Minerals Formed by OrganismsHeinz A. Lowenstam

In the past two decades our knowledgeof the diversity of biomineralizationproducts and processes has increasedbeyond expectation. The vast array oforganisms now known to precipitateminerals, the numerous cell and tissuesites at which they form and reform, andthe diverse functions which they fulfillwere to a large extent unknown until re-

cent years (1, 2). In this article I attemptto document much of this diversity and,where possible, to begin the task of iden-tifying underlying trends and associa-tions. I emphasize the properties of the

ment. Calcium carbonate buildup has de-fined the morphology of continental mar-

-gins over vast areas. By presenting thismaterial, I hope to reemphasize the needfor improving our understanding of thisimportant biological process.

Diversity of Biogenic Minerals

A total of 31 different biogenic miner-als have been identified to date (2). Fig-ure 1 lists these minerals and shows theirdistribution in the phyla of the five king-

Summary. Organisms are capable of forming a diverse array of minerals, some ofwhich cannot be formed inorganically in the biosphere. The initial precipitates maydiffer from the form in which they are finally stabilized, or during development of theorganism one mineral may substitute for another. Biogenic minerals commonly haveattributes which distinguish them from their inorganic counterparts. They fulfill impor-tant biological functions. They have been formed in ever-increasing amounts duringthe last 600 million years and have radically altered the character of the biosphere.

biogenic minerals themselves and in-troduce the subject of biomineralizationprocesses, which are generally less fa-miliar.

In the broader context, understandingbiomineralization goes beyond the im-portant need to identify the fundamentalmechanisms involved. Throughout geo-

logical time the formation of minerals byorganisms has increasingly modified thechemical and physical nature of the bio-sphere (3). Aspects of the chemistry ofthe oceans have been radically changed.Biogenic minerals have contributed sig-nificantly to the sedimentary environ-

The author is a professor of paleoecology in theDivision of Geological and Planetary Sciences, Cali-fornia Institute ofTechnology, Pasadena 91125. Thisarticle is based in part on a talk presented at the Con-ference on Advances in the Biogeochemistry ofAmino Acids held 29 October to I November 1978 atAirlie House, Warrenton, Virgma. The conferencewas jointly sponsored by the Carnegie Institution ofWashington and the National Science Foundationand marked the 25th anniversary of Philip H. Abel-son's discovery of amino acids in fossils and the oc-casion of his retirement as president of the CarnegieInstitution of Washington. Parts of the talk havebeen published in Biogeochemistry ofAmino Acids,P. E. Hare, T. C. Hoering, K. King, Jr., Eds.(Wiley, New York, 1980).

doms. A nearly threefold increase in thenumber of identified biomineralizationproducts has occurred since a similarcompilation was made 17 years ago (4).-In particular, additional phosphates, ironoxides, iron sulfides, and manganeseoxides have been discovered (1, 2). Wecan now discern three dominant traits inbiomineralization products: (i) two-thirds are calcium minerals, (ii) almosttwo-thirds contain H20 or OH, and (iii)one-fourth consist of colloidal materials(2).

The distribution of biogenic mineralsbetween the five kingdoms shows that 25are synthesized by animals, 11 by pro-toctists, 8 by monerans, 7 by vascularplants, and 4 by fungi (2). Animals are

thus singled out as the most versatilemineralizers, although this may, in part,be due to the fact that they have beenmore extensively -investigated than theothers. The distribution of biogenic min-erals within the phyla, as shown in Fig.1, confirms the long-established fact thatcarbonate minerals are by far the most

0036-8075/81/0313-1126$01.50/0 Copyright 0 1981 AAAS

widely utilized bioinorganic constitu-ents. Carbonate minerals such as vater-ite, monohydrocalcite, and amorphouscalcium carbonate cannot now be con-sidered biological oddities, inasmuch asthey are found in species of many genera(Fig. 1). Opal emerges as the secondmost extensively formed biogenic miner-al. Ferrihydrite and related ferric oxideminerals rank third, and magnetite(Fe3O4), first discovered as a biomineral-ization product in chiton teeth (5) andnow known to be synthesized by magne-totactic bacteria (6, 7), honey bees (8),and homing pigeons (9), may well proveto be the fourth most extensively formedbiogenic mineral.

