Exp. Eye Res. (1992)

Human

55, 413-417

Retinal

Pigment

Epithelial Cells Receptors VINH

Department of Ophthalmology,

Possess

P,-adrenergic

T.TRAN

University of Southern California School of Medicine, and Doheny Eye Institute, Los Angeles, CA 90033, U.S.A.

(Received Houston 76 September 1991 and accepted in revised form 27 November 7997) ,!6adrenergicreceptorson cultured human retinal pigmentepitheliumwere demonstratedby the binding of [‘“51]cyanopindolol.Its pharmacologicspecificity was also examined. Specific [1251]cyanopindolol bindingwassaturable,with a dissociationconstant of 130 PM and a receptor density of 12 fmol per onehalf million

cells, which is equivalent

to 14 000 receptor sites per retinal pigment epithelial

cell. Agonists

competedfor specific[1251]cyanopindolol binding, with the following rank order of potencies: ( - Iisoproterenol> ( - )-epinephrine> ( - )-norepinephrine.p,-selective antagonist ICI-118551 was approximately 3 log orders more potent than the p,-selective antagonist, betaxolol. These receptors were also coupledto an adenylatecyclase.Theseresultssuggeststrongly that cultured human retinal pigment epithelial cells possess ,8,-adrenergic receptors. The potential significance of these findings with regard to retinal pigment epithelial functions is discussed herein.

Key words: human; retinal pigment epithelium; P-adrenergic receptors: adenylate cyclase; ICI118551 : betaxolol. 1. Introduction The retinal pigment epithelium (RPE) is a specialized neuroepithelium that interposes between the neuroretina and the choriocapillaris. Neurotransmitters released by the retina and other hormones or neuromodulators that are carried by blood through the choriocapillaris potentially can affect the physiologic functions of the RPE by interacting with specific receptors. The P-adrenergic (Koh and Chader, 1984), HZ-histamine (Koh and Chader, 1983), and VIP receptors (Koh, 1989) have been demonstrated in cultured chick RPE, whereas the a,-adrenergic receptor has been demonstrated in rabbit RPE (Frambach, Valentine and Weiter, 1988). Several lines of evidence suggest that human RPE also possesses P-adrenergic receptors. The human RPE monolayer in culture respondsto isoproterenol stimulation with an increase in transepithelial potential and current (Frambach et al., 1990), and P-adrenergic agents stimulate adenylate cyclase activity in cultured human RPE (Friedman, Hackett and Campochiaro, 1987). However, the specific P-receptors on cultured human RPE have not been characterized biochemically. In the studies described herein, we examined the binding of [1251]cyanopindolol([lz51]CYP) to cultured human RPE and its competition by specific /J-adrenergic agonists and antagonists.

2. Materials and Methods Culture of Human RPE Cells The techniques for isolating and culturing human RPE were adapted from Kirchhof and Sorgente (1989). 00144835/92/090413+05

$08.00/O

RPE cells were harvested from human eyes obtained through the Lions Doheny Eye Bank (LosAngeles, CA) within 24 hr post-mortem. The eyecups were prepared by opening the globe circumferentially 2 mm posterior to the limbus. The anterior segmentswere discarded. Retina was teased from the underlying RPE layer, dissectedfree and discarded. The eyecups were washed with Dulbecco’s modified minimal essential medium (DMEM, Irvine Scientific, Santa Ana, CA) and incubated at 37°C with 0.05 % trypsin and 0.02 % ethylenediaminetetraacetic acid (0.05 y0 trypsin0.02 % EDTA, Irvine Scientific). After 1 hr, the trypsin solution was aspirated, discarded and replaced with DMEM supplemented with 10% fetal bovine serum (FBS, Hyclone, Logan, UT). The RPE cellswere released by gentle pipetting and transferred to 75 cm2 plastic flasks containing DMEM supplemented with 10% FBS and penicillin and streptomycin antibiotics. The cells were cultured at 3 7°C in a humidified 5 % CO,/9 5 % air atmosphere. The cells were fed twice per week. Upon reaching confluence at 4 weeks, cells were passedby trypsinization. RPE cellswere usedat second to fourth passages. Binding of [1251]cyanopindolol to Human RPE Cells Human RPE cells in their second or fourth passage were grown to confluence in 75 cm* flasks. The cells were scraped from the flasks, dissociatedby tituration, and collected by centrifugation. Binding of [‘251]CYPto RPE cells was carried out according to the methods of Kobilka et al. (1987). Aliquots of one-half million RPE cells were incubated for 2 hr at 24°C in 0.5 ml of Tris-HCl buffer (50 mM, pH 7.4) containing 0.01% ascorbic acid with or without 100 mM NaCl. Reaction 0 1992 Academic Press Limited

