Exp. Eye Res.(1992) 55. 443-450

Alpha,-adrenoceptors RONALD

in the Cornea1 Endothelium

GUO-SUI YE”, PETER S. REINACHC J. WALKENBACHa,b,“, AND FRANCES BONEY”

aThe Missouri Lions Eye Research Foundation, 404 Portland St., Columbia, MO, 65201, bDepartments of Pharmacology and Ophthalmology, University of Missouri-Columbia School of Medicine, Columbia, MO 65212, and c Department of Physiology and Endocrinology, Medical College of Georgia, Augusta, GA 30912, U.S.A. (Received Chicago 3 September 1991 and accepted in revised form 9 December 1991). The potent cr,-adrenoceptorantagonist, [3H]prazosin,exhibited high affinity, specific and reversible binding to intact rabbit, bovine and human cornea1endothelialcellsin culture. The binding of 1 nM [3H]prazosinto rabbit cellsreacheda steady-statelevel within 10 min at 37°C. Undertheseconditions, approximately 50% of the [3H]prazosinbound wasspecific.The level of specific[3H]prazosinbinding was concentration-dependent,but Rosenthalanalysisindicatedthat [3H]prazosinbound to at leasttwo sites. One site exhibited a high affinity for [3H]prazosin(K, = 0.2 nM). but a relatively low binding capacity (4mx = 175 fmol boundmg-l protein); the other siteshoweda relatively low affiity for the radioligand (K, = 85 nM), but a much higher binding capacity (1280 fmol mg-I). Severalknown a,-adrenoceptor antagonistsand agonistscompetitively inhibited [3H]prazosinbinding at the high affinity site when incubated with the radioligand. The relative potenciesof these competing ligands were generally consistentwith their binding tinities for a,-adrenoceptorsin other tissues.Phenylephrinestimulatedthe rate of hydrolysis of phosphatidylinositol4,5-bisphosphateto inositol 1,4,Wrisphosphateby 63% in these cells. This stimulation was inhibited by 52% if phentolamine was also present during the incubation. Thesedata indicatethat cornea1endothelialcellshave cc,-adrenoceptors which can modulate polyphosphoinositide turnover in this tissue. Key words:eye: cornea: endothelium: cc,-adrenoceptors : ol,-adrenergicreceptors: inositolphosphate: phosphoinositide : rabbit; bovine; human.

natriuretic

1. Introduction

Alpha,-adrenoceptors are expressed by a wide variety of cells in biological tissues(Bylund, 19 8 7 ; Minneman, 1988). but their existence and possible role in cornea1 endothelial physiology have not yet been determined. Several preliminary reports which failed to detect these receptors in particulate fractions of this tissue may have delayed progress in this area (Neufeld, Olsen and Zowistowski, 1978; Walkenbach et al., 1983). Furthermore, the lack of established hormonal effects on cornea1 deturgescence may account for the limited general knowledge concerning hormonal regulation of the cornea1 endothelium. There is, however, substantial evidence for the existence of cornea1 endothelial /3-adrenoceptors (Walkenbach and LeGrand, 1982; Chao and Walkenbach, 1985: Walkenbach et al., 1985). There is also evidence for

peptide receptors (Korenfeld. Ye and

Walkenbach, 1990). In our studies of cornea1 hormonal

receptors, we

made the unusual finding that hormonal receptors in

the cornea could often be identified only if intact cells of cultured or native tissue were utilized. Receptors appeared Co be destroyed or masked if native or cultured cornea1 cells were broken to produce particulate fractions. This observation was made in studies characterizing muscarinic cholinoceptors in cornea1 endothelium (Walkenbach and Ye, 1990) and epithelium (Walkenbach and Ye, 19 9 1) as well as for cornea1 epithelial a,-adrenoceptors (Ye, Walkenbach and Reinach, 1990).

endothelial cells, such as adenosine receptors (Walkenbach and Chao, 1985), prostaglandin E, receptors (Neufeld et al., 1986; Jumblatt and Paterson, 1990), muscarinic cholinoceptors (Walkenbach and

