Introduction
Although the concept of panspermia (life everywhere) had its origins in the late 19th century (Hoyle et al. Reference Hoyle, Wickramasinghe and Hoyle1981), it was ignored at a 1957 international symposium on The Origin of Life on the Earth, held in Moscow (Oparin et al. Reference Oparin, Braunshtein, Pasynskii and Pavlovskaya1957). Its revival had however already begun in correspondence between Hoyle and Haldane, leading to Hoyle's 1953 book Frontiers of Astronomy (Wickramasinghe Reference Wickramasinghe and Burdyuzha2006). Support for Hoyle's contentions was subsequently provided by the detection of organic matter in meteorites (e.g. Brooks & Shaw Reference Brooks and Shaw1969) and interstellar dust (e.g. Wickramasinghe et al. Reference Wickramasinghe, Brooks, Shaw and Hoyle1977; Allen & Wicramasinghe Reference Allen and Wicramasinghe1981; Hoyle et al. Reference Hoyle, Wickramasinghe, Al-Mufti, Olavesen and Wickramasinghe1982). The proposition that bacterial cells were amplified within comets, to then be constantly fed to the top of the Earth's atmosphere (Hoyle & Wickramasinghe Reference Hoyle and Wickramasinghe1980, Reference Hoyle and Wickramasinghe1999a), was also vindicated by infrared absorption spectra showing a good fit with those for freeze-dried bacteria (e.g. Hoyle & Wickramasinghe Reference Hoyle and Wickramasinghe1999b).
In recent times the mechanisms of microbial dispersion, transport and protection over interstellar distances have been examined by a range of authors (e.g. Secker et al. Reference Secker, Wesson and Lepock1996; Orgel Reference Orgel1998; Horneck et al. Reference Horneck, Rettberg, Reitz, Wehner, Eschweiler, Strauch, Panitz, Starke and Baumstark-Khan2001; Wickramasinghe et al. Reference Wickramasinghe, Wainwright, Narlikar, Rajaratnam, Harris and Lloyd2003; Burchell Reference Burchell2004; Coulson Reference Coulson2004; Napier Reference Napier2004; Wallis & Wickramasinghe Reference Wallis and Wickramasinghe2004), while dramatic direct evidence for panspermia has been revealed by electron-microscopic and spectroscopic analysis of carbonaceous meteorites, with the detection of filamentous microfossils embedded in freshly fractured surfaces (Hoover Reference Hoover2006). Some of these fossils were identifiable with known cyanobacteria from their imprints of specialized structures of sheaths, trichomes, heterocysts, akinetes and hormogonia.
There have been considerable recent advances in our knowledge of the timeframe for life to evolve on Earth, the chemical or biological constraints imposed on this event and the early appearance of complex life on Earth. We now have a good understanding of conditions pertaining to Earth since its formation, and of the signatures and fossils of early life on Earth. That the timeframe available for life to colonize was exceedingly brief, as noted by Brooks & Shaw (Reference Brooks, Shaw and Noda1978) and Hoyle & Wickramasinghe (Reference Hoyle and Wickramasinghe1980), has been strongly supported by the recent evidence described in this article. There have also been major advances in our knowledge of the phylogeny of life on Earth, particularly following the advent of computer analysis of nucleic acid and protein sequences. The resolution of this phylogeny continues to increase as new data are forthcoming, but already significant implications can be drawn relating to panspermia.
The synthesis of this disparate information as it relates to panspermia is the focus of this article.
