| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Physiology and Biophysics, State University of New York at Buffalo, Buffalo, New York, USA
1Correspondence: Department of Physiology and Biophysics SUNY at Buffalo 124 Sherman Hall, 3435 Main St. Buffalo, NY 14214, USA. E-mail: awayda{at}buffalo.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
. We propose that AKAP15 regulates ENaC via a novel PKA-independent pathway.—Bengrine, A., Li, J., Awayda, M. S. The A-kinase anchoring protein 15 regulates feedback inhibition of the epithelial Na+ channel.
Key Words: PKA • ENaC activity • intracellular Na+ • AKAP15
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
. This apical membrane-located channel is expressed in many epithelia such as kidney collecting duct, alveolar epithelium, distal colon, salivary duct, and sweat glands (see ref. 2
). The homologous subunits , ß, and 
constitute a highly regulated, low-conductance (5 pS), high Na+ selectivity channel (3
4
5)
. It is well accepted that this channel is rate-limiting to transepithelial Na+ reabsorption.
Channel activity is under tight control by many hormonal and humoral factors, including PKA and PKC. These kinases are implicated in a variety of signaling cascades that range from PKA mediating the actions of antidiuretic hormone (alcohol dehydrogenase, or ADH) to PKC mediating the effects of feedback inhibition by intracellular Na+ (6)
. However, although links between these physiological processes and the actions of such kinases have been well established for the native Na+ channel, the mechanisms of such regulation and the effect on the cloned ENaC are only partially understood. For example, the effects of accessory proteins on the vesicular trafficking of ENaC are emerging (7
8
9)
. However, the roles of anchoring proteins in the response to PKA or PKC and the potential cross-talk between these kinases remain undetermined. The roles of other signaling cascades in channel regulation by feedback inhibition are also poorly understood.
The second messenger regulation of the cloned ENaC is not well understood. We previously reported that activation of PKC in oocytes inhibits ENaC via specific effects on the channel as well as nonspecific membrane internalization (10)
. The PKC isoforms implicated in this response were not examined. However, we subsequently established a specific function for PKC in the process of feedback regulation of ENaC (11)
. We also proposed that this isoform may play a role in feedback inhibition by intracellular Na+. Similarly, regulation of the cloned channel by PKA is controversial, with evidence for (12
, 13)
and against (14
, 15)
regulation by this kinase. These issues remain unresolved and require better understanding, as their outcome can elucidate the roles of accessory proteins in interacting with the cloned channel to convey the complex regulation observed in native tissues.
It is now emerging that the interaction between protein kinases and membrane-bound integral proteins (e.g., ion channels) requires recruitment of anchoring proteins that stabilize the kinases at the plasma membrane. Indeed, RACKs are known to participate in the membrane effect of PKC whereas AKAPs participate in the membrane effects of PKA. Moreover, some degree of cross-reactivity is now emerging whereby anchoring proteins previously thought to exclusively interact with one type of kinase are now shown to interact with others. Examples of this function include the roles of AKAP12 in PKC-mediated inactivation of the KCNQ potassium channel (16)
and AKAP79 in the PKC-mediated pepsinogen secretion in gastric chief cells (17)
. In some instances, these anchoring proteins also bind to and interact directly with integral membrane-bound ion channels, as is the case for AKAP15 and the L-type Ca2+ channel from skeletal muscle (18
19
20)
.
In epithelial tissue a variety of AKAPs are known to affect membrane-bound channel activity, some in a PKA-dependent manner. For example, stimulation of the water channel aquaporin-2 in renal principal cells involves activation of PKA and its tethering to a subcellular compartment by AKAP15 (21)
(also known as AKAP18). Ali et al. (22)
also demonstrated that AKAP79 augmented the PKA-mediated stimulation of ROMK1, a renal apical membrane potassium channel.
