Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

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 Computational and Integrative Biology, Massachusetts General Hospital, Boston, Massachusetts 02114 Correspondence: [email protected]

Understanding the origin of cellular life on Earth requires the discovery of plausible pathways for the transition from complex prebiotic chemistry to simple biology,defined asthe emergence of chemical assemblies capable of Darwinian evolution. We have proposed that a simple primitive cell, or protocell, would consist of two key components: a protocell membrane that defines a spatially localized compartment, and an informational polymer that allows for the replication and inheritance of functional information. Recent studies of vesicles composed of fatty-acid membranes have shed considerable light on pathways for protocell growth and division, 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 chemical genome replication and compatibility with membrane encapsulation. The integration of a dynamic fatty-acid compartment with robust, generalized genetic polymer replication wouldyield a laboratory model of a protocell with the potential forclassical Darwinian biologi- cal evolution, and may help to evaluate potential pathways for the emergence of life on the early Earth. Here we discuss efforts to devise such an integrated protocell model.

he emergence of the first cells on the early membrane-encapsulated nucleic acids, and the TEarth was the culmination of a long history chemical and physical processes that allowed of prior chemical and geophysical processes. for their replication. We will discuss recent Although recognizing the many gaps in our progress toward the construction of laboratory knowledge of prebiotic chemistry and the early models of a protocell (Fig. 1), evaluate the planetary setting in which life emerged, we remaining steps that must be achieved before a will assume for the purpose of this review that complete protocell model can be constructed, the requisite chemical building blocks were and consider the prospects for the observation available, in appropriate environmental set- of spontaneous Darwinian evolution in labo- tings. This assumption allows us to focus on ratory protocells. Although such laboratory the various spontaneous and catalyzed assem- studies may not reflect the specific pathways bly processes that could have led to the that led to the origin of life on Earth, they are formation of primitive membranes and proving to be invaluable in uncovering surprising early genetic polymers, their coassembly into and 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 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

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

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

us to reconstruct plausible pathways and scenar- tial function of creating an internal environment ios for the origin of life. within which genetic materials can reside and The term protocell has been used loosely metabolic activities can take place without being to refer to primitive cells or to the first cells. lost to the environment. Modern cell mem- Here we will use the term protocell to refer spe- branes are composed of complex mixtures of cifically to cell-like structures that are spatially amphiphilic molecules such as phospholipids, delimited by a growing membrane boundary, sterols, and many other lipids as well as diverse and that contain replicating genetic informa- proteins that perform transport and enzymatic tion. A protocell differs from a true cell in that functions. Phospholipid membranes are stable the evolution of genomically encoded advanta- under a wide range of temperature, pH, and geous functions has not yet occurred. With a salt concentration conditions. Such membranes genetic material such as RNA (or perhaps one are extremely good permeability barriers, so that of many other heteropolymers that could pro- modern cells have complete control over the vide both heredity and function) and an appro- uptake of nutrients and the export of wastes priate environment, the continued replication through the specialized channel, pump and of a population of protocells will lead inevitably pore proteins embedded in their membranes. to the spontaneous emergence of new coded A great deal of complex biochemical machinery functions by the classical mechanism of evolu- is also required to mediate the growth and divi- tion through variation and natural selection. sion of the cell membrane during the cell cycle. Once such genomically encoded and therefore The question of how a structurally simple proto- heritable functions have evolved, we would cell could accomplish these essential membrane consider the system to be a complete, living bio- functions is a critical aspect of understanding logical cell, albeit one much simpler than any the origin of cellular life. modern cell (Szostak et al. 2001). Vesiclesformed by fatty acids have long been studied as models of protocell membranes BACKGROUND (Gebicki and Hicks 1973; Hargreaves and Deamer 1978; Walde et al. 1994a). Fatty acids Membranes as compartment boundaries are attractive as the fundamental building All biological cells are membrane-bound com- block of prebiotic membranes in that they are partments. The cell membrane fulfills the essen- chemically simpler than phospholipids. Fatty

2 Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

Origins of Cellular Life

acids with a saturated acyl chain are extremely structures and/or transient complexes with stable compounds and therefore might have charged solutes, which facilitate transport across accumulated to significant levels, even given a the membrane. The subject of the permeability relatively slow or episodic synthesis. Moreover, of fatty-acid based membranes is dealt with in the condensation of fatty acids with glycerol greater detail by Mansy (2010). to yield the corresponding glycerol esters pro- Prebiotic vesicles were almost certainly vides a highly stabilizing membrane component composed of complex mixtures of amphiphiles. (Monnard et al., 2002). Finally, phosphory- Amphiphilic molecules isolated from meteor- lation and the addition of a second acyl chain ites (Deamer 1985; Deamer and Pashley 1989) yields phosphatidic acid, the simplest phospho- as well as those synthesized under simulated lipid, thus providing a conceptually simple prebiotic conditions (McCollum et al. 1999; pathway for the transition from primitive to Dworkin et al. 2001; Rushdi and Simoneit more modern membranes. The prebiotic chem- 2001) are highly heterogeneous, both in terms istry leading to the synthesis of fatty acids and of acyl chain length and head group chemistry. other amphiphilic compounds is treated in Membranes composed of mixtures of amphi- more detail in Mansy (2010). philes often have superior properties to those The best reason for considering fatty acids composed of single pure species. For example, as fundamental to the nature of primitive cell mixtures of fatty acids together with the corre- membranes is not, however, their chemical sponding alcohols and/or glycerol esters gene- simplicity. Rather, fatty-acid molecules in mem- rate vesicles that are stable over a wider range branes have dynamic properties that are essential of pH and ionic conditions (Monnard et al. forbothmembranegrowthandpermeability. 2002), and are more permeable to nutrient Because fatty acids are single chain amphiphiles molecules including ions, sugars and nucleo- with less hydrophobic surface area than phospho- tides (Chen et al. 2004; Sacerdote and Szostak lipids, they assemble into membranes only at 2005; Mansy et al. 2008). This is in striking much higher concentrations. This equilibrium contrast to the apparent requirement for homo- property is mirrored in their kinetics: Fatty acids geneity in the nucleic acids, where even low are not as firmly anchored within the membrane levels of modified nucleotides can be destabiliz- as phospholipids; they enter and leave the mem- ing or can block replication. brane on a time scale of seconds to minutes (Chen and Szostak 2004). Fatty acids can also exchange between the two leaflets of a bilayer RECENT RESULTS membrane on a subsecond time scale. Rapid flip- Pathways for Vesicle Growth flop is essential for membrane growth when new amphiphilic molecules are supplied from the Fatty-acid vesicle growth has been shown to environment. New molecules enter the mem- occur through at least two distinct pathways: braneprimarilyfromtheoutsideleaflet,andflip- growth through the incorporation of fatty acids flop allows the inner and outer leaflet areas to from added micelles, and growth through equilibrate, leading to uniform growth. fatty-acid exchange betweenvesicles. The growth Considering that protocells on the early of membrane vesicles from micelles has been Earth did not, by definition, contain any com- observed following the addition of micelles or plex biological machinery, they must have relied fatty-acid precursors to pre-formed vesicles on the intrinsic permeability properties of their (Walde et al. 1994a, 1994b; Berclaz et al. 2001). membranes. Membranes composed of fatty When initially alkaline fatty-acid micelles are acids are in fact reasonably permeable to small mixed with a buffered solution at a lower pH, polar molecules and even to charged species the micelles become thermodynamically unsta- such as ions and nucleotides (Mansy et al. ble. As a consequence, the fatty-acid molecules 2008). This appears to be largely a result of the can either be incorporated into pre-existing ability of fatty acids to form transient defect membranes, leading to growth (Berclaz et al.

Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 3 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

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

2001), or can self-assemble into new vesicles A second, distinct pathway for vesicle (Blochliger et al. 1998; Luisi et al. 2004). These growth involves fatty-acid exchange between pioneering studies were done by cryo-TEM, vesicles. Under certain conditions this exchange which does not allow growth to be followed in can lead to growth of a subpopulation of vesicles real time, and by light scattering, which is diffi- at the expense of their surrounding neigh- cult to interpret in the case of samples with het- bors. Within populations of osmotically relaxed erogeneous size distributions. We therefore vesicles, such exchange processes do not result adapted a fluorescence assay based on FRET in significant changes in size distribution with (Fo¨rster resonance energy transfer) to measure time. Similarly, a population of uniformly changes in membrane area in real time. This osmotically swollen vesicles does not change in assay is based on the distance dependence of size distribution, but such vesicles are in equil- energy transfer between donor and acceptor ibrium with a lower solution concentration of fluorescent dyes; thus when a membrane grows fatty acids because the tension in the membrane in area by incorporating additional fatty-acid of the swollen vesicles makes it more ener- molecules, the dyes are diluted and the efficiency getically favorable for fatty-acid molecules to of FRET decreases. Studies on small (typically reside in membrane. When osmotically swollen 100 nm in diameter) unilamellar fatty-acid vesicles are mixed with osmotically relaxed vesicles using this assay showed that the slow (isotonic) vesicles, rapid fatty-acid exchange addition of fatty-acid micelles led to vesicle processes result in growth of the swollen vesicles growth with an efficiency of 90% (Hanczyc and corresponding shrinkage of the relaxed et al. 2003). vesicles (Chen et al. 2004). Because vesicles The real-time FRET assay allowed for a can be osmotically swollen as a result of the kinetic dissection of the growth process, reveal- encapsulation of high concentrations of nucleic ing a surprisingly complex series of events after acids such as RNA, this process allows for the the rapid addition of micelles (Chen and Szo- growth of vesicles containing genetic polymers stak 2004). Two major processes were observed. at the expense of empty vesicles (or vesicles The first fast phase resulted in membrane area that contain less internal nucleic acid). Because growth that was limited to 40% increase in faster replication would increase the internal area, independent of the amount of added nucleic acid concentration, this pathway of micelles. A second much slower phase led to a competitive vesicle growth provides the poten- further increase in membrane area that varied tial for a direct physical link between the rate with the amount of added micelles. We inter- of replication of an encapsulated genetic poly- preted the fast phase as reflecting the rapid mer and the rate of growth of the protocell as assembly of a layer of adhering micelles around a whole. the pre-formed vesicles, with rapid monomer Assuming that the division of osmotically exchange resulting in the efficient incorporation swollen vesicles could occur either stochastically of this material into the pre-formed membrane. or at some threshold size, protocells that de- We interpreted the slow phase as the conse- veloped some heritable means of faster replica- quence of micelle–micelle interactions leading tion and growth would have a shorter cell cycle, to the assembly of intermediate structures on average, and would therefore gradually take that could partition between two pathways— over the population. This simple physical mech- with some monomers dissociating and contri- anism might therefore lead to the emergence buting to membrane growth and the remainder of Darwinian evolution by competition at the ultimately assembling into new membrane cellular level. However, if replication is limited vesicles. Although these interpretations are con- by the rapid reannealing of complementary sistent with our data, the experiments are strands (see later discussion), it may be difficult rather indirect, and further exploration of the to reach osmotically significant concentrations. mechanism of membrane growth is certainly Furthermore, osmotically swollen vesicles are desirable. intrinsically difficult to divide owing to the

4 Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

Origins of Cellular Life

energetic cost of reducing the volume of a parental vesicle. Over time, this thin membrane spherical vesicle to that of two daughter vesicles tubule elongates and thickens, gradually incor- of the same total surface area. One possibility porating more and more of the parental vesicle, is that osmotically driven competitive growth until eventually the entire vesicle is transformed might alternate with the faster membrane into a long, hollow threadlike vesicle (Fig. 2). growth that follows micelle addition. If new This pathway occurs with vesicles ranging in fatty-acid material was only available sporadi- size from 1 to at least 10 mmindiameter,com- cally, rapid membrane growth might follow an posed of a variety of different fatty acids and influx of fresh fatty acids, facilitating division related amphiphiles. Only multilamellar vesicles (see following discussion). grow in this manner and only when vesicle vol- All of the experiments discussed earlier were ume increases slowly (relative to surface area done with small unilamellar vesicles prepared growth) because of a relatively impermeable buf- by extrusion through 100 nm pores in filters. fer solute. Confocal microscopy has provided In contrast, fatty-acid vesicles that form sponta- insight into the mechanism of this mode of neously by rehydrating dry fatty-acid films tend growth: The outermost membrane layer grows to be several microns in diameter and multila- first, and because there is little volume between mellar (Hargreaves and Deamer 1978; Hanczyc it and the next membrane layer, and that volume et al. 2003). Such large multilamellar fatty-acid cannot increase on the same time scale, the extra vesicles are so heterogeneous that quantitative membrane area is forced into the form of a thin studies of growth and division are difficult. We tubule. Over time, this tubule grows, and as a have recently developed a simple procedure result of poorly understood exchange processes, for the preparation of micron-sized, monodis- the entire original vesicle is ultimately trans- perse (homogeneous in size) multilamellar formed into a long thread-like hollow vesicle vesicles by large-pore dialysis (Zhu and Szostak (Fig. 3). 2009b). The preparation of large monodisperse multilamellar vesicles has allowed us to directly Pathways for Vesicle Division observe an unusual mode of vesicle growth (Zhu and Szostak 2009a). We showed that feed- Vesicle division by the extrusion of large ing a micron-sized multilamellar fatty-acid vesicles through small pores is a way in which vesicle with fatty-acid micelles results in the for- mechanical energy can be used to drive division mation of a thin membranous protrusion which (Hanczyc et al. 2003). Vesicle growth by micelle extends from the side of the initially spherical feeding followed by division by extrusion can

