The Genetic Basis of Addiction

February 1, 2002
Eric J. Nestler, MD, PhD
Volume 19, Issue 2

While progress in identifying the specific genetic variations that help determine an individual's vulnerability to addiction has been slow, a great deal has been uncovered about delineating the role that numerous genes and their protein products play in mediating the development of addiction.

Genetic factors play a significant role in addiction. Epidemiological studies have long established that alcoholism, for example, is familial, with estimates that 40% to 60% of the risk for this disorder is genetic (Kendler et al., 2000; Reich et al., 1998; Tsuang et al., 1998). Other studies have suggested similar rates of heritability for other drug addictions, such as to opiates and cocaine (Kendler et al., 2000; Kendler and Prescott, 1998; Tsuang et al., 2001). Numerous genetic linkage and association studies are now underway to identify the specific genes that comprise this risk. While investigators have identified several relatively large chromosomal regions as being possibly involved, no specific genetic polymorphism has yet been tied to addiction vulnerability with certainty. The one exception is the genetic defects found in certain East Asian populations in enzymes (e.g., alcohol and aldehyde dehydrogenases) that metabolize alcohol (Chen et al., 1999). These defects dramatically increase side effects of acute alcohol intake, thereby reducing the individual's vulnerability to alcoholism.

It is well-established that inbred strains of mice and rats show robust differences in behavioral and biochemical responses to drugs of abuse (Berrettini et al., 1994; Brodkin et al., 1998; Crabbe et al., 1999; McBride and Li, 1998). In addition, lines of rodents have been selectively bred for alcohol (or other drug) responsiveness. While the genetic variations that occur in animal models may be different from those in humans, identification of genes will provide insight into mechanisms underlying the addiction process. No specific genetic polymorphism has yet been identified in these animal models.

The difficulty in finding genes that contribute to risk for addiction parallels the difficulty in finding genes for other psychiatric disorders and, in fact, for most common diseases. Among the many reasons for this difficulty is the fact that addiction is a complex trait with many genes possibly involved. Thus, any single gene might produce a relatively small effect and would, therefore, be difficult to detect experimentally. It is also possible that variants in different genes may contribute to addiction in each family or rodent model.

Of course, vulnerability to addiction is only partly genetic; nongenetic factors -- which may include stochastic or random events during development or a host of environmental factors throughout life -- are also important (Jacob et al., 2001). Such environmental factors remain only vaguely defined (e.g., poverty, crime, delinquency). In animals, environmental factors such as stress can interact with an animal's genotype to determine its ultimate responses to a drug of abuse. As a result, delineating the mechanisms by which specific genetic variations and specific environmental factors interact is an important focus of investigation.

A gene could contribute to addiction vulnerability in several ways. A mutant protein (or altered levels of a normal protein) could change the structure or functioning of specific brain circuits either during development or in adulthood. These altered brain circuits could change the responsiveness of the individual to initial drug exposure or the adaptations that occur in the brain after repeated drug exposure. Likewise, environmental stimuli could affect addiction vulnerability by influencing these same neural circuits. Perhaps combining genetic approaches with more narrowly defined phenotypes would facilitate the identification of addiction vulnerability genes.

Genetic Dissection of Behavior

In contrast to the difficulty in finding genetic factors that control individual risk for addiction, great strides have been made in demonstrating the role of specific gene products in the complex behavior of addiction as assessed in animal models. The general strategy is to modify the amount of a particular gene product, or in some cases to modify the product itself, and to characterize the consequences of such modifications in behavioral tests. The genetic approaches used most often are constitutive mutations in mice (knockouts and overexpressors); such mice continue to provide important insight into drug mechanisms. In more recent years, mice with inducible and tissue-specific mutations have been used increasingly to overcome some of the limitations of the first-generation mutants. Other genetic approaches include viral-mediated gene transfer, intracerebral infusions of antisense oligonucleotides and mutations in nonmammalian model organisms.

Behavioral Tests for Addiction

Animals with altered levels of a particular gene product in the brain are then subjected to a variety of behavioral tests to assess their responses to drugs of abuse (Nestler, 2000). The most widely used tests are measures of locomotor activity (most drugs of abuse increase activity in rodents when given acutely) and the progressive increase in locomotor activity (locomotor sensitization) that occurs with repeated drug exposure. While the relationship between locomotor responses and drug reward and addiction remains a matter of some debate, locomotor responses are mediated by the mesolimbic dopamine system, which is also implicated in reward and addiction.

A more direct measure of drug reward is conditioned place preference, where an animal learns to prefer an environment that was paired with drug exposure. Conditioned place preference is also mediated partly by the mesolimbic dopamine system and is thought to model some of the powerful conditioning effects of drugs that are seen in humans. Place-conditioning assays, like measures of locomotor activity, are amenable to relatively high throughput, which explains their wide use. However, neither test directly measures the behavioral abnormalities (i.e., compulsive drug-seeking and drug-taking) that are the core features of human addiction.

