The direct vascular relaxing action of betaxolol, carteolol and timolol in porcine long posterior ciliary artery

SURVEY OF OPHTHALMOLOGY

VOLUME 38. SUPPLEMENT - MAY 19%

The Direct Vascular Relaxing Action of Betaxolol, Carteolol and Timolol in Porcine Long Posterior Ciliary Artery R. KELLY HESTER, PHD,**~ ZHOU CHEN, BS,2 ELIZABETH J. BECKER, MARSHA MCLAUGHLIN, PHD,~ AND LOUIS DESANTIS, PHD~

BS,

‘Department ofMedical Pharmacology and Toxicolog~v and ‘Microcirculatzon Keseorch Institute, Colle e of Medicine, Texas AUM University Health Science Center, College Stotim. aud 9;Alcon Laboratories, Research nrhd Development, Fort Worth, Texas

Abstract. The vascular relaxing properties of three beta adrenoceptor

antagonists, betaxo101, carteolol and timolol, currently used in the treatment of glaucoma, were characterized, compared and contrasted in the porcine long posterior ciliary artery. Isolated arterial ring segments precontracted with increased extracellular KC1 (plus 40 mM) or the thromboxane analog, U-46619 (3 x lo-’ M), were relaxed in a concentration-dependent fashion by betaxolol, carteolol, timolol or nitroprusside. In vessel segments depolarized with increased extracellular KCl, EC50 values indicated that the intrinsic relaxant sensitivitv to betaxolol was equal to that of nitroprusside, six-fold greater than that of carteolol, and ienfold greater than that of timolol. Similarly, the maximum relaxation occurring at equimolar concentrations (lo-” M) for the beta adrenoceptor antagonists was betaxolol > carteolol = timolol. Qualitatively similar results were noted in ring segments of the rabbit external iliac artery precontracted with increased extracellular KCI (plus 30 mM). Underconditions in which specific receptor-linked events are absent and voltage-gated (:a+ ’ entry is maximized, the Ca+ + concentration response relationship in porcine long posterior ciliary artery was shifted to the right in an apparent competitive manner by betaxolol, reflecting a 5.6-fold reduction in the sensitivity to Ca++. Conversely, nitroprusside reduced the Ca++ sensitivity three-fold in a noncompetitive fashion; not only shifting the Ca+ + concentration response relationship to the right, but also depressing the maximum by 57%. Porcine long posterior ciliary arterial segments precontracted to a similar degree with U-46619, in which voltage-gated Ca++ entry is only one component of many specific cell signalling transduction mechanisms contributing to the precontraction. exhibited a sensitivity to betaxolol that was six-fold less than to nitroprusside, but two-fold greater than to timolol and 20-fold greater than to carteolol. These results are consistent with an obvious direct vascular relaxing capacity for beta adrenoceptor antagonists that primarily represents a capacity for inhibiting voltage-gated Ca’ + entry in vascular smooth muscle. Additionally, the differential potencies of these three beta adrenoceptor antagonists characterized in this study suggests that this property is much more likely to contribute to an) potentially beneficial effects of betaxolol than carteolol or timolol. (Surv Ophthalmol 38 [Suppl, May]:S125-SlY4. 1994)

Key words. beta adrenoceptor long posterior ciliary artery voltage-gated Ca + + entry

l

antagonist nitroprusside

l

betaxolol timolol l

l

carteolol glaucoma vascular relaxant . l

l

l

laxation. 1~ uitro studies in canine coronary arteries”.‘” clearly demonstrated that proprano101 actually had obvious vascular relaxant properties. These properties resembled those of the <:a+ + channel blockers, and it was further suggested that these effects might contribute to the

A little over a decade ago, several studies focused on the potential for propranolol, when used in the treatment of angina, to promote coronary vasospasm as a result of vasoconstrictive mechanisms being unopposed in the absence of any beta adrenoreceptor-mediated res125

S126

Surv Ophthalmol

HESTER ET AL

38 (Suppl) May 1994

therapeutic efficacy of this compound in the treatment of angina. Over the years similar relaxing properties have consistently been confirmed for propranolol, and also have been noted for other beta adrenoceptor antagonists in other species and several other peripheral vessels.‘~“~‘“~‘g However, all beta adrenoceptor antagonists are not equally effective vascular relaxants. Generally speaking, propranolol consistently causes significant relaxation, regardless of species or vessel type.“*‘“.‘g,2”*‘6Additionally, betaxolol, bunitrolol and tertatolol have been shown to have obvious vascular relaxant effects in rat aorta’ and rat femoral, mesenteric or renal arteries.‘.lg Conversely, in comparison, ateno101,‘~practolol,‘” and timolol’” rarely show much, if any, vascular relaxing capacities. During this same decade, the topical use of beta adrenoceptor antagonists in the treatment of glaucoma to lower intraocular pressure (IOP) became well established. However, alterations in IOP do not appear to be the only beneficial effects noted in the treatment of glaucoma with beta adrenoceptor antagonists.4.xs’7.‘8.” Potential beneficial effects on blood flow might contribute to the therapeutic efficacy as well. Any concomitant and dynamic interactions or interplay between potential effects on ocular blood flow and the known effects on IOP may be more important in terms of overall therapeutic benefits than effects on IOP alone. In fact, a recent studyI clearly demonstrated that both perfusion pressure and IOP may interact dynamically and affect the ability of ocular vessels to autoregulate blood flow. Thus, since the complete mechanism involved in the effectiveness of beta adrenoceptor antagonists in the treatment of glaucoma has not yet been fully delineated, the following study was initiated. The primary purpose was to determine in a comparative study the intrinsic vascular relaxant capacities of three beta adrenoceptor blocking drugs currently used clinically in the treatment of glaucoma - betaxolol, carteolol and timolol - utilizing a relevant ocular vessel. It was reasoned that by determining these properties in a mechanistic and comparative manner the results might provide some insight or imply suggestions that would allow resolution of whether these effects might be relevant and could possibly contribute to the therapeutic effectiveness of these agents. The following study characterizes the apparent direct vascular relaxing properties of these three beta adrenoceptor antagonists and compares and contrasts these

properties among these three beta adrenoceptor antagonists as well as with those of the potent nitrovasodilator, nitroprusside.

Methods Porcine eyes were supplied by Animal Technologies, Inc. (Tyler, Texas). Eyes were rapidly removed and placed in a cold Tris-buffered physiological saline solution (PSS) within 20 min after death. They were then shipped in PSS on ice to College Station, TX