2 May 2007. Genetic association and expression studies have implicated the neuregulin1 growth factor (NRG1) and its ErbB receptors, particularly ErbB4, in raising the risk of schizophrenia (Stefansson et al., 2002). A paper out May 1 in PNAS online from Gabriel Corfas and colleagues from Harvard Medical School in Boston, Massachusetts, adds substantial evidence to previous suggestions that defects in NRG1/ErbB4 signaling could underlie white matter defects in schizophrenia. The study finds that blocking NRG1/ErbB4 signaling in oligodendrocytes in mice causes a defect in myelination and slower nerve transmission in the brain. Other changes that they report, including enhanced dopaminergic signaling, decreased activity, impaired social interactions, and sensitivity to amphetamine, may model some of the manifestations of schizophrenia.
A second paper, led by the Icelandic company that originally linked neuregulin to schizophrenia, looks at how the loss of the NRG1/ErbB pathway affects neurons and synaptic function. In the April 25 Journal of Neuroscience, a research team led by Mark Gurney at deCODE Genetics in Reykjavik, Iceland, and their collaborators at Roche Palo Alto in California report that NRG1 regulates the function of NMDA-type glutamate receptors, thereby exerting a direct effect on neurotransmission through glutamatergic pathways, another locus of aberration in schizophrenia (see Current Hypothesis paper by B. Moghaddam).
To find out if NRG1/ErbB signaling contributes to CNS myelination, and whether that might play a role in psychiatric disorders, joint first authors Kristine Roy and Joshue Murtie and the Harvard team created transgenic mice with a targeted block of ErbB signaling only in oligodendrocytes (OL), the cells that wrap CNS axons with myelin. The ErbB family of receptors features extracellular ligand binding domains linked to cytoplasmic tyrosine kinase signaling domains. Expression of a truncated receptor lacking the cytoplasmic domain has been shown to block signaling through endogenous receptors in a dominant negative fashion. By making transgenic mice with a truncated ErbB4 gene under control of an OL-specific promoter, the investigators produced mice that expressed the dominant negative protein only in OLs and their precursors. Because the truncated form can dimerize with other ErbB family members, Roy and colleagues detected inhibition of signaling not just through the ErbB4 receptor, but the related -B2 and -B3 receptors as well.
Since NRG1/ErbB was believed to be required to maintain the OL lineage, the investigators thought they might find fewer OLs in the transgenic mice, but that was not the case. OL number actually increased by 40 percent, but morphology was not abnormal. The cells were smaller, and their processes had fewer branches. These simplified, smaller OLs would be expected to cover less axon area, and indeed the researchers found thinner myelin sheaths in the optic nerve and the corpus callosum of transgenic mice.
In young mice, the changes in myelination resulted in slower nerve conduction speeds and changes in behavior. Transgenic mice explored in the open field less, displayed increased anxiety, and had abnormal social interactions. The mice were also hypersensitive to repeated doses of amphetamine, a sign of defects in dopamine function seen in schizophrenia. The researchers took a closer look at the dopamine system in the transgenic mice, and found that levels of dopamine transporters and dopamine receptor type 1-like binding increased in several regions of the brain. Moreover, stimulation with either a dopamine receptor agonist or amphetamine induced greater responses in gene expression, dopamine release, and behavioral measures than in control mice.
“Our findings indicate that defects in OL structure/function can cause alterations in neurotransmission that are relevant to psychiatric diseases,” the authors conclude. The study may help to resolve an outstanding question from NRG1/ErbB4 genetics work, that is, whether the loss of the pathway or a gain of function contributes to the features of schizophrenia. This study suggests a loss-of-function model for these genes. It also suggests that one lesion could contribute to both positive symptoms (as evidenced by increased sensitivity to amphetamines in this study) and negative ones (as evidenced by hypoactivity, and social withdrawal).
In a press release accompanying the paper, Corfas also speculated that the involvement of white matter could help explain why schizophrenia often sets in during adolescence or early adulthood, a time of active myelination of the prefrontal cortex.
NRG1 and neurons
The NRG1/ErbB pathway could contribute to some of the symptoms of schizophrenia through effects in neurons as well, says the second report. In that paper, the researchers show that interfering with NRG1 signaling (they use NRG1 or ErbB knockout mice) leads to reductions in NMDA receptor phosphorylation and changes in its function (see SRF related news story). Treatment with the atypical antipsychotic drug clozapine reversed the receptor hypophosphorylation and improved behavioral abnormalities in the mice. The results indicate that alterations in the NRG1/ErbB pathway and resulting alterations of NMDA receptor function might cause some of the pathophysiology of schizophrenia.
Joint first authors Maria Bjarnadottir and Dinah Misner started out looking for downstream effectors of the NRG1/ErbB4 pair. A yeast two-hybrid assay with the ErbB4 cytoplasmic tail and a mass spectrometry analysis of ErbB4 protein complexes immunoprecipitated from brain identified the cytoplasmic tyrosine kinases Fyn and Pyk2, respectively, as binding partners for the receptor. In cells overexpressing ErbB4 and Fyn, the researchers found that NRG1 caused Fyn activation.
Both kinases can phosphorylate a regulatory site on the NR2A subunit of the NMDA receptor, and the researchers determined that NRG1 treatment of human neuroblastoma cells resulted in an increase in NR2A phosphorylation at Y1472. Conversely, the same site is hypophosphorylated in either NRG1 or ErbB4 heterozygous knockout mice, suggesting the pathway functions in vivo.
Consistent with the idea that NRG1/ErbB4 regulates NMDA receptor activity, the knockout mice showed changes in some measures of synaptic plasticity. Depending on the stimulus, they had defects in LTP and changes in short-term synaptic plasticity. Treating hippocampal slices from NRG1+/- animals with exogenous neuregulin reversed some effects. The dose-response curves were complicated, suggesting that there may be an optimal level of NRG1 required for proper synapse formation.
The deCODE group previously showed that their NRG1+/- mice have behavioral abnormalities that could be reversed by treatment with clozapine (Stefansson et al., 2002). The current study extends this by showing that the same treatment also restores normal levels of NMDA receptor phosphorylation.
“These data suggest to us that NRG1-associated susceptibility to schizophrenia is at least partly associated with hypofunction of NRG1 signaling through ErbB4, Fyn, and other associated kinases such as Pyk2, that phosphorylate regulatory sites on NMDAR subunits, resulting in abnormal modulation of excitatory glutamatergic neurotransmission,” the authors write. These data are consistent, they say, with the glutamatergic hypothesis of schizophrenia, and studies implicating several other genes involved in glutamatergic signaling in the disease.
Both reports suggest potential new avenues for treatment, including restoring oligodendrocyte function by augmenting the NRG1/ErbB4 signaling pathway, or normalizing NMDA receptor phosphorylation and function.—Pat McCaffrey.
Roy K, Murtie JC, El-Khodor BF, Edgar N, Sardi SP, Hooks BM, Benoit-Marand M, Chen C, Moore H, O'Donnell P, Brunner D, Corfas G. Loss of erbB signaling in oligodendrocytes alters myelin and dopaminergic function, a potential mechanism for neuropsychiatric disorders. 2007 May 1; PNAS Early Edition.
Bjarnadottir et al. Neuregulin1 (NRG1) signaling through Fyn modulates NMDA receptor phosphorylation: Differential synaptic function in NRG+/- knock-outs compared with wild-type mice. J. Neurosci. 2007 April 25; 27:4519-4529. Abstract