the origins of cellular life - newcastle university · the origins of cellular life jason p....

19, 20102010; doi: 10.1101/cshperspect.a002212 originally published online MayCold Spring Harb Perspect Biol

Jason P. Schrum, Ting F. Zhu and Jack W. Szostak The Origins of Cellular Life

Subject Collection The Origins of Life

Constructing Partial Models of Cells

et al.Norikazu Ichihashi, Tomoaki Matsuura, Hiroshi Kita,

The Hadean-Archaean EnvironmentNorman H. Sleep

RibonucleotidesJohn D. Sutherland

An Origin of Life on MarsChristopher P. McKay

Characterize Early Life on EarthHow a Tree Can Help−−Deep Phylogeny

RandallEric A. Gaucher, James T. Kratzer and Ryan N.

Primitive Genetic PolymersAaron E. Engelhart and Nicholas V. Hud

Medium to the Early EarthCosmic Carbon Chemistry: From the Interstellar

Pascale Ehrenfreund and Jan Cami

Membrane Transport in Primitive CellsSheref S. Mansy

Origin and Evolution of the RibosomeGeorge E. Fox

The Origins of Cellular LifeJason P. Schrum, Ting F. Zhu and Jack W. Szostak

BiomoleculesPlanetary Organic Chemistry and the Origins of

Kim, et al.Steven A. Benner, Hyo-Joong Kim, Myung-Jung

From Self-Assembled Vesicles to ProtocellsIrene A. Chen and Peter Walde

the Origins of LifeMineral Surfaces, Geochemical Complexities, and

Robert M. Hazen and Dimitri A. Sverjensky

The Origin of Biological HomochiralityDonna G. Blackmond

Historical Development of Origins ResearchAntonio Lazcano

Earth's Earliest AtmospheresKevin Zahnle, Laura Schaefer and Bruce Fegley

http://cshperspectives.cshlp.org/cgi/collection/ For additional articles in this collection, see

Copyright © 2010 Cold Spring Harbor Laboratory Press; all rights reserved

on April 17, 2014 - Published by Cold Spring Harbor Laboratory Press http://cshperspectives.cshlp.org/Downloaded from on April 17, 2014 - Published by Cold Spring Harbor Laboratory Press http://cshperspectives.cshlp.org/Downloaded from

http://cshperspectives.cshlp.org/cgi/collection/



http://cshperspectives.cshlp.org/cgi/adclick/?ad=42786&adclick=true&url=http%3A%2F%2Fwww.genelink.com

http://cshperspectives.cshlp.org/




The Origins of Cellular Life

Jason P. Schrum, Ting F. Zhu, and Jack W. Szostak

Howard Hughes Medical Institute, Department of Molecular Biology and the Center for Computationaland Integrative Biology, Massachusetts General Hospital, Boston, Massachusetts 02114

Correspondence: [email protected]

Understanding the origin of cellular life on Earth requires the discovery of plausible pathwaysfor the transition fromcomplex prebiotic chemistry to simple biology, defined as the emergenceof chemical assemblies capable of Darwinian evolution. We have proposed that a simpleprimitive cell, or protocell, would consist of two key components: a protocell membranethat defines a spatially localized compartment, and an informational polymer that allows forthe replication and inheritance of functional information. Recent studies of vesicles composedof fatty-acid membranes have shed considerable light on pathways for protocell growth anddivision, as well as means by which protocells could take up nutrients from their environment.Additional work with genetic polymers has provided insight into the potential for chemicalgenome replication and compatibility with membrane encapsulation. The integration ofa dynamic fatty-acid compartment with robust, generalized genetic polymer replicationwouldyield a laboratorymodelof a protocell with the potential forclassical Darwinian biologi-cal evolution, and may help to evaluate potential pathways for the emergence of life on theearly Earth. Here we discuss efforts to devise such an integrated protocell model.

The emergence of the first cells on the earlyEarth was the culmination of a long history

of prior chemical and geophysical processes.Although recognizing the many gaps in ourknowledge of prebiotic chemistry and the earlyplanetary setting in which life emerged, wewill assume for the purpose of this review thatthe requisite chemical building blocks wereavailable, in appropriate environmental set-tings. This assumption allows us to focus onthe various spontaneous and catalyzed assem-bly processes that could have led to theformation of primitive membranes andearly genetic polymers, their coassembly into

membrane-encapsulated nucleic acids, and thechemical and physical processes that allowedfor their replication. We will discuss recentprogress toward the construction of laboratorymodels of a protocell (Fig. 1), evaluate theremaining steps that must be achieved before acomplete protocell model can be constructed,and consider the prospects for the observationof spontaneous Darwinian evolution in labo-ratory protocells. Although such laboratorystudies may not reflect the specific pathwaysthat led to the origin of life on Earth, they areproving to be invaluable in uncovering surprisingand unanticipated physical processes that help

Editors: David Deamer and Jack W. Szostak

Additional Perspectives on The Origins of Life available at www.cshperspectives.org

Copyright # 2010 Cold Spring Harbor Laboratory Press; all rights reserved; doi: 10.1101/cshperspect.a002212

Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212

1

on April 17, 2014 - Published by Cold Spring Harbor Laboratory Press http://cshperspectives.cshlp.org/Downloaded from



us to reconstruct plausible pathways and scenar-ios for the origin of life.

The term protocell has been used looselyto refer to primitive cells or to the first cells.Here we will use the term protocell to refer spe-cifically to cell-like structures that are spatiallydelimited by a growing membrane boundary,and that contain replicating genetic informa-tion. A protocell differs from a true cell in thatthe evolution of genomically encoded advanta-geous functions has not yet occurred. With agenetic material such as RNA (or perhaps oneof many other heteropolymers that could pro-vide both heredity and function) and an appro-priate environment, the continued replicationof a population of protocells will lead inevitablyto the spontaneous emergence of new codedfunctions by the classical mechanism of evolu-tion through variation and natural selection.Once such genomically encoded and thereforeheritable functions have evolved, we wouldconsider the system to be a complete, living bio-logical cell, albeit one much simpler than anymodern cell (Szostak et al. 2001).

BACKGROUND

Membranes as compartment boundaries

All biological cells are membrane-bound com-partments. The cell membrane fulfills the essen-

tial function of creating an internal environmentwithin which genetic materials can reside andmetabolic activities can take place without beinglost to the environment. Modern cell mem-branes are composed of complex mixtures ofamphiphilic molecules such as phospholipids,sterols, and many other lipids as well as diverseproteins that perform transport and enzymaticfunctions. Phospholipid membranes are stableunder a wide range of temperature, pH, andsalt concentration conditions. Such membranesare extremely good permeability barriers, so thatmodern cells have complete control over theuptake of nutrients and the export of wastesthrough the specialized channel, pump andpore proteins embedded in their membranes.A great deal of complex biochemical machineryis also required to mediate the growth and divi-sion of the cell membrane during the cell cycle.The question of how a structurally simple proto-cell could accomplish these essential membranefunctions is a critical aspect of understandingthe origin of cellular life.