To test whether AKAPs are involved in regulating ENaC, we individually expressed human ENaC with AKAP79, AKAP-KL, and AKAP15. We report that coexpression with AKAP-KL and AKAP79 were without effect on ENaC. On the other hand, AKAP15 reduced ENaC activity independent of PKA activation. Using a cytoplasmic AKAP15 construct (AKAP15m), we demonstrated that membrane targeting of AKAP15 was necessary for its inhibitory effect on ENaC. We also demonstrated an additional interaction between cytoplasmic AKAP and PKC, providing an explanation for the role of [Na+] in the effect of AKAP15m on ENaC. Altogether, our data uncover a novel mode of channel regulation that also plays a role in feedback regulation by high Na+ concentrations.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
, ß, and ENaC constructs (accession nos. NM_001038, NM_000336, and NM_001039) as described previously (23)
. These are referred to as ENaC. The DNA for AKAP15 (accession nos. NM_004842 and NM_138633) and AKAP79 (accession #NM_004857) were a gift from J. D. Scott (Vollum Institute, Portland, OR, USA). AKAP-KL (accession #BC003735) and PKC (accession #NM_002737) were purchased from (American Type Culture Collection, Manassas, VA, USA). AKAP15 and AKAP-KL were subcloned into PGEM-HE, as described previously for , ß, and ENaC (23)
. The AKAP15 myristoylation site mutant (referred to as AKAP15m) was generated with the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA) using the following primers: 15 mF 5'CAGTGCCATGGCCCAGCTTTCCTCCTTTCCTTTCTC3' and 15 mR 5'GAGAAAGGAAAGGAGGAAAGCTGGGCCATGGCACTG3'.
Xenopus oocytes
Experiments on animals were carried out in accordance with protocols approved by the University at Buffalo Animal Care and Use Committee. Oocytes were prepared as described previously (10)
. Briefly, surgically removed oocytes were defolliculated in Ca2+-free buffer containing 1 mg/ml collagenase (type 1A, Sigma Chem. Co., St. Louis MO, USA). After an overnight recovery, oocytes were injected with the appropriate cRNA. Electrophysiological recordings were carried out 1 to 4 days after injection. Oocytes used for the binding experiments utilized expression of an hemagglutinin (HA)-tagged ENaC construct (extracellular loop tag). ENaC cRNA was injected at 1–2 ng for each subunit; PKC and the various AKAP cRNAs were injected at 5–20 ng.
Oocyte incubation and recording solutions have been described (10)
. Briefly, the normal recording solution (ND94) was composed of 94 mM NaCl, 2 mM KCl, 1.8 mMCaCl2, 1 mM MgCl2, and 5 mM HEPES, pH 7.4. This is referred to as high Na+ solution. For the low sodium solution, 89 mM of NaCl from the previous solution was replaced by 89 mM of NMDG-Cl. Prolonged incubation in the low Na+ solution results in a nearly 10-fold change of [Na+]i and affects feedback inhibition of ENaC (24)
. Amiloride was obtained from Merck Pharmaceuticals (Rahway, NJ, USA) and was used at 10 µM to inhibit ENaC currents. Forskolin and IBMX were used at 5 µM and 0.5 mM, respectively, to activate PKA, whereas 100 µM of the AKAP peptide inhibitor Ht31 was used to inhibit the interaction of AKAP and PKA. Experiments also used injection of purified catalytic subunit of PKA as well as the specific PKA-inhibiting protein PKI (EMD Biosciences, San Diego, CA, USA). Additional purified catalytic subunit of PKA was obtained from Promega (Milwaukee, WI, USA). Oocytes in these experiments were impaled by a third pipette, which was used for intracellular delivery of reagents. This setup allows the recording of control and experimental channel activities in the same group of oocytes, as already described (24)
, and monitoring of potential effects on channel activity within fractions of a minute (see Results).
Oocytes fractionation, immunoprecipitation, and Western blot
Injected oocytes were washed three times in ND94 solution at room temperature, then transferred to 5 µl per oocyte of homogenization solution (80 mM sucrose, 5 mM MgCl2, 5 mM NaH2PO4, 1 mM Tris, pH 7.4) supplemented with a protease inhibitor cocktail (Sigma-Aldrich Co., St. Louis, MO, USA). Oocytes were broken by trituration with 25 gauge and 27 gauge needles. Yolk and nuclei were pelleted by centrifugation at low speed (100 g) for 5 min. Supernatants were spun at high speed (14000 g) for 20 min at 4°C, then separated into soluble and insoluble components. The insoluble component was dissolved in SDS electrophoresis sample buffer or IP buffer (80 mM sucrose, 5 mM MgCl2, 5 mM NaH2PO4, 0.5% sodium deoxycholate, 1% Nonidet P-40, 1 mM Tris, pH 7.4) and spun for 5 min at high speed to obtain the membrane fraction. The soluble component was further extracted with 5 µl per oocyte of Freon (trichlorotrifluoroethane) to remove the remaining yolk proteins. The aqueous upper phase (cytoplasmic component) was mixed with 6x SDS electrophoresis sample buffer or 10x IP buffer. Proteins in either fraction were then used for Western blot or for immunoprecipitation, followed by Western blot.
