SRD5A2 is associated with increased cortisol metabolism in schizophrenia spectrum disorders

Methods A case–control sample of subjects with BD (N = 213), SCZ (N = 274) and healthy controls (N = 370) from Oslo, Norway, were included and genotyped from 2003 to 2008. A sub-sample (healthy controls: N = 151; SCZ: N = 40; BD: N = 39) had estimated enzyme activities based on measurements of urinary free cortisol, urinary free cortisone and metabolites. A total of 102 single nucleotide polymorphisms (SNPs) in the SRD5A1 , SRD5A2 , AKR1D1 , HSD11B1 and HSD11B2 genes were genotyped, and significant SNPs analyzed in the sub-sample. Results There was a significant association of rs6732223 in SRD5A2 (5α-reductase) with SCZ (p = 0.0043, Bonferroni corrected p = 0.030, T risk allele). There was a significantly increased 5α-reductase activity associated with rs6732223 (T allele) within the SCZ group (p = 0.011). Conclusions The present data suggest an interaction between SCZ and SRD5A2 variants coding for the enzyme 5α-reductase, giving rise to increased 5α-reductase activity in SCZ. The findings may have implications for cortisol metabolizing enzymes as possible drug targets. Research Highlights ► Data suggesting an interaction between SCZ and the gene SRD5A2 . ►Variants of SRD5A2 associated with increased 5α-reductase activity in SCZ. ► Supports cortisol metabolism as part of the HPA axis pathophysiology in SCZ. Abbreviations ACTH adrenocorticotrophic hormone aTHF allo-tetrahydrocortisol BD bipolar disorder BMI body mass index CDSS Calgary Depression Scale for Schizophrenia CHISQ Chi-squared statistic CRH corticotrophin releasing hormone DF degrees of freedom 11β-HSD1 11β-hydroxysteroid dehydrogenase type 1 11β-HSD2 11β-hydroxysteroid dehydrogenase type 2 GR glucocorticoid receptor HPA hypothalamic-pituitary-adrenal HWD Hardy–Weinberg disequilibrium IDS-C Inventory of Depressive Symptoms LD linkage disequilibrium MAF minor allele frequency PANSS Positive and Negative Syndrome Scale PRIME-MD Primary Care Evaluation of Mental Disorders SCID 1 The Structured Clinical Interview for DSM-IV Axis I Disorders SCZ schizophrenia (spectrum) SNP single-nucleotide polymorphism THE tetrahydrocortisone THF tetrahydrocortisol TOP Thematically Organized Psychosis UFE urinary free cortisone UFF urinary free cortisol YMRS Young Mania Rating Scale Keywords 5α-Reductase Bipolar disorder HPA Schizophrenia Severe mental disorders 1 Introduction Bipolar disorder (BD) and schizophrenia (SCZ) are severe mental disorders with lifetime prevalences of 1–2% ( Merikangas et al., 2007; Perala et al., 2007 ). They cause severe individual disability as well as a major economical burden to the society ( Kleinman et al., 2003; Mangalore and Knapp, 2007 ). BD and SCZ spectrum disorders are considered part of a common psychosis spectrum, sharing both etiological components and symptom patterns ( Craddock et al., 2009 ). The combination of several gene variants, each conferring a minor risk, probably adds up to a high heritability with estimates of 80% ( Cichon et al., 2009 ). Several candidate genes have been suggested ( Barnett and Smoller, 2009; Owen et al., 2009 ), but the genetic and environmental causes and pathophysiological mechanisms remain largely unknown. The hypothalamic-pituitary-adrenal (HPA) axis is an important response system for both physical and mental stressors ( Biondi and Picardi, 1999 ), and a well described biological correlate for the stress-vulnerability hypothesis of several mental disorders ( de Kloet, 2008 ). In general, the hypothalamus receives both stimulatory and inhibitory signals originating from the limbic system, regulating corticotrophin releasing hormone (CRH). CRH stimulates adrenocorticotrophic hormone (ACTH) from the anterior pituitary, and ACTH acts on the adrenal cortex for cortisol secretion. Cortisol has regulatory negative feedback effects on the axis ( Herman et al., 2005 ). Limbic abnormalities and dysfunction of the axis are well documented in BD ( Blumberg et al., 2003; Daban et al., 2005 ) and SCZ ( Ebdrup et al., 2010; Walker et al., 2008 ). On the basis of findings including abnormal activity in euthymia ( Watson et al., 2004 ) and in close relatives ( Mondelli et al., 2008 ), it has been argued that the HPA axis is involved in the pathophysiology of these disorders. Several trials with drugs targeting the HPA axis have been initiated and with some promising results ( Gallagher et al., 2008; Marco et al., 2002; Young et al., 2004 ). However, despite extensive research, the pathophysiology underlying the dysfunction is still unclear. The genetics of the HPA axis in BD and SCZ has been explored, but not with uniform results. Studies of genes coding for CRH ( Stratakis et al., 1997 ), the CRH receptor ( De Luca et al., 2007 ) and the glucocorticoid receptor (GR) ( Moutsatsou et al., 2000 ) are generally negative in BD. Studies of the FK506 binding protein 5 gene, a gene involved in GR functioning, are negative in SCZ ( Fallin et al., 2005 ) and inconclusive in BD ( Fallin et al., 2005; Willour et al., 2009 ). However, there are indications that both disorders are associated with variants of the 14-3-3 eta chain gene ( Duan et al., 2005; Grover et al., 2009; Wong et al., 2003 ) which is involved in GR turnover. Recently we found increased systemic activity of cortisol metabolizing enzymes (5α-reductase, 5β-reductases and 11β-hydroxysteroid dehydrogenase type 2 [11β-HSD2]) in BD and SCZ ( Steen et al., in press ). This was the first time that abnormal metabolism of cortisol was implicated as a potential factor underlying the HPA axis pathology in these disorders. However, studies on other mental disorders including chronic fatigue syndrome ( Jerjes et al., 2006 ), posttraumatic stress syndrome ( Yehuda et al., 2009a, b ), eating disorders ( Poor et al., 2005 ) and depression ( Poor et al., 2004; Raven and Taylor, 1996; Raven and Taylor, 1998; Romer et al., 2009; Weber et al., 2000 ) have previously addressed the issue of cortisol metabolism. Cortisol metabolism has also been subject to interest in various somatic disorders including obesity ( Andrew et al., 1998 ) and metabolic syndrome ( Anagnostis et al., 2009 ), conditions associated with BD and SCZ ( Birkenaes et al., 2007 ). The findings of increased systemic cortisol metabolism in BD and SCZ, together with several reports of the metabolism influencing HPA axis functioning ( Paterson et al., 2007; Rasmuson et al., 2001; Stewart et al., 1990 ), suggest cortisol metabolism as a part in the mechanism of the HPA axis dysfunction. This indicates that abnormal cortisol metabolism is linked to the pathophysiology of these disorders. To the best of our knowledge there are not yet any studies on candidate genes coding for enzymes in the cortisol metabolism in BD and SCZ. Our objective in the present study was to investigate if variations in genes coding for cortisol metabolizing enzymes could explain the findings of increased activity in these enzymes in BD and SCZ ( Steen et al., in press ). Our specific aims to investigate were 1) if the distribution of single-nucleotide polymorphism (SNP) genotypes in genes coding for 5α-reductase, 5β-reductases, 11β-HSD1 and 11β-HSD2 in BD, SCZ and healthy control groups indicated SNPs for analyses with enzyme data; 2) if SNPs suggested in 1) were associated with an altered enzyme activity; and 3) if the combination of diagnosis and SNPs in genes coding for 5α-reductase, 5β-reductases, 11β-HSD1 and 11β-HSD2 were associated with an altered enzyme activity. These aims were approached combining genetic data with a reanalysis of our enzyme data ( Steen et al., in press ). 2 Methods 2.1 Subjects The study is based on the case–control sample of the ongoing Thematically Organized Psychosis (TOP) Study that is carried out by the University Hospitals of Oslo, Norway. Subjects were included and genotyped from 2003 to 2008, and constitute a total sample of 213 patients with BD, 274 patients with SCZ and 370 healthy controls. Characteristics of the sample are presented in Table 1 . Patients were included according to the following criteria: Caucasian ethnicity; being registered as in- or out-patients in the psychiatric services of any one of the four hospitals in Oslo; age 18 to 65 years; meeting DSM-IV criteria for schizophrenia spectrum disorders (schizophrenia, schizophreniform and schizoaffective disorder), in the following termed “schizophrenia (SCZ)”, or bipolar disorder (bipolar I disorder, bipolar II disorder and bipolar disorder not otherwise specified), in the following termed “bipolar disorder (BD)”; and being willing and able to give written, informed consent of participation. Exclusion criteria were: history of moderate or severe head injury and neurological disorders. Inclusion and diagnostic interviews were done by trained clinical research personnel using The Structured Clinical Interview for DSM-IV Axis I Disorders, SCID 1 ( First et al., 1995 ). Inter-rater reliability was good, with an overall kappa score of 0.77 (95% C.I: 0.60–0.94). For more details, see ( Steen et al., in press ). A representative healthy control group was randomly selected from statistical records from the same catchment area as the patient groups, and the subjects were contacted by letter inviting them to participate. The healthy control group was screened, including the use of Primary Care Evaluation of Mental Disorders (PRIME-MD) ( Spitzer et al., 1994 ), to ensure that the subjects or any of their close relatives did not have a lifetime history of a severe psychiatric disorder. About 90% of the patients and 85% of controls had both of their parents born in Norway, while the rest had one parent from another European country. The homogeneity of the Norwegian sample is further indicated by Djurovic et al., 2009 . The investigation was carried out in accordance with the latest version of the Declaration of Helsinki. After complete description of the study to the subjects, written informed consent was obtained. The Regional Ethics Committee and The Data Inspectorate approved the study. The biobank was approved by the Norwegian Directorate of Health. 2.2 SNP selection/genotyping On the basis of systemic effects ( Vogeser et al., 2002 ) and findings in our previous study ( Steen et al., in press ), the genes for 5α-reductase ( SRD5A1 and SRD5A2 ), 5β-reductases ( AKR1D1 ), 11β-hydroxysteroid dehydrogenase type 1 (11β-HSD1) ( HSD11B1 ) and 11β-HSD2 ( HSD11B2 ) were selected. A total of 102 SNPs associated with these genes were genotyped using Affymetrix Genome-Wide Human SNP Array 6.0. 