Selection on X-Linked Genes during Speciation in the Drosophila Athabasca Complex

We present the results of a restriction site survey of variation at five loci in Drosophila athabasca, complimenting a previous study of the period locus. There is considerably greater differentiation between the three semispecies of D. athabasca at the period locus and two other X-linked genes (no-on-transient-A and E74A) than at three autosomal genes (Xdh, Adh and RC98). Using a modification of the HKA test, which uses fixed differences between the semispecies and a test based on differences in Fst among loci, we show that the greater differentiation of the X-linked loci compared with the autosomal loci is inconsistent with a neutral model of molecular evolution. We explore several evolutionary scenarios by computer simulation, including differential migration of X and autosomal genes, very low levels of migration among the semispecies, selective sweeps, and background selection, and conclude that X-linked selective sweeps in at least two of the semispecies are the best explanation for the data. This evidence that natural selection acted on the X-chromosome suggests that another X-linked trait, mating song differences among the semispecies, may have been the target of selection.     

Fine Mapping of Dominant X-Linked Incompatibility Alleles in Drosophila Hybrids

Sex chromosomes have a large effect on reproductive isolation and play an important role in hybrid inviability. In Drosophila hybrids, X-linked genes have pronounced deleterious effects on fitness in male hybrids, which have only one X chromosome. Several studies have succeeded at locating and identifying recessive X-linked alleles involved in hybrid inviability. Nonetheless, the density of dominant X-linked alleles involved in interspecific hybrid viability remains largely unknown. In this report, we study the effects of a panel of small fragments of the D. melanogaster X-chromosome carried on the D. melanogaster Y-chromosome in three kinds of hybrid males: D. melanogaster/D. santomea, D. melanogaster/D. simulans and D. melanogaster/D. mauritiana. D. santomea and D. melanogaster diverged over 10 million years ago, while D. simulans (and D. mauritiana) diverged from D. melanogaster over 3 million years ago. We find that the X-chromosome from D. melanogaster carries dominant alleles that are lethal in mel/san, mel/sim, and mel/mau hybrids, and more of these alleles are revealed in the most divergent cross. We then compare these effects on hybrid viability with two D. melanogaster intraspecific crosses. Unlike the interspecific crosses, we found no X-linked alleles that cause lethality in intraspecific crosses. Our results reveal the existence of dominant alleles on the X-chromosome of D. melanogaster which cause lethality in three different interspecific hybrids. These alleles only cause inviability in hybrid males, yet have little effect in hybrid females. This suggests that X-linked elements that cause hybrid inviability in males might not do so in hybrid females due to differing sex chromosome interactions.     

Characterization of X-Linked Recessive Lethal Mutations Affecting Embryonic Morphogenesis in Drosophila Melanogaster

This study attempted to assay the zygotic contribution of X chromosome genes to the genetic control of embryonic morphogenesis in Drosophila melanogaster. A systematic screen for X-linked genes which affect the morphology of the embryo was undertaken, employing the phenotype of whole mount embryos as the major screening criterion. Of 800 EMS-induced lethal mutations analyzed, only 14% were embryonic lethal, and of these only a minority affected embryonic morphogenesis. By recombination and complementation analyses, the mutations that affected embryonic morphogenesis were sequestered into 26 complementation groups. Fourteen of the loci correspond to genes previously identified in a large-scale screen in which fixed cuticles were examined, and 12 new loci have been identified. Most of the mutations which disrupt embryonic morphology had specific and uniform mutant phenotypes. Mutations were recovered which disrupt major morphogenetic events such as gastrulation, germ band retraction and head involution. No mutations were found which arrest the embryos prior to blastoderm formation. However, a novel class was found, one comprised of mutations which interfere with the development of internal structures but not cuticular structures. Nevertheless, saturation of the X chromosome for genes important for embryonic morphogenesis is probably incomplete.     

X-Linked Elements Associated with Negative Segregation Distortion in the Sd System of Drosophila Melanogaster

Three elements, M(1), M(2) and M(3), found in a special X chromosome, supp-X(SD), modify the degree and direction of segregation distortion in the SD system of Drosophila melanogaster. The first element, M(1), is located between the y and the cv loci, probably close to the y locus. The second element, M(2), is located near the cv locus and the third element, M(3), is located between the y and the car loci. The M(1) element appears to cause a relatively small amount of reduction in the rate of recovery of the SD-72, but not the cn bw, chromosome from SD-72/ cn bw males, when raised at 27.5°. The M(2) and the M(3) elements cause considerable decrease in the recovery rate of the SD-72 chromosome, whereas they increase the recovery rate of the cn bw chromosome. The amount of decrease is nearly the same as the amount of increase for each element. Some type of ``switch' mechanism in the directions of distortion is suggested for each of these two elements and their effects appear to be approximately additive.     

Selection and recombination in Drosophila melanogaster

Summary Artificial selection for wing length in Drosophila melanogaster resulted in changed crossing-over frequencies between three marker genes on the 2nd chromosome, b, cn and vg.The results suggest that artificial selection is a causal agent in producing the observed changes; moreover it is suggested that the modifications in cross-over frequency are controlled by extra-nuclear factors.Research supported by C.N.R. Consiglio Nazionale delle Ricerche Roma, Grant n. 115.2298.4791.     

