Molecular Genetics and the Treatment of Addiction

Overview

Substance use disorders represent complex phenotypes that result from the intricate interplay of genetic variation, neurobiological mechanisms, psychosocial variables, and environmental variables. To date, one of the least-studied factors has been genetic variation. However, basic research on the human genome is progressing at a rapid pace, and investigations of genetic factors that influence the etiology and treatment of substance use disorders are now much more common. The promise of this research is that it may help scientists optimize the success of treatments by matching specific treatments with individuals who have specific genetic vulnerabilities. The ability to match a specific treatment with an individual who is most likely to benefit from that treatment is especially exciting because, while a number of treatment alternatives exist, the overall effectiveness of these treatments is quite modest and there are currently no objective criteria that can be used to match an individual with the treatment that is most likely to be effective. It is only a matter of time before much of the genetic variation that contributes to the risk of addiction is uncovered and, likewise, only a matter of time before clinicians begin to utilize genetic information to match individuals with the treatment that is safest and most likely to benefit them.

This chapter provides a critical review of the expanding literature with respect to molecular genetics and the treatment of addiction. First, we present a brief overview of key concepts in the genetics of addictions. Second, we provide a more extended review and discussion of pharmacogenetics and pharmacogenomics applied to addiction medicine. Recent studies of genetic differences and responses to pharmacological, and to a lesser degree psychosocial, treatments for addictions will be reviewed for various substances of abuse, including alcohol, nicotine, cocaine, and opiates. Third, we discuss practical and ethical issues in the translation of pharmacogenetics science into clinical practice. Fourth, we outline several future directions for the field of molecular genetics applied to the treatment of addictions. Finally, we present a summary and some concluding remarks.

Genetics and the Treatment of Addictions

Twin and adoption studies have suggested that approximately 50% of the variance in risk for developing alcohol dependence can be explained by genetic factors. Likewise, studies have demonstrated that genetic factors account for a significant portion of variance in drug use, abuse, and dependence. ^a

a References 9, 43, 44, 66, 68, 138, 139.

The progression from initial use to abuse or dependence for substances such as marijuana and cocaine also appears to be largely due to genetic factors.

Just as the etiology of substance use disorders appears to be under moderate genetic control, so does the response to pharmacotherapies. Broadly speaking, evidence for heritability of medication effects in psychiatry dates as far back as 1967, when heritable variation in plasma concentration of the tricyclic antidepressants, desipramine and nortriptyline, were first shown in twin and family studies. In addition, more recent research has documented the heritability of response to typical antipsychotics, including differences in antipsychotic response among ethnic groups. As discussed in detail below, genetic factors also seem to play a role in response to pharmacotherapies, and perhaps psychosocial treatments, for substance use disorders.

After determining that genetic variation plays a substantial role in the etiology of addictive disorders and the response to treatment through family, twin, and adoption studies, the next step in genetics research often consists of identifying specific genetic variations that contribute to the etiology and response to treatment for these disorders. In many ways, research on genetics of addiction has already transitioned from establishing that genetic variables contribute to the variance in a disorder to identifying the specific genetic variables that actually contribute to the disorder.

Currently, there are two basic approaches to the identification of genetic variations that influence substance use disorders and/or treatment outcomes. The first is a hypothesis-driven approach, in which investigators develop a priori hypotheses based on what is known about the genetic variation and the neurobiology of the disorder or the mechanism of action for a specific treatment. For example, one might hypothesize that a specific genetic variation that influences the mu opioid receptor expression might also predict acute responses to alcohol and the effects of a medication (e.g., naltrexone) that targets this receptor.

