JOURNAL OF OCULAR PHARMACOLOGY Volume 1, Number 4, 1985 Mary Ann Liebert, Inc., Publishers Pp. 343-352 Pharmacological Modification of Corneal Endothelial Intracellular pH, Intracellular Electrical Potential Difference, and Corneal Swelling and Deswelling Rates R. DAVID ELIJAH,1 KENNETH E. CHEEKS,1 GERALD J. ROBERT W. ROBERSON,1 and KEITH GREEN12 BOHN,1 1Department of Ophthalmology and ^Department of Physiology, Medical College of Georgia, Augusta, Georgia ABSTRACT Rabbit corneal endothelial intracellular pH and electrical potential difference were measured using radioactive tracer techniques. A variety of drugs and chemicals were used to modify specific aspects of cell membranes in order to determine, in this steady state system, whether a close relationship existed Corneal swelling and deswelling rates between intracellular pH and potential. were also determined in the presence and absence of the drugs. Longer times of immersion of tissue with drugs (3 hours) tended to provide more support for a link between intracellular pH and the inability of the cornea to maintain a constant thickness. Drugs that altered corneal thickness without altering either intracellular pH or potential difference presumably acted on extracellular No conclusive evidence was obtained for a direct link between pathways. intracellular pH and the intracellular electrical potential difference of the endothelial cell. INTRODUCTION ambient bathing solution at pH 7.5 and a bicarbonate concentration of rabbit corneal endothelial cells have an average resting membrane potential of about -35 mV and an intracellular pH of 7.1, determined using the bromide tracers radioactive (TPPBr) and tetraphenylphosphonium dimethyloxazolidine-2,4-dione (DMO), respectively (1). Of the many perturbations or imposed on the endothelLum, only external pH variations, bathing in Na M ouabain for 1 M ouabain for 3 hours or 10 or 10 HCO, -free solutions, 10 hour, or 60 mM bicarbonate, altered the intracellular pH, while increasing extracellular K , 10 M ouabain for 3 hours and 1 mM acetazolamide altered the intracellular electrical potential difference (1). These were initial attempts to obtain information regarding the interaction between pH and cellular electrical potential difference (PD), in order to determine any relationship between these parameters. A net bicarbonate transport has been demonstrated across the corneal endothelium (2,3); but, of course, this In 25 an mM, - 343 ELIJAH ET AL. 344 could either be bicarbonate moving from the stromal to aqueous-facing surface or a proton moving in the opposite direction. In order to accentuate differences and examine relationships between intracellular pH and cellular PD, various pharmacological agents were used which possess specific properties related to cellular membranes. These agents were used to modify cell membrane properties in order to further study the possible relationship between intracellular pH and PD. That ambient pH influences various transendothelial ion fluxes is known from data showing an effect on sodium fluxes across the endothelium (4). Thus, ambient pH The present studies were can influence sodium fluxes and perhaps cellular PD. made to examine the cellular pH/PD relationship. MATERIALS AND METHODS Intracellular pH and PD The techniques for isolation of the corneas from 2 to 3 kg albino rabbits, incubation in media containing resolubilized C-DMO (dimethyloxazolC-TPPBr (tetraphenylphosphonium bromide) for the measureidine-2,4-dione) and ment of intracellular pH and PD, respectively, and the composition of Krebs bicarbonate Ringer have all been given previously (1) The calculations for endothelial intracellular pH and PD have been discussed