Implication of OPRM1 A118G Polymorphism in Opioids Addicts in Pakistan: In vitro and In silico Analysis

Single nucleotide polymorphism in OPRM1 gene is associated with hedonic and reinforcing consequences of opioids. Risk and protective alleles may vary in different populations. One hundred healthy controls and 100 opioids (predominantly heroin) addicts from Pakistani origin were genotyped for A118G (N40D) polymorphism in OPRM1. Structural and functional impact of the polymorphism on encoded protein was predicted by in silico analysis. Results show significant association between homozygous GG genotype and opioid addiction in Pakistani population (p value = 0.016). In silico analysis by SIFT (TI = 0.61), PolyPhen (PISC = 0.227), PANTHER (subPSEC = −1.7171), and SNP effect predicted this SNP benign for encoded protein. Superimposing wild-type and mutated proteins by MODELLER shows no change (RMSD = 0.1) in extracellular ligand binding domain of μ-opioid receptor. However, Haploreg and RegulomeDB predicted OPRM1 gene repression by chromatin condensation and increased binding affinity of RXRA transcription factor that may reduce protein translation and hence the number of available receptors to bind with drugs, which may trigger underlying mechanisms for opioids addiction. Thus, this study outlines causal relationship between opioids addiction and genetic predisposition in Pakistani population.


Introduction
Drug addiction, characterized by repetitive and compulsive use of drugs to attain euphoria (Deroche-Gamonet et al. 2004), predisposes the addict to tolerance and elevate the risk of withdrawal symptoms upon reducing the intake of drug (Volkow and Li 2004). Persistent drug abuse triggers neurological changes leading to psychological and physical dependence, craving, and relapse (Camí and Farré 2003). Deregulation of endogenous opioid and dopamine systems mediate the hedonic and reinforcing consequences of addictive drugs (Camí and Farré 2003). It burdens an individual with the high costs associated with medical treatment, injuries, drug-related complications, crimes, incarceration, and time lost from work and social welfare programs.
Multiple factors including pharmacological and physicochemical properties of drugs, psychiatric discords, riskseeking and novelty-seeking traits, and stressed life and dominantly genetic makeup may provoke an individual to abuse drugs (Crabbe 2002;FARRÉ and CAMÍ 1991;Helmus et al. 2001). A study evidenced higher incidence of alcohol addiction in individuals born to alcohol addict parents despite being adopted and raised by non-addict parents (Schuckit and Smith 2001). Genetic polymorphism in endogenous opioid system is associated with drug abuse (Kosten et al. 1986). Single nucleotide polymorphism (SNP) in OPRM1 gene encoding μopioid receptor is significantly associated with opioid dependence and heterogeneous response to various ligands including β-endorphin, enkephalins, morphine, heroin, methadone, cocaine, and alcohol (Nestler 2001;Uhl et al. 1999). A missense SNP rs1799971 (A118G) located in exon 1 (Asn40Asp) has been extensively studied for its association with substance dependence in Caucasian (Bergen et al. 1997), European American, African American (Crowley et al. 2003), German (Mistry et al. 2014), and Japanese (Gelernter et al. 1999) populations.
A118G polymorphism removes a highly conserved Nglycosylation site in protein's extracellular domain (Bergen et al. 1997) that may hamper pain perception in chronic diseases (Fillingim et al. 2005;Janicki et al. 2006), reduce response towards analgesic drugs (Oertel et al. 2009;Oertel et al. 2006), and tend to increase administration of opioids (Chou et al. 2006;Sia et al. 2008).Another missense SNP C17T in exon 1 (Ala6Val) (Zhang et al. 2007) was significantly linked with heroin addiction (Rommelspacher et al. 2001). Several rare variants such as Ser147Cys in exon 2 and Ile292Val in exon 3 of OPRM1 gene have also been reported but their functional significance is not elucidated (Bergen et al. 1997). Moreover, OPRM1 SNPs in intron 1 are associated with euphoric response to heroin in Chinese population (Zhang et al. 2007) and cocaine and opioid drugs' dependence in European Americans (Zhang et al. 2006). Another study documented combined association of SNPA118G and SNP C1031G in intron 2 of OPRM1 gene with heroin addiction (Szeto et al. 2001).
In Pakistan, drug abusers have raised from 4.1 million to 6.45 million. A recent report documented opioid abuse in 5.8% population aged 15-64 years in past 12 months (Organization 1992). It is imperative to identify genetic predisposition to opioid addiction in Pakistani population. The present study determines the association of drug addiction with SNP A118G in OPRM1 gene in Pakistan. It also predicts implications of single nucleotide polymorphism on encoded protein structure and function, allele-specific transcription factor interactions, and chromatin structure to elucidate molecular mechanism underlying A118G association with drug addiction. The aim of the current study is to identify genetic predisposition to opioids addiction in Pakistani population that may enable us to take pre-emptive and preventive measure to reduce prevalence of opioids addiction in genetically vulnerable individuals. Moreover, it may also enable us to retrieve novel molecular and genetic targets to design safe and effective treatment for opioids addicts.