In many cases, it is more common to work with a gene for which function variations have yet to be identified. In this situation, a variation on the approach described earlier is to hypothesize that a gene is related to a specific aspect of a substance use disorder or the effects of a medication and then use special analytic approaches to probe genetic variation across the entire gene. This approach is commonly known as a haplotype-based approach, and is designed to capture most of the genetic variation across the gene even when the function variations have not been identified. To that end, “tag single nucleotide polymorphisms” are often used, as they allow scientists to capture genetic variation in various loci by genotyping fewer, but informative, markers. More specifically, tag single nucleotide polymorphisms are selected on the basis of patterns of linkage disequilibrium that indicate whether several polymorphisms are highly correlated, in which case, instead of having to genotype all markers, scientists can identify a few that strongly predict genetic variation in a given area or locus. This approach has become increasingly accessible due to the availability of bioinformatics resources, most notably results from the International HAPMAP Project, which have been made publicly available in the user-friendly HAPMAP Project website ( http://www.hapmap.org/ ). The next step after identifying tag single nucleotide polymorphisms for areas of interest is often to build haplotypes, which describe common patterns of DNA sequence variation. In fact, the objective of the HAPMAP Project is to develop a haplotype map of the human genome, which in turn can aid scientists in finding genes affecting health, disease, and responses to medications and environment. A detailed review of haplotype-based techniques in pharmacogenetics is beyond the scope of this chapter and can be found elsewhere.

Finally, a more recent approach is to conduct exploratory genome-wide analyses to identify genetic variation that influences substance use disorders or responses to medications. The genome-wide association study is currently in vogue and represents one of the most cutting-edge approaches in terms of identifying sources of genetic variation that may eventually be used to predict response to treatment. A genome-wide association study utilizes a high-density single nucleotide polymorphism array to generate data on more than one million genetic markers (e.g., using the Illumina 1 M array). This vast array of genetic data can then be analyzed in combination with a set of phenotypes. A number of reviews have been published recently on the advantages, disadvantages, limitations, and recommendations associated with this approach. One obvious problem with this approach is the sheer number of statistical tests and the resulting increase in type I error that may lead to false positives. A corollary is the requirement of strict statistical corrections and the need for massive sample sizes. To date, there have been a number of genome-wide association study reports in the psychiatric genetics literature, including major depressive disorder, bipolar disorder, and schizophrenia, but only one that involves a substance abuse disorder, namely nicotine dependence. Although genome-wide association is the approach du jour and has generated much excitement in the field, it is important to note that genome-wide association studies represent a transition to even more difficult and time-consuming work. Once new genetic variations are identified, models will need to be developed and hypothesis-driven research will be needed to translate the effect of genetic variation uncovered in the genome-wide association studies regarding the effect of the genetic variations from the molecular level, to the cellular level, to the systems level, and to the behavioral level in order to understand the implications of these findings for the etiology, prevention, and treatment of substance use disorders. This translation will likely lead to new findings on as-yet unknown neuronal mechanisms that influence the development of substance use disorders and lead to new targets for pharmacotherapies as well as generate information about which individuals will be most likely to respond to those new pharmacotherapies.

Pharmacogenetics and Pharmacogenomics

Pharmacogenetics is a field of research that seeks to understand individual differences in the metabolism and efficacy of medications. As described by Vogel, pharmacogenetics is the study of heritable differences in the metabolism and activity of exogenous agents, including medications and environmental toxins. Current pharmacogenetics research focuses on identifying genetic factors that account for variability in pharmacotherapy effects, in terms of both pharmacodynamics and efficacy. In recent years, the term pharmacogenomics has been defined as the application of genomics to the study of pharmacogenetics. In brief, the distinction between the two terms refers to the methodological and theoretical approach such that pharmacogenetics investigations are generally hypothesis driven and focus on a few loci at a time. Conversely, pharmacogenomics investigations include the use of high-throughput genotyping and genome-wide association approaches to understanding genetic determinants of pharmacotherapy response. In essence, the objective of pharmacogenomics is the same as pharmacogenetics, which is to elucidate genetic variants that influence the efficacy and safety of pharmacotherapies. For simplicity, we will refer simply to pharmacogenetics in this chapter.

