- Top of page
Aim: The dysbindin gene (dystrobrevin binding protein 1: DTNBP1) is a susceptibility gene for schizophrenia. Susceptibility genes for schizophrenia have been hypothesized to mediate liability for the disorder at least partly by influencing cognitive performance. This report investigated the relationship between cognitive function and the dysbindin gene.
Methods: The possible association between a single nucleotide polymorphism (SNP) of DTNBP1 (rs2619539: P1655), which is a risk-independent SNP for schizophrenia in Japanese populations, and memory and IQ was investigated in 70 schizophrenia patients and 165 healthy volunteers in a Japanese population.
Results: This SNP was associated with two memory scales, verbal memory and general memory, on the Wechsler Memory Scale–Revised (WMS-R), and three subcategories of the Wechsler Adult Intelligence Scale–Revised (WAIS-R), vocabulary, similarities and picture completion in healthy subjects. The SNP, however, did not influence either the indices of WMS-R, IQ or subcategories of WAIS-R in schizophrenia patients.
Conclusion: A risk-independent SNP in DTNBP1 may have an impact on cognitive functions such as memory and IQ in healthy subjects.
SCHIZOPHRENIA IS A complex genetic disorder characterized by profound disturbances of cognition, emotion and social functioning. It affects approximately 1% of the general population worldwide. A recent study implicated a gene on chromosome 6p, dystrobrevin binding protein 1 (DTNBP1; dysbindin, Online Mendelian Inheritance in Man [OMIM] 607145; National Center for Biotechnology Information [NCBI] gene ID 84062), as a susceptibility locus in Irish pedigrees.1 Since then a significant association between schizophrenia and genetic variations in DTNBP1 has been reported in various populations from Ireland, Wales, Germany/Hungary/Israel, Sweden, Bulgaria, USA, China, and Japan,2–11 and only a few studies did not support this association.12,13 Post-mortem brain studies have indicated reduced expression of the DTNBP1 mRNA in hippocampus and prefrontal cortices and DTNBP1 protein in the hippocampus of schizophrenia patients.14–16 Long-term treatment of mice with typical or atypical antipsychotics did not alter the mRNA expression levels or protein levels of dysbindin in the frontal cortex and hippocampus,16,17 suggesting that the prior evidence of decreased expression of DTNBP1 in the post-mortem brains of schizophrenia patients is not likely to be a simple artifact of ante-mortem drug treatment.
Schizophrenia is a neuropsychiatric disorder characterized by cognitive dysfunction. The heritability study of a collection of endophenotypes for schizophrenia suggests that endophenotypes including cognitive function are important measures to consider in characterizing the genetic basis of schizophrenia.18 Susceptibility genes for schizophrenia have been hypothesized to mediate liability for the disorder at least partly by influencing cognitive performance.19 There have been few studies, however, on the relationship between cognitive function and the dysbindin gene. It is reported that a DTNBP1 risk haplotype for schizophrenia was associated with general cognitive ability (g) in both schizophrenia patients and healthy controls.20 The same group subsequently reported the association between the risk haplotype and cognitive decline in schizophrenia.21 Another preliminary study assessed the association between another DTNBP1 risk haplotype for schizophrenia and verbal and spatial memory, working memory, attentional control, and premorbid IQ in schizophrenia patients.22 Patients carrying the dysbindin risk haplotype had significantly lower spatial working memory performance than patients who were non-risk carriers. Zinkstok et al. reported an association between genetic variations in DTNBP1 and intelligence, IQ.23 Recently, we reported the memory and learning impairment in sandy (sdy) mutant mice with a deletion in the dysbindin gene such as long-term memory retention and working memory,24 supporting roles of dysbindin in cognitive function. In the present study we examined a possible association between a genetic variant of DTNBP1, which was not associated with schizophrenia in Japanese populations, and memory function and general cognitive ability.