The timeframe available for life to arise on Earth or Mars
The transfer of live microbes between Mars and Earth appears probable (e.g. Castresana & Moreira Reference Castresana and Moreira1999; Mileikowski et al. Reference Mileikowsky, Cucinotta, Wilson and Gladman2000), and that from the moons of the outer planets to the inner planets is also possible. However, the transport of life within the Solar System becomes largely irrelevant since the additional time gained by invoking such scenarios amounts to only ∼200–300 million years (Line Reference Line2002), a paltry addition in evolutionary terms. The accretion of the Earth and Moon was completed by ∼4.5 Gya (Halliday Reference Halliday2001) and the first evidence of the primordial ocean appears as early as 4.2–4.3 Gya (Mojzsis et al. Reference Mojzsis, Harrison and Pidgeon2001). However, this ocean was likely to have been evaporated, possibly on several occasions by the Late Heavy Meteorite Bombardment between 4.0 and 3.8 Gya (Schoenberg et al. Reference Schoenberg, Kamber, Collerson and Moorbath2002), in which case the surface of the Earth (and Mars) may have been uninhabitable for the duration (Cohen et al. Reference Cohen, Swindle and Kring2000; Yeomans Reference Yeomans2005). Nonetheless, we cannot rule out the possibility of life surviving this bombardment, because as noted by Davies (Reference Davies1998), microbes may have survived in orbit following splash-expulsion from the planet's surface or in transit from Mars. Indeed the primordial ocean on Mars is expected to have condensed shortly after the planet formed (Clifford & Parker Reference Clifford and Parker2001), and Mars appears to have provided a more conducive habitat for life due to its greater abundance of oxygen, providing both a terminal electron acceptor for respiration and an ozone shield of the planet's surface (Kirschvink & Weiss Reference Kirschvink and Weiss2001). The suggestion that the Lunar bombardment may have been a local event reflecting collisions of the receding Moon with a ringlet of primordial satellites of the Earth (Arrhenius & Lepland Reference Arrhenius and Lepland2000) is not supported by more recent reports that the bombardment resulted from the outward passage of Neptune disturbing planetesimals in the main asteroid belt (Strom et al. Reference Strom, Malhotra, Ito, Yoshida and Kring2005; Yeomans Reference Yeomans2005).
In view of this evidence, the earliest time when life might have evolved on either Earth or Mars is likely to be ∼4.3–4.4 Gya. While the origin of life on one of the outer moons of the larger planets of the Solar System with splash-transport to the inner planets is not discounted, the physical constraints applying to Earth and Mars are generally also applicable to them.
The closing of the timeframe for life to evolve on Earth comes from evidence of fossils and shifts of isotopic ratios (particularly depletion of 13C over 12C and 34S over 32S) within ancient sedimentary rocks. Claims of fossil stromatolites dated to 3.5–3.3 Gya and of isotopic shifts in Greenland rocks dated to 3.85 Gya have been disputed (Dalton Reference Dalton2002; Lepland et al. Reference Lepland, van Zuilen, Arrhenius, Whitehouse and Fedo2005), as has a report by Schopf et al. (Reference Schopf, Kudryavtsev, Agresti, Wdowiak and Czaja2002) of 3.46 Gya bacteria-like fossils from Western Australia (Brasier et al. Reference Brasier, Green, Jephcoat, Kleppe, Van Kranendonk, Lindsay, Steele and Grassineau2002). McCollom & Seewald (Reference McCollom and Seewald2006) have also shown that abiotic synthesis of organic compounds under conditions representing those found in hydrothermal vents can produce 13C-depletion to a degree typically ascribed to biological processes, meaning that evidence of 13C-depletion alone from samples originating in the sea-vent habitat is insufficient to indicate biogenic origin.
With these qualifications in mind, support remains for the biogenic origin of mineralized tubes in pillow larva (South Africa) dated at ∼3.5 Gya (Furnes et al. Reference Furnes, Banerjee, Muehlenbachs, Staudigel and de Wit2004) and of 3.4 Gya fossil microbial mats (apparently photosynthetic) from South African chert (Tice & Lowe Reference Tice and Lowe2004). Complex stromatolite laminations in rock formations dated at 3.43 Gya are also reported from the Pilbara Craton, Australia, with abiogenic hypotheses for their origin being reviewed and rejected (Allwood et al. Reference Allwood, Walter, Kamber, Marshall and Burch2006). Further fossil and isotopic evidence for ancient (probably photosynthetic) microbial mats is reported for 3.2 Gya sandstone of the Moodies Group, South Africa (Noffke et al. Reference Noffke, Eriksson, Hazen and Simpson2006).