The field of pharmacogenetics has grown rapidly and has benefited greatly from advancements in molecular genetics tools for identifying gene polymorphisms, developments in bioinformatics and functional genomics, and new findings from the human genome project. The foremost goal of this line of research is to optimize pharmacotherapy by identifying genetic factors that predict who is more likely to respond to certain pharmacotherapies and who will not respond, thereby matching individuals to medications on the basis of genetic factors. Genetic factors can account for individual differences in medication toxicity and response in many ways. Genetic polymorphisms may lead to functional differences in medication metabolism and disposition, such as functional differences in enzyme activity or medication transporters. Alternatively, genetic polymorphisms may impact the target of a medication, such as a particular receptor. An example of the first case is a polymorphism of the CYP2D6 gene, which is involved in the availability of specific medication-metabolizing enzymes associated with one’s response to opioid painkillers, such as codeine or morphine. Individuals who are homozygous for the nonfunctional CYP2D6 alleles were found to be resistant to the analgesic effects of opioid painkillers. It is important to note that this response in poor metabolizers may be moderated by ethnicity. Specifically, individuals who carry the CYP2D6∗10 allele, which is most common among Chinese individuals, have reduced metabolism of opioid analgesics, such as codeine or tramadol. Furthermore, treatments such as methadone maintenance therapy can inhibit the metabolism of tramadol to O -desmethyltramadol by interfering with CYP2D6 activity. This has important implications for therapy as those individuals receiving methadone may not experience the same degree of opioid analgesia, as those receiving other treatment, such as buprenorphine maintenance. Thus both ethnic variation in allele types and treatment considerations are important when determining the potential efficacy of opioid analgesics.

On the other hand, genetic polymorphisms involved in a medication’s target may also impact one’s response to pharmacotherapy. For example, polymorphisms of the dopamine D4 receptor gene ( DRD4 ) have been associated with differential response to antipsychotic medications. For example, a recent finding suggests that for the DRD4 exon III repeat polymorphism, the 4R allele is associated with better clozapine response, but other individual single nucleotide polymorphisms (SNPs) in this study showed no association with clozapine response. Similarly, when individual SNPs of DRD4 (rs1800955 and rs4646984) were analyzed, they did not show any associations with antipsychotic drug response. Thus future studies may consider further exploring the efficacy of the 48-bp repeat polymorphism of the 4R allele. There is also a growing literature on the pharmacogenetics of antidepressant medications. Specifically, research has suggested that the functional polymorphism of the serotonin transporter gene located in the 5′ upstream regulatory region consisting of a 44-bp insertion/deletion, which results in a long or short variant, predicts response to various selective serotonin reuptake inhibitors, including fluoxetine, fluvoxamine, and paroxetine. Carriers of the long allele of the serotonin transporter promoter polymorphism have better clinical response to antidepressant medications compared with individuals who are homozygous for the short allele, which results in twofold decreased expression and transport activity of the receptor in vitro. These results suggest that pharmacogenetics may soon inform a more targeted use of antidepressant medications. In addition to medications’ efficacy, pharmacogenetics research has focused on identifying susceptibility loci contributing to adverse effect profiles and medications’ toxicity, thereby enhancing the safety profile of pharmacotherapies. Next we review pharmacogenetic studies in the field of addictions to various substances of abuse.

