Abstract
It was previously shown that cyclosporine A (CsA) increases transepithelial resistance in MDCK cells. Activation of the extracellular signal–regulated kinase 1/2 (ERK1/2) mitogen-activated protein kinase (MAPK) cascade seems to be pivotal to the CsA-induced increase in transepithelial electrical resistance (TER). This study examined the role played by TGF-β in mediating the CsA-induced activation of ERK1/2 and the resulting increase in TER in MDCK cells. Paracellular permeability across MDCK monolayers after various treatments was assessed by measurement of TER. TGF-β secretion was measured by Western blot and ELISA. Activation of the ERK1/2 pathway and tight junction protein expression were also assessed by Western blot analysis. CsA increased production and secretion of TGF-β and expression of the TGF-β receptor II. Exogenous addition of TGF-β1 activated ERK1/2 and increased TER across MDCK monolayers, both of which were attenuated by the MEK inhibitor U0126. Neutralizing antibodies against TGF-β1 and the TGF-β receptor II significantly reduced the CsA-induced increase in TER. Both CsA and TGF-β1 increased expression of tight junction proteins claudin-1 and zonula occludens 2. Inhibition of the p38 MAPK pathway also attenuated the TGF-β1–induced increase in TER. The results presented here suggest that the CsA-induced modulation of paracellular permeability may be mediated, at least in part, by an increase in TGF-β production.
Cyclosporine A (CsA) is an immunosuppressive drug that is used in organ transplantation and in the treatment of some primary renal and nonrenal immune-mediated diseases. However, its clinical use is limited by adverse effects, the most important of which is nephrotoxicity (1). This condition is manifested by renal insufficiency as a result of glomerular and vascular changes, abnormalities of tubular function, and hypertension. Renal biopsy reveals arteriolar hyalinosis, focal interstitial fibrosis, tubular atrophy, and glomerulosclerosis with focal atrophy. Mechanisms of CsA-induced nephrotoxicity were thought to be hemodynamic in origin, but there is accumulating evidence for a direct tubular effect. CsA-induced tubular damage is not well understood; however, alterations in tubular morphology have been seen in acute studies in rats (2–4).
Epithelial tissues serve as selective permeability barriers, separating fluid compartments with different chemical compositions. Renal epithelium tight junctions mediate this barrier role and seal cells together to impede the paracellular leakage of small molecules. In renal tubules, the paracellular pathway plays an important role in vectorial transport with some selectivity for transported ions such as magnesium, calcium, sodium, and chloride. In mammalian kidney, the complexity of the tight junction increases from the proximal to the collecting tubule. In mammalian kidney, the transepithelial resistance (TER) and complexity of the tight junctions increase from the proximal tubule along the nephron to the distal tubule. Examination of the differential expression of tight junction proteins along isolated rabbit renal tubules showed that the levels of zonula occludens 1 (ZO-1), ZO-2, and occludin increased from the proximal tubules to the distal tubules, mirroring the increase in complexity of the junctions (5). Tight junctions are subject to physiologic and pharmacologic regulation/modulation. We previously showed that CsA alters the barrier function of renal epithelial cells, and the mechanism may involve activation of the extracellular signal–regulated kinase 1/2 (ERK1/2) mitogen-activated protein kinase (MAPK) pathway (6).
CsA stimulates renal and systemic overproduction of TGF-β, which is thought to be an important factor in the development of CsA chronic nephrotoxicity (7–9). More than 70% of the renal biopsies from CsA-treated transplant patients with chronic allograft fibrosis expressed high levels of TGF-β (10). Using Affymetrix gene microarrays, we demonstrated upregulation of TGF-β in human renal tubular cells that were treated with CsA (11). TGF-β has been implicated in the modulation of paracellular permeability. TGF-β1 enhanced the barrier function of intestinal epithelial monolayers and promoted intestinal epithelial restitution (12–16). A recent study demonstrated that TGF-β lowered endothelial permeability in mouse brain capillary endothelial cells, suggesting that cellular constituents that produce TGF-β in the brain may keep the blood–brain barrier functioning (17). TGF-β1 not only upregulates colonic epithelial barrier function but also curtails the effects of barrier-reducing cytokines such as IFN-γ, IL-4, and IL-10 (12). In contrast, TGF-β1 prevented glucocorticoid-stimulated tight junction formation and reduced TER in murine mammary epithelial cells (18). Therefore, it is probable that TGF-β1 may produce converse effects on epithelial barrier function in various tissues.
