The etiology of schizophrenia (SCZ) is not yet known, but the genetic factor has an essential role. Its heritability is estimated up to 85 percent. This study aimed to determine whether the selected brain-derived neurotropic factor (BDNF) gene variant (rs6265) is associated with SCZ, its psychopathology and intelligence quotient (IQ).
MethodsRs6265 was genotyped in 159 participants, including 71 unrelated patients with SCZ (males = 40 [56%]) and 88 healthy controls (males = 35 [40%]). Psychopathology assessment was done by using positive and negative symptoms scale (PANSS), and IQ was measured by Wechsler adult intelligence scale (WAIS). COCAPHASE and CLUMP22 softwares were used to compare allele and genotype frequencies respectively. Two-way multivariate ANOVA was done using SPSS22 to evaluate the effects of the group (patients and healthy control) and the genotype on the test scores, including PANSS and WAIS scores, considering sex and smoking status as covariates.
ResultsThe study showed that the BDNF rs6265 C allele and CC genotypes, psychopathology, including PANSS scores, and also IQ were significantly different in all, male and female patients compared to healthy participants. For rs6265, we found significantly higher frequencies of the C allele (P = 0.04) in all participants with SCZ.
The findings also revealed that BDNF polymorphism (rs6265) increased significantly the risk of SCZ in different genetic models.
ConclusionThe results showed that the BDNF rs6265 may be associated with the risk of SCZ development in an Iranian group of patients with SCZ. These results may be helpful in better understanding the role of the BDNF gene in the pathogenesis of the disorder.
Schizophrenia (SCZ)[OMIM: 181500] is a serious mental disorder and has a heterogeneous nature. Although its etiology is not yet known, but the genetic factor has an essential role.1–5 Also, the neurodevelopmental hypothesis is one of the most important explanations for its pathogenesis in which it is claimed that any impairments occurring in neuronal migration, connections, plasticity, and their related brain structural abnormalities play a significant role in the development of SCZ.6 Brain-derived neurotropic factor (located on the short arm of chromosome 11 [11p14.1; GRCh38/hg38])7 causes neurons to be produced and survive.8 It is activated according to the type of brain region.9 The BDNF plays an important role in the development of the central nervous system (CNS). It has an impact on the serotonergic signaling, glial cells,10 hippocampus neurons, and the brain cortex.11 It plays a critical role in the process of neurogenesis, neuronal survival, synaptogenesis, synaptic plasticity, cognitive functions, and influence on the functioning of dopaminergic neurons.12,13 The BDNF interacts with the membrane receptor tropomyosin-related kinase B (TrkB), followed by a cascade of protein phosphorylation inside the cell. This leads to diverse effects, such as changes in gene expression.14 Evidence shows alterations in BDNF signaling in SCZ.13
The most commonly studied single nucleotide polymorphism (SNP) of the BDNF is rs6265 (C > T) the results of which indicate the presence of amino acids valine or methionine in the 66th position of the polypeptide.15,16 It was observed, however, that Val66Met affects the release of neurotrophin depending on the excitation of the cell.12 Studies showed that the Val allele is related to a higher risk of SCZ.17,18 The genotype Val/Val correlates with a lower volume of the hippocampus and impaired memory.19 However, the results of studies are controversial.20,21 The BDNF variants also affect the age of onset, symptoms, therapeutic responsiveness, neurocognitive function, and brain morphology of SCZ.15,22
It should be noted that the data coming from studies on different ethnicities may lead to different interpretations and results.8 However, the involvement of BDNF in the development of the CNS, the activity of dopaminergic neurons, synapses, and the neurogenesis process suggest its role in the etiology of SCZ. The purpose of this study was to develop the insight into the role of BDNF SNP rs6265 and SCZ, its psychopathology, and IQ in a sample of the Iranian population.
Materials and methodsParticipantsOne hundred and fifty-nine men and women participated in this study. The participants consisted of 71 unrelated participants with SCZ, with a mean age of 37.74 ± 10.14, including 40 male [56%], with a mean age of 38.00 ± 10.44, and 31 female with a mean age of 37.37 ± 9.95, and also 88 healthy controls with a mean age of 34.46 ± 7.46, including 35 male [40%], with a mean age of 38.79 ± 8.91, and 31 female with a mean age of 33.02 ± 7.63. Schizophrenia patients were recruited from the Roozbeh psychiatry hospital, Tehran University of Medical Sciences (TUMS), Tehran, Iran. In the case group, the inclusion criteria were the SCZ diagnostic criteria based on the fifth edition of the diagnostic and statistical manual of mental disorders (DSM-V), and no other mental disorder. Healthy controls were recruited from the same geographical area with the inclusion criteria of not suffering from any mental disorders based on DSM-V criteria and not having first-degree relatives with any mental disorder. The exclusion criteria for both groups of participants were IQ score below 70, history of any other medical disorders, substance addiction or use in the past year, and severe head trauma. Written informed consent was obtained from all participants. This study was carried out following the approved guidelines of the Ethical Committee of TUMS. Demographic features are shown in Supplementary Table 1.