These experiments, using intact cultured and native cornea1 endothelial cells from several species,provide evidence for the existence of cl,-adrenoceptors on these cells which regulate the rate of polyphosphoinositide metabolism. Alpha,-adrenoceptor stimulation of the second messengers inositol 1,4,5-trisphosphate and 1,2-diacylglycerol are known to be physiologically relevant in many other tissues (Berridge, 1987). This system may also prove to be important in the

Ye, 1990). EGF receptors (Raymond et al., 1986;

regulation of cornea1 endothelial homeostasis.

the existence of other hormonal receptors on cornea1

Yoshida et al., 1989; Junquero et al., 1990) and atria1 * For correspondence and reprintrequests. 0014-4835/92/090443+08

$08.00/O

0 1992 AcademicPressLimited

444

R. J. WALKENBACH

TABLE

[“H]prazosin

ET AL.

I

binding to cormal endothelial tissues [YH]prazosin bound (fmol mg ’ protein)

Species

Tissue preparation

Rabbit

Cultured, intact cell Cultured, part. fract. Native, part. fract. Cultured, intact cell Cultured, part. fract. Native, part. fract. Cultured, intact cell

Bovine Human

Total (n)

Non-specific (n 1

Specific

353.0-&28.8 (21) 125.3k9.8 (15) 146.2 k 16.3 (20) 191.9 + 14-8 (12) 87.2kl3.3 (12) 1050$13.1 (12) 506.6 k 30-9 (I 5 1

1794k13.6 (21) 137.4k7.7 (15) 142.9 * 11.7 (20) 129.9T9.0 (12) 98.5-J IO.1 (12) 108.5+8.8 (12, 373.3+ 16.4 (15)

173.2* < 0 3.3 62-O* < 0 < 0 133.3*

Cornea1 endothelial tissues were incubated with 1 nM [“H]prazosin (total binding) or with 1 nM [“H]prazosin+ 0.1 mM phenylephrine (nonspecific binding) and processed as described in Materials and Methods. Data indicate the means k SAM. of the number of assays indicated in parentheses. * Significant differences (P < 0.01) by Student’s t-test.

2. Materials

and Methods

Tissue Preparations

Rabbit and bovine eyes were obtained from local packing houses within 2 hr after slaughter. The eyes were kept on ice for up to four additional hours until further processing occurred. Rabbit and bovine cornea1 endothelial cells were cultured in multiwell plates (Falcon, Be&on Dickinson & Co., Lincoln Park, NJ) of 24-wells (binding experiments) or 12-wells (polyphosphoinositide turnover experiments) as described previously (Walkenbach et al., 1985). Briefly, whole eyes were liberally rinsed with a phosphate-buffered saline solution (PBS) containing 10 mM K,HPO, in 0.9 % NaCI, pH 7.4. Endothelia were scraped from the corneas and cultured in Dulbecco’s Modified Eagle Medium (MEM), with other agents and culture conditions as described previously. The medium was changed after 1 week of culture and twice weekly thereafter. Each well typically contained 100 pug (24-well plates) or 300 ,ug (12-well plates) of ceilular protein when used for experiments after 34 weeks of culture. Particulate fractions of native and cultured cornea1 endothelium were prepared as described previously (Walkenbach et al., 1985). Human corneas (donor ages 75-85 years) were gifts from the Missouri Lions Eye Tissue Bank. The corneas (with 2 mm of sclera) were isolated from the rest of the eye within 12 hr post-mortem and stored in Dexsol (Chiron Ophthalmics, Irvine, CA) or a similar medium using M-199 tissue culture medium and supplemented with 1.3 5 Y. chondroitin sulfate, 1% dextran (40 000 kDa), 17 mM Na bicarbonate and 20 mM Hepes buffer. The medium was then adjusted to a final pH of 7.4 with 1 N NaOH and sterilized by ultrafiltration. Endothelial cells from human corneas were cultured using significant modifications to the method described by Yue et al. (1989). Cultures were established within