Mineral-Forming Processes

Two fundamentally different process-es of mineral formation can be distin-guished. The first, exemplified by numer-ous animals, is an "organic matrix-medi-ated" process. In general, the organismconstructs an organic framework .ormold into which the appropriate ions areactively introduced and then induced tocrystallize and grow. The mineral type,orientation of crystallographic axes, andmicroarchitectures are under geneticcontrol.The second basic process of mineral

formation, exemplified by some bacterialspecies as well as various green andbrown algae, is characterized by bulk ex-tracellular and/or intercellular mineralformation, without the elaboration of or-ganic matrices. This "biologically in-duced" mineralization results in the min-erals having crystal habits similar to thoseproduced by precipitation from inorganicsolutions. Resulting polycrystalline ag-gregates-found, for example, in the in-tercellular spaces of some green algae-show random orientation of their mineralconstituents (10). The bacterial mineralprecipitates form as a result of the inter-action between the biogenically formedgases and metal ions present in the ex-ternal medium (11). In the algae, mineralformation results from the reduction ofCO2 in the external medium due tophotosynthetic activity. As the water inwhich these algae live is generally super-saturated with respect to calcium car-bonate, the CO2 reduction induces theminerals to form (12). These biologicallyinduced mineral-forming processes arefar less rigorously controlled than organ-ic matrix-mediated mineralization, andthey appear to constitute an earlier,more ""primitive" stage in the evolutionof biogenic mineral formation.

SCIENCE, VOL. 211, 13 MARCH 19811126

on

Nov

embe

r 14

, 200

7 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

Structural Complexity and Sites ofDeposition

Biogenic minerals may be amorphous,paracrystalline, or crystalline (2). At agiven site, they may occur as a singleunit (such as a single crystal), numerousindividual units, or aggregates. The lat-ter, in turn, are known to form structuresof varying degrees of complexity. Miner-alized skeletal hard parts are examples ofsuch aggregates. The aggregated unitsare usually arranged in an orderly fash-ion, and when crystalline, the crystal-lographic axes are partially or fullyaligned (13, 14). These ordered struc-tures are generally differentiated into anumber of microarchitectural units, eachof which is enveloped by an organic ma-

trix sheath (Fig. 2a) (15). Compound ag-gregates with randomly oriented crystal-lographic axes, such as are found insome algae (Fig. 2b), are less common.In exoskeletal structures an additionalorganic veneer may protect the mineral-ized hard parts from the external envi-ronment. The organic matrix structuralframework and the surficial organic ve-neers appear to be absent in aggregateswith randomly oriented mineral constitu-ents, such as some green algae. Thestructural complexity of biogenic miner-als, ranging from single crystals to com-pound aggregates, can be convenientlysubdivided into six categories (2).

Mineral deposits are formed either in-tracellularly, intercellularly, or extra-cellularly (16). The intracellular mineral-

ization site is by far the most common. Itoccurs in a few ifonerans (bacteria) andis abundant in three of the four eu-karyotic kingdoms (Protoctista, Ani-malia, and Plantae). Intracellular depos-its are located within organelles or invesicles. The most commonly identifiedsites are the mitochondria, the Golgicomplex, or vesicles, which are in-timately associated with the latter. Inter-cellular mineralization occurs in somemulticellular filamentous algae, but isnot, as yet, known in plants or animals.Extracellular mineralization sites arewidespread and well known from all fivekingdoms.Monerans, most protoctists, and fungi

have only one mineralized site, whichmay be intra-, inter-, or extracellular. In

KINGDOM

ICL-

MONER/A4 PRO'OCTSfTA IFU11.4N'LANM4ALI/A TPeLASTAf,1 ,

l - T - r

K BD.IZ~I

..2:11- k".;z -.t'Rt

S llq.!-,i IR t%

csZ%- :-,.2-., -z .k i-.1.-4 -1-.11-1 1- :!<

t-Z., L..CC ..

"I

iS~x44%t.;,

CARBONATES:____ llI_______CAL ITE ~~+ --+++ +4- ++++ +++++ + +

ARAGONITE +__ _++ + + T-+ + ++ ++ + +VATERITE _ _ + +1+ + +MONOHYDROCALCITE + + __I_ __AMORPH. HYDR. CARB. +_I + ++ +

PHOSPHATES: -----_ +-TDAH TUE_______ - --- +FRANCOLITE + + +3Mg31?Q44 + I I _SRtIST-~ ~~ ___T ___ T+T__II_ _

AMORPH. DAHLLITE PRECURSOR - - - + + +

AMORPH. WHITLOCKITE PRECURSOR _I_ + +_ IAMORPH. HYDR. FERRIC FHOSPHATE - _ - _ ++ +

HALIDES:FLUORITE ++ +AMORPH. FLUORITE PRECURSOR + l+ li_

[XALATES:WHEWELLITE - -I + ? - + + +WEDDELITE + ? + ++

SULFATES:GYPSUM _ +T -? +CELESTITE __ ___BARITE II _

|SILICA: l-- + |2 |||| | | || 2-OPAL ______ ++

Fe-OXIDES: l______MAGNETITE + +1+ +--MAGHEMITE _______GOETHITE _ +LEPIDOCROCITE - +FERRIHYDRITE +- + 4-__ +4 + + TjAMORPH. FERRIHYDRATES ______ +

_Mn-OXIDES:_ T- T - __T___ _ _Fe-SULFIDES: __

PYRITE _ _T_HYDROTROILITE = __

Fig. 1. Diversity and phylum distribution of biomineralization products in extant organisms.

