Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The pigmented rabbit conjunctiva is a moderately tight epithelial tissue capable of active Cl secretion into the tear-side, with a potential difference (PD) of ~18 mV (tear-side negative), a short-circuit current (Isc) of ~15 µA/cm², and a transepithelial electrical resistance (TEER) of ~1.3 k · cm² (Kompella et al., 1993). About 60 to 80% of the Isc in the rabbit conjunctiva can be accounted for by net Cl secretion (Kompella et al., 1993; Shi and Candia, 1995), which is subject to modulation by cAMP, Ca²⁺, and protein kinase C (Shiue et al., 1998). Recent reports suggest that extracellular adenosine nucleotides and nucleosides may regulate a variety of biological processes, including cellular ion transport and secretory activity (Harden et al., 1995). These effects are thought to be mediated by P1- and P2-purinergic receptors. P1-purinergic receptors are activated primarily by adenosine, whereas P2-purinergic receptors are activated by ATP, ADP, and UTP with a general rank order potency of ATP > ADP > AMP > adenosine. P2-purinergic receptors are further classified into P2X and P2Y superfamilies (Fredholm et al., 1997). The P2X is a ligand-gated ion channel family with the P2Y being G protein-coupled, metabotropic receptors (Boyer et al., 1997; Ralevic and Burnstock, 1998). Seven P2X (P2X_1-7) receptors have been identified, together with five P2Y receptors (P2Y₁, P2Y₂, P2Y₄, P2Y₆, P2Y₁₁; Ralevic and Burnstock, 1998; King et al., 1998). Of the latter, P2Y₂, P2Y₄, and P2Y₆ are sensitive to uridine nucleotides, e.g., UTP (Boeynaems et al., 1996; Suh et al., 1997). Extracellular ATP appears to stimulate epithelial Cl secretion mainly via the P2Y₂ receptor (Harden et al., 1995; Hwang et al., 1996). Moreover, UTP, ATP, and ATPS at 1 to 100 µM were recently reported to stimulate mucin secretion in a concentration-dependent manner in the rabbit and human conjunctiva (Jumblatt and Jumblatt, 1998).

The purpose of the present study was to characterize the role of extracelluar nucleotides in active ion transport across the pigmented rabbit conjunctiva. The focus was on UTP, the most potent agonist for the P2Y₂ purinergic receptor (Harden et al., 1995). To that end, we measured the Isc and ³⁶Cl and ²²Na fluxes in response to extracellular UTP and other nucleotides, both in the presence and absence of ion transport inhibitors and suramin, a P2-purinergic receptor antagonist (Ralevic and Burnstock, 1998) as well as a UTP-sensitive receptor antagonist (Charlton et al., 1996).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals. Male Dutch-belted pigmented rabbits, weighing 2.5 to 3.0 kg, were purchased from Irish Farms (Los Angeles, CA). The investigations using rabbits described in this report conformed to the Guiding Principles in the Care and Use of Animals (Department of Health, Education and Welfare Publication, National Institutes of Health 80-23) and the Association for Research in Vision and Ophthalmology Statement on the Use of Animals in Ophthalmic and Vision Research.

Chemicals. UTP, UDP, UMP, ATP, ADP, AMP, ATPS, ,-methylene ATP, BzATP, bumetanide, and suramin were obtained from Sigma Chemical Co. (St. Louis, MO). UTP was also obtained from ICN Biochemicals (Costa Mesa, CA). 2-(Methylthio)-adenosine 5'-triphosphate (2-MeSATP), 2-(methylthio)-adenosine 5'-diphosphate (2-MeSADP), pyridoxal-phosphate-6-azophenyl-2',4'-disulfonic acid tetrasodium (PPADS), 8-cyclopentyl-1,3-dipropylxanthine (DPCPX), and 3,7-dimethyl-1-propargylxanthine (DMPX) were purchased from Research Biochemicals International (Natick, MA). N-phenylanthranilic acid (NPAA) was purchased from Fluka Chemicals (Ronkonkoma, NY). Na³⁶Cl (1.38 µCi/mg Cl) and ²²NaCl (666 mCi/mg Na) were obtained from Amersham Co. (Downers Grove, IL). D-[³H]Mannitol (19.7 Ci/mmol) was obtained from Dupont-NEN (Boston, MA).