There is evidence linking activation of the MAPK pathways to the TGF-β–induced modulation of tight junctions and paracellular permeability. Co-stimulation of thyroid epithelial cells with EGF and TGF-β1 drastically reduced TER and increased paracellular flux of inulin (19). Reduced levels of claudin-1 and occludin accompanied the loss of barrier function. The MEK inhibitor U0126 prevented residual ERK phosphorylation and abrogated the synergistic responses to TGF-β1 and EGF. TGF-β enhanced epithelial barrier function of T84 and HT-29 monolayers up to three-fold (16). The barrier function could be restored by treatment with SB203580, an inhibitor of the p38 MAPK pathway, but not by inhibitors of c-Jun N-terminal kinase, ERK1/2 MAPK, or phosphatidylinositol 3-kinase (16). In contrast, TGF-β3 reduced Sertoli cell tight junction function (20). This TGF-β3–mediated effect on tight junction barrier and the TGF-β3–induced phospho-p38 MAPK production could be blocked by a specific p38 MAPK inhibitor but not by a specific MEK1/2 kinase inhibitor (21). These results demonstrate that TGF-β3 uses the p38 MAPK pathway to regulate Sertoli cell tight junction dynamics. In this study, we examined whether the CsA-induced effects on paracellular permeability are mediated by TGF-β.
Materials and Methods
Cell Culture and Treatment
MDCK II cells (22) that were obtained from the American Type Culture Collection (Middlesex, UK) were cultured as before (6). CsA was prepared as a stock solution (4.2 mM) in 100% EtOH. U0126 and SB203580 were prepared as stock solutions of 10 mM in DMSO. TGF-β1 was reconstituted with 4 mM HCl that contained 1 mg/ml BSA to a stock of 5 μg/ml. TGF-β1–neutralizing antibody (nAb) was prepared as a stock solution of 500 μg/ml in PBS. TGF-β receptor II (TGF-β-RII) nAb was prepared as a stock solution of 0.25 mg/ml in PBS. Cells were preincubated for 1 h with inhibitors before treatment with CsA.
TER
The intactness of paracellular pathways of the MDCK monolayers to small ions was monitored by measurement of TER using the REMS automated system (World Precision Instruments, Stevenage, UK). Cells were grown on Costar HTS-Transwell (Corning, NY) cell culture inserts (pore size 0.4 μm). MDCK cells that were seeded at a density of 1 × 105 cells/ml were seen to reach a stable TER, representing a confluent monolayer, 5 d after seeding. Cells were exposed, on the apical side only, to the various treatments. The point of drug addition was taken as time 0, and TER was monitored over a range of times. TER was normalized to the area of the filter after removal of background resistance of a blank filter that contained only medium. TER was thus measured as ohms × cm2 (Ω·cm2) and results were expressed as the change in TER with respect to time matched controls [ΔTER (Ω·cm2)].
FITC-Dextran Flux Measurements
The intactness of the paracellular pathway across monolayers was assessed by measurement of FITC-dextran flux. MDCK cells were seeded at a density of 1 × 105 cells/ml onto Costar Transwell filters. After treatment, medium was replaced with serum-free medium that contained FITC-dextran on the apical surface. After an appropriate incubation period, the FITC fluorescence in each well of the base plate was determined by using a Wallac Victor V plate reader (Perkin Elmer Life Sciences, Waltham, MA). The results were expressed as mean FITC flux (ng/cm2 per h).
Cellular Viability
Confluent monolayers were treated for appropriate periods of time. Monolayers were incubated in 10% AlamarBlue reagent (Biosource, Nivelles, Belgium) at 37°C for 2 h. Fluorescence was read at 545 nm on a Wallac Victor V multiwell plate reader. The amount of fluorescence detected is proportional to the percentage of viable cells.