2.3 Sub-sample 2.3.1 Subjects From the total group of patients and healthy controls referred to the TOP study, a sub-sample of subjects included from 2006 to 2008 had their urine sampled for estimation of systemic cortisol metabolizing activity. Criteria for not estimating enzyme activities were a diagnosis of hepatic- or renal disorder, thyroid dysfunction, Addison's disease, Cushing's syndrome or use of corticosteroid medications. The resulting sub-sample constituted consecutively referred subjects with both measurements of urinary cortisol metabolites and genotyping, consisting of a total of 151 healthy controls and 79 patients; N = 40 with SCZ (schizophrenia [n = 34], schizophreniform [n = 2] and schizoaffective disorder [n = 4]) and N = 39 with BD (bipolar I disorder [n = 24], bipolar II disorder [n = 10] and bipolar disorder not otherwise specified [n = 5]). Urinary samples were originally obtained from 156 patients and 169 healthy controls (for details see Steen et al., in press ); however, after matching for genotype analysis, the sample size was reduced. This was due to genotyping errors, and because only patients with Caucasian ethnicity were genotyped. Thus, the current study includes a reanalysis of parts of the enzyme data ( Steen et al., in press ). 2.3.2 Symptoms Symptom assessments of the patients in the sub-sample were done with Inventory of Depressive Symptoms (IDS-C) ( Rush et al., 1996 ), Calgary Depression Scale for Schizophrenia ( Addington et al., 1990 ) (CDSS, for n = 9 subjects with SCZ), Young Mania Rating Scale (YMRS) ( Young et al., 1978 ) and Positive and Negative Syndrome Scale (PANSS) ( Kay et al., 1987 ). 2.3.3 Sample characteristics Age, gender, body mass index (BMI), urinary creatinine, time of sampling and medication were recorded. BMI was missing in 18 healthy controls and 3 subjects with BD, and was imputed with the respective group means. Urinary creatinine was missing in one subject in each of the patient groups and in two healthy controls, and was imputed with the sample mean. There was no significant differences in gender ([% males]: BD [44%], SCZ [65%], healthy controls [59%], p = 0.13), but significant differences in age (years) ([median/interquartile range]: BD [39.0/20.0], SCZ [29.5/14.0], healthy controls [31.0/13.0], p = 0.008) between the diagnostic groups. Total score of IDS-C (BD [12.0/20.0], SCZ [19.0/13.0]) did not differ significantly between groups (CDSS of n = 9 subjects with SCZ [5.0/7.0]), however, there was a difference in the total score of PANSS (BD [43.0/10.0], SCZ [65.0/23.0], p < 0.001) and total score of YMRS (BD [2.0/6.0], SCZ [8.0/9.0], p = 0.037). BMI (kg/m 2 ), urinary creatinine (mmol/l) and time of sampling differed significantly between groups (BMI: BD [26.4/3.6], SCZ [26.2/6.8], healthy controls [24.9/3.9], p = 0.005, and urinary creatinine: BD [8.5/11.1], SCZ [11.3/11.9], healthy controls [6.4/8.0], p = 0.008, and time of sampling: BD [10:20 AM/2:10 (h:min)], SCZ [11:30 AM/2:36], healthy controls [12:45 AM/6:10], p < 0.001). In the SCZ group, n = 34 (85%) used one or more antipsychotics, n = 9 (23%) used antidepressants and n = 6 (15%) used mood stabilizers. In the BD group, n = 20 (51%) used one or more antipsychotics, n = 15 (38%) used antidepressants, and n = 21 (54%) used mood stabilizers. For characteristics of the original sample with enzyme data, see Steen et al., in press . 2.3.4 Cortisol measurements Within 2 weeks after clinical assessments, patients underwent neuropsychological testing and routine blood withdrawal. On this occasion, urine was sampled and immediately frozen for later analyses of urinary free cortisol (UFF, reflecting systemic free cortisol), urinary free cortisone (UFE), allo-tetrahydrocortisol (aTHF), tetrahydrocortisol (THF) and tetrahydrocortisone (THE). The healthy controls underwent the same urine sampling, neuropsychological testing and blood withdrawal procedure. Allo-THF, THF, THE, cortisol and cortisone in urine were measured with liquid chromatography tandem mass spectrometry as described in Steen et al., in press . 2.3.5 Indexes of enzyme activities Activity of 5α-reductase was estimated with the aTHF/UFF index ( Ulick et al., 1992 ). Subjects with two urinary samples (n = 7) had indexes computed from their average values. In order to control for variation in urine concentration, urinary creatinine was measured (Jaffé-reaction, Cobas Integra, Roche Diagnostics GmbH, Mannheim, Germany). 2.4 Statistical analysis 2.4.1 Total sample Before analyzing our total sample, we removed all SNPs in high linkage disequilibrium (r 2 > 0.8) with other SNPs in our analysis. All SNPs were tested for deviation from Hardy–Weinberg equilibrium (HWE) in the controls using PLINK (version 1.04; http://pngu.mgh.harvard.edu/purcell/plink/ ) ( Purcell et al., 2007 ). SNPs with P < 0.001 in controls were considered in Hardy–Weinberg disequilibrium (HWD) and excluded. SNPs with minor allele frequency (MAF) below 0.01, SNPs with genotype missingness rate above 5% and individuals with genotype missingness rate above 5% were also excluded. Single SNP association tests ( Aim 1 ) were performed in PLINK, investigating best-fitting model and inheritance pattern (dominant gene action test, recessive gene action test, genotypic test, allelic test and Cochran–Armitage trend test). We also did gender specific analyses. Correction for multiple testing was done with Bonferroni for the number of SNPs tested in each gene. 2.4.2 Sub-sample PASW Statistics 18.0 (SPSS, Inc., Chicago, Ill., USA) was used for analyses of characteristics and association analyses of systemic cortisol metabolism and genotypes in the sub-sample. Indexes of enzyme activities had skewed distribution and were logarithmically transformed for comparisons between groups with