Buffer Solutions. Unless otherwise indicated, all experiments were conducted in the bicarbonated Ringer's solution maintained at 37°C and pH 7.4 under 95% air/5% CO₂. The bicarbonated Ringer's solution contained 111.5 mM NaCl, 4.8 mM KCl, 29.2 mM NaHCO₃, 0.75 mM NaH₂PO₄, 1.04 mM CaCl₂, 0.74 mM MgCl₂, and 5 mM D-glucose. Na⁺-free Ringer's solution was prepared by equimolar replacement of NaCl, NaH₂PO₄, and NaHCO₃ with choline chloride, KH₂PO₄, and choline bicarbonate, respectively. Cl-free Ringer's solution was prepared by equimolar replacement of NaCl, KCl, CaCl₂, and MgCl₂ with sodium isethionate, potassium isethionate, calcium gluconate, and magnesium gluconate, respectively. The osmolality of these solutions was adjusted to 300 mOsm/kg H₂O with D-mannitol.

Tissue Preparation. We have previously reported the detailed procedure for excising the pigmented rabbit conjunctiva for Ussing chamber studies (Kompella et al., 1993). Briefly, rabbits were euthanized with an injection of 85 mg/kg sodium pentobarbital solution into a marginal ear vein. The entire eyeball was removed from the orbit, taking care not to damage the conjunctival epithelium. Within 15 min of surgery, the excised conjunctiva was mounted carefully onto a tissue adapter, which has a circular aperture of 1.0 cm². The adapter-tissue assembly was then placed in a modified Ussing chamber maintained at 37°C by circulating water bath. The bathing solution (6 ml on each side) of the tissue was bubbled with 95% air/5% CO₂ to maintain the pH 7.4 and to provide adequate agitation of the solution.

Bioelectric Parameter Measurements. All experiments were performed under short-circuit conditions with the use of an automatic voltage clamp device (558C-5; Bioengineering Department, University of Iowa, Iowa City, IA). PD was measured with two matched calomel electrodes. Two polyethylene (PE 90) bridges (containing 4% agar in 3 M KCl), which had tips located near the center of the tissue surfaces, were used to connect the bathing fluids electrically to the electrode wells. The electrical output of the calomel electrodes was amplified by the voltage-clamp unit. Direct current flowing across the tissue was sent with a pair of matched Ag/AgCl electrodes with conducting agar bridges, with its tips positioned away from the tissue surfaces at the far ends of the two reservoirs. The Isc flowing in the bath-tissue-bath circuit was monitored with a strip chart recorder (Kipp and Zonen, Delft, the Netherlands). At 60-s intervals, a 2-mV voltage pulse (V) was imposed for 3 s across the short-circuited tissue to estimate the TEER as a surface area normalized ratio of applied voltage pulse to the resultant current (I) response flowing on top of Isc [TEER = (V/I)A, where A is the nominal surface area of the Ussing chamber opening]. Before each experiment, the solution resistance (<100  · cm²) was compensated for by the automatic voltage clamp unit (Kompella et al., 1993). The baseline PD of 14.6 ± 0.5 mV (tear-side negative), Isc of 12.8 ± 0.2 µA/cm², and TEER of 1,151 ± 37  · cm², observed in 208 conjunctival tissues used in this study, were comparable with previously reported values (Kompella et al., 1993; Hosoya et al., 1996).

Cl and Na⁺ Flux Measurement. Unidirectional Cl or Na⁺ fluxes across the conjunctiva were determined separately using ³⁶Cl (0.5 µCi/ml) or ²²Na (1 µCi/ml), respectively. D-[³H]Mannitol at 10 µCi/ml was used concurrently for monitoring the integrity of the paracellular pathway. At predetermined times, 500-µl samples were collected from the receiver fluid, and the aliquot removed was immediately replenished with an equal volume of fresh buffer. Sample radioactivity was assayed in a liquid scintillation counter (LS1801; Beckman, Fullerton, CA). Unidirectional flux (J) for ³⁶Cl, ²²Na, or D-[³H]mannitol was estimated from the steady-state rate of the respective radioactivity appearing in the receiver fluid as a function of time.

