John C. Crabbe - University of Stirling · 2006-05-29 · John C. Crabbe Portland Alcohol Research...

29 Nov 2001 11:41 AR AR146-16.tex AR146-16.SGM LaTeX2e(2001/05/10)P1: GSR

Annu. Rev. Psychol. 2002. 53:435–62

GENETIC CONTRIBUTIONS TO ADDICTION∗

John C. CrabbePortland Alcohol Research Center, Department of Behavioral Neuroscience,Oregon Health & Science University, and VA Medical Center, Portland, Oregon;e-mail: [email protected]

Key Words substance abuse, animal models, QTL, gene-environment interaction,transgenic

■ Abstract Even the most extreme environmentalists along the nature-nurture con-tinuum in psychology now acknowledge that genes often contribute to individual dif-ferences in behavior. Behavioral traits are complex, reflecting the aggregate effects ofmany genes. These genetic effects are interactive,inter seand with the environmentsin which they are expressed. Human studies of addictive behaviors have clearly impli-cated both environmental and genetic influences. This review selects drug dependenceas a paradigmatic addiction, and further, concentrates on the extensive literature withgenetic animal models. Both traditional studies with inbred strains and selected linesand studies exploiting the new molecularly based technologies of the genomics era arediscussed. Future directions for further contribution of animal models studies to ourunderstanding of the brain dysregulations characteristic of addictions are identified.

CONTENTS

DEFINING ADDICTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436GENETIC APPROACHES TO ADDICTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437

Genetics and Genomics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437Human Genetic Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438Genetic Animal Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440Expression Arrays/Gene Chips. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446

GENES AND THEIR ENVIRONMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447Gene-Environment Interaction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 447Gene-Environment Correlation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449Epistasis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450

TAKE HOME MESSAGES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451Interpretation of Genetic Differences. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451What Specific Contributions Can AnimalGenetics Make to the Addictions?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452

Addiction in the Postgenomics World. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453

∗The US Government has the right to retain a nonexclusive, royalty-free license in and toany copyright covering this paper.

435

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


436 CRABBE

DEFINING ADDICTION

Addiction is a lay term, so everyone assumes he or she knows what it means. How-ever, serious discussions of the basis for the motivational dysregulation of behaviorthat is its core feature must navigate between languages seeking to describe phys-iological/pharmacological sources of influence and those focused on intrapsychicevents whose basis is unspecified. An interesting review of the field that evaluatedthis distinction forthrightly concluded that both orientations contribute to our un-derstanding of addiction, and that both structural frameworks can offer predictivevalue (Davies 1998). This review takes up the recent neurobiological evidence. Inaddition, it focuses on genetic animal models, which have contributed much of ourknowledge of addiction biology.

Which behaviors fall into the category of “addiction” is a matter of debate,but genetic studies have helped in some cases to clarify the issues. For example,whereas all would agree that obesity has a behavioral component, the multiplepotential sources of dysregulation that lead to obesity (e.g., in appetite, overeating,and physiological processing of foods) can support multiple opinions about theessential pathology. Recent animal studies have identified multiple genes influenc-ing each of these component processes. A review of these individual genes’ effectsin mice demonstrates that we still cannot explain all forms of obesity simply withthese genes, but there are specific behaviors and/or physiological events affectedby each (Wahlsten 1999). Armed with this knowledge, we can more readily assessthe extent to which certain cases of obesity may represent addiction.

Most would agree that overuse of alcohol and other abused drugs representclear examples of addictions. Although other addictive behaviors are occasionallydiscussed, this review concentrates on drug addiction for three principal reasons.First, there is a wealth of data on the genetics of drug dependence. Second, manyof the key features of addiction have been modeled successfully in mice and rats.Finally, drugs can be studied from the framework provided by their pharmacology.Drug receptors are localized in the brain, and drug effects are often local to specificbrain regions. Specific tools (e.g., antagonist drugs) are often available. Further-more, the experimenter can arrange access to and delivery of drugs to the subject.These and other specific features of drug dependence, coupled with a populationincidence in US adult males diagnosed as alcoholic and/or drug dependent of morethan 10%, have led to intense study in this area.

To understand the genetics of individual differences in susceptibility to abusedrugs, we need to consider several aspects of drug response to be comprehensive,including (a) susceptibility to an initial challenge with a drug; (b) neuroadaptationthat occurs with chronic drug administration, represented as reduced (tolerance) orincreased (sensitization) sensitivity; (c) dependence, as inferred from the presenceof withdrawal symptoms when the drug is removed; (d) the reinforcing effects ofthe drug, which may be positive or negative, and in humans are often character-ized as craving; and (e) the efficiency of metabolism and elimination of the drug.In addition to these issues, drug studies have wrestled with modeling presumptively

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


GENETICS AND ADDICTION 437

important, related neurobiological factors such as impulsivity/disinhibition/loss of control, and various neurobiological and behavioral aspects of stressresponses.

The neurobiological bases of drug addiction have stimulated several theoreticalapproaches. Most evidence historically has centered on explanations that seekto identify dysregulation in the brain circuits underlying reward more generally,primarily in the dopamine systems of the basal forebrain. The various approachesand neurobiological studies testing these hypotheses are reviewed elsewhere (Wise1996, Self & Nestler 1995, Altman et al. 1996, O’Brien et al. 1998, Robbins &Everitt 1999, Freeman et al. 2001, van Ree et al. 1999, Tzschentke 1998). Geneticstrategies have provided much important evidence (Nestler 2000): Before takingup the genetic evidence, a few terms and concepts are introduced. More descriptioncan also be found in a recent review of behavior genetics (Wahlsten 1999).

GENETIC APPROACHES TO ADDICTION

Genetics and Genomics

Even the most extreme environmentalists along the nature-nurture continuum inpsychology now acknowledge that genes often contribute to individual differencesin behavior. No reasonable person would argue that genes determine behavioraloutcomes, but understanding their influence is important. Each individual pos-sesses two alleles at each gene, one inherited from each parent. When differentalleles at a gene are circulating in a population, the gene is said to be polymorphic,and these different alleles are represented as differences in the base sequence of theDNA coding for the gene product, a protein. So, the first level of genetic variationthat can give rise to individual behavioral differences is due to DNA sequencedifferences.

However, complexity is introduced by several other aspects of gene structureand function. A gene directs production of a protein via the intermediary molecule,RNA. When an RNA is synthesized, or transcribed, many DNA sequences (calledintrons) are ignored, and the RNA (which will be translated into protein) representsonly some of the original genomic DNA, specifically that from the “coding regions”(exons). However, some DNAs lead to multiple splice variants of RNA, hence tomultiple proteins. Furthermore, all cells contain identical DNA sequences, butthe genes are not expressed all the time. Expression of each gene is regulateddifferentially among body tissues and organs, at developmental periods throughoutlife, and even in different regions of the brain. Some genes are not expressed at allin certain tissues.

Finally, behavioral traits are complex. They are rarely affected by only a singlegene. Indeed, they have been characterized as multigenic or polygenic in recogni-tion of the fact that any given gene is likely to contribute only a small influence onphenotypic (behavioral) variance. The complexity is also the result of pleiotropy,the term geneticists use for the impact of a single gene on multiple behavioral traits.

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


438 CRABBE

As discussed below, the ultimate effects of genes may prove to be very differentin different environments.

The recent strides forward in our understanding of genetic influences on behav-ior have been elegantly reviewed (Wahlsten 1999), and some specific implicationsof the genomics revolution discussed (McGuffin et al. 2001). The principles ofgene action sketched above apply equally to human and nonhuman animal studies.However, the methods for performing experiments and statistically analyzing out-comes are radically different. Because of the much greater experimental controlover animal studies, our review concentrates on them. We also follow the lead ofother recent reviews (Wahlsten 1999) and concentrate on the complex interactionsamong genetic and environmental conditions in an attempt to address the problemof addictions from a behavioral genomics perspective, i.e., one that emphasizesthe effects of genes on behavioral functions of the whole organism (Plomin &Crabbe 2000). This perspective may be contrasted with the much more prevalentfunctional genomics and proteomics efforts to identify for each relevant gene theways in which that gene’s protein product affects cellular function.

Human Genetic Studies

TWIN/FAMILY/ADOPTION STUDIES The classical approaches to complex trait ge-netics in psychology have been the examination of co-occurrence of or comorbidityfor the trait in monozygotic vs dizygotic twins, reared together or apart, and inanalogous family studies with other sorts of biological relatives. These studies,particularly when coupled with genetic epidemiological analyses (Merikangas &Swendsen 2001), have provided solid evidence of genetic influence on addictions.Discussions of the older human behavior genetic studies relating to addictions canbe found in an earlier review that considers personality, cognitive, and a broad rangeof psychopathological traits in addition to the addictions (Rose 1995). This reviewalso offers an excellent survey of the complex interpretive issues surrounding fam-ily and twin genetic studies, as well as many other helpful insights into such issuesas shared versus nonshared environmental effects and the chronic misperceptionof genetic data by the popular press.

Recent twin studies have explored such issues as the nature of the environ-mental contributions to alcohol abuse and dependence (Prescott & Kendler 1999;van den Bree et al. 1998a; Johnson et al. 1996b, 1998), caffeine dependence(Kendler & Prescott 1999), and comorbidity of alcohol abuse and depression orother psychiatric disorders (Prescott et al. 2000, Pickens et al. 1995, van den Breeet al. 1998b). An enduring issue in the addiction literature is the potential role of ge-netic influences in the substantial comorbidity for abuse of alcohol and other drugs.Alcoholism and smoking are highly genetically comorbid (Rose 1995, Maddenet al. 1999, Stallings et al. 1999), and alcoholism and drug dependence also sharecommon genetic influences (Rose 1995; Pickens et al. 1991, 1995; Tsuang et al.1996). However, each disorder also reflects independent genetic influences as well(Enoch & Goldman 2001).

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.



An issue related to comorbidity is whether or not certain personality traits arerisk factors for addictions, including substance abuse. This notion has a very longhistory (Marlatt et al. 1988), but has always faced serious criticism (Nathan 1988).A systematic attempt to identify “addictive” personality traits failed to do so (Rozin& Stoess 1993). Nonetheless, the extensive genetic comorbidity among addictivedisorders contributes to the idea’s longevity (Patton et al. 1994, Williams 1996).The basic problem in testing the hypothesis of an addictive personality type is thatthe addictions are diagnosed post hoc, making it difficult to ascribe personalitycharacteristics to the cause or effect column of the ledger. For example, depressivesymptoms in many alcoholics resolve after abstinence (Schuckit et al. 1997).

Another example is the potential relationships among pathological gambling,substance abuse disorders, and personality characteristics. A thoughtful reviewexplored these issues and noted the comorbidity of gambling and drug abuse, butdid not address potential genetic bases for this relationship (Murray 1993). Morerecent studies, including a twin study, have also noted this pattern of comorbidity,a notion supported by theDSM-III-R andDSM-IVcriteria for problem gambling(Slutske et al. 2000). A recent review raised the possibility that genetic studies mayin fact help to elucidate the nature of excessive gambling behavior and discussed theimplications of characterizing problem gambling as an addiction (Shaffer 1999).All investigators agree that it shares with the addictive disorders the feature ofdiagnostic heterogeneity.

Finally, one set of genetic findings has proven to be especially intriguing. Thoseindividuals with a positive family history for alcoholism have been known todisplay reduced sensitivity to certain of alcohol’s acute effects, such as body swayand subjective intoxication, compared with those who are family-history negative.A cohort of young men was ascertained in the late 1970s and followed prospectivelyas they entered the age of maximum risk for alcoholism. Family history wasexpected to predict later risk of alcohol dependence and alcoholism, consistentwith a genetic contribution. Family history was indeed found to be predictive ofalcoholism in a follow-up 8 years later, but level of initial response to drug challengewas even more highly associated with later alcoholism, and initial responsiveness(which itself is likely to reflect genetic influences) appeared to account for most ofthe variability in susceptibility to alcoholism (Schuckit & Smith 1996, Schuckit1999). Other biological correlates of genetic risk for alcoholism have been recentlyreviewed (Begleiter & Porjesz 1999, Schuckit 2000).

