An amino acid polymorphism in the couch potato gene forms the basis for climatic adaptation in Drosophila melanogaster
Most organisms are faced with dealing with seasonal variations in environmental conditions. As winter approaches, physiological changes need to be implemented: deciduous trees drop their leaves, mammals may hibernate, and so forth. In the case of many insects, the strategy is to move into a diapause state. This may be in any of the life stages - pupal diapause, larval diapause etc, and in the case of Drosophila melanogaster reproductive diapause, in which ovarian activity is shut down in response to a combination of short day length and low temperature. My interests in diapause are two-fold in origin: firstly, diapause appears to have an impact on lifespan in Drosophila, and secondly, my father identified the existence of reproductive diapause in Drosophila back in 1989.
D. melanogaster is a human commensal that originated in sub-Saharan Africa, and which has spread alongside humans (due in large part, perhaps, to the legacy of T. H. Morgan) over the last 5,000-16,000 years. What's quite interesting here is that there is a latitudinal cline of diapause incidence in the eastern USA. Given that Drosophila is a relatively recent immigrant to the New World, this offers an attractive insight into novel physiological adaptation to environmental stress. This paper looks into the genetic basis (or, more correctly, one of the genetic bases) for variation in diapause incidence.
Prior research had shown that a major factor influencing expression of the diapause characteristic lies on the third chromosome. In this paper, a set of 250 recombinant inbred lines generated by recombination between the VT46 strain (which does not show diapause) and the 6326 (which does show diapause). Te recombinant lines were subjected to QTL analysis using an initial set of 15 SNP markers on the third chromosome, which revealed a strong QTL in 90D-92E (for a description of polytene chromosome map cartography, see this prior post), as shown in the figure below (which may require subscription to PNAS).
Further mapping localised the QTL to a 248 kb region, more specifically in a candidate gene, couch potato (cpo). cpo mutants show a variety of phenotypes consistent with a neurodevelopmental role, and cpo is known to be expressed inthe principal neuroendocrine organ, the ring gland. Independent evidence of the role of cpo in diapause was provided by a detailed examination of the effect of several cpo alleles on diapause. cpo is expressed at a much reduced level in the VT46 strain than in the 6326 strain.
Next up was to investigate whether a polymorphism in cpo was associated with diapause. 3.5kb encompassing cpo exon 5 was sequenced from 35 third chromosomes isolated from a single natural population, the rather splendidly named Davis Peach Farm strain (DPF): these chromosomes had all been placed in a 6326 genetic background (recall that 6326 is a diapausing strain). The only polymorphic sites associated with diapause phenotype lie towards the 3' end of exon 5, and reflect the two major haplotypes identified in prelininary experiments. Two polymorphisms resulted in amino acid substitutions: Alanine>Valine ay position 356 and Isoleucine>Lysine at position 462, with I462L having the strongest association (see the figure below).
Following up the latitudinal cline data, the authors looked for latitudinal clines in polymorphism frequency. Here, my lack of statistical experience bites somewhat, but it appears the 356 polymorphism does show a latitudinal cline, as does the 462 polymorphism (though this seems to be partly in inference due to it being in linkage disequilibrium with the 356 polymorphism).
The evidence seems clear that cpo plays a role in diapause in Drosophila. A number of observations are consistent: expression in the ring gland, and ecdysone-responsive elements associated with cpo. Interestingly recently published work from Williams et al (2006) [PNAS 103;15911-15915] polymorphism in dp110 is responsible for natura variation in diapause. Thiswas interesting because of dp110's role as an insulin-regulated PI-3 kinase. (Insulin signalling is linked to longevity and to dauser formation in C. elegans - dp110 is the Drosophila homologue of the C. elegans gene age-1). dp110 maps at 92F3, not terribly far from cpo. Williams' work suggest that there may be a link between aspects of diapause induction and insulin signalling.
On the evolutionary origin of diapause, it's interesting to note that African populations of D. melanogaster, and its sibling species Drosophila simulans lack a diapause response: either diapause is absent in those populations, or the genetic factors that enable it are rare in those populations. The detailed function of cpo related to diapause remain to be elucidated - there are six transcripts produced via differential splicing, and all but one of the ORFs in these transcriots posses an RRM RNA recognition motif, and this transcript is the only one affected by the residue 462 polymorphism. In the protein encoded by that splice variant, the RNA binding domain is replaced by a highly basic arginine/lysine rich terminus, and it is unclear what its function might be.
Finally, one might observe that diapause is likely to be a polygenic trait and other genetic components of diapause induction undoubtedly remain to be discovered.
The paper does, however, raise a few questions in my mind (which I suppose is desirable!). If the original D. melanogaster populations lacked diapause, but modern dispersed populations have managed to acquire the diapause characteristic via single gene alterations (e.g. in cpo and dp110), then it seems to me that either the manifestion of diapause is physiologically simple, or African strains of D. melanogaster already had diapause systems in place, but not used.