionary theory in ways that are not anticipated in the Modern Evolutionary Synthesis even though they are consistent with a more broadly conceived Darwinism [33,34].

Toward an Expanded Darwinian Synthesis and a General Biology

More innovative approaches to catch evolutionary phenomena in a expanded syn-thesis have relied upon a variety of tools from the still developing sciences of complexity. One example is that of Daniel Brooks and E.O. Wiley who, along with John Collier and Jonathan Smith, have sought to expand the evolutionary synthesis by introducing concepts from information theory and non-equilibrium thermody-namics to robustly account for the appearance of new biological information and pattern as well as natural selection itself via a process of ‘infodynamics’ [35–42; see also 43]. Using non-equilibrium thermodynamics in a more conventional usage Jeffrey Wicken sought to “expand the Darwinian program” not only to account for the emergence of new information in biological systems but to extend a kind of Darwinian approach to the problem of the origin, or more properly the emergence, of life [44]. Stuart Kauffman applied concepts of non-linear dynamics and self-organization to both developmental genetic systems and to the problem of the origin of life, to the latter of which he also brought in non-equilibrium ther-modynamic considerations as well as consideration of the emergence of ‘agency’ [45–47]. I will return to the issue of the origin of life below. With regard to the inclusion of developmental biology into evolutionary theory, Depew and I have argued that the shift to such systems dynamics employing insights from the behav-ior of complex systems can provide the conceptual context within which a synthe-sis both can be effected while staying within in the Darwinian Research Tradition, if not narrowly formulated versions of neo-Darwinism as espoused by Richard Dawkins, for example [48–50]. One attempt to forge such a synthesis is known as Developmental Systems Theory (see contributions in [51] as well as in [52]). It shows a range of commitment from some form of Darwinism (see [53]) all the way to embracing instead an alternative research tradition, such as the Lamarckian [54–57] or the Geoffroyean [58–60]. Jablonka and Lamb argue that since in later editions of On the Origin of Species Darwin’s hypothetical mechanism of inherit-ance had a Lamarckian character their inclusion of epigenetic factors could be considered as a recovery of Darwin’s original vision [56–57]. A recent review of developmental genetics and epigenetics by Robert Reid argues for an evolutionary theory that is in his own terms outside the Darwinian tradition but more at home in a Lamarckian or Geoffroyean one [61].

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

538 B. H. Weber


A current research program, which we might denote as ‘emergentist’ as a con-venient label, has the goal of developing a theory of general biology, that is, a theory of structural and functional complexity and the emergence of novel struc-ture/function as well as new information and phenomena [45–47,62–77]. This is a program very much in its early stages, but one that holds the promise of eventually developing a theory of biological organization that would hold not only for terrene biology but also for possible biological phenomena elsewhere in the universe. Such a general biology would be part of a more general theory of emergence (see contributions to [66]).

Cauti onary Considerati ons and a Perspecti ve on Emergence

When we are evaluating the sufficiency or inadequacy of the Modern Evolutionary Synthesis, or of Darwinism more generally as a research tradition in some new synthesis, or of rival naturalistic research traditions, or of theories such as intelli-gent design that posit sources of order and information outside of natural pro-cesses, it is important that we take care in being explicit about what we are discussing. Some evolutionary thinkers, such as Gould or Corning, see their approaches, for all the new empirical and theoretical content, as closer in concep-tual stance to Darwin’s original Darwinism than to a narrowly construed Modern Evolutionary Synthesis. Others, such as Deacon, Depew, Kauffman, Wicken and myself, see the deployment of the new complex systems dynamics leading to a totally new version of Darwinism, but still a research program within the Darwinian Research Tradition. Critics of Darwinism, such as Stanley Salthe, Eva Jablonka, and Robert Reid, are not rejecting evolutionary phenomena nor are they calling for sources of order outside nature. Rather, they are arguing for a different set of naturalist assumptions and dynamics that they regard to be better suited to guide future research. As a commitment to methodological naturalism does not logically entail a commitment to philosophical materialism, so we should not take any version of Darwinism as being a synonym or a placeholder for philosophical materialism, unless such a move is self avowed or can be demonstrated, as is the case in writers such as Dawkins and Dennett.

In what follows, I am going to examine current research on emergence theory as well as current work on emergence of life. Even though this issue of the origin of life historically lies outside the orbit of the Darwinian Research Tradition, I will take the cue from Wicken, as well as Kauffman, and Terrence Deacon that the processes and phenomena are rightfully the topic of a general biology and can and should be incorporated in any expanded version or new synthesis of Darwinism. I will assess the value of any theoretical approach in terms of its potential

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.



fecundity and robustness in the development of such a new synthesis and theory of general biology and of emergence. This means I am viewing science not as a body of established facts only, but rather as a process of exploring nature and deepening our understanding of natural phenomena.

Emergence of Emergence as Paradigm

The latter part of the twentieth century saw the rise of a new way of understanding nature, employing complex systems dynamics to explore and explain phenomena of self-organization and emergence (for an overview see for example [2,3,45,46,65–68,71–73,78–88]). Self-organization, or more properly systems-organization, in which the interaction of the system and its environment under particular initial and boundary conditions leads to the emergence of novel order and structure, occurs widely in nature as well as under laboratory conditions and can be consid-ered as a natural phenomenon [89,90]. Developing a theory of such emergent organization has as its goal providing natural explanations for such phenomena. This is very much a work still in progress but the insights gained so far provide a conceptual framework for thinking about and guiding research on the problem of the origin of life.

I define emergence as the appearance of novel properties, structures, and/or patterns in a system that are not present in the constituent components or easily predicted (weak form) or explained (strong form) from the laws and processes affecting the constituents of the system. The new level of phenomena and the lower level of constituents have mutual constraints and the arrows of causal expla-nation point in both directions. If we are tracking the process of the appearance of the new phenomena we are speaking of diachronic emergence in which the lower-level causality exceeds that of the upper level, but when the system has settled to a steady state we than have an instance of synchronic emergence in which the constraints fully mutual. In any event, the emergentist view is that the new, upper-level structure/properties/processes/phenomena represent real natural phenomena and not epiphenomena. In reductionism the lower level is the locus of causality and the upper-level properties are regarded as merely epiphenomenal, that is, with-out causality; in holism the upper level has the causality and the lower levels are epiphenomenal.

It is the strong form of emergence that will be of concern here, especially with regard to the emergence of life. In strong emergence, the emergent phenomena are novel in that they have properties not contained in the components, and are irreduc-ible in sense that the emergent phenomena are not identical to their composition. Emergent systems exhibit a kind of holism in that the emergent phenomena cannot

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

540 B. H. Weber


be analyzed into their parts without losing sight of their essential character. Further, in strong emergence the emergent phenomena obey laws that rely, in at least part, on their novel properties, that is, some of the processes and laws them-selves are emergent, even as the process of their emergence itself operates under general natural laws (including for example a putative ‘fourth of thermodynamics’ in addition to other natural laws [45,46]). Finally, in strong emergence the emer-gent phenomena can impose conditions on their constituents that depend on the nature of the identity of the emergent phenomena, that is, such systems can exhibit downward causation.

