Background potassium channels determine membrane potential and input resistance and serve as prominent effectors for modulatory regulation of cellular excitability. TREK-1 is a two-pore domain background K ؉ channel (KCNK2, K2P2.1) that is sensitive to a variety of physicochemical and humoral factors. In this work, we used a recombinant expression system to show that activation of G␣ q -coupled receptors leads to inhibition of TREK-1 channels via protein kinase C (PKC), and we identified a critical phosphorylation site in a key regulatory domain that mediates inhibition of the channel. In HEK 293 cells co-expressing TREK-1 and either the thyrotropin-releasing hormone receptor (TRHR1) or the Orexin receptor (Orx1R), agonist stimulation induced robust channel inhibition that was suppressed by a bisindolylmaleimide PKC inhibitor but not by a protein kinase A blocker ((R p )-cAMP-S). Channel inhibition by agonists or by direct activators of PKC (phorbol dibutyrate) and PKA (forskolin) was disrupted not only by alanine or aspartate mutations at an identified PKA site (Ser-333) in the C terminus, but also at a more proximal regulatory site in the cytoplasmic C terminus (Ser-300); S333A and S300A mutations enhanced basal TREK-1 current, whereas S333D and S300D substitutions mimicked phosphorylation and strongly diminished currents. When studied in combination, TREK-1 current density was enhanced in S300A/S333D but reduced in S300D/S333A mutant channels. Channel mutants were expressed and appropriately targeted to cell membranes. Together, these data support a sequential phosphorylation model in which receptor-induced kinase activation drives modification at Ser-333 that enables subsequent phosphorylation at Ser-300 to inhibit TREK-1 channel activity.

To examine the hypothesis that physical features of the membrane contribute to protein kinase C activation, phosphatidylcholine/phosphatidylserine/diolein (70:25:5) vesicles of defined acyl chain composition were tested for their ability to activate the enzyme. Maximal activation was found to correlate with the mole percent unsaturation in the system. Unsaturation could be provided by either the phosphatidylserine or the phosphatidylcholine component. Vesicles containing 5 mol% diolein but lacking any unsaturation in the phospholipid did not support activity, indicating that acidic head groups alone are not sufficient for activity. The saturated lipid vesicles could be rendered effective but only at very high (25 mol%) concentrations of diolein. The degree of acyl chain unsaturation and the positioning of the double bond had little effect on the activity, suggesting that the effect of the unsaturation is due to some physical property of the lipid rather than to a specific lipid-protein interaction. Addition of cholesterol to both saturated and unsaturated systems indicated that fluidity, as assessed by fluorescence anisotropy, did not correlate with activity. These results suggest that a physical property of the membrane other than fluidity is important for the activation of protein kinase C. A model for protein kinase C activation involving phase separation and/or head group spacing is discussed.

To test the hypothesis that the activation of protein kinase C (PKC) is influenced by lateral heterogeneities of the components of the lipid bilayer, the thermotropic phase behavior of dimyristoylphosphatidylcholine (DMPC)/dimyristoylphosphatidylserine (DMPS)/dioleoylglycerol (DO) vesicles was compared with the activation of PKC by this system. Differential scanning calorimetry (DSC) and Fourier transform infrared (FTIR) spectroscopy were used to monitor the main transition (i.e., the gel-to-fluid phase transition) as a function of mole fraction DO (chi(DO)) in DMPC/DO, DMPS/DO, and [DMPC/DMPS (1:1, mol/mol)]/DO multilamellar vesicles (MLVs). In each case, when chi(DO) < or approximately 0.3, DO significantly broadened the main transition and shifted it to lower temperatures; but when chi(DO) > approximately 0.3, the main transition became highly cooperative, i.e., narrow, again. The coexistence of overlapping narrow and broad transitions was clearly evident in DSC thermograms from chi(DO) approximately 0.1 to chi(DO) approximately 0.3, with the more cooperative transition growing at the expense of the broader one as chi(DO) increased. FTIR spectroscopy, using analogs of DMPC and DMPS with perdeuterated acyl chains, showed that the melting profiles of all three lipid components in [DMPC/DMPS (1:1, mol/mol)]/DO MLVs virtually overlay when chi(DO) = 0.33, suggesting that a new type of phase, with a phospholipid/DO mole ratio near 2:1, is formed in this system. Collectively, the results are consistent with the coexistence of DO-poor and DO-rich domains throughout the compositions chi(DO) approximately 0.1 to chi(DO) approximately 0.3, even at temperatures above the main transition. Comparison of the phase behavior of the binary mixtures with that of the ternary mixtures suggests that DMPS/DO interactions may be more favorable than DMPC/DO interactions in the ternary system, especially in the gel state. PKC activity was measured using [DMPC/DMPS (1:1, mol/mol)]/DO MLVs as the lipid activator. At 35 degrees C (a temperature above the main transition of the lipids), PKC activity increased gradually with increasing chi(DO) from chi(DO) approximately 0.1 to chi(DO) approximately 0.4, and activity remained high at higher DO contents. In contrast, at 2 degrees C (a temperature below the main transition), PKC activity exhibited a maximum between chi(DO) approximately 0.1 and chi(DO) approximately 0.3, and at higher DO contents activity was essentially constant at 20-25% of the activity at the maximum. We infer from these results that the formation of DO-rich domains is related to PKC activation, and when the lipid is in the gel state, the coexistence of DO-poor and DO-rich phases also contributes to PKC activation.

No abstract

Cyclic stretch of cultured urinary tract smooth muscle cells has been used to mimic some of the events that occur with bladder obstruction. The stretch stimulus induces production of nerve growth factor (NGF), which has been implicated in changes in bladder innervation. Stretch-induced NGF production was blocked by actinomycin. Involvement of protein kinase C (PKC) in the stretch-induced NGF production is strongly suggested by the following observations. Phorbol ester activators of PKC mimicked the stretch response as did platelet-derived growth factor (PDGF), which acts, in part, through generation of endogenous diacylglycerols. Both stretch- and PDGF-induced NGF production were blocked by prolonged incubation with phorbol ester to downregulate PKC. Western blot analysis confirmed partial downregulation of the Ca(2+)-dependent PKC-alpha and PKC-beta 1 and near complete downregulation of the Ca(2+)-independent PKC isozymes delta, epsilon, and zeta. The involvement of PKC in transducing a physical stimulus (stretch) into a biochemical response (NGF production) has implications for novel types of therapeutic intervention in ailments such as bladder obstruction.