Vesicles formed by fatty acids have long beenstudied as models of protocell membranes(Gebicki and Hicks 1973; Hargreaves andDeamer 1978; Walde et al. 1994a). Fatty acidsare attractive as the fundamental buildingblock of prebiotic membranes in that they arechemically simpler than phospholipids. Fatty

Figure 1. A simple protocell model based on a replicating vesicle for compartmentalization, and a replicatinggenome to encode heritable information. A complex environment provides lipids, nucleotides capable ofequilibrating across the membrane bilayer, and sources of energy (left), which leads to subsequent replicationof the genetic material and growth of the protocell (middle), and finally protocellular division throughphysical and chemical processes (right). (Reproduced from Mansy et al. 2008 and reprinted with permissionfrom Nature Publishing #2008.)

J.P. Schrum, T.F. Zhu, and J.W. Szostak

2 Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212




acids with a saturated acyl chain are extremelystable compounds and therefore might haveaccumulated to significant levels, even given arelatively slow or episodic synthesis. Moreover,the condensation of fatty acids with glycerolto yield the corresponding glycerol esters pro-vides a highly stabilizing membrane component(Monnard et al., 2002). Finally, phosphory-lation and the addition of a second acyl chainyields phosphatidic acid, the simplest phospho-lipid, thus providing a conceptually simplepathway for the transition from primitive tomore modern membranes. The prebiotic chem-istry leading to the synthesis of fatty acids andother amphiphilic compounds is treated inmore detail in Mansy (2010).

The best reason for considering fatty acidsas fundamental to the nature of primitive cellmembranes is not, however, their chemicalsimplicity. Rather, fatty-acid molecules in mem-branes have dynamic properties that are essentialfor both membrane growth and permeability.Because fatty acids are single chain amphiphileswith less hydrophobic surface area than phospho-lipids, they assemble into membranes only atmuch higher concentrations. This equilibriumproperty is mirrored in their kinetics: Fatty acidsare not as firmly anchored within the membraneas phospholipids; they enter and leave the mem-brane on a time scale of seconds to minutes(Chen and Szostak 2004). Fatty acids can alsoexchange between the two leaflets of a bilayermembrane on a subsecond time scale. Rapid flip-flop is essential for membrane growth when newamphiphilic molecules are supplied from theenvironment. New molecules enter the mem-brane primarily from the outside leaflet, and flip-flop allows the inner and outer leaflet areas toequilibrate, leading to uniform growth.

Considering that protocells on the earlyEarth did not, by definition, contain any com-plex biological machinery, they must have reliedon the intrinsic permeability properties of theirmembranes. Membranes composed of fattyacids are in fact reasonably permeable to smallpolar molecules and even to charged speciessuch as ions and nucleotides (Mansy et al.2008). This appears to be largely a result of theability of fatty acids to form transient defect

structures and/or transient complexes withcharged solutes, which facilitate transport acrossthe membrane. The subject of the permeabilityof fatty-acid based membranes is dealt with ingreater detail by Mansy (2010).

Prebiotic vesicles were almost certainlycomposed of complex mixtures of amphiphiles.Amphiphilic molecules isolated from meteor-ites (Deamer 1985; Deamer and Pashley 1989)as well as those synthesized under simulatedprebiotic conditions (McCollum et al. 1999;Dworkin et al. 2001; Rushdi and Simoneit2001) are highly heterogeneous, both in termsof acyl chain length and head group chemistry.Membranes composed of mixtures of amphi-philes often have superior properties to thosecomposed of single pure species. For example,mixtures of fatty acids together with the corre-sponding alcohols and/or glycerol esters gene-rate vesicles that are stable over a wider rangeof pH and ionic conditions (Monnard et al.2002), and are more permeable to nutrientmolecules including ions, sugars and nucleo-tides (Chen et al. 2004; Sacerdote and Szostak2005; Mansy et al. 2008). This is in strikingcontrast to the apparent requirement for homo-geneity in the nucleic acids, where even lowlevels of modified nucleotides can be destabiliz-ing or can block replication.

RECENT RESULTS

Pathways for Vesicle Growth

Fatty-acid vesicle growth has been shown tooccur through at least two distinct pathways:growth through the incorporation of fatty acidsfrom added micelles, and growth throughfatty-acid exchange between vesicles. The growthof membrane vesicles from micelles has beenobserved following the addition of micelles orfatty-acid precursors to pre-formed vesicles(Walde et al. 1994a, 1994b; Berclaz et al. 2001).When initially alkaline fatty-acid micelles aremixed with a buffered solution at a lower pH,the micelles become thermodynamically unsta-ble. As a consequence, the fatty-acid moleculescan either be incorporated into pre-existingmembranes, leading to growth (Berclaz et al.

Origins of Cellular Life

Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 3




2001), or can self-assemble into new vesicles(Blochliger et al. 1998; Luisi et al. 2004). Thesepioneering studies were done by cryo-TEM,which does not allow growth to be followed inreal time, and by light scattering, which is diffi-cult to interpret in the case of samples with het-erogeneous size distributions. We thereforeadapted a fluorescence assay based on FRET(Forster resonance energy transfer) to measurechanges in membrane area in real time. Thisassay is based on the distance dependence ofenergy transfer between donor and acceptorfluorescent dyes; thus when a membrane growsin area by incorporating additional fatty-acidmolecules, the dyes are diluted and the efficiencyof FRET decreases. Studies on small (typically100 nm in diameter) unilamellar fatty-acidvesicles using this assay showed that the slowaddition of fatty-acid micelles led to vesiclegrowth with an efficiency of �90% (Hanczycet al. 2003).

The real-time FRET assay allowed for akinetic dissection of the growth process, reveal-ing a surprisingly complex series of events afterthe rapid addition of micelles (Chen and Szo-stak 2004). Two major processes were observed.The first fast phase resulted in membrane areagrowth that was limited to �40% increase inarea, independent of the amount of addedmicelles. A second much slower phase led to afurther increase in membrane area that variedwith the amount of added micelles. We inter-preted the fast phase as reflecting the rapidassembly of a layer of adhering micelles aroundthe pre-formed vesicles, with rapid monomerexchange resulting in the efficient incorporationof this material into the pre-formed membrane.We interpreted the slow phase as the conse-quence of micelle–micelle interactions leadingto the assembly of intermediate structuresthat could partition between two pathways—with some monomers dissociating and contri-buting to membrane growth and the remainderultimately assembling into new membranevesicles. Although these interpretations are con-sistent with our data, the experiments arerather indirect, and further exploration of themechanism of membrane growth is certainlydesirable.

A second, distinct pathway for vesiclegrowth involves fatty-acid exchange betweenvesicles. Under certain conditions this exchangecan lead to growth of a subpopulation of vesiclesat the expense of their surrounding neigh-bors. Within populations of osmotically relaxedvesicles, such exchange processes do not resultin significant changes in size distribution withtime. Similarly, a population of uniformlyosmotically swollen vesicles does not change insize distribution, but such vesicles are in equil-ibrium with a lower solution concentration offatty acids because the tension in the membraneof the swollen vesicles makes it more ener-getically favorable for fatty-acid molecules toreside in membrane. When osmotically swollenvesicles are mixed with osmotically relaxed(isotonic) vesicles, rapid fatty-acid exchangeprocesses result in growth of the swollen vesiclesand corresponding shrinkage of the relaxedvesicles (Chen et al. 2004). Because vesiclescan be osmotically swollen as a result of theencapsulation of high concentrations of nucleicacids such as RNA, this process allows for thegrowth of vesicles containing genetic polymersat the expense of empty vesicles (or vesiclesthat contain less internal nucleic acid). Becausefaster replication would increase the internalnucleic acid concentration, this pathway ofcompetitive vesicle growth provides the poten-tial for a direct physical link between the rateof replication of an encapsulated genetic poly-mer and the rate of growth of the protocell asa whole.

