Reciprocal Regulation between Proinflammatory Cytokine-induced Inducible NO Synthase (iNOS) and Connexin43 in Bladder Smooth Muscle Cells*

Gap junctions (GJs) play an important role in the control of bladder contractile response and in the regulation of various immune inflammatory processes. Here, we investigated the possible interaction between inflammation and GJs in bladder smooth muscle cells (BSMCs). Stimulation of BSMCs with IL1β and TNFα increased connexin43 (Cx43) expression and function, which was associated with increased phosphorylation of vasodilator-stimulated phosphoprotein. Inhibition of PKA with H89 or down-regulation of CREB with specific siRNAs largely abolished the Cx43-elevating effect. Further analysis revealed that IL1β/TNFα induced NFκB-dependent inducible NO synthase (iNOS) expression. Inhibition of iNOS with G-nitro-l-arginine methyl ester abrogated and an exogenous NO donor mimicked the effect of the cytokines on Cx43. Intraperitoneal injection of LPS into mice also induced bladder Cx43 expression, which was largely blocked by an iNOS inhibitor. Finally, the elevated Cx43 was found to negatively regulate iNOS expression. Dysfunction of GJs with various blockers or down-regulation of Cx43 with siRNA significantly potentiated the expression of iNOS. Fibroblasts from Cx43 knock-out (Cx43−/−) mice also displayed a significantly higher response to the cytokine-induced iNOS expression than cells from Cx43 wild-type (Cx43+/+) littermates. Collectively, our study revealed a previously unrecognized reciprocal regulation loop between cytokine-induced NO and GJs. Our findings may provide an important molecular mechanism for the symptoms of bladder infection. In addition, it may further our understanding of the roles of GJs in inflammatory diseases.

Acute cystitis is a common disease afflicting millions of people each year and is usually caused by the infection of the Gramnegative bacteria Escherichia coli (1). Patients with cystitis often complain of urinary frequency and urgency. Although the pathological basis for these symptoms is still poorly under-stood, it is thought to be related to the altered bladder microenvironment subsequent to the local influx of inflammatory cells and the production of inflammatory mediators (2,3).
Urinary frequency and urgency, representing the elevated detrusor excitability, are also typical symptoms of overactive bladder (4). In overactive bladder, these urodynamic dysfunctions are associated with the increased expression and function of gap junctions (GJs) 2 in the bladder (5)(6)(7). GJs are clusters of transmembrane channels that permit the direct intercellular exchange of ions, secondary messengers, and small signaling molecules. Gap junctional intercellular communication (GJIC) is thought to play an important role in the control of a variety of cellular functions, including cell growth, migration, differentiation, and electric coupling (8,9). GJs are formed by a family of special proteins termed connexins (Cx). To date, more than 20 different isoforms of Cx molecules have been identified. Among them, connexin 43 (Cx43) has been extensively investigated because of its ubiquitous expression in most cell types. Bladder smooth muscle cells (BSMCs) highly express Cx43 (10). In the normal bladder, Cx43-forming channels provide a pathway for the transmission and propagation of electrical signals, thus contributing to the coordinated contraction and relaxation responses required for normal bladder emptying and filling (5)(6)(7). In overactive bladder, Cx43 is abnormally up-regulated and has been shown to be implicated in the enhanced modular autonomous activity and detrusor overactivity (11). In this context, it is likely that the symptoms of bladder infection might also be related to the altered GJs in the bladder. As the first step toward demonstrating this, we examined the changes of GJs during bladder inflammation.
GJs are known to be critically involved in various immune inflammatory situations. On the one hand, they participate in the regulation of almost every step of the inflammatory response, including antigen presentation, chemokine and cytokine production, inflammatory cell migration, adhesion, and activation (12)(13)(14)(15)(16)(17). On the other hand, the proinflammatory mediators LPS, TNF␣, and NO have been identified as potent regulators of GJs in a variety of cell types (14, 18 -23). There appears to exist an interactive regulation loop between inflammation and GJs. However, this possibility has not yet been addressed. Another purpose of this study was to test this hypothesis.