96 hr of cornea1 storage, typically using groups of six to ten corneas. Each cornea was dissected within the corneal-scleral border while submerged in storage medium. The cornea was then rinsed with sterile PBS and placed epithelial side down in a well of 24-well culture plate. The endothelial surface of the cornea was filled with approximately 1 ml of 0.5 % trypsin, 16 mM EDTA in Hank’s balanced salt solution (BSS) and incubated for 45 min at 3 7°C. The cellular suspension was then withdrawn and the endothelial surface rinsed with 15 ml of MEM containing 20% horse serum. The cellular suspensions and rinse medium from each treated cornea were pooled and centrifuged at 1500 rpm for 10 min at room temperature. The supernatant was decanted and the cellular contents resuspended with MEM containing 1.5 Y0 chondroitin sulfate, 10% newborn calf serum, 1 mM sodium pyruvate, 2 mM I-glutamine, 2 5 mM Hepes buffer and an antibiotic mixture of 100 U ml-’ penicillin, 0.1 mg ml-l streptomycin and 250 ng ml-’ amphotericin B. The cells were gently agitated to promote complete isolation and then cultured in wells of a 12-well plate, at a density of one cornea per well. Incubation conditions and medium changing patterns were identical to those for rabbit and bovine cultures. Human cultures were used for binding experiments 34 weeks after the start of culture. At that time, cells typically covered SO-75 % of the wells and produced loo-150 pg of cellular protein per well. Assay Protocols

The potent ol,-adrenoceptor antagonist, [3H]prazosin NEN Research Products, Boston, MA) was used to assess receptor binding activity in these tissue preparations. The protocols were analogous to those described for [3H]quinuclidinyl benzilate binding to muscarinic cholinoceptors in cornea1 endothelial (87 Ci mmol-I:

tissue

(Walkenbach

and Ye, 1990).

binding of the radioligand fractions

Briefly,

to endothelial

was assessed by incubating

total

particulate

the tissue with

a,-ADRENOCEPTORS

IN CORNEAL

ENDOTHELIUM

FIG. 1. Phase contrast micrographs of adult human cornea1 endothelial cells in culture. A, Endothelial cells early in the second week of culture. Small groups of cells begin to form mosaic-like patterns. B, During the second and third weeks of culture, the size of the groups of cells exhibiting a mosaic-pattern enlarges. Cell size, however, becomes smaller. C, During the fourth week of culture, numerous large fields of endothelial cells in a mosaic-pattern are commonly seen. x 100; actual field size = 0% x 1.5 mm.

the indicated concentration of [3H]prazosin in 25 mM glycylglycine buffer, pH 7.6, at 37’C for 30 min (unless otherwise indicated). The assays were then filtered and washed free of unbound radioligand with three S-ml aliquots of ice-cold buffer. Each filter was then counted by standard liquid scintillation techniques. Non-specific binding of [3H]prazosin was measured by running parallel assays with 0.1 mM

phenylephrine added to the reaction mixtures during the incubation. Specific binding of [3H]prazosin was defined as the difference between the measured total [3H]prazosin bound and the non-specific level bound in each set of parallel assays. Binding protocols using intact cultured cells were identical except that incubations were performed in MEM without serum and incubations were terminated