Assuming that the division of osmoticallyswollen vesicles could occur either stochasticallyor at some threshold size, protocells that de-veloped some heritable means of faster replica-tion and growth would have a shorter cell cycle,on average, and would therefore gradually takeover the population. This simple physical mech-anism might therefore lead to the emergenceof Darwinian evolution by competition at thecellular level. However, if replication is limitedby the rapid reannealing of complementarystrands (see later discussion), it may be difficultto reach osmotically significant concentrations.Furthermore, osmotically swollen vesicles areintrinsically difficult to divide owing to the






energetic cost of reducing the volume of aspherical vesicle to that of two daughter vesiclesof the same total surface area. One possibilityis that osmotically driven competitive growthmight alternate with the faster membranegrowth that follows micelle addition. If newfatty-acid material was only available sporadi-cally, rapid membrane growth might follow aninflux of fresh fatty acids, facilitating division(see following discussion).

All of the experiments discussed earlier weredone with small unilamellar vesicles preparedby extrusion through 100 nm pores in filters.In contrast, fatty-acid vesicles that form sponta-neously by rehydrating dry fatty-acid films tendto be several microns in diameter and multila-mellar (Hargreaves and Deamer 1978; Hanczycet al. 2003). Such large multilamellar fatty-acidvesicles are so heterogeneous that quantitativestudies of growth and division are difficult. Wehave recently developed a simple procedurefor the preparation of micron-sized, monodis-perse (homogeneous in size) multilamellarvesicles by large-pore dialysis (Zhu and Szostak2009b). The preparation of large monodispersemultilamellar vesicles has allowed us to directlyobserve an unusual mode of vesicle growth(Zhu and Szostak 2009a). We showed that feed-ing a micron-sized multilamellar fatty-acidvesicle with fatty-acid micelles results in the for-mation of a thin membranous protrusion whichextends from the side of the initially spherical

parental vesicle. Over time, this thin membranetubule elongates and thickens, gradually incor-porating more and more of the parental vesicle,until eventually the entire vesicle is transformedinto a long, hollow threadlike vesicle (Fig. 2).This pathway occurs with vesicles ranging insize from 1 to at least 10 mm in diameter, com-posed of a variety of different fatty acids andrelated amphiphiles. Only multilamellar vesiclesgrow in this manner and only when vesicle vol-ume increases slowly (relative to surface areagrowth) because of a relatively impermeable buf-fer solute. Confocal microscopy has providedinsight into the mechanism of this mode ofgrowth: The outermost membrane layer growsfirst, and because there is little volume betweenit and the next membrane layer, and that volumecannot increase on the same time scale, the extramembrane area is forced into the form of a thintubule. Over time, this tubule grows, and as aresult of poorly understood exchange processes,the entire original vesicle is ultimately trans-formed into a long thread-like hollow vesicle(Fig. 3).

Pathways for Vesicle Division

Vesicle division by the extrusion of largevesicles through small pores is a way in whichmechanical energy can be used to drive division(Hanczyc et al. 2003). Vesicle growth by micellefeeding followed by division by extrusion can

A B

Figure 2. Vesicle shape transformations during growth. All vesicles are labeled with 2 mM encapsulated HPTS, awater-soluble fluorescent dye, in their internal aqueous space. (A) 10 min and (B) 30 min after the addition offive equivalents of oleate micelles to oleate vesicles (in 0.2 M Na-bicine, pH 8.5). Scale bar: 50 mm. (Reproducedfrom Zhu and Szostak 2009a and reprinted with permission from ACS Publications #2009.)






be performed repetitively, resulting in cycles ofgrowth and division in which both membranematerial and vesicle contents are distributedto daughter vesicles in each cycle. However,division by extrusion results in the loss of30%–40% of the encapsulated vesicle contentsto the environment during each cycle (Hanczycet al. 2003; Hanczyc and Szostak 2004). Mostof this loss is a result of the unavoidable geo-metric constraint of dividing a spherical vesicleinto two spherical (or subspherical) daughtervesicles with conservation of surface area;some additional loss may occur as a result ofpressure-induced membrane rupture. Althoughextrusion is a useful laboratory model forvesicle division, an analogous extrusion processappears unlikely to occur in a prebiotic scenarioon the early Earth because vesicle extrusionfrom the flow of suspended vesicles through aporous rock would require both the absence ofany large pores or channels and a very highpressure gradient (Zhu and Szostak 2009a).

The above problems stimulated a search fora more realistic pathway for vesicle division. Thepossible spontaneous division of small uni-lamellar vesicles after micelle addition has beenreported (Luisi et al. 2004; Luisi 2006), andelectron microscopy has revealed structuresthat are possible intermediates in growthand division, notably pairs of vesicles joinedby a shared wall (Stano et al. 2006). However,the mechanism of the proposed division aswell as the nature of the energetic drivingforce remain unclear. Additional studies arerequired to clearly distinguish between thevesicle-stimulated assembly of new vesicles,

and the more biologically relevant processes ofgrowth and division.

We have recently found that the growth oflarge multilamellar vesicles into long threadlikevesicles, described above, provides a pathway forcoupled vesicle growth and division (Zhu andSzostak 2009a). The long threadlike vesiclesare extremely fragile, and divide spontaneouslyinto multiple daughter vesicles in response tomodest shear forces. In an environment of gen-tle shear, growth and division become coupledprocesses because only the filamentous vesiclescan divide (Fig. 3). If the initial parental vesiclecontains encapsulated genetic polymers such asRNA, these molecules are distributed randomlyto the daughter vesicles and are thus inherited.The robustness and simplicity of this pathwaysuggests that similar processes might haveoccurred under prebiotic conditions. The mech-anistic details of this mode of division remainunclear. One possibility, supported by somemicroscopic observations, is that the long thinmembrane tubules are subject to the “pearlinginstability” (Bar-Ziv and Moses 1994), and min-imize their surface energy by spontaneouslytransforming from a cylindrical shape to a stringof beads morphology. The very thin tether join-ing adjacent spherical beads may be a weak pointthat can be easily disrupted by shear forces.

RNA-Catalyzed RNA Replication on the EarlyEarth and the Modern Laboratory

A core assumption of the RNAworld hypothesisis that the RNA genomes of primitive cells werereplicated by a ribozyme RNA polymerase

Add micelles Growth

Agitate

DivisionRepeated

cycles

Figure 3. Schematic diagram of coupled vesicle growth and division. (Reproduced from Zhu and Szostak 2009aand reprinted with permission from the Journal of the American Chemical Society #2009.)






