Neuregulin 1 transcripts are differentially expressed in schizophrenia and regulated by 5' SNPs associated with the disease

Abstract

Genetic variation in neuregulin 1 (NRG1) is associated with schizophrenia. The disease-associated SNPs are noncoding, and their functional implications remain unknown. We hypothesized that differential expression of the NRG1 gene explains its association to the disease. We examined four of the disease-associated SNPs that make up the original risk haplotype in the 5' upstream region of the gene for their effects on mRNA abundance of NRG1 types I-IV in human postmortem hippocampus. Diagnostic comparisons revealed a 34% increase in type I mRNA in schizophrenia and an interaction of diagnosis and genotype (SNP8NRG221132) on this transcript. Of potentially greater interest, a single SNP within the risk haplotype (SNP8NRG243177) and a 22-kb block of this core haplotype are associated with mRNA expression for the novel type IV isoform in patients and controls. Bioinformatic promoter analyses indicate that both SNPs lead to a gain/loss of putative binding sites for three transcription factors, serum response factor, myelin transcription factor-1, and High Mobility Group Box Protein-1. These data implicate variation in isoform expression as a molecular mechanism for the genetic association of NRG1 with schizophrenia.

Conflict of interest statement

Conflict of interest statement: No conflicts declared.

Figures

Fig. 1.

NRG1 types I–IV mRNA expression in the hippocampus of schizophrenic patients and control individuals. Quantitative RT-PCR analysis of NRG1 types I–IV normalized to PBGD. (A–C) n = 53 normal control subjects (NC) and 38 patients with schizophrenia (SZ). (D) n = 46 normal controls subjects and 28 patients with schizophrenia. Box represents the proportion of the distribution falling between the 25th and 75th percentiles. Bars outside the box represent the SD. Bar inside represents the mean. Significant differences were found between controls and patients for type I NRG1 expression. ∗∗, significant differences (P < 0.01)

Fig. 2.

Association between SNP8NRG122132 and type I NRG1 mRNA in normal controls (NC) and schizophrenics (SZ). A significant interaction of genotype and diagnosis was observed on normalized type I mRNA expression (P < 0.003). Post hoc t tests revealed a main effect of genotype on type I NRG1 expression in controls, with A allele carriers having increased levels compared with homozygous G individuals (P < 0.003). Schizophrenic patients homozygous for the G allele had higher levels of type I mRNA compared with controls with the same genotype (P < 0.001). Two individuals were excluded from analysis because of genotyping failure. Box represents the proportion of the distribution falling between the 25th and 75th percentiles. Bars outside the box represent the SD. The bar inside represents the mean.

Fig. 3.

Association between SNP8NRG243177 and type IV NRG1 mRNA expression. (A) In the whole cohort, a main effect of genotype was observed (ANOVA; P = 0.04) An allele dose-dependent effect is suggested, with individuals homozygous for the risk allele having the highest levels of type IV NRG1 mRNA (P = 0.05). (B) Data parsed by diagnosis. No genotype × diagnosis interaction was observed.

Fig. 4.

Association between diplotypes containing the deCODE risk haplotype (hap4) and type IV NRG1 mRNA. Individuals were divided according to diplotype into two groups, non-hap4 carriers (haplotypes 1/1; 1/2, 1/3, 2/2, 3/3, and 2/3) and hap4 carriers (haplotypes 1/4, 2/4, 3/4, and 4/4). A main effect of diplotype was observed on normalized type IV NRG1 mRNA levels (ANOVA; P = 0.04). (A) Individuals carrying the hap4 risk haplotype had increased levels compared with individuals not carrying hap4. (B) Effect of diplotype on type IV NRG1 mRNA levels in controls and patients. Eleven individuals were not included in the diplotype analysis because of either the failure of genotyping at one or more of the SNPs or low probability (<93%) of diplotype assignment according to snphap.