Following Deacon’s analysis I will further distinguish three types of emergence: first-order or supervenience, second-order or self-organization, and third order or evolution [67]. In supervenience, the higher-order properties of an aggregate are determined by the statistical or stochastic properties of the ensemble. For example, the liquid properties of water are said to supervene on the properties of individual water molecules. Second-order emergence, or self-organization, occurs on a higher hierarchical level than first-order emergence but as in all hierarchical systems the lower level continues to operate. In self-organization the configurations of individ-ual components and the unique interactions in the system exert an organizing effect on the entire ensemble. Initial conditions and outliers can strongly affect the ensem-ble properties. Self-organization occurs in systems open to matter/energy flows that keep the systems away from equilibrium, resulting in macroscopic structures such as convection cells. Second order emergence also includes phenomena associated with nonlinearity and chaos. It is characteristic of all second-order emergent sys-tems that they have a spatially distributed re-entrant causality that allows microstate variation to amplify and influence macrostate development, even as the macro-relationships undermine, constrain and bias micro-relationships. Snowflakes, Benard convection cells, tornados, chemical waves in the Belousov-Zhabotinskii reaction are examples of such second-order emergence. Self-organizing systems that generate and store information that is useful for system stability and survival evolve. Such informational memory produces recursive, self-referential self-organ-ization that exerts a causal, cumulative (over time) influence over the future of the system. Fitness, function, and natural selection itself can be seen as examples of third-order emergence. Third-order emergence biases across iterations or genera-tions, as in biological development or biological evolution, and can be viewed as an autopoiesis of autopoieses. “So life, even in its simplest forms, is third-order emer-gent. That is why its products cannot be fully understood apart from either historical or functional concerns” [67, p 300]. Both second and third order emergence exhibit a diachronic symmetry breaking not seen in first order emergence. Although higher levels in the hierarch are based upon the lower ones they can exhibit properties not seen at the lower levels because of this symmetry breaking.

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.



The formation of Benard convection cells is an example of a self-organizing process in which the macroscopic structure of the convection flow allows for more efficient dissipation of the energy gradient, giving a thermodynamic “reward” for the production of structure. The process of formation of such convection cells involves a type of selection process working with self-organization. Rod Swenson has shown that the initial formation of convection cells produces macroscopic structures of various sizes and shapes, but that the system quickly settles down into a pattern of hexagonal cells of uniform size [91,92]. Thus there is a sorting or selection process working with self-organization. Brian Goodwin saw the shape and size selection as an instance of physical selection for the most stable [80,85]. To this Swenson added selection of the most dissipatively efficient. For complex chemical systems exhibiting self-organization there is additionally selection for the catalytically efficient, in addition to that for thermodynamic efficiency and physical stability. Thus, even before there is biological selection for the reproduc-tively fit, emerging with the emergence of life, there exists in nature interplay of self-organization and selection at the level of physical and chemical phenomena [2–4,45,46,68,69,71,92].

Is the Origin of Life a Darwinian Problem?

Darwin himself carefully avoided the issue of the origin of life since he was con-cerned with explaining how living beings and their lineages changed over time and how novelties could arise through the action of natural selection upon heritable variation. For example, “How a nerve becomes sensitive to light hardly concerns us more than how life itself originated” [93, p187] was consistent with his accept-ing that life was “breathed into a few forms or into one” [93, p490] (Darwin [1859] 1964, 490). This position served to distinguish Darwin’s theory of evolu-tion from Lamarck’s in which “active matter” spontaneously and continuously generated life [see 94–96]. Privately, Darwin was willing to speculate about the origin of life, as he did in a letter to Joseph Hooker in 1871, “But if (and oh what a big if) we could conceive in some warm little pond with all sorts of ammonia and phosphoric salts, light, heat, electricity and etc., present, that a protein com-pound was chemically formed, ready to undergo still more complex changes” (Cambridge University Library Manuscript Collection: DAR 94: 188–89).

Herbert Spencer argued that biological evolution is a part of a general, cosmic process of the universe becoming less homogeneous and more complex in which the origin of life was a specific instance [97]. Josiah Royce reasserted the more narrow claims of Darwinism as distinguished from those of the Spencerians [98]. With the rise of the Modern Evolutionary Synthesis, the demarcation of the

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

542 B. H. Weber


problem of the origin of life from matters Darwinian was reasserted and continues today in mainstream evolutionary discourse [99,100].

However, one of the founders of the Modern Evolutionary Synthesis, J.B.S. Haldane, along with Alexander Oparin and J.D. Bernal (Marxists all), argued that advances in biochemistry and geochemistry meant that serious scientific study of the origin of life is possible, even if not required by the theories of the Darwinian Research Tradition [101–105]. They recognized that from their commitment to philosophical materialism it was necessary that the origin of life be the result of natural processes only. Opponents of Darwinism and also of philosophical mate-rialism similarly argue that the origin of life is conflated with Darwinian theories [106–110]. Indeed, some neo-Darwinian advocates, such as Richard Dawkins, accept this conflation. In order to reduce biological phenomena to “selfish genes” Dawkins assumes that, however improbable, all that was needed for the appear-ance of life was to get a nucleic acid molecule that could replicate itself, although later this “naked replicator” decorated itself over time with proteins, lipids, etc. to produce better “survival machines” [49,50]. Alex Rosenberg attempts to achieve reduction of all biology to molecular genetics by a slightly different move at the origin of life [111]. He argues that natural selection has to be grounded in chemi-cal and physical selection during the process of life’s origin. During the process of life’s origins, I agree; but this attempt at reduction points instead toward an emer-gentist account [112,113,118]. In what follows, I will consider experimental and theoretical approaches to the emergence of life as well as the implications of the dynamics of emergent complexity for our understanding of biological organiza-tion and how it arises.

Current Perspecti ves on the Emergence of Life

Whether a reductionist or emergentist approach is taken to the origin of life, the pos-sible reactions and routes to the organized complexity of living entities is con-strained by the properties of matter and the laws of chemistry and physics [43,113–118]. Not all types of bonding arrangements and compounds are possible [119]. In aqueous environments, for example, phosphate has unique properties that make it essential for life and even for proto-life. Only phosphoanhydrides had the needed mix of thermodynamic instability and kinetic stability to serve as an inter-mediate for capturing and providing energy. One consequence is that polypeptides can be synthesized abiotically from amino acids, polyphosphate (a phosphoanhy-dride) and magnesium cation [120]. Of course, life may be possible using non-aqueous chemistry, and such possibilities should be explored in a theory of general biology. Steven Benner has suggested that what is essential for the emergence of life

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.



is some sort of solvent system, the chemical elements carbon, hydrogen, nitrogen, sulfur, phosphorous, and oxygen, along with thermodynamic disequilibrium and temperatures consistent of chemical bonding [121].

However, for the present it is a sufficient challenge to address what might have happened during the emergence of life on earth. Given that, we can proceed with the understanding that the possibility space of chemical reactions in living systems is not unconstrained, nor random, but rather reflects in part structural, thermody-namic, kinetic, and combinatorial constraints. Overall, the transition to life and the subsequent evolution of living systems involves retention of reduced compounds in the presence of the resulting ever more oxidizing environment [114]. With an on-going influx of energy and matter the complexity of chemical reactions would be expected to increase as well as non-sequence specific macromolecules under pre-biotic conditions [44].