Twin and family studies will continue to contribute to our understanding of thegenetic etiology of addictive behavior. However, such studies cannot easily provideevidence of either the number, location, or identity of the responsible genes unlesscoupled with molecular strategies described in the next section.

MOLECULAR GENE-FINDING METHODS: ASSOCIATION AND LINKAGE STUDIES Oneapproach to understanding genetic influences on addictions is to relate geneticmarkers (i.e., specific sequences of DNA in the genome) to the phenotype acrossindividuals. A cogent introduction to these methods for the nonspecialist can be

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


440 CRABBE

found elsewhere (Gelernter 1999). Studies of this sort have provided a steadystream of reported localizations for candidate genes or for markers at particulargenomic loci (see for example Foroud et al. 1998, 2000; Loh et al. 2000). Manyalcoholism and substance abuse–related loci have been discussed elsewhere (Reichet al. 1999, Chen et al. 1999). The low sensitivity to alcohol’s effects discussedabove has recently been mapped to specific markers as well (Schuckit et al. 2001).One interesting recent study suggests that impulsive-aggressive alcoholics wereidentified by a marker near the serotonin 5-HT1B receptor gene in two populations(Lappalainen et al. 1998). Much neurobiological evidence links serotonin dys-function, alcoholism, and aggression (Le Marquand et al. 1994a,b). Another studyfound evidence for quantitative trait loci associations common to both alcoholconsumption and smoking (Bergen et al. 1999).

Although association and linkage methods can sometimes provide strong ev-idence that there must be a gene near the linked marker that affects the trait, themethods are inherently limited by relatively weak effects of specific genes. Moreimportantly, false positive associations are common, largely because populationsthat appear genetically rather uniform may in fact show a great deal of variationin gene frequencies for genes in the associated region (Gelernter 1999). It is oftendifficult to avoid situations in which the control and addicted groups are in factdrawn from two genetically distinct populations, a condition called stratification.The transmission disequilibrium test and variants thereof can mitigate this dif-ficulty somewhat. These methods compare marker frequencies among relativeswithin family groups who share the trait with those who do not. Because all familymembers are by definition drawn from the same population genetic stratum, dif-ferences in some markers but not others are less likely to represent false-positiveassociations (see Long & Langley 1999, Uhl 1999 for discussion).

In the end, even very strong association and linkage data for markers near thegene for Huntington’s disease required many years of additional work before thegene itself was isolated (MacDonald et al. 1993), and this is a single-gene, virtuallyall-or-none disorder. Final proof that a candidate association has truly captured therelevant gene or genes requires a collection of converging evidence drawn from awide range of genetic and nongenetic techniques (Belknap et al. 2001).

Genetic Animal Models

Compared to genetic studies of other areas of psychopathology, it is in the area ofgenetic animal model development and utilization that genetic studies of alcoholand substance addictions are the most advanced. This is due to two factors. First,serious attempts to study voluntary alcohol drinking in rats began in the late 1940sand were followed by studies in mice in the 1960s. These studies had an explicitgenetic orientation that has been sustained ever since. The second factor was raisedat the beginning of this chapter: Drugs can be administered systematically by theexperimenter or the experimental animal, and pharmacology provides a theoreticalframework within which attempts to understand their effects can be organized.

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.



INBRED STRAINS The simplest genetic animal model system is the study of exist-ing genetic variation. In 1959 McClearn & Rodgers studied several inbred strainsof mice by offering them a choice between a bottle filled with tap water and onecontaining alcohol. Same-sex members of an inbred strain are essentially geneticclones owing to many generations of brother-sister matings, which reduces allelicvariation at each gene until it eventually disappears entirely. They found that thedifferences among strains in preference for ethanol far exceeded the within-straindifferences (McClearn & Rodgers 1959). This demonstrates a significant geneticcontribution to the trait, because differences among strains assessed in as invariantan environment as possible can only arise from the underlying genetic differ-ences. In particular, C57BL/6-strain mice were high preferrers, whereas DBA/2-strain mice were nearly complete abstainers, and other strains showed intermediatepreference.

This pioneering study has been followed by dozens of other studies comparingstrains for alcohol and drug sensitivity, tolerance, dependence/withdrawal severity,and propensity to self-administer drugs (Crawley et al. 1997, Marks et al. 1989,Stitzel et al. 2000, Seale et al. 1984; for reviews, see Crabbe & Harris 1991, Mogilet al. 1996). A major advantage of the inbred strain work is that the genotypesremain stable over time: Studies of C57BL/6 and DBA/2 mice performed in the1990s have been compared directly with those from the 1960s, and the result hasbeen a rich accumulation of knowledge about a few strains of mice. A disad-vantage, however, is that the specific genes responsible for the strain differencesare anonymous. Nonetheless, comparisons among characteristic strain mean re-sponses through correlational analysis have taught us much about codeterminationof genetic influence. For example, mouse strains that are high alcohol preferrerstend to be those that show minimal withdrawal severity when the drug is removed.The genetic correlation between these responses has been estimated to be as highas r= −0.65 across 15 inbred strains (Metten et al. 1998). A similar analysis ofresults from 13 inbred strains of mice demonstrated that efficiency of response in-hibition assessed in a signaled nosepoke task was highly predictive of low ethanolconsumption (Logue et al. 1998). Together, these studies suggest that strains ge-netically predisposed to experience severe withdrawal and to be able to inhibitresponding are those who elect not to self-administer alcohol when it is offered.

There are more than 100 inbred mouse strains available. A new initiative calledthe Mouse Phenome Project has been undertaken to support systematic collectionof behavioral and physiological data in a number (up to 40) of inbred strainsand is assembling a relational database for centralizing access to such geneticrelationships (Paigen & Eppig 2000). There are also many rat inbred strains, butthey are in general less systematically characterized for traits related to addiction.

SELECTED LINES The oldest technique in behavioral genetics is that of artificialselection. By arranging matings such that extreme responders are mated, linesof mice or rats have been selected to differ genetically in sensitivity, tolerance,dependence, and preference for alcohol and several other drugs of abuse. The

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


442 CRABBE

principles were derived from agricultural genetics, in which crops and animals arebred for favorable traits. When such selected lines are compared for other traits,they are often found to differ. Ideally, this is because of the pleiotropic effects of thegenes underlying the selected trait, but care must be taken to insure that the changein genes caused by the limited population sizes one can actually maintain are notthe cause of the differences in correlated responses to selection. Such studies havebeen of immeasurable value in advancing our understanding of the neurobiologicalbasis for individual differences in drug responses.

The first selected lines relevant to addictions were developed in Chile in thelate 1940s, where UChB rats were bred for high and UChA rats for low alcoholpreference (Mardones & Segovia-Riquelme 1983). Mardones’ studies, and theidentification of the propensity of C57BL/6 inbred mice to prefer drinking alcoholsolutions mentioned above (McClearn & Rodgers 1959), stimulated the first mod-ern systematic studies of genetic determinants of alcohol and drug responsiveness.

Studies with these selected lines have been reviewed (Crabbe & Li 1995,Eriksson 1972, Crabbe et al. 1994, Li et al. 1994). Most lines have been selected forresponses to alcohol, but selection has also been applied for opioid drugs (Mogilet al. 1995, Belknap et al. 1983), nicotine (Schechter et al. 1995), and cocaine(Marley et al. 1998), among other drugs (for reviews, see chapters in Crabbe &Harris 1991, Mogil et al. 1995, Mohammed 2000).

Perhaps the best known of these selected lines are the Preferring and Nonpre-ferring lines of rats and two additional pairs of lines subsequently derived for thesame alcohol preference trait, High Alcohol-Drinking and Low Alcohol-Drinkingrats. Many correlated responses have emerged that differentiate these animals, andonly a few highlights are summarized here. Under some conditions, Preferringand High Alcohol-Drinking rats self-administer enough alcohol to achieve bloodalcohol concentration levels of 200 mg% or greater (Murphy et al. 1986), but moregenerally, Preferring rats appear to drink for the pharmacological effects and willstop self-administration when blood levels reach 50–70 mg% (Waller et al. 1982a).These levels correspond to 0.05 and 0.07%—most US states now outlaw drivingat either the 0.05 or 0.08% level. Preferring rats develop metabolic and neuronaltolerance (Lumeng & Li 1986) and dependence (Waller et al. 1982b) with chronicfree-choice ethanol drinking.

Another widely used set of selected lines was bred for the severity of withdrawalsymptoms when chronic alcohol exposure was discontinued. Alcohol withdrawalconvulsions have been reported to occur in all animal species, including humans.Duplicate lines of mice Withdrawal Seizure-Prone or -Resistant to alcohol with-drawal convulsions following a period of chronic alcohol vapor inhalation weredeveloped (Crabbe et al. 1985). These lines did not differ in sensitivity to severaleffects of ethanol, or in the magnitude of tolerance development, but WithdrawalSeizure-Prone and -Resistant mice differed in withdrawal severity from diazepam,phenobarbital, and other sedative-hypnotic drugs (Belknap et al. 1987, 1988, 1989).Furthermore, Withdrawal Seizure-Resistant mice were found to drink more alco-hol than Withdrawal Seizure-Prone mice in a preference test (Kosobud et al. 1988),

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.



consistent with the negative genetic relationship seen in inbred strains, discussedearlier (Metten et al. 1998). They also differed in sensitivity to ethanol place andtaste conditioning (Chester et al. 1998). Many other differences between theselines have been reviewed (Crabbe 1996). Although the results of these lines arenot obviously consistent with the low response in subjects at risk for alcoholismreported by Schuckit (2000), they suggest that this set of selected lines might offera model for a genetic propensity to polydrug abuse.

Selection remains a powerful tool in the arsenal of the genetic animal modeler.In the case of drug dependence, many of the contributing factors obvious from apharmacological perspective have been modeled in selected lines. A particularlyuseful future strategy might be to breed selectively for and against expression ofother traits with presumptive relevance for the intrapsychic effects of drugs, and/orfor those personality and behavioral traits found to be comorbid with addictionrisk in human populations, such as impulsivity, antisocial behavior, and depression.Highly impulsive mice could then be compared with a low-impulsivity line to seewhether they also displayed differences in abusive drug self-administration. Suchtraits are, of course, intrinsically more difficult to model convincingly in rodents(Altman et al. 1996), but some attempts have been made, particularly for depression(Weiss et al. 1998, Overstreet 1993) and for anxiety, assessed in an elevated plusmaze (Liebsch et al. 1998a,b). Rats bred for High-Anxiety-Related Behavior havebeen found to drink less alcohol than those bred for Low-Anxiety-Related Behavior(Henniger et al., submitted).

QUANTITATIVE TRAIT LOCUS MAPPING Recent advances in molecular biologyhave led to an unprecedented explosion in knowledge about the physical aspects ofour genes. This has allowed neuroscientists for the first time to begin to translatestatistical statements about genetic risk into knowledge of the specific regions onspecific chromosomes where genes of importance have been localized. This is ahuge first step toward the ultimate identification of those genes and ascertainmentof their function. The implications of these technologies for psychology have beendiscussed in more detail elsewhere (Wahlsten 1999, Plomin & Crabbe 2000).