in detail elsewhere (1,5), together with detailed accounts of the washout procedure and technical limitations (1). Because some agents caused changes in either swelling or deswelling rates (see Results) only after 1-1/2 hours of incubation, experiments using these particular agents were performed with 3 hour incubations as well as the usual 1 hour. their . Specular Microscopy removed from eyes of 2 to 3 kg New Zealand albino rabbits, overdose of sodium pentobarbital, and mounted in specular microscopes as described previously (6-8). Normal, paired corneas were allowed to equilibrate with bicarbonate Ringer solution for 1 hour before one of the pair was perfused with Ringer plus drug or Ringer plus vehicle. Preswollen corneas were de-epithelialized prior to mounting in the specular microscope with bicarbonate Ringer on the denuded epithelial-facing stromal surface for 15 to 20 minutes. At this time Ringer was replaced with oil and a further hour allowed for deswelling to begin and achieve a steady state. One of the pair of corneas then perfused with Ringer plus drug or Ringer plus vehicle. Corneal was thickness for both normal and preswollen corneas was measured every 30 minutes for the 3 hours following the initial hour equilibration period of deswelling. Swelling and deswelling rates were calculated by linear least squares regression analysis (8) since the rates of swelling or deswelling were approximately linear (by eye). A comparison of experimental and control regression lines was made by an analysis of covariance. Corneas killed with were an Drugs their Drugs employed, and their dose levels, are given in Table 1, together with effects. carbonylcyanide Nigericin, anticipated physiological m-chlorophenylhydrazone (CCCP), valinomycin, verapamil, and gramicidin were first dissolved in methanol prior to addition to the Ringer solution; the final methanol concentration was 2 pl/ml Ringer. Appropriate controls with Ringer containing methanol alone were also used. DRUG MODIFICATION OF CORNEAL FUNCTION 345 TABLE 1. List of Drugs Employed in Study, Drug Dose Dose Valinomycin 10 2,4, dinitrophenol 10 Nigericin 20 Gramicidin 1 4,4'-diisothiocyano-2,2' 10 5 Level, and Mode of Action Action Reference M K ionophore 9-11 M Protonophore 12, 13 ug/ml Na+ 14 Ug/ml Equalizes cellular and ambient pH 15 M Anion transport blocker 16, and Na 3 -4 and K ionophore disulfonic acid stilbene 17 (DIDS) 10 Harmaline 5 Carbonylcyanide m-chlorophenylhydrazone (CCCP) 10 Verapamil 10 3 x -4 10~3 transport inhibitor 18, 19 M M Protonophore M Ca channel blocker 13, 20, 21 22 RESULTS Intracellular pH The data are shown in Table 2; the control series compares favorably with previous controls incubated in 25 mM bicarbonate Ringer for 1 hour (1). In addition, the methanol controls also compare well with both series of normal controls. Dinitrophenol, 10 _,M, had no effect on intracellular pH, whereas harmaline (5 x 10 and 10 M), DIDS (10 M), valinomycin, nigericin, and verapamil caused intracellular alkalinization to a value closer to that of the bathing solution. Gramicidin and CCCP had no effect on intracellular pH. The 3 hour incubation (Table 2) control values tended to be slightly higher than the 1 hour incubations and DIDS, nigericin and gramicidin all caused significantly increased alkalinization of the cells. 