Peroxisome proliferator-activated receptors α and γ are linked with alcohol consumption in mice and withdrawal and dependence in humans

Abstract

Background: Peroxisome proliferator-activated receptor (PPAR) agonists reduce voluntary ethanol (EtOH) consumption in rat models and are promising therapeutics in the treatment for drug addictions. We studied the effects of different classes of PPAR agonists on chronic EtOH intake and preference in mice with a genetic predisposition for high alcohol consumption and then examined human genomewide association data for polymorphisms in PPAR genes in alcohol-dependent subjects.

Methods: Two different behavioral tests were used to measure intake of 15% EtOH in C57BL/6J male mice: 24-hour 2-bottle choice and limited access (3-hour) 2-bottle choice, drinking in the dark. We measured the effects of pioglitazone (10 and 30 mg/kg), fenofibrate (50 and 150 mg/kg), GW0742 (10 mg/kg), tesaglitazar (1.5 mg/kg), and bezafibrate (25 and 75 mg/kg) on EtOH intake and preference. Fenofibric acid, the active metabolite of fenofibrate, was quantified in mouse plasma, liver, and brain by liquid chromatography tandem mass spectrometry. Data from a human genome-wide association study (GWAS) completed in the Collaborative Study on the Genetics of Alcoholism (COGA) were then used to analyze the association of single nucleotide polymorphisms (SNPs) in different PPAR genes (PPARA, PPARD, PPARG, and PPARGC1A) with 2 phenotypes: DSM-IV alcohol dependence (AD) and the DSM-IV criterion of withdrawal.

Results: Activation of 2 isoforms of PPARs, α and γ, reduced EtOH intake and preference in the 2 different consumption tests in mice. However, a selective PPARδ agonist or a pan agonist for all 3 PPAR isoforms did not decrease EtOH consumption. Fenofibric acid, the active metabolite of the PPARα agonist fenofibrate, was detected in liver, plasma, and brain after 1 or 8 days of oral treatment. The GWAS from COGA supported an association of SNPs in PPARA and PPARG with alcohol withdrawal and PPARGC1A with AD but found no association for PPARD with either phenotype.

Conclusions: We provide convergent evidence using both mouse and human data for specific PPARs in alcohol action. Reduced EtOH intake in mice and the genetic association between AD or withdrawal in humans highlight the potential for repurposing FDA-approved PPARα or PPARγ agonists for the treatment of AD.

Keywords: C57BL/6J; Fenofibrate; Fenofibric Acid; Genome-Wide Association Study; Pioglitazone; Tesag-litazar; Two-Bottle Choice.

Figure 1. Effects of PPAR agonists on ethanol intake after the first 6 hours in the 24-hour two-bottle choice test in C57BL/6J male mice

After at least 3 weeks of 15% ethanol consumption and after stable intake was reached, ethanol (EtOH) consumption was measured (g/kg/6 hours) after 2 days of saline administration (day 2 in graph) and mice were grouped to provide similar levels of ethanol intake and preference. Beginning on day 3, saline or drug was administered and intake averaged over 2-day periods using different bottle positions (see Methods for details). A. Pioglitazone (n=13) B. Fenofibrate (n=6) C. GW0742 (n=6) D. Tesaglitazar (n=6) E. Bezafibrate (n=6). Data were analyzed by two-way repeated measures ANOVA followed by Bonferroni’s test for multiple comparisons. *p<0.05; **p<0.01; ***p<0.001 compared to control.

Figure 2. Effects of PPAR agonists on ethanol intake during limited access (3-hour) two-bottle choice Drinking in the Dark test in C57BL/6J male mice

After at least 3 weeks of 15% ethanol consumption and after stable intake was reached, ethanol (EtOH) consumption was measured (g/kg/3 hours) after 2 days of saline administration (day 2 in graph) and mice were grouped to provide similar levels of ethanol intake and preference. Beginning on day 3, saline or drug was administered and intake averaged over 2-day periods using different bottle positions (see Methods for details). A. Pioglitazone B. Fenofibrate C. GW0742 D. Tesaglitazar E. Bezafibrate. Data were analyzed by Student's t-test or two-way repeated measures ANOVA followed by Bonferroni’s test for multiple comparisons. *p<0.05; **p<0.01; ***p<0.001 compared to control. (n=6 for all groups)

Figure 3. Association results from the Collaborative Study on the Genetics of Alcoholism (COGA)

A. PPARA and withdrawal B. PPARG and withdrawal C. PPARGC1A and AD. Y-axis denotes the –log₁₀ (p-value) for association. X-axis is the physical position on the chromosome (Mb). The most significantly associated SNP is denoted with a purple symbol, and the SNP name is shown below the color scale. The extent of linkage disequilibrium (LD, as measured by r²) between each SNP and the most significantly associated SNP with the lowest p-value within the gene is indicated by the color scale. Larger values of r² indicate greater LD. Association results with genotyped SNPs are shown as a circle while association results with imputed SNPs are shown as a square.

Figure 3. Association results from the Collaborative Study on the Genetics of Alcoholism (COGA)

A. PPARA and withdrawal B. PPARG and withdrawal C. PPARGC1A and AD. Y-axis denotes the –log₁₀ (p-value) for association. X-axis is the physical position on the chromosome (Mb). The most significantly associated SNP is denoted with a purple symbol, and the SNP name is shown below the color scale. The extent of linkage disequilibrium (LD, as measured by r²) between each SNP and the most significantly associated SNP with the lowest p-value within the gene is indicated by the color scale. Larger values of r² indicate greater LD. Association results with genotyped SNPs are shown as a circle while association results with imputed SNPs are shown as a square.

Figure 3. Association results from the Collaborative Study on the Genetics of Alcoholism (COGA)

A. PPARA and withdrawal B. PPARG and withdrawal C. PPARGC1A and AD. Y-axis denotes the –log₁₀ (p-value) for association. X-axis is the physical position on the chromosome (Mb). The most significantly associated SNP is denoted with a purple symbol, and the SNP name is shown below the color scale. The extent of linkage disequilibrium (LD, as measured by r²) between each SNP and the most significantly associated SNP with the lowest p-value within the gene is indicated by the color scale. Larger values of r² indicate greater LD. Association results with genotyped SNPs are shown as a circle while association results with imputed SNPs are shown as a square.