Studies with genetic animal models in multiple laboratories have established adense genetic map of distinct DNA sequences scattered throughout mouse and ratchromosomes. Because of our evolutionarily shared ancestor, humans and miceshare approximately 80% of these sequence juxtapositions. Practically, this meansthat when a specific gene’s location has been identified in mice, the location of thehomologous gene is known in humans 80% or more of the time. During the past10 years many studies have demonstrated that individual differences in geneticresponse to drugs of abuse can be reliably associated with particular regions of thegenome [termed “quantitative trait loci” (QTLs)]. Whenever the degree of drugsensitivity is reliably associated with a QTL, this implies that a specific allelicform of a specific gene or genes in that region leads to altered drug sensitivity.

There are now more than 30 QTLs mapped for drug response traits in mice, usinginbred strains, their F2 and backcross generations, recombinant inbred strains,

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


444 CRABBE

selected lines, and congenic strains (Crabbe et al. 1999a). The responses mappedrange from drug sensitivity (Deitrich et al. 2000, Gehle & Erwin 2000, Jones et al.1999), tolerance development (Gehle & Erwin 2000, Deitrich et al. 2000), with-drawal severity (Buck et al. 1997, 1999), reinforcement (Risinger & Cunningham1998), and tendency toward self-administration. QTL studies are beginning to inte-grate behavioral and neurobiological analyses. For example, QTLs for the severityof withdrawal from alcohol and pentobarbital have been identified in largely thesame regions of mouse chromosomes (Buck et al. 1997, 1999), and one region in-cludes genes that code for several of the subunits of the GABAA receptor. Furtherstudies show significant association between one variant form of the GABAAγ2subunit gene and alcohol withdrawal severity across a panel of inbred strains (Hood& Buck 2000). Although this does not prove that theγ2 subunit gene is actuallyresponsible for the original QTL association, it is promising that this gene can-not be excluded and supports further efforts to test this candidate gene. It is alsoencouraging that some human QTL studies with alcohol-dependent subjects havefound evidence for an association with this cluster of GABAA receptor subunitgenes (Sander et al. 1999; Loh et al. 1999, 2000; Iwata et al. 2000).

The largest group of studies has mapped genes related to alcohol preferencedrinking in mice. These studies have been reviewed elsewhere (Crabbe et al. 1999a,Phillips et al. 1998a), but the genetic locations identified have been very similaracross laboratories, despite the use of different specific tests of alcohol preferenceand different mapping populations. Newer studies have also found similar maplocations (Vadasz et al. 2000, Whatley et al. 1999). These studies have suggestedseveral candidate genes, some of which have been tested (see next section).

Many studies are now accumulating that do not target addiction-related re-sponses directly but may be of some relevance owing to the comorbidity of anxiety-related disorders and addictions. QTLs for activity in an open field (Gershenfeldet al. 1997; Gershenfeld & Paul 1997, 1998; Turri et al. 1999) and for contextualfear conditioning (Caldarone et al. 1997, Wehner et al. 1997) have been mapped inmice. An interesting, related project in rats has explored the use of factor analysesof data from several behavioral tasks including open field activity to derive factorsthought to reflect anxiety. These studies have used multiple rat inbred strains aswell as crosses. QTLs for these factors have then been identified. The studies haveshown relationships between the serotonin system, stress-related responses, andanxiety-like behavior, reflected as overlapping QTL regions and as co-contributorsto specific factors (Castanon et al. 1995; Courvoisier et al. 1996; Kulikov et al.1995; Moisan et al. 1996; Ramos et al. 1997, 1998, 1999; Ramos & Mormède1998). Relevance to addictions was addressed in a study that factor analyzed 13behavioral variables in a number of rat lines selected for high or low alcoholconsumption. Although not all variables usually taken to reflect anxiety were cor-related across lines, for some variables there was a clear negative association withalcohol preference. That is, high-preferring genotypes loaded lower on factors re-flecting anxiety than low-preferring genotypes (Overstreet et al. 1997). And, asnoted above, selection for high anxiety was related to high alcohol preference(Henniger et al., submitted).

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.



The current problem is refining these genetic maps to the point that only a fewgenes are contained in the QTL confidence interval. No group has yet successfullyidentified a gene through QTL mapping for an addictive drug response, but this hasbeen achieved for some other traits. For example, one QTL for hypertension in ratswas subsequently demonstrated to be the gene encoding angiotensin-convertingenzyme, known to be important in regulating blood pressure (Jacob et al. 1991).Several groups around the world are using these methods to close in on genes im-portant for addictions. Fehr et al. (submitted) have used specially bred, congenicstrains to reduce the confidence interval surrounding a barbiturate withdrawalQTL to a region containing fewer than 20 genes. In parallel, human associationand linkage studies are seeking analogous statistical evidence for QTLs relatedto alcoholism, depression, substance abuse, and other related traits (Foroud et al.1998), although these studies are intrinsically much more difficult in human pop-ulations for a number of technical reasons (see Gelernter 1999).

TARGETED MUTAGENESIS Sometimes a QTL confidence interval contains a genewhose function appears to be highly relevant to the trait being mapped, as in theexample of the GABAA receptor subunit genes and drug withdrawal given above.Either in pursuit of the genes responsible for QTLs or because a particular geneproduct is implicated in an addictive behavior based on other neurobiological evi-dence, investigators have turned to the study of targeted mutants to explore the roleof particular proteins. A specific gene can now be inserted into a mouse’s germ line,and an over or underexpression transgenic animal studied. Alternatively, the genecan be disrupted or deleted entirely, creating a null mutant or knockout. Many suchmutant mice have been shown to display altered drug responses (for review, seeBuck et al. 2000). For example, considering only alcohol preference drinking, thegenes thus far effectively targeted, or implicated by mapping strategies, include thegene for the serotonin-1B receptor subtype (Crabbe et al. 1996; but see Phillipset al. 1999, Crabbe et al. 1999b), the dopamine D1 and D2 receptor subtypes(El Ghundi et al. 1998, Phillips et al. 1998b), the neuropeptide Y2 receptor gene(Thiele et al. 1998), the protein kinase A gene (Thiele et al. 2000), the proteinkinase C epsilon gene (Hodge et al. 1999), theβ-endorphin gene (Grisel et al.1999), and others. In each of these cases, a significant difference was reported inalcohol preference drinking between the null mutant and its control. However, inmuch the same way that lesioning a brain area and finding a subsequent differ-ence in behavior does not identify that brain area unequivocally as the biologicalsource of the behavior, results from null mutants must be interpreted cautiously. Aprimary source of caution is the fact that such mutants experience their entire devel-opmental course lacking the deleted gene product, and the highly plastic brain hassought to compensate in unpredictable ways for whatever functions were disrupted(see Wehner & Bowers 1995, Uhl 1999, Gerlai 1996).

RANDOM MUTAGENESIS Mutations can also be induced at random throughout thegenome through X-irradiation of mice (an older technology) or through treatmentwith a mutagenic chemical such as N-ethyl-N-nitrosourea. A number of large-scale

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


446 CRABBE

projects have recently begun in which thousands of mice are mutated and their off-spring screened for behavioral and neurobiological abnormalities (Nolan et al.1997, 2000; de Angelis et al. 2000). If a phenotypically deviant mouse is thenbred, and proves fertile, its offspring should also carry the mutated gene, whichcan then be mapped and identified using the methods alluded to in the QTL map-ping section. These methods are too new for it to be known whether they will beefficacious in identifying relevant mutated genes for complex traits, but they arecurrently highly touted (Nadeau & Frankel 2000). A more balanced assessment oftheir prognosis suggests that it may be difficult to apply them to the case of complextraits, in which multiple, small gene effects are the rule (Belknap et al. 2001).

Expression Arrays/Gene Chips

Current technology has provided the genetic research community with the powerto ask which genes are more or less active in directing synthesis of their proteinproducts (Watson & Akil 1999). With the recent near-completion of the mapcontaining all human genes by the Human Genome Project (Lander et al. 2001,Venter et al. 2001) have come the current generation of gene chips. Using oneof several technologies, snippets representing many thousands of individual genesequences have been bonded to tiny chips (e.g., glass plates). When a sampleof DNA is applied, those genes actively expressed in the sample bind to theirembedded ligand, and the resulting interaction is visualized. At least 6000 mousebrain DNA probes are available on chips, and the first studies are beginning toidentify genes differentially expressed in brain tissue from alcoholics vs controls(Lewohl et al. 2000) and in adrenal tissue from mice acutely withdrawing fromethanol (Thibault et al. 2000). (For review, see Reilly et al. 2001.)

Another study showed a specific pattern of gene-expression changes in nucleusaccumbens tissue from primates exposed to cocaine for over a year, includingprotein kinases and other cell regulatory genes (Freeman et al. 2001). A recentreview summarizes the roles that changes in gene expression are likely to play inthe addictive process (Nestler 2000). One goal driving a great deal of the interestin gene expression profiling work is the hope that new genes will be identifiedthat will lead to the development of novel drugs useful in therapy (Hefti 2001).Drugs could also be tailored to maximize an individual’s response by using specificknowledge of an individual’s genotype.

Use of these techniques will increase exponentially for the next several years,and they bring a new challenge—that of making sense of the data. A typicalgene-chip expression array analysis identifies dozens of genes whose expressionis increased or decreased as a function of the diagnostic or treatment group com-pared with controls. Occasionally, expression is drastically altered [e.g., exposureof cells to alcohol chronically led to a 20-fold increase in expression of dopaminebeta hydroxylase (Thibault et al. 2000)]. Much more common, though, is the find-ing of numerous genes whose expression is changed about 100–200%. Numerousstatistical problems attend the analysis of these studies, including detecting which

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.



are true gene expression differences and which are false positive changes, detectingchanges in genes with intrinsically low expression levels, and foremost, determin-ing what the pattern of changes in expression of all those diverse genes means tothe organism’s function (i.e., the behavioral genomics issue).

The current state of the art for such analyses classifies genes according to broadly(and ill-defined) functions such as “cellular metabolism” or “cellular signaling.”The categories reveal the intrinsic orientation toward a reductionist and proteomicsperspective: They are oriented toward explaining the functions of the protein in theimmediate environment in which it is synthesized and acts. A behavioral genomicsperspective will be useful here, where the pattern of gene expression is related notonly to cellular function per se but also to the behavioral functions in the wholeorganism that attended the original treatment or diagnostic comparison.

GENES AND THEIR ENVIRONMENTS

The existing genetic animal models have taught us a great deal about the basicpharmacology, physiology, and biochemistry of drugs’ effects on the nervous sys-tem. They have unequivocally proven that a substantial proportion of individualdifferences in response to or avidity for drugs of abuse is genetically influenced.Recent studies have begun to isolate the genes responsible for such individualdifferences. In addition to (and in some cases building on) these gene mappingefforts, candidate gene approaches have also implicated many specific genes asimportant for drug responses. Three sources of complexity beyond that introducedby pleiotropy need to be considered.

Gene-Environment Interaction

By definition, the behavior of individuals with particular genotypes can only beassessed in an environment, and systematic changes in the environment clearlyaffect the behavioral outcome. Abundant evidence reveals that different geno-types respond differentially to environmental manipulation. This is termed “gene-environment interaction,” and is crucial to the reasonable interpretation of therange and limits of genetic influences on behavior. Ninety-eight recombinantinbred Drosophila strains were reared in three environmental conditions, stan-dard medium at two temperatures, and one medium-temperature combination withethanol added. QTL effects on fitness (reproductive success) could be estimatedfrom the strain means, and the effects of the QTLs depended on the rearing medium(Fry et al. 1998). In another example, inbred strains of mice showed differing de-grees of willingness to drink an offered alcohol solution (McClearn & Rodgers1959). When increasing concentrations of alcohol were offered, some strains(e.g., C57BL/6) continued to self-administer the drug to water, whereas others(e.g., A/HeJ) began to reject it at higher concentrations (Rodgers & McClearn1962, Belknap et al. 1993). Genotypic differences in alcohol self-administrationwere also shown to differ in different cage types and according to how food was

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


448 CRABBE

presented to the rat strains (Adams et al. 2000). Animal-model research in addic-tions has provided many such examples.