To get closer to such abnormalities, operant tests must be used. In self-administration tests, animals work (press a lever) to give themselves an intravenous or oral dose of a drug of abuse. The paradigm can also be used to study incentive motivation for drug and relapse after a period of abstinence. In intracranial self-stimulation, animals work to electrically stimulate a particular brain area (e.g., mesolimbic dopamine system). This test is thought to measure the affective state of an animal and the sensitivity of brain reward pathways to drugs of abuse. In conditioned reinforcement, animals work to obtain a previously neutral conditioned stimulus (e.g., light) that has been paired with a natural reinforcer (e.g., water). Drugs of abuse potently stimulate incentive motivation for the conditioned reinforcer. However, these tests generally are far more complicated to perform, particularly in mice. Thus, to date, the tests have been applied to only a small number of genetically altered animals. A major challenge for the field is to devise schemes that make application of these behavioral tests more widely available.

Confirming Initial Drug Targets

One of the most straightforward ways that genetic tools have been used in the addiction field is to confirm the initial protein target for a drug of abuse. While classic pharmacological approaches have revealed many initial drug targets, they often have failed, for example, to identify which of several receptor subtypes is most important. In this way, tests of knockout mice lacking the

opioid receptor, dopamine transporter, CB1 cannabinoid receptor or ß2 nicotinic acetylcholine receptor have confirmed that these are the initial targets that mediate the acute rewarding and other effects of opiates, stimulants, cannabinoids or nicotine, respectively.

Implicating Other Neurotransmitter Systems

Genetic tools are also providing evidence for the numerous neurotransmitters and receptors (and postreceptor signaling pathways) beyond the initial drug target that can modify responses to acute and chronic drug exposure. In some cases, the information obtained has confirmed earlier evidence from pharmacological approaches with receptor agonists and antagonists. In other cases, genetic studies have yielded fundamentally new insight into mechanisms of drug action.

As just one example, mice lacking the serotonin 5-HT1B receptor show enhanced responsiveness to cocaine and alcohol in several behavioral paradigms. Most importantly, the mice self-administer the drugs at higher levels, compared to wildtype controls (Crabbe et al., 1996; Rocha et al., 1998). The mice also show higher levels of

FosB, a Fos-like transcription factor implicated in addiction (see below), under basal conditions. These data suggest particular mechanisms through which serotonergic systems may be involved in addiction.

Nonmammalian model organisms are also proving useful in identifying novel biochemical pathways related to the actions of drugs of abuse. In one recent study, Drosophila melanogaster lacking the clock gene (the master regulator of circadian rhythms) showed reduced locomotor sensitization to cocaine (Andretic et al., 1999). This evidence raises the possibility that circadian genes in mammals may contribute to the mechanisms by which the brain responds to cocaine exposure.

At first it may seem implausible that something as complex as addiction can be modeled in flies; however, it is important to clarify two points. First, the locomotor responses of flies to cocaine (acting via dopamine pathways) are remarkably similar to that seen in mammals, which means that the use of dopaminergic neurons in motor circuits was obligated over 1 billion years ago in evolution. Second, even if flies may not develop the more complex (e.g., cognitive, emotional) aspects of addiction seen in mammals, they certainly can be used to identify the types of genetic and biochemical pathways that are perturbed as nerve cells adapt to a drug of abuse over long periods of time.

Identifying Transcriptional Mechanisms

The stability of the behavioral abnormalities that characterize addiction suggests that drug-induced changes in gene expression may be one important mechanism involved. Since classic pharmacological agonists and antagonists are not yet available for most proteins involved in gene regulation, genetic tools have provided the best approach to explore such processes in addiction.

One such mechanism is related to the Fos-like transcription factor

FosB (Nestler et al., 2001). It accumulates in the nucleus accumbens (a target of the mesolimbic dopamine system) after chronic, but not acute, exposure to any of several drugs of abuse, including opiates, cocaine, amphetamine, alcohol, nicotine and phencyclidine (PCP). This is in contrast to all other Fos-like proteins, which are induced only transiently after acute drug administration. The

FosB protein accumulates because it is highly stable, unlike other Fos-like proteins. Consequently,

FosB persists in the nucleus accumbens long after drug-taking ceases and thereby provides one mechanism by which changes in gene expression (and resulting changes in neural function and behavior) can be relatively stable.

Such a role for

FosB has been confirmed recently in transgenic mice, in which

FosB can be induced selectively within the same subset of nucleus accumbens neuron where it is normally induced by drug administration. Such mice show enhanced responsiveness to cocaine in several animal models, including locomotor activity, place conditioning, self-administration and relapse assays, which suggests that

FosB may function as a relatively sustained molecular switch that contributes to a state of addiction.

