Cell, Vol. 65, 531-533,
May 17, 1991, Copyright
0 1991 by Cell Press
Origin of Life- Facing Up to the Physical Setting Norman FL Pace Department of Biology Indiana University Bloomington, Indiana 47405
There has been an enormous amount of speculation on the origin of life, with little heed to constraints that might be imposed by the physical setting. Indeed, few data have been available on which to base constraints. Results and ideas over the past decade have, however, begun to sharpen our image of that distant past. The purpose of this article is to review what can be inferred about the environment in which life arose. Lines of inference stem from two types of records. A physical record of geological and astronomical data provides information on the character of Earth’s surface during the epoch in which life arose. A biological record of quantitative evolutionary relationships among modern organisms documents early Earth’s environment because it points to some properties of the first cells. It seems reasonable that the environment in which life arose was also that of the earliest organisms. The lines of inference from these recordsare beginning toconverge. The emerging picture is that of a harsh, high-temperature environment. Speculation on the origin of life must accommodate this physical setting. When Did Life Begin? Earth was well formed by 4.5 billion years ago, but the time at which life began remains uncertain. The evidence rests on two types of fossil record in ancient geological strata. Hard fossils include microscopic structures that resemble organisms in thin sectionsof rocks, and macroscopic‘stromatolites,” laminated mounds with structures of a type often associated with mats of microorganisms (Schopf and Walter, 1983). Another type of fossil record relies on the fact that biological reactions commonly select for lighter isotopes of carbon (reviewed by Schidlowski, 1987). Therefore, the enrichment of 12C/‘3C in organics extracted from ancient rocks, relative to abiogenic carbon, is some evidence for the occurrence of life. The very early traces of life are sketchy, however, because there are only a few remnants of the early crust in which to prospect. The action of plate tectonics has erased, by subduction, most early strata. Carbon isotope fractionation studies show enrichment for YY3C in kerogens (carbon-containing polymers) extracted from some of the oldest known rocks, about 3.8 billion years in age, from the lsua supracrustal geologic formation (Greenland). Such evidence for life is compromised, however, because thermal processes also can cause isotopic fractionation, and the lsua rocks have been deeply buried and heated at least once. Putative microfossils and stromatolites of the 3.5billion-year-old Warrawoona (Australia) and Swaziland (South Africa) geological formations are generally considered more convincing as evidence for early life (reviewed by Schopf and Walter, 1983). This evidence, however, is also compromised. Based on the character of kerogen
Minireview
associated with putative microfossils and the morphology of stromatolites, there is reasonable doubt as to the biological origin of those features (Buick, 1990). The earliest unambiguous fossil microorganisms and stromatolites are in rocks of about 3.1 billion years in age (Mason and Von Brunn, 1977). Our notion as to the time of life’s emergence clearly is tenuous. Certainly it occurred before 3-3.5 billion years ago; however, 4 billion years ago or even earlier seems possible, depending on the compatibility of the physical conditions at that time. What Was the Character of Earth at the Tim of the Origfn of Life? It is now clear that the surface of Earth was molten during the late stages of its formation, say, 4.2-4.5 billion years ago. The character of the surface would have evolved dramatically as the crust formed and cooled during the first 0.5-l billion years. Few data regarding that scenario are available, but theoretical studies have provided credible models for the character of the late prebiotic atmosphere (Chang et al., 1983). The details of calculations depend on the model, but the temperature (500°C-1000%) and pressure (e.g., 500 atm H20, 40 atm CO/C02, 10 atm Hz) would have been high. The surface of Earth was reducing ([Fe*] > > [Fe”]), but the atmosphere would have been mildly oxidizing ([CO/CO,] > > [CH,], regardless of model). The most life-forbidding of these conditions would be high temperature: it seems unlikely that complex organic molecules could develop at 500%. Earth’s surface presumably had to cool to some extent before organic complexity could develop, but it