Data Analysis. The concentration-response parameters for nucleotide effects in the conjunctiva were estimated by nonlinear least-squares regression analysis of the data for Isc (nucleotide-induced changes in Isc) versus nucleotide concentration using the NFIT software (Island Products, Galveston, TX):
where C is nucleotide concentration, Isc_min and Isc_max are minimal and maximal values of Isc, respectively, EC₅₀ is the effective half-maximal concentration, and n is a Hill coefficient (Segel, 1976).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of UTP on Short-Circuit Current in Conjunctiva. UTP (purchased from Sigma), when applied to the mucosal side of the conjunctiva at 10 µM, transiently increased the Isc by 7.3 ± 0.5 µA/cm² and decreased the TEER by 233 ± 72 ·cm² (Fig. 1). Peak Isc was achieved within 2 min. Both Isc and TEER gradually returned to baseline within 60 min. UTP (97-99% purity; obtained from ICN) at 10 µM afforded a Isc of 8.0 ± 1.7 µA/cm², which is not statistically different from that afforded by Sigma's UTP, presumably of lower purity. This indicates that the impurities contained in either commercial source did not participate significantly in the observed Isc. We used the Sigma product in all subsequent experiments.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Time courses of Isc and TEER in the pigmented rabbit conjunctiva following the mucosal addition of 10 µM UTP. Each point represents mean ± S.E.M. (n = 8).

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Concentration-response curves for the induced short-circuit current (Isc) in the pigmented rabbit conjunctiva as a function of nucleotide concentration both with and without 1 mM suramin Each point represents mean ± S.E. (n = 3-14). m, mucosal; s, serosal.

View this table: [in this window] [in a new window]   TABLE 1 Effect of purinergic receptor inhibitors on Isc induced by 10 µM UTP in the pigmented rabbit conjunctiva Data represent mean ± S.E. (n = 3-14).

Effect of Ion Transport Inhibitors on Short-Circuit Current Elicited by Mucosally Added 10 µM UTP. The serosal absence of Cl, serosal presence of 10 µM bumetanide, and mucosal presence of 0.3 mM NPAA all reduced the Isc elicited by UTP by 78, 77, and 42%, respectively (p < .05; Fig. 3). Whereas the mucosal absence of Na⁺ did not affect the Isc induced by UTP (p > .05), the serosal presence of 0.5 mM ouabain inhibited the UTP-stimulated Isc by 97%. The time course of changes in Isc elicited by 10 µM UTP under the serosal Cl- and mucosal Na⁺-free conditions is shown in Figs. 4, A and B, respectively. As can be seen, the ability of UTP to stimulate Isc was substantially reduced under the serosal Cl-free condition (Fig. 4A), but not affected under the mucosal Na⁺-free condition (Fig. 4B).

View larger version (34K): [in this window] [in a new window]   Fig. 3.   Effect of various ion transport inhibitors on the short-circuit current changes (Isc) in the pigmented rabbit conjunctiva elicited by 10 µM UTP. Each point represents mean ± S.E. (n = 3-14). *, significantly different from the control.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Time courses of Isc (% of initial Isc) in the pigmented rabbit conjunctiva under serosal Cl-free (A) and mucosal Na+-free (B) conditions. Each data point represents mean ± S.E. (n = 4). m, mucosal.

Effect of 10 µM UTP on Net Cl and Na⁺ Transport in Conjunctiva. Unidirectional Cl flux (J) in the serosal-to-mucosal (sm) direction (J^sm) was significantly increased by mucosal 10 µM UTP (p < .05), whereas that in the mucosal-to-serosal (ms) direction (J^ms) was unaffected (Table 2). The net result was stimulation of net Cl secretion (J^net) of 0.17 µEq/cm²/h, representing about 70% of the Isc elicited by 10 µM UTP (Table 2). In contrast, Na⁺ flux and D-mannitol permeability were unaffected (p > .05; Table 2).

