Alleles of Chaser, a dominant modifier of the Drosophila melanogaster foraging gene, are consistent with variegating alleles of the heterochromatic gene spookier

Abstract

The relationship between genes and quantitative behavioral traits involves complex regulatory networks. Identifying genes that operate in these regulatory pathways can be challenging, especially when dealing with dominant genetic factors. Our work has focused on a naturally occurring behavioral polymorphism in larval foraging behavior in Drosophila melanogaster. This polymorphism in larval foraging behavior arises from variation in the foraging (for) gene with its rover and sitter naturally occurring variants. The dominant rover allele (forR) results in larvae which move longer distances while foraging compared with larvae with the recessive sitter (fors) alleles. In this article, we report the successful mapping of the Chaser (Csr) gene, a dominant modifier of larval foraging behavior which makes sitter larvae behave in a rover-like manner. We localized Csr by first mapping recessive phenotype tags closely linked to Csr. These phenotype tags mapped to the centromeric heterochromatin on the right arm of chromosome 3. We showed through a combination of deletion mapping, qRT-PCR and feeding of ecdysone hormone to larvae during development that the alleles of Csr are consistent with variegating alleles of the gene spookier (spok). With spok being an essential gene in the synthesis of the molting hormone ecdysone, we have established a link between ecdysone signaling and larval foraging behavior in D. melanogaster.

Keywords: Chaser gene; foraging gene; cGMP-dependent protein kinase; ecdysone; heterochromatin; larval foraging behavior.

Fig. 1.

A new recessive phenotype tag in the Csr^RV lines produces abdominal banding abnormalities. Examples of variability in D. melanogaster abdominal banding abnormalities in female progeny of crosses involving the Csr^RV2 line (a and b) and wild-type c) are shown. a): minor abnormality with one side of bands A2 and A3 coming together at the dorsal midline with only a single band on the side opposite. b): a more severe abnormality with one half of band A4 missing and the other half of A4 curving upwards to meet band A3 at the dorsal midline. Band A5 is thin and band A6 is bulbous-looking. c): wild-type banding in a for^s female. d): representative drawings of additional banding patterns observed with crosses involving the Csr^RV2 line during the initial investigation of the phenotype (with wild-type for comparison on the left hand side).

Fig. 2.

D. melanogaster salivary gland polytene squash from a Csr²/ry  ^{+  5} heterozygous larvae reveals the presence of a heterochromatic break in the Csr² line. A heterochromatic insertion at band 36B, shown by the arrow, is visible on the left arm of chromosome 2 (2L) in this polytene squash from a Csr²/ry  ^{+  5} heterozygous larvae. It is a typical observation with heterochromatic insertions into euchromatin to see the ends of paired chromosomes unzip distal to the heterochromatic insertion during slide preparation, as seen here. The thin line extending from 36B on the Csr² chromosome arm is heterochromatin and leads back to the chromocenter at the bottom right of the slide. Although it is difficult to distinguish thin heterochromatic strands from ectopic pairing in individual nuclei, the features described here, including the thin line extending from 36B, were consistent in multiple nuclei from independent squashes.

Fig. 3.

Deletion analysis in 3Rh narrowed the candidate region for Csr to 3 genes: spok, PARP1 and alg-2. Known or predicted coding regions in 3Rh are shown as vertical lines. The distance between genes is not to scale and genes known to be essential are highlighted. Deletions are represented by horizontal lines and the centromere by a circle. Both the pupal lethal tag and the abdominal banding abnormalities were localized to 3Rh using Df(3Rh)8740#18, with their location narrowed down to the region spanning from spok to alg-2 based on a failure to complement Df(3Rh)e-1676. Only the pupal lethal tag was uncovered by Df(3R)PARP1-1, localizing the pupal lethal tag to the essential gene PARP1. Abdominal banding abnormalities were uncovered by Df(3Rh)e-204, mapping the abdominal banding phenotype to spok.

Fig. 4.

Sequence alignment of exon 1 in D. melanogaster's spok gene between for^s and Csr lines, showing a single nucleotide polymorphism was present in Csr³ and the Csr^RV lines derived from Csr³. Sequence comparison in spok exon 1 between for^s and multiple Csr lines (Csr², Csr³, Csr^RV1, Csr^RV2 and Csr^RV5) is shown. The Csr² line contained no nucleotide differences and matched the for^s genetic background. Csr³ and the Csr^RV lines (derived from Csr³) all share a single A to G) nucleotide polymorphism at location +241 in the first exon of spok.

Fig. 5.

A comparison of expression levels of 3 D. melanogaster Csr candidate genes, using qRT-PCR, uncovered a significant difference in spok expression between lines with sitter behavior (for^s) and rover behavior (for^R and Csr²). qRT-PCR was conducted on for^R, for^s and Csr² third instar foraging larvae (96 h old ±2 h). Expression levels are shown as the concentration of mRNA relative to the Actin reference sample. Each qRT-PCR used 10 larvae and was replicated 5 times per line. Significant differences in expression levels between lines were only found with the spok gene. for^R and Csr² had similar expression levels in spok (0.0022 ± 0.0003 and 0.0030 ± 0.0012), but both had expression levels in spok significantly higher than for^s (0.0015 ± 0.0001).

Fig. 6.

Foraging path lengths of sitter larvae (for^s) are rover-like when raised in food laced with 10 μM 20-hydroxyecdysone (20E). Path lengths (cm) of D. melanogaster for^R and for^s control lines differed significantly (5.59 ± 0.27 cm and 4.36 cm ± 0.36 cm, respectively; n = 40 per line). Path lengths for for^R and for^s larvae raised on 10 μM 20E laced food (5.67 ± 0.32 cm and 5.77 ± 0.33 cm, respectively; n = 40 per line) were significantly longer than the for^s control, but not significantly different than each other or the for^R control. A SNK a posteriori analysis (P < 0.05) revealed 2 groups: group A consisted of the for^R rover control, for^s raised on 10 μM 20E and for^R raised on 10 μM 20E while group B consited of the for^s sitter control.

Fig. 7.

Visual representation of D. melanogaster chromosomes showing possible rearrangements causing spok to variegate. C is the centromere, s represents spok and L represents the pupal lethal tag. Numbers above the chromosome represent cytological bands. Squiggly arrows show the location of X-ray induced chromosome breaks determined by polytene analysis. a) Creation of the Csr³ chromosome from a simple pericentric inversion in for^s. b) Creation of the Csr^RV1 and Csr^RV5 chromosomes from Csr³. c) Creation of the Csr² chromosome by a direct insertion into 2L euchromatin.