Genomic consequences of background effects on scalloped mutant expressivity in the wing of Drosophila melanogaster

Abstract

Genetic background effects contribute to the phenotypic consequences of mutations and are pervasive across all domains of life that have been examined, yet little is known about how they modify genetic systems. In part this is due to the lack of tractable model systems that have been explicitly developed to study the genetic and evolutionary consequences of background effects. In this study we demonstrate that phenotypic expressivity of the scalloped(E3) (sd(E3)) mutation of Drosophila melanogaster is background dependent and is the result of at least one major modifier segregating between two standard lab wild-type strains. We provide evidence that at least one of the modifiers is linked to the vestigial region and demonstrate that the background effects modify the spatial distribution of known sd target genes in a genotype-dependent manner. In addition, microarrays were used to examine the consequences of genetic background effects on the global transcriptome. Expression differences between wild-type strains were found to be as large as or larger than the effects of mutations with substantial phenotypic effects, and expression differences between wild type and mutant varied significantly between genetic backgrounds. Significantly, we demonstrate that the epistatic interaction between sd(E3) and an optomotor blind mutation is background dependent. The results are discussed within the context of developing a complex but more realistic view of the consequences of genetic background effects with respect to mutational analysis and studies of epistasis and cryptic genetic variation segregating in natural populations.

Figure 1.—

Phenotypic consequences of genetic background on the expressivity of the sd^E3 allele in the wing. In the Oregon-R background (A) this is similar to severe hypomorphic alleles of sd, whereas in the Samarkand genetic background (B) the effect resembles a weak to moderate allele. (C) The distribution of the wing sizes for the sd^E3 allele in each wild-type genetic background is completely nonoverlapping. The distribution of the F₂ population shows clear evidence of bimodality, consistent with at least one modifier of large effect segregating between the two wild-type backgrounds.

Figure 2.—

The expression of the sd transcript is reduced in sd^E3 mutants, independent of genetic background. Relative abundance of sd transcript is reduced in sd^E3 individuals as measured using microarrays. There is no statistical support for an interaction between genetic background and the mutation with respect to the abundance of sd transcript, and results from quantitative RT–PCR also failed to support an interaction (not shown). Error bars are ±1 SE.

Figure 3.—

Background-specific spatial and quantitative patterns of gene expression in Sd-regulated genes. The left column shows the reaction norm plot of relative transcript abundance as monitored using the DGRC arrays for each of four genes (vg, Dll, omb/bi, and wg). Consistent with the patterns of phenotypic expressivity between genetic backgrounds, we observed a significant decrease in gene expression, as monitored with in situ hybridization of wing imaginal discs, as well as from the microarrays. Row 1 shows that expression of vg transcript is reduced in a genetic background-specific manner. While there is no difference between the two wild-type genetic backgrounds, they do differ significantly under sd^E3 (±1 SE). Consistent with this, the spatial domain of vg transcript is reduced in the ORE sd^E3 background relative to SAM sd^E3. Both ORE sd^E3 and SAM sd^E3 show spatial restriction relative to wild-type expression patterns for vg. Similar patterns of expression were observed for a number of genes including Dll (row 2) and omb/bi (row 3). Interestingly, two candidate genes (wg, row 4; and salm, not shown) show clear expression differences between genetic backgrounds with sd^E3, but with little evidence for expression differences in the array data.

Figure 4.—

Background-dependent genetic interaction between mutations in omb and sd. Double-mutant combinations between the omb^md and sd^E3 alleles were made in both the Samarkand and the Oregon-R genetic backgrounds. (A) In the Samarkand background the double-mutant combination (observed in hemizygous males) shows a strong enhancement of the sd^E3 phenotype (compare with Figure 1B). However, in the Oregon-R background (B) the double-mutant combination is indistinguishable from the qualitative and quantitative range of ORE sd^E3 single mutants (compare with Figure 1A).

Figure 5.—

Consequences of the sd^E3 mutation on the wing imaginal disc transcriptome are background dependent. The majority of differentially expressed genes between sd^E3 and wild type are the result of a reduction of expression in the sd^E3 mutant condition (A–C), as demonstrated by the strong asymmetries. These effects appear to be background independent (B vs. C). In contrast (D–F), fewer genes show evidence for differences in expression between backgrounds; however, the magnitude of the differences tends to be larger than in the comparison of wild type to sd^E3.