A novel approach to detecting and measuring recombination: new insights into evolution in viruses, bacteria, and mitochondria.

An accurate estimate of the extent of recombination is important whenever phylogenetic methods are applied to potentially recombining nucleotide sequences. Here, data sets from viruses, bacteria, and mitochondria were examined for deviations from clonality using a new approach for detecting and measuring recombination. The apparent rate heterogeneity (ARH) among sites in a sequence alignment can be inflated as an artifact of recombination. However, the composition of polymorphic sites will differ in a data set with recombination-generated ARH versus a clonal data set that exhibits the equivalent degree of rate heterogeneity. This is because recombinant data sets, encompassing regions of conflicting phylogenetic history, tend to yield "starlike" trees that are superficially similar to those inferred from clonal data sets with weak phylogenetic signal throughout. Specifically, a recombinant data set will be unexpectedly rich in conflicting phylogenetic information compared with clonally generated data sets supporting the same tree shape. Its value of q-defined as the proportion of two-state parsimony-informative sites to all polymorphic sites-will be greater than that expected for nonrecombinant data. The method proposed here, the informative-sites test, compares the value of q against a null distribution of values found using Monte Carlo-simulated data evolved under the null hypothesis of clonality. A significant excess of q indicates that the assumption of clonality is not valid and hence that the ARH in the data is at least partly an artifact of recombination. Investigations of the procedure using simulated sequences indicated that it can successfully detect and measure recombination and that it is unlikely to produce "false positives." Simulations also showed that for recombinant data, naïve use of maximum-likelihood models incorporating rate heterogeneity can lead to overestimation of the time to the most recent common ancestor. Application of the test to real data revealed for the first time that populations of viruses, like those of bacteria, can be brought close to complete linkage equilibrium by pervasive recombination. On the other hand, the test did not reject the hypothesis of clonality when applied to a data set from the coding region of human mitochondrial DNA, despite its high level of ARH and homoplasy.

Evidence for the non-quasispecies evolution of RNA viruses [corrected].

The quasispecies model of RNA virus evolution differs from those formulated in conventional population genetics in that neutral mutations do not lead to genetic drift of the population, and natural selection acts on the mutant distribution as a whole rather than on individual variants. By computer simulation, we show that this model could be inappropriate for many RNA viruses because the neutral sequence space may be too large to allow the formation of a quasispecies distribution. This view is supported by our analysis of gene sequences from vesicular stomatitis virus, which is considered a prototype RNA virus quasispecies. Our results are relevant to the evolution of RNA systems in general.

Homologous recombination in GB virus C/hepatitis G virus.

Analysis of 33 GB virus C/hepatitis G virus (GBV-C/HGV) full or nearly full genome sequences revealed several putative inter- and intrasubtype recombinants. The breakpoints of the recombinant regions were mapped using a maximum-likelihood method, and the statistical significance for each region was tested using Monte Carlo simulation. The results were highly significant and provided evidence for the existence of complex mosaic genomes showing as many as nine recombination events, with breakpoints in the 5' UTR and in all of the coding regions except the short NS4b gene. Recombination was confirmed by separate phylogenetic analysis of the various recombinant regions and by Sawyer's runs test. Taken together, these findings demonstrate for the first time that recombination is common in natural populations of GBV-C and that it takes place both within and between subtypes. The wide-ranging implications of such nonclonal history for reconstructing the spread and timescale of GBV-C evolution are discussed.

Extensive homologous recombination among widely divergent TT viruses.

Analyses of a collection of full-length TT virus genomes showed nearly half of them to be recombinant. The results were highly significant and revealed homologous recombination both within and among genotypes, often involving extremely divergent lineages. Recombination breakpoints were significantly more common in the noncoding region of the TT virus genome than in the coding region.

Widespread intra-serotype recombination in natural populations of dengue virus.

Diversity analysis of 71 published dengue virus gene sequences revealed several strains that appeared to be mosaics comprising gene regions with conflicting evolutionary histories. Subsequent maximum likelihood breakpoint estimation identified seven recombinants, including members of three of the four dengue virus serotypes, with breakpoints in the premembrane/membrane gene, the envelope gene, and at the junction of the envelope and first nonstructural genes. Many of the individual recombinants contain sequence representing separate genetic subtypes. The results were highly statistically significant and were confirmed by phylogenetic analysis of the regions of interest. These findings indicate that recombination may play a very significant role in shaping genetic diversity in dengue virus and, as such, have important implications for its biology and its control.

Phylogenetic evidence for recombination in dengue virus.

A split decomposition analysis of dengue (DEN) virus gene sequences revealed extensive networked evolution, indicative of recombination, among DEN-1 strains but not within serotypes DEN-2, DEN-3, or DEN-4. Within DEN-1, two viruses sampled from South America in the last 10 years were identified as recombinants. To map the breakpoints and test their statistical support, we developed a novel maximum likelihood method. In both recombinants, the breakpoints were found to be in similar positions, within the fusion peptide of the envelope protein, demonstrating that a single recombination event occurred prior to the divergence of these two strains. This is the first report of recombination in natural populations of dengue virus.

Phylogenetics of social behavior in Australian gall-forming thrips: evidence from mitochondrial DNA sequence, adult morphology and behavior, and gall morphology.

Six species of Australian gall-forming thrips (Insecta: Thysanoptera) on Acacia exhibit soldier castes, individuals with reduced wings and enlarged forelegs that defend their gall against interspecific invaders. We used data from two mitochondrial genes (cytochrome oxidase I and 16S rDNA), adult morphology and behavior, and gall morphology to infer a phylogeny for Acacia gall-forming thrips with and without soldiers, and we used this phylogeny to evaluate hypotheses concerning soldier evolution. Phylogenies inferred from each data set analyzed separately yielded large numbers of most-parsimonious trees and weak support for most nodes. However, when analyzed together the data sets complemented and reinforced one another in such a way as to yield a well-resolved phylogeny. Our phylogeny implies that soldiers originated once or twice early in the history of this clade, that soldiers were lost once or twice, and that soldiers evolved from winged dispersers rather than from nonsoldier within-gall reproductive offspring of foundresses. The phylogeny also provides evidence for long-term morphological stasis, an ancient split between eastern and western gall thrips species, and a high degree of conservatism in host-plant affiliations.