ANCOVA. Explanatory variables included in the analysis for Aim 2 were “genotype category”, “diagnostic group” (BD or SCZ, or BD and SCZ combined, and healthy controls) and adjustments with backward elimination for age, gender, BMI, urinary creatinine and time of sampling as main effects. Only significant SNPs and diagnostic groups from Aim 1 were analyzed, and the genotypes were categorized to fit the relevant genetic model from the PLINK analyses (category of combined genotypes indicated with “genotype + genotype” and category of single genotype indicated with “genotype”). In the analyses for Aim 3 , significant SNPs from aim 1 were analyzed with a similar model as in Aim 2, with the addition of the “genotype category x diagnostic group” interaction. Pairwise comparisons based on estimated marginal means with Bonferroni correction were used to compare enzyme activities between genotype categories (differences indicated with “genotype category > genotype category”) within the “diagnostic group” categories. For testing group differences in age, BMI, urinary creatinine, time of sampling and total scores of PANSS, IDS-C and YMRS, we used Kruskal–Wallis H test/Mann–Whitney U test. Differences in gender were analyzed with the chi-square test. For all statistical analyses the significance level was set at 0.05 (two-tailed). 3 Results 3.1 Total sample 3.1.1 Genotyping After pruning out SNPs in high linkage disequilibrium (LD) (r 2 > 0.8) with other SNPs in our analysis, 46 out of 102 SNPs were included in the BD case–control sample; total genotyping rate for the remaining individuals was 99.1%. 50 out of 102 SNPs were included in the SCZ case–control sample; total genotyping rate for the remaining individuals was 99.2%. 1 marker failed the HWE test in controls. 3.1.2 Diagnosis association analysis rs6732223 7 kb downstream of SRD5A2 (5α-reductase) was significantly associated with SCZ with healthy controls as reference group, using a dominant model with T as dominant allele (nominal p = 0.0043, Bonferroni corrected p = 0.030, risk allele: T, Table 2 ). There were other SNPs nominally associated with SCZ, BD or the combined BD and SCZ group, but none withstanding Bonferroni correction, and no significant findings in males and females separately. 3.2 Sub-sample 3.2.1 Enzyme activity association analysis There was no significant difference in 5α-reductase activity measured as allo-THF/UFF when comparing genotype categories CC and CT + TT (dominant T allele model) of SNP rs6732223 (F = 0.41, DF = 1, p = 0.52) in the sub-sample of subjects with SCZ and healthy controls. The “genotype category × diagnostic group” interaction was significant for rs6732223 with the genotype categories CC and CT + TT; there was a significant difference in enzyme activities between the genotype categories within the SCZ group, and no significant difference between genotype categories within the healthy control group (interaction: F = 6.40, df = 1, p = 0.012; pairwise comparison within SCZ: CT + TT > CC, F = 6.61, df = 1, p = 0.011, Fig. 1 ). 4 Discussion This is the first study to suggest an interaction between SCZ and SRD5A2 coding for the steroid metabolizing enzyme 5α-reductase, which converts testosterone into 5α-dihydrotestosterone and progesterone and cortisol into their respective 5α-3-oxo-steroids; the increase in estimated 5α-reductase activity measured as urinary allo-THF/UFF ratio in SZC patients with the T allele, was indicated by a significant association of rs6732223 with SCZ. This is in line with recent findings of increased cortisol metabolism in psychosis spectrum disorders ( Steen et al., in press ). The genes selected in the present study all code for cortisol metabolizing enzymes. 5α-reductase type 1 ( SRD5A1 ), 5β-reductase ( AKR1D1 ), 11β-HSD1 ( HSD11B1 ) and 11β-HSD2 ( HSD11B2 ) are known cortisol metabolizing enzymes ( Tomlinson and Stewart, 2001 ), while 5α-reductase type 2 ( SRD5A2 ) traditionally is associated with conversion of testosterone to 5α-dihydrotestosterone in reproductive tissues ( Penning, 2010 ). However, there is evidence for 5α-reductase type 2 involvement in human cortisol metabolism; it is expressed in hepatic tissue ( Russell and Wilson, 1994 ) where it seems the most important of the two isoenzymes ( Tomlinson et al., 2008 ), subjects with a 5α-reductase type 2 deficiency have altered urinary cortisol metabolite excretion ratios ( Peterson et al., 1985 ), and finasteride, a selective inhibitor of this isoenzyme, alters urinary cortisol metabolite excretion ratios even in low doses ( Imperato-McGinley et al., 1990 ). rs6732223 is located 7 kb downstream of SRD5A2 , thus the mechanism of its impact is probably through the expression of SRD5A2 . However, the molecular mechanism of this SNP is not known. The indexes of urinary measurements estimating cortisol metabolizing enzyme activities are widely used, and correlate with absolute measurements of urinary metabolites ( Romer et al., 2009; Yehuda et al., 2009a ). This strongly suggests that reported significant differences in urinary ratios reflect differences in enzyme activities. The current association of SRD5A2 with cortisol metabolism in SCZ, is in accordance with our recent findings of increased systemic enzyme activities of 5α-reductase, 5β-reductase and 11β-HSD2 in BD and SCZ compared to healthy controls ( Steen et al., in press ). The reason for lack of significant associations with BD in the present analyses could be due to