21 December 2010. Researchers have identified 1,461 proteins in the human cortical post-synaptic density (PSD), a multi-protein complex found on the receiving end of synapses in the brain. As reported in Nature Neuroscience on December 16, mutations in the genes encoding some of these proteins result in 133 human nervous system diseases, highlighting the importance of the PSD and its potential involvement in other brain disorders, including schizophrenia.
Named for the darkly staining region on the post-synaptic side of synapses visible by electron microscopy, the PSD comprises an interconnected network of proteins that includes neurotransmitter receptors, cell adhesion molecules, and scaffolding and signaling proteins. Once thought of as a static structure, the PSD is now recognized as a dynamic machine that can shuffle around some of its component parts according to a cell's needs. The new study throws some light on this complicated machine by essentially coming up with its parts list. It also found a connection between mutations in these parts and human disease and phenotypes, arguing that the PSD is a place of particular interest when looking for the origins of brain disorders.
"We've developed a kind of template or roadmap for investigating human synapse function," Seth Grant, senior author of the study and professor at the Wellcome Trust Sanger Institute in the United Kingdom, told SRF. "This could help drug discovery for treatment of various human diseases."
Though no one doubts that synapses are vital for the brain to work properly, no one had ever tallied up the number of diseases stemming from malfunctions in synaptic components. "Now we've put a number down on the table for PSD diseases, though we expect that number to rise with time," said Grant.
First authors Álex Bayés and Louie van de Lagemaat and colleagues took a different approach from typical genetic studies of disease, which often deliver a list of seemingly unrelated genes. Instead, the team started with proteins that are already known to function together and asked which of them, when mutated, resulted in disease. They began by extracting the PSD proteins from neocortical tissue taken from nine adults undergoing brain surgery. The nine samples were pooled and divided into three groups. Using mass spectrometry, the researchers identified 1,461 proteins total, each encoded by a different gene. Of these, 748 proteins were found in all three groups, representing a "consensus" protein set. These numbers are similar in magnitude to those found for proteomic characterizations of the mouse PSD (Collins et al., 2006).
Though the vast majority of the proteins on this list (see the Genes2Cognition website) had never been connected to the PSD before, many are familiar. Some, like GABA receptors, are embedded in the PSD, whereas others, like MAP kinase, have additional jobs beyond the PSD in other parts of the neuron and in other cell types. Some of these PSD genes already have a reputation in human disease, including HTT for Huntington's and ApoE in Alzheimer's. NRXN1, a gene whose deletion is associated with schizophrenia and autism, also turned up.
Cross-reference to disease
This list allowed the researchers to systematically track the effects of mutation in each of these PSD genes on human health. The researchers limited themselves to diseases in the Online Mendelian Inheritance in Man (OMIM) database, which catalogs monogenic diseases. They found that 269 diseases in the OMIM stemmed from mutations to 199 of these PSD genes, and 133 of these diseases were nervous system disorders. Similar proportions resulted when looking only at mutations in the consensus PSD. These included neurodegenerative diseases like Alzheimer's, as well as mental retardation, motor disorders, and epilepsy. Though the number of diseases linked to mutations in the human PSD was higher than expected, this number is likely to increase once complex psychiatric diseases like schizophrenia are included, Grant told SRF.
The team also tried to get beyond disease classifications by describing in phenotypic terms what these PSD proteins might be doing. For this, they turned to the Human Phenotype Ontology (HPO) database, which categorizes the constellation of signs and symptoms associated with different diseases into discrete phenotypes. Mutations to the human PSD were associated more often with cognitive and motor phenotypes than were mutations to other neuronal genes, arguing that the PSD is particularly enriched for susceptibility to these kinds of symptoms. This analysis also revealed that multiple genes contributed to particular phenotypes; for example, 40 PSD genes were related to mental retardation and 20 to muscle spasticity.
Future work will have to fill in the picture of exactly how these many proteins work together. From a therapeutic standpoint, it will be particularly interesting to see which ones act through a common mechanism within the PSD. These subsets of proteins may bridge the gap between the encouraging sign that many diseases stem from malfunctions in the same synaptic structure and the distress at seeing its very long parts list. Multiple proteins that converge on the same machinery within the PSD could reveal new targets for drug design, potentially providing a way of normalizing PSD function and of treating multiple diseases.—Michele Solis.
Bayés Á, van de Lagemaat LN, Collins MO, Croning MDR, Whittle IR, Choudhary JS, Grant SGN. Characterization of the proteome, diseases and evolution of the human post-synaptic density. Nat Neurosci. 2010 Dec 16. Abstract