The utility of animal models for exploring gene-environment interaction isespecially high because both genotype and environment can be manipulated ex-perimentally. However, even under this very high level of control the specificenvironmental variables that are potent in differentially affecting genotypes can behard to identify. A recent study asked whether inbred strain differences in behaviorwould be the same in three different laboratories when as many environmental andgenetic variables as possible were rigorously equated (Crabbe et al. 1999b). Micefrom several inbred strains and one null mutant were tested at exactly the sameages on exactly the same days on a battery of six behaviors. The animal husbandrywas nearly identical (same laboratory chow, cages changed on the same days,etc.). The apparatus (e.g., elevated plus mazes, water mazes) and test protocols(including how animals were handled) were nearly identical. However, there weresome variables that it was impractical to standardize completely, such as the localwater and air in the animal facility, and the individual experimenters in the threelocations.

As expected, the strains differed a great deal in all behaviors, and for sometests (such as alcohol preference drinking) there were no strong indications ofdifferences in the strain pattern of alcohol preference across sites. However, forsome tests (such as the tendency of mice to venture onto the open arms of anelevated plus maze, generally taken as an index of anxiety), there were significantstrain X laboratory interactions. In general, the weaker the overall genetic influenceon a trait (i.e., the lower the heritability), the more likely there was to be a genotypeX–environment interaction.

Do the results of this study imply that behavioral tests in laboratory mice areintrinsically unreliable? Reliability was generally high at each site. The alternativeinterpretation is that even differences in environmental test situations that are notobvious to the experimenter may have great importance for the animal and mayaffect animals with different genotypes differently (Crabbe et al. 1999b).

One implication of this finding is that the particular behavioral test employedto assay a given behavioral domain may affect the interpretation of the results.This is potentially a large problem, particularly for experiments characterizingnull mutants. For example, an early study with a null mutant lacking one of themultiple variants (5-HT1B) of receptors for the neurotransmitter serotonin foundthat the knockouts were much less susceptible to the intoxicating effects of anacute alcohol injection than were the wild-type control mice in a test called thegrid test (Crabbe et al. 1996). However, a subsequent study tested these animals foralcohol’s incoordinating effects using several other tasks, including frequently usedtasks such as the rotarod. Null mutants were found to be less sensitive to alcoholon some, but not all, of these tasks. This implies that these different behaviorsmust not all represent a single, monolithic domain, and a careless investigatormight conclude that the 5-HT1B receptor gene was important for “alcohol-inducedataxia,” whereas a more careful analysis would reveal that it affects some but not

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.



all of the contributing behaviors (e.g., balance, intact proprioceptive feedback,patterned gait, muscle strength, etc.) (Boehm et al. 2000).

Although more difficult to demonstrate in humans, in whom the genotype ismuch more difficult to control rigorously, gene-environment interaction clearlyexists. For example, a classic study of the genetic susceptibility to alcoholism andrelated behaviors compared Scandinavian men at (or not at) genetic risk (basedon diagnosis of close relatives) for one of two broadly defined variants of al-coholism. Type I alcoholism is characterized by relatively mild abuse, minimalcriminality, and passive-dependent personality variables, whereas Type II alco-holism is characterized by early onset, violence, and criminality, and is largelylimited to males (Cloninger 1987). Multiple variables in the rearing environmentswere assessed, and individuals were subsequently classified as having been raisedin either risk-promoting or protective environments. Individuals at genetic risk forType I alcoholism were more often diagnosed, demonstrating genetic influence,but this tendency was much more pronounced when they also had higher-riskenvironments. For Type II alcoholism, genetic loading also increased diagnoses,but there was little further elevation if the rearing environment was also risky(Cloninger et al. 1981). This outcome illustrates the concept that the same envi-ronmental risk factors can play a very different role depending on an individual’sgenotype.

On at least three counts, this is a great oversimplification of a very complexanalysis. First, whereas many different diagnostic typologies have been proposedfor alcoholism, nearly all support the existence of at least two broadly differentiablevariants of the disease (Johnson et al. 1996a, Litt et al. 1992, Babor et al. 1992).Second, multivariate statistical methods were used to categorize variables in thesubjects’ environmental background as risk-promoting or protective. Finally, asimilar analysis of a sample of female alcoholics provided a somewhat differentoutcome (Bohman et al. 1981). Nonetheless, the interaction seen was substantial.

Gene-Environment Correlation

Genotypes are often not randomly represented in environments. A frequent contrib-utor to relapse to substance abuse is thought to be succumbing to the environmentaltriggers represented by myriad cues in the patient’s environment, e.g., seeing thehouse where he habitually purchased drugs, hanging around with other drug users.Studies have demonstrated the potency of exposure to previously drug-related cuesin eliciting both craving for drugs and increases in physiological responses suchas heart rate and pupil diameter (Childress et al. 1999, O’Brien et al. 1998). Atleast some contribution to substance abuse is likely to be the tendency of geneti-cally susceptible individuals to remain in the risk-promoting environment, therebypotentiating their overall risk. Indeed, many therapies strongly advocate makingradical changes in the day-to-day living situation of recovering addicts, a prescrip-tion that is unfortunately difficult for many to follow due to limited socioeconomicchoices (Budney & Higgins 1998).

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


450 CRABBE

It is obviously difficult to conduct controlled studies of gene-environment cor-relation with humans. However, one underlying principle is that drug responsescan often depend upon conditioning to environmental cues. Many animal studieshave documented genetic differences in sensitivity to drug-conditioned responses,such as place and taste conditioning (Broadbent et al. 1996, Cunningham 1995,Cunningham et al. 2000, Risinger & Cunningham 1998). The role of conditioningin drug abuse has been reviewed elsewhere (Altman et al. 1996, O’Brien et al.1998, Robbins & Everitt 1999).

Epistasis

Whenever more than one gene affects a trait, epistatic interactions may be at work.Complex behaviors, including those contributing to addictions, are influenced bymany genes, each with relatively small independent effect. Epistasis is the statisticalinteraction of such individual gene effects. The simplest case is that in which thepresence or absence of a particular allele at a second gene significantly modulatesthe effect of allelic differences at a gene of interest (Browman & Crabbe 1999). Arecent example is not strictly about addictive behavior, but rather anxiety, whichis extensively comorbid with the addictions. Three groups interested in the stressaxis recently independently produced mice in which the corticotropin-releasinghormone (CRH) receptor-2 gene (Crhr2) had been deleted. Because of the well-established role of CRH in modulating anxiety-like responses (Weninger et al.1999, Skutella et al. 1994), all three groups used the elevated plus-maze anxietytest as well as the classic open field test, and each tested both sexes. Many additionalvariables were also assessed in each study. Although the findings were internallyconsistent within each group’s results, they differed markedly in their conclusionsabout the role of the CRH-R2 receptor gene in anxiety. One group (Coste et al.2000) saw no effects on anxiety-related behavior in either sex, whereas the secondgroup (Bale et al. 2000) found greater anxiety in both male and female knockouts.The third group (Kishimoto et al. 2000) saw greater anxiety in male knockoutsonly, and in only one test.

The source of these differences could simply be that different apparatus, light-ing, handling conditions, and other test procedures were used, idiosyncratic to eachgroup. We have already seen that even when such variables are carefully standard-ized, genetic differences play out differently in different environments. Anotherpossibility is that the CRH-R2 receptor gene has no consistent role in modulatinganxiety, but each group used multiple, putative tests of anxiety and each obtainedlargely consistent results across tests.

It seems more likely that epistasis was at work. The genetically engineeredconstructs inserted into the embryonic stem cells to produce the gene deletionwere different in each laboratory, and each carried a relatively long piece of DNAalong with the targeted gene. Thus, other closely linked genes could have beenintroduced to the recipient mice, and these necessarily differed from laboratoryto laboratory. These “passenger genes” could have been interacting epistatically

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.



with the CRH-R2 receptor gene to influence behavioral outcomes, such that agiven anxiety response was only expressed in the presence of a specific, additionalgene in addition to the loss-of-function variant of the CRH-R2 receptor gene.Furthermore, each group introduced its null mutant into a different substrain of129 inbred mice (these 129 mice served as the source for the embryonic stem cellsinto which the null mutation was introduced). Different substrains of 129 are verysimilar genetically, but not identical (Simpson et al. 1997). The mice tested alsohad varying percentages of the C57BL/6 inbred strain genome, so the effects of thetargeted gene could also have been interacting with genes differing in the geneticbackground, as well as with passenger genes in the construct (Gerlai 1996). Thereare other possible contributors to these behavioral differences, discussed morefully in the three papers mentioned above and elsewhere (Crabbe 2001).

Researchers are beginning to look for these sorts of interactions in their genemapping efforts, and it is not surprising that epistasis appears to occur frequently.In a QTL analysis from our group, we were able to demonstrate a significantdifference in acute pentobarbital withdrawal severity between mice homozygousfor DBA/2J strain alleles in a region of chromosome 11 and those homozygousfor C57BL/6J alleles (Buck et al. 1999). This indicated the presence of a genein this chromosomal region where the DBA/2J-specified gene tended to reducewithdrawal as compared with the C57BL/6J gene. A recent analysis of epistaticinteractions, however, showed that this was only true when the animals had DBA/2Jalleles in a second region, on the distal end of chromosome 1. If mice had C57BL/6Jalleles on distal chromosome 1, there was no difference in withdrawal betweenmice with C57BL/6J and DBA/2J genomes on chromosome 11 (Hood et al. 2001).

Genes can obviously interact in multiples greater than two. The field is just nowbeginning to study such interactions, and extension of such analyses to multigenicinteractions will require a daunting degree of statistical power. This is becausepower to detect interactions requires much greater numbers of subjects than arerequired to detect main effects (Wahlsten 1990). Nonetheless, understanding thecomplexity of genetic interactions will be crucial. After all, the relevant clinicaltraits are extensively comorbid (e.g., alcoholism, other substance abuse, impulsiv-ity, attention deficit/hyperactivity disorder, depression, etc.), and their comorbidityis likely to represent a mixture of genetic and nongenetic sources (Crabbe 1999).

TAKE HOME MESSAGES

Interpretation of Genetic Differences

The ubiquitous influence of genes on addiction-related traits must not be over-simplified. “The gene for. . .” syndrome understandably infects the popular press(although scientists have a clear responsibility to lobby strenuously against thiskind of reporting). Unfortunately, for many molecularly oriented neuroscientists, alimited result identifying a specific gene with a specific behavioral outcome oftenalso leads to over-naive interpretation. For complex traits, the general rule seems to

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


452 CRABBE

be that the aggregate contribution of the many genes contributing toward individualdifferences is no more than 50% of the variance—the rest is explicitly not genetic.Whereas some would argue that teasing apart the relative contributions of genesand environments is a near impossibility (Gottlieb 1998), the general view is thatspecific domains of genetic influence can, with care, be identified. Nonetheless, itis important to remember that many genes contribute to complex traits, interactingwith each other as well as the environments in which they are expressed.

What Specific Contributions Can AnimalGenetics Make to the Addictions?

One suggestion raised earlier in this review was that insufficient use has beenmade of perhaps the most powerful of all behavior genetics techniques, artificialselection. If one wishes to deconstruct the contributions of disinhibition to drug-seeking behavior, it would be a straightforward matter to breed mice for highor low “novelty-seeking” responses and then see whether they differed in drugself-administration. It has been shown that when rats are exposed to an openfield, a situation whose novel features cause mild stress, they display differentlevels of locomotion. High responders can then be shown subsequently to self-administer psychostimulant drugs to a greater degree than low responders. Reviewsof these and related studies discuss the neuroendocrine and behavioral profilescharacterizing these groups (Piazza & Le Moal 1996, Bardo et al. 1996).

