Molecular Genetics and New Medication Strategies for Opioid Addiction

Individual differences in genetics play an important role in a person’s vulnerability to developing opioid use disorder. It is estimated, on the basis of twin studies, that approximately 50% of the variation in opioid addiction is attributable to genetic factors (12, 13). Although an individual’s genetic makeup is not deterministic in the development of a substance use disorder, especially if a person is never exposed to an agent, knowledge of genetic vulnerability can help provide important insight pertaining to the underlying neurobiology of substance use disorders, reveal novel biological targets for potential therapeutic development, and, potentially, optimize personalized medication therapy. The OPRM1 gene on chromosome 6 that encodes the mu opioid receptor has logically been a high-priority candidate in studies investigating disease risk and pharmacogenomic factors associated with opioid use. The locus of the OPRM1 gene that has received the most attention is the common missense single-nucleotide polymorphism (SNP) A118G rs1799971, a nonsynonymous point mutation that changes the amino acid sequence of the protein (14). The OPRM1 variants have been shown to have functional relevance with regard to in vitro mu opioid receptor binding and signaling (15–17), in vivo mu opioid receptor binding (18, 19), mu opioid receptor signaling in human postmortem specimens (20–22), and opioid neuropeptide gene expression levels relevant to addiction in the human brain (23). Most of the findings suggest reduced mu opioid receptors in subjects with the A118G SNP. Other OPRM1 variants have been investigated with regard to heroin addiction (24–27) and the functional relationship to mu opioid receptor signaling and downstream transcriptional regulation (21).

Multiple studies have addressed the relationship of the rs1799971 polymorphism to heroin and opioid abuse (15, 23, 28–30). Not surprisingly, to date, results for OPRM1 from candidate gene studies have been equivocal, due in part to low sample sizes as well as differences in the race and ethnicity of the study subjects or differences in potential phenotype and environmental variables, among other factors. Meta-analyses of opioid use disorder that attempted to increase statistical power by combining the results from multiple investigations have also been inconclusive (31, 32) but suggest a contribution to addiction liability shared across different addictive substances (33). Additionally, there is research implicating the rs1799971 allele in naltrexone response in the treatment of alcohol use disorder (34, 35). Given the multifaceted nature of addiction, it is evident that a single-gene focus is an extremely limited strategy for demonstrating conclusive genetic contributions. Indeed, a large, comprehensive replication study demonstrated that the rs1799971 SNP was only associated with heroin addiction in the presence of another SNP (rs3778150), which was identified as a disease-associated expression quantitative trait locus that influenced OPRM1 expression in the human prefrontal cortex (26). These findings may explain some of the discrepant literature regarding the association between the rs1799971 genotype and heroin and opioid addiction, and they highlight the importance of haplotype strategies for complex disorders such as addiction, in which the combination of alleles that are inherited together has stronger statistical power in associating a genetic link with the phenotype.

An important question for guiding future clinical care is whether documented functional differences of OPRM1 variants could be leveraged to improve the pharmacological response in patients undergoing opioid treatment (e.g., methadone) and to prevent adverse effects, including addiction vulnerability in healthy individuals who are prescribed opioid analgesics. Determining the effective individual dosage for methadone is often clinically challenging, because underdosing can lead to craving and relapse, and high doses can induce euphoria and sedation as well as other side effects. Implementing an agnostic genome-wide association study approach, Smith et al. (36) recently identified one statistically significant region in the genome that was associated with higher daily methadone dosing in African American (but not Caucasian) patients with opioid dependence. Interestingly, the region was on chromosome 6, with the lead SNP rs73568641 localized in the OPRM1 gene. The authors replicated the finding showing the SNP to be associated with increased morphine dosage requirements for pain relief in an independent sample of African American children treated for surgical pain. Significant research is needed to determine whether the rs73568641 SNP has a causal relationship to the expression or function of the mu opioid receptor. Nevertheless, these findings represent a critical step forward, suggesting that OPRM1 genetics could be useful clinically in determining appropriate opioid medication dosages. Results from a meta-analysis (37) and other studies (38, 39) also suggest that the A118G rs1799971 allele variant can influence opioid pain management in individuals carrying the A118G rs1799971 allele who require higher opioid doses than A118A carriers. The fact that OPRM1 may hold promise as a genetic predictor of opioid medication dosage in the setting of addiction treatment and in analgesia could be helpful in identifying drug-naive individuals without dependence who may have a potential risk for addiction when treated with opioid prescription medications. However, large-scale investigations are needed before individual OPRM1 genetics can be incorporated into the clinical formula for setting optimal opioid treatment dosages for opioid use disorder and pain management.