13 MARCH 1981 1127

on

Nov

embe

r 14

, 200

7 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

I

Fig. 2. (a) Example of organic matrix-mediated biomineralization process: calcitic prismaticlayer of the bivalve Mytilus californianus, showing crystal enveloping organic matrix. (b) Ex-ample of biologically induced mineralization process: intercellular array of randomly orientedaragonite crystals in the algae Galaxaura obtusata.

many tissue-grade eukaryotes, mineralsform at all three sites. Coelomate ani-mals show the greatest diversity in min-eral deposition sites. Many of these con-tain intracellular minerals at a number oftissue localities, and in some cases theintracellular deposits are monomineralicwhereas the extracellular deposits may

contain a few different mineral types.These are usually located in separate dis-crete microarchitectural units.

Minerals formed at a particular sitemay be (i) retained in place, (ii) trans-ferred intact to other sites, (iii) excreted,(iv) dissolved and replaced continuously,periodically, or only occasionally, or (v)continuously reconstituted. Mineral de-posits which are subject to dissolutionand replacement function, in part, as res-

ervoirs for cations or anions to be usedfor a variety of vital functions. Transferof ions destined for mineral formationmay involve a single or a multistep dis-solution-reprecipitation process. At theterminal deposition site, the mineral ei-ther is stabilized or undergoes continu-ous reconstitution.

Examples of Biomineralization Processes

The following examples of biomineral-ization processes both illustrate the var-

ious properties described above and are

not generally familiar. Magnetotacticbacteria contain intracellular chains ofsingle magnetite (Fe304) crystals whichare bound to the membrane. An as yetunidentified ferric iron oxide-bearingcompound may be the precursor of themagnetite crystals (7). The magnetitecrystals are stabilized in place by an

elongated sheath of organic material.1128

The swimming behavior of these bacteriain a magnetic field indicates that the mag-netite serves as a biomagnetic compass(17). In the planktonic Coccolithophor-idae, the calcitic platelets are formed in-tracellularly in the Golgi cisternae and,upon completion, are extruded to the cellsurface (18). The gastropod Pomaceapaludosa has an aragonitic shell. It formsegg capsules containing vaterite, anotherpolymorph of calcium carbonate. Thesources of calcium and carbonate for theshell are amorphous calcium carbonateor vaterite granules, located in the Golgiapparatus of the mantle epithelial cells.It has been suggested that granules local-ized in the foot, which are composed ofamorphous CaCO3 and trace amounts ofvaterite, are used for large-scale suddenmobilization of calcium during shell re-

generation. Normally these granulesserve as a reservoir for the calcium thatregulates muscle activity in the foot. Yeta third reservoir of calcium is located inthe albumen-capsule gland of the fe-males. These granules of vaterite pro-vide calcium for mineralization of theegg capsules (19).

Perhaps one of the most complex se-

quences of mineralization known in-volves the formation of magnetite crys-

tals in the tooth denticles of a group ofmollusks called chitons (20). Iron is ex-

tracted from the food and transported inthe bloodstream, apparently in the formof a ferrihydrite (5Fe2O3'9H20) (21) core

surrounded by the protein apoferritin(20, 22). The iron from the complex (fer-ritin) is removed from the blood (23) tothe superior epithelial cells of the radula,which adhere tightly to the tooth den-ticles. These cells accumulate large con-

centrations of ferritin, which is the stor-

age site for the iron mineralization of thetooth. From here the iron is transferredinto the preformed organic matrix frame-work of the teeth (20, 24). As there is noevidence of intact transfer of ferritin tothe tooth denticles, the iron must betransported in the soluble ferrous formacross the cell membrane. In fact, theferritin micelles are observed to be invarious stages of demineralization (20).The iron is then redeposited in specificareas on the surface of the tooth in theform of ferrihydrite. Within 1 or 2 days,however, the ferrihydrite is converted tomagnetite, by an as yet undeterminedbiochemical process (24). The unmineral-ized portions of the organic matrix arelater infilled by other minerals such asamorphous hydrous ferric phosphate (5)or lepidocrocite and francolite (25). Thehardness of the magnetite affords the chi-tons an advantage over other grazers byenabling them to recover the algae em-bedded within the rocks (5). Mineralizedhard parts can have a number of func-tions other than structural and mechani-cal ones. It is well known that the endo-skeletons of mammals act as a reservoirof calcium and phosphorus and are in-timately linked to the organism's metab-olism. It is perhaps less widely appreci-ated that many cold-blooded vertebratesas well as invertebrates utilize their hardparts for a similar purpose. In the gastro-pod genera Conus and Cypracea, thesurface-exposed aragonitic shell wallsare always significantly thicker than theinner shell walls. The inner shell wall cal-cium carbonate is dissolved and trans-ferred, in solution, to the growing shelledge. This process continues throughoutpostlarval development (26). Other gas-tropods limit their recycling activity tomodify their aperture-restricting orna-mentations (27, 28). It has been observedthat these species also deposit new shellin a discontinuous manner. Shoal-waterbivalves, which are forced to keep theirvalves closed for extended periods dueto storms or unusual tides, are known to"etch" the internal surfaces of theirshells. The mobilized carbonate buffersthe enclosed fluids and prevents the pHfrom dropping due to the production ofmetabolic CO2 and organic acids (29).Another example of ion mobilization un-der stress was recently noted in the blueshark, Prionace glauca. The aragoniticotoconia (mineralization in the vestibularapparatus) of a pregnant female capturedclose to the end of gestation showed no-ticeable surface resorption, suggestingthat calcium was mobilized from the oto-conia to the rapidly mineralizing em-bryos (Fig. 3) (1).

