CA2074519C - Allelic association of the dopamine (d2) receptor gene in compulsive disorders such as alcoholism - Google Patents

Abstract

In an important embodiment, the present invention concerns a method for diagnosing compulsive disease predisposition of an individual. The method comprises initially obtaining a DNA sample of said individual and then determining the presence or absence of a particular human D2 receptor gene allele in said sample.
Detection of said allele in the sample is indicative of predilection to compulsive disease. A most preferred embodiment is to detect a susceptibility to alcoholism, particularly be-cause said allele has been found to be present in a majority of clinically diagnosed alcoholics. The human D2 receptor gene A1 allele is most preferably detected in said sample.

Description

~, WO 91/12339 PCT/US91/0085s 20 7~5~

ALLELIC ASSOCIATION OF THE HUMAN DOPAMINE (D2) RECEPTOR GENE IN COMPUhSIVE DISORDERS
SUCH AS ALCOHOLISM
The present invention relates to the first molecular genetic evidence, through the use of RFLP analysis, that an allele in the human dopamine (D2) receptor gene is more significantly associated with human brain tissue from alcoholics than with brain tissue obtained from non-alcoholics. The occurrence of this disease-associated polymorphism has a high predictive value in the classification of, at least, one probable subtype of alcoholics.
The identification of a genetic marker that is closely linked to alcoholism means that the gene's inheritance can be followed, leading to simple tests for diagnosing carriers and future disease victims, and potential gene therapy.
The tendency of certain individuals to display com-pulsive behavior patterns is well. known and includes individuals with an excessive desire for substances classed as psychoactive drugs including, but not limited to alcohol, opiates, and food. ~Thether alcoholism is a psychiatric illness or a biological disease has been a controversial question, but there is some agreement that there are probably similar biochemical mechanisms for WO 91/12339 PCT/US91/0085~
_2_ 2~"~!~ il°~
alcohol and opiates in terms of behavioral and pharmacological activities. (1) Some authors believe that dopaminergic cells are implicated in the rewarding action of alcohol (23), opiates (2) and cocaine (2). In contrast, others (3) argue that at least alcohol/opiates/cocaine and alcohol reinforcing effects are mediated primarily by nonadrenergic and not dopaminergic systems in the brain.
Whether or not multiple systems exist, the question of several parallel reward mechanisms, or a very few, even one, is yet to be fully resolved. The cause and effect of compulsive behavior diseases, including alcoholism, appears to be biogenic. Regardless of the number of systems involved, the ability to identify an allelic gene segment associated with specific compulsive behavior is a significant step forward in developing predictive tests for compulsive behavior patterns.
Alcoholism is a major and devastating health problem with an unknown etiological basis. The question of whether environment or heredity is the prime determinant for the development of alcoholism continues to receive extensive attention throughout the world, and has recently involved the Supreme Court of the United States (4).
However, family, twin, and adoption studies (5) are pointing to genetic factors as significant contributors to alcoholism. These studies also demonstrate that other forms of mental illness such as schizophrenia and other major psychoses are not found at higher frequencies in families of alcoholics compared with the general population. This would suggest that alcoholism is a primary disease, although many disagree.

.~.WO 91/12339 PC1'/US91/00855 Alcoholism currently is viewed as a heterogeneous entity arising from a combination of biopsychosocial factors (6). In regard to biological factors, an extensive literature reveals a wide range of potential physiological (7) and biochemical (8) markers in the risk for alcoholism. Moreover, family pedigree linkage analysis has implicated chromosomes 4, 6 and 11, but not specific gene markers, in the genetic risk for alcoholism (9) .
Restriction Fragment Length Polymorphism (RFLP) offers a powerful molecular genetic tool for the direct analysis of the human genome to determine elements that provide predisposition to genetic diseases (10). This technique has been used to demonstrate a structural mutation in the gene that codes for an enzyme involved in alcohol metabolism (aldehyde dehydrogenase) which leads to the loss of this enzyme's ability to metabolize acetaldehyde. This altered gene is prevalent among Orientals (1l) and may explain the well-known alcohol-flush syndrome as a protective factor in this population.
However, no specific gene abnormality has been identified thus far which could regulate alcohol-seeking behavior, or is associated with alcoholism in humans.
Numerous studies indicate that, in animals, genetic control of neurotransmitter synthesis, metabolism, regulation, and receptor activity mediates reward in the meso-limbic circuitry of the brain (12), as well as drug (era., ethanol) - seeking behavior (13). In this regard, the dopamine2 (D2) receptor has been implicated as a prime target site in the N_. accumbens and hippocampal CA1 cluster cells of the brain reward system (14). Three major dopaminergic systems in the human brain have been identified. The nigrostriatal i.~ involved in the initiation and execution of movement; the WO 91 / 12339 PCT/US91 /0085=
_4_ ' ,.
2U74~19 tuberoinfundibular is responsible for the regulation of peptide secretion from the pituitary; and the mesolimbic tract controls emotional stability and affect. Mediating these effects of dopamine are two receptor subtypes, D1 and'D2~; each of which is coupled to different second ines'senger systems. The D1 receptor has been implicated in the sleep disorder, insomnia. Most recently, a D3 receptor has been found (47) and is also implicated in limbic system function.
Important clinically relevant studies on the pharmacology of D2 receptors indicated that antipsychotic drugs display high affinities for the receptor. Other work suggested that the DZ receptor is involved in movement disorders, i.e., Parkinson's disease and tardive dyskinesia, tumors of the pituitary, and compulsive disease.
A cDNA encoding for rat dopamine (D2) receptor has been isolated (15). This receptor has been implicated in the pathophysiology of certain diseases, including drug addiction. The same laboratory localized the receptor gene to chromosome 11 (19). Partial sequence analysis revealed that the genomic clone lambda-hD2G1 (~lhDZGl) contains the last coding exon of the DZ receptor and 16.5 kb of 3-prime flanking sequence. When this clone was hybridized to human metaphase chromosomes and DNA from rodent-human hybrid cells, the data were consistent with a single human dopamine D2 receptor gene which mapped to the q22-q23 region of chromosome 11. This previous work provides a research tool to begin a molecular analysis of the human D2 receptor in alcoholism.
Access to sequence variation in the human genome now allows construction of genetic linkage maps through the technique of RFLPs (restriction fragment length polymorph-~~WO 91 / 12339 PCT/ US91 /00855 isms). This technique provides probes which are isolated from chromosome specific phage libraries constructed to contain some portion of human DNA (16). With this tool in hand, the analysis of human gene segments is possible.
5 The identification of an apparent. gene abnormality in the tissue of alcoholics is an important advance in the art and of potential value in objectively identifying individuals who are genetically predisposed to alcoholism.
The need for differential diagnosis and the ability to l0 identify genetic predisposition for susceptibility to compulsive diseases such as alcoholism has been recognized at the national level (17).
In the present invention, th.e DNA probe ~lhD2Gl, a 1.6 kb fragment of ~lhDZGl, and likely a 30 bit (base pair) fragment of JlhD2G1 effectively visualize the human dopamine (D2) receptor gene. This permits evaluation of polymorphisms on the gene in a region close to the gene which could modify the function of the gene as a valuable predictor of alcoholism or other compulsive disorders.
In an important embodiment, the present invention concerns a method for diagnosing susceptibility to compulsive disease. The method comprises initially obtaining DNA from an individual and determining the presence of a particular human D2 receptor gene allele.
Detection of said allele's presence in the sample is indicative of susceptibility to compulsive disease, particularly because said allele has been found to be present in a large majority of clinically diagnosed victims of compulsive disease. In a most preferred embodiment, the method is used to indicate a susceptibility to alcoholism, particularly because said allele has been found to be present in a large majority of clinically diagnosed severe alcoholics. The human D2 WO 91/12339 PCT/US91/0085' receptor gene A1 allele is most preferably detected in said DrT,A.
The allele is readily detected by digesting DNA with Tag I restriction endonuclease and detecting the presence or absence of human D2 receptor gene A1 allele in the hydrolyzed DNA. This method involves separating, preferably according to their size, restriction fragments from the digest. The separated fragments are then probed with labelled lambda-hDZGl or fragment thereof such as a Bam H1-generated 1.6 kb fragment or a synthetic 30 bit fragment of .thD2G1 (30 base pair oligomer), for example, to specifically detect the presence of human DZ receptor gene A1 allele. The presence of a 6.6 kb fragment in said sample, said 6.6 kb fragment being representative of the human DZ receptor gene A1 allele, is indicative of suscep-tibility to compulsive disease.
An object of the invention is to provide a safe and reliable method to diagnose alcohol and/or other drug risk at the prenatal and postnatal level.
The above described method may also be of value in detecting the predisposition toward other compulsive-obsessive behavior patterns including but not limited to overeating or substance abuse such as seen with nicotine, narcotics and other abused drugs. In a particular embodiment, the above described method may also be used to detect a susceptibility to attention deficit disorder with hyperactivity (ADDH) in children. In a more particular embodiment, this method may be correlated to the presence of ADDH and/or to detect susceptibility to alcoholism with greater reliability as well as other genetic diseases such as Tourette Syndrome (48, 49, 50). This genetic disorder has been linked to a severe form of alcoholism possibly "'" WO 91 / 12339 PCT/US91 /OORS:

caused by a disinhibition of the limbic system (48, 49, 50) .
Figure 1(A) shows the hybridization pattern of Tai I-digested DNA isolated from a heterozygous individual. The hybridization probe is the full-:Length lambda-hdZGl which hybridizes with the 6.6 kb fragment associated with the A1 allele plus the 3.7 kb and the 2"9 kb bands associated with the A2 allele. In addition" the probe also hybridizes with two constant bands, 10.5 and 2.3 kb in length.
Figure 1(B) shows the hybridization patterns of Tai I-digested DNA isolated from a nonalcoholic (homozygous for the A2 allele) and an alcoholic (heterozygous for the A1 and A2 alleles) individual. The hybridization probe is a 1.6 kb Bam H1 fragment isolated from lambda-hDzGl. Note that the smaller probe does not hybridize to the 2.9 and 2.3 kb Tag I fragments of the human dopamine D2 receptor gene. Lambda-hDZGi is a genomic F~iBL 3 phage containing approximately 18 kb of human leukocyte DNA.
Figure 1(C) schematically shows the production of the 1.5 kb probe for the dopamine D2 receptor gene obtained from chromosome 11.
Figure 2 shows a Southern blot analysis of human DNA
from brain tissue grouped according to presence of A1 allele (6.6 kb band) of the human dopamine DZ receptor (D2HR) gene. Predictive value in our sample size in correctly identifying alcoholics is 77% (Chi-square=9.32, DF=1, P=0.002). Note that the A2 allele (band 3.7 kb) is missing from samples 8, 13, 24, 27, and indicating that these DNAs are homozygous for the A1 allele.