Here, we present our evidence showing the existence of an NO-centered regulatory loop between inflammatory cytokines and GJs. Our findings thus provide potentially important insights into the molecular mechanisms of GJs in inflammatory disorders.
Cell Culture-BSMCs were established from the bladder of female Sprague-Dawley rats as described previously (24). For maintenance, the cells were cultured in DMEM/F-12 containing 10% FBS. For experiments, they were cultured in DMEM/F-12 containing 1% FBS with or without the indicated stimulants. Mouse embryonic fibroblasts were derived from the fetal offspring of mating pairs of heterozygous Cx43 knock-out mice (B6, 129-GjalϽtm 1 Kdr ϩ/Ϫ ; The Jackson Laboratory, Bar Harbor, ME), using a method described by Ehlich et al. (25) with minor modifications (26 -28). Briefly, both mouse forelimbs were taken from fetuses at day 18 of gestation, minced, and digested in DMEM/ F-12 containing 0.1% collagenase for 30 min. Freed cells were collected and cultured in DMEM/F-12 medium containing 15% FBS. Cells at passages 5-15 were used for this study. Genotypes of individual mice were determined by PCR.
Animals-Adult female C57BL/6J mice weighing 25-30 g were purchased from Japan SLC, Inc. (Hamamatsu, Japan). All animal experimental procedures were approved by the Animal Experimental Committee of Yamanashi University. Mice were housed in containment facilities of the Animal Center and maintained on a regular 12:12-h light/dark cycle with food and water.
Western Blot Analysis-Western blot was performed using an enhanced chemiluminescence system (22,23). Briefly, equal amounts of extracted cellular proteins were separated by 10% SDS-polyacrylamide gels and electrotransferred onto polyvinylidene difluoride membranes. After blocking with 3% BSA in PBS, the membranes were incubated with primary antibody. After washing with PBS, 0.1% Tween 20, filters were probed with horseradish peroxidase-conjugated sheep anti-rabbit IgG or rabbit anti-mouse IgG (Cell Signaling; Beverly, MA). Immunoreactivity was detected by an enhanced chemiluminescence system (Amersham Biosciences). The chemiluminescent signal was captured with a Fujifilm luminescent image LAS-4000 analyzer (Fujifilm, Tokyo, Japan). To confirm equal loading per lane, membranes were treated with 2% SDS and 100 mM ␤-mercaptoethanol in 62.5 mM Tris-HCl (pH 6.8) for 30 min at 60°C and reprobed for ␤-actin (dilution 1:30,000; Sigma). Data shown are representative of at least three independent experiments with similar results.
Northern Blot Analysis-BSMCs were treated with various agents for 12 h. Equal amounts of RNA (5 g) extracted from cells were separated by electrophoresis and transferred onto nylon membranes (Hybond N ϩ ; Amersham Biosciences). The level of Cx43 mRNA was examined as described previously using the entire coding sequence of the rat Cx43 cDNA as a probe (provided by Drs. G. Olbina and W. Eckhart, Molecular and Cell Biology Laboratory, The Salk Institute for Biologic Studies, San Diego). The staining of 28 S and 18 S ribosomal RNA by ethidium bromide was used as a loading control (22,23).
Transient Transfection of BSMCs with siRNA-Cells were transiently transfected with siRNA specifically targeting CREB (Hs_CREBBP_2 FlexiTube siRNA, Qiagen, Japan), Cx43 (Mm_Gja 1_2 HP siRNA, Qiagen, Japan), or a negative control siRNA (AllStars Negative Control siRNA) at a final concentration of 20 nM using Hyperfect transfection reagent for 24 h. After transfection, cells were left untreated or exposed to IL1␤/ TNF␣ for the indicated time. Cellular protein was extracted and subjected to Western blot analysis of the targeted proteins (Cx43, iNOS, and CREB).