is unclear how much it was necessary to cool in order to provide an environment suitable for the origin of life. The only clue to the rate of cooling from the geological record is that some of the earliest crustal remnants, such as the Isua rocks, are sediments and therefore there must have been liquid water by 3.8 billion years ago. Indeed, it seems likely that Earth’s entire surface was covered with water. However, the occurrence of liquid water reveals little about temperature. Depending on the atmospheric pressure, the temperature of that water may have been substantially greater than 100%. The hydrosphere would have been in dynamic interaction with the hot, forming crust, through convective hydrothermal circulation. A reasonable uppertemperature at which life could arise would be the upper temperature at which modern organisms occur. That boundary is not yet established. Several organisms have been grown in the laboratory at lOO“C110% (reviewed by Stetter et al., 1990) and there is evidence for the occurrence of microbial ecosystems at somewhat higher temperatures (possibly as high as 140°C-150%). If such extremely thermophilic organisms were the earliest life-forms, as now seems likely (see below), then it is also reasonable to conclude that life arose at temperatures far higher than that of today’satmosphere. What Was the Nature of the Ear/lest Cells? Insights into the nature of the first organisms are emerging from evolutionary studies of modern organisms. Our un-
Cell 532
derstanding of the evolutionary course of life has improved enormously over the past few decades. In good part this has been due to studies in molecular phylogeny, establishing relationships of organisms by the comparison of macromolecular sequences: amino acids in proteins or nucleotides in DNA or RNA. The number of differences in homologous sequences from different organisms is a measure of the evolutionary distance that separates the organisms. The ribosomal RNAs (rRNAs) have been especially useful for molecular phylogeny because they are of fundamental importance and so occur in all organisms (reviewed by Pace et al., 1986). Phylogenetic trees based on rRNA or protein sequences show that all known organisms are related, establishing that there must have been a singular (not necessarily cellular) common ancestor of all modern life. Extant organisms are seen to derive from three primary lineages, previously dubbed eukaryotes, eubacteria and archaebacteria, and recently proposed to be designated as Eucarya, Bacteria, and Arches, respectively (Woese et al., 1990). The exact “root” of the universal tree, the position of the hypothetical common ancestor, cannot be identified with confidence because there is no “outgroup” with which to determine the deepest branching. However, recent comparisons of molecules that likely arose from gene duplications before the last common ancestor suggest that the root of the tree lies on the eubacterial branch (reviewed by Woese et al., 1990). Although the phylogenetic record is based on modern organisms, some properties of hypothetical ancestors may be inferred because an ancestor is expected to have had properties that occur in all (or most) of its descendants. We can therefore infer that the most recent common ancestor of all known life had DNA, ATP, an essentially modern translation apparatus, and all such other properties that occur in all modern organisms. However, using the universal properties of all extant life-forms tells us nothing about specific qualities of the common ancestor that might reflect the nature of the environment it occupied. One way to go further with only the perspective offered by modern organisms is to consider the properties of the phylogenetic group that seems most closely related to the earliest organisms. It may be possible to identify this group from consideration of rates of evolutionary change. Sequence comparisons show that different evolutionary lineages have not evolved at the same rates. Since their last common ancestor, archaebacterial rRNAson average have accumulated substantially fewer mutations than the rRNAs of either eukaryotes or eubacteria (Woese, 1987). Therefore, in terms of rRNA sequence divergence, archaebacteria are more closely related to the common ancestor than are eukaryotes or eubacteria. Some lineages of archaebacteria have evolved particularly slowly to the present day. Their modern representatives, such genera as Pyrodictium, Thermoproteus, and Pyrobaculum (Stetter et al., 1990) thus would be the most closely related of extant organisms to the common ancestor of all life. All of these evolutionarily primitive organisms are similar in their requirement for extremely high growth temperatures, utilization of geochemical energy sources
