- Email: [email protected]
Adenosine induces dephosphorylation of myosin II regulatory light chain in cultured bovine corneal endothelial cells
Experimental Eye Research 79 (2004) 543–551 www.elsevier.com/locate/yexer
Adenosine induces dephosphorylation of myosin II regulatory light chain in cultured bovine corneal endothelial cells S.P. Srinivasa,*, M. Satpathya, P. Gallagherc, E. Larivie`reb, W. Van Driesscheb b
a School of Optometry, Indiana University, 800 Atwater Ave, Bloomington, IN 47405, USA Laboratory of Physiology, K.U. Leuven, Campus Gasthuisberg O/N, B-3000 Leuven, Belgium c Indiana University School of Medicine, Physiology, Indianapolis, IN, USA
Received 25 January 2004; accepted in revised form 25 June 2004 Available online 24 August 2004
Abstract Purpose: Dephosphorylation of the myosin II regulatory light chain (MLC) promotes barrier integrity of cellular monolayers through relaxation of the actin cytoskeleton. This study has investigated the influence of adenosine (ADO) on MLC phosphorylation in cultured bovine corneal endothelial cells (BCEC). Methods: MLC phosphorylation was assessed by urea-glycerol gel electrophoresis and immunoblotting. Elevation of cAMP in response to agonists of A2b receptors (subtype of P1 purinergic receptors) was confirmed by phosphorylation of the cAMP response element binding protein (CREB), which was determined by Western blotting. Activation of MAP kinases (i.e. activated ERK1 and ERK2) was assessed by Western blotting to examine their influence on MLC phosphorylation. Transepithelial electrical resistance (TER) of cells grown on porous filters was measured to assess the altered barrier integrity. Results: Exposure to ADO (200 mM; 30 min) and N-ethyl (carboxamido) adenosine (NECA; 50 mM; 30 min), known agonists of A2b receptors, induced phosphorylation of CREB similar to forskolin (FSK, 20 mM; 30 min), a direct activator of adenylate cyclase. Exposure to ADO, NECA, and FSK led to dephosphorylation of MLC by 51, 40, and 47%, respectively. ADO-induced dephosphorylation was dosedependent with as much as 31% dephosphorylation at 1 mM ADO. CGS-21680, a selective A2a agonist, neither induced MLC dephosphorylation nor CREB phosphorylation. ADO phosphorylated MAP kinases which could be prevented by exposure to the MAP kinase-specific inhibitor, U0126 (10 mM). NECA and FSK also induced ERK1 and ERK2 activation similar to ADO. Exposure to U0126 inhibited MLC phosphorylation under basal conditions by 17%. ADO-induced MLC dephosphorylation was enhanced by a simultaneous exposure to U0126 (25% increase in dephosphorylation). Exposure to ADO caused an increase in TER from 17 to 22 ohms cm2. Conclusions: (1) CREB phosphorylation in response to ADO and NECA, which indicates activation of the cAMP-PKA axis, suggests expression of A2b receptors in BCEC. (2) ERK1 and ERK2, activated by cAMP and A2b receptors, promote MLC phosphorylation. However, the net result of cAMP elevation is MLC dephosphorylation, presumably because the competing pathways involving inactivation of MLCK and/or ROCK are dominant (Rho-associated coiled coil-containing protein kinase or Rho kinase). (3) Consistent with MLC dephosphorylation, exposure to ADO increases TER, which suggests increased barrier integrity. q 2004 Elsevier Ltd. All rights reserved. Keywords: corneal endothelium; adenosine; myosin light chain kinase; cAMP; MLC phosphorylation
1. Introduction The corneal endothelium (CE), a monolayer on the posterior surface of the cornea, maintains the transparency
* Corresponding author. Address: Dr Srinivas, S.P. School of Optometry, Indiana University, 800 Atwater Ave, Bloomington, IN 47405, USA. E-mail address: [email protected] (S.P. Srinivas). 0014-4835/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. DOI:10.1016/j.exer.2004.06.027
of the tissue by controlling the hydration of the corneal stroma (Fischbarg et al., 1985; Riley, 1985). The hydrophilic glycosaminoglycans bound to stromal collagen imbibe water and thereby induce a fluid leak across the endothelium. This leakage is counterbalanced by an endothelial ‘fluid pump’ directed from the stroma to the aqueous humor. The tight junctions (TJs) of CE, although very leaky (trans-endothelial electrical resistance !25 ohms cm2) (Noske et al., 1994a,b), restrain excessive
544
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
fluid leak and thus form the basis for the barrier integrity of CE. The ‘fluid pump’ function of CE is sustained by a host of polarized and active ion transport mechanisms which couple net ionic fluxes to fluid transport through an obligatory osmotic coupling (Bonanno, 2003). Although much is known about the identity of ion transport mechanisms in CE (Bonanno, 2003), the mechanisms of cell signaling that regulate ion transport and barrier integrity are just beginning to be understood (Riley et al., 1996, 1998; Wigham et al., 2000; Zhang et al., 2002; Bonanno, 2003). Forskolin (FSK)-induced elevation of cAMP has been shown to increase conductance for ClK and HCOK through ion channels, including cystic fibrosis 3 transmembrane regulator (CFTR) (Bonanno and Srinivas, 1997; Sun et al., 2003). Riley et al. showed that adenosine (ADO) and FSK promote deturgescence of swollen rabbit corneas (Riley et al., 1996). Wigham et al. discovered that exposure of rabbit CE to a selective inhibitor of cAMPdependent phosphodiesterase (isoform PDE4), rolipram, leads to stromal thinning (Wigham et al., 2000). Apart from these effects on ion transport and corneal hydration, stimulation of the cAMP-PKA axis is also known to enhance barrier integrity (i.e. a decrease in paracellular permeability) of CE. Specifically, Riley et al. (1996, 1998) suggested that ADO, an endogenous agonist for P1 purinergic receptors, enhances the rate of deswelling of pre-swollen corneas through a cAMP-mediated increase in barrier integrity. This finding is also reflected in studies with the PDE4 inhibitor, rolipram (Wigham et al., 2000). The physiological role of the cAMP-PKA axis in CE has gained further prominence through recent studies showing expression of HCOK 3 -sensitive adenylate cyclase (the soluble isoform) (Sun et al., 2003). These observations on cAMP have prompted us to seek further elaboration on the putative role of the second messenger on the barrier integrity of CE. Several recent studies of the vascular endothelium (Garcia et al., 1995; Patterson et al., 2000; Stevens et al., 2000; van Hinsbergh and van Nieuw Amerongen, 2002) and certain epithelial monolayers (Turner et al., 1997; Turner, 2000) have demonstrated that contractility of the actin cytoskeleton plays an important role in the regulation of barrier integrity. Specifically, when a cell signaling cascade activates the motor protein myosin II by phosphorylating its regulatory light chain (MW: 20 kD; also referred to as myosin light chain or MLC), the resulting contractions of the actin cytoskeleton, anchored at TJs and adherent junctions (AJs), generate a centripetal force (Dudek and Garcia, 2001, 2002). This force opposes the intercellular tethering forces which, in turn, break down the cell–cell apposition necessary for interactions of the transmembrane proteins at the TJs and AJs. In the absence of these interactions, the occlusion of the paracellular pathway is not complete, and hence, barrier integrity is disrupted. In this study, we demonstrate that in cultured bovine corneal endothelial cells (BCEC), ADO and NECA, both agonists
for the A2b receptors, induce MLC dephosphorylation and increase barrier integrity.
2. Materials and methods 2.1. Cell culture Primary cultures of BCEC were established from fresh cow eyes as described earlier (Bonanno and Srinivas, 1997) in Dulbecco’s Modified Eagle’s Medium (supplemented with 10% fetal calf serum and an antibiotic–antimycotic mixture consisting of penicillin at 100 U mlK1, streptomycin at 100 mg mlK1 and fungizone at 0$25 mg mlK1) at 378C in a humidified atmosphere containing 5% CO2 and 95% air. The cells were fed every 2–3 days. Second or third passage cultures were grown to confluence on Transwelle filters (Corning Inc., Coburn, MA) or Petri dishes. 2.2. Quantitative estimation of MLC phosphorylation MLC phosphorylation was assayed with cells grown on Petri dishes by urea-glycerol gel electrophoresis followed by immunoblotting (Garcia et al., 1995). This protocol has been used in a number of studies involving vascular smooth muscle cells, trabecular meshwork endothelial cells, and vascular endothelial cells (Rao et al., 2001; Wang et al., 2001; Blue et al., 2002). After exposure to the specified agents, the reaction was terminated by the addition of 1 ml of ice-chilled PBS containing trichloroacetic acid and dithiothreitol (DTT). Cells were then scraped off and centrifuged for 10 min at 10 000g. The resulting pellet was washed with chilled acetone and dried. The precipitated protein was dissolved in a urea sample buffer containing 8 M urea, 10 mM DTT, 20 mM Tris-base, 23 mM glycine and 0$04% bromophenol blue. After a brief spin, the ureasolubilized sample was electrophoresed for 3 hr in 1$5 mm polyacrylamide slabs containing acrylamide, bis-acrylamide, glycerol, Tris-base, and glycine, at pH 8$6. Prior to loading the samples (25 mg of protein per lane), gels were subjected to pre-electrophoresis for 1$5 h. The reservoir buffer consisted of 20 mM Tris-base, 23 mM glycine, 2$96 mM thioglycolic acid, and 3 mM DTT. The separated phosphorylated and non-phosphorylated MLC were detected by immunoblotting using a polyclonal anti-MLC antibody (E201; 1:3000 dilution). In these gels, the diphosphorylated form (identified as ‘PP’ on gel photographs) migrated faster than the mono-phosphorylated form (identified as ‘P’ on gel photographs). Furthermore, the mono-phosphorylated form migrated faster than the nonphosphorylated form (identified as ‘NP’ on gel photographs). These migration patterns are typical of urea-glycerol gels (Garcia et al., 1995). Blots were washed with TBST three times (15 min each) then visualized using a peroxidase conjugated secondary antibody and an enhanced chemiluminescence kit (Amersham). The bands were
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
scanned into TIF images and band intensities were quantified using a custom-made software program. The program subtracted the background before quantifying the pixel intensity of each band associated with non-phosphorylated (denoted by iNP), mono-phosporylated (iP) and di-phosphorylated (iPP) forms. The fraction of the MLC phosphorylated form (i.e. moles of MLC phosphorylated/ total moles of MLC) was calculated by dividing the total MLC phosphorylated (equal to iPC[2!iPP]) by the total MLC (equal to iPCiPPCiNP) (Garcia et al., 1995; Wang et al., 2001). 2.3. CREB phosphorylation by Western blotting Phosphorylation of the cyclic AMP response element binding protein (CREB) was assessed by Western blotting. Endothelial cells from 100 mm Petri dishes were treated with desired agents and then lysed with an ice-cold lysis buffer containing 50 mM Tris HCl (pH 7$6), 150 mM NaCl, 1% Triton X-100, 2 mM EGTA (pH 8$0), EDTA, DTT, PMSF, NaF, and a protease inhibitor cocktail (containing leupeptin, pepstatin, chymostatin, and aprotinin). After a pre-clearing centrifugation step (14,000 rpm for 15 min at 48C), the whole lysate was subjected to SDS-PAGE (100 mg protein per lane) and immunoblotting. For experiments with CGS-21,680, nuclear extracts were prepared using Active Motif (Carlsbad, CA; Cat. No. 40010). Membranes were blocked with TBST containing 5% nonfat dry milk for 1 hr at room temperature, and subsequently incubated with primary antibody that had been diluted in TBST and dry milk overnight at 48C. Blots were washed with TBST three times (15 min each) and visualized using a peroxidase conjugated secondary antibody and an enhanced chemiluminescence kit (Amersham).