Immunocytochemical Analysis-Immunocytochemical staining for Cx43 was performed as reported previously (22,23). In brief, cultured BSMCs were fixed in 2% paraformaldehyde in PBS for 15 min and permeabilized with 1% Triton X-100 before staining. The cells were incubated overnight with anti-Cx43 antibody (diluted 1:200 in 1% FBS in PBS; 4°C). After rinsing with PBS, the appropriate secondary antibody (diluted in 1% FBS in PBS; 37°C) was added for 2 h before final washing. The slides and sections were covered with Tris-buffered moviol (pH 8.6), and microscopy was performed with an Olympus BX50 microscope with a 40ϫ Planapo and 570-nm emission filter. Immunofluorescent images were captured using a CCD camera attached to the microscope.
Immunohistochemical Analysis-Bladder cryosections of about 6 m thickness were routinely rehydrated in PBS (pH 7.4) and stained for Cx43 and phalloidin. The secondary antibody for Cx43 was Alexa Fluor 488 donkey anti-rabbit IgG. Briefly, the serial sections were treated with 1% hydrogen peroxide in PBS to inhibit enzyme activity of endogenous peroxidase and additionally incubated in 10% BSA. They were then incubated in primary antibodies at 4°C overnight and corresponding secondary antibodies at room temperature for 1 h. All immunostained sections were embedded in Vectashield (Vector Laboratories) and observed on a confocal laser-scanning microscope (FV1000; Olympus, Tokyo, Japan).
Scrape Loading Dye Transfer (SLDT) Assay-The SLDT assay was used to assess GJIC (27). Cells were exposed to culture medium containing 0.1% LY and 0.05% RD. A scrape line on the monolayer was made with a surgical blade. After washing out background fluorescence, the cells were fixed, and dye transfer results were examined using an Olympus BX50 microscope with a 40ϫ Planapo and FITC (green) or TRITC (red) filter. Immunoflu-orescence was photographed using a CCD camera attached to the microscope. The distances of LY diffusion from the scrape line were counted for statistical analysis. Because of its large size (10,000 Da), the RD cannot pass through GJs and was used as marker of the scratch-loaded cells. GJIC was evaluated by comparing the diffusion of the LY to the nondiffusable RD.
Measurement of Nitric Levels-NO released into the culture medium was measured by evaluating nitrite accumulation using Griess reagent (29,30). Briefly, 100 l of conditioned medium was mixed with an equal amount of Griess reagent (a solution containing 1% sulfanilamide, 0.1% naphthylethylenediamine in 2 M HCl) and allowed to stand at room temperature FIGURE 1. Effects of proinflammatory cytokines on Cx43 expression and GJIC. Effects of IL1␤ and TNF␣ on Cx43 expression at the protein (A and B) and mRNA levels (C and D). BSMCs were exposed to 2 ng/ml IL1␤ or 20 ng/ml TNF␣ alone or in combination for 24 (A) or 12 h (C). Cellular protein or RNA was extracted and subjected to Western (A) or Northern blot analysis (C) for Cx43. ␤-Actin and ribosomal RNA 18 S and 28:S were used as a loading control. P0, P1, and P2 in A denote the nonphosphorylated, phosphorylated, and hyper-phosphorylated Cx43 form, respectively. B and D, densitometric analysis of Cx43 expression shown in A and C. Results were expressed as induction relative to the basal level of Cx43 (mean Ϯ S.D., n ϭ 3). #, p Ͻ 0.01 versus control. *, p Ͻ 0.05 versus single stimulation. E, time-dependent effect of IL1␤ and TNF␣ on Cx43 protein levels. BSMCs were treated with 2 ng/ml IL1␤ and 20 ng/ml TNF␣ in combination for the indicated duration. F, immunofluorescent staining of Cx43 in cultured BSMCs. BSMCs were treated with or without 2 ng/ml IL1␤ and 20 ng/ml TNF␣ for 24 h and subjected to immunofluorescent staining of Cx43 (magnification ϫ400). G and H, SLDT assay for GJIC. BSMCs were treated the same as above. LY (green) diffusion into the cellular monolayer after scrape loading is shown. The nondiffusable RD (red) was used to indicate the scratch-loaded cells. The merged image of LY and RD is also shown. H, distance of LY diffusion in G was counted, and the values are expressed as cell layer (mean Ϯ S.D., n ϭ 3). #, p Ͻ 0.01 versus untreated control.
for 10 min. The absorbance at 550 nm was then measured using a microtiter plate reader. Nitrite levels were expressed in nanomoles of NO 2 /g of total cellular protein.
