Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Nov 15.
Published before final editing as: Genes Brain Behav. 2021 May 12:e12751. doi: 10.1111/gbb.12751

Systems genetic analysis of binge-like eating in a C57BL/6J x DBA/2J-F2 cross

Emily J Yao 1, Richard K Babbs 1, Julia C Kelliher 1, Kimberly P Luttik 1, Kristyn N Borrelli 1,2,3,4, M Imad Damaj 5, Megan K Mulligan 6, Camron D Bryant 1,2,3,4,*
PMCID: PMC9361732  NIHMSID: NIHMS1820672  PMID: 33978997

Abstract

Binge eating is a heritable trait associated with eating disorders and refers to the rapid consumption of a large quantity of energy-dense food that is associated with loss of control and negative affect. Binge Eating Disorder is the most common eating disorder in the US; however, the genetic basis is unknown. We previously identified robust mouse inbred strain differences between C57BL/6J and DBA/2J in binge-like eating of sweetened palatable food in an intermittent access, conditioned place preference paradigm (PF-CPP). To map the genetic basis of changes in body weight and binge-like eating and to identify candidate genes, we conducted quantitative trait locus (QTL) analysis in 128 C57BL/6J x DBA/2J-F2 mice combined with PheQTL and trait covariance analysis in GeneNetwork2 using legacy BXD-RI trait datasets. We identified a QTL on chromosome 18 influencing changes in body weight across days in females (LOD=6.3; 1.5-LOD: 3-12 cM) that contains the candidate gene Zeb1. We also identified a sex-combined QTL influencing initial palatable food intake on chromosome 5 (LOD = 5.8; 1.5-LOD: 21-28 cM) that contains the candidate gene Lcorl and a second QTL influencing escalated palatable food intake on chromosome 6 in males (LOD = 5.4; 1.5-LOD: 50-59 cM) that contains the candidate genes Adipor2 and Plxnd1. Finally, we identified a suggestive QTL in females for slope of binge-like eating on distal chromosome 18 (LOD=4.1; p=0.055; 1.5-LOD: 23-35 cM). Future studies will use BXD-RI strains to fine map loci and support candidate gene nomination for gene editing.

Keywords: GWAS, PheWAS, eQTL, BXD-RI, bulimia nervosa, PGC-ED, anthropometric, psychiatric genetics, SABV, sex differences

INTRODUCTION

Binge eating is a heritable complex trait present within the spectrum of eating disorders, including binge eating disorder, bulimia nervosa, and anorexia nervosa. Binge eating is defined by repeated bouts of ingesting large quantities of food intake over a short time period (typically less than two hours) that is associated with a loss of control, anxiety, guilt, remorse, and depression (Wolfe et al. 2009). While binge eating quantity, duration, and frequency are deemed important characteristics (Johnson et al. 2000), the severity of loss of control [inability to eat the amount of food that was intended or planning a binge in violation of normal dietary standards (Johnson et al. 2000)] can best predict clinical impairment and psychiatric dysfunction (Vannucci et al. 2013).

Binge eating is associated with behavioral, malnutritional, metabolic and psychiatric dysfunction, including aberrant and compensatory restrictive eating, obesity and associated health risks (da Luz et al. 2018), negative valence (Vannucci et al. 2015), mood disorders (Guerdjikova et al. 2019), and substance use disorders (Munn-Chernoff & Baker 2016). Genetic and environmental factors contribute to susceptibility to binge eating (Bulik et al. 2003). Genome-wide association studies of eating disorders have identified significant risk loci for anorexia nervosa, including loci near genes related to metabolic and psychiatric dysfunction (Watson et al. 2019). However, GWAS of binge eating, Binge Eating Disorder, and Bulimia Nervosa are lacking (Hübel et al. 2018) .

We developed a binge-like eating procedure in mice to measure escalation in the consumption of sweetened palatable food over time in an intermittent, limited access conditioned place preference (CPP) paradigm and subsequent compulsive-like intake in a light/dark conflict procedure (Kirkpatrick et al. 2017). We identified multiple novel genetic factors contributing to binge-like eating. First, using a reduced complexity cross between closely related substrains of C57BL/6 mice (Bryant et al. 2018, 2020a), we mapped a major-effect quantitative trait locus (QTL) near a proposed gain-of-function missense mutation in Cyfip2 (Kumar et al. 2013) that influenced binge-like eating (Kirkpatrick et al. 2017). Cyfip2 +/− mice showed reduced binge-like eating on the binge-prone C57BL/6NJ background (Kirkpatrick et al. 2017). A subsequent study showed that haploinsufficiency of the closely related gene Cyfip1 also modulated binge-like eating but in a complex manner that depended on C57BL/6 genetic background, sex, and parent-of-origin (Babbs et al. 2019). Finally, knockout mice for casein kinase 1-epsilon (Csnk1e), a gene whose deletion enhances behavioral sensitivity to the stimulant, rewarding, and reinforcing responses to opioids and psychostimulants (Bryant et al. 2012b; Wager et al. 2014) and is associated with opioid dependence in humans (Levran et al. 2008, 2015), showed a robust, female-specific induction of binge-like eating on the binge-resistant C57BL/6J background (Goldberg et al. 2017). These studies illustrate the utility of our binge-like eating paradigm in identifying genetic factors exerting pleiotropic influence on addiction traits and binge-like eating that could have clinical relevance in humans.

To expand our efforts in gene discovery of binge-like eating, we identified a robust genetic difference in binge-like eating between the binge-resistant C57BL/6J inbred strain which showed very little binge-like eating of sweetened palatable food in our intermittent, limited access CPP (Goldberg et al. 2017; Kirkpatrick et al. 2017) versus the DBA/2J inbred strain which showed robust binge-like eating (Babbs et al. 2018). In that study, we generated a small cohort of C57BL/6J x DBA/2J-F2 mice and tested candidate loci based on the prior QTL literature regarding sweet taste (chromosome 4) and bitter taste (chromosome 6) between C57BL/6J and DBA/2J strains (Blizard et al. 1999). We found a significant association between binge-like eating in males and a polymorphic marker within the Tas2r locus on chromosome 6 (133 Mb) containing bitter taste receptors (Babbs et al. 2018) that was previously associated with variation in quinine (bitter) taste between C57BL/6J and DBA/2J (Blizard et al. 1999; Nelson et al. 2005). However, there are several remaining questions from these findings. First, is the association between Tas2r and binge-like eating significant at the genome-wide level? Second, can the peak for the QTL on chromosome 6 be localized to the Tas2R gene cluster (~132.5-133.5 Mb, mm10) that contains bitter taste receptors? Third, is there functional evidence for candidate genes underneath this QTL that could modulate binge-like eating? And fourth, can we identify additional genome-wide significant loci linked to binge-like eating as well as concomitant changes in body weight?

To answer these questions, we genotyped the same cohort of C57BL/6J x DBA/2J-F2 mice (Babbs et al. 2018) genome-wide to conduct QTL analysis of binge-like eating and determine if we could replicate candidate loci at the genome-wide level and identify new QTLs underlying binge-like eating. Our results confirmed a major genome-wide significant, male-sensitive QTL on chromosome 6. We also identified a QTL on proximal chromosome 18 influencing percent body weight gain in females. Furthermore, we identified an additional QTL on chromosome 5 affecting initial palatable food intake in both sexes and a nearly significant female-sensitive QTL for the slope of escalation in binge-like eating across training days on chromosome 18. We employed GeneNetwork (Chesler et al. 2004; Mulligan et al. 2017) to identify candidate genes exhibiting functional evidence for polymorphisms between C57BL/6J and DBA/2J that influence gene expression and to further hone this list of candidates by examining their correlation with other GeneNetwork2 traits exhibiting peak QTLs within our loci. GeneNetwork2 (www.genenetwork.org/) is an online data repository and tool for analyzing thousands of historical gene expression, physiological, and behavioral traits in the BXD recombinant inbred panel that segregates C57BL/6J and DBA/2J alleles (Chesler et al. 2004; Mulligan et al. 2017).

METHODS

Mice

All experiments were conducted in accordance with the NIH Guidelines for the Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee at Boston University (AN-15403). Seven-week old, C57BL/6J x DBA/2J-F1 mice (15 breeder pairs) were purchased from Jackson Laboratory (JAX; Bar Harbor, ME) and were habituated in the colony for one week prior to breeding in-house to generate 128 C57BL/6J x DBA/2J -F2 mice for experimental testing. Estrus cycle of females was not monitored during the course of the study. Mice were run in three cohorts over 4 months. Cohort 1 contained 22 females and 25 males; cohort 2 contained 19 females and 25 males; cohort 3 contained 25 females and 15 males.

Home cage diet and experimental palatable food pellets

Chow (Teklad 18% Protein Diet, Envigo, Indianapolis, IN, USA) and tap water were provided in the home cage ad libitum throughout the entire study. In training for binge-like eating, sweetened palatable food (PF) pellets (TestDiet; 20 mg each; 5TUL diet; MO, USA) contained a metabolizable energy density of 3.4 kcal/g (21% from protein, 13% from fat, 67% from carbohydrates) were provided in an intermittent, limited access model of binge-like eating as described below.

Binge-like eating (BLE) and compulsive-like eating

F2 mice were previously trained in an intermittent, limited access binge-like eating procedure in a PF conditioned place preference (PF-CPP) paradigm over 22 days (Babbs et al. 2018) as originally described (Kirkpatrick et al. 2017). On Day (D) 1, mice were assessed for initial preference for the palatable food-paired side (right side) over 30 min. On D2, D4, D9, D11, D16, and D18, mice were confined to the right side of the apparatus for 30 min and allowed access to palatable food. Food pellets were weighed both before and after each 30 min session. On D3, D5, D10, D12, D17, and D19, a clean, empty food bowl was provided in the corner on the left side and mice were confined there for 30 min. on D22, mice were provided with open access to both sides (clean, empty food bowls on both sides) and were re-assessed for time spent on the PF-paired side. PF-CPP was assessed by calculating the change in time (s) spent on the PF-paired side between D1 and D22 (D22-D1).