Additionally, it is important to reemphasize that it is unlikely that only the OPRM1 gene will be able to inform and improve clinically relevant treatment on the basis of genetics. Functional genetic variations of other genes, such as those involved with liver metabolic enzyme activity, were recently reported to be associated with the steady-state plasma concentration of methadone enantiomers, which provide a measure of methadone metabolism and are used clinically as an index of treatment response and efficacy of methadone therapy (40, 41). If replicated, such strategies will help to individualize treatment to achieve dosage optimization for patients with opioid use disorder, to reduce and avert the onset of withdrawal symptoms, and to optimize opioid pain management for persons without dependence.

DNA sequence variations and the mechanism of their regulation of gene expression and disease phenotype are complex and not well understood; however, multiple processes have begun to be explored as potential targets for medication development. Alternative splicing of genes is an efficient means of generating variation in protein function and thus has been of growing interest in attempts to personalize and optimize pharmacological therapies. Splicing determines which exons of a gene that code for its amino acid product (e.g., the mu opioid receptor) are used or not used to synthesize the final receptor. As a result, there can be multiple subtypes of the mu receptor, based on differences in splicing. Not surprisingly, the development of novel medications based on molecular genetics has involved consideration of the multiple isoforms of the mu opioid receptor. An array of mu opioid receptor variants is produced by alternative pre-mRNA splicing of the single copy of the OPRM1 gene (42, 43). The extensive alternative splicing of OPRM1 creates at least three structurally distinct classes of splice variants that are conserved from rodent to human, thus improving the possibility for preclinical scientific studies to better inform human investigations. Animal studies have shown, for example, that the different truncated variants at the C-termini generated from 3′-alternative splicing of the OPRM1 gene do not substantially affect morphine analgesia but differentially alter morphine-induced tolerance, physical dependence, and reward behavior (44). Additionally, whereas normal analgesia is maintained for morphine and methadone analgesia in variants within exon 11 of the OPRM1 gene, the analgesic actions of heroin and fentanyl are markedly decreased (45). Thus, developing opioid analgesics that lack the side effects of traditional opioids may be possible by targeting truncated splice variants of the mu opioid receptor (46, 47). Altogether, research efforts to dissociate the desirable analgesic properties of opioids from the undesirable side effects of addiction may be possible. Targeting specific regions of the mu opioid receptor could be an effective therapeutic strategy to reduce the abuse and addiction liability of opioids while maintaining analgesic properties.

The recent selective molecular targeting of the mu opioid receptor through biased agonism, although not a genetic approach, represents a significant advancement in the ability to selectivity target specific downstream signal transduction pathways in the same G-protein-coupled receptor for medication development (48–50). In contrast to the classic categorization of ligands as full, partial, or inverse agonists or antagonists, biased agonism leverages the capability of G-protein–coupled receptors to stabilize receptor conformation to regulate different signaling pathways. Agonists have thus been designed to deliver different physiologic outcomes by biasing a selective downstream signal transduction pathway (such as G-protein signaling, beta-arrestin recruitment, and receptor internalization) mediated by the same receptor. This strategy significantly expands the repertoire for drug discovery for ligands targeting mu opioid receptor signaling to potentially have analgesic properties (such as those recruiting beta-arrestin proteins) while avoiding tolerance or other opioid adverse effects (linked to G-protein signaling) (51, 52). Clearly, the fact that individual variation exists for genes aligned to distinct G-protein–coupled receptor pathways indicates that genetic factors could dictate which individuals might respond to certain biased agonists.