g _ Figure 3 shows a Southern blot analysis of human DNA
from brain tissue grouped according to absence of A1 allele of the human dopamine D, receptor gene. Predictive value in correctly identifying nonalcoholics i.s 720 (Chi-square=7.41, DF=1, P=0.002).
Figure 4 shows saturation curves of [BH] spiperone binding to D2DR in caudates of (A) a nonalcoholic subject homozygous for the A2 allele (A2Aa?) and (B) a nonalcoholic subject heterozygous for A1 and A2 (A1A2). Data points are means of duplicate determinations. L)etails for the binding studies are given in Example 2. Figure 4AA and 4BB:
Scatchard analysis of [3H] spiperone binding.
Figure 5 shows the 1.6 kb probe doublet (Doublet 1.73 kb = 1.6 kb) which results from the ciigestion of ~,hD2G1 with BamHl. Initially this probe was thought to consist of 1.5 kb (personal communication D.K Grandy). After sequencing the doublet was estimated as 7L.73 kb. Subsequent separation of the doublet into clone 9 and clone 16 singlets, revealed, through sequencing, an actual size for the clone 9 fragment of 1.6 kb. Figure 7 illustrates the fragments from clone 9 and _L6. For purposes of this application, reference to the 1.73 kb probe is equivalent to the 1.6 kb probe. 'Twenty ~zg (ug) of t:he parent clone, ~.hD2Gl, was digested with ~L8 unite of BamH1 for two hours at 37°C in Buffer C (IBI), loaded onto a 0.8o agarose gel, run overnight at 23 volts, and vi:~ualizecl wit:h ethidium bromide staining. The adjacent gel indicate: DNA fragments of known molecular weight as st=andards .
Figure 6 shows th~? hybridization patterns of the 1.6 kb probe after doublet separ_at:ion by subcloning. The 1.6 kb fragment (doublet) was ligated by following the procedure given by Sea Plaque GTG agarose and cloned as ~~ WO 91/12339 PCT/US91/00855 described in Example 4. DNA from antibiotic resistant clones was digested with Ana H1 and separated by gel electrophoresis as in Figure 5 to identify the presence of the 1.6 kb fragment. Other dige.aions of clone DNA with different endonucleases (Hinf-1, ~S,g-1, ~aq,-1, BamHl, and Hind III) revealed differences in DNA patterns following separation on gel electrophoresis:. In this manner, clones 9 and 16 were selected and grown.. The 1.6 kb fragment from BamHl-digested DNA from clone 9 and clone 16 was radiolabeled and hybridized with fag I-digested human genomic DNA using the same procedure as in Figure 1 above.
The genomic DNA was from two individuals, one homozygous for the AZ allele, indicated here: as A2, and one heterozygous, AlA2, indicated as Al.
Figure 7 shows the nucleotide sequence and the hybridization of the 30 Base Pair Oligomer (30 bit fragment of xhD2G1) to the 1.6 kY> probe obtained from Clone 9. The 30 base pair oligomer was custom synthesized to correspond to the last 30 base pairs of the cDNA of the 7th exon of the DZ receptor. This oligomer was radiolabeled and hybridized to the 1.6 kb fragment from clone 9 but not clone 16.
LI8T OF ABHRE'NIATIONB
IBI International Biotechnologies, Inc.
AlA2 heterozygous for the A1 and A2 allele A2A2 homozygous for the A2 allele lambda-hD2G1 =
JIhDZGl 18 kb probe used to detect A1 allele.
dCTP deoxycytosine triphosp~hate SSC saline solution containing sodium citrate SSDNA Salmon sperm DNA
DZDR Dopamine D2 Receptor DSM -IV-R (criteria) diagnostic standard manual Bm"~ Number of receptors WO 91/12339 PCT/US91/0085~

Kd Dissociation constant TE Tris-EDTA Buffer kb ='Kb Kilobase or Kilobit v ' ~P Probability pM picomolar ( 10-1ZM) fmol femtomole ( 10-15 mole) RFLP Restriction Fragment Length Polymorphism ug = ~g microgram ETOH Ethanol ERAMPhE I
The present invention demonstrates the first allelic association, namely of the dopamine DZ receptor gene, with alcoholism. DNA, from matched alcoholic and nonalcoholic brain samples, was digested with restriction endonucleases and probed with the human D2 receptor gene (lambda-hDZGl).
The presence of the A1 (6.6 kb band) allele of the dopamine (D2) receptor (also abbreviated D2DR) gene correctly predicts 77% of alcoholics, and its absence is predictive of 72%_of nonalcoholics. The polymorphic pattern of this receptor gene suggests that the abnormality in at least one form of alcoholism is located on the q22-q23 region of chromosome 11 with a co-dominant Mendelian mode of inheritance. The allelic association of the dopamine (D2) receptor gene with alcoholism has a high predictive value in the classification of one probable alcoholic subtype. This subtype may represent a virulent form of alcoholism.
Seventy frozen brain samples were thawed and processed for high molecular weight genomic DNA. The tissue was homogenized in 0.25 M sucrose and a nuclear pellet prepared. Next the pellets were incubated at 37'C
for three hrs in 0.05% SDS and proteinase K and the DNA
was extracted with phenol, followed by extraction with chloroform:isoamyl alcohol. The_DNA was then spooled out, '~°" WO 91/12339 PCT/US91/00855 . -11 2~ ~45I9 washed with ethanol, and stored in TE at 4'C. When all the DNfis were isolated, aliquots (20 ug DNA) were digested separately with four different restriction endonucleases (i.e., Tai I, Mso I, Eco RI and fist I) at approximately two units enzyme/ug DNA, run on agarose gels, Southern-transferred to nylon membranes, and hybridized with different DNA probes using standard methods (18). In the present experiment, the DNA samples, after digestion with the four restriction enzymes, were hybridized with a number of probes involved with either ethanol metabolism or neurotransmitter regulation of reward, including the human dopamine D2 receptor gene (lambda-hD2G1) to determine polymorphism. This procedure used to generate the human DZ receptor gene (~lhD2G1) is as follows: A
human genomic library was screened with the rat dopamine D2 receptor cDNA. The human geno;mic library (Clonetech) in EMBL3 was prepared from normal. male leukocyte DNA and screened with a nick-translated probe containing portions of the cDNA for the coding region of the rat D2 receptor.
One clone, lambda-hD2Gl, with an 18 kb insert was identified and characterized. This clone was found to contain the entire 3' coding exon, the polyadenylation signal, and approximately 16.4 kb of noncoding 3' sequence. Twenty ~cg of the parent clone (lambda-hDZGl) was digested with 48 units of $~n_ Hl for two hrs at 37'C
in Buffer C (IBI), loaded on 1% agarose gel (Sea Plaque), and run overnight at 23 volts. When digested under these conditions, several fragments were generated, including a 1.6 kb fragment and a second band of 1.5 kb, made up of two 1.5 kb fragments. The 1.5 kb band was cut, heated to 68'C, diluted by a factor of three with TE buffer, and stored at 4'C. The diluted gel was placed in boiling water for three min and then incubated for ten min at 37'C. A 25 u1 aliquot was then removed and labelled to a specific activity of 1 x 109 pm/~g with (32P]-dCTP
according to the oligolabeling k:it (Pharmacia). The 50 u1 - 1.2 -(microliter) incubation mixi~ur_e was then chromatographed through a G-50 SEPHADE:X column an~~ the eluant used for hybridization. The Taa I digested DNAs were then transferred to Nytran (TM) rnembra:nes and hybridized with the labelled insert in 50o formamide, 5 x SSC, 1 x Denhart' s, 20 mM NaH2P04, 200 ~~g/ml of_ SSDNA, 0 . 1 o SDS, 10 0 dextran sulfate, 0.250 dry milk, and incubated overnight at 42°C. The filters were then washed 2 x with SSC, O.lo SDS
at 55°C, and radioautographed overnight. The only endonuclease to show polymorphism with lambda-hD2Gl was T.aa I (vide infra).
In previous studies (19), where the lambda-hD2G1 was used to probe digests of human ge:nomi.c DNA, it was found that only Tai I, but not dige:>ts from 30 other endonucleases, revealed a frequent two allele RFLP. Allele A1 = 6.6 kb and allele A2 - 3..7 ~ 2.9 kb with constant bands at 10.5 and 2.3 kb. Allele frequencies were measured in 43 unrelated Caucasians and calculated to have a frequency of A1 = 0.24, A2 =- 0.76. C:o-dominant Mendelian inheritance was observed in four informative families with a total of 39 children. Addit=ion~slly, the human dopamine receptor gene wa:~ mapped on t=he q22-q23 region of chromosome 11 (19).
Figure 1 illustrates the polymorphic: pattern of the human dopamine ( Dz ) receptor gene . Figure 1A depicts the polymorphic allelic pattern for t:he lambda-hD2Gl gene clone.
Figure 1B shows the allelic pattern using a BamHl 1.5 kb subclone which reduced overal7_ ba~~kground and still was informative as to the :presence of alleles A1 and A2.
However, the smaller probe did not hybridize to the 2.9 and 2.3 kb Taa I fragments of the hum~~n dopamine (D2) gene. For illustrative purposes only, the p~~lymorphic patterns are labelled according to their highly significant allelic association with either alcoholics (A1 .~~.w~ 91/12339 PCT/US91/00855 allele) or nonalcoholics (the absence of A1 allele), respectively labelled A1/A2 and A2/A2. The schematic production of the 1.5 kb ("1.5 kb''' illustrated here later was determined to actually be "1.n kb" - See Example III
for an explanation) subclone probe from chromosome 1l is shown in Figure 1(C).
Table 1 illustrates the polymorphic pattern of the dopamine DZ receptor gene with DNA obtained from alcoholic and nonalcoholic subjects following three independent hybridizations. The A1 allele is associated with 24 of 35 (69%) known alcoholics, but it associated with only 7 of 35 (20%) nonalcoholics. In contrast, the absence of the A1 allele is associated with 28 out of 35 (80%) of nonalcoholics and with only 11 of 35 (31%) alcoholics.
The proportion of the presence of the Al allele to the absence of this allele is significantly different in alcoholics as compared to nonalcoholics (Pates Chi-square corrected for continuity equals to 14.8, DF=l,P<0.001).

WO 91/12339 PCT/US91/0085~

Table 1. Polymorphic pattern of the 1.5 kb fragment ( lambda-hD2G1 ) of the dopamine DZ receptor gene in nonalcoholic and alcoholic brain tissue.
DNA Type Absence of A1 Allele Presence of A1 Alleles Control (N = 35) 28 (80.0)2 7 (20.0) Alcoholic (N = 24) 11 (31.4) 24 (68.6) 15 The race of subject populations is an important determinant in allelic patterns. Recently, Kidd et a1.3 reported that at some loci, alleles that are infrequent in Caucasians are common in other populations. As the present brains were derived from both Caucasians and 20 Blacks, the allelic frequency of the dopamine D2 receptor gene was analyzed in these two racial groups.
Table 2 illustrates the Polymorphic Pattern of the Dopamine DZ Receptor Gene with DNA obtained from alcoholic and non-alcoholic Caucasians and Blacks following three 25 independent hybridizations.
lAi allele = 6.6 kb ZValues in parenthesis represent percent of nonalcoholics or Alcoholics showing absence of the A1 allele. The proportion of the presence of the A1 allele to the absence of this allele is significantly different in alcoholics compared to nonalcoholics (Pates Chi-square corrected for continuity equals 14.8, Df=1, P<.0001).
3K. K. Kidd a al., Genome Mapping and Sequencing (Cold Spring Harbor Meeting, NY, 1989), pp. 66.

..n WO 91/12339 PCT/US91/OO85s -15- 20 74 ~~ y Table 2. Polymorphic Pattern of the Dopamine D2 Receptor Gene (lambda-hD2G1) in Hrain Tissute of Nonalcoholics and Alcoholics.
Absence of Prasanes of DNA Source A1 Allsls A1 Allale4 Nonalcoholic (N~35 28 (80x)5 7 (20Z) Alcohol c (N~35) 11 (31x) 24 (69Z) Nonalcoholic Caucasians (N=24) 20 (83Z) 4 (17Z) Alcoholic Caucasians (N~lI) B (36Z) 14 (64Z) Nonalcoholic Hlacks (N~11) 8 (73Z) 3 (27Z) Alcoholic Hlecka (N~13) 3 (23Z) 10 (77Z) 15 The A1 allele is found to be associated with 14 of 22 (64%) Caucasian alcoholics, but it associated with only 4 of 24 (17%) Caucasian nonalcoholics (16). The proportion of the presence of the A1 allele to the absence of the allele in Caucasian alcoholics compared to Caucasian 20 nonalcoholics is highly significant (Pates Chi-square=8.75, DF=1, P=0.003). In Blacks, the Al allele is associated with 10 of 13 (77%) alcoholics, but it associated with only 3 of 11 (27%) nonalcoholics. The proportion of the presence of the A1 allele to the absence 25 of this allele is also significantly different in Black alcoholics compared to Black nonalcoholics (Pates Chi-square=5.92, DF=1, P=0.015). Thus, in the present sample, the results favor the view that A:l allelic association is based on whether or not an individual is an alcoholic, 30 rather than the individual's racial background.
4A1 allele = 6.6 kb.
sValues in parentheses represent percent of nonalcoholics or alcoholics showing absence or presence of 35 the A1 allele.

