Neuregulin-1 immunoreactivity in peripheral plasma is associated with rs6982890 polymorphism-mediated psychotic symptoms in schizophrenia

,


Introduction
Schizophrenia is a severe mental illness that includes positive symptoms (hallucinations, delusions), negative symptoms (poor speech, social dysfunction) and cognitive dysfunction (Jauhar et al., 2022).
NRG1 is a risk gene for schizophrenia and has been confirmed to be associated with schizophrenia in many countries and regions (Shiota et al., 2008).Studies have found that NRG1 gene polymorphisms are significantly associated with clinical symptoms in patients with schizophrenia (Jagannath et al., 2017) (Kastin et al., 2004).Here, we investigated whether peripheral plasma NRG1 immunoreactivity is regulated by polymorphisms, which affect schizophrenia symptoms in a northern Chinese Han population.
NRG1 is a member of the epidermal growth factor family.It functions through the ERBB receptor tyrosine kinase and participates in the migration of cortical excitatory neurons, synaptic plasticity, myelination and neurotransmitter transmission (Fischbach, 2007).Studies have shown that NRG1 SNP rs7014762 is related to schizophrenia.It not only regulates the low expression of NRG1 in suballelic carriers but also predicts the expression of NRG1 in the human brain after death (Nicodemus et al., 2009).Clinical studies have shown that the protein concentration and mRNA levels of NRG1 in the peripheral plasma of schizophrenia patients are significantly lower than those of healthy controls, and they are significantly increased after treatment with anti-schizophrenia drugs (Mostaid et al., 2017;Yang et al., 2022;Zhang et al., 2019).And level of NRG1 in brain tissue is down-regulated in animal models of schizophrenia, which is then up-regulated after antipsychotic treatment and accompanied by improvement of cognitive function (Li et al., 2016).Many studies have demonstrated a correlation between NRG1 SNPs and cognitive function in schizophrenia patients (Nawaz et al., 2014;Shibuya et al., 2010;Yokley et al., 2012;Yoshimi et al., 2016).However, whether NRG1 SNPs affect the response of patients to antipsychotic drugs and how they affect clinical symptoms need to be further studied.
Here, we validated the allele and genotype frequency of NRG1 in the Han population of northern China.Peripheral plasma NRG1 immunoreactivity were measured in 40 healthy subjects and 91 schizophrenia patients at baseline and after 8 weeks of antipsychotic treatment, clinical symptoms and cognitive function were also recorded.We then genotyped the above 84 schizophrenia patients and evaluated the effects of SNPs on NRG1 immunoreactivity, symptoms, and cognitive function in schizophrenia patients.

Subjects
We recruited subjects from 2010 to 2022 in a northern Chinese Han population.A total of 1304 patients with schizophrenia and 871 healthy controls were genotyped, and NRG1 SNPs analysis was subsequently conducted.After obtaining consent from the patients, we selected 91 of them and matched them with 40 healthy controls to detect peripheral plasma NRG1 immunoreactivity.Peripheral plasma NRG1 immunoreactivity were measured by ELISA in 91 patients with schizophrenia, both at baseline and after 8 weeks of treatment, as well as in 40 healthy controls at baseline.
Each patient was diagnosed with schizophrenia by a consensus of more than two experienced psychiatrists according to the Diagnostic and Statistical Manual of Mental Disorders Fourth Edition (DSM-IV) diagnostic criteria.All patients were assessed with the PANSS for psychiatric symptoms and the MCCB for cognitive function at baseline and after the 8-week treatment with antipsychotics (SR et al., 1987).We evaluated a total of eight indicators of the MCCB, including the Trail Making Test (TMT), Symbol Coding Subtest (SC), Category Fluency Test (CF), Hopkins Verbal Learning Test-Revised (HVLT-R), Spatial Span (SS), Mazes subtest (MAZES), Brief Visuospatial Memory Test-Revised (BVMT-R) and Continuous Performance Test-Identical Pairs (CPT-IP) (Shi et al., 2015).Individuals with hypertension, diabetes, infectious diseases or physical diseases were excluded.40 healthy subjects without psychiatric, somatic or organic brain diseases were recruited as normal controls.All subjects in this study cooperated to complete the evaluation content of the project and signed an informed consent.