(Gilbert 1986). The idea of RNA-catalyzed RNAreplication provides a solution to the apparentparadox of DNA replication catalyzed byproteins that are encoded by DNA. Thissimplification of early biochemistry gainedinstant plausibility from the discovery of cata-lytic RNAs almost 30 years ago (Kruger et al.1982; Guerrier-Takada et al. 1983). In the timeafter the discovery of the first ribozymes, theRNA World hypothesis has continued to gainsupport, most dramatically from the discoverythat the ribosome is a ribozyme (Nissen et al.2000), and that all proteins are assembledthrough the catalytic activity of the ribosomalRNA. Support has also come from the in vitroevolution of a wide range of new ribozymes,including bona fide RNA polymerases madeof RNA (Johnston et al. 2001). On the otherhand, no ribozyme polymerase yet comes closeto being a self-replicating RNA, like the repli-case envisaged as the core of the RNA Worldbiochemistry.

Why has the in vitro evolution of an RNAreplicase been so much more difficult thanoriginally expected? It is clear that the problemis not with catalysis of the chemical step, evenwith catalytically demanding triphosphatesubstrates. Evolutionarily optimized versionsof the Class I ligase carry out multiple-turnoverligation reactions at .1 s21, with over 50,000turnovers overnight (Ekland et al. 1995), andoptimized versions of the smaller, simplerDSL ligase carry out sustained multiple turn-over ligation reactions at rates .1/min (Voytekand Joyce 2007). These catalytic rates are morethan sufficient to carry out the replication of a100–200 nt ribozyme in minutes to hours, ifthese rates could be maintained in the contextof a polymerase reaction using monomer sub-strates. However, even the best available poly-merase ribozyme requires 1–2 days to copy10–20 nucleotides of a template strand, ap-parently as a consequence of poor binding toboth the ribonucleoside triphosphate (NTP)monomers and the primer-template substrate.The need to overcome the electrostatic repul-sion between negatively charged NTP andRNA substrates and ribozyme is thought to con-tribute to the very high Mgþþ requirement for

the polymerase ribozyme (Glasner et al. 2000).Such high levels of Mgþþ lead to hydrolyticdegradation of the ribozyme, and are also notcompatible with known fatty-acid based mem-branes because of crystallization of the fattyacid-magnesium salt. In addition, fatty-acidmembranes are almost impermeable to NTPs(Mansy et al. 2008).

The incompatibility between currently avail-able ribozyme polymerases and fatty-acidbased vesicles suggests either that early repli-cases were quite different, or that RNA repli-cation in early cells proceeded in a very differentmanner. For example, less charged and moreactivated nucleotides might be easier for aribozyme polymerase to bind, with little orno Mgþþ. Many potential leaving groups,such as imidazole, adenine or 1-Me-adeninehave been examined in template-directed andnontemplated polymerization reactions, buthave not yet been tested as substrates for ribo-zyme polymerases (Prabahar and Ferris 1997;Huang and Ferris 2006). The tethering of ribo-zyme and primer-template to hydrophobicaggregates has been examined as a way ofincreasing local substrate concentration (Mullerand Bartel 2008). However, this approach didnot lead to a dramatic improvement in theextent of template copying, apparently as aresult of ribozyme inhibition at high effectiveRNA concentrations. It may be possible toevolve polymerases that operate well at highRNA concentrations, but membrane localiza-tion by chemical derivatization adds furthercomplexity to a replication pathway because ofthe need for a specific catalyst for the derivatiza-tion step. Alternative means of facilitating theinteraction of a ribozyme with a primer-templatesubstrate, such as the presence of basic peptidesor other cofactors, might overcome this problem.Finally it is noteworthy that very small self-aminoacylating ribozymes have been obtainedusing RNA libraries that have an unconstrainedsequence at their 30-end (Chumachenko et al.2009). This strategy may also be fruitful in selec-tions for ribozyme polymerases.

Given the above constraints and uncertain-ties, what can we say about the emergence ofRNA-catalyzed RNA replication in the origin






of life? It is still possible that under the properconditions, and using the right substrates, asmall simple ribozyme could effectively catalyzeRNA replication. However, a replicase must domore than catalyze a simple phospho-transferreaction. Binding in a nonsequence-specificmanner to a primer-template complex, facilitat-ing binding of the proper incoming monomer,catalyzing primer extension, and repeating thisprocess until the end of the template (or setof templates for a multi-component replicase)is reached might require a complex replicasestructure. Such a replicase would presumablybe rare in collections of random RNA se-quences. If life required a very special sequenceto get started, then the origin of life on earthcould have been a low probability event andlife on other earth-like planets might be veryrare. If, on the other hand, RNA-catalyzed RNAreplication could have emerged gradually in aseries of simpler steps, it might have been easierand thus more likely for life to begin, and lifeelsewhere might be common. For this reason,we now turn to a consideration of nonenzy-matic template-directed replication chemistry.

Chemical Template Replication Revisited

The nonenzymatic template-directed polymeri-zation of activated ribonucleotides was studiedin depth by Leslie Orgel, together with his stu-dents and colleagues, over a period of severaldecades (Orgel 2004). Here, the template itselfacts as a catalyst by helping to align and orientthe monomers so that they are pre-organizedfor polymerization. The main lesson from thiswork is that spontaneous chemical copyingof RNA sequences is indeed possible, but issubject to several important constraints andlimitations. The constraints make template-directed RNA copying incompatible with cur-rently available membrane vesicle systems, andthe limitations have, so far at least, made itimpossible to obtain repeated cycles of RNAreplication through chemical copying.

To obtain reasonable reaction rates, Orgelmade use of nucleoside monophosphates acti-vated with a good leaving group such as imida-zole. The ribonucleotide 50-phosphorimidazolides

spontaneously assemble on a template oligonu-cleotide and polymerize over several days, gen-erating a complementary strand. However,monomer binding to RNA templates is weakand concentrations on the order of 0.1 M wererequired for optimal copying. In addition, aMgþþ concentration of �0.1 M, which asnoted above is incompatible with the presenceof fatty-acid based membranes, is required foroptimal polymerization. Even under theserather extreme conditions, polymerization pro-ceeds at only 1–2 nucleotides per day, and istherefore limited by monomer hydrolysis.Beyond these constraints, three aspects of thecopying reaction present major hurdles to mul-tiple rounds of replication. First, the chemicalstructure of RNA results in a problem of regio-specificity, because new linkages can be either3–50 or 20 –50 phosphodiester bonds. Surpris-ingly, under most conditions, it is the 20 –50

phosphodiester bonds that are most commonin polymerization products. This problem canbe ameliorated by the choice of ions and leavinggroups; for unknown reasons, Zn2þ ions andthe 2-methylimidazole leaving group favor thesynthesis of 30 –50 linkages (Lohrmann et al.1980; Inoue and Orgel 1982). Studies of oligo-nucleotide ligation showed that the helical con-text of an extended RNA duplex also favors theformation of 30 –50 linkages (Rohatgi et al.1996). However, it remains difficult to obtaina homogeneous RNA backbone without enzy-matic catalysis. Second, adenine (A) residuesin the template are difficult to copy, and twoor more As in succession block chain growth(Inoue and Orgel 1983), presumably as a resultof the poor base-stacking propensity of theincoming U monomers. It has recently beenfound that template copying at subzero temper-atures can proceed past multiple A residues inthe template (Vogel and Richert 2007), butthis required sequential additions of oxyazaben-zotriazole-activated monomer together with aseries of helper oligos. Third, there is the issueof fidelity. The copying of G and C residues isremarkably accurate, with error rates estimatedat 0.5% or less. However, the addition of Aand U residues causes problems, most signifi-cantly the formation of G:U wobble base-pairs






(Wu and Orgel 1992), which would lead to sig-nificant error rates in a four base system. Mightan alternative genetic polymer, perhaps even aclose relative of RNA, overcome these problemsand enable chemical replication? The identifica-tion of such a system might ultimately lead tothe discovery of plausible progenitors of RNA,or, alternatively, to the discovery of new replica-tion strategies that allow for the chemical repli-cation of RNA itself.