"'~ "' "'339 PCT/US91/0085 To determine the relationship between alcoholism and the A1 allele controlling for race, we used the Mantel-Hazenzel test was used. This test evaluates the relationship between two variables, while controlling for a third. Since Chi-square=14.20 with P<0.001, there is a highly significant association between alcoholism and the A1 allele was found, even after controlling for race.
In the present sample, this test also suggested that the odds ratio of finding the A1 allele in alcoholics is 8.8 times as large as that for nonalcoholics.
Figures 2 and 3 show the samples grouped according to whether or not the A1 allele was present. This grouping allowed a classification of samples based on their unique allelic association with alcoholism. Figure 2 represents 31 brain samples which possess the A1 allele (6.6).
Twenty-four out of 31 DNAs that had the A1 allele were from alcoholics. This suggests, that in our sample, the predictive value of this test in correctly identifying alcoholics is 77%. Figure 3 represents 39 brain samples which did not possess the A1 allele. Since 28 out of 39 samples did not have the A1 allele and were from nonalcoholics, this suggests that the predictive value of this test in correctly identifying nonalcoholics is 72%.
To evaluate the hypothesis that the presence or absence of the Ai allele was distributed between the alcoholic or nonalcoholic groups at other than equal probabilities, a single sample Chi-square analysis with assigned expected value of 0.50 was used. When this expected value was assigned, Chi-square analysis revealed no significant difference from the expected probability for the A1 allele in nonalcoholics (Chi-square=0.47, DF-1, P=0.50). In contrast, observed frequencies were significantly different from the expected probability of .~ WO 91/12339 PCT/US91/00855 2r74~~9 0.25 for the A1 allele in alcoholics (Chi-square equals 35.4, DF=1, P<0.001).
To evaluate the hypothesis that the presence of the A1 allele was distributed between the alcoholic or nonalcoholic groups at other than equal probabilities, a single sample Chi-square analysis with assigned expected value of 0.25 was used to approximate the frequency of the A1 allele in the general population. When this expected value was assigned, Chi-square analysis revealed no significant difference from the expected probability for the A1 allele in nonalcoholics (Chi-square=0.47, DF=1, P=0.50). In contrast, observed frequencies were significantly different from the expected probability of 0.25 for the A1 allele in alcoholics (Chi-square=35.4, DF=1, P<0.001).
Of the total alleles in the present sample, the frequency of the A1 allele was 25% and that for the A2 allele was 75%. In the samples of nonalcoholics, A1 and A2 allelic frequencies were 13% and 87% respectively. The allelic frequencies in the samples of alcoholics were:
A1=37% and A2=63%. The frequency of the A1 allele in samples of nonalcoholics and alcoholics were significantly different (Yates Chi-square=9.75, DF=1, P=0.002).
To determine ability to correctly classify the alcoholic or nonalcoholic in this. sample, according to the presence or absence of the A1 allele distributed between the two groups at a better than Chance probability, a Chi-square analysis with assigned expected value of 0.50 was used. Observed values were significantly different from expected probability for the A1 allele (Chi-square=9.32, DF=1, P=0.002) and for the absence of the A1 allele (Chi-square=9.26, DF=1, P=0.002). These findings, taken u'~ ~' '12339 PCT/US91100855 20'7 4519 together, suggest a strong allelic association of the dopamine D2 receptor gene with alcoholism.
To determine the association of other putative genes .kith alcoholism, a number of additional candidate probes ~~were used. Unlike lambda-hDZGl, none of these probes revealed a polymorphic pattern of association with alcoholism.
Nuclear DNA was isolated from the matched brain samples as previously described for lambda-hD2G1 probe.
Twenty micrograms of DNA was digested with one of the four restriction endonucleases. The resulting DNA fragments were separated according to size by electrophoresis in 1%
agarose gel, transferred to nitrocellulose membranes, fixed, and hybridized with phosphorus 3ZP-labelled probes.
Washing of filters and autoradiography were carried out as described previously in this paper. A number of probes were employed, including alcohol dehydrogenase (pADH/3), protein kinase-C (phPKC), carboxypeptidase-A (CPA), pro-enkephalin (pHPE9), tryptophan hydroxylase (TPH479), tyrosine hydroxylase (BTH4), monoamine oxidase B (MAOB), transferrin (TF) and others (See Table 3). Evaluation of the data (Table 3) revealed that none of these DNA probes utilizing four restriction endonucleases, (which endonucleases, to date, are responsible for about two-thirds of all known polymorphisms) are associated With alcoholism. The cDNA probe for alcohol dehydrogenase, an enzyme involved in the metabolism of alcohol, displays a polymorphism using MSp I, but the polymorphism is not linked to alcoholism. The cDNA probe for transferrin, a protein involved in hemoglobin synthesis, displays polymorphism using Eco RI, but again, this polymorphism is not associated with alcoholism. Other probes used were:
protein kinase-C, involved in second messenger coupling mechanisms for neurotransmitters; carboxypeptidase-A, ~. WO 91 /12339 PCT/US91 /00855 involved in the metabolism of the opioid peptide enkephalin; pro-enkephalin, the precursor protein for the synthesis of enkephalin; the enzyme tryptophan hydroxylase, involved in the regulation of serotonin synthesis; tyrosine hydroxylase, the rate-limiting enzyme in the synthesis of dopamine; and transferrin, a protein involved in hemoglobin dynamics. This latter group of probes as well as others displayed no polymorphism with Tai I, Mso I, Eco RI, and Pst I restriction endonucleases.
Thus the only probe that showed polymorphism associated with alcoholism was lambda-DZG1.

WO 91/12339 PCT/US91/0085~
207419 . -20-Table 3. Bvaluatioas of Polymorphisms of DNA Probes with Various Endoaucleases Alcoholism DNA Probe Msp Eco a I Pst I Xbal Association II RI

Alcohol Dehydrogenase(pADH/3)yes no no no - no Protein Kinase-C(phPKC) no no no no - no Tryptophan Hydroxy,lase(TPH479)no no no no - no Pro-enkephalin(pHPE9) no no no no - no Monoamine Oxidase(MAOB) no no no no - no Carboxypeptidase A(CPA) no no no no - no Transferrin(TF) no yes no no - no Tyrosine Hydroxylase(BTH4) no no no no - no Choline Acetyl Traneferase no no no no - no (Chat) Serotonin (SHTlA) receptor no no no no - no (GZ1) Catecholamine Receptors 22 no no no no - no B1 no no no no - no B2 no no no no - no GABA Receptors 21 no no no no - no 24 no no no no - no B1 no no no no - no Dopamine B-Hydroxylase no no no no yes no Dopamine (D2) Receptor lambda-hD2G1 no no yes no - yea ~., WO 91/12339 PGT/US91/00855 20?~~1:9 Over the past three decades, research concerns with the interaction of genetic and environmental factors in the development of alcoholism shows that the risk for this behavior is determined by genetic: as well as by environ-s mental factors (20). However, the conclusion that there is a significant genetic component to alcoholism has led to the realization that individuals who are at risk of becoming alcoholic, because of inherited factors, are biologically different from individuals who have few or no inherited factors that predispose them to alcoholism.
This notion has stimulated an extensive search for alcoholism genes (alcogenes) or markers to identify individuals at increased risk for alcoholism, a concept elaborated from studies of inbred strains of mice, C57 and DBA, with a differing predilection to alcohol (21).
It is theoretically possible that the polymorphism of the dopamine (D2) receptor gene in the brains of alcoholics is due to alcohol-induced alteration in DNA
(22); hence, the polymorphism observed might be a consequence of prolonged alcohol consumption by the alcoholic and thus represent a state marker instead of a trait marker. This possibility is unlikely, given the fairly wide prevalence (24%) of the A1 allele in the general population. Moreover, the presence of the A1 allele and its co-dominant Mendelian inheritance (19) in alcohol-naive children indicate that alcohol per se was not responsible for this genetic variation. It is of further interest to note that naive inbred alcohol-preferring rats show a significantly lower dopamine (D2) receptor binding activity than naive alcohol-avoiding rats (23), suggesting an abnormality in this gene or in its expression. These observations support the idea that the allelic association of the dopamine (D2) receptor gene, or a gene close to it, in brain tissue of alcoholics is a WO 91/12339 PCT/US91/0085~
20'~451~ -22-likely candidate trait marker for, at least, one important subtype of potential alcoholism.
Given the evidence that children of alcoholics are at a greater risk of developing alcoholism than children of nonalcoholics (24), it may be predicted that the prevalence of a candidate trait marker would be significantly greater in a population of subjects who have a positive rather than a negative family history of alcoholism.
In the present sample, derived from 70 deceased individuals, a strong association between alcoholism and the A1 of a Taa I polymorphism close to the dopamine Dz receptor gene has been found. That this association prevailed in a subsample of Caucasians and Blacks raises interesting questions about the prevalence of the A1 allele in other samples of alcoholics. It is, however, important to note that a large majority of alcoholics in the present study had experienced repeated treatment failures in their alcoholic rehabilitation and whose cause of death was primarily attributed to the chronic damaging effects of alcohol on their bodily systems. It is possible then that the A1 allele found in this study may be associated with a particular subgroup of virulent alcoholism. Besides these molecular genetic studies, we have also carried out, in the same brain samples as above, the actual characteristics of the dopamine D2 receptor using [3H]spiperone (a dopamine [D2] receptor antagonist ligand). The data show that the affinity of the dopamine (D2) receptor ligand is significantly different in subjects having the A1 allele compared to those having the A2 allele. Thus, the evidence, put together, shows not only that a strong association is found between the A1 allele and alcoholism, but that the A1 and A2 allele ,~d WO 91/12339 PCT/US91/00855 23 2074~I9 express themselves in different dopamine (D2) characteristics in the brain.
Unlike genetic diseases such as Auntington's chorea and cystic fibrosis (l0), where a single gene is responsible for its expression, the heterogeneous nature of alcoholism may not allow for the generation of a single marker that can identify all individuals at risk.
Given that there are various subtypes of alcoholics (51, 52), it would have been surprising if a 100%
association was found between the A1 allele and alcoholism. In this regard, the 31% of alcoholics in this study which did not associate with the dopamine (D2) receptor gene polymorphism suggests some interesting possibilities: 1) environmental (25) rather than genetic factors contributed to their alcoholism; 2) other genes may be critical for the predisposition and subsequent expression of alcohol-seeking behavior. This possibility is intriguing, since it suggests that gene-specific subtypes of alcoholism could now be identified through RFLP analysis and provide the basis for multiple etiologies; and 3) there may be only partial linkage disequilibrium between the RFLP and the gene responsible for the disease. This could occur because of occasional crossover between marker and gene.
Support for alcoholism subtypes can be found in various neurochemical hypothesis, including: 1) individual differences in nerve cell membrane sensitivity to ethanol (26); 2) inherited variations in the sensitivity of sodium-potassium A,TPase inhibition to ethanol (27); 3) inherited variations in neurotransmitter release and uptake systems involved in a reward cascade of events (28); 4) inherited variations in the production of abnormal amounts of tetrahydroisoquinolines (29); 5) WO 91/12339 PCT/US91/0085s 2~0 7 4 5 ~.-9 -24- ' inherited variations in the neuroadaptive mechanisms for reinforcing certain behaviors (30); and 6) inherited variation in second messenger response coupling mechanisms (31) .
EBAMPLE II
;The allelic association of the human D2 dopamine receptor (D2DR) gene with the binding characteristics of the D2DR was determined in 66 brains of alcoholic and nonalcoholic subjects. In a blinded experiment, the DNA
from these samples was treated with the restriction endonuclease TaaI and probed with a 1.5 kb fragment from a BamHl digest of ~lhD2Gl. As discussed in Example I, .lhD2G1 contains the entire last exon of the D2DR gene, the polyadenylation signal, and approximately 16.4 kilobases of noncoding 3' sequence of the human D2DR gene. The binding characteristics (Kd, binding affinity and Bmax, maximum number of binding sites) of the D2DR were determined in the caudate nucleus from these brains using (3H)spiperone, a dopamine receptor (D2DR) antagonist ligand. Log Kd was significantly lower in alcoholic compared to nonalcoholic subjects. Moreover, a linear relationship in reduced Bmax was found respectively in A2A2, AlA2, and A1A1 allelic subjects. In individuals with the A1 allele, where a high association with alcoholism was found, the maximum number of D2DRs (Bm"~) was significantly reduced when compared to the Bm~ in individuals with the A2 allele (53).
METHODS
Brain samples Tissues from 33 alcoholic and 33 nonalcoholic subjects was obtained from the National Neurological Research Bank at the VA Medical Center, Wadsworth, Los Angeles. Frontal grey cortex and caudate nucleus were removed from the brain at autopsy by a neuropathologist .,~ WO 91/12339 PCT/US91/00855 -2 5- . ~~~0 ~~4 ~'I 9 and immediately frozen at -70°C until used. The 66 brains analyzed consisted of the 70 that: were previously studied in Example I (32); 4 caudates were not available. The ages (average ~ SEM) of the alcoholics and nonalcoholics, respectively, were 50.4 ~ 2.3 years and 53.2 ~ 2.6 years.
The racial distribution of alcoholics included 21 whites and 12 blacks, and there were 24 white and 9 black nonalcoholics. The sex distribution of alcoholics included 30 males and 3 females, and there were 29 male and 4 female nonalcoholics. The autolysis times (average ~ SEM) of the alcoholics' and nonalcoholics' brair. samples were, respectively, 23.0 ~ 1.5 hours and 22.6~ 1.7 hours.
alcoholic (Alcohol Dependence and Alcohol Abuse, using DSM-III-R criteria [33]) and nonalcoholic diagnoses were made independently by two trained psychiatrists, through examination of medical and autopsy records, interviews of treatment center personnel and relatives and alcohol consumption data. There was a 1010% concordance in diagnosing alcoholic and nonalcoholic subjects between these two assessments. Examination of medical records and/or analysis of body fluids at: autopsy did not reveal any of the subjects to have used neuroleptics. The cause of death included: accidents, gun-shot wounds, myocardial infarction, heart failure, cancer, gastrointestinal bleeding, suicide, and pneumonia. Informed consent was obtained from next of kin to carry out the present study.
DNA Probe The DNA probe, as previously used in Example I (32), was a 1.73 kb band obtained as a doublet from a ,~~nHl digest of a human genomic fragment, ~LhD2Gl, provided by 0.
Civelli. This fragment contains the last coding (7th) exon of the D2DR gene and part of 16.5 kb of 3' flanking sequence (19,34). The 1.5 kb (doublet) probe was labeled by random-priming with [32P]dCTP 1;18) to a specific activity of 1 x 10° cpm/~g.