Nicotine

Currently, there are two non-nicotine pharmacotherapies approved by the FDA for the treatment of nicotine dependence, namely bupropion hydrochloride and varenicline. In addition, there are five FDA-approved nicotine replacement therapies, which vary mostly in terms of their delivery kinetics; these include transdermal patch, gum, lozenge, inhaler, and nasal spray. Several candidate genes have been subjected to pharmacogenetic studies, mostly those of nicotine replacement therapies and bupropion, as described in recent reviews of the pharmacogenetics of smoking cessation. Specifically, pharmacogenetic studies of nicotine dependence have examined genes underlying the metabolism of nicotine, focusing primarily on the cytochrome P450 (CYP)2A6 gene ( CYP2A6 ). This gene codes for the primary enzyme that converts nicotine to cotinine and cotinine to 3-hydroxycotinine. In a study of transdermal patch and nasal spray nicotine replacement therapies, at the same levels of nicotine replacement, carriers of CYP2A6 alleles coding for a slower metabolism were found to have higher plasma nicotine concentrations following 1 week of the nicotine patch than normal metabolizers. Those differences were not seen using the nasal spray, and at 6-month follow-up, slow metabolizers had higher quit rates in the transdermal patch condition, as compared with normal metabolizers. A community sample of treatment-seeking nicotine-dependent patients who received transdermal nicotine therapy were also found to benefit more from the treatment if they were slow metabolizers, whereas another study found that extended release transdermal nicotine therapy (24 weeks) was more effective than standard therapy (8 weeks) in slow metabolizers. Other important considerations may be examining whether interactions between the CYP2A6 gene and other genes predict quit success rates. For example, a study investigating the role of serotonergic systems in treatment response found that slow metabolizer patients with the 5-HTTL allele or HTR2A-1438GG alleles benefited more from nicotine replacement therapy, suggesting reduced levels of synaptic serotonin may be a mechanism by which slow metabolizers may show better success for smoking cessation. However, the results have not been consistent, and a study found that slow metabolizers had higher relapse rates when treated with the nicotine patch, whereas another study reported that nicotine replacement therapy was more effective in fast metabolizers. It is important to note that findings may also depend on the treatments being tested. For example, in one study, bupropion was shown to result in 1.7 times longer abstinence in slow metabolizers than fast metabolizers. More research is needed to identify the reason for discrepancies across studies, but the CYP2A6 gene appears to be a promising target for translating personalized treatment to the clinic.

In addition, nicotinic receptor genes have been subjected to pharmacogenetic investigations. Nicotine binds to nicotinic acetylcholine receptors, which are ligand-gated ion channels for which there are several subunits. Allelic variation in the gene coding for the nicotinic acetylcholine receptor’s α4 subunit (CHRNA4) has been associated with nicotine dependence. More recent molecular work has suggested that certain single nucleotide polymorphisms in the CHRNA4 gene are functional and related to smoking cessation during nicotine replacement therapy and varenicline treatment. Although promising, these findings await replication. Likewise, a series of studies have examined the role of functional genetic variation in the DRD2 in response to bupropion and nicotine replacement therapy. Results revealed that the DRD2–141C Ins/Del genotype was associated with treatment response to bupropion, such that smokers homozygous for the Ins C allele had a more favorable response to treatment compared with those carrying the Del C allele. Conversely, regardless of nicotine replacement therapy type, those carrying the Del C allele had higher quit rates from nicotine replacement therapy compared with those homozygous for the Ins C allele. Additional polymorphisms that have received attention as putative genetic moderators of smoking cessation in response to nicotine replacement therapies and bupropion include dopaminergic genes (e.g., the Val/Met single nucleotide polymorphism of the catechol- O -methyltransferase gene), opioidergic genes (e.g., Asn40Asp single nucleotide polymorphism of the OPRM1 gene), and serotonergic genes (e.g., the serotonin transporter promoter polymorphism).

Perhaps one of the more exciting new developments in the pharmacogenetics of smoking cessation is a series of genome-wide association studies of smoking cessation with bupropion and nicotine replacement therapy. These studies revealed that genetic variants in quit-success were likely to alter cell adhesion, enzymatic, transcriptional, structural, and protein-handling functions. The genes identified through these genome-wide association studies had modest overlap with genes associated with addictions and memory processes. Clearly, as noted previously, there are limitations to the genome-wide association approach, and these results should be interpreted with caution until replicated in an independent sample.