Map: Research / Early Earth / Genomics & Geology	< Back		Next >

3.6 Genomics, geology, and the tree of life (Awramik, Fitz-Gibbon, House, Lake, Rivera, Runnegar)

Figure 7.6.1. The distribution of environmental and genomic parameters during prokaryotic evolution as calculated from extant taxa. The parameters illustrated include minimum (Tmin), optimal (Topt), and maximum (Tmax) growth temperature, genome size, genome composition (%GC), and salinity (Water). Clockwise from lower left to lower right, the taxa are: Escherichia coli (g proteobacterium), Bacillus subtilis (gram positive eubacterium), Synechocystis (cyanobacterium), Aquifex (eubacterial extreme thermophile), Pyrococcus (archaebacterial extreme thermophile), Methanococcus (thermophilic methanogen), Methanobacterium (moderately thermophilic methanogen), and Archaeoglobus (a thermophilic archaebacterium).

Phylogenetic trees can serve as DNA-derived windows into Earth's past. They offer a framework upon which one can map the atmospheric, paleontological, geological, and climatological records and thereby test scenarios for evolution on a planetary scale. Here, phylogenetic reconstructions will be used to address a central set of questions relating to life's early history: What were environmental conditions like early in the evolution of life on Earth? Did life start in a hot environment? What compounds were being made in the first half billion years of life on earth? What was the role of horizontal gene transfer during these early times? Was methanogenesis developed quickly and only once? How did environmental variables such as temperature, pH, sulfur availability, and oxygen affect the evolution of life? Did heterotrophy evolve before autotrophy, or vice versa? The research proposed in this section will focus on these, and related, questions using a combined geological and genomics approach.

3.6.1 Mapping microbial metabolisms on to the tree of life

Parsimony analyses of phylogenetic trees, can allow one to extrapolate from the environments in which extant prokaryotes are currently living to the environments in which their distant ancestors lived. This process (Williams and Fitch 1989) is illustrated in Figure 3.6.1 using a standard reference tree of eight prokaryotes of diverse ancestry. Six environmental parameters of present day organisms are indicated by the colors at the tips of the branches and the inferred parameter values are shown by colors continuously distributed across the trees. This example is inconclusive because few taxa have been sampled, but serves to illustrate the method. Using trees constructed in this way, but which contain many prokaryotic taxa, it will be possible to trace back the evolution of different environmental parameters into the interior of the tree and thereby relate the tree of life to the geological record.

Phylogenetic reconstruction is central to interpreting the geological record, but obtaining realistic reconstructions of the prokaryotic branches of the tree of life is a daunting task. At present, prokaryotic relationships remain poorly resolved. There is now an emerging consensus that prokaryotic trees, whether based upon ribosomal RNA gene sequences or upon other molecules, fail at depths greater than the "phylum" level (Garrity and Holt 2001). Although it is usually known to which phylum any particular bacterium belongs, it is generally not known how the various phyla are related to each other. This deficiency has obvious implications for understanding the origins and evolution of planetary-scale metabolic activities such as, for example, oxygenic photosynthesis.

For example, at present, the relationships among the various kinds of photosynthetic prokaryotes are unknown. Solving this problem is important if we are to understand fully the early evolution of life on Earth. Here, several steps towards obtaining better, more robust trees are proposed.

New computational tools and approaches for genome analysis, that specifically address horizontal gene transfer, have and are being developed (Lake and Moore 1998). HGT can distort phylogenetic relationships since different genes within a given genome will have different histories and therefor give rise to different trees when compared across organisms.

Thus, gene trees do not necessarily correspond to organismal trees (Brown and Doolittle 1997; Feng et al. 1997; Koonin et al. 1997; Rivera et al., 1998). However, more recent studies suggest that horizontal gene transfer, while rampant, is not random, since it predominantly affects only "operational" genes, those primarily involved in housekeeping functions (Lake et al. 1999). By using only "informational genes", which are less affected by horizontal gene transfer (e.g., those involved in protein synthesis and RNA transcription), one should be able to derive correct organismal trees and also resolve deep phylogenetic divergences among prokaryotes. Ribosomal RNA genes belong to the informational class, but because they are short (2,000 to 4,000 nucleotides) they cannot accurately place prokaryotic phyla relative to each other. However, sequences constructed from concatenated informational protein genes (>100,000 nucleotides) obtained from public genome databases, are expected to be able to resolve these deep prokaryotic divergences.

Lake and Rivera will use these methods to greatly refine the prokaryotic (bacterial and archaebacterial) sectors of the tree of life and then to use these refined trees to explore the history of environmental tolerances, as described above. This information will be integrated with the results of geological and geochemical studies ( §3.4, 3.5, 4.2) that provide evidence for the time of appearance of key metabolic activities.

< Previous Section	Next Section >