Phosphoramidate Nucleic Acids

Early studies of template copying using morereactive nucleotide derivatives were performedby Orgel et al., who examined both 20- and30-amino ribonucleotide 50-phosphorimidazo-lides (Lohrmann and Orgel 1976; Zielinksiand Orgel 1985; Tohidi et al. 1987). Polymeriza-tion of these nucleotides yields phosphorami-date nucleic acids, which are generally similarto standard phosphodiester linked nucleicacids except that the phosphoramidate linkageis more acid labile. As expected, replacing asugar hydroxyl in the monomer with a morenucleophilic amino group resulted in a largeincrease in monomer reactivity. The activated30-amino ribonucleotides participated in rapidcopying of short oligonucleotide templates5-13 nucleotides in length, yielding N30!P50

linked complementary oligonucleotides (Tohidiet al. 1987). The increased reactivity also led tofaster intramolecular monomer cyclization,which depleted the template copying reactionsof activated substrate molecules (Hill et al.1988).

More recently, the Richert group has begunto explore the potential of 30-amino- nucleotideanalogs in template-directed condensation reac-tions. Deoxyribonucleotide monomers, acti-vated with an oxyazabenzotriazole leavinggroup, completed a template-directed reactionwith a 30-amino-terminated primer in seconds(Rothlingshofer et al. 2008). The fidelity ofthis reaction is sufficient to allow for sequencingby nonenzymatic primer extension. However,the monomer is rapidly consumed by internalcyclization.

The high intrinsic reactivity of the amino-sugar modified nucleotides suggested to usthat an alternative phosphoramidate nucleicacid might act as a good platform for chemicalself-replication. Our group has therefore startedto study a series of phosphoramidate nucleicacids with sugar phosphate backbones thatvary in their degree of conformational flexibilityor constraint. We are currently focusing on the20-50 linked phosphoramidate analog of DNA,and the corresponding monomers, the 20-amino dideoxyribonucleotide 50-phosphorimi-dazolides (Fig. 4). The 20-amino ImpddNs areadvantageous as monomers because they can-not undergo intramolecular cyclization becauseof the steric constraint of the ribose ring; theyare only depleted during polymerization reac-tions by competing hydrolysis.

Our first experiments with 20-amino ImpddGled to rapid and efficient primer-extensionacross a dC15 template, generating full-lengthproduct in �6 hour (Mansy et al. 2008). En-couraged by the rapid copying of oligo-dCtemplates by 20-amino ImpddG, we performeda more extensive study of the copying of tem-plates with differing sugar-phosphate back-bones, lengths and sequences (Schrum et al.2009). The most important property contribu-ting to good template activity appears to be pre-organization in the form of an A-type helix.Thus, RNA templates were uniformly superior

O O

O P

O O

O– N

N 2′-NP-DNA

NH

O

N O

N N

NH2

N P O O O–

O P

O

O– NH

O P

O

O– NH

Figure 4. Structures of 20-50 phosphoramidate DNA,and the corresponding activated monomers.






to DNA templates of the same sequence, andLNA (locked nucleic acid) templates, whichare chemically locked in a C30-endo sugar con-formation, were superior to RNA templates.This result is consistent with previous obser-vations that under most conditions RNA-template directed polymerization of ribonu-cleotides leads to a majority of 20-50 linkages;it appears that an A-type helical geometry gen-erally favors polymerization through attack bya 20 nucleophile. A second key factor is thatenhanced monomer affinity for the templateincreases the reaction efficiency. Thus G:C base-pairs lead to efficient copying, whereas A:Ubase-pairs were very poorly copied. ReplacingAwith D (diaminopurine) results in a D:U base-pair with three hydrogen bonds, and slightlyimproved primer-extension. However, thepoor base stacking of U residues must also beimproved to obtain efficient template copying.When we replaced U with C5-propynyl-U, theresulting A:Up base-pairs led to improved copy-ing, but the D:Up combination had a clearlysynergistic effect (Fig. 5). In the context of aDNA duplex, the D:Up base-pair has previously

been shown to be energetically almost equiva-lent to a C:G base-pair (Chaput et al. 2002).When we used G, C, D, and Up as the fourmonomer and template bases, we were able tocopy mixed-sequence RNA templates over 15nts in length in about a day. This represents asignificant step toward developing a robust,generalized chemical replication system.

This system is, however, far from ideal, andthere are strong indications that the fidelity oftemplate copying becomes an issue when allfour nucleotides are present, largely because ofthe formation of G:U wobble base-pairs. Giventhat primer-extension on 20 –50 linked DNAtemplates is approximately similar to that onthe corresponding RNA templates, we are cur-rently synthesizing 20 –50 linked phosphorami-date DNA templates to assess self-replicationin this system. In light of the efficient templatingof LNA oligonucleotides, we are also interestedin exploring prebiotically plausible nucleicacids that are more conformationally constrai-ned than RNA. A particularly interesting candi-date is TNA (threose nucleic acid) (Schoninget al. 2000) and its 20-amino substitutedphosphoramidate version, NP-TNA (Wu et al.2002). These are both base-pairing systemsthat form standard Watson-Crick duplexes,despite having only five atoms per backbonerepeat unit (vs. six for RNA and DNA). It willbe of great interest to see if the resulting decreasein flexibility leads to increased fidelity intemplate copying reactions.

Protocell Assembly

In principle, protocell-like objects could formspontaneously as new membranes self-assembleand encapsulate genetic molecules in solution.Recently, simple physical processes that wouldenhance the efficiency of the coassembly ofnucleic acids and membrane vesicles havebeen proposed. One such alternative scenariois based on the fact that the clay mineralmontmorillonite is not only a catalyst of RNApolymerization (Ferris et al. 1996; Huang andFerris 2006) but also catalyzes membraneassembly (Hanczyc et al. 2003). Experimentswith clay particles containing surface-adsorbed

A - U

D - Up

N O

O

N

N

N

N

N N

N

N

NH2

HN

O

N

NH2

NH2 O

HN

Figure 5. Watson Crick base-pairs. Top: Standard A:Ubase-pair. Bottom: Alternative diaminopurine (D):C5-propynyl-uracil (Up) base-pair.






RNA showed that such particles stimulatedvesicle assembly, and frequently becametrapped inside the vesicles whose assemblythey had catalyzed. Thus, a common mineralcan catalyze both the assembly of a genetic pol-ymer and the assembly of a membrane vesicle,and bring these two components together togenerate a protocell-like structure consisting ofa genetic polymer trapped within a membranecompartment. Although the effectiveness ofthis process is attractive, some means of releas-ing at least some of the bound nucleic acid fromthe mineral surface, and/or replicating it on thesurface, would be necessary for subsequentreplication to occur.

