18 June 2006. Finding a link between a disease and one of the 30,000 or so human genes is one thing, but finding that the disease is linked to two genes that code for interacting proteins seems highly significant, because it pinpoints a specific biological pathway that might be involved in pathology. So when researchers recently found that variations in the gene for the kinase ErbB4 increase the risk for schizophrenia only when they are inherited together with specific variants of the neuregulin 1 (NRG1) gene, it strengthened the case for dysfunctional NRG1-ErbB4 signaling in the disease (see SRF related news story). But despite this, and numerous reports linking neuregulin polymorphisms to schizophrenia (see SRF related news story), no consensus has emerged as to how the NRG1-ErbB4 signaling pathway might be disrupted. The main stumbling block seems to be that those genetic variations, which lie outside the coding region of the genes, have inconsistent, or very subtle, effects on mRNA and protein levels (see SRF related news story). But in the June 12 Nature Medicine online, researchers led by Chang-Gyu Hahn, University of Pennsylvania, Philadelphia, reported that NRG1-ErbB4 signal may, in fact, be altered in schizophrenia patients, even though levels of the two proteins appear normal.
Hahn, together with joint first author and Hoau-Yan Wang and colleagues, tested the neuregulin-ErbB4 pathway by challenging postmortem brain slices with exogenous NRG1, then measuring ErbB4 signaling in extracts prepared so that protein-protein interactions are maintained. Though the authors admit that it is unclear how such measurements relate to what goes on in vivo, they found significant differences between brain samples taken from people who had had schizophrenia and those who had not.
The first difference the authors noted was in activation of ErbB4 itself. Despite the fact that levels of NRG1 and ErbB4 were no different between prefrontal cortex (PFC) slices from schizophrenia patients and matched control samples, the authors found that on addition of exogenous NRG1, twice as much phosphorylated ErbB4 could be detected in PFC slices from schizophrenia subjects. Phosphorylation of the downstream kinases Erk-2 and Akt was also increased relative to control samples, though only slightly.
Why exogenous NRG1 should enhance ErbB4 signaling in schizophrenia brain samples is uncertain, but the downstream effects seem related to one of the major schizophrenia hypotheses, namely that the brain suffers from lack of N-methyl-D-aspartate receptor (NMDAR) activity (see glutamate hypothesis paper by Bita Moghaddam). When the authors stimulated PFC slices with NMDA they found classic signs of NMDAR activation, including increased phosphorylation of subunit 2A of the receptor and recruitment of phosphatidyl inositol phospholipase C-γ1 (PIPLC-γ1) by the NMDAR1 subunit. However, in PFC slices from postmortem schizophrenia brain, both these telltale modifications were significantly lower than in control tissue. The addition of NRG1 attenuated NMDAR activation even further. The latter effect may be related to increased coupling of ErbB4 to postsynaptic density 95 (PSD-95), a synapse protein that has a crucial role in activation of ErbB4 because though the authors found that PSD-95 levels in schizophrenia PFC samples were normal, much more of the protein coimmunoprecipitated with ErbB4 in schizophrenia samples than from control samples. “Our observation that the ErbB4-PSD-95 association was distinctly increased in schizophrenia, with PSD-95 protein levels unaltered, highlights the protein-protein interactions of PSD-95 as a potentially important mode of dysregulation in this disorder,” write the authors.
Though there is accumulating evidence that supports the NMDAR hypofunction hypothesis, “to our knowledge, this is the first evidence of decreased NMDA receptor function in postmortem brains from schizophrenia individuals,” write the authors about their findings. While dysfunctional NRG1-ErbB4 signaling could undoubtedly explain poor NMDAR activity, it is still unclear how non-coding polymorphisms in the two genes may increase the risk for the disease. Nonetheless, these latest findings add weight to both the NRG1-ErbB4 link and to the NMDAR hypothesis.—Tom Fagan.
Hahn C-G, Wang H-Y, Cho D-S, Talbot K, Gur RE, Berrettini WH, Bakshi K, Kamins J, Borgmann-Winter KE, Siegel SJ, Gallop RJ, Arnold SE. Altered neuregulin 1-erbB4 signaling contributes to NMDA receptor hypofunction in schizophrenia. Nature Medicine. June 12, 2006, advanced online publication. Abstract