Genetic variants that modify neuroendocrine gene expression and foraging behavior of C. elegans

Abstract

The molecular mechanisms underlying diversity in animal behavior are not well understood. A major experimental challenge is determining the contribution of genetic variants that affect neuronal gene expression to differences in behavioral traits. In Caenorhabditis elegans, the neuroendocrine transforming growth factor-β ligand, DAF-7, regulates diverse behavioral responses to bacterial food and pathogens. The dynamic neuron-specific expression of daf-7 is modulated by environmental and endogenous bacteria-derived cues. Here, we investigated natural variation in the expression of daf-7 from the ASJ pair of chemosensory neurons. We identified common genetic variants in gap-2, encoding a Ras guanosine triphosphatase (GTPase)-activating protein homologous to mammalian synaptic Ras GTPase-activating protein, which modify daf-7 expression cell nonautonomously and promote exploratory foraging behavior in a partially DAF-7-dependent manner. Our data connect natural variation in neuron-specific gene expression to differences in behavior and suggest that genetic variation in neuroendocrine signaling pathways mediating host-microbe interactions may give rise to diversity in animal behavior.

Fig. 1.. Natural variation in daf-7 gene expression from the ASJ neurons in wild strains of C. elegans.

(A) daf-7p::GFP expression pattern in N2 (left) and MY18 (right) genetic backgrounds. Filled triangles indicate the ASI neurons; dashed triangles indicate the ASJ neurons. Scale bar, 20 μm. (B) Maximum fluorescence values of daf-7p::GFP in the ASJ neurons in the genetic background of indicated wild strains. Each dot represents an individual animal, and error bars indicate SDs. Statistical test result with unpaired t test with N2 is in table S1.

Fig. 2.. Identification of a genomic locus on chromosome X that affects daf-7 expression in the ASJ neurons.

(A) Diagram for constructing RILs with strains derived from N2 and MY18 backgrounds. (B) Maximum fluorescence values daf-7p::GFP in the ASJ neurons of 123 RILs and parental strains derived from N2 and MY18. The color of the bar represents the genotype at the QTL on chromosome X. (C) Linkage mapping results for daf-7p::GFP expression in the ASJ neurons. X-axis tick marks denote every 5 Mb. Significant QTL is denoted by a red triangle at the peak marker, and blue shading shows 95% confidence interval around the peak marker. The 5% genome-wide error rate logarithm of the odds (LOD) threshold is represented as a dashed horizontal line. (D) Maximum fluorescence values daf-7p::GFP in the ASJ neurons of NILs. Each dot represents an individual animal, and error bars indicate SDs. ****P < 0.0001 as determined by an unpaired two-tailed t test compared to N2.

Fig. 3.. Natural variants in gap-2 modulate daf-7 expression in the ASJ neurons.

(A) Genomic organization of multiple splice isoforms of gap-2. Deletion alleles of gap-2 and the S11L change in gap-2(syb6873) and S64T change in gap-2(syb4046) are depicted. (B) Representative images of daf-7p::GFP expression in N2 and gap-2(syb4046). Filled triangles indicate the ASI neurons; dashed triangles indicate the ASJ neurons. Scale bar, 20 μm. (C to H) Maximum fluorescence values daf-7p::GFP in the ASJ neurons of indicated strains. HET is heterozygous of N2 and NIL, ZD2627. Each dot represents an individual animal, and error bars indicate SDs. ****P < 0.0001 and **P < 0.01 as determined by an unpaired two-tailed t test. Each genotype was compared to N2.

Fig. 4.. gap-2 variants affect foraging behavior.

(A to I) Scatter plot of speed and curvature of each indicated strain. Each dot represents average speed and body curvature during 10 s. The orange line was determined on total N2 data and used for analyses of all the strains. Chromosome diagrams indicate gap-2 allele and genetic background (yellow: N2, blue: MY18). (J) Roaming ratio of N2, gap-2(syb4046), gap-2(syb6873), MY18, gap-2(syb4684), daf-7(ok3125), and daf-7(ok3125); gap-2(syb4046). Each dot represents average roaming ratio of one animal. ****P < 0.0001 as determined by an unpaired two-tailed t test. (K) Roaming ratio of daf-7 floxed gap-2(syb4046) with or without trx-1p::Cre transgene expression. Each dot represents average roaming ratio of one animal. *P < 0.05 as determined by an unpaired two-tailed t test.

Fig. 5.. Natural variants in gap-2 act in the ADE pair of neurons to modulate daf-7 ASJ expression and foraging behavior.

(A) Diagram of the gap-2 genomic locus with C-terminal GFP tagging and stop codon insertion engineered to facilitate the determination of expression pattern. (B) GFP tagged GAP-2 G-/J- isoform expression of the gap-2(syb4587 syb5834) strain (left) and Nomarski image (right). Scale bar, 10 μm. Arrows point to sensillar endings of ADE neurons. (C) Colocalization of GAP-2 G-/J- isoform and dat-1p::mCherry reporter in the cell bodies of the ADE neuron pair. Scale bar, 10 μm. (D) Maximum fluorescence values daf-7p::GFP in the ASJ neurons of pan-neuronal GAP-2(64T) expressing transgenic strains. Each dot represents an individual animal, and error bars indicate SDs. ****P < 0.0001 as determined by an unpaired two-tailed t test. Each genotype was compared to wild type (WT). (E) Maximum fluorescence values daf-7p::GFP in the ASJ neurons of ADE neuronal GAP-2(64T) expressing transgenic strains. Each dot represents an individual animal, and error bars indicate SDs. ****P < 0.0001 and **P < 0.01 as determined by an unpaired two-tailed t test. Each genotype was compared to WT. (F) Roaming ratios of pan-neuronal GAP-2(64T) expressing transgenic strains. Each dot represents an individual animal, and error bars indicate SDs. ****P < 0.0001 as determined by an unpaired two-tailed t test. (G) Roaming ratios of ADE neuronal GAP-2(64T) expressing transgenic strain. Each dot represents an individual animal, and error bars indicate SDs. ****P < 0.0001 as determined by an unpaired two-tailed t test. (H to J) Scatter plot of speed and curvature of each indicated strain. Each dot represents average speed and body curvature during 10 s. Orange line has been determined on total N2 data and used for analysis of all the strains.

Fig. 6.. Environmental enrichment of the gap-2 64T allele is observed in natural populations.

(A) gap-2 haplotypes of wild C. elegans strains are represented by colored shapes (32 unique haplotypes across 550 strains) and are plotted by their sampling location. Shapes distinguish the presence or absence of a naturally occurring variant predicted to affect the splicing or protein coding sequence of gap-2 g/j. (B) Substrate enrichment of the S64T across wild strains. A Fisher’s exact test was used to evaluate significance. ***P < 0.001, **P < 0.01.