Immunoprecipitation was carried out using a myc-tagged AKAP15 construct. To IP AKAP15, 5 µg of a horseradish peroxidase (HRP) -conjugated anti-myc antibody (Roche, Indianapolis IN, USA) was added to 250 µl of the soluble fraction or 100 µl of the membrane fraction (equivalent of 50 oocytes) and incubated at 4°C overnight with rotation (50 rpm). The next day, 100 µl of ImmunoPure immobilized protein A slurry (Pierce, Rockford, IL, USA) was used to precipitate the antigen-antibody complex. After 2 h incubation with rotation (50 rpm) at room temperature, the beads were washed six times with IP buffer. The complex was then eluted twice using 25 µl of SDS loading buffer for 5 min at 95°C.
Proteins were separated by SDS-PAGE and transferred to nitrocellulose membranes. Blots were blocked in Tris-buffered saline containing 0.05% Tween 20 (TBS-T) and 5% of nonfat dry milk at room temperature. For AKAP79, AKAP-KL, and PKC, blots were incubated in the same blocking solution containing an anti- AKAP79 antibody (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA) at 1/100 dilution, an anti-AKAP-KL antibody (BD Biosciences, San Jose, CA, USA) at dilution 1/1000, or an anti-PKC antibody (BD Biosciences) at dilution 1/1000. For AKAP15, the primary antibody (a gift from W. A. Catterall, Washington University, St. Louis, MO, USA; ref. 25
was used at 1/500 dilution in TBS-T containing 2% of nonfat dry milk at room temperature for 1 h. After four washes of 5 min each, AKAP79 and AKAP-KL blots were incubated in TBST containing an anti-rabbit secondary antibody coupled to HRP at 1:20000 dilution as recommended by the manufacturer (Pierce). PKC blots were incubated in TBS-T containing an anti-mouse antibody (Pierce) coupled to HRP at 1:20000 dilution. AKAP15 blots were incubated with HRP-conjugated protein A (Pierce) at 1:20000 dilution in TBS-T. After 1 h incubation at room temperature, blots were washed extensively with TBS-T for 20 min. The bound antibodies were visualized by enhanced chemiluminescence (Super Signal ECL, Pierce).
ENaC surface expression
Cell surface expression of HA-tagged ENaC was measured in groups of 25 oocytes (each group was regarded as a single experiment and was obtained from oocytes from separate frogs) as described previously (26)
. In this study, oocytes were washed three times with ND94, then transferred to tubes containing an anti-HA antibody coupled to HRP (Roche) at a final concentration of 0.357 µg/ml in ND94, supplemented with 1% BSA. After 90 min incubation at room temperature with rotation, oocytes were washed six times with ice-cold ND-94. Assaying the HRP content of the last wash solution allowed us to determine the effectiveness of the washes. Bound antibodies were detected indirectly by measuring HRP activity using Pierce Turbo-TMB HRP substrate according to the manufacturer’s instructions. Experiments were always conducted using ENaC-injected and noninjected oocytes to assess nonspecific binding. The ability of the assay to detect changes of surface binding was tested in oocytes injected with varying levels of ENaC cRNA (26)
.
Impedance analysis
Dual electrode voltage clamp was carried out using a TEV-200 oocyte clamp (Dagan Corp., Minneapolis MN, USA). Electrodes were constructed of borosilicate glass and used with a resistance of 1–5 mOhms. Impedance was recorded at 10 s intervals throughout the experiments, as described (10)
. The measurement consisted of five sequential discrete sine wave signals ranging from 55 to 390 Hz. Impedance analysis allows the determination of membrane capacitance (Cm) and conductance (gm).
Patch clamp analysis
Conditions were as described previously (27)
. Briefly, oocytes were shrunk in hypertonic solution, followed by manual removal of the vitelline membrane. Experiments were carried out in the cell-attached patch clamp configuration, with electrode resistances in the range of 10 MOhms. Electrodes were filled with modified ND94 lacking K+. Only seals with resistances of >40 GOhms were used. These were stable over prolonged periods (30 min or more), with no visible rundown. Data acquisition and analyses used Axon Instruments’ (Carson City, CA, USA) suite of DigiData 1322A, Axopatch 200B, and pClamp 9. Recordings 5–10 min in length were used to analyze NPo or channel activity. Shorter recordings (2–5 min) were used to determine the current/voltage relationship.
