Genetic and epigenetic factors determine the functional properties of the brain and skeletal muscles during development and can be altered in disease. Our laboratory focuses on elucidating the functions of a factor known as Neuregulin-1 (NRG-1) and its cognate receptor ErbB4 in the developing brain. NRG-1 and ErbB4 have been genetically identified as schizophrenia "at-risk" genes. We recently demonstrated that NRG-1 and its receptor ErbB4 regulate synaptic plasticity in the hippocampus by modulating long-term potentiation, a cellular mechanism implicated in learning, memory, and cognition. Our more recent studies demonstrated that NRG-1 regulates gamma rhythm activity in the brain and uncovered an important relationship between the NRG-1, glutamate, and dopamine-signaling pathways. We also found that ErbB4 is not detectable in excitatory pyramidal cells; instead, the receptor is expressed in three classes of inhibitory interneurons in the hippocampus. Moreover, we found that, within these interneurons, ErbB4 is targeted to glutamatergic postsynaptic sites. In our most recent studies, we found that in the frontal cortex of mice as well as in the cortex of primates ErbB4 expression is also confined to GABAergic interneurons. The conserved regional and subcellular expression of ErbB4 from rodents to primates validates the use of mice to investigate the cellular mechanisms that may be altered in psychiatric disorders. Our findings provide an important and novel framework within which to test how these signaling systems, which include the NRG/ErbB network and various neurotransmitters, all implicated either genetically or pharmacologically in schizophrenia, may contribute to psychiatric disorders.
The contractile properties of slow-twitch (red) and fast-twitch (white) muscles are also determined by genetic and epigenetic factors. Genetic cues are important during early development to differentiate muscle types; later in development, activity, in the form of exercise, can modify muscle types. Our group identified DNA–regulatory sequences and transcription factors that modify the contractile properties of skeletal muscles during development and, in an activity-dependent fashion, in the adult. Interestingly, the factors regulate different contractile genes in response to distinct frequencies of muscle depolarization. Our goal is to uncover how these transcription factors can "sense" slow and fast patterns of motor neuron depolarization and consequently intraconvert the contractile properties of slow and fast muscles.
Polymorphisms in genes encoding NRG1 and its receptor ErbB4 are associated with risk for developing schizophrenia. A single nucleotide polymorphism (SNP) in NRG1, designated SNP8NRG243177 (T/T), is associated with reduced prefrontal cortical function, working memory, myelination, and premorbid IQ. Because SNP8NRG243177 (T/T) has characteristics of a functional polymorphism and maps close to DNA sequences encoding NRG1 type IV, which is 1 of 7 NRG1 mRNA isoforms, our goal was to precisely map the type IV transcription initiation site and to investigate the properties and subcellular distribution of NRG1 type IV protein. We mapped a novel type IV transcription initiation site and isolated two full-length mRNAs encoding type IV proteins. Using our type IV–specific antiserum, we found that pro–NRG1 type IV is targeted to the plasma membrane, where it is proteolitically released through a PKC–dependent mechanism, similar to other NRG1 isoforms. However, unlike NRG1 type III, which is expressed in the somato-dendritic and axonal compartments of neurons, NRG1 type IV and its close homolog NRG1 type I are excluded selectively from axons. These results constitute an important step toward understanding how alterations in NRG1 type IV expression levels associated with SNP8NRG243177 (T/T) could selectively modify signaling from NRG1 released from somato-dendritic compartments.
Knowledge of the cellular and subcellular localization of the ErbB4 receptor is important for understanding how NRG1 regulates neuronal network activity and behavior, and it is thus critical to resolve where ErbB4 receptors are expressed in the primate cortex by using reagents with demonstrated specificity. Using a combination of electrophysiological profiling of pyramidal and GABAergic cells and single-cell RT-PCR in mice, we demonstrated, in collaboration with Chris McBain's group, that ErbB4 transcripts are expressed selectively in interneurons in the frontal cortex. Using highly specific monoclonal antibodies we had developed and carefully characterized, we showed that ErbB4 protein is undetectable in pyramidal cells of rodent, rhesus monkey, and human frontal cortex. In collaboration with David Lewis's group, we found that almost all interneurons positive for parvalbumin, calretinin, or cholecystokinin but only a minority of calbindin-positive cells coexpress ErbB4. Notably, we detected no presynaptic ErbB4 expression in any species. Our data validate the use of rodents to analyze cellular and neural-circuit effects of abnormal ErbB4 function as a means to model endophenotypes pertinent to psychiatric disorders.