545
purchased from Amersham-Pharmacia Biotech (Piscataway, NJ). All other drugs/reagents were purchased from Sigma (St Louis, MO). 2.6. Data analysis Unpaired t-tests were used to compare the effects of various drugs at p!0$05. Data are expressed as meansGSE.
3. Results 3.1. Adenosine-induced CREB phosphorylation Among A1, A2a, A2b, and A3 subtypes of ADO-sensitive receptors, A2a and A2b subtypes are coupled to the stimulatory G-protein (Gas) (Feoktistov and Biaggioni, 1997; Fredholm et al., 2001; Rees et al., 2003). To identify the receptor subtype(s) expressed in BCEC, we investigated phosphorylation of CREB, a transcription factor phosphorylated by protein kinase A (PKA) (Collins et al., 1990; Rehfuss et al., 1991; Fitzgerald et al., 1999; Servillo et al., 2002). Cells were serum-starved overnight and exposed for 30 min to ADO (200 mM) and NECA (50 mM, selective A2b agonist) (Grant et al., 2001). The whole-cell protein extracts were then subjected to Western blotting using an antibody to detect phosphorylated CREB (p-CREB). However, the antibody also detected the activated transcription factor 1 (ATF-1), as noted by the manufacturer. Fig. 1A is a typical gel showing CREB and ATF-1 activation (phosphorylation) in response to ADO (Lane 3) and NECA (Lane 4). As expected, exposure to FSK, a direct activator of adenylate
2.4. Barrier integrity by impedance analysis Impedance analysis was used to determine the resistance of the BCEC monolayer grown on Transwell porous supports, as described earlier (Van Driessche et al., 1999; Robaye et al., 2003). The experimental procedure consisted of mounting the porous support with cultured BCEC between two halves of an Ussing chamber (Robaye et al., 2003). KCl electrodes, connected to the solution via short agar bridges, were used for measuring the potential difference and injecting current. The volume of each compartment bathing the cells was 2 ml. A HCOK 3 -free Ringers at 378C and pH 7$4 was perfused through each compartment. The Ringers was composed of 140 mM NaC, 4 mM KC, 1$2 mM Mg2C, 1$2 mM Ca2C, 120 mM ClK, 2$8 mM PO2K 4 , 25 mM gluconate, and 6$5 mM glucose. 2.5. Drugs and chemicals Bovine calf serum was purchased from Hyclone (Logan, UT). An enhanced chemiluminescence reagent was
Fig. 1. A2b agonists, ADO and NECA, phosphorylate PKA-sensitive transcription factors CREB and ATF-1: Western blotting analysis of CREB and ATF-1 phosphorylation in response to ADO (200 mM, 30 min), NECA (50 mM, 30 min) and CGS-21680 (CGS; 50 mM; 30 min). As a positive control, response to forskolin (FSK; 30 mM, 30 min) is also shown. Western blotting was performed with an antibody specific for Ser133 phosphorylated CREB (p-CREB) that also reacted with ATF-1. Panel A: Cells were treated with agents specified and whole cell extracts (100 mg per lane) were electrophoresed on SDS-PAGE. Panel B: Cells were treated with the agents specified and nuclear extracts (100 mg per lane) were electrophoresed on SDS-PAGE. Lanes ‘MC’ and ‘MC(EGF/FSK)’ are positive controls for p-CREB/ATF-1 provided by the manufacturer. Lane ‘C’ represents BCEC under basal conditions.
546
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
cyclase, also led to CREB activation (Lane 5). Lane 1 (identified as MC(EGF)) represents positive control from the manufacturer in which the cells were treated with EGF. Lane 2 (identified as MC), represents cells without EGF treatment, also supplied by the manufacturer. Activation of CREB and ATF-1 in response to both ADO and NECA, similar to FSK, is consistent with activation of cAMP-PKA axis by A2b receptors in BCEC. The gel in Panel B obtained from nuclear extracts shows a lack of activation of CREB or ATF-1 in response to CGS-21680 (A2a selective agonist). 3.2. cAMP-mediated MLC dephosphorylation Myosin light chain kinase (MLCK), a dedicated protein kinase for MLC phosphorylation, (Stull et al., 1991) is phosphorylated by PKA, leading to its inactivation (Garcia et al., 1997) and reduced MLC phosphorylation. In addition, Rho kinase (ROCK; Rho-associated coiled coil-containing protein kinase) activity is reduced by PKA activation through inhibition of the dissociation of RhoA from guanine nucleotide dissociation inhibitor (GDI) (Qiao et al., 2003). Since Rho kinase phosphorylates the myosin light chain phosphatase (MLCP) and promotes its inactivation, the net effect of PKA activation is MLC
Fig. 2. MLC dephosphorylation in response to P1 purinergic agonists: Panel A: Cells were treated with ADO (100 mM; 30 min), NECA (20 mM; 18 min), and CGS-21680 (CGS; 10 mM; 18 min). FSK (30 mM, 18 min) was used as a positive control (Lane ‘CC’). Basal phosphorylation in untreated cells is shown in Lane ‘C’. CGS-21680 (CGS), selective for A2a purinergic receptors, did not have a significant effect on MLC phosphorylation. Panel B: Densitometric analysis of data from experiments similar to those shown in Panel A. The histogram shows the meanGSE from 5 to 12 independent experiments. Notes: Nonphosphorylated MLC; P: Monophopshorylated MLC; PP: Diphosphorylated MLC. ‘*’ denotes p!0$05 for comparison of mean values with respect to the control. Note CGS-21680-induced dephosphorylation is not significant when compared to the control (pZ0$23).
dephosphorylation (Qiao et al., 2003). Since activation of the cAMP-PKA axis was shown above, we investigated the influence of ADO and NECA on the phosphorylation status of MLC in BCEC by using urea-glycerol gel electrophoresis. Similar to CREB phosphorylation experiments, cells were serum-starved overnight and then exposed to the ADO and NECA for 18 min. Both agonists induced an apparent MLC desphosphorylation (Lanes 2 and 4, respectively, in Fig. 2, labeled ADO and NECA) compared to cells under basal conditions (Lane 1, labeled C). FSK also induced dephosphorylation (Lane 3, Fig. 2, and labeled FSK). Similar experiments with the A2a agonist CGS-21680 (100 mM; 18 min) did not show significant dephosphorylation (Fig. 2; Lane 5; labeled CGS) compared to the control. The histogram in Fig. 2B, obtained by densitometric analysis of gels similar to that shown in Fig. 2A, clearly demonstrates that A2b agonists induce significant MLC dephosphorylation. The extent of desphosphorylation by A2b agonists is similar to that by FSK. On the other hand, dephosphorylation induced by CGS-21680 is not significant compared to untreated controls. The results in Fig. 2B also suggest that NECA is more potent than ADO. Fig. 3 is a summary of experiments showing time-dependence of ADO-induced MLC dephosphorylation. Maximum dephosphorylation was reached at about 18 min and then a reversal towards the baseline was noted (see Lane 5 in Fig. 3A; labeled
Fig. 3. Time course of ADO-induced MLC dephosphorylation: Cells were treated with ADO (100 mM) for 2, 5, 10, 18, and 30 min. Densitometric analysis of the data is plotted on the Y-axis, indicating the extent of MLC phosphorylation. The histogram shows the meanGSE from four independent experiments. MLC phosphorylation status of serum-starved cells was taken as 100%. ‘*’ denotes p!0$05 for comparison of mean values with respect to the control. ‘**’ denotes that ADO-induced dephosphorylation is not significant when compared to the control (pZ0$46).
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
547
30 min and Fig. 3B). Fig. 4 is the summary of doseresponse data obtained with ADO at three concentrations. Clearly, the extent of dephosphorylation increased with the doses of greater concentrations (Fig. 4B). 3.3. ADO-mediated phosphorylation of ERK1 and ERK2 ERK1 and ERK2 promote MLC phosphorylation by increasing the catalytic efficiency of MLCK (Stull et al., 1991; Klemke et al., 1997). To confirm these in BCEC, we first investigated NECA and ADO-induced ERK1 and ERK2 activation. Consistent with previous reports (Gao et al., 1999; Grant et al., 2001), ADO and NECA induced ERK1 and ERK2 activation after 18 min of exposure (Fig. 5A). This activation is inhibited by U0126 (10 mM; Fig. 5B), a selective inhibitor of MEK which is upstream from ERK1 and ERK2 activation. U0126 also reduced basal MLC phosphorylation by 17% (Fig. 6A; Lane 4, labeled U). Exposure to ADO in the presence of U0126 caused dephosphorylation of MLC to a greater extent than with ADO alone (Fig. 6A, Lane 3; labeled UCADO). A summary of similar experiments is given in Fig. 6B. 3.4. ADO-induced increase in barrier integrity In order to examine the effect of ADO on barrier integrity, we measured TER using impedance analysis. The analysis was based on modeling the monolayer as a parallel
Fig. 4. Dose response of adenosine-induced MLC dephosphorylation: Cells were treated with ADO for 10 min at 1, 10, and 100 mM. The gels resulting from urea-glycerol gel electrophoresis and Western blotting were subjected to densitometric analysis. The resulting data are shown as a histogram with the Y-axis representing the extent of MLC phosphorylation (meanGSE) from five independent experiments. Note that the phosphorylation under basal conditions (i.e. serum starved cells) was taken as 100%. ‘*’ denotes p!0$05 for comparison of mean values with respect to the control.
Fig. 5. Phosphorylation of ERK1 and ERK2 in response to ADO. Cells were treated with ADO (100 mM) and NECA (10mM) for 30 min. FSK (10 mM; 30 min) was used as the positive control. Equal amounts of protein were loaded on SDS-PAGE and phosphorylated ERK1 and ERK2 were examined by Western blotting. The gel at the top shows enhanced phosphorylation of both ERK1 and ERK2 in response to ADO, FSK, and NECA. The gel at the bottom of the figure shows sensitivity of the phosphorylation to MAPK inhibitor U0126. Note that U0126 (10 mM) inhibited activation of ERK1 and ERK2 under basal conditions as well as in the presence of ADO. The results shown are typical of five independent experiments.
circuit of a capacitance (C) and resistance (TER) in series with the solution resistance (Rsol) between the voltage electrodes (Van Driessche et al., 1999; Robaye et al., 2003). The corresponding impedance spectrum is a semicircle in
Fig. 6. Effect of ERK1 and ERK2 activation on MLC phosphorylation: Cells starved of serum overnight were treated with ADO (100 mM, 18 min). In some experiments, cells were pre-treated with U0126 alone or U0126 (10 mM, 10 min) followed by ADO (100 mM, 18 min). MLC phosphorylation was further reduced when cells were exposed to the specific MAP kinase inhibitor, U0126. ‘*’ denotes p!0$05 for comparison of mean values with respect to the control. ‘**’ denotes p!0$05 for comparison of mean values between ADO and ‘ADOCU0126’. The results shown are typical of 3–4 independent experiments.
548
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
Fig. 7. Increase in transendothelial electrical resistance after exposure to ADO: BCEC, grown on porous matrices and mounted in an Ussing chamber, were continuously perfused with HCOK 3 -free Ringers. The steadystate TER of the monolayer was measured by impedance analysis under resting conditions and subsequently after exposure to ADO (200 mM). The results shown are a summary from five independent experiments. The increase in resistance at 10, 20 and 30 min after exposure to ADO is statistically significant (p!0$05) with respect to resting conditions.
a Nyquist plot (Robaye et al., 2003). In such a plot, the intercepts of the semicircle at high and low frequencies correspond to Rsol and RsolCTER, respectively. We recorded impedance spectra before and after exposure to ADO. As described by Van Driessche et al. (1999) and Robaye et al. (2003), the Cole–Cole equation was fitted to the data with a custom-made computer program and extrapolation led to an estimate of TER. Consistent with previous studies (Riley et al., 1998; Wigham et al., 2000), exposure to ADO resulted in a significant increase in TER from 17 to 22 ohms cm2. Time course of the change in TER is summarized from six independent experiments in Fig. 7. The increase in TER was rapid and reached a maximum in about 15 min. The increase in TER persisted in the presence of ADO (observed for !30 min). After removal of ADO from the bath, TER showed a tendency to decrease (data not shown).