such as sulfur and molecular hydrogen, fixation of Con, their fundamentally anaerobic nature, and the occupancy of geothermal environments. It is a leap from phylogeny to physiology, but following the logic that commonalities in descendents were properties of the ancestor, these common properties of the least-evolved archaebacteria also would be properties of the earliest archaebacteria. Stretching that logic further suggests that those properties also pertained to the most recent common ancestor of all modern life (Woese, 1987). This description of the universal ancestor as a chemosynthetic thermophile differs fundamentally from classic notions, which have presumed the first organisms to have been heterotrophic (relying on reduced carbon compounds), at low temperatures. If the environment in which life arose was similar to that occupied by the most recent common ancestor, then life probably arose in a geothermal environment at relatively high temperature. Because of the extensive modification of the surface of Earth by biological as well as geological processes, it seems unlikely that there is a modern remnant of the environment in which life arose. However, modern geothermal settings-for example, submarine hydrothermal vents associated with crustal spreading-are common and possibly mimic the prebiotic environment in many ways (Baross and Hoffman, 1985; Corliss, 1990). Little is known about the microbiology and organic chemistry of these crustal plumbing systems. High-Temperature and Chemical Models for the Origin of Life The origin of life at high temperature places constraints on the nature of the chemistry involved. It is reasonable to presume that life was a consequence of organic chemistry. Modern cells rely upon organic chemistry, and evolution builds upon the preexisting. Early Earth’s surface would have been fairly rich in organics, contributed by infalling matter as the planet coalesced (Anders, 1989). It also has long been recognized that simple organic molecules of the types found in modern cells, for instance, amino acids, nucleic acid bases, and sugars, form readily under aqueous conditions from compounds that likely existed on early Earth (Miller and Orgel, 1974). The origin of life, however, required the polymerization of those simple compounds into functional macromolecules such as nucleic acids and proteins. There has been some success at synthesis of macromolecules in the laboratory under aqueous conditions using exotic condensing reagents (reviewed by Joyce, 1989). but such chemistry is far from the facile reactions that likely sparked life. A major stumbling block is the fact that biological polymers generally are formed by dehydration, the removal of water, to form a peptide bond in the case of protein or a phosphodiester bond in the case of nucleic acids. However, because of the high concentration of water (55 M) in a fully aqueous environment, hydrolysis, not polymerization, is overwhelmingly favored. The origin of life seems inconsistent with fully aqueous chemistry, particularly at high temperature. The inconsistency of aqueous solution chemistry and the origin of life is exemplified by the properties of RNA.
Minireview 533
Current speculation on the nature of the earliest genetic information focuses on RNA because that molecule can serve as an information repository and also can catalyze certain chemical reactions, including formation of internucleotide bonds (Cech, 1987). The discovery of RNA catalysis overcame, in principle, a chicken-and-egg paradox: it could be imagined that the earliest genome, if composed of RNA, could replicate itself without a complex proteinsynthesizing machinery to produce a replicase composed of protein. This notion elicited visions of an “RNA world,” a time during the evolutionary genesis of living things when RNA molecules rather than proteins carried out protobiological reactions, and were replicated by other RNAs (Gilbert, 1988; Weiner and Maizels, 1987). If it ever existed, however, the RNA world was not exposed to free solution. The ribose 2’9H group that renders RNA, in contrast to DNA, catalytic also renders the RNA chain particularly susceptible to hydrolysis. The RNA phosphodiester in aqueous solution is highly labile to hydrolysis promoted by the ribose 2’-CH group, which forms the 2’,3’-cyclic phosphate and breaks the RNA chain. This reaction is accelerated by high pH, high temperature, and the presence of divalent (or other multivalent) cations, which bind to