UKPMC Funders Group Author Manuscript Mol Psychiatry. Author manuscript; available in PMC 2009 January 22.

Transcription

1 UKPMC Funders Group Author Manuscript Published in final edited form as: Mol Psychiatry July ; 13(7): doi: /mp D-Amino acid oxidase (DAO) activity and expression are increased in schizophrenia PWJ Burnet, SL Eastwood, GC Bristow, BR Godlewska, P Sikka, M Walker, and PJ Harrison Department of Psychiatry, University of Oxford, Warneford Hospital, Oxford, OX3 7JX, United Kingdom. Keywords Glutamate; bipolar disorder; DAAO; NMDA receptor; D-serine; polymorphism D-amino acid oxidase (DAO, DAAO) metabolises the NMDA receptor (NMDAR) modulator D-serine. 1 Enhanced DAO activity is thus a potential cause of reduced D-serine and thence impaired NMDAR functioning in schizophrenia, 2 and an explanation for the genetic contribution that DAO may make to the disorder. 3,4 Here we report that DAO activity, and expression, are increased in the cerebellum in schizophrenia, but not in bipolar disorder, and are not related to SNPs in DAO or its putative activator gene G72/G30. 3 We measured DAO enzyme activity, and DAO mrna by qrt-pcr, using established methods, in cerebellar tissue from subjects with schizophrenia, bipolar disorder, and controls, from the Stanley Microarray Collection (Supplementary Table 1). We genotyped subjects for two schizophrenia-associated tag SNPs in DAO and G72/G30. We used cerebellar tissue from rats administered haloperidol or clozapine for 14 days to investigate antipsychotic effects on DAO activity. For demographic and methodological details, see Supplementary Materials. As shown in Fig. 1A, DAO activity differed between groups (F 2,100 =4.79, p=0.010), being increased in schizophrenia compared to controls (+37%, p=0.027) and compared to bipolar disorder (+57%, p=0.004). The latter groups did not differ (p=0.45). DAO activity was not correlated with post mortem interval, brain ph, or antipsychotic exposure (Supplementary Fig. 1) or age (R=0.03, p=0.79). DAO activity was unrelated to alcohol or substance misuse history (using a five-point scale), smoking, sex, or suicide. DAO activity increased with duration of schizophrenia (R=0.34, p=0.05, n=35). Within the bipolar disorder group, DAO activity did not vary according to a history of psychotic symptoms. DAO activity correlated with DAO mrna (R=0.43, p=0.0001, n=97). DAO affinity showed no group differences (Fig. 1B). DAO activity in rat cerebellum was unaffected by antipsychotics (Supplementary Table 2). Normalised DAO mrna differed between groups (F 2,88 =3.84, p=0.025), being increased compared to controls in schizophrenia (p=0.01) and bipolar disorder (p=0.04; Fig. 1C). DAO mrna was not related to antipsychotic exposure, post mortem interval, brain ph, or RIN, but was affected by death-to-refrigeration interval (Supplementary Fig. 1). Neither DAO activity nor DAO mrna were influenced by the DAO or G72/G30 SNPs (Supplementary Table 3), apart from a trend for higher DAO activity in allele 2 carriers of Correspondence: Dr Harrison. paul.harrison@psych.ox.ac.uk Tel: Fax

