TISSUE-SPECIFIC EXPRESSION OF MESSENGER-RNA ENCODING RAT-KIDNEY WATER CHANNEL CHIP28K BY INSITU HYBRIDIZATION

The tissue distribution of mRNA encoding rat kidney water channel CHIP28k was determined by in situ hybridization. cDNA encoding rat kidney CHIP28k was isolated by homology to human erythrocyte CHIP28 (G. M. Preston and P. Agre. Proc. Natl. Acad. Sci. USA 88: 11110-11114, 199 1) and used to construct 155-base 35S-labeled cRNA sense and antisense probes corresponding to base pair 7-162. Fixed and frozen tissue were cut in 6- to 12-mum sections, hybridized with probes at 55-degrees-C for 16 h, and exposed for 5-9 days. In renal cortex, CHIP28k mRNA was detected intensely on proximal tubule epithelial cells but not in glomeruli or collecting duct. Hybridization to proximal tubule was strongest in deep renal cortex. In no study was there significant hybridization of sense cRNA probe. In renal papilla, CHIP28k mRNA was detected in only a fraction of tubules corresponding to thin limbs of Henle. Hybridization in spleen was observed in red splenic pulp containing erythroid precursors but not in white pulp. In colon, there was selective hybridization in crypt epithelial cells but not in villus epithelial cells or nonepithelial structures. In lung, hybridization was observed in alveolar epithelial cells. In eye, there was selective hybridization in corneal endothelium and ciliary body. No hybridization was observed in any cell types in liver. Northern analysis revealed a 2.8-kilobase mRNA encoding CHIP28k in kidney cortex and papilla but not in brain, skeletal muscle, and liver. These results indicate a wide and highly selective tissue distribution of mRNA encoding the CHIP28k water channel. The high water permeability in tissues expressing the CHIP28k water channel is likely to have physiological importance in salt and fluid movement.