Cellular Proliferation
Confluent monolayers were treated for appropriate periods of time. Cell numbers were determined using the CyQUANT cell proliferation assay (Molecular Probes, Eugene, OR) according to the supplied protocol.
Preparation of Whole Extracts for SDS-PAGE
Whole-cell extracts were prepared as before (6). Briefly, after appropriate treatment, cells were scraped into radioimmunoprecipitation assay buffer (20 mM Tris [pH 7.4], 50 mM NaCl, 5 mM EDTA, 50 mM NaF, 20 mM sodium pyrophosphate, 1 mM orthovanadate, 1% [vol/vol] Triton X-100, 0.1% [wt/vol] SDS, 1 mM PMSF, 1 μg/ml pepstatin, 0.5 μg/ml leupeptin, and 0.5 μg/ml aprotinin). Protein concentration was determined using a BCA protein assay kit (Pierce, Rockford, IL).
Preparation of Samples for TGF-β Protein
Supernatants were collected, and 9 μl of Strataclean resin was added to 1 ml of each supernatant sample. After centrifugation, supernatants were discarded and the pellets were resuspended in 10 μl of 3× sample buffer. Proteins were denatured by boiling at 100°C for 3 min.
SDS-PAGE and Western Blotting
Equal amounts of cell extracts were electrophoresed on SDS-polyacrylamide gels, and proteins were transferred to nitrocellulose membrane. For ensuring equal loading, membranes were stained with Ponceau-S. Membranes were blocked with 5% (wt/vol) milk proteins/Tris-buffered saline and incubated overnight at 4°C with the primary antibody: rabbit anti-TGF-β (Pan specific), mouse anti-TGF-β-RII, rabbit anti-phospho-ERK1/2, rabbit anti-occludin, rabbit anti–junctional adhesion molecule-A (JAM-A), rabbit anti–atypical PKC isotype-specific interacting protein/partitioning defective 3 (APIS/PAR3), rabbit anti–claudin-1, rabbit anti–claudin-2, rabbit anti–claudin-3, rabbit anti–claudin-4, rabbit anti–claudin-16, rabbit anti–ZO-1, or rabbit anti–ZO-2. Bound antibody was detected with appropriate secondary antibodies and enhanced chemiluminescence.
TGF-β1 ELISA
After treatment of confluent monolayers, 100 μl of supernatant was transferred to a 96-well microtiter plate. This plate had been prepared by coating it with a monoclonal mouse TGF-β1 antibody and blocking with 4% (wt/vol) BSA/PBS overnight at 4°C. The supernatant or TGF-β1 standards (recombinant human TGF-β1) were incubated for 2 h. After washing, each well was incubated with the biotinylated secondary antibody, the streptavidin-conjugated horseradish peroxidase detection reagent, and the substrate solution. The reaction was halted by the addition of 1.8 M H2SO4. Absorbance was measured at 450 nm using a Wallac Victor V plate reader. The mean absorbance of each standard was plotted against TGF-β1 concentration to allow quantification of TGF-β1 (pg/ml).
Statistical Analyses
Statistical analyses were performed using the statistical program GraphPad Prism 2.1 (GraphPad, San Diego, CA). Data were analyzed by one-way ANOVA, and multiple comparisons between control and treatment groups were made using the Dunnett posttest. Comparisons between treatments were made using the Bonferroni posttest. Alternatively, the unpaired t test was used to test for statistical significance. Results were expressed as means ± SEM. P ≤ 0.05 was deemed statistically significant.
Results
Effect of CsA on TGF-β Secretion and TGF-β-RII Expression
Treatment of MDCK cells with 4.2 μM CsA resulted in a significant increase in TGF-β secretion at 24, 48, and 72 h (Figure 1A), as demonstrated by Western blotting using a PAN-specific antibody. There was also a significant increase in TGF-β1 secretion, as determined by a specific ELISA (Figure 1B). Because production of the TGF-β1 isoform was increased by CsA treatment, we investigated the effect of CsA on TGF-β-RII expression: Treatment with 4.2 μM CsA significantly increased TGF-β-RII expression after 24 and 72 h (Figure 1C).