A B

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

Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 5 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

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

Add micelles Growth

Agitate

Repeated cycles Division

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

be performed repetitively, resulting in cycles of and the more biologically relevant processes of growth and division in which both membrane growth and division. material and vesicle contents are distributed We have recently found that the growth of to daughter vesicles in each cycle. However, large multilamellar vesicles into long threadlike division by extrusion results in the loss of vesicles, described above, provides a pathway for 30%–40% of the encapsulated vesicle contents coupled vesicle growth and division (Zhu and to the environment during each cycle (Hanczyc Szostak 2009a). The long threadlike vesicles et al. 2003; Hanczyc and Szostak 2004). Most are extremely fragile, and divide spontaneously of this loss is a result of the unavoidable geo- into multiple daughter vesicles in response to metric constraint of dividing a spherical vesicle modest shear forces. In an environment of gen- into two spherical (or subspherical) daughter tle shear, growth and division become coupled vesicles with conservation of surface area; processes because only the filamentous vesicles some additional loss may occur as a result of can divide (Fig. 3). If the initial parental vesicle pressure-induced membrane rupture. Although contains encapsulated genetic polymers such as extrusion is a useful laboratory model for RNA, these molecules are distributed randomly vesicle division, an analogous extrusion process to the daughter vesicles and are thus inherited. appears unlikely to occur in a prebiotic scenario The robustness and simplicity of this pathway on the early Earth because vesicle extrusion suggests that similar processes might have from the flow of suspended vesicles through a occurred under prebiotic conditions. The mech- porous rock would require both the absence of anistic details of this mode of division remain any large pores or channels and a very high unclear. One possibility, supported by some pressure gradient (Zhu and Szostak 2009a). microscopic observations, is that the long thin The above problems stimulated a search for membrane tubules are subject to the “pearling a more realistic pathway for vesicle division. The instability” (Bar-Ziv and Moses 1994), and min- possible spontaneous division of small uni- imize their surface energy by spontaneously lamellar vesicles after micelle addition has been transforming from a cylindrical shape to a string reported (Luisi et al. 2004; Luisi 2006), and of beads morphology. The very thin tether join- electron microscopy has revealed structures ing adjacent spherical beads may be a weak point that are possible intermediates in growth that can be easily disrupted by shear forces. and division, notably pairs of vesicles joined by a shared wall (Stano et al. 2006). However, the mechanism of the proposed division as RNA-Catalyzed RNA Replication on the Early well as the nature of the energetic driving Earth and the Modern Laboratory force remain unclear. Additional studies are A core assumption of the RNAworld hypothesis required to clearly distinguish between the is that the RNA genomes of primitive cells were vesicle-stimulated assembly of new vesicles, replicated by a ribozyme RNA polymerase

6 Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

Origins of Cellular Life

(Gilbert 1986). The idea of RNA-catalyzed RNA the polymerase ribozyme (Glasner et al. 2000). replication provides a solution to the apparent Such high levels of Mgþþ lead to hydrolytic paradox of DNA replication catalyzed by degradation of the ribozyme, and are also not proteins that are encoded by DNA. This compatible with known fatty-acid based mem- simplification of early biochemistry gained branes because of crystallization of the fatty instant plausibility from the discovery of cata- acid-magnesium salt. In addition, fatty-acid lytic RNAs almost 30 years ago (Kruger et al. membranes are almost impermeable to NTPs 1982; Guerrier-Takada et al. 1983). In the time (Mansy et al. 2008). after the discovery of the first ribozymes, the Theincompatibilitybetweencurrentlyavail- RNA World hypothesis has continued to gain able ribozyme polymerases and fatty-acid support, most dramatically from the discovery based vesicles suggests either that early repli- that the ribosome is a ribozyme (Nissen et al. cases were quite different, or that RNA repli- 2000), and that all proteins are assembled cation in early cells proceeded in a very different through the catalytic activity of the ribosomal manner. For example, less charged and more RNA. Support has also come from the in vitro activated nucleotides might be easier for a evolution of a wide range of new ribozymes, ribozyme polymerase to bind, with little or including bona fide RNA polymerases made no Mgþþ. Many potential leaving groups, of RNA (Johnston et al. 2001). On the other such as imidazole, adenine or 1-Me-adenine hand, no ribozyme polymerase yet comes close have been examined in template-directed and to being a self-replicating RNA, like the repli- nontemplated polymerization reactions, but case envisaged as the core of the RNA World have not yet been tested as substrates for ribo- biochemistry. zyme polymerases (Prabahar and Ferris 1997; Why has the in vitro evolution of an RNA Huang and Ferris 2006). The tethering of ribo- replicase been so much more difficult than zyme and primer-template to hydrophobic originally expected? It is clear that the problem aggregates has been examined as a way of is not with catalysis of the chemical step, even increasing local substrate concentration (Mu¨ller with catalytically demanding triphosphate and Bartel 2008). However, this approach did substrates. Evolutionarily optimized versions not lead to a dramatic improvement in the of the Class I ligase carry out multiple-turnover extent of template copying, apparently as a ligation reactions at .1s21, with over 50,000 result of ribozyme inhibition at high effective turnovers overnight (Ekland et al. 1995), and RNA concentrations. It may be possible to optimized versions of the smaller, simpler evolve polymerases that operate well at high DSL ligase carry out sustained multiple turn- RNA concentrations, but membrane localiza- over ligation reactions at rates .1/min (Voytek tion by chemical derivatization adds further and Joyce 2007). These catalytic rates are more complexity to a replication pathway because of than sufficient to carry out the replication of a the need for a specific catalyst for the derivatiza- 100–200 nt ribozyme in minutes to hours, if tion step. Alternative means of facilitating the these rates could be maintained in the context interaction of a ribozymewith a primer-template of a polymerase reaction using monomer sub- substrate, such as the presence of basic peptides strates. However, even the best available poly- or other cofactors, might overcome this problem. merase ribozyme requires 1–2 days to copy Finally it is noteworthy that very small self- 10–20 nucleotides of a template strand, ap- aminoacylating ribozymes have been obtained parently as a consequence of poor binding to using RNA libraries that have an unconstrained both the ribonucleoside triphosphate (NTP) sequence at their 30-end (Chumachenko et al. monomers and the primer-template substrate. 2009). This strategy may also be fruitful in selec- The need to overcome the electrostatic repul- tions for ribozyme polymerases. sion between negatively charged NTP and Given the above constraints and uncertain- RNA substrates and ribozyme is thought to con- ties, what can we say about the emergence of tribute to the very high Mgþþ requirement for RNA-catalyzed RNA replication in the origin

Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 7 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

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

of life? It is still possible that under the proper spontaneously assemble on a template oligonu- conditions, and using the right substrates, a cleotide and polymerize over several days, gen- small simple ribozyme could effectively catalyze erating a complementary strand. However, RNA replication. However, a replicase must do monomer binding to RNA templates is weak more than catalyze a simple phospho-transfer and concentrations on the order of 0.1 M were reaction. Binding in a nonsequence-specific required for optimal copying. In addition, a manner to a primer-template complex, facilitat- Mgþþ concentration of 0.1 M, which as ing binding of the proper incoming monomer, noted above is incompatible with the presence catalyzing primer extension, and repeating this of fatty-acid based membranes, is required for process until the end of the template (or set optimal polymerization. Even under these of templates for a multi-component replicase) rather extreme conditions, polymerization pro- is reached might require a complex replicase ceeds at only 1–2 nucleotides per day, and is structure. Such a replicase would presumably therefore limited by monomer hydrolysis. be rare in collections of random RNA se- Beyond these constraints, three aspects of the quences. If life required a very special sequence copying reaction present major hurdles to mul- to get started, then the origin of life on earth tiple rounds of replication. First, the chemical could have been a low probability event and structure of RNA results in a problem of regio- life on other earth-like planets might be very specificity, because new linkages can be either rare. If, on the other hand, RNA-catalyzed RNA 3–50 or 20 –50 phosphodiester bonds. Surpris- replication could have emerged gradually in a ingly, under most conditions, it is the 20 –50 series of simpler steps, it might have been easier phosphodiester bonds that are most common and thus more likely for life to begin, and life in polymerization products. This problem can elsewhere might be common. For this reason, be ameliorated by the choice of ions and leaving we now turn to a consideration of nonenzy- groups; for unknown reasons, Zn2þ ions and matic template-directed replication chemistry. the 2-methylimidazole leaving group favor the synthesis of 30 –50 linkages (Lohrmann et al. 1980; Inoue and Orgel 1982). Studies of oligo- Chemical Template Replication Revisited nucleotide ligation showed that the helical con- The nonenzymatic template-directed polymeri- text of an extended RNA duplex also favors the zation of activated ribonucleotides was studied formation of 30 –50 linkages (Rohatgi et al. in depth by Leslie Orgel, together with his stu- 1996). However, it remains difficult to obtain dents and colleagues, over a period of several a homogeneous RNA backbone without enzy- decades (Orgel 2004). Here, the template itself matic catalysis. Second, adenine (A) residues acts as a catalyst by helping to align and orient in the template are difficult to copy, and two the monomers so that they are pre-organized or more As in succession block chain growth for polymerization. The main lesson from this (Inoue and Orgel 1983), presumably as a result work is that spontaneous chemical copying of the poor base-stacking propensity of the of RNA sequences is indeed possible, but is incoming U monomers. It has recently been subject to several important constraints and found that template copying at subzero temper- limitations. The constraints make template- atures can proceed past multiple A residues in directed RNA copying incompatible with cur- the template (Vogel and Richert 2007), but rently available membrane vesicle systems, and this required sequential additions of oxyazaben- the limitations have, so far at least, made it zotriazole-activated monomer together with a impossible to obtain repeated cycles of RNA series of helper oligos. Third, there is the issue replication through chemical copying. of fidelity. The copying of G and C residues is To obtain reasonable reaction rates, Orgel remarkably accurate, with error rates estimated made use of nucleoside monophosphates acti- at 0.5% or less. However, the addition of A vated with a good leaving group such as imida- and U residues causes problems, most signifi- zole. The ribonucleotide 50-phosphorimidazolides cantly the formation of G:U wobble base-pairs

8 Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

Origins of Cellular Life

(Wu and Orgel 1992), which would lead to sig- The high intrinsic reactivity of the amino- nificant error rates in a four base system. Might sugar modified nucleotides suggested to us an alternative genetic polymer, perhaps even a that an alternative phosphoramidate nucleic close relative of RNA, overcome these problems acid might act as a good platform for chemical and enable chemical replication? The identifica- self-replication. Our group has therefore started tion of such a system might ultimately lead to to study a series of phosphoramidate nucleic the discovery of plausible progenitors of RNA, acids with sugar phosphate backbones that or, alternatively, to the discovery of new replica- vary in their degree of conformational flexibility tion strategies that allow for the chemical repli- or constraint. We are currently focusing on the cation of RNA itself. 20-50 linked phosphoramidate analog of DNA, and the corresponding monomers, the 20- amino dideoxyribonucleotide 50-phosphorimi- dazolides (Fig. 4). The 20-amino ImpddNs are Phosphoramidate Nucleic Acids advantageous as monomers because they can- Early studies of template copying using more not undergo intramolecular cyclization because reactive nucleotide derivatives were performed of the steric constraint of the ribose ring; they by Orgel et al., who examined both 20- and are only depleted during polymerization reac- 30-amino ribonucleotide 50-phosphorimidazo- tions by competing hydrolysis. lides (Lohrmann and Orgel 1976; Zielinksi Our first experiments with 20-amino ImpddG and Orgel 1985; Tohidiet al. 1987). Polymeriza- led to rapid and efficient primer-extension tion of these nucleotides yields phosphorami- across a dC15 template, generating full-length date nucleic acids, which are generally similar product in 6 hour (Mansy et al. 2008). En- to standard phosphodiester linked nucleic couraged by the rapid copying of oligo-dC acids except that the phosphoramidate linkage templates by 20-amino ImpddG, we performed is more acid labile. As expected, replacing a a more extensive study of the copying of tem- sugar hydroxyl in the monomer with a more plates with differing sugar-phosphate back- nucleophilic amino group resulted in a large bones, lengths and sequences (Schrum et al. increase in monomer reactivity. The activated 2009). The most important property contribu- 30-amino ribonucleotides participated in rapid ting to good template activity appears to be pre- copying of short oligonucleotide templates organization in the form of an A-type helix. 5-13 nucleotides in length, yielding N30 !P50 Thus, RNA templates were uniformly superior linked complementaryoligonucleotides (Tohidi et al. 1987). The increased reactivity also led to faster intramolecular monomer cyclization, 2′-NP-DNA which depleted the template copying reactions N of activated substrate molecules (Hill et al. O O 1988). NH More recently, the Richert group has begun – to explore the potential of 30-amino- nucleotide O P O N O analogs in template-directed condensation reac- O tions. Deoxyribonucleotide monomers, acti- NH vated with an oxyazabenzotriazole leaving O – N N O P O group, completed a template-directed reaction N N P O O O 0 O with a 3 -amino-terminated primer in seconds O–

(Rothlingshofer et al. 2008). The fidelity of NH2 NH this reaction is sufficient to allow for sequencing O P O– by nonenzymatic primer extension. However, O the monomer is rapidly consumed by internal Figure 4. Structures of 20-50 phosphoramidate DNA, cyclization. and the corresponding activated monomers.

Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 9 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

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

to DNA templates of the same sequence, and been shown to be energetically almost equiva- LNA (locked nucleic acid) templates, which lent to a C:G base-pair (Chaput et al. 2002). are chemically locked in a C30-endo sugar con- When we used G, C, D, and Up as the four formation, were superior to RNA templates. monomer and template bases, we were able to This result is consistent with previous obser- copy mixed-sequence RNA templates over 15 vations that under most conditions RNA- nts in length in about a day. This represents a template directed polymerization of ribonu- significant step toward developing a robust, cleotides leads to a majority of 20-50 linkages; generalized chemical replication system. it appears that an A-type helical geometry gen- This system is, however, far from ideal, and erally favors polymerization through attack by there are strong indications that the fidelity of a20 nucleophile. A second key factor is that template copying becomes an issue when all enhanced monomer affinity for the template four nucleotides are present, largely because of increases the reaction efficiency. Thus G:C base- the formation of G:U wobble base-pairs. Given pairs lead to efficient copying, whereas A:U that primer-extension on 20 –50 linked DNA base-pairs were very poorly copied. Replacing templates is approximately similar to that on Awith D (diaminopurine) results in a D:U base- the corresponding RNA templates, we are cur- pair with three hydrogen bonds, and slightly rently synthesizing 20 –50 linked phosphorami- improved primer-extension. However, the date DNA templates to assess self-replication poor base stacking of U residues must also be in this system. In light of the efficient templating improved to obtain efficient template copying. of LNA oligonucleotides, we are also interested When we replaced U with C5-propynyl-U, the in exploring prebiotically plausible nucleic resulting A:Up base-pairs led to improved copy- acids that are more conformationally constrai- ing, but the D:Up combination had a clearly ned than RNA. A particularly interesting candi- synergistic effect (Fig. 5). In the context of a date is TNA (threose nucleic acid) (Scho¨ning DNA duplex, the D:Up base-pair has previously et al. 2000) and its 20-amino substituted phosphoramidate version, NP-TNA (Wu et al. 2002). These are both base-pairing systems that form standard Watson-Crick duplexes, A - U despite having only five atoms per backbone repeat unit (vs. six for RNA and DNA). It will N NH2 O be of great interest to see if the resulting decrease in flexibility leads to increased fidelity in N N HN template copying reactions. N N O Protocell Assembly In principle, protocell-like objects could form D - Up spontaneously as new membranes self-assemble and encapsulate genetic molecules in solution.

N NH2 O Recently, simple physical processes that would enhance the efficiency of the coassembly of N nucleic acids and membrane vesicles have N HN been proposed. One such alternative scenario NNis based on the fact that the clay mineral

NH2 O montmorillonite is not only a catalyst of RNA polymerization (Ferris et al. 1996; Huang and Figure 5. Watson Crick base-pairs. Top: Standard A:U Ferris 2006) but also catalyzes membrane base-pair. Bottom: Alternative diaminopurine (D): assembly (Hanczyc et al. 2003). Experiments C5-propynyl-uracil (Up) base-pair. with clay particles containing surface-adsorbed

10 Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

Origins of Cellular Life

RNA showed that such particles stimulated more challenging question is, how could such vesicle assembly, and frequently became a structure replicate? We have already consid- trapped inside the vesicles whose assembly ered the replication of the protocell membrane they had catalyzed. Thus, a common mineral and the genetic material as separate entities. can catalyze both the assembly of a genetic pol- To address the question of their replication as ymer and the assembly of a membrane vesicle, a combined structure we must consider in and bring these two components together to more detail the molecular constituents of the generate a protocell-like structure consisting of protocell membrane and the molecular nature a genetic polymer trapped within a membrane of the encapsulated genetic material. compartment. Although the effectiveness of Genome replication within a protocell can this process is attractive, some means of releas- only occur if the building blocks used to copy ing at least some of the bound nucleic acid from template strands are able to enter the fatty-acid the mineral surface, and/or replicating it on the vesicle compartment. Early work using phos- surface, would be necessary for subsequent pholipid-based vesicles and protein enzymes replication to occur. showed the feasibility of constructing primitive More recent experiments suggest a very cell-like compartments (Chakrabarti et al. 1994; different geochemical scenario leading to the Luisi et al 1994). More recent permeability stud- assembly of similar protocell-like structures. ies (Mansy et al. 2008; see also Mansy 2010) The hollow channels within the rocks of the showed that nucleotides could spontaneously alkaline off-axis hydrothermal vents provide a diffuse across simple fatty-acid membranes, protected compartmentalized environment but that net negative charge is a critical determi- where it has been suggested that primitive meta- nant of permeability. Thus, nucleotides that are bolic activities might have originated (Martin chemically activated, e.g., by conversion of and Russell 2003). Recent theoretical studies the 50-phosphate to a 50-phosphorimidazolide, suggested that the strong thermal gradients equilibrate across vesicle membranes much present in hydrothermal vents, together with more rapidly on account of the reduction of the thin channels produced by mineral precipi- the net negative charge. In addition, mixtures tation, could greatly concentrate small organic of fatty acids with their glycerol esters generate molecules such as nucleotides as well as larger membranes that are more permeable to polar nucleic acids from a very dilute external reser- and charged molecules. By combining these voir (Baaske et al. 2007). Work from our labo- observations, we were able to show that activated ratory (Budin et al. 2009) has confirmed the 20-amino-20,30-dideoxyguanosine-50-phosphor- predicted concentration effect, and has also imidazolide, the same nucleotide previously shown that subcritical concentrations of fatty shown to rapidly copy oligo-dC templates, acids can be concentrated to the extent that could be added to the outside of fatty-acid they self-assemble into vesicles at the bottom vesicles containing an encapsulated primer- of the capillary channels. Moreover, DNA oligo- template complex, and copy the internal dC15 nucleotides can also be greatly concentrated template. Copying of the encapsulated dC15 and can become encapsulated within the template by primer-extension reached .95% vesicles, resulting in the spontaneous assembly completion in 12–24 hour (Mansy et al. of protocell-like structures. 2008), compared with 6–12 hour in free solu- tion. The longer time required for copying encapsulated templates reflects the time Encapsulated Template Replication: required for entry of external nucleotides to Emergence of a Protocell the interior of the vesicles. Importantly, the The experiments discussed earlier suggest that presence of high concentrations (5–10 mM) the assembly of protocell-like structures is not of highly reactive activated nucleotides did that difficult, because it appears to be possible not have any disruptive effects on the integrity through multiple distinct mechanisms. The of the vesicle membrane, as no leakage of

Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 11 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

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

encapsulated primer-template complexes was ing genome, and protect the nascent evolving observed. It is also important to note that system from parasitic polymers. control experiments with phospholipid mem- branes showed no copying of internal template, because the activated nucleotides could not CHALLENGES AND FUTURE RESEARCH enter the vesicle; similarly, “modern” activated DIRECTIONS nucleotides such as nucleoside triphosphates Prospects for a Complete Protocell Model cannot cross fatty-acid based membranes. Successful copying of encapsulated templates Although considerable progress has been made therefore requires both “primitive” nucleotides toward the assembly of model protocells, several with reduced charge, and “primitive” mem- remaining issues must be solved before multiple branes composed of single chain amphiphiles. cycles of protocell replication can be achieved The copying of a genetic polymer inside a in the laboratory. These factors are also relevant membrane compartment is an important step to protocell replication on the early Earth. The toward the realization of a self-replicating most important factor at this time appears to system capable of Darwinian evolution. What, be the competition between strand reannealing then, are the remaining barriers to the assembly and strand copying, after thermal strand separa- of such a system? The copying of a single- tion. PCR reactions generally plateau at about stranded template produces a double-stranded 1-mM DNA strand concentration, which is the product; these strands would have to separate concentration at which strand reannealing and before a second cycle of genome replication strand copying occur on a similar time scale could begin. Separate follow-up experiments of about 1 minute However, nonenzymatic by our group showed that some fatty-acid based template copying requires on the order of a vesicles are able to retain encapsulated DNA and day for completion, which implies that either RNA oligonucleotides over a temperature range template copying must be much faster, or of 0 8C to 100 8C (Mansy and Szostak 2008). reannealing must be much slower. One way to As with permeability, mixtures of amphiphiles make reannealing sufficiently slow is to keep lead to improved thermostability, with glycerol strand concentrations subnanomolar. Low esters being particularly stabilizing, possibly strand concentrations are possible in large because of the additional hydrogen bond vesicles, but it is hard to see how a few molecules donors and acceptors provided by the glycerol of a genetic polymer could have any significant head group. Furthermore, we found that en- phenotypic effect on a large vesicle composed of capsulated double-stranded DNA could be millions of amphiphilic molecules. The emer- denatured at elevated temperatures, with the gence of metabolic ribozymes would be more strands reannealing once the temperature was plausible if nucleic acid strand concentrations lowered (Mansy and Szostak 2008). This implies were much higher, so that a catalyst of modest that thermal fluctuations could provide a mech- efficiency could generate enough product to anism for strand separation that is compatible influence cell properties. with the integrity of fatty-acid vesicles, poten- What other factors might affect the rate of tially allowing for complete cycles of replication strand reannealing? Perhaps the most obvious of encapsulated genetic polymers. The mutual possibility is that secondary structure, which compatibility of nucleic acid replication and can form extremely rapidly because the inter- fatty-acid compartment growth is very encour- actions are intramolecular, could greatly slow aging because it alleviates concerns related to down strand annealing. This phenomenon is the permeability and stability of membrane essential to the viability of single-stranded vesicles. Vesicles therefore do seem to be a RNA phage such as Qb† (Axelrod et al. 1991). physically plausible way to segregate and spa- Significant intrastrand secondary structure tially localize genomes, keep emergent catalytic would also be an expected consequence of selec- polynucleotides physically close to their encod- tion for sequences that fold into functional

12 Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

Origins of Cellular Life

shapes with catalytic activity. On the other divalent cations such as Ca2þ. Systematic efforts hand, chemical replication through dense sec- to measure vesicle fusion frequencies under dif- ondary structure would probably be much ferent environmental conditions could there- slower than replication of an unfolded, open fore be quite useful. Finally, it is possible that template. The outcome of simultaneous selec- this problem could be circumvented entirely if tion for an open, accessible template sequence, early life was discontinuous (Budin et al. and a folded functional structure remains un- 2009). For example, protocells could be occa- clear. An alternative but even more speculative sionally disrupted by drying, or simply dissolve solution might result from the rapid binding by as a result of dilution with water to a level below base-pairing of short oligonucleotides or even the critical aggregate concentration; subsequent monomers to freshly separated strands. If a tem- re-hydration or concentration would result in plate strand was largely occupied by monomers reformation of vesicles encapsulating genomic or short oligomers, even if these were in rapid nucleic acids, thus generating a new “random- exchange, the strand might be prevented from ized” set of protocells. As long as such events annealing to a complementary strand. This pos- were fairly uncommon, and assuming that sibility hasthe advantage that it need not blockor genomic replication had kept ahead of vesicle slow the copying reaction, however its effective- replication so that each vesicle contained multi- ness remains to be tested experimentally. ple genome copies prior to disruption, this Another challenge faced by replicating process of disruption and reformation would protocells, whether on the early earth or in the lead to the spread of evolving genomic sequen- modern laboratory, is the continuous dilution ces through the “new” vesicles. of protocells through the competing formation Laboratory models of protocell systems of new empty vesicles. When new fatty acids are should be helpful in modeling many of the supplied as micelles, the efficiency of incorpora- above scenarios. Assuming that protocell repro- tion into pre-formed vesicles (or protocells) can duction can be achieved, and made efficient be quite high, but some new vesicles are always enough to continue through many generations, formed. Thus over time, the descendants of a it should then be possible to observe the spon- given protocell will gradually be diluted out by taneous evolution of adaptive innovations in the continuous formation of these new vesicles. this relatively simple chemical system. The To avoid extinction by dilution, the protocells nature of such adaptations may provide clues must out-compete other vesicles either by hav- as to how modern cells evolved from their ear- ing a more rapid cell cycle, thereby generating liest ancestors. Ultimately this line of research more progeny during division, or by surviving may also tell us whether the conserved bio- destructive processes more efficiently. We have chemistry of life is driven by chemical necessity, previously proposed (Chen et al. 2004) that or whether biochemically very different forms faster growth, driven by the osmotic pressure of life are also possible. of encapsulated nucleic acid, could lead to an effectively shorter cell cycle for protocells that contain high copy numbers of their replicating ACKNOWLEDGMENTS genome. However, in light of the problems asso- ciated with this approach, it is of considerable We thank Itay Budin, Matt Powner, and interest to explore new ways in which a protocell other members of the Szostak lab for helpful genome could lead to faster growth or growth discussions. that occurs at the expense of empty vesicles. An alternative strategy for surviving dilution would be for a protocell genome to colonize REFERENCES empty vesicles. This could occur through a Axelrod VD, Brown E, Priano C, Mills DR. 1991. Coliphage low level of stochastic vesicle–vesicle fusion Q b RNA replication: RNA catalytic for single-strand events, possibly catalyzed by low levels of release. Virology 184: 595–608.

Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 13 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

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

Baaske P,WeinertFM, Duhr S, Lemke KH, Russell MJ, Braun Hanczyc MM, Szostak JW. 2004. Replicating vesicles as D. 2007. Extreme accumulation of nucleotides in simu- models of primitive cell growth and division. Curr lated hydrothermal pore systems. Proc Natl Acad Sci Opin Chem Biol 8: 660–664. 104: 9346–9351. Hanczyc MM, Fujikawa SM, Szostak JW. 2003. Experimen- Bar-Ziv R, Moses E. 1994. Instability and “pearling” states tal models of primitive cellular compartments: Encapsu- produced in tubular membranes by competition of cur- lation, growth, and division. Science 302: 618–622. vature and tension. Phys Rev Lett 73: 1392–1395. Hargreaves WR, Deamer DW. 1978. Liposomes from ionic, Berclaz N, Muller M, Walde P, Luisi PL. 2001. Growth and single-chain amphiphiles. Biochemistry 17: 3759–3768. transformation of vesicles studied by ferritin labeling Hill AR Jr, Nord LD, Orgel LE, Robins RK. 1988. Cyclization and cryotransmission electron microscopy. J Phys Chem of nucleotide analogues as an obstacle to polymerization. B 105: 1056–1064. J Mol Evol 28: 170–1. Blochliger E, Blocher M, Walde P, Luisi PL. 1998. Huang W, Ferris JP.2006. One-step, regioselective synthesis Matrix effect in the size distribution of fatty acid vesicles. of up to 50-mers of RNA oligomers by montmorillonite J Phys Chem B 102: 10383–10390. catalysis. J Am Chem Soc 128: 8914–9. Inoue T, Orgel LE. 1982. Oligomerization of (guanosine Budin I, Bruckner R, Szostak JW. 2009. Formation 0 of protocell-like vesicles in a thermal diffusion column. 5 -phosphor)-2-methylimidazolide on poly(C). An J Am Chem Soc 131: 9628–9629. RNA polymerase model. J Mol Biol 162: 201–17. Inoue T, Orgel LE. 1983. A nonenzymatic RNA polymerase Chakrabarti AC, Breaker RR, Joyce GF, Deamer DW. 1994. model. Science 219: 859–62. Production of RNA by a polymerase protein encapsu- lated within phospholipid vesicles. J Mol Evol 39: 555–9. Johnston WK, Unrau PJ, Lawrence MS, Glasner ME, Bartel DP.2001. RNA-catalyzed RNA polymerization: Accurate Chaput JC, Sinha S, Switzer C. 2002. 5-propynyluracil.dia- and general RNA-templated primer extension. Science minopurine: An efficient base-pair for non-enzymatic 292: 1319–25. transcription of DNA. Chem Commun 15: 1568–9. Kruger K, Grabowski PJ, Zaug AJ, Sands J, Gottschling DE, Chen IA, Roberts RW, Szostak JW. 2004. The emergence Cech TR. 1982. Self-splicing RNA: Autoexcision and of competition between model protocells. Science 305: autocyclization of the ribosomal RNA intervening 1474–1476. sequence of Tetrahymena. Cell 31: 147–57. Chen IA, Szostak JW. 2004. A kinetic study of the growth of Lohrmann R, Orgel LE. 1976. Template-directed synthesis fatty acid vesicles. Biophys J 87: 988–998. of high molecular weight polynucleotide analogues. Chumachenko NV, Novikov Y, Yarus M. 2009. Rapid and Nature 261: 342–344. simple ribozymic aminoacylation using three conserved Lohrmann R, Bridson PK, Orgel LE. 1980. Efficient metal- nucleotides. J Am Chem Soc 131: 5257–63. ion catalyzed template-directed oligonucleotide synthe- Deamer DW. 1985. Boundary structures are formed by sis. Science 208: 1464–5. organic-components of the Murchison carbonaceous Luisi PL. 2006. The emergence of life: From chemical chondrite. Nature 317: 792–794. origins to synthetic biology, Cambridge University Press, Deamer DW, Pashley RM. 1989. Amphiphilic components Cambridge. of the Murchison carbonaceous chondrite: Surface prop- Luisi PL, Walde P, Oberholzer T. 1994. Enzymatic RNA erties and membrane formation. Orig Life Evol Biosph 19: synthesis in self-reproducing vesicles: An approach to 21–38. the construction of a minimal synthetic cell. Ber Bunsenges Phys Chem 98: 1160–5. Dworkin J, Deamer D, Sandford S, Allamandola L. 2001. Self-assembling amphiphilic molecules: Synthesis in Luisi PL, Stano P,Rasi S, Mavelli F.2004. A possible route to simulated interstellar/precometary ices. Proc Natl Acad prebiotic vesicle reproduction. Artif Life 10: 297–308. Sci 98: 815–819. Mansy SS. 2010. Membrane transport in primitive cells. Ekland EH, Szostak JW, Bartel DP. 1995. Structurally com- Cold Spring Harb Perspect Biol 2: a002188. plex and highly active RNA ligases derived from random Mansy SS, Szostak JW.2008. Thermostability of model pro- RNA sequences. Science 269: 1319–25. tocell membranes. Proc Natl Acad Sci 105: 13351–13355. Ferris JP,Hill AR Jr, Liu R, Orgel LE. 1996. Synthesis of long Mansy SS, Schrum JP,Krishnamurthy M, Tobe´ S, Treco DA, prebiotic oligomers on mineral surfaces. Nature 381: Szostak JW. 2008. Template-directed synthesis of a 59–61. genetic polymer in a model protocell. Nature 454: 122–125. Gebicki JM, Hicks M. 1973. Ufasomes are stable particles Martin W, Russell MJ. 2003. On the origins of cells: A surrounded by unsaturated fatty acid membranes. Nature hypothesis for the evolutionary transitions from abiotic 243: 232–234. geochemistry to chemoautotrophic prokaryotes, and Gilbert W. 1986. The RNAWorld. Nature 319: 618. from prokaryotes to nucleated cells. Philos Trans R Soc Glasner ME, Yen CC, Ekland EH, Bartel DP.2000. Recogni- Lond B Biol Sci 358: 59–83. tion of nucleoside triphosphates during RNA-catalyzed McCollom TM, Ritter G, Simoneit BR. 1999. Lipid synthesis primer extension. Biochemistry 39: 15556–62. under hydrothermal conditions by Fischer-Tropsch-type Guerrier-Takada C, Gardiner K, Marsh T, Pace N, Altman S. reactions. Orig Life Evol Biosph 29: 153–166. 1983. The RNA moiety of ribonuclease P is the catalytic Monnard PA, Apel CL, Kanavarioti A, Deamer DW. 2002. subunit of the enzyme. Cell 35: 849–57. Influence of ionic inorganic solutes on self-assembly