On D23, mice were tested in a light/dark conflict test for compulsive-like eating (Kirkpatrick et al. 2017) in which palatable food was available in the light side of the box which is considered anxiety-provoking and aversive. Palatable food intake was calculated as % body weight consumed [g consumed / body weight (g) * 100]. All behaviors were recorded with infrared cameras (Swann Communications U.S.A. Inc., Santa Fe Springs, CA, USA) and video-tracked with ANY-maze video tracking software (Stoelting Co., Wood Dale, IL, USA).

Genotyping in C57BL/6J × DBA/2J-F2 mice

DNA was extracted from tail snips using a salting out procedure. DNA was shipped for genome-wide genotyping on the MiniMUGA array (Neogen GeneSeek Operations, Lincoln, NE, USA). There are 3314 polymorphic SNP markers between C57BL/6J and DBA/2J on this array that can identify parental inheritance of recombinant, chromosomal regions. Marker positions were converted from bp to sex-averaged cM prior to mapping, using the JAX Mouse Map Converter (http://cgd.jax.org/mousemapconverter).

Data analysis

Prior to QTL analysis, F2 mice were analyzed irrespective of genotype in SPSS (version 27.0) using mixed model RM ANOVAs with Sex as a factor and Day as a categorical repeated measure followed by Tukey’s post-hoc comparisons of individual days in the case of an interaction. Mauchly’s sphericity test was used to determine whether the data in the mixed model ANOVAs violated the assumption of sphericity which assumes equal variances between all combinations of group differences. If this assumption was violated (as indicated by the reported epsilon value), a Greenhouse-Geisser correction estimate was generated and multiplied by the degrees of freedom to adjust the p-value of the F ratio statistic. In the case of a Sex x Day interaction, post-hoc comparisons were conducted using Tukey’s honestly significant difference. Slopes for body weight gain and increased food intake during BLE training were calculated for each individual sample using all six datapoints (D2, D4, D9, D11, D16, and D18) with the “SLOPE” function in Excel. The function returns the slope of the linear regression line for known x and y values and indicates the rate of change in phenotype (body weight gain or PF intake) across training days.

Quality checking of genotypes and QTL analysis were performed in R (https://www.r-project.org/) using R/bestNormalize (https://github.com/petersonR/bestNormalize) and R/qtl (Broman et al. 2003). Phenotypes were assessed for normality using the Shapiro-Wilk Test. Because the data residuals sometimes deviated significantly from normality, we used the orderNorm function to perform Ordered Quantile normalization (Peterson & Cavanaugh 2019) on all phenotypes. Effect plots for the raw data for QTLs generated from the quantile-normalized data are provided in Supplementary Information. We also report all significant QTLs with the raw, un-normalized data in the Supplementary Information. For sex-stratified analyses, datasets were quantile-normalized separately for females and males. Genotype calls were quality checked to identify possible errors and to ensure that the markers were reliable. We removed markers with a call rate of less than 95%. We used the countXO function, to examine the number of crossovers per individual and removed three outlier subjects showing more than 1000 crossovers each. Finally, we identified double-crossover genotyping errors by running the calc.errorlod function in R/qtl with an assumed genotyping error rate of 0.05. Markers with log of the odds (LOD) scores greater than 5 were removed. After QC, 2994 markers and phenotypes from 128 F2 samples were used in QTL analysis.

QTL analysis was performed using the “scanone” function and Haley-Knott (HK) regression. “Cohort” was included as an additive covariate and “Sex” was included as an interactive covariate in the QTL model. For separate female and male analyses, “Cohort” was included as an additive covariate. Permutation analysis (perm = 1000) was used to compute genome-wide suggestive (p < 0.63) and significance (p < 0.05) thresholds. For each significant QTL, we calculated both the Bayes credible interval and 1.5 LOD drop intervals from the peak-associated marker. QTL intervals are reported in both cM and Mb. The physical interval (Mb) was conservatively derived by expanding both proximally and distally from the cM boundaries to the nearest markers. Percent phenotypic variance explained by each QTL was calculated using the “fitqtl” function.

Systems genetic analysis in GeneNetwork2 and identification of candidate genes

To identify candidate genes within our QTLs, we employed a bioinformatic approach. First, using the Sanger Institute Mouse Genomes Project (https://www.sanger.ac.uk/science/data/mouse-genomes-project) website that contains gene annotations of inbred mouse strains, we identified and filtered for genes containing high-impact, polymorphic variants (defined in Supplementary Information) between C57BL/6J and DBA/2J within the QTL intervals. Next, we used GeneNetwork2, an online analysis tool and data repository containing legacy SNP and transcriptome datasets to explore gene regulatory networks (Chesler et al. 2004; Mulligan et al. 2017). We conducted both eQTL and PheQTL-eQTL network analysis using several BXD RI gene expression datasets from multiple brain regions (datasets documented in Supplementary Information) and using the entirety of > 7,000 BXD Published Phenotypes deposited in GeneNetwork2 [BXDPublish; GN602]. Genes were considered candidates if they contained at least two cis-eQTLs across brain regions that peaked within our defined QTL intervals (LRS > 13.8; LOD > 3) and if they correlated with at least two GeneNetwork2 PheQTLs that also peaked within our intervals (Spearman’s rank r > 0.7; r < −0.7; behavioral and physiological GeneNetwork2 traits; LRS > 13.8; LOD > 3). PheQTL-eQTL network graphs were generated for select candidate genes to visualize correlations.

Power analysis

Knowing that we had limited power to detect QTLs with a sample size of 128 F2 mice unless a locus was of large magnitude, we used the R package QTLdesign with the “detectable” function on D23 intake data to generate a plot representing power versus variance explained for an additive QTL (p<0.05) (Sen et al. 2007).

RESULTS

Changes in body weight, palatable food intake, PF-CPP, and power analysis in C57BL/6J x DBA/2J-F2 mice.

A schematic of the behavioral paradigm for binge-like eating in the conditioned place preference paradigm (Kirkpatrick et al. 2017) and the bioinformatics pipeline for systems genetic analysis are provided in Figure 1. In examining % change in body weight from Day(D)1 across binge-like eating training (D2-D18) and compulsive-like eating assessment (D23), males showed a significantly greater percent increase in BW and a greater slope of increase in BW gain compared to females (Figure 2A-C). Despite gaining more weight, males consumed less palatable food than females (Figure 2D, F); however, there was no significant sex difference in the slope of increase in palatable food intake (Figure 2E: p = 0.1).

Figure 1. (A): Schematic of binge-like eating and compulsive-like eating.

Figure 1.

Open in a new tab

D = Day of training. The main phenotypes that were subjected to QTL mapping include % weight gain relative to D1 (for D2, D4, D9, D11, D16, D18), palatable food intake (% body weight for the particular day of assessment (PF; D2, D4, D9, D11, D16, D18), conditioned place preference for the palatable food (PF)-paired side was measured on D22 [D22-D1 (s), food-paired side; “right side”], and compulsive-like eating (red dotted tracing) that was measured on D23 (palatable food intake, % body weight) in the light/dark box by providing access to PF in the light side of the arena. The white circle in the conditioned place preference chambers and in the ligh/dark box denotes the food bowl that either contains palatable food pellets (right side of preference apparatus or light side of light/dark box) or is devoid of any food (left side of preference apparatus). (B): Schematic of the bioinformatics pipeline for identifying positional, functional candidate genes for palatable food intake and associated phenotypes.

Figure 2. Change in body weight and palatable food (PF) intake during binge-like eating, PF-CPP, and power analysis in C57BL/6J x DBA/2J-F2 mice.

Figure 2.

Open in a new tab

(A): Change in body weight [BW; % of Day (D)1] across binge-like eating training days from D2 through D18; prior to assessment of compulsive-like eating on D23. RM ANOVA indicated a main effect of Sex (F1,129 = 24.96, p < 0.0001), Day [Greenhouse-Geisser-adjusted (epsilon for “Day” = 0.53): F2.65,341.85 = 187.53, p < 0.0001], and an interaction (Greenhouse-Geisser-adjusted: F2.65,341.85 = 14.91, p <0.0001). Tukey’s post-hoc comparison indicated that males gained significantly more %BW compared to females on D11, D16, and D18 (*all psadjusted < 0.05). (B): Slope of % increase in body weight (BW) from D2 through D18 in F2 females (n = 66) and in F2 males (n = 65). Unpaired Student’s t-test revealed a significantly steeper rise in % increase in BW over binge-like eating training days in males compared to females (t129 = −5.13, *p = 1.05 x 10−6). (C): In assessing change in BW (% of D1) on D23, unpaired Student’s t-test revealed significantly greater body weight (BW) gain in the males compared to females (t129 = −5.58, *p = 1.39 x 10−07). (D): Palatable food intake (%BW) over binge-like eating training D2 through D18 in F2 females and F2 males. RM ANOVA revealed a main effect of Sex (F1,129 = 4.11, *p = 0.045), indicating an overall increase in PF intake in females compared to males. There was also an effect of Day [Greenhouse-Geisser-adjusted (epsilon for “Day” = 0.74): F3.80,490.43 = 71.32, p < 0.0001); however, the Sex x Day interaction was not statistically significant (Greenhouse-Geisser-adjusted: F3.80,490.33 = 1.98; p = 0.10). (E): Analysis of slope of intake (%BW) from D2 through D18 in F2 females and F2 males using an unpaired Student’s t-test indicated no significant difference (p = 0.10). (F): Palatable food intake (%BW) on D23 (compulsive-like eating assessment) in F2 females and F2 males. Unpaired Student’s t-test revealed significantly less intake in the males compared to the females (t129 = 2.18, *p = 0.03). (G): Time spent on the palatable food (PF)-paired side on D2 prior to binge-like eating training and on D22 post-binge-like eating training. There was a main effect of Day (F1,129 = 14.66, *p = 2 x 10−4) which indicated significant PF-CPP, but no effect of Sex (p = 0.619), and no interaction (p = 0.37). (H): Analysis of the difference in time spent on the palatable food-paired side between D2 and D22 (D22-D2, seconds) in females and males confirmed no significant difference (t129 < 1). (I): Power versus effect size (% variance explained) for an additive QTL model and a sample size of 128 F2 mice (p < 0.05). 0.2, 0.4, 0.6, and 0.8 power is achieved with an observed effect size of 5%, 8%, 10%, and 13% variance explained, respectively.

