Association of 5′ end neuregulin-1 (NRG1) gene variation with subcortical medial frontal microstructure in humans

Abstract

Animal data suggest that the gene neuregulin-1 (NRG1) is involved in neuronal myelination. A haplotype (deCODE) in the 5′ end region of the gene was described to double the risk for schizophrenia in an Icelandic population (Stefansson, H., Sigurdsson, E., Steinthorsdottir, V., Bjornsdottir, S., Sigmundsson, T., Ghosh, S., Brynjolfsson, J., Gunnarsdottir, S., Ivarsson, O., Chou, T.T., Hjaltason, O., Birgisdottir, B., Jonsson, H., Gudnadottir, V.G., Gudmundsdottir, E., Bjornsson, A., Ingvarsson, B., Ingason, A., Sigfusson, S., Hardardottir, H., Harvey, R.P., Lai, D., Zhou, M., Brunner, D., Mutel, V., Gonzalo, A., Lemke, G., Sainz, J., Johannesson, G., Andresson, T., Gudbjartsson, D., Manolescu, A., Frigge, M.L., Gurney, M.E., Kong, A., Gulcher, J.R., Petursson, H., Stefansson, K. 2002. Neuregulin-1 and susceptibility to schizophrenia. Am. J. Hum. Genet. 71, 877–892). Of note, there is now increasing evidence of disturbed myelination in this illness—particularly in subcortical frontal lobe white matter (Konrad, A., Winterer, G. 2008. Disturbed structural connectivity in schizophrenia—primary factor in pathology or epiphenomenon? Schiz. Bull. [Electronic publication ahead of print]). Therefore, we investigated with diffusion tensor imaging (DTI) the impact of a tagging single nucleotide polymorphism (SNP) from the deCODE haplotype, i.e., SNP8NRG221533, on fractional anisotropy (FA), which reflects structural integrity of white matter. SNP8NRG221533 was selected because it gave the single best uncorrected association with schizophrenia in the original report by Stefansson et al. (Stefansson, H., Sigurdsson, E., Steinthorsdottir, V., Bjornsdottir, S., Sigmundsson, T., Ghosh, S., Brynjolfsson, J., Gunnarsdottir, S., Ivarsson, O., Chou, T.T., Hjaltason, O., Birgisdottir, B., Jonsson, H., Gudnadottir, V.G., Gudmundsdottir, E., Bjornsson, A., Ingvarsson, B., Ingason, A., Sigfusson, S., Hardardottir, H., Harvey, R.P., Lai, D., Zhou, M., Brunner, D., Mutel, V., Gonzalo, A., Lemke, G., Sainz, J., Johannesson, G., Andresson, T., Gudbjartsson, D., Manolescu, A., Frigge, M.L., Gurney, M.E., Kong, A., Gulcher, J.R., Petursson, H., Stefansson, K. 2002. Neuregulin-1 and susceptibility to schizophrenia. Am. J. Hum. Genet. 71, 877–892). As predicted, we found medial frontal FA to be significantly associated with this NRG1 gene variation. Using voxel-based morphometry (VBM), we could largely exclude the possibility that this genotype effect is indirectly caused by genotype-dependent effects on brain volume. This is the first demonstration that SNP8NRG221533 of the NRG1 gene affects medial frontal white matter microstructure in humans. As the degree of neuronal myelination contributes to structural integrity, our finding further supports a potential role of NRG1 in neuronal myelination in the human brain. By extension, our findings suggest that SNP8NRG221533 may contribute to the risk for the complex polygenic illness schizophrenia via its impact on myelination in frontal lobe white matter.

Introduction

The human neuregulin-1 (NRG1) gene on chromosome 8p13 is approximately 1.4 Mb long and encodes a family of signaling proteins in various tissues of the body with NRG1 expression being highest in the brain (Corfas et al., 2004, Harrison and Law, 2006). In the nervous system, NRG1 proteins have been implicated in numerous functions, including neuronal migration, synapse formation and receptor expression as well as myelination by regulating oligodendrocyte proliferation and differentiation (Corfas et al., 2004, Steinthorsdottir et al., 2004, Gu et al., 2005). However, there is currently only incomplete knowledge on the specific roles of NRG1 in human central nervous system (CNS) because most studies have been conducted in rodents or cell lines, which mostly focused on the neuromuscular junction or signaling between axons and the peripheral counterparts of oligodendrocytes, i.e., Schwann cells (Chen et al., 2006, Harrison and Law, 2006).