More recent experiments suggest a verydifferent geochemical scenario leading to theassembly of similar protocell-like structures.The hollow channels within the rocks of thealkaline off-axis hydrothermal vents provide aprotected compartmentalized environmentwhere it has been suggested that primitive meta-bolic activities might have originated (Martinand Russell 2003). Recent theoretical studiessuggested that the strong thermal gradientspresent in hydrothermal vents, together withthe thin channels produced by mineral precipi-tation, could greatly concentrate small organicmolecules such as nucleotides as well as largernucleic acids from a very dilute external reser-voir (Baaske et al. 2007). Work from our labo-ratory (Budin et al. 2009) has confirmed thepredicted concentration effect, and has alsoshown that subcritical concentrations of fattyacids can be concentrated to the extent thatthey self-assemble into vesicles at the bottomof the capillary channels. Moreover, DNA oligo-nucleotides can also be greatly concentratedand can become encapsulated within thevesicles, resulting in the spontaneous assemblyof protocell-like structures.

Encapsulated Template Replication:Emergence of a Protocell

The experiments discussed earlier suggest thatthe assembly of protocell-like structures is notthat difficult, because it appears to be possiblethrough multiple distinct mechanisms. The

more challenging question is, how could sucha structure replicate? We have already consid-ered the replication of the protocell membraneand the genetic material as separate entities.To address the question of their replication asa combined structure we must consider inmore detail the molecular constituents of theprotocell membrane and the molecular natureof the encapsulated genetic material.

Genome replication within a protocell canonly occur if the building blocks used to copytemplate strands are able to enter the fatty-acidvesicle compartment. Early work using phos-pholipid-based vesicles and protein enzymesshowed the feasibility of constructing primitivecell-like compartments (Chakrabarti et al. 1994;Luisi et al 1994). More recent permeability stud-ies (Mansy et al. 2008; see also Mansy 2010)showed that nucleotides could spontaneouslydiffuse across simple fatty-acid membranes,but that net negative charge is a critical determi-nant of permeability. Thus, nucleotides that arechemically activated, e.g., by conversion ofthe 50-phosphate to a 50-phosphorimidazolide,equilibrate across vesicle membranes muchmore rapidly on account of the reduction ofthe net negative charge. In addition, mixturesof fatty acids with their glycerol esters generatemembranes that are more permeable to polarand charged molecules. By combining theseobservations, we were able to show that activated20-amino-20,30-dideoxyguanosine-50-phosphor-imidazolide, the same nucleotide previouslyshown to rapidly copy oligo-dC templates,could be added to the outside of fatty-acidvesicles containing an encapsulated primer-template complex, and copy the internal dC15

template. Copying of the encapsulated dC15

template by primer-extension reached .95%completion in 12–24 hour (Mansy et al.2008), compared with 6–12 hour in free solu-tion. The longer time required for copyingencapsulated templates reflects the timerequired for entry of external nucleotides tothe interior of the vesicles. Importantly, thepresence of high concentrations (5–10 mM)of highly reactive activated nucleotides didnot have any disruptive effects on the integrityof the vesicle membrane, as no leakage of






encapsulated primer-template complexes wasobserved. It is also important to note thatcontrol experiments with phospholipid mem-branes showed no copying of internal template,because the activated nucleotides could notenter the vesicle; similarly, “modern” activatednucleotides such as nucleoside triphosphatescannot cross fatty-acid based membranes.Successful copying of encapsulated templatestherefore requires both “primitive” nucleotideswith reduced charge, and “primitive” mem-branes composed of single chain amphiphiles.

The copying of a genetic polymer inside amembrane compartment is an important steptoward the realization of a self-replicatingsystem capable of Darwinian evolution. What,then, are the remaining barriers to the assemblyof such a system? The copying of a single-stranded template produces a double-strandedproduct; these strands would have to separatebefore a second cycle of genome replicationcould begin. Separate follow-up experimentsby our group showed that some fatty-acid basedvesicles are able to retain encapsulated DNA andRNA oligonucleotides over a temperature rangeof 0 8C to 100 8C (Mansy and Szostak 2008).As with permeability, mixtures of amphiphileslead to improved thermostability, with glycerolesters being particularly stabilizing, possiblybecause of the additional hydrogen bonddonors and acceptors provided by the glycerolhead group. Furthermore, we found that en-capsulated double-stranded DNA could bedenatured at elevated temperatures, with thestrands reannealing once the temperature waslowered (Mansy and Szostak 2008). This impliesthat thermal fluctuations could provide a mech-anism for strand separation that is compatiblewith the integrity of fatty-acid vesicles, poten-tially allowing for complete cycles of replicationof encapsulated genetic polymers. The mutualcompatibility of nucleic acid replication andfatty-acid compartment growth is very encour-aging because it alleviates concerns related tothe permeability and stability of membranevesicles. Vesicles therefore do seem to be aphysically plausible way to segregate and spa-tially localize genomes, keep emergent catalyticpolynucleotides physically close to their encod-

ing genome, and protect the nascent evolvingsystem from parasitic polymers.

CHALLENGES AND FUTURE RESEARCHDIRECTIONS

Prospects for a Complete Protocell Model

Although considerable progress has been madetoward the assembly of model protocells, severalremaining issues must be solved before multiplecycles of protocell replication can be achievedin the laboratory. These factors are also relevantto protocell replication on the early Earth. Themost important factor at this time appears tobe the competition between strand reannealingand strand copying, after thermal strand separa-tion. PCR reactions generally plateau at about1-mM DNA strand concentration, which is theconcentration at which strand reannealing andstrand copying occur on a similar time scaleof about 1 minute However, nonenzymatictemplate copying requires on the order of aday for completion, which implies that eithertemplate copying must be much faster, orreannealing must be much slower. One way tomake reannealing sufficiently slow is to keepstrand concentrations subnanomolar. Lowstrand concentrations are possible in largevesicles, but it is hard to see how a few moleculesof a genetic polymer could have any significantphenotypic effect on a large vesicle composed ofmillions of amphiphilic molecules. The emer-gence of metabolic ribozymes would be moreplausible if nucleic acid strand concentrationswere much higher, so that a catalyst of modestefficiency could generate enough product toinfluence cell properties.

What other factors might affect the rate ofstrand reannealing? Perhaps the most obviouspossibility is that secondary structure, whichcan form extremely rapidly because the inter-actions are intramolecular, could greatly slowdown strand annealing. This phenomenon isessential to the viability of single-strandedRNA phage such as Qb† (Axelrod et al. 1991).Significant intrastrand secondary structurewould also be an expected consequence of selec-tion for sequences that fold into functional






shapes with catalytic activity. On the otherhand, chemical replication through dense sec-ondary structure would probably be muchslower than replication of an unfolded, opentemplate. The outcome of simultaneous selec-tion for an open, accessible template sequence,and a folded functional structure remains un-clear. An alternative but even more speculativesolution might result from the rapid binding bybase-pairing of short oligonucleotides or evenmonomers to freshly separated strands. If a tem-plate strand was largely occupied by monomersor short oligomers, even if these were in rapidexchange, the strand might be prevented fromannealing to a complementary strand. This pos-sibility has the advantage that it need not block orslow the copying reaction, however its effective-ness remains to be tested experimentally.

