Role of the SNARE protein SNAP23 on cAMP-stimulated renin release in mouse juxtaglomerular cells

Abstract

Renin, the rate-limiting enzyme in the formation of angiotensin II, is synthesized and stored in granules in juxtaglomerular (JG) cells. Therefore, the controlled mechanism involved in renin release is essential for the regulation of blood pressure. Exocytosis of renin-containing granules is likely involved in renin release; a process stimulated by cAMP. We found that the “soluble NSF (N-ethylmaleimide-sensitive factor) attachment protein receptor” (SNARE) protein VAMP2 mediates cAMP-stimulated renin release and exocytosis in JG cells. To mediate exocytosis, VAMP2 must interact with a synaptosome-associated protein (SNAP). In the renal cortex, the isoform SNAP23 is abundantly expressed. We hypothesized that SNAP23 mediates cAMP-stimulated renin release from primary cultures of mouse JG cells. We found that SNAP23 protein is expressed and colocalized with renin-containing granules in primary cultures of mouse JG cell lysates. Thus, we then tested the involvement of SNAP23 in cAMP-stimulated renin release by transducing JG cells with a dominant-negative SNAP23 construct. In control JG cells transduced with a scrambled sequence, increasing cAMP stimulated renin release from 1.3 ± 0.3 to 5.3 ± 1.2% of renin content. In cells transduced with dominant-negative SNAP23, cAMP increased renin from 1.0 ± 0.1 to 3.0 ± 0.6% of renin content, a 50% blockade. Botulinum toxin E, which cleaves and inactivates SNAP23, reduced cAMP-stimulated renin release by 42 ± 17%. Finally, adenovirus-mediated silencing of SNAP23 significantly blocked cAMP-stimulated renin release by 50 ± 13%. We concluded that the SNARE protein SNAP23 mediates cAMP-stimulated renin release. These data show that renin release is a SNARE-dependent process.

in mammals, renin is stored in electron-dense core granules in the juxtaglomerular (JG) cells of the kidney. Renin is the rate-limiting enzyme in the formation of angiotensin II (ANG II), therefore playing a dominant role in the control of blood pressure. Some of the stimuli leading to renin release from JG cells are low NaCl (sensed by the macula densa) and neurohormonal pathways (via β-adrenergic stimulation). These stimuli lead to increases in the intracellular second messenger cAMP in JG cells (10). However, the intermediate steps between cAMP synthesis and the release of renin from storage granules in the JG cells remain unclear.

Renin is stored in dense core granules where it is processed from its pro-form, pro-renin (10). Taugner et al. (33) found that mature electron-dense renin granules most likely fuse with the plasma membrane to release renin from JG cells. In addition, membrane capacitance and imaging studies showed increased exocytosis via a cAMP/protein kinase A (PKA) signaling mechanism (4, 26). However, there is no information on the molecular mechanisms and proteins involved in exocytosis of renin-containing granules.

In secretory cells, the process of exocytosis and membrane fusion is mediated by “soluble NSF (N-ethylmaleimide-sensitive factor) attachment protein receptors” (SNAREs) proteins (2), whose assembly and disassembly are further modulated by additional regulatory proteins (31). Once secretory cells are stimulated, a minimum of three SNARE family members interact to form a stable four-helix complex: one vesicle-associated membrane protein (VAMP), one syntaxin, and one synaptosome-associated protein (SNAP) (36). In neurons (30) and neuroendocrine cells, such as pancreatic beta cells (7, 15, 17, 37), VAMP-2, Syntaxin 1A, and SNAP25 are the putative SNARE proteins.

In nonneuronal secretory cells, VAMP2 remains the dominant secretory granule SNARE protein to mediate exocytosis. However, SNAP25 is replaced by several other SNAP isoforms in mediating exocytosis, which are ubiquitously expressed (3, 12, 13, 29). The isoform SNAP23 mediates the final step in regulated granule fusion with the plasma membrane in many nonneuronal secretory cells such as pancreatic acinar cells (6, 16), mast cells (9), and fat cells (21, 27). The renal cortex shows high levels of SNAP23 expression (20). Specifically, SNAP23 mediates cAMP-stimulated exocytosis in principal cells (20) and collecting ducts of the kidney (8). However, it is not known whether SNAP23 is present and mediates cAMP-stimulated renin release in JG cells.