4. Discussion ADO stimulates fluid transport (Fischbarg et al., 1977) and promotes barrier integrity in CE (Riley et al., 1996, 1998). These effects contribute to an enhancement in stromal hydration control (Riley et al., 1996). ADO-induced stimulation of fluid transport may involve anion channels sensitive to cAMP (Bonanno and Srinivas, 1997; Zhang et al., 2002; Bonanno, 2003). However, a mechanistic basis for the enhancement of barrier integrity has not been ascertained. This study investigated MLC phosphorylation in response to ADO, since increased MLC phosphorylation is associated with a breakdown of barrier integrity and,
dephosphorylation is known to overcome barrier dysfunction (Turner et al., 1997; Patterson et al., 2000; Stevens et al., 2000; Turner, 2000; van Hinsbergh and van Nieuw Amerongen, 2002). This phenomenon is attributed to the fact that elevated MLC phosphorylation increases the contractility of the actin cytoskeleton and leads to the development of a centripetal force that opposes the forces of tethering between neighboring cells. These forces are necessary for interactions of the transmembrane proteins at tight and AJs (Dudek and Garcia, 2001; Dull and Garcia, 2002). Experiments in this study show that (a) ADO stimulates the cAMP-PKA axis in BCEC as evidenced by phosphorylation of CREB, (b) ADO induces significant MLC dephosphorylation, (c) NECA, a potent agonist for the A2b receptors, also induces CREB phosphorylation as well as MLC dephosphorylation, and (d) ADO increases the TER of BCEC, indicating an enhancement in barrier integrity. 4.1. Significance of adenosine-induced CREB phosphorylation Expression of ADO-sensitive receptors in CE has been suspected for a long time. Walkenbach showed cAMP elevation in BCEC when challenged by ADO at 1 mM to 0$5 mM (Walkenbach and Chao, 1985). Riley et al. noted the elevation of cAMP in response to ADO in rabbit CE (Riley et al., 1996). Despite these observations, further confirmation of its identity at the mRNA level and pharmacological profiling has been elusive. CREB phosphorylation is a hallmark of elevated cAMP and the subsequent activation of PKA. Accordingly, CREB phosphorylation has been used previously, for example, to demonstrate the activation of b-adrenergic receptors (Collins et al., 1990; Fitzgerald et al., 1999). Thus, our findings in Fig. 1 showing CREB phosphorylation in response to ADO and NECA suggest the expression of A2b receptors in BCEC. Our protocols were validated using FSK and positive controls provided by the manufacturer of the p-CREB antibody. Activation of the transcription factor ATF-1 is also consistent with activation of the cAMP-PKA axis in response to ADO (Rehfuss et al., 1991; Servillo et al., 2002). The fact that NECA also induced CREB phosphorylation further indicates the expression of A2b receptors in BCEC since NECA is a selective agonist of the receptor. 4.2. The mechanism underlying MLC dephosphorylation MLCK is expressed as smooth muscle (SM-MLCK; 130 kD) and vascular endothelial isoforms (EC-MLCK; 220 kD) in BCEC (Satpathy et al., 2004). EC-MLCK is known to possess a number of regulatory sites, including a consensus sequence for phosphorylation by PKA. In fact, in vitro studies have shown inactivation of EC-MLCK upon phosphorylation by PKA (Garcia et al., 1997). Accordingly, dephosphorylation of MLC has been reported in response to
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
549
activation of the cAMP-PKA axis in vascular endothelium (Essler et al., 2000; Patterson et al., 2000; van Hinsbergh and van Nieuw Amerongen, 2002). In our experiments, as summarized in Fig. 2, agents that elevated cAMP in BCEC, including FSK, ADO and NECA, induced MLC dephosphorylation. As shown in Figs. 3 and 4, dephosphorylation was rapid and dose-dependent, lasting for at least 18 min. In addition to the inactivation of MLCK, recent studies also indicate that PKA upregulates the activity of the MLCP which opposes MLCK-mediated MLC phosphorylation (Kamm and Stull, 2001). Specifically, Qiao et al. (2003) showed that PKA/cAMP could inhibit activation of RhoA upstream of Rho kinase, which phosphorylates the catalytic subunit of MLCP and down-regulates the activity of the enzyme. Thus, it is possible that the MLC dephosphorylation observed in this study is due to the inactivation of Rho kinase secondary to inhibition of activation of RhoA. Our previous study on thrombin-mediated MLC phosphorylation showed that the expression of Rho kinase (ROCK-1 isoform) in BCEC and, moreover, that the inhibition of Rho kinase by its selective inhibitor Y-27632 was sufficient to block MLC phosphorylation significantly under basal conditions (Satpathy et al., 2004). 4.3. Significance of adenosine-induced phosphorylation of ERK1 and ERK2 Despite MLC dephosphorylation, as discussed above, experiments illustrated in Figs. 5 and 6 unraveled activation of another opposing pathway in response to ADO. As noted earlier, activation of ERK1 and ERK2 is a hallmark of a number of ADO-sensitive receptors, including the A2b subtype (Feoktistov and Biaggioni, 1997; Gao et al., 1999; Grant et al., 2001). Although an exact mechanism underlying activation of ERK1 and ERK2 is not revealed by our experiments, it is clear that the elevation of cAMP itself is sufficient since FSK also induced activation of ERK1 and ERK2 (Fig. 5). In regard to the positive influence of ERK1 and ERK2 on MLC phosphorylation, our results are consistent with previous reports (Klemke et al., 1997; Kamm and Stull, 2001). Specifically, Fig. 6 clearly shows that U0126, a selective inhibitor of MAPK kinase, reduced MLC phosphorylation of cells under resting conditions. However, in the presence of ADO, activation of the cAMPPKA axis overwhelms the positive influences of ERK1 and ERK2 in promoting MLC phosphorylation (Fig. 6). In Fig. 8, we present a summary of the pathways likely to be involved in ADO-induced MLC dephosphorylation. 4.4. Adenosine-induced increase in barrier integrity The tight junctional permeability, a principal determinant of barrier integrity, is regulated by a number of protein kinases: (a) directly through phosphorylation of the transmembrane proteins (i.e. occludins and claudins and b-catenin), and/or (b) indirectly through reorganization of
Fig. 8. Schematic of the cell signaling mediated by ADO in CE for the control of barrier integrity: Activation of A2b receptors is coupled to the elevation of intracellular cAMP, leading to the activation of PKA. Direct phosphorylation of MLCK leads to its inactivation and hence would reduce MLC phosphorylation. Alternatively, MLC dephosphorylation could be induced by an inhibition of RhoA activation leading reduced Rho kinase activity. The reduced extent of MLC phosphorylation induces relaxation of the actin cytoskeleton, leading to an enhancement in barrier integrity.
the actin cytoskeleton (Yuan, 2000). The latter, which has received much attention recently, is achieved by altering the phosphorylation status of MLC (Turner et al., 1997; Stevens et al., 2000; Turner, 2000; Turner et al., 2000; van Hinsbergh and van Nieuw Amerongen, 2002). Apart from a few exceptions noted to date (Yuan, 2000), an increase in the contractility of the actin cytoskeleton is associated with a breakdown of barrier integrity (Turner et al., 1997; Stevens et al., 2000; Turner, 2000; Turner et al., 2000; van Hinsbergh and van Nieuw Amerongen, 2002). Dephosphorylation, on the other hand, has been shown to cause relaxation of the actin cytoskeleton and to overcome barrier dysfunction concomitantly (Noll et al., 1999, 2000) when induced by agents such as thrombin, histamine, and lipopolysaccharides (Essler et al., 2000). This is consistent with observations on rabbit CE (Riley et al., 1998; Wigham et al., 2000). BCEC also showed an increase in TER in response to ADO (Fig. 7). The increase in resistance was rapid and matched the time course of MLC phosphorylation (Fig. 3). The average TER under resting conditions (17 ohms cm2; Fig. 7) is similar to values reported earlier by Noske in bovine eyes
550
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
(20 ohms cm2) (Noske et al., 1994b). Furthermore, our findings and those by Riley et al. (1998) and Wigham et al. (2000) contradict those reported by Le Varlet et al. (1995) who claimed an increase in paracellular permeability and a decrease in TER of BCEC in response to dibutyryl-cAMP and FSK. 4.5. Significance to stromal hydration Since there is no apparent interconnection between regulation of barrier integrity and fluid transport in CE, it appears that hyper-permeability of the endothelium at a constant imbibition pressure would result in uncontrolled stromal hydration. This is evident in an increase in corneal thickness when inflammation occurs (Watsky et al., 1996; Behar-Cohen et al., 1998; Edelhauser, 2000; Yi and Dana, 2002) and, indirectly, when the net hydraulic pressure favoring fluid influx across the endothelium is increased by ocular hypertension (Herndon et al., 1997; Wolfs et al., 1997; Bron et al., 1999; Copt et al., 1999). Under these conditions, it may be therapeutically useful to pharmacologically alter endothelial barrier dysfunction and overcome corneal swelling. To this end, the results from this study, taken together, suggest that pharmacological approaches capable of elevating intracellular cAMP in CE would be useful. In summary, this study has demonstrated that ADO and its A2b-sensitive analogs induce MLC dephosphorylation in CE. Concomitant with this dephosphorylation, an increase in barrier integrity is promoted. These findings offer a potential strategy for using A2b receptor agonists in preventing corneal swelling, especially for counteracting enhanced MLC phosphorylation induced by inflammatory mediators.
Acknowledgements Supported by NIH grant EY11107 and EY14415 (SPS).
References Behar-Cohen, F.F., Savoldelli, M., Parel, J.M., Goureau, O., Thillaye-Goldenberg, B., Courtois, Y., Pouliquen, Y., de Kozak, Y., 1998. Reduction of corneal edema in endotoxininduced uveitis after application of L-NAME as nitric oxide synthase inhibitor in rats by iontophoresis. Invest. Ophthalmol. Vis. Sci. 39, 897–904. Blue, E.K., Goeckeler, Z.M., Jin, Y., Hou, L., Dixon, S.A., Herring, B.P., Wysolmerski, R.B., Gallagher, P.J., 2002. 220and 130-kDa MLCKs have distinct tissue distributions and intracellular localization patterns. Am. J. Physiol. Cell Physiol. 282, C451–C460. Bonanno, J.A., 2003. Identity and regulation of ion transport mechanisms in the corneal endothelium. Prog. Retin. Eye Res. 22, 69–94. Bonanno, J.A., Srinivas, S.P., 1997. Cyclic AMP activates anion channels in cultured bovine corneal endothelial cells. Exp. Eye Res. 64, 953–962.