TGF-β expression and TGF-β receptor II (TGF-β-RII) expression in MDCK cells. (A) Western blot analysis of TGF-β secretion after cyclosporin A (CsA) treatment. Confluent monolayers were treated with vehicle or 4.2 μM CsA for 24, 48, and 72 h. Supernatants were collected, and proteins were separated electrophoretically on SDS-polyacrylamide gel and transferred to nitrocellulose membrane. Expression of TGF-β was identified using a polyclonal rabbit anti–TGF-β (Pan-specific) antibody and appropriate secondary reagents. (B) TGF-β1 secretion as measured by ELISA. Confluent monolayers were treated with vehicle (□) or 4.2 μM CsA (□) for 72 h. Supernatants were collected at various time points and assayed for TGF-β1. *P < 0.05, **P < 0.01, and ***P < 0.001 versus time-matched control. (C) Confluent monolayers were treated with vehicle or 4.2 μM CsA for 24 and 72 h. Whole-cell lysates were then made using radioimmunoprecipitation assay (RIPA) lysis buffer. Cellular proteins were separated electrophoretically on SDS-polyacrylamide gel and transferred to nitrocellulose membrane. Expression of TGF-β-RII was identified using a polyclonal mouse anti–TGF-β-RII antibody and appropriate secondary reagents.
Effect of TGF-β1 on TER
Exposure of MDCK cells to TGF-β1 (5 or 10 ng/ml) significantly increased TER at 24 h after treatment, and this effect was maintained up to 72 h (Figure 2A). Treatment with a TGF-β1–neutralizing antibody (10, 20, or 30 μg/ml) resulted in no significant changes of basal TER over 72 h (data not shown). However, co-treatment of the monolayers with 4.2 μM CsA plus 30 μg/ml TGF-β1 nAb significantly reduced the CsA-induced increase in TER at 24 h, and this effect was maintained up to 72 h (Figure 2B). Treatment with a TGF-β-RII nAb (10, 50, or 100 μg/ml) did not significantly alter the basal TER (data not shown). However, co-treatment with CsA (4.2 μM) and TGF-β-RII nAb (100 μg/ml) resulted in significantly lower TER readings at 24, 48, and 72 h compared with treatment with CsA alone (Figure 2C). Treatment with the TGF-β-RII nAb did not abolish the CsA-induced increase in TER but only partially ameliorated the CsA-induced effect.
Effect of TGF-β1 on transepithelial electrical resistance (TER). (A) Confluent monolayers were treated, on the apical side alone, with increasing concentrations of TGF-β1. (B) Confluent monolayers were pretreated for 1 h with 30 μg/ml TGF-β1–neutralizing antibody (nAb) before treatment with 4.2 μM CsA, from the apical side. (C) Confluent monolayers were pretreated for 1 h with 100 μg/ml TGF-β-RII nAb before apical treatment with 4.2 μM CsA. TER was monitored over 72 h. Results are expressed as the change in TER (ΔTER) compared with time-matched control filters and are given as means ± SEM of five independent experiments, each performed in duplicate. *P < 0.05, **P < 0.01, and ***P < 0.001 versus time-matched control or between different treatments, when indicated.
Effect of CsA on FITC-Dextran Flux, Cell Proliferation, and Cell Viability
When confluent monolayers were exposed to CsA for periods up to 72 h from the apical side only, there was no significant change in the rate of flux of 4000, 10,000, or 20,000 FITC-dextran across the monolayers (Figure 3A).
Effect of CsA on MDCK cell monolayers. (A) Flux of FITC-dextran across intact monolayers was assessed after CsA treatment. After treatment, medium on the apical side was replaced with 20 μg/ml 4000, 10,000, or 20,000 FITC-dextran, and the fluorescence levels of the basolateral medium were read after 90 min. Results are expressed as ΔFITC flux (ng/cm2 per h) compared with time-matched control filters and are given as means ± SEM of three independent experiments performed in triplicate. (B) Cellular proliferation was assessed using the CyQUANT cell proliferation assay. (C) Cellular viability was assessed using the AlamarBlue assay. Results are expressed as percentage of control (100%) and are given as means ± SEM of three independent experiments, each performed in triplicate. 95% confidence intervals were constructed, and a t test was performed to assess statistical significance.