We embarked on identifying the components of the exocytic machinery mediating renin-dense core granule exocytosis in JG cells. Recently, we identified VAMP2 as the VAMP isoform that mediates cAMP-stimulated renin release in JG cells (23). In the present study, we hypothesized and demonstrated that SNAP23 is present and involved in regulated release of renin in primary cultures of JG cells.

MATERIALS AND METHODS

Isolation and culture of mouse primary JG cells.

Primary cultures of mouse JG cells were prepared as described and characterized before (23, 25). C57/BL6 mice (8 to 9 wk old, Jackson Laboratories) were killed by cervical dislocation. Kidneys were removed, decapsulated, and the renal cortex was dissected. Cortical tissue from four mice was minced and digested as we described previously (23). Cells were cultured in DMEM containing 100 U/ml penicillin, 100 μg/ml streptomycin, and 5% fetal calf serum (Hyclone) at 37°C/5% CO₂ in poly-d-lysine-coated plates (0.1 mg/ml; Millipore). All protocols were approved by the Institutional Animal Care and Use Committee of Henry Ford Hospital and in accord with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Western blot.

To study the expression of SNAP23, JG cells were lysed in a buffer containing (in mM) 150 NaCl, 50 HEPES (pH 7.5), 1 EDTA (pH 8), 2% Triton X-100, 0.2% SDS, and a protease inhibitor cocktail (1). Protein content was measured by colorimetric assay (BSA, Pierce). Proteins were resolved on 12% SDS-PAGE and transferred to a PVDF membrane (Amersham). Membranes were incubated first in blocking buffer containing 50 mM Tris, 500 mM NaCl, 0.1% Tween 20 (TBS-T), and 5% nonfat dried milk for 60 min and then with a primary antibody (1/15,000 rabbit SNAP23; Synaptic Systems) in blocking buffer for 60 min. Membranes were washed in TBS-T and incubated with a secondary antibody conjugated to horseradish peroxidase (1/4,000 anti-rabbit; Amersham). For SNAP25, a monoclonal antibody was used at 1/5,000 dilution (Covance). VAMP2 (monoclonal), VAMP3, and VAMP4 antibodies (polyclonal) were used at a 1/3,000 dilution (Synaptic Systems). As an internal loading control, membranes were reblotted with an antibody against the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (1/100,000 GAPDH; Millipore). Signals were detected with a chemiluminescence kit (Amersham).

Immunofluorescence and confocal microscopy.

JG cells were grown on poly-d-lysine-coated coverslips. Following fixation with 4% paraformaldehyde in PBS (pH 7.4) for 30 min, membranes were permeabilized with 0.1% Triton X-100 in PBS for 10 min and blocked in TBS-T/5% albumin for 30 min. Cells were incubated first with a FITC-labeled antibody against renin (1/250, Innovative Research) for 1 h. Then, cells were labeled for 1 h with the primary rabbit antibody for SNAP23 (1/100, Synaptic System), followed by 1-h incubation with an anti-rabbit secondary antibody (1/200 Alexa Fluor 568) and mounted with Fluoromount-G (SouthernBiotech). Images were obtained using a laser-scanning confocal imaging system (Visitech International) with 488-nm (renin) and 568-nm (SNAP23) excitation lasers. Fluorescence was filtered with 525- and 590-nm emission filters, respectively, and images were acquired with a ×100 oil immersion lens (1.33 NA) at a pixel size of 0.05 μm/pixel in serial 0.3-μm optical sections in the z-axis plane of the cells. Samples incubated in the absence of primary antibodies were used as a control for nonspecific binding. In addition, an extra sample without antibodies was mounted as a control for autofluorescence. Images were acquired and saved as Tagged image file format. To quantify colocalization, optical sections on z-axis from the top, middle, and bottom planes of individual JG cells (separated by 3 μm) were minimally deconvolved with Autoquant software (Media Cybernetics) using two-dimensional blind deconvolution. Images from both channels were aligned; pixel-by-pixel colocalization measured using a miminum Mander's overlap coefficient of 0.95 and an image for overlapping pixels was generated. The number of renin-positive granules that contained pixels colocalizing with SNAP23 was counted manually or using segmentation software. Results are expressed as percentage of renin granules showing colocalization with SNAP23. This analysis was performed in multiple cells from two independent preparations.