Another challenge faced by replicatingprotocells, whether on the early earth or in themodern laboratory, is the continuous dilutionof protocells through the competing formationof new empty vesicles. When new fatty acids aresupplied as micelles, the efficiency of incorpora-tion into pre-formed vesicles (or protocells) canbe quite high, but some new vesicles are alwaysformed. Thus over time, the descendants of agiven protocell will gradually be diluted out bythe continuous formation of these new vesicles.To avoid extinction by dilution, the protocellsmust out-compete other vesicles either by hav-ing a more rapid cell cycle, thereby generatingmore progeny during division, or by survivingdestructive processes more efficiently. We havepreviously proposed (Chen et al. 2004) thatfaster growth, driven by the osmotic pressureof encapsulated nucleic acid, could lead to aneffectively shorter cell cycle for protocells thatcontain high copy numbers of their replicatinggenome. However, in light of the problems asso-ciated with this approach, it is of considerableinterest to explore new ways in which a protocellgenome could lead to faster growth or growththat occurs at the expense of empty vesicles.An alternative strategy for surviving dilutionwould be for a protocell genome to colonizeempty vesicles. This could occur through alow level of stochastic vesicle–vesicle fusionevents, possibly catalyzed by low levels of

divalent cations such as Ca2þ. Systematic effortsto measure vesicle fusion frequencies under dif-ferent environmental conditions could there-fore be quite useful. Finally, it is possible thatthis problem could be circumvented entirely ifearly life was discontinuous (Budin et al.2009). For example, protocells could be occa-sionally disrupted by drying, or simply dissolveas a result of dilution with water to a level belowthe critical aggregate concentration; subsequentre-hydration or concentration would result inreformation of vesicles encapsulating genomicnucleic acids, thus generating a new “random-ized” set of protocells. As long as such eventswere fairly uncommon, and assuming thatgenomic replication had kept ahead of vesiclereplication so that each vesicle contained multi-ple genome copies prior to disruption, thisprocess of disruption and reformation wouldlead to the spread of evolving genomic sequen-ces through the “new” vesicles.

Laboratory models of protocell systemsshould be helpful in modeling many of theabove scenarios. Assuming that protocell repro-duction can be achieved, and made efficientenough to continue through many generations,it should then be possible to observe the spon-taneous evolution of adaptive innovations inthis relatively simple chemical system. Thenature of such adaptations may provide cluesas to how modern cells evolved from their ear-liest ancestors. Ultimately this line of researchmay also tell us whether the conserved bio-chemistry of life is driven by chemical necessity,or whether biochemically very different formsof life are also possible.

ACKNOWLEDGMENTS

We thank Itay Budin, Matt Powner, andother members of the Szostak lab for helpfuldiscussions.

REFERENCES

Axelrod VD, Brown E, Priano C, Mills DR. 1991. ColiphageQ b RNA replication: RNA catalytic for single-strandrelease. Virology 184: 595–608.






Baaske P, Weinert FM, Duhr S, Lemke KH, Russell MJ, BraunD. 2007. Extreme accumulation of nucleotides in simu-lated hydrothermal pore systems. Proc Natl Acad Sci104: 9346–9351.

Bar-Ziv R, Moses E. 1994. Instability and “pearling” statesproduced in tubular membranes by competition of cur-vature and tension. Phys Rev Lett 73: 1392–1395.

Berclaz N, Muller M, Walde P, Luisi PL. 2001. Growth andtransformation of vesicles studied by ferritin labelingand cryotransmission electron microscopy. J Phys ChemB 105: 1056–1064.

Blochliger E, Blocher M, Walde P, Luisi PL. 1998.Matrix effect in the size distribution of fatty acid vesicles.J Phys Chem B 102: 10383–10390.

Budin I, Bruckner R, Szostak JW. 2009. Formationof protocell-like vesicles in a thermal diffusion column.J Am Chem Soc 131: 9628–9629.

Chakrabarti AC, Breaker RR, Joyce GF, Deamer DW. 1994.Production of RNA by a polymerase protein encapsu-lated within phospholipid vesicles. J Mol Evol 39: 555–9.

Chaput JC, Sinha S, Switzer C. 2002. 5-propynyluracil.dia-minopurine: An efficient base-pair for non-enzymatictranscription of DNA. Chem Commun 15: 1568–9.

Chen IA, Roberts RW, Szostak JW. 2004. The emergenceof competition between model protocells. Science 305:1474–1476.

Chen IA, Szostak JW. 2004. A kinetic study of the growth offatty acid vesicles. Biophys J 87: 988–998.

Chumachenko NV, Novikov Y, Yarus M. 2009. Rapid andsimple ribozymic aminoacylation using three conservednucleotides. J Am Chem Soc 131: 5257–63.

Deamer DW. 1985. Boundary structures are formed byorganic-components of the Murchison carbonaceouschondrite. Nature 317: 792–794.

Deamer DW, Pashley RM. 1989. Amphiphilic componentsof the Murchison carbonaceous chondrite: Surface prop-erties and membrane formation. Orig Life Evol Biosph 19:21–38.

Dworkin J, Deamer D, Sandford S, Allamandola L. 2001.Self-assembling amphiphilic molecules: Synthesis insimulated interstellar/precometary ices. Proc Natl AcadSci 98: 815–819.

Ekland EH, Szostak JW, Bartel DP. 1995. Structurally com-plex and highly active RNA ligases derived from randomRNA sequences. Science 269: 1319–25.

Ferris JP, Hill AR Jr, Liu R, Orgel LE. 1996. Synthesis of longprebiotic oligomers on mineral surfaces. Nature 381:59–61.

Gebicki JM, Hicks M. 1973. Ufasomes are stable particlessurrounded by unsaturated fatty acid membranes. Nature243: 232–234.

Gilbert W. 1986. The RNAWorld. Nature 319: 618.

Glasner ME, Yen CC, Ekland EH, Bartel DP. 2000. Recogni-tion of nucleoside triphosphates during RNA-catalyzedprimer extension. Biochemistry 39: 15556–62.

Guerrier-Takada C, Gardiner K, Marsh T, Pace N, Altman S.1983. The RNA moiety of ribonuclease P is the catalyticsubunit of the enzyme. Cell 35: 849–57.

Hanczyc MM, Szostak JW. 2004. Replicating vesicles asmodels of primitive cell growth and division. CurrOpin Chem Biol 8: 660–664.

Hanczyc MM, Fujikawa SM, Szostak JW. 2003. Experimen-tal models of primitive cellular compartments: Encapsu-lation, growth, and division. Science 302: 618–622.

Hargreaves WR, Deamer DW. 1978. Liposomes from ionic,single-chain amphiphiles. Biochemistry 17: 3759–3768.

Hill AR Jr, Nord LD, Orgel LE, Robins RK. 1988. Cyclizationof nucleotide analogues as an obstacle to polymerization.J Mol Evol 28: 170–1.

Huang W, Ferris JP. 2006. One-step, regioselective synthesisof up to 50-mers of RNA oligomers by montmorillonitecatalysis. J Am Chem Soc 128: 8914–9.

Inoue T, Orgel LE. 1982. Oligomerization of (guanosine50-phosphor)-2-methylimidazolide on poly(C). AnRNA polymerase model. J Mol Biol 162: 201–17.