In examining palatable food conditioned place preference (PF-CPP), there was a main effect of Day (Figure 2G: *p = 2.0 x 10−4) but no interaction with Sex, indicating significant PF-CPP, regardless of Sex. Analysis of the change in preference (D22-D1) confirmed no significant sex difference (Figure 2H).

Figure 2I shows power versus effect size (% variance explained) for an additive QTL with our sample size (128 F2 mice; p < 0.05). Twenty, 40, 60, and 80% power can be achieved with an effect size of 5%, 8%, 10%, and 13% phenotypic variance explained, respectively.

Identification of QTLs underlying changes in body weight and palatable food intake but not PF-CPP

Table 1 summarizes all of the QTLs discussed below. Supplementary Figure 1 and 2 provide visual heat maps of the QTLs for % change in body weight gain and palatable food intake, respectively, via R/qtlcharts (Broman 2015). To facilitate visualization of the effect sizes, for all QTLs reported in Figures 2 through 5, we also provide effect plots for the raw, unnormalized data in Supplementary Figure 3.

Table 1. QTLs for percent change in body weight from D1 and PF intake (quantile-normalized).

D = Day of protocol; cM = centimorgans; Mb = megabases. Bolded rows indicate overlapping QTLs with the raw, un-normalized data (see Supplementary Table 1).

Dataset Trait Chr. Peak, cM
(Mb)		 P-value LOD 1.5 LOD
interval, cM
(Mb)		 Bayes
interval, cM
(Mb)		 %
variance
explained
Combined D16 %ΔBW 18 3 cM (5 Mb) 0.004 6.3 3-12 cM (5-21 Mb) 3-9 cM (5-16 Mb) 41%
Combined D2 intake 5 26 cM (46 Mb) 0.008 5.8 21-28 cM (40-53 Mb) 13-61 cM (26-124 Mb) 19%
Combined D23 intake 6 53 cM (114 Mb) 0.026 5.4 50-59 cM (110-125 Mb) 47-72 cM (93-144 Mb) 21%
 
Males D11 %ΔBW 7 83 cM (138 Mb) 0.042 4 64-89 cM (119-145 Mb) 64-88 cM (118-144 Mb) 40%
Males D2 intake 5 25 cM (45 Mb) 0.031 4.4 12-66 cM (25-128 Mb) 8-66 cM (17-128 Mb) 16%
Males D16 intake 6 53 cM (114 Mb) 0.033 4.3 43-60 cM (93-127 Mb) 30-67 cM (65-136 Mb) 23%
Males D18 intake 5 11 cM (24 Mb) 0.023 4.2 7-25 cM (16-46 Mb) 6-25 cM (14-45 Mb) 24%
Males D23 intake 6 53 cM (114 Mb) 0.002 5.9 52-59 cM (111-125 Mb) 47-72 cM (98-141 Mb) 32%
 
Females D2 %ΔBW 16 51 cM (88 Mb) 0.022 4.5 49-57 (86-97 Mb) 50-57 cM (86-97 Mb) 27%
Females D16 %ΔBW 18 3 cM (5 Mb) 0.002 5.6 3-12 cM (5-21 Mb) 3-9 cM (5-16 Mb) 48%
Females D16 %ΔBW 18 8 cM (15 Mb) 0.002 5.5 3-12 cM (5-21 Mb) 3-9 cM (5-16 Mb) 28%
Females D18 %ΔBW 18 3 cM (5 Mb) 0.038 4.1 3-16 cM (5-31 Mb) 3-13 cM (5-25 Mb) 40%
Females D18 %ΔBW 18 8 cM (15 Mb) 0.034 4.1 3-16 cM (5-31 Mb) 3-13 cM (5-25 Mb) 25%
Females D23 %ΔBW 13 35 cM (68 Mb) 0.049 4.0 29-53 cM (55-100 Mb) 31-52 cM (58-99 Mb) 40%
Females Slope intake 18 24 cM (45 Mb) 0.055 4.1 23-35 cM (43-66 Mb) 23-33 cM (43-66 Mb) 23%
Open in a new tab

Figure 5. Male QTLs influencing palatable food intake on chromosomes 5 and 6.

Figure 5.

Open in a new tab

(A): Genome-wide significant QTLs were also identified on chromosomes 5 and 6 for the males-only dataset. Chromosome 5 contained QTLs for D2 intake and D18 intake. Chromosome 6 contained a QTL for D16 intake and D23 intake. The solid horizontal line for panels A, B, and D indicates the significance threshold (p < 0.05) and the dotted horizontal line indicates the suggestive threshold (p < 0.63). (B): The QTL plot for chromosome 5 shows significant QTLs for D2 intake (red trace) and D18 intake (blue trace). (C): The effect plots for chromosome 5 at the peak loci for D2 intake and D18 intake show increased normalized intake with increasing D alleles. (D): The chromosome 6 QTL plot shows a significant QTL for intake on D16 (light blue trace) and D23 (black trace). (E): Effect plots for the chromosome 6 QTLs at the peak locus shows an increasing effect of the D allele on intake for D16 and D23. For each QTL, see Table 1 for LOD scores, p-values, peak locations, intervals, and % variance explained.

We identified a genome-wide significant QTL on chromosome 18 for D16 % change in body weight (Figure 3A). There was a day-dependent increase in linkage that was significant by D16 (Figure 3B). Sex-specific analyses revealed that only females showed a significant peak (Figure 3C). The effect plots for the sex-combined and females-only dataset indicated that the DBA/2J allele was associated with reduced body weight gain (Figure 3D, E). The null effect plots for males is shown in Figure 3F. We also identified a genome-wide significant QTL for % change in body weight for males on chromosome 7 (Table 1); however, unlike the chromosome 18 QTL in females, the chromosome 7 QTL in males was not significant with the raw dataset (Supplementary Table 1).

Figure 3. Emergence of a genome-wide significant QTL on proximal chromosome 18 influencing % change in body weight during binge-like eating.

Figure 3.

Open in a new tab

(A): Genome-wide QTL plot for % change in body weight (BW) across training days (D) for palatable food intake revealed a significant QTL on chromosome 18 for D16. The solid horizontal lines for panels A-C indicate significance threshold (p < 0.05) and the dotted horizontal lines indicate the suggestive threshold (p < 0 .63). (B): Chromosome 18 QTL plot of % change in body weight (BW) for each day of binge-like eating assessment. (C): Female and male chromosome 18 QTL plots for % change in body weight (BW) on D16. Females showed a highly significant peak at 3 cM (5 Mb), indicating that females drove the overall sex-combined QTL signal. (D): Sex-combined effect plot at the peak locus for D16 % change in body weight. (E): Female effect plot at the peak chromosome 18 locus for D16 % change in body weight. (F): Male effect plot of the peak locus on chromosome 18 for D16 body weight. For each QTL, see Table 1 for LOD scores, p-values, peak locations, intervals, and % variance explained.

QTL analysis of sex-combined palatable food intake revealed genome-wide significant QTLs on chromosomes 5 and 6 for initial intake on D2 and for D23 intake, respectively (Figure 4A). For chromosome 5, the effect plot of the peak locus showed an increase in D2 intake with increasing DBA/2J alleles (Figure 4B,C). For chromosome 6, the DBA/2J allele was also associated with increased intake (Figure 4D,E). The chromosome 6 QTL clearly showed a progressive increase in linkage across intake assessment days (Figure 4D), indicating that the strength of genetic linkage reflects the strength of increased intake.

Figure 4. Sex-combined QTLs influencing palatable food intake on chromosomes 5 and 6.

Figure 4.

Open in a new tab

(A): Genome-wide QTL plot revealed significant QTLs for palatable food intake [% body weight (BW) consumed] on chromosomes 5 and 6. The chromosome 5 QTL was significant for D2 intake. The chromosome 6 QTL was significant for D23 intake. The solid horizontal lines for panels A, B, and D indicate significance threshold (p < 0.05) and the dotted horizontal line indicates the suggestive threshold (p < 0.63). (B): The chromosome 5 QTL plot shows a significant QTL for D2 intake (red trace). (C): The effect plot of the QTL peak for chromosome 5 shows an increase in quantile-normalized D2 intake with each copy of the DBA/2J (D) allele. BB = homozygous for C57BL/6J allele; BD = heterozygous; DD = homozygous for DBA/2J allele. (D): The chromosome 6 QTL plot shows a significant QTL for compulsive-like, palatable food intake on D23 in the light/dark box. (E): The effect plot of the QTL peak for chromosome 6 shows an increase in quantile-normalized D23 intake with each copy of the D allele. For each QTL, see Table 1 for LOD scores, p-values, peak locations, intervals, and % variance explained.