Stimulation of renin release.

JG cells were serum deprived for 2 h by replacing the medium with serum-free DMEM (DMEM-SF) containing 100 U/ml penicillin and 100 μg/ml streptomycin. Renin release was stimulated by increasing intracellular levels of cAMP with forskolin (10 μM) plus 3-isobutyl-1-methylxanthine (IBMX; 0.5 mM) for 1 h. Following treatment, the medium was centrifuged to remove cellular debris. Supernatants were collected in fresh tubes and stored at −20°C until processing. To measure renin content in cells, after collection of cell culture medium, 0.5 ml of 0.1% Triton X-100 in PBS was added. Plates were rocked for 10 min at room temperature. Cells were scraped, collected, and spun at 16,100 g in a countertop centrifuge and supernatants were stored at −20°C until processed.

Measurement of renin release and renin content.

Supernatants and cell lysates were incubated with excess of rat angiotensinogen at 37°C for 3 h as we described previously (23). To terminate the reaction, the samples were boiled for 10 min followed by centrifugation at 16,100 g. Generated ANG I was measured using a gamma coat RIA kit (DiaSorin). Values for renin concentration (ng ANG I generated per h incubation) in cell culture medium (renin release) were normalized to renin concentration in cell lysates (renin content) and expressed as a percentage of renin content (renin release/renin content × 100). Under the incubation conditions, <10% of the substrate was consumed and ANG I production over time remained linear.

Construction of adenoviral vectors.

Adenovirus-cytomegalovirus (AdCMV)-hemagglutinin (HA)-tagged SNAP-23ΔC8 (Ad-dn-SNAP23) was generated as reported previously (5, 21). For construction of adenoviral particles encoding short hairpin silencing RNA (shRNA) against SNAP23, we first tested four annealed double-strand nude oligonucleotide sequences (sequences #1–4) against mouse SNAP23. Sequence information was obtained from SA Biosciences and oligonucleotide sense and antisense synthesized by Eurofins MWG Operon: sequence #1: 5′-ACAACUCACCUAGCAAUGUTT-3′, sequence #2: 5′-AGGAGAUUUCCUCAAAGUUTT-3′, sequence #3: 5′-AGGUUCUUGGAUCCAGUUUTT-3′, and sequence #4: 5′-GCUUCUGUUGACAUUAAAUTT-3′.

For adenoviral particles production encoding the shRNA against SNAP23 (Ad-si-SNAP23), an oligonucleotide fragment encoding 19 nucleotides (nt) of mouse sense SNAP23 (sequence #4) followed by a loop region (TTCAAGAGA) and the antisense of the 19 nt was subcloned into the 5′ AflII and 3′ SpeI sites of the adenovector pMIGHTY (Viraquest, North Liberty, IA). Oligonucleotides encoded AflII and SpeI sites at the 5′ and 3′ ends, respectively, for easier insertion into the adenovector. The control construct (Ad-si-Cont) was generated similarly using a scrambled sequence (5′-TTCTCCGAACGTGTCACGT-3′). Constructs were sequenced before production of viral particles.

Transient transfection of mouse cell line and adenoviral transduction of primary cultures of JG cells.

A mouse cell line (MS1, ATCC) was cultured in DMEM containing 100 U/ml penicillin, 100 μg/ml streptomycin, and 5% fetal calf serum. For transient transfections, MS1 cells were seeded in six-well plates at a density of 0.3 × 10⁶ cells/well. After 24 h in culture, cells were transfected with 80 nM oligonucleotides plus Lipofectamine 2000 (Invitrogen) following the manufacturer's instructions. After 48-h posttransfection, cells were lysed, 1.25 μg protein was resolved by SDS-PAGE on 12% polyacrylamide gels, and SNAP23 protein expression was detected using an antibody from Synaptic Systems (1/15,000). As an internal loading control, membranes were reblotted with an antibody against GAPDH (Millipore).