One challenge of current research is to identify target genes through which

FosB, as well as other transcription factors, produce their behavioral effects. Two general approaches have been used. One approach considers particular candidate genes that contain putative response elements for the transcription factor in question or whose products are implicated in drug mechanisms within the brain region of interest. The other approach is more open-ended and involves analysis of differential gene expression in certain brain regions (e.g., nucleus accumbens) under control and drug-treated conditions. For example, there is currently a great deal of excitement in the use of DNA array technology to identify genes involved in addiction. In encouraging news, the use of various arrays (filter-, glass- and chip-based) in preliminary studies has led to the identification of thousands of potential drug-regulated genes. The daunting news is that the field needs to learn how to better evaluate this vast amount of new information (beyond evaluating single genes with traditional approaches) to identify which genes truly are drug-regulated and contribute to addiction.

Conclusions

In contrast to the difficulty in identifying genes that underlie individual differences in vulnerability to addiction, genetic tools have been invaluable in increasing our understanding of neurobiological mechanisms involved in the addiction process.

One weakness of the field is that, in some cases, a genetic mutation is shown to result in altered behavioral responses to a drug of abuse, but there is no plausible scheme explaining how the mutation actually causes the abnormal behavior. However, the increasing sophistication of genetic tools and the increasing predictive value of animal models of addiction make it increasingly feasible to fill in the missing pieces and to understand the cellular mechanisms and neural circuitry that ultimately connect molecular events with complex behavior.

Acknowledgement

This essay is based on a recent review (Nestler, 2000).

References:

References


1.

Andretic R, Chaney S, Hirsh J (1999), Requirement of circadian genes for cocaine sensitization in Drosophila. Science 285(5430):1066-1068.

2.

Berrettini WH, Ferraro TN, Alexander RC et al. (1994), Quantitative trait loci mapping of three loci controlling morphine preference using inbred mouse strains. Nat Genet 7(1):54-58 [see comment pp3-4].

3.

Brodkin ES, Carlezon WA Jr, Haile CN et al. (1998), Genetic analysis of behavioral, neuroendocrine, and biochemical parameters in inbred rodents: initial studies in Lewis and Fischer 344 rats and in A/J and C57BL/6J mice. Brain Res 805(1-2):55-68.

4.

Chen YC, Lu RB, Peng GS et al. (1999), Alcohol metabolism and cardiovascular response in an alcoholic patient homozygous for the ALDH2*2 variant gene allele. Alcohol Clin Exp Res 23(12):1853-1860.

5.

Crabbe JC, Phillips TJ, Buck KJ et al. (1999), Identifying genes for alcohol and drug sensitivity: recent progress and future directions. Trends Neurosci 22(4):173-179.

6.

Crabbe JC, Phillips TJ, Feller DJ et al. (1996), Elevated alcohol consumption in null mutant mice lacking 5-HT1B serotonin receptors. Nat Genet 14(1):98-101.

7.

Jacob T, Sher KJ, Bucholz KK et al. (2001), An integrative approach for studying the etiology of alcoholism and other addictions. Twin Res 4(2):103-118.

8.

Kendler KS, Karkowski LM, Neale MC, Prescott CA (2000), Illicit psychoactive substance use, heavy use, abuse, and dependence in a US population-based sample of male twins. Arch Gen Psychiatry 57(3):261-269.

9.

Kendler KS, Prescott CA (1998), Cocaine use, abuse and dependence in a population-based sample of female twins. Br J Psychiatry 173:345-350.

10.

McBride WJ, Li TK (1998), Animal models of alcoholism: neurobiology of high alcohol-drinking behavior in rodents. Crit Rev Neurobiol 12(4):339-369.

11.

Nestler EJ (2000), Genes and addiction. Nat Genet 26(3):277-281 [see comment p263].

12.

Nestler EJ, Barrot M, Self SDW (2001), DFosB: a sustained molecular switch for addiction. Proc Natl Acad Sci U S A 98(20):11042-11046.

13.

Reich T, Edenberg HJ, Goate A et al. (1998), Genome-wide search for genes affecting the risk for alcohol dependence. Am J Med Genet 81(3):207-215.

14.

Rocha BA, Scearce-Levie K, Lucas JJ et al. (1998), Increased vulnerability to cocaine in mice lacking the serotonin-1B receptor. Nature 393(6681):175-178 [see comment pp118-119].

15.

Tsuang MT, Bar JL, Harley RM, Lyons MJ (2001), The Harvard twin study of substance abuse: what we have learned. Harv Rev Psychiatry 9(6):267-279.

16.

Tsuang MT, Lyons MJ, Meyer JM et al. (1998), Co-occurrence of abuse of different drugs in men: the role of drug-specific and shared vulnerabilities. Arch Gen Psychiatry 55(11):967-972 [see comment pp964-965].