Pauli J. Ojala
Member # 794
posted 24. June 2003 10:22
I'm a poor Finnish biochemist in trouble with my transcript for a PhD. Could you give me some feed back criticism on this part?
Horizontal gene transfer
Cell is the bioscientific publication carrying the highest impact factor in any journal in the field. Palaeontologist and molecular biologist Simon Conway Morris had the honour to write the first article of the first issue of the 100th volume of the Cell at the beginning of the new millennium. His Introduction was as follows:
“When discussing organic evolution the only point of agreement seems to be: ‘It happened.‘ Thereafter, there is little consensus, which at first sight must seem rather odd. Towering majestically over the citadel is the figure of Darwin. In squares and piazzas the other heroes of evolution stand in marmoreal splendor: Bateson, Morgan, Dobzhansky, Simpson, and, just completed, Lewis and Nüsslein-Vollhard. These are the grand architects of the evolutionary synthesis, and together they provide a narrative for everything from the study of variation and the genetic structure of populations to the remarkable discoveries of homeotic genes. Given, therefore, this history and the most recent and spectacular advances in molecular biology, it may seem curmudgeonly, if not perverse, to even hint that our understanding of evolutionary processes and mechanisms is incomplete. Yet, this review has exactly that intention.
There is a paradox in as much as sensible advances are usually only possible under a severely reductionist program, whereas questions basic to our understanding of evolution demand an encyclopedic knowledge of the science combined with an unprecedented skill in distillation and synthesis. Of these questions, perhaps the most fundamental is to explain the immense diversity of life despite its deep and pervasively similar molecular architecture. One answer is to invoke the presence of new genes, and then introduce them to a constantly shifting ecological and physical environment. In itself, however, this sort of answer is hardly sufficient. Here I wish to tackle four major questions that inevitably emphasize present ignorance but also, hopefully, indicate future possibilities: (1) What is the connection between phenotype and genotype? (2) Evolution has a temporal framework, but molecular ''clocks'' now plot a history of life seriously at odds with the fossil record. Which is correct? (3) Do data from molecular biology offer insights into mechanisms of macroevolution? (4) Constructing phylogenies is central to the evolutionary enterprise, yet rival schemes are often strongly contradictory. Can we really recover the true history of life?“ (Morris 2000).
It is the SSU rRNA (16S rRNA) present in all organisms, that still constitutes the single most comprehensive database available for phylum-level systematics. Because scientific enterprise should not acknowledge authorities, Carl Woese could be called an expert, when he writes in 2002:
"In the late 1990s it became possible to infer universal phylogenetic trees other than rRNA. Disturbingly, the majority of these yielded universal trees that differ significantly in topology from the rRNA tree."
The tripartite division of life into the three superkingdoms of Eukarya, Eubacteria and Archaebacteria was based on Woese's own rRNA work, and substituted the old division to prokaryotes and eukaryotes. It is said, that the kingspin on much of the current discussion of uprooting the tree of life is Woese, who has stirred the biotic soap for over 30 years.
The current conflict has been explained e.g. by the currently available protein data plagued by uneven taxonomic sampling, inadequate characterization, and proposed failure of the evolutionary mutation clock that is turning up along with the genome projects (Baldauf et al 2000).
In what seems to be days of trouble in many fronts, Woese belongs to the growing group of phylogenetists, who call upon the Horizontal Gene Transfer (HGT) to deliver the field. The deep lineage protein trees contradicting with rRNA differ in topology also from one another, and parsimony points out to a ancient "supercell", which is considered as a sign of HGT.
"What does it mean, then, to speak of an organismal genealogy when nearly all of the genes in the cell - genes that give its general character - do not share a common history? This question again goes beyond the classical Darwinian context." (Woese 2002, emphasis in the original.)
The new universal tree does not have any root in the classical sense, and even the unrooted tree -representation used in the three published neighbour joining/quartet puzzle trees of mine (Paunola et al 2002; Palmgren et al 2002; Vartiainen et al 2002; in the former two I am not included among the authors) deviates it in the dawn of life.
The evolutionary literature is in turmoil, and phrases used for the phenomenon include "bush of life", "chaos of the gene-exchange pool", or "evolutionary fluid conditions" before the "Darwinian Thresholds" of speciation. Doolittle writes (2000):
“Homology is a matter of quality, not quantity and the oxymoronic term ‘percent homology’ is seldom seen these days… It’s a short step from here back to ‘percent homology’. It is ironic that the words we seem to need in order to think productively about biology, words such as ‘homology’, ‘individual’, ‘organism’ and ‘species’, have no precise meaning”.
It seems that in many cases the narrative of a gene family is different than the narrative of the organism, and apart from the essay of Woese (2002), I suggest reading Bunk (2001), Forterre & Philippe (1999), or Martin & Muller (1998) to get a glimpse of the risen controversy.
