Adh Nucleotide Variation in Drosophila willistoni: High Replacement Polymorphism in an Electrophoretically Monomorphic Protein

Drosophila willistoni was the subject of intensive allozyme studies and the locus coding for alcohol dehydrogenase (Adh) was found to be virtually monomorphic. DNA sequence analysis of 18 alleles throughout the distribution of the species has revealed six replacement polymorphisms. The ratio of replacement to silent polymorphisms is higher in D. willistoni than in any other Drosophila species studied for Adh nucleotide variation. Also in contrast to other species, the variation in introns and noncoding DNA is about the same as in the coding region. We speculate that both these differences indicate D. willistoni has historically had a small population size possibly related to Pleistocene refugia in the Neotropics. Received: 5 August 1996 / Accepted: 12 April 1997     

Estimates of Gene Flow in Drosophila Pseudoobscura Determined from Nucleotide Sequence Analysis of the Alcohol Dehydrogenase Region

The genetic structure of Drosophila pseudoobscura populations was inferred from a nucleotide sequence analysis of a 3.4-kb segment of the alcohol dehydrogenase (Adh) region. A total of 99 isochromosomal strains collected from 13 populations in North and South America were used to determine if any population departed from a neutral model and to estimate levels of gene flow between populations. This study also included the nucleotide sequences from two sibling species, D. persimilis and D. miranda. We estimated the neutral mutation parameter, 4N mu, in synonymous and noncoding sites for 17 subregions of Adh in each of nine populations with sample sizes greater than three. The nucleotide diversity data in the nine populations was tested for departures from an equilibrium neutral model with two statistical tests. The Tajima and the Hudson, Kreitman, Aguade tests showed that each population fails to reject a neutral model. Tests for genetic differentiation between populations fail to show any population substructure among the North American populations of D. pseudoobscura. The nucleotide diversity data is consistent with direct and indirect measures of gene flow that show extensive dispersal between populations of D. pseudoobscura.     

Estimates of Linkage Disequilibrium and the Recombination Parameter Determined from Segregating Nucleotide Sites in the Alcohol Dehydrogenase Region of Drosophila Pseudoobscura

The alcohol dehydrogenase (Adh) region of Drosophila pseudoobscura, which includes the two genes Adh and Adh-Dup, was used to examine the pattern and organization of linkage disequilibrium among pairs of segregating nucleotide sites. A collection of 99 strains from the geographic range of D. pseudoobscura were nucleotide-sequenced with polymerase chain reaction-mediated techniques. All pairs of the 359 polymorphic sites in the 3.5-kb Adh region were tested for significant linkage disequilibrium with Fisher's exact test. Of the 74,278 pairwise comparisons of segregating sites, 127 were in significant linkage disequilibrium at the 5% level. The distribution of five linkage disequilibrium estimators D(ij), D(2), r(ij), r(2) and Dij were compared to theoretical distributions. The observed distributions of D(ij), D(2), r(ij) and r(2) were consistent with the theoretical distribution given an infinite sites model. The observed distribution of Dij differed from the theoretical distribution because of an excess of values at -1 and 1. No spatial pattern was observed in the linkage disequilibrium pattern in the Adh region except for two clusters of sites nonrandomly associated in the adult intron and intron 2 of Adh. The magnitude of linkage disequilibrium decreases significantly as nucleotide distance increases, or a distance effect. Adh-Dup had a larger estimate of the recombination parameter, 4Nc, than Adh, where N is the effective population size and c is the recombination rate. A comparison of the mutation and recombination parameters shows that 7-17 recombination events occur for each mutation event. The heterogeneous estimates of the recombination parameter and the inverse relationship between linkage disequilibrium and nucleotide distance are no longer significant when the two clusters of Adh intron sites are excluded from analyses. The most likely explanation for the two clusters of linkage disequilibria is epistatic selection between sites in the cluster to maintain pre-mRNA secondary structure.     