346 ELIJAH ET AL. TABLE 2. Effect of Various Inhibitors Bathing on Corneal Endothelial Intracellular Solution Intracellular pH Controls from previous studies Control, 25 mM 7.10 bicarbonate, Dinitrophenol, 10 DIDS, 10" M Harmaline, 10~3 Harmaline, 5 Control, 25 2 x pH -3 1 hour M M 10~3 M (14) 7.19 ± 0.01 (13) ( 7) 7.23 ± 0.03 7.30 ± 0.01* 7.38 ± 0.02* 7.33 ± 0.01*( 7) 7.22 ± 0.03 methanol 10~5 0.02 hr, mM bicarbonate 1 ul/ml ± (12) 7.34 ± 0.02*(12) Nigericin, 20 ul/ml 7.41 ± 0.05* Gramicidin, 1 ug/ml 7.24 + 0.04 Valinomycin, CCCP, 10~7 7.29+0.04 M -4 Verapamil, 10 Control, 25 Control, 25 2 M 7.34 M mM bicarbonate, 3 mM bicarbonate, 3 hr. hr. ± 0.03* 7.28+0.06 7.20 ± 0.02 ul/ml methanol 7.36 ± 0.04 ( 6) 7.46 ± 0.03* Nigericin, 20 ug/ml 7.57 + 0.02*( 6) Gramicidin, 1 ug/ml 7.62 ± 0.02*( 6) Dinitrophenol DIDS, 10~4 M are the means + SEM of 8 determinations (each consisting of 4 pooled endothelia), except where shown in parenthesis. All incubations were made at an ambient pH of 7.5. First control values compared to an n of 14 from earlier studies (1). *, indicates significantly different from its control (i.e., Values methanol or methanol-free); P < 0.05. Intracellular PD The data are shown in Table 3. Both the normal and methanol control compare with previous control values (1). A low concentration of harmaline (10 DRUG MODIFICATION OF CORNEAL FUNCTION 347 had no .effect on PD but 5 x 10 M caused a marked depolarization to -13.6 mV. DIDS (10 M) caused a significant hyperpolarization by about 9 mV. No other agent employed had an effect on PD when compared to its respective control (i.e., M) methanol-containing or not). TABLE 3. Effect of Various Inhibitors Bathing on Corneal Endothelial Cellular Potential Intracellular Potential (mV) Solution Controls from previous studies 25 mM bicarbonate, 1 hour 10 Dinitrophenol, DIDS, 10" -3 -33.5 ± 4.0 -33.3 ± 1.8 -33.0 ± 5.0 M -42.0 ± 2.5* M Harmaline, 10 Harmaline, 5 -3 M 10 x -3 M 25 mM bicarbonate 1 hr, 2 yl/ml methanol 10 Valinomycin, -33.6 ± 3.7 -13.6 ± 2.4* -37.5 ± 3.4 -45.6 ± 3.3 M yg/ml -35.1 ± 1.4 Gramicidin, 1 yg/ml -41.6 ± 1.8 Nlgericin, CCCP, 10~7 Verapamil, 25 mM 20 -41.7 ± 1.9 M 10 -4 -43.9 M bicarbonate, 3 hours ± 1.8 -46.1 ± 2.0 (12) 2.4 ( 5) 25 mM bicarbonate, 3 hours, 2 yl/ml methanol -44.4 Dinitrophenol -42.1 ± 0.9 DIDS, 10" -43.9 M Nigericin, 20 yg/ml Gramicidin, ( 8) 1 yg/ml ± ± 2.4 -27.7 t 2.4*( 5) -58.8 1.7*(12) ± the means ± SEM of 6 endothelia unless otherwise indicated in All incubations were First control value obtained from Ref. 1. made at an ambient pH of 7.5. *, indicates significantly different from its own control (i.e., methanol or methanol-free); P < 0.05. Values are parenthesis. ELIJAH ET AL. 348 The 3 hour incubation (Table 3) control values (both normal and methanolcontaining Ringer) were higher than previous controls by about 10 mV. Nigericin caused a depolarization of the PD, while gramicidin hyperpolarized the potential. Swelling and Deswelling Rates The average swelling rate of control corneas was 9.5 ± 1.3 um/hr. The average deswelling rate of control preswollen corneas was 17.5 + 1.7 um/hr (Table 4). The lower deswelling rate, compared to values in the literature (2,6), is related both to the initial 1 hour delay in recording after oil replacement of Rir.ger and to the lower starting thickness. Gramicidin, 2,4,-dinitrophenol, and nigericin caused an increased swelling rate compared to paired controls (Table 4). Gramicidin, 2,4,-dinitrophenol, nigericin, harmaline, valinomycin, and DIDS all inhibited the deswelling rate (Table 4). The effects of 2,4,-dinitrophenol and harmaline were evident within 30 minutes of drug perfusion, but the effects of gramicidin, nigericin, valinomycin, and DIDS were not significantly different from the paired eye until 1-1/2 hours after initiation of drug perfusion. The separate values for each 1-1/2 hour period are given in Table 4 where marked differences can be seen for these drugs in the initial and final perfusion periods. DISCUSSION The basis for the use of the tracer techniques employed herein was established