When the human genome draft was finished, into the highlights was cast a bold announcement that we (or one of our vertebrate ancestors) have got not only viral, but also more than 200 bacterial genes directly from bacteria. The original report identified a set of 223 proteins as having “significant similarity to proteins from bacteria, but no comparable similarity to proteins from yeast, worm, fly and mustard weed…” (Lander et al 2001).
The public HGP included this “most parsimonious” explanation among its 7-11 major conclusions, and some newspapers made headlines of this particular announcement just under the media celebration of the human genome release itself. It is somewhat unfortunate, that the collaborators of the commercial Celera company at TIGR were the winners in this main difference of the results of the two projects, as it shot down HGP’s claim almost immediately in HGP's prominent journal (Salzberg et al 2001; Andersson et al 2001). Salzberg et al politely referred to inadequate invertebrate database searches and differences in evolutionary rates among lineages. What concerned these critics, rather amazing to me, was not the evolutionary request, but that
“...it’s a particularly dangerous claim they’re making, because it’s been picked up in the press that bacteria transfer genes into us, and people might speculate - although no evidence exists to support this - that GMOs [genetically modified organisms] could transfer their genes into humans” (GenomeWeb news article http://www.genomeweb.com/articles/view.asp?Article=2001320111458).
Empirically, HGP seems to be an attribute of unicellular organisms, most familiar for as E. coli on transformations for the protein preps of reprints, naturally. No evident “sex with bacteria“ in the case of us, despite the apparently great plead of the noisy audience.
The case of actin is interesting in the light of the “Complexity Hypothesis“, as introduced by Jain et al (1999) for the HGT. That is: genes of the operational class are transferred most frequently, whereas transfer of informational genes is hindered by the multiple intermolecular interactions. Operational genes cover metabolism etc., informational ones refer to genes usually associated to transcription, translation, replication and GTPases. One way to define the "informational" class is the greater amount of genetic interactions (Eisen 2000). Now, if something is manifested in particularly many interactions, it is actin, for which empirical reports of HGT is hard to find. I have been forced to extract actin from different tissues (yeast, muscles, spleen, thymus), but it would be interesting to test the recombinant over-production of prokaryotic "actins" in E. coli, and I am very interested on the possibility of a second actin for some bacteria in the databanks.
When demonstrating the tree-reconstruction artefacts, Philippe et al (2000) write:
"First, many genes used for inferring the eukaryotic phylogeny (e.g. rRNA, actin, tubulin and elongation factor 1) are highly mutationally saturated. Furthermore, saturated genes make believe in a molecular clock even if evolutionary rates are actually highly variable. This explains why rRNA has incorrectly been considered a good molecular chronometer and why LBA [long branching artefact] is so difficult to detect. Second, the evolutionary rates within different eukaryotic phyla have been estimated for several genes and it has been shown that the faster a phylum evolves the earlier it emerges (for example, euglenozoans for rRNA and ciliates for actin). This result is in agreement with our hypothesis that the order of emergence [in a tree] is dictated by the evolutionary rate (through the LBA artefact) and not by the historical pattern. Third, the addition of new sequences in phylogenetic analyses, which is known to reduce the impact of the LBA artefact, results in an upward movement of the early-branching species in the tree. This is also congruent with the fact that these early-branching species are, in fact, fast-evolving ones".
The ratio of number of ligands per isoformity could be an inhibitor to the change of platform. Recalling the misery of buying the correct dust-bags for vacuum cleaners, I again suggest that the paradigm of actin standardity would help to clarify the issue.
"The time has come for biology to go beyond the Doctrine of Common Descent" is the strongest formulation of the present crisis that I have found (Woese 2002). Woese is a microbiologist, and although he is advertizing his own subject, I am afraid he has simultaneously provided confessions to a more general discussion for ambitious novices.
If solid facts need to be distinguished from uncertainties, it can be stated that analogs of actin and tubulin come according to their kind, and that they are found ubiquitously in the biosphere. After they are established by whatever means, the actin is not prone to mutate quickly or to be transferred via HGT. The skeletal muscle actin in warm-blooded vertebrates is 100% identical in amino acid level.
I have published my arguments for the usage of the abbreviation Mryr (contra the absolute Myr), millions of radioactive years, in the World Wide Web: http://www.helsinki.fi/~pjojala/Mryr. The more precise term is defended mainly on the basis of contradicting results from genetic mutation rates and the brief history of biological material, like the alleged survival of 250 Myr old bacteria that was popularized as a proof for panspermia in the year 2000.
I have also published some case studies concerning the consensus problem associated with the evolutionary molecular clock hypothesis: http://www.helsinki.fi/~pjojala/Mryr. The appendixes have not gone through a peer review process and should be evaluated with extra caution.