Background Eating disorders are lethal and heritable; however, the underlying genetic factors are unknown. Binge eating is a highly heritable trait associated with eating disorders that is comorbid with mood and substance use disorders. Therefore, understanding its genetic basis will inform therapeutic development that could improve several comorbid neuropsychiatric conditions. Methods We assessed binge eating in closely related C57BL/6 mouse substrains and in an F2 cross to identify quantitative trait loci (QTL) associated with binge eating. We used gene targeting to validated candidate genetic factors. Finally, we used transcriptome analysis of the striatum via mRNA sequencing (RNA-seq) to identify the premorbid transcriptome and the binge-induced transcriptome to inform molecular mechanisms mediating binge eating susceptibility and establishment. Results C57BL/6NJ but not C57BL/6J mice showed rapid and robust escalation in palatable food consumption. We mapped a single genome-wide significant QTL on chromosome 11 (LOD=7.4) to a missense mutation in cytoplasmic FMR1-interacting protein 2 (Cyfip2). We validated Cyfip2 as a major genetic factor underlying binge eating in heterozygous knockout mice on a C57BL/6N background that showed reduced binge eating toward a wild-type C57BL/6J-like level. Transcriptome analysis of premorbid genetic risk identified the enrichment terms “morphine addiction” and “retrograde endocannabinoid signaling” whereas binge eating resulted in the downregulation of a gene set enriched for decreased myelination, oligodendrocyte differentiation, and expression. Conclusions We identified Cyfip2 as a major significant genetic factor underlying binge eating and provide a behavioral paradigm for future genome-wide association studies in populations with increased genetic complexity.

Psychostimulant addiction is a heritable substance use disorder; however its genetic basis is almost entirely unknown. Quantitative trait locus (QTL) mapping in mice offers a complementary approach to human genome-wide association studies and can facilitate environment control, statistical power, novel gene discovery, and neurobiological mechanisms. We used interval-specific congenic mouse lines carrying various segments of chromosome 11 from the DBA/2J strain on an isogenic C57BL/6J background to positionally clone a 206 kb QTL (50,185,512–50,391,845 bp) that was causally associated with a reduction in the locomotor stimulant response to methamphetamine (2 mg/kg, i.p.; DBA/2J < C57BL/6J)—a non-contingent, drug-induced behavior that is associated with stimulation of the dopaminergic reward circuitry. This chromosomal region contained only two protein coding genes—heterogeneous nuclear ribonucleoprotein, H1 (Hnrnph1) and RUN and FYVE domain-containing 1 (Rufy1). Transcriptome analysis via mRNA sequencing in the striatum implicated a neurobiological mechanism involving a reduction in mesolimbic innervation and striatal neurotransmission. For instance, Nr4a2 (nuclear receptor subfamily 4, group A, member 2), a transcription factor crucial for midbrain dopaminergic neuron development, exhibited a 2.1-fold decrease in expression (DBA/2J < C57BL/6J; p 4.2 x 10−15). Transcription activator-like effector nucleases (TALENs)-mediated introduction of frameshift deletions in the first coding exon of Hnrnph1, but not Rufy1, recapitulated the reduced methamphetamine behavioral response, thus identifying Hnrnph1 as a quantitative trait gene for methamphetamine sensitivity. These results define a novel contribution of Hnrnph1 to neurobehavioral dysfunction associated with dopaminergic neurotransmission. These findings could have implications for understanding the genetic basis of methamphetamine addiction in humans and the development of novel therapeutics for prevention and treatment of substance abuse and possibly other psychiatric disorders.

This tutorial is a learning resource that outlines the basic process and provides specific software tools for implementing a complete genome‐wide association analysis. Approaches to post‐analytic visualization and interrogation of potentially novel findings are also presented. Applications are illustrated using the free and open‐source R statistical computing and graphics software environment, Bioconductor software for bioinformatics and the UCSC Genome Browser. Complete genome‐wide association data on 1401 individuals across 861,473 typed single nucleotide polymorphisms from the PennCATH study of coronary artery disease are used for illustration. All data and code, as well as additional instructional resources, are publicly available through the Open Resources in Statistical Genomics project: http://www.stat-gen.org. © 2015 The Authors. Statistics in Medicine Published by John Wiley & Sons Ltd.

Objective To investigate the feasibility of microRNA (miRNA) levels in CSF as biomarkers for prodromal Huntington disease (HD).Methods miRNA levels were measured in CSF from 60 PREDICT-HD study participants using the HTG protocol. Using a CAG-Age Product score, 30 prodromal HD participants were selected based on estimated probability of imminent clinical diagnosis of HD (i.e., low, medium, high; n = 10/ group). For comparison, participants already diagnosed (n = 15) and healthy controls (n = 15) were also selected. ResultsA total of 2,081 miRNAs were detected and 6 were significantly increased in the prodromal HD gene expansion carriers vs controls at false discovery rate q < 0.05 (miR-520f-3p, miR-135b-3p, miR-4317, miR-3928-5p, miR-8082, miR-140-5p). Evaluating the miRNA levels in each of the HD risk categories, all 6 revealed a pattern of increasing abundance from control to low risk, and from low risk to medium risk, which then leveled off from the medium to high risk and HD diagnosed groups.