Clinical assessmentsThe clinical diagnosis of SCZ was made independently by two psychiatrists according to the DSM-V criteria. The psychopathology, including positive and negative symptoms, general psychopathology, and total scores of the patients were assessed according to the PANSS.
Cognitive assessmentsIntelligence quotient assessment was done in both case and control groups using the WAIS.
DNA preparation, SNP selection, and genotypingBlood samples were taken by vacuum tubes pre-filed with the anticoagulant EDTA. Genomic DNA was prepared from venous blood using the salting out procedure.23 Rs6265 was genotyped by using the tetra-primer amplification refractory mutation system polymerase chain reaction (T-ARMS-PCR),24 and DNA sequencing. Genotyping was performed blind to status. Primers were designed using the Primer3. Primer sequences and PCR product sizes are shown in Supplementary Table 2.
The annealing temperature was 60 °C. All of the genotypes were callable, and minor allele frequency was greater than 2%. The T-ARMS-PCR fragments were fractionated on 1% agarose gel. Also, to eliminate the genotyping errors, all raw genotyping data were independently read by two experts and the suspect genotypes were retyped. Furthermore, a random group of samples was re-genotyped by direct sequencing to confirm the genotyping results.
Statistical analysisThe Hardy–Weinberg equilibrium for genotypic distribution was tested using the χ2 goodness-of-fit test. The odds ratio and 95% confidence interval (CI) were calculated to evaluate the effect of different genotypes. The allelic association was estimated using the COCAPHASED (UNPHASED) program. The genotype frequencies were compared using the CLUMP22 software by running 1000 Monte Carlo permutation.25 The genotypic association was tested by the chi-square test to assess the significance of the results. The independent-samples t-test was used to compare the means of the age and quantitative test scores. We used two-way multivariate ANOVA (MANOVA) to evaluate the effects of the group (patients and healthy control) and the genotype as independent factors on the test scores as dependent factors, considering sex and smoking status as covariates. The significance level for all statistical tests was P < 0.05. The power analysis was %80.
ResultsClinical assessmentsThe scores of PANSS were increased significantly in all male and female SCZ patients vs. healthy controls (Table 1).
PANSS analysis of patients with schizophrenia compared to healthy controls.
Clinical assessments (PANSS scores) | CaseN = 71 | ControlN = 88 | Statistics | |
---|---|---|---|---|
F | P value | |||
All | ||||
| 21.41 ± 7.44 | 7.13 ± 0.33 | 131.683 | <0.001* |
| 23.70 ± 11.39 | 7.21 ± 1.22 | 138.626 | <0.001* |
| 39.02 ± 11.59 | 15.91 ± 12.63 | 8.011 | <0.001* |
| 84.12 ± 26.07 | 31.76 ± 1.86 | 83.969 | <0.001* |
Male | ||||
| 21.03 ± 3.95 | 7.08 ± 0.28 | 47.933 | <0.001* |
| 21.21 ± 10.92 | 7.44 ± 1.80 | 43.595 | <0.001* |
| 37.14 ± 11.16 | 13.72 ± 18.78 | 0.019 | <0.001* |
| 79.38 ± 24.54 | 31.62 ± 1.88 | 33.053 | <0.001* |
Female | ||||
| 21.90 ± 8.18 | 7.16 ± 0.37 | 100.131 | <0.001* |
| 26.96 ± 11.41 | 7.03 ± 0.18 | 199.521 | <0.001* |
| 41.50 ± 11.93 | 17.68 ± 1.8 | 40.439 | <0.001* |
| 90.36 ± 27.26 | 31.87 ± 1.86 | 60.271 | <0.001* |
The scores of WAIS were decreased significantly in all male and female patients with SCZ vs. healthy controls (Table 2).
WAIS analysis patients with schizophrenia compared to healthy controls.
