Molecular Psychiatry (1999) 4, 284–285  1999 Stockton Press All rights reserved 1359–4184/99 $12.00 ORIGINAL RESEARCH ARTICLE Investigation of cholecystokinin system genes in panic disorder JL Kennedy 1 , J Bradwejn 2 , D Koszycki 2 , N King 1 , R Crowe 3 , J Vincent 1 and O Fourie 1 1 Neurogenetics Section; 2 Anxiety Disorders Program, Clarke Division, Centre for Addiction and Mental Health, University of Toronto, Ontario, M5T 1R8, Canada; 3 Department of Psychiatry, University of Iowa College of Medicine, Iowa City, USA Keywords: panic disorder; cholecystokinin; CCK-BR; CCK- AR; genetic association study There is evidence for the role of the cholecystokinin (CCK) neurotransmitter system in the neurobiology of panic disorder (PD). 1 The CCK receptor agonist, CCK- tetrapeptide (CCK-4) fulfills criteria for a panicogenic agent 1 and there is evidence that PD might be associa- ted with an abnormal function of the CCK system. For example, PD patients show an enhanced sensitivity to CCK-4, and exhibit lower CSF and lymphocyte CCK con- centration as compared to healthy controls (reviewed by Bradwejn et al. 2 ). Also, untreated PD patients display an increased CCK-4-induced intracellular Ca 2+ mobilization in T cells relative to treated PD, depression and schizo- phrenia. 3 The CCK receptors have been classified into two subtypes: CCK-A and CCK-B. We report here a study of polymorphisms in the CCK pre-pro hormone gene (CCK), CCK-AR, and CCK-BR in DSM-IV panic patients (n = 99) vs controls matched for gender and eth- nicity. The CCK polymorphism revealed no association with PD. We identified a new polymorphism for the CCK- A receptor gene, and tested it in our sample, with nega- tive results. A single nucleotide polymorphism has been found in the coding region of the CCK-B receptor gene 4 (CCK-BR) and D Collier (personal communication) ident- ified a highly polymorphic dinucleotide (CT) n microsatel- lite in the 59 regulatory region. For the CCK-B receptor gene polymorphism, PD patients showed a significant association. Our genetic dissection of the CCK system thus far suggests that the CCK-B receptor gene vari- ation may contribute to the neurobiology of panic dis- order. Non-significant P-values were obtained for the CCK peptide gene (genotype: P = 0.915; allele: P = 0.87) and CCK-A gene (genotype: P = 0.502; allele: P = 0.58) mar- kers. The CCK-B polymorphism showed a significant P-value for association with panic disorder, using our a priori design of grouping alleles into four categories (x 2 = 13.8, 3 d.f., P = 0.004). An analysis using all alleles (see Table 1) was also significant (x 2 = 26.0, 15 d.f., P = 0.038). Simulations using the Monte Carlo method 5 supported this positive association. Panic patients showed an excess of alleles 6 and 7, with odds ratios of 2.3 (95% CI 1.3–3.9) and 1.7 (95% CI 1.0–2.7), respectively. For these two risk alleles, the sum chi- square is 12.38, 2 d.f., P-value 0.002. Our study showed no evidence that panic disorder is associated with the cholecystokinin pre-pro-hormone gene polymorphism, results that are contradictory to the recent study by Wang and colleagues 6 who reported significant findings in two populations. We also developed and examined a DNA polymorphism for CCK-AR (see Methods) and found negative results. We did find evidence supporting an association between PD and CCK-BR, suggesting a role for this gene as a modifying factor in conferring susceptibility to the dis- order. In view of the location of the marker in the pro- moter region, variation in gene expression may cause dysregulation of the CCK system, which could result in a risk factor for panic attacks. A limitation of this study is the use of cases and con- trols. This methodology is susceptible to spurious posi- tive results due to population stratification. We attempted to prevent this by matching carefully for eth- nic background, age, and sex. The large number of alleles for CCK-BR makes it more difficult to determine the role of each. It is possible for example that one or more of the alleles have a protective effect, and we cur- rently do not have power to detect this. A final con- sideration is that we tested three loci and thus theoreti- cally should correct for this by multiplying the P- values by the number of tests. In argument against this correction is the fact that the CCK system genes have a very good biological hypothesis behind them, and could thus be considered distinct tests that do not need correction for other tests. In any case, even with correc- tion, our findings for CCK-BR remain significant under the a priori four-allele category design. Future work in this project will include collection of probands and both their parents in order to attempt a replication of our findings using family-based controls. Methods Subjects Subjects were recruited through referrals or advertising through the Mood and Anxiety Division of the Depart- ment of Psychiatry, University of Toronto. They were invited to participate if they fulfilled criteria for PD as substantiated by a psychiatric interview and by a struc- tured interview based on the DSM-IV (SCID). Comor- bidity with other anxiety disorders or depression was allowed as long as the diagnosis of PD was primary and predominant. PD subjects (63 females and 36 males) were closely matched with controls on the basis of age, sex and ethnic ancestry. Ethnicity in both patients and controls was determined based on primary language spoken, religious affiliation and birthplace of both the proband and his or her parents. The control group con- sisted of healthy individuals responding to local adver- tisements. After complete description of the study to the subjects, written informed consent was obtained. Genomic DNA extraction was performed using stan- dard high salt methods.