Genotyping
The genomic DNA of 84 schizophrenia patients and 40 healthy controls was isolated using a plasma genomic DNA extraction kit produced by Beijing Tiangen Biochemical Technology Co, and genotyping was performed using the Illumina BeadStation 500 G platform and GoldenGate technology according to the operation manual.We selected six previously identified SNPs in the human NRG1 genome for analysis: rs12547858, rs6982890, rs7005288, rs35753505, rs6994992 and rs3924999 (Cao et al., 2014;Kang et al., 2012;Kéri et al., 2009;Walker et al., 2010;Wen et al., 2016;Yang et al., 2021).The genes of these six SNPs were also determined using TaqMan assay.To evaluate the quality of genotyping, 100 randomly selected samples were genotyped repeatedly by Sanger sequencing.

Enzyme-linked immunosorbent assay
Fifteen milliliters of peripheral plasma from fasting patients and healthy controls were taken in the morning and transported using an ice box.The upper plasma was taken after centrifugation at 12000 rpm and then stored in a refrigerator at − 80 • C for subsequent experiments.Plasma NRG1 immunoreactivity were measured using an ELISA kit (Elabscience, Wuhan, China; No: E-EL-H6092), according to the manufacturer's instructions.The sensitivity of the kit for the detection of NRG1 immunoreactivity in human plasma was 9.38 pg/ml, and the protein concentration in plasma was pg/ml.

Statistical analysis
All data were analyzed using SPSS 26.0 and presented as quartiles.We plotted NRG1 immunoreactivity data using GraphPad Prism 8.0.The Kolmogorov-Smirnov test for normal distribution was used for continuous variables.Differences between groups were determined by t-test (normally distributed data) or the Mann-Whitney U test (non-normally distributed data).The Wilcoxon rank sum test was used for pairwise comparison.The Chi-square test was used to compare all categorical data.Partial correlation analysis was used to test the relationship between related variables after controlling for potential confounding variables.Multiple linear regression analysis was used to analyze NRG1 immunoreactivity and its influencing factors in peripheral plasma.Age, sex, years of education, duration of illness and antipsychotic use at baseline were included as covariates.PLINK software was used to analyze SNP genotyping, minimal allele frequency (MAF) and deviation from Hardy-Weinberg equilibrium (HWE) results.Two-sided p values < 0.05 were considered statistically significant.Multiple testing was performed using Bonferroni correction to exclude Type I errors.

Association of NRG1 with schizophrenia
Genotype data were obtained for all 6 SNP loci, and allelic and genotypic frequencies are shown in Table 1.Analysis results showed that only the rs6982890 genotype and allele frequencies were statistically significant between schizophrenia patients and healthy controls (P = 0.047; P = 0.037 after Bonferroni correction).

The demographic and clinical characteristics of schizophrenia patients with SNP rs6982890 genotypes at baseline and after treatment
There was no significant difference in sex and years of education between patients with the CC genotype and CT genotype.However, patients with the CT genotype were older and had a longer disease duration.At baseline, there was no significant difference between the two genotype in any item of the MCCB except the CF on.After 8 weeks of drug treatment, PANSS scores were significantly improved in both groups, and patients with the CT genotype had better positive symptoms than those with the CC genotype after treatment.Meanwhile, patients with the CC genotype had significant improvement in all items of the MCCB except the SC, while patients with CT genotype had no significant improvement in the SC, SS or CPT-IP (Table 2).

Peripheral plasma NRG1 immunoreactivity in schizophrenia and healthy controls
The changes in peripheral plasma NRG1 immunoreactivity at baseline and after treatment in schizophrenia patients and healthy subjects are shown in Fig. 1.Peripheral plasma NRG1 immunoreactivity of schizophrenia patients were significantly lower than healthy controls (P < 0.0001, Fig. 1A).Peripheral plasma NRG1 immunoreactivity of schizophrenia patients significantly increased after 8 weeks of antipsychotic treatment (P = 0.003, Fig. 1A).In Fig. 1B, patients were grouped according to rs6982890 genotypes.Peripheral plasma NRG1 immunoreactivity of patients with the CC genotype increased significantly after treatment (P = 0.006), while no significant change occurred in patients with the CT genotype.Peripheral plasma NRG1 immunoreactivity of patients with the CT genotype were significantly higher than those of patients with CC genotype at baseline (P = 0.011).