The minimal elements that need to be considered in any account of the emer-gence of life are:

• An energy source (gradient) and a mechanism to capture energy such that the entropy of the ‘system’ decreases even as the entropy of the system + environment increases

• Abiotically produced component molecules (subsequently produced by autocatalytic networks in proto-cells, and later in cellular metabolism

• Autocatalytic sets of catalysts (polypeptides, polynucleotides)• Closure in both the sense of physical closure (an osmotic barrier) that

separates the system from everything else, and chemical or catalytic closure

• Some means of reproduction and variation at the level of autocatalytic sets and thermodynamic cycles

• Templates for replication and for coding for catalysts.

It is an open question as to which of these steps must be prior to others or if some ensemble of factors is needed before the transition to life could occur. In an emergentist approach it would be expected that several steps could arise concur-rently and act synergistically to give rise to more complex structures and phenom-ena, among which would be included natural selection [43,113,122].

Stanley Miller, working in the laboratory of Harold Urey, demonstrated that a number of amino acids could be produced via chemical processes that might have occurred on the primitive earth [123]. Although the atmosphere globally might not have been as reducing as Miller assumed, mainly due to escape of hydrogen gas, there would be local regions that were, such as near volcanoes or deep-ocean hydrothermal vents [124]. Alternative pathways to amino acids are plausible from

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

544 B. H. Weber


carbon dioxide and from hydrogen cyanide [124]. Further, the presence of amino acids in the interior of meteorites indicates that they can be produced elsewhere in the universe by natural processes; indeed, extraterrestrial sources of organic com-pounds might have been up to three orders of magnitude greater than terrestrial ones for the primitive earth [117 p49,125]. Further, similar such putative processes involving electrical discharge and/or solar-driven photochemical reactions involv-ing hydrogen cyanide, formic acid, hydrogen sulfide, and methane have been shown to produce sugars and purine and pyrimidine bases [for reviews see 113,124,126–129]. Chirality in such momoners could arise in a geologically short period of time due to asymmetry in cosmic radiation that was bombarding the earth [130]. Such monomers could polymerize to form polypeptides and proteins under plausible ambient temperatures [129,131]. Alternatively, hydrogen cyanide polymers form spontaneously when hydrogen cyanide is exposed to an electrical discharge; when such polymers react with water they yield polypeptides, and even polynucleotides [132–134]. Yet another alternative for generating such polymers is considered below involving chemiosmotic-type mechanisms.

Theorizing about the abiotic generation of the organic molecules that are the building blocks of living entities has given rise to a “prebiotic soup” model of increasingly complex molecules, driven by energy flows, from which macromol-ecules arise allowing the emergence of directed synthesis of catalysts, from which protocells would eventually be possible, followed by metabolism in true cells [44,135]. Alternative approaches follow a “metabolism first” approach, harkening back to Haldane, Oparin, and Bernal, often invoking the catalytic capacities of clays [136–138]. A third group of approaches assumes the early presence of some sort of encapsulating barrier, a “proto-cell first” model in which chemical pro-cesses occur in high and sequestered concentrations, within which emerge the catalytic polymers and ultimately directed synthesis [77,139,140]. In this scenario the mutual interaction of catalytic macromolecules and the reactions of a proto-metabolism within an osmotic barrier provides the “theatre” within which speci-fied information can emerge.

Regardless of the approach, at some point catalytic polymers would be expected to emerge and open new chemical possibilities. Polypeptides and proteins pro-duced abiotically would initially have a random sequence [44]. But such sequences have a high probability (at least 25%) of assuming a compact, globular tertiary structure and can exhibit some weak catalytic activity [117,141]. Given that many different sequences of amino acids fold up into the same or similar three dimen-sional structure, the number of such possible folds is a relatively rather small number [142]. Further, completely different and unrelated sequences can produce the same active-site geometry and catalytic function, that is they overlap in the map of catalytic task space [143]. Thus a highly specified informational content is

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.



not necessary for a polypeptide to serve as a catalyst. However, when such a speci-fication process became available via nucleic acid templates, there would be an enormous advantage to such specified information, selected on the basis of cata-lytic and thermodynamic efficiency.

The “hard problem” in origin-of-life research is not so much how the mono-mers and even polymers might have arisen by physical and chemical pro-cesses, but rather how it came to be that a digital-type code in nucleic acids came to specify the analogical information in the thousands of proteins that catalyze metabolism and are involved in signally and information processing [43,45–47,69,108,109,113,118,144–148]. It is here that the new sciences of complexity can have their greatest impact.

The Complex Systems View of the Emergence of Life

As Kauffman has analyzed in his simulations, “protein sequence space” can cover what he terms the “catalytic task space” of all possible chemical reactions that can be catalyzed by polypeptides [45]. Thus, even an ensemble of random peptides would be able to provide such coverage. Such an ensemble can be self-sustaining when it can catalyze the formation of more such catalytic polymers in what is called an autocatalytic cycle. When such a set of autocatalytic cycles can produce their components such that they are self-sustaining, a condition termed catalytic closure is said to obtain. Such catalytically closed, autocatalytic cycles can be maintained, grow, and complexify if they also have some mechanism by which they can tap available energy gradients so as to drive the ensemble away from chemical equilibrium [44,46]. In such emergent systems there would be physical selection of clusters of amino acid sequences that are soluble in water and more stable in an aqueous environment since the less stable structures would tend to degrade and less soluble to precipitate. There would also be a chemical selection of those sequences that were more efficient catalysts or which more efficiently contributed to the autocatalytic cycles and/or more efficiently extracted energy from ambient gradients as the ensembles to which they occur would tend to persist longer. Kauffman, who suspects that such an emergence of organization and com-plexity, an emergence of life, would be an expected consequence of natural law, possibly a fourth law of thermodynamics, writes: “We can think of the origin of life as an expected emergent collective property of a modestly complex mixture of catalytic polymers” [45, xvi, emphasis in original]. Such ensembles of catalytic polymers would be expected to show weak inheritance due to the action of physi-cal and chemical selection. Such systems as those modeled by Kauffman currently are being experimentally studied by Reza Ghardiri to document their dynamics as

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

546 B. H. Weber


compared to those shown in computer simulations (Kauffman, personal commu-nication). These experiments could be enhanced through incorporating thermody-namic work cycles in their action to make them more realistic. We are moving from theoretical speculation and computer simulations to experimental testing of approaches based upon complex systems dynamics.

In such autocatalytic ensembles, possibly encapsulated in ensembles of proto-cells (see below), would be catalyzing not only their assembly but could catalyze, if weakly, chemical reactions to produce component monomers as well as the processes by which energy is extracted from the environment. These ensembles could grow and reproduce themselves even in the absence of central templates coding for such catalytic sets. Not only does Kauffman see an innate holism dur-ing the emergence of life, but he concludes that “the routes to life are broader than imagined” [45, p. 330]. Nevertheless, a crucial event during the emergence of life was the appearance of nucleic acids.

Although an “ RNA World” is a popular scenario for the emergence of life, since RNA can both code and serve as a limited catalyst, there are problems with this approach because of the difficulty of abiotically adding purine and pyrimidine bases to ribose phosphate to form nucleosides and nucleotides. However, some speculative proposals still need exploration [109,149,150]. Such a problem could easily be overcome if there were some sort of proto-metabolism catalyzed by an ensemble of polypeptides that covered catalytic task space. This would be particu-larly so if there were an ensemble of proto-cells in which the Kauffman catalytic sets were sequestered.