414

was terminated by filtration under vacuum over Whatman GF/C glass fiber filters. The filters were washed three times with 5 ml of ice-cold Tris buffer. Radioactivity that bound to RPE cells collected on the filters was quantified by a gamma counter. Nonspecific binding was determined in the presence of l.O,u~ propanolol. [lz51]CYP was added to a final concentration of 2 5-1200 pM for saturation experiments to determine the total number of binding sites. Competition for [1”51]CYP binding by different drugs was carried out with 50 PM concentrations of [‘“51]CYP and appropriate concentrations of inhibitors.

V. T. TRAN

a one-site model, with a dissociation constant K,, of 130 pM k 10 and total binding sites (H,,,,,) of 12 f 1 fmol per 5 x 10” cells (Fig. 1). The Hill coefficient was 0.94. Thus, each RPE cell exhibited 1.4 x 10” binding sites for [‘““IICYP. RPE cells in their fourth passage had only one-half to one-third the number of binding sites that were present in cells in their second passage (data not shown).

14

/3-adrenergic Stimulation of Adenylate Cyclase in Human RPE Adenylate cyclase activity of cultured human RPE was assayed according to a modified version of the technique developed by Hackett, Friedman and Campochiaro (1986). RPE cells in their fourth passage were plated into each 35 mm well of a Corning 6-well culture plate (Corning City, NY) at a density of 2 x lo5 cells and grown to confluence. On the day of assay, culture media were aspirated and the wells washed with DMEM without serum. Fresh DMEM (2 ml) containing 0.1 mM of isobutylmethylxanthine (IBMX) that had been prewarmed and equilibrated in the incubator was then added to each well. Cells were allowed to equilibrate for 15 min in the incubator at 37°C. after which test agents were added. Cyclase activities were measured for 20 min at 37°C. Basal levels were estimated in the absence of test substances other than IBMX. Cytoplasmic cyclic adenosine monophosphate (CAMP) was extracted with 1 ml of 1 mM HCl in ethanol after the aspiration of incubation media. CAMP was lyophilized and dissolved in 0.5 ml of Tris-HCl buffer (50 mM, pH 7.5) containing 4 mM EDTA. Fifty-microliter aliquots were tested for CAMP using a radioimmunoassay kit from Amersham (Arlington Heights, IL) according to the directions of the manufacturer. Protein concentrations were determined by the methods of Lowry et al. (19 5 1) after dissolving RPE monolayers with 1 ml of 0.2 N NaOH overnight at room temperature. Neurochemicals were purchased Gem RBI (Natick, MA). [‘251]CYP (2000 Ci mmol-I) was purchased from DuPont-NEN (Detroit, MI). Betaxolol was a generous gift from the Alcon Laboratory (Fort Worth, TX), and ICI-1 18 5 5 1 was provided by ICI Pharmaceuticals (Wilmington, DE). 3. Results Cultured human RPE cells in their second passage exhibited high levels of specific [lz51]CYP binding. Such binding was saturable with increasing concentrations of the radiolabeled ligand. Non-linear regression analysis (Munson and Rodbard, 1980) of the saturation isotherm indicated that the data were best fit to

0

200

400

600

800

Concentration

1000

1200

1400

(PM)