Statistical Analysis-Values are expressed as means Ϯ S.D. Comparisons of two populations were made using Student's t test. For multiple comparisons with a single control, one-way analysis of variance followed by Dunnett's test was employed. Both analyses were carried out using SigmaStat statistical soft-ware (Jandel Scientific). p Ͻ 0.05 was considered to be a statistically significant difference.

Proinflammatory Cytokines Increase Cx43 Expression and
Function-To test whether bladder inflammation could affect GJs, we examined Cx43 protein levels after incubation of BSMCs with the major proinflammatory cytokines IL1␤ and . BSMCs were exposed to 2 ng/ml IL1␤ and 20 ng/ml TNF␣ for the indicated duration. Cellular protein was subjected to Western blot analysis for phosphorylated VASP at serine 157. ␤-Actin was used as a loading control. B, blockade of the effects of IL1␤/TNF␣ on Cx43 and pVASP by the PKA inhibitor H89. BSMCs were pretreated with or without 20 M H89 for 30 min and then exposed to 2 ng/ml IL1␤ and 20 ng/ml TNF␣ for 24 h. C and D, densitometric analysis of Cx43 and pVASP shown in B. Results are expressed as induction relative to the basal level of Cx43 and pVASP, respectively (mean Ϯ S.D., n ϭ 3). #, p Ͻ 0.01. E, effect of H89 on the cytokine-induced GJIC. Cells were treated the same as in B. GJIC was evaluated by SLDT assay. The distance of LY diffusion was counted and is expressed as cell layer (mean Ϯ S.D., n ϭ 3). #, p Ͻ 0.01. F, inhibition of the IL1␤/TNF␣-elicited increase in Cx43 protein levels by specific siRNA against CREB. BSMCs were transfected with CREB siRNA or control siRNA for 24 h. The cells were then exposed to IL1␤/TNF␣ for an additional 24 h. Cellular proteins were subjected to Western blot analysis for Cx43 and CREB. G and H, densitometric analysis of CREB and Cx43 levels shown in D. Results are expressed as induction relative to the basal level of CREB and Cx43 (mean Ϯ S.D., n ϭ 3). #, p Ͻ 0.01. Fig. 1, A-D, IL1␤ or TNF␣ alone caused a moderate increase in Cx43 protein and mRNA levels, while in combination, they exerted a significantly more potent effect. Therefore, the combined stimulation was used for all of the following experiments.

TNF␣. As shown in
In Western blots, Cx43 protein was detected as three bands (P0, P1, P2; Fig. 1A), corresponding to the nonphosphorylated (P0), phosphorylated (P1), and hyperphosphorylated forms (P2), respectively. IL1␤ and TNF␣ induced the native (P0) and phosphorylated forms of Cx43 protein. Time course analysis revealed that the effects of IL1␤/TNF␣ on Cx43 were time-dependent (Fig. 1E). The increased Cx43 levels could be detected 6 h after the stimulation and persisted for at least 48 h.
The elevating effect of IL1␤/TNF␣ on Cx43 expression was confirmed by immunofluorescence staining. Treatment of BSMCs with IL1␤/TNF␣ markedly increased the size and intensity of the punctate Cx43 staining. Normally, Cx43 was mainly localized in the perinuclear region. In the presence of the cytokines, more Cx43 was detected at the region of cell-tocell contacts (Fig. 1F).