Because we identified sex-dependent candidate loci for binge-like eating (Babbs et al. 2018), we also conducted separate QTL analyses for males and females. For males, we identified QTLs on chromosomes 5 and 6 that mirrored the sex-collapsed results (Figure 5A). For chromosome 5, a significant QTL was again identified for initial palatable food intake on D2 (Figure 5B). The DBA/2J allele was associated with increased intake (Figure 5C).

A significant chromosome 6 QTL was also identified for palatable food intake on D16 in males (Figure 5D) and a similar, yet more robust peak for palatable food intake on D23 during assessment of compulsive-like eating (Figure 5D). The DBA/2J allele was associated with increased palatable food intake (Figure 5E).

For the females-only analysis, there was a nearly significant QTL on chromosome 18 (p = 0.055) underlying the slope of escalation in palatable food intake (Figure 6A,B). Interestingly, the effect plot of the peak locus showed evidence for an overdominance effect, with heterozygotes showing the greatest normalized escalation compared to homozygous genotypes (Figure 6C).

Figure 6. Suggestive female QTL on chromosome 18 for slope of escalation in palatable food intake during binge-like eating.

Figure 6.

Open in a new tab

(A): There was one, nearly significant QTL on chromosome 18 in females (p = 0.055) for the slope of intake across palatable food training days. The chromosome 18 QTL explained 23% of the variance in the slope of intake across days. Solid horizontal line for panels A and B indicates significance threshold (p < 0.05), dotted horizontal line indicates suggestive threshold (p < 0.63). (B): Chromosome 18 QTL plot shows a peak on the medial portion. (C): Chromosome 18 effect plot at the peak locus (24 cM, 45 Mb) shows evidence for an overdominance effect whereby the heterozygous BD genotype displays the greatest normalized slope value. For each QTL, see Table 1 for LOD scores, p-values, peak locations, intervals, and % variance explained.

For QTL analysis of PF-CPP, there were no genome-wide significant QTLs for conditioned place preference for the palatable food-paired side [D22-D1, right side time (s); Figure 1A; Figure 2G,H] in the sex-combined dataset or in the males-only dataset (Supplementary Figure 4A,B). For the females-only dataset, there was a genome-wide significant QTL on distal chromosome 6 for initial time spent on the PF-paired side on D1 that thus, confounded the subtraction measure (D22-D1, s) for change in time spent-on the PF-paired side (Supplementary Figure 4C-E).

To summarize, we identified QTLs influencing % change in body weight and PF intake, some of which were driven primarily by males or females.

eQTL and PheQTL-eQTL networks identify and visualize candidate genes underlying initial and escalated palatable food intake

We employed the bioinformatic pipeline described in the Methods section and illustrated in Figure 1B to identify positional candidate genes for QTLs detected in both the quantile-normalized and raw data (Table 1; Supplementary Table 1), including the female-selective chromosome 18 locus influencing % change in body weight on D16 relative to D1 (3 candidate genes; Table 2), the sex-combined chromosome 5 locus influencing palatable food intake on D2 (2-52 candidate genes; Table 2), and the male-selective chromosome 6 locus influencing compulsive-like palatable food intake on D23 (5-16 candidate genes; Table 2). All PheQTLs (LRS > 13.8; LOD > 3) that peaked within our QTL intervals are provided in Supplementary Tables 2 through 6.

Table 2. Candidate genes within the QTL intervals for change in body weight, initial palatable food intake during binge-like and compulsive-like eating.

Pheno Chr. Interval Candidate genes
D16 %ΔBW (females) 18 1.5 LOD (5-21 Mb) Zeb1, Dsc3, Cul2
D2 intake (sex-combined) 5 1.5 LOD: 40-53 Mb Lcorl, Prom1
D2 intake (sex-combined) 5 Bayes: 26-124 Mb 4930519G04Rik, Afp, Anxa3, Asahl, C530008M17Rik, Cdkl2, Chek2, Cit, Cnot6l, Cops4, Coq2, Coro1c, Cox18, Diablo, Dynll1, Evi5, Fras1, Fryl, Gabrb1, Gcn1l1, Glmn, Gm8730, Hfm1, Idua, Klhl5, Lcorl, Limch1, Lrrc8c, Mapk10, Mpa2, Mtf2, Mlr1, Nos1, Nsun7, Ociad1, Ppat, Prom1, Psmd9, Pxmp2, Rasl11b, Rbpj, Rgs12, Rpap2, Rpl5, Sdad1, Sgsm1, Slit2, Tec, Ttc28, Ung, Usp30, Usp46, Wsb2, Zar1
D23 PF intake (males) 6 1.5 LOD: 111-125 Mb Adipor2, Ogg1, Plxnd1, Setmar, Tmcc1
D23 PF intake (males) 6 Bayes: 98-141 Mb A2ml1, Acrbp, Adipor2, Arl8b, Atf7ip, Emp1, Eps8, Foxp1, Grm7, Itpr1, Lrp6, Ogg1, Plxnd1, Rad18, Setmar, Smim10l1, Tigar, Tmcc1, Tspan11, Ttll3, Wnk1, Zfp9
Open in a new tab

Details regarding the criteria used to identify candidate genes are provided in the Methods and in the Supplementary Information. The physical positions that define the QTL intervals are from Table 1 and are based on the cM positions expanded both proximally and distally to the nearest markers. See Supplementary Tables 2 through 6 for PheQTLs that showed peak LRS scores of > 13.8 (LOD > 3) within these intervals.

For the female-driven chromosome 18 QTL influencing % change in body weight, we identified three candidate genes that met our filtering criteria, including Zeb1, Dsc3, and Cul2 (Table 2). Notably, Zeb1 (a.k.a. Tcf8) codes for a transcription factor that represses adiposity in female C57BL/6 mice (Saykally et al. 2009) and is a central regulator of gene expression in adipose tissue (Gubelmann et al. 2014); thus Zeb1 is a plausible candidate gene for female-specific differences in body weight gain. A complete list of PheQTL traits identified within the chromosome 18 locus is provided in Supplementary Table 2.

For the chromosome 5 QTL influencing initial palatable food intake, we identified 2 candidate genes within the 1.5-LOD interval that met our filtering criteria (Lcorl, Prom1) and 54 candidate genes within the Bayes interval that met our filtering criteria (Table 2). We generated PheQTL-eQTL network diagrams for two strong candidate genes (Lcorl, Prom1) based on position, PheQTL-eQTL connections, and/or function. For Lcorl, 10 traits are shown, along with their strength of correlation, including Lcorl expression in the ventral tegmental area, nucleus accumbens, hypothalamus, and amygdala (Figure 7A). For Prom1, 12 traits are shown, along with their strength of correlation, including Prom1 expression in ventral tegmental area, nucleus accumbens, hypothalamus, amygdala, and striatum (Figure 7B). Complete lists of PheQTL traits within the 1.5-LOD and Bayes intervals are provided in Supplementary Tables 3 and 4.

Figure 7. PheQTL-eQTL networks for the chromosome 5 QTL (1.5-LOD) influencing initial palatable food intake (Lcorl, Prom1).

Figure 7.

Open in a new tab

Diagrams for Lcorl (panel A) and Prom1 (panel B) were generated via GeneNetwork2. Correlated traits, including gene expression (brain region indicated), behavioral, and physiological traits were identified from the GeneNetwork2 database using the pipeline outlined in Figure 1B and described in the Methods section and Supplementary Information. GeneNetwork2 identification numbers are shown for behavioral and physiological traits. A list of all phenotypic traits included in this analysis is provided in Supplementary Table 3. VTA = ventral tegmental area; NAc = nucleus accumbens. For Spearman’s rank correlation coefficients (r), blue = −1 to −0.7; red = +1 to +0.7; green = −0.7 to −0.5; orange = +0.7 to +0.5.

For the chromosome 6 QTL influencing escalated palatable food intake in males, we identified 5 genes within the 1.5-LOD interval and 22 genes within the Bayes interval that met our filtering criteria (Table 2). We generated PheQTL-eQTL network diagrams for two strong candidate genes (Adipor2, Plxnd1) based on position, PheQTL-eQTL connections, and function. For Adipor2, 9 traits are shown, along with their strength of correlation, including Adipor2 expression in the ventral tegmental area and nucleus accumbens (Figure 8A). For Plxnd1, 9 traits are shown, along with their strength of correlation, including Plxnd1 expression in the hypothalamus, amygdala, and striatum (Figure 8B). Complete lists of PheQTL traits within the 1.5-LOD and Bayes intervals are provided in Supplementary Tables 5 and 6.

Figure 8. PheQTL-eQTL networks for the chromosome 6 QTL (1.5-LOD) influencing escalated palatable food intake in males (Adipor2, Plxnd1).

Figure 8.

Open in a new tab

Diagrams for Adipor2 (panel A) and Plxnd1 (panel B) were generated via GeneNetwork2. Correlated traits, including gene expression (brain region indicated), behavioral, and physiological traits were identified from the GeneNetwork2 database using the pipeline outlined in Figure 1B and described in the Methods section and Supplementary Information. GeneNetwork2 identification numbers are shown for behavioral and physiological traits. A list of all phenotypic traits included in this analysis is provided in Supplementary Table 5. HFD = high fat diet; VTA = ventral tegmental area; NAc = nucleus accumbens. For Spearman’s rank correlation coefficients (r), blue = −1 to −0.7; red = +1 to +0.7; green = −0.7 to −0.5; orange = +0.7 to +0.5.