and polarize phosphates, enhancing reactivity. The early hydrosphere would have contained substantial concentrations of many multivalent cations. RNA molecules of much complexity could not have survived if exposed to solution in that environment, particularly if at high temperatures. The RNA-based world would have required a special environment, in the least as protection from hydrolysis. This environment need not have been low in water concentration, only low in water chemical activity. It therefore seems unlikely that the prebiotic milieu was an organicscontaining aqueous soup, as commonly supposed. Rather, life more likely arose in an organic scum that could provide an environment of low water activity. As in modern cells (Fulton, 1982). water comprising the prebiotic milieu could have been immobilized by hydration of polar molecules: the effective water concentration (chemical activity) under such conditions could be far lower than in aqueous solution. We have little grasp on the nature of the organic chemistry that gave rise to the first self-replicating entities and their environment. Classical experimentation along these lines has focused on aqueous chemistry, which seems unlikely to have been exclusively relevant in the prebiotic milieu. Organic surface chemistry is crucial to consider. The involvement of surface chemistry, for instance, on clays (Cairns-Smith, 1985), has long been invoked as potentially important in the prebiotic setting. Pyrite (Wachtershauser, 1988) and basaltic glasses also would have been abundant on early Earth, and could provide surface support for complex chemistry. A novel, theoretically attractive collection of hypothetical reactions between phosphorylated organics associated with surfaces has recently been articulated by Wbhtershluser (1988). This chemistry could result in the propagation and accumulation of biologically noteworthy compounds, and provide for an environment of low water activity in which macromolecular
chemistry could occur. The environment of low water activity probably could accommodate biologically relevant organic chemistry in the temperature range 100°C-200%. Conclusion Findings over the past several years and new approaches (e.g., phylogenetic inferences) promise an increasingly sound experimental base from which to speculate on the origin of life. At the least, models and experimental approaches must be framed in the context of acredible physical and chemical setting. The possibility that life arose in a relatively high temperature, geothermal setting places constraints on speculation and poses new questions. The RNA world, for instance, was possibly a step in the invention of genetic information, an invention necessary for cellularization. However, in the geothermal setting an RNAbased world would have required, for utter survival, a special environment. The discoveries of the double helix and of RNA catalysis solved in principle the problem of replication, but we are left with the next question: How did the instructions for that special environment become encoded in genotype? References Anders,
E. (1989).
Baross.
J. A., and Hoffman,
Buick,
Ft. (199Q.
Nature
342, 255-257.
Palaios
S. E. (1985).
Origins
Cairns-Smith, A. G. (1985). Seven Clues bridge: Cambridge University Press). Cech,
T. R. (1987).
of Life l&327-345.
5, 441-459.
Science
to the Origin
of Life (Cam-
236, 1532-1539.
Chang, S., DesMarais, D., Mack, Ft., Miller, S. L., and Strathern, G. E. (1983). In Earths Earliest Biosphere: Its Origin and Evolution, J. W. Schopf, ed. (Princeton, New Jersey: Princeton University Press), pp. 51-92. Corliss,
J. B. (1990).
Nature
Fulton,
A. B. (1982).
Cell 30, 345-347.
Gilbert, Joyce, Mason,
W. (1988).
Nature
G. F. (1989).
Nature
347, 824. 379, 818. 338, 217-224.
T. Ft., and Von Brunn,
V. (1977).
Nature
Miller, S. L., and Orgel, L. E. (1974). The Origins (Englewood Cliffs, New Jersey: Prentice-Hall). Pace,
N. R., Olsen,
Schidlowski,
G. J., and Woese,
M. (1987).
Annu.
Rev.
266, 47-49. of Life on the Earth
C. R. (1988). Planet.
Cell 45325328.
Sci. 75, 47-72.
Schopf, J. W., and Walter, M. R. (1983). In Earth’s Earliest Biosphere: Its Origin and Evolution, J. W. Schopf, ed. (Princeton, New Jersey: Princeton University Press), pp. 214-239. Stetter, K. O., Fiala, G., Huber, G., Huber, FEMS Microbial. Rev. 75, 117-124. WBchtersh&ser,
G. (1988).
Weiner, A. M., and Maizels, 7383-7387. Woese,
C. R. (1987).
Microbial. N. (1987).
Microbial.
R., and Segerer,
A. (199@.
Rev. 52, 452-484. Proc.
Natl. Acad.
Sci. USA 84,
Rev. 51, 221-271.
Woese, C. R., Kandler, O., and Wheelis, Sci. USA 87.4578-4579.
M. L. (1990).
Proc. Nab. Acad.