Using the highly specific monoclonal antibodies described above, we found that ErbB4 levels are high in association cortices, intermediate in sensory cortices, and relatively low in motor cortices. The overall immunoreactivity in the hippocampal formation is intermediate, but is high in a subset of interneurons. We detected the highest overall immunoreactivity in distinct locations of the ventral hypothalamus, medial habenula, intercalated nuclei of the amygdala and structures of the ventral forebrain, such as the islands of Calleja, olfactory tubercle, and ventral pallidum. While this pattern is generally consistent with ErbB4 mRNA expression data, further investigations are needed to identify the exact cellular and subcellular sources of mRNA and protein expression in these areas. We detected only low levels of ErbB4 immunoreactivity in dopaminergic nuclei of the mesencephalon but additional medium-to-high expression in the dorsal striatum and high expression in the ventral forebrain, suggesting that most ErbB4 protein in dopaminergic neurons could be transported to axons. Based on these findings, we conclude that the NRG-ErbB4 signaling pathway can potentially influence many functional systems throughout the brain of primates, and suggest that major sites of action are areas of the "corticolimbic" network. This interpretation is functionally consistent with the genetic association of NRG1 and ERBB4 with schizophrenia.
Previous studies demonstrated that NRG1 affects plasticity at glutamatergic synapses in principal glutamatergic neurons in the hippocampus and frontal cortex; however, these effects are indirect because expression of ErbB4 is confined to GABAergic interneurons in these brain regions. In collaboration with Catherine Fenster, we analyzed the effects of NRG1 on developing cultured cerebellar granule cells (CGCs). The cultures offer the advantage that they are relatively homogenous and consist primarily of granule neurons that express ErbB4. We found that NRG1 does not affect whole-cell AMPAR– or NMDAR–mediated currents or the frequency or amplitude of spontaneous NMDAR– or AMPAR–mediated miniature excitatory post-synaptic currents, in baseline conditions of CGCs grown for 10–12 days in vitro. However, we found that high glycine induces a form of chemical potentiation (chemLTP) in CGCs characterized by an increase in AMPAR-mEPSC frequency that is reduced by NRG1 treatment. In our culture conditions, CGCs express very low levels of the AMPA receptor GluR1 subunit; thus our data suggest that the NRG1 effect could be mediated via GluR4 subunits, which are highly expressed. The study provides the first evidence that, in CGCs, high glycine can induce plasticity at glutamatergic synapses and that acute NRG1/ErbB-signaling can regulate glutamatergic plasticity. Taken together with previous reports, our results suggest that, similar to Schaffer collateral–CA1 synapses, NRG1 effects are activity-dependent and mediated via modulation of synaptic AMPARs.
Most studies on activity-dependent regulation of muscle genes have focused on genes encoding slow contractile proteins in response to slow-patterned motoneuron depolarization (approximately 10–20 Hz). However, as we and others have shown, fast-patterned (greater than 50 Hz) activity regulates the expression of genes encoding muscle proteins specific to fast-twitch muscles. In collaboration with Kristian Gundersen, we used cDNA microarrays to analyze the effects of skeletal muscle denervation and stimulation of denervated muscles with either slow or fast patterns of activity on myofiber type–specific gene expression. We analyzed expression of mRNA at multiple time points following the stimulation of muscles to analyze temporal expression profiles of transcription factors and myofibril proteins. Interestingly, we found that members of the ets family of transcription regulators, which are sensitive to calcium levels, are selectively regulated by fast-patterned activity shortly after stimulation commences. Utilizing the Troponin I fast (TnIf) intronic regulatory element (FIRE), we identified sequences that are necessary to regulate transcription in response to fast-patterned activity and that correspond to potential ets consensus binding sites. siRNA–mediated knockdown experiments confirmed the specificity and importance of these factors to regulate TnIf in response to fast-depolarization patterns.
All related news