DISCUSSION

We identified major QTLs underlying changes in body weight and palatable food intake during binge-like eating and compulsive-like eating (Table 1), including a female-driven chromosome 18 QTL for % change in body weight (Figure 3), a sex-combined chromosome 5 QTL for initial palatable food intake (Figure 4B,C), a male-driven chromosome 6 locus for escalated palatable food intake (Figure 4D,E; Figure 5D,E), and a nearly significant, female-selective chromosome 18 QTL that contributed to the slope in escalation of palatable food intake (Figure 6). The female chromosome 18 QTL for body weight was more proximally localized and did not overlap with the female chromosome 18 QTL for slope of palatable food intake; thus, they are genetically distinct loci.

Bioinformatic analysis via GeneNetwork2 of the chromosome 5 1.5-LOD QTL interval influencing initial palatable food intake (Figure 4A-C) revealed two strong candidate genes near the QTL peak, (46 Mb), including Lcorl (46 Mb) and Prom1 (44 Mb) (Table 2; Figure 7). Lcorl (ligand-dependent nuclear receptor coprepressor-like) codes for a transcription factor involved in spermatogenesis and is well-documented to influence body size and to a lesser extent, eating traits. In humans, genetic variants in LCORL have been associated with a number of anthropometric traits, including skeletal size and adult height, growth rate in infants, infant birth weight, and birth length (Takasuga 2016). Interestingly, the untranslated region of LCORL overlaps with the well-established FTO-IRX3 locus for obesity (Smemo et al. 2014), suggesting a potential contribution of LCORL. In animal livestock, the LCORL locus accounts for a large portion of the variance in body stature, including body weight, carcass weight, body length, muscle area, adipose tissue, and organ size (Lindholm-Perry et al. 2013; Takasuga 2016). Of relevance to our study, the LCORL locus was previously associated with both food intake and body weight gain in cattle (Lindholm-Perry et al. 2011; Snelling et al. 2011). Fine mapping and testing Lcorl knockout or gene-edited Lcorl mice are logical future directions in pursuit of this translationally relevant locus.

We resolved the male-selective chromosome 6 locus for D23 palatable food intake (Babbs et al. 2018) to a region that peaked at 114 Mb and spanned 111-125 Mb (1.5-LOD: Table 1; Figure 5D). TheTas2r locus containing bitter taste receptors (132.5-133.5 Mb) lies distally from the 1.5-LOD support interval but is contained within the more conservatively estimated Bayes interval (98-141 Mb). Although these observations do not rule out the Tas2r cluster as a source of the QTL, they do call into question whether Tas2r is the primary contributor to escalated palatable food intake. The original QTL for bitter (quinine) taste sensitivity using the same F2 cross was more distal and peaked much closer to Tas2r at 129 Mb (62 cM; D6Mit338) and spanned 112-139 Mb (49-67 cM: D6Mit287-D6Mit198) (Blizard et al. 1999). Other studies identified the same locus for bitter taste sensitivity in other crosses with C57BL/6J, including sucrose octaacetate with NZB/BINJ spanning 87 Mb (38 cM: D6Mit9) to 146 Mb (78 cM: D6Mit14) (Le Roy et al. 1999). Subsequent analysis of quinine sensitivity in the BXD recombinant inbred (RI) strain panel (comprising fixed alleles from C57BL/6J and DBA/2J) resolved the locus to a sharp peak squarely flanking Tas2r (D6Mit13; 132.6 Mb), spanning 125.4 Mb (D6Mit254) to 134.2 Mb (D6Mit374) (Nelson et al. 2005). High resolution mapping confirmed the same peak marker and interval for sucrose octaacetate taste aversion (Bachmanov et al. 2001). To summarize, we located a more proximal chromosome 6 peak and locus for binge-like eating compared to the historical bitter taste locus, suggesting additional genetic factors besides the Tas2r locus contribute to variance in binge-like eating.

What are the causal genetic factor(s) upstream of Tas2r on chromosome 6 that underlie binge-like eating in males? To identify positional candidate genes based on functional evidence and correlations with historical phenotypes, systems genetic analysis using legacy BXD RI datasets from GeneNetwork (Mulligan et al. 2017) identified Adipor2 (adiponectin receptor 2, 119 Mb) as a top candidate gene (Table 2), which codes for a seven transmembrane domain cell-surface receptor for the protein adiponectin, an adipokine secreted by white adipose tissue that regulates the metabolism of lipids and glucose and insulin sensitivity (Yamauchi et al. 2014). Adiponectin acts on Adipor2 and Adipor1 in the periphery and in the brain (hypothalamus, brainstem, pituitary, cortex) to regulate energy homeostasis and other processes such as synaptic plasticity and neurogenesis by signaling through AMPK, p38 MAPK, JNK, PPARα, and NF-kB (Yamauchi et al. 2007; Thundyil et al. 2012; Bloemer et al. 2018). There are 212 variants within Adipor2 that distinguish the DBA/2J strain from the C57BL/6J strain (https://www.sanger.ac.uk), including one 5’ UTR SNP, three, 3’ UTR SNPs, and over 60 nonsense mediated decay intronic variants. Serum levels of adiponectin are inversely correlated with BMI and risk for diabetes and are decreased in patients with Binge Eating Disorder and increased in patients with anorexia nervosa (Khalil & El Hachem 2014). ADIPOR2 is implicated in diabetes, obesity, high-fat feeding, and metabolism (Yamauchi et al. 2014). Adipor2 knockout mice show lower body fat and increased resistance to high-fat diet-induced obesity, along with improved glucose tolerance, increased locomotor activity, and energy metabolism whereas Adipor1 knockouts showed largely opposite phenotypes (Bjursell et al. 2007). In addition, high-fat feeding is associated with lower adiponectin and higher levels of both Adipor1 and 2 (Bullen et al. 2007). Interestingly, adult males (mice and humans) show lower plasma adiponectin than females (Arita et al. 1999; Gui et al. 2004), providing evidence that sex differences in the adiponectin system could underlie male-selective genetic effects of the chromosome 6 locus containing Adipor2 on binge-like eating (Figures 4-5).

Although Adipor2 is an interesting candidate gene, our binge-like eating regimen is relatively abbreviated and is unlikely to induce much metabolic dysfunction as evidenced by a lack of increased weight gain relative to control chow pellet training (Babbs et al. 2018) and a genetic dissociation between QTLs for BW (Figure 3) and binge-like eating (Figures 4-6). Therefore, could Adipor2 dysfunction contribute to earlier physiological processes that initiate progression to BLE? Adipor1, Adipor2, and T-cadherin transcripts (third adiponectin receptor) can be detected in taste receptor cells (Crosson et al. 2019), suggesting that saliva-derived adiponectin could impact taste processing to influence eating. Also, adiponectin-induced activation of Adipor1 expressed on dopamine neurons in the VTA decreased spontaneous neuronal activity and firing and reversed stress-induced increase in dopamine neuron firing and anxiety-like behavior and Adipor1 haploinsufficiency increased dopamine neuron firing and anxiety-like behavior (Sun et al. 2019). These findings suggest that the adiponectin system, traditionally thought to regulate metabolic/homeostatic functions, could communicate taste information (e.g., hedonic versus aversive) to the mesolimbic dopaminergic reward system to influence development of BLE (Figure 7).

A second candidate gene based on GeneNetwork2 analyses was Plxnd1 (Table 2; Supplementary Figure 5) which codes for plexin D1, a cell surface receptor for class 3 semaphorins that regulates migration of cell types, including neuronal axon guidance and synapse formation [e.g., in striatum (Ding et al. 2011)] and vascular development (Oh & Gu 2013). There are 119 variants in Plxnd1 distinguishing DBA/2J from C57BL/6J (https://www.sanger.ac.uk), including three, 3’ UTR variants, two splice site variants, one 5’ UTR variant, and one missense variant. Semaphorin 3E/plexin D1 also mediates macrophage recruitment to visceral adipose tissue during obesity to promote cytokine expression, inflammation and insulin resistance (Schmidt & Moore 2013). PLXND1 has been associated with body fat distribution in humans (Justice et al. 2019) and a nominal genetic association was identified with lipolysis (Strawbridge et al. 2016) – the hydrolysis of lipids to fatty acids in adipocytes that contributes to metabolic dysfunction and obesity. Plxnd1 function is required for normal adipocyte morphology and number, body fat distribution, and insulin sensitivity (Minchin et al. 2015). Thus, Plxnd1 is a reasonably strong, second candidate gene within the chromosome 6 locus that could underlie male-selective differences in palatable food intake during binge-like eating.

There are several limitations to this study. First, our sample size was only powered to detect QTLs of large magnitude (> 13%; Figure 2I). A larger F2 sample size will permit detection of smaller-effect QTLs and explain additional variance. A second limitation is that QTL resolution in F2 mice is notoriously poor. Large-effect QTLs are almost certainly mediated by multiple smaller-effect loci that are in linkage disequilibrium. We previously observed large-effect behavioral QTLs dissolve into smaller and smaller effect sizes that can nevertheless still tractable (Bryant et al. 2012b; Yazdani et al. 2015; Ruan et al. 2020) but in some cases, can completely disappear, presumably due to localized epistasis where closely linked loci are required to detect a phenotype (Bryant et al. 2012a). Advanced intercrosses can narrow the regions to resolve QTLs (Parker et al. 2012) and reduced complexity crosses (Bryant et al. 2018, 2020b) can simplify the genetic complexity of the QTL regions to more readily identify causal genes and nucleotides (Kumar et al. 2013; Kirkpatrick et al. 2017; Mulligan et al. 2019). Careful selection of a subset of BXD-RI strains could potentially help us fine-map the QTLs reported here. For example, there are 54 BXD-RI strains containing at least one historical recombination event within the 111-125 Mb interval on chromosome 6 (https://www.genenetwork.org). Another limitation is that we limited our bioinformatics exercise to genes containing Ensembl-defined high impact variants that were associated with eQTLs in multiple brain tissues. Causal variants could lie within other genes that lack polymorphisms with “high impact” designation, within genes that change protein function without modulating transcript levels, or within intergenic regions not assigned to any nearby genes. A final limitation of this study is that we only assessed consumption of one palatable food diet that was essentially sweetened miniature chow pellets. Although the current dataset cannot speak to the specificity of the observed QTLs for the sweetened component of the chow, note that in our prior study, we found very little parental strain differences in control, miniature chow pellet intake between C57BL/6J and DBA/2J (Babbs et al. 2018).