446

by aspirating the medium and quickly rinsing three times with 1 ml of ice-cold PBS. After the final PBS rinse, 1 ml of 1 N NaOH was added to each well and digestion occurred overnight at room temperature before measuring for protein and radioactivity. Polyphosphoinositide turnover in cultured rabbit cornea1 endothelial cells was measured using the technique described by Martin (1983 ). The growth medium was removed and the cells washed several times with inositol-free Minimum Essential Medium (IFMEM). This medium was prepared by mixing Hank’s BSS (Sigma Chemical Co., St Louis, MO: H138 7) with MEM amino acid mixture (Sigma M7020) and adding the individual vitamins as listed by the MEM formulation. The cells were then cultured for 48 hr with 3 ml of fresh IFMEM. supplemented with 2 % dialyzed sterile calf serum (Sigma C5542) and 33 nM myo-[3H]inositol (15.6 Ci mmol-’ ; NEN Research Products) in order to generate an endogenous pool of radiolabeled phosphatidylinositol 4,5bisphosphate. The labeling medium was then removed, the cells were washed with PBS and then preincubated for 5 min at 37°C in serum-free IFMEM with 10 mM LiCl to block dephosphorylation of inositol-l-PO, to inositol (Berridge, Downes and Hanley, 1982). The cells were then incubated for 1.0 min with fresh serum-free IFMEM. LiCl and the drugs indicated in Table II. Reactions were terminated by aspirating the incubation medium and adding 1 ml of ice-cold 10 y0 HClO, to each well. The cells in the wells were frozen, thawed, and kept on ice for 30 min to complete the extraction of [3H]inositol phosphates from the cells, then decanted. The residue in each well was saved for subsequent protein assays ; the extract was neutralized using approximately 1 ml of 75 mM Hepes in 2 N KOH. The neutralized extract was centrifuged at 2000 rpm for 2 min at 4°C. One milliliter of each supernatant fraction was applied to columns with a 3 x 0.7 cm diameter bed of Dowex 1 resin. [3H]inositol, [3H]inositol- l(1X8-200) phosphate (IP,), [3H]inositol-l,4-bisphosphate (IP,), and [“Hlinositol- 1,4,5-trisphosphate (IP,) were isolated from each column by retaining the effluents after sequential applications of: 6 ml of 0.1 M formic acid : 10 ml of 0.2 M ammonium formate in 0.1 M formic acid: 10 ml of 0.6 M ammonium formate in 0.1 M formic acid: and 7.5 ml of 1.0 M ammonium formate in 0.1 M formic acid, respectively. A 2.5-ml sample of each effluent was mixed with 10 ml of Scintiverse BOA cocktail, stored in the dark for 3 1 hr and the radioactivity measured using a liquid scintillation counter. In preliminary experiments, standards of [3H]inositol and each of the above-mentioned [3H]inositol phosphate derivatives (NEN Research Products) were used to determine optimal isolation conditions. Protein was measured using the method of Lowry et al. (19 5 1) for all tissue preparations. Unless otherwise indicated, the data are presented as the means f S.E.M.

I? J. WALKENBACH

ET AL.

of three experiments. each using triplicate assays per experimental condition. Unless otherwise indicated. all reagents were purchased from Sigma Chemical Co., St Louis, MO and other supplies were obtained from Fisher Scientific, Springfield, NJ. 3. Results The levels of [3H]prazosin binding to various tissue preparations of rabbit, bovine and human cornea1 endothelium are compared in Table I. Intact cells from rabbit and bovine cornea1 endothelial cultures exhibited significant levels of specific [3H]prazosin binding. However, particulate fractions from endothelia of the same culture dates prepared using typical techniques (Bylund, 198 7) failed to exhibit specific [3H]prazosin binding. Furthermore, particulate fractions from native rabbit or bovine cornea1 endothelium failed to exhibit significant levels of specific [3H]prazosin binding. This lack of binding activity in cornea1 endothelial broken cell preparations may explain the earlier negative findings concerning CLadrenoceptors in this tissue (Neufeld et al., 1978; Walkenbach et al., 1983). Because of this observation, subsequent experiments employed intact cells. Soluble fractions of these cornea1 endothelial tissue preparations also showed no specific [“Hlprazosin binding activity (data not shown). Adult human cornea1 endothelium in primary culture also expressed significant levels of specific [3H]prazosin binding. Under these culture conditions, human cornea1 endothelial cells appeared very flat and amorphous during the first week of culture. During the second week, however, small groups of cells which resembled the characteristic endothelial mosaic pattern began to form [Fig. l(A)]. During the third and fourth weeks, the groups of cells became larger in size, while the density of cells mrne2 also increased [Fig. 1 (B)]. Human endothelial cultures typically reached quiescence after 4 weeks with fairly large areas of cells in a mosaic pattern. Their cell densities within the mosaic areas were approximately 1200 cells rnrnm2 [Fig. 1 (C)l. The characteristics of [3H]prazosin binding were further investigated using rabbit cornea1 endothelial cell cultures. Figure 2 shows that specific binding of [3H]prazosin to endothelial cells increased with assay time for up to 10 min ; at this point a steady state was achieved which remained stable for at least 50 additional minutes. Under these conditions, specific [3H]prazosin binding comprised approximately onehalf of the radioligand’s total binding. The relationship between [3H]prazosin assay concentration and the level of radioligand binding to endothelial cells is portrayed in Fig. 3. In general, both the total and non-specific [3H]prazosin binding levels rose over the indicated range of radioligand concentrations. However, binding was not always commensurate with radioligand concentration. The level

