Problems in Estimating Horizontal Gene Transfer

When the human genome was sequenced, several hundred human genes were at first attributed to horizontal transfer from bacteria. However, later analyses indicated that very few of these were genuine cases of horizontal transfer (see Ch. 24). Several factors have contributed to such over-estimates of horizontal transfer, both for the human genome and in other cases.

a. Sampling Bias. Relatively few eukaryotic genomes have been fully sequenced whereas hundreds of bacterial genomes have been sequenced. Thus the absence of sequences homologous to a human gene from a handful of other eukaryotes is insufficient evidence for an external (bacterial) origin. As more eukaryotic sequence data has become available many genes supposedly of "bacterial" origin have in fact been found in other eukaryotes.

b. The loss of homologs in related lineages may suggest that a gene originated externally to the group of organisms that retain it. As in the related case (a) above, the solution to this artifact is the collection of more sequence data from many related lineages.

c. Gene duplication followed by rapid divergence may give rise to apparently novel genes that are missing from the direct vertical ancestor of a group of organisms.

d. Intense evolutionary selection for a particular gene may result in a greatly increased rate of sequence alteration. Those genes that evolve faster than normal will tend to be misplaced when sequence comparison is used to construct evolutionary trees.

e. The ease of horizontal transfer of genetic information by plasmids, viruses, transposons under laboratory conditions is misleading. Under natural conditions there are major barriers to such movements. Furthermore, the results of horizontal transfer are often only temporary. Newly acquired genes, especially those on plasmids, transposons, etc., are easily lost. Such genes tend to be acquired in response to selection such as antibiotic resistance and, conversely, they will be lost when the original selective conditions disappear.

f. Experimental problems such as DNA contamination. Bacterial and viral parasites are associated with esentially all higher organisms and completely purifying the eukaryotic DNA is not always easy.

Many of the originally proposed examples of widespread horizontal gene transfer have been severely compromised by the above factors. However, some examples do seem to be valid. One of the most interesting is the recent finding of relatively frequent horizontal gene transfer between the mitochondrial genomes of flowering plants. The genes for certain mitochondrial ribosomal proteins have apparently been transferred from an early monocotyledonous lineage to several different dicotyledonous lineages. Examples include transfer of the rps2 gene to kiwifruit (Actinidia) and the rps11 gene to bloodroot (Sanguinaria).

0 0

Post a comment