Molecular Population Genetics of Drosophila Immune System Genes

A striking aspect of many vertebrate immune system genes is the exceptionally high level of polymorphism they harbor. A convincing case can be made that this polymorphism is driven by the diversity of pathogens that face selective pressures to evade attack by the host immune system. Different organisms accomplish a defense against diverse pathogens through mechanisms that differ widely in their requirements for specific recognition. It has recently been shown that innate defense mechanisms, which use proteins with broad-spectrum bactericidal properties, are common to both primitive and advanced organisms. In this study we characterize DNA sequence variation in six pathogen defense genes of Drosophila melanogaster and D. mauritiana, including Andropin; cecropin genes CecA1, CecA2, CecB, and CecC; and Diptericin. The necessity for protection against diverse pathogens, which themselves may evolve resistance to insect defenses, motivates a population-level analysis. Estimates of variation levels show that the genes are not exceptionally polymorphic, but Andropin and Diptericin have patterns of variation that differ significantly from neutrality. Patterns of interpopulation and interspecific differentiation also reveal differences among the genes in evolutionary forces.     

Reexamination of Alcohol Dehydrogenase Structural Mutants in Drosophila Using Protein Blotting

Using protein blotting and an immuno-overlay procedure, we have reexamined the cross-reacting material produced by ADH null-activity mutants generated with ethyl methanesulfonate (EMS). Of the 13 mutants, 11 have an immunodetectable polypeptide of wild-type size. The native and urea denatured isoelectric points (pI) establish that 7 of 13 of the mutations have no effect on protein charge. The electrophoretic mobilities of each variant on increasing percent acrylamide gels (Ferguson analysis), reveal that 9 of the 11 immunodetectable mutants have retained the ability to form dimers under native conditions. None of the inactive mutant proteins has the ability to form the "adduct-bound" isozyme. We have found no correlation between protein pI and in vivo stability. The observed frequencies of specific charge class alterations do not dispute the propensity of G:A transitions previously found for EMS mutagenesis.     

Molecular Analysis of an Allozyme Cline: Alcohol Dehydrogenase in Drosophila Melanogaster on the East Coast of North America

Clines may either be selectively maintained or be the by-product of nonadaptive processes related to population structure and history. Drosophila melanogaster populations on the east coast of North America show a latitudinal cline in the frequencies of two common electrophoretically distinguishable alleles at the alcohol dehydrogenase locus (Adh), designated Adh-S and Adh-F. This cline may either be adaptive or an artifact of a possible recent dual founding of North American D. melanogaster populations in which frequencies of Adh alleles differed between founder populations. By means of a high resolution restriction-mapping technique, we studied the distribution of 113 haplotypes derived from 44 polymorphic DNA markers within the Adh region in 1533 individuals from 25 populations throughout the cline. We found significant clinal differentiation at the polymorphisms determining the mobility-difference causing amino acid replacement between Adh-F and Adh-S alleles. Hitchhiking was limited, despite extensive linkage disequilibrium, and other sites did not vary clinally. Such a pattern of differentiation implies that selection is responsible for the cline. To investigate whether selection acts only on the Adh-F/S site, we performed a ``selective equivalence'''' test under the assumption that all variability within the specified allelic class is selectively neutral. This revealed selective equivalence among Adh-S-bearing haplotypes, whose frequencies showed no differentiation throughout the cline, implying high levels of frequency-homogenizing gene flow. Geographical heterogeneity among Adh-F-bearing haplotypes implied the action of selection on one or more additional variants in linkage disequilibrium with Adh-F. In a further study of a subset of the data (n = 1076 from 18 populations), we found a combined insertion/deletion polymorphism, designated &1, located in the 5'' adult intron and in linkage disequilibrium with Adh-F, to show more marked clinal variation than Adh-F/S. Although the unequivocal identification of the precise target(s) of selection requires further study, we suggest that clinal selection may be acting epistatically on the Adh-F/S and &1 polymorphisms.     

