The present study focused on the association between OPRD1 genetic variants and cocaine and opioid addiction in EA and AA populations. In the EA population, a nominally significant association of the synonymous SNP rs2234918 with CA was observed, as well as sex-specific nominal associations for several other OPRD1 SNPs. In the EA population with OA, a nominal sex-specific association of the non-synonymous SNP rs1042114 was identified. Associations of rs581111 with OA and rs2236855 with CA in the AA population were also nominally significant. Further, the intronic SNP, rs678849, was found to be strongly associated with CA and confirmed in an independent set of samples. This is the first study to report an association of OPRD1 with CA in an AA population.
Other studies have analyzed OPRD1
polymorphisms and substance dependence. A large candidate gene study of Australian heroin addicted cases and controls analyzed SNPs in OPRD1
. They found rs2236857 to be associated with heroin addiction (p=2.9×10−4
, OR=1.25; Nelson et al., 2012
). The present study genotyped rs2236855, which is in complete linkage disequilibrium with rs2236857. We observed no association between rs2236855 and OA in the EA population, although rs2236855 was nominally associated with CA in the AA population (allelic p=0.03, OR=1.24). The Australian study also identified an association between a haplotype block of rs2236857 and rs581111 and heroin addiction (Nelson et al., 2012
). Interestingly, rs581111 was found to be nominally associated with CA in EA (genotypic p=0.05, OR=1.16) and OA in AA (allelic p=0.02, OR=1.26) in our study. Zhang et al. genotyped 1063 EA substance addicted cases and 443 controls and found a 6 SNP haplotype to be associated with alcohol, cocaine and opioid addiction (Zhang et al., 2008
). This haplotype contained rs1042114 and rs2234918, both of which displayed nominal significance with opioid and cocaine addiction in our study. A large scale candidate gene study analyzing 1350 variants in 412 cases and 184 controls found 3 intronic SNPs in OPRD1
to be associated with heroin addiction (Levran et al., 2008
). Although the same SNPs were not genotyped in our study, tag SNPs rs1075331 and rs2236855 are in LD with these variants and they were found to be nominally associated with cocaine and opioid addiction, respectively.
The strongest association reported in the present study was between rs678849 and CA in the AA population. The minor allele (T) was overrepresented in the controls compared to cases, indicative of a protective effect. Following the initial association of rs678849 with CA in the Group 1 samples, the association was confirmed in an independent set of samples (Group 2). Two additional SNPs in high LD with rs678849 (D’>0.88) were genotyped to determine the source of the association signal. The results revealed that rs678849 had the strongest association with CA, supporting the surrounding intronic segment as a genomic region of interest in addiction. At this time, we cannot conclude whether rs678849 is driving the association as un-typed variants genetically linked to the SNP may also be responsible for the signal. However, previous studies have implicated rs678849 in addiction susceptibility and treatment efficacy. In a pharmacogenetic study of naltrexone treatment in alcohol addiction, individuals carrying the “T” allele of rs678849 had a significantly lower relapse rate after treatment compared to matched placebo controls (Gelernter et al., 2007
). Another recent study by Luo et al. reported an association between a haplotype including rs678849 and drug addiction in a mixed population of EA and AA cases (Luo et al., 2008
). In conjunction with our findings, these results suggest that future studies designed to test the pharmacogenetic effects of rs678849 in AA CA populations are warranted.
We cannot exclude the possibility that associations in the AA population are due to population stratification. According to data from the International Hapmap Project (www.hapmap.org
), the minor allele of rs678849 (T) has a frequency of 26% in a population with African ancestry from the Southwest USA (ASW). In a population of Utah residents with Northern and Western European ancestry (CEU), however, the ‘T’ allele has a frequency of 53%. Due to this large difference in minor allele frequency between the two populations, different degrees of population admixture between AA cases and controls may have contributed to the association found in the present study. In support of our findings, African genetic heritage was not found to be associated with OA or CA in a previous study (Ducci et al., 2009
) and the significance level of the rs678849 allele remains significant when comparing our control data to a larger independent control group. Further work using ancestry informative markers is needed to determine whether rs678849 is relevant for CA in AA.
The present study, in combination with the current literature, suggests that DOR may be a potential therapeutic target for addiction. OPRD1
knock-out mice demonstrate anxiogenic-like and depressive-like behaviors, suggesting a role for DOR in modulating emotional state (Filliol et al., 2000
; Konig et al., 1996
; Ragnauth et al., 2001
). Decreased DOR signaling resulting from treatment with the DOR antagonist naltrindole also causes anxiogenic-like effects (Marin et al., 2003
). Mediation of anxiety makes DOR an appealing target for alleviation of cocaine withdrawal-induced anxiety. SNC-80, a DOR agonist, produced anxiolytic-like effects in males similar to the effects of classical therapeutics for anxiety (Perrine et al., 2006
) and was able to decrease anxiety-like behavior following withdrawal from chronic cocaine in male (Perrine et al., 2008
) and female rats (Ambrose-Lanci, 2009
). Non-peptide DOR agonists, such as SNC-80, have greater bioavailability than peptide agonists and more readily cross the blood-brain-barrier. These non-peptide agonists may have clinical potential for the treatment of not only anxiety but also depression, pain and other neurological disorders. Despite some potential limitations related to convulsant effects, the use of DOR agonists is encouraging since they are reported to have fewer negative side effects and lower abuse potential compared to MOR agonists (reviewed in Jutkiewicz, 2006
; Pradhan et al., 2011
The present study provides an important step in understanding the genetic influence of OPRD1 on both CA and OA. In the OPRD1 gene, nominally significant associations were reported as well as a highly statistically significant association an intronic SNP (rs678849) and CA in AA. In light of previous findings implicating the DOR system in both emotional state and addiction withdrawal, our data support further research to fully determine the role of DOR in cocaine addiction.