May 17, 1991, Copyright
0 1991 by Cell Press
Origin of Life- Facing Up to the Physical Setting Norman FL Pace Department of Biology Indiana University Bloomington, Indiana 47405
There has been an enormous amount of speculation on the origin of life, with little heed to constraints that might be imposed by the physical setting. Indeed, few data have been available on which to base constraints. Results and ideas over the past decade have, however, begun to sharpen our image of that distant past. The purpose of this article is to review what can be inferred about the environment in which life arose. Lines of inference stem from two types of records. A physical record of geological and astronomical data provides information on the character of Earth’s surface during the epoch in which life arose. A biological record of quantitative evolutionary relationships among modern organisms documents early Earth’s environment because it points to some properties of the first cells. It seems reasonable that the environment in which life arose was also that of the earliest organisms. The lines of inference from these recordsare beginning toconverge. The emerging picture is that of a harsh, high-temperature environment. Speculation on the origin of life must accommodate this physical setting. When Did Life Begin? Earth was well formed by 4.5 billion years ago, but the time at which life began remains uncertain. The evidence rests on two types of fossil record in ancient geological strata. Hard fossils include microscopic structures that resemble organisms in thin sectionsof rocks, and macroscopic‘stromatolites,” laminated mounds with structures of a type often associated with mats of microorganisms (Schopf and Walter, 1983). Another type of fossil record relies on the fact that biological reactions commonly select for lighter isotopes of carbon (reviewed by Schidlowski, 1987). Therefore, the enrichment of 12C/‘3C in organics extracted from ancient rocks, relative to abiogenic carbon, is some evidence for the occurrence of life. The very early traces of life are sketchy, however, because there are only a few remnants of the early crust in which to prospect. The action of plate tectonics has erased, by subduction, most early strata. Carbon isotope fractionation studies show enrichment for YY3C in kerogens (carbon-containing polymers) extracted from some of the oldest known rocks, about 3.8 billion years in age, from the lsua supracrustal geologic formation (Greenland). Such evidence for life is compromised, however, because thermal processes also can cause isotopic fractionation, and the lsua rocks have been deeply buried and heated at least once. Putative microfossils and stromatolites of the 3.5billion-year-old Warrawoona (Australia) and Swaziland (South Africa) geological formations are generally considered more convincing as evidence for early life (reviewed by Schopf and Walter, 1983). This evidence, however, is also compromised. Based on the character of kerogen
Minireview
associated with putative microfossils and the morphology of stromatolites, there is reasonable doubt as to the biological origin of those features (Buick, 1990). The earliest unambiguous fossil microorganisms and stromatolites are in rocks of about 3.1 billion years in age (Mason and Von Brunn, 1977). Our notion as to the time of life’s emergence clearly is tenuous. Certainly it occurred before 3-3.5 billion years ago; however, 4 billion years ago or even earlier seems possible, depending on the compatibility of the physical conditions at that time. What Was the Character of Earth at the Tim of the Origfn of Life? It is now clear that the surface of Earth was molten during the late stages of its formation, say, 4.2-4.5 billion years ago. The character of the surface would have evolved dramatically as the crust formed and cooled during the first 0.5-l billion years. Few data regarding that scenario are available, but theoretical studies have provided credible models for the character of the late prebiotic atmosphere (Chang et al., 1983). The details of calculations depend on the model, but the temperature (500°C-1000%) and pressure (e.g., 500 atm H20, 40 atm CO/C02, 10 atm Hz) would have been high. The surface of Earth was reducing ([Fe*] > > [Fe”]), but the atmosphere would have been mildly oxidizing ([CO/CO,] > > [CH,], regardless of model). The most life-forbidding of these conditions would be high temperature: it seems unlikely that complex organic molecules could develop at 500%. Earth’s surface presumably had to cool to some extent before organic complexity could develop, but it is unclear how much