a,-ADRENOCEPTORS

IN CORNEAL

ENDOTHELIUM

447

9 0,

60

-

c

0

-

I

I

20

40

60

20-

-9

-8

-7

Assay time (mm) FIG. 2. Kinetics of [3H]prazosin binding to rabbit cornea1 endothelial cultures. Cells were incubated with 1 nM [3H]prazosin alone (-a-) or with 1 nM [3H]prazosin + 0.1 mM phenylephrine (-+--) at 3 7°C for the times indicated. Specific binding of [3H]prazosin (-A -) was calculated as described in Materials and Methods.

of specific [3H]prazosin

-6

Log unlabeled

binding plateaued between

-3

agents : phentolamine

(-4--)

: yohimbine

(--*-)

:

phenylephrine (-A-); methoxamine (-+--) ; or norepinephrine (- 0 -) under otherwise standard conditions. Control was defined as the quantity of [3H]prazosin bound using radioligand alone.

1

Bound

-4 (M)

FIG. 4. Inhibition of [3H]prazosin binding to rabbit cornea1 endothelium by unlabeled alpha-adrenoceptor antagonists and agonists. Cells were incubated with 1 nM [3H]prazosin and the indicated concentration of one of the following

and 3 nM radioligand before increasing again at higher concentrations. Rosenthal analysis (196 7) of these data (Fig. 3, inset) indicated two sites of [3H]prazosin binding to cornea1 endothelial cells under the experimental conditions employed. One binding site exhibited a dissociation constant (I$) of 0.2 nM and a maximal binding capacity (B,,,) of 175 fmol [3H]prazosin mg-’ protein. The lower affinity site showed a K, of 85 nM and an apparent B,,, of 1280 fmol mg-‘. Unlabeled ligands with known potency as CQadrenoceptor antagonists and agonists were able to

50

-5 llgand

compete with [3H]prazosin for binding to cornea1 endothelial cells. Figure 4 summarizes the experiments in which varying concentrations of unlabeled ligands were incubated with endothelial cultures in the presence of 1 nM [3H]prazosin. The a-adrenoceptor antagonist, phentolamine, was the most potent competitor tested: requiring O-5 ,UM to inhibit 50% of [3H]prazosin binding (IC,,). Yohimbine, a selective a,-adrenoceptor antagonist, also competed with [3H]prazosin but exhibited an IC,, of 0.8 PM, which is

too

150

(PM)

log [3Hlprazosm

(nM)

FIG. 3. Dose-response relationship of [3H]prazosin binding to rabbit cornea1 endothelial cultures. Cells were incubated with the indicated concentration of [3H]prazosin alone (---a---) or with radioligand +O.l mu phenylephrine (-+-). Specific [3H]prazosin binding (---A---) was calculated as described previously. Inset, Rosenthal (1967) graphica analysis of these data.

448

R. J. WALKENBACH

60 -

‘-

401 0

4

8 Time

12

16

(mln)

FIG. 5. Reversibility of [3H]prazosinbinding to rabbit cornea1endothelialcultures.All cellswere incubatedfor 30 min at 37°C with 1 nM [3H]prazosin. Unlabeled phenylephrine wasthen addedto 0.1 mM and the cellscontinued to incubate for the indicated period of time before termination of the assays.The control value wasdefinedasthe level of [3H]prazosinbound just prior to phenylephrine addition. approximately 1000 times more than needed to interact with a,-adrenoceptors (Bylund, Ray-Prenger and Murphy, 1988). The a,-adrenoceptor agonists phenylephrine, methoxamine and norepinephrine also competed with [3H]prazosin at somewhat higher concentrations. Their IC,, values were 10, 25 and 50 ,uM, respectively. The /3-adrenoceptor agonist, isoproterenol, did not compete with [3H]prazosin binding using concentrations up to 1 mM (data not shown). The reversibility of [3H]prazosin binding is depicted in Fig. 5. Cornea1endothelial cells were first incubated with 1 nM [3H]prazosin for 30 min in order to reach a steady state as seen in Fig. 2. At this point, all wells received 0.1 mM phenylephrine and binding reactions were terminated at the times indicated after phenylephrine addition. Excess unlabeled phenylephrine caused a time-dependent decrease in [3H]prazosin bound for 10 min, after which a new steady state was established at a non-specific level of [3H]prazosin binding. The relationship between n,-adrenoceptors in cor-