In summary, we employed a systems genetic approach to identify candidate genes underlying changes in body weight and palatable food intake during binge-like eating in mice, including a chromosome 5 QTL influencing initial palatable food intake containing the candidate gene Lcorl and a QTL on chromosome 6 underlying escalated palatable food intake in males containing the candidate genes Adipor2 and Plxnd1. Phenotyping and fine mapping in a population with more recombination events like the BXD RI panel will reduce the number of candidate genes and variants. Gene/variant editing and validation will permit the study of gene function in the context of multiple eating disorder models, including additional diets (e.g., high fat diet), regimens (cycles of food restriction and binge eating, stress) and comorbidity with other eating disorder models (e.g., activity-based anorexia) and other psychiatric disorders (mood, substance use).

Supplementary Material

Supplementary Table 2
Supplementary Table 1
Supplementary Table 3
Supplementary Information
Supplementary Table 4
Supplementary Table 5
Supplementary Table 6

ACKNOWLEDGMENTS

We thank Dr. Robert Williams and Zachary Sloan at University of Tennessee for their assistance in dataset management and programming in Genenetwork2.

FUNDING

This work was funded by R21DA038738 (C.D.B.), R01DA039168 (C.D.B.), U01DA050243 (C.D.B.), and R01CA221260 (M.I.D., C.D.B.)

Footnotes

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

CONSENT TO PUBLICATION

All authors have read and approved of the final version of the manuscript that has been submitted for publication.

DATA AVAILABILITY STATEMENT

All data in its raw and processed forms will be made immediately available upon request.

