The fine-structural distribution of G-protein receptor kinase 3, beta-arrestin-2, Ca2+/calmodulin-dependent protein kinase II and phosphodiesterase PDE1C2, and a Cl(-)-cotransporter in rodent olfactory epithelia

The sequentially activated molecules of olfactory signal-onset are mostly concentrated in the long, thin distal parts of olfactory epithelial receptor cell cilia. Is this also true for molecules of olfactory signal-termination and -regulation? G-protein receptor kinase 3 (GRK3) supposedly aids in signal desensitization at the level of odor receptors, whereas beta-arrestin-2, Ca2+/calmodulin-dependent protein kinase II (CaMKII) and phosphodiesterase (PDE) PDE1C2 are thought to do so at the level of the adenylyl cyclase, ACIII. The Na+, K(+)-2Cl(-)-cotransporter NKCC1 regulates Cl(-)-channel activity. In an attempt to localize the subcellular sites olfactory signal-termination and -regulation we used four antibodies to GRK3, two to beta-arrestin-2, five to CaMKII (one to both the alpha and beta form, and two each specific to CaMKII alpha and beta), two to PDE1C2, and three to Cl(-)-cotransporters. Only antibodies to Cl(-)-cotransporters labeled cytoplasmic compartments of, especially, supporting cells but also those of receptor cells. For all other antibodies, immunoreactivity was mostly restricted to the olfactory epithelial luminal border, confirming light microscopic studies that had shown that antibodies to GRK3, beta- arrestin-2, CaMKII, and PDE1C2 labeled this region. Labeling did indeed include receptor cell cilia but occurred in microvilli of neighboring supporting cells as well. Apical parts of microvillous cells that are distinct from supporting cells, and also of ciliated respiratory cells, immunoreacted slightly with most antibodies. When peptides were available, antibody preabsorption with an excess of peptide reduced labeling intensities. Though some of the antibodies did label apices and microvilli of vomeronasal (VNO) supporting cells, none immunoreacted with VNO sensory structures.

The interaction of Bex and OMP reveals a dimer of OMP with a short half-life

Olfactory marker protein (OMP) participates in the olfactory signal transduction pathway. This is evident from the behavioral and electrophysiological deficits of OMP-null mice, which can be reversed by intranasal infection of olfactory sensory neurons with an OMP-expressing adenovirus. Bex, brain expressed X-linked protein, has been identified as a protein that interacts with OMP. We have now further characterized the interaction of OMP and Bex1/2 by in vitro binding assays and by immuno-coprecipitation experiments. OMP is a 19 kDa protein but these immunoprecipitation studies have revealed the unexpected presence of a 38 kDa band in addition to the expected 19 kDa band. Furthermore, the 38 kDa form was preferentially co-immunoprecipitated with Bex from cell extracts. In-gel tryptic digestion, mass spectrometry, and two-dimensional gel electrophoresis indicate that the 38 kDa protein behaves as a covalently cross-linked OMP-homodimer. The 38 kDa band was also identified in western blots of olfactory epithelium demonstrating its presence in vivo. The stabilities and subcellular localizations of the OMP-monomer and -dimer were studied in transfected cells. These results demonstrated that the OMP-dimer is much less stable than the monomer, and that while the monomer is present both in the nuclear and cytosolic compartments, the dimer is preferentially located in a Triton X-100 insoluble cytoskeletal fraction. These novel observations led us to hypothesize that regulation of the level of the rapidly turning-over OMP-dimer and its interaction with Bex1/2 is critical for OMP function in sensory transduction.