ET AL.

neal endothelial cells and the turnover of inositol polyphosphoinositides is shown in Table II. After culturing cells with myo-[3H]inositol to adequately radiolabel phosphatidylinositol 4,5-bisphosphate, differences in the rate of hydrolysis to inositol 1,4,5trisphosphate and its metabolites were estimated by incubating cells in control medium, with 1 mM phenylephrine or with phenylephrine + 0.1 mM phentolamine. Phenylephrine caused a 63 % increase in inositol 1,4,5-trisphosphate formation, while increasesover control were also seenin its metabolites. When phentolamine was also present during the incubation, phenylephrine stimulated inositol 1,4,5trisphosphate formation by only 30 %, indicating that a,-adrenoceptors modulate polyphosphoinositide turnover in the cornea1 endothelium. 4. Discussion These experiments offer several lines of evidence for the existence of ol,-adrenoceptors in cornea1 endothelial cells. [“Hlprazosin binds specifically to intact rabbit, bovine and human cornea1 endothelial cells (Table I). The non-specific binding of [3H]prazosin was relatively high, but this is not an unusual finding when intact cells are employed (Bylund, 1987). Any radioligand which diffused into the cells or was trapped in the pericellular space at the termination of an incubation would have been measured as non-specific binding. Even though these factors make it more difficult to show statistical differences between total and non-specific [3H]prazosin binding levels, there remains unequivocal evidence for the presence of specific z,-adrenoceptors. Intact cells were needed in order to show specific binding of [“Hlprazosin to endothelial cells because cellular disruption appeared to destroy or mask the radioligand’s ability to bind specifically to particulate fractions of rabbit or bovine cornea1 endothelium (Table I). The loss of CI,adrenoceptor binding ability because of cellular disruption is an unusual finding, but it is consistent with our laboratory’s previous results with LX,adrenoceptors in human cornea1epithelium (Ye et al., 1990) as well as for muscarinic cholinoceptors in

TABLE II

Polyphosphoinositide turnover in cultured rabbit cornea1endothelium Phosphoinositide metabolites(cpm rng-’ protein) Condition Control 1 mM phenylephrine 1 mM phenylephrine+ 0.1 mM phentolamine The data indicate incubation with the significantly higher significantly less IP,

IPI

IP,

8037+687 15820+2013 12124flOl

2521f153 2103&90 2633+437

IP3 2132f56 3485k193 2776$240

the means k S.E.M. of each inositol fraction from Dowex chromatography after preincubation with myoj3H]inositol and above drugs, as described in Materials and Methods. The quantity of IP, in cells treated with phenylephrine was than control cells (P < 0.001). Furthermore, the cells incubated with phenylephrine and phentolamine contained than cells treated with phenylephrine alone (P < 0.05).

a,-ADRENOCEPTORS

IN

CORNEAL

cornea1 endothelium (Walkenbach and Ye, 1990) and epithelium (Walkenbach and Ye, 1991). Particulate fractions of cornea1 endothelium will exhibit specific [3H]prazosin binding if incubated at 2°C (data not shown). It seems plausible that cornea1 cells may use proteolysis or other degradative activity upon cellular disruption as a physiological defense mechanism against invasion by foreign organisms. The characteristics of [3H]prazosin binding in cultured cornea1 endothelial cells were typical of the radioligand’s binding to a,-adrenoceptors in many tissues (Minneman, 1988) in terms of its association kinetics (Fig. 2); its competitive inhibition by known a-adrenoceptor agents (Fig. 4) and its reversibility, as shown by displacement of the radioligand with excess unlabeled phenylephrine in Fig. 5. The [3H]prazosin binding affinity and saturation characteristics (Fig. 3) were complicated by the apparent presence of more than one binding site for the radioligand. The higher

affinity

449

ENDOTHELIUM

(K,, = 0.2 nM) and lower binding

capacity

= 175 fmol rng-‘) site is in the range of binding affinity and binding capacity for a,-adrenoceptors in a wide variety of tissues (Bylund, 1987). The charac(B,,,

