PRIORITY COMMUNICATION Glycogen Synthase Kinase-3 Overexpression Replicates Electroretinogram Anomalies of Offspring at High Genetic Risk for Schizophrenia and Bipolar Disorder Joëlle Lavoie, Marc Hébert, and Jean-Martin Beaulieu Background: Electroretinogram (ERG) anomalies occur in patients with psychiatric disorders and represent potential biomarkers for diagnosis. For instance, decreased rod ERG (b-wave amplitude at Vmax) is a biological endophenotype in young offspring at high genetic risk (HR) for schizophrenia (SZ) and bipolar disorder (BD). Also, a decrease in cone a-wave and rod a- and b- wave was observed in SZ patients. However, the biological underpinning of these anomalies remains unknown. Several genetic variants associated with enhanced risk for SZ and/or BD can activate glycogen synthase kinase-3 isozymes (GSK3α and β). Here we examined the potential contribution of GSK3α and β in the modulation of the ERG. Methods: Cone and rod ERGs were recorded in mice having increased (prpGSK3β mice) or reduced (GSK3β +/– mice) GSK3β expression and in GSK3α knockout (KO) mice. Results: In prpGSK3β mice, we observed a decrease in rod b-wave amplitude at Vmax, whereas enhanced b-wave amplitude at Vmax was found in GSK3β +/– mice. An increase in cone a- and b-wave amplitude at Vmax and in rod b-wave amplitude at Vmax was observed in GSK3α-KO mice. Conclusions: GSK3 expression modulates some ERG parameters. The phenotype observed in prpGSK3β mice is consistent with observations made in HRs. ERG anomalies observed in GSK3β +/– and GSK3α-KO mice confirm an association between the rod and cone b-wave amplitude and the expression of GSK3 isozymes. Changes in GSK3 expression or activity may explain some ERG anomalies in HRs and patients, thus supporting the biological validity of ERG measurements as a valuable biomarker for psychiatric research. Key Words: Biomarker, bipolar disorder, electroretinogram, glycogen synthase kinase-3, risk factor, schizophrenia P sychiatry research and clinical practice are limited by difficulties in obtaining quantitative biological data from the living brain. There is therefore a need to identify reliable surrogate biomarkers to better characterize the pathophysiology of mental illnesses, allow evidence-based diagnosis, and develop personalized therapies. Because it is part of the central nervous system, the retina has been widely used as a site of investigation for understanding psychiatric disorders. Several studies reported retinal anomalies measured with the flash electroretinogram (ERG), and these anomalies may serve as biomarkers in patients with seasonal affective disorder (1–4), schizophrenia (SZ) (5,6), autism (7), and drug addiction (8). This noninvasive technique records the light- evoked biopotential originating from the retina in response to standardized flash stimuli. Using the ERG technique, cone (day vision) and rod (night vision) function can be assessed separately. Recently, a biological endophenotype consisting in a decreased rod ERG (b-wave amplitude at Vmax) and trends for a longer rod b-wave implicit time and a decreased cone a-wave amplitude were observed in young nonaffected offspring at high genetic risk (HR) of SZ or bipolar disorder (BD) (9). These ERG anomalies were independent of the diagnosis of the parents (either SZ or BD) and the age of the HRs. This biological signature was also observed in patients with SZ (among other ERG anomalies such as a decreased rod and cone a-wave amplitude) (6), suggesting that ERG profiles represent biomarkers of risk to develop major neuropsychiatric disorders, but their biological origin remains unknown. The signaling module composed by the serine/threonine kinases Akt and glycogen synthase kinase-3 (GSK3) has been associated with monoamine neurotransmission, psychiatric drug action, and several genetic risk factors for the development of SZ and BD, suggesting a crucial role of these molecules in the pathophysiology of mental disorders (10,11). Mammalian genomes encode three Akt (Akt1, 2, and 3) as well as two GSK3 (GSK3α and β) isozymes. Akt kinases are activated by phosphor- ylation in response to phosphoinositide 3-kinase mediated signal- ing (12). GSK3 isozymes are believed to be constitutively active and are inactivated by Akt following the phosphorylation of their amino-terminal regulatory domain (Ser21 for GSK3α and Ser9 for GSK3β) by Akt and other kinases (13,14). Pharmacologic agents used for the management of psychiatric disorders act on monoamines neurotransmission, especially dop- amine and serotonin (15–18). Among several mechanisms, part of their action on behavior is exerted by modulating the Akt–GSK3 signaling pathway (10). In short, both typical and atypical antipsychotics as well as lithium and several antidepressants have been shown to inhibit brain GSK3 activity in rodents (19–23). Conversely, several genetic variants associated with enhanced risk for SZ and/or BD can result in either direct or indirect GSK3 activation as a consequence of Akt inhibition (24,25). Indeed, Akt1 From the Centre de Recherche de l’Institut Universitaire en Santé Mentale de Québec (JL, MH, J-MB) and Departments of Psychiatry and Neuroscience (JL, J-MB) and Ophthalmology, Otorhinolaryngology, and Cervico-Facial Surgery (JL, MH), Faculty of Medicine, Université Laval, Quebec City, Canada. Address correspondence to Jean-Martin Beaulieu, Ph.D., 2601, Chemin de la Canadrière, F-6500, Quebec City, Quebec, Canada G1J 2G3; E-mail: martin.beaulieu@crulrg.ulaval.ca. Received Aug 5, 2013; revised Aug 26, 2013; accepted Aug 27, 2013. 0006-3223/$36.00 BIOL PSYCHIATRY 2013;]:]]]–]]] http://dx.doi.org/10.1016/j.biopsych.2013.08.035 & 2013 Society of Biological Psychiatry