SCIENCE, VOL. 211

on

Nov

embe

r 14

, 200

7 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

In situ Maturation and OntogeneticChanges

In only a few studies has mineral-ization been traced in situ from its initia-tion to the completion of the final prod-uct. The iron oxide minerals in chitonteeth illustrate the most radical changenoted to date in maturation in situ. Min-eralization of the major lateral teethstarts by deposition of ferrihydrite.Transformation to magnetite is initiatedwithin the ferrihydrite deposits and isfollowed by mass conversion in onetooth row. It continues thereafter on themagnetite surfaces of subsequent toothrows until the magnetite mineralizationphase of the teeth is complete (30).The mature shells of land and fresh-

water gastropods are composed of arago-nite. However, the newly formed cal-cium carbonate deposits are reported tobe vaterite (31). It is not yet knownwhether a similar process occurs in ma-rine gastropods. In mammalian bone anumber of precursor minerals have beensuggested, including amorphous calciumphosphate, tricalcium phosphate, octa-calcium phosphate, and brushite (32).Although many groups of organisms

begin to form mineralized hard parts atthe embryonic stage, there have beenvery few studies of these materials.Among them are several that show on-togenetic changes in hard-part mineral-ogy. In the bivalves studied the larvalshell is mineralized by aragonite. In thepostlarval shell of Crassostrea virginica,however, the mantle-secreted layerscontain calcite and the myostracal lay-ers, formed by the cells of the adductorand auxiliary muscles, are mineralizedby aragonite (33). Mineralogical changesduring- ontogeny have also been ob-served in the pearl oyster Pinctada (34).The otoconia in the two auditory cham-bers of the labyrinth in some sharksshow a different mode of mineralogicalchange during ontogeny (35). Newlyformed otoconia of a different mineral-ogy are added to the initially formed de-posits during ontogeny. In the blue sharkthe otoconia of the embryos are mineral-ized by amorphous hydrous calciumphosphate (ACP), whereas those formedat the juvenile and adult stages are com-posed of aragonite (35). Theref,re, theotoconia consist entirely of ACP in theembryos, ofACP plus aragonite in the ju-veniles, and of aragonite with a trace ofACP in the adults. This ACP fraction canbe detected only in scanning electron mi-crographs by its distinct botryoidal sur-face geometry. In some other species ofsharks belonging to the same superorder,13 MARCH 1981

Fig. 3. (a) Aragonitic otoconia from adult shark Prionace glauca. (b) Aragonite otoconia frompregnant female of P. glauca, showing distinct resorption features.

the otoconia of the embryos are alsocomposed of ACP, whereas the materialformed after the embryonic stage isaragonite. In this case, however, theACP fraction cannot be located in themature adult otoconia and has apparent-ly been resorbed (32). A most spectacu-lar mode of mineralogical change occursin the holothurian species Molpadia in-termedia (36). The mesodermal depositsin early postlarval individuals consist ofcalcitic spicules, similar to those in otherspecies of this class. During subsequentgrowth, most spicules outside the oraland caudal regions are either resorbedcompletely or resorbed and replaced bythe minerals amorphous hydrous ferricphosphate and opal. Thus during ontog-eny these organisms switch over fromcalcium carbonate deposition to ferricphosphate and amorphous silica deposi-tion.