The cell-first, or proto-cell first, scenarios mentioned above have a potential advantage over the chemistry of dilute solution. David Deamer has shown that amphiphilic molecules, those with a hydrophobic or “water-hating” end and a hydro-philic or “water-loving” end, though not lipids per se, can be extracted from carbo-naceous chondrites (meteors containing carbon compounds) and that these molecules spontaneously form bilayered vesicles [151,152]. Other amphiphilic molecules of terrestrial origin similarly show the spontaneous formation of vesicles [153]; also photochemical routes to lipid molecules have been documented [117]. Further, vesi-cles of generic amphipiles and/or lipids show an autocatalytic self-replication [117,154]. Such a proto-membrane would have provided not only a way of localizing the chemistry in an ensemble of such vesicles or proto-cells, but provide surfaces at which additional chemistry could occur [117]. More importantly, membranes allow for important energy transduction reactions, driven either chemically or photochemi-cally. Such chemiosmotic reactions, as they are called, use proton gradients across, and possibly within, the membrane to energize movement of molecules across the membrane as well as to form phosphoanhydrides — ATP in modern cells — but likely polyphosphate in early proto-cells [115,155–158]. Indeed, such chemiosmotic

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.



mechanisms are probably one of the most ancient of the characteristics of life [159]. When vesicles of amphiphiles derived from a meteorite are supplemented with poly-cyclic hydrocarbons also extracted from meteorites have light shinned upon them they pump protons across the membrane [160]. Thus such vesicles could not only have provided the cradle for life to emerge but also an energy-capture mechanism, which, polyphophates (and later ATP) could power polymerization reactions of amino acids and nucleotides. Alternatively, iron-sulfur membranes could have formed in the ocean of early earth near thermal vents, for which there is geological evidence as well as experimental replication in the laboratory [161]. In either possi-bility, the chemistry within such membranes would facilitate the actions of autocata-lytic polypeptide sets and the reactions needed to generate nucleic acids, as well as the proto-metabolism in which true lipid components for membranes could have been made. What we have here is a scenario in which the elements of a complex system are emerging together and articulating with each other.

In such a case, the role of nucleic acids may have come later rather than sooner. Once both protein and nucleic acid polymers were present, though not yet in a coding relationship, there would be interactions between these types of macro-molecules, possibly initially providing mutual stabilization of these polymers against hydrolysis and such interactions have been proposed as having to potential to lead to specific templating and ultimately the genetic code [44,162]. The cru-cial consequence of such a template coding of nucleic acids for protein sequences would be that the nucleic acids would stabilize the metabolic and autocatalytic cycle information that were more stable and efficient. Pier Luigi Luisi has esti-mated that such a minimal proto-cell with its osmotic barrier, from which true cells could have emerged, would probably have required around fifty to one hun-dred nucleic acid templates, or genes, in order to sustain viability rather than the thousands now present in the simplest bacterial cell. From such an emergence of proto-cells would arise true biological or natural selection of the reproductively fit [43]. With this type of perspective made available through the application of complex systems theory, it is possible to develop experimental plans using com-puter simulations and laboratory experiments to explore how such a process might have occurred. The hard problem is still hard but it is amenable to scientific inquiry.

Drawing upon empirical data and deploying computer models as well as experi-mental studies, emergentists are seeking to develop a theory that encompasses the problems of the origin of life itself, of biological information and of natural selec-tion that is general in its principles, incorporating life as we know it but also life as it might be. Kauffman assumes that the universe is not a closed system and thus is not fully determined by initial and boundary conditions, but rather is open and has a possibility so enormous that fifteen billion years has been sufficient for

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

548 B. H. Weber


exploration of only a small subset of the possible patterns of organization [46]. When a sufficiently complex organization emerges, not only does natural selection arise, but also the autonomous agency exhibited by living entities. He seeks a pos-sible fourth law of thermodynamics that would account for the emergence of life and new organization. Deacon seeks to develop a broader theory of general biol-ogy through expanding our conception of organism [69]. His autaea are the chemical systems that exhibit autonomous self-maintenance, in contrast to all other configurations of matter, and include autocells. Autocells have coherent and integrated organization as well as self-reproduction in that they can reproduce by direct morphological means. Such morphota would include not only autocells, but also bilayer vesicles capable of reproduction or reproduction of autocatalytic sets. The transition to life comes when it is possible to transmit information of repre-sentation via genetic coding, so living things as we know them are also examples of semeota. The criteria Deacon develops for these categories and the specific example he explores can give us insight as to how to frame questions as to whether some entity encountered elsewhere in the universe is living or to delineate the logi-cal requirements for the emergence of life. In Deacon’s view, as in that of Weber and Depew, natural selection emerges as a phenomenon along with the phenom-enon of the emergence of life, which in turn is a specific instance of the interaction of self-organizational principles with each other and with general selectional principles [3,43,67,69,71,113,118,163,164].

Implicati ons of an Emerging Emergence Paradigm

We are in the very early stages of the development of the emergentist research program. If successful and if widely adopted such theories of emergent organiza-tion and general biology may in time become a new paradigm. Even in these early years it is generating new theoretical and experimental approaches that are par-ticularly relevant to the problem of the emergence of life. When a more complete picture of how life might have emerged is available and we see how it fits into a broader theory of general biology, it will be time to assess whether the Darwinian Research Tradition, if not the Modern Evolutionary Synthesis, can encompass such insights, or if some new conceptual synthesis will be required. At this point we can acknowledge that Conrad Waddington’s intuitions were fecund but needed the developments in biochemistry, molecular biology, developmental genetics, computer simulations, and complexity theory to be cashed out.

The complexification of abiotic chemical reactions is driven primarily by non-equilibrium thermodynamics, exploring state space in an ergodic fashion. When the transition occurs to living systems, a much larger state space of combinatorial

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.



possibilities, provided by catalytic (and templating) polymers, is explored by a combination of self-organizing and selecting processes via what Kauffman terms the “adjacent possible” [45,46]. Though thermodynamics provides the driving force for self-organization, it is the kinetic mechanisms that afford the pathways of emergence. With the emergence of life there is a shift to an extreme expression of kinetic control in which thermodynamic requirements play a supporting rather than directing role. Replication is an instance of this kinetic control. From this emerges the teleonomic and semiotic character of living entities.

In the emergentist perspective, organisms are begotten not made, that is they are the result of developmental processes individually and of evolving lineages. In both cases these phenomena are viewed the result of an on-going interplay of selection and self-organization. What organisms, or their constituent parts, are not, are artifacts. Although emergentist and reductionist approaches to biology share a commitment to methodological naturalism, they view organisms differently in this sense of the importance of epigenetic processes. What the reductionist version of the Modern Evolutionary Synthesis and proponents of intelligent design theory share is a view of biological traits and molecules as artifacts, something made by a designer or by the process of random variation and selection. Emergentists argue that natural and artifactual systems should not be conflated; by anchoring the emergence of life and natural selection in natural laws and processes of thermody-namics and kinetics, a conceptual wedge is driven between natural organization and design.