Hence, the timeframe for the origin of life on either Earth or Mars is defined by the boundaries described above of ∼4.4 to ∼3.6 Gya. However, during perhaps 200 000 years of this (during the Late Heavy Bombardment) life, if it existed, would be merely surviving rather than evolving, leaving perhaps ∼600 000 years for life to develop in the Earth–Mars system.
Could life have developed to the complexity inferred at 3.6 Gya or earlier within the timeframe available?
Having established a maximal timeframe for life to evolve on Earth or Mars of ∼600 000 years, we can ask whether this allows sufficient time for life to evolve on Earth or Mars to the complexity of life apparent at 3.6 Gya or earlier. Hoyle & Wickramasinghe (Reference Hoyle and Wickramasinghe1980) mooted that the timeframe was too short for complex life to appear, and this contention is being increasingly supported by the synthesis of genomic, proteomic and phylogenetic evidence relating to the LCC.
Some reports arguing that life could develop very rapidly have become untenable, or were intrinsically questionable at the outset. For example, Lazcano & Miller (Reference Lazcano and Miller1994) argued that the development of a cyanobacterial-like genome could be evolved in 7×106 years or less given current rates of gene duplication. However, genome duplication alone is insufficient to develop complex biochemical pathways or novel cell structures. It is not the size of the genome that is important but its content, and there is no evidence that gene duplication was the dominant limiting factor for early evolutionary development. The rate of evolution is also determined by rates of point mutation, lateral transfer, selective pressure and difficulties associated with the development of novel biochemical pathways in the face of competitive selection against intermediate steps that confer no advantage to the host.
There are two overarching reasons for thinking that life could not develop to the apparent complexity that is evident within a 600 000-year timeframe, each of which are detailed below.
The problem of life evolving from inorganic chemicals
Chemical constraints on the origin of life anywhere are formidable and have been widely canvassed (e.g. Wächtershäuser Reference Wächtershäuser1997; Nisbet & Sleep Reference Nisbet and Sleep2001; Chyba & Hand Reference Chyba and Hand2005; Maurel & Haenni Reference Maurel and Haenni2006; Ourisson & Nakatani Reference Ourisson and Nakatani2006). Although incremental progress is being made, major hurdles remain relating to our understanding of the development of cellular life from a racemic mix of inorganic chemicals.
While possible Earth-located ‘cradles of life’ have extended from Darwin's ‘warm pond’ to include marine hydrothermal vents (Nisbet & Sleep Reference Nisbet and Sleep2001) and the lithosphere (Trevors Reference Trevors2002, Reference Trevors2003), none is free of problems. The foremost of these is perhaps the low availability of pre-biotic ribose sugars, soluble phosphate (Deamer Reference Deamer1997) and nucleotides, especially cytosine (Shapiro Reference Shapiro1999). Proposed mechanisms that may alleviate such low availabilities include evaporation from a lagoon (Dworkin et al. Reference Dworkin, Lazcano and Miller2003), solute exclusion in freezing water (Bada et al. Reference Bada, Bigham and Miller1994) or adsorption onto graphite surfaces (Liu & Fraser Reference Liu and Fraser2006). While the improved chemical kinetics provided by high pressure and temperature and the abundance of metal catalysts and sources of energy might favour the deep-sea vent as a cradle of life, this habitat appears unsuited to the polymerization of RNA or DNA. Furthermore calcium and magnesium in seawater will precipitate anions including phosphate, while simple lipids (a likely requisite to membrane enclosures) would also aggregate into insoluble curds in seawater (Deamer Reference Deamer1997).