FIG. 1. Typical saturation binding of [‘251]CYP to dissociated cultured human RPE cells. Human RPE cells in their second passage were grown in DMEM supplemented with 10% fetal calf serum. Upon reaching confluence, cells were scraped and aliquoted. One-half million cells were incubated in 0.5 ml of Tris-HCI buffer (50 mM, pH 7.4) in the presence of 0.01% ascorbic acid. [1251]CYP was added to a final concentration of 25-1200 PM. Radioactivity bound to the cells collected onto GF/C filters by filtration was quantified by a gamma counter. Non-specific binding was determined in the presence of 1.0 ,u~ propanolol. Data were analysed by LIGAND, a reiterative non-linear least-square regression computer program (Munson and Rodbard, 1980). Experiments were carried out in triplicate and were repeated three times. K,= 130+10pM; B,,,= 12+1 fmol; Hill coefficient = 0.94. TABLE

I

Pharmacologicalspecificity of [‘251]cyanopindololbinding to human RPE cells I&I bf) With Na+ (100 rnM)

Drug ( - )-Isoproterenol ( - )-Epinephrine ( -)-Norepinephrine Betaxolol ICI-118551

Without Na+ (100

rnM)

6.2Ifr 1.6 x lo-’ 8.6k2.3 x lo-” 2~5~2.1~10~~ 2.5fl.OxlO-’ 4.5+2.0~10-~ 3.2fl.5~10-~ 3.6k1.2 x 10m63.7k1.5 x 1Om6 2.5fl.O~ 1O-9 2.2k1.2 x lOmy

Specific[“‘I]CYP bindingwasdetermined in the presence of variousconcentrations of inhibitors.The percentages of specific bindingthat remained wereplottedagainst the concentration of the displacers in a log-probitplot.TheI&s (50%inhibitionconstants) wereextrapolated fromeachgraph.Theexperiments were performed in triplicateand repeated three times. Sodium was added to a final concentration

of 100 mM in the form of NaCl.

&RECEPTORS

ON

HUMAN

415

RPE

BO-

D

-8

‘--IO

-4

-6 Log concentration

-2

(M)

FIG. 2. Displacement curves of [1251]CYP binding to cultured human RPE by agonists. Binding was carried out as described for Fig. 1, except that a 50 PM concentration of [1251]CYPwas added in the presence of 100 mM NaCl. Onehundred percent binding was defined as the radioactivity bound after subtracting non-specificbinding determinedin the presenceof 1.0 ,uM concentration of propanolol. The coordinate

represents the percentages of the specific binding

that remained,and the abscissa represents the logarithm of the concentration of the displacers.This plot allows us to determine

the relative

potencies

of different agonists. (0)

Isoproterenol; (0) epinephrine; ( n ) norepinephrine.