Dominant-negative or shRNA SNAP23 was delivered to primary cultures of mouse JG cells by incubating them with DMEM-SF containing adenoviral particles (100 PFU/cell). After 3 h, fetal calf serum to reach a 5% concentration was added for 24 h in the case of Ad-dn-SNAP23 and 28 h for Ad-si-SNAP23. JG cells were then stimulated with F/IBMX for 1 h as described above.

Cleavage of SNAP23 with botulinum neurotoxin E.

Botulinum neurotoxin E (BotE) cleaves the COOH-terminal portion of mouse SNAP23 at position ∼185 amino acids (aa), inactivating it (34, 35). Intact JG cells were preincubated in DMEM-SF with either vehicle or 10–60 nM BotE (Metabiologics, Madison, WI) for 19 h. Then, cells were lysed, resolved by SDS-PAGE on 12% polyacrylamide gels, and SNAP23 was detected by Western blot. SNAP23 protein expression was studied with an antibody directed against the COOH-terminal 9 aa (Synaptic Systems). Therefore, a decrease in SNAP23 protein expression indicates cleavage. The cleaved ∼26-aa fragment of SNAP23 (∼3 kDa) runs out of the gel and it is not detected. For renin release studies, JG cells were preincubated with either vehicle or 10–60 nM BotE as described above. The medium was then changed to fresh DMEM-SF for 2 h and then stimulated with F/IBMX for 1 h.

Reagents.

Fetal calf serum was obtained from Hyclone, and DMEM culture medium and antibiotics from Invitrogen. Forskolin, IBMX, Percoll, and protease inhibitors were from Sigma. BotE was from Metabiologics (Madison, WI). Poly-d-lysine from Millipore and the RIA kits used to measure ANG I from DiaSorin (Stillwater, MN).

Statistical analysis.

Data were expressed as means ± SE and subjected to statistical analysis by t-test with correction of the rejection level using Hochberg's method or one-way ANOVA with multiple comparisons made by the Student-Newman-Keuls method. A value of P < 0.05 was considered significant.

RESULTS

SNAP23 is expressed in renin-containing secretory granules in primary cultures of mouse JG cells.

Consistent with our recent report showing SNAP23 mRNA expression in JG cells (23), we show that SNAP23 protein is abundantly expressed in primary cultures of JG cells. By Western blot, we detect a band corresponding to the predicted molecular weight of 23 kDa similar to that in brain homogenate used as a positive control (Fig. 1A; n = 4). Although SNAP25 is mainly neuronal, it is expressed in other endocrine organs and plays a role in the regulated exocytic pathway (16). We found that SNAP25 is not detectable in JG cells. However, we can detect a clear band at the expected molecular weight in a brain homogenate used as a positive control (Fig. 1B).

Fig. 1.

Expression and subcellular localization of synaptosome-associated protein (SNAP)23 in primary cultures of mouse juxtaglomerular (JG) cells. A: representative Western blot showing expression of SNAP23 (23 kDa). Lane 1 is JG cell lysate (2.5 μg); lane 2 is brain homogenate (7.5 μg) used as a positive control (n = 4). B: representative Western blot showing that SNAP25 is not expressed in JG cell lysates (25 kDa). SNAP25 and GAPDH protein are clearly detected in 2-μg brain homogenate (lane 1) but not in JG cell lysates even after overexposure of the film (lanes 2–5). GAPDH was used as a loading control (n = 4). C: immunofluorescence labeling and confocal microscopy of a single mouse JG cell. Left: stacked 3-dimensional (3D) image projection of 4 confocal slices in the middle of a single JG cell labeled with an antibody against renin (green). Dense core renin-containing granules can be observed. Granule size average 0.8- to 1.5-μm size. Middle: SNAP23 labeling (red). Right: pixel-by-pixel colocalization of renin and SNAP23 (blue). Bottom left: merged image illustrating colocalization of renin with SNAP23 (yellow-orange color). Bottom right: inset of higher magnification from merged panel.