genotype interactions or insufficient sample size. Furthermore, the same polymorphisms do not necessarily affect HPA axis regulation in SCZ and BD. Generally, numerous genetic variants can potentially contribute to disease susceptibility in converging biological pathways; all enzymes included in this study are able to modulate the activity of the HPA axis ( Johnstone et al . , 2004; Morita et al . , 2004; Vogeser et al . , 2002 ), thus, variants of the investigated polymorphisms could contribute to HPA axis dysregulation in these disorders. In line with this, several SNPs were nominally associated with the diagnostic groups, although not withstanding Bonferroni correction. Thus, the present study provides some support for genetic increased cortisol metabolism across the psychosis spectrum. However, due to small effects, huge samples are needed to confirm genetic associations in psychiatric disorders ( Purcell et al . , 2009; Shi et al . , 2009; Stefansson et al . , 2009 ). The present study sample is relatively small and was used to indicate which SNPs to analyze with enzyme data. A dominant T allele model fits with the results of increased 5α-reductase activity in the combined CT + TT genotype category relative to the CC genotype category within the SCZ group. We found no significant impact of the risk genotype category (CT + TT) as main effect on enzyme activity, only an effect in the SCZ group; this indicates that the effect is an interaction of rs6732223 with specific biological alterations in the schizophrenia spectrum group, as shown for other genetic variations (e.g. Prata et al., 2009a, b; Wirgenes et al . , 2010 ). Thus, the present data suggest a role of 5α-reductase in psychosis spectrum disorders pathophysiology. The HPA axis is thoroughly studied in BD and SCZ. Both disorders seem to have a similar dysfunction of the axis; a hyperdrive is documented with several methods including the dexamethasone/corticotropin-releasing hormone test, basal measurements of cortisol and imaging of HPA axis components ( Daban et al., 2005; Pariante et al., 2005; Takahashi et al., 2009; Walker et al., 2008 ). The mechanism of HPA axis dysfunctions in these and especially affective disorders has been the focus of research for many years ( Holsboer, 2000 ). The HPA axis is a fine balanced equilibrium, and Plotsky et al. (1998) argues that dysfunctions may arise from alterations at any level of the axis. The association of genotype and enzyme activity in the present study relates to systemic cortisol metabolism ( Tomlinson and Stewart, 2001; Ulick et al., 1992 ). Cortisol metabolizing enzymes are widespread in the body ( Tomlinson and Stewart, 2001 ), and it is likely that local action of enzymes within the brain affects negative feedback ( Harris et al., 2001 ). However, recent evidence indicates that also systemic cortisol metabolism influences HPA axis activity ( Paterson et al., 2007 ). The present association of a gene variant in a cortisol metabolizing enzyme with enzyme activity within SCZ, is thus in accordance with previous HPA axis findings and the increased systemic enzyme activity reported in BD and SCZ ( Steen et al., in press ). The effect of genotype category might be considered a secondary effect of other biological alterations in SCZ, however, genotype seems important regardless of this due to the metabolisms' influence on the axis. The study has some limitations. Rare genotypes make the study vulnerable for missing true differences (type II errors) between groups, and may explain negative results in BD. It would be preferable (but not possible) to analyze enzyme activities separately for each of the two isoenzymes, but despite interference from isoenzyme type 1, we were able to detect suggested differences in enzyme activities in different genotype categories in a region close to isoenzyme type 2, indicating an even more significant effect if the isoenzymes had been separately analyzed. Furthermore, few subjects in the CC genotype category is a disadvantage, however, the enzyme data fitted the dominant model from the single SNP association test, and there was enough power to reveal a significant difference. To the best of our knowledge, there are no other studies regarding the functional activity of rs6732223, thus the reported associations currently lack support from data on molecular mechanisms. The use of psychopharmacological agents and differences in symptom severity might interfere with estimated effects, however, the previous report on enzyme activities with larger patient groups, did not find significant effects of total scores of symptom scales or various psychopharmacological agents on enzyme activity ( Steen et al., in press ). Finally, the imputation carries risk to underestimate variances, however, the method is easy to comprehend, preserves data and is used in the psychiatric literature (e.g. Levinson et al . , 2010; Wakschlag et al . , 2009 ). 