NRG1 has been identified as one of the leading risk genes for schizophrenia illness (Badner and Gershon, 2002, Lewis et al., 2003, Stefansson et al., 2002, Stefansson et al., 2003, Harrison and Weinberger, 2005, Munafo et al., 2006, Harrison and Law, 2006). The majority of these studies found genetic associations with the illness in the region at the 5′ end around the first two exons in close proximity to the deCODE schizophrenia risk haplotype. However, the impact of these gene variations on brain structure and function in schizophrenia is still largely unknown. There is some evidence from postmortem studies of schizophrenic patients that NRG1 mRNA expression might be altered and that these gene variations may possibly influence mRNA expression in various brain regions including the prefrontal cortex (Harrison and Law, 2006). One very recent magnetic resonance study of US-American schizophrenia high-risk subjects of mixed ethnicity also suggests that genetic variation in the NRG1, i.e., the microsatellite 420M9-1395 contained in the deCODE haplotype, may influence brain development (Addington et al., 2007). In this study, an association of schizophrenia risk was also described with SNP8NRG221533, which gave the single best uncorrected association with schizophrenia in the original report by Stefansson et al. (2002). However, an overtransmission of the ‘T’ allele rather than of the ‘C’ allele was observed. Another functional magnetic resonance imaging (fMRI) study in Scottish schizophrenia high-risk families demonstrated that another SNP from the deCODE haplotype, i.e., SNP8NRG243177, is associated with diminished activation in medial prefrontal cortex and at the right temporo-occipital junction (Hall et al., 2006), whereas no association was observed, however, with SNP8NRG221533. In a follow up study using diffusion tensor imaging (DTI), the same group reported that SNP8NRG243177 is associated with reduced white matter density in the anterior limb of the internal capsule in a sample of healthy subjects providing evidence of reduced structural connectivity in the same region (McIntosh et al., 2007).

There is now growing evidence from postmortem and neuroimaging studies that subcortical white matter myelination might be disturbed in schizophrenia—particularly in the frontal lobe (for a review, see Konrad and Winterer, 2008). Since animal data suggest NRG1 to be involved in axonal myelination (Corfas et al., 2004, Steinthorsdottir et al., 2004, Gu et al., 2005), it is conceivable that the NRG1 gene variations that have been associated with schizophrenia may have an impact on white matter neuronal myelination in human brain.

Here, we addressed the question whether the NRG1 tagging SNP (SNP8NRG221533) from the deCODE haplotype affects white matter neuronal myelination in a sample of healthy young Caucasian subjects. SNP8NRG221533 was selected because it gave the strongest association with schizophrenia in the original report of Stefansson et al. (2002) and because it has been associated most consistently – although not always – with schizophrenia illness (Harrison and Law, 2006, Munafo et al., 2006). Young healthy subjects instead of schizophrenia patients were investigated for several reasons. First, schizophrenia is a complex polygenic disorder whereby each of multiple gene variations that are common in the general population is thought to contribute only a minor fraction to the overall risk (Harrison and Weinberger, 2005). Therefore, the effects on brain structure and function of specific gene variations can be investigated in the general population as well, i.e., in healthy subjects – even in the absence of overt behavioral abnormalities seen in patients – provided the employed phenotypic measure is sensitive enough to detect genetic effects (Winterer et al., 2005). While this approach does not allow to study the interaction between the gene variation of interest and any other illness-related factors including additional risk genes, which would only be possible by investigating patients, it avoids that the obtained findings are confounded by treatment-related factors. For instance, it has been reported that chronic exposure to antipsychotic drugs can contribute to glial cell alterations and – by extension – to myelination (Fernø et al., 2005, Konopaske et al., 2007). In order to assess white matter myelination in human brain, we used DTI. DTI is an imaging technique which allows to assess the structural integrity of axonal connections in white matter by measuring fractional anisotropy (FA). FA provides an index of directional selectivity of water diffusion. Currently, the preferred notion for anisotropic (directed) water diffusion is the hindrance of perpendicular water diffusion by the myelin sheath encasing the axons and it has been argued that FA is primarily related to myelination properties and fiber organization (Beaulieu, 2002). However, axonal diameter, fiber density and the ratio of intracellular/extracellular space may also contribute to differences in FA (Schmithorst et al., 2002, Schmithorst et al., 2007).