Selection for malathion-resistance in Drosophila melanogaster

Results from a long-term selection experiment for malathion-resistance in Drosophila melanogaster are described. A polled population of 40 locally-caught, iso-female lines was exposed to increasing concentrations of malathion in the food at both a high selection intensity (MH) and a lower intensity (ML). The response was consistent with a polygenic system. Both adult and larval resistance increased in parallel. Changes in the dose-response curve of adults could be approximately described by a dose-modification factor. Larval resistance was more complex; both selected populations showed a maternal effect which could not be explained by sex-linked genes. Larval resistance in the selected populations behaved as a co-dominant trait with respect to the susceptible controls. Adult resistance was dominant in the ML and co-dominant in the MH population, suggesting that different genes conferring resistance were selected. The selection procedure also produced a developmental delay in both populations, dependent on malathion concentration, but present even in its absence.     

In Situ Hybridisation Analysis of the X-Linked Genes in the Species of the Virilis Group of Drosophila

When estimating the level of DNA sequence variation within and between populations or when planning QTL analysis, it is essential to know the location of the genes under study. In the present work, five X chromosomal genes, earlier localised in Drosophila virilis and D. littoralis, were mapped by in situ hybridisation on the larval polytene chromosomes of four other virilis group species, D. a. americana, D. flavomontana, D. lacicola and D. montana. Conjugation of X chromosomes of the most interesting species pairs was studied in interspecific hybrids. Three of the marker genes were used as RFLP markers to examine the occurrence of recombination in D. flavomontana and D. montana hybrid females. The gene arrangement of all species studied, appeared to be different at the proximal end of the X chromosome, which prevented normal conjugation along the most part of the X chromosome. The data illustrating the locations of five X chromosomal marker genes are presented for D. a. americana, D. flavomontana, D. lacicola and D. montana.     

Selection on codon usage in Drosophila americana

Synonymous codons are not used at random, significantly influencing the base composition of the genome. The selection-mutation-drift model proposes that this bias reflects natural selection in favor of a subset of preferred codons. Previous estimates in Drosophila of the intensity of selective forces involved seem too large to be reconciled with theoretical predictions of the level of codon bias. This probably results from confounding effects of the demographic histories of the species concerned. We have studied three species of the virilis group of Drosophila, which are more likely to satisfy the assumptions of the evolutionary models. We analyzed the patterns of polymorphism and divergence in a sample of 18 genes and applied a new method for estimating the intensity of selection on synonymous mutations based on the frequencies of unpreferred mutations among polymorphic sites. This yielded estimates of selection intensities (N(e)s) of the order of 0.65, which is more compatible with the observed levels of codon bias. Our results support the action of both selection and mutational bias on codon usage bias and suggest that codon usage and genome base composition in the D. americana lineage are in approximate equilibrium. Biased gene conversion may also contribute to the observed patterns.     

Selection on Wing Allometry in Drosophila Melanogaster

Five bivariate distributions of wing dimensions of Drosophila melanogaster were measured, in flies 1) subjected to four defined environmental regimes during development, 2) taken directly from nature in seven U.S. states, 3) selected in ten populations for change in wing form, and 4) sampled from 21 long inbred wild-type lines. Environmental stresses during development altered both wing size and the ratios of wing dimensions, but regardless of treatment all wing dimensions fell near a common allometric baseline in each bivariate distribution. The wings of wild-caught flies from seven widely separated localities, and of their laboratory-reared offspring, also fell along the same baselines. However, when flies were selected divergently for lateral offset from these developmental baselines, response to selection was rapid in every case. The mean divergence in offset between oppositely selected lines was 14.68 SD of the base population offset, after only 15 generations of selection at 20%. Measurements of 21 isofemale lines, founded from wild-caught flies and maintained in small populations for at least 22 years, showed large reductions in phenotypic variance of offsets within lines, but a large increase in the variance among lines. The variance of means of isofemale lines within collection localities was ten times the variance of means among localities of newly established wild lines. These observations show that much additive genetic variance exists for individual dimensions within the wing, such that bivariate developmental patterns can be changed in any direction by selection or by drift. The relative invariance of the allometric baselines of wing morphology in nature is most easily explained as the result of continuous natural selection around a local optimum of functional design.     

Selection for pupation height in Drosophila melanogaster

Divergent directional selection for high and low pupation height was practiced in D. melanogaster. A quick response was observed in the two directions of selection. This is the first time selection for low pupation sites was successful. Realized heritabilities were 18% and 13% for the high and low lines. Reciprocal crosses between divergent lines showed little or no dominance for low pupation sites. The need for a strict control of environmental factors when measuring pupation height is emphasized.     

Size Dependent Sexual Selection in Drosophila ananassae

Mate choice based on body size is widespread and can have numerous consequences. We present data, which show the effect of male and female body size on sexual selection in Drosophila ananassae. The relationships between wing size, locomotor activity, mating latency, courtship pattern, fertility and mating success were studied. Mating latency was negatively correlated with wing length and with locomotor activity, while wing length and locomotor activity was positively correlated in males as well as in females. In female- and male-choice, we found that mate choice influenced size-assortative mating by: (1) large and small males preferring to mate with large females, (2) large males successfully competing for large females, leaving small males to mate with small females. Males increased their reproductive success by mating with large and more fecund females. In addition, in pairs of long/short winged flies, long winged flies courted and mated more successfully than short winged flies and they also have longer duration of copulation and more progeny than short winged flies. We found sterile mating in pairs of small winged males and females.     

Selection components in background replacement lines of Drosophila

Selection components analysis was performed in lines of Drosophila melanogaster at three times during substitution backcrossing. The initial two lines were inbred isofemale lines from natural populations in California, and one had the spread wing mutation eagle. The selection components analysis revealed aspects of the genetic structure of the determinants of fitness by demonstrating changes in the marginal fitnesses of the eagle locus. Differences among backgrounds essentially disappeared by the 20th generation of backcrossing, suggesting that the previously observed differences were attributable to linkage disequilibrium. The method of bootstrapping was used as a novel means of determining statistical confidence in selection components.