5 Conclusion This is the first study to suggest that the gene SRD5A2 coding for the steroid metabolizing enzyme 5α-reductase is involved in SCZ pathophysiology. The relationship was hypothesized from previous work on cortisol metabolizing enzyme activities in BD and SCZ ( Steen et al., in press ). The current results of an increase in estimated 5α-reductase activity with the risk allele, indicated by a significant SNP association, support that genetically determined increased cortisol metabolism is part of the HPA axis pathophysiology in SCZ. The findings of abnormal cortisol metabolism are new in psychosis spectrum disorders. If replicated in larger studies, the findings could have an impact on the understanding of the role of the HPA axis in disease mechanisms, and may serve as a basis for novel drug targets in these disorders. Acknowledgements The study was supported by an award from The Lundbeck Foundation, Lysaker, Norway , and by grants to the TOP study group from the University of Oslo, Oslo, Norway ; Oslo University Hospital, Oslo, Norway ; the Research Council of Norway, Oslo, Norway (# 167153/V50 , # 163070/V50 ) and the South - Eastern Norway Regional Health Authority, Hamar, Norway ( # 2004-123 , # 2008-039 ). The authors thank the patients and controls for participating in the study, and the TOP study group members for contributing with data collection. References Addington et al., 1990 D. Addington J. Addington B. Schissel A depression rating scale for schizophrenics Schizophr Res 3 1990 247 251 Anagnostis et al., 2009 P. Anagnostis V.G. Athyros K. Tziomalos A. Karagiannis D.P. Mikhailidis Clinical review: the pathogenetic role of cortisol in the metabolic syndrome: a hypothesis J Clin Endocrinol Metab 94 2009 2692 2701 Andrew et al., 1998 R. Andrew D.I. Phillips B.R. Walker Obesity and gender influence cortisol secretion and metabolism in man J Clin Endocrinol Metab 83 1998 1806 1809 Barnett and Smoller, 2009 J.H. Barnett J.W. Smoller The genetics of bipolar disorder Neuroscience 164 2009 331 343 Biondi and Picardi, 1999 M. Biondi A. Picardi Psychological stress and neuroendocrine function in humans: the last two decades of research Psychother Psychosom 68 1999 114 150 Birkenaes et al., 2007 A.B. Birkenaes S. Opjordsmoen C. Brunborg J.A. Engh H. Jonsdottir P.A. Ringen The level of cardiovascular risk factors in bipolar disorder equals that of schizophrenia: a comparative study J Clin Psychiatry 68 2007 917 923 Blumberg et al., 2003 H.P. Blumberg J. Kaufman A. Martin R. Whiteman J.H. Zhang J.C. Gore Amygdala and hippocampal volumes in adolescents and adults with bipolar disorder Arch Gen Psychiatry 60 2003 1201 1208 Cichon et al., 2009 S. Cichon N. Craddock M. Daly S.V. Faraone P.V. Gejman J. Kelsoe Genomewide association studies: history, rationale, and prospects for psychiatric disorders Am J Psychiatry 166 2009 540 556 Craddock et al., 2009 N. Craddock M.C. O'Donovan M.J. Owen Psychosis genetics: modeling the relationship between schizophrenia, bipolar disorder, and mixed (or “schizoaffective”) psychoses Schizophr Bull 35 2009 482 490 Daban et al., 2005 C. Daban E. Vieta P. Mackin A.H. Young Hypothalamic-pituitary-adrenal axis and bipolar disorder Psychiatr Clin North Am 28 2005 469 480 de Kloet, 2008 E.R. de Kloet About stress hormones and resilience to psychopathology J Neuroendocrinol 20 2008 885 892 De Luca et al., 2007 V. De Luca S. Tharmalingam J.L. Kennedy Association study between the corticotropin-releasing hormone receptor 2 gene and suicidality in bipolar disorder Eur Psychiatry 22 2007 282 287 Djurovic et al., 2009 S. Djurovic A.K. Kahler B. Kulle E.G. Jonsson I. Agartz H.S. Le A possible association between schizophrenia and GRIK3 polymorphisms in a multicenter sample of Scandinavian origin (SCOPE) Schizophr Res 107 2009 242 248 Duan et al., 2005 S. Duan R. Gao Q. Xing J. Du Z. Liu Q. Chen A family-based association study of schizophrenia with polymorphisms at three candidate genes Neurosci Lett 379 2005 32 36 Ebdrup et al., 2010 B.H. Ebdrup B. Glenthoj H. Rasmussen B. Aggernaes A.R. Langkilde O.B. Paulson Hippocampal and caudate volume reductions in antipsychotic-naive first-episode schizophrenia J Psychiatry Neurosci 35 2010 95 104 Fallin et al., 2005 M.D. Fallin V.K. Lasseter D. Avramopoulos K.K. Nicodemus P.S. Wolyniec J.A. McGrath Bipolar I disorder and schizophrenia: a 440-single-nucleotide polymorphism screen of 64 candidate genes among Ashkenazi Jewish case–parent trios Am J Hum Genet 77 2005 918 936 First et al., 1995 M. First R. Spitzer M. Gibbon J.B.W. Williams Structured Clinical Interview for DSM-IV Axis I Disorders: Patient Edition (SCID-P), Version 2 1995 Biometrics Research, New York State Psychiatric Institute New York Gallagher et al., 2008 P. Gallagher S. Watson D.C. Elizabeth A.H. Young F.I. Nicol Persistent effects of mifepristone (RU-486) on cortisol levels in bipolar disorder and schizophrenia J Psychiatr Res 42 2008 1037 1041 Grover et al., 2009 D. Grover R. Verma F.S. Goes P.L. Mahon E.S. Gershon F.J. McMahon Family-based association of YWHAH in psychotic bipolar disorder Am J Med Genet B Neuropsychiatr Genet 150B 2009 977 983 Harris et al., 2001 H.J. Harris Y. Kotelevtsev J.J. Mullins J.R. Seckl M.C. Holmes Intracellular regeneration of glucocorticoids by 11beta-hydroxysteroid dehydrogenase (11beta-HSD)-1 plays a key role in regulation of the hypothalamic-pituitary-adrenal axis: analysis of 11beta-HSD-1-deficient mice Endocrinology 142 2001 114 120 Herman et al., 2005 J.P. Herman M.M. Ostrander N.K. Mueller H. Figueiredo Limbic system mechanisms of stress regulation: hypothalamo-pituitary-adrenocortical axis Prog Neuropsychopharmacol Biol Psychiatry 29 2005 1201 1213 Holsboer, 2000 F. Holsboer The corticosteroid receptor hypothesis of depression Neuropsychopharmacology 23 2000 477 501 Imperato-McGinley et al., 1990 J. Imperato-McGinley C. Shackleton S. Orlic E. Stoner C19 and C21 5 beta/5 alpha metabolite ratios in subjects treated with the 5 alpha-reductase