Coupling of epithelial Na+ and Cl- channels by direct and indirect activation by serine proteases

Abstract

The mammalian collecting duct (CD) is continuously exposed to urinary proteases. The CD expresses an epithelial Na(+) channel (ENaC) that is activated after cleavage by serine proteases. ENaC also exists at the plasma membrane in the uncleaved form, rendering activation by extracellular proteases an important mechanism for regulating Na(+) transport. Many exogenous and a small number of endogenous extracellular serine proteases have been shown to activate the channel. Recently, kallikrein 1 (KLK1) was shown to increase γENaC cleavage in the native CD indicating a possible direct role of this endogenous protease in Na(+) homeostasis. To explore this process, we examined the coordinated effect of this protease on Na(+) and Cl(-) transport in a polarized renal epithelial cell line (Madin-Darby canine kidney). We also examined the role of native urinary proteases in this process. Short-circuit current (I(sc)) was used to measure transport of these ions. The I(sc) exhibited an ENaC-dependent Na(+) component that was amiloride blockable and a cystic fibrosis transmembrane conductance regulator (CFTR)-dependent Cl(-) component that was blocked by inhibitor 172. Apical application of trypsin, an exogenous S1 serine protease, activated I(ENaC) but was without effects on I(CFTR). Subtilisin an exogenous S8 protease that mimics endogenous furin-type proteases activated both currents. A similar activation was also observed with KLK1 and native rat urinary proteases. Activation with urinary proteases occurred within minutes and at protease concentrations similar to those in the CD indicating physiological significance of this process. ENaC activation was irreversible and mediated by enhanced cleavage of γENaC. The activation of CFTR was indirect and likely dependent on activation of an endogenous apical membrane protease receptor. Collectively, these data demonstrate coordinated stimulation of separate Na(+) and Cl(-) transport pathways in renal epithelia by extracellular luminal proteases. They also indicate that baseline urinary proteolytic activity is sufficient to modify Na(+) and Cl(-) transport in these epithelia.

Fig. 1.

Representative time-control experiment demonstrating stability and reproducibility of the current and resistance (I_sc and R_T). A: both I_sc and R_T were stable for over 45 min allowing us to use each tissue as its control where feasible. Note the reversible increase in R_T following addition of 10 μM amiloride demonstrating the electrical “tightness” of the tissue. B: summary of the amiloride (10 μM)-sensitive I_sc as a function of time. A, w/o, and NS indicate amiloride, washout from amiloride, and not significant as determined statistically; n = 6.

Fig. 2.

Trypsin stimulates the amiloride sensitive current. A: representative example demonstrating acute stimulation of amiloride-sensitive short-circuit current (I_ENaC) by 10 μg/ml trypsin. Effects of apical trypsin on the I_ENaC are summarized in B. Ratio indicates the value of the amiloride sensitive current after trypsin to that before trypsin. NS indicates not statistically significant; A and w/o indicate apical amiloride and washout from amiloride; n = 6.

Fig. 3.

Forskolin stimulates a cystic fibrosis transmembrane conductance regulator (CFTR) current that is trypsin insensitive. A: representative example demonstrating acute stimulation of an amiloride-insensitive current by 5 μM forskolin (I_CFTR). Apical trypsin was without effect on this current. B: representative example demonstrating the absence of an effect of 10 μM CFTR inhibitor 172 (inh₁₇₂) on the trypsin response. C: summary of the effects of 10 μg/ml trypsin on the current observed in the presence of amiloride and forskolin. No difference were observed with and without inh₁₇₂ indicating the absence of changes to I_CFTR with trypsin. NS indicates not statistically significant; A indicates apical amiloride; Tryp + inh₁₇₂ indicates apical trypsin and CFTR inhibitor 172; n = 6 in each group.

Fig. 4.

Subtilisin irreversibly stimulates ENaC. A: S8 protease subtilisin stimulated the amiloride-sensitive I_sc (I_ENaC). B: this effect was irreversible consistent with channel proteolysis. C: summary of the effects of 10 μg/ml subtilisin indicating a 4.3-fold stimulation of ENaC by this protease. Ratio indicates the value of the amiloride sensitive current after subtilisin to that before subtilisin. A and w/o indicate apical amiloride and washout from amiloride in A and washout of amiloride and subtilisin in B; n = 6 in each group.

Fig. 5.

