In glutamate receptor, instead of binding to a ligand, this domain appears to be involved in protein interaction and allosterically modulates channel open probability and determines the subcellular localization. When the NR3B PI-103 insCGTT type was expressed in heterologous cells, much of the protein was trapped in PLX-4720 intracellular organelles, likely at the endoplasmitic reticulum or Golgi apparatus in a form that cannot be readily extracted under alkaline conditions that normally extract luminal proteins. Also the NR3B insCGTT type associates with the cell surface, to approximately 50% of the major type protein. Given that this variation does not have a transmembrane or membrane associated domain, the remaining extracellular domain in the NR3B insCGTT type most likely accounts for this property. Co-immunoprecipitation data demonstrated that NR3B insCGTT type can interact with the NR1 subunit and suggested that NR1 may associate with NR3B insCGTT on the ER membrane, and a part of the complex can leave the ER then stably localize in the cell membrane. Alternatively, a binding protein functionally similar to pentraxin for AMPA type glutamate receptors may exist for NR3B and limit the dissociation from the cellular membrane. However, this remaining portion of the protein is apparently non-functional as demonstrated with electrophysiological recordings. It remains to be shown whether such translation products exist or not in the neuronal tissue of carriers. It is also possible that the insCGTT type mRNA is subject to nonsense- mediated mRNA decay. In either case, in vivo, the NR3B insCGTT type results in a functionally null allele in carriers. Among schizophrenia patients and healthy controls, the NR3B insCGTT type was significantly overrepresented in the patients compared with controls. Also, NR3B insCGTT type was associated with impaired performance in schizophrenia related phenotypes��SPQ, PPI and WCST. We previously found that the Grin3b knockout mouse have a phenotype suggestive of increased anxiety, such as decreased entry and time spent in the open arm of the elevated plus maze task. The animals showed an overall tendency for a reduction in pre-pulse inhibition compared with wild type animals, which is consistent with our finding in schizophrenia patients, though this effect did not reach statistical significance. However, the mice showed behavior suggestive of increased social interaction in the home cage. While this indicates that NR3B plays a role in the regulation of social behavior, it is not completely consistent with the schizophrenia phenotype seen in humans.