The First Cell has been defined by Koch (Reference Koch1994) as ‘the first membrane-bound entity that could replicate, mutate, and replicate in the mutated state’, yet the requirement that it grows and divides has been hypothesized to involve at least seven quite different processes that must have arisen independently and initially served functions other than cell-wall synthesis (Koch Reference Koch2000). A container for life to develop appears essential to the conversion of energy flux into useful chemical reactions, for the concentration of reactants and for cell replication (Deamer Reference Deamer1997). As the serendipitous entrapment of a self-replicating genome coding for a primitive energy-conversion pathway that performs beneficial metabolic functions in a growing membrane enclosure (which selectively takes up nucleosides and other essential nutrients) remains improbable, alternatives have been sought that avoid this scenario. Suggestions include the emergence of life from iron monosulphide bubbles (Russell & Hall Reference Russell and Hall1997), the prior establishment of biochemical pathways on metal catalysts (Wächtershäuser Reference Wächtershäuser2000), the evolution of biochemical pathways by template- and sequence-directed reactions (Lahav et al. Reference Lahav, Nir and Elitzur2001; Szabó et al. Reference Szabó, Scheuring, Czárán and Szathmáry2002), spontaneously formed lipid vesicles that become increasingly complex by controlled peptide binding and polymerization reactions (Luisi et al. Reference Luisi, Walde and Oberholzer1999; Segré et al. Reference Segré, Ben-Eli, Deamer and Lancet2001), a simplified linkage between genome growth and cell membrane growth (Chen et al. Reference Chen, Roberts and Szostak2004) and primitive cell division resulting from abiotic shear or thermodynamic forces (Szostak et al. Reference Szostak, Bartel and Luigi2001).
As for the abiotic development of a replicator molecule, RNA is known to have considerable catalytic ability and synthetic RNA molecules have been shown to replicate other RNA molecules up to 14 nucleotides in length (Johnston et al. Reference Johnston, Unrau, Lawrence, Glasner and Bartel2001). However, apart from the low availability of nucleotides mentioned above, there remain major problems relating to RNA as the first genome. Templating of nucleotides on, e.g., montmorillonite results only in highly ordered sequences such as polyadenine which lack information content, and an unequal availability of nucleotides in a primordial mix would further reduce the chance-complexity of randomly bound bases (Abel & Trevors Reference Abel and Trevors2005). Not least, the chirality problem (homochirality of biomolecules being considered essential to life; Cohen (Reference Cohen1995) ) continues to confound us, despite some intriguing suggestions to solve it (Bonner Reference Bonner1995; Keszthelyi Reference Keszthelyi2001; Ribó et al. Reference Ribó, Crusats, Sagués, Claret and Rubires2001; Sandars Reference Sandars2003; Cline Reference Cline2005; Toxvaerd Reference Toxvaerd2005; Cronin & Reisse Reference Cronin and Reisse2006). Even the subsequent transition from an RNA world to a DNA world is non-trivial, seemingly requiring a cascade of gene duplications and arbitrary fortuitous evolutionary steps, all without the organisms being selectively disadvantaged (Line Reference Line2005).
Other chemical problems relating to the origin of life from abiotic chemical reactions have been reviewed by Ourisson & Nakatani (Reference Ourisson and Nakatani2006). While incremental progress is being made on all fronts, we appear to be far from a solution.
The complex character of life at the time of the LCC
Clarity is being increasingly shed on the complexity of the LCC from genome- and protein-sequence analysis across the three domains using computer programs that clear the fog of lateral gene transfer. There is good evidence of an RNA world preceding the DNA/protein world (e.g. Orgel Reference Orgel1986), yet prior to the advent of the LCC the DNA/protein world familiar to us had been optimized (Freeland et al. Reference Freeland, Knight, Landweber and Hurst2000) and major biochemical pathways established for the fixation of carbon dioxide and nitrogen, and for nitrate, sulphate, sulphur and possibly oxygen respiration. The scant evidence of the de-novo development of major biochemical pathways in the interval following the LCC provides the contrast. Methanogenesis is an example of one such pathway; photosynthesis is not, because it is derived from nitrogen fixation (Burke et al. Reference Burke, Hearst and Sodow1993) which, owing to its complexity, most likely had an earlier function (Fani et al. Reference Fani, Gallo and Liò2000).