Molecular Population Genetics of the Distal Portion of the X Chromosome in Drosophila: Evidence for Genetic Hitchhiking of the Yellow-Achaete Region

We have estimated DNA sequence variation and differentiation within and between Drosophila melanogaster and its sibling species, Drosophila simulans, using six-cutter restriction site variation at yellow-achaete (y-ac), phosphogluconate dehydrogenase (Pgd), and period (per). These three gene regions are of varying distance from the telomere of the X chromosome and range from very low to moderate rates of recombination in D. melanogaster. According to Tajima's test of neutrality, the Pgd region has been influenced by balancing selection in D. melanogaster. This is consistent with previous data suggesting the allozyme polymorphism at this locus is visible to selection. The Hudson, Kreitman, Aguadé test of neutrality reveals a significant departure from neutrality for the y-ac region compared to the per or rosy regions in D. simulans. There is also a significant departure for the y-ac region compared to the Adh 5' flanking region in D. melanogaster. In both species the departure appears to be due to reduced variation at y-ac compared to that expected from divergence between D. simulans and D. melanogaster. We conclude that recent hitchhiking associated with the selective fixation of one or more advantageous mutants in the y-ac region is the best explanation for reduced variation at y-ac.     

The Genetics of a Small Autosomal Region of Drosophila Melanogaster Containing the Structural Gene for Alcohol Dehydrogenase. VII. Characterization of the Region around the Snail and Cactus Loci

The genetic interval 35C to 36A on chromosome arm 2L of Drosophila melanogaster has been saturated for mutations with visible or lethal phenotypes. 38 loci have been characterized, including several maternal-effect lethals (vasa, Bic-C, chiffon, cactus and cornichon) and several early embryonic lethals, including snail and fizzy. About 130 deletions have been used to order these loci. Complex interactions between mutant alleles have been uncovered in the immediate genetic environs of the snail gene, as has further evidence for an interaction between this region and that including the nearby genes no-ocelli and elbow.     

The Effect of an Intronic Polymorphism on Alcohol Dehydrogenase Expression in Drosophila Melanogaster

Several lines of evidence indicate that natural selection controls the frequencies of an allozyme polymorphism at the alcohol dehydrogenase (Adh) locus in Drosophila melanogaster. However, because of associations among sequence polymorphisms in the Adh region, it is not clear whether selection acts directly (or solely) on the allozymic site. This problem has been approached by using in vitro mutagenesis to distinguish among the effects on Adh expression of individual polymorphisms. This study shows that a polymorphism within the first Adh intron ( &1) has a significant effect on the level of ADH protein. Like the allozyme, & shows a geographic cline in frequency, indicating that it may also be a target of natural selection. These results suggest that multisite selection models may be required to understand the evolutionary dynamics of individual loci.     

Post-Translational Control of Alcohol Dehydrogenase Levels in Drosophila melanogaster

A trans-acting regulatory gene that alters in vivo protein levels of alcohol dehydrogenase (ADH) has been mapped to a region of the third chromosome of Drosophila melanogaster. The gene has been found to affect the in vivo stability of ADH protein. It was not found to alter levels of total protein of two other enzymes assayed. The action of the gene over development and its possible mode of control are discussed.     

Chemical Selection of Alcohol Dehydrogenase Negative Mutants in Drosophila

We describe a selection procedure which utilizes the vapor from an unsaturated alcohol, 1-pentene-3-ol, for the detection and isolation of mutant flies with little or no alcohol dehydrogenase activity. ADH-negative flies are unaffected by exposure to the unsaturated alcohol, but ADH positives (wild-types) die after short exposure. The technique can be used to select rare ADH-negative individuals from large populations of wild-type flies.     

Molecular Basis of Polymorphism at the Esterase-5b Locus in Drosophila Pseudoobscura

Sequence variation was studied in a 2.2-kb region encompassing the esterase-5B locus in Drosophila pseudoobscura from two California populations. In these populations, two common electrophoretic classes and many less frequent variants occur, and it was formerly shown by KEITH (1983) that allele frequencies differed from random distribution under an infinite allele model. Nucleotide polymorphisms were determined in 16 sequences representing 14 electrophoretic classes. There was no significant sequence differentiation between populations, and both synonymous and nonsynonymous polymorphisms are distributed homogeneously along the sequence. The data show that the two major electrophoretic classes are heterogeneous at the amino acid level with no diagnostic amino acid(s) distinguishing them. At the nucleotide level, members of one major class are more similar to members of other electrophoretic classes than they are to each other. It appears that random combinations of the neutral amino acid polymorphisms and other undefined physical properties of the proteins generate the different electrophoretic classes and maintain considerable variation at Est-5B.