Consistent with the increased localization of Cx43 at the cellular membrane, the GJIC was also increased as evaluated by the SLDT assay. As shown in the Fig. 1, G and H, the distance of LY diffusion (green) from the scrape line in IL1␤/TNF␣-treated cells was significantly further than that in control cells. Because of the large size, RD (Fig. 1, G and H, red) could not pass through GJs and served as an indicator of the scratch-loaded cells. The merged images are also shown, in which colocalization of LY (Fig. 1, G and H, green) and RD (red) yields yellow. These results indicate that proinflammatory cytokines increase Cx43 expression and promote GJIC in BSMCs.
cAMP Signaling Pathway Mediates IL1␤/TNF␣-elicited Cx43 Expression-Because cAMP is a well characterized second messenger in the induction of Cx43 and promotion of GJIC (22,31,32), one possible mechanism underlying the effect of IL1␤/TNF␣ could be the activation of this pathway. Indeed, exposure of BSMCs to IL1␤/TNF␣ caused a time-dependent increase in the phosphorylation of VASP at serine 157 ( Fig. 2A), a validated substrate of cAMP-dependent protein kinase, indicating the activation of PKA (33). In addition, the inhibition of PKA with the PKA inhibitor H89 largely suppressed the cytokine-induced phosphorylation of VASP and Cx43 (Fig. 2, B-D). Consistent with the above finding, H89 also inhibited GJIC, as  Activated PKA subsequently phosphorylates CREB, leading to the activation of genes that have cAMP-response elementbinding sites. Because there are cAMP-response element sites in the Cx43 gene (8,22), we evaluated their involvement by using CREB siRNA. As shown in Fig. 2, F-H, treatment of BSMCs with the siRNA effectively down-regulated the protein levels of cAMP-response element and concomitantly abolished the increase in Cx43. Therefore, these results indicate that the cAMP signaling pathway mediates the cytokine-induced upregulation of Cx43.
NFB Is Required for IL1␤/TNF␣-induced Cx43 Expression-NFB plays a pivotal role in the cellular inflammatory response (34). Therefore, we examined the possible involvement of NFB in the Cx43-elevating effect of the cytokines. As shown in Fig. 3, A and B, the inhibition of NFB with SC514 (50 M) largely blocked the effect of the cytokines on Cx43 expression (relative levels of Cx43, IL1␤/TNF␣, 2.69 Ϯ 0.10; IL1␤/TNF␣ ϩ SC514, 1.31 Ϯ 0.14; mean Ϯ S.D.; n ϭ 3). The enhanced GJIC was also suppressed by SC-514 ( Fig. 3C and supplemental Fig.  1A). These results indicate that the effect of IL1␤/TNF␣ on Cx43 occurs through the activation of NFB.
NO Contributes to the Cytokine-induced Cx43 Elevation-Among NFB-regulated gene products, iNOS has been reported to elevate Cx43 expression and function through the cAMP signaling pathway (22,35). Therefore, we examined the possible involvement of NO. Consistent with the cooperative effects of the cytokines on Cx43, IL1␤ and TNF␣ also synergistically elicited iNOS expression (Fig. 4A). In addition, iNOS expression was under the control of NFB. Inhibition of NFB with SC-514 completely abrogated iNOS expression (Fig. 4, B and C).

Reciprocal Regulation between iNOS and Connexin43
DECEMBER 2, 2011 • VOLUME 286 • NUMBER 48 TNF␣. Diphenyleneiodonium, a broad spectrum flavoprotein inhibitor whose targets also include nitric-oxide synthase (37,38), also significantly inhibited the Cx43-elevating effect. In addition, iNOS inhibitor L-NAME also significantly suppressed the GJIC, as evaluated by SLDT (Fig. 5C and supplemental Fig.  1A). Contrary to the suppressive effects of these inhibitors, the exogenous NO donor SNAP elevated VASP phosphorylation and Cx43 expression in a time-dependent manner, which was associated with an increased GJIC (Fig. 5, D and E, and supplemental Fig. 1B). These results indicate an involvement of NFB-regulated iNOS in the cytokine-induced up-regulation of Cx43.
Induction of Bladder Cx43 by Intraperitoneal Injection of LPS-To test whether bladder infection and inflammation alter GJs in vivo, we examined the levels of mouse bladder Cx43 after intra-peritoneal injection of LPS. As shown in Fig. 6, A and B, LPS induced a time-and dosage-dependent increase in Cx43 protein levels, together with enhanced expression of iNOS. Inhibition of iNOS with L-NAME also significantly blunted the elevation of Cx43 in vivo (Fig. 6, C and D).