Evolutionary Aspects of

Biomineralization

Recent discoveries of fossils in Pre-cambrian rocks have extended the docu-mentation of life to about 3500 millionyears ago (37). Up to late Vendian time(about 600 million years ago), the fossilsare all soft-bodied organisms. This hadled to the startling realization that miner-alized hard-part formation, so wide-spread in present-day life, is a late devel-opment. However, it was certainly pre-ceded by intracellular mineralization andby extracellular biologically inducedmineral precipitation by at least 2000 mil-lion years. This may be discerned fromthe 4S/32S values of many sedimentarypyrites, which show isotopic fractiona-tion similar to that produced by extantsulfate-reducing bacteria (38), and from

the presence of bacteria that are morpho-logically indistinguishable from extantstrains which precipitate manganeseoxide minerals (39).The inception of mineralized hard-part

formation enormously enhanced thechance of life being preserved as fossils.It is generally recognized that the rapidspread of this innovation to a wide rangeoforganisms and its continuing prolifera-tion thereafter account for the wealth ofbiological records preserved during thepast 600 million years (the Phanerozoicperiod). This has enabled paleontologiststo define the phylogenies of the groups oforganisms that acquired mineralizedhard parts, on the basis of morphologicalchanges of skeletal designs and skeletalmicroarchitectures. Few attempts havebeen made to extract evolutionary infor-mation from the preserved organic con-stituents of fossils, which may provideinsight into underlying processes of bio-chemical evolution (40).Few evolutionary changes in the types

of minerals formed by organisms duringthe Phanerozoic can be recognized atpresent (41). The best documented caseoccurred at the onset of hard-part miner-alization, approximately 600 millionyears ago (2). Initially, two-thirds of theorganisms formed phosphatic (42) andone-third formed calcareous hard parts.Within less than 20 million years (LowerTommotian) about half the organismsformed calcareous hard parts, and in factever since, calcium carbonate poly-morphs have proved to be the mineralsof choice for most organisms. At the out-set calcite was more widely utilized thanaragonite.The anthozoan corals exemplify this

trend. The Paleozoic Rugosa had calciticskeletons, whereas the Scleractinia,which evolved in Triassic time, have

1129

on

Nov

embe

r 14

, 200

7 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

aragonitic skeletons (4, 43). Similartrends may have occurred in theBryozoa and possibly in the Nautiloidea(13). Thus the fossil record indicates atrend in the evolution of calcareousskeletons from phosphates to carbon-ates, and within the carbonates a tenden-cy to replace calcite by aragonite.

Impact on the Biosphere

Magnetite is a common mineralformed at elevated temperatures andpressures in igneous and metamorphicrocks. No magnetite is known to beformed inorganically in the biosphere.Yet organisms ranging from bacteria tovertebrates are capable of forming mag-netite under atmospheric conditions.The magnetite of chiton teeth is a prod-uct of the transformation of a hydrousferric oxide precursor mineral called fer-rihydrite; Even though this is a commonproduct of both biological and inorganicprocesses in the biosphere, the chitonssomehow biochemically mediate thistransformation to magnetite even at at-mospheric temperatures and pressures.The mineral fluorite is widely formed

at high temperatures in hydrothermal sys-tems and is occasionally found in sedi-mentary rocks as a result of diagenesis.Like magnetite, it is not formed in-organically in marine or freshwater envi-ronments. It, too, is synthesized in thehydrosphere by two groups of organisms(Fig. 1). The statoliths or balancing or-gans of a wide range of opossumshrimps (Mysidacea), which occupy ma-rine to freshwater environments, containfluorite (44). Fluorite is also one of themineralization products of the gizzardplates of the gastropods Scaphander andMeloscaphander (45). The amorphousprecursor of this mineral (also not knownto be formed inorganically in the hydro-sphere) is found in the spicules of certaincommon marine nudibranch gastropods,where it occurs together with amorphouscalcium carbonate (44). Several- abun-dant marine protoctists use celestite(SrSO4) to form their skeletons. As sea-water is greatly undersaturated with re-spect to SrSO4, this mineral will cer-tainly not form inorganically, and in factthe skeletons rapidly dissolve after death(46).One important consequence of the

mass production of certain biogenic min-erals is that these minerals are preventedfrom forming inorganically in the hydro-sphere. Silica in the form of opal is per-haps the most striking example. No few-er than 10 (possibly 11) different groupsof organisms are now known to utilize

1130

opal in forming their hard parts (Fig. 1).The diatoms, silicoflagellates, and radio-larians are primarily responsible for thehuge quantities of biogenic opal formedin the oceans. The result is that seawateris undersaturated with respect to opal(3). In their absence, inorganic opal pre-cipitation would take place in theoceans. A less extreme but quan-titatively very important example of bio-genic mineral formation in under-saturated waters is calcium carbonate inthe form of aragonite and calcite. Onlythe upper portions of the ocean are satu-rated with respect to calcite and arago-nite (47). Organisms, however, synthe-size aragonite and calcite at all depths inthe oceans, even in the deepest trenches.In all these cases it appears that the or-ganic sheaths that envelop the mineralconstituents prevent their dissolutionduring life (48).