View this table: [in this window] [in a new window]   TABLE 2 Cl and Na+ fluxes (J), and apparent permeability coefficient (Papp) of D-mannitol across the pigmented rabbit conjunctiva in the absence and presence of 10 µM UTP added mucosally Data represent mean ± S.E. Numbers in parentheses are number of tissues used. Transport of 36Cl (0.5 µCi/ml), 22Na (1 µCi/ml), and D-[3H]mannitol (10 µCi/ml) was studied in the mucosal absence and presence of 10 µM UTP under short-circuit condition. Jnet equals Jsm minus Jms. JIsc,eq is net flux based on Isc measurements.

Effect of Adenosine and Nucleotides on Conjunctival Short-Circuit Current. At 10 µM, the rank order of Isc induced by adenosine and nucleotides was UTP  ATP > ATPS = ADP = AMP = adenosine > 2-MeSADP = 2-MeSATP = UDP > BzATP > UMP > , -methylene ATP (Fig. 5). The Isc observed for UTP and ATP was 1.8 to 36 times higher than that for the others, including 2-MeSATP [the most potent P2Y₁ agonist (Ralevic and Burnstock, 1998), 1.7 ± 0.3 µA/cm²], 2-MeSADP [the most potent P2Y_ADP (ADP-sensitive P2Y receptor) agonist (Ralevic and Burnstock, 1998), 1.9 ± 0.5 µA/cm²], , -methylene ATP [a potent P2X agonist (North and Barnard, 1997), 0.2 ± 0.1 µA/cm²], and BzATP [the most potent P2X₇ agonist (North and Barnard, 1997), 1.0 ± 0.4 µA/cm²]. The above rank order of Isc for an abbreviated list of nucleotides over the 0.0001 to 1 mM concentration range, Isc_max, in this case, remained essentially unchanged (Table 3): UTP = ATP = ADP > AMP > 2-MeSADP = 2-MeSATP. In terms of EC₅₀, the rank order appeared to be UTP = ATP = 2-MeSATP = 2-MeSADP < ADP < AMP (Table 3).

View larger version (32K): [in this window] [in a new window]   Fig. 5.   Changes in the short-circuit current (Isc) of the pigmented rabbit conjunctiva elicited by 10 µM adenosine and various nucleotides. Each point represents mean ± S.E. (n = 4-8). *, significantly different from that observed with UTP.

View this table: [in this window] [in a new window]   TABLE 3 EC50 and lscmax elicited by selected nucleotides in the pigmented rabbit conjunctiva

View this table: [in this window] [in a new window]   TABLE 4 Effect of pretreatment with nucleotides on Isc (µA/cm2) elicited in the pigmented rabbit conjunctiva by UTP, ATP, 2-MeSATP, and 2-MeSADP Data represent mean ± S.E. (n = 4-14).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We have demonstrated that UTP increases net Cl secretion in the pigmented rabbit conjunctiva (Table 2) in a concentration-dependent manner with an EC₅₀ of 11.4 µM (Fig. 2). Because serosally added UTP (up to 1 mM) exerted very little effect in Isc (Fig. 2), the UTP binding sites probably are located mainly at the mucosal surface of the conjunctiva. Such a possibility was suggested by the observation of Jumblatt and Jumblatt (1998) that conjunctival mucin secretion was stimulated by UTP and ATP with an associated EC₅₀ of 10 µM for ATP in the rabbit and of 5 and 8 µM for UTP and ATP, respectively, in the human conjunctiva. The effect of UTP on conjunctival Isc was abated by 0.05 mM (Table 1) and 1 mM suramin (Fig. 2), which is known to antagonize P2 responses at concentrations varying from 0.01 to 1 mM (van Rhee et al., 1994; Piper and Hollingsworth, 1995; Henning et al., 1996). Charlton et al. (1996) showed that suramin at 30 to 300 µM causes a rightward shift of the concentration-response curve for UTP in human P2Y₂-transfected 1321N1 cells. Although the precise mechanism of suramin in antagonizing P2 receptors is not clear (Ralevic and Burnstock, 1998), the findings with suramin shown in Table 1 and Fig. 2 do suggest the involvement of mainly P2Y purinergic receptors in our conjunctival studies. The inability of 50 µM PPADS to inhibit Isc elicited by UTP (Table 1) suggests that P2X and P2Y₁, as well as UTP-insensitive P2Y receptor subtypes may not be involved.