- Top of page
We examined the associations between an SNP (P1655) in DTNBP1 and two cognitive tests in 70 schizophrenia patients and 165 healthy controls. As expected, schizophrenia patients performed significantly worse than controls in all cognitive tests (all P < 0.00001). There were huge differences in cognitive performance between patients and controls (an average difference of the means is around two SD; e.g. verbal memory: patients, 79.1 ± 19.5; controls, 111.1 ± 13.4). Thus, we analyzed the effect of genotype in patients and controls, separately.
The characteristics of subjects are presented in Table 1. There was no significant difference between the genotype group in any of the variables, including illness features in schizophrenia patients. The genotype groups in healthy controls did not significantly differ in gender or education years, but there was a significant difference in age (F = 4.23, P = 0.016, post hoc G/G vs G/C, P = 0.012).
Table 1. Subject details
|(n = 39)||(n = 27)||(n = 4)||(n = 69)||(n = 80)||(n = 16)|
|Age (years)||45.7 ± 12.2||43.6 ± 15.0||49.0 ± 7.3||0.68||34.1 ± 10.9||39.8 ± 12.8||38.4 ± 11.7||0.016|
|Education years||13.5 ± 2.7||12.7 ± 3.6||14.3 ± 2.9||0.45||16.0 ± 2.6||16.3 ± 3.3||15.9 ± 3.0||0.79|
|Family history of psychiatric diseases (Yes/No)||15/24||7/20||2/2||0.53|| || || || |
|Age at onset (years)||24.5 ± 8.2||24.6 ± 10.0||29.5 ± 16.2||0.29|| || || || |
|Duration of illness (years)||21.6 ± 14.0||18.0 ± 13.8||16.8 ± 14.5||0.54|| || || || |
|CPZeq of total antipsychotic drugs (mg/day)||806 ± 658||793 ± 612||850 ± 918||0.99|| || || || |
We assessed the effects of the SNP on the WMS-R and the WAIS-R scores in schizophrenia patients and control subjects (Table 2). Significant effects of the SNP were found in verbal memory (F = 3.24, P = 0.042), general memory (F = 3.28, P = 0.040), vocabulary (F = 3.71, P = 0.027), similarities (F = 3.74, P = 0.026) and picture completion (F = 9.53, P = 0.00012) in healthy controls. In contrast, no effect of the genotype on the results of cognitive tests was observed in schizophrenia patients.
Table 2. Cognitive test results and a genetic variation in DTNBP1 in schizophrenia patients (mean ± SD)
|WMS-R||Verbal Memory||77.4 ± 19.1||82.4 ± 20.8||74.0 ± 15.3||0.27||113.0 ± 12.3||110.9 ± 14.0||103.4 ± 13.0||0.042|
|Visual memory||79.3 ± 21.9||81.7 ± 20.7||78.5 ± 26.5||0.68||109.8 ± 9.0||110.2 ± 8.8||107.7 ± 12.3||0.66|
|General memory||75.5 ± 18.9||81.2 ± 20.6||70.0 ± 21.7||0.16||113.5 ± 11.2||112.6 ± 11.8||104.9 ± 12.7||0.040|
|Attention/Concentration||89.0 ± 17.6||87.9 ± 17.3||89.5 ± 13.2||0.97||105.3 ± 13.8||104.7 ± 14.9||100.8 ± 10.4||0.45|
|Delayed recall||75.4 ± 20.4||80.5 ± 22.2||81.3 ± 14.5||0.40||113.2 ± 11.3||112.0 ± 12.3||108.8 ± 11.8||0.52|
