Bayesian Mapping of Protein Kinases to Vasopressin-Regulated Phosphorylation Sites in Renal Collecting Duct.

Vasopressin controls water permeability in the renal collecting duct by regulating the water channel protein, aquaporin-2 (AQP2). Phosphoproteomic studies have identified multiple proteins that undergo phosphorylation changes in response to vasopressin. The kinases responsible for phosphorylation of most of these sites have not been identified. Here, we use large-scale Bayesian data integration methods to predict the responsible kinases for 51 phospho-proteomically identified vasopressin-regulated phosphorylation sites in the renal collecting duct. To do this, we applied Bayes' Rule to rank the 515 known mammalian protein kinases for each site. Bayes' rule was applied recursively to integrate each of seven independent data sets, each time using the posterior probability vector of a given step as the prior probability vector of the next step. 30 of the 33 phosphorylation sites that increase with vasopressin were predicted to be phosphorylated by protein kinase A catalytic subunit-a (PKA), consistent with prior studies implicating PKA in vasopressin signaling. Eighteen of the vasopressin-regulated phosphorylation sites were decreased in response to vasopressin and all but three of these sites were predicted to be targets of extracellular signal-regulated kinases, ERK1 and ERK2. This result implies that ERK1 and ERK2 are inhibited in response to vasopressin V2 receptor occupation, secondary to PKA activation. The six phosphorylation sites not predicted to be phosphorylated by PKA or ERK1/2 are potential targets of other protein kinases previously implicated in aquaporin-2 regulation, including cyclin-dependent kinase 18 (CDK18), calmodulin-dependent kinase 2d (CAMK2D). AMP-activated kinase catalytic subunit a-1 (PRKAA1) and CDC42 binding protein kinase beta (CDC42BPB).

Database for renal collecting duct regulatory and transporter proteins.

The mammalian kidney collecting duct plays an important role in the fine regulation of Na, K, water, and acid-base balance. Functional genomic and proteomic studies of the kidney offer new opportunities in the understanding of renal physiology and pathophysiology, and the collecting duct is an appropriate target tissue because of the relative simplicity of its cells and the ease of isolating or culturing large numbers of collecting duct cells. Study of the collecting duct includes assessment of gene expression and protein regulation and abundance. For example, DNA and protein microarrays can be used to quantitate gene expression and protein regulation and abundance under varying physiological conditions. An Internet-accessible database has been devised for major collecting duct proteins involved in transport and regulation of cellular processes. The individual proteins included in this database are those culled from literature searches and from previously published studies involving cDNA arrays and serial analysis of gene expression (SAGE). Design of microarray targets for the study of kidney collecting duct tissues is facilitated by the database, which includes links to curated base pair and amino acid sequence data, relevant literature, and related databases. Use of the database is illustrated by a search for water channel proteins, aquaporins, and by a subsequent search for vasopressin receptors. Links are shown to the literature and to sequence data for human, rat, and mouse, as well as to relevant web-based resources. Extension of the database is dynamic and is done through a maintenance interface. This permits creation of new categories, updating of existing entries, and addition of new ones.

Immunomorphometric study of rat renal inner medulla.

We utilized immunofluorescent immunolabeling of renal tissue sections to identify and count tubules at specified depths of the rat renal inner medulla. We used primary antibodies to aquaporin-1 (AQP1; labeling thin descending limbs), aquaporin-2 (AQP2; labeling inner medullary collecting ducts), the rat kidney-specific chloride channel (ClC-K1; labeling thin ascending limbs), and von Willebrand factor (labeling descending vasa recta). Secondary antibodies conjugated to different fluorophores were used, giving up to a three-color display. Labeled structures were then identified and counted. At each level sampled in the inner medulla, many more thin limbs were labeled by ClC-K1 than AQP1. In addition, thin limbs were found to label with antibodies to ClC-K1 on both sides of their hairpin turns. We conclude that the descending thin limbs shift from expressing AQP1 to expressing ClC-K1 some distance before the point where they turn and begin to ascend. Mathematical models can use our quantitative data to explore implications for the urine-concentrating mechanism.