Inoue T, Orgel LE. 1983. A nonenzymatic RNA polymerasemodel. Science 219: 859–62.

Johnston WK, Unrau PJ, Lawrence MS, Glasner ME, BartelDP. 2001. RNA-catalyzed RNA polymerization: Accurateand general RNA-templated primer extension. Science292: 1319–25.

Kruger K, Grabowski PJ, Zaug AJ, Sands J, Gottschling DE,Cech TR. 1982. Self-splicing RNA: Autoexcision andautocyclization of the ribosomal RNA interveningsequence of Tetrahymena. Cell 31: 147–57.

Lohrmann R, Orgel LE. 1976. Template-directed synthesisof high molecular weight polynucleotide analogues.Nature 261: 342–344.

Lohrmann R, Bridson PK, Orgel LE. 1980. Efficient metal-ion catalyzed template-directed oligonucleotide synthe-sis. Science 208: 1464–5.

Luisi PL. 2006. The emergence of life: From chemicalorigins to synthetic biology, Cambridge University Press,Cambridge.

Luisi PL, Walde P, Oberholzer T. 1994. Enzymatic RNAsynthesis in self-reproducing vesicles: An approach tothe construction of a minimal synthetic cell. BerBunsenges Phys Chem 98: 1160–5.

Luisi PL, Stano P, Rasi S, Mavelli F. 2004. A possible route toprebiotic vesicle reproduction. Artif Life 10: 297–308.

Mansy SS. 2010. Membrane transport in primitive cells.Cold Spring Harb Perspect Biol 2: a002188.

Mansy SS, Szostak JW. 2008. Thermostability of model pro-tocell membranes. Proc Natl Acad Sci 105: 13351–13355.

Mansy SS, Schrum JP, Krishnamurthy M, Tobe S, Treco DA,Szostak JW. 2008. Template-directed synthesis of agenetic polymer in a model protocell. Nature 454:122–125.

Martin W, Russell MJ. 2003. On the origins of cells: Ahypothesis for the evolutionary transitions from abioticgeochemistry to chemoautotrophic prokaryotes, andfrom prokaryotes to nucleated cells. Philos Trans R SocLond B Biol Sci 358: 59–83.

McCollom TM, Ritter G, Simoneit BR. 1999. Lipid synthesisunder hydrothermal conditions by Fischer-Tropsch-typereactions. Orig Life Evol Biosph 29: 153–166.

Monnard PA, Apel CL, Kanavarioti A, Deamer DW. 2002.Influence of ionic inorganic solutes on self-assembly






and polymerization processes related to early formsof life: Implications for a prebiotic aqueous medium.Astrobiology 2: 139–52.

Muller UF, Bartel DP. 2008. Improved polymerase ribozymeefficiency on hydrophobic assemblies. RNA 14: 552–62.

Nissen P, Hansen J, Ban N, Moore PB, Steitz TA. 2000.The structural basis of ribosome activity in peptidebond synthesis. Science 289: 920–30.

Orgel LE. 2004. Prebiotic chemistry and the origin of theRNA world. Crit Rev Biochem Mol Biol 39: 99–123.

Prabahar KJ, Ferris JP. 1997. Adenine derivatives as phos-phate-activating groups for the regioselective formationof 3’,5’-linked oligoadenylates on montmorillonite: Pos-sible phosphate-activating groups for the prebiotic syn-thesis of RNA. J Am Chem Soc 119: 4330–7.

Rohatgi R, Bartel DP, Szostak JW. 1996. Nonenzymatic,template-directed ligation of oligoribonucleotides ishighly regioselective for the formation of 30-50 phospho-diester bonds. J Am Chem Soc 118: 3340–4.

Rothlingshofer M, Kervio E, Lommel T, Plutowski U,Hochgesand A, Richert C. 2008. Chemical primer exten-sion in seconds. Angew Chem Int Ed Engl 47: 6065–8.

Rushdi AI, Simoneit BR. 2001. Lipid formation by aqueousFischer-Tropsch-type synthesis over a temperature rangeof 100 to 400 degrees C. Orig Life Evol Biosph 31:103–118.

Sacerdote MG, Szostak JW. 2005. Semi-permeable lipidbilayers exhibit diastereoselectivity favoring ribose;implications for the origins of life. Proc Natl Acad Sci102: 6004–6008.

Schoning K, Scholz P, Guntha S, Wu X, Krishnamurthy R,Eschenmoser A. 2000. Chemical etiology of nucleic acidstructure: The a-threofuranosyl-(30!20) oligonucleo-tide system. Science 290: 1347–51.

Schrum JP, Ricardo A, Krishnamurthy K, Blain JC, Szostak JW.2009. Efficient and rapid template-directed nucleic acidcopying using 20-amino-20, 30-dideoxyribonucleoside-50-phosphorimidazolide monomers. J Am Chem Soc 31:14560–14570.

Stano P, Wehrli E, Luisi PL. 2006. Insights into the self-reproduction of oleate vesicles. J Phys: Condens Matter18: S2231–S2238.

Szostak JW, Bartel DP, Luisi PL. 2001. Synthesizing life.Nature 409: 387–390.

Tohidi M, Zielinski WS, Chen CH, Orgel LE. 1987.Oligomerization of 30-amino-30deoxyguanosine-50phos-phorimidazolidate on a d(CpCpCpCpC) template. J MolEvol 25: 97–99.

Vogel SR, Richert C. 2007. Adenosine residues in thetemplate do not block spontaneous replication steps ofRNA. Chem Commun 19: 1896–8.

Voytek SB, Joyce GF. 2007. Emergence of a fast-reactingribozyme that is capable of undergoing continuousevolution. Proc Natl Acad Sci 104: 15288–93.

Walde P, Wick R, Fresta M, Mangone A, Luisi PL. 1994a.Autopoietic self-reproduction of fatty acid vesicles.J Am Chem Soc 116: 11649–11654.

Walde P, Goto A, Monnard P-A, Wessicken M, Luisi PL.1994b. Oparin’s reactions revisited: Enzymatic synthesisof poly(adenylic acid) in micelles and self-reproducingvesicles. J Am Chem Soc 116: 7541–7547.

Wu T, Orgel LE. 1992. Nonenzymatic template-directedsynthesis on hairpin oligonucleotides. 3. Incorporationof adenosine and uridine residues. J Am Chem Soc 114:7963–9.

Wu X, Guntha S, Ferencic M, Krishnamurthy R,Eschenmoser A. 2002. Base-pairing systems relatedto TNA: a-Threofuranosyl oligonucleotides containingphosphoramidate linkages. Org Lett 4: 1279–82.

Zhu TF, Szostak JW. 2009a. Coupled growth and division ofmodel protocell membranes. J Am Chem Soc 131: 5705–5713.

Zhu TF, Szostak JW. 2009b. Preparation of large monodis-perse vesicles. PLoS ONE 4: e5009.

Zielinski WS, Orgel LE. 1985. Oligomerization of activatedderivatives of 30-amino-30-deoxyguanosine on poly(C)and poly(dC) templates. Nucleic Acids Res 13: 2469–2484.






the origins of cellular life - newcastle university · the origins of cellular life jason p....

Documents