inhibitor finasteride: comparison of male pseudohermaphrodites with inherited 5 alpha-reductase deficiency J Clin Endocrinol Metab 70 1990 777 782 Jerjes et al., 2006 W.K. Jerjes T.J. Peters N.F. Taylor P.J. Wood S. Wessely A.J. Cleare Diurnal excretion of urinary cortisol, cortisone, and cortisol metabolites in chronic fatigue syndrome J Psychosom Res 60 2006 145 153 Johnstone et al., 2004 A.M. Johnstone P. Faber R. Andrew E.R. Gibney M. Elia G. Lobley Influence of short-term dietary weight loss on cortisol secretion and metabolism in obese men Eur J Endocrinol 150 2004 185 194 Kay et al., 1987 S.R. Kay A. Fiszbein L.A. Opler The positive and negative syndrome scale (PANSS) for schizophrenia Schizophr Bull 13 1987 261 276 Kleinman et al., 2003 L. Kleinman A. Lowin E. Flood G. Gandhi E. Edgell D. Revicki Costs of bipolar disorder Pharmacoeconomics 21 2003 601 622 Levinson et al., 2010 D. Levinson M.D. Lakoma M. Petukhova M. Schoenbaum A.M. Zaslavsky M. Angermeyer Associations of serious mental illness with earnings: results from the WHO World Mental Health surveys Br J Psychiatry 197 2010 114 121 Mangalore and Knapp, 2007 R. Mangalore M. Knapp Cost of schizophrenia in England J Ment Health Policy Econ 10 2007 23 41 Marco et al., 2002 E.J. Marco O.M. Wolkowitz S. Vinogradov J.H. Poole J. Lichtmacher V.I. Reus Double-blind antiglucocorticoid treatment in schizophrenia and schizoaffective disorder: a pilot study World J Biol Psychiatry 3 2002 156 161 Merikangas et al., 2007 K.R. Merikangas H.S. Akiskal J. Angst P.E. Greenberg R.M. Hirschfeld M. Petukhova Lifetime and 12-month prevalence of bipolar spectrum disorder in the National Comorbidity Survey replication Arch Gen Psychiatry 64 2007 543 552 Mondelli et al., 2008 V. Mondelli P. Dazzan A. Gabilondo K. Tournikioti M. Walshe N. Marshall Pituitary volume in unaffected relatives of patients with schizophrenia and bipolar disorder Psychoneuroendocrinology 33 2008 1004 1012 Morita et al., 2004 H. Morita Y. Isomura T. Mune H. Daido R. Takami N. Yamakita Plasma cortisol and cortisone concentrations in normal subjects and patients with adrenocortical disorders Metabolism 53 2004 89 94 Moutsatsou et al., 2000 P. Moutsatsou A. Tsolakidou G. Trikkas C. Troungos C.E. Sekeris Glucocorticoid receptor alpha and beta isoforms are not mutated in bipolar affective disorder Mol Psychiatry 5 2000 196 202 Owen et al., 2009 M.J. Owen H.J. Williams M.C. O'Donovan Schizophrenia genetics: advancing on two fronts Curr Opin Genet Dev 19 2009 266 270 Pariante et al., 2005 C.M. Pariante P. Dazzan A. Danese K.D. Morgan F. Brudaglio C. Morgan Increased pituitary volume in antipsychotic-free and antipsychotic-treated patients of the AEsop first-onset psychosis study Neuropsychopharmacology 30 2005 1923 1931 Paterson et al., 2007 J.M. Paterson M.C. Holmes C.J. Kenyon R. Carter J.J. Mullins J.R. Seckl Liver-selective transgene rescue of hypothalamic-pituitary-adrenal axis dysfunction in 11beta-hydroxysteroid dehydrogenase type 1-deficient mice Endocrinology 148 2007 961 966 Penning, 2010 T.M. Penning New frontiers in androgen biosynthesis and metabolism Curr Opin Endocrinol Diabetes Obes 17 2010 233 239 Perala et al., 2007 J. Perala J. Suvisaari S.I. Saarni K. Kuoppasalmi E. Isometsa S. Pirkola Lifetime prevalence of psychotic and bipolar I disorders in a general population Arch Gen Psychiatry 64 2007 19 28 Peterson et al., 1985 R.E. Peterson J. Imperato-McGinley T. Gautier C. Shackleton Urinary steroid metabolites in subjects with male pseudohermaphroditism due to 5 alpha-reductase deficiency Clin Endocrinol (Oxf) 23 1985 43 53 Plotsky et al., 1998 P.M. Plotsky M.J. Owens C.B. Nemeroff Psychoneuroendocrinology of depression. Hypothalamic-pituitary-adrenal axis Psychiatr Clin North Am 21 1998 293 307 Poor et al., 2004 V. Poor S. Juricskay A. Gati P. Osvath T. Tenyi Urinary steroid metabolites and 11beta-hydroxysteroid dehydrogenase activity in patients with unipolar recurrent major depression J Affect Disord 81 2004 55 59 Poor et al., 2005 V. Poor A. Bufa I. Biro E. Telegdy T. Tenyi A. Gati Urinary steroid measurements in some endocrine and psychiatric diseases Curr Med Chem 12 2005 1339 1342 Prata et al., 2009a D.P. Prata A. Mechelli C.H. Fu M. Picchioni T. Toulopoulou E. Bramon Epistasis between the DAT 3′ UTR VNTR and the COMT Val158Met SNP on cortical function in healthy subjects and patients with schizophrenia Proc Natl Acad Sci USA 106 2009 13600 13605 Prata et al., 2009b D.P. Prata A. Mechelli M.M. Picchioni C.H. Fu T. Toulopoulou E. Bramon Altered effect of dopamine transporter 3′UTR VNTR genotype on prefrontal and striatal function in schizophrenia Arch Gen Psychiatry 66 2009 1162 1172 Purcell et al., 2007 S. Purcell B. Neale K. Todd-Brown L. Thomas M.A. Ferreira D. Bender PLINK: a tool set for whole-genome association and population-based linkage analyses Am J Hum Genet 81 2007 559 575 Purcell et al., 2009 S.M. Purcell N.R. Wray J.L. Stone P.M. Visscher M.C. O'Donovan P.F. Sullivan Common polygenic variation contributes to risk of schizophrenia and bipolar disorder Nature 460 2009 748 752 Rasmuson et al., 2001 S. Rasmuson R. Andrew B. Nasman J.R. Seckl B.R. Walker T. Olsson Increased glucocorticoid production and altered cortisol metabolism in women with mild to moderate Alzheimer's disease Biol Psychiatry 49 2001 547 552 Raven and Taylor, 1996 P.W. Raven N.F. Taylor Evidence for independent modulation of human 11-HSD and 5 alpha/5 beta reductase activities Endocr