Material and Methods
Blood samples from 100 healthy individuals (mean age 33.58 years) without history of psychotic disorder, drug abuse, and dependence and 100 opioid addicts (predominantly heroin; mean age 34.35 years) undergoing detoxification therapy at rehabilitation centers of Faisalabad and Lahore, Pakistan, were collected with the consent of test subjects and according to human ethical protocols provided by Declaration of Helsinki (World Medical Association 2001). Drug addicts were selected based on criteria for drug dependence described in diagnostic and statistical manual of mental disorders (American Psychiatric Association 2013). Structured questionnaires were used to gather demographic data, medical history, family history, and socioeconomic status of the individuals.

Collection of Blood Samples
Blood samples were collected in EDTA impregnated vacutainers (3 ml, 13 × 75 mm, Ayset, Turkey) and stored at 4°C until further analysis.

SNP Genotyping
SNP genotyping was done by restriction fragment length polymorphism (RFLP) and confirmed by sequencing. Purified DNA samples from control and experimental groups were digested with restriction enzyme Bsh 12361 (BstU1, #ER0921, 10 units/μl, Fermentas, Lithuania) and analyzed by Agilent Bioanalyzer 2100 (Agilent, USA) according to manufacturer's instructions. Twenty-five samples from each group were randomly selected and sent for sequencing to Eurofins MWG Operon (Huntsville, AL). Sequencing was performed by dideoxy chain termination method (Sanger et al. 1977) using ABI 3730XL sequencer.

Statistical Analysis
Genotype and SNP association with opioid (predominantly heroin) addiction was analyzed by chi square test supplemented by power analysis and determination of odd ratios (OD). Statistical analysis was done using statistical software R version 3.0.1.

