Opposing effects of podocin on the gating of podocyte TRPC6 channels evoked by membrane stretch or diacylglycerol

Abstract

Gain-of-function mutations in the transient receptor potential (TRP) cation channel subfamily C member 6 (TRPC6) gene and mutations in the NPHS2 gene encoding podocin result in nephrotic syndromes. The purpose of this study was to determine the functional significance of biochemical interactions between these proteins. We observed that gating of TRPC6 channels in podocytes is markedly mechanosensitive and can be activated by hyposmotic stretch or indentation of the plasma membrane. Stretch activation of cationic currents was blocked by small interfering RNA knockdown of TRPC6, as well as by SKF-96365 or micromolar La(3+). Stretch activation of podocyte TRPC6 persisted in the presence of inhibitors of phospholipase C (U-73122) and phospholipase A2 (ONO-RS-082). Robust stretch responses also persisted when recording electrodes contained guanosine 5'-O-(2-thiodiphosphate) at concentrations that completely suppressed responses to ANG II. Stretch responses were enhanced by cytochalasin D but were abolished by the peptide GsMTx4, suggesting that forces are transmitted to the channels through the plasma membrane. Podocin and TRPC6 interact at their respective COOH termini. Knockdown of podocin markedly increased stretch-evoked activation of TRPC6 but nearly abolished TRPC6 activation evoked by a diacylglycerol analog. These data suggest that podocin acts as a switch to determine the preferred mode of TRPC6 activation. They also suggest that podocin deficiencies will result in Ca(2+) overload in foot processes, as with gain-of-function mutations in the TRPC6 gene. Finally, they suggest that mechanical activation of TRP family channels and the preferred mode of TRP channel activation may depend on whether members of the stomatin/prohibitin family of hairpin loop proteins are present.

Keywords: TRPC6; glomerular filtration; mechanosensitive; podocyte; slit diaphragm.

Fig. 1.

Hyposmotic stretch activates currents through transient receptor potential (TRP) cation channel subfamily C member 6 (TRPC6) channels in immortalized mouse podocyte cell lines. A: representative example of currents in a single cell evoked by ramp voltage commands (−80 to +80 mV over 2.5 s from a holding potential of −40 mV) in normal external saline (100%, left), in hyposmotic saline (70%, middle), and after return to normal saline (100%, right). Note reversible increase in outwardly rectifying cationic currents. In these and subsequent experiments, except as noted (e.g., Fig. 3), extracellular Ca²⁺ concentration = 5.4 mM, which markedly increases stability of whole cell recordings during application of large voltage ramps but also results in reduced amplitude of inward currents. B: hyposmotic stretch responses in a control cell and in a different cell in the presence of 10 μM SKF-96365. Currents recorded in normal and hyposmotic solutions are superimposed. C: hyposmotic stretch responses in a control cell and in a different cell in the presence of 50 μM La³⁺. D: summary of results from several cells. Values (means ± SE) represent difference current (70% saline − 100% saline) at +80 mV in control (Con) conditions and in the presence of 10 μM SKF-96365 (SKF) or 50 μM La³⁺; n = 10 cells in each group. *P < 0.05 vs. control (Con). E: immunoblot showing markedly reduced expression of TRPC6 in podocytes transiently transfected with TRPC6 small interfering RNA (siRNA) compared with cells treated with control siRNA. F: summary of effects of TRPC6 siRNA knockdown on stretch-evoked cationic currents compared with control siRNA. Values are means ± SE; n = 10 cells in each group. *P < 0.05 vs. Con siRNA.

Fig. 2.

