Posted: Mar 18, 2011 2:40 am
Indeed! HGT mainly occurs if a single gene codes for a single trait. This is why anti-biotic resistance is easily transfer between non-related bacteria, and sometimes between bacteria and eucaryotes, such as Wolbachia species infecting arthropods. The Wolbacia case is particularly interesting because it interferes with the reproduction of it's host.
Werren, J. H., W. Zhang, et al. (1995). "Evolution and phylogeny of Wolbachia: reproductive parasites of arthropods." Proceedings: Biological Sciences 261(1360): 55-63.
http://www.rochester.edu/college/bio/la ... lution.pdf
Because traits are often under the control of complex modular gene regulatory networks with multiple signaling pathways, such traits are not usually transferred by HGT. In other words, the signaling and network machinery is too large to be transferred in one 'go", although it does happen. So basically the 'tree" concept is basically sound. If it were not, if the tree of life were in fact a tangled web, then you would get 'crocoducks". So the tree of life is best seen still as a tree, with lots of thin creepers connecting the branches. Organisms maintain their "identity" even if they experience extensive HGT. Thus no 'fly-men" or human flies. This is because development is hard to change early on, but easier to change in later development. The evolutionary consequence of this is that innovation mainly occurs in late development and early development is highly conserved.
Cohen, O., U. Gophna, et al. (2011). "The Complexity Hypothesis Revisited: Connectivity Rather Than Function Constitutes a Barrier to Horizontal Gene Transfer." Molecular Biology and Evolution 28(4): 1481-1489.
Horizontal gene transfer (HGT) is a prevalent and a highly important phenomenon in microbial species evolution. One of the important challenges in HGT research is to better understand the factors that determine the tendency of genes to be successfully transferred and retained in evolution (i.e., transferability). It was previously observed that transferability of genes depends on the cellular process in which they are involved where genes involved in transcription or translation are less likely to be transferred than metabolic genes. It was further shown that gene connectivity in the protein–protein interaction network affects HGT. These two factors were shown to be correlated, and their influence on HGT is collectively termed the “Complexity Hypothesis”. In this study, we used a stochastic mapping method utilizing advanced likelihood-based evolutionary models to quantify gene family acquisition events by HGT. We applied our methodology to an extensive across-species genome-wide dataset that enabled us to estimate the overall extent of transfer events in evolution and to study the trends and barriers to gene transferability. Focusing on the biological function and the connectivity of genes, we obtained novel insights regarding the “complexity hypothesis.” Specifically, we aimed to disentangle the relationships between protein connectivity, cellular function, and transferability and to quantify the relative contribution of each of these factors in determining transferability. We show that the biological function of a gene family is an insignificant factor in the determination of transferability when proteins with similar levels of connectivity are compared. In contrast, we found that connectivity is an important and a statistically significant factor in determining transferability when proteins with a similar function are compared.
This is another interesting article about phylogenetic relationships by Cavalier-Smith:-
Cavalier-Smith, T. (2010). "Deep phylogeny, ancestral groups and the four ages of life." Philosophical Transactions of the Royal Society B: Biological Sciences 365(1537): 111-132.
Organismal phylogeny depends on cell division, stasis, mutational divergence, cell mergers (by sex or symbiogenesis), lateral gene transfer and death. The tree of life is a useful metaphor for organismal genealogical history provided we recognize that branches sometimes fuse. Hennigian cladistics emphasizes only lineage splitting, ignoring most other major phylogenetic processes. Though methodologically useful it has been conceptually confusing and harmed taxonomy, especially in mistakenly opposing ancestral (paraphyletic) taxa. The history of life involved about 10 really major innovations in cell structure. In membrane topology, there were five successive kinds of cell: (i) negibacteria, with two bounding membranes, (ii) unibacteria, with one bounding and no internal membranes, (iii) eukaryotes with endomembranes and mitochondria, (iv) plants with chloroplasts and (v) finally, chromists with plastids inside the rough endoplasmic reticulum. Membrane chemistry divides negibacteria into the more advanced Glycobacteria (e.g. Cyanobacteria and Proteobacteria) with outer membrane lipolysaccharide and primitive Eobacteria without lipopolysaccharide (deserving intenser study). It also divides unibacteria into posibacteria, ancestors of eukaryotes, and archaebacteria—the sisters (not ancestors) of eukaryotes and the youngest bacterial phylum. Anaerobic eobacteria, oxygenic cyanobacteria, desiccation-resistant posibacteria and finally neomura (eukaryotes plus archaebacteria) successively transformed Earth. Accidents and organizational constraints are as important as adaptiveness in body plan evolution.
Thus a single trait [such as antibiotic resistance] can be tranferred easily between -say- E.coli and Psuedomonas.sp, but E.col remain E.coli, and psuedomonas remain pseudomonas. At the eukaryote level Humans remain humans even after a retrovirus like HIV adds its code tothe human genome. Of course, this does affect the evolution of the host, and many endogenous retroviruses are known to "infect" the human genome. But HGT cannot turn a human into a bat or fly. The embryological development has diverged too significantly between taxa to allow this to happen. Although all metazoans share many Hox genes, the expression networks of these Hox clusters and other regulatory genes is vastly different as is the signaling pathways that connect these modules. HGT can never transfer a Hox cluster. It is just too big. Hence, no 'crocoducks".