|WAIS-R||Information||8.7 ± 3.9||7.4 ± 3.5||10.0 ± 3.6||0.49||10.1 ± 2.9||10.9 ± 2.8||10.1 ± 3.3||0.27|
|Digit Span||8.2 ± 2.7||8.2 ± 3.7||8.5 ± 2.1||0.75||11.1 ± 3.2||11.0 ± 2.6||10.5 ± 2.5||0.65|
|Vocabulary||8.7 ± 3.3||7.2 ± 3.6||8.5 ± 2.9||0.36||10.4 ± 3.1||11.8 ± 2.8||11.6 ± 2.8||0.027|
|Arithmetic||7.5 ± 3.1||7.2 ± 3.0||9.5 ± 3.8||0.54||10.7 ± 3.3||11.8 ± 2.9||10.9 ± 3.2||0.14|
|Comprehension||7.2 ± 3.9||6.5 ± 3.4||7.3 ± 3.1||0.95||10.7 ± 3.0||10.9 ± 2.8||10.9 ± 2.0||0.94|
|Similarities||9.4 ± 3.3||8.6 ± 3.6||11.5 ± 4.0||0.53||11.5 ± 2.5||12.5 ± 2.1||12.4 ± 1.6||0.026|
|Picture Completion||8.1 ± 3.3||7.5 ± 3.6||9.8 ± 3.6||0.57||9.0 ± 2.3||10.6 ± 2.3||10.8 ± 1.7||0.00012|
|Picture Arrangement||7.4 ± 3.2||6.9 ± 3.2||8.8 ± 5.5||0.82||11.3 ± 2.5||11.5 ± 2.2||10.9 ± 2.4||0.64|
|Block Design||8.7 ± 4.2||8.1 ± 3.8||10.0 ± 2.3||0.86||12.4 ± 3.1||12.7 ± 2.1||13.0 ± 2.9||0.63|
|Object Assembly||7.7 ± 3.6||7.6 ± 4.0||8.8 ± 2.2||0.92||11.7 ± 3.1||11.7 ± 2.7||11.8 ± 2.8||0.99|
|Digit Symbol||6.9 ± 2.8||6.2 ± 3.3||8.5 ± 1.7||0.62||12.7 ± 2.8||12.8 ± 2.9||13.8 ± 2.5||0.40|
|Verbal IQ||89.2 ± 17.4||84.4 ± 18.4||95.0 ± 17.8||0.81||104.9 ± 14.5||109.6 ± 12.3||107.1 ± 10.6||0.17|
|Performance IQ||84.3 ± 18.3||80.1 ± 18.3||93.8 ± 18.6||0.58||109.2 ± 13.3||111.9 ± 10.0||113.1 ± 10.4||0.29|
|Full scale IQ||86.0 ± 18.0||80.9 ± 19.2||94.3 ± 19.4||0.65||107.2 ± 12.9||111.6 ± 11.0||110.4 ± 8.3||0.12|
We focused on the effects of the SNP on verbal memory in the WMS-R scores, because general memory is a component of verbal memory, and on picture completion in the WAIS-R scores, because the effect of the SNP was much stronger than that in vocabulary or similarities in controls (Fig. 1). On post-hoc analysis subjects carrying the G/G genotype had significantly higher verbal memory scores than those with the C/C genotype (P = 0.025; Fig. 1a). Conversely, subjects with the G/G genotype performed significantly worse in picture completion tasks than subjects with the G/C or C/C genotype (P = 0.00015, P = 0.015; Fig. 1b). Similar effects of the SNP were observed in general memory of the WMS-R (higher performance in G/G genotype compared with C/C genotype, P = 0.028) and in vocabulary and similarities of the WAIS-R (lower performance in G/G genotype compared with G/C genotype, P = 0.014 and P = 0.017).
Figure 1. Association between cognitive functions and a single nucleotide polymorphism (SNP) in dystrobrevin binding protein 1 (DTNBP1). An SNP in DTNBP1 (P1655) was associated with (a) verbal memory and (b) picture completion in control subjects. G/G, G/C, and C/C represent genotypes of P1655. Data given as mean ± SE. *P < 0.05; ***P < 0.001 compared with the G/G genotype.