Res 22 1996 811 815 Raven and Taylor, 1998 P.W. Raven N.F. Taylor 11Beta-HSD and 17beta-HSD as biological markers of depression: sex differences and correlation with symptom severity Endocr Res 24 1998 659 662 Romer et al., 2009 B. Romer S. Lewicka D. Kopf F. Lederbogen B. Hamann M. Gilles Cortisol metabolism in depressed patients and healthy controls Neuroendocrinology 90 2009 301 306 Rush et al., 1996 A.J. Rush C.M. Gullion M.R. Basco R.B. Jarrett M.H. Trivedi The Inventory of Depressive Symptomatology (IDS): psychometric properties Psychol Med 26 1996 477 486 Russell and Wilson, 1994 D.W. Russell J.D. Wilson Steroid 5 alpha-reductase: two genes/two enzymes Annu Rev Biochem 63 1994 25 61 Shi et al., 2009 J. Shi D.F. Levinson J. Duan A.R. Sanders Y. Zheng I. Pe'er Common variants on chromosome 6p22.1 are associated with schizophrenia Nature 460 2009 753 757 Spitzer et al., 1994 R.L. Spitzer J.B. Williams K. Kroenke M. Linzer F.V. deGruy III S.R. Hahn Utility of a new procedure for diagnosing mental disorders in primary care. The PRIME-MD 1000 study JAMA 272 1994 1749 1756 Steen et al., in press Steen NE, Methlie P, Lorentzen S, Hope S, Barrett EA, Larsson S, et al. Increased systemic cortisol metabolism in patients with schizophrenia and bipolar disorder; a mechanism for increased stress vulnerability? J Clin Psychiatry (in press), doi:10.4088/JCP.10m06068yel . Stefansson et al., 2009 H. Stefansson R.A. Ophoff S. Steinberg O.A. Andreassen S. Cichon D. Rujescu Common variants conferring risk of schizophrenia Nature 460 2009 744 747 Stewart et al., 1990 P.M. Stewart C.H. Shackleton G.H. Beastall C.R. Edwards 5 alpha-reductase activity in polycystic ovary syndrome Lancet 335 1990 431 433 Stratakis et al., 1997 C.A. Stratakis N.J. Sarlis W.H. Berrettini J.A. Badner G.P. Chrousos E.S. Gershon Lack of linkage between the corticotropin-releasing hormone (CRH) gene and bipolar affective disorder Mol Psychiatry 2 1997 483 485 Takahashi et al., 2009 T. Takahashi G.S. Malhi S.J. Wood M. Walterfang M. Yucel V. Lorenzetti Increased pituitary volume in patients with established bipolar affective disorder Prog Neuropsychopharmacol Biol Psychiatry 33 2009 1245 1249 Tomlinson and Stewart, 2001 J.W. Tomlinson P.M. Stewart Cortisol metabolism and the role of 11beta-hydroxysteroid dehydrogenase Best Pract Res Clin Endocrinol Metab 15 2001 61 78 Tomlinson et al., 2008 J.W. Tomlinson J. Finney B.A. Hughes S.V. Hughes P.M. Stewart Reduced glucocorticoid production rate, decreased 5alpha-reductase activity, and adipose tissue insulin sensitization after weight loss Diabetes 57 2008 1536 1543 Ulick et al., 1992 S. Ulick R. Tedde J.Z. Wang Defective ring A reduction of cortisol as the major metabolic error in the syndrome of apparent mineralocorticoid excess J Clin Endocrinol Metab 74 1992 593 599 Vogeser et al., 2002 M. Vogeser V. Gulberg T. Waggershauser A.L. Gerbes In-vivo quantification of hepatic 11beta-hydroxysteroid dehydrogenase type I activity—a preliminary study Clin Biochem 35 2002 655 657 Wakschlag et al., 2009 L.S. Wakschlag E.O. Kistner D.S. Pine G. Biesecker K.E. Pickett A.D. Skol Interaction of prenatal exposure to cigarettes and MAOA genotype in pathways to youth antisocial behavior Mol Psychiatry 2009 Mar 3. [Epub ahead of print] Walker et al., 2008 E. Walker V. Mittal K. Tessner Stress and the hypothalamic pituitary adrenal axis in the developmental course of schizophrenia Annu Rev Clin Psychol 4 2008 189 216 Watson et al., 2004 S. Watson P. Gallagher J.C. Ritchie I.N. Ferrier A.H. Young Hypothalamic-pituitary-adrenal axis function in patients with bipolar disorder Br J Psychiatry 184 2004 496 502 Weber et al., 2000 B. Weber S. Lewicka M. Deuschle M. Colla P. Vecsei I. Heuser Increased diurnal plasma concentrations of cortisone in depressed patients J Clin Endocrinol Metab 85 2000 1133 1136 Willour et al., 2009 V.L. Willour H. Chen J. Toolan P. Belmonte D.J. Cutler F.S. Goes Family-based association of FKBP5 in bipolar disorder Mol Psychiatry 14 2009 261 268 Wirgenes et al., 2010 K.V. Wirgenes S. Djurovic K. Sundet I. Agartz M. Mattingsdal L. Athanasiu Catechol O-methyltransferase variants and cognitive performance in schizophrenia and bipolar disorder versus controls Schizophr Res 2010 May 31. [Epub ahead of print] Wong et al., 2003 A.H. Wong F. Macciardi T. Klempan W. Kawczynski C.L. Barr S. Lakatoo Identification of candidate genes for psychosis in rat models, and possible association between schizophrenia and the 14-3-3eta gene Mol Psychiatry 8 2003 156 166 Yehuda et al., 2009a R. Yehuda L.M. Bierer R. Andrew J. Schmeidler J.R. Seckl Enduring effects of severe developmental adversity, including nutritional deprivation, on cortisol metabolism in aging Holocaust survivors J Psychiatr Res 43 2009 877 883 Yehuda et al., 2009b R. Yehuda L.M. Bierer C. Sarapas I. Makotkine R. Andrew J.R. Seckl Cortisol metabolic predictors of response to psychotherapy for symptoms of PTSD in survivors of the World Trade Center attacks on September 11, 2001 Psychoneuroendocrinology 34 2009 1304 1313 Young et al., 1978 R.C. Young J.T. Biggs V.E. Ziegler D.A. Meyer A rating scale for mania: reliability, validity and sensitivity Br J Psychiatry 133 1978 429 435 Young et al., 2004 A.H. Young P. Gallagher S. Watson D. Del-Estal B.M. Owen I.N. Ferrier Improvements in neurocognitive function and mood following adjunctive treatment with mifepristone (RU-486) in bipolar disorder Neuropsychopharmacology 29 2004 1538 1545