In order to evaluate the impact of thebiomineralization process on the bio-sphere, it is important to be able to dis-tinguish biogenic from inorganicallyformed minerals. Furthermore, this willimprove our ability to recognize biogenicminerals in the sedimentary record andperhaps even allow us to detect extra-terrestrial life. In certain cases, the merepresence of the mineral itself is an in-dication that it is of biogenic origin. Mag-netite is an unequivocal example. Bio-genic minerals usually adopt character-istic forms defined by the geometry oftheir organic matrix frameworks. Thefluorite crystals in mysid statoliths are al-ways acicular in shape, whereas in-organic precipitates are distinctly dif-ferent (3). In fact, minute inorganicallyformed acicular fluorite crystals have on-ly once been observed, in hydrothermalvug fillings (49). Perhaps the most tellingdifference between biological and in-organically formed minerals is their so-called geochemical properties. Orga-nisms are capable of controlling the traceelement and isotopic compositions of theminerals they form, so that these are outof equilibrium with the environment (50).Examples are the strontium and magne-sium contents of most mollusk shells (51)and the stable oxygen and carbon isotop-ic compositions of coral skeletons (52).

Concluding Remarks

The rapid progress made in this fieldduring the past decade or so only empha-sizes the fact that much more has yet tobe discovered. We have hardly begun toinvestigate biomineralization processesin the monerans and the fungi. To date,there are only a few well-documented

examples where the initial mineralizationdeposit differs from the mature form, orwhere changes in mineral type occurduring development. We are now awarethat such processes do occur, but wehave yet to determine whether they arewidespread.

This article demonstrates that theproducts of biomineralization are unex-pectedly diverse. They may be formed,resorbed, modified, or redeposited. Thesheer complexity raises the question ofwhether underlying common strategieshave been adopted by different groups oforganisms. The widespread use of a pre-constructed organic framework in whichthe minerals grow is in itself an in-dication of the existence of a commonapproach toward biomineralization.Properties of these organic matrices maywell provide insights into such basicmechanisms.

References and Notes

1. H. A. Lowenstam, in Biogeochemistry ofAminoAcids, P. E. Hare, T. C. Hoering, K. King, Eds.(Wiley, New York, 1980), p. 3.

2. __ and L. Margulis, BioSystems 12, 27(1980).

3. H. A. Lowenstam, in The Sea, E. D. Goldberg,Ed. (Wiley, New York, 1974). vol. 5, p. 715.

4. __ , in The Earth Sciences: Problems andProgress in Current Research, T. W. Donnelly,Ed. (Univ. of Chicago Press, Chicago, 1963), p.137.

5. __ , Geol. Soc. Am. Bull. 73, 435 (1962).6. R. Blakemore, Science 190, 377 (1975).7. R. B. Frankel, R. P. Blakemore, R. S. Wolfe,

ibid. 203, 1355 (1979).8. J. L. Gould, J. L. Kirschvink, K. S. Deffeyes,

ibid. 201, 1026 (1978).9. C. Walcott, J. L. Gould, J. L. Kirschvink, ibid.

205, 1027 (1979).10. M. A. Borowitzka, A. W. D. Larkum, C. E.

Nockolds, Phycologia 13, 195 (1974); L. Bohmand D. Futterer, J. Phycol. 14, 486 (1978).

11. L. G. M. Baas-Becking, Ann. Bot. (London) 39,613 (1925); M. B. Goldhaber and I. R. Kaplan, inThe Sea, E. D. Goldberg, Ed. (Wiley, NewYork, 1974), vol. 5, p. 569; K. N. Nealson, inEnvironmental Biogeochemistry and Geomicro-biology, W. E. Krumbein, Ed. (Ann Arbor Sci-ence, Ann Arbor, Mich., 1978), vol. 3, p. 847.

12. W. E. Krumbein, in Biogeochemical Cycling ofMineral-Forming Elements, P. A. Trudinger andD. J. Swain, Eds. (Elsevier, Amsterdam, 1979),p. 47.

13. 0. B. Boggild, K. Dan. Vidensk. Selsk. Skr. Na-turvidensk. Math. Afd. 2, 233 (1930).

14. S. Wise, Eclogae Geol. Helv. 63, 775 (1970).15. Ch. Gregoire, J. Biophys. Biochem. Cytol. 3,

797 (1957).16. See review by G. Krampitz and W. Witt, Top.

Curr. Chem. 78, 58 (1979).17. R. P. Blakemore, D. Maratea, R. S. Wolfe, J.

Bacteriol. 140, 720 (1979).18. K. M. Wilbur and N. Watabe, Ann. N. Y. Acad.

Sci. 109, 82 (1963).19. N. Watabe, V. R. Meenakshi, P. L. Black-

welder, E. M. Kurtz, D. G. Dunkelberger, inThe Mechanisms of Mineralization in the In-vertebrates and Plants, N. Watabe and K. M.Wilbur, Eds. (Univ. of South Carolina Press,Columbia, 1974), p. 283.

