ICOSR 2007—Synaptic Structural Plasticity and Schizophrenia
Editor's Note: On Friday, 30 March 2007, at the International Congress on Schizophrenia Research, James Bibb of the University of Texas Southwestern convened a session entitled, “Regulation of cytoskeletal dynamics and synaptic structural plasticity in schizophrenia.” ICOSR Young Investigator awardee Vibeke S. Catts of the University of Queensland, Australia, was there to report on it for you:
13 April 2007. The reduced neuropil hypothesis, which proposes that dendritic arborization of large neurons is reduced in brains from patients with schizophrenia (Selemon and Goldman-Rakic, 1999) has been supported by a number of human postmortem Golgi-staining studies that show reduced dendritic length and reduced spine density of layer 3 pyramidal neurons in the dorsolateral prefrontal cortex (e.g., Garey et al., 1998; Glantz and Lewis, 2000). In light of these observations, the symposium on synaptic plasticity was of considerable interest to attending researchers. Clinicians interested in early psychosis would be particularly interested in the observations concerning increased synaptic spine plasticity in the adolescent period described by James Bibb and Wen-Biao Gan, as these observations may relate to the “critical period” of intervention.
James Bibb, who chaired the symposium, described the work of Huttenloch and Dabholkar (1997) demonstrating regional differences in the production of synapses during infancy and subsequent elimination during childhood and adolescence. A recent paper by Bibb’s group (Hayashi et al., 2006) reported a similar pattern of age-related changes in the phosphorylation of stathmin, a tubulin-binding protein important in structural plasticity, in mice. These workers also observed altered stathmin phosphorylation in postmortem tissue from patients with schizophrenia, consistent with altered neuroplasticity. This example of bringing together structural and molecular findings concerning neural plasticity from the field of basic science with those from schizophrenia research illustrates the progress now possible in the elucidation of the molecular underpinnings of the reduced neuropil in schizophrenia.
Wen-Biao Gan from New York University followed with a description of in vivo stability of spines on apical dendrites from layer 5 pyramidal cells in adolescent (1 month old) and adult (>4 months old) mice. He reported that spine elimination is significantly higher than spine formation in late postnatal life, with greater instability of spines during adolescence compared with adulthood (see Zuo et al., 2005a ). Sensory deprivation by whisker trimming during adolescence prevents net spine loss by reducing the rate of ongoing spine elimination, an effect not apparent when whiskers were trimmed during adulthood, suggesting that the effect of sensory deprivation on spine elimination decreases as mice reach adulthood (Zuo et al., 2005b ). Gan concluded his talk by showing that in mice with impaired sensory gating (lacking the voltage-gated potassium channel, Kv3.2), there is increased spine elimination, which can be rescued by sensory deprivation during adolescence (unpublished results).
David Lewis, University of Pittsburgh, described a series of findings which support the reduced neuropil hypothesis of schizophrenia, focusing specifically on those pertaining to decreased somal volume. He brought to the attention of the audience a recent paper from his group which demonstrated a potential confound of immunoperoxidase cell labeling techniques, namely, that these techniques are associated with an overestimation of the volume of labeled neurons in postmortem tissue, and that this overestimation is more pronounced in tissue from subjects with schizophrenia. This confound, therefore, has the potential to obscure differences in somal volume between clinical groups (Maldonado-Avilés et al., 2006). Lewis concluded his talk with a review of data from his group concerning alterations in the signaling pathways regulating spine dynamics in schizophrenia. mRNA expression levels of Cdc42, a RhoGTPase regulating spine dynamics, and Duo, a protein involved in spine formation and maintenance, were found to be decreased in postmortem tissue from patients with schizophrenia, and both were strongly correlated with the spine density observed in the tissue examined. Analysis of brain tissue from macaque monkeys treated with antipsychotic medication suggested that these observed changes were not due to medication effects (Hill et al., 2006; see SRF news story).
The symposium’s next speaker, Angus Nairn from Yale University, took the audience on a tour of some of the molecular machinery underlying spine morphological changes. First, he described how lack of functional WAVE1 protein resulted in a predominance of immature spines in mouse brain in vivo. Furthermore, phosphorylation of the WAVE1 protein by Cdk5 regulates actin polymerization and spine morphology (Kim et al., 2006). Actin is the principal cytoskeletal protein of dendritic spines, and the regulation of actin assembly and disassembly is thought to underlie spines’ ability to change shape. Spinophilin is one of many transmembrane proteins with a PDZ domain that anchors these proteins to the cytoskeleton and holds together signaling complexes. Phosphorylation of spinophilin modulates this protein’s association with actin, but not with protein phosphatase 1 (PP1; Hsieh-Wilson et al., 2003). Spinophilin has been demonstrated also to interact with AMPA and NMDA receptor subunits, suggesting that spinophilin is important for targeting PP1 to phosphorylation sites on AMPA and NMDA receptor subunits (Kelker et al., 2007). Further unraveling the machinery underlying spine morphological changes, Nairn introduced the audience to another molecular player, Lcf (named after lbc [lymphoid blast crisis]’s first cousin), a Rho-specific guanine nucleotide exchange factor (GEF). Lcf normally resides in the shaft of dendrites in association with microtubules (dendritic cytoskeletal proteins), but upon neuronal stimulation, Lcf translocates to spines. Lcf interacts with spinophilin and provides a link between microtubule and actin cytoskeletons (Ryan et al., 2005) and is important for regulation of spine size.
The last speaker of the morning, Andrew Matus from the Friedrich Miescher Institute in Switzerland, described the role of the cytoskeleton in synaptic plasticity. Specifically, he described how profilin, a regulator of actin polymerization, is targeted to spine heads in response to postsynaptic NMDA receptor activation in in vitro cultures of hippocampal neurons (Ackermann and Matus, 2003) and how this molecule remains in the activated spines for extended periods of time. Similar results have been observed in vivo, where fear conditioning in rats leads to the movement of profilin into dendritic spines in the lateral amygdala (Lamprecht et al., 2006). Interestingly, profilin is also found in the cell nucleus, where it functions as the nucleotide exchange factor for actin and as a cofactor for actin export from the nucleus, thereby suppressing actin polymerization in the nucleus (Stüven et al., 2003). Neuronal activity in vitro has been observed to result in dissociation of actin and profilin in the cell body, and leads to profilin entering the nucleus (Birbach et al., 2006), where it can regulate the activity of p42[POP], a novel Myb-related transcription factor (Lederer et al., 2005). Together, these observations provide a potential mechanism for the “synaptic tagging” and signaling from synapse to nucleus required for the establishment of long-term potentiation as proposed by Frey and Morris (1997).
Taken together, the talks delivered during the symposium were a “tour de force” of the molecular regulation of synaptic structural plasticity in health and as it may apply to schizophrenia. The symposium provided schizophrenia researchers with the rare privilege of hearing, in a single thematic session, directly from the basic scientists who have developed the techniques and made the major discoveries. In particular, the symposium detailed two novel techniques, namely, labeling of a subset of cells by ballistic delivery of fluorescent indicators into multiple cell types, providing rapid labeling of neurons and their processes both in vivo and in vitro (Grutzendler et al., 2003); and the generation of transgenic animals in which fluorescent proteins are selectively expressed in neurons (Feng et al., 2000), permitting real-time visualization of cell generation, migration, differentiation, and growth. Combining these techniques with the development of transgenic animals modeling aspects of schizophrenia, and with confirmatory postmortem human tissue studies, has the potential to significantly enhance our understanding of the molecular mechanisms underlying the reduced neuropil observed in patients with schizophrenia.—Vibeke Catts.