The ribosome that is responsible for high-fidelity translation of mRNA into protein has changed little since the LCC. This machine with moving parts has never been supplanted in any lineage and has been described, together with its accessories, as probably the most sophisticated machine ever made (Garrett Reference Garrett1999). From a study of genome sequences coding for 246 universal proteins ranged across the domains of life having a known function, Kyrpides et al. (Reference Kyrpides, Overbeek and Ouzounis1999) concluded that ‘the nature of the Last Universal Common Ancestor is now revealed to be even more advanced and complex than previously believed. … It seems that it contained metabolic enzymes and genetic systems similar to those of extant unicellular organisms.’. Likewise, Baymann et al. (Reference Baymann, Lebrun, Brugna, Schoepp-Cothenet, Giudici-Orticoni and Nitschke2003) found that the redox proteins from diverse enzymes involved in bioenergetic electron transport were constructed from a limited set of building blocks, the origin of which lay prior to the divergence of the primary domains, and Ranea et al. (Reference Ranea, Sillero, Thornton and Orengo2006) identified 140 ancestral protein domains that were highly likely to have been present in the LCC. The latter authors again concluded that the LCC had all of the essential functional systems present in extant organisms.
The dating of characters that most probably developed following the divergence of the LCC give an indication of the upper time limit on the age of this community. Isotopic depletion of 34S in sedimentary barites from the Pilbara Craton indicates mesophilic sulphate reduction at 3.47 Gya (Shen et al. Reference Shen, Buick and Canfield2001) and possibly (in view of the report by McCollom & Seewald (Reference McCollom and Seewald2006) mentioned previously) methanogenesis prior to 3.46 Gya, based on the evidence of 13C-depletion of methane relative to that of co-existing CO2 (Ueno et al. Reference Ueno, Yamada, Yoshida, Maruyama and Isozaki2006). It has been claimed that anaerobic photosynthesis was almost certainly established by 3.5–3.4 Gya (e.g. Tice & Lowe Reference Tice and Lowe2004). Anaerobic photosynthesis and mesophilic sulphate reduction are found only in present-day domain Bacteria, while methanogenesis is found only in present-day Archaea. Hence, the most parsimonious explanation of these findings might be that the three domains of life existed ∼3.5 Gya placing the LCC earlier still. A deeply located LCC is also favoured by genomic and proteomic evidence indicating that the eukaryotic cell cannot be deconstructed into features inherited directly from Archaea and Bacteria (Kurland et al. Reference Kurland, Collins and Penny2006), although there is still considerable debate relating to the advent of the eukaryotic lineage.
From this information the placement of the origin of life earlier than 4.0 Gya would be very reasonable, but for the present purpose a limit of ∼3.6 Gya is proposed for both the LCC and the origin of life.
Conclusions
The origin of life from inorganic chemicals within any timeframe appears, in light of the present knowledge, most improbable. Although some progress is being made, we have not come anywhere near achieving the synthesis of a living cell from inorganic chemicals, even under the most contrived conditions imaginable. On top of this we are faced with an apparently modern (in biochemical terms) LCC which existed, at the latest, at a remarkably early stage in Earth's history and shortly after the planet became habitable. More plausible is the greatly extended timeframe available to panspermia for life to evolve to the modernity evident in the LCC.