REFERENCES

  1. Arita Y, Kihara S, Ouchi N, Takahashi M, Maeda K, Miyagawa J, Hotta K, Shimomura I, Nakamura T, Miyaoka K, Kuriyama H, Nishida M, Yamashita S, Okubo K, Matsubara K, Muraguchi M, Ohmoto Y, Funahashi T & Matsuzawa Y (1999) Paradoxical decrease of an adipose-specific protein, adiponectin, in obesity. Biochem Biophys Res Commun 257, 79–83. [DOI] [PubMed] [Google Scholar]
  2. Babbs RK, Beierle JA, Ruan QT, Kelliher JC, Chen MM, Feng AX, Kirkpatrick SL, Benitez FA, Rodriguez FA, Pierre JJ, Anandakumar J, Kumar V, Mulligan MK & Bryant CD (2019) Cyfip1 Haploinsufficiency Increases Compulsive-Like Behavior and Modulates Palatable Food Intake in Mice: Dependence on Cyfip2 Genetic Background, Parent-of Origin, and Sex. G3 (Bethesda) 9, 3009–3022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Babbs RK, Kelliher JC, Scotellaro JL, Luttik KP, Mulligan MK & Bryant CD (2018) Genetic differences in the behavioral organization of binge eating, conditioned food reward, and compulsive-like eating in C57BL/6J and DBA/2J strains. PhysiolBehav 197, 51–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bachmanov AA, Li X, Li S, Neira M, Beauchamp GK & Azen EA (2001) High-resolution genetic mapping of the sucrose octaacetate taste aversion (Soa) locus on mouse Chromosome 6. Mamm Genome 12, 695–699. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bjursell M, Ahnmark A, Bohlooly-Y M, William-Olsson L, Rhedin M, Peng X-R, Ploj K, Gerdin A-K, Arnerup G, Elmgren A, Berg A-L, Oscarsson J & Lindén D (2007) Opposing effects of adiponectin receptors 1 and 2 on energy metabolism. Diabetes 56, 583–593. [DOI] [PubMed] [Google Scholar]
  6. Blizard DA, Kotlus B & Frank ME (1999) Quantitative trait loci associated with short-term intake of sucrose, saccharin and quinine solutions in laboratory mice. ChemSenses 24, 373–385. [DOI] [PubMed] [Google Scholar]
  7. Bloemer J, Pinky PD, Govindarajulu M, Hong H, Judd R, Amin RH, Moore T, Dhanasekaran M, Reed MN & Suppiramaniam V (2018) Role of Adiponectin in Central Nervous System Disorders. Neural Plast 2018, 4593530. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Broman KW (2015) R/qtlcharts: interactive graphics for quantitative trait locus mapping. Genetics 199, 359–361. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Broman KW, Wu H, Sen S & Churchill GA (2003) R/qtl: QTL mapping in experimental crosses. Bioinformatics (Oxford, England) 19, 889–90. [DOI] [PubMed] [Google Scholar]
  10. Bryant CD, Ferris MT, De Villena FPM, Damaj MI, Kumar V & Mulligan MK (2018) Reduced complexity cross design for behavioral genetics. In Gerlai RT (ed), Molecular-Genetic and Statistical Techniques for Behavioral and Neural Research, pp. 165–190. [Google Scholar]
  11. Bryant CD, Kole LA, Guido MA, Sokoloff G & Palmer AA (2012a) Congenic dissection of a major QTL for methamphetamine sensitivity implicates epistasis. Genes Brain Behav 11, 623–632. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Bryant CD, Parker CC, Zhou L, Olker C, Chandrasekaran RY, Wager TT, Bolivar VJ, Loudon AS, Vitaterna MH, Turek FW & Palmer AA (2012b) Csnk1e is a genetic regulator of sensitivity to psychostimulants and opioids. Neuropsychopharmacology 37, 1026–1035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Bryant CD, Smith DJ, Kantak KM, Nowak TS, Williams RW, Damaj MI, Redei EE, Chen H & Mulligan MK (2020a) Facilitating Complex Trait Analysis via Reduced Complexity Crosses. Trends Genet. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Bryant CD, Smith DJ, Kantak KM, Nowak TS, Williams RW, Damaj MI, Redei EE, Chen H & Mulligan MK (2020b) Facilitating Complex Trait Analysis via Reduced Complexity Crosses. Trends Genet 36, 549–562. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Bulik CM, Sullivan PF & Kendler KS (2003) Genetic and environmental contributions to obesity and binge eating. IntJEatDisord 33, 293–298. [DOI] [PubMed] [Google Scholar]
  16. Bullen JW, Bluher S, Kelesidis T & Mantzoros CS (2007) Regulation of adiponectin and its receptors in response to development of diet-induced obesity in mice. Am J Physiol Endocrinol Metab 292, E1079–1086. [DOI] [PubMed] [Google Scholar]
  17. Chesler EJ, Lu L, Wang J, Williams RW & Manly KF (2004) WebQTL: rapid exploratory analysis of gene expression and genetic networks for brain and behavior. Nature neuroscience 7, 485–6. [DOI] [PubMed] [Google Scholar]
  18. Crosson SM, Marques A, Dib P, Dotson CD, Munger SD & Zolotukhin S (2019) Taste Receptor Cells in Mice Express Receptors for the Hormone Adiponectin. Chem Senses 44, 409–422. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Ding JB, Oh W-J, Sabatini BL & Gu C (2011) Semaphorin 3E-Plexin-D1 signaling controls pathway-specific synapse formation in the striatum. Nat Neurosci 15, 215–223. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Goldberg LR, Kirkpatrick SL, Yazdani N, Luttik KP, Lacki OA, Babbs RK, Jenkins DF, Johnson WE & Bryant CD (2017) Casein kinase 1-epsilon deletion increases mu opioid receptor-dependent behaviors and binge eating1. Genes Brain Behav 16, 725–738. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gubelmann C, Schwalie PC, Raghav SK, Röder E, Delessa T, Kiehlmann E, Waszak SM, Corsinotti A, Udin G, Holcombe W, Rudofsky G, Trono D, Wolfrum C & Deplancke B (2014) Identification of the transcription factor ZEB1 as a central component of the adipogenic gene regulatory network. eLife 3, e03346. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Guerdjikova AI, Mori N, Casuto LS & McElroy SL (2019) Update on Binge Eating Disorder. Med Clin North Am 103, 669–680. [DOI] [PubMed] [Google Scholar]
  23. Gui Y, Silha JV & Murphy LJ (2004) Sexual dimorphism and regulation of resistin, adiponectin, and leptin expression in the mouse. Obes Res 12, 1481–1491. [DOI] [PubMed] [Google Scholar]
  24. Hübel C, Leppä V, Breen G & Bulik CM (2018) Rigor and reproducibility in genetic research on eating disorders. Int J Eat Disord 51, 593–607. [DOI] [PubMed] [Google Scholar]
  25. Johnson WG, Boutelle KN, Torgrud L, Davig JP & Turner S (2000) What is a binge? The influence of amount, duration, and loss of control criteria on judgments of binge eating. Int J Eat Disord 27, 471–479. [DOI] [PubMed] [Google Scholar]
  26. Justice AE, Karaderi T, Highland HM, Young KL, Graff M, Lu Y, Turcot V, Auer PL, Fine RS, Guo X, Schurmann C, Lempradl A, Marouli E, Mahajan A, Winkler TW, Locke AE, Medina-Gomez C, Esko T, Vedantam S, Giri A, Lo KS, Alfred T, Mudgal P, Ng MCY, Heard-Costa NL, Feitosa MF, Manning AK, Willems SM, Sivapalaratnam S, Abecasis G, Alam DS, Allison M, Amouyel P, Arzumanyan Z, Balkau B, Bastarache L, Bergmann S, Bielak LF, Blüher M, Boehnke M, Boeing H, Boerwinkle E, Böger CA, Bork-Jensen J, Bottinger EP, Bowden DW, Brandslund I, Broer L, Burt AA, Butterworth AS, Caulfield MJ, Cesana G, Chambers JC, Chasman DI, Chen Y-DI, Chowdhury R, Christensen C, Chu AY, Collins FS, Cook JP, Cox AJ, Crosslin DS, Danesh J, de Bakker PIW, Denus S. de, Mutsert R. de, Dedoussis G, Demerath EW, Dennis JG, Denny JC, Di Angelantonio E, Dörr M, Drenos F, Dubé M-P, Dunning AM, Easton DF, Elliott P, Evangelou E, Farmaki A-E, Feng S, Ferrannini E, Ferrieres J, Florez JC, Fornage M, Fox CS, Franks PW, Friedrich N, Gan W, Gandin I, Gasparini P, Giedraitis V, Girotto G, Gorski M, Grallert H, Grarup N, Grove ML, Gustafsson S, Haessler J, Hansen T, Hattersley AT, Hayward C, Heid IM, Holmen OL, Hovingh GK, Howson JMM, Hu Y, Hung Y-J, Hveem K, Ikram MA, Ingelsson E, Jackson AU, Jarvik GP, Jia Y, Jørgensen T, Jousilahti P, Justesen JM, Kahali B, Karaleftheri M, Kardia SLR, Karpe F, Kee F, Kitajima H, Komulainen P, Kooner JS, Kovacs P, Krämer BK, Kuulasmaa K, Kuusisto J, Laakso M, Lakka TA, Lamparter D, Lange LA, Langenberg C, Larson EB, Lee NR, Lee W-J, Lehtimäki T, Lewis CE, Li H, Li J, Li-Gao R, Lin L-A, Lin X, Lind L, Lindström J, Linneberg A, Liu C-T, Liu DJ, Luan J, Lyytikäinen L-P, MacGregor S, Mägi R, Männistö S, Marenne G, Marten J, Masca NGD, McCarthy MI, Meidtner K, Mihailov E, Moilanen L, Moitry M, Mook-Kanamori DO, Morgan A, Morris AP, Müller-Nurasyid M, Munroe PB, Narisu N, Nelson CP, Neville M, Ntalla I, O’Connell JR, Owen KR, Pedersen O, Peloso GM, Pennell CE, Perola M, Perry JA, Perry JRB, Pers TH, Ewing A, Polasek O, Raitakari OT, Rasheed A, Raulerson CK, Rauramaa R, Reilly DF, Reiner AP, Ridker PM, Rivas MA, Robertson NR, Robino A, Rudan I, Ruth KS, Saleheen D, Salomaa V, Samani NJ, Schreiner PJ, Schulze MB, Scott RA, Segura-Lepe M, Sim X, Slater AJ, Small KS, Smith BH, Smith JA, Southam L, Spector TD, Speliotes EK, Stefansson K, Steinthorsdottir V, Stirrups KE, Strauch K, Stringham HM, Stumvoll M, Sun L, Surendran P, Swart KMA, Tardif J-C, Taylor KD, Teumer A, Thompson DJ, Thorleifsson G, Thorsteinsdottir U, Thuesen BH, Tönjes A, Torres M, Tsafantakis E, Tuomilehto J, Uitterlinden AG, Uusitupa M, van Duijn CM, Vanhala M, Varma R, Vermeulen SH, Vestergaard H, Vitart V, Vogt TF, Vuckovic D, Wagenknecht LE, Walker M, Wallentin L, Wang F, Wang CA, Wang S, Wareham NJ, Warren HR, Waterworth DM, Wessel J, White HD, Willer CJ, Wilson JG, Wood AR, Wu Y, Yaghootkar H, Yao J, Yerges-Armstrong LM, Young R, Zeggini E, Zhan X, Zhang W, Zhao JH, Zhao W, Zheng H, Zhou W, Zillikens MC, Rivadeneira F, Borecki IB, Pospisilik JA, Deloukas P, Frayling TM, Lettre G, Mohlke KL, Rotter JI, Kutalik Z, Hirschhorn JN, Cupples LA, Loos RJF, North KE, Lindgren CM, CHD Exome+ Consortium, Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium, EPIC-CVD Consortium, ExomeBP Consortium, Global Lipids Genetic Consortium, GoT2D Genes Consortium, InterAct, ReproGen Consortium, T2D-Genes Consortium, & MAGIC Investigators. (2019) Protein-coding variants implicate novel genes related to lipid homeostasis contributing to body-fat distribution. Nat Genet 51, 452–469. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Khalil RB & El Hachem C (2014) Adiponectin in eating disorders. Eat Weight Disord 19, 3–10. [DOI] [PubMed] [Google Scholar]
  28. Kirkpatrick SL, Goldberg LR, Yazdani N, Babbs RK, Wu J, Reed ER, Jenkins DF, Bolgioni AF, Landaverde KI, Luttik KP, Mitchell KS, Kumar V, Johnson WE, Mulligan MK, Cottone P & Bryant CD (2017) Cytoplasmic FMR1-Interacting Protein 2 Is a Major Genetic Factor Underlying Binge Eating. BiolPsychiatry 81, 757–769. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kumar V, Kim K, Joseph C, Kourrich S, Yoo SH, Huang HC, Vitaterna MH, de Villena FP, Churchill G, Bonci A & Takahashi JS (2013) C57BL/6N mutation in Cytoplasmic FMRP interacting protein 2 regulates cocaine response. Science 342, 1508–1512. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Le Roy I, Pager J & Roubertoux PL (1999) Genetic dissection of gustatory sensitivity to bitterness (sucrose octaacetate) in mice. C R Acad Sci III, Sci Vie 322, 831–836. [DOI] [PubMed] [Google Scholar]