Bron, A.M., Creuzot-Garcher, C., Goudeau-Boutillon, S., d’Athis, P., 1999. Falsely elevated intraocular pressure due to increased central corneal thickness. Graefes Arch. Clin. Exp. Ophthalmol. 237, 220–224. Collins, S., Altschmied, J., Herbsman, O., Caron, M.G., Mellon, P.L., Lefkowitz, R.J., 1990. A cAMP response element in the beta 2-adrenergic receptor gene confers transcriptional autoregulation by cAMP. J. Biol. Chem. 265, 0–5. Copt, R.P., Thomas, R., Mermoud, A., 1999. Corneal thickness in ocular hypertension, primary open-angle glaucoma, and normal tension glaucoma. Arch. Ophthalmol. 117, 14–16. Dudek, S.M., Garcia, J.G., 2001. Cytoskeletal regulation of pulmonary vascular permeability. J. Appl. Physiol. 91, 1487–1500. Dull, R.O., Garcia, J.G., 2002. Leukocyte-induced microvascular permeability: how contractile tweaks lead to leaks. Circ. Res. 90, 1143–1144. Edelhauser, H.F., 2000. The resiliency of the corneal endothelium to refractive and intraocular surgery. Cornea 19, 263–273. Essler, M., Staddon, J.M., Weber, P.C., Aepfelbacher, M., 2000. Cyclic AMP blocks bacterial lipopolysaccharide-induced myosin light chain phosphorylation in endothelial cells through inhibition of Rho/Rho kinase signaling. J. Immunol. 164, 6543–6549. Feoktistov, I., Biaggioni, I., 1997. Adenosine A2B receptors. Pharmacol. Rev. 49, 381–402. Fischbarg, J., Hernandez, J., Liebovitch, L.S., Koniarek, J.P., 1985. The mechanism of fluid and electrolyte transport across corneal endothelium: critical revision and update of a model. Curr. Eye Res. 4, 351–360. Fischbarg, J., Lim, J.J., Bourguet, J., 1977. Adenosine stimulation of fluid transport across rabbit corneal endothelium. J. Membr. Biol. 35, 95–112. Fitzgerald, L.R., Mannan, I.J., Dytko, G.M., Wu, H.L., Nambi, P., 1999. Measurement of responses from Gi-, Gs-, or Gq-coupled receptors by a multiple response element/cAMP response element-directed reporter assay. Anal. Biochem. 275, 54–61. Fredholm, B.B., AP, I.J., Jacobson, K.A., Klotz, K.N., Linden, J., 2001. International Union of Pharmacology. XXV. Nomenclature and classification of adenosine receptors. Pharmacol. Rev. 53, 527–552. Gao, Z., Chen, T., Weber, M.J., Linden, J., 1999. A2B adenosine and P2Y2 receptors stimulate mitogen-activated protein kinase in human embryonic kidney-293 cells. Cross-talk between cyclic AMP and protein kinase c pathways. J. Biol. Chem. 274, 5972–5980. Garcia, J.G., Davis, H.W., Patterson, C.E., 1995. Regulation of endothelial cell gap formation and barrier dysfunction: role of myosin light chain phosphorylation. J. Cell. Physiol. 163, 510–522. Garcia, J.G., Lazar, V., Gilbert-McClain, L.I., Gallagher, P.J., Verin, A.D., 1997. Myosin light chain kinase in endothelium: molecular cloning and regulation. Am. J. Respir. Cell Mol. Biol. 16, 489–494. Grant, M.B., Davis, M.I., Caballero, S., Feoktistov, I., Biaggioni, I., Belardinelli, L., 2001. Proliferation, migration, and ERK activation in human retinal endothelial cells through A(2B) adenosine receptor stimulation. Invest. Ophthalmol. Vis. Sci. 42, 2068–2073. Herndon, L.W., Choudhri, S.A., Cox, T., Damji, K.F., Shields, M.B., Allingham, R.R., 1997. Central corneal thickness in normal, glaucomatous, and ocular hypertensive eyes. Arch. Ophthalmol. 115, 1137–1141. Kamm, K.E., Stull, J.T., 2001. Dedicated myosin light chain kinases with diverse cellular functions. J. Biol. Chem. 276, 4527–4530. Klemke, R.L., Cai, S., Giannini, A.L., Gallagher, P.J., de Lanerolle, P., Cheresh, D.A., 1997. Regulation of cell motility by mitogen-activated protein kinase. J. Cell. Biol. 137, 481–492. Le Varlet, B., Ducroc, R., Dagonet, F.B., Pouliquen, Y., Vandewalle, A., Hirsch, M., 1995. Dibutyryl cyclic adenosine monophosphate and forskolin alter the paracellular pathway in cultured corneal endothelial cells. Invest. Ophthalmol. Vis. Sci. 36, 2503–2513.
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551 Noll, T., Holschermann, H., Koprek, K., Gunduz, D., Haberbosch, W., Tillmanns, H., Piper, H.M., 1999. ATP reduces macromolecule permeability of endothelial monolayers despite increasing [Ca2C]i. Am. J. Physiol. 276, H1892–H1901. Noll, T., Schafer, M., Schavier-Schmitz, U., Piper, H.M., 2000. ATP induces dephosphorylation of myosin light chain in endothelial cells. Am. J. Physiol. Cell Physiol. 279, C717–C723. Noske, W., Fromm, M., Levarlet, B., Kreusel, K.M., Hirsch, M., 1994a. Tight junctions of the human corneal endothelium: morphological and electrophysiological features. Ger. J. Ophthalmol. 3, 253–257. Noske, W., Levarlet, B., Kreusel, K.M., Fromm, M., Hirsch, M., 1994b. Tight junctions and paracellular permeability in cultured bovine corneal endothelial cells. Graefes Arch. Clin. Exp. Ophthalmol. 232, 608–613. Patterson, C.E., Lum, H., Schaphorst, K.L., Verin, A.D., Garcia, J.G., 2000. Regulation of endothelial barrier function by the cAMP-dependent protein kinase. Endothelium 7, 287–308. Qiao, J., Huang, F., Lum, H., 2003. cAMP-dependent protein kinase inhibits RhoA activation: a protection mechanism against endothelial barrier dysfunction. Am. J. Physiol. Lung Cell. Mol. Physiol. 2003;. Rao, P.V., Deng, P.F., Kumar, J., Epstein, D.L., 2001. Modulation of aqueous humor outflow facility by the Rho kinase-specific inhibitor Y-27632. Invest. Ophthalmol. Vis. Sci. 42, 1029–1037. Rees, D.A., Scanlon, M.F., Ham, J., 2003. Adenosine signalling pathways in the pituitary gland: one ligand, multiple receptors. J. Endocrinol. 177, 357–364. Rehfuss, R.P., Walton, K.M., Loriaux, M.M., Goodman, R.H., 1991. The cAMP-regulated enhancer-binding protein ATF-1 activates transcription in response to cAMP-dependent protein kinase A. J. Biol. Chem. 266, 18431–18434. Riley, M., 1985. Pump and leak in regulation of fluid transport in rabbit cornea. Curr. Eye Res. 4, 371–376. Riley, M.V., Winkler, B.S., Starnes, C.A., Peters, M.I., 1996. Adenosine promotes regulation of corneal hydration through cyclic adenosine monophosphate. Invest. Ophthalmol. Vis. Sci. 37, 1–10. Riley, M.V., Winkler, B.S., Starnes, C.A., Peters, M.I., Dang, L., 1998. Regulation of corneal endothelial barrier function by adenosine, cyclic AMP, and protein kinases. Invest. Ophthalmol. Vis. Sci. 39, 2076–2084. Robaye, B., Ghanem, E., Wilkin, F., Fokan, D., Van Driessche, W., Schurmans, S., Boeynaems, J.M., Beauwens, R., 2003. Loss of nucleotide regulation of epithelial chloride transport in the jejunum of P2Y4-null mice. Mol. Pharmacol. 63, 777–783. Satpathy, M., Gallagher, P., Lizotte-Waniewski, M., Srinivas, S.P., 2004. Thrombin-induced inter-endothelial gap formation in bovine corneal endothelial cells. Exp. Eye Res. 2004; in press (doi: 10.1016/j.exer.2004.06.018). Servillo, G., Della Fazia, M.A., Sassone-Corsi, P., 2002. Coupling cAMP signaling to transcription in the liver: pivotal role of CREB and CREM. Exp. Cell. Res. 275, 143–154. Stevens, T., Garcia, J.G., Shasby, D.M., Bhattacharya, J., Malik, A.B., 2000. Mechanisms regulating endothelial cell barrier function. Am. J. Physiol. Lung Cell. Mol. Physiol. 279, L419–L422.
551
Stull, J.T., Tansey, M.G., Word, R.A., Kubota, Y., Kamm, K.E., 1991. Myosin light chain kinase phosphorylation: regulation of the Ca2Csensitivity of contractile elements. Adv. Exp. Med. Biol. 304, 129–138. Sun, X.C., Zhai, C.B., Cui, M., Chen, Y., Levin, L.R., Buck, J., Bonanno, J.A., 2003. HCO(3)(K)-dependent soluble adenylyl cyclase activates cystic fibrosis transmembrane conductance regulator in corneal endothelium. Am. J. Physiol. Cell Physiol. 284, C1114–C1122. Turner, J.R., 2000. Putting the squeeze on the tight junction: understanding cytoskeletal regulation. Semin. Cell Dev. Biol. 11, 301–308. Turner, J.R., Black, E.D., Ward, J., Tse, C.M., Uchwat, F.A., Alli, H.A., Donowitz, M., Madara, J.L., Angle, J.M., 2000. Transepithelial resistance can be regulated by the intestinal brush-border Na(C)/ H(C) exchanger NHE3. Am. J. Physiol. Cell Physiol 279, C1918– C1924. Turner, J.R., Rill, B.K., Carlson, S.L., Carnes, D., Kerner, R., Mrsny, R.J., Madara, J.L., 1997. Physiological regulation of epithelial tight junctions is associated with myosin light-chain phosphorylation. Am. J. Physiol. 273, C1378–C1385. Van Driessche, W., De Vos, R., Jans, D., Simaels, J., De Smet, P., Raskin, G., 1999. Transepithelial capacitance decrease reveals closure of lateral interspace in A6 epithelia. Pflugers Arch. 437, 680–690. van Hinsbergh, V.W., van Nieuw Amerongen, G.P., 2002. Intracellular signalling involved in modulating human endothelial barrier function. J. Anat. 200, 549–560. Walkenbach, R.J., Chao, W.T., 1985. Adenosine regulation of cyclic AMP in corneal endothelium. J. Ocul. Pharmacol. 1, 337–342. Wang, P., Verin, A.D., Birukova, A., Gilbert-McClain, L.I., Jacobs, K., Garcia, J.G., 2001. Mechanisms of sodium fluoride-induced endothelial cell barrier dysfunction: role of MLC phosphorylation. Am. J. Physiol. Lung Cell. Mol. Physiol. 281, L1472–L1483. Watsky, M.A., Guan, Z., Ragsdale, D.N., 1996. Effect of tumor necrosis factor alpha on rabbit corneal endothelial permeability. Invest. Ophthalmol. Vis. Sci. 37, 1924–1929. Wigham, C.G., Turner, H.C., Swan, J., Hodson, S.A., 2000. Modulation of corneal endothelial hydration control mechanisms by Rolipram. Pflugers Arch. 440, 866–870. Wolfs, R.C., Klaver, C.C., Vingerling, J.R., Grobbee, D.E., Hofman, A., de Jong, P.T., 1997. Distribution of central corneal thickness and its association with intraocular pressure: The Rotterdam Study. Am. J. Ophthalmol. 123, 767–772. Yi, D.H., Dana, M.R., 2002. Corneal edema after cataract surgery: incidence and etiology. Semin. Ophthalmol. 17, 110–114. Yuan, S.Y., 2000. Signal transduction pathways in enhanced microvascular permeability. Microcirculation 7, 395–403. Zhang, Y., Xie, Q., Sun, X.C., Bonanno, J.A., 2002. Enhancement of HCO(3)(K) permeability across the apical membrane of bovine corneal endothelium by multiple signaling pathways. Invest. Ophthalmol. Vis. Sci. 43, 1146–1153.