The potency rank order of Isc increases caused by various nucleotides, as shown in Fig. 5 and Table 3, is consistent with an agonist for UTP-sensitive P2Y-type purinergic receptors including P2Y₂, P2Y₄, and P2Y₆, toward which ATP, ADP, and UTP are equipotent (Dubyak and El-Moatassim, 1993; Boyer et al., 1997; Jumblatt and Jumblatt, 1998; Ralevic and Burnstock, 1998). Such a possibility is indirectly supported by the lack of a marked effect on the UTP-induced Isc upon pretreatment of the conjunctiva with 2-MeSATP or 2-MeSADP (Table 4). P2Y₆ contribution was possibly negligible, because Isc was much lower for UDP (24%) and UMP (6%) than for UTP. Pending molecular identification, it follows that either P2Y₂ or P2Y₄ receptor is most likely involved in UTP-induced Cl secretion in the conjunctiva (Harden et al., 1995; Ralevic and Burnstock, 1998).

It is possible that adenosine nucleotide-specific A₁ subtype receptor may be involved in the ATP effect on the conjunctival Isc. This is because 50 µM DPCPX reduced the Isc elicited by 10 µM ATP by 52% (Table 1), and because adenosine at 10 µM also induced the conjunctival Isc by 4.3 ± 0.5 µA/cm² (Fig. 5). Indeed, A₁ receptor is known to be involved in the stimulation of Cl conductance in human airway epithelial cells in a cAMP-dependent manner (McCoy et al., 1995). This, however, is not the case in the UTP effect in conjunctival Isc, given that the Isc elicited by 10 µM UTP after 1 mM 8Br-cAMP pretreatment (6.4 ± 1.3 µA/cm²) was not significantly different from UTP treatment alone (7.3 ± 0.5 µA/cm²). Because DMPX failed to affect the ATP-induced Isc (Table 1), A₂ type purinergic receptor probably is not involved.

Extracellular membrane-bound ectonucleotidases are known to sequentially dephosphorylate nucleoside triphosphate to nucleoside diphosphate, nucleoside monophosphate, and nucleoside (Gleeson et al., 1989; Guibert et al., 1998), thereby terminating the action of nucleotides (Westfall et al., 1996). The half-life (T_1/2) of ATP is ~10 min in whole blood ex vivo (Trams et al., 1980) and ~40 min in folliculated Xenopus oocytes (Zigansin et al., 1996). Lazarowski et al. (1997) also reported that UTP at the mucosal surface of human nasal epithelium was hydrolyzed by ectonucleotidases with a T_1/2 of 14 min. It is tempting to attribute the transient nature of the UTP effect on Isc, as shown in Fig. 1, to ectonucleotidase action. However, such a possibility is not supported by the observation that when the conjunctiva was pretreated with 10 µM UTP for 30 min, the Isc induced by a second application of 10 µM UTP was much lower than the first treatment with UTP (Table 4). Signal transduction events, including activation of protein kinase C and Ca²⁺ mobilization after receptor stimulation, may account for the transient effect, as suggested by Ko et al. (1997). Although further study is necessary to measure UTP stability in tear fluid and nucleotidase activity in the conjunctiva, UTP binding to the receptor in the conjunctiva probably would take place within ~2 min, if it is instilled in the eye. The maximal increase in Isc caused by UTP was equally fast (Fig. 1).