  31. Levran O, Londono D, O’Hara K, Nielsen DA, Peles E, Rotrosen J, Casadonte P, Linzy S, Randesi M, Ott J, Adelson M & Kreek MJ (2008) Genetic susceptibility to heroin addiction; a candidate-gene association study. Genes, brain, and behavior. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Levran O, Randesi M, da Rosa JC, Ott J, Rotrosen J, Adelson M & Kreek MJ (2015) Overlapping dopaminergic pathway genetic susceptibility to heroin and cocaine addictions in African Americans. AnnHumGenet 79, 188–198. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Lindholm-Perry AK, Kuehn LA, Oliver WT, Sexten AK, Miles JR, Rempel LA, Cushman RA & Freetly HC (2013) Adipose and muscle tissue gene expression of two genes (NCAPG and LCORL) located in a chromosomal region associated with cattle feed intake and gain. PLoS One 8, e80882. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Lindholm-Perry AK, Sexten AK, Kuehn LA, Smith TPL, King DA, Shackelford SD, Wheeler TL, Ferrell CL, Jenkins TG, Snelling WM & Freetly HC (2011) Association, effects and validation of polymorphisms within the NCAPG - LCORL locus located on BTA6 with feed intake, gain, meat and carcass traits in beef cattle. BMC Genet 12, 103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. da Luz FQ, Hay P, Touyz S & Sainsbury A (2018) Obesity with Comorbid Eating Disorders: Associated Health Risks and Treatment Approaches. Nutrients 10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Minchin JEN, Dahlman I, Harvey CJ, Mejhert N, Singh MK, Epstein JA, Arner P, Torres-Vázquez J & Rawls JF (2015) Plexin D1 determines body fat distribution by regulating the type V collagen microenvironment in visceral adipose tissue. Proc Natl Acad Sci USA 112, 4363–4368. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Mulligan MK, Abreo T, Neuner SM, Parks C, Watkins CE, Houseal MT, Shapaker TM, Hook M, Tan H, Wang X, Ingels J, Peng J, Lu L, Kaczorowski CC, Bryant CD, Homanics GE & Williams RW (2019) Identification of a functional non-coding variant in the GABAA receptor α2 subunit of the C57BL/6J mouse reference genome: Major implications for neuroscience research. Frontiers in Genetics 540211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Mulligan MK, Mozhui K, Prins P & Williams RW (2017) GeneNetwork: A Toolbox for Systems Genetics. Methods MolBiol 1488, 75–120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Munn-Chernoff MA & Baker JH (2016) A Primer on the Genetics of Comorbid Eating Disorders and Substance Use Disorders. EurEatDisordRev 24, 91–100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Nelson TM, Munger SD & Boughter JD Jr. (2005) Haplotypes at the Tas2r locus on distal chromosome 6 vary with quinine taste sensitivity in inbred mice. BMC Genet 6, 32-2156–6-32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Oh W-J & Gu C (2013) The role and mechanism-of-action of Sema3E and Plexin-D1 in vascular and neural development. Semin Cell Dev Biol 24, 156–162. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Parker CC, Cheng R, Sokoloff G & Palmer AA (2012) Genome-wide association for methamphetamine sensitivity in an advanced intercross mouse line. Genes Brain Behav 11, 52–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Peterson R & Cavanaugh J (2019) Ordered quantile normalization: a semiparametric transformation built for the cross-validation era. Journal of Applied Statistics 1–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Ruan QT, Yazdani N, Reed ER, Beierle JA, Peterson LP, Luttik KP, Szumlinski KK, Johnson WE, Ash PEA, Wolozin B & Bryant CD (2020) 5’ UTR variants in the quantitative trait gene Hnrnph1 support reduced 5’ UTR usage and hnRNP H protein as a molecular mechanism underlying reduced methamphetamine sensitivity. FASEB J 34, 9223–9244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Saykally JN, Dogan S, Cleary MP & Sanders MM (2009) The ZEB1 transcription factor is a novel repressor of adiposity in female mice. PLoS One 4, e8460. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Schmidt AM & Moore KJ (2013) The Semaphorin 3E/PlexinD1 axis regulates macrophage inflammation in obesity. Cell Metab 18, 461–462. [DOI] [PubMed] [Google Scholar]
  47. Sen S, Satagopan JM, Broman KW & Churchill GA (2007) R/qtlDesign: inbred line cross experimental design. MammGenome 18, 87–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Smemo S, Tena JJ, Kim K-H, Gamazon ER, Sakabe NJ, Gómez-Marín C, Aneas I, Credidio FL, Sobreira DR, Wasserman NF, Lee JH, Puviindran V, Tam D, Shen M, Son JE, Vakili NA, Sung H-K, Naranjo S, Acemel RD, Manzanares M, Nagy A, Cox NJ, Hui C-C, Gomez-Skarmeta JL & Nóbrega MA (2014) Obesity-associated variants within FTO form long-range functional connections with IRX3. Nature 507, 371–375. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Snelling WM, Allan MF, Keele JW, Kuehn LA, Thallman RM, Bennett GL, Ferrell CL, Jenkins TG, Freetly HC, Nielsen MK & Rolfe KM (2011) Partial-genome evaluation of postweaning feed intake and efficiency of crossbred beef cattle. J Anim Sci 89, 1731–1741. [DOI] [PubMed] [Google Scholar]
  50. Strawbridge RJ, Laumen H, Hamsten A, Breier M, Grallert H, Hauner H, Arner P & Dahlman I (2016) Effects of Genetic Loci Associated with Central Obesity on Adipocyte Lipolysis. PLoS ONE 11, e0153990. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Sun F, Lei Y, You J, Li C, Sun L, Garza J, Zhang D, Guo M, Scherer PE, Lodge D & Lu X-Y (2019) Adiponectin modulates ventral tegmental area dopamine neuron activity and anxiety-related behavior through AdipoR1. Mol Psychiatry 24, 126–144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Takasuga A (2016) PLAG1 and NCAPG-LCORL in livestock. Anim Sci J 87, 159–167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Thundyil J, Pavlovski D, Sobey CG & Arumugam TV (2012) Adiponectin receptor signalling in the brain. Br J Pharmacol 165, 313–327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Vannucci A, Nelson EE, Bongiorno DM, Pine DS, Yanovski JA & Tanofsky-Kraff M (2015) Behavioral and neurodevelopmental precursors to binge-type eating disorders: support for the role of negative valence systems. Psychol Med 45, 2921–2936. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Vannucci A, Theim KR, Kass AE, Trockel M, Genkin B, Rizk M, Weisman H, Bailey JO, Sinton MM, Aspen V, Wilfley DE & Taylor CB (2013) What constitutes clinically significant binge eating? Association between binge features and clinical validators in college-age women. Int J Eat Disord 46, 226–232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Wager TT, Chandrasekaran RY, Bradley J, Rubitski D, Berke H, Mente S, Butler T, Doran A, Chang C, Fisher K, Knafels J, Liu S, Ohren J, Marconi M, DeMarco G, Sneed B, Walton K, Horton D, Rosado A & Mead A (2014) Casein kinase 1δ/ε inhibitor PF-5006739 attenuates opioid drug-seeking behavior. ACS Chem Neurosci 5, 1253–1265. [DOI] [PubMed] [Google Scholar]
  57. Watson HJ, Yilmaz Z, Thornton LM, Hübel C, Coleman JRI, Gaspar HA, Bryois J, Hinney A, Leppä VM, Mattheisen M, Medland SE, Ripke S, Yao S, Giusti-Rodríguez P, Anorexia Nervosa Genetics Initiative, Hanscombe KB, Purves KL, Eating Disorders Working Group of the Psychiatric Genomics Consortium, Adan RAH, Alfredsson L, Ando T, Andreassen OA, Baker JH, Berrettini WH, Boehm I, Boni C, Perica VB, Buehren K, Burghardt R, Cassina M, Cichon S, Clementi M, Cone RD, Courtet P, Crow S, Crowley JJ, Danner UN, Davis OSP, de Zwaan M, Dedoussis G, Degortes D, DeSocio JE, Dick DM, Dikeos D, Dina C, Dmitrzak-Weglarz M, Docampo E, Duncan LE, Egberts K, Ehrlich S, Escaramís G, Esko T, Estivill X, Farmer A, Favaro A, Fernández-Aranda F, Fichter MM, Fischer K, Föcker M, Foretova L, Forstner AJ, Forzan M, Franklin CS, Gallinger S, Giegling I, Giuranna J, Gonidakis F, Gorwood P, Mayora MG, Guillaume S, Guo Y, Hakonarson H, Hatzikotoulas K, Hauser J, Hebebrand J, Helder SG, Herms S, Herpertz-Dahlmann B, Herzog W, Huckins LM, Hudson JI, Imgart H, Inoko H, Janout V, Jiménez-Murcia S, Julià A, Kalsi G, Kaminská D, Kaprio J, Karhunen L, Karwautz A, Kas MJH, Kennedy JL, Keski-Rahkonen A, Kiezebrink K, Kim Y-R, Klareskog L, Klump KL, Knudsen GPS, La Via MC, Le Hellard S, Levitan RD, Li D, Lilenfeld L, Lin BD, Lissowska J, Luykx J, Magistretti PJ, Maj M, Mannik K, Marsal S, Marshall CR, Mattingsdal M, McDevitt S, McGuffin P, Metspalu A, Meulenbelt I, Micali N, Mitchell K, Monteleone AM, Monteleone P, Munn-Chernoff MA, Nacmias B, Navratilova M, Ntalla I, O’Toole JK, Ophoff RA, Padyukov L, Palotie A, Pantel J, Papezova H, Pinto D, Rabionet R, Raevuori A, Ramoz N, Reichborn-Kjennerud T, Ricca V, Ripatti S, Ritschel F, Roberts M, Rotondo A, Rujescu D, Rybakowski F, Santonastaso P, Scherag A, Scherer SW, Schmidt U, Schork NJ, Schosser A, Seitz J, Slachtova L, Slagboom PE, Slof-Op ’t Landt MCT, Slopien A, Sorbi S, Świątkowska B, Szatkiewicz JP, Tachmazidou I, Tenconi E, Tortorella A, Tozzi F, Treasure J, Tsitsika A, Tyszkiewicz-Nwafor M, Tziouvas K, van Elburg AA, van Furth EF, Wagner G, Walton E, Widen E, Zeggini E, Zerwas S, Zipfel S, Bergen AW, Boden JM, Brandt H, Crawford S, Halmi KA, Horwood LJ, Johnson C, Kaplan AS, Kaye WH, Mitchell JE, Olsen CM, Pearson JF, Pedersen NL, Strober M, Werge T, Whiteman DC, Woodside DB, Stuber GD, Gordon S, Grove J, Henders AK, Juréus A, Kirk KM, Larsen JT, Parker R, Petersen L, Jordan J, Kennedy M, Montgomery GW, Wade TD, Birgegård A, Lichtenstein P, Norring C, Landén M, Martin NG, Mortensen PB, Sullivan PF, Breen G & Bulik CM (2019) Genome-wide association study identifies eight risk loci and implicates metabo-psychiatric origins for anorexia nervosa. Nat Genet 51, 1207–1214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Wolfe BE, Baker CW, Smith AT & Kelly-Weeder S (2009) Validity and utility of the current definition of binge eating. IntJEatDisord 42, 674–686. [DOI] [PubMed] [Google Scholar]
  59. Yamauchi T, Iwabu M, Okada-Iwabu M & Kadowaki T (2014) Adiponectin receptors: a review of their structure, function and how they work. Best Pract Res Clin Endocrinol Metab 28, 15–23. [DOI] [PubMed] [Google Scholar]
  60. Yamauchi T, Nio Y, Maki T, Kobayashi M, Takazawa T, Iwabu M, Okada-Iwabu M, Kawamoto S, Kubota N, Kubota T, Ito Y, Kamon J, Tsuchida A, Kumagai K, Kozono H, Hada Y, Ogata H, Tokuyama K, Tsunoda M, Ide T, Murakami K, Awazawa M, Takamoto I, Froguel P, Hara K, Tobe K, Nagai R, Ueki K & Kadowaki T (2007) Targeted disruption of AdipoR1 and AdipoR2 causes abrogation of adiponectin binding and metabolic actions. Nat Med 13, 332–339. [DOI] [PubMed] [Google Scholar]
  61. Yazdani N, Parker CC, Shen Y, Reed ER, Guido MA, Kole LA, Kirkpatrick SL, Lim JE, Sokoloff G, Cheng R, Johnson WE, Palmer AA & Bryant CD (2015) Hnrnph1 Is A Quantitative Trait Gene for Methamphetamine Sensitivity. PLoS Genet 11, e1005713. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Table 2
NIHMS1820672-supplement-Supplementary_Table_2.xlsx (10.6KB, xlsx)
Supplementary Table 1
NIHMS1820672-supplement-Supplementary_Table_1.docx (15.7KB, docx)
Supplementary Table 3
NIHMS1820672-supplement-Supplementary_Table_3.xlsx (12.8KB, xlsx)
Supplementary Information
NIHMS1820672-supplement-Supplementary_Information.docx (1.8MB, docx)
Supplementary Table 4
NIHMS1820672-supplement-Supplementary_Table_4.xlsx (21.8KB, xlsx)
Supplementary Table 5
NIHMS1820672-supplement-Supplementary_Table_5.xlsx (13.4KB, xlsx)
Supplementary Table 6
NIHMS1820672-supplement-Supplementary_Table_6.xlsx (17.6KB, xlsx)

Data Availability Statement

All data in its raw and processed forms will be made immediately available upon request.

RESOURCES