previously (1). The present data confirm our earlier estimates of normal intracellular pH and PD (1) for 1 hour incubations. As discussed previously (1), the values found here reflect those of entire endothelia in which The 3 hour a range of values are found by direct impalement (24-26). incubations, made at a different time of the year, show a greater PD more closely allied to the average PD's found recently in endothelia (2,24-26). These tracer techniques, however, can only be used to examine the status of the cell at specific times, and transient effects that could provide valuable information regarding ion movements cannot be observed. Intracellular pH became more alkaline in the presence of DIDS, harmaline, valinomycin, nigericin, and verapamil. Intracellular PD, however, was affected only by DIDS and 5 x 10 M harmaline. Each of these agents acts by a different mechanism on cell membranes (Table 1) It would be anticipated that alkalinization of the cell would be accompanied by a depolarization of the cell membrane (24). DIDS acts by inhibiting anionic transport systems (16,17), and the present results imply that inhibition of such systems in the endothelial cell causes either [H ] to decrease or [HCO. ] to increase under cellular alkalinization, together with the enhancement of the intracellular PD. DIDS probably inhibits HCO. exit from the cells (24) Harmaline, 5 x 10 M, caused a cellular depolarization and a cellular alkalinization, which would be anticipated given that harmaline inhibits active ion pump activity (18,19) and would not alter passive membrane conductance. Some of the present data support the notion that cellular pH and PD are related and indicate the possibility of H -cation exchanges across the endothelial cell membrane. The protonophores, dinitrophenol (12,13) and CCCP (27), however, did not alter either cellular pH or PD, but DIDS altered both pH and PD. That changes in intracellular pH are not necessarily accompanied by changes in PD at steady state is illustrated by the effects of 10 M harmaline, valinomycin, nigericin, and verapamil. These compounds caused alkalinization of the intracellular pH without altering PD. This data, contrary to that of DIDS, indicates that cellular pH can apparently change independently of the PD, at least at steady state. These observations stand in contradistinction to those of Jentsch et al. (24) who, by examination of cultured bovine endothelial cells using microelectrodes, concluded that a fall in cellular pH consistently hyperpolarized the cells whereas movement of intracellular pH in the opposite direction caused a depolarization. Our studies were all conducted at steady . . DRUG MODIFICATION OF CORNEAL FUNCTION 349 co co o o V V o ^ co co 55 =3 m o -j- o \D o -* o m CN CO 0\ en m o o t-H + I .-t -it m o o o m o H inn r- o CN o r-^ co >X3 >JD I o m en co O xJ-] cn m O o-> •<r o r-t r-~ >í <t O st •~\ i-H I + O o O r~-* <t en vO I O ^O r- co ct\ en en m co \o O CO O ^D i-H o S en H 0° ELIJAH ET AL. 350 Thus, state whereas the microelectrode studies were made with transients (24). different conclusions could be reached due to long term equilibration in our studies. Nevertheless, a link between pH and PD is not directly apparent. That verapamil influenced pH without affecting PD is interesting because of the known effects of calcium deprivation on the corneal endothelium (28,29). It would have been anticipated that calcium effects would be more directed towards the ionic permeability of the cell membrane, but no PD change occurred. Three hour incubations with the agents indicate a closer correlation between the intracellular parameters and corneal swelling or deswelling since disturbance of