Hyposmotic stretch evokes cationic currents in primary rat podocytes in glomerular explants. Whole cell recordings are from podocytes on the external margin of the preparation, which could be identified by larger primary processes emanating from the cell body. A: representative recording from a single cell showing currents evoked by ramp voltage commands in normal saline (100%), hyposmotic saline (70%), and hyposmotic saline containing 50 μM La³⁺. B: transient transfection of glomerular explants with the same TRPC6 siRNA used in Fig. 1 caused nearly complete inhibition of the hyposmotic stretch effect on cationic currents in these cells. Stretch responses were robust in podocytes in glomeruli transfected with the control siRNA. C: summary of results from the experiment in B. Values are means ± SE; n = 7 cells in each group. Note that stretch responses in glomerular explants are qualitatively similar to those in podocyte cell lines, but currents are substantially larger. *P < 0.05 vs. control siRNA.

Fig. 3.

Currents evoked by pressure pulses in mouse podocyte cell lines. A: schematic diagram showing first experimental design. Normal saline containing 60% sucrose to increase viscosity is applied to the cell body by pressophoresis (5 psi = 34 kPa) from an adjacent micropipette. This stimulus produces a distinct inward current in normal saline that was abolished after application of external saline containing 50 μM La³⁺ (B) or 10 μM SKF-96365 (C). Holding potential was −60 mV. Similar responses were seen in 8 other cells. D: mechanical stimuli could also be delivered by touching the podocyte with a blunt fire-polished glass probe controlled by a micromanipulator. E: inward currents evoked at −60 mV by the procedure described in D. In B, C, and E, extracellular Ca²⁺ concentration was reduced to 0.8 mM to facilitate detection of inward currents.

Fig. 4.

Flow can activate cationic currents in rat podocytes in glomerular explants. A: schematic diagram showing experimental design to examine the effect of bath flow on podocytes. Gravity-fed bath perfusion could be started or stopped by use of valves while current was monitored at a holding potential (V_M) of −60 mV. B: cationic currents repeatedly evoked by flow of bath saline in podocytes on the external margin of an isolated glomerulus. Apparent desensitization of the response in B was not observed consistently. C: recording showing complete block of flow-evoked current after introduction of 50 μM La³⁺ into the bath.

Fig. 5.

Hyposmotic stretch responses in mouse podocyte cell lines persist after inhibition of phospholipases. A: representative responses to hyposmotic stretch in control cells, cells treated with the pan-phospholipase A₂ inhibitor ONO-RS-082 (ONO), and cells treated with the pan-phospholipase C inhibitor U-73122. B: summary of several repetitions of experiments described in A. Values are means ± SE; n = 8 cells in each group. ns, Not significant.

Fig. 6.

Responses to stretch after inhibition of G protein signaling in podocytes in isolated glomerulus explant preparation. A: whole cell recording made using an electrode containing 50 μM guanosine 5′-O-(2-thiodiphosphate) (GDP-βS), a GDP analog that cannot be hydrolyzed or phosphorylated. Recordings from a single cell show reversible response to hyposmotic stretch but no increase in TRPC6 current following subsequent exposure to 10 nM ANG II. B: whole cell recording showing increase in cationic current evoked by 1 nM ANG II using normal electrode solution. C: summary showing mean increases in current at +80 mV evoked by hyposmotic stretch using normal electrode solutions and electrode solutions containing 50 μM GDP-βS. Values are means ± SE; n = 5 cells in each group.

Fig. 7.

Evidence that forces are transmitted to TRPC6 channels through the lipid bilayer in mouse podocyte cell lines. A: phalloidin staining showing that cytochalasin D (Cyto D) treatment for 20 min caused major disruption of F-actin filaments in podocytes. B: representative examples of stretch-evoked cationic currents in a control cell and in a different podocyte that had been treated with cytochalasin D for 20 min. C: summary of several repetitions of experiment shown in B. Note marked increase in stretch-evoked currents after cytochalasin D. Values are means ± SE; n = 8 cells in each group. D: recording from a single cell in control saline (100%), during hyposmotic stretch (70%), and during complete blockade of the hyposmotic response within minutes of addition of 2.5 μM GsMTx4 (70% + 2.5 μM GsMTx4). GsMTx4 was added to the cell after full development of the stretch response. E: summary of results of experiment in which stretch responses were recorded in control cells and in different cells pretreated with 2.5 μM GsMTx4. Note essentially complete inhibition of hyposmotic stretch response. Values are means ± SE; n = 6 cells in each group. *P < 0.05 vs. Con.