Download figure to PowerPoint
- Top of page
In the present study we evaluated the relationship between an SNP in DTNBP1 and several domains of memory performance measured on the WMS-R, IQ score, and its subscales measured on the WAIS-R in schizophrenia patients and healthy volunteers. Results indicated that this SNP was associated with two memory scales, verbal memory and general memory, and three subcategories of the WAIS-R, vocabulary, similarities and picture completion in control subjects. These results suggest that DTNBP1 may be a candidate gene for human memory performance and cognitive ability. We have first reported the association between memory performance and the dysbindin gene in healthy subjects. Our data and a previous preliminary study did not find an association between the dysbindin gene and such memory performance, verbal memory, in schizophrenia patients.22 Taken together, the effects of the dysbindin gene on memory performance could be affected by the disease and/or medication. Although one study reported an association between a risk haplotype of DTNBP1 and general cognitive ability (g)20 and another study found an association between several SNPs in DTNBP1 and IQ in a Caucasian sample,23 we could not find an association between the SNP and IQ in the present Japanese sample. Three subcategories of the WAIS-R, however, were associated with the SNP in control subjects. This inconsistency could be due to several reasons, such as the use of differential genetic variations in the three studies; allelic heterogeneity; false-negatives in the present study; false-positives in the previous studies; and ethnic difference. For example, the G allele of P1635, which is significantly in excess in Japanese schizophrenia patients (3.0%),3 was also overtransmitted in Irish samples (10.2%),1 but undertransmitted in German samples (17.6%),10 suggesting that this SNP might be a marker rather than a polymorphism responsible for susceptibility. Taken together, the genetic variation in DTNBP1 might be a marker that is differentially associated with IQ among different populations. Thus, further examination such as association analysis using the same genetic variation studied in the previous study and the present study, and an independent study with a new cohort, are needed to draw any conclusions.
We observed that healthy subjects with the G/G genotype performed better in verbal memory tests and worse in several WAIS-R scores than those with the C/C genotype. These data suggest that this genetic variation in DTNBP1 could contribute to the variation in human cognitive ability in both positive and negative ways. Although it apparently seems to be inconsistent, these results could explain the diversity of human cognitive domains. It is well known that each individual subject has strong points and weak points in specific cognitive functions. Subjects with the G/G genotype might have strengths in verbal memory and weaknesses in vocabulary, similarities and picture completion. And even though we used the Bonferroni correction for multiple testing, we could not exclude the possibility that these data were false-positive results.
The mechanisms underlying the effect of genetic variations in DTNBP1 on cognitive function are unknown. No genetic variant in DTNBP1 has produced direct evidence of functional effects. But DTNBP1 is widely distributed in several brain regions, including the frontal cortex, temporal cortex, hippocampus, caudate, putamen, nucleus accumbens, amygdala, thalamus and midbrain.15 A reduction in the expression of DTNBP1 in the hippocampus and dorsolateral prefrontal cortex, known to be important areas for cognitive function, has been reported.14–16DTNBP1 plays roles in neurotransmission,3,30,31 cellular signaling3,32 and neuronal survival.3 The reduced expression of DTNBP1 could be related to the reduced release of glutamate and the increased release of dopamine.3,31 Reductions of dopamine content in sdy mice, which lack dysbindin-1 owing to a deletion in the DTNBP1 gene, have been reported.33,34 Reduced dysbindin-1 protein increased surface expression of dopamine D2 receptor and blocked dopamine-induced internalization of dopamine receptor D2 (DRD2) in SH-SY5Y cells.35 Taken together, the reduced expression of DTNBP1 could be related to impairment of glutamatergic and dopaminergic systems, which are implicated in the neuropathology in schizophrenia.36 The association of the dysbindin gene with cognitive functions might be related to the effects of the dysbindin gene on glutamatergic and/or dopaminergic systems.
- Top of page
The authors thank Tomoko Shizuno for technical assistance. This work was supported in part by Grants-in-Aid from the Japanese Ministry of Health, Labor and Welfare (H17-kokoro-007 and H18-kokoro-005), the Japanese Ministry of Education, Culture, Sports, Science and Technology, CREST (Core research for Evolutional Science and Technology) of JST (Japan Science and Technology Agency), Grant-in-Aid for Scientific Research on Priority Areas – Research on Pathomechanisms of Brain Disorders – from the Ministry of Education, Culture, Sports, Science and Technology of Japan (18023045) and Japan Foundation for Neuroscience and Mental Health.