Molecular Psychiatry (2003) 8, 485–487 & 2003 Nature Publishing Group All rights reserved 1359-4184/03 $25.00 www.nature.com/mp
IMMEDIATE COMMUNICATION
Support for genetic variation in neuregulin 1 and susceptibility to schizophrenia NM Williams, A Preece, G Spurlock, N Norton, HJ Williams, S Zammit, MC O’Donovan and MJ Owen Department of Psychological Medicine, University of Wales College of Medicine, Heath Park, Cardiff, Wales CF14 4XN, UK Recently, it has been reported that genetic variants around the gene neuregulin 1 are associated with schizophrenia in an Icelandic sample. Of particular interest was the presence of a single-risk haplotype that was significantly over-represented in schizophrenic individuals compared to controls (15.4 : 7.5%, P ¼ 6.7 106). We have attempted to replicate this result in our large collection of 573 schizophrenia cases and 618 controls. We found that the risk haplotype was more common in cases than controls (9.5 : 7.5%; P ¼ 0.04), and especially in our subset of 141 cases with a family history of schizophrenia (11.6%; P ¼ 0.019). Our results therefore replicate the Icelandic findings in an out-bred Northern European population, although they suggest that the risk conferred by the haplotype is small. Molecular Psychiatry (2003) 8, 485–487. doi:10.1038/sj.mp.4001348 Keywords: schizophrenia; neuregulin 1, genetics
Schizophrenia is complex genetic disorder characterized by profound disturbances of cognition, emotion and social functioning. It carries a surprisingly uniform lifetime morbid risk of approximately 1% across different populations with different cultures. A number of systematic linkage scans have been performed reporting positive linkage findings at several genomic regions. One such region that has yielded positive linkage scores from a number of different groups is 8p22–p11 (locus SCZD6 (MIM 603013)).1–4 A recent study has implicated a gene from chromosome 8p, neuregulin 1 (NRG1) as a schizophrenia susceptibility locus.1 Following their initial evidence for linkage to schizophrenia to 8p12–21 in a large number of multigeneration families from Iceland, Stefansson and co-workers performed case/control and family-based association analyses and obtained evidence for association between schizophrenia and several haplotypes at the NRG1 locus. The entire exonic sequence of the NRG1 gene was extensively screened for mutations in 184 schizophrenics, but despite identifying a large number of variants, it was not possible to find any that were unambiguously associated with schizophrenia. The most important finding, however, was the identification of a single ‘core at-risk haplotype’ which was significantly overrepresented in schizophrenic individuals compared to controls (15.4 : 7.5%, P ¼ 6.7 106). The first replication of this finding has recently been reported,5
Correspondence: M Owen, M O’Donovan, Department of Psychological Medicine, University of Wales College of Medicine, Heath Park, Cardiff CF14 4XN, UK. E-mail:
[email protected];
[email protected] Received 23 December 2002; revised 6 February 2003; accepted 9 February 2003
in which the at-risk haplotype was again found to be more prevalent in 609 unrelated cases compared to 618 matched controls (10.2 : 5.9%, P ¼ 3.1 104). Interestingly the sample for this study was derived from Scotland suggesting that the risk haplotype is not specific to the Icelandic population. We have attempted to perform a fully independent replication of these results in our large collection of 709 cases with DSM IV schizophrenia and 710 controls. All subjects included in our case control sample were unrelated Caucasians born in the UK or Ireland, and did not overlap with samples reported in the earlier studies. The core at-risk haplotype reported by Stefanson et al1 is defined by a minimal haplotype of one SNP and two microsatellites, SNP8NRG221533, 478B14-848 and 420M9-1395. All three markers were genotyped in our sample. No single marker achieved a significant allelic association (w2 ¼ 0.011, P ¼ 0.92 (1 d.f.); w2 ¼ 5.1, P ¼ 0.75 (8 d.f.); w2 ¼ 3.04, P ¼ 0.96 (9 d.f.) for markers SNP8NRG221533, 478B14-848 and 420M9-1395, respectively). Tests for haplotypic association with schizophrenia were calculated from a total of 573 cases and 618 controls for which we had complete genotype data. The haplotype in our sample constructed from alleles 1, 2, and 4 of markers SNP8NRG221533, 478B14-848 and 420M9-1395, respectively, corresponds to the at-risk haplotype defined as alleles 1, 0, 0 in the original study.1 This haplotype was estimated by Stefansson co-workers to confer a relative risk of 2.2, an effect size that our sample has over 99% power to detect at a nominal significance level of 5%. We found that the at-risk haplotype was in excess in our cases when compared to our controls (Table 1; P ¼ 0.04). However, the estimated odds ratio (OR; Table 1) was 1.25 indicating a much smaller effect
Genetic variation in neuregulin 1 NM Williams et al
486
Table 1 Test of the ‘at-risk’ haplotype of markers SNP8NRG221533, 478B14-848 and 420M9-1395 Test Cases (n=573) Control (n=618) Familial cases (n=141) Nonfamilial cases (n=432)
Frequency (%) 9.5 7.5 11.6 8.8
size than that reported by Stefansson. However, the original study was based on individuals derived from large multiply affected families from Iceland. Therefore, we stratified our cases into those with and without a family history of schizophrenia. We found that the risk haplotype was enriched in the 141 familial cases (11.6%), the estimated OR was increased (1.63; P ¼ 0.019) and the upper end of the 95% confidence interval overlapped with the estimate of Stefansson et al.1,5 We therefore conclude that we have achieved the first fully independent replication of association between the haplotype that has been reported in two previous studies to be a risk haplotype for schizophrenia.1,5 Moreover, the effect size appears to be stronger in our subset of 141 familial cases. Taken together, the three studies provide strong evidence therefore that NRG1 is a susceptibility gene for schizophrenia. The estimated frequency of the NRG1 risk haplotype is similar in our sample to those from Scotland (10.2%), and is somewhat less than was found in Iceland (15.4%). However, our estimate of control haplotype frequency (7.5%) is similar to the Icelandic sample both of which are greater than that observed in the Scottish controls (5.9%). Thus, each sample has somewhat different point estimates for the OR. Nevertheless, our 95% CI for the OR extends as low as 1.08 even in the familial sample. This suggests that large samples may be required for further replication, a factor that will be important for others to bear in mind when further studies, which are clearly warranted, are reported.
