Laboratory evolution of adenylyl cyclase independent learning in Drosophila and missing heritability

Abstract

Gene interactions are acknowledged to be a likely source of missing heritability in large-scale genetic studies of complex neurological phenotypes. However, involvement of rare variants, de novo mutations, genetic lesions that are not easily detected with commonly used methods and epigenetic factors also are possible explanations. We used a laboratory evolution study to investigate the modulatory effects of background genetic variation on the phenotypic effect size of a null mutation with known impact on olfactory learning. To accomplish this, we first established a population that contained variation at just 23 loci and used selection to evolve suppression of the learning defect seen with null mutations in the rutabaga adenylyl cyclase. We thus biased the system to favor relatively simplified outcomes by choosing a Mendelian trait and by restricting the genetic variation segregating in the population. This experimental design also assures that the causal effects are among the known 23 segregating loci. We observe a robust response to selection that requires the presence of the 23 variants. Analyses of the underlying genotypes showed that interactions between more than two loci are likely to be involved in explaining the selection response, with implications for the missing heritability problem.

Keywords: Adenylyl cyclase; Drosophila; cAMP; cryptic variation; epistasis; laboratory evolution; learning; memory; missing heritability; selective breeding.

Figure 1. Multi-generational selection of improved learning in rut¹ mutants

6 populations (A,B) are rut¹homozygous mutant - and heterogeneous at 23 loci (see text). Three of these (MOR, MUL, LEW) underwent selection, three (BRI, DOB and STU) were allowed to drift. Two control populations are rut- and do not contain any of the 23 transposon insertions (C). All populations are in the same inbred bacground (methods). (A) Learning Performance Index (Gravner et al.) of individual selected populations [Morgan (dark blue), Lewis (teal), Muller (light blue)] and individual unselected populations [Bridges (light red), Dobzhansky (dark red), and Sturtevant (orange)]. N=4 PI measurements per population at each time-point. (B) Mean Performance Index of 3 selected populations Morgan, Muller, Lewis (blue) and 3 unselected populations Dobzhansky, Sturtevant, Bridges (red). (C) Performance indices of selected (blue) and unselected (Red) controls on an inbred background. (N = 4 PI measurements per group).

Figure 2. Genotype data and Multivariate analyses identify 8 loci underlying selection response

(A) Heat map of genotypes are shown for flies sampled from all populations at generations 11 and 25. Individual samples are arranged by population on the Y-axis with selected groups (Mor, Mul and Lew) and unselected (Dob, Stu and Bri). Each of the 23 loci are shown on the Y axis. Black ticks denote homozygosity for the wild type allele. Yellow ticks denote heterozygosity. Red denotes homozygosity for the mutant allele. Missing data values were ‘filled in’ by imputation (see methods). Shown is the result of a single imputation. (B) Difference between mean dosages in selected and unselected groups error bars are 95% confidence intervals obtained from a combination jackknife/bootstrap procedure to account for group substructure.(C) SVD analysis (see methods) of individual fly genotypes found in (A). Selected flies in (blue) and unselected in (red) are plotted by principle component 1 (PC1) on the X-axis, and principle component 2 (PC2) on the Y- axis. Again, data are from a single imputation (corresponding to the matrix in A) (D) Plot of the discrimination vector with the alleles (X-axis) sorted by contribution (Y- axis). Mean values of the 100 imputations.

Figure 3. Effects on rut¹ memory performance single, double and triple heterozygous combinations among alleles identified in by SVD/LDA

Memory performance of animals that are hemizygous for the rut¹ mutation and heterozygous for E3272 or D0077 (A) or E3945 (B) relative to performance of rut¹ hemizygous or rut+ males. E3272, but not D0077 or E3945, yields a partial but significant (N = 17, Tukey HSD) suppression of the rut¹ memory defect. Effects on the rut¹ memory levels of each of the 28 heterozygous di-allele combinations amongst top 8 alleles also were tested for memory performance (C). In 19 di-allele combinations, no significant impact on the memory performance of rut¹ was observed (yellow boxes in top half matrix). 9 combinations significantly enhanced the learning defect of rut¹ (blue), i.e. they exhibited lower levels of learning [Tukey HSD]. These effect sizes are shown in the bottom half matrix as a fraction of rut¹ performance levels. Animals that were rut¹ hemizygous and D0077/+, E3272/+, E3945/+ triple heterozygous exhibit memory performance that is not significantly different from that of rut¹ (D).