Unless otherwise noted, data are reported as mean ± SE; significance was determined using Student’s t test (paired or unpaired as applicable at the 0.01 level).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
, whereas AKAP79 has been shown to affect the activity of another renal apical membrane channel, ROMK1 (22
; see also ref. 29
for a review).
As shown in Fig. 1
, coexpression of ENaC and AKAP79 was without effect on either the conductance (Fig. 1A
) or capacitance (Fig. 1B
). Similarly, AKAP-KL coexpression was without effect on ENaC activity or capacitance (Fig. 1C, D
). No effect of either AKAP was observed irrespective of whether oocytes were equilibrated in high or low Na+ conditions, a maneuver that attenuates ENaC baseline expression by 3- to 4-fold (see Fig. 1
). This lack of an effect cannot be attributed to the absence of these AKAPs, as protein expression was verified by Western blot with AKAP79 and AKAP-KL-specific antibodies (see inset to Fig. 1A, C
).
|
The absence of an effect of these AKAPs on ENaC activity was not modified by stimulation of endogenous PKA with forskolin and IBMX (data not shown). Indeed, ENaC currents remained insensitive to PKA stimulation, similar to those observed in oocytes expressing ENaC alone (see below and ref. 14
). These results indicate that AKAP79 and AKAP-KL do not modify baseline ENaC activity nor do they directly confer PKA-mediated regulation to the cloned channel.
AKAP15 reduces ENaC conductance
Experiments then focused on the potential roles of AKAP15. This protein, also known as AKAP18, is expressed in native kidney cells and, more specifically, in collecting duct principal epithelial cells, where it colocalizes in membrane vesicles with the aquaporin2 water channel (30)
. This AKAP is also recruited to these vesicles after activation of PKA by arginine vasopressin (30)
. AKAP15 is also endogenously expressed in a cultured immortalized renal epithelial cell line: MDCK (31)
. This AKAP exhibits at least two splice isoforms— and ß—that are specifically targeted to the basolateral and apical membranes of epithelial cells (31)
. Our experiments with AKAP15 used both the ß and isoforms of this protein. Expression with either isoform produced similar results in the oocyte system, and the data were combined. Methods of whole-cell impedance were used to examine the effects of this AKAP on ENaC activity or the amiloride-sensitive, whole-cell conductance, as well as on membrane capacitance—a parameter related to membrane area. In the experiments shown in Fig. 2
, coexpression of ENaC with AKAP15 led to a marked decrease of the amiloride-sensitive ENaC conductance (gNa) by 80 and 91% in high and low Na+ conditions, respectively. In either condition, these effects could not be explained by changes of membrane area as both AKAP-expressing groups exhibited a higher (rather than lower) membrane capacitance (Fig. 2B
). These effects on Cm rule out a nonspecific decrease of membrane area as an explanation for the reduction in ENaC conductance with AKAP15 expression. This conclusion is strengthened by results of the binding studies and by the patch clamp data shown below indicating a reduced ENaC activity in AKAP15-coexpressing oocytes.
|
To assess the expression and distribution of AKAP15, membrane and cytoplasmic protein fractions were probed with an AKAP15-specific antibody. As shown in Fig. 2C
, AKAP15 exhibited a nearly equal distribution between the membrane and the soluble cytoplasmic fractions. As the majority of ENaC expressed in oocytes is also localized to the membrane fraction (data not shown), these results indicate partial overlap between the ENaC and AKAP15 pools. The presence of AKAP15 in the membrane is consistent with its known distribution in MDCK cells (31)
. The mechanism of ENaCs inhibition by this protein is explored in the following sections.
AKAP15 inhibition of ENaC is not PKA mediated
A potential explanation for the effect of AKAP15 on ENaC is that this anchoring protein alters PKA-mediated regulation of ENaC. We have demonstrated that rat ENaC is insensitive to stimulation with forskolin and IBMX (14)
. This cocktail is known to activate CFTR (cystic fibrosis transmembrane conductance regulator) expressed in oocytes in a PKA-dependent manner (32)
, and presumably its lack of effect on ENaC indicates an absence of PKA sensitivity. However, it remains to be determined whether use of this cocktail presents an adequate means of altering PKA activity to specifically test its effects on ENaC (as recently reported for cpt-cAMP; refs. 12
, 15
) and whether human ENaC behaves in a manner similar to that of the rat.