20. M. H. Nesson, thesis, California Institute ofTechnology, Pasadena (1968).

21. This mineral was identified by K. M. Towe andW. F. Bradley [1. Colloid Interface Sci. 24, 384(1967)] and named by F. V. Chukhrov [Int.Geol. Rev. 16, 1131 (1974)]. A different chemicalformula for mammalian blood ferritin micellehas been given by P. M._Harrison, F. A. Fish-bach, T. G. Hay, and G. H. Haggis [Nature(London) 216, 1188 (1967)].

22. K. M. Towe, H. A. Lowenstam, M. H. Nesson,Science 142, 63 (1963).

23. Part of the blood ferritin supplies iron to themyoglobin of the radular musculature. How-ever, the oxygen carrier of the blood is hemo-cyanin.

SCIENCE, VOL. 211

on

Nov

embe

r 14

, 200

7 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

24. K. M. Towe and H. A. Lowenstam, J. Ultra-struct. Res. 17, 1 (1967).

25. H. A. Lowenstam, Science 156, 1373 (1967).26. __, personal observations.27. , personal observations.28. G. E. MacGinitie and N. MacGinitie, Natural

History of Marine Animals (McGraw-Hill, NewYork, 1949), p. 354.

29. M. A. Crenshaw and J. M. Neff, Am. Zool. 9,881 (1969).

30. J. L. Kirschvink and H. A. Lowenstam, EarthPlanet. Sci. Lett. 44, 193 (1979).

31. K. G. Levetzow, Jena. Z. Naturwiss. 68, 41(1932); E. Kessel, Z. Morphol. Oekol. Tiere 27,9 (1933).

32. G. L. Wendt and A. H. Clarke, J. Am. Chem.Soc. 45, 881 (1923); W. E. Brown, J. R. Lehr, J.P. Smith, A. W. Frazier, ibid. 79, 5318 (1957); E.D. Eanes, I. H. Gillessen, A. S. Posner, Nature(London) 208, 365 (1965); E. D. Eanes, J. D.Termine, M. U. Nylen, Calcif. Tissue Res. 12,143 (1973).

33. H. B. Stenzel, Science 142, 232 (1963); ibid. 145,155 (1964).

34. I. Kobayashi, in The Mechanisms ofBiomineral-ization in Animals and Plants, M. Omori and N.Watabe, Eds. (Tokai Univ. Press, Tokyo, 1980),p. 145.

35. H. A. Lowenstam and J. E. Fitch, Geol. Soc.Am. Abstr. 10, 114 (1978).

36. H. A. Lowenstam and G. R. Rossman, Chem.Geol. 15, 15 (1975).

37. S. M. Awramik, J. W. Schopf, M. R. Walter, R.Buick, Science, in press.

38. J. Monster, P. W. U. Appel, H. G. Thode, M.Schidlowski, C. M. Carmichael, D. Bridgewa-ter, Geochim. Cosmochim. Acta 43,405 (1979).

39. M. D. Muir, in Environmental Biogeochemistryand Geomicrobiology, W. E. Krumbein, Ed.

(Ann Arbor Science, Ann Arbor, Mich., 1978),vol. 3, p. 937.

40. S. Weiner, H. A. Lowenstam, L. Hood, Proc.Natl. Acad. Sci. U.S.A. 73, 2541(1976); S. Wei-ner and H. A. Lowenstam, in BiogeochemistryofAmino Acids, P. E. Hare, T. C. Hoering, K.King, Eds. (Wiley, New York, 1980), p. 95; S.Weiner, H. A. Lowenstam, B. Taborek, L.Hood, Paleobiology 5, 144 (1979); P. West-broek, P. H. van der Meide, J. A. van der Wey-Kloppers, R. J. van der Sluis, J. W. de Leeuw,E. W. de Jong, ibid., p. 151.

41. For a more detailed discussion of the evolution-ary trends in biogenic mineral formation, see (2,3).

42. This trend can be clearly seen in individual lin-eages within a phylum such as the Ostracoda,where substitution of calcite for carbonate apa-tite occurred some 500 million years ago (Cam-brian-Ordovician boundary) [H. A. Lowenstam,Geol. Soc. Am. Abstr. 5, 719 (1972)]. At the phy-lum level, the first coelenterates to acquire min-eralized hard parts (Conulariina) utilized a phos-phate, whereas all subsequently evolving groupsformed, and still do form, carbonate skeletons.

43. F. G. Stehli, Science 123, 1031 (1956).44. H. A. Lowenstam and D. McConnell, ibid. 162,

1496 (1968).45. H. A. Lowenstam, Chem. Geol. 9, 153 (1972).46. W. H. Berger, Science 159, 1237 (1968); E. M.

Bottazzi, B. Schreiber, V. T. Bowen, Limnol.Oceanogr. 16, 677 (1971).

