2002 Global Researcher Conference Proceeding

April 26 - 28, 2002

Conference: 2002 Global Researcher Conference
Title: Impaired routing of AQP2 to late endosomes/lysosomes following heterotetramerization with AQP2-E258K is likely to explain dominant nephrogenic diabetes insipidus
Authors: Savelkoul, Paul J.M.; Kamsteeg, Erik-Jan; Konings, Irene B. M.; Slot, Jan-Willem; van der Sluijs, Peter; Deen, Peter M.T.
Institutions: NHGRI, University of Nijmegen, Radboud University Nijmegen Medical Center, Univ. Utrecht, Utrecht University

Binding of vasopressin to its V2-receptor in renal collecting duct cells increases intracellular cAMP and calcium levels, resulting in the phosphorylation of Aquaporin-2 (AQP2) at S256 by protein kinase A. Subsequently, vesicles containing AQP2 fuse with the apical membrane (AM), rendering these cells water permeable. Also, phosphorylation of S256 by the Golgi casein kinase (GCK) followed by dephosphorylation has been suggested to be essential for exit of AQP2 from the Golgi complex1.

The first case of dominant NDI was caused by a missense mutation in the AQP2 gene, encoding the mutant AQP2-E258K. In oocytes, this mutant was retained in the Golgi complex region2,3. Since this mutant was also able to form heteroligomers with wild type AQP2 (wt-AQP2), dominant NDI was explained by the impaired further routing of complexed wt-AQP2 to the plasma membrane. However, since in this mutant the recognition sequence for GCK is destroyed, it was further suggested that the inability of GCK to phosphorylate AQP2-E258K at S256 might explain its localization.

Therefore, we investigated the role of S256 phosphorylation in the trafficking of AQP2-E258K. In transfected MDCK cells, the vasopressin-regulated shuttling of AQP2 is as in vivo4. MDCK cells were generated, stably-expressing this mutant. Confocal laser scanning and electron microscopy revealed that, with or without forskolin stimulation, AQP2-E258K was localized in subapical vesicular structures. Also, AQP2-E258K did not colocalize with Golgi marker proteins (Giantin, Mannosidase II, Golgi 58K protein), but showed a profound colocalization with LAMP1, which is a marker protein for late endosomes/lysosomes. Also, immunoprecipitation experiments following radioactive labelling of the cells with 32P revealed that AQP2-E258K was phosphorylated in MDCK cells and that this fosforylation could be increased with forskolin. Besides, phosphorylation of S256 in AQP2-E258K, but not wt-AQP2, was also increased following incubation with the protein kinase C activator PMA, which is in line with the introduction of a protein kinase C phosphorylation site with the E258K mutation.

Taken together, our data reveal that in MDCK cells, (de)phosphorylation of AQP2 is not required for exit from the Golgi complex and therefore excludes a role for GCK in AQP2 routing. In addition, retention of wt-AQP2/AQP2-E258K heterotetramers in late endosomes/lysosomes is likely to explain dominant NDI in this particular family.

  1. Procino et al., J. Am Soc. Nephrol 10, 23A.
  2. Mulders et al., J Clin Invest. 1998 Jul 1;102(1):57-66.
  3. Kamsteeg et al., EMBO J. 1999 May 4;18(9):2394-400.
  4. Deen et al., J Am Soc Nephrol. 1997 Oct;8(10):1493-501.

Aquaporin 2 (AQP2), when positioned in the apical cell membrane of the kidney collecting duct cells, allows water to pass through the cells into another part of the kidney. The AQP2s must move from the cell interior to the cell membrane to do this. A series of molecular events needs to happen for this movement to take place. It had been hypothesized that a phosphate group had to first be attached then detached at a specific amino acid residue of the AQP2 protein, (specifically, S256) for the AQP2 to exit from a portion of the cell interior called the Golgi Complex. Golgi casein kinase (GCK) was the enzyme that was supposed to catalyze the process of phosphate group attachment (a process called phosphorylation).

Savelkoul, et al., studied a mutant AQP2, AQP2-E258K. Researchers speculated that the reason AQP2-E258K was unable to leave the Golgi complex in oocyte studies was because GCK could not recognize the site for phosphorylation in AQP2 and therefore was unable to phosphorylate it at S256. They studied the movement of the mutant AQP2-E258K in laboratory cell cultures. The researchers found the mutant in the cell interior instead of the cell membrane clustered with cell structures called late endosomes and lysosomes (in contrast to a Golgi localization in oocytes). However, this mutant AQP2 had been phosphorylated. This suggests that the phosphorylation of AQP2 is not required for it to be able to exit the Golgi complex. Thus, it appears that GCK does not play a role in the movement of AQP2 to the cell membrane and that the lysosomal localization of AQP2-E258K explains dominant NDI.