MDCK cells were counted using the CyQUANT cell proliferation assay after treatment with 4.2 μM CsA (Figure 3B). There was a small but significant decrease in cell proliferation in CsA-treated cells at all time points: 24, 48, and 72 h. We also noted that TGF-β1 (5 and 10 ng/ml) significantly reduced MDCK cell proliferation at 48 and 72 h (data not shown).
The viability of MDCK cells was analyzed after treatment for 24, 48, and 72 h with 4.2 μM CsA (Figure 3C). CsA had no significant effect on the cellular viability at any time points. Cell viability was significantly reduced at 72 h after treatment with TGF-β1 (5 and 10 ng/ml; data not shown).
Effect of CsA on ERK1/2 Activation
Treatment of MDCK cells with CsA (4.2 μM), for periods up to 72 h, resulted in significant activation of ERK1/2 signaling as demonstrated by Western blotting using activated ERK-specific antibody (Figure 4). Significant activation of ERK1/2 was observed at 3, 24, and 72 h of exposure (Figure 4).
Western blot analysis of phosphorylated extracellular signal–regulated kinase 1/2 (ERK 1/2) expression in MDCK II cells in the presence or absence of CsA. MDCK II cells were grown to confluence on Falcon six-well plates and treated as described for up to 72 h. Cells were treated with 4.2 μM CsA. Whole-cell lysates were then made, using RIPA lysis buffer, at 30 min (A), 3 h (B), 24 h (C), and 72 h (D) after treatment. Cellular proteins were separated electrophoretically on SDS-polyacrylamide gel and transferred to nitrocellulose membrane. Expression of phospho-ERK1/2 was identified using a polyclonal rabbit anti–phospho-ERK1/2 antibody and an anti-rabbit horseradish peroxidase (HRP)-conjugated secondary antibody. Representative blots from one of three separate experiments are shown. Lane 1, control; lane 2, 4.2 μM CsA.
TGF-β1 Activates ERK1/2
Both CsA (4.2 μM) and TGF-β1 (5 ng/ml) activated ERK1/2 (Figure 5). The MEK inhibitor U0126 reduced basal activation of ERK1/2 and also prevented the CsA-induced and the TGF-β1–induced activation of ERK1/2 (Figure 5). No effects of ERK1/2 activation were seen after treatment with a control IgG antibody for a period up to 72 h (data not shown).
Western blot analysis of phospho-ERK1/2 expression after treatment with CsA or TGF-β1. Monolayers were pretreated for 1 h with U0126 (10 μM) or TGF-β1 nAb (30 μg/ml) before treatment with 4.2 μM CsA or 5 ng/ml TGF-β1 for periods up to 24 h. Whole-cell lysates were made using RIPA lysis buffer, and proteins were separated electrophoretically on SDS-polyacrylamide gel and transferred to nitrocellulose membrane. Expression of phospho-ERK1/2 was identified using a polyclonal rabbit anti–phospho-ERK1/2 antibody and appropriate secondary reagents. Lane 1, control; lane 2, 4.2 μM CsA; lane 3, 5 ng/ml TGF-β1; lane 4, 10 μM U0126; lane 5, 10 μM U0126 plus 4.2 μM CsA; lane 6, 10 μM U0126 plus 5 ng/ml TGF-β1.
Effect of the MEK Inhibitor U0126 and the p38 Inhibitor SB203580 on TER
U0126 (10 μM) induced a significant decrease in TER. Co-treatment of the MDCK cells with 5 ng/ml TGF-β1 plus 10 μM U0126 significantly reduced the TGF-β1–induced increase in TER that was observed when cells were treated with TGF-β1 alone (Figure 6A).