47. W. H. Berger, Mar. Geol. 8, 11 (1970).48. __, Science 156, 383 (1967).49. R. Jahns, personal communication.50. Factors known to account for this phenomena

are (i) the metabolism and in particular the min-eral metabolism of a particular species, (ii) theparticular biochemical processes involved in

biomineralization, (iii) ecological factors such astemperature, salinity, water chemistry of the ex-ternal environment, and possibly hydrostaticpressure, (iv) food sources, and (v) metabolicwaste products of symbionts, where present.

51. D. J. J. Kinsman, J. Sediment. Petrol. 39, 486(1969); H. A. Lowenstam, in Isotopic and Cos-mic Chemistry, H. Craig, S. L. Miller, G. J.Wasserburg, Eds. (North Holland, Amsterdam,1964), p. 114; H. A. Lowenstam, in Recent Re-searches in the Fields of Hydrosphere, Atmo-sphere and Nuclear (Geochemistry, Y. Miyakeand T. Koyama, Eds. (Maruzen, Tokyo, 1964),p. 373. For more information see J. D. Milliman,Marine Carbonates, Recent Sedimentary Car-bonates (Springer-Verlag, New York, 1974),part 1, p. 138.

52. H. Craig, Geol. Soc. Am. Mem. 67, 239 (1957);J. N. Weber and P. M. J. Woodhead, Chem.Geol. 6, 93 (1970); J. Geophys. Res. 77, 463(1971); J. Erez, Nature (London) 273, 199(1978).

53. I am deeply indebted to S. Weiner for many sug-gestions and in particular his skillful editing ofthe manuscript. P. A. Sandberg and S. Weinermade available the scanning electron micro-grphs shown in Fig. 1. H. Lemche, K. Towe;K. Nealson, L. Margulis, and J. Oro providedliterature references. M. Dekker's skill in pre-paring samples has for many years contributedto my identification of numerous minerals. Datawhich I obtained were largely by-products of in-vestigations supported by National ScienceFoundation grants. Parts of this article werewritten during an extended visit to the Weiz-mann Institute of Science, Rehovot, Israel. Con-tribution No. 3530, Division of Geological andPlanetary Sciences, California Institute-of Tech-nology.

I wish to address three major topics.First, I want to share with you somerecent landmark achievements in funda-mental scientific research. At the sametime, I hope to show that gathering newknowledge is not only rewarding in itselfbut represents an investment; the aggre-gate payoff will surely be great, eventhough the future applications are some-times hard to foresee. I will put the caseas strongly as I can that continuous andsubstantial support for basic research isa necessity for all great nations. At thesame time, I will assert my own convic-tion that support of the best research andeducation in the sciences and engineer-ing is and must remain the central pur-pose of the National Science Foundation(NSF).Next, I will take up the role of re-

search that directly addresses some ofthe pressing needs of our society, suchSCIENCE, VOL. 211, 13 MARCH 1981

as energy, natural resources, and pro-ductivity. Much of the research that ad--dresses such "real-world" problems isquite fundamental in character. We maycall it applied research because we seeclearly the problem- it addresses, butwhether the distinction between basicand applied research represents in everyinstance a real difference I will leave toyour judgment.

Finally, I will turn to two areas that Ibelieve require closer attention and moreintense effort in the 1980's. One of theseis engineering science, that is, the theo-retical body of knowledge and techniqueunderlying the practice of engineering.Much of this work is every bit as funda-mental in character as correspondingwork in the disciplinary sciences. Theother area is education in the science andengineering disciplines. It is apparentthat we will face increasing difficulties in

securing talented and well-trained mindsfor the professional disciplines unless-veact promptly to improve our educationprocesses, not only at the universitylevel but in the elementary and second-ary schools as well.

Recent Scientific Achievements

In considering major achievements offundamental s7cientific inquiry in recentyears, I have chosen a handful of exam-ples that I am familiar with because ofNSF's involvement or interest in them. Icould as easily have chosen many oth-ers, equally exciting, if there were spaceto discuss them. Science is virtually ex-ploding with new developments, and it isa great privilege and responsibility to berejoining NSF during these heady times.

Certainly a historic advance in ourtimes has been the recent progress to-ward the theoretical unification of theseveral forces in nature. Today we havean extremely persuasive conceptual ar-gument, supported by experimental evi-dence, that the electromagnetic andweak nuclear forces are different expres-sions of a single phenomenon. More-over, we sense a possibility that thisunification can be extended to the strongnuclear force as well. Gravity-the

The author is director of the National ScienceFoundation, Washington, D.C. 20550. This article isbased on his address at the AAAS Annual Meeting,Toronto, Canada, 3 to 8 January 1981.

0036-8075/81/0313-1131$01.50/0 Copyright C 1981 AAAS

The National Science FoundationLooks to the Future

John B. Slaughter

1131

on

Nov

embe

r 14

, 200

7 w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from