Consistent with the elevated Cx43 protein levels, immunofluorescent analysis revealed an increased number of punctate Cx43 in the outer layer and smooth muscle layer of the bladder from LPS-treated mice (Fig. 6, E and F). These in vivo results indicate that NO mediates inflammation-triggered up-regulation of Cx43.
Inhibition of GJs Potentiates the Cytokine-induced Expression of iNOS in BSMCs-GJs have been implicated in the regulation of various inflammatory responses (12,15,17,20). We therefore asked whether the NO-mediated increase of Cx43 could The control mice were injected with saline. Bladder proteins were subjected to Western blot analysis for Cx43 and ␤-actin. C, inhibition of iNOS on Cx43 expression in vivo. Before LPS administration, mice were injected intraperitoneally with the iNOS inhibitor 20 mg/kg L-NAME every 12 h three times. D, densitometric analysis of Cx43 levels shown in C. Results are expressed as induction relative to the control (mean Ϯ S.D., n ϭ 3). #, p Ͻ 0.01. E, immunohistochemical staining of Cx43 in the bladder. Mice were injected intraperitoneally with 20 mg/kg LPS for 24 h. Bladder cryosections were subjected to immunofluorescent staining of Cx43 (green), phalloidin (red), and nuclei (blue). Note the obvious increased number and intensity of punctate Cx43 in the smooth muscle layer (long arrow) as well as bladder outer layer (short arrow) of LPS-treated mice. F, quantitative analysis of Cx43 density shown in E. The Cx43 density is expressed as the average number of Cx43 spots per field (magnification, ϫ1200) from four separate experiments (mean Ϯ S.E., n ϭ 4; #, p Ͻ 0.01 versus control). For each experiment, an average of five fields were counted. reciprocally affect iNOS expression and NO production. To this end, we examined the change of iNOS and NO after GJ dysfunction caused by three structurally different GJ blockers. As shown in Fig. 7, GJ blockers ␣-GA, lindane, and heptanol all significantly potentiated the cytokine-triggered expression of iNOS and production of NO. This observation thus suggests that GJs negatively regulate iNOS expression under inflammatory conditions. Cx43 ϩ/ϩ and Cx43 Ϫ/Ϫ Fibroblasts Display Different Responses to IL1␤/TNF␣-induced iNOS Expression-To further establish the role of Cx43-forming channels in the regulation of iNOS, we have compared the cytokine-induced iNOS expression between fibroblasts derived from Cx43 wild-type (Cx43 ϩ/ϩ ) and knock-out (Cx43 Ϫ/Ϫ ) littermates. In previous reports, we and others have described that Cx43 ϩ/ϩ fibroblasts expressed abundant Cx43 protein and were functionally coupled by GJs, whereas Cx43 Ϫ/Ϫ fibroblasts had neither Cx43 nor functional GJIC (25)(26)(27)(28).
First, we confirmed our findings in Cx43 ϩ/ϩ fibroblasts. As shown in Fig. 8A, IL1␤/TNF␣ also increased Cx43 expression in fibroblasts. This effect of the cytokines was similarly associated with a parallel increase in the protein levels of iNOS and phosphorylated VASP. In addition, dysfunction of Cx43-forming channels in these cells also influenced the cytokine-elicited expression of iNOS. As shown in Fig. 8, B and C, treatment of Cx43 ϩ/ϩ fibroblasts with GJ blockers or specific siRNA amplified the effect of the cytokines on iNOS expression. The effectiveness of Cx43 siRNA in the down-regulation of Cx43 protein levels is shown in Fig. 8C, middle panel. Finally, we compared the iNOS expression and NO production between Cx43 ϩ/ϩ and Cx43 Ϫ/Ϫ fibroblasts. Consistent with the above results, IL1␤/TNF␣ induced a significantly higher level of iNOS and NO in Cx43 Ϫ/Ϫ than Cx43 ϩ/ϩ fibroblasts (Fig. 8, D-F). Collectively, these results indicate that Cx43-forming channels negatively regulate cytokine-induced iNOS expression.