DNA Isolation and Southern Blot Analysis The 66 frozen brain cortical samples were coded without reference to their group identity (alcoholic and nonalcoholic). They were then thawed and processed for high molecular weight. genom:ic DNA and hybridized by established procedures (18), as previously detailed in Example I (32).
D2 Dopamine Receptor Assay Sixty-six frozen caudate nuclei, of the same brains from which cerebral cortex DNAs were isolated, were also coded without reference to their group identity, and assayed during a one month period for D2DR characteristics.
The frozen samples were routinely ground into a fine powder in liquid nitrogen using a mortar and pestle and stored at -70°C from which small amounts of homogeneous powdered tissue could be used for assays at di_ffer.ent times. A
sample of 200-300 mg of the powered tissue was homogenized in 30 mL of ice--cold buffer (50 mmol/L Tris-HC1, pH 7.4;
120 mmol/L NaCl; 2 mmol/L MgCI,) with a Brinkman Polytron (TM). The homogenate was cent=rifuged at 35,000 x g for 20 minutes. The pellet was resuspended in 30 mL of buffer and again centrifuged at 35,000 x g for 20 minutes. The final pellet was resin>pended i.n 30 mL of buffer. for the binding assay.
D2DR binding was :measured by a ~>light modification of previously established procedures (3_'>-37). Saturation curves (Figure 4) were obtained using 12 duplicate increasing concentrations (:10--1000 pmol/L) of [~H] spiperone (32.4 Ci/mmol; New England Nuclear) i.n buffer containing 50 mmol/L Tris-HC1, pH 7.4; 121) rnmol/L TIaCl; 2 mmol/L MgCl2.
To measure nonsaturable binding, S-(-~)-sulp hide was added to a final concentration of 10 umol/h. Binding was initiated by the addition o:E membrane preparation (250-350 ug protein), and the samples were ..yVO 91/12339 PCT/US91/00855 ~0~4~19 incubated in the dark at 20°C for' 2 hours. Final assay volume was 1 mL. The samples were then rapidly filtered through GF/B glass fiber filters with a Brandel cell harvester. The filters were washed twice with 2 mL of ice-cold assay buffer and placed in scintillation minivials with 4 mL of scintillation fluid (National Diagnostics) for counting. Protean concentrations were determined using bovine serum albumin as the standard (38). Maximum number of binding sites (Bmu) and equilibrium dissociation constant's (Kd) were estimated using the weighted nonlinear least-squares curve-fitting (39) program LIGAND. Data were fit for both one-site and two-site models with the two-site model accepted only if a stitistically significant improvement was obtained over the one-site model.
statistical Analysis The frequency distributions for B~, Kd and Bm,x/Ka were examined for departures from normality. Correlations were estimated for Bm,x with age, Log Kd with Bm,=, and Bm,x/Kd with age. Mean differences between groups for Bm"~, Log Kd and B,m,x/Kd were tested using two factor ANCOVA to determine the statistical significance of the main effects and the interactions of allele and the presence or absence of alcoholism. Measures for Bm,x were covariate-adjusted for and and Log Kd while Log Kd measures were covariate-adjusted for Bm,z. A one-tailed alpha criteria of < 0.05 was used to evaluate all effects upon measures of Bm"~ and Log Kd. In this study, the more powerful one-sided tests were used when the sample mean fell in the expected direction (40). No directional expectations were established for Hm"~/Kd, therefore, alpha criteria were set at alpha = 0.05, two-tailed.
The expression of the A1 allele consists of both homozygote and heterozygote individuals. These two groups 2Q'~ 4519 were labeled AlAl for homozygotes and AlA2 for heterozygotes and were compared with A2A2 homozygote individuals using single factor ANOVA and polynomial tests for: trends of Bm,x, Log Kd and Bm~/Kd.

DNA Analysis In the previous hybridization study in Example I
(32), using TagI-digested human DNA, we have shown that a BamHl 1.6 kb doublet of the D2DR gene (~hD2G1) reduced overall background and was still informative as to the presence of D2DR gene alleles. Three bands were obtained:
a constant 10.5 kb band, a 6.6 kb A1 allele and a 3.7 kb A2 allele. Using the BamHl l.6kb doublet as probe, TagI
digested DNAs from the caudate nucleus of alcoholic and nonalcoholic brains in the present study were subjected to two independent hybridizations. The results are shown in Tables 4A and 4B. Table 4A presents the 29 DNAs that show the presence of the A1 allele (A1+). Twenty-two (76%) of these DNAs associated with alcoholic subjects and 7 (24%) associated with nonalcoholics. Table 4b shows the 37 DNAs that exhibit the absence of the A1 allele (A1-). Eleven (30%) of these DNAs associated with alcoholic subjects and 26 (70%) associated with nonalcoholics. The proportion of the presence of the A1 allele to the absence of this allele is significantly different in alcoholics compared to nonalcoholics (Yates _XZ corrected for continuity =
12.06, df = 1, P < .001). This observation, as expected, is similar to our previous study of 70 brains in Example I(32), of which 66 brains were again probed in the present investigation.
Da Dopamine Receptor Saturation curves, using [3H]spiperone as antagonist ligand and S-(-)-sulpiride to measure nonspecific binding, of each of the 66 caudates studied complied with a single-..CVO 91/12339 PCT/US91/0085s -29- ~0~4~1~
;r~ :.
.r model binding site (39). To eva.uate interassay reliability, replicate determinations, on different days for the same tissue, were also made on a subset of 10 caudates from the present brain samples. Using 100% as the assigned Kd or Bm,= value for the first determination, the second assay revealed Kd or B~ value for the first determination, the second assay revealed Kd = 104 ~ 7.5%
and Bm~ = 103 ~ 4.3%. Paired t-tests showed no significant difference between the two determinations.
to The Spearman rank-order correlation coefficients for Kd and Bm~ were 0.72 and 0.95 respecaively. The two-tailed significance levels for Kd and B~ were P < .009 and P <
.001, respectively. These data i:ogether reveal no statistically significant differences between the two independent assays, suggesting interassay reliability.
Similar good interassay reliability for the binding characteristics of the D2DR in human caudates frozen at postmortem has been observed by others (36).
Figure 4 shows examples of saturation curves of [3H]spiperone and Scatchard analysis in caudate tissue:
(A) non-alcoholic subject with A2A2 allele and (B) nonalcoholic subject with AlA2 allele.
The Ka and Bm"~, for each individual caudate in the present study, are shown in Table 4. Table 4A shows the binding characteristics in A1' samples (presence of the 6.6 kb band) and Table 4B depict:a the values in A1-samples (absence of the 6.6 kb band). The range of values are consistent with reported results for the binding characteristics of the D2DR in human caudate tissue (42,37,41).

