Abstract
Background
Pharmacological, neurochemical and electrophysiological studies provide compelling evidence that N-methyl-D-aspartate-receptor (NMDAR) hypofunction is a pathologic feature of schizophrenia (SZ). GWAS studies highlighted risk genes such as Serine Racemase (SRR), which synthesizes D-Serine (D-Ser), the co-agonist of glycine site at NMDARs, and Serine Hydroxymethyltransferase (SHMT1) which synthesizes L-Serine (L-Ser), the substrate of SRR. Around 30% of patients do not respond to dopamine modulation and are considered to suffer from treatment resistant SZ (TRS). While the exact cause of TRS remains unclear, multiple lines of evidence suggest the involvement of a dysregulation of the Glutamate (Glu) neurotransmission. To test the hypothesis whether Glu system dysregulation mediated by NMDAR hypofunction is an underlying mechanism of TRS, we investigate the Glu and D-Ser pathways in TRS and treatment responder (RESP) early psychosis patients (EPP).
Methods
From a total of 621 EPP aged 18 to 35, included in the TIPP (early Intervention Program in Lausanne, Baumann & al., 2013), 225 EPP were classified as TRS (n=33) or RESP (n=192) according to the strict Treatment Response and Resistance in Psychosis (TRRIP) criteria (Howes & al., 2017), with compliance ascertained by antipsychotic plasma levels. A matched healthy control (HC) group (N=114) was also recruited (DIGS criteria). No patient was taking clozapine at baseline. Clinical data was collected over a 3-year period. At baseline, following systems were assessed: 1) D-Ser pathway: plasmatic D-Ser, L-Ser and Glycine by HPLC (Hashimoto & al., 2016), protein levels of SRR and SHMT1 by ELISA; 2) Glu pathway: Glu and glutamine in plasma (HPLC) and prefrontal cortex (magnetic resonance spectroscopy, Xin & al., 2016).
Results
D-Ser pathway: in TRS, SRR levels were decreased by 56% as compared to RESP. Interestingly, we observed a positive correlation between plasma levels of D-Ser (SRR metabolite) and L-Ser (SRR substrate) in the TRS (r= 0.58; p =0.0015) but not in the RESP group, suggesting that SRR dysregulation might be a limiting factor in TRS patients. Moreover, in TRS patients, SHMT1 levels were decreased by 15% as compared to RESP, with a positive correlation between the substrate and metabolite of SHMT1, glycine and L-Ser (r =0.48; p =0.011). Dysregulation of SHMT1 might thus be a limiting factor in the TRS group. As compared to HC, L-Ser and D-Ser were significantly increased in patients (p <0.001 for L-Ser, p =0.0001 for D-Ser). However, no difference was observed in D-Ser, L-Ser and glycine in TRS as compared to RESP, although L-Ser tended to be higher in male TRS patients (p =0.06).
Glu pathway: comparing TRS with RESP patients, plasma Glu levels were increased in the TRS group (p <0.0001), whereas they were higher in both patient groups compared to HC (p <0.0001). Interestingly, plasma and brain Glu levels showed a negative correlation in EPP, mostly driven by RESP (r = -0.42; p =0.035), a correlation which was absent in HC.
Global Assessment of Functioning (GAF): at baseline, TRS and RESP displayed the same range of GAF. After a 3-year follow-up, TRS patients had poorer functioning as compared to RESP group (p <0.0001).
Discussion
Taken together, our results suggest that the TRS group, in which the levels of SRR and SHMT1 were lower and Glu plasma levels were higher, display a different regulation of the synthesis, degradation and/or accumulation of D-Ser and Glu as compared to the RESP group. However, replication in larger groups is needed. Moreover, our findings highlighted a dysregulation of D-Ser and Glu pathways in TRS patients in their early phase of psychosis. On the clinical side, our results confirm the significantly poorer functioning outcome in TRS patients.
Six months functional response to early psychosis intervention program best predicts outcome after three years