Elsewhere I address my more general philosophical problems with design argu-ments [165–167]. Here I am attempting only to argue that whereas the emerging theory of general biology is generating novel theoretical insights, predictions, and experimental approaches by which we can deepen our understanding of the emer-gence of life, ID theory does not suggest how to proceed theoretically or experi-mentally as to how life originated, other than to place the causes outside of scientific scrutiny. ID seems to me to provide only a negative capability by criticiz-ing proposed naturalistic and emergentist explanations for the origin of life. Good critics are always helpful in the process of scientific research, but any research program worth its salt also has to guide in the generation of new experiments and theories. The latter is being achieved by those, such as Deacon, Deamer, Ghardiri, Kauffman, Luisi, Morowitz, and Wicken among others, seeking to understand the emergence of life, but not yet substantially by those advocating design arguments.4

4 ID advocates would, of course, dispute these assessments, arguing that intelligent causes can reli-ably be distinguished from unintelligent (undirected natural) causes, and that intelligent causation therefore forms a significant part of our understanding of the cause-and-effect structure of the world under uniformitarian assumptions and constraints. As noted above, emergentists would argue against

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

550 B. H. Weber


Through processes of emergence, life itself may be viewed as begotten, not made, from underlying natural laws and a dialectic of self-organization and selection.

What Might We Expect from a Theory of General Biology About the Origin of Life?

We not only have to acknowledge the difficulty of the problem of how life might have emerged here on earth, let alone how it might emerge and instantiate else-where in the universe, but we need to accept that we should not expect a single narrative trajectory for life’s emergence. Not only would the earliest true living beings destroy the traces of earlier transitional forms, but the action of living systems alters in fundamental ways the chemistry of their environments. Thus, we can only hope to elucidate plausible pathways of emergence, tested by simula-tions, experiments, and what geological data is available. This is not unlike the point Keith Miller makes about the paleontological record, in which we do not have all the details but do have some general patterns to explain [168]. Thus, we need to explore all possible routes of chemistry and proto-biochemistry to develop a range of plausible scenarios for life’s emergence on earth and to elimi-nate those that are unlikely, through theoretical analysis, computer simulations, and experimentation.

In complex systems not only is the whole defined by closure conditions (physi-cal and catalytic) but there is redundancy and parallelism. Thus even weakly insipient functional patterns of structure and interaction can persist due to greater stability and/or efficiency. With functionality comes pressure for improved struc-tures/stability/efficiency, through an on-going process of selection and self-organization. Thus in the origin of life, we should not expect one function to be perfected, say replication, before others appear, but that there would be an inherent holism in the process by which cellular life arose [43,45,46,113,118,140,147].

If there is not grandeur in this view of the emergence of life at least there is a reasonable hope for progress, through application of the tools of complex systems dynamics, towards developing a theory of emergence and of general biology.

this conflation of natural and artifactual systems. To be fair to ID advocates, however, a more sub-stantial ID research program seems to be brewing as of late, as evidenced in the research being done through the Evolutionary Informatics Lab (http://www.evoinfo.org) and in the work of Biologic Institute (http://www.biologicinstitute.org) and its journal BIO-Complexity (http://bio-complexity.org/ojs/index.php/main/index). Indeed, this present volume is part of that general trend. The only thing that can be said is that we must wait and see whether these efforts will go anywhere. For a broader discussion of these issues from a variety of perspectives, both supportive of ID and critical, see Gordon and Dembski [169].

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.



References

1. Moorehead PS, Kaplan MM (eds) (1967) The Mathematical challenges to the Neo-

Darwinian interpretation of evolution. Wisar Institute Press, Philadelphia, PA.

2. Depew DJ, Weber BH (1995) Darwinism Evolving: Systems Dynamics and the

Genealogy of Natural Selection. MIT Press, Cambridge, MA.

3. Weber BH, Depew DJ (1996) Natural selection and self-organization: Dynamical

models as clues to a new evolutionary synthesis. Biol and Phil 11: 33–65.

4. Weber BH, Depew DJ (1999a), Does the second law of thermodynamics refute the

neo-Darwinian synthesis? In: Koslowski P (ed) Sociobiology and Bioeconomics:

The Theory of Evolution in Biological and Economic Theory, pp. 50–75. Springer

Verlag, Berlin.

5. Weber BH, Depew DJ (1999b) The modern evolutionary synthesis and complex

systems dynamics: Prospects for a new synthesis. In: Taborsky E (ed), Semiosis,

Evolution, Energy: Towards a Reconceptualization of the Sign, pp. 263–281. Saker

Verlag, Aachen.

6. Weber BH (2007c) Fact, phenomenon and theory in the Darwinian research tradi-

tion. Biological Theory 2: 168–178.

7. Fisher RA (1930) The Genetical Theory of Natural Selection. Oxford University

Press, Oxford.

8. Jepsen GL, Mayr E, Simpson GG (eds) (1949) Genetics, Paleontology, and

Evolution. Princeton University Press, Princeton, NJ.

9. Mayr E, Provine WB (1980) The Evolutionary Synthesis: Perspective on the

Unification of Biology. Harvard University Press, Cambridge, MA.

10. Nowak MA (2006) Evolutionary Dynamics: Exploring the Equations of Life.

Harvard University Press, Cambridge, MA.

11. Mayr E (1988) Toward a New Philosophy of Biology: Observations of an

Evolutionist. Cambridge MA: Harvard University Press.

12. Waddington CH (1940) Organizers and Genes. Cambridge University Press,

Cambridge.

13. Waddington CH (ed) (1968–72) Towards a Theoretical Biology. Aldine, Chicago.

14. Waddington CH (1975) The Evolution of an Evolutionist. Cornell University Press,

Ithaca NY.

15. Lotka AJ (1924) Elements of Physical Biology. Williams and Wilkins, Baltimore, MD.

16. Odam E (1969) The strategy of ecosystem development. Science 164: 262–270.

17. Odam HT (1988) Self-organization, transformity, and information. Science 242:

1132–1139.

18. Ulanowicz RE (1986) Growth and Development: Ecosystems Phenomenology.

Springer Verlag, New York.

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

552 B. H. Weber


19. Ulanowicz RE (1997) Ecology, The Ascendent Perspective. Columbia University

Press, New York.

20. Ulanowicz RE (2007) Emergence, naturally! Zygon 42: 945–960.

21. Nei M (1987) Molecular Evolutionary Genetics. Columbia University Press, New York.

22. Gillespie JH (1991) The Causes of Molecular Evolution. Oxford University Press,

New York.

23. Lewin R (1999) Patterns in Evolution: The New Molecular View. Scientific American

Library, New York.

24. Gould SJ (1982) Darwinism and the expansion of evolutionary theory. Science 216:

380–387.

25. Gould SJ (2002) The Structure of Evolutionary Theory. Harvard University Press,

Cambridge, MA.

26. Eldredge N (1985) Unfinished Synthesis: Biological Hierarchies and Modern

Evolutionary Thought. Oxford University Press, New York.

27. Gilbert SF (2006a) Developmental Biology, 8th edn. Sinauer, Sunderland, MA.

28. Gilbert SF (2006b) The generation of novelty: The province of developmental biology.

Biological Theory 1: 209–212.

29. Gilbert SF, Opitz JM, Raff RA (1996), Resynthesizing evolutionary theory and

developmental biology. Devel Biol 173: 357–372.

30. Carroll SB, Grenier JK, Weatherbee SD (2001), From DNA to Diversity: Molecular

Genetics and the Evolution of Animal Design. Blackwell Science, Malden MA.

31. Carroll SB (2005) Endless Forms Most Beautiful: The New Science of Evo-Devo.

Norton, New York.

32. Forgas G, Newman SA (2005) Biological Physics of the Developing Embryo.

Cambridge University Press, Cambridge.

33. West-Eberhard MJ (2003) Developmental Plasticity and Evolution. Oxford University

Press, New York.