Intelligence assessments (WAIS scores) | CaseN = 71 | ControlN = 88 | Statistics | |
---|---|---|---|---|
F | P value | |||
All | ||||
| 86.63 ± 14.65 | 94.86 ± 14.83 | 0.322 | 0.019* |
| 84.90 ± 15.51 | 99.92 ± 29.16 | 1.282 | 0.012* |
| 85.48 ± 13.60 | 98.30 ± 15.12 | 0.452 | <0.001* |
Male | ||||
| 89.50 ± 14.70 | 98.42 ± 12.89 | 0.423 | 0.052** |
| 89.63 ± 14.56 | 101.21 ± 20.80 | 1.362 | 0.054** |
| 88.63 ± 12.61 | 98.70 ± 17.20 | 1.625 | 0.045* |
Female | ||||
| 80.90 ± 13.44 | 92.60 ± 15.73 | 0.020 | 0.04* |
| 75.90 ± 13.67 | 104.55 ± 15.94 | 0.533 | <0.001* |
| 80.50 ± 14.15 | 98.30 ± 15.12 | 0.452 | <0.001* |
VIQ = verbal intelligence quotient, PIQ = performance intelligence quotient, TIQ = total intelligence quotient.
The rs6265 was polymorphic and included in the analysis. We found a significant increase in C allele frequency in patients with SCZ rather than healthy controls in all subjects (P = 0.04). We also found a trend towards the significant increase of C allele frequency in patients rather than healthy controls in male subjects (P = 0.05) (Table 3).
Allele frequency analysis between patients with schizophrenia and healthy controls.
Additionally, the findings also revealed that BDNF polymorphism (rs6265) increased significantly the risk of SCZ in different genetic models in total subjects (co-dominant [CC vs. CT or TT; P < 0.0001, OR = 3.530, 95% CI = 1.79−6.96], dominant [CC vs. CT + TT; P = 0.0007, OR = 3.120, 95% CI = 1.60−6.10], and recessive [CC + CT vs. TT; P = 0.01, OR = 0.00, 95% CI = 0.00-NA]), in male subjects (co-dominant [CC vs. CT or TT; P = 0.002, OR = 4.630, 95% CI = 1.71−12.52], and dominant [CC vs. CT + TT; P = 0.003, OR = 4.080, 95% CI = 1.53−10.88]), and in female subjects (co-dominant [CC vs. CT or TT; P = 0.03, OR = 2.310, 95% CI = 0.87−6.12], and recessive [CC + CT vs. TT; P = 0.04, OR = 0.00, 95% CI = 0.00–NA]) (Table 4), (Fig. 1).
Genotype frequency analysis of BDNF polymorphism (rs6265) in patients with schizophrenia and healthy controls.
Groups | Genotypes | Genotype number (frequency) | Statistics | ||
---|---|---|---|---|---|
Case (N = 71) | Control (N = 88) | OR (95% CI) | P value | ||
Total | |||||
Co-dominant | CC | 37 (52) | 22 (25) | 3.530 (1.79−6.96) | <0.0001* |
CT | 30 (42) | 66 (75) | |||
TT | 4 (6) | 0 (0) | |||
Dominant | CC | 37 (52) | 22 (25) | 3.120 (1.60−6.10) | 0.0007* |
CT + TT | 34 (48) | 66 (75) | |||
Recessive | CC + CT | 67 (94) | 88 (25) | 0.00 (0.00–NA) | 0.01* |
TT | 4 (6) | 0 (0) | |||
Male | |||||
Co-dominant | CC | 25 (59) | 9 (26) | 4.630 (1.71−12.52) | 0.002* |
CT | 15 (36) | 25 (74) | |||
TT | 2 (5) | 0 (0) | |||
Dominant | CC | 25 (59) | 9 (26) | 4.080 (1.53−10.88) | 0.003* |
CT + TT | 17 (41) | 25 (74) | |||
Recessive | CC + CT | 40 (95) | 34 (100) | 0.00 (0.00–NA) | 0.12 |
TT | 2 (5) | 0 (0) | |||
Female | |||||
Co-dominant | CC | 12 (39) | 13 (25) | 2.310 (0.87−6.12) | 0.03* |
CT | 17 (55) | 40 (75) | |||
TT | 2 (1) | 0 (0) | |||
Dominant | CC | 12 (39) | 13 (25) | 2.050 (0.78−5.37) | 0.14 |
CT + TT | 19 (61) | 40 (75) | |||
Recessive | CC + CT | 29 (94) | 53 (100) | 0.00 (0.00–NA) | 25 (74) |
TT | 2 (6) | 0 (0) |
We used two-way MANOVA to evaluate the effects of the group (patients and healthy control) and the genotype as independent factors on the PANSS and WAIS scores, considering sex and smoking status as covariates. The results showed that only the group factor has significant effects on PANSS positive, negative, general psychopathology, and total, and also performance, and total IQ scores. The results also showed that the genotype factor has a trend towards the significant effect on PANSS negative symptoms score (Table 5).