In silico Analysis of SNPs
In silico analysis of OPRM1 SNP rs1799971 was done to predict its impact on encoded protein's structure and function. SIFT (http://sift.bii.a-star.edu.sg/) predicts the phenotypic effect of mutation on protein structure, based on sequence homology and physical properties of amino acids (Ng and Henikoff 2003) and calculates tolerance index (TI) that is classified as intolerant (0.00-0.05), potentially intolerant (0. 051-0.10), and tolerant (0.201-1.00). The higher the TI, the lesser the functional impact of mutation. Query was submitted as SNP Id with default settings (3.00 median conservation score, remove sequences > 90% identical to query sequence) and selecting SWISS-PROT and TrEMBL databases. Risk associated with amino acid substitution (AAS) was predicted by PolyPhen (Ramensky 2002). PolyPhen (http://coot.embl. de/PolyPhen/) predicts effect of AAS based on evolutionary conservations, physiochemical differences, and substitution vicinity to structural features of protein. Query was submitted as protein sequence (FASTA) along with substituted amino acids and their position.
SNPeffect (http://snpeffect.switchlab.org/) (Reumers et al. 2006) annotate the variant by algorithms like TANGO (predicts aggregation regions), WALTZ (for amylogenic region prediction), and FoldX (analyzes effect on structure stability) [34]. PANTHER (http://www.pantherdb.org/ tools/csnpScoreForm.jsp) (Reumers et al. 2006) uses hidden Markov model (HMM)-based statistical methods to predict deleterious variants by calculating substitution position-specific evolutionary conservation (subPSEC) score ranging between 0 (neutral) and − 10 (most likely to be deleterious). P deleterious determined the probability of given variant to cause a deleterious effect (Mi et al. 2013). The higher the value of P deleterious score, the severe the impact of a variant on protein function. Protein sequence was used as input for SNP prediction.
To visualize the impact of mutation, A118G SNP was mapped on 3D structure of OPRM1. Since, the crystal structure of human OPRM1 is not determined; therefore, we used 3D structure of Mus musculus OPRM1 for homology-based prediction. The protein template (PDB ID: 4DKL) was selected from basic local alignment search tool (BLAST) with highest sequence identity and smallest distance on the phylogenetic tree. MODELLER (Martí-Renom et al. 2000) was used for prediction of 3D structure of OPRM1. RMSD between the mutant and wild-type protein structure was calculated to check the effect of mutation on stability of protein structure.
In addition, we used RegulomeDB (Boyle et al. 2012) and Haploreg V2 (Ward and Kellis 2012) to determine the effect of OPRM1 SNP on chromatin structure and allele-specific transcription factor binding. HaploregV2 (http://www. broadinstitute.org/mammals/haploreg/haploreg.php) discovers variants present on haplotype blocks and explores their regulatory nature and linkage with disease associated loci (Ward and Kellis 2012). We used European and American population groups to retrieve the SNPs present in LD with our lead SNP. RegulomeDB (http://regulomedb.org) utilized CHIP-seq data and chromatin state information across many cell types as well as expression quantitative trait loci (eQTL) information for functional annotation of variants (Boyle et al. 2012). This data was retrieved by using SNP Ids. RegulomeDB scores the variants based on predicted potential effects caused by the variant residing in a functionally important region of the genome. The lower the score, the higher is the effects on protein binding and expression of target gene (Boyle et al. 2012).

Association of OPRM1 SNP with Opioid Addiction
Genotyping of OPRM1 SNP rs1799971 by RFLP and sequencing methods show significant association of A118G variant with addiction in Pakistani population. Among 35 females and 165 males, homozygous genotype A/A was found with highest frequency in both control and experimental groups while heterozygous genotype A/G was more frequent in control group (Table 1). Moreover, frequency of mutated Gallele was higher in addicts as compared to control group (Table 1) indicating a higher risk for the opioids addiction. The genotypic and allelic frequencies of A118G variant (Figs. 1 and 2) depict significant association of opioid addiction with GG genotype in addicts (Table 2). Genotype distribution in control group and drug dependents was not in Hardy-Weinberg equilibrium (p < 0.05). The +118AA genotype was taken as reference due to its higher frequency in control group. Power analysis of these results is significant for the strength of performed tests (Table 2).

Functional Annotation of rs1799971
In silico analysis of SNP rs1799971 by SIFT and PolyPhen characterized mutation N40D as tolerant and benign (sensitivity, 0.91, specificity = 0.88) respectively to protein structure. Similar effect was predicted by PANTHER. Furthermore, physicochemical properties of mutated protein analyzed by SNPeffect predicted change neither in protein aggregation   (Table 3). FoldX did not provide output due to unavailability of OPRM1 protein 3D structure.
3D structure of OPRM1 protein extracellular domain was predicted by MODELLER using energy score of models and ERRATscore (Fig. 3a). No change was observed in the structural domains of protein (RMSD = 0.1) after superimposing the native and mutated protein models and inducing mutation at corresponding position (N40D) by PyMOL viewer (Fig. 3b).
Furthermore, Regulome DB score 4 shows minimal binding evidence (Table 3) while the chromatin is condensed in different cell types (Table 4) indicating reduced expression of OPRM1 gene in drug addicts by A118G polymorphism. Haploreg V2 shows no other SNP in LD with rs1799971. Moreover, A to G polymorphism creates canonical motif BAGARGGCG^for retinoid X receptor alpha (RXRA), whereas affinity of hypermethylated in cancer 1 (Hic 1) transcription factor is reduced due to loss of its motif BNBBRTGCCAMCCNRHH^ (Table 3). These modifications repress transcription of OPRM1 gene.