34. West-Eberhard MJ (2007) Dancing with DNA and flirting with the ghost of Lamarck.

Biol Phil 22: 439–451.

35. Brooks DR, Wiley EO (1986) Evolution as Entropy: Toward a Unified Theory of

Biology, 2nd edn. University of Chicago Press, Chicago.

36. Brooks DR (1998) The unified theory and selection processes. In: Van De Vijver G,

Sathe SN, Delpos M (eds) Evolutionary Systems: Biological and Epistemological

Perspectives on Selection and Self-Organization, pp 113–128. Kluwer, Dordrecht.

37. Brooks DR, McClennan, DA (1999) The nature of the organisms and the emergence of

selection processes and biological signals. In: Taborsky E (ed) Semiosis, Evolution,

Energy: Towards a Reconceptualization of the Sign, pp. 185–218. Aachen: Saker Verlag.

38. Collier J (1986) Entropy in evolution. Biol Phil 1: 5–24.

39. Collier J (1998) Information increase in biological systems: How does adaptation fit?

In: Van De Vijver G, Sathe SN, Delpos M (eds) Evolutionary Systems: Biological

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.



and Epistemological Perspectives on Selection and Self-Organization, pp 129–139.

Kluwer, Dordrecht.

40. Collier J (1999) The dynamical basis of information and the origins of semiosis. In:

Taborsky E (ed) Semiosis, Evolution, Energy: Towards a Reconceptualization of the

Sign, pp. 111–138. Saker Verlag, Aachen.

41. Smith JDH (1998) Canonical ensembles, evolution of competing species, and the

arrow of time. In: Van De Vijver G, Sathe SN, Delpos M (eds) Evolutionary Systems:

Biological and Epistemological Perspectives on Selection and Self-Organization,

pp 141–153. Kluwer, Dordrecht.

42. Smith JDH (1999) On the evolution of semiotic capacity. In: Taborsky E (ed) Semiosis,


Verlag, Aachen.

43. Weber BH (1998) Emergence of life and biological selection from the perspective of

complex systems dynamics. In: Van De Vijver G, Sathe SN, Delpos M (eds)

Evolutionary Systems: Biological and Epistemological Perspectives on Selection

and Self-Organization, pp 59–66. Kluwer, Dordrecht.

44. Wicken JS (1987) Evolution, Information and Thermodynamics: Extending the

Darwinian program. Oxford University Press, New York.

45. Kauffman SA (1993) The Origins of Order: Self-Organization and Selection in

Evolution. Oxford University Press, New York.

46. Kauffman SA (2000) Investigations. Oxford University Press, New York.

47. Kauffman SA (2004), Autonomous agents. In: Barrow JD, Davies, PCW, Harper Jr. CL

(eds) Science and Ultimate Reality: Quantum Theory, Cosmology, and Complexity,

pp. 654–666. Cambridge University Press, Cambridge.

48. Weber BH, Depew DJ (2001) Developmental systems, Darwinian evolution, and the

unity of science. In: Oyama S, Griffiths PE, Gray RD (eds) Cycles of Contingency:

Developmental Systems and Evolution. MIT Press, Cambridge, MA.

49. Dawkins R (1976) The Selfish Gene. Oxford University Press, Oxford.

50. Dawkins R (1989) The Selfish Gene. Oxford University Press, Oxford.

51. Oyama S, Griffiths PE, Gray RD (eds) (2001) Cycles of Contingency: Developmental

Systems and Evolution. MIT Press, Cambridge, MA.

52. Weber. BH, Depew DJ (2003) Evolution and Learning: The Baldwin Effect

Reconsidered. MIT Press, Cambridge, MA.

53. Griffiths PE (2007) The Phenomena of Homology. Biol Phil 22: 643–658.

54. Jablonka E, Lamb MJ (1995) Epigenetic Inheritance and Evolution: The Lamarckian

Dimension. Oxford University Press, Oxford.

55. Jablonka E, Lamb MJ (1998) Bridges between development and evolution. Biol Phil

13: 119–145.

56. Jablonka E, Lamb MJ (2005), Evolution in Four Dimensions: Genetic, Epigenetic,

Behavioral ad Symbolic Variation in the History of Life. MIT Press, Cambridge, MA.

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

554 B. H. Weber


57. Jablonka E, Lamb MJ (2007), The expanded evolutionary synthesis — a reply to

Godfrey-Smith, Haig, and West-Eberhard. Biol Phil 22: 453–472.

58. Salthe SN (1993) Development in Evolution: Complexity and Change in Biology.

MIT Press, Cambridge, MA.

59. Salthe SN (1998) The role of natural selection theory in understanding evolutionary

systems. In: Van De Vijver G, Sathe SN, Delpos M (eds) Evolutionary Systems:

Biological and Epistemological Perspectives on Selection and Self-Organization,

pp. 13–20. Kluwer, Dordrecht.

60. Salthe SN (1999) Energy, Development and Semiosis. In: Taborsky E (ed.) Semiosis,


Verlag, Aachen.

61. Reid R (2007) Biological Emergences: Evolution by Natural Experiment. MIT Press,

Cambridge, MA.

62. Kauffman SA (2007) Beyond reductionism: Reinventing the sacred. Zygon 42:

903–914.

63. Kauffman SA, Clayton P (2006) On emergence, agency, and organization. Biol Phil

21: 501- 521.

64. Kauffman SA, Logan RK, Este R, Goebel R, Hobill D, Shmulevich I (2008)

Propagating organization: an enquiry. Biol Phil 23: 27–45.

65. Clayton, P. (2004) Mind and Emergence: From Quantum to Consciousness. Oxford

University Press, Oxford.

66. Clayton, P. and P. Davies (2006) The Re-emergence of Emergence: The Emergentist

Hypothesis from Science to Religion. Oxford University Press, New York.

67. Deacon TW (2003) The hierarchic logic of emergence: Untangling the interdepen-

dence of evolution and self-organization. In: Weber BH, Depew DJ (eds) Evolution

and Learning: The Baldwin Effect Reconsidered, pp. 273–308. MIT Press,

Cambridge, MA.

68. Deacon TW (2006) Emergence: The hole at the wheel’s hub. In: Clayton P, Davies

P (eds) The Re-emergence of Emergence: The Emergentist Hypothesis from Science

to Religion, pp. 111–150. Oxford University Press, New York.

69. Deacon TW (2006) Reciprocal linkage between self-organizing processes is suffi-

cient for self-reproduction and evolvability. Biological Theory 1: 136–149.

70. Sherman J, Deacon TW (2007) Teleology for the perplexed: How matter began to

matter. Zygon 42: 873–901.

71. Weber BH, Deacon TW (2000) Thermodynamic cycles, developmental systems, and

emergence. Cybernetics and Human Knowing 7: 21–43.

72. Johnson S (2001) Emergence: The Connected Lives of Ants, Brains, Cities, and

Software. Scribner, New York.

73. Salthe SN (2004) The spontaneous origin of new levels in a scalar hierarchy. Entropy

6: 327–343.

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.



74. Schneider ED, Sagan D (2005) Into the Cool: Energy Flow Thermodynamics and

Life. University of Chicago Press, Chicago.

75. Corning P (2003) Nature’s Magic: Synergy in Evolution and the Fate of Humankind.

Cambridge University Press, Cambridge.

76. Corning P (2005) Holistic Darwinism: Synergy, Cybernetics, and the Bioeconomics

of Evolution. University of Chicago Press, Chicago.