intracellular characteristics occurred with 3 hour but not 1 hour incubation with nigericin, gramicidin and DIDS (the DIDS effects on pH were greater at 3 than at 1 hour). Nigericin, gramicidin, and DIDS all caused cellular alkallnization, to the level of, if not greater than, the ambient solution. Nigericin depolarized and gramicidin hyperpolarized the PD, indicating that a different final status was achieved compared to the 1 hour incubation data. These changes occurred despite alkalinization of the cell, again indicating that pH differences across the cell membrane are not the primary determinants of PD. Comparisons of drug effects of cellular pH or PD and swelling or deswelling indicated temporal dissociation of these effects. Gramicidin and nigericin had no effect on corneal swelling rate until 1-1/2 hours after perfusion but intraThis data cellular characteristics were altered at 1 hour of incubation. indicates that changes in cellular composition can precede alterations in endothelial function. Correlations between drug effects on swelling/deswelling and intracellular pH or PD are not evident (Tables 2, 3, and 4). Some drugs (DIDS and harmaline) altered both pH and PD as well as swelling or deswelling rate, while dinitrophenol altered corneal hydration without inducing any change in pH or PD. In summary, the present data indicate that compounds that caused a change in intracellular pH also caused an alteration in the ability of the cornea to either maintain a normal thickness or deswell from a swollen state. Agents that affected corneal thickness yet had no effects on intracellular pH or PD must obviously alter some aspect of endothelial permeability that is unrelated to these characteristics. ACKNOWLEDGEMENTS This work was supported, in part, by Public Health Service Research Grant EY04558 from the National Eye Institute (to KG); in part, by a Core Grant for Vision Research, EY04636, from the National Eye Institute; in part, by a departmental award from Research to Prevent Blindness, Inc.; and, in part, by the We thank American Medical Association, Education and Research Foundation (GJB). Mrs. Sylvia Catravas for her excellent secretarial assistance. REFERENCES 1. K., Bowman, Elijah, and pH of 3:991-1000, 1984. potential 2. corneal Intracellular K.E. and Green, K. Curr. Eye Res. , endothelial cells. The bicarbonate ion pump in the endothelium which of rabbit cornea. J. Physiol., 263:561-577, 1976. Hodson, S. and Miller, F. regulates 3. Cheeks, R.D., rabbit Huff, the hydration Characterization of bicarbonate, J.W. and Green, K. fluxes in the rabbit corneal endothelium. Exp. chloride sodium and Eye Res. , 36:607-615, 1983. 4. Green, K. and Thomas, L. ion fluxes. Invest. Effect of pH and [HCO. ] on corneal endothelial Vis. Sei., 25 (suppl.):103, 1984. Ophthalmol. 351 DRUG MODIFICATION OF CORNEAL FUNCTION 5. Green, K., Simon, S., Kelly, G.M. and Bowman, K.A. Effect of [Na ], [Cl~], carbonic anhydrase, and intracellular pH on corneal endothelial bicarbonate transport. Invest. Ophthalmol. Vis. Sei., 21: 586-594, 1981. 6. The metabolic basis to the fluid pump in Dikstein, S. and Maurice, D.M. the J. cornea. Physiol., 221:29-41, 1972. Functional and B.E., Edelhauser, H.F. and Van Horn, D.L. structural changes in the corneal endothelium during ±n vitro perfusion. Invest. Ophthalmol., 121:410-417, 1973. 7. McCarey, 8. Hull, E.C. and Green, K. endothelial cell function. 1979. cornea Sei., 18:1226-1231, 9. Henn, F.A. and Thompson, T.E. Rev. Biochem., 38:241-262, 1969. 10. Bangham, A.D., Standish, Synthetic lipid bilayer M.M. and Watkins, J.C. the lamellae of swollen phospholipids. across Photodynamically induced Invest. Ophthalmol. Vis. Strickland, D.S., alteration of membranes. Ann. Diffusion of univalent ions J. Mol. Biol., 13:238-252, 1965. 