The groups (SCZ vs. healthy Controls) and genotype factors effects on PANSS and WAIS scores.
Source | Dependent variable | df | F | P value | Partial eta squared |
---|---|---|---|---|---|
Sex | PANSS Positive | 1 | 0.659 | 0.420 | 0.010 |
PANSS Negative | 1 | 0.663 | 0.418 | 0.010 | |
PANSS General | 1 | 0.074 | 0.787 | 0.001 | |
PANSS Total | 1 | 0.045 | 0.832 | 0.001 | |
Verbal IQ | 1 | 2.781 | 0.100** | 0.042 | |
Performance IQ | 1 | 0.000 | 0.990 | 0.000 | |
Total IQ | 1 | 0.816 | 0.370 | 0.013 | |
Smoking | PANSS Positive | 1 | 0.067 | 0.797 | 0.001 |
PANSS Negative | 1 | 3.719 | 0.058** | 0.055 | |
PANSS General | 1 | 3.135 | 0.081 | 0.047 | |
PANSS Total | 1 | 2.228 | 0.140 | 0.034 | |
Verbal IQ | 1 | 0.002 | 0.961 | 0.000 | |
Performance IQ | 1 | 0.000 | 0.997 | 0.000 | |
Total IQ | 1 | 0.005 | 0.947 | 0.000 | |
Group | PANSS Positive | 1 | 144.048 | 0.000* | 0.692 |
PANSS Negative | 1 | 101.357 | 0.000* | 0.613 | |
PANSS General | 1 | 131.903 | 0.000* | 0.673 | |
PANSS Total | 1 | 161.004 | 0.000* | 0.716 | |
Verbal IQ | 1 | 3.043 | 0.086 | 0.045 | |
Performance IQ | 1 | 11.408 | 0.001* | 0.151 | |
Total IQ | 1 | 7.456 | 0.008* | 0.104 | |
Genotype | PANSS Positive | 2 | 0.290 | 0.749 | 0.009 |
PANSS Negative | 2 | 2.197 | 0.119** | 0.064 | |
PANSS General | 2 | 0.213 | 0.809 | 0.007 | |
PANSS Total | 2 | 0.340 | 0.713 | 0.011 | |
Verbal IQ | 2 | 0.561 | 0.573 | 0.017 | |
Performance IQ | 2 | 0.230 | 0.795 | 0.007 | |
Total IQ | 2 | 0.318 | 0.728 | 0.010 | |
Group X Genotype | PANSS Positive | 1 | 0.390 | 0.534 | 0.006 |
PANSS Negative | 1 | 2.470 | 0.121 | 0.037 | |
PANSS General | 1 | 0.182 | 0.671 | 0.003 | |
PANSS Total | 1 | 0.068 | 0.795 | 0.001 | |
Verbal IQ | 1 | 0.043 | 0.837 | 0.001 | |
Performance IQ | 1 | 0.313 | 0.578 | 0.005 | |
Total IQ | 1 | 0.091 | 0.764 | 0.001 |
We investigated the association between BDNF SNP rs6265 and SCZ, its psychopathology, and IQ. We found that rs6265 has a significant effect on the risk of SCZ development. This effect was detected in all male and female participants. The results showed that the C allele may be the risk allele with an increased risk of SCZ development throughout the sample. In the male sample, the C allele showed just a trend towards a significant effect. We also found that the CC genotype increases significantly in SCZ patients in all, male and female samples.