Before determining whether the effects of AKAP15 on ENaC involve changes to the channel’s regulation by PKA, we verified that ENaC expressed in oocytes is indeed PKA insensitive. Experiments were carried out with direct injection of purified catalytic subunit of PKA. As shown in Fig. 3
A, injection of the catalytic subunit of this kinase was without effect on the gNa in either high or low Na+ conditions. Injection of PKI, a specific PKA-inhibiting protein, was also without effect on ENaC (Fig. 3B
). Thus, these results demonstrate that the lack of a PKA response is not due to a maximally PKA-activated ENaC under baseline conditions.
|
The ability of either PKA or PKI to alter the activity of CFTR, an unrelated channel known to respond to PKA by stimulation by forskolin and IBMX, is shown in Fig. 3C, D
. Thus, our results demonstrate that the lack of an effect of PKA on ENaC in oocytes is not a limitation when stimulating this kinase with forskolin and IBMX. Our data also indicate that human and rat ENaC behave in a similar manner with respect to PKA in this system. These conclusions are consistent with those reported earlier using forskolin and IBMX as an indirect means of stimulating oocyte endogenous PKA (14
, 15
, 33)
.
Given the results of the above experiments, we tested whether AKAP15 affects ENaC by conferring PKA-mediated regulation. As shown in Fig. 4
, activating PKA with forskolin and IBMX was without significant effect on ENaC irrespective of whether AKAP15 is coexpressed and irrespective of whether oocytes were studied under low or high Na+ conditions. Thus, the decrease in ENaC activity by this anchoring protein is unlikely to be related to its interaction with PKA.
|
This conclusion is further strengthened by the results summarized in Fig. 5
. These experiments used an agent that disrupts the interaction between AKAP15 and its PKA substrate. Indeed, injection of Ht31, a peptide known to block the interaction between PKA and AKAP15 (34)
, did not affect baseline or AKAP15-reduced ENaC currents under either high or low Na+ conditions. Altogether, these results demonstrate that the effects of AKAP15 on ENaC activity are not PKA mediated.
|
The presence of AKAP15 at the membrane is required for ENaC inhibition
The above experiments rule out a role of PKA and indicate an alternate or unconventional role for AKAP15 in reducing ENaC activity. Given that AKAP15 is thought to exert its actions at the plasma membrane and that this protein partitions nearly equally between a particulate and soluble fraction, experiments were carried out to determine whether membrane localization was essential to its novel effects on ENaC. To alter the membrane targeting of AKAP15, glycine 2, and cysteins at positions 5 and 6 were mutated to alanine (Fig. 6
A). Mutagenesis of this membrane-targeting domain has been shown to remove the N-terminal myristoylation sequences of AKAP15 and to inhibit its plasma membrane targeting (35)
. This effect was indeed verified by Western blot (Fig. 6B
), indicating that the bulk of AKAP15m is now contained in the cytoplasmic fraction.
|
The functional effects of coexpressing AKAP15m are summarized in Fig. 6C, D
. In the low Na+ solution, a condition that results in intracellular Na+ activity in the range of 
5 mEq (24)
, reduction of ENaC activity by AKAP15 is now eliminated. This indicates that inhibition of ENaC by AKAP15 requires localization of this AKAP to the membrane fraction. On the other hand, the gNa in the high Na+ solution, a condition that results in an intracellular activity of 50 mEq (24)
, was higher than that observed in ENaC-expressing oocytes. Moreover, gNa was increased to levels not different from those observed in oocytes expressing ENaC alone studied under low Na+ conditions.
These results indicate that the stimulation of ENaC by AKAP15m is likely an indirect process given the poor overlap between these two proteins (soluble vs. particulate fractions). Given the role of [Na+] in this stimulation, these results indicate that cytoplasmically located AKAP15 may interact with components of the process of Na+-mediated feedback inhibition of ENaC. This hypothesis is further examined in the ensuing sections.
Membrane capacitance was unchanged between ENaC- and ENaC+AKAP15m-expressing oocytes in both high and low Na+ conditions (Fig. 6D
). The absence of a change of Cm indicates that the elevated conductance observed in the ENaC+AKAP15m-expressing oocytes under high Na+ is not due to a change of membrane area. Of interest is the observation that the mutation of AKAP15 membrane targeting eliminated the increase in Cm. The significance of this effect is examined in the Discussion section.
Inhibition of membrane-resident channels
The inhibitory effects of AKAP15 on ENaC can, in principle, be mediated via a decrease of any of the following parameters: channel density (NT), channel open probability (Po), and single-channel conductance (g). The absence of a decrease of Cm indicates that NT is likely unchanged, leaving the possibility of effects on g or Po. This hypothesis was independently tested by examining ENaC protein surface density and channel activity.