Subtilisin stimulated I_CFTR. A: representative example demonstrating acute stimulation of I_CFTR by 10 μM subtilisin. Similar to stimulation with forskolin, this effect was biphasic with a large initial peak followed by a smaller sustained stimulation. B: stimulation of I_sc was prevented in the presence of 10 μM CFTR inh₁₇₂ demonstrating that both the initial peak and sustained responses were due to CFTR. C: summary of the changes of the amiloride insensitive current indicating inhibition of the peak stimulation by inh₁₇₂. Ratio indicates the value of the amiloride-insensitive current before and after subtilisin demonstrating stimulation of both the peak and sustained responses by subtilisin. A indicates apical amiloride; subt + inh₁₇₂ indicates apical subtilisin and CFTR inhibitor 172; n = 6 in each group.

Fig. 6.

Stimulation of ENaC by urinary proteases. A: representative example demonstrating stimulation of ENaC mediated I_sc by rat urinary proteases at a final concentration of 10% of that found in the urine (see materials and methods). B: this effect was absent in heat inactivated urine (inactivated at 85°C for 15 min). C: summary data indicate a 2.6-fold increase of I_ENaC by rat urinary proteases and a marked attenuation of the response in heat-inactivated urine. A and w/o indicate apical amiloride and washout from amiloride; n = 6 in each group.

Fig. 7.

Stimulation of CFTR by rat urinary proteases. A: representative example demonstrating stimulation of CFTR mediated I_sc by rat urinary proteases at the same concentration as that which stimulated I_ENaC. B: this effect was absent in the presence of CFTR inh₁₇₂ demonstrating the nature of the current. C: summary of the changes of current indicating significant differences in the presence of inh₁₇₂. Urinary proteases stimulated both the peak and sustained phases of I_CFTR. A indicates apical amiloride; urine + inh₁₇₂ indicates apical rat urinary proteases and CFTR inhibitor 172; n = 6 in each group.

Fig. 8.

Kallikrein 1 (KLK1) stimulates ENaC. A: time course of the effect of 8 μg/ml KLK1. B: summary of the increase of I_ENaC with KLK1 which averaged 2.6-fold. No effect of subtilisin was observed following KLK1 treatment indicating overlapping mechanism of channel stimulation by these proteases. Ratio indicates the values of I_ENaC in KLK1 to those in control or I_ENaC in subtilisin and KLK1 to those in KLK1. A and w/o indicate apical amiloride and washout from amiloride; n = 6 in each group.

Fig. 9.

Stimulation of CFTR by KLK1. A: representative example demonstrating biphasic stimulation of current by KLK1. B: these changes were absent in the presence of CFTR inh₁₇₂. C: summary of the changes of current indicating a block of this stimulation by inh₁₇₂, and the CFTR nature of this current. KLK1 caused a peak and sustained stimulation of 7- and 2-fold, respectively. A indicates apical amiloride; KLK1 + inh₁₇₂ indicates apical KLK1 and CFTR inhibitor 172; n = 6 in each group.

Fig. 10.

Enhanced γ ENaC cleavage by activating proteases. A: representative blot demonstrating the presence of cleaved and uncleaved γ at both the membrane (M) and total (T) fractions. The majority of this subunit was already in the cleaved form as expected given that intracellular furin mediated cleavage was intact and unblocked. To eliminate variability, cleaved γ levels at the membrane were normalized to total γ levels in the cells and more specifically to total uncleaved γ levels as this represents the maximal pool reactive to proteolysis and activation. All channel activating proteases were accompanied by increased cleaved γ. Top and bottom boxes denote uncleaved and cleaved forms. Data represent 4 blots. B: summary of the changes observed to γ ENaC proteolysis; n = 3–4 in each group except for trypsin where n = 2.

Fig. 11.

KLK1 cleaves γ ENaC sequences in vitro. A: representative example indicating much higher cleavage activity toward the 1st RKRR site than the 2nd RKRK site. B: summary of the reaction rates (calculated as the initial slope shown by the line in A) of both peptides in response to KLK1. Final KLK1 levels in both reactions were 1 μg/ml. Rat urine indicates activity toward RKRR at its final diluted concentration as used in Figs. 6 and 7; a.u., arbitrary units; n = 4.

Fig. 12.

PAR1 and 2 independent activation of CFTR by subtilisin. A and B: activation of PAR2 by a peptide with sequences from the human tethered ligand did not modify the subsequent response to subtilisin. A indicates apical amiloride; A+forskolin+hPAR2 indicates forskolin and apical amiloride and human PAR2 tethered ligand; +subtilisin indicates addition of apical subtilisin to A+forskolin+hPAR2. C and D: neomycin at 1 mM blocked the effects of subtilisin on I_CFTR. A indicates apical amiloride; A+forskolin indicates apical amiloride and forskolin; +N indicates addition of neomycin to A+forskolin; +subtilisin indicates addition of apical subtilisin to +N. E and F: summary data indicating that prior stimulation of PAR1 or PAR2 did not affect the response to subtilisin, while treatment with neomycin markedly attenuated this response; n = 4–6 in each group.