teristics of the lower affinity (K, = 85 nM) and higher capacity (B,,,, = 12 80 fmol mg-‘) site is not consistent with characteristics of physiological a,-adrenoceptors

in other tissues. The observation of multiple binding sites is not unusua1, especially in experiments using intact cells (Skomedal. Aass and Osnes, 1984). The expression of substantial levels of [3H]prazosin specific binding to rabbit, bovine as well as human

cornea1 endothelial cells indicates that these receptors may be a characteristic of cornea1 endothelial cells in general and raises the clinical relevance of a,adrenoceptors in this tissue. Although culture of adult

cornea1 endothelium has been recently reported in other laboratories (Yue et al., 1989; Bergsma et al., 199 1 ; Wilson and Lloyd, 199 1). the simplified method described herein appears to produce more consistent

results and larger yields using older donor tissue. Stimulation of polyphosphoinositide turnover in these cells (Table II) by an a,-adrenoceptor agonist (phenylephrine) and its inhibition by the addition of an X, -adrenoceptor antagonist (phentolamine) offers independent evidence for the presence of functional a,-adrenoceptors in cornea1 endothelium. Alpha,adrenoceptor regulation of phosphatidylinositol 4,5bisphosphate hydrolysis to inositol 1,4,5-trisphosphate and IJdiacylglycerol are thought to be important second messengers in the regulation of many cell types (Berridge, 1987; Minneman, 1988). The physiological role of x,-adrenoceptors and their

related second messengers in the cornea1 endothelium is unknown. Our laboratory has failed to detect any consistent change in cornea1 deturgescence rates in

response to x,-adrenergic agents (data not shown). It is possible that these receptors are involved in more subtle control of endothelial physiology and/or pathology, such as in the development of Fuchs dystrophy

or bullous keratopathy. Additional studies will be needed to define more clearly the physiological relevance of a,-adrenoceptors in the cornea1 endothelium.

Acknowledgements This work was supportedby NIH grants EY 02 597 and EY 04795.

References Bergsma,D. R.. Lopez, J. G.. Insler, M. S. and Malter. J. S. (1991). Flow cytometric DNA analysis of cultured human cornea1endothelium. Invest. Ophthalmol.Vis. Sci. 32 (Suppl.). 1178. Berridge, M. J. (1987). Inositol trisphosphateand diacylglycerol: two interacting secondmessengers. Arm. Rev. Biochem. 56. 159-93. Berridge, M. J., Downes,P. C. and Hanley, M. R. (1982). Lithium amplifies agonist-dependentphosphatidylinositolresponsein brain and salivary glands.Biochrm. J. 206, 587-95. Bylund, D. B. (1987). Biochemistry and pharmacology of the alpha-adrenergicreceptor.In The Alpha-I Adrenergic Receptors. (Ed. Ruffolo.

R. R. Jr). Pp. 19-69.

The

Humana Press:Clifton, NJ. Bylund, D. B.. Ray-Prenger, C. and Murphy. T. J. (1988). Alpha-2A and alpha-2B adrenergicreceptorsubtypes: antagonistbinding in tissuesand cell linescontaining only one subtype.J. Pharmacol. Exp. Ther. 245, 600-7. Chao, W. T. H. and Walkenbach, R. J. (1985). Desensitization of ,&adrenergic receptors in cornea1endothelium. 1. OcuJ. Pharmacol. 1, 263-8. Jumblatt. M. M. and Paterson,C. A. (1990). PGE,effectson cultured cornea1endothelium: cyclic AMP synthesis and cell shapeare regulatedby an EP, receptor.Invest. Ophthalmol. Vis. Sci. 31 (Suppl.).249. Junquero.D., Modat. G.. Coquelet.C. and Bonne,C. (1990). Retinoids enhance the number of EGF receptors in cornea1endothelialcells.Exp. Eye Res. 51, 49-53. Korenfeld. M. S.. Ye, G. S. and Walkenbach, R. J. (1990). Atria1 natriuetic peptide:cornea1endothelialreceptors in three species.Invest.Ophthalmol. KS. Sri. 31 (Suppl.), 145. Lowry, 0. H., Rosebrough,N. J.. Farr, A. L. and Randall, R. J. (1951). Protein measurementwith the Folin phenolreagent. I. Biol. Chem. 193, 265-i5. Martin, T. F. (198 3). Thyrotropin-releasinghormonerapidly activatesthe phosphodiester hydrolysisof polyphosphoinositidesin GH, pituitary cells. I. Rio!. Chem. 258, 14816-22.