Most importantly, we determined the subcellular localization of SNAP23 in JG cells by immunofluorescence and confocal microscopy. Double immunofluorescence labeling of JG cells with antibodies for renin (green) and SNAP23 (red) showed abundance of SNAP23 in renin-containing large secretory granules. Quantitative analysis of colocalizing granules revealed that 84 ± 4% of renin-labeled granules was also positive for SNAP23 (Fig. 1C).

We then proceeded to unequivocally demonstrate the function of SNAP23 in renin release employing several strategies.

Dominant-negative SNAP23 mutant protein partially blocks cAMP-stimulated renin release in primary cultures of mouse JG cells.

The first strategy we employed to test whether SNAP23 is involved in cAMP-stimulated renin release is by a dominant-negative approach. We previously showed that transduction of a COOH terminus (8 aa) truncated SNAP23-inhibited secretion in pancreatic acinar cells (5). We transduced JG cells with adenovirus encoding a dominant-negative SNAP23 (Ad-dn-SNAP23) lacking a portion of the COOH terminus essential for SNAP23-VAMP2 protein binding (21). We previously showed that adenoviral transduction itself did not affect basal or cAMP-stimulated renin release when compared with nontransduced cells (23). In JG cells transduced with an adenovirus encoding cytomegalovirus (Ad-CMV), cAMP stimulated renin release from 1.27 ± 0.26 to 5.33 ± 1.19% of renin content. However, in JG cells transduced with Ad-dn-SNAP23, cAMP stimulation of renin release was significantly impaired by ∼50% (from 1.0 ± 0.12 to 3.05 ± 0.6% of renin content, P < 0.05; Fig. 2A). The inhibitory effect of Ad-dn-SNAP23 did not likely affect processing of prorenin to renin since total renin values remained unchanged between groups (n = 12; P = not significant; Fig. 2B).

Fig. 2.

Dominant-negative SNAP23 impairs cAMP-stimulated renin release in primary cultures of mouse JG cells. A: renin release. Primary cultures of mouse JG cells were transduced with a control vector (Ad-CMV; black bars) or adenovector encoding dominant-negative SNAP23 (Ad-dn-SNAP23; gray bars). Twenty-four-hour postinfection, both groups were treated with vehicle or forskolin plus 3-isobutyl-1-methylxanthine (IBMX) for 1 h (n = 10; *P < 0.03 vs. Ad-CMV + F/IBMX; #P < 0.01 vs. control). B: total renin. Transduction of Ad-dn-SNAP23 (gray bars) into JG cells did not affect renin content compared with JG cells transduced with Ad- CMV (black bars). Renin release and content in JG cells were determined as described in materials and methods. Renin content from Control Ad-CMV was arbitrarily set to 100. Data are expressed as means ± SE [n = 11; P = not significant (N.S.)].

Cleavage of SNAP23 with BotE blocks cAMP-stimulated renin release in primary cultures of mouse JG cells.

A second strategy to examine SNAP23 function is by enzymatic cleavage of endogenous SNAP23 with botulinum neurotoxins (19). BotE specifically cleaves and inactivates SNAP23 (21) and SNAP25 (15, 17). Since SNAP25 is absent in JG cells (Fig. 1B), we first optimized the conditions to efficiently cleave SNAP23 in intact JG cells. After 19-h incubation with BotE (60 nM), cells were lysed and resolved on SDS-PAGE. Blots were probed with an antibody against SNAP23. As shown in Fig. 3, A and B, SNAP23 protein was significantly decreased by 39 ± 5% in BotE-treated JG cells (n = 7; P < 0.05). We then repeated the above protocol to study the effect of inactivation of SNAP23 on cAMP-stimulated renin release. We found that increasing cAMP with F/IBMX for 1 h stimulated renin release to 5 ± 0.7% of renin content. However, in JG cells pretreated with two increasing concentrations of BotE (10 and 60 nM), cAMP-stimulated renin release was inhibited by 51 and 67% respectively [from 5 ± 0.7 to 2.45 ± 0.48 (Bot 10 nM) and 1.67 ± 0.95 (Bot 60 nM)]. As a control, to ensure that the effect of BotE was due to its enzymatic activity, we then repeated the above protocol with boiled-heat-inactivated BotE. We found that the inhibitory effect of BotE on cAMP-stimulated renin release was not further observed (5.7 ± 0.5% of renin content n = 3; P = N.S. vs. F/IBMX; Fig. 3C). No significant differences between groups were found on total cellular renin values, suggesting that SNAP23 does not affect renin granule maturation or processing (data not shown).