DISCUSSION
In this study, we showed for the first time that the proinflammatory mediators (IL1␤, TNF␣, and LPS) potently increased Cx43 expression and function in BSMCs. In addition, we established Cx43-forming channels as a negative feedback mechanism in the control of iNOS expression. The major findings and signaling mechanisms involved are schematically depicted in Fig. 9. We have chosen the bladder and cultured BSMCs as a model system for this investigation because the presence of GJs in the bladder and their roles in the regulation of bladder contractility have been well documented. In addition, bladder infection is a common disease. Unraveling the interactive regulation between inflammation and GJs in the bladder could have important clinical and scientific implications.
Exposure of the bladder to LPS markedly elevated Cx43 expression. This effect was similarly achieved by IL1␤ and TNF␣ in cultured BSMCs. These observations indicate that bladder infection and inflammation alter GJs in BSMCs. Most of the cellular inflammatory responses are known to be governed by the transcription factor NFB (34). It was also true for the cytokine-induced expression of Cx43, as shown in this study. Inhibition of NFB with SC514 completely abolished the increase in Cx43 expression. Several previous papers described the presence of an NFB-binding site in the promoter region of the Cx43 gene (39,40). However, the functional role of this site has not been firmly established. In this study, the NFB-mediated regulation of Cx43 was due to the NFB-controlled gene products. This notion is supported by the observation that the inhibition of NFB or the NFB-regulated gene product iNOS significantly abrogated the Cx43-elevating effects.
The induction of iNOS under inflammatory situations and the implication of NO in the up-regulation of Cx43 have been previously documented (22,35). Therefore, an involvement of iNOS in elevating Cx43 expression under inflammatory situations is not surprising. The evidence supporting the mediating role of NO also includes that the following. 1) iNOS was under the control of NFB in BSMCs.
2) The level of iNOS induced by IL1␤/TNF␣ was in parallel to their potency in inducing Cx43 expression.
3) The inhibition of iNOS significantly attenuated the effect of the cytokines. 4) The exog- M lindane and 4 mM heptanol for 30 min and exposed to 2 ng/ml IL1␤ and 20 ng/ml TNF␣ for an additional 24 h. Cellular proteins were subjected to Western blot analysis for iNOS and ␤-actin. B, densitometric analysis of iNOS expression shown in A. Results are expressed as relative levels compared with the control (mean Ϯ S.D., n ϭ 3). *, p Ͻ 0.05, and #, p Ͻ 0.01 versus control. C, potentiation of IL1␤/TNF␣-elicited NO release by GJ blockers. BSMCs were treated the same as above. The conditioned media were harvested and measured for nitrite concentration. Results are expressed as relative levels compared with the basal NO 2 concentration (mean Ϯ S.D., n ϭ 4). *, p Ͻ 0.05, and #, p Ͻ 0.01 versus IL1␤/TNF␣. enous NO donor SNAP similarly induced the phosphorylation of VASP and Cx43.
In addition to BSMCs, IL1␤/TNF␣ also elevated Cx43 protein levels in glomerular mesangial cells, renal tubular epithelial cells (data not shown), and fetal forearm fibroblasts. In addition, Cx43 in the bladder epithelial cells was also induced by intraperitoneal injection of LPS. Interestingly, the combined stimulation of DCs and macrophages with LPS, TNF␣, and IFN-␥ also increased Cx43 expression in these cells (12,41). It is possible that similar regulatory mechanisms operate in these cells. Of note, in contrast to our results, there are also reports documenting a suppressive effect of proinflammatory mediators (LPS, TNF␣, and NO) on GJs (42,43). The reasons for this discrepancy are presently unclear. It could be due to the different cell types and experimental settings used for the investigation. More detailed analysis to account for the difference may be needed in the future.
The increased Cx43 expression by NO could contribute to the altered sensitivity of bladder detrusor to stimuli. In a recent report, E. coli-induced bladder inflammation was associated with an increased bladder contractile response, which was mediated by the NO-induced activation of PKC (44). Given that GJs are required for the coordinated contractile response (45), the elevation of Cx43 by NO in BSMCs could exaggerate the abnormal bladder contractile response under inflammatory situations.