77. Morowitz H, Smith E (2007) Energy flow and the organization of life. Complexity

13: 51–59.

78. Casti JL (1994) Complexification: Explaining a Paradoxical World through the

Science of Surprise. HarperCollins, New York.

79. Crofts AR (2007) Life, information, entropy, and time. Complexity 13: 14–50.

80. Goodwin, BC (1994) How the Leopard Changed Its Spots: The Evolution of

Complexity. London: Weidenfeld and Nicolson.

81. Juarrero, A. (1999), Dynamics in Action: Intentional Behavior as a Complex System.

MIT Press, Cambridge, MA.

82. Morowitz, H. (2002), The Emergence of Everything: How the World Became

Complex. Oxford University Press, New York.

83. Schneider, ED, Sagan, D (2005), Into the Cool: Energy Flow Thermodynamics and

Life. University of Chicago Press, Chicago.

84. Silberstein M (2002) Reduction, emergence and explanation. In: Machamer P,

Silberstein M (eds). The Blackwell Guide to the Philosophy of Science, pp. 80–107.

Blackwell, Malden, MA.

85. Solé R, Goodwin BC (2000) Signs of Life: How Complexity Pervades Biology.

Basic Books, New York.

86. Taylor MC (2003) The Moment of Complexity: Emerging Network Cultures.

University of Chicago Press, Chicago.

87. Ulanowicz RE (2002) Ecology, a dialog between the quick and the dead. Emergence

4: 34–52.

88. Wheeler TJ (2007) Analysis, modeling, emergence & integration in complex systems:

a modeling and integration framework & system biology. Complexity 13: 60–75.

89. Camazine S, Deneubourg J-L, Franks NR, Sneyd J, Theraulaz G, Bonabeau E (2001)

Self-Organization in Biological Systems. Princeton University Press, Princeton NJ.

90. Milsteli T (2001) The concept of self-organization in cellular architecture. Journal of

Cell Biology 155: 181–185.

91. Swenson R (1989) Emergent attractors and the law of maximum entropy production:

Foundations to a theory of general evolution. Systems Research 6: 187–197.

92. Swenson, R. (1998) Spontaneous order, evolution, and autocatakinetics: The

nomoloical basis for the emergence of meaning. In: Van De Vijver G, Sathe SN,

Delpos M (eds) Evolutionary Systems: Biological and Epistemological Perspectives

on Selection and Self-Organization, pp. 155–180. Kluwer, Dordrecht.

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

556 B. H. Weber


93. Darwin CR ([1859] 1964) On the Origin of Species by Means of Natural Selection,

or the Preservation of Favoured Races in the Struggle for Life. Facsimile of the

English first edition published by John Murray; with introduction by Ernst Mayr,

Harvard University Press, Cambridge, MA.

94. Desmond A (1989) The Politics of Evolution. University of Chicago Press, Chicago.

95. Fry I. (2000) The Emergence of Life on Earth: A Historical and Scientific Overview.

Rutgers University Press, New Brunswick, NJ.

96. Strick JE (2000) Sparks of Life: Darwinism and the Victorian Debates over

Spontaneous Generation. Harvard University Press, Cambridge, MA.

97. Spencer H (1864), The Principles of Biology. Williams and Norgate, London.

98. Royce J (1904) Herbert Spencer: An Estimate and Review. Collier, New York.

99. Miller KR (1999) Finding Darwin’s God: A Scientist’s Search for Common Ground

between God and Evolution. Harper Collins, New York.

100. Scott EC (2004) Evolution vs. Creationism: An Introduction. Greenwood, Westport CT.

101. Haldane JBS ([1929] 1967) The origin of life. Rationalist Animal. Reprinted in: The

Origin of Life, Bernal JD (ed), pp. 242–249. World, Cleveland OH.

102. Oparin AI (1924) Proiskhozhdenie zhizy. Moskovski Rahochii, Moscow. (English

translation reprinted as: The origin of life, in Bernal JD (ed) The Origin of Life,

pp. 199–241. World, Cleveland, OH.

103. Oparin AI (1938) The Origin of Life. Macmillan, London.

104. Bernal JD (1951) The Physical Basis of Life. Routledge and Kegan Paul, London.

105. Bernal JD (ed) (1967) The Origin of Life. World, Cleveland, OH.

106. Behe MJ (1996) Darwin’s Black Box: The Biochemical Challenge to Evolution. Free

Press, New York.

107. Behe MJ (2007) The Edge of Evolution: The Search for the Limits of Darwinism.

Free Press, New York.

108. Meyer SC (2003) DNA and the origin of life, information, specification, and expla-

nation. In: Campbell JH and S.C. Meyer SC (eds) Darwinism, Design, and Public

Education, pp. 223–285. Michigan State University Press, Lansing, MI.

109. Meyer SC (2009) Signature in the Cell: DNA and the Evidence for Intelligent

Design. HarperOne, San Francisco.

110. Bradley WL (2004) Information, entropy, and the origin of life. In: Dembski WA,

Ruse M (eds) Debating Design: From Darwin to DNA, pp. 331–351. Cambridge

University Press, Cambridge.

111. Rosenberg A (2006) Darwinian Reductionism: Or, How to Stop Worrying and Love

Molecular Biology. University of Chicago Press, Chicago.

112. Weber BH (2007) Back to basics. Nature 445: 601.

113. Weber BH (2007) Emergence of life. Zygon 42: 837–856.

114. Williams RJP, Fraústo Da Silva JJR (2003) Evolution was chemically constrained.

J Theor Biol 220: 323–343.

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.



115. Williams RJP, Fraústo Da Silva JJR (2006) The Chemistry of Evolution. Elsevier,

Amsterdam.

116. Lurquin PF (2003) The Origins of Life and the Universe. Columbia University Press,

New York.

117. Luisi PL (2006) The Emergence of Life: From Chemical Origins to Synthetic

Biology. Cambridge University Press, Cambridge.

118. Weber BH (2009) On the emergence of living systems. Biosemiotics 2: 343–359.

119. Deamer DW (2011) First Life: Discovering the Connections between Stars, Cells,

and How Life Began. University of California Press, Berkeley.

120. Yamagata Y, Inomata K (1997) Condensation of glycylclycine to oligoglycines with

trimetaphosphate in aqueous solution II: Catalytic effect of magnesium ion. Origins

of Life and Evolution of the Biosphere 27:339–344.

121. Benner SA, Ricardo A, Carrigan MA (2004) Is there a common chemical model for

life in the universe? Curr Opin Chem Biol 8: 672–689.

122. Shapiro R (2007) A simpler origin of life. Scientific American 296(6): 46–53.

123. Miller SL (1953) A production of amino acids under possible primitive earth condi-

tions. Science 117: 528–529.

124. Bada JL (2004) How life began on earth: A status report. Earth and Planetary Science

Letters 226: 1–15.

125. Ehrenfreund P, Irvine W, Becker L (2002) Astrophysical and astrochemical insights

into the origin of life. Reports of Progress in Physics 65: 1427–1487.

126. Lahav N (1999) Biogenesis: Theories of Life’s Origin. Oxford University Press,

New York.

127. Wills C, Bada J (2000) The Spark of Life: Darwin and the Primeval Soup. Perseus,

Cambridge, MA.

128. Bada JL, Lazcano A (2003) Prebiotic soup — revisiting the Miller experiment.

Science 300: 745–746.