The neurodevelopmental hypothesis is well established in SCZ, and BDNF may be involved in the pathogenesis of this disorder.26 Rs6265 is identified as a promising SCZ risk variant.27 Although the results of some studies could not confirm its role in SCZ development,28 other studies confirm this role.16,17 The BDNF gene is essential for brain development and plasticity and is sensitive to environmental stressors and harbors the functional SNP rs6265 which creates or abolishes a CpG dinucleotide for DNA methylation. It has been concluded that rs6265 methylation interacts with genotype to bridge early environmental exposures to adult phenotypes, relevant for SCZ.16 Zakharyan and Boyajyan29 report that carriers of the rs6265 minor allele (T/C + T/T) which is significantly more frequent in SCZ patients than in controls, had decreased BDNF levels. Although Suchanek et al.30 could not found any association between rs6265 and the development of paranoid SCZ, they could find an association between rs6265 and age at onset of SCZ. Men with the Val/Met genotype had an earlier age at onset.30,31 Li et al. could not find any significant association between rs6265 and SCZ in the Chinese Han population, but they found a significant difference in haplotype frequencies, with more frequent haplotype ATC of rs6265-rs12273539-rs10835210 in the patients with SCZ than in controls. Besides, these authors reported that the interaction of BDNF and neurotrophic tyrosine kinase receptor 2 (NTRK2; the high-affinity receptor of BDNF), genes polymorphisms (BDNF-rs6265, NTRK2-rs1387923, and NTRK2-rs2769605) may be involved in the susceptibility to paranoid SCZ.32
We could not find significant associations between the rs6265 and PANSS scores. Suchanek et al.30 could found a significant association between Val66Met polymorphism and the psychopathology of paranoid SCZ. Men with the Val/Val genotype predisposed to more severe symptoms. The analysis of PANSS single items has shown that patients with the BDNF Val66Met had higher scores on positive symptoms,33 especially on a hallucinatory behavior item.30 It has been postulated that the correlation between rs6265 and positive symptoms of SCZ might be associated with the increased dopamine function caused by BDNF signaling.34,35 Zhang et al.36 found the association of the two haplotypes including the rs6265 Val allele with positive symptoms, but they did not find a direct association between Val66Met and clinical symptoms. Chang et al.37 found that Val/Val patients displayed the highest negative symptom scores among the Han Chinese population. Li et al.26 showed a significant association between the BDNF polymorphisms, including rs6265 (A)-rs12273539 (C)-rs10835210 (A) haplotype and just negative symptom scores. The discrepancies among the studies may be attributable to several reasons. The BDNF genotypes can be regarded as trait-dependent features, while the assessment of SCZ severity is state-dependent. So, it can be postulated that a comparison of genotypes with the severity of disorder could be different at any given time. Also, inconsistency can arise from differences in clinical subtypes of SCZ, and in stages of disorder progression (e.g., acute, chronic, active phase, or in remission), in disorder courses, and exposure to different types, dosages, and durations of antipsychotic medications.36
We could not find any significant associations between the rs6265 and WAIS scores. In their two separate meta-analyses to investigate the association between the Val66Met polymorphism and neurocognition in patient with SCZ, Ahmed et al.38 also found a non-significant differences between the genotype groups on most neurocognitive domains. However, in a few studies, different domains of IQ have been reported to be associated with rs6265.39 Moreover, it has been reported that the IQ in patients with the first-episode of SCZ is associated with the Val66Met polymorphism of the BDNF gene.39 In their study in a group of paranoid SCZ patients, Mak et al.40 reported that the patients with rs6265 G/G genotype started the rehabilitation program from a higher level of difficulty than patients with the A/G genotype. They explained it as a possible negative influence of Val66Met on cognitive functioning.
A possible explanation for all of the inconsistencies in genetic studies of SCZ may be in the diversity of ethnic background. For example, whereas in our study the Met allele frequency of rs6265 was 37%, was around 50%41 in the Chinese population, and, around 20% in Caucasian subjects.42 Additionally, it should be noted that the SCZ is a heterogeneous disorder, and leading to inconsistent results. So, haplotypes may be more specific risk markers than single SNPs, and their use reduces false-positive associations.43 Also, the interaction among multi-genes is probably playing the main role in the pathogenesis of SCZ. So, due to study differences and contrasting findings, further research will be required to understand in better detail the association between BDNF SNPs and SCZ.
There were limitations to the current study. The first was the small sample size which may impact the statistical power of our study. The second was that all participants with SCZ were recruited from hospitals. So there was an inherent selecting bias that was an unavoidable problem. The other limitation was that we matched the patients and healthy groups just based on age and handedness, but could not match them on sex and ethnicity. The last limitation was that although it may be convenient to obtain a large sample, but studying a sample of SCZ patients at any time of their treatment is not as strong as studying first-episode patients.
ConclusionAccording to our knowledge, this is for the first time that the association of BDNF SNP rs6265 is reported with the risk of SCZ, its psychopathology, and IQ in an Iranian sample. Our findings were consistent with the theory stating that the BDNF gene variants may mediate the risk of SCZ. However, further studies in different ethnic groups are needed.
Ethical considerationsThis study was carried out fallowing the approved guidelines of the Ethical Committees of Tehran University of Medical Sciences, Tehran, Iran. All procedures performed in studies involving human participants were under the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Written informed consent was obtained from all participants.
FundingThe research used no funding or financial support at any stage.
Conflict of interestThe authors declare there are no conflicts of interest.
We thank the staff of the Department of Genomic Psychiatry and Behavioral Genomics (DGPBG) at the Roozbeh Psychiatry Hospital in Tehran University of Medical Sciences (TUMS). The authors would like to thank Mrs. Abdi, and Mr. Eliasvand for their cooperation too.