Discussion
Drug addiction is a developmental and neurological disorder influenced by genetic, behavioral, and environmental factors (Deroche-Gamonet et al. 2004;Volkow and Li 2004;FARRÉ and CAMÍ 1991;Crabbe 2002;Camí and Farré 2003;Helmus et al. 2001). In present study, we found significant association of SNP rs1799971 (A118G) with opioid addiction in Pakistan. We also predicted that A118G polymorphism does not change encoded protein structure; rather, it may repress OPRM1 transcription and number of available receptors to bind with drugs by chromatin condensation and allele-specific transcription factor binding.
Results of this study reveal that frequency of mutated Gallele in our study is higher than Caucasian, European American, African American, German, and Swedish populations (Crowley et al. 2003;Barr et al. 2001;Gelernter et al. 1999;Bart et al. 2004;Bart et al. 2005;Hernandez-Avila et al. 2003), while it is lower as compared to Japanese, Chinese, Malay, Korean, Taiwanese, and Indian Asian populations  AAS amino acid substitution, NIC number of independent counts, PSIC position-specific independent counts, subPSEC substitution position-specific evolutionary conservation, TI tolerance index, Ref reference allele, Alt alternative allele; (-) indicates data no found (Kim et al. 2004;Li et al. 2000;Loh el et al. 2004;Nagaya et al. 2012;Nishizawa et al. 2006;Tan et al. 2003). This SNP is also associated with reduced effectiveness of morphine as an analgesic (Janicki et al. 2006). Nevertheless, a study has also reported no association of A118G polymorphism with drug addiction (Coller et al. 2009). This indicates that there is variability in implications of OPRM1 gene polymorphism on drug addiction among different populations. Several factors contribute to false positive results in studies; for example, utilization of subjects belonging to mixed populations with craving for various substances (Bond et al. 1998) and small samples size (Nagaya et al. 2012). However, strong power analysis in our study supports significance of relationship between A118G polymorphism and opioids addiction. SNPs also affect pharmacological response of drugs. Studies have documented better response of opioid antagonist naloxone with good hypothalamic-pituitary-adrenal (HPA) axis activity in individual carrying 118G-allele (Hernandez-Avila et al. 2003;Wand et al. 2002). We adopted most commonly used computational tools (Hou and Zhao 2013) for predicting the impact of A118G polymorphism in opioids addicts individuals. A recent review discussed the merits and demerits of these computational tools in detail (Nishizaki and Boyle 2017).
Computational analysis predicted A118G mutation nondamaging to extracellular domain of encoded protein indicating that it may not affect ligand binding affinity for μ-opioid receptor. Our results are in accordance with a previous study (Beyer et al. 2004), which reported unaltered binding affinity of morphine, morphine-6-glucuronide, and β-endorphin for both wild-type and mutated (N40D) μ-opioid receptors and demonstrated reduced expression of mutated receptors. This indicates that A118G mutation in OPRM1 gene declined the expression of μ-opioid receptor in drug addicts decreasing the number of receptors available to interact with drugs. This is proved in our study by in silico analysis that A118G polymorphism causes chromatin condensation ultimately reducing transcription of OPRM1 gene. Moreover, the absence of a ligand RXRA forms heterodimer with retinoic acid receptor alpha (RARA) associating with transcription co-repressor complex that causes chromatin condensation and transcriptional repression (Kastner et al. 1995;Mangelsdorf and Evans 1995). It is previously reported that RXRA-RARA complex repress transcription of multidrug resistanceassociated protein (MRP3) (Chen et al. 2007). This shows that in the presence of opioids, RXR-RAR complex will activate μ-opioid receptor to attain hedonic effects while it modulate drug-seeking behavior and compulsive drug administration by repressing OPRM1 gene in the absence of ligand.
Our study demonstrates significant association of opioid addiction with A118G polymorphism in Pakistani population. This SNP repress OPRM1 transcription by chromatin condensation and altering regulatory motifs in an allele-specific manner. This corresponds to reduced μ-opioid receptor expression decreasing the number of receptors available to interact with drugs in opioid addicts. This study provides significant causal relationship between opioid addiction and genetic predisposition.

Compliance with Ethical Standards
Blood samples from 100 healthy individuals were collected with the consent of test subjects and according to human ethical protocols provided by Declaration of Helsinki.

Conflict of Interest
The authors declare that they have no conflict of interest.
Open Access This article is distributed under the terms of the Creative Comm ons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.