it was necessary to cool in order to provide an environment suitable for the origin of life. The only clue to the rate of cooling from the geological record is that some of the earliest crustal remnants, such as the Isua rocks, are sediments and therefore there must have been liquid water by 3.8 billion years ago. Indeed, it seems likely that Earth’s entire surface was covered with water. However, the occurrence of liquid water reveals little about temperature. Depending on the atmospheric pressure, the temperature of that water may have been substantially greater than 100%. The hydrosphere would have been in dynamic interaction with the hot, forming crust, through convective hydrothermal circulation. A reasonable uppertemperature at which life could arise would be the upper temperature at which modern organisms occur. That boundary is not yet established. Several organisms have been grown in the laboratory at lOO“C110% (reviewed by Stetter et al., 1990) and there is evidence for the occurrence of microbial ecosystems at somewhat higher temperatures (possibly as high as 140°C-150%). If such extremely thermophilic organisms were the earliest life-forms, as now seems likely (see below), then it is also reasonable to conclude that life arose at temperatures far higher than that of today’satmosphere. What Was the Nature of the Ear/lest Cells? Insights into the nature of the first organisms are emerging from evolutionary studies of modern organisms. Our un-
Cell 532
derstanding of the evolutionary course of life has improved enormously over the past few decades. In good part this has been due to studies in molecular phylogeny, establishing relationships of organisms by the comparison of macromolecular sequences: amino acids in proteins or nucleotides in DNA or RNA. The number of differences in homologous sequences from different organisms is a measure of the evolutionary distance that separates the organisms. The ribosomal RNAs (rRNAs) have been especially useful for molecular phylogeny because they are of fundamental importance and so occur in all organisms (reviewed by Pace et al., 1986). Phylogenetic trees based on rRNA or protein sequences show that all known organisms are related, establishing that there must have been a singular (not necessarily cellular) common ancestor of all modern life. Extant organisms are seen to derive from three primary lineages, previously dubbed eukaryotes, eubacteria and archaebacteria, and recently proposed to be designated as Eucarya, Bacteria, and Arches, respectively (Woese et al., 1990). The exact “root” of the universal tree, the position of the hypothetical common ancestor, cannot be identified with confidence because there is no “outgroup” with which to determine the deepest branching. However, recent comparisons of molecules that likely arose from gene duplications before the last common ancestor suggest that the root of the tree lies on the eubacterial branch (reviewed by Woese et al., 1990). Although the phylogenetic record is based on modern organisms, some properties of hypothetical ancestors may be inferred because an ancestor is expected to have had properties that occur in all (or most) of its descendants. We can therefore infer that the most recent common ancestor of all known life had DNA, ATP, an essentially modern translation apparatus, and all such other properties that occur in all modern organisms. However, using the universal properties of all extant life-forms tells us nothing about specific qualities of the common ancestor that might reflect the nature of the environment it occupied. One way to go further with only the perspective offered by modern organisms is to consider the properties of the phylogenetic group that seems most closely related to the earliest organisms. It may be possible to identify this group from consideration of rates of evolutionary change. Sequence comparisons show that different evolutionary lineages have not evolved at the same rates. Since their last common ancestor, archaebacterial rRNAson average have accumulated substantially fewer mutations than the rRNAs of either eukaryotes or eubacteria (Woese, 1987). Therefore, in terms of rRNA sequence divergence, archaebacteria are more closely related to the common ancestor than are eukaryotes or eubacteria. Some lineages of archaebacteria have evolved particularly slowly to the present day. Their modern representatives, such genera as Pyrodictium, Thermoproteus, and Pyrobaculum (Stetter et al., 1990) thus would be the most closely related of extant organisms to the common ancestor of all life. All of these evolutionarily primitive organisms are similar in their requirement for extremely high growth temperatures, utilization of geochemical energy sources