In addition to genetics, susceptibility to opioid addiction is known to be strongly influenced by environmental factors. Thus, epigenetics (biological mechanisms that mediate genetic control of gene expression without a change in DNA sequence) could be of significant importance for understanding individual vulnerability to addiction and response to treatment. The epigenetic mechanisms that turn genes on and off to set the state of gene expression patterns and thus cellular function include methylation of DNA and modifications (e.g., methylation, acetylation, and phosphorylation) of histones. Epigenetics has emerged as an important biological driver of addiction pathology (53–56). To date, most epigenetic studies that are relevant to opioid use disorder have focused on DNA methylation. A number of investigations have reproducibly observed that chronic exposure to opioids (e.g., in patients with chronic pain being treated with opioids, active heroin users, and former heroin users receiving methadone maintenance treatment) induces epigenetic changes in peripheral marks (lymphocyte and blood), including increased methylation of the OPRM1 gene (57, 59–61). The hypermethylation of DNA located in the OPRM1 promoter appears to block the binding of transcription activators such as Sp1, which ultimately leads to silencing of the OPRM1 (62). Reduced mu opioid receptor expression that has been detected in various brain regions in individuals who abuse heroin (21, 63, 64) may relate to their increased opioid requirement. Consistent with this, pain relief in patients with cancer has been shown to correlate with methylation of the OPRM1 promoter, with high-dose opioid use associated with OPRM1 hypermethylation (57). These and other studies suggest that DNA methylation in peripheral blood samples, and thus a potential proxy for CNS mu opioid receptor function, could provide a biomarker for OPRM1 function that could aid in determining dosage. However, it is important to emphasize the cell-specific nature of epigenetic mechanisms where clear DNA methylation differences have, for example, been shown in different neuronal and glial populations in the prefrontal cortex in heroin users (65), and thus what specific CNS function that any alterations of peripheral OPRM1 methylation would predict remains unclear. Furthermore, although the OPRM1 is a rational target for research in guiding future clinical care, the gene list needs to be expanded by gathering genome-wide unbiased data from large-scale clinical studies to more efficiently direct pharmacoepigenetic approaches.

A critical aspect of epigenetics that makes it an intriguing strategic therapeutic target is that the modifications are reversible. Moreover, multiple families of proteins are involved in adding (writers), recognizing (readers), or removing (erasers) epigenetic marks (66, 67). This plethora of proteins provides a diverse system to tweak the tone of gene expression and thus cellular functions and phenotypes relevant to addiction. The importance of epigenetics to opioid use disorder was highlighted in a recent postmortem study of the human striatum in persons who abused heroin (53). Epigenetic disturbances were observed to correlate with alterations of genes relevant to glutamatergic function and synaptic plasticity, impairments of which are well acknowledged as a hallmark of addiction pathology (68, 69). Interestingly, enhanced histone acetylation levels (and specifically acetylation of the histone H3 protein, lysine 27) in the striatum of persons who abused heroin correlated significantly with years of heroin use. It is well known that acetylated lysine residues on histones are specifically recognized and “read” by the bromodomains and extraterminal subfamily of proteins. Bromodomains and extraterminal inhibitors have become a favored strategy, developed as anticancer medications that could provide novel agents to repurpose as potential treatments for opioid use disorder (53). A small molecular bromodomain and extraterminal inhibitor, JQ-1, reduced heroin self-administration and heroin-seeking behavior in a rodent model, thus setting the stage for these inhibitors to be investigated in clinical trials of persons with opioid use disorder. The wide range of epigenetic molecules being developed for many clinical symptoms and diseases opens a treasure trove of compounds that could be examined in relation to epigenetic pathologies in addiction.