From a genetic perspective, these studies offer no evidence that the behavioraldifferences are genetic as opposed to environmental. However, simply attemptingto select for the novelty response would rapidly answer the question. Of course,the studies reviewed above suggest that it would not necessarily be easy to do thisexperiment. For example, breeding for high vs low scores on a signaled nosepoketask (Logue et al. 1998) might, or might not, lead to parallel divergence in scores ina delay or probability discounting task (Richards et al. 1997) or in a delayed rein-forcement of low rate operant task, even though all three tasks are thought to assayimpulsivity. Still, such experiments would be a worthwhile undertaking and couldoffer much useful insight to the predisposition/comorbidity issues surroundinghuman genetic studies.

A second area of contribution is the identification of specific genes for risk foror protection from addictive behavior (see “Quantitative Trait Locus Mapping”).The rapid pace of technological development in genetic markers [e.g., the devel-opment of many thousands of new genetic markers, much more densely spaced, byascertaining single nucleotide polymorphisms (Lindblad-Toh et al. 2000, Cargillet al. 1999)] will make the path from QTL to responsible gene much easier in thenear future.

Will the proliferation of genetic studies in animals resolve all the most vexing is-sues facing addiction studies? Almost certainly not. Even this brief review has madeit clear that any single addiction diagnosis is etiologically heterogeneous, whichprobably means that it is both genetically and environmentally heterogeneous

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.



as well (not to mention heterogeneous at the intersection of genotype and environ-ment, and so on). Certainly, the history of genetic animal-model research suggeststhat such studies have great potential for furthering our knowledge of how drugswork in the brain. There is hope among behavioral geneticists that use of geneticinformation could lead to better diagnostic approaches (Plomin & Crabbe 2000),but the principal, and probably insurmountable, problem is the small effect size ofmost specific genes of importance. And there is a strong probability that behav-ioral genomics studies in animals will lead to better therapeutic agents than thosecurrently available. It may not be necessary to identify all the influential genes todevise novel strategies for prevention and treatment of complex disease traits.

Addiction in the Postgenomics World

The rapid proliferation of genetic data–gathering capability has found many a sci-entist in possession of the DNA from many patients/subjects. Because it is now socomparatively sraightforward to genotype those samples, the potential for misuseof genetic information is great. A fraction of the resources of the Human GenomeProject has been devoted to the study of the Ethical, Legal and Social Implications(ELSI) Program and others with similar goals (see http://www.lbl.gov/Education/ELSI/ELSI.html). Privacy and confidentiality issues and their implications for em-ployment and insurance are an extremely complex area. In addition, patients whoseDNA leads to patentable discoveries are beginning to sue for a share of the profits.

The need for serious ethical discussions is clear, but the answers will not besimple. As has been the central message of this review, at the heart of the problemis the small effect size of any individual gene contributing to complex traits. Theethical issues surrounding a diagnosis of Huntington’s disease are difficult enough,and this is a single-gene disorder for which a yes/no answer to genetic risk canbe given. What does it mean to know that an individual has 3 of the 15 (or isit 30?) “bad” genes, e.g., those predisposing to alcoholism, versus having 8 ofthe 15? Obviously, everyone would prefer the former diagnosis to the latter, butabsent knowledge of whether alcoholism is a threshold character or a continuoustrait and how the various risk-promoting genes interact with each other in specificcombinations, and without environment-related information that is as sophisticatedand articulated as the genetic information on risk, it is difficult to see what theappropriate ethical choices are. This does not mean that regulation will not beattempted before the science is clear. Several states have introduced laws regulatinggenetic privacy.

I am not a bioethicist, but I have been repeatedly exposed to ethical issuesduring a career devoted to chasing genetic sources of influence on complex traits.In my perusals of the literature relating to ethical decision-making vis à vis ge-netics, I have been struck by the persistent tendency to raise, rather than answer,questions such as those raised above. The reader is directed to a recent reviewfor other sources of relevance to the ethical questions, where we perpetuatedthis tendency (Crabbe & Belknap 1998). In addition, the National Institutes of

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


454 CRABBE

Health maintains several links to web sites of relevance to ethnicity and genet-ics, gene patenting, genetic testing/counseling, and gene therapy/gene transfer(http://www.nih.gov/sigs/bioethics/). I suspect that no one with more than a littleknowledge in the related scientific areas feels capable of prescribing ethical guide-lines. Nearly all scientists I know would agree that it seems unfair, immoral, or aviolation of privacy for an insurance company to obtain access to a patient’s DNAinformation without explicit permission and decide that a high risk for a diseasejustified a higher premium. However, they are much more divided on the issue ofwhether a patient should retain privacy rights blocking the use of his or her DNAinformation, freely given with informed consent for a particular genetic linkagestudy, in a future genetic linkage study for a different trait, in which personalpatient-identifying information is doubly blinded. Entire books have been devotedto these complex issues, and it is simply beyond the scope of this review to pursuethem in any reasonable depth. The interested reader may also find useful the text ofa February 2000 ELSI Research Planning and Evaluation Group Report coveringthe first 10 years of the ELSI Programs and future plans, and the links cited therein(http://www.nhgri.nih.gov/ELSI/erpgreport.html).

ACKNOWLEDGMENTS

Thanks to Tamara Phillips and Chris Cunningham for their comments, and the VA,NIAAA, and NIDA for support.

Visit the Annual Reviews home page at www.AnnualReviews.org

LITERATURE CITED

Adams N, Hannah JA, Henry W. 2000. Envi-ronmental influences on the failure to drink ininbred rats with an ethanol preference.Phys-iol. Behav.69:563–70

Altman J, Everitt BJ, Glautier S, Markou A,Nutt D, et al. 1996. The biological, social andclinical bases of drug addiction: commentaryand debate.Psychopharmacology125:285–345

Babor TF, Hofmann M, DelBoca FK, Hessel-brock V, Meyer RE, et al. 1992. Types ofalcoholics. I. Evidence for an empirically de-rived typology based on indicators of vulner-ability and severity.Arch. Gen. Psychiatry49:599–608

Bale TL, Contarino A, Smith GW, Chan R,Gold LH, et al. 2000. Mice deficient forcorticotropin-releasing hormone receptor-2

display anxiety-like behaviour and are hy-persensitive to stress.Nat. Genet.24:410–14

Bardo MT, Donohew RL, Harrington NG.1996. Psychobiology of novelty seeking anddrug seeking behavior.Brain Behav. Res.77:23–43

Begleiter H, Porjesz B. 1999. What is inher-ited in the predisposition toward alcoholism?A proposed model.Alcohol: Clin. Exp. Res.23:1125–35

Belknap JK, Crabbe JC, Laursen SE. 1989.Ethanol and diazepam withdrawal convul-sions are extensively codetermined in WSPand WSR mice.Life Sci.44:2075–80

Belknap JK, Crabbe JC, Young ER. 1993.Voluntary consumption of ethanol in 15inbred mouse strains.Psychopharmacology112:503–10

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.



Belknap JK, Danielson PW, Lamé M, CrabbeJC. 1988. Ethanol and barbiturate withdrawalconvulsions are extensively codetermined inmice.Alcohol5:167–71

Belknap JK, Haltli NR, Goebel DM, Lamé M.1983. Selective breeding for high and lowlevels of opiate-induced analgesia in mice.Behav. Genet.13:383–96

Belknap JK, Hitzemann R, Crabbe JC, PhillipsTJ, Buck KJ, Williams RW. 2001. QTL anal-ysis and genome-wide mutagenesis in mice:complementary genetic approaches to thedissection of complex traits.Behav. Genet.31:5–15

Belknap JK, Laursen SE, Crabbe JC. 1987.Ethanol and nitrous oxide produce withdraw-al-induced convulsions by similar mecha-nisms in mice.Life Sci.41:2033–40

Bergen AW, Korczak JF, Weissbecker KA,Goldstein AM. 1999. A genome-wide searchfor loci contributing to smoking and alcoho-lism. Genet. Epidemiol.17(Suppl 1):S55–60

Boehm II SL, Schafer GL, Phillips TJ, Brow-man KE, Crabbe JC. 2000. Sensitivityto ethanol-induced motor incoordination in5-HT1B receptor null mutant mice is task-dependent: implications for behavioral as-sessment of genetically altered mice.Behav.Neurosci.114:401–9

Bohman M, Sigvardsson S, Cloninger CR.1981. Maternal inheritance of alcohol abuse.Cross-fostering analysis of adopted women.Arch. Gen. Psychiatry38:965–69

Broadbent J, Linder HV, Cunningham CL.1996. Genetic differences in naloxone en-hancement of ethanol-induced conditionedtaste aversion. Psychopharmacology126:147–55

Browman KE, Crabbe JC. 1999. Alcohol andgenetics: new animal models.Molec. Med.Today5:310–18

Buck K, Metten P, Belknap J, Crabbe J. 1999.Quantitative trait loci affecting risk for pen-tobarbital withdrawal map near alcohol with-drawal loci on mouse chromosomes 1, 4, and11.Mamm. Genome10:431–37

Buck KJ, Crabbe JC, Belknap JK. 2000. Alco-

hol and other abused drugs. InGenetic Influ-ences on Neural and Behavioral Functions,ed. DW Pfaff, WH Berrettini, TH Joh, SCMaxson, pp. 159–83. Boca Raton, FL: CRCPress

Buck KJ, Metten P, Belknap JK, Crabbe JC.1997. Quantitative trait loci involved in ge-netic predisposition to acute alcohol with-drawal in mice.J. Neurosci.17:3946–55

Budney AJ, Higgins ST. 1998.A Commu-nity Reinforcement Plus Vouchers Approach:Treating Cocaine Addiction. Rockville, MD:USDHHS-NIH-NIDA

Caldarone B, Saavedra C, Tartaglia K, WehnerJM, Dudek BC, Flaherty L. 1997. Quanti-tative trait loci analysis affecting contextualconditioning in mice.Nat. Genet.17:335–37

Cargill M, Altshuler D, Ireland J, Sklar P,Ardlie K, et al. 1999. Characterization ofsingle-nucleotide polymorphisms in codingregions of human genes.Nat. Genet.22:231–38

Castanon N, Perez-Diaz F, Mormède P. 1995.Genetic analysis of the relationships betweenbehavioral and neuroendocrine traits inRoman High and Low Avoidance rat lines.Behav. Genet.25:371–84

Charney DS, Nestler EJ, Bunney BS, eds.1999.Neurobiology of Mental Illness. NewYork: Oxford Univ. Press

Chen CC, Lu RB, Chen YC, Wang MF,Chang YC, et al. 1999. Interaction betweenthe functional polymorphisms of the alcohol-metabolism genes in protection against alco-holism.Am. J. Hum. Genet.65:795–807

Chester JA, Risinger FO, Cunningham CL.1998. Ethanol reward and aversion in micebred for sensitivity to ethanol withdrawal.Al-cohol: Clin. Exp. Res.22:468–73

Childress AR, Mozley PD, McElgin W, Fitz-gerald J, Reivich M, O’Brien CP. 1999. Lim-bic activation during cue-induced cocainecraving.Am. J. Psychiatry156:11–18

Cloninger CR. 1987. Neurogenetic adaptivemechanisms in alcoholism.Science236:410–16

Cloninger CR, Bohman M, SigvardssonS. 1981. Inheritance of alcohol abuse.

