Low serum truncated-BDNF isoform correlates with higher cognitive impairment in schizophrenia

https://doi.org/10.1016/j.jpsychires.2010.06.012Get rights and content

Abstract

Brain-derived neurotrophic factor (BDNF) is a key factor in learning and memory. Altered BDNF-signalling is thought to contribute to the pathogenesis of schizophrenia (SZ) especially in relation to cognitive deficits. However, analysis of serum BDNF as a potential biomarker in schizophrenia has provided controversial data. We hypothesized that these confounding results might be due to a differential regulation of BDNF precursor pro-BDNF (32 KDa) and proteolytic products mature (mat-BDNF; 14 KDa), and truncated-BDNF (28 KDa). Accordingly, we investigated the serum abundance of these BDNF isoforms and its relationship with cognitive impairment in schizophrenia. Schizophrenia was diagnosed with PANSS test. Abbreviated cognitive assessment included tests for attention, perceptual-motor skills, processing speed and memory. Using an ELISA assay, we found a slight reduction in serum BDNF levels in SZ patients (n = 40) with respect to healthy controls (HC, n = 40; p = 0.018). Western-blot analysis revealed increased serum pro-BDNF and mat-BDNF and reduced truncated-BDNF (p < 0.001) in SZ with respect to HC. Patients with an increase in pro-BDNF (n = 15/40) or mat-BDNF (n = 9/40) higher than the HC mean + 2 Standard Deviations (SD) also had >2SD reduction of truncated-BDNF (n = 27/40). Reduced truncated-BDNF correlated significantly with higher positive and lower negative PANNS scores and a worst performance in all cognitive assays but not with antipsychotic type. Measurement of serum truncated-BDNF abundance predicted for high cognitive deficits with sensitivity = 67.5%, specificity = 97.5%, Negative Predictive Value = 75% and Positive Predictive Value = 96.4%. These results suggest deficiency in pro-BDNF processing as a possible biological mechanism underlying schizophrenia with cognitive impairment.

Introduction

The neurotrophin Brain-Derived Neurotrophic Factor (BDNF) plays a key role in neuronal survival and plasticity (Huang and Reichardt, 2001). Like other secreted proteins, the functionally active BDNF protein, or mature BDNF (mat-BDNF; MW = 14 kDa), is obtained from the proteolytic cleavage of a precursor (pro-BDNF; MW = 32 kDa) (Mowla et al., 2001). Through a different cleavage, pro-BDNF can produce also another proteolytic BDNF isoform of 28 kDa (truncated-BDNF), which is not further cleaved (Seidah et al., 1999). However, pro-BDNF precursor is not just an inactive intermediate of the biosynthesis of BDNF as it shows distinct and opposite functions with respect to mat-BDNF. Specifically, pro-BDNF induces apoptosis and long-term depression in cultured neurons while mat-BDNF promotes survival and long-term potentiation (Teng et al., 2005, Woo et al., 2005). The role of the truncated-BDNF isoform is still unknown.

A number of studies investigated the role of BDNF in the pathophysiology of schizophrenia (Buckley et al., 2007, Shoval and Weizman, 2005, Thome et al., 1998). Schizophrenia (SZ) is a chronic and debilitating, clinically heterogeneous mental illness with a complex pathogenesis resulting from multiple genetic and environmental interactions (Ashe et al., 2001, van Os and Kapur, 2009). Recent studies have shown that mRNA levels for BDNF and its receptor TrkB are decreased in the prefrontal cortex of patients with schizophrenia (Weickert et al., 2003, Weickert et al., 2005). In addition, mat-BDNF levels were found to be decreased in the hippocampus and increased in the neocortex of affected patients (Durany et al., 2001).

Analysis of BDNF concentration in the serum as a potential biomarker in schizophrenia has provided controversial data. Indeed, some studies reported that BDNF levels in the serum of SZ patients were not different from those of healthy volunteers (Huang and Lee, 2006, Jockers-Scherubl et al., 2004, Shimizu et al., 2003), while other studies found a significant decrease (Grillo et al., 2007, Pirildar et al., 2004, Tan et al., 2005, Toyooka et al., 2002) or even an increase (Gama et al., 2007). A recent study suggested that quantification of serum BDNF may be helpful to measure the success of a computer-assisted training to improve cognitive functions in SZ patients (Vinogradov et al., 2009). However, before training, there was no difference in total serum BDNF levels between SZ patients and healthy donors. Thus, at present, there are no empirical methods to identify any specific features in SZ patients and in particular, to determine their cognitive status. In this study we determined the serum abundance of the different BDNF isoforms in order to establish a relationship with different pathological parameters including cognitive performance in chronic schizophrenic patients.

Section snippets

Subjects

The study was approved by the local Ethics Committee. Written informed consent was obtained from all participants or, their first-degree relatives. Subjects were Caucasians chronic outpatients with a Diagnostic and Statistical Manual of mental disorders Version-IV (DSM-IV) diagnosis of schizophrenia and fulfilled the criteria for the Structured Clinical Interview for DSM disorders version-I (SCID-I). Patients with a history of substance or alcohol abuse, personality disorder, medical or

Results

We studied a group (n = 40) of chronic SZ patients in comparison with a demographically matched healthy controls group (HC, n = 40) with comparable male/female ratio (20/20), age, years of education, BMI, and smoking habits (Table 1). SZ subjects enrolled in the study had evidence of continued illness (mean illness duration 23.05 years ± 10.99), a score on Positive and Negative Syndrome Scale (PANNS; Kay et al., 1989), and significantly reduced IQ with respect to HC (Table 1). All SZ patients

Discussion

In this study we provide evidence that reduced levels of serum truncated-BDNF/total BDNF ratio correlate with worst PANSS negative and positive symptoms and poorer neurocognitive performance. Instead, measurement of total serum BDNF levels resulted poorly informative, even if we found a small decrease in the whole population of schizophrenic patients. We further show that when using a cut-off at mean − 2SD from the healthy group, measurement of relative amount of serum truncated-BDNF provides a

Role of financial support

This study was supported by the SISTER project of the Government of the Region Friuli Venezia Giulia (to ET) a public project to support translational research.

Contributions

Davide Carlino enrolled the patients, made the diagnosis and part of the cognitive assessment, collected the blood and wrote the paper; Emiliano Leone and Francesco Di Cola performed the Western-blot analysis; Gabriele Baj performed the ELISA quantification and analysed the data; Raffaella Marin made the cognitive assessement, Giacomo Dinelli enrolled the patients and made the diagnosis; Enrico Tongiorgi projected the experiments, analysed the data and wrote the paper; Maurizio De Vanna

Conflict of interests disclosure

M. De Vanna acts as a consultant for Ely-Lilly, Italy for projects not related to this study. The other authors have no financial interests or potential conflicts of interest.

Acknowledgements

The authors thank all the study participants and staff from the UCO of Clinical Psychiatry of the University of Trieste who helped enrol the study sample. We also thank Monica Baiano (Centro Disturbi Alimentari, Portogruaro, Italy) and Massimo Borelli (University of Trieste, Italy) for critically reviewing the manuscript and help with statistical analysis and Yuri Bozzi (University of Trento, Italy) for helpful discussions.

References (37)

Cited by (0)

1

Equal contributing first authors.

2

Equal senior authors.

View full text