Fig. 3.

Cleavage of SNAP23 with botulinum neurotoxin E (BotE) blocks cAMP-stimulated renin release. A: representative Western blot showing the effect of incubating intact JG cells with either vehicle (lane 1) or BotE (lane 2) on SNAP23. Note that cleaved SNAP23 with BotE is observed as a decrease on the 23-kDa band intensity, since the COOH-terminal portion recognized by the antibody is removed by BotE. The cleaved ∼26-aa fragment of SNAP23 (∼3 kDa) runs out of the gel and it is not detected. After treatment, JG cells were lysed and SDS-resolved in 12% polyacrylamide gels as described in materials and methods. Membranes were reblotted with an antibody against GAPDH as an internal loading control (∼35 kDa). GAPDH values were not changed (P = N.S.). B: optical density quantitation from SNAP23 before (vehicle) and after (60 nM BotE) treatment. Vehicle-treated O.D. values were arbitrarily set to 1 and the means ± SE of at least 7 independent experiments were determined (*P < 0.05). C: effect of SNAP23 cleavage with BotE on cAMP-stimulated renin release. The protocol described for A was repeated for secretion studies. JG cells were pretreated with either vehicle or BotE for 19 h. Graph shows stimulated renin release (F/IBMX-stimulated renin release − vehicle-treated renin release) expressed as percentage of renin content. Black bar is JG cells in the absence of BotE (n = 7). Gray bars are JG cells pretreated with 10 (n = 7, *P < 0.02 vs. black bar) or 60 nM BotE (n = 10, #P < 0.03 vs. black bar). Striped gray bar represents JG cells pretreated with boiled/inactive BotE as a control. Basal renin release values were not significantly different: No-BotE = 2.0 ± 0.3; 10 nM BotE = 3.3 ± 0.7; 60 nM Bot E = 4.6 ± 0.7 (P = N.S.).

Construction and testing of adenovirus silencing SNAP23.

To independently assess the role of SNAP23 on stimulated renin release, we used the silencing RNA approach. We first tested the efficiency of four different nude sequences to knockdown SNAP23 endogenous protein expression in a mouse cell line (MS1, ATCC) as evaluated by Western blot. We found that 48-h posttransfection of nude oligos #1–4 in MS1, sequence #4 (line 6) showed the greatest SNAP23 knockdown when compared with nontransfected (line 1) or cells transfected with a scrambled sequence (line 2; Fig. 4A).

Fig. 4.

Knockdown of SNAP23 protein expression by adenoviral delivery of short hairpin silencing in JG cells. A: noncoding (lane 2) and 4 SNAP23 double-strand oligonucleotide sequences (lanes 3–6) were transfected into mouse cell line (MS1) cells with liposomes as described in materials and methods. Forty-eight-hour posttransfection, cells were lysed, and their efficiency was tested by Western blot using SNAP23 antibody. Lane 1 shows an MS1 lysate without any treatment. B: sequence #4 was subcloned in adenovector and viral particles were tested in JG cells. JG cells were transduced with a scrambled sequence (lane 1) or silencing SNAP23 adenovirus (lane 2). After 28-h incubation, JG cell lysates were lysed and resolved in 12% SDS-PAGE for immunoblotting for detection of SNAP23 knockdown efficiency. Membranes were blotted for GAPDH as loading control. Knockdown of SNAP23 did not affect VAMP2, VAMP3, and VAMP4 protein levels. C: optical density quantitation from SNAP23. Band was normalized to GAPDH band in each experiment, and the means ± SE of at least 4 independent experiments were determined (*P < 0.05).

To ensure the efficacy of sequence #4 in JG cells, we then subcloned sequence #4 into an adenovector and viral particles were produced and tested. Transduction of JG cells for 28 h with adenovirus silencing SNAP23 (Ad-si-SNAP23) resulted in a ∼50% reduction in SNAP23 protein compared with adenovirus-scrambled sequence (Ad-si-Cont; n = 3; P < 0.05) without affecting VAMP2, VAMP3, or VAMP4 expression levels (Fig. 4B).