WO 91/1?,339 PCT/US91/008~~
~0~45i9 w r vo ov co r r e~~ ~ er o ao ao y cr~d~rr ~r~NO . . .p . .c~~aoNO r~ervo . . .r .p,nrao ~ao ~M
U ~ . . . . . ,n . . . .-, ,n r ~ r ~ . . . . p . . . ap ,.~ op . op . . . .
r, . p m 'O L1 ~o~O~oMOrwO~o.-~a».iMrro.~eho~0o~at.~.~NUwGwrNOd~r~
.,., a4 -- r omn awo .-i ~ ~ r .~-~ .1 .-r ~ ~ .-~ r ~ ~ r ~ iwo a~ .-, .1 .-r r N wo o~ a~ .-i ao .-wn m O
c al E a .C ~ p m O ,1 ~"~ N 1D O O O~ O
U ~ tD .1 u1 O O O r1 d' d' N O N O t0 M d, a0 ~r1 d' r1 N O~ O v0 ~ O n1 ef C1 eh OD r1 O O~ tn ~
!A t0 ~ N It1 m r f~f P"1 N t1' Il1 tn tf' N 01 r ri f' N O O !~ .-i .-i f"1 t!) 10 U1 m r O~ N ~' N 1f1 w0 O~ N .-1 O r N N U7 .1 .1 .1 v0 .-i r1 O~ t0 d' tn O 1ff 1~ c~1 V e1 r Nf O
e'~'1 ~D O O~ O~ d' ~D d' a " ..
b W z a o b ~, a w ~=
U ~ O t -ni O CJ ,~ v p it ~' ii' ~ a,' R' R', ~' ~' st a.' R.' ~' ii' ~' ~' R.' ~ ~' R' 0.' ~' ~' ~C ~
~ y2 Q Z x ~ ~ Z ~ x x sC x 2 2 ~ Z x Z ~ ~ x x x x ri iC x x x ~ x x ~ Z x x Z Z Z
U .rJ
t b U N
W 41 0l CI
O a -r-~
d tr1 ~ a ' O ~1 N M d' Ifl t0 r GO Ov O .-1 N ~°1 ~ 1f1 t0 r fn C~ O
r1 N f'1 V' Il1 v0 Q V7 '1. .1 N f'~1 d' in t0 r O OW1 .1 r1 r-1 r1 n1 .1 .1 r1 n-1 N N N N N N N
N N N M ~"1 A1 Al M P7 ('1 O
b O O
by a ~
a o ro uu m aw a o ar d a .c g r N r o ~~woM.-~Na~~ ...,NU,~o~~ .N~co~ ~ .NO ~roM~r w O .14 L7m1 O M ~O t0 ~O v0 f~ r N N Q~ O~ .-1 ~' N N .-1 N N ~ d' If1 1I1 O N
O~ ~O r W
1i'O " d'InInlOMt(7~Nlnrir~V't0r10.1t11V'r~-leill101'-110101nO
m ~N O
U A O
b L~ o O N
.a y "~ O
p O 'd m E C
w b~ ~
a m E o t'''1t0111NOtDNfDri~lGd'sPO00~~tONNIIINoNe~d'~Ot'1 b Wi O W 1.a M CD 10 d' ~D tn tn P t0 t~1 O~ N O~ r N t~1 .-I Q1 1l) t(1 47 P ~O
O O~ N ~ r O
a .C ...p,~Ne.~aoerM~Nm.mvnMwom.-ic~mnrrwvorwo.wo y, ~. W
p, O C
U ~t ~ ~C U C
O ~ U 'a r U w O
r"~ 07 ~.1 O.C
p0 O U
C
GL
0 0 ~ o z ~t d a ~c .s .e .s ~o .c '~ x " '~ '~ '~ '~ '~ '~ '~ '~ x ~ ~c x r~ .s x ~ .x x ~ z ~ ~ x ~c ~ .e x ~ .s N
O
U f.i V V '~
Q ~ O ~ N M '? 1n ~D r ~D O~ O .1 N M d' If1 1O r O O~ y ,.pp (n x .-1 N f'~1 d' 1!f ~O r O O~ r1 .1 r1 r1 r1 .1 ~-1 ~I ~1 ~1 N N N N N
N N N N N r Q
a w E CL