Methods Subjects All subjects included in our case control sample were unrelated Caucasians born in the UK or Ireland. All 709 cases (483 males and 226 females with mean age of 45712 years) met DSM-IV6 criteria for schizophrenia with consensus diagnosis being determined following a semistructured interview, SCAN or PSE,7,8 and examination of case notes. Local Research Ethics Committee approval was obtained, and subjects gave written informed consent to participate. In total, 710 control individuals (474 male and 236 female subjects with mean age of 43712 years) were group matched to cases for age, sex and ethnicity from more than 1400 blood donors recruited from the local Molecular Psychiatry
P-value (one tailed)
w
OR (95% CI)
3.07
0.04
1.25 (0.94–1.67)
4.33 1.23
0.019 0.13
1.63 (1.08–2.47) 1.19 (0.88–1.63)
branch of the National Blood Transfusion Service (Wales). Controls were not screened for the absence of psychiatric illness; however, the policy of the National Blood Transfusion Service is not to receive from individuals who are on medication. The subset of familial cases (n ¼ 141, consisting of 47 female and 94 male subjects with a mean age of 49.05712.86) were all ascertained for a sib-pair linkage study9 from the same sources, and assessed and diagnosed using the same procedures. The nonfamilial sample was ascertained without reference to family history. Clearly, some of these will have an affected firstdegree relative but as no other family members were interviewed, their diagnostic category is considered unknown. Genotyping SNP8NRG221533 was genotyped using a fluorescence polarization-based primer extension assay,10 using the AcycloPrime Kit (Perkin Elmer) according to the manufacturer’s instructions and an Analyst genotyping platform (LJL Biosystems). The microsatellites 478B14-848 and 420M9-1395 were genotyped following analysis on an ABI3100 capillary sequencer. Statistical analysis Tests for single marker allelic association in our case/ control sample were performed by w2 analysis. Tests for haplotypic association with schizophrenia were performed using the software EHPLUS11,12 with a permutation test to obtain the empirical significance levels PMPLUS.11 The P-values reported are one tailed.
References 1 Stefansson H, Sigurdsson E, Steinthorsdottir V, Bjornsdottir S, Sigmundsson T, Ghosh S et al. Neuregulin 1 and susceptibility to schizophrenia. Am J Hum Gen 2002; 71: 877–892. 2 Blouin JL, Dombroski BA, Nath SK, Lasseter VK, Wolyniec PS, Nestadt G et al. Schizophrenia susceptibility loci on chromosomes 13q32 and 8p21. Nat Genet 1998; 20: 70–73. 3 Kendler KS, MacLean CJ, ONeill FA, Burke J, Murphy B, Duke F et al. Evidence for a schizophrenia vulnerability locus on chromosome 8p in the Irish study of high-density schizophrenia families. Am J Psychiatry 1996; 153: 1534–1540. 4 Kaufmann CA, Suarez B, Malaspina D, Pepple J, Svrakic D, Markel PD et al. NIMH Genetics Initiative Millennium Schizophrenia Consortium: linkage analysis of African-American pedigrees. Am J Med Gen 1998; 81: 282–289. 5 Stefansson H, Sarginson J, Kong A, Yates P, Steinthorsdottir V, Gudfinnsson E et al. Association of Neuregulin 1 with schizo-
Genetic variation in neuregulin 1 NM Williams et al
487 phrenia confirmed in a Scottish population. Am J Hum Genet 2002; 72: 83–87. 6 American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th edn. American Psychiatric Association: Washington DC, 1994. 7 Wing JK, Babor T, Brugha T. SCAN: schedules for the clinical assessment in neuropsychiatry. Arch Gen Psychiatry 1990; 47: 589–593. 8 Wing JK, Cooper JE, Satorius N. The Measurement and Classification of Psychiatric Illness. Cambridge University Press: Cambridge, 1974.
9 Williams NM, Rees MI, Holmans P, Norton N, Cardno AG, Jones LA et al. A two-stage genome scan for schizophrenia susceptibility genes in 196 affected sibling pairs. Hum Mol Genet 1999; 8: 1729– 1739. 10 Chen X, Levine L, Kwok P-Y. Fluorescence polarization in homogenous nucleic acid analysis. Genome Res 1999; 9: 492–498. 11 Zhao J, Curtis D, Sham PC. Model-free analysis and permutation tests for allelic associations. Hum Hered 2000; 50: 133–139. 12 Xie X, Ott J. Testing linkage disequilibrium between a disease gene and marker loci. Am J Hum Genet 1993; 53: 1107.
Molecular Psychiatry