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


456 CRABBE

Cross-fostering analysis of adopted men.Arch. Gen. Psychiatry38:861–68

Coste SC, Kesterson RA, Heldwein KA, Ste-vens SL, Heard AD, et al. 2000. Abnormaladaptations to stress and impaired cardiovas-cular function in mice lacking corticotropin-releasing hormone receptor-2.Nat. Genet.24:403–9

Courvoisier H, Moisan MP, Sarrieau A, Hend-ley ED, Mormede P. 1996. Behavioral andneuroendocrine reactivity to stress in theWKHA/WKY inbred rat strains: a multi-factorial and genetic analysis.Brain Res.743:77–85

Crabbe JC. 1996. A genetic animal model ofalcohol withdrawal.Alcohol: Clin. Exp. Res.20:96A–100A

Crabbe JC. 1999. Molecular genetics of ad-diction. See Charney et al. 1999, pp. 591–600

Crabbe JC. 2001. Use of genetic analyses torefine phenotypes related to alcohol toler-ance and dependence.Alcohol: Clin. Exp.Res.25:288–92

Crabbe JC, Belknap JK. 1998. Ethical conse-quences of mapping QTLs for complex hu-man traits. InMolecular Dissection of Com-plex Traits, ed. AH Paterson, pp. 279–85.Boca Raton, FL: CRC Press

Crabbe JC, Belknap JK, Buck KJ. 1994.Genetic animal models of alcohol and drugabuse.Science264:1715–23

Crabbe JC, Harris RA, eds. 1991.The GeneticBasis of Alcohol and Drug Actions.NewYork: Plenum Press

Crabbe JC, Kosobud A, Young ER, Tam BR,McSwigan JD. 1985. Bidirectional selec-tion for susceptibility to ethanol withdrawalseizures inMus musculus. Behav. Genet.15:521–36

Crabbe JC, Li T–K. 1995. Genetic strategiesin preclinical substance abuse research. InPsychopharmacology: A Fourth Generationof Progress, ed. FE Bloom, DJ Kupfer,69:799–811, New York: Raven Press

Crabbe JC, Phillips TJ, Buck KJ, CunninghamCL, Belknap JK. 1999a. Identifying genesfor alcohol and drug sensitivity: recent

progress and future directions.Trends Neu-rosci.22:173–79

Crabbe JC, Phillips TJ, Feller DJ, Hen R,Wenger CD, et al. 1996. Elevated alcoholconsumption in null mutant mice lacking5–HT1B serotonin receptors.Nat. Genet.14:98–101

Crabbe JC, Wahlsten D, Dudek BC. 1999b. Ge-netics of mouse behavior: interactions withlaboratory environment.Science284:1670–22

Crawley JN, Belknap JK, Collins A, CrabbeJC, Frankel W, et al. 1997. Behavioral phe-notypes of inbred mouse strains: implicationsand recommendations for molecular studies.Psychopharmacology132:107–24

Cunningham CL. 1995. Localization of genesinfluencing ethanol-induced conditionedplace preference and locomotor activity inBXD recombinant inbred mice.Psychophar-macology120:28–41

Cunningham CL, Howard MA, Gill SJ, Rubin-stein M, Low MJ, Grandy DK. 2000. Etha-nol-conditioned place preference is re-duced in dopamine D2 receptor-deficientmice.Pharmacol. Biochem. Behav.67:693–99

Davies JB. 1998. Pharmacology versus socialprocess: competing or complementary viewson the nature of addiction?Pharmacol. Ther.80:265–75

Deitrich RA, Bludeau P, Erwin VG. 2000. Phe-notypic and genotypic relationships betweenethanol tolerance and sensitivity in mice se-lectively bred for initial sensitivity to ethanol(SS and LS) or development of acute toler-ance (HAFT and LAFT).Alcohol: Clin. Exp.Res.24:595–604

El Ghundi M, George SR, Drago J, FletcherPJ, Fan T, et al. 1998. Disruption of dopa-mine D1 receptor gene expression attenuatesalcohol-seeking behavior.Eur. J. Pharmacol.353:149–58

Enoch MA, Goldman D. 2001. The geneticsof alcoholism and alcohol abuse.Curr. Psy-chiat.3:144–51

Eriksson K. 1972. Behavioral and physiolog-ical differences among rat strains specially

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.



selected for their alcohol consumption.Ann.NY Acad. Sci.197:32–41

Fehr C, Belknap JK, Crabbe JC, Buck KJ.2001. Congenic mapping of alcohol and pen-tobarbital withdrawal liability quantitativetrait loci to a 0. 9 cM region of mouse chro-mosome 4: identification ofMpdzas a can-didate gene. Submitted

Foroud T, Bucholz KK, Edenberg HJ, Goate A,Neuman RJ, et al. 1998. Linkage of analcoholism-related severity phenotype tochromosome 16.Alcohol: Clin. Exp. Res.22:2035–42

Foroud T, Edenberg HJ, Goate A, Rice J,Flury L, et al. 2000. Alcoholism susceptibil-ity loci: confirmation studies in a replicatesample and further mapping.Alcohol: Clin.Exp. Res.24:933–45

Freeman WM, Nader MA, Nader SH, Robert-son DJ, Gioia L, et al. 2001. Chronic cocaine-mediated changes in non-human primatenucleus accumbens gene expression.J. Neu-rochem.77:542–9

Fry JD, Nuzhdin SV, Pasyukova EG, MackayTF. 1998. QTL mapping of genotype-environment interaction for fitness inDroso-phila melanogaster. Genet. Res.71:133–41

Gehle VM, Erwin VG. 2000. The genetics ofacute functional tolerance and initial sensitiv-ity to ethanol for an ataxia test in the LSxSSRI strains.Alcohol: Clin. Exp. Res.24:579–87

Gelernter J. 1999. Clinical molecular genetics.See Charney et al. 1999, pp. 108–20

Gerlai R. 1996. Gene-targeting studies ofmammalian behavior: Is it the mutation orthe background genotype?Trends Neurosci.19:177–81

Gershenfeld HK, Neumann PE, Mathis C,Crawley JN, Li X, Paul SM. 1997. Mappingquantitative trait loci for open-field behaviorin mice.Behav. Genet.27:201–10

Gershenfeld HK, Paul SM. 1997. Mappingquantitative trait loci for fear-like behaviorsin mice.Genomics46:1–8

Gershenfeld HK, Paul SM. 1998. Towards a ge-netics of anxious temperament: from mice to

men.Acta Psychiatr. Scand.393(Suppl.):56–65

Gottlieb G. 1998. Normally occurring environ-mental and behavioral influences on gene ac-tivity: from central dogma to probabilisticepigenesis.Psychol. Rev.105:792–802

Grisel JE, Mogil JS, Grahame NJ, Rubin-stein M, Belknap JK, et al. 1999. Ethanoloral self-administration is increased in mu-tant mice with decreasedβ-endorphin ex-pression.Brain Res.835:62–67

Hefti F. 2001. From genes to effective drugs forneurological and psychiatric diseases.TrendsPharmacol. Sci.22:159–60

Henniger MSH, Spanagel R, Wigger A, Land-graf R, Holter SM. 2001. Alcohol self-administration in two rat lines selectivelybred for extremes in anxiety-related behav-ior. Submitted

Hodge CW, Mehmert KK, Kelley SP, McMa-hon T, Haywood A, et al. 1999. Supersensi-tivity to allosteric GABAA receptor modula-tors and alcohol in mice lacking PKCε. Nat.Neurosci.2:997–1002

Hood H, Crabbe JC, Belknap JK, Buck KJ.2001. Epistatic interaction between quantita-tive trait loci affects the genetic predisposi-tion for physical dependence on pentobarbi-tal in mice.Behav. Genet.31:93–100

Hood HM, Buck KJ. 2000. Allelic varia-tion in the GABAA receptor γ2 subunitis associated with genetic susceptibility toethanol-induced motor incoordination andhypothermia, conditioned taste aversion, andwithdrawal in BXD/Ty recombinant inbredmice.Alcohol: Clin. Exp. Res.24:1327–34

Hrabe de Angelis MH, Flaswinkel H, FuchsH, Rathkolb B, Soewarto D, et al. 2000.Genome-wide, large-scale production of mu-tant mice by ENU mutagenesis.Nat. Genet.25:444–47

Iwata N, Virkkunen M, Goldman D. 2000.Identification of a naturally occurringPro385-Ser385 substitution in the GABAA

receptorα6 subunit gene in alcoholics andhealthy volunteers.Mol. Psychiatry5:316–19

Jacob HJ, Lindpaintner K, Lincoln SE,

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


458 CRABBE

Kusumi K, Bunker RK, et al. 1991. Geneticmapping of a gene causing hypertension inthe stroke-prone spontaneously hypertensiverat.Cell 67:213–24

Johnson EO, van den Bree MBM, GupmanAE, Pickens RW. 1998. Extension of a typo-logy of alcohol dependence based on relativegenetic and environmental loading.Alcohol:Clin. Exp. Res.22:1421–29

Johnson EO, van den Bree MBM, Pickens RW.1996a. Subtypes of alcohol-dependent men:a typology based on relative genetic and envi-ronmental loading.Alcohol: Clin. Exp. Res.20:1472–80

Johnson EO, van den Bree MBM, Uhl GR,Pickens RW. 1996b. Indicators of genetic andenviromental influences in drug abusing in-dividuals.Drug Alcohol Depend.41:17–23

Jones BC, Tarantino LM, Rodriguez LA, ReedCL, McClearn GE, et al. 1999. Quantitative-trait loci analysis of cocaine-related be-haviours and neurochemistry.Pharmacoge-netics9:607–17

Kendler KS, Prescott CA. 1999. Caffeine in-take, tolerance, and withdrawal in women: apopulation-based twin study.Am. J. Psychi-atry 156:223–28

Kishimoto T, Radulovic J, Radulovic M, LinCR, Schrick C, et al. 2000. Deletion ofcrhr2reveals an anxiolytic role for corticotropin-releasing hormone receptor-2.Nat. Genet.24:415–19

Kosobud A, Bodor AS, Crabbe JC. 1988. Vol-untary consumption of ethanol in WSP, WSCand WSR selectively bred mouse lines.Phar-macol. Biochem. Behav.29:601–7

Kulikov A, Castanon N, Mormède P, Chaou-loff F. 1995. Cerebral tryptophan hydroxy-lase activity, and 5-HT1A receptor, 5-HT2A

receptor, and 5-HT transporter binding ingrouped and isolated Roman RHA andRLA rats: relationships with behaviours intwo models of anxiety.Psychopharmacology121:385–95

Lander ES, Linton LM, Birren B, NusbaumC, Zody MC, et al. 2001. Initial sequencingand analysis of the human genome.Nature409:860–921

Lappalainen J, Long JC, Eggert M, Ozaki N,Robin RW, et al. 1998. Linkage of antisocialalcoholism to the serotonin 5-HT1B receptorgene in 2 populations.Arch. Gen. Psychiatry55:989–94

Le Marquand D, Pihl RO, Benkelfat C. 1994a.Serotonin and alcohol intake, abuse, and de-pendence: clinical evidence.Biol. Psychiatry36:326–37

Le Marquand D, Pihl RO, Benkelfat C. 1994b.Serotonin and alcohol intake, abuse, and de-pendence: findings of animal studies.Biol.Psychiatry36:395–421

Lewohl JM, Wang L, Miles MF, Zhang L,Dodd PR, Harris RA. 2000. Gene expres-sion in human alcoholism: microarray analy-sis of frontal cortex.Alcohol: Clin. Exp. Res.24:1873–82

Li T-K, Lumeng L, McBride WJ, Murphy JM.1994. Genetic and neurobiological basis ofalcohol-seeking behavior.Alcohol 29:697–700