100

.\

-0

‘m \

ya 0

80- -\

$ : a;; ;0 0 “p m

60-

n

4. Discussion

0

\ -

Human

n

0 40-

\. \ 20 -

0 \

-10

-9

-8

-7 Log

FIG. 3. Displacement

concentration

-6

RPE in culture expresses high levels of

P-adrenergic receptors that can be labeled with

\

8

ment of specific [1251]CYP binding by these agonists followed a shallow curve (Fig. 2). Sodium ions at a concentration of 100 mM decreased the estimated IC,,s by approximately ten-fold (Table I). By comparison, the /3-specific antagonists betaxolol and ICI118551 [erythro-3-isopropylaminel-( 7-methyl-4indanyloxy)-2-butanol hydrochloride] competed with specific [‘251]CYP binding with steeper curves (Fig. 3). The Hill coefficient estimated from the log-probit plots was approximately 1 (data not shown). The I&s for ICI-l 1855 1 and betaxolol were 2.5 nM and 3.6 ,UM, respectively (Table I). The IC,,s of the antagonists did not appear to be affected by sodium ions. In order to ascertain whether ,&adrenergic receptors on human RPE cells are coupled to adenylate cyclase, we examined the accumulation of CAMP by cultured human RPE cells when challenged by different concentrations of isoproterenol, epinephrine and norepinephrine. The duration of incubation was 20 min. In preliminary experiments, we determined that the accumulation of CAMP was linear for up to 30 min under our experimental conditions. In the presence of IBMX, CAMP accumulation was 3.2 pmol mg-’ protein. ( - )-Isoproterenol caused a robust stimulation of CAMP production, with a maximal accumulation (B,,,) of 2320 pmol mg-’ protein. (- )-Epinephrine and ( - )-norepinephrine were less effective, inducing Bmaxsof 1450 and 1200 pmol mg-’ protein, respectively (Fig. 4). The rank order of potency for the stimulation of adenylate cyclase was isoproterenol > epinephrine > norepinephrine.

-5

-4

-3

(M )

of specific [‘TJCYP

binding

to

cultured human RPE by antagonists.The displacementof [‘zSI]CYPwas determinedas describedfor Fig. 2. Betaxolol (0) is a /3,-selectiveantagonist; ICI-118551 (0) is a p,selectiveantagonist. The pharmacologic specificity of [lz51]CYP binding was examined. Binding of [1251]CYP was stereoselective, with the (-)-isomer of isoproterenol being approximately lOO-fold more potent than the ( + )-

isomer (data not shown). The rank order of potencies for adrenergic agonists were ( - )-isoproterenol > ( - )epinephrine > ( - )-norepinephrine, with 50 % inhibition concentrations (I&,$) of 8.6 x lOmE, 2.5 x lo-‘, and 3.2 x lo+ M, respectively (Table I). The displace-

high affinity by [1251]CYP. The rank order of potency of agonists for the displacement of specific [1251]CYP binding was isoproterenol > epinephrine > norepinephrine. The &specific antagonist ICI118 5 5 1 (Kobilka et al., 198 7) was more potent than was the /&selective antagonist betaxolol in competing for the specific binding of [1251]CYP. These lines of evidence suggest strongly that cultured human RPE cells express P,-adrenergic receptors. Agonists competed for the binding of [1251]CYP to /3,receptors in human RPE with shallow displacement curves. This observation has been made with ,3-

receptors on other tissues, such as rat brain. The shallow slopes of the displacement curves have been attributed to different conformational states of the receptors rather than to molecular heterogeneity of the receptors. Indeed, when human P,-receptors were cloned and expressed in Xenopus oocytes, the slopes of the displacement curves for agonists remained shallow (Kobilka et al., 1987). The monovalent cation ions, lithium and sodium, have been shown to decrease the affinities of agonists in competing for the binding of antagonists to the P-receptors in tissue homogenates.

416

V. T. TRAN

p 2ooo/ y7‘7 L I 2 1” I 4cl -0 IO00

I

O-0-0 ./. /.’