Effects on membrane-bound ENaC density were examined using an immunobinding assay that allows measurement of relative changes of ENaC surface expression, thereby providing an index of changes of NT (26)
. The binding data are summarized in Fig. 7
and indicate that despite the marked reduction of channel activity observed with AKAP15 expression, no accompanying changes in surface expression are observed. The absence of an effect is not due to assay limitations, as this technique can be used to detect small (25%) changes in ENaC surface density (26)
. Thus, AKAP15 coexpression is likely reducing g or Po of membrane-resident ENaC.
|
We used patch clamp analysis to test the effects on g. Experiments used the cell attached patch clamp configuration in ENaC-expressing and ENaC+AKAP15-coexpressing oocytes. A recording from an ENaC- and ENaC+AKAP15-expressing oocyte is shown in Fig. 8
A, B. Both recordings were measured at an applied voltage of 100 mV (pipette with respect to the cell). Lower channel activity is observed in the AKAP15-coexpressing oocyte, but no major differences in the single-channel currents could be observed. This is also evident from the current/voltage relationship summarized in Fig. 8C
indicating similar currents in the voltage range of –40 to –120 mV. The single-channel conductance was calculated from linear regression of the currents in Fig. 8C
; it was unchanged and averaged 4.64 ± 0.11 (n=5) and 4.45 ± 0.32 (n=5) pS for ENaC and ENaC+AKAP15 oocytes, respectively.
|
The effect of AKAP15 coexpression on channel activity is summarized in Fig. 8D
. Data are summarized as NPo and indicate a nearly 4-fold lower activity in the ENaC+AKAP15 group compared with that in the ENaC group. This decrease in channel activity is similar to that observed at the whole-cell level examining the amiloride-sensitive conductance. Given the absence of a change of g and NT, the above data indicate a decrease in the activity of membrane-resident channels and likely single-channel Po. Although this parameter normally can be directly assessed from cell-attached patch clamp data, studies with ENaC pose additional technical problems that preclude an accurate assessment of Po given the channel’s long open and closed times, especially in experiments with membrane patches with markedly different activities (as observed between ENaC and ENaC+AKAP15 groups). This is one of the main advantages of pursuing the binding studies as an independent method of assessing NT.
Interaction with PKC
The role of [Na+]i in the response to AKAP15m coexpression indicates an interaction with a regulatory component that mediates channel feedback inhibition. A potential candidate for this interaction would be PKC, which is involved in the feedback inhibition of ENaC by elevated [Na+]i. More specifically, PKC was previously implicated in the feedback regulation of ENaC (11)
. This isoform is endogenously expressed in oocytes; it is inhibitory of ENaC (see below and ref. 11
), making it an appropriate candidate for interaction with AKAP15. Such an interaction would also be consistent with the emerging role of AKAPs as mediators of multiple kinase signaling cascades.
To test this hypothesis, we examined the effects of PKC coexpression. As shown in Fig. 9
A, coexpression with this kinase isoform reduced ENaC activity to 28% of baseline, consistent with the premise that conventional PKC isoforms inhibit ENaC (11)
. In the presence of AKAP15m, this inhibition is attenuated to 72% of baseline, indicating that channel inhibition by PKC and presumably feedback inhibition can indeed be prevented by AKAP15 expression. These results establish a functional link between cytoplasmic AKAP15 and PKC, a link that was previously unknown.
|
Physical interaction between AKAP15 and PKC was confirmed in coimmunoprecipitation experiments. As shown in Fig. 9B
, pull down of AKAP15 is accompanied by PKC, indicating a direct interaction between these proteins. This effect was observed only in the cytoplasmic fraction and was absent from the membrane fraction. One possibility for the absence of "direct" interaction between membrane AKAP15 and PKC is that their modification at the membrane—myristoylation for AKAP15 and phosphorylation for PKC—may limit their physical interaction.
These results indicate that the binding of AKAP15 to cytoplasmic PKC is likely sequestering this kinase isoform, leading to stimulation of ENaC by eliminating the effects of this kinase as a mediator of Na+ feedback inhibition. Our data also rule out direct interaction between AKAP15 and PKC at the membrane as the mechanism leading to channel inhibition. In co-IP experiments, we detect no direct interaction between any of the ENaC subunits and AKAP15 (data not shown). However, these results do not rule out an indirect mechanism such as that mediated by membrane lipids, as potential lipid-mediated interaction would be lost by the co-IP solubilization protocol.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
. These results establish a novel regulatory pathway that may play a role in regulating ENaC activity in a variety of cells.