Adenosine induces dephosphorylation of myosin II regulatory light chain in cultured bovine corneal endothelial cells S.P. Srinivasa,*, M. Satpathya, P. Gallagherc, E. Larivie`reb, W. Van Driesscheb b
a School of Optometry, Indiana University, 800 Atwater Ave, Bloomington, IN 47405, USA Laboratory of Physiology, K.U. Leuven, Campus Gasthuisberg O/N, B-3000 Leuven, Belgium c Indiana University School of Medicine, Physiology, Indianapolis, IN, USA
Received 25 January 2004; accepted in revised form 25 June 2004 Available online 24 August 2004
Abstract Purpose: Dephosphorylation of the myosin II regulatory light chain (MLC) promotes barrier integrity of cellular monolayers through relaxation of the actin cytoskeleton. This study has investigated the influence of adenosine (ADO) on MLC phosphorylation in cultured bovine corneal endothelial cells (BCEC). Methods: MLC phosphorylation was assessed by urea-glycerol gel electrophoresis and immunoblotting. Elevation of cAMP in response to agonists of A2b receptors (subtype of P1 purinergic receptors) was confirmed by phosphorylation of the cAMP response element binding protein (CREB), which was determined by Western blotting. Activation of MAP kinases (i.e. activated ERK1 and ERK2) was assessed by Western blotting to examine their influence on MLC phosphorylation. Transepithelial electrical resistance (TER) of cells grown on porous filters was measured to assess the altered barrier integrity. Results: Exposure to ADO (200 mM; 30 min) and N-ethyl (carboxamido) adenosine (NECA; 50 mM; 30 min), known agonists of A2b receptors, induced phosphorylation of CREB similar to forskolin (FSK, 20 mM; 30 min), a direct activator of adenylate cyclase. Exposure to ADO, NECA, and FSK led to dephosphorylation of MLC by 51, 40, and 47%, respectively. ADO-induced dephosphorylation was dosedependent with as much as 31% dephosphorylation at 1 mM ADO. CGS-21680, a selective A2a agonist, neither induced MLC dephosphorylation nor CREB phosphorylation. ADO phosphorylated MAP kinases which could be prevented by exposure to the MAP kinase-specific inhibitor, U0126 (10 mM). NECA and FSK also induced ERK1 and ERK2 activation similar to ADO. Exposure to U0126 inhibited MLC phosphorylation under basal conditions by 17%. ADO-induced MLC dephosphorylation was enhanced by a simultaneous exposure to U0126 (25% increase in dephosphorylation). Exposure to ADO caused an increase in TER from 17 to 22 ohms cm2. Conclusions: (1) CREB phosphorylation in response to ADO and NECA, which indicates activation of the cAMP-PKA axis, suggests expression of A2b receptors in BCEC. (2) ERK1 and ERK2, activated by cAMP and A2b receptors, promote MLC phosphorylation. However, the net result of cAMP elevation is MLC dephosphorylation, presumably because the competing pathways involving inactivation of MLCK and/or ROCK are dominant (Rho-associated coiled coil-containing protein kinase or Rho kinase). (3) Consistent with MLC dephosphorylation, exposure to ADO increases TER, which suggests increased barrier integrity. q 2004 Elsevier Ltd. All rights reserved. Keywords: corneal endothelium; adenosine; myosin light chain kinase; cAMP; MLC phosphorylation
1. Introduction The corneal endothelium (CE), a monolayer on the posterior surface of the cornea, maintains the transparency
* Corresponding author. Address: Dr Srinivas, S.P. School of Optometry, Indiana University, 800 Atwater Ave, Bloomington, IN 47405, USA. E-mail address: [email protected] (S.P. Srinivas). 0014-4835/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. DOI:10.1016/j.exer.2004.06.027
of the tissue by controlling the hydration of the corneal stroma (Fischbarg et al., 1985; Riley, 1985). The hydrophilic glycosaminoglycans bound to stromal collagen imbibe water and thereby induce a fluid leak across the endothelium. This leakage is counterbalanced by an endothelial ‘fluid pump’ directed from the stroma to the aqueous humor. The tight junctions (TJs) of CE, although very leaky (trans-endothelial electrical resistance !25 ohms cm2) (Noske et al., 1994a,b), restrain excessive
544
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
fluid leak and thus form the basis for the barrier integrity of CE. The ‘fluid pump’ function of CE is sustained by a host of polarized and active ion transport mechanisms which couple net ionic fluxes to fluid transport through an obligatory osmotic coupling (Bonanno, 2003). Although much is known about the identity of ion transport mechanisms in CE (Bonanno, 2003), the mechanisms of cell signaling that regulate ion transport and barrier integrity are just beginning to be understood (Riley et al., 1996, 1998; Wigham et al., 2000; Zhang et al., 2002; Bonanno, 2003). Forskolin (FSK)-induced elevation of cAMP has been shown to increase conductance for ClK and HCOK through ion channels, including cystic fibrosis 3 transmembrane regulator (CFTR) (Bonanno and Srinivas, 1997; Sun et al., 2003). Riley et al. showed that adenosine (ADO) and FSK promote deturgescence of swollen rabbit corneas (Riley et al., 1996). Wigham et al. discovered that exposure of rabbit CE to a selective inhibitor of cAMPdependent phosphodiesterase (isoform PDE4), rolipram, leads to stromal thinning (Wigham et al., 2000). Apart from these effects on ion transport and corneal hydration, stimulation of the cAMP-PKA axis is also known to enhance barrier integrity (i.e. a decrease in paracellular permeability) of CE. Specifically, Riley et al. (1996, 1998) suggested that ADO, an endogenous agonist for P1 purinergic receptors, enhances the rate of deswelling of pre-swollen corneas through a cAMP-mediated increase in barrier integrity. This finding is also reflected in studies with the PDE4 inhibitor, rolipram (Wigham et al., 2000). The physiological role of the cAMP-PKA axis in CE has gained further prominence through recent studies showing expression of HCOK 3 -sensitive adenylate cyclase (the soluble isoform) (Sun et al., 2003). These observations on cAMP have prompted us to seek further elaboration on the putative role of the second messenger on the barrier integrity of CE. Several recent studies of the vascular endothelium (Garcia et al., 1995; Patterson et al., 2000; Stevens et al., 2000; van Hinsbergh and van Nieuw Amerongen, 2002) and certain epithelial monolayers (Turner et al., 1997; Turner, 2000) have demonstrated that contractility of the actin cytoskeleton plays an important role in the regulation of barrier integrity. Specifically, when a cell signaling cascade activates the motor protein myosin II by phosphorylating its regulatory light chain (MW: 20 kD; also referred to as myosin light chain or MLC), the resulting contractions of the actin cytoskeleton, anchored at TJs and adherent junctions (AJs), generate a centripetal force (Dudek and Garcia, 2001, 2002). This force opposes the intercellular tethering forces which, in turn, break down the cell–cell apposition necessary for interactions of the transmembrane proteins at the TJs and AJs. In the absence of these interactions, the occlusion of the paracellular pathway is not complete, and hence, barrier integrity is disrupted. In this study, we demonstrate that in cultured bovine corneal endothelial cells (BCEC), ADO and NECA, both agonists
for the A2b receptors, induce MLC dephosphorylation and increase barrier integrity.
2. Materials and methods 2.1. Cell culture Primary cultures of BCEC were established from fresh cow eyes as described earlier (Bonanno and Srinivas, 1997) in Dulbecco’s Modified Eagle’s Medium (supplemented with 10% fetal calf serum and an antibiotic–antimycotic mixture consisting of penicillin at 100 U mlK1, streptomycin at 100 mg mlK1 and fungizone at 0$25 mg mlK1) at 378C in a humidified atmosphere containing 5% CO2 and 95% air. The cells were fed every 2–3 days. Second or third passage cultures were grown to confluence on Transwelle filters (Corning Inc., Coburn, MA) or Petri dishes. 2.2. Quantitative estimation of MLC phosphorylation MLC phosphorylation was assayed with cells grown on Petri dishes by urea-glycerol gel electrophoresis followed by immunoblotting (Garcia et al., 1995). This protocol has been used in a number of studies involving vascular smooth muscle cells, trabecular meshwork endothelial cells, and vascular endothelial cells (Rao et al., 2001; Wang et al., 2001; Blue et al., 2002). After exposure to the specified agents, the reaction was terminated by the addition of 1 ml of ice-chilled PBS containing trichloroacetic acid and dithiothreitol (DTT). Cells were then scraped off and centrifuged for 10 min at 10 000g. The resulting pellet was washed with chilled acetone and dried. The precipitated protein was dissolved in a urea sample buffer containing 8 M urea, 10 mM DTT, 20 mM Tris-base, 23 mM glycine and 0$04% bromophenol blue. After a brief spin, the ureasolubilized sample was electrophoresed for 3 hr in 1$5 mm polyacrylamide slabs containing acrylamide, bis-acrylamide, glycerol, Tris-base, and glycine, at pH 8$6. Prior to loading the samples (25 mg of protein per lane), gels were subjected to pre-electrophoresis for 1$5 h. The reservoir buffer consisted of 20 mM Tris-base, 23 mM glycine, 2$96 mM thioglycolic acid, and 3 mM DTT. The separated phosphorylated and non-phosphorylated MLC were detected by immunoblotting using a polyclonal anti-MLC antibody (E201; 1:3000 dilution). In these gels, the diphosphorylated form (identified as ‘PP’ on gel photographs) migrated faster than the mono-phosphorylated form (identified as ‘P’ on gel photographs). Furthermore, the mono-phosphorylated form migrated faster than the nonphosphorylated form (identified as ‘NP’ on gel photographs). These migration patterns are typical of urea-glycerol gels (Garcia et al., 1995). Blots were washed with TBST three times (15 min each) then visualized using a peroxidase conjugated secondary antibody and an enhanced chemiluminescence kit (Amersham). The bands were
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
scanned into TIF images and band intensities were quantified using a custom-made software program. The program subtracted the background before quantifying the pixel intensity of each band associated with non-phosphorylated (denoted by iNP), mono-phosporylated (iP) and di-phosphorylated (iPP) forms. The fraction of the MLC phosphorylated form (i.e. moles of MLC phosphorylated/ total moles of MLC) was calculated by dividing the total MLC phosphorylated (equal to iPC[2!iPP]) by the total MLC (equal to iPCiPPCiNP) (Garcia et al., 1995; Wang et al., 2001). 2.3. CREB phosphorylation by Western blotting Phosphorylation of the cyclic AMP response element binding protein (CREB) was assessed by Western blotting. Endothelial cells from 100 mm Petri dishes were treated with desired agents and then lysed with an ice-cold lysis buffer containing 50 mM Tris HCl (pH 7$6), 150 mM NaCl, 1% Triton X-100, 2 mM EGTA (pH 8$0), EDTA, DTT, PMSF, NaF, and a protease inhibitor cocktail (containing leupeptin, pepstatin, chymostatin, and aprotinin). After a pre-clearing centrifugation step (14,000 rpm for 15 min at 48C), the whole lysate was subjected to SDS-PAGE (100 mg protein per lane) and immunoblotting. For experiments with CGS-21,680, nuclear extracts were prepared using Active Motif (Carlsbad, CA; Cat. No. 40010). Membranes were blocked with TBST containing 5% nonfat dry milk for 1 hr at room temperature, and subsequently incubated with primary antibody that had been diluted in TBST and dry milk overnight at 48C. Blots were washed with TBST three times (15 min each) and visualized using a peroxidase conjugated secondary antibody and an enhanced chemiluminescence kit (Amersham).