Minneman, K. P. (1988). a,-Adrenergic receptor subtypes, inositolphosphates,and sources of cell Ca2+. Pharmacol. Rev. 40, 87-119.

Neufeld.A. H.. Jumblatt,M. M.. Markin, E.D. and Raymond, G. M. (1986). Maintenance of cornea1endothelialcell shape by prostaglandinE,: effects of EGF and indomethacin. Invest. OphthaJmoJ. Vis. Sci. 27, 143742. Neufeld, A. H.. Olsen,J. S. and Zawistowski,K. A. (1978). Autonomic receptorsin the rabbit cornea.Ass. Res. Vis. OphthaJmoJ. Abs. Pp. 189. Raymond,G. M.. Jumblatt,M. M., Bartels.S. P. and Neufeld, A. H. (1986). Rabbit cornea1endothelialcellsin vitro: effectsof EGF. Invest. Ophthalmol. Vis. Sri. 27, 474-9. Rosenthal,

H. E. (1967).

Graphic

method

for the deter-

450 tnination and presentation of binding parameters in a complex system. Anal. Biochem. 20. 525-32. Skomedal. T., Aass, H. and Osnes, J.-B. (1984). Specific binding of [3H]prazosin to myocardial cells isolated from adult rats. Biochern. Pharmacol. 33, 1897-906. Walkenbach, R. J., Bylund. D. B., Chao, W. T. H. and Gibbs. S. R. (1983). Adrenergic and cholinergic receptors in the cornea. Invest. Ophthalmol. KS. Sci. 24 (Suppl.). 198. Walkenbach, R. J.. Bylund, D. B., Chao, W. T. H. and Gibbs, S. R. (198 5). Characterization of /3-adrenergic receptors in fresh and primary cultured bovine cornea1 endothelium. Exp. Eye Res. 40, 15-2 1. Walkenbach, R. J. and Chao. W. T. H. (1985). Adenosine regulation of cyclic AMP in cornea1 endothelium. 1. Ocular Pharmacol. 1. 33 742. Walkenbach, R. J. and LeGrand, R. D. (1982). Adenylate cyclase activity in bovine and human cornea1endothelium. Invest. Ophthalmol. Vis. Sci. 22, 120-4. Walkenbach, R. J. and Ye, G. S. (1990). Muscarinic receptors

R. J. WALKENBACH

ET AL.

and their regulation of cyclic GMP in cornea1 endothelial cells. Invest. Ophthalmol. Vis. Sri. 31, 702-7. Walkenbach, R. J. and Ye, G. S. (1991). Muscarinic cholinoceptor regulation of cyclic guanosine monophosphate in human cornea1 epithelium. Invest. Ophthnlmol. Vis. Sci. 32, 610-I 5. Wilson, S. E. and Lloyd, S. A. (1991). Growth factor and growth factor receptor mRNA production by cultured human cornea1 endothelial cells. Invest. OphthnlmoI. Vis. Sci. 32 (Suppl.), 953. Ye, G. S.. Walkenbach. R. J. and Reinach. P. S. (1990). Alpha,-adrenergic receptors in human cornea1 epithelium. Invest. Ophthalmol. Vis. Sci. 31 (Suppl.), 466. Yoshida, A., Laing, R. A., Joyce, N. C. and Neufeld. A. H. (1989). Effects of EGF and indomethacin on rabbit

cornea1endothelialwound closurein excisedcorneas. Invest. Ophthalmol. Vis. Sci. 30, 1991-6. Yue. B. Y. J. T.. Sugar, J.. Gilboy, J. E. and Elvart. J. L. (1989). Growth of human cornea1 endothelial cells in culture. Invest. Ophthalmol. Vis. Sci. 30, 248-53.