Can molecular biology and bioinformatics be used to probe an evolutionary pathway?

Williamson has proposed an interesting hypothesis regarding the evolution of caterpillars from onycophorans by hybridogenesis “as an addition to Darwin's great insight into the generation of diversity ” (1). He has proposed some molecular biology-/ bioinformatics-based experiments to test his theory of hybridogenesis. Although his theory is interesting, we think that the experiments he proposes are too simplistic and will not necessarily shed much light on the topic.

Contrary to his suggestion, the insects that undergo development via a larval stage do not necessarily need to have a larger genome as compared to the ones that lack a larval stage during their development. Besides, even if the insects with a larva show a bigger genome, this does not show that they have a different evolutionary path as suggested by the author. The insects with larva will presumably have a more intricate physiology, and hence its genome will have many coding as well as noncoding (including repetitive) sequences, leading to a bigger genome size. Hence, the relative genome sizes might be indicative of the complexity of these organisms and not be related to their evolution. Similarly, the onychophoran genome being smaller than those of the holometabolan insects does not prove his hypothesis. It is important to note that genome size does not need to correlate with the number of genes or the complexity of the organism (2).

Williamson's other proposal is that holometalobolans will have “more base-pairs of protein coding chromosomal DNA” as compared to the exopterzygote insects (1). This is also not required. It is possible (in fact probable) that the protein diversity generated in holometalobolans is a function of alternative splicing and/or other recombination/processing events. In this case, these species do not need to have more base pairs to code for their proteins (3).

Williamson (1) proposes that two distinct set of genes should be detected in the genomes of all insects with caterpillar grub- or maggot-like larvae (a set of onychophoran-derived genes encoding for the larval stage and insect genes for the adult proteins), and this is the acid-test of his theory. We feel that this is a more complex phenomenon. In fact, if hybridogenesis is indeed the process of evolution, then it is quite possible that there has been a recombination of the set of genes, and thus the distinctness of the individual sets of genes will not be obvious from a bioinformatics study.

We would, however, encourage a laboratory hybridization study between an onychophore and a cockroach, as proposed by Williamson (1). Our proposal to test the hypothesis would be a rigorous study of the “sequence similarity” between an extant onychophoran and a holometabolan insect at the genome level. If a satisfactory degree of homology is found, then the metabolically essential proteins of onychophores should be tested for their sequence similarity with the holometabolans. A high degree of homology should help advance the theory proposed by Williamson (1).