Recent attention has focused on “medical marijuana” as a potential nonconventional strategy. Several recent epidemiological studies, although still in their infancy in data collection, suggest that in states with existing medical marijuana laws, there has been a reduction in opioid-related deaths, opioid prescriptions, and opioid-related motor vehicle fatalities (70–74). Many factors—even those unrelated to the pharmacological effects of cannabis on brain function relevant to opioid use—may account for the apparent associations. However, broad use of the term “medical marijuana” (often confused with conventional recreational marijuana) clearly ignores the complex nature of the plant, which contains hundreds of cannabinoids and other entourage compounds that are essential to consider in the development of a clinically useful medication. What is known from a number of preclinical studies is that different cannabinoids can have adverse or beneficial effects on opioid sensitivity. For example, whereas tetrahydrocannabinol, the psychoactive component of cannabis, can enhance reward sensitivity to opioids (75–78), exposure to cannabidiol, a nonrewarding cannabinoid, reduces the reward-facilitating effect of morphine (79) and reduces cue-induced heroin-seeking behavior, even weeks after the last cannabidiol exposure (80). Cannabidiol normalizes glutamatergic impairments induced by heroin self-administration (80). Such findings have set in motion many research studies examining not only opioids but other drugs of abuse in relation to the potential impact of cannabidiol. Moreover, results from pilot clinical studies have suggested replication of animal findings of cannabidiol reducing cue-induced cravings, as well as anxiety, in individuals who are abstinent from heroin use (81). Intriguingly, similar to the rodent model, cannabidiol resulted in a maintained reduction in heroin craving even a week after the last administration. Cannabidiol’s protracted action may be of particular benefit in a successful therapeutic strategy for opioid use disorder, because the protective effects in reducing craving, and thus risk of relapse, could be maintained even if the individual has missed a daily dose. Importantly, cannabidiol lacks any rewarding effects (79, 82–85) and has a wide safety margin (86–88), and thus would not require the restrictive governmental regulations associated with opioid agonist medications that have abuse potential and are diverted to the black market. However, cannabidiol is still currently under the cannabis umbrella of a schedule I drug. As additional clinical trials are conducted, the knowledge gained will, hopefully, help revise the federal regulations so that a full battery of research can be explored to determine the potential of cannabidiol for opioid use disorder treatment. As with all novel strategies, future application of cannabidiol for opioid use disorder should determine what specific aspect of the complex clinical spectrum of the disorder (e.g., craving, acute reward substitution) would most optimally be targeted by this approach.

Although not new, another “outside the box” approach, originally considered nearly 40 years ago, involves the development of antidrug vaccines. The first vaccine was designed to target opioids (89). Vaccines for the treatment of other drugs of abuse, such as nicotine (90–92) and cocaine (93–96), have been tested to a greater extent in human subjects, with mixed results, seeming to depend on individual variability in antibody titer levels. The challenge of raising sufficiently high antibody titers has been recently addressed with a novel strategy to develop a more efficacious heroin conjugate vaccine in combination with specific carrier proteins and adjuvants (97, 98). This antiheroin vaccine approach was recently evaluated in preclinical models (with mice and nonhuman primates) and resulted in a significant 15-fold reduction of heroin operant responding for 8 months in nonhuman primates (99). Future clinical studies will help to determine whether the promise of antiheroin vaccines can indeed achieve their long promise.

We cannot address the current opioid epidemic with old tools, including declarations of an opioid “war” and harsh judicial ramifications, as previously employed over the past century. These approaches failed in the past, and they exacerbated psychosocial pathologies that persist today. Instead, it is essential that the education of prescribing physicians and of the general public about the benefits and dangers of opioids be complemented with knowledge of the rapid development and translation of novel strategies to expand currently available medications. To meet an epidemic, a different mentality needs to be employed in which specific paths are created at the level of the federal and state governments to mobilize the efforts of scientists and clinicians to advance care, prevention, and ultimately treatments. Strategies should span the improvement of current opioid treatments by leveraging genetic and epigenetic factors as well as the development of new therapies, such as medical cannabinoids and innovative medications that could specifically strengthen impaired synaptic plasticity in the management of opioid use disorder. These approaches might also be employed to reduce the transition to addiction in patients without dependence who are treated with opioids for chronic pain. Moreover, updating medical school curricula with information regarding evidence-based treatments for pain and opioid use vulnerability would be beneficial.

What continues to be missing in the development of novel medications, especially in consideration of personalized medicine and the complex nature of addiction disorders, is the structured clinical phenotyping of patients that is critical to integrate with genetic and epigenetic data. Such knowledge can provide a strong biological foundation on which to develop better targeted personalized medication strategies. Nevertheless, irrespective of developing the most effective innovative medication for opioid use disorder, supportive social services must go hand in hand with drug development. There will not be a miracle therapeutic strategy. The science-based future medication approaches discussed above and elsewhere are interesting, but even the most promising strategies will fail to be realized without fast-track transition of preclinical and early-stage phase 1 clinical studies to full clinical trials and an “all hands on board” approach that involves input from patients and their families. There is much to be learned after 175 years that will help transform the clinical toolbox in the coming years.