Role of ERK1/2 and p38 mitogen-activated protein kinase (MAPK) pathways in TGF-β1–induced modulation of TER. (A) Confluent monolayers were pretreated for 1 h with 10 μM U0126 (MEK1/2 inhibitor) before apical treatment with 5 ng/ml TGF-β1. (B) Confluent monolayers were pretreated for 1 h with 10 μM SB203580 (p38 MAPK inhibitor) before apical treatment with 5 ng/ml TGF-β1. TER was monitored over 72 h, and results are expressed as ΔTER compared with time-matched control filters, given as means ± SEM of five independent experiments, each performed in duplicate. *P < 0.05, **P < 0.01, and ***P < 0.001 versus time-matched control or between different treatments, when indicated.
The p38 inhibitor SB203580 (10 μM) did not significantly alter basal TER. However, co-treatment with 5 ng/ml TGF-β1 and 10 μM SB203580 significantly reduced the TGF-β1–induced increase in TER at 48 and 72 h (Figure 6B).
Effect of CsA on the Expression of Tight Junction–Associated Proteins
The effect of CsA (4.2 μM) on expression of various tight junction–associated proteins is shown in Figures 7 and 8. Western blot analysis showed increased occludin expression after 24 and 72 h of treatment (Figure 8). Expression of JAM-A significantly increased after 72 h of treatment, whereas there was no significant change in the level of ASIP/PAR3 after treatment for 72 h (Figure 7). Expression levels of claudin-1 (Figure 8) and claudin-3 significantly increased after treatment with CsA for 24 h (Figure 7). However, there was no detectable change in expression of claudin-2, claudin-4, or claudin-16 after 24 h of CsA treatment. ZO-1 expression levels did not change after 24 h of CsA treatment; however, levels of ZO-2 were significantly increased after 72 h (Figure 7).
Expression of tight junction proteins after CsA treatment. Confluent monolayers were treated with CsA (4.2 μM) for periods of 24 or 72 h. Whole-cell lysates were made using RIPA lysis buffer. Cellular proteins were separated electrophoretically on SDS-polyacrylamide gel and transferred to nitrocellulose membrane. Expression of tight junction proteins was identified using specific antibodies. A representative blot from one of three separate experiments is shown.
(A) Effect of CsA on occludin, claudin-1, and claudin-2 expression in whole-cell MDCK II cellular lysates at 24 h. Cells were grown on Falcon petri dishes. When confluent, they were exposed to vehicle (EtOH) or 4.2 μM CsA for 24 h. Whole-cell lysates were then made using RIPA lysis buffer. Cellular proteins were separated electrophoretically on SDS-polyacrylamide gel and transferred to nitrocellulose membrane. Expression of the proteins was identified using a polyclonal rabbit anti–claudin-1 or -2 or occludin antibody and an anti-rabbit HRP-conjugated secondary antibody. (i) Representative blot
Effect of TGF-β1 on Tight Junction Protein Expression
Having demonstrated that CsA altered expression of some but not all tight junction–associated proteins, we decided to examine the effect of TGF-β1 on ZO-2 and claudin-1 expression (Figure 9). Treatment of MDCK cells for 72 h with 5 ng/ml TGF-β1 significantly increased the expression of the integral tight junction–associated protein claudin-1 and the peripheral tight junction–associated protein ZO-2 (Figure 9).
Tight junction protein expression after treatment with TGF-β1. Confluent monolayers were treated with 4.2 μM CsA or 5 ng/ml TGF-β1 for 72 h. Whole-cell lysates were then made using RIPA lysis buffer. Cellular proteins were separated electrophoretically on SDS-polyacrylamide gel and transferred to nitrocellulose membrane. Expression of claudin-1 was identified using a polyclonal rabbit anti–claudin-1 antibody, whereas expression of zonula occludens 2 (ZO-2) was identified using a polyclonal rabbit anti–ZO-2 antibody. Band intensity was quantified using densitometric analysis (n = 3). **P < 0.01 and ***P < 0.001 versus time-matched control or between different treatments, when indicated.
Discussion
Regulation of cellular barrier permeability is a vital and complex process that involves intracellular signaling and rearrangement of tight junction proteins. We investigated the signaling mechanism that is triggered by CsA and results in enhanced tight junction barrier function. Our data suggest that the CsA-induced increase in TER may be mediated, at least in part, by an increase in TGF-β production.