That the LCC existed before Earth became habitable is almost a necessity if the evidence of cyanobacteria in carbonaceous meteors (Hoover Reference Hoover2006) is accepted, because the phylogenetic tree of life indicates that cyanobacteria have only ever coexisted with (rather than preceded) the other primary domains. In that case, the Earth must then have been subsequently seeded by representatives of each of the three domains. A further ramification of the existence of cyanobacteria in comets would be the early potential for microaerobic metabolism to evolve. Such a possibility is supported by the existence of elements of aerobic respiration deep in the microbial phylogeny, at or below the depth of the LCC.
Regardless of the question of the existence of the LCC prior to Earth's habitability, it is notable that the Gram-positive lineage (representatives of which are endowed with long-lived radiation-resistant spores well suited to space travel) has in recent years been consistently indicated to lie at the deepest level of domain Bacteria phylogeny (Koch Reference Koch2003; Griffiths & Gupta Reference Griffiths and Gupta2004; Ciccarelli et al. Reference Ciccarelli, Doerks, von Mering, Creevey, Snel and Bork2006). While we have insufficient information to positively identify the domain of extant life to which the LCC was most closely allied, Cavalier-Smith (Reference Cavalier-Smith2002) has provided strong evidence that the Gram-positive lineage gave rise to the Archaea/Eukaryota branches, thereby indicating the LCC to be Gram positive.
A brief comment is warranted on the microbiological ramifications of the proposition by Hoyle and co-workers (e.g. Hoyle & Wickramasinghe Reference Hoyle and Wickramasinghe1980; Wickramasinghe et al. Reference Wickramasinghe, Wainwright, Narlikar, Rajaratnam, Harris and Lloyd2003) that the upper atmosphere is continuously seeded with bacteria and viruses of cometary origin. Inevitably these would add to Earth's gene pool, although from an evolutionary perspective the gene flow would be one-way, at least for comets of interstellar origin. If cometary biota is effectively modern as argued above, then newcomers to Earth would only appear as diversity in the existing phylogeny, although such added diversity would be largely restricted to habitats that matched those found on comets. That survival of such cometary biota would be possible in terrestrial habitats is indicated by the very high diversity of biota found in soil, showing that competitive exclusion is not a significant player in such heterogeneous, low-nutrient environments. Furthermore the equilibrium conditions conducive to competitive exclusion are never attained, and the survival of apparently maladapted microbes such as thermophiles in Antarctic soil habitats has previously been noted by Hoyle & Wickramasinghe (Reference Hoyle and Wickramasinghe1980). That said, the attribution of the evolutionary leaps seen in the phylogeny of life to an influx of biota from comets may be misplaced. Although there are instances of evolutionary leaps apparent in the phylogenetic tree which remain puzzling (most notably the appearance of the eukaryotic cell; see Kurland et al. (Reference Kurland, Collins and Penny2006) and Embley & Martin (Reference Embley and Martin2006) ), such instances appear to be more likely to be resolved in terms of gene duplication, cell fusion or lateral gene transfer. Likewise, no consideration for an exotic origin of extremophiles on Earth is warranted, because phylogenetic analysis indicates that thermophily (Cavalier-Smith, Reference Cavalier-Smith2002), psychrophily (Bowman, personal communication) and halophily (particularly Archaeal representatives) are of relatively recent origin, arising in an orderly fashion from non-extremophile antecedents.
It should also be noted that from a microbiological perspective, bacteria or viruses of cometary origin would make poor pathogens of plants or animals, because such pathogens (with a few readily explicable exceptions) have necessarily evolved in intimate association with their respective hosts, each driving the further evolution of the other.
As an aside, it is a curious coincidence (perhaps just one of the many that allow life to exist) that the dimensions of particles necessary for rapid expulsion from a stellar system very closely match those of small microbial clumps, or individual bacterial spores coated with a thin protective layer of carbon or graphite.
Acknowledgements
The author is indebted to the School of Agricultural Science and the University of Tasmania for support on a project totally unrelated to agriculture, as well as to Tom Head and Christopher Line for helpful discussions on this topic.