545
purchased from Amersham-Pharmacia Biotech (Piscataway, NJ). All other drugs/reagents were purchased from Sigma (St Louis, MO). 2.6. Data analysis Unpaired t-tests were used to compare the effects of various drugs at p!0$05. Data are expressed as meansGSE.
3. Results 3.1. Adenosine-induced CREB phosphorylation Among A1, A2a, A2b, and A3 subtypes of ADO-sensitive receptors, A2a and A2b subtypes are coupled to the stimulatory G-protein (Gas) (Feoktistov and Biaggioni, 1997; Fredholm et al., 2001; Rees et al., 2003). To identify the receptor subtype(s) expressed in BCEC, we investigated phosphorylation of CREB, a transcription factor phosphorylated by protein kinase A (PKA) (Collins et al., 1990; Rehfuss et al., 1991; Fitzgerald et al., 1999; Servillo et al., 2002). Cells were serum-starved overnight and exposed for 30 min to ADO (200 mM) and NECA (50 mM, selective A2b agonist) (Grant et al., 2001). The whole-cell protein extracts were then subjected to Western blotting using an antibody to detect phosphorylated CREB (p-CREB). However, the antibody also detected the activated transcription factor 1 (ATF-1), as noted by the manufacturer. Fig. 1A is a typical gel showing CREB and ATF-1 activation (phosphorylation) in response to ADO (Lane 3) and NECA (Lane 4). As expected, exposure to FSK, a direct activator of adenylate
2.4. Barrier integrity by impedance analysis Impedance analysis was used to determine the resistance of the BCEC monolayer grown on Transwell porous supports, as described earlier (Van Driessche et al., 1999; Robaye et al., 2003). The experimental procedure consisted of mounting the porous support with cultured BCEC between two halves of an Ussing chamber (Robaye et al., 2003). KCl electrodes, connected to the solution via short agar bridges, were used for measuring the potential difference and injecting current. The volume of each compartment bathing the cells was 2 ml. A HCOK 3 -free Ringers at 378C and pH 7$4 was perfused through each compartment. The Ringers was composed of 140 mM NaC, 4 mM KC, 1$2 mM Mg2C, 1$2 mM Ca2C, 120 mM ClK, 2$8 mM PO2K 4 , 25 mM gluconate, and 6$5 mM glucose. 2.5. Drugs and chemicals Bovine calf serum was purchased from Hyclone (Logan, UT). An enhanced chemiluminescence reagent was
Fig. 1. A2b agonists, ADO and NECA, phosphorylate PKA-sensitive transcription factors CREB and ATF-1: Western blotting analysis of CREB and ATF-1 phosphorylation in response to ADO (200 mM, 30 min), NECA (50 mM, 30 min) and CGS-21680 (CGS; 50 mM; 30 min). As a positive control, response to forskolin (FSK; 30 mM, 30 min) is also shown. Western blotting was performed with an antibody specific for Ser133 phosphorylated CREB (p-CREB) that also reacted with ATF-1. Panel A: Cells were treated with agents specified and whole cell extracts (100 mg per lane) were electrophoresed on SDS-PAGE. Panel B: Cells were treated with the agents specified and nuclear extracts (100 mg per lane) were electrophoresed on SDS-PAGE. Lanes ‘MC’ and ‘MC(EGF/FSK)’ are positive controls for p-CREB/ATF-1 provided by the manufacturer. Lane ‘C’ represents BCEC under basal conditions.
546
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
cyclase, also led to CREB activation (Lane 5). Lane 1 (identified as MC(EGF)) represents positive control from the manufacturer in which the cells were treated with EGF. Lane 2 (identified as MC), represents cells without EGF treatment, also supplied by the manufacturer. Activation of CREB and ATF-1 in response to both ADO and NECA, similar to FSK, is consistent with activation of cAMP-PKA axis by A2b receptors in BCEC. The gel in Panel B obtained from nuclear extracts shows a lack of activation of CREB or ATF-1 in response to CGS-21680 (A2a selective agonist). 3.2. cAMP-mediated MLC dephosphorylation Myosin light chain kinase (MLCK), a dedicated protein kinase for MLC phosphorylation, (Stull et al., 1991) is phosphorylated by PKA, leading to its inactivation (Garcia et al., 1997) and reduced MLC phosphorylation. In addition, Rho kinase (ROCK; Rho-associated coiled coil-containing protein kinase) activity is reduced by PKA activation through inhibition of the dissociation of RhoA from guanine nucleotide dissociation inhibitor (GDI) (Qiao et al., 2003). Since Rho kinase phosphorylates the myosin light chain phosphatase (MLCP) and promotes its inactivation, the net effect of PKA activation is MLC
Fig. 2. MLC dephosphorylation in response to P1 purinergic agonists: Panel A: Cells were treated with ADO (100 mM; 30 min), NECA (20 mM; 18 min), and CGS-21680 (CGS; 10 mM; 18 min). FSK (30 mM, 18 min) was used as a positive control (Lane ‘CC’). Basal phosphorylation in untreated cells is shown in Lane ‘C’. CGS-21680 (CGS), selective for A2a purinergic receptors, did not have a significant effect on MLC phosphorylation. Panel B: Densitometric analysis of data from experiments similar to those shown in Panel A. The histogram shows the meanGSE from 5 to 12 independent experiments. Notes: Nonphosphorylated MLC; P: Monophopshorylated MLC; PP: Diphosphorylated MLC. ‘*’ denotes p!0$05 for comparison of mean values with respect to the control. Note CGS-21680-induced dephosphorylation is not significant when compared to the control (pZ0$23).
dephosphorylation (Qiao et al., 2003). Since activation of the cAMP-PKA axis was shown above, we investigated the influence of ADO and NECA on the phosphorylation status of MLC in BCEC by using urea-glycerol gel electrophoresis. Similar to CREB phosphorylation experiments, cells were serum-starved overnight and then exposed to the ADO and NECA for 18 min. Both agonists induced an apparent MLC desphosphorylation (Lanes 2 and 4, respectively, in Fig. 2, labeled ADO and NECA) compared to cells under basal conditions (Lane 1, labeled C). FSK also induced dephosphorylation (Lane 3, Fig. 2, and labeled FSK). Similar experiments with the A2a agonist CGS-21680 (100 mM; 18 min) did not show significant dephosphorylation (Fig. 2; Lane 5; labeled CGS) compared to the control. The histogram in Fig. 2B, obtained by densitometric analysis of gels similar to that shown in Fig. 2A, clearly demonstrates that A2b agonists induce significant MLC dephosphorylation. The extent of desphosphorylation by A2b agonists is similar to that by FSK. On the other hand, dephosphorylation induced by CGS-21680 is not significant compared to untreated controls. The results in Fig. 2B also suggest that NECA is more potent than ADO. Fig. 3 is a summary of experiments showing time-dependence of ADO-induced MLC dephosphorylation. Maximum dephosphorylation was reached at about 18 min and then a reversal towards the baseline was noted (see Lane 5 in Fig. 3A; labeled
Fig. 3. Time course of ADO-induced MLC dephosphorylation: Cells were treated with ADO (100 mM) for 2, 5, 10, 18, and 30 min. Densitometric analysis of the data is plotted on the Y-axis, indicating the extent of MLC phosphorylation. The histogram shows the meanGSE from four independent experiments. MLC phosphorylation status of serum-starved cells was taken as 100%. ‘*’ denotes p!0$05 for comparison of mean values with respect to the control. ‘**’ denotes that ADO-induced dephosphorylation is not significant when compared to the control (pZ0$46).
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
547
30 min and Fig. 3B). Fig. 4 is the summary of doseresponse data obtained with ADO at three concentrations. Clearly, the extent of dephosphorylation increased with the doses of greater concentrations (Fig. 4B). 3.3. ADO-mediated phosphorylation of ERK1 and ERK2 ERK1 and ERK2 promote MLC phosphorylation by increasing the catalytic efficiency of MLCK (Stull et al., 1991; Klemke et al., 1997). To confirm these in BCEC, we first investigated NECA and ADO-induced ERK1 and ERK2 activation. Consistent with previous reports (Gao et al., 1999; Grant et al., 2001), ADO and NECA induced ERK1 and ERK2 activation after 18 min of exposure (Fig. 5A). This activation is inhibited by U0126 (10 mM; Fig. 5B), a selective inhibitor of MEK which is upstream from ERK1 and ERK2 activation. U0126 also reduced basal MLC phosphorylation by 17% (Fig. 6A; Lane 4, labeled U). Exposure to ADO in the presence of U0126 caused dephosphorylation of MLC to a greater extent than with ADO alone (Fig. 6A, Lane 3; labeled UCADO). A summary of similar experiments is given in Fig. 6B. 3.4. ADO-induced increase in barrier integrity In order to examine the effect of ADO on barrier integrity, we measured TER using impedance analysis. The analysis was based on modeling the monolayer as a parallel
Fig. 4. Dose response of adenosine-induced MLC dephosphorylation: Cells were treated with ADO for 10 min at 1, 10, and 100 mM. The gels resulting from urea-glycerol gel electrophoresis and Western blotting were subjected to densitometric analysis. The resulting data are shown as a histogram with the Y-axis representing the extent of MLC phosphorylation (meanGSE) from five independent experiments. Note that the phosphorylation under basal conditions (i.e. serum starved cells) was taken as 100%. ‘*’ denotes p!0$05 for comparison of mean values with respect to the control.
Fig. 5. Phosphorylation of ERK1 and ERK2 in response to ADO. Cells were treated with ADO (100 mM) and NECA (10mM) for 30 min. FSK (10 mM; 30 min) was used as the positive control. Equal amounts of protein were loaded on SDS-PAGE and phosphorylated ERK1 and ERK2 were examined by Western blotting. The gel at the top shows enhanced phosphorylation of both ERK1 and ERK2 in response to ADO, FSK, and NECA. The gel at the bottom of the figure shows sensitivity of the phosphorylation to MAPK inhibitor U0126. Note that U0126 (10 mM) inhibited activation of ERK1 and ERK2 under basal conditions as well as in the presence of ADO. The results shown are typical of five independent experiments.
circuit of a capacitance (C) and resistance (TER) in series with the solution resistance (Rsol) between the voltage electrodes (Van Driessche et al., 1999; Robaye et al., 2003). The corresponding impedance spectrum is a semicircle in
Fig. 6. Effect of ERK1 and ERK2 activation on MLC phosphorylation: Cells starved of serum overnight were treated with ADO (100 mM, 18 min). In some experiments, cells were pre-treated with U0126 alone or U0126 (10 mM, 10 min) followed by ADO (100 mM, 18 min). MLC phosphorylation was further reduced when cells were exposed to the specific MAP kinase inhibitor, U0126. ‘*’ denotes p!0$05 for comparison of mean values with respect to the control. ‘**’ denotes p!0$05 for comparison of mean values between ADO and ‘ADOCU0126’. The results shown are typical of 3–4 independent experiments.
548
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
Fig. 7. Increase in transendothelial electrical resistance after exposure to ADO: BCEC, grown on porous matrices and mounted in an Ussing chamber, were continuously perfused with HCOK 3 -free Ringers. The steadystate TER of the monolayer was measured by impedance analysis under resting conditions and subsequently after exposure to ADO (200 mM). The results shown are a summary from five independent experiments. The increase in resistance at 10, 20 and 30 min after exposure to ADO is statistically significant (p!0$05) with respect to resting conditions.
a Nyquist plot (Robaye et al., 2003). In such a plot, the intercepts of the semicircle at high and low frequencies correspond to Rsol and RsolCTER, respectively. We recorded impedance spectra before and after exposure to ADO. As described by Van Driessche et al. (1999) and Robaye et al. (2003), the Cole–Cole equation was fitted to the data with a custom-made computer program and extrapolation led to an estimate of TER. Consistent with previous studies (Riley et al., 1998; Wigham et al., 2000), exposure to ADO resulted in a significant increase in TER from 17 to 22 ohms cm2. Time course of the change in TER is summarized from six independent experiments in Fig. 7. The increase in TER was rapid and reached a maximum in about 15 min. The increase in TER persisted in the presence of ADO (observed for !30 min). After removal of ADO from the bath, TER showed a tendency to decrease (data not shown).