Regulation by PKA
A hallmark of ENaC regulation in renal epithelia is the PKA-mediated stimulation by ADH. This effect is robust, and in many Na+-absorbing epithelia (e.g., frog skin) can result in as much as a 10-fold increase of membrane-bound channel density (48)
. Clearly such responses are not observed with the cloned ENaC. Indeed, bypassing the steps leading to PKA activation by simply injecting the purified kinase is also without effect on the cloned ENaC (see Fig. 3
). Thus, we establish the first direct evidence for the lack of effect of PKA on the cloned ENaC and also rule out AKAPs as direct mediators between PKA and ENaC. These results indicate that the oocyte system is ideal for heterologous expression of PKA-interacting proteins to test their roles in PKA-mediated regulation of ENaC.
AKAPs
The protein kinase A anchoring proteins are a group of structurally diverse proteins that have a common function of binding protein kinases. The initial or classical function of AKAPs was binding to the regulatory subunit of protein kinase A, thereby confining this kinase to discrete cellular compartments. This function has now been expanded to include other binding partners. Indeed, AKAP79, also known as AKAP5, associates with the ß-adrenergic receptor, which is coupled downstream to a G-protein. This association involves electrostatic interaction with a G-protein-coupled receptor as well as with the inner leaflet of the cell membrane (36)
. AKAP79 was also shown to associate with PKC and to organize a scaffold with PKA and protein phosphatases (37)
. This interaction was coupled to the underlying cytoskeleton effects exhibited on epithelial cell integrity, differentiation, and polarization through binding to cadherin (37)
. Similarly, AKAP15 was shown to organize binding between PKA, aquaporin2 water channel, and phosphodiesterase 4 in renal collecting duct principal cells (38)
. Our data are consistent with this emerging function of AKAPs, where we demonstrate effects of AKAP15 on ENaC and protein-protein interaction between AKAP15 and PKC.
Effects of AKAP15
The anchoring protein AKAP15 was originally cloned from rabbit skeletal muscle, where it was shown to mediate binding between an L-type Ca2+ channel and PKA (18
, 39)
. Fraser et al. also cloned this AKAP from a human fetal brain library and referred to it as AKAP18. They also demonstrated that this AKAP regulated L-type Ca2+ channel activity when heterologously expressed in cultured cells (35)
. These actions were dependent on an interaction with PKA at the cell membrane. Reports have demonstrated that AKAP15 is endogenously expressed in cultured renal epithelial cells (31)
. More recently, this AKAP was also found in native renal epithelial cells, where it was shown to regulate aquaporin 2 in a PKA-dependent manner (30
, 40)
. Both native and cultured AKAP15-expressing renal epithelia express ENaC, allowing the potential for endogenous interaction between these proteins.
We demonstrate a reduction in ENaC activity when coexpressed with AKAP15. These effects were specific to ENaC, as AKAP15 expression was without effect on ROMK1 activity (41)
and led to a stimulation of an oocyte endogenous Cl– current (data not shown). The specificity of the effects on ENaC was also confirmed by the absence of a parallel decrease in membrane capacitance and, presumably, area. The reduction in ENaC activity was not due to saturation of protein expression because 1) our experiments were carried out in cells injected with subsaturating levels of cRNA, 2) AKAP15m expressed equally well but did not inhibit ENaC activity, and 3) membrane-bound ENaC levels, which are a reflection of those in the entire oocyte, were unchanged.
Using patch clamp analysis, we demonstrated that AKAP15 coexpression did not alter the ENaC single-channel current or conductance. However, coexpression with this AKAP led to a nearly 4-fold decrease in NPo or channel activity (see Fig. 8
). Given the results of binding experiment indicating no change of N (Fig. 7)
, these data indicate a decrease in activity of membrane-resident ENaC, consistent with a decrease in Po.
Our data indicate an interaction between AKAP15 and ENaC at the membrane. Other AKAPs have been shown to directly interact with membrane-bound ion channels and transporters (18
, 38
, 40
, 42)
. AKAP15 was also demonstrated to directly bind a skeletal muscle Ca2+ channel (19
, 20)
. Therefore, such an interaction remains a possibility for ENaC. Using co-IP experiments, we did not observe pull down of any of the ENaC subunits with AKAP15 (data not shown), ruling out a direct electrostatic interaction between these proteins. However, these co-IP experiments do not rule out all forms of potential interaction between ENaC and AKAP15, especially the lipid-mediated ones, as they would be lost by the necessary solubilization protocol. AKAP15 is known to associate with the membrane through its lipid binding domain; since such an interaction is obligatory for the membrane localization of this AKAP and its inhibitory effects on ENaC (see Fig. 6
), and as ENaC has been shown to directly bind lipids (43
, 44)
, a lipid-mediated interaction cannot be excluded.
