- Evolutionary Relationships and Taxonomy
As a bacteriologist interested in how evolutionary understanding molds bacterial taxonomy, I am concerned that Ludwig and Schleifer (ASM News, November 1999, p. 752-757) paid minimal attention to approaches other than those derived from rRNA sequences. They state, without going further, that the conflicting conclusions arising from other molecular markers raise difficult questions about relationships of primary domains and that determination of branching order on trees requires paralogous data. True enough, but there are enough problems in unraveling evolution, establishing phylogenetic order, and in the interpretations required for making trees to make it truly important to be precise about the alternatives based on proteins.
Their objection to protein data as stated seems to be that position effects and the accumulation of mutations would lead to conflicting and confusing interpretations. This would be true if the longer-range conclusions were based on total amino acid sequences of even the most conserved of polypeptides. But the most convincing and thought- stimulating conclusions based on highly conserved proteins come from the work of Gupta and colleagues (recently reviewed in Microbiol. Mol. Biol. Rev. 62:1435-1491, 1998) based on the lineal inheritance of clearly homologous insertions and deletions (``indels'') in their sequences. There are inherited differences in the macromolecular form. Following the succession of these indels provides an inescapable and convincing mark for some major lineages; provide strong suggestion of lineage associations that seem unlikely in interpreting rRNA data; provide evidence for establishing the order of branching of some major lineages; and in places lend point to the consideration of uncertainties in strongly held opinions.
The validity of alternative approaches in phylogeny has been discussed in a set of technical comments from four groups of colleagues considering the detection of lateral gene transfer ( Science 286:1443a et seq., 19 November 1999). Two of them (J. W. Stiller and B. S. Hall, and W. F. Doolittle) argue that signatures of the type provided by indels may offer a basis for constructing more reliable overall phylogenies.
Research in evolutionary biology is certainly opening our eyes to exciting facets of biology and a new aspect to taxonomy, but we have to beware of excluding well-founded novelties that are counter to current and strongly held interpretations. It is worth remembering that science is an ever-moving set of approximations to the truths we strive to discover.
Robert G. E. Murray
University of Western Ontario