,~..7~V0 91/12339 PCT/US91/00855 In the total sample of caudates, skewness and kurtosi~~ values for the distribution of Kd were outside the limits of ~ 1, so the values were re-expressed as their natural logarithms. The lag-transformed values for Kd distribution were within acceptable limits for the measures of skewness and kurtosi~c as were the non-trans formed measures f or Bm,x and B~/ Kd .
Correlations between outcome measures were examined to determine if they were non-independent with each other, with age or with autolysis time. Linear correlations were r = -0.37 (P = 0.0023, two-tailed) for Bm,: with age, and r - 0.68 (P < 0.0001, two-tailed) for Log Kd with Bm~.
Therefore, age and Log Kd were used as covariates to remove their effects from the measures of Bm,=. Likewise, the effects of Bm"~ were removed from the effects of the Log Kd measures. No correlations were evident between BmaX
or Log Kd and autolysis time.
Table 5 compares the unadjusted and adjusted Kd and Bm,x in caudates of alcoholic and nonalcoholic individuals and that of A1+ and A1- allelic subjects. In Table 4a, the mean Log Kd in samples of the alcoholic group, after covariate adjustment for Bm~, was found to be significantly lower than that of the nonalcoholic group when tested using a two-factor ANCOVA (P = 0.023, one-tailed) . No significant Bm,~ differences were found among these two groups. Table 5b shows the unadjusted Bm,x of A1+ allelic subjects to be significantly lower (P < .008, one-tailed) than that of A1' allelic subjects. Further, Bm,x measures were covariate-adjusted for age as well as Log Kd values and also tested using two-factor ANCOVA.
The main effect for allele was still evident (P < 0.01, one tailed), with the adjusted B,m~ mean of the Al+ group being smaller than that of the A1' group. No significant WO 91/12339 PCT/US91/008~~
~~~~ , _32-differences were found in Kd measures when classification was based on the presence or absence of the A1 allele.
Covariate-adjusted means for B~ and Log Kd of the D2DR Were also compared in four subgroups of 66 caudates derived from: A1+ non-alcoholics (n=7), A1+ alcoholics (n=22), A1- nonalcoholics (n=26) and A1- alcoholics (n=11). The differences between the alleles for the Bm measures were parallel between alcoholics and nonalcoholics, thus no factor interaction was detected.
Comparisons of binding measures among the A2A2, AlA2 and AlAl groups indicated differences only for measures of Bm~ (P = 0.034, two-tailed). A test for polynomial trends among the three groups demonstrated a linear relationship (P = 0.01, two-tailed) with the highest means for A2A2 followed by the AlA2 samples and with the lowest for the AlAl group. Fisher's LSD post hoc test found a significant (P < 0.05) difference between the A2A2 and AlA2 groups. While the AlAl mean was the lowest, no significant differences were obtained with a sample size of four when compared with either of the two other groups.

x~JVO 91/12339 PCT/US91/00855 -33- zo~~~s Table 5. Binding characteristics of the D2 dopamine receptor in caudate of alcoholic and nonalcoholic subjects (a) and as a fuaction of the pras~~noe (+) or absence (-) of the A1 all~lo (b).
a. Alcoholic and nonalcoholic subjects Alcoholics (n=33) Nonalcoholica In=331 iCd (pM) Unadjusted 71.0 z S.2 NS 92.8 ~ 7.6 Log 1Cd 1 4.19 ~_ 0.06 P < 0.023 4.38 t 0.07 Bm~ (fmol/mg protein) Unadjusted 60.8 ~_ S.6 NS 69.8 ~ 6.0 Adjusted2 67.9 ~_ 4.4 NS S9S ~_ 4.9 b. A1+ and A1- allelic subiects Al+ fn=291 Al- (n=371 Kd (PM) Unadjusted 70.9 . S.4 NS 90S ~_ 7.1 Log ECdl 431 ~_ 0.07 NS 4.27 ~_ 0.06 Bm~ (fmol/mg pivtein) Unadjusted 53.7 ~_ 4.9 P < O.OOB 74.4 ~_ 5.9 Adjusted2 55.8 ~_ 5.2 P < 0.01 71.7 ~_ 4.2 1 Log-transfoneed and covariate-adjusted for Bmax by Ieaet-equate estimations.
2 Cavariate-adjusted for Log ICd and age by least-square estitnationc Significance obtained by one-tailed test;
NS + not significant.
~agI digests of human DNA, probed with a clone of a human genomic fragment of the D2DR gene (~lhD2G1), reveal two alleles: A1 and A2 (19j. Here, like in our previous study discussed in Example 1 (32), the present inventors have shown that digests of DNA, obtained from the cerebral cortex of alcoholics and nonalcoh.olics, when probed with the 1.6 kb doublet of ~lhD2Gl, revealed the Al allele to be associated with alcoholics and its absence to be WO 91/12339 PCT/US91/0085~

associated with non-alcoholics. Since the caudate nucleus is among brain regions with the highest expression of the D2DR, the question raised herein is whether a relationship e~tists among the binding characteristics of this receptor ~ and the polymorphic pattern of the D2DR gene in caudates of alcoholic and nonalcoholic subjects.
It should be noted that the alcoholics of the present study had a very severe type of alcoholism (43). Detailed clinical records, interviews of next of kin and examination at autopsy (both macroscopic and microscopic) revealed that these alcoholics not only had a history of heavy alcohol consumption and multiple failures in their alcoholic rehabilitation but also, in a majority of them, the cause of death was attributed to the damaging effects of alcohol on their bodily systems.
That severity of alcoholism is an important determinant in A1 allelic association is shown in an analysis of a recent study by Bolos et a1. (44) which used a different and less severe alcoholic population than the inventors' sample. In that study of blood obtained from living Caucasian subjects, patients were divided into two groups: less severe and more severe alcoholics. Using the inventors' two Caucasian groups (nonalcoholics and very severe alcoholics) and the three groups of Bolos et a1. (44) (CEPH 'controls'[alcoholics not excluded], less severe and more severe alcoholics) a~2 test for linear trend was conducted for prevalence of the A1 allele.
Increasing degree of alcoholism severity was found to correspond with significant increase (P = .0002) of A1 allele prevalence (45). Moreover, Cloninger et a1 (46), utilizing a population of very severe Caucasian alcoholics similar to inventors' and controls (alcoholics rigorously excluded), found 60% of the former group and 20% of the latter group to have the A1 allele. This distribution of ",~WO 91/12339 PCT/US91/00855 the A1 allele in their two groups closely corresponds to the original observation (15).
The polymorphic pattern of this gene and its 5 differential expression of receptors suggests the involvement of the dopaminergic system in conferring susceptibility to at least one subtype of alcoholism.
ERAMPLE II_~
10 Following the same procedure: described in Example I, the parent clone, lambda-hD2Gl, was obtained and digested with BamHl and run on 0.8% agarose (not Sea Plaque) and the bands visualized with ethidium bromide staining.
Figure 5 indicates a doublet in the gel on the right 15 labeled "Doublet 1.73 kb" with DZ~A molecular weight standards shown in the gel on the: left. For equal amounts of different, unrelated fragments., one would expect the smaller fragments to stain with less intensity since fewer nucleotides would be present. Since the two bands labeled 20 Doublet 1.73 indicated less staining intensity in the upper, larger kb band than in the lower, smaller kb band and since only one band was sometimes observed, it was suspected that the 1.73 kb fragment may consist of two, perhaps related, fragments. Hence, the 1.73 fragment was 25 labeled Doublet 1.73 kb.
This 1.73 kb fragment was originally estimated by gel electrophoresis to 1.6 kb (personal communication from D.K. Grandy). Subsequent sequencing of the doublet 30 determined 1.73 kb. As described. in Example IV, this doublet was separated into two different fragments in clones 9 and 16. Clone 9 was sequenced and found to actually consist of 1.6 kb, which. includes about 200 bases of an intron 5' to the 7th exon of the D2Dr gene, all of 35 the 7th exon, and about 1.3 kb of the noncoding region adjoining the 3' end of the 7th axon. Thus, what is .. -36-indicated in this example and in Figure 5 as 1.73 kb is actually: 1.6 kb. Any reference in this application to Doublet 1.73 or a 1.73 kb fragment or probe should be considered equivalent to Doublet 1.6 or a 1.6 kb or l.5kb fragment or probe. Therefore, for discussion purposes in this application, 1.73 kb, 1.6 kb, and 1.5 kb refer to essentially the same fragment, which has now been determined to be 1.6 kb.
ERAMPLE IV
To check the possibility of the 1.6 kb fragment being a doublet, a ligation was performed using this fragment by following the procedure given by Sea Plaque GTG agarose (FMC BioProducts, 5 Maple Street, Rockland, ME 04841-2994 USA). Transformation of DHSa cells was performed following the Hanahan method (Hanahan, J. Mol. 8iol. 166 (1983):557-580).
In order to differentiate between different clones, mini-preps were performed and the DNA cut with a series of enzymes (Hinf-1, MSp-l, Tai-1, $amHi, and Hind III) establishing two definite populations containing inserts of the appropriate size (1.6 kb). These two subclones, named #9 and ,~16, were grown and the DNA purified. The 1.6 kb fragments from each clone were labeled with 32P and used as probes to hybridize onto human genomic DNA cut with Ta~l using the same procedure described in Example 1.
Figure 6 shows that the 1.6 kb probe from clone hybridized with human genomic DNA cut with ~gl at 6.6 kb and 3.7 kb, whereas the 1.6 kb probe from clone ,~16 hybridized with a band at 10.5 kb. Through this subcloning of the 1.6'kb fragment, two 1.6 kb fragments were separated since the original 1.6 kb probe, suspected of comprising a doublet, hybridized with human genomic DNA

~,WO 91/12339 PCT/US91/00855 _37_ . $~~:~~~19 cut with T_aal gave three bands lacated at 10.5 kb, 6.6 kb, and 3.7 kb. Thus, the 1.6 kb fragment from clone 9 is informative as a probe for the presence of the A1 allele in human genomic DNA.
ERAMPhE 'P
After the close association of alcoholism with the A1 allele was discovered, a synthetic probe was developed to be used in the screening of genomic DNA for the presence or absence of the A1 allele. They inventors had GENETIC
DESIGNS, INC., 7505 S. Main Street, Houston, Texas, custom synthesize a oligomer of 30 base pairs corresponding to the last 30 base pairs of the cDNA for the 7th exon of the DZ receptor (19). The oligomer was radiolabelled at its 5' terminal by incubating at 37'C: for 1 hour: 2~1 of T4 Polynucleotide Kinase, 3u1 32yATP (6000 Ci/mM), l~l oligomer (100 ng/~cl), 2~1 of lOx Polynucleotide Kinase buffer (700mM Tris-HCL, pH 7.6; 100mM MgCl2; 50mM
dithiothreitol), 121 H20. After 1 hour of incubation, this mixture was passed over a lml G-50 column (gravity flow) to separate labelled 30 base pair oligomer (30 bit fragment of 7lhD2G1) from free nucleotide.
To test this 30 base pair o7.igomer (30 bit fragment of ~hD2G1) as a probe which would be informative for the A1 allele, it was hybridized to t:he 1.6 kb DNA fragments (e. g. probes) from clones 9 and 16 described in Example IV. Figure 7 shows the nucleotide sequence of this 30 base pair oligomer and the results of the hybridization 3o with the 1.6 kb probes from clones 9 and 16. The radiolabeled, synthetic oligomer was found to hybridize to clone 9 but not to clone 16. Since the 1.6 kb probe of clone 9 hybridizes to the 6.6 kb fragment of human genomic DNA, it is likely that this 30 base pair oligomer would hybridize to the 6.6 kb fragment of human genomic DNA and 274519 . -38-similarly function as an effective probe for detecting the presence of the A1 allele.
Another 30 bit oligomer corresponding to the first 30 ba~sr~ pairs of the cDNA coding for the 7th exon of the D2 receptor was tested as a potential probe for the A1 allele. In contrast to the oligomer consisting of the last 30 base pairs of the 7th exon shown in Figure 7, this oligomer hybridized to the 1.6 kb ~mHl fragments (probes) from clones 9 and 16. Therefore, it does not recognize the polymorphism associated with the A1 allele.
Although the 30 bit probe indicated in Figure 7 was hybridized to the 1.6 kb fragment from clone 9 and not the 6.6 kb fragment from a human genomic DNA, those of skill in the art will recognize that the 30 bit probe will also bind to the 6.6 kb fragment since it comprises a nucleotide sequence which is complementary to the 30 bit probe. Stated another way, since the 1.6 kb probe hybridizes with the A1 allele 6.6 kb fragment from human genomic DNA, a subfragment of the 1.6 kb probe would also be expected to hybridize to complementary sequences of the A1 allele. The 30 bit probe, being a complementary subfragment of the 1.6 kb probe, would be recognized by those skilled in the art as capable of hybridizing with a complementary sequence on the 6.6 kb fragment from human genomic DNA thereby being informative as a probe for the A1 allele.
Sequences in the 7th exon of the D2DR gene have been identified that are informative for the A1 allele through their ability to hybridize with the 6.6 kb fragment of human genomic DNA. Those sequences are shown in the present invention to be found in a ~LhDZGl probe, a 1.6 kb probe of clone 9 and a 30 bit fragment which encodes for the terminal portion of the 7th exon of the D2DR gene.

,.,CVO 91/12339 PCT/US91/00855 Since the 30 bit fragment sequence is common to both the .lhDZG1 probe and the 1.6 kb probe of clone 9, it is likely that analogous sequences within ~,hD2G1 would also be informative as probes for the A1 allele. Certainly one would expect other subfragments of these three probes to be capable of hybridizing to the 6.6 kb fragment which identifies the A1 allele in human genomic DNA.