Cytokines, such as TGF-β, are known to perturb tight junctions in retinal endothelial cells, 31EG4 cells, and Sertoli cells in vitro (23). Chronic kidney fibrosis has been associated with the ability of CsA to increase production of TGF-β1 (9). We have demonstrated that CsA increased TGF-β production and TGF-β1 secretion by MDCK cells. Although control cells seemed to secrete relatively high levels of TGF-β1, we observed that treatment with CsA significantly upregulated TGF-β-RII expression, and this may increase the overall available number of receptors for TGF-β1 to bind to in CsA-treated cells. We subsequently showed that the CsA-induced increase in TER could be partially prevented by simultaneously blocking the bioactivity of TGF-β1 or blocking the bioactivity of the TGF-β-RII with neutralizing antibodies. This finding indicates that TGF-β1 plays a part in the CsA-induced increase in tight junction barrier function.
Two parameters of tight junction permeability, TER and paracellular flux of uncharged molecules, are considered to measure similar characteristics of tight junctions. Although we observed an increase in TER (decrease in paracellular permeability to small ions) upon treatment with CsA, there was no significant change in the rate of FITC-dextran flux across the monolayers. However, paracellular flux and TER are not always inversely correlated, especially in leaky epithelia (24). As cell density increases, there is an increase in intracellular space per unit area of monolayer. There have been several conflicting theories regarding the relationship between tight junction function and cell proliferation. Some investigators reported that a rise in cell number could generate an increase in TER (25–28). However, contrasting reports suggested that overgrowth of cells could result in enhanced transepithelial permeability. For example, the tumor-promoting phorbol ester increased the passage of a range of solutes across the epithelium with a concurrent increase in cell proliferation (26). Because CsA seemed to induce a small reduction in cell number with a concomitant decrease in transepithelial permeability, it is conceivable that there were fewer tight junctions, which subsequently restricted transepithelial electrical conductance. However, the viability assay showed no significant effects of CsA on MDCK cells. Therefore, other explanations for the increase in TER were investigated.
Because TGF-β1 has been reported to activate ERK1/2 in epithelial cells (29) and we previously demonstrated upregulation of TGF-β in response to CsA in renal tubular cells (11), we investigated the possibility that CsA activates ERK1/2 via the TGF-β1 isoform in our model system. TGF-β1 activated ERK1/2 within 30 min of treatment. Because TGF-β1 secretion was activated within 3 h of exposure to CsA, this supports our hypothesis that CsA may be altering tight junction barrier function, at least partially, via the TGF-β1-ERK1/2 route. TGF-β1 also significantly increased the TER across MDCK monolayers. The magnitude of this increase was similar to that observed when MDCK cells were treated with CsA. The TGF-β1–induced increase confirms that TGF-β1 is a promoter of tight junction integrity in MDCK cells and that CsA is potentially acting via this TGF-β isoform. Like the CsA-induced increase in TER, the TGF-β1–induced increase in TER was attenuated when cells were co-treated with U0126. This suggests that both CsA and TGF-β1 are mediating their effects via the ERK1/2 MAPK signaling pathway.
Recent in vivo studies revealed that TGF-β3 is a crucial regulator of blood–testis barrier (BTB) dynamics in the rat (21,30). For example, when CdCl2 was administered to adult rats, the disruption of Sertoli cell tight junction at the BTB and the subsequent germ cell loss from the epithelium were associated with a transient surge in TGF-β3 (21). This subsequently activated the p38 MAPK downstream as well as disruption and loss of occludin and actin filaments at the BTB (31). Therefore, we investigated whether TGF-β1 was partly mediating its effects on TER via the p38 MAPK pathway. Inhibition of this pathway significantly attenuated the TGF-β1–induced increase in TER after 48 and 72 h of treatment. However, we observed no significant change in the CsA-induced effect when we inhibited the p38 MAPK pathway (6), which signifies that TGF-β1 may also be mediating its effect on barrier function via the p38 MAPK pathway, but this particular signaling route is not triggered by exposure to CsA.