4. Discussion ADO stimulates fluid transport (Fischbarg et al., 1977) and promotes barrier integrity in CE (Riley et al., 1996, 1998). These effects contribute to an enhancement in stromal hydration control (Riley et al., 1996). ADO-induced stimulation of fluid transport may involve anion channels sensitive to cAMP (Bonanno and Srinivas, 1997; Zhang et al., 2002; Bonanno, 2003). However, a mechanistic basis for the enhancement of barrier integrity has not been ascertained. This study investigated MLC phosphorylation in response to ADO, since increased MLC phosphorylation is associated with a breakdown of barrier integrity and,
dephosphorylation is known to overcome barrier dysfunction (Turner et al., 1997; Patterson et al., 2000; Stevens et al., 2000; Turner, 2000; van Hinsbergh and van Nieuw Amerongen, 2002). This phenomenon is attributed to the fact that elevated MLC phosphorylation increases the contractility of the actin cytoskeleton and leads to the development of a centripetal force that opposes the forces of tethering between neighboring cells. These forces are necessary for interactions of the transmembrane proteins at tight and AJs (Dudek and Garcia, 2001; Dull and Garcia, 2002). Experiments in this study show that (a) ADO stimulates the cAMP-PKA axis in BCEC as evidenced by phosphorylation of CREB, (b) ADO induces significant MLC dephosphorylation, (c) NECA, a potent agonist for the A2b receptors, also induces CREB phosphorylation as well as MLC dephosphorylation, and (d) ADO increases the TER of BCEC, indicating an enhancement in barrier integrity. 4.1. Significance of adenosine-induced CREB phosphorylation Expression of ADO-sensitive receptors in CE has been suspected for a long time. Walkenbach showed cAMP elevation in BCEC when challenged by ADO at 1 mM to 0$5 mM (Walkenbach and Chao, 1985). Riley et al. noted the elevation of cAMP in response to ADO in rabbit CE (Riley et al., 1996). Despite these observations, further confirmation of its identity at the mRNA level and pharmacological profiling has been elusive. CREB phosphorylation is a hallmark of elevated cAMP and the subsequent activation of PKA. Accordingly, CREB phosphorylation has been used previously, for example, to demonstrate the activation of b-adrenergic receptors (Collins et al., 1990; Fitzgerald et al., 1999). Thus, our findings in Fig. 1 showing CREB phosphorylation in response to ADO and NECA suggest the expression of A2b receptors in BCEC. Our protocols were validated using FSK and positive controls provided by the manufacturer of the p-CREB antibody. Activation of the transcription factor ATF-1 is also consistent with activation of the cAMP-PKA axis in response to ADO (Rehfuss et al., 1991; Servillo et al., 2002). The fact that NECA also induced CREB phosphorylation further indicates the expression of A2b receptors in BCEC since NECA is a selective agonist of the receptor. 4.2. The mechanism underlying MLC dephosphorylation MLCK is expressed as smooth muscle (SM-MLCK; 130 kD) and vascular endothelial isoforms (EC-MLCK; 220 kD) in BCEC (Satpathy et al., 2004). EC-MLCK is known to possess a number of regulatory sites, including a consensus sequence for phosphorylation by PKA. In fact, in vitro studies have shown inactivation of EC-MLCK upon phosphorylation by PKA (Garcia et al., 1997). Accordingly, dephosphorylation of MLC has been reported in response to
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
549
activation of the cAMP-PKA axis in vascular endothelium (Essler et al., 2000; Patterson et al., 2000; van Hinsbergh and van Nieuw Amerongen, 2002). In our experiments, as summarized in Fig. 2, agents that elevated cAMP in BCEC, including FSK, ADO and NECA, induced MLC dephosphorylation. As shown in Figs. 3 and 4, dephosphorylation was rapid and dose-dependent, lasting for at least 18 min. In addition to the inactivation of MLCK, recent studies also indicate that PKA upregulates the activity of the MLCP which opposes MLCK-mediated MLC phosphorylation (Kamm and Stull, 2001). Specifically, Qiao et al. (2003) showed that PKA/cAMP could inhibit activation of RhoA upstream of Rho kinase, which phosphorylates the catalytic subunit of MLCP and down-regulates the activity of the enzyme. Thus, it is possible that the MLC dephosphorylation observed in this study is due to the inactivation of Rho kinase secondary to inhibition of activation of RhoA. Our previous study on thrombin-mediated MLC phosphorylation showed that the expression of Rho kinase (ROCK-1 isoform) in BCEC and, moreover, that the inhibition of Rho kinase by its selective inhibitor Y-27632 was sufficient to block MLC phosphorylation significantly under basal conditions (Satpathy et al., 2004). 4.3. Significance of adenosine-induced phosphorylation of ERK1 and ERK2 Despite MLC dephosphorylation, as discussed above, experiments illustrated in Figs. 5 and 6 unraveled activation of another opposing pathway in response to ADO. As noted earlier, activation of ERK1 and ERK2 is a hallmark of a number of ADO-sensitive receptors, including the A2b subtype (Feoktistov and Biaggioni, 1997; Gao et al., 1999; Grant et al., 2001). Although an exact mechanism underlying activation of ERK1 and ERK2 is not revealed by our experiments, it is clear that the elevation of cAMP itself is sufficient since FSK also induced activation of ERK1 and ERK2 (Fig. 5). In regard to the positive influence of ERK1 and ERK2 on MLC phosphorylation, our results are consistent with previous reports (Klemke et al., 1997; Kamm and Stull, 2001). Specifically, Fig. 6 clearly shows that U0126, a selective inhibitor of MAPK kinase, reduced MLC phosphorylation of cells under resting conditions. However, in the presence of ADO, activation of the cAMPPKA axis overwhelms the positive influences of ERK1 and ERK2 in promoting MLC phosphorylation (Fig. 6). In Fig. 8, we present a summary of the pathways likely to be involved in ADO-induced MLC dephosphorylation. 4.4. Adenosine-induced increase in barrier integrity The tight junctional permeability, a principal determinant of barrier integrity, is regulated by a number of protein kinases: (a) directly through phosphorylation of the transmembrane proteins (i.e. occludins and claudins and b-catenin), and/or (b) indirectly through reorganization of
Fig. 8. Schematic of the cell signaling mediated by ADO in CE for the control of barrier integrity: Activation of A2b receptors is coupled to the elevation of intracellular cAMP, leading to the activation of PKA. Direct phosphorylation of MLCK leads to its inactivation and hence would reduce MLC phosphorylation. Alternatively, MLC dephosphorylation could be induced by an inhibition of RhoA activation leading reduced Rho kinase activity. The reduced extent of MLC phosphorylation induces relaxation of the actin cytoskeleton, leading to an enhancement in barrier integrity.
the actin cytoskeleton (Yuan, 2000). The latter, which has received much attention recently, is achieved by altering the phosphorylation status of MLC (Turner et al., 1997; Stevens et al., 2000; Turner, 2000; Turner et al., 2000; van Hinsbergh and van Nieuw Amerongen, 2002). Apart from a few exceptions noted to date (Yuan, 2000), an increase in the contractility of the actin cytoskeleton is associated with a breakdown of barrier integrity (Turner et al., 1997; Stevens et al., 2000; Turner, 2000; Turner et al., 2000; van Hinsbergh and van Nieuw Amerongen, 2002). Dephosphorylation, on the other hand, has been shown to cause relaxation of the actin cytoskeleton and to overcome barrier dysfunction concomitantly (Noll et al., 1999, 2000) when induced by agents such as thrombin, histamine, and lipopolysaccharides (Essler et al., 2000). This is consistent with observations on rabbit CE (Riley et al., 1998; Wigham et al., 2000). BCEC also showed an increase in TER in response to ADO (Fig. 7). The increase in resistance was rapid and matched the time course of MLC phosphorylation (Fig. 3). The average TER under resting conditions (17 ohms cm2; Fig. 7) is similar to values reported earlier by Noske in bovine eyes
550
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551
(20 ohms cm2) (Noske et al., 1994b). Furthermore, our findings and those by Riley et al. (1998) and Wigham et al. (2000) contradict those reported by Le Varlet et al. (1995) who claimed an increase in paracellular permeability and a decrease in TER of BCEC in response to dibutyryl-cAMP and FSK. 4.5. Significance to stromal hydration Since there is no apparent interconnection between regulation of barrier integrity and fluid transport in CE, it appears that hyper-permeability of the endothelium at a constant imbibition pressure would result in uncontrolled stromal hydration. This is evident in an increase in corneal thickness when inflammation occurs (Watsky et al., 1996; Behar-Cohen et al., 1998; Edelhauser, 2000; Yi and Dana, 2002) and, indirectly, when the net hydraulic pressure favoring fluid influx across the endothelium is increased by ocular hypertension (Herndon et al., 1997; Wolfs et al., 1997; Bron et al., 1999; Copt et al., 1999). Under these conditions, it may be therapeutically useful to pharmacologically alter endothelial barrier dysfunction and overcome corneal swelling. To this end, the results from this study, taken together, suggest that pharmacological approaches capable of elevating intracellular cAMP in CE would be useful. In summary, this study has demonstrated that ADO and its A2b-sensitive analogs induce MLC dephosphorylation in CE. Concomitant with this dephosphorylation, an increase in barrier integrity is promoted. These findings offer a potential strategy for using A2b receptor agonists in preventing corneal swelling, especially for counteracting enhanced MLC phosphorylation induced by inflammatory mediators.
Acknowledgements Supported by NIH grant EY11107 and EY14415 (SPS).
References Behar-Cohen, F.F., Savoldelli, M., Parel, J.M., Goureau, O., Thillaye-Goldenberg, B., Courtois, Y., Pouliquen, Y., de Kozak, Y., 1998. Reduction of corneal edema in endotoxininduced uveitis after application of L-NAME as nitric oxide synthase inhibitor in rats by iontophoresis. Invest. Ophthalmol. Vis. Sci. 39, 897–904. Blue, E.K., Goeckeler, Z.M., Jin, Y., Hou, L., Dixon, S.A., Herring, B.P., Wysolmerski, R.B., Gallagher, P.J., 2002. 220and 130-kDa MLCKs have distinct tissue distributions and intracellular localization patterns. Am. J. Physiol. Cell Physiol. 282, C451–C460. Bonanno, J.A., 2003. Identity and regulation of ion transport mechanisms in the corneal endothelium. Prog. Retin. Eye Res. 22, 69–94. Bonanno, J.A., Srinivas, S.P., 1997. Cyclic AMP activates anion channels in cultured bovine corneal endothelial cells. Exp. Eye Res. 64, 953–962.