Considering the 30 bit fragment of ~IhDZGI and the known art for hybridization specificity with nucleic acid oligomers, it is likely that subfragments of the 30 bit fragment would demonstrate specific hybridization to the 6.6 kb fragment of human genomic DNA and thereby be informative as probes for the A1 allele.
Thus, an embodiment of the present invention would utilize a subfragment of the 30 bit fragment comprising about 10 base pairs as a probe. In further embodiments, subfragments comprising about 15, 20, or 25 nucleotide bases of the 30 bit fragment sequence shown in Figure 7 would be utilized as probes to be: informative for the A1 allele in human genomic DNA.
ERAMPLE V~
Individuals categorized as non-alcoholic, alcoholic, children of alcoholics (COA), or drug abuser (DA) were tested for the presence of the A1. allele. Subjects were assigned to a particular category after they filled out a standard chemical identification diagnostics form for substance abuse developed by Kenneth Glum, E.P. Noble and 3o associates as well as DSI~2-3 criteria assessed by a clinician (DSI~t-3 form) . The 1.6. kb fragment from clone 9 was used as a hybridization probe for detecting the presence or absence of the A1 allele in the subject's DNA
isolated from lymphocytes. Genomic DNA from lymphocytes was obtained by the following methodology.

WO 91/12339 PCT/US91/008~~
2074519 . -40-1), Collect blood in (2) capped tubes without heparin.
2) Put 10-15 mls of whole blood into a polycarbonate tube (50 ml) and fill the remainder of the tube with DNA isolation solution [0.3M sucrose, lOmM Tris (pH
7.5), 5mM MgCl2, 1% Triton X-100]. Mix by inversion.
3) Centrifuge at 4,000 rpm at 4°C for 10 minutes.
Aspirate the supernatant into bottle containing bleach. Wash the pellet once again with the same solution. Repeat if needed until no red blood cells are apparent.

4) quickly resuspend pellet in 4.7 mls with TE and transfer into one 30 ml corex tube.

5) Slowly add 250 ~1 of Proteinase K (10 mg/ml in TE).

Gently mix.

6) Incubate 3 hours to overnight at 37C in a water bath. Swirl the viscous solution periodically.

7) Gently extract the DNA 1 time with an equal volume of phenol (pH 8.0) by mixing and immediately centrifuging for 10' at 5,000 rpm. Remove aqueous layer (top) carefully. Gently extract aqueous phase with 1 vol. of chloroform: isoamyl alcohol (24:1) mix gently and centrifuge immediately. Centrifuge for 10' at 5,000 rpm. Transfer aqueous layer to polypropylene tube.

8) Add 1/10 volume (0.5 ml) of 3 M sodium acetate plus total volume (5.5 ml) of cold isopropanol. Invert slowly to mix. DNA should fall out immediately.

9. Remove DNA with looped pipet. Wash DNA with 70%

ETOH, air dry briefly.

10) Resuspend in 400 to 600 u1 TE (pH 7.6) in 1.5 ml flip cap tube. Store at 4C.

11) Measure the exact concentration of the DNA and analyze an aliquot by electrophoresis through a 0.3%

agarose gel. The DNA should be greater than 100 kb -41- _ 20~45~9 in size and should migrate more slowly than a marker of intact bacteriophage. Store DNA at 4°C.
Hybridizations using the 1.6 kb probe of clone 9 were carried out utilizing the hybridization conditions described in Example 1. Results of these tests are summarized in Table 6.
Table 6 under "TOTAL", which summarizes the results for those subjects with and without a family history of alcoholism, Table 6 shows that there is a good correlation between the findings of Example 7:, where the A1 allele was detected in brain samples from dsaceased subjects, and the findings of this study, where the A1 allele was detected in living subjects. The previous study of Example 1, summarized in Table 2, indicated nonalcoholics with 20%
A1+ (having the A1 allele) and 80% A1- (without the A1 allele), whereas Table 5 indicates nonalcoholics with 22%
A1+ and 78% A1-. Similarly, the alcoholics were found in the brain tissue study of ExamplE~ I to have 69% A1+ and 31% A1-, whereas in the present study with lymphocyte DNA, alcoholics were found to have 63% A1+ and 37% A1-. These findings fit very well with other recent studies such as those of Cloninger et al. (46), where a much larger number of living alcoholics and nonalcoholics were tested for the presence of the A1 allele. In the Cloninger et al. study, the A1 allele was found to be present in 20% of the nonalcoholics versus 60% of the alcoholic (Table 6: 22 %
nonalcoholics versus 63% alcoholics).
In view of interviews With the subjects in this study, it is likely that the A1 allele may be predictive of attention deficit disorder with hyperactivity (ADDH).
Further, coupling of a determination for ADDH with the presence of the A1 allele may allow for a more reliable detection of alcoholism susceptibility or may be of WO 91/12339 PCT/US91/0085~

benefit in identifying a certain subtype of alcoholism.
The coupling of other indicators with the presence of the A1 allele may offer a significant advance in the detection, prediction or diagnosis of a susceptibility for other compulsive disorders.

-43- .

2~ ~45~ 9 H

W ~ h a ~ ' $ g ?' ~' a z s x . a i~
A <

F-:. b~ bsb~b~

a N ~Oh N

a +
~

~ a ~ ' ~ r c V

a ar a ~ ~ < ~ _ 4 v t H

a ~ ~ h N

v t v~O <

a ~, iZ .Y
a N ~ _ G ~ ~

~ N ~ ~ .~ 9 C C n .

a o ~ ~ g zs ~ ~

.
a i~ tz. N ~ ~ ~ ~ C v a 9 E ~' E 3 _>_;

' v ,~ 8 " ~
a ~

p i~

m C ~o ~ o o ~ ~ a ~ m a ~ : ,., '~

o >1 .~

.
z a a ~ Y= ~ E .
~ o ~ a$
~.o f~ P ~ o a, E
P' ~

U - E = 8 b o Q
W

~ a '~ E
9 O v ~
L

Ia '~ '~
C

'p .

r1 ~ O . C
~

iC H ' ' ~. Z.
o : , a s a ~

~ ,8 O a 9 'C ~ v ~ 4 m z U D

b ~

m~
gin' c'.a a.
H ~ n v'~ .

b~

o a m v o ~ ..
N O . . +
. : v ~
. .
,p Although the data for drug abusers in this particular study did not show a dominant presence of the A1 allele, it is believed that other restriction endonucleases or related probes would show some relationship to these 5, particular chronic drug abusers.
It is viewed as possible that the dopamine (D2) receptor gene polymorphism observed herein may also be associated with predilection to other addictive diseases, such as those relating to nicotine, narcotics or other drugs.
It is believed that research dealing with the exploration of various candidate gene probes which encode elements related to the synthesis, metabolism, storage, release, and receptor activity of neurotransmitters and neuropeptides involved in brain reward might ultimately lead to multigene trait markers which can detect susceptibility of individuals with a family history of alcoholism.
It is well known that relapse is often accompanied by stressful situations. It is of interest that in an experiment using brain tissues that had been classified as having the A1 or A2 allele in a prior study, the inventors attempted to answer the question of whether the presence of the A1 allele leads to an altered number of D2 receptors in the brain. 8y measuring the number of dopamine D2 receptors in the caudate nucleus, an area that normally has the highest density of these receptors, the inventors found that individuals having the A1 allele had approximately 30 percent fewer D2 receptors than those with the A2 allele. Since the dopamine D2 receptor gene controls the production of these receptors, this suggests that the A1 allele causes a reduction in the number of receptors.

,~1'O 91/12339 PCT/US91/00855 This finding suggests an interesting hypothesis. It is known that dopamine acts to reduce stress. When stress occurs in an individual with a normal number of dopamine receptors, dopamine is released, all of the receptors are filled, and equilibrium is restored. In an individual who has the A1 allele, however, the shortage of dopamine receptors interferes with this process and equilibrium is not restored. This person may seek alcohol or other substances or stimuli that release dopamine, in the attempt to find relief and pleasure. The desired effects do not come, however, because of the shortage of receptors, and the attempt is repeated, leading to aberrant pleasure-seeking behavior. The inventors call this concept the stress-dopamine-genotype hypothesis of craving. Taken together, the findings of a high association of the A1 allele of the DZ receptor gene as well as a 30% reduction of the DZ receptors in A1 allele carriers suggest the importance of diagnosing alcoholism and related behaviors including stress for risk potential.
A polymorphism of the dopamine receptor gene associated with alcoholism and its potential association with other addictive diseases is taught by the present invention. The use of the ~hD2Gl,, 1.6 kb, or 30 bit probes for detecting the A1 allele polymorphism of ~g I
digests are believed to exemplify one approach to detecting the D2DR gene polymorphism associated with alcoholism. The specific approach used in the present application by way of example does not preclude alternative approaches to polymorphism detection in this gene. Basically, the present disclosure teaches a polymorphism for the dopamine receptor gene associated with alcoholism which may be detected using other approaches, in addition to that exemplified iw the present disclosure by the A1 allele polymorphism in ~,g I digests of human genomic DNA. For example, Bolos et al. detect a ._ 4 6 -polymorphism by amplifying a 3' noncoding region of the dopamine receptor gene sequence with PCR (polymerase chain reaction) and separating the amplified fragments by electrophoresis under nondenaturing conditions. This approach is stated to reveal polymorphisms that affect the secondary structure of the single DNA strands which are amplified. Thus, it is intended that: this invention encompass alternative methods of detecting polymorphisms in the dopamine receptor gene which have been shown by the instant invention to be associated with alcoholism in humans.
At this time, the present findings of an allelic association of t:he dopamine (D~) ~°eceptor gene with alcoholism suggest that a do=fect in this gene, or in another gene with linkage d:isequilibrium with it, may cause susceptibility t:o, at least, one type of alcoholism.
Still, this finding may hold promise for specifically focused treatment and prevention strategies. Clearly, application of the discover_ie:~ and methods described herein should have great benefit for the 28 million children of alcoholics who are potentially at risk for this disease.
Finally, this research, as well as other work along similar lines, should result in the desti~~mat.izat:ion of alcoholism, and ensure that the erroneous view of it as a moral weakness should no longer be accepted by society.

-47- .207451 REFERENCES
to 1. Blum et al., U.S. Patent No. 4,761,429 (August 2, 1988) .
2. Wise, R.A. and Bozarth, N.A., Pharmacol Biochem J
Behavior ~7, 239-243 (1982) "Action of Drugs of Abuse on Brain Reward Systems: An Update with Specific Attention to Opiates".
3. Amit, Z. and Brown, Z.W. Pharmacol. Biochem J
Behavior 17, 233-238 (1982) "Actions of Drugs of Abuse on Brain Reward Systems: A Reconsideration with Specific Attention to Alcohol".
4. Supreme Court of the United States. Traynor v.
Turnage, Administrator, Veterans Administration ~t al., and McKelvey v. Turnage, Administrator, Veterans Administration et al. (Argued Dec. 7, 1987; decided April 20, 1988). Syllabus no. 86-22 and no. 86-737 (Washington, DC, 1988).
5. L. Kai, Alcoholism in Twins° Studies on th Ft;~~ogy and Sequelae of Abuse of Alcohol (Almqvist and' Wiksell Publ., Stockholm, 1960); D.S. Goodwin, Arch.
Gen. Psychiatry ~5, 545 (1971); D.S. Goodwin, ibid.
3~, 57 (1979); C.R. Cloninger, M. Bohman, S.
Sigvardsson, ibid. 38, 861 1;1981).
6. American Psychiatric Association, Diagnostic and Statistical Manual of Menta7~,Disorders, 3rd edition (American Psychiatric Association, Washington, DC, 1987).
7. H. Begleiter, B. Porjesz, C.L. Chow, Science 2~, 1064 (1981); Gabrielli e~ ~., Psychophysiolocrv 19, 404 (1982); Pollock et ~., Arch. Gen. Psychiatrv 40, 857 (1983); H. Begleiter, B.. Porjesz, B. Bihari, B.
Kissin, Science ~, 1493 (1984); S. O'Connor, V.
Hesselbrock A. Tasman Proct.
NeurQpsychopharmacol Biol. Psychiatry 10, 211 (1986); S.C Whipple, E.S.
Parker, E.P. Noble, J. Stud. Alcohol 49, 240 (1988).
8. S. Takahashi, N. Tani, H. Yamane, F~lia Psychiat.
Neurol. Jpn. ~Q, 455 (1976); A. Wiberg, C.G.
Gottfried, L. Oreland, Med. Biol. ,~, 181 (1977);
J.L. Sullivan, J.O. Cavenar, Jr., A.A. Maltbie, P.
Lister, W.W.K. Zung, Biol. Psychiatry ,~, 385 (1979);
C.J. Fowler, K.F. Tipton, A.V.P. MacKay, M.B.H.
Youdim, Neu;~oscience 7, 1577 (1982); L. Oreland et ., J. Neural Transm. ~6, 73 (1983); G. Alexopoulos, K.W. Lieberman, R.J. Frances, Am. J. Psychiatry X40, 1501 (1983); A.L. yon Knorring, M. Bohman, L. yon WO 91/12339 PCT/US91/0085~
20'~ 4519 Knorring, L. Oreland, Acta Psychiat. Scand. 72, 51 (1985); I. Diamond, B. Wrubel, W. Estrin, A. Gordon, Proc. Natl. Acad. Sci. USA ~, 1413 (1987); B.
Tabakoff et ~., N. Ena. J. Med. ~$, 134 (1988); D.
Mochly-Rosen et al., Nature 333, 848 (1988).
9. S. Hill, D.W. Goodwin, R. Cadoret, C.K. Osterland, S.M. Doner, J. Stud. Alcohol ~ø, 981 (1975); Y.
Shigeta et al., Pharmacol. Biochem. Behav. 13 (Suppl.
1), 89 (1980); S.Y. Hill, J. Armstrong, S.R.
Steinhauer, T. Baughman, J. Zubin, Alcoholism (NY) 11, 345 (1987).
10. Y.W. Kan., A.M. Dosy, Proc. Natl. Acad. Sci. USA 75, 5631 (1978); J.S. Gusella et ~., ature 306, 234 (1983); D.S. Gerhard et al., Am. J. Hum. Genet. ~6_, 3S (1984); M.J.M. Saraiva, P.P. Costa, D.S. Goodman, Neurolocty 36, 1413 (1986); R.J. Bartlett et al., Science 235,5, 1648 (1987); P.H. St. George-Hyslop et al., ibid, 885; M. Barrow _e~ ~., ature 326, 289 (1987); J.M. Rommens et al., Science 245, 1059 (1989); J.R. Riordan et al., ibid, 1066; B. Kerem et al., ibid, 1073.
11. W.F. Bosron, T-K. Li, Biochem Bio~hys Res. Commun.
91, 1549 (1979); D.P. Agarwal, S. Harada, H.W.
Goedde, Alcoholism fNYl 5, 12 (1981); W.F. Bosron, L.J. Magnes, T-K. Li, Biochem. Genet. ~, 735 (1983);