2 Burnet et al. Page 2 G72/G30 rs (p=0.08 vs 1/1 homozygotes). There were no genotype-by-diagnosis interactions. Our data show an elevation of DAO activity in schizophrenia accompanied by increased gene expression. The activity increase confirms the results of a pilot study, 5 and the elevated DAO mrna replicates findings in a separate cohort. 6 There were no correlations of DAO activity or expression with medication exposure, and no effect of haloperidol or clozapine on DAO activity in rat brain. Other confounders (e.g. smoking) did not have a demonstrable influence either. Hence, within the constraints of a post mortem study, our findings appear related to the diagnosis of schizophrenia. We focused on the cerebellum because DAO is abundant and active therein; it will be important to ascertain whether DAO activity is affected elsewhere in schizophrenia, but problematic since forebrain DAO activity is extremely low, despite robust expression. Parenthetically, this discrepancy is complemented by a difference in the cellular localization of DAO: it is glial in the cerebellum, but mainly neuronal in the cerebral cortex. 6 Elevated DAO activity will presumably enhance metabolism of D-serine and, other things being equal, contribute to a reduced synaptic availability of D-serine, potentially impairing NMDAR function. 1 As such, our findings support the hypothesized involvement of DAO, and thence D-serine, in the NMDAR hypofunction of schizophrenia and its potential therapeutic amelioration. 2,7,8 Clearly any such conclusions are tentative, since many other factors likely also influence D-serine availability, e.g. its synthesis by serine racemase, 6,9 and its release and reuptake. 10 Alterations in these processes could counteract or exacerbate enhanced DAO activity. Moreover, since DAO also metabolises other D-amino acids, 1 D-serine is not the only substrate that might be affected by an increase in DAO activity. For example, D-alanine is present in the cerebellum, is an NMDAR modulator, and may be therapeutically beneficial in schizophrenia. 7 Overall, whilst a primary effect on D-serine, and thence NMDARs, is an attractive interpretation of the DAO increase in schizophrenia, further studies are needed to confirm the biochemical consequences, as well as the causes, of the elevation. (See Supplementary Materials for further discussion). As the DAO and G72/G30 SNPs did not influence DAO activity or expression, our data provide no evidence that they are functional (nor indexing SNPs that are), nor have we identified a mechanism whereby variation in these genes might confer schizophrenia susceptibility. There may be other SNPs that are relevant in this respect; alternatively, the DAO SNPs might operate at other places and times, whilst G72/G30 SNPs could act via a DAO-independent pathway. In summary, DAO activity and expression are increased in schizophrenia, but not related to DAO or G72/G30 genotype. Any pathophysiological consequences of increased DAO activity seem most likely to impact on D-serine metabolism and thence NMDAR function, but this remains to be established. Supplementary Material References Refer to Web version on PubMed Central for supplementary material. 1. Mothet JP, Parent AT, Wolosker H, Brady RO, Linden DJ, Ferris CD, et al. Proc Natl Acad Sci USA 2000;97: [PubMed: ] 2. Tsai GC, Coyle JT. Annu Rev Pharmacol Toxicol 2002;42: [PubMed: ] 3. Chumakov I, Blumenfeld M, Guerrasimenko O, Cavarec L, Palicio M, Abderrahim H, et al. Proc Natl Acad Sci USA 2002;99: [PubMed: ] 4. Boks MPM, Rietkerk T, van de Beek MH, Sommer IE, de Koning TJ, Kahn RS. Eur Neuropsychopharmacol 2007;17: [PubMed: ]

3 Burnet et al. Page 3 5. Kapoor R, Lim KS, Cheng A, Garrick T, Kapoor V. Brain Res 2006;1106: [PubMed: ] 6. Verrall L, Walker M, Rawlings N, Benzel I, Kew JNC, Harrison PJ, Burnet PJW. Eur J Neurosci 2007;26: [PubMed: ] 7. Javitt D. Biol Psychiatry 2008;63:6 8. [PubMed: ] 8. Adage T, Trillat A-C, Quattropani A, Perrin D, Cavarec L, Shaw J, et al. Eur Neuropsychopharmacol 2008;18: [PubMed: ] 9. Bendikov I, Nadri C, Amar S, Panizzutti R, De Miranda J, Wolosker H, Agam G. Schizophr Res 2007;90: [PubMed: ] 10. Burnet PJW, Hutchinson L, von Hesling M, Gilbert E-J, Harrison PJ. Schizophr Res. in press

4 Burnet et al. Page 4 Figure 1. Cerebellar DAO activity and expression in 35 control subjects (CON, white bars), 35 with schizophrenia (SCZ, black bars), and 34 with bipolar disorder (BPD, shaded bars). A) DAO activity (Vmax, nmol D-proline/min/mg protein) is higher in subjects with schizophrenia compared to controls and to subjects with bipolar disorder. Bipolar disorder subjects have lower DAO activity than those with schizophrenia. B) DAO affinity (Km, nm) is not different between groups. C) DAO mrna, measured by Taqman qrt-pcr and normalised to the geometric mean of four housekeeping genes (β2m, GAPDH, GUSB, TFRC; see Supplementary Materials), is increased in subjects with schizophrenia and in bipolar disorder. 5 subjects (1 CON, 1 SCZ, 3 BPD) were excluded due to amplification failure. The death-to-

5 Burnet et al. Page 5 refrigeration interval was included as a covariate in this analysis as it correlated negatively with DAO mrna (Supplementary Figure 1D). Repeating the ANOVA without this covariate retained the overall diagnostic effect (F 2,95 =3.422, p=0.037) and the increase in schizophrenia (p=0.01) but rendered the bipolar disorder increase non-significant (p=0.22).