Positive selection of haplotypes in the middle of the X chromosome and reproductive isolation between members of melanogaster species group
As limited evidence is available to support the hypothesis presented below, I would like to take the opportunity of this forum to exchange views on it, hoping that someone is working on the Rab GTPase genes clustered in cytological region 9 of D. melanogaster.
The Hmr gene, involved in both inviability and sterility of hybrids between members of the Drosophila melanogaster species group, has remained consistently linked to the Rab9D gene. In the middle of the X chromosome synteny has been conserved for at least 45 million years between these two genes separated by only one kb of genomic, and each of them has recently been shown to be subject to particularly strong positive selection. Besides, Rab9D has five paralogs in region 9, three of which (namely Rab9Db, Rab9E and Rab9Fa) encode proteins which share 98-99% amino acid identity with RAB9D protein. The precise functions of these four monoexonic Rab GTPases paralogs are not known, but a BLAST search indicate that they are most similar to RAB8 and RAB13 GTPases. These two proteins coordinate the assembly of adherens junctions and tight junctions, and at least in human they influence endosomal cholesterol efflux. Both Hmr1 and Hmr2/In(1)AB mutants in D. melanogaster rescue otherwise inviable hybrids between D. melanogaster and its most closely related species. Aruna et al (Genetics 181: 1437-1450, 2009) recently reported that the In(1)AB chromosome is mutant for Hmr (i.e.Hmr2) function, which is associated with a regulatory protein with homology to MADF- and MYB-related DNA binding transcriptional regulators. Inactivating only one (Rab9D, 1 kb away from Hmr) of the four nearly identical Rab paralogs, appears to have a weak effect in hybrids as slightly improved hybrid viability, while having a significantly deleterious effect as decreased viability in non-hybrid flies (Hutter, Advances in Genetics, 2007). With regard to this, I have discussed the possibility of positively selected associations between Hmr and nearby genes involved in intracellular traffic, across 500 kb of DNA (Dev. Genes Evol, 2002). I would like to further comment here on the observation of concomitant alterations affecting these X-linked genes.
In D. melanogaster, the X chromosome carrying the hybrid rescuing mutant Hmr1 also harbors an allele of Rab9D (1 kb apart) encoding a 197-amino acid protein with four amino acid substitutions, and corresponding to an haplotype so far only observed on this Hmr1 chromosome. The RAB9D alterations lie within the effector interaction domain and within the Switch I and II regions of the GTPase. Similarly, I observed that the distal breakpoint (in 9E) of In(1)AB inversion present on the Hmr2/In(1)AB chromosome, lies 406 base pairs before the 3’ end of the four annotated transcripts from the CG15211 gene, which encodes a protein containing a MARVEL domain, predicted to have a four transmembrane-helix architecture. Thus, CG15211 function appears to be related to cholesterol-rich membrane apposition events in cellular processes such as biogenesis of vesicular transport carriers, or tight junction regulation. Moreover, in the above 2007 paper as well as in recent personal communications to Flybase, I have reported that segments of a variable 360 bp satellite repeat from D. melanogaster, present at many X linked sites, are over-represented near the 5’ and 3’ ends of Rab9D and of each of its five neighboring paralogs (namely RabX2, Rab9Db, Rab9E, Rab9Fa and Rab9Fb), near the 3’ end of the rox1 gene (involved in dosage compensation of X-linked genes), as well as at the Dox and Nmy genes (involved in sex ratio distortion) from D. simulans and D. mauritiana. I also emphasized that the above 360 bp repeat is partly transcribed from the CG42611 (=SOSIE) gene, pointing to a possible regulation by RNA interference. When considering a hypothetical link between CG15211 (presumably involved in cholesterol traffic) and nearby Rab GTPases/Hmr in region 9, it is worth noting that CG42611-SOSIE has a low-density lipophorin receptor activity related to the cell surface receptors involved in cholesterol trafficking. Taken together, these observations underline concomitant alterations on the two main hybrid male rescuing X chromosomes known in D. melanogaster, carrying either Hmr2/In(1)AB and truncated CG15211, or Hmr1/multivariant Rab9D-and its linked 360 bp repeats possibly targeted by CG42611-SOSIE RNA. Functional association of mutations clustered in positively selected haplotypes at the middle of the X chromosome may deserve further attention, in relation to the process of reproductive isolation. Cooperation between alterations reported here may point to a pathway involving cholesterol, which is typically circulated to fat body via lipophorin in a process in which endocytosed proteins are transported to various intracellular compartments, with Rab GTPases being critical for coordinating vesicle formation, transport and fusion with target membranes. Drosophila is a cholesterol auxotroph, since a dietary supply of cholesterol is an absolute requirement for viability and development. Diverging diets between incipient species may occasionally promote the evolution of specific components of finely orchestrated traffic of sterols and of fatty acid metabolism. In a hybrid genotype, impairment of this intracellular transport or defective target membrane recognition may become critical when a high amount of lipid is needed, for instance for hormone biosynthesis at the onset of metamorphosis. Whenever abnormal traffic of dietary sterols may lead to functional incompatibility compromising metamorphosis in hybrids, cholesterol would stand as a favorite candidate as a causative agent, since it is the most frequently used precursor of ecdysteroid moulting hormones, such as beta-ecdyson. It is interesting to point out here that man has largely exploited agonists of ecdysteroids (e.g. neutralizing ecdyson receptors) to combat insect pests in a species specific manner. The phenotype of very fat male hybrid larvae rescued by Hmr1 or Hmr2/In(1)AB, but otherwise unable to enter metamorphosis, appears to be compatible with defects in sterol trafficking, maybe in relation to adipose tissue hypoxia. With regard to the above mentioned genes involved in rescue of inviable hybrid male larvae (Hmr), in meiotic drive (Dox and Nmy) and in dosage compensation of X-linked genes (rox1), it may be worth bearing in mind that cholesterol is also a precursor of sex hormones. The widespread role of ecdysteroids as protection agents against parasites might have occasionally conferred a role in the evolution of sex on ecdysteroids, following the scenario referred to as the Red Queen hypothesis. An X linked cluster of genetic factors whose combined mutations would result in a selective advantage with respect to both the evolution of sex through genetic recombination, and the differentiation of sex chromosomes (processes somehow associated with dosage compensation and control of sex ratio), might represent a most wanted candidate for speciation. Indeed, the above sex-related characteristics are deemed to recurrently generate intra- and inter organism genetic conflicts, thus exhibiting genetic incompatibilities between carriers of diverging haplotypes in the middle of the X (as discussed in my above 2002 paper). The aforementioned 360 bp satellite repeats in D. melanogaster may bring about genetic incompatibilities for instance through extended modification of chromatin regulation, differently affecting the expression of X linked genes comprised within distinct haplotypes. With regard to the above speculations, it would seem worth characterizing the functions of the four Rab GTPase paralogs from region 9 in D. melanogaster, as well as searching for additional alterations in the three remaining paralogs (Rab9Db, Rab9E and Rab9Fa) which are nearly identical to Rab9D, from the chromosomes carrying either Hmr1 or Hmr2 mutant.