such as sulfur and molecular hydrogen, fixation of Con, their fundamentally anaerobic nature, and the occupancy of geothermal environments. It is a leap from phylogeny to physiology, but following the logic that commonalities in descendents were properties of the ancestor, these common properties of the least-evolved archaebacteria also would be properties of the earliest archaebacteria. Stretching that logic further suggests that those properties also pertained to the most recent common ancestor of all modern life (Woese, 1987). This description of the universal ancestor as a chemosynthetic thermophile differs fundamentally from classic notions, which have presumed the first organisms to have been heterotrophic (relying on reduced carbon compounds), at low temperatures. If the environment in which life arose was similar to that occupied by the most recent common ancestor, then life probably arose in a geothermal environment at relatively high temperature. Because of the extensive modification of the surface of Earth by biological as well as geological processes, it seems unlikely that there is a modern remnant of the environment in which life arose. However, modern geothermal settings-for example, submarine hydrothermal vents associated with crustal spreading-are common and possibly mimic the prebiotic environment in many ways (Baross and Hoffman, 1985; Corliss, 1990). Little is known about the microbiology and organic chemistry of these crustal plumbing systems. High-Temperature and Chemical Models for the Origin of Life The origin of life at high temperature places constraints on the nature of the chemistry involved. It is reasonable to presume that life was a consequence of organic chemistry. Modern cells rely upon organic chemistry, and evolution builds upon the preexisting. Early Earth’s surface would have been fairly rich in organics, contributed by infalling matter as the planet coalesced (Anders, 1989). It also has long been recognized that simple organic molecules of the types found in modern cells, for instance, amino acids, nucleic acid bases, and sugars, form readily under aqueous conditions from compounds that likely existed on early Earth (Miller and Orgel, 1974). The origin of life, however, required the polymerization of those simple compounds into functional macromolecules such as nucleic acids and proteins. There has been some success at synthesis of macromolecules in the laboratory under aqueous conditions using exotic condensing reagents (reviewed by Joyce, 1989). but such chemistry is far from the facile reactions that likely sparked life. A major stumbling block is the fact that biological polymers generally are formed by dehydration, the removal of water, to form a peptide bond in the case of protein or a phosphodiester bond in the case of nucleic acids. However, because of the high concentration of water (55 M) in a fully aqueous environment, hydrolysis, not polymerization, is overwhelmingly favored. The origin of life seems inconsistent with fully aqueous chemistry, particularly at high temperature. The inconsistency of aqueous solution chemistry and the origin of life is exemplified by the properties of RNA.
Minireview 533
Current speculation on the nature of the earliest genetic information focuses on RNA because that molecule can serve as an information repository and also can catalyze certain chemical reactions, including formation of internucleotide bonds (Cech, 1987). The discovery of RNA catalysis overcame, in principle, a chicken-and-egg paradox: it could be imagined that the earliest genome, if composed of RNA, could replicate itself without a complex proteinsynthesizing machinery to produce a replicase composed of protein. This notion elicited visions of an “RNA world,” a time during the evolutionary genesis of living things when RNA molecules rather than proteins carried out protobiological reactions, and were replicated by other RNAs (Gilbert, 1988; Weiner and Maizels, 1987). If it ever existed, however, the RNA world was not exposed to free solution. The ribose 2’9H group that renders RNA, in contrast to DNA, catalytic also renders the RNA chain particularly susceptible to hydrolysis. The RNA phosphodiester in aqueous solution is highly labile to hydrolysis promoted by the ribose 2’-CH group, which forms the 2’,3’-cyclic phosphate and breaks the RNA chain. This reaction is accelerated by high pH, high temperature, and the presence of divalent (or other multivalent) cations, which bind to and