Bron, A.M., Creuzot-Garcher, C., Goudeau-Boutillon, S., d’Athis, P., 1999. Falsely elevated intraocular pressure due to increased central corneal thickness. Graefes Arch. Clin. Exp. Ophthalmol. 237, 220–224. Collins, S., Altschmied, J., Herbsman, O., Caron, M.G., Mellon, P.L., Lefkowitz, R.J., 1990. A cAMP response element in the beta 2-adrenergic receptor gene confers transcriptional autoregulation by cAMP. J. Biol. Chem. 265, 0–5. Copt, R.P., Thomas, R., Mermoud, A., 1999. Corneal thickness in ocular hypertension, primary open-angle glaucoma, and normal tension glaucoma. Arch. Ophthalmol. 117, 14–16. Dudek, S.M., Garcia, J.G., 2001. Cytoskeletal regulation of pulmonary vascular permeability. J. Appl. Physiol. 91, 1487–1500. Dull, R.O., Garcia, J.G., 2002. Leukocyte-induced microvascular permeability: how contractile tweaks lead to leaks. Circ. Res. 90, 1143–1144. Edelhauser, H.F., 2000. The resiliency of the corneal endothelium to refractive and intraocular surgery. Cornea 19, 263–273. Essler, M., Staddon, J.M., Weber, P.C., Aepfelbacher, M., 2000. Cyclic AMP blocks bacterial lipopolysaccharide-induced myosin light chain phosphorylation in endothelial cells through inhibition of Rho/Rho kinase signaling. J. Immunol. 164, 6543–6549. Feoktistov, I., Biaggioni, I., 1997. Adenosine A2B receptors. Pharmacol. Rev. 49, 381–402. Fischbarg, J., Hernandez, J., Liebovitch, L.S., Koniarek, J.P., 1985. The mechanism of fluid and electrolyte transport across corneal endothelium: critical revision and update of a model. Curr. Eye Res. 4, 351–360. Fischbarg, J., Lim, J.J., Bourguet, J., 1977. Adenosine stimulation of fluid transport across rabbit corneal endothelium. J. Membr. Biol. 35, 95–112. Fitzgerald, L.R., Mannan, I.J., Dytko, G.M., Wu, H.L., Nambi, P., 1999. Measurement of responses from Gi-, Gs-, or Gq-coupled receptors by a multiple response element/cAMP response element-directed reporter assay. Anal. Biochem. 275, 54–61. Fredholm, B.B., AP, I.J., Jacobson, K.A., Klotz, K.N., Linden, J., 2001. International Union of Pharmacology. XXV. Nomenclature and classification of adenosine receptors. Pharmacol. Rev. 53, 527–552. Gao, Z., Chen, T., Weber, M.J., Linden, J., 1999. A2B adenosine and P2Y2 receptors stimulate mitogen-activated protein kinase in human embryonic kidney-293 cells. Cross-talk between cyclic AMP and protein kinase c pathways. J. Biol. Chem. 274, 5972–5980. Garcia, J.G., Davis, H.W., Patterson, C.E., 1995. Regulation of endothelial cell gap formation and barrier dysfunction: role of myosin light chain phosphorylation. J. Cell. Physiol. 163, 510–522. Garcia, J.G., Lazar, V., Gilbert-McClain, L.I., Gallagher, P.J., Verin, A.D., 1997. Myosin light chain kinase in endothelium: molecular cloning and regulation. Am. J. Respir. Cell Mol. Biol. 16, 489–494. Grant, M.B., Davis, M.I., Caballero, S., Feoktistov, I., Biaggioni, I., Belardinelli, L., 2001. Proliferation, migration, and ERK activation in human retinal endothelial cells through A(2B) adenosine receptor stimulation. Invest. Ophthalmol. Vis. Sci. 42, 2068–2073. Herndon, L.W., Choudhri, S.A., Cox, T., Damji, K.F., Shields, M.B., Allingham, R.R., 1997. Central corneal thickness in normal, glaucomatous, and ocular hypertensive eyes. Arch. Ophthalmol. 115, 1137–1141. Kamm, K.E., Stull, J.T., 2001. Dedicated myosin light chain kinases with diverse cellular functions. J. Biol. Chem. 276, 4527–4530. Klemke, R.L., Cai, S., Giannini, A.L., Gallagher, P.J., de Lanerolle, P., Cheresh, D.A., 1997. Regulation of cell motility by mitogen-activated protein kinase. J. Cell. Biol. 137, 481–492. Le Varlet, B., Ducroc, R., Dagonet, F.B., Pouliquen, Y., Vandewalle, A., Hirsch, M., 1995. Dibutyryl cyclic adenosine monophosphate and forskolin alter the paracellular pathway in cultured corneal endothelial cells. Invest. Ophthalmol. Vis. Sci. 36, 2503–2513.
S.P. Srinivas et al. / Experimental Eye Research 79 (2004) 543–551 Noll, T., Holschermann, H., Koprek, K., Gunduz, D., Haberbosch, W., Tillmanns, H., Piper, H.M., 1999. ATP reduces macromolecule permeability of endothelial monolayers despite increasing [Ca2C]i. Am. J. Physiol. 276, H1892–H1901. Noll, T., Schafer, M., Schavier-Schmitz, U., Piper, H.M., 2000. ATP induces dephosphorylation of myosin light chain in endothelial cells. Am. J. Physiol. Cell Physiol. 279, C717–C723. Noske, W., Fromm, M., Levarlet, B., Kreusel, K.M., Hirsch, M., 1994a. Tight junctions of the human corneal endothelium: morphological and electrophysiological features. Ger. J. Ophthalmol. 3, 253–257. Noske, W., Levarlet, B., Kreusel, K.M., Fromm, M., Hirsch, M., 1994b. Tight junctions and paracellular permeability in cultured bovine corneal endothelial cells. Graefes Arch. Clin. Exp. Ophthalmol. 232, 608–613. Patterson, C.E., Lum, H., Schaphorst, K.L., Verin, A.D., Garcia, J.G., 2000. Regulation of endothelial barrier function by the cAMP-dependent protein kinase. Endothelium 7, 287–308. Qiao, J., Huang, F., Lum, H., 2003. cAMP-dependent protein kinase inhibits RhoA activation: a protection mechanism against endothelial barrier dysfunction. Am. J. Physiol. Lung Cell. Mol. Physiol. 2003;. Rao, P.V., Deng, P.F., Kumar, J., Epstein, D.L., 2001. Modulation of aqueous humor outflow facility by the Rho kinase-specific inhibitor Y-27632. Invest. Ophthalmol. Vis. Sci. 42, 1029–1037. Rees, D.A., Scanlon, M.F., Ham, J., 2003. Adenosine signalling pathways in the pituitary gland: one ligand, multiple receptors. J. Endocrinol. 177, 357–364. Rehfuss, R.P., Walton, K.M., Loriaux, M.M., Goodman, R.H., 1991. The cAMP-regulated enhancer-binding protein ATF-1 activates transcription in response to cAMP-dependent protein kinase A. J. Biol. Chem. 266, 18431–18434. Riley, M., 1985. Pump and leak in regulation of fluid transport in rabbit cornea. Curr. Eye Res. 4, 371–376. Riley, M.V., Winkler, B.S., Starnes, C.A., Peters, M.I., 1996. Adenosine promotes regulation of corneal hydration through cyclic adenosine monophosphate. Invest. Ophthalmol. Vis. Sci. 37, 1–10. Riley, M.V., Winkler, B.S., Starnes, C.A., Peters, M.I., Dang, L., 1998. Regulation of corneal endothelial barrier function by adenosine, cyclic AMP, and protein kinases. Invest. Ophthalmol. Vis. Sci. 39, 2076–2084. Robaye, B., Ghanem, E., Wilkin, F., Fokan, D., Van Driessche, W., Schurmans, S., Boeynaems, J.M., Beauwens, R., 2003. Loss of nucleotide regulation of epithelial chloride transport in the jejunum of P2Y4-null mice. Mol. Pharmacol. 63, 777–783. Satpathy, M., Gallagher, P., Lizotte-Waniewski, M., Srinivas, S.P., 2004. Thrombin-induced inter-endothelial gap formation in bovine corneal endothelial cells. Exp. Eye Res. 2004; in press (doi: 10.1016/j.exer.2004.06.018). Servillo, G., Della Fazia, M.A., Sassone-Corsi, P., 2002. Coupling cAMP signaling to transcription in the liver: pivotal role of CREB and CREM. Exp. Cell. Res. 275, 143–154. Stevens, T., Garcia, J.G., Shasby, D.M., Bhattacharya, J., Malik, A.B., 2000. Mechanisms regulating endothelial cell barrier function. Am. J. Physiol. Lung Cell. Mol. Physiol. 279, L419–L422.
551
Stull, J.T., Tansey, M.G., Word, R.A., Kubota, Y., Kamm, K.E., 1991. Myosin light chain kinase phosphorylation: regulation of the Ca2Csensitivity of contractile elements. Adv. Exp. Med. Biol. 304, 129–138. Sun, X.C., Zhai, C.B., Cui, M., Chen, Y., Levin, L.R., Buck, J., Bonanno, J.A., 2003. HCO(3)(K)-dependent soluble adenylyl cyclase activates cystic fibrosis transmembrane conductance regulator in corneal endothelium. Am. J. Physiol. Cell Physiol. 284, C1114–C1122. Turner, J.R., 2000. Putting the squeeze on the tight junction: understanding cytoskeletal regulation. Semin. Cell Dev. Biol. 11, 301–308. Turner, J.R., Black, E.D., Ward, J., Tse, C.M., Uchwat, F.A., Alli, H.A., Donowitz, M., Madara, J.L., Angle, J.M., 2000. Transepithelial resistance can be regulated by the intestinal brush-border Na(C)/ H(C) exchanger NHE3. Am. J. Physiol. Cell Physiol 279, C1918– C1924. Turner, J.R., Rill, B.K., Carlson, S.L., Carnes, D., Kerner, R., Mrsny, R.J., Madara, J.L., 1997. Physiological regulation of epithelial tight junctions is associated with myosin light-chain phosphorylation. Am. J. Physiol. 273, C1378–C1385. Van Driessche, W., De Vos, R., Jans, D., Simaels, J., De Smet, P., Raskin, G., 1999. Transepithelial capacitance decrease reveals closure of lateral interspace in A6 epithelia. Pflugers Arch. 437, 680–690. van Hinsbergh, V.W., van Nieuw Amerongen, G.P., 2002. Intracellular signalling involved in modulating human endothelial barrier function. J. Anat. 200, 549–560. Walkenbach, R.J., Chao, W.T., 1985. Adenosine regulation of cyclic AMP in corneal endothelium. J. Ocul. Pharmacol. 1, 337–342. Wang, P., Verin, A.D., Birukova, A., Gilbert-McClain, L.I., Jacobs, K., Garcia, J.G., 2001. Mechanisms of sodium fluoride-induced endothelial cell barrier dysfunction: role of MLC phosphorylation. Am. J. Physiol. Lung Cell. Mol. Physiol. 281, L1472–L1483. Watsky, M.A., Guan, Z., Ragsdale, D.N., 1996. Effect of tumor necrosis factor alpha on rabbit corneal endothelial permeability. Invest. Ophthalmol. Vis. Sci. 37, 1924–1929. Wigham, C.G., Turner, H.C., Swan, J., Hodson, S.A., 2000. Modulation of corneal endothelial hydration control mechanisms by Rolipram. Pflugers Arch. 440, 866–870. Wolfs, R.C., Klaver, C.C., Vingerling, J.R., Grobbee, D.E., Hofman, A., de Jong, P.T., 1997. Distribution of central corneal thickness and its association with intraocular pressure: The Rotterdam Study. Am. J. Ophthalmol. 123, 767–772. Yi, D.H., Dana, M.R., 2002. Corneal edema after cataract surgery: incidence and etiology. Semin. Ophthalmol. 17, 110–114. Yuan, S.Y., 2000. Signal transduction pathways in enhanced microvascular permeability. Microcirculation 7, 395–403. Zhang, Y., Xie, Q., Sun, X.C., Bonanno, J.A., 2002. Enhancement of HCO(3)(K) permeability across the apical membrane of bovine corneal endothelium by multiple signaling pathways. Invest. Ophthalmol. Vis. Sci. 43, 1146–1153.