Vitamin B6 deficiency hyperactivates the noradrenergic system, leading to social deficits and cognitive impairment

Abstract

We have reported that a subpopulation of patients with schizophrenia have lower levels of vitamin B₆ (VB6) in peripheral blood than do healthy controls. In a previous study, we found that VB6 level was inversely proportional to the patient's positive and negative symptom scale (PANSS) score for measuring symptom severity, suggesting that the loss of VB6 might contribute to the development of schizophrenia symptoms. In the present study, to clarify the relationship between VB6 deficiency and schizophrenia, we generated VB6-deficient (VB6(-)) mice through feeding with a VB6-lacking diet as a mouse model for the subpopulation of schizophrenia patients with VB6 deficiency. After feeding for 4 weeks, plasma VB6 level in VB6(-) mice decreased to 3% of that in control mice. The VB6(-) mice showed social deficits and cognitive impairment. Furthermore, the VB6(-) mice showed a marked increase in 3-methoxy-4-hydroxyphenylglycol (MHPG) in the brain, suggesting enhanced noradrenaline (NA) metabolism in VB6(-) mice. We confirmed the increased NA release in the prefrontal cortex (PFC) and the striatum (STR) of VB6(-) mice through in vivo microdialysis. Moreover, inhibiting the excessive NA release by treatment with VB6 supplementation into the brain and α2A adrenoreceptor agonist guanfacine (GFC) suppressed the increased NA metabolism and ameliorated the behavioral deficits. These findings suggest that the behavioral deficits shown in VB6(-) mice are caused by enhancement of the noradrenergic (NAergic) system.

Fig. 1. Generation of VB6-deficient mouse model and their behavioral abnormalities.

VB6(−) mice were generated by feeding with a VB6-lacking diet from 8 to 12 weeks of age. A Changes in body weight during feeding with a VB6-lacking diet are shown. Two-way ANOVA with repeated measurements: F_{Interaction(4,132)} = 7.15, p < 0.001; F_Day(4,132) = 46.3, p < 0.001; F_VB6(1,33) = 9.82, p < 0.01 (Control: n = 18; VB6(−): n = 17). *p < 0.05, **p < 0.01, and ***p < 0.001 using Bonferroni’s multiple comparison test. B Food intake of a control and VB6-lacking diet were measured during the first 1 week of the feeding. n.s. (not significant) using Student’s t-test. C VB6 level in mouse plasma was determined after feeding for 4 weeks. ***p < 0.001 by Student’s t-test. D PLP and E PMP in the brains of VB6(−) mice were quantified by HPLC. Two-way ANOVA: D F_{Interaction(3,30)} = 3.58, p < 0.05; F_Area(3,30) = 38.0, p < 0.001; F_VB6(1,10) = 76.3, p < 0.001 and E F_{Interaction(3,30)} = 0.53, p > 0.05; F_Area(3,30) = 30.8, p < 0.001; F_VB6(1,10) = 1.32, p > 0.05 (n = 6). n.s., *p < 0.05, **p < 0.01, and ***p < 0.001 using Bonferroni’s multiple comparison test. In social interaction test, F time spent in the chamber and G interaction time were measured (Control: n = 18; VB6(−): n = 17). In the novel object recognition test, H exploratory time and I exploratory preference were determined (Control: n = 20; VB6(−): n = 20). Two-way ANOVA: F F_{Interaction(1,33)} = 4.07, p > 0.05; F_{Session(1,33)} = 19.9, p < 0.001; F_VB6(1,33) = 4.28, p < 0.05, G F_{Interaction(1,33)} = 4.94, p < 0.05; F_{Session(1,33)} = 39.2, p < 0.001; F_VB6(1,33) = 2.88, p > 0.05, H F_{Interaction(1,38)} = 8.10, p < 0.01; F_{Session(1,38)} = 18.2, p < 0.001; F_VB6(1,38) = 5.41, p < 0.05, and I F_{Interaction(1,38)} = 0.02, p > 0.05; F_{Session(1,38)} = 217, p < 0.001; F_VB6(1,38) = 0.55, p > 0.05. *p < 0.05 and ***p < 0.05 using Bonferroni’s multiple comparison test. ^###p < 0.001 and ^$p < 0.05 compared with control and VB6(−) mice in the habituation session respectively, by Bonferroni’s multiple comparison test. The data were represented as mean ± standard error of the mean (SEM). PFC prefrontal cortex, NAC nucleus accumbens, STR striatum, HIP hippocampus.