Conjunctival Cl secretion is known to be regulated by cAMP (Kompella et al., 1996). Cl enters the conjunctival epithelial cells via serosal Na⁺-(K⁺)-Cl cotransport and exits into the tear fluid via Cl conductive pathway (Kompella et al., 1993). Consistent with this model of Cl transport, the serosal absence of Cl, serosal presence of 10 µM bumetanide, and mucosal presence of 0.3 mM NPAA all inhibited the Isc elicited by UTP (Fig. 3). The serosal presence of 0.5 mM ouabain abolished the Isc elicited by 10 µM mucosal UTP, suggesting that Na⁺/K⁺-ATPase provides the gradient for Cl entry via Na⁺-(K⁺)-Cl cotransport (Fig. 3). Net secretory Cl flux (0.17 µEq/cm²/h) stimulated by 10 µM UTP was about 70% of the corresponding Isc (Table 2). Although 20% of the Isc elicited by 10 µM UTP was independent of Cl secretion (Figs. 3 and 4), it cannot be caused by Na⁺ absorption. This is because UTP exerted no effect on Na⁺ transport (Table 2). As ATP treatment has been reported to increase Ca²⁺ influx into fibroblasts (Fine et al., 1989) and activate Ca²⁺-dependent K⁺ channels in Madin-Darby canine kidney (MDCK) cells (Friedrich et al., 1989), other ion transport processes in the conjunctival epithelial cells may similarly be affected by nucleotides.

Extracellular ATP or UTP is a major stimulus for cAMP-independent Cl secretion in human airway epithelial cells (Hwang et al., 1996), human intestinal goblet cell line HT29-Cl.16E (Merlin et al., 1996), and cultured pig aorta smooth muscle cells (Droogmans et al., 1991). Because transepithelial fluid movement in the airway epithelia is coupled to active ion transport (Knowles et al., 1995), stimulation of noncystic fibrosis transmembrane conductance regulator Cl channels in cystic fibrosis (CF) airways has been suggested as a way to restore fluid secretion that has been impaired from the lack of functional CF transmembrane conductance regulator type Cl channel activities (Knowles et al., 1995; Hwang et al., 1996). When placed in the context of the conjunctiva, this possible therapeutic measure, coupled with observation of ATP- and UTP-activated mucin secretion (Jumblatt and Jumblatt, 1998), is attractive for alleviating the symptoms in keratoconjunctivitis sicca and dry-eye patients (Morkeberg et al., 1995). Indeed, work in progress in our laboratory has revealed stimulation of fluid secretion in the conjunctiva by 10 µM UTP from 4.3 ± 0.2 (n = 27) to 9.8 ± 0.6 µl/h/cm² (n = 6) (M. H. I. Shiue, K.-J.K., and V.H.L., unpublished observations).

Presently, neither the source nor concentration of ATP, UTP, and other nucleotides in the tear fluid is known. In human lens, UTP and ATP levels are 106 and 1100 nmol/g wet tissue, respectively (Deussen and Pau, 1989). In ocular ciliary epithelial cells, ATP release via autocrine and/or paracrine mechanisms was reported to be stimulated upon lowering the tonicity of the bathing fluid (Mitchell et al., 1998). This was also the case in non-CF bronchial, submucosal gland, and airway epithelial cells (Taylor et al., 1998). It would be interesting to determine whether the symptomatic relief, offered by hypotonic solutions of 75-225 mOsm/l, in keratoconjunctivitis sicca patients (Gilbard and Kenyon, 1985) may be attributed to stimulated ATP release from conjunctival epithelial cells and the subsequent nucleotide-mediated stimulation of transconjunctival Cl and fluid flow.

In conclusion, the findings in the present study are consistent with the mucosal presence of purinergic receptors, the precise identity of which must await further molecular characterization. UTP, acting through these receptors (most likely of P2Y₂ and/or P2Y₄ subtype), stimulates net Cl secretion, but not Na⁺ transport. This raises the possibility that nucleotides may be used therapeutically to augment cAMP-independent Cl and fluid secretion in the conjunctiva. It remains to be seen whether exogenous nucleotides can be used to provide symptomatic relief in dry eye patients.