Correlation analysis of NRG1 immunoreactivity with PANSS and MCCB in patients with CC and CT genotypes
In patients with the CC genotype, baseline NRG1 immunoreactivity were significantly positively correlated with PANSS positive symptoms (r = 0.324, P = 0.025) and negatively correlated with changes in the TMT-A and BVMT-R in the MCCB (r = − 0.348, P = 0.035: r = − 0.378, P = 0.021).In patients with the CT genotype, baseline NRG1 immunoreactivity were significantly positively correlated with PANSS negative symptoms and the change of the HVLT-R in the MCCB (r = 0.431, P = 0.028: r = 0.600, P = 0.003).The other correlation analysis date can be found in the supplementary materials (supplementary 1 and 2).

Discussion
In this study, the genotype and allele frequencies of SNP rs6982890 were found to be significant in schizophrenia patients.After 8 weeks of treatment, the peripheral plasma NRG1 immunoreactivity of schizophrenia patients increased significantly but was lower than those of healthy controls.The NRG1 immunoreactivity was also found to be correlated with symptoms and cognitive function.
There are few studies on this locus in the Chinese Han population (Wen et al., 2016).SNP rs6982890 is known to be located in the intronic region of the chromosome.We found that it is associated with NRG1 immunoreactivity in the National Center for Biotechnology Information annotation of this loci.SNP rs6982890 is known to induce intron variation and gene transcription variation, thus affecting NRG1 immunoreactivity.We also searched the expression quantitative trait loci of this loci and found that it was significantly correlated with NRG1 immunoreactivity in plasma (http://www.mulinlab.org/qtlbase/index.html).Based on the above data, we speculate that the mutation of this loci may increase NRG1 immunoreactivity, and it may no longer affected by other gene or drug regulation.Several studies have reported that NRG1 immunoreactivity can be regulated by SNPs (Nicodemus et al., 2009;Shibuya et al., 2010) and patients with the CT genotype had less cognitive improvement after treatment than those with the CC genotype.Interestingly, patients with the CC genotype had longer disease duration and were less likely to show improvement in symptoms.Indeed, studies have shown that NRG1 immunoreactivity is negatively correlated with disease duration (Yang et al., 2022), and therefore may be a potential factor affecting NRG1 expression.After stratification of patients, the baseline peripheral plasma NRG1 immunoreactivity was significantly positively correlated with PANSS positive symptoms in patients with the CC genotype, a result that was not in agreement with our hypothesis.However, NRG1 immunoreactivity increased significantly after treatment and the correlation disappeared, which may indicate that lower NRG1 immunoreactivity predict the symptom severity.In addition, NRG1 immunoreactivity was significantly positively correlated with the changes of negative symptoms in CT genotype, indicating the influence of genetic variation on patients' symptoms.There was a significant correlation between the changes in cognitive function in patients as assessed by the MCCB and the two genotypes, indicating that NRG1 immunoreactivity has a certain impact on the improvement of cognitive function in patients.There is substantial evidence that genetic variants in NRG1 are associated with an increased genetic risk for schizophrenia (Yoosefee et al., 2016;Yoshimi et al., 2016).It is not clear though whether the risk haplotype can lead to altered NRG1 immunoreactivity in the brain, which then affect the psychotic symptoms of schizophrenia patients.
NRG1 immunoreactivity is significantly lower in schizophrenia patients than in healthy controls, which is consistent with previous studies (Li et al., 2018;Wang et al., 2015;Zhang et al., 2020a).And we found that the peripheral plasma NRG1 immunoreactivity of male patients were lower than that of female patients before treatment, although the difference was not significant (P=0.0516).However, after 8 weeks of treatment, male peripheral plasma NRG1 immunoreactivity increased significantly, but not in females.At the same time, there was no difference in the peripheral plasma NRG1 immunoreactivity between male and female both in patients and healthy controls.Studies have shown that the NRG1 immunoreactivity in peripheral plasma is sex-specific, and is lower in male patients than in female patients, and the clinical symptoms are more severe, which may be due to the role of estrogen (Olaya et al., 2018;Shibuya et al., 2010;Zhang et al., 2020b).And we found no correlation between NRG1 immunoreactivity and drug concentrations (translated to chlorpromazine equivalents) (supplementary 3).
We evaluated the cognitive function of patients and found a correlation between impaired cognitive function and NRG1 immunoreactivity, which was not addressed in previous studies.The levels of NRG1 mRNA in peripheral plasma of schizophrenia patients are higher than those of healthy controls (Mostaid, Md Shaki et al., 2017), the expression of NRG1 in the hippocampus of schizophrenia patients is up-regulated (Chen et al., 2021), and in treatment-resistant schizophrenia patients is generally increased (Mostaid et al., 2017).These inconsistencies may be related to the demographic differences of the subjects (age, gender, years of education and region), disease duration and medication dosage.Nevertheless, our study suggests that increased NRG1 immunoreactivity strongly predict improvement in clinical symptoms and cognitive function in schizophrenia patients.
Although the use of peripheral plasma NRG1 immunoreactivity to predict symptoms and cognitive function in schizophrenia is attractive for practical reasons, peripheral proteins are not fully predictive of central nervous system (CNS) disease status.However, both a clinical study using antipsychotic drugs and a basic study showed that treatment with antipsychotic drugs increased NRG1 mRNA and protein levels (Li et al., 2016;Nawwar et al., 2022;Yang et al., 2022).Treatment with antipsychotic drugs increased NRG1 expression in the mouse brain (Chana et al., 2009).This also suggests that NRG1 immunoreactivity in peripheral plasma can partly reflect levels in the CNS and disease status.NRG1 plays a key role in the development of vertebral neurons and migration and myelination of interneurons in the CNS (Dang et al., 2016).Studies suggest that NRG1 affects synaptic plasticity and neural cycling mainly through glutamatergic or GABAergic systems in schizophrenia, and NRG1 promotes GABA release in the cortex and hippocampus through ErbB4 (Agarwal et al., 2014;Deng et al., 2019).In vitro, NRG1 was able to promote cell proliferation (Jarde et al., 2020).Another study indicated that NRG1 could be used to dedifferentiate cells back to a stem-cell-like state by activating the WNT pathway, which aided in regeneration (Jarde et al., 2016).Taken together, the results indicate that high NRG1 immunoreactivity in the peripheral plasma may play a role in the protection and repair of the damaged CNS, which is supported by the present study.
Our study has some limitations.First, the sample size was small, and the results should be further verified with a large sample size in future studies.Second, we did not restrict the use of antipsychotic drugs in our schizophrenia cohort.It will be necessary to analyze the effect of different psychotropic drugs on NRG1 immunoreactivity.Lastly, we  showed the effect of only one SNP, and future studies should examine the synergistic effect of multiple SNPs on NRG1 immunoreactivity.