Effects of N-terminal myristoylation AKAP15 mutant
Mutation of the N-terminal myristoylation site is known to disrupt plasma membrane localization of AKAP15 without affecting total protein levels (31
, 35)
. Consistent with these effects, we demonstrate a loss of protein distribution in the membrane fraction with AKAP15m. Expression of this construct also eliminated the reduction observed in ENaC activity, indicating that such an effect required the presence of both proteins at the membrane. This is consistent with the effects of AKAP15 on other ion channels that require an intact membrane-targeting sequence (31
, 35)
and is consistent with the absence of effects of AKAP15m coexpression on an oocyte endogenous Cl– channel (data not shown).
Additional effects of AKAP15m were observed when present in the cytoplasm. AKAP15m coexpression stimulated ENaC under high Na+ conditions. As elevated Na+ activity has been shown to induce feedback inhibition of ENaC (24)
, we proposed that soluble or cytoplasmic AKAP15 may inhibit this process, leading to stimulation of ENaC. Such a mechanism would predict that only oocytes with an activated feedback inhibition pathway (i.e., those in Na+-loaded conditions) would respond to cytoplasmic AKAP15 with increased ENaC activity. Moreover, this activity should approach levels observed in the absence of feedback inhibition (i.e., those under low Na+ conditions). As shown in Fig. 6
, this is precisely what we observed.
This hypothesis was further validated by demonstrating functional and physical interaction between AKAP15 and PKC (see Fig. 9
). The role of PKC follows from the role of this kinase in feedback inhibition of ENaC (6
, 11
, 24)
. This isoform was chosen for three reasons. First, PKC is known to be involved in the feedback regulation of ENaC (11)
. Second, this ubiquitous isoform is endogenously expressed in many cell types, including oocytes (11)
. Third, activation of this kinase isoform also leads to marked inhibition of ENaC (10)
. Given the absence of an interaction between membrane AKAP15 and PKC, our data indicate that binding of cytoplasmic AKAP15 to PKC likely precludes membrane targeting of this kinase isoform, thereby relieving its role in the feedback inhibition of ENaC. This interpretation is consistent with the translocation of activated PKC from the cytoplasm to the membrane as a requisite for its actions on ENaC (11)
and the novel roles of other AKAPs in binding PKC (45
46
47)
.
Effects on Cm
The group of oocytes expressing AKAP15 exhibited an elevated Cm in both high and low Na+ conditions (see Fig. 2
). This effect was specific to the AKAP15 construct that reached the membrane and was not observed with AKAP79 or AKAP-KL. Two mechanisms may explain this increase of Cm. First, AKAP15 may have caused a small increase of membrane area. In this case, because of the large decrease of ENaC activity, this effect may have gone unnoticed. A second, more likely, explanation is that AKAP15 expression may have altered the membrane dielectric properties. We previously demonstrated that temperature, Gd+3, and chlorpromazine (23)
can cause large changes in the membrane dielectric properties leading to appreciable effects on Cm. Note that oocytes expressing AKAP15m, a construct that failed to reach the membrane, did not exhibit this increase in Cm. Moreover, it is interesting that the other AKAPs tested, which are not known to interact directly with or bind to lipids, did not exhibit this increase of Cm either. Thus, AKAP15 may alter the membrane dielectric properties through its myristoylation. This hypothesis remains to be tested.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
. It is expected that this process would have roles in addition to its effects on ENaC, as many of the cellular processes affected by the recruitment of PKC to the membrane would also be affected by this AKAP. Our data indicate that the distribution of AKAP15 between the membrane and cytoplasmic pools creates a dynamic equilibrium in which its effects on ENaC can be either stimulatory or inhibitory. Moreover, processes that regulate this distribution would also control ENaC activity.
|
ACKNOWLEDGMENTS |
|---|
construct and W. A. Catterall (Washington University) for the AKAP15 antibody. We also thank I. Vukojicic and W. Shao for superb technical assistance. This work was supported by National Institutes of Health grant DK55626. Received for publication March 22, 2006. Accepted for publication November 21, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
