Gene flux from bacteria to nematodes has been proposed to have played a role in the evolution of parasitism, and we developed scripts to flag Meloidogyne genes more similar to bacterial sequences than to sequences in the C. elegans or D. melanogaster genomes. Using filters with low threshold values, we have identified numerous potential candidates which will be subjected to two types of analysis: 1) a Markov chain Monte Carlo technique developed for studying co-speciation, which is a statistically similar problem. 2) phylogenetic analysis using both sequence and information on the absence of genes in genomes due to gene loss. The idea is that the topology and branch lengths of an evolutionary tree affect the number of times that gene loss must be postulated as well as the probabilities of these gene loss events. Conventional methods for statistical analysis of phylogenies ignore information on the presence/absence of genes in genomes. We have five strong candidates, of which our best to date is the Meloidogyne nodL gene.

BLAST analysis showed that the deduced product of M. incognita NemaGene cluster ID#1450 is most similar to the rhizobial-specific, NODL O-acetylase (score = 194; P = 5e-18). Experimental demonstration of a trans-spliced leader on the nematode nodL mRNA, and the presence of introns in the gene confirm this is not a bacterial contaminant. Maximum parsimony of ClustalW-aligned sequences confirmed the relationship (some values from 1,000 bootstraps are shown).

We have confirmed the presence of nodL in M. javanica and M. arenaria and are examining other Meloidogyne spp. We have been unable to identify this gene in the cyst nematode Heterodera glycines and are examining other members of the family.