I/ II0

/

7 WI -9

-8

-7

I -6

Log concentration

I

I -5

I

I -4

I -3

(M 1

FIG. 4. Stimulation of adenylate cyclase activity in cultured human RPE by agonists. Human RPE cells were cultured in 35mm wells of a Corning 6-well culture plate and exposed to various concentrations of different agonists for 20 min in DMEM in the presence of 0.1 mM isobutylmethyl-xanthine. CAMP was extracted with 1 mM HCl in ethanol, then lyophilized and reconstituted in 50 mM Tris-HCl containing 4 mM EDTA. CAMP was assayed by radioimmunoassay with a commercial kit from Amersham. Protein was solubilized from the same plates with 0.2 N NaOH and its concentration determined by the methods of Lowry et al. (1951). Results represent the means of three separate experiments. (0) Isoproterenol ; (0) epinephrine ; ( n ) norepinephrine. In our hands, sodium ions at 100 mM concentrations decreased the affinities or increased the IC,,s by approximately ten-fold. Conceivably, sodium increased the proportion of p,-receptors in the lower affinity states, as suggested by UPrichard, Bylund and Snyder (1978). As is the case in other cells and tissues, &receptors on human RPE are coupled to adenylate cyclase. Specifically, isoproterenol stimulated a rapid and robust accumulation of CAMP at concentrations of or below 1 ,UM. Frambach et al. (1990) demonstrated that isoproterenol at 0.1 ,UM elicited a rapid, transient rise in transepithelial voltage and short-circuit current across the cultured human RPE monolayer. The effects of isoproterenol could be mimicked partially by exogenous dibutyryl CAMP, indicating that the effects of isoproterenol were mediated by the activation of adenylate cyclase. Our results suggest that /&receptors mediate the effects of isoproterenol in altering the electrical properties of cultured human RPE. By inference, P,-receptors could affect transepithelial fluid movement across the RPE and across Bruch’s membrane. However, the direction of fluid movement induced by P-agonists or circulating catecholamines would ultimately depend on the nature of ionic conductance. The effects of catecholamines and P-agonists upon the RPE under the normal physiologic conditions of the human eye are still poorly understood. Some insight, however, can be derived from a known

clinical entity that is thought to affect the transport and barrier functions of the RPE. Central serous chorioretinopathy (CSR) is a disease that affects retinal function as a consequence of the accumulation of fluid in the subretinal space, which can be shown by fluorescein angiography to result from leakage from the choriocapillaris through the RPE layer. leading to pooling in the subretinal space (Gass, 1967). Clinical observations and case-control studies have indicated a correlation between CSR, stress and the type-A personality (Lipowski and Kiriakos, 19 71; Yannuzzi, 1987). By inference, therefore, these patients have higher circulating levels of catecholamines. Other investigators have successfully created an experimental model of CSR in monkeys by the administration of exogenous epinephrine (Yoshioka, Katsume and Akune, 1982). The observations made in this experimental model lend credence to the theory that catecholamines play an important role in the genesis of CSR. We would like to suggest that the effect of epinephrine was mediated by its interaction with ,/I,receptors on the monkey RPE. Although it is possible that the effect of epinephrine was mediated by an SIreceptor, we believe this is unlikely since we have been unable to detect either the zl- or a,-receptor by the binding of [3H]prazocin or [3H]rauwolscine (unpubl. res.). Furthermore, the a-agonist phenylephrine was shown to be loo-fold less potent than isoproterenol in altering transepithelial properties of the RPE (Frambath et al., 1990). Thus, existing evidence supports a ,4’,-adrenergic mechanism in CSR. P,-antagonists may, therefore, be the therapeutic agent of choice for the treatment of CSR. CAMP has been shown also to inhibit rod outer segment phagocytosis (Edwards and Bakshian, 1980: Ogino et al., 19 8 3) by the RPE as well as the migration (Hackett et al., 1986) of human RPE in culture. Many therapeutic agents, such as antihypertensive and antiasthmatic agents, are known to interact with /I,receptors, and potentially can affect the ability of the RPE to metabolize and phagocytose the shed rod disc membranes. A defect in the RPE’s ability to phagocytose shed rod outer segments has led to the degeneration of photoreceptors in RCS rats (Bok and Hall, 19 71). It is conceivable, therefore, if only at a theoretical level, that many pharmacotherapeutic agents may secondarily affect the photoreceptor outer segment renewal processes in humans as a result of their effects on the RPE.

Acknowledgments The author thanks Siobhan Duffy for excellent technical assistance, Diane Gegala for invaluable assistance in manuscript preparation, and Ann Dawson for expert editorial assistance. This study was supported, in part, by grants from the Margaret and Herbert Hoover Foundation, the Visionaries/Weikersheimer Estate, and a Core Grant for Vision Research (EY03040) from the National Institutes of Health. A portion of this study was presented at the annual

&RECEPTORS

ON

HUMAN

RPE

meeting of the Association for Research in Vision and Ophthalmology, held in Sarasota, Florida, April 29-May 4, 1990.