Fig. 2. Monoamine levels in the brain of VB6-deficient mice.

A DA, B 5-HT, and C NA contents, D DA, E 5-HT, and F NA turnover and G MHPG contents in various regions of the mouse brain were determined. *p < 0.05, **p < 0.01 and ***p < 0.001 using Bonferroni’s multiple comparison test (n = 6). The data were represented as the mean ± SEM values. n.d. not detected.

Fig. 3. NA release in the PFC and STR of VB6-deficient mice.

Basal levels of NA in A the PFC and C the STR are shown. (n.s. using Student’s t-test. n = 7). NA release in B the PFC and D the STR were measured. High K⁺ stimulation was performed at the time point of 0 min (Two-way ANOVA with repeated measures: B F_{Interaction(8,96)} = 1.95, p > 0.05; F_Time(8,96) = 46.8, p < 0.001; F _VB6(1,12) = 3.49, p > 0.05, D F_{Interaction(8,88)} = 3.98, p < 0.001; F_Time(8,88) = 19.1, p < 0.001; F _VB6(1,11) = 3.98, p > 0.05; **p < 0.01, ***p < 0.001 using Bonferroni’s multiple comparison test. n = 7). The data were represented as the mean ± SEM values.

Fig. 4. Rescue of behavioral deficits in VB6-deficient mice by PLP supplementation into the brain.

A Changes in body weight after the implantation of osmotic pump are shown (Two-way ANOVA with repeated measures: F_{Interaction(4,50)} = 2.38, p > 0.05; F_Day(2,50) = 3.31, p < 0.05; F_Group(2,25) = 11.6, p < 0.05. B NA turnover in the PFC were determined by HPLC. One-way ANOVA: F_(2,25) = 15.7, p < 0.001. C Exploratory preference in the novel object recognition test and D time spent in the chamber in the social interaction test were measured. Two-way ANOVA: C F_{Interaction(2,25)} = 6.10, p < 0.01; F_{Session(1,25)} = 38.0, p < 0.001; F_Group(2,25) = 1.67, p > 0.05, D F_{Interaction(2,25)} = 7.67, p < 0.01; F_{Session(1,25)} = 16.5, p < 0.001; F_Group(2,25) = 1.16, p > 0.05. *p < 0.05, **p < 0.01, ***p < 0.001 (vs. control/SAL) and ^#p < 0.05, ^##p < 0.01 (vs. VB6(−)/SAL) using Bonferroni’s multiple comparison test (n = 9–10). The data were shown as mean ± SEM values.

Fig. 5. Improvement of behavioral deficits in VB6-deficient mice by guanfacine treatment.

VB6(−) mice were administrated GFC during the last of 4 weeks of feeding with VB6-lacking diet. A NA turnover in the PFC were determined by HPLC. Two-way ANOVA: F_{Interaction(1,28)} = 7.62, p < 0.05; F_GFC(1,28) = 17.3, p < 0.001; F_VB6(1,28) = 58.2, p < 0.001. ***p < 0.001, ^###p < 0.001, and ^%p < 0.05 using Bonferroni’s multiple comparison test. In social interaction test, B traveled distance and C time spent in the chamber were measured (n = 12). In the novel object recognition test, D exploratory time and E exploratory preference were determined (n = 12). Two-way ANOVA: B F_{Interaction(3,44)} = 6.86, p < 0.001; F_{Session(1,44)} = 147.8, p < 0.001; F_Group(3,44) = 54.9, p < 0.001, C F_{Interaction(3,44)} = 10.51, p < 0.001; F_{Session(1,44)} = 71.7, p < 0.001; F_Group(3,44) = 1.30, p > 0.05, D F_{Interaction(3,44)} = 4.78, p < 0.01; F_{Session(1,44)} = 111.6, p < 0.001; F_Group(3,44) = 17.5, p < 0.001, and E F_{Interaction(3,44)} = 1.07, p > 0.05; F_{Session(1,44)} = 25.8, p < 0.001; F_Group(3,44) = 4.00, p < 0.05. *p < 0.05, ***p < 0.001, ^#p < 0.05, ^##p < 0.01, ^%%p < 0.01, ^%%%p < 0.001, and ^&&&p < 0.001 using Bonferroni’s multiple comparison test. ^$$p < 0.01 compared with control/SAL mice between the habituation and the training session by Bonferroni’s multiple comparison test. The data were shown as mean ± SEM values.