Adenoviral-mediated delivery of silencing-SNAP23 partially blocks cAMP-stimulated renin release in primary cultures of mouse JG cells.

We then employed a third and last approach to demonstrate the involvement of SNAP23 on cAMP-stimulated renin release by knockdown expression of endogenous SNAP23 using the Ad-si-SNAP23. We found that in JG cells transduced with a scrambled sequence (Ad-si-Cont) cAMP stimulated renin release from 1.07 ± 0.12 to 2.82 ± 0.43% of renin content. However, in JG cells transduced with adenovirus silencing SNAP23 (Ad-si-SNAP-23), cAMP-stimulated renin release was impaired by ∼55% (from 1.13 ± 0.22 to 1.9 ± 0.22% of renin content; Fig. 5A). Total renin content from silencing SNAP23 was unaffected compared with the scrambled-transduced group (P = N.S.; n = 7; Fig. 5B), indicating that SNAP23 is not likely involved in renin-containing granule maturation. These results taken together indicate that SNAP23 is implicated in stimulated renin release.

Fig. 5.

Silencing SNAP23 blocks cAMP-stimulated renin release in mouse JG cells. A: renin release. After transduction of JG cells for 28 h, cells were serum starved for 2 h and treated for 1 h with F/IBMX (10 μM/0.5 mM) or vehicle according to the description in materials and methods. Black bars are JG cells transduced with a scrambled sequence (Ad-si-Cont). Gray bars are JG cells transduced with silencing SNAP23 (Ad-si-SNAP23; n = 7; #P < 0.01 vs. vehicle-treated Ad-si-Cont; *P < 0.03 vs. Ad-si-Cont + F/IBMX). B: total renin. Total renin content values are corrected by protein concentration (ng ANG I·h incubation⁻¹·mg protein⁻¹). Renin content from Ad-si-Cont vehicle-treated was arbitrarily set to 100. Data are expressed as means ± SE (n = 7; P = N.S.).

DISCUSSION

Renin is the rate-limiting enzyme in the generation of ANG II, which has been long-lasting identified and targeted to regulate blood pressure (reviewed in Ref. 10). In addition, a detrimental role of renin independent of its enzymatic activity (18) has renewed an interest in directly targeting renin per se. It would seem even more attractive and efficient to prevent the release of renin by targeting the proteins involved in its exocytosis. However, fundamental questions on how renin is released from JG cells and the proteins and molecular mechanisms involved have not been explored to date.

Since renin is stored in large dense core granules in JG cells (10) and undergo exocytosis (4, 10, 23, 26, 33), these secretory granules very likely employ the SNARE membrane fusion machinery (2). In fact, we just identified VAMP2 to be one of the three minimally required SNARE family members that mediate cAMP-stimulated renin release in JG cells (23). However, the SNAP isoform mediating renin release is not known. The current work identified the second putative SNARE protein showing that SNAP23, and not SNAP25, is expressed in mouse JG cells, and it mediates cAMP-stimulated renin release from mouse JG cells.

Our immunofluorescence labeling and confocal imaging showed that most (84%) renin-containing dense core granules also contained SNAP23. However, we also observed that some granules containing renin did not express SNAP23. The antibody used to label renin recognizes both, renin and its proform prorenin. Thus, it is possible that renin-labeled granules that colocalized with SNAP23 might contain the mature/active form of renin, whereas those granules that do not posses SNAP23 presumably contain the immature proform, prorenin. Although speculative, this might point to SNAP23 being involved in a specific step of granule fusion (presumably with the plasma membrane) rather than an involvement in renin maturation. Our results are in agreement with those of others that demonstrated that SNAP23 is expressed not only at the plasma membrane but in vesicle and large granule membranes (16, 20).