References Bok, D. and Hall, M. 0. (19 71). The role of the pigment epithelium in the etiology of inherited retinal dystrophy in the rat. 1. Cell Biol. 49, 664-82. Edwards, R. B. and Bakshian, S. (1980). Phagocytosis of outer segments by cultured rat pigment epithelium: reduction by cyclic AMP and phosphodiesterase inhibitors. Invest. Ophthalmol. KS. Sci. 19, 1184-8. Frambach. D. A., Fain, G. L.. Farber, D. B. and Bok, D. (1990). Beta adrenergic receptors on cultured human retinal pigment epithelium. Invest. Ophthalmol. Vis. Sci.

31, 1767-72. Frambach, D. A., Valentine, J. L. and Weiter, J. J. (1988). Alpha- 1 adrenergic receptors on rabbit retinal pigment epithelium. Invest. Ophthalmol. Iris. Sci. 29, 737-41. Friedman, Z., Hackett, S. F. and Campochiaro. P. A. (1987). Characterization of adenylate cyclase in human retinal pigment epithelial cells in vitro. Exp. Eye Res. 44, 471-9.

Gass, J. D. M. (1967). Pathogenesis of disciform detachment of the neuroepithelium. II. Idiopathic central serous choroidopathy. Am. 1. Ophthalmol. 63, 587-615. Hackett, S., Friedman, Z. and Campochiaro, P. A. (1986). Cyclic 3’,5’-adenosine monophosphate modulates retinal pigment epithelial cell migration in vitro. Arch. Ophthalmol. 104, 1688-92. Kirchhof, B. and Sorgente, N. (1989). Pathogenesis of proliferative vitreoretinopathy : modulation of retinal pigment epithelial cell functions by vitreous and macrophages. Dev. Ophthalmol. 16, l-53. Kobilka, B. K., MacGregor, C.. Daniel, K., Kobilka, T. S..

417

Caron, M. G. and Lefkowitz, R. J. (1987). Functional activity and regulation of human /f,-adrenergic receptors expressed in Xenopus oocytes. J. Biol. Chem. 262, 15796-802. Koh, S.-W. M. (1989). Modulation of a 190-kD microtubuleassociated protein in pigment epithelium by VIP. Peptides 10, 1089-99. Koh, S.-W. M. and Chader, G. J. (1983). Chick retinal pigment epithelium : a new culture system for studying Hz-histamine receptors. FEBS Lett. 164, 277-80. Koh. S.-W. M. and Chader, G. J. (1984). Retinal pigment epithelium in culture demonstrates a distinct fiadrenergic receptor. Exp. Eye Res. 38, 7-l 3. Lipowski, Z. J. and Kiriakos, R. Z. (1971). Psychosomatic aspects of central serous retinopathy : a review and case report. Psychosomatics 12, 398401. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (19 5 1). Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193. 265-7.5. Munson, P. J. and Rodbard, D. (1980). LIGAND: a versatile computerized approach for characterization of ligandbinding systems. Anal. Biochem. 107. 220-39. Ogino. N.. Matsumura, M., Shirakawa. H. and Tsukahara, I. ( 198 3). Phagocytic activity of cultured retinal pigment epithelial cells from chick embryo: inhibition by melatonin and cyclic AMP, and its reversal by taurine and cyclic GMP. Ophthalmic Res. 15, 72-89. LJ’Prichard. D. C., Bylund. D. B. and Snyder, S. H. (1978). ( + )-[“Hlepinephrine and ( - )-[3H]dihydroalprenolol binding to p,- and ,!3,-noradrenergic receptors in brain, heart. and lung membranes. 1. Biol. Chem. 253, 5090-l 02. Yannuzzi, L. A. (1987). Type-A behavior and central serous chorioretinopathy. Retina 7, 11 l-30. Yoshioka, H.. Katsume. Y. and Akune. H. (1982). Experimental central serous chorioretinopathy in monkey eyes : fluorescein angiographic findings. Ophthalmologica 185,

168-78.

EER

55