Sensitivity of Allelic Divergence to Genomic Position: Lessons from the Drosophila tan Gene

Abstract

To identify genetic variants underlying changes in phenotypes within and between species, researchers often utilize transgenic animals to compare the function of alleles in different genetic backgrounds. In Drosophila, targeted integration mediated by the ΦC31 integrase allows activity of alternative alleles to be compared at the same genomic location. By using the same insertion site for each transgene, position effects are generally assumed to be controlled for because both alleles are surrounded by the same genomic context. Here, we test this assumption by comparing the activity of tan alleles from two Drosophila species, D. americana and D. novamexicana, at five different genomic locations in D. melanogaster We found that the relative effects of these alleles varied among insertion sites, with no difference in activity observed between them at two sites. One of these sites simply silenced both transgenes, but the other allowed expression of both alleles that was sufficient to rescue a mutant phenotype yet failed to reveal the functional differences between the two alleles. These results suggest that more than one insertion site should be used when comparing the activity of transgenes because failing to do so could cause functional differences between alleles to go undetected.

Keywords: allele-specific; pigmentation; position effect; tan; transgenic analysis.

Figure 1

Body color of D. americana and D. novamexicana. D. novamexicana (right) has evolved lighter body pigmentation since it diverged from the common ancestor shared with D. americana (left). D. americana has retained the darker body pigmentation shared by all other members of the virilis group.

Figure 2

Measurements of pigmentation intensity in a control sample varied slightly during image collection. Raw median pigmentation intensity in tergite A4 (insert) is plotted against imaging order for the reference cuticle (open circles). All images were taken during in a single sitting without adjustment of lighting, focus, or other imaging parameters; the small (β = −0.0075), yet significant (p-value = 0.008), downward trend in pigmentation intensity as the imaging progressed, presumably as a result of changes in ambient lighting or other uncontrolled imaging variables. An imaging order correction was therefore applied to all measurements, as described in the Materials and Methods section. Corrected median pigmentation intensity values for the same images are also plotted against imaging order (closed circles) to show the effects of this correction.

Figure 3

Genomic location can impact the relative difference in pigmentation caused by D. americana and D. novamexicana tan alleles. Box plots of pigmentation intensity summarize, for each genotype, the range of pigmentation phenotypes observed. The median (center line), first quartile (bottom of box), third quartile (top of box), and ± 1.5 × the interquartile range (whiskers) are shown for each genotype examined. Yellow boxes along the x-axis represent the D. novamexicana allele and brown boxes represent the D. americana allele. Significant increases in pigmentation from the control were detected for all genomic locations except 102D (Table 2). Three of the other four landing sites (58A, 86Fa, and 86Fb) showed significant differences in pigmentation driven by the D. americana and D. novamexicana tan alleles, whereas the fourth landing site (51C) did not show a detectable difference in pigmentation between flies carrying the two species’ alleles (Table 2). Representative images from the 25th percentile (first quartile), median, and 75th percentile (third quartile) are shown below the box plot for each genotype. The most striking differences between alleles are seen in the anterior regions outside the dorsal midline stripe.

Figure 4

Genomic location impacts relative expression of the D. americana tan transgene in D. melanogaster. Expression of the D. americana tan transgene inserted at each of the five landing sites tested is shown relative to its expression when inserted in the 86Fa landing site. Circles indicate mean expression among replicate samples and the error bars show the 95% C.I. of the estimates. Note that the relative expression level of D. americana tan among landing sites correlates with the ability to detect differences in abdominal pigmentation (Figure 3). The D. americana tan transgene inserted at 58A, 86Fa, and 86Fb all showed similar expression as well as similar pigmentation phenotypes. The D. americana tan transgene inserted at 51C had a level of expression between these lines and the line with the transgene inserted at 102D, as well as pigmentation that was intermediate between these lines and 102D. The D. americana tan transgene inserted at 102D had the lowest transgene expression and failed to increase dark pigmentation relative to the tan mutant phenotype.