Employing three strategies, we found that SNAP23 mediates cAMP-stimulated renin release in primary cultures of mouse JG cells. Our results show that blockade of renin release is partial after shRNA reduction of SNAP23 expression or proteolytic destruction with BotE, but nonetheless, the reduced secretion (∼50%) was proportionate to the reduction in endogenous levels of SNAP23 (∼50%). Although it is desirable to have complete ablation of SNAP23 expression, these results strongly suggest that SNAP23 is the dominant SNAP isoform that mediates cAMP-evoked renin release. Attempts to increase reduction of SNAP23 levels beyond ∼50% unfortunately resulted in cell damage, observed morphologically and also by a higher basal renin release. Another strategy would be by SNAP23 genetic deletion, but such SNAP23-null mice were recently reported to be not viable (32). We found that silencing or ablating SNAP23 did not affect the amount of total active renin in JG cells (renin content), suggesting a primary role of SNAP23 in the final steps that mediate renin release from granules rather than in renin maturation or the regulation of the total pool of renin.

In brain and neuroendocrine pancreatic islet beta cell exocytosis, SNAP25 is the main isoform mediating exocytosis (12), whereas SNAP23 is the major isoform in most nonneuronal secretory cells, such as pancreatic exocrine cells (16). Consistently, SNAP23 but not SNAP25 is expressed in nonneuronal JG cells. As SNAP25 could be functionally redundant to SNAP23 (28), the absence of SNAP25 in JG cells could explain the toxic effects of knocking down SNAP23 protein since SNAP25 cannot compensate for the lack of SNAP23. Botulinum neurotoxins have been widely used to identify the SNAREs mediating vesicle fusion in many mammalian cells studied (19) and exhibit target specificity. Although BotE can specifically cleave murine SNAP23 and SNAP25 (34, 35), we found that SNAP25 is not expressed in JG cells. Thus, it is unlikely that the blockade of renin release by BotE is due to SNAP25 cleavage.

The SNAP23 homolog SNAP29 has also been described to be ubiquitously expressed in mammalian cells (13). However, SNAP29 has been shown to preferentially bind to syntaxin isoforms involved in endosomal and trans-Golgi fusion events (13) and is expressed almost exclusively in organelles mediating postendocytic trafficking, whereas SNAP23 has been shown to mediate the ultimate step in regulated granule fusion at the plasma membrane in nonneuronal secretory cells (9, 27). Much less is known about the role of the isoform SNAP47 (14). While our data demonstrate that SNAP23 mediates most of the cAMP-stimulated renin release, a potential partial role of SNAP29 in JG cells cannot be ruled out.

The molecular mechanisms by which cAMP increases renin release via SNAP23 are not known. Previous studies showed that the SNARE complex can be regulated by modulatory binding partners and kinase proteins in other secretory cells (22). Neuronal SNAP25 can be phosphorylated by PKA, which modifies the size of the releasable pool of secretory vesicles (24). Perhaps a similar PKA phosphorylation site or other such sites may be present in JG cell SNAP23. However, phosphorylation of SNAP23 has opposing effects on exocytosis in mast cells (11). Differential regulation of exocytosis by such distinct SNAP23 phosphorylation events might partly explain cell-type specificity as well as stimulus dependency. What seems to be in common is that phosphorylation critically affects the ability of SNAP23 to interact with their cognate SNAREs to mediate exocytosis.

The precise mechanisms that mediate renin release have not been fully elucidated. Our previous and current work demonstrated that fusion proteins, namely VAMP2 (23) and SNAP23, are involved in stimulated renin release. Future studies will be directed at identifying the cognate syntaxin(s). It is possible that deregulated renin secretion causing hypertension could be enhanced by pathologic coupling of cAMP signaling to this exocytic complex via SNAP23 or that this fusion complex assembly is itself perturbed. Thus, the JG renin granule exocytic complex might present a provocative alternative target for drug development to control blood pressure.

GRANTS

The work was supported by a National Kidney Foundation fellowship grant and National Institutes of Health-National Research Service Award F32 to M. Mendez.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: M.M. conception and design of research; M.M. performed experiments; M.M. analyzed data; M.M. and H.Y.G. interpreted results of experiments; M.M. prepared figures; M.M. drafted manuscript; M.M. and H.Y.G. edited and revised manuscript; M.M. and H.Y.G. approved final version of manuscript.