Systems genetics of sensation seeking

Abstract

Sensation seeking is a multifaceted, heritable trait which predicts the development of substance use and abuse in humans; similar phenomena have been observed in rodents. Genetic correlations among sensation seeking and substance use indicate shared biological mechanisms, but the genes and networks underlying these relationships remain elusive. Here, we used a systems genetics approach in the BXD recombinant inbred mouse panel to identify shared genetic mechanisms underlying substance use and preference for sensory stimuli, an intermediate phenotype of sensation seeking. Using the operant sensation seeking (OSS) paradigm, we quantified preference for sensory stimuli in 120 male and 127 female mice from 62 BXD strains and the C57BL/6J and DBA/2J founder strains. We used relative preference for the active and inactive levers to dissociate preference for sensory stimuli from locomotion and exploration phenotypes. We identified genomic regions on chromosome 4 (155.236-155.742 Mb) and chromosome 13 (72.969-89.423 Mb) associated with distinct behavioral components of OSS. Using publicly available behavioral data and mRNA expression data from brain regions involved in reward processing, we identified (a) genes within these behavioral QTL exhibiting genome-wide significant cis-eQTL and (b) genetic correlations among OSS phenotypes, ethanol phenotypes and mRNA expression. From these analyses, we nominated positional candidates for behavioral QTL associated with distinct OSS phenotypes including Gnb1 and Mef2c. Genetic covariation of Gnb1 expression, preference for sensory stimuli and multiple ethanol phenotypes suggest that heritable variation in Gnb1 expression in reward circuitry partially underlies the widely reported relationship between sensation seeking and substance use.

Keywords: BXD; Gnb1; Mef2c; addiction; alcohol; mouse; operant sensation seeking.

Figure 1

Operant sensation seeking in the BXD recombinant inbred mouse panel. A, B, Mice rapidly learned to press the active lever to receive a combination auditory‐tactile‐visual reward. Males learned to discriminate between the active and inactive levers more rapidly than females. C, Males received significantly more auditory‐tactile‐visual rewards than females during many of the testing sessions. D, Preference for the active lever increased significantly across sessions until it stabilized on sessions 11 to 19 in both males and females. E‐H, Response patterns on the final session indicate modest within‐session variation across blocks but do not indicate within‐session habituation of the reinforcing effectiveness of the auditory‐tactile‐visual stimuli. *P < 0.05; #P < 0.10

Figure 2

Oss1 behavioral QTL associated with preference for an auditory‐tactile‐visual stimulus in an operant sensation seeking assay. A, Genome‐wide significant behavioral QTL on chromosome 4 (Oss1) associated with active lever preference on the final testing session of an operant sensation seeking assay. B, The 2‐LOD confidence interval for Oss1 was 506 kilobases and contained 16 protein coding genes. C, Oss1 accounted for 18% of the variance on active lever preference during the final session. D, E, Across sessions, mice with the B6 allele at Oss1 received significantly more rewards and exhibited significantly greater active lever preference relative to mice with the D2 allele. *P < 0.05; #P < 0.10

Figure 3

Expression of Gnb1 in reward‐related brain regions is associated with preference for an auditory‐tactile‐visual stimulus and multiple ethanol phenotypes. A‐C, Gnb1 is a positional candidate for Oss1. Genome‐wide significant cis‐eQTL associated with expression of Gnb1, which encodes the G protein β1 subunit, were identified in reward‐related brain regions including the ventral tegmental area, nucleus accumbens, hippocampus and prefrontal cortex (prefrontal cortex not shown). The peak marker for these cis‐eQTL is located within Gnb1 and is identical to the peak marker for the Oss1 behavioral QTL (Figure 2). D‐I, Expression of Gnb1 in these brain regions is genetically correlated with preference for an auditory‐tactile‐visual stimulus and multiple ethanol phenotypes. Data points represent BXD strain means. Black and gray colors signify the B6 and D2 alleles, respectively, at the peak of the Oss1 QTL. **P < 0.01; *P < 0.05

Figure 4

Oss2 behavioral QTL associated with number of rewards but not active lever preference in an operant sensation seeking assay. A, Genome‐wide significant behavioral QTL on chromosome 13 (Oss2) associated with number of rewards, active levers presses or inactive lever presses on multiple sessions (1, 2, 3, 4, 5 and 9) of an operant sensation seeking assay when mice were still learning the response‐reward contingencies. Data from rewards on session 5 is shown. B, The 2‐LOD confidence interval for Oss2 spanned 16.5 Mb of chromosome 13 and contained 58 protein coding genes. C, Oss2 accounted for 25% of the variance on number of rewards. D, E, As seen with Oss1, mice with the B6 allele at Oss2 received significantly more rewards relative to mice with the D2 allele. In contrast to Oss1, the allele at Oss2 did not significantly affect active lever preference. *P < 0.05

Figure 5

Expression of Mef2c is associated with number of rewards in an operant sensation seeking assay. Mef2c is a positional candidate for Oss2. A, Genome‐wide significant cis‐eQTL associated with expression of Mef2c, which encodes myocyte enhancer factor 2C, was identified in the hippocampus. B, Expression of Mef2c mRNA in the hippocampus was positively genetically correlated with number of rewards on an operant sensation seeking assay. Data points represent BXD strain means. Black and gray colors signify the B6 and D2 alleles, respectively, at the peak of the Oss2 QTL. *P < 0.05