Conclusions
In conclusion, we showed that after 8 weeks of treatment, NRG1 immunoreactivity in the peripheral plasma of schizophrenia patients were increased, and clinical symptoms and cognitive function were improved.NRG1 is associated with neurodevelopment, and high NRG1 immunoreactivity may plays a positive role in the functional recovery in schizophrenia patients.Genetic factors are associated with developmental abnormalities in schizophrenia, and we have found that genetic factors may influence both the treatment and functional recovery of schizophrenia patients.

Table 1
Genotype and allele frequencies of six SNPs in the NRG1 gene in schizophrenia patients and healthy controls.Major/minor alleles, where D and d denote major and minor alleles, respectively.2. P values in parentheses are Bonferroni corrected P values and P < 0.05 is indicated in bold.3. HWE: Hardy-Weinberg equilibrium, Hardy-Weinberg equilibrium; MAF: minor allele frequency, minor allele frequency.

Table 2
The demographic and clinical characteristics of schizophrenia with different SNP rs6982890 genotypes at baseline and after treatment.

Table 2
(continued ) : NA, not applicable; y, year; M, male; F, female; CPZ, chlorpromazine equivalents; PANSS, Positive and Negative Syndrome Scale.*P < 0.05 indicates significant differences between baseline and the 8-week time point. Notes