14 Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

Origins of Cellular Life

and polymerization processes related to early forms Stano P, Wehrli E, Luisi PL. 2006. Insights into the self- of life: Implications for a prebiotic aqueous medium. reproduction of oleate vesicles. J Phys: Condens Matter Astrobiology 2: 139–52. 18: S2231–S2238. Mu¨ller UF,Bartel DP.2008. Improved polymerase ribozyme Szostak JW, Bartel DP, Luisi PL. 2001. Synthesizing life. efficiency on hydrophobic assemblies. RNA 14: 552–62. Nature 409: 387–390. Nissen P, Hansen J, Ban N, Moore PB, Steitz TA. 2000. Tohidi M, Zielinski WS, Chen CH, Orgel LE. 1987. The structural basis of ribosome activity in peptide Oligomerization of 30-amino-30deoxyguanosine-50phos- bond synthesis. Science 289: 920–30. phorimidazolidate on a d(CpCpCpCpC) template. JMol Orgel LE. 2004. Prebiotic chemistry and the origin of the Evol 25: 97–99. RNA world. Crit Rev Biochem Mol Biol 39: 99–123. Vogel SR, Richert C. 2007. Adenosine residues in the Prabahar KJ, Ferris JP. 1997. Adenine derivatives as phos- template do not block spontaneous replication steps of phate-activating groups for the regioselective formation RNA. Chem Commun 19: 1896–8. of 3’,5’-linked oligoadenylates on montmorillonite: Pos- Voytek SB, Joyce GF. 2007. Emergence of a fast-reacting sible phosphate-activating groups for the prebiotic syn- ribozyme that is capable of undergoing continuous thesis of RNA. J Am Chem Soc 119: 4330–7. evolution. Proc Natl Acad Sci 104: 15288–93. Rohatgi R, Bartel DP, Szostak JW. 1996. Nonenzymatic, Walde P, Wick R, Fresta M, Mangone A, Luisi PL. 1994a. template-directed ligation of oligoribonucleotides is Autopoietic self-reproduction of fatty acid vesicles. highly regioselective for the formation of 30-50 phospho- J Am Chem Soc 116: 11649–11654. diester bonds. J Am Chem Soc 118: 3340–4. Walde P, Goto A, Monnard P-A, Wessicken M, Luisi PL. Ro¨thlingsho¨fer M, Kervio E, Lommel T, Plutowski U, 1994b. Oparin’s reactions revisited: Enzymatic synthesis Hochgesand A, Richert C. 2008. Chemical primer exten- of poly(adenylic acid) in micelles and self-reproducing sion in seconds. Angew Chem Int Ed Engl 47: 6065–8. vesicles. J Am Chem Soc 116: 7541–7547. Rushdi AI, Simoneit BR. 2001. Lipid formation by aqueous Wu T, Orgel LE. 1992. Nonenzymatic template-directed Fischer-Tropsch-type synthesis over a temperature range synthesis on hairpin oligonucleotides. 3. Incorporation of 100 to 400 degrees C. Orig Life Evol Biosph 31: of adenosine and uridine residues. J Am Chem Soc 114: 103–118. 7963–9. Sacerdote MG, Szostak JW. 2005. Semi-permeable lipid Wu X, Guntha S, Ferencic M, Krishnamurthy R, bilayers exhibit diastereoselectivity favoring ribose; Eschenmoser A. 2002. Base-pairing systems related implications for the origins of life. Proc Natl Acad Sci to TNA: a-Threofuranosyl oligonucleotides containing 102: 6004–6008. phosphoramidate linkages. Org Lett 4: 1279–82. Scho¨ning K, Scholz P, Guntha S, Wu X, Krishnamurthy R, Zhu TF,Szostak JW. 2009a. Coupled growth and division of Eschenmoser A. 2000. Chemical etiology of nucleic acid model protocell membranes. J Am Chem Soc 131: 5705– structure: The a-threofuranosyl-(30 !20) oligonucleo- 5713. tide system. Science 290: 1347–51. Zhu TF, Szostak JW. 2009b. Preparation of large monodis- Schrum JP,Ricardo A, Krishnamurthy K, Blain JC, Szostak JW. perse vesicles. PLoS ONE 4: e5009. 2009. Efficient and rapid template-directed nucleic acid Zielinski WS, Orgel LE. 1985. Oligomerization of activated copying using 20-amino-20,30-dideoxyribonucleoside-50- derivatives of 30-amino-30-deoxyguanosine on poly(C) phosphorimidazolide monomers. JAmChemSoc31: and poly(dC) templates. Nucleic Acids Res 13: 2469– 14560–14570. 2484.

Cite this article as Cold Spring Harb Perspect Biol 2010;2:a002212 15 Downloaded from http://cshperspectives.cshlp.org/ on October 1, 2021 - Published by Cold Spring Harbor Laboratory Press

The Origins of Cellular Life

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

Cold Spring Harb Perspect Biol 2010; doi: 10.1101/cshperspect.a002212 originally published online May 19, 2010

Subject Collection The Origins of Life

Constructing Partial Models of Cells The Hadean-Archaean Environment Norikazu Ichihashi, Tomoaki Matsuura, Hiroshi Kita, Norman H. Sleep et al. Ribonucleotides An Origin of Life on Mars John D. Sutherland Christopher P. McKay Deep Phylogeny−−How a Tree Can Help Primitive Genetic Polymers Characterize Early Life on Earth Aaron E. Engelhart and Nicholas V. Hud Eric A. Gaucher, James T. Kratzer and Ryan N. Randall Cosmic Carbon Chemistry: From the Interstellar Membrane Transport in Primitive Cells Medium to the Early Earth Sheref S. Mansy Pascale Ehrenfreund and Jan Cami Origin and Evolution of the Ribosome The Origins of Cellular Life George E. Fox Jason P. Schrum, Ting F. Zhu and Jack W. Szostak Planetary Organic Chemistry and the Origins of From Self-Assembled Vesicles to Protocells Biomolecules Irene A. Chen and Peter Walde Steven A. Benner, Hyo-Joong Kim, Myung-Jung Kim, et al. Mineral Surfaces, Geochemical Complexities, and The Origin of Biological Homochirality the Origins of Life Donna G. Blackmond Robert M. Hazen and Dimitri A. Sverjensky Historical Development of Origins Research Earth's Earliest Atmospheres Antonio Lazcano Kevin Zahnle, Laura Schaefer and Bruce Fegley

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

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