There is growing evidence indicating a critical involvement of GJs in immune inflammatory processes. Depending on the cell types, isoforms of Cxs and employed stimuli, GJs could either promote or suppress the inflammatory responses (20,46,47). For example, in DC cells, we have demonstrated that GJs contribute to their activation and production of inflammatory mediators (14). In contrast, Cx32 in endothelial cells has been recently documented to suppress TNF␣-induced production of FIGURE 8. Different response to IL1␤/TNF␣-induced iNOS expression in Cx43 ؉/؉ and Cx43 Ϫ/Ϫ fibroblasts. A, time-dependent effect of IL1␤/TNF␣ on Cx43, iNOS, and pVASP expression in Cx43 ϩ/ϩ fibroblasts. Cx43 ϩ/ϩ fibroblasts were exposed to 2 ng/ml IL1␤ and 20 ng/ml TNF␣ for the indicated duration. Cellular protein was subjected to Western blot analysis for iNOS, pVASP, and Cx43. ␤-Actin was used as a loading control. B, effects of various GJ blockers on iNOS expression. Cx43 ϩ/ϩ fibroblasts were pretreated with 20 M ␣-GA, 100 M lindane, and 4 mM heptanol for 30 min and exposed to 2 ng/ml IL1␤ and 20 ng/ml TNF␣ for an additional 24 h. The cellular proteins were extracted and assayed for iNOS and ␤-actin levels by Western blot analysis. C, effect of Cx43 siRNA on iNOS expression. Cx43 ϩ/ϩ fibroblasts were transfected with control or Cx43 siRNA for 24 h. The cells were then exposed to IL1␤/TNF␣ for an additional 24 h. Cellular proteins were analyzed for iNOS, Cx43, and ␤-actin. D, different induction of iNOS by IL1␤/TNF␣ in Cx43 ϩ/ϩ and Cx43 Ϫ/Ϫ fibroblasts. Cx43 ϩ/ϩ and Cx43 Ϫ/Ϫ fibroblasts were exposed to 2 ng/ml IL1␤ and 20 ng/ml TNF␣ for the indicated duration. The cellular proteins were assayed for iNOS, Cx43, and ␤-actin levels by Western blot analysis. E, densitometric analysis of iNOS expression shown in D. Results were expressed as levels relative to the zero point control (mean Ϯ S.D., n ϭ 3). #, p Ͻ 0.01. F, induction of NO release by IL1␤/TNF␣ in Cx43 ϩ/ϩ and Cx43 Ϫ/Ϫ fibroblasts. Cx43 ϩ/ϩ and Cx43 Ϫ/Ϫ fibroblasts were exposed to 2 ng/ml IL1␤ and 20 ng/ml TNF␣ for the indicated duration. The conditioned media were harvested and assayed for nitrite concentration. Results are expressed as level relative to the basal NO 2 concentration in Cx43 ϩ/ϩ cells (mean Ϯ S.D., n ϭ 4). #, p Ͻ 0.01. IL-6 and MCP-1 (48). Here, we established Cx43-forming channels as a negative feedback mechanism for cytokine-induced iNOS expression. Currently, the molecules and signaling mechanisms involved in the regulatory effects of GJs on iNOS are still unclear. GJs are known to regulate cell functions through GJICdependent or -independent actions. The suppressive effect of GJs on iNOS could be related to the increased intercellular communication because the GJ inhibitor heptanol that disrupts GJIC by decreasing membrane fluidity without affecting Cx43 (49,50) also significantly potentiated the cytokine-induced iNOS expression. The molecules transmitted by GJs remain to be characterized. The NO-centered regulatory loop between inflammation and GJs, as revealed in this study, might provide a molecular link between intercellular and paracrine pathways under inflammatory situations, which may have an important impact on the initiation and development of inflammatory disorders.
Collectively, we established a reciprocal regulation loop between iNOS and GJs in BSMCs. Our findings thus provide a potentially important molecular mechanism for urgency symptoms of bladder infection. In addition, it may open a new window toward our understanding of the role of GJs in the pathogenesis of inflammatory disorders.