polarize phosphates, enhancing reactivity. The early hydrosphere would have contained substantial concentrations of many multivalent cations. RNA molecules of much complexity could not have survived if exposed to solution in that environment, particularly if at high temperatures. The RNA-based world would have required a special environment, in the least as protection from hydrolysis. This environment need not have been low in water concentration, only low in water chemical activity. It therefore seems unlikely that the prebiotic milieu was an organicscontaining aqueous soup, as commonly supposed. Rather, life more likely arose in an organic scum that could provide an environment of low water activity. As in modern cells (Fulton, 1982). water comprising the prebiotic milieu could have been immobilized by hydration of polar molecules: the effective water concentration (chemical activity) under such conditions could be far lower than in aqueous solution. We have little grasp on the nature of the organic chemistry that gave rise to the first self-replicating entities and their environment. Classical experimentation along these lines has focused on aqueous chemistry, which seems unlikely to have been exclusively relevant in the prebiotic milieu. Organic surface chemistry is crucial to consider. The involvement of surface chemistry, for instance, on clays (Cairns-Smith, 1985), has long been invoked as potentially important in the prebiotic setting. Pyrite (Wachtershauser, 1988) and basaltic glasses also would have been abundant on early Earth, and could provide surface support for complex chemistry. A novel, theoretically attractive collection of hypothetical reactions between phosphorylated organics associated with surfaces has recently been articulated by Wbhtershluser (1988). This chemistry could result in the propagation and accumulation of biologically noteworthy compounds, and provide for an environment of low water activity in which macromolecular
chemistry could occur. The environment of low water activity probably could accommodate biologically relevant organic chemistry in the temperature range 100°C-200%. Conclusion Findings over the past several years and new approaches (e.g., phylogenetic inferences) promise an increasingly sound experimental base from which to speculate on the origin of life. At the least, models and experimental approaches must be framed in the context of acredible physical and chemical setting. The possibility that life arose in a relatively high temperature, geothermal setting places constraints on speculation and poses new questions. The RNA world, for instance, was possibly a step in the invention of genetic information, an invention necessary for cellularization. However, in the geothermal setting an RNAbased world would have required, for utter survival, a special environment. The discoveries of the double helix and of RNA catalysis solved in principle the problem of replication, but we are left with the next question: How did the instructions for that special environment become encoded in genotype? References Anders,
E. (1989).
Baross.
J. A., and Hoffman,
Buick,
Ft. (199Q.
Nature
342, 255-257.
Palaios
S. E. (1985).
Origins
Cairns-Smith, A. G. (1985). Seven Clues bridge: Cambridge University Press). Cech,
T. R. (1987).
of Life l&327-345.
5, 441-459.
Science
to the Origin
of Life (Cam-
236, 1532-1539.
Chang, S., DesMarais, D., Mack, Ft., Miller, S. L., and Strathern, G. E. (1983). In Earths Earliest Biosphere: Its Origin and Evolution, J. W. Schopf, ed. (Princeton, New Jersey: Princeton University Press), pp. 51-92. Corliss,
J. B. (1990).
Nature
Fulton,
A. B. (1982).
Cell 30, 345-347.
Gilbert, Joyce, Mason,
W. (1988).
Nature
G. F. (1989).
Nature
347, 824. 379, 818. 338, 217-224.
T. Ft., and Von Brunn,
V. (1977).
Nature
Miller, S. L., and Orgel, L. E. (1974). The Origins (Englewood Cliffs, New Jersey: Prentice-Hall). Pace,
N. R., Olsen,
Schidlowski,
G. J., and Woese,
M. (1987).
Annu.
Rev.
266, 47-49. of Life on the Earth
C. R. (1988). Planet.
Cell 45325328.
Sci. 75, 47-72.
Schopf, J. W., and Walter, M. R. (1983). In Earth’s Earliest Biosphere: Its Origin and Evolution, J. W. Schopf, ed. (Princeton, New Jersey: Princeton University Press), pp. 214-239. Stetter, K. O., Fiala, G., Huber, G., Huber, FEMS Microbial. Rev. 75, 117-124. WBchtersh&ser,
G. (1988).
Weiner, A. M., and Maizels, 7383-7387. Woese,
C. R. (1987).
Microbial. N. (1987).
Microbial.
R., and Segerer,
A. (199@.
Rev. 52, 452-484. Proc.
Natl. Acad.
Sci. USA 84,
Rev. 51, 221-271.
Woese, C. R., Kandler, O., and Wheelis, Sci. USA 87.4578-4579.
M. L. (1990).
Proc. Nab. Acad.