11. cells. 12. The effect of HK and LK sheep red Tosteson, D.C., Cook, P., Andreoli, T. and Tieffenberg, M. valinomycin on J. Gen. potassium and sodium permeability of Physiol., 50:2513-2525, 1967. Effect of electrochemical gradients on and Dixon, T.E. Current Topics in Membranes and transport in an epithelium. Al-Awqati, Q., active H Transport, 16:163-174, 1982. 13. 14. McLaughlin, S.G.A. and Dilger, J.P. Transport of protons by weak acids. Physiol. Revs., 60:825-863, 1980. Graves, C. and Sachs, G.E. using 15. a carbocyanine dye. Sachs, G., Faller, L.D. and intestinal and epithelia. Rothstein, A., Cabantchik, 17. Cabantchik, membranes Quantitation of corneal endothelial potentials Biophys. Acta, 685:27-31, 1982. Biochim. Raison, E. J. Membrane Proton/hydroxyl transport in gastric Biol., 64:123-135, 1982. Z.I. and Knauf, P. in red blood cells: role of membrane proteins. 16. across Mechanism of anion transport Fed. Proc., 35:3-10, 1976. Membrane proteins related to anion and Rothstein, A. of human red blood cells as determined by studies with disulfonic stilbene derivatives. J. Membrane Biol., 15:207-226, 1974. Z.I. permeability 18. De Sousa, R.C. and Grosso, A. harmaline drug 40:77-94, 1978. - on sodium Vasopressin-like and water effect of transport. J. an hallucinogenic Biol., Membrane - Na , Li , and Cl J. Membrane Biol., transport by brush border 19. Günther, R.D. and Wright, E.M. membranes from rabbit jejunum. 20. effect of uncouplers of oxidative phosphorylation on J. membranes: carbonylcyanide-m-chlorophenylhydrazone. Membrane Biol., 4:227-251, 1971. LeBlanc, O.H. lipid 21. The bilayer Liberman, Y.A. and Topaly, V.P. and classification of Biofizika, 13:1025-1035, 1968. branes 74:85-94, 1983. Transfer of ions across bimolecular memof oxidative phosphorylation. uncouplers 352 22. ELIJAH ET AL. Bakker, E.P., Van den Heuvel, E.J., Wiechman, A.H.C.A. and the of effectiveness comparison between uncouplers phosphorylation systems. 23. in Biochim. mitochondria and in different Dam, K. A of oxidative artificial membrane Van Biophys. Acta, 292:78-87, 1973. P.G. and Krishtal, O.A. Ionic and gating currents in the membrane cells. In Membrane Transport Processes, Vol. 2, Tosteson, D.C., Ovchinnikov, Y.A., and Latorre, R. , eds. Raven Press, New York, 1978, pp. 101-123. Kostyuk, of nerve 24. Effect of Jentsch, T.J., Koch, M. , Bleckmann, H. and Weiderholt, M. bicarbonate, pH, methazolamide and stilbenes on the intracellular potentials of corneal endothelial cells. Membrane Biol., J. cultured bovine 78:103-117, 1984. 25. Lim, 26. Weiderholt, 27. Felle, H. and Bentrup, F.W. Intracellular potential J.J. and Fischbarg, J. endothelial cells. Exp. Eye Res., 28:619-626, 1979. M. and Koch, M. and human corneal endothelium. carbonyl of rabbit corneal Intracellular potentials of isolated rabbit Exp. Eye Res., 27:511-518, 1978. A study of the primary effect of the uncoupler m-chlorophenylhydrazone on membrane potential and Riccia fluitans. Biochim. Biophys. Acta, 464:179-187, 1977. cyanide conductance in 28. Kaye, G.I., Mishima, S., Cole, J.D. and Kaye, N.W. Studies on the cornea. VII. free medium on the corneal Effects of perfusion with a Ca endothelium. Invest. Ophthalmol., 7:53-56, 1968. 29. Edelhauser, H.F., Van Horn, D.L., Schultz, R.O. and Comparative toxicity of intraocular irrigating solutions endothelium. Am. J. Ophthalmol., 81:473-481, 1976. Received: Accepted October 30, 1985 for publication: November Reprint requests: 21, 1985 Keith Green, Ph.D., D.Sc. Department of Ophthalmology Medical College of Georgia MCG Box 3059 Augusta, Georgia 30912-0805 USA R.A. corneal Hyndiuk, on the