Fig. 8.

Podocin interacts with TRPC6 channels. A: coimmunoprecipitation (IP) of endogenously expressed podocin and TRPC6 from mouse podocyte cell lines. Immunoblots (IB) show podocin that was immunoprecipitated with goat anti-TRPC6 (left) and TRPC6 (detected with rabbit anti-TRPC6) that was immunoprecipitated using anti-podocin (right). Note that the interaction is reciprocal. These proteins were not immunoprecipitated by IgG. B: GST pull-down assay in which heterologously expressed hemagglutinin (HA)-tagged TRPC6 fusion proteins were pulled out of a HEK-293 cell lysate using a GST-podocin fusion protein. COOH terminus of podocin (GST-Pod-C) was able to pull the COOH terminus of TRPC6 (HA-TRPC6-C) out of HEK-293 cell lysates, but the NH₂ terminus of podocin (GST-Pod-N) and GST alone were ineffective (top). Lane on left shows that HA-TRPC6-C was present in the lysates. By contrast, neither of the GST-fusion proteins nor GST alone could pull HA-TRPC6-N out of HEK-293 cell lysates (bottom). Therefore, the COOH terminus of TRPC6 is able to interact with the COOH terminus of podocin.

Fig. 9.

Podocin knockdown by transient siRNA procedures affects mechanical activation of cationic currents in podocytes. A: immortalized mouse podocyte cells were transfected with a control siRNA and a siRNA directed against podocin, resulting in marked decrease in expression of podocin but no effect on total expression of TRPC6. B: treatment described in A did not produce major effects on organization of actin microfilaments, as assessed by phallodin staining. C: representative examples of hyposmotic stretch responses in control and podocin-knockdown cells. Note very large amplitude of cationic currents after podocin knockdown. D: summary of several repetitions of experiment described in C showing 4- to 5-fold increase in stretch-evoked currents. Values are means ± SE; n = 12 cells in each group. *P < 0.05 vs. Con siRNA.

Fig. 10.

Large stretch-evoked currents in podocin-knockdown cells have properties of TRPC6. A: podocytes transfected with podocin siRNA, as described in Fig. 9 legend, show very large stretch-evoked cationic currents that are completely inhibited after superfusion of 50 μM La³⁺ while maintaining the stretch. B: large stretch-evoked currents in podocin siRNA-knockdown cells are also blocked by 10 μM SKF-96365. C: experiments in podocytes stably expressing short hairpin RNA (shRNA) directed against podocin. Stretch-activated currents are very large, comparable to transient podocin knockdown, and much larger than normal podocytes. These currents are not affected by subsequent transient transfection with control siRNA (left). By contrast, transient transfection with siRNA directed against TRPC6 causes nearly complete loss of stretch-evoked cationic currents (right). D: summary of results showing that stable podocin shRNA knockdown causes a large increase in stretch-evoked cationic currents compared with cells stably expressing control shRNA. In addition, a subsequent transient siRNA knockdown of TRPC6 causes a nearly complete loss of stretch-activated current in the podocin-knockdown cells. Values are means ± SE; n = 8 cells in each group. *P < 0.05 vs. control shRNA.

Fig. 11.

Effects of podocin knockdown on responses to the membrane-permeable diacylglycerol (DAG) analog 1-oleoyl-2-acetyl-sn-glycerol (OAG). Traces show currents in single cells before and after bath application of 100 μM OAG. A: in cells transiently transfected with control siRNA, OAG evokes a marked increase in cationic currents. B: responses to OAG are nearly eliminated after transient podocin knockdown using siRNA (see Figs. 9 and 10). Dashed lines allow comparison of amplitudes of the responses evoked by OAG in the 2 experiments. C: summary of several repetitions of this experiment showing nearly complete block of OAG responses in podocin-knockdown cells. Values are means ± SE; n = 8 cells in each group. *P < 0.05 vs. Con siRNA.