129. Leman L, Orgel L, Ghadiri MR (2004) Carbonylsulfide-mediated prebiotic forma-

tion of peptides. Science 306: 283–286.

130. Kondepudi D (1988) Parity violation and the origin of biomolecular chirality. In:

Weber BH, Depew DJ, Smith JD (eds) Entropy, Information, and Evolution: New

Perspectives on Physical and Biological Evolution, pp. 41–50. MIT Press, Cambridge, MA.

131. Kumar JK, Oliver JS (2002) Proximity effects in monolayer films: Kinetic analysis

of amide bond formation in the air-water interface using proton NMR spectroscopy.

J Am Chem Soc 124: 11307–11314.

132. Matthews CN, Moser RE (1967) Peptide synthesis from hydrogen cyanide and

water. Nature 215: 1230–1234.

133. Liebman SA, Pesce-Rodriguez RA, Matthews CN (1995) Organic analysis of hydro-

gen cyanide polymers: Prebiotic and extratrerrestrial chemistry. Advances in Space

Research 15(3): 71–80.

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

558 B. H. Weber


134. Minard RD, Hatcher PG, Gourley RC, Matthews CN (1998) Structural investigations

of hydrogen cyanide polymers: New insights using TMAH Thermochemolysis/

GC-MS. Origin of Life and Evolution of the Biosphere 28: 461–473.

135. Fox SW (1988) The Emergence of Life. Academic Press, New York.

136. Wächterhäuser G (1988) Before enzymes and templates: Theory of surface metabo-

lism. Microbiological Reviews 52: 452–484.

137. Wächterhäuser G (1992) Groundworks for an evolutionary biochemistry: The iron-

sulfur world. Progress in Biophysics and Moleculr Biology 58: 85–201.

138. Wächterhäuser G (1997) The origin of life and its methodological challenge. J Theor

Biol 187: 483–494.

139. Morowitz H, Deamer DW, Smith T (1991) Biogenesis as an evolutionary process.

J Mol Evol 33: 207–208.

140. Morowitz H (1992) Beginnings of Cellular Life: Metabolism Recapitulates

Biogenesis. Yale University Press, New Haven.

141. Shakhnovich EI, Gutin AM (1990) Implications of Thermodynamics of Protein

Folding for Evolution of Primary Sequences. Nature 346: 773–775.

142. Bowie JU (1990) Deciphering the message in protein sequences: tolerance of amino

acid substitutions. Science 247: 1306–1310.

143. Hazen RM, Griffin PL, Carothers JM, Szostak JW (2007) Functional information

and the emergence of biocomplexity. Proc Natl Acad Sci USA 104: 8574–8531.

144. Axe DD (2010) The case against a Darwinian origin of protein folds. BIO-

Complexity 2010(1):1–12. doi:10.5048/BIO-C.2010.1

145. Fox RF (1997) The origins of life: What one needs to know. Zygon 32: 393–406.

146. Meyer SC, Nelson PA (2011) Can the origin of the genetic code be explained by

direct RNA templating? BIO-Complexity 2011(2): 1–10.

147. Weber BH (2010) What is Life? Defining life in the context of emergent complexity.

Origins of Life and Evolution of Biospheres 40: 221–229.

148. Yockey HP (2005) Information Theory, Evolution, and the Origin of Life. Cambridge

University Press, Cambridge.

149. Joyce GF, Orgel, LE (1993), Prospects for understanding the origin of the RNA

world. In: R.F. Gesteland RF, Atkins JF (eds) The RNA World, pp. 1–25. Cold Spring

Harbor Press, Cold Spring Harbor MA.

150. Schwartz AW (1997) Speculation on the RNA precursor problem. J Mol Biol 187:

523–527.

151. Deamer DW, Pashley RM (1989) Amphiphilic components of the Murchison carbo-

naceous chondrite: Surface properties and membrane formation. Origin of Life and

Evolution of the Biosphere 19: 21–38.

152. Deamer DW, Dworkin JP, Sandford SA, Bernstein MP, Allamandola LJ (2002), The

first cell membranes. Astrobiology 2: 371–381.

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.



153. Ourisson G, Nakatani Y (1994) The terpenoid theory of the origin of cellular life:

The evolution of terpenoids to cholesterol. Curr Biol 1: 11–23.

154. Bachmann PA, Luisi PL, Lang J (1992) Autocatalytic self-replication micelles as

modes for prebiotic structures. Science 357: 57–58.

155. Mitchell PD (1961) Coupling of phosphorylation to electron and hydrogen transfer

by a chemiosmotic type of mechanism. Nature 191: 144–148.

156. Williams RJP (1961) Possible functions of chains of catalysts. J Theoret Biol 1: 1–17.

157. Weber BH, Prebble JN (2006) An issue of originality and priority: The correspon-

dence and theories of oxidative phosphorylation of Peter Mitchell and Robert J. P.

Williams, 1961–1980. Journal of the History of Biology 39: 125–163.

158. Prebble J, Weber BH (2003) Wandering in the Gardens of the Mind: Peter Mitchell

and the Making of Glynn. Oxford University Press, New York.

159. Lane N (2006) Power, Sex, Suicide: Mitochondria and the Meaning of Life. Oxford

University Press, Oxford.

160. Deamer DW (1992) Polycyclic aromatic hydrocarbons: Primitive pigment systems

in the prebiotic environment. Advances in Space Research 12: 1–4.

161. Martin W, Russell MT (2003) On the origins of cells: A hypothesis for the evolution-

ary transitions from abiotic geochemistry to chemoautotrophic prokaryotes, and

from prokaryotes to nucleated cells. Phil Trans Royal Society of London (Series B)

358: 59–85.

162. Carter CW, Kraut J (1974) A proposed model for interaction of polypeptides with

RNA. Proc Natl Acad Sci USA 71: 283–287.

163. Weber BH (2000) Closure in the Emergence and Evolution of Life. NY Acad Sci

901: 132–138.

164. Weber BH (2003a), Emergence of mind and the Baldwin effect. In: B.H. Weber BH,

Depew DJ (eds) Evolution and Learning: The Baldwin Effect Reconsidered,

pp. 309–326.

165. Weber BH (2003b) Biochemical complexity: Emergence of design? In: J.A. Campbell JA,

Meyer SC (eds) Darwinism, Design, and Public Education, pp. 455–462. Michigan

State University Press, East Lansing, MI.

166. Weber BH, Depew DJ (2004) Darwinism, Design, and Complex Systems Dynamics.

In: Dembski WA, Ruse M (eds) The Appearance of Design in Nature, pp. 173–190.

Cambridge University Press, Cambridge UK.

167. Weber BH (2011) Design and its discontents. Synthese 178 (2): 271–289.

168. Miller KB (2003) Common descent, transitional forms, and the fossil record. In:

Miller KB (ed) Perspectives on an Evolving Creation, pp. 152–181. Eerdmans,

Grand Rapids MI.

169. Gordon BL, Dembski WA (eds) (2011) The Nature of Nature: Examining the Role

of Naturalism in Science. ISI Books, Wilmington, DE.

Bio

logi

cal I

nfor

mat

ion

Dow

nloa

ded

from

ww

w.w

orld

scie

ntif

ic.c

omby

69.

170.

92.2

43 o

n 06

/10/

13. F

or p

erso

nal u

se o

nly.

a simpler origin for life - scientific american …researchers use the term "rna world" to...

Documents