A. Yoshida, G. Wang, V. Dave, Am. J. Hum. Genet. 35, 1107 (1983); A. Yoshida, I-Y. Huang, M. Ikawa, Proc Natl. Acad. Sci. USA 81, 258 (1984); H.W. Goedde, D.P. Agarwal, in Genetics of Alcoholism, H.W. Goedde, D.P. Agarwal, Eds. (Alan Liss, NY, 1987), pp. 3-20.

12. A.S. Lippa, S.M. Antelman, A.E. Fisher, D.R.
Canfield, Pharmacol. Biochem. Behav. _1, 23 (1973); L.
Ahtee, K. Eriksson, Acta Ptlysiol. Scand. 93, 563 (1975); M.D. Dibner, N.R. Zahniser, B.B. Wolfe, R.A.
Rabin, P.B. Molinoff, Pharmacol. Biochem. Behav. 12, 509 (1980); R.A. Wise, , ~ (Suppl. 1), 213 (1980); M.L. Barbaccia, A. Reggiani, P.F. Spano, M.
Trabucchi, Ps~rchopharmacoloqy ~, 260 (1981).

13. M.K. Ticku, T. Burch, J. Neurochem. ~, 417 (1980);
S.F.A. Elston, K. Blum, L. DeLallo, A.H. Briggs, #~harmacol. Biochem. Behav. ~,, 13 (1982); K. Blum, S.F.A. Elston, L. DeLallo, A.H. Briggs, J.E. Wallace, Proc. Natl. Acad. Sci. USA 80, 6510 (1983); K. Blum, A.H. Briggs, J.E. Wallace, C.W. Hall, M.C.
Trachtenberg,. Exoerientia ~, 408 (1986); C.
Gianoulakis, A. Gupta, ~.fe Sci. ~, 2315 (1986); K.
Blum, H. Topel, Funct. Neurol. ~, 71 (1986); J.M.
Murphy, W.J. McBride, L. Lumeng, T-K. Li, Pharmacol.

"~WO 91/12339 PCT/US91/00855 49 ' Biochem. Behav. 26, 389 (1987); J.M. Murphy et al., Alcohol ;~, 283 ( 1988 ) .

14. S. Liljequist, Acta Pharmacol. Toxi~nl_ ~~ fig (1978); S.A. Newlin, J. Mancillas-Trevino, F.E.
Bloom, Brain Res. 209, 113 (1981); G. Mereu, F.
Fadda, G.L. Gessa, Brain Res. ~,, 63 (1984); L.
Stein, J. Belluzzi, Clin. Neur2nharmacol. 9_ (suppl.
4), 205 (1986); S. Govoni g~ ~., Brain Res. 38~, 138 (1986); P. Valverius, P.L. Hoffman, B. Tabakoff, T~.
Neurochem. ~?, 492 (1989); F. Fadda, E. Mosca, G.
Columbo, G.L. Gessa, Life Sc,~ g~, 281 (1989).

15. Bunzow et al. Nature, 336, 783-787 (1988) "Cloning and Expression of a Rat (D2) Dopamine Receptor cDNA".

16. Maslen, et al. Genomics, 2_, 66 (1988).

17. Tabakoff, g~ ~. Public Health Re orts, Vol. ~(6), 690-698 (1988) "Genetics and Biological Markers of Risk for Alcoholism".

18. T. Maniatis, E.F. Fritsch, J. Sambrook, Eds., Molecular Cloning. A Laboratory Manual (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982).

19. D.K. Grandy et al., Am. J. Hum. Genet., 4~, 778-785 (1989).

20. C.R. Cloninger, T-K. Li, Alcoholism: An Inhe fired Disease (U. S. Government Printing Office, Washington, DC, 1985), DHHS Publ. No. (ADM)85-1426; C.R.
Cloninger, Psychiat. Dev. ~, 167 (1986).

21. R.D. Myers, in Aldehvde Adducts in Alcoholism, M.A.
Collins, Ed. (Alan R. Liss, NY, 1985), pp. 201-220.

22. G. Obe, H. Ristow, J. Herha, in Biochemistry and Pharmacology of Ethanol~ Vol I, E. Majchrowicz, E.P. Noble, Eds. (Plenum Press, NY, 1979), pp. 659-676.

23. E.R. Korpi, J.D. Sinclair, O. Malminen, Pharmacol.
Toxicol. ~~, 94 (1987).

24. M.A. Schuckit, in Psychiatry Update: Vol. III, L.
Grinspoon, Ed. (American Psychiatric Press, Washington, DC, 1986), pp. 320-328.

25. Blocking of dopamine (D2) receptors reduces reward-seeking behaviors in animals (C. R. Gallistel, A.J.
Davis, Pharmacol Biochem. Behav. ~, 867 [1983]).
Acute alcohol administration increases pleasureful behavior by releasing dopamine (A. Imperato, G.

20745x9 Chiara, J. Pharmacol. Exp. Ther. 239, 219 [1986]) through stimulation of dopaminergic neurons in the ventral tegmental area of the brain (G.L. Gessa, F.
Muntoni, M. Collu, L. Vargiu, G. Mereu, Brain Res.
348, 201 [2985]). However, chronic ethanol ingestion reduces the responsiveness and number of dopamine (D2) receptors in striatal membranes of rats (L.
Lpcchi, R.M. Moresco, S. Govoni, M. Trabucci, Brain Res. 449, 337 [1988]). The consequence of decreased ~ activity of this reward system may lead to a compensatory increase in alcohol-seeking behavior.
This could be one nongenetic mechanism for developing alcoholism.

26. D.B. Goldstein, J.H. Chin, R.C. Lyon, Proc. Natl.
Acad. Sci. USA 79, 4231 (1982).

27. Y. Israel, H. Kalant, I. Lanfer, Biochem. Pharmacol.
14, 1803 (1965); A.L. Swann, J. Pharmacol. Exp. Ther.
232, 475 (1985); P.T. Nhamburo, B.P. Salafsky, B.
Tabakoff, P.L. Hoffman, Biochem. Pharmacol. 36, 2027 (1987).

28. F.E. Bloom, Advanc. Pharmacol. Ther. 2, 205 (1979);
K. Blum, A.H. Briggs, M.C. Trachtenberg, ~erientia 45, 444 (1989); K. Blum, Integ~r. Psychiat. 6, 199 (1989).

29. G. Siggins, T. Berger, E.D. French, T. Shier, F.E.
Bloom, roc. Clin. Biol. Res. ~Q, 275 (1982); R.D.
Myers, Experientia 45, 436 (1989); R.D. Myers, T.H.
Privette, ~3rain Res. Bull. ~, 899 (1989); B. Faraj, V.M. Camp, D.C. Davis, J.D. Lenton, K. Kutner, Alcoholism 13, 155 (1989).

30. C.R. Cloninger, M. Bohman, S. Sigvardsson, Arch. Gen.
Psychiatry 38, 861 (1981); C.R. Cloninger, Science ø3 ,, 410 (1987); D.W. Goodwin, y7. Stud. on Alc. 50, 397 (1989).

31. J.H. Allison, T.J. Cicero, .T- Pharmacol. Exn. Ther.
213, 24 (1980); P.L Hoffman, F. Moses, G.R. Luthin, B. Tabakoff, Mol. Pharmacol. 3~(,, 13 (1986); G.Y. Sun, H-M. Huang, R. Chandrasekhar, D.Z. Lee, A.Y. Sun, J.
Neurochem. 48, 974 (1987); T. Ritchie, H-S. Kim, R.
Cole, J. de Vellis, E.P. Noble, Alcohol ~, 183 (1988); D.H. Ross, Experientia ~, 407 (1989).

32. Blum K, Noble EP, Sheridan PJ, Montgomery A, Ritchie T, Jagadeeswaran P, Nogami H, Briggs AH, Cohn JB.
Allelic association of human dopamine DZ receptor gene in alcoholism. J_AMA. 263:2055-2060 (1990).

.~WO 91/12339 PCT/US91/00855 51 24~~~19 33. Diagnostic and Statistical Manual f Mental Disorders, Third Edition. Revised. Washington, D.C..: American Psychiatric Association; (1987).

34. Grandy DK, Marchionni MA, Makam H, Stofko RE, Alfano M, Frothingham L, Fischer J, Burke-Howie KJ, Bunzow JR, Server AC, Civelli O. Cloning of the cDNA and gene for a human DZ dopamine receptor. Proc Natl.
Acad. Sci. USA. 86:9762-9766 (1989).

35. Seeman P, Grigoriadis D. Dopamine D2 receptor dissociation constant for spiperone: Identical value using 3H-labeled agonist or ~'H-labeled antagonist.
Biochem. Pharmacol. 34:4065-4066 (1985).

36. Seeman P, Bzowej NH, Guan H-C, Bergeron C, Becker LE, Reynolds GP, Bird ED, Riederer P, Jellinger K, Watanabe S, Tourtellotte WW. Human brain dopamine receptors in children and aging adults. Synapse.
1:399-404 (1987).

37. De Keyser J, Walraevens H, De Backer J-P, Ebinger G, Vauquelin G. DZ dopamine receptors in the human brain: heterogeneity based on differences in guanine nucleotide effect on agonist binding, and their presence on corticostriatal nerve terminals. Brain Res. X84:36-42 (1989.

38. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ.
Protein measurement with the folin phenol reagent.
J. Biol Chem. 193:265-275 (1951).

39. Munson PJ, Rodbard D. LIGAND: A versatile computerized approach for characterization for ligand-binding systems. F~1. Biochem.
x:220-239 (1980).

40. Rosner B. Fundamentals of Biostatistics. 2nd ed.
Boston, MA: Duxbury Press; 1980:195.

41. Severson JA, Marcusson J, Winblad B, Finch CE. Age-correlated loss of dopaminergic binding sites in human basal ganglia. T~. Neurochem. 39:1623-1631 (1982).

42. Seeman P, Ulpian C, Bergeron C, Riederer P. Jellinger K, Gabriel E, Reynolds GP, Tourtellotte WW. Bimodal distribution of dopamine receptor densities in brains of schizophrenics. Science. X5:728-730 (1984).

43. Blum K, Noble EP, Sheridan P~J, Montgomery A, Ritchie T, Jagadeeswaran P, Nogami H, Briggs AH, Cohn JB.
Allelic association of human dopamine DZ receptor gene in alcoholism. . ;~,~~:2055-2060 (1990).

44. Bolos AM, Dean M, Lucas-Der~~e S, Ramsburg M, Brown GL, Goldman D. Population and pedigree studies reveal a lack of associat;i.on between the dopamine D~ receptor gene and a:Lcoholi.sm. JAMA. 264:3156-3160 (1990).

45. Noble, EP, Blum, K. better t.o the Editor. JAMA.
264(13):1547-1548 (1993).

46. Cloninger et al., Arc:h. Gen. Ps;ych., 250(12):975-990 (1993) .

47. Sokoloff P, Giro~> B, Martre~s M-P, Bouthenet M-L, Schwartz JC. Mo1_ecu~_ar cloning and characterization of a novel doparni.ne x,eceptor (D3) as a target for neuroleptics. Nature (London) 347:146-151 (1990).

48. Comings DE, Comings E3G. A controlled study of Tourette syndrome. 7.. Attention-deficit disorder, learning disorders, and school problems. Am. J. Hum.
Genet. 41:701-747_ (1~~87) .

49. Comings, DE. A c:ontr_olled ~>tudy of Tourette syndrome.
VII. Summary: A common genetic disorder causing disinhibition of the limbic system. Am. J. Hum.
Genet. 41:839-86Ei (1987).

50. Comings, DE, Comings BG Cont:rolled family history study of Tourette's syndromE~. II: alcoholism, drug abuse and obesity. ~T. Clin. Psychiatry. 51:281-287 (1990) .

51. Pickens RW, Svikis DS, McGue M, Lykken DT, Heston LL, Clayton PJ. Heterogeneity in the inheritance of alcoholism. Arch. Gen. Psvc:hiat~ 48:19-28 (1991).

52. Tarter RE, McBr:ide H,. Buonp~rne N, Schneider DU.
Differentiation of alcoholics - childhood history of minimal brain dy~~function, family history, and drinking pattern.. Arch. Gen. Psychiatry 34:761-768 (1977).

53. Noble EP, Blum K, Rit=chie T, Montgomery A, Sheridan, PJ. Allelic associat:i.on of the D~ dopamine receptor gene with receptor binding c:haracteri.stics in alcoholism, Arch. Gen. Fsychiat~, 48(7):648-654(1991).
* *~ * * * * * -k yr Changes may be made .in the processes of the invention described herein without departing from the concept and scope of the invention as define<~ in the following claims.

Claims (15)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. A method for determining susceptibility to alcoholism, the method comprising:
obtaining a DNA containing sample from a subject; and determining whether human dopamine D2 receptor gene alleles are present in said sample;
wherein detection of an allele A1 of the human dopamine D2 receptor indicates susceptibility to alcoholism.

2. The method for determining susceptibility to alcoholism, according to claim 1, the method further comprising:
determining that a human dopamine D2 receptor gene allele is present in the sample.

3. The method for determining susceptibility to alcoholism according to claim 1 or claim 2, the method comprising:
digesting DNA from a subject with a restriction endonuclease; and determining human dopamine D2 receptor gene allele present in said digested DNA;
a susceptibility to alcoholism being diagnosed when allele A1 of human dopamine D2 receptor gene is present, which is found in a majority of clinically diagnosed alcoholics.

4. The method for determining susceptibility to alcoholism according to claim 3, the method comprising:
digesting DNA from a subjects with Taq I restriction endonuclease;
determining the presence of human dopamine D2 receptor gene alleles in said digested DNA;
wherein a susceptibility to alcoholism may be diagnosed when allele A1 of human dopamine D2 receptor gene, which is found in a majority of clinically diagnosed alcoholics, is present.

5. The method for determining susceptibility to alcoholism according to claim 3, the method comprising:
digesting DNA from a subject with Taq I restriction endonuclease;
separating DNA fragments found in the digest;
probing separated fragments with labelled .lambda.D2G1 or fragments thereof having about 30 bases to determine the presence of human dopamine D2 receptor gene allele in said sample; and where susceptibility to alcoholism may be diagnosed when allele A1 of human dopamine D2, characterized as being found in a majority of clinically diagnosed alcoholics is present.

6. The method for determining susceptibility to alcoholism according to any one of claims 1, 3, or 4, wherein the human D2 receptor gene allele is human dopamine D2 receptor gene Al receptor gene allele.

7. The method for determining susceptibility to alcoholism according to claim 5, wherein said separated fragments are probed with a Bam H1-generated 1.6 kb fragment of .lambda.hD2G1.

8. The method for determining susceptibility to alcoholism according to claim 5, wherein said separated fragments are probed with a 30 base fragment of .lambda.hD2G1.

9. A method for determining susceptibility to alcoholism according to claim 3, the method comprising:
digesting DNA from a subject with Taq I restriction endonuclease;
separating DNA fragments obtained from the hydrosylate according to fragment size;

probing separated DNA fragments with labelled .lambda.hD2G1 or a portion of labelled .lambda.hD2G1 having a size of at least 30 bases;
determining the presence of a 6.6 kb DNA fragment in said separated fragments which is labelled by the probe; and where susceptibility to alcoholism may be diagnosed when said labelled 6.6 kb fragment is found.

10. The method for determining susceptibility to alcoholism as recited in claim 9, wherein the separated DNA
fragments are probed with a labelled Bam H1-generated 1.6 kb fragment of .lambda.hD2G1.

11. The method for determining susceptibility to alcoholism as recited in claim 9, wherein the separated fragments are probed with a labelled 30 base fragment of .lambda.hD2G1.

12. A process for determining susceptibility to alcoholism comprising:
detecting in a sample from an individual dopamine D2 receptor gene alleles, wherein detection of an allele A1 of the human dopamine D2 receptor gene indicates susceptibility to alcoholism.

13. A labelled probe comprising a nucleic acid molecule which specifically binds a human dopamine D2 receptor A1 gene allele, for use in a diagnosis of susceptibility to alcoholism.

14. A labelled probe according to claim 13 which specifically binds an allele of human dopamine D2 receptor gene obtained after Taq I hydrolysis of human DNA, for use in a diagnosis of susceptibility to alcoholism.

15. A labelled 1.6 kb probe comprising a nucleic acid molecule which specifically binds a human dopamine D2 receptor A1 gene allele obtained after Taq I hydrolysis of human DNA, for use in a diagnosis of susceptibility to alcoholism.