Identification Of Genes Associated With Heroin Dependence By An Endophenotype-based Method

Object: To evaluate the association between opioid receptor genes (μ,δandκreceptor gene: OPRM1, OPRD1, OPRK1, respectively) and heroin-inducedsubjective responses, and cue-elicited heroin craving.Method: We studied 336 Chinese Han heroin addicts recruited in Shanghai anddivided heroin addicts into two groups (positive vs. negative) according to theself-reporting feeling on first use. Association analyses with the genotypes and allelesin nine tagging single nucleotide polymorphisms (tSNPs) in OPRM1, five tSNPs withOPRD1, and six tSNPs in OPRK1 with subjective responses were performed. Similaranalysis with haplotypes of these tSNPs was also performed. 308 of them wereexposed to heroin related cue. Cue-elicited heroin craving was evaluated by LikertScale. Similarly, we also tested the hypothesis that opioid receptor genes wereassociated with cue-elicited heroin craving. For detection of gene-gene interactions,we used the multifactor dimensionality reduction (MDR) method and a statistic basedon linkage disequilibrium (LD).Results: Allele frequencies of three tSNPs were significantly different between thepositive and negative groups. They were rs696522 (odds ratio [OR]=3.06, p=.0013),rs1381376 (OR=3.16, p=0.0008), and rs3778151 (OR=3.12, p=0.0004). Suchassociation remained after adjustment for demographic covariates and for multipletesting. Multinomial non-conditional logistic analysis revealed that the genotype CTat rs12404612 and the C allele of rs1691808 were associated with heroin-inducedsubjective responses with p value being 0.03 and 0.048, respectively. The associationvanished after considering multiple tests. A strong interaction was detectedOPRMI-OPRD1 and OPRMI-OPRK1 (rs3778151 and rs482387: 0.000101;rs3778151 and rs16918941: 0.000321). The subjects with heroin-induced positiveresponses on first use consumed more drugs than the negative group (Mann-WhitneyU=224.0, Wilcoxon W=16334.0, p≤0.0001).No significance was observed in cue-elicited heroin craving (Kruskal-Wallis test,P＞0.05). Heroin addicts were divided into two groups (high-level and low-level)according to the center of cue-elicited heroin craving. The allele frequency ofrs1799971 differed markedly between two groups. Such association remained afteradjustment for confounding factors and disappeared after multiple test correction.Logistic regression indicated the genotype CT of rs678849 was associated withcue-elicited heroin craving; however, the association vanished after considering multiple tests. No association was found between OPRK1 and cue-elicited heroincraving. A strong interaction was detected between rs508448 and rs16918938 (p=0.000269).Conclusions: Self-reported positive responses on first use of heroin were found to beassociated with OPRM1. The findings suggested that heroin-induced positiveresponses were likely associated with more heroin consumption. No strong evidencewas presented to indicate an association between OPRD1 or OPRK1 andheroin-induced subjective responses. Strong interactions were detected inOPRM1-OPRD1 and OPRMI-OPRK1 in subjective responses on first heroin use. Inaddiction, we found no strong evidence to support an association between singleopioid receptor gene (OPRM1 or OPRD1 or OPRK1) and cue-elicited heroin craving.A strong interaction was observed between OPRD1 and OPRK1 in cue-elicited heroincraving.