Liebsch G, Montkowski A, Holsboer F, Land-graf R. 1998a. Behavioural profiles of twoWistar rat lines selectively bred for high orlow anxiety-related behaviour.Behav. BrainRes.94:301–10

Liebsch G, Linthorst ACE, Neumann ID, ReulJMHM, Holsboer F, Landgraf R. 1998b. Be-havioral, physiological, and neuroendocrinestress responses and differential sensitivity todiazepam in two Wistar rat lines selectivelybred for High- or Low-Anxiety-related Be-havior.Neuropsychopharmacology19:381–96

Lindblad-Toh K, Winchester E, Daly MJ,Wang DG, Hirschhorn JN, et al. 2000. Large-scale discovery and genotyping of single-nucleotide polymorphisms in the mouse.Nat.Genet.24:381–86

Litt MK, Babor TF, DelBoca FK, Kadden RM,Cooney NL. 1992. Types of alcoholics. II.Application of an empirically derived typol-ogy to treatment matching.Arch. Gen. Psy-chiatry49:609–14

Logue SF, Swartz RJ, Wehner JM. 1998.Genetic correlation between performance onan appetitive-signaled nosepoke task and

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.



voluntary ethanol consumption.Alcohol:Clin. Exp. Res.22:1912–20

Loh EW, Higuchi S, Matsushita S, Murray R,Chen CK, Ball D. 2000. Association analysisof the GABAA receptor subunit genes clus-ter on 5q33-34 and alcohol dependence in aJapanese population.Mol. Psychiatry5:301–7

Loh E-W, Smith I, Murray R, McLaughlin M,McNulty S, Ball D. 1999. Association be-tween variants at the GABAAβ2, GABAAα6and GABAAγ2 gene cluster and alcohol de-pendence in a Scottish population.Mol. Psy-chiatry4:539–44

Long AD, Langley CH. 1999. The power of as-sociation studies to detect the contribution ofcandidate genetic loci to variation in complextraits.Genome Res.9:720–31

Lumeng L, Li TK. 1986. The development ofmetabolic tolerance in the alcohol-preferringP rats: comparison of forced and free-choicedrinking of ethanol.Pharmacol. Biochem.Behav.25:1013–20

MacDonald ME, Ambrose CM, Duyao MP,Myers RH, Lin C, et al. 1993. A novel genecontaining a trinucleotide repeat that is ex-panded and unstable on Huntington’s diseasechromosomes.Cell 72:971–83

Madden PA, Heath AC, Pedersen NL, KaprioJ, Koskenvuo MJ, Martin NG. 1999. Thegenetics of smoking persistence in men andwomen: a multicultural study.Behav. Genet.29:423–31

Mardones J, Segovia-Riquelme N. 1983.Thirty-two years of selection of rats byethanol preference: UChA and UChB strains.Neurobehav. Toxicol. Teratol.5:171–78

Marks MJ, Stitzel JA, Collins AC. 1989. Ge-netic influences on nicotine responses.Phar-macol. Biochem. Behav.33:667–78

Marlatt GA, Baer JS, Donovan DM, KivlahanDR. 1988. Addictive behaviors: etiology andtreatment.Annu. Rev. Psychol.39:223–52

Marley RJ, Arros DM, Henricks KK, Mar-ley ME, Miner LL. 1998. Sensitivity to co-caine and amphetamine among mice selec-tively bred for differential cocaine sensitivity.Psychopharmacology140:42–51

McClearn GE, Rodgers DA. 1959. Differencesin alcohol preference among inbred strainsof mice.Q. J. Stud. Alcohol20:691–95

McGuffin P, Riley B, Plomin R. 2001. Ge-nomics and behavior. Toward behavioral ge-nomics.Science291:1232–49

Merikangas KR, Swendsen JD. 2001. Contri-butions of epidemiology to the neurobiologyof mental illness. See Charney et al. 1999,pp. 100–7

Metten P, Phillips TJ, Crabbe JC, TarantinoLM, McClearn GE, et al. 1998. High geneticsusceptibility to ethanol withdrawal predictslow ethanol consumption.Mamm. Genome9:983–90

Mogil JS, Flodman P, Spence MA, SternbergWF, Kest B, et al. 1995. Oligogenic deter-mination of morphine analgesic magnitude:a genetic analysis of selectively bred mouselines.Behav. Genet.25:397–406

Mogil JS, Sternberg WF, Marek P, SadowskiB, Belknap JK, Liebeskind JC. 1996. Thegenetics of pain and pain inhibition.Proc.Natl. Acad. Sci. USA93:3048–55

Mohammed AH. 2000. Genetic dissection ofnicotine-related behaviour: a review of an-imal studies.Behav. Brain Res.113:35–41

Moisan MP, Courvoisier H, Bihoreau MT,Gauguier D, Hendley ED, et al. 1996. A ma-jor quantitative trait locus influences hyper-activity in the WKHA rat. Nat. Genet.14:471–73

Murphy JM, Gatto GJ, Waller MB, McBrideWJ, Lumeng L, Li TK. 1986. Effects ofscheduled access on ethanol intake by thealcohol-preferring (P) line of rats.Alcohol3:331–36

Murray JB. 1993. Review of research on patho-logical gambling.Psychol. Rep.72(3, Pt. 1):791–810

Nadeau JH, Frankel WN. 2000. The roads fromphenotypic variation to gene discovery: mu-tagenesis versus QTLs.Nat. Genet.25:381–84

Nathan PE. 1988. The addictive personality isthe behavior of the addict.J. Consult. Clin.Psychol.56:183–88

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


460 CRABBE

Nestler EJ. 2000. Genes and addiction.Nat.Genet.26:277–81

Nolan PM, Kapfhamer D, Bucan M. 1997.Random mutagenesis screen for dominantbehavioral mutations in mice.Methods13:379–95

Nolan PM, Peters J, Strivens M, Rogers D,Hagan J, et al. 2000. A systematic, genome-wide, phenotype-driven mutagenesis pro-gramme for gene function studies in themouse.Nat. Genet.25:440–43

O’Brien CP, Childress AR, Ehrman R, Rob-bins SJ. 1998. Conditioning factors in drugabuse: Can they explain compulsion?J. Psy-chopharmacol.12:15–22

Overstreet DH. 1993. The Flinders sensitiveline rats: a genetic animal model of de-pression.Neurosci. Biobehav. Rev.17:51–68

Overstreet DH, Halikas JA, Seredenin SB,Kampov-Polevoy AB, Viglinskaya IV, et al.1997. Behavioral similarities and differ-ences among alcohol-preferring and -non-preferring rats: confirmation by factor anal-ysis and extension to additional groups.Alcohol: Clin. Exp. Res.21:840–48

Paigen K, Eppig JT. 2000. A mouse phenomeproject.Mamm. Genome11:715–17

Patton D, Barnes GE, Murray RP. 1994. Thereliability and construct validity of two mea-sures of addictive personality.Int. J. Addict.29:999–1014

Phillips TJ, Belknap JK, Buck KJ, Cunning-ham CL. 1998a. Genes on mouse chromo-somes 2 and 9 determine variation in ethanolconsumption.Mamm. Genome9:936–41

Phillips TJ, Brown KJ, Burkhart-Kasch S,Wenger CD, Kelly MA, et al. 1998b. Alco-hol preference and sensitivity are markedlyreduced in mice lacking dopamine D2 recep-tors.Nat. Neurosci.1:610–15

Phillips TJ, Hen R, Crabbe JC. 1999. Compli-cations associated with genetic backgroundeffects in research using knockout mice.Psy-chopharmacology147:5–7

Piazza PV, Le Moal M. 1996. Pathophysiolog-ical basis of vulnerability to drug abuse: roleof an interaction between stress, glucocorti-

coids, and dopaminergic neurons.Annu. Rev.Pharmacol. Toxicol.36:359–78

Pickens RW, Svikis DS, McGue M, LaBudaMC. 1995. Common genetic mechanisms inalcohol, drug, and mental disorder comorbid-ity. Drug Alcohol Depend.39:129–38

Pickens RW, Svikis DS, McGue M, LykkenDT, Heston LL, Clayton PJ. 1991. Hetero-geneity in the inheritance of alcoholism.Arch. Gen. Psychiatry48:19–28

Plomin R, Crabbe J. 2000. DNA.Psychol. Bull.126:806–28

Prescott CA, Aggen SH, Kendler KS. 2000.Sex-specific genetic influences on the comor-bidity of alcoholism and major depressionin a population-based sample of US twins.Arch. Gen. Psychiatry57:803–11

Prescott CA, Kendler KS. 1999. Genetic andenvironmental contributions to alcohol abuseand dependence in a population-based sam-ple of male twins.Am. J. Psychiatry156:34–40

Ramos A, Berton O, Mormède P, Chaouloff F.1997. A multiple-test study of anxiety-related behaviours in six inbred rat strains.Behav. Brain Res.85:57–69

Ramos A, Mellerin Y, Mormède P, ChaouloffF. 1998. A genetic and multifactorial analysisof anxiety-related behaviours in Lewis andSHR intercrosses.Brain Behav. Res.96:195–205

Ramos A, Moisan MP, Chaouloff F, Mor-mede C, Mormède P. 1999. Identificationof female-specific QTLs affecting an emo-tionality-related behavior in rats.Mol. Psy-chiatry4:453–62

Ramos A, Mormede P. 1998. Stress and emo-tionality: a multidimensional and geneticapproach.Neurosci. Biobehav. Rev.22:33–57

Reich T, Hinrichs A, Culverhouse R, BierutL. 1999. Genetic studies of alcoholism andsubstance dependence.Am. J. Hum. Genet.65:599–605

Reilly MT, Fehr C, Buck KJ. 2001. Alcohol andgene expression in the central nervous sys-tem. InNutrient-Gene Interactions in Healthand Disease, ed. N Moussa-Moustaid, CD

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.



Berdanier, 7:131–62. Boca Raton, FL: CRCPress

Richards JB, Mitchell SH, de Wit H, SeidenLS. 1997. Determination of discount func-tions in rats with an adjusting-amount proce-dure.J. Exp. Anal. Behav.67:353–66

Risinger FO, Cunningham CL. 1998. Ethanol-induced conditioned taste aversion in BXDrecombinant inbred mice.Alcohol: Clin. Exp.Res.22:1234–44

Robbins TW, Everitt BJ. 1999. Drug addiction:bad habits add up [news].Nature398:567–70

Rodgers DA, McClearn GE. 1962. Mousestrain differences in preference for variousconcentrations of alcohol.Q. J. Stud. Alco-hol 23:1:26–33

Rose RJ. 1995. Genes and human behavior.Annu. Rev. Psychol.46:625–54

Rozin P, Stoess C. 1993. Is there a general ten-dency to become addicted?Addict. Behav.18:81–87

Sander T, Ball D, Murray R, Patel J, Samo-chowiec J, et al. 1999. Association analysisof sequence variants of GABAAα6, α2, andγ2 gene cluster and alcohol dependence.Al-cohol. Clin. Exp. Res.23:427–31

Schechter MD, Meehan SM, Schechter JB.1995. Genetic selection for nicotine activ-ity in mice correlates with conditioned placepreference.Eur. J. Pharmacol.279:59–64

Schuckit MA. 1999. New findings in the genet-ics of alcoholism.JAMA281:1875–76

Schuckit MA. 2000. Biological phenotypes as-sociated with individuals at high risk for de-veloping alcohol-related disorder. Part 2.Ad-dict. Biol.5:23–36

Schuckit MA, Edenberg HJ, Kalmijn J, FluryL, Smith TL, et al. 2001. A genome-widesearch for genes that relate to a low level ofresponse to alcohol.Alcohol: Clin. Exp. Res.25:323–29

