Rapidly Evolving Rab GTPase Paralogs and Reproductive Isolation in Drosophila

Pierre Hutter    Division of Genetics, Institut Central des Hôpitaux Valaisans, Avenue Grand-Champsec 86, 1951 Sion, Switzerland

I Introduction

It has often been argued that Darwin did not solve the issue of “The Origin of Species” because small genetic variations can only account for differences between populations within a species and not account for divergence between species by macroevolution. Findings discussed here support the view that gene duplication, an event relevant to macroevolution, together with classical Darwinian microevolution, may be instrumental in the process of reproductive isolation. Indeed, a high rate of evolution of genes involved in reproductive isolation can be facilitated when these involve paralogs that are subject to less selective constraints than the parental form.

According to the biological species concept formulated by Mayr (1943), the creation of a new species requires that members of a population build up their own gene pool after becoming reproductively isolated from other members of the parental species. In organisms produced by crosses between populations in the state of incipient speciation, genetics posits that barriers to gene flow, such as lethality or sterility, result from genetic factors that exhibit functional incompatibilities. In hybrids between sibling species that have since long become distinct biological entities, it is difficult to identify, a posteriori, which of the genes associated with hybrid incompatibility (HI) actually reflect the primary factors that were instrumental to the initiation of speciation.

While both Cordon and Kosswig proposed as early as 1927 the concept of genetic HI in the fish genus Xiphophorus, entry points for experimental studies on this major biological issue only emerged over the last 25 years (Cordon, 1927; Kosswig, 1927). A handful of observations have indicated that single genetic changes can drastically influence reproductive isolation between most closely related species. The first mutations that can break through reproductive isolation were reported in organisms amenable to genetic analysis, such as Lhr and mhr in Drosophila simulans (Sawamura et al., 1993a; Watanabe, 1979), as well as Hmr and Zhr in D. melanogaster (Hutter and Ashburner, 1987; Hutter et al., 1990; Sawamura et al., 1993b). These genetic changes are sufficient to restore full viability in otherwise completely inviable hybrids, and some of these alterations also partly restore female fertility. Similarly, mutations in otherwise fully fertile flies were found to play a significant role in causing sterility of hybrids between Drosophila species (see Wu and Ting, 2004 for a review). Because genetic alterations responsible for these rescues have no phenotypic effect within a given species, they are thought to reflect incompatibilities in hybrid genotypes having to do with genes that have diverged functionally.

As Coyne and Orr (2004) emphasized, the first molecular studies of these genes strongly suggested that they represent the actual genes that cause the death or the sterility of hybrids rather than being second-site suppressors that might ameliorate effects of the loci causing hybrid problems. Thus, these genetic alterations influencing HI are our best candidates to explore the scenario of reproductive isolation based on a minimum of two genes, as originally envisioned by Dobzhansky (1937) and Müller (1940). Noteworthy, in D. melanogaster all mutations that rescue otherwise lethal hybrids were found in genes of the X chromosome, and previous studies in other Drosophila species have also indicated a major effect of the X chromosome on hybrid inviability (reviewed in Hutter, 1997). Genes that cause HI as a result of positive selection are predicted to be often X-linked, as recessive X-linked alleles can become fixed more quickly than autosomal genes and are likely to evolve rapidly (Charlesworth et al., 1987).

Under laboratory conditions, D. melanogaster can hybridize with its three most closely related species, D. simulans, D. mauritiana, and D. sechellia (hereafter referred to as the sibling species), even though the latter species have diverged from an ancestor of D. melanogaster approximately 2–3 million years (Myr) ago (Lachaise et al., 1988). However, crosses between D. melanogaster females and males of any of its three sibling species produce hybrid daughters that are viable but sterile at low temperatures and lethal at high temperatures. Hybrid sons invariably die as late larvae or pseudopupae and never metamorphose (Hadorn, 1961; Hutter et al., 1990; Lachaise et al., 1988; Sturtevant, 1920). Three D. melanogaster mutants—Hmr1, In(1)AB, and Df(1)EP307–1-2—were found to rescue otherwise lethal hybrids, and all three mutations map to cytological interval 9D–9E in the middle of the X chromosome (Barbash et al., 2000, 2004b; Hutter and Ashburner, 1987; Hutter et al., 1990). The first two mutations not only rescue lethal hybrids, but also contribute to restore fertility in otherwise sterile hybrid females (Barbash and Ashburner, 2003).

During recent years, three studies have addressed the molecular basis of the genetic factors capable to suppress the invariant lethality of hybrids. The first report postulated that a cluster of six paralogs in D. melanogaster, which encode RAB GTPase proteins, is involved in HI between members of the D. melanogaster subgroup species (Hutter, 2002). Sexually antagonistic coevolution between the Rab paralogs and extranuclear components were hypothesized to result in fast evolution of genes involved in vesicle trafficking and cell signaling. The second molecular study, based on transgenesis experiments, identified a single gene as Hmr (Barbash et al., 2003), lying in the middle of the above cluster of six Rab paralogs. Hmr encodes a regulatory protein with homology to a family of MADF- and MYB-related DNA-binding transcriptional regulators and was shown to be evolving particularly fast as a result of strong positive selection (Barbash et al., 2003, 2004a). Indeed, an unusually high average divergence rate of 7.7% for nonsynonymous nucleotide substitutions [leading to amino acid (AA) replacements] was observed in HMR protein between D. melanogaster and its sibling species. Nonetheless, in In(1)AB and Df(1)EP307–1-2 D. melanogaster mutants which rescue the same interspecific hybrids as Hmr1 does, no causative mutation has yet been attributed to Hmr. This gene appears normally transcribed in both above-mentioned mutants by Northern analysis (Barbash et al., 2003). The third molecular study on HI in Drosophila, based on complementation mapping of HI genes between D. melanogaster and D. simulans, identified a gene called Nup96, on the third chromosome of...