Autoinhibition and Chloride Sensing in the WNK1 Kinase

Protein kinases control diverse cellular pathways and have are the subject of intensive study regarding how they maintain specificity toward sub- strates. The research presented here focuses on a 230kDa serine/threonine protein kinase known as WNK1 (with no lysine {k}). The protein was first cloned by Melanie Cobb's laboratory, and its isoforms have been associated with a monogenic form of hypertension as well as with breast and prostate cancer. Recent data have also shown that WNK1 is necessary for maintaining spindle polarity in mitosis and plays a role in post-mitotic abscission. The function of WNK1 is most commonly associated with the regulation of CCCs via the activation of the WNK1 substrates OSR/SPAK. Prior investigation of the system has demonstrated that CCCs are activated by increasing con- centrations of extracellular salt and by intracellular phosphorylation from the OSR/SPAK kinases. Due to it ubiquity in mammalian cell types, a question has arisen as to how the pathway responds to changes in osmolarity. Further, because of the involvement of WNK1 in a diverse set of cellular mechanisms, how is WNK1 activity and substrate specificity controlled? An autoinhibitory domain of WNK1 was characterized by the Cobb Lab regarding its ability to inhibit the kinase in cis and in trans. In this study, we find that the solution structure of the autoinhibitory domain retains a conserved RFXV binding site from the PASK/FRAY homology 2 (PF2) domain present in OSR/SPAK. Titration data shows that incubation with a 5-mer and a 20-mer peptide derived from the WNK1 kinase domain displays extensive chemical shift perturbation as assessed by 1H,15N-HSQC. Expression of this autoinhibitory domain in cis with the WNK1 kinase domain followed by size-exclusion chromatography shows substantial confor- mational changes when dialized from high to low salt. A measurement of the activity of the WNK1 kinase domain in the presence of increasing amounts of sodium chloride indicate an IC50 of 130mM. Further biophysical investigation using differential scanning fluorimetry with the kinase domain shows that the domain undergoes substantial increases in domain stabilization as the concen- tration of salt is increased. Continued analysis of this phenomena has pointed toward evidence of anion sensing by the WNK1 kinase domain. Other protein kinases studied in our lab do not exhibit this salt sensitivity. To determine the binding site of chloride in the WNK1 kinase domain, the inactive WNK1 kinase was cocrystallized in the presence of sodium bro- mide. A dataset was collected using the bromine anomalous edge (0.92 ̊A). The anomalous difference fourier map was calculated and a 5.2 σ peak was observed at the N-terminus of the 3.10 helix present in the DLG motif of the activation loop. To corroborate these data, the structure of the inactive kinase domain previously crystallized in sodium chloride was re-refined. A similar binding site corroborated by a 2mFo − DFc peak of 5.5 σ was observed in subunit A near the N-terminus of the 3.10 helix. When the structure was refined in with a chloride ion placed in the observed density, similar hydrogen bonding interactions between the amide backbone and the chloride ion were observed compared to that in the bromide-soaked structure. The presence of this chloride ion appears to favor sequestration of E268 in αC and R348 in the catalytic loop and promotes an inactive kinase structure. Finally, the crystal structure of the activated WNK1 kinase domain was determined under low-salt conditions. The term 'activated' and not active is used to describe this structure of the WNK1 kinase domain because, although it is phosphorylated at S378 and S382 in the activation-loop during expression, the structure adopts an inactive conformation due to the placement E268 in helix C. The structure displays disorder of many key structural elements such as the N-terminus of αC. A key observation is the lack of a 3.10 helix in the N-terminus of the activation loop and the lack of water or any atom that could be chloride near the amide backbone near the chloride binding site. Based upon the literature surrounding the activation of WNK1 and the data presented in this thesis, we predict a three-tiered regulation of WNK1 driven by a) autophosphorylation b) chloride binding and c) autoinhibitory domain occlusion of the nucleotide and/or docking interfaces present in the WNK1 kinase domain. The coupling of the information that we have gath- ered on the autoinhibitory and kinase domains appear to point to an overall mechanism of salt sensing and self-contained signaling control in the WNK1 kinase cascade.