Schuckit MA, Smith TL. 1996. An 8-yearfollow-up of 450 sons of alcoholic and con-trol subjects.Arch. Gen. Psychiatry53:202–10

Schuckit MA, Tipp JE, Bergman M, Reich W,Hesselbrock VM, Smith TL. 1997. Compar-ison of induced and independent major de-

pressive disorders in 2,945 alcoholics.Am. J.Psychiatry154:948–57

Seale TW, Johnson P, Carney JM, RennertOM. 1984. Interstrain variation in acute toxicresponse to caffeine among inbred mice.Pharmacol. Biochem. Behav.20:576–73

Self DW, Nestler EJ. 1995. Molecular mech-anisms of drug reinforcement and addiction.Annu. Rev. Neurosci.18:463–95

Shaffer HJ. 1999. Strange bedfellows: a criticalview of pathological gambling and addiction.Addiction94:1445–48

Simpson EM, Linder CC, Sargent EE, Davis-son MT, Mobraaten LE, Sharp JJ. 1997. Ge-netic variation among 129 substrains and itsimportance for targeted mutagenesis in mice.Nat. Genet.16:19–27

Skutella T, Montkowski A, Stohr T, ProbstJC, Landgraf R, et al. 1994. Corticotropin-releasing hormone (CRH) antisense oligo-deoxynucleotide treatment attenuates socialdefeat-induced anxiety in rats.Cell. Mol.Neurobiol.14:579–88

Slutske WS, Eisen S, True WR, Lyons MJ,Goldberg J, Tsuang M. 2000. Common ge-netic vulnerability for pathological gamblingand alcohol dependence in men.Arch. Gen.Psychiatry57:666–73

Stallings MC, Hewitt JK, Beresford T, HeathAC, Eaves LJ. 1999. A twin study of drinkingand smoking onset and latencies from firstuse to regular use.Behav. Genet.29:409–21

Stitzel JA, Lu Y, Jimenez M, Tritto T, Col-lins AC. 2000. Genetic and pharmacologi-cal strategies identify a behavioral functionof neuronal nicotinic receptors.Behav. BrainRes.113:57–64

Thibault C, Lai C, Wilke N, Duong B, OliveMF, et al. 2000. Expression profiling of neu-ral cells reveals specific patterns of ethanol-responsive gene expression.Mol. Pharma-col. 58:1593–600

Thiele TE, Marsh DJ, Ste Marie, Bernstein IL,Palmiter RD. 1998. Ethanol consumptionand resistance are inversely related to neu-ropeptide Y levels.Nature396:366–69

Thiele TE, Willis B, Stadler J, Reynolds JG,Bernstein IL, McKnight GS. 2000. High

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.


462 CRABBE

ethanol consumption and low sensitivity toethanol-induced sedation in protein kinaseA-mutant mice.J. Neurosci.0:RC75(1–6)

Tsuang MT, Lyons MJ, Eisen SA, GoldbergJ, True W, et al. 1996. Genetic influenceson DSM-III-R drug abuse and dependence: astudy of 3,372 twin pairs.Am. J. Med. Genet.67:473–77

Turri MG, Talbot CJ, Radcliffe RA, WehnerJM, Flint J. 1999. High-resolution mappingof quantitative trait loci for emotionality inselected strains of mice.Mamm. Genome10:1098–101

Tzschentke TM. 1998. Measuring reward withthe conditioned place preference paradigm:a comprehensive review of drug effects, re-cent progress and new issues.Prog. Neuro-biol. 56:613–72

Uhl GR. 1999. Molecular genetics of substanceabuse vulnerability: a current approach.Neu-ropsychopharmacology20:3–9

Vadasz C, Saito M, Balla A, Kiraly I, GyetvaiB, et al. 2000. Mapping of quantitative traitloci for ethanol preference in quasi-congenicstrains.Alcohol20:161–71

van den Bree MBM, Johnson EO, Neale MC,Pickens RW. 1998a. Genetic and environ-mental influences on drug use and abuse/dependence in male and female twins.DrugAlcohol Depend.52:231–41

van den Bree MBM, Svikis DS, Pickens RW.1998b. Genetic influences in antisocial per-sonality and drug use disorders.Drug Alco-hol Depend.49:177–87

van Ree JM, Gerrits MA, Vanderschuren LJ.1999. Opioids, reward and addiction: an en-counter of biology, psychology, and medi-cine.Pharmacol. Rev.51:341–96

Venter JC, Adams MD, Myers EW, Li PW,Mural RJ, et al. 2001. The sequence of thehuman genome.Science291:1304–51

Wahlsten D. 1990. Insensitivity of the analysisof variance to heredity-environment interac-tion. Behav. Brain Sci.13:109–20

Wahlsten D. 1999. Single-gene influences on

brain and behavior.Annu. Rev. Psychol.50:599–624

Waller MB, McBride WJ, Lumeng L, Li TK.1982a. Effects of intravenous ethanol andof 4-methylpyrazole on alcohol drinkingin alcohol-preferring rats.Pharmacol. Bio-chem. Behav.17:763–68

Waller MB, McBride WJ, Lumeng L, Li TK.1982b. Induction of dependence on ethanolby free-choice drinking in alcohol-preferringrats.Pharmacol. Biochem. Behav.16:501–7

Watson SJ, Akil H. 1999. Gene chips and ar-rays revealed: a primer on their power andtheir uses.Biol. Psychiatry45:533–43

Wehner JM, Bowers BJ. 1995. Use of transgen-ics, null mutants, and antisense approaches tostudy ethanol’s actions.Alcohol: Clin. Exp.Res.19:811–20

Wehner JM, Radcliffe RA, Rosmann ST,Christensen SC, Rasmussen DL, et al. 1997.Quantitative trait locus analysis of contex-tual fear conditioning in mice.Nat. Genet.17:331–34

Weiss JM, Cierpial MA, West CHK. 1998.Selective breeding of rats for high and lowmotor activity in a swim test: toward anew animal model of depression.Pharma-col. Biochem. Behav.61:49–66

Weninger SC, Dunn AJ, Muglia LJ, DikkesP, Miczek KA, et al. 1999. Stress-inducedbehaviors require the corticotropin-releasinghormone (CRH) receptor, but not CRH.Proc.Natl. Acad. Sci. USA96:8283–88

Whatley VJ, Johnson TE, Erwin VG. 1999.Identification and confirmation of quantita-tive trait loci regulating alcohol consumptionin congenic strains of mice.Alcohol: Clin.Exp. Res.23:1262–71

Williams TG. 1996. Substance abuse and ad-dictive personality disorders. InHandbookof Relational Diagnosis and DysfunctionalFamily Patterns, ed. FW Kaslow, pp. 448–62. New York: Wiley & Sons

Wise RA. 1996. Neurobiology of addiction.Curr. Opin. Neurobiol.6:243–51

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.

P1: FDS

December 4, 2001 14:39 Annual Reviews AR146-FM

Annual Review of PsychologyVolume 53, 2002

CONTENTS

Frontispiece—Endel Tulving xvi

PREFATORY

Episodic Memory: From Mind to Brain, Endel Tulving 1

GENETICS OF BEHAVIOR

Genetic Contributions to Addiction, John C. Crabbe 435

BRAIN IMAGING/COGNITIVE NEUROSCIENCE

Child-Clinical/Pediatric Neuropsychology: Some Recent Advances,Byron P. Rourke, S. A. Ahmad, D. W. Collins, B. A. Hayman-Abello,S. E. Hayman-Abello, and E. M. Warriner 309

AUDITION AND ITS BIOLOGICAL BASES

Change Detection, Ronald A. Rensink 245

MEMORY

Remembering Over the Short-Term: The Case Against the StandardModel, James S. Nairne 53

JUDGMENT AND DECISION MAKING

Rationality, Eldar Shafir and Robyn A. LeBoeuf 491

BIOLOGICAL AND GENETIC PROCESSES IN DEVELOPMENT

Gene-Environment Interplay in Relation to Emotional andBehavioral Disturbance, Michael Rutter and Judy Silberg 463

DEVELOPMENT IN SOCIETAL CONTEXT

Socioeconomic Status and Child Development, Robert H. Bradleyand Robert F. Corwyn 371

MOOD DISORDERS

Depression: Perspectives from Affective Neuroscience, Richard J.Davidson, Diego Pizzagalli, Jack B. Nitschke, and Katherine Putnam 545

PSYCHOPATHOLOGY: VARIOUS DISORDERS

Causes of Eating Disorders, Janet Polivy and C. Peter Herman 187

vi

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.

P1: FDS


CONTENTS vii

Insomnia: Conceptual Issues in the Development, Maintenanceand Treatment of Sleep Disorder in Adults, Colin A. Espie 215

CLINICAL ASSESSMENT

Clinical Assessment, James M. Wood, Howard N. Garb,Scott O. Lilienfeld, and M. Teresa Nezworski 519

ADULT CLINICAL NEUROPSYCHOLOGY

Adult Clinical Neuropsychology: Lessons from Studies of theFrontal Lobes, Donald T. Stuss and Brian Levine 401

SELF AND IDENTITY

Self and Social Identity, Naomi Ellemers, Russell Spears,and Bertjan Doosje 161

ALTRUISM AND AGGRESSION

Human Aggression, Craig A. Anderson and Brad J. Bushman 27

INTERGROUP RELATIONS, STIGMA, STEREOTYPING, PREJUDICE,DISCRIMINATION

Intergroup Bias, Miles Hewstone, Mark Rubin, and Hazel Willis 575

CULTURAL INFLUENCES

Cultural Influences on Personality, Harry C. Triandisand Eunkook M. Suh 133

ORGANIZATIONAL PSYCHOLOGY OR ORGANIZATIONAL BEHAVIOR

Organizational Behavior: Affect in the Workplace, Arthur Briefand Howard Weiss 279

LEARNING AND PERFORMANCE IN EDUCATIONAL SETTINGS

Motivational Beliefs, Values, and Goals, Jacquelynne S. Ecclesand Allan Wigfield 109

PSYCHOBIOLOGICAL FACTORS IN HEALTH

Emotions, Morbidity, and Mortality: New Perspectives fromPsychoneuroimmunology, Janice K. Kiecolt-Glaser, LynanneMcGuire, Theodore F. Robles, and Ronald Glaser 83

PSYCHOPHYSIOLOGICAL DISORDERS AND PSYCHOLOGICAL EFFECTS

ON MEDICAL DISORDERS

Effects of Psychological and Social Factors on Organic Disease:A Critical Assessment of Research on Coronary Heart Disease,David S. Krantz and Melissa K. McCeney 341

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.

P1: FDS


viii CONTENTS

ANALYSIS OF LATENT VARIABLES

Latent Variables in Psychology and the Social Sciences,Kenneth A. Bollen 605

INDEXES

Author Index 635Subject Index 679Cumulative Index of Contributing Authors, Volumes 43–53 705Cumulative Index of Chapter Titles, Volumes 43–53 709

ERRATA

Online log of corrections to the Annual Review of Psychology corrections:Steven Regeser Lopez and Peter J. GuarnacciaCultural Psychopathology: Uncovering the Social Worldof Mental IllnessAnnu. Rev. Psychol. 2000, Vol. 51: 571–598.http://psych.annualreviews.org/errata.shtml

Ann

u. R

ev. P

sych

ol. 2

002.

53:4

35-4

62. D

ownl

oade

d fr

om a

rjou

rnal

s.an

nual

revi

ews.

org

by U

nive

rsity

of

Stir

ling

on 1

1/03

/05.

For

per

sona

l use

onl

y.

Date post:	10-Apr-2020
Category:	Documents
Upload:	others
View:	1 times
Download:	0 times

John C. Crabbe - University of Stirling · 2006-05-29 · John C. Crabbe Portland Alcohol Research...

Documents