Treatment with CsA induced significant changes in the expression of several tight junction components. Although CsA increased occludin expression, it remains unclear whether occludin is actually a key player in regulating tight junctions. There are many conflicting reports regarding the significance of occludin in the maintenance of a functioning tight junction complex, and evidence suggests that it is the claudin proteins that play a pivotal role in the regulation of tight junctions. After exposure to CsA, increased expression of claudin-1 and claudin-3 was observed, both of which are implicated in the maintenance of tight junction integrity (32–34) No alterations in claudin-2 expression were detected after CsA treatment. Claudin-2 is associated with increased “leakiness” across epithelial cells. Significantly, claudin-2 has been detected in the low-resistance MDCK II cells but is absent from the high-resistance MDCK I cell strain, and introduction of claudin-2 into the MDCK I cells results in a drop in resistance (35). It has been reported that activation of ERK1/2 results in a reduction in claudin-2 expression in MDCK cells, and this is accompanied by an increase in resistance (33,36). In our system claudin-2 expression was unaltered after treatment with CsA despite activation of ERK1/2, suggesting that alterations in claudin-2 expression may be stimulus dependent. Exposure of MDCK cells to TGF-β1 significantly increased claudin-1 and ZO-2 expression, correlating with the effect of CsA on these tight junction components, which further strengthens our hypothesis that CsA regulates tight junction function, at least partially, via this cytokine. Another study (37) demonstrated that TGF-β upregulated claudin-1 expression two- to three-fold in T84 monolayers, but expression of claudin-2 and claudin-4 was unaltered. In addition, we found that treatment of MDCK cells with the MEK 1 inhibitor U0126 decreased claudin-1 expression compared with control level (data not shown). Therefore, it is possible that in MDCK II cells, CsA targets claudin-1 via the TGF-β1–ERK1/2 route.
Both CsA and TGF-β1 upregulated expression of ZO-2. The ZO family consists of three peripheral membrane proteins that are members of the MAGUK family. ZO-1, ZO-2, and ZO-3 are located in the underlying tight junction plaque and link to the actin cytoskeleton. ZO-1 and ZO-2 seem to interact directly with each other, occludin, the claudins, and the actin-based cytoskeleton through their PDZ domains. It is reported that subcellular localization of ZO-2 is sensitive to the state of cell–cell contact that is exhibited by the epithelial monolayer. In sparse epithelial cultures, ZO-2 accumulates in clusters at the nucleus. Shuttling of ZO-2 between the tight junction region and nucleus might be accomplished by the occurrence of putative nuclear localization and exportation signals on its sequence (38). The functional significance of the nuclear distribution of ZO-2 remains to be determined.
What is the significance of establishing the CsA-induced signaling mechanism? Enhancement of tight junction barrier function as a result of treatment with CsA may be a cytoprotective response by the cells to protect against injury. Determining how this “tightening” is triggered could be beneficial in the development of therapeutics for tight junction–related diseases or as a mechanism for drug delivery. Researchers are attempting to identify active compounds or excipients that can reversibly open tight junctions, thereby permitting drugs to pass through.
Disclosures
None.
Acknowledgments
This project was supported by the Conway Institute of Biomolecular and Biomedical Research and the Dublin Molecular Medicine Centre, under the Programme for Research in Third Level Institutions administered by the Higher Education Authority.
Footnotes
Published online ahead of print. Publication date available at www.jasn.org.
G.F. and B.K. contributed equally to this work.
B.K.'s current affiliation is Cork Cancer Research Centre, BioSciences Institute and Mercy University Hospital, University College Cork, Cork, Ireland.
See the related editorial, “Opening Pandora's Box in the Tight Junction,” on pages 1624–1625.
- © 2007 American Society of Nephrology
References
In this issue
Jump to section
More in this TOC Section
Cited By...
- Transforming Growth Factor-{beta}, a Whey Protein Component, Strengthens the Intestinal Barrier by Upregulating Claudin-4 in HT-29/B6 Cells
- Upregulation of ZO-1 in Cultured Human Corneal Epithelial Cells by a Peptide (PHSRN) Corresponding to the Second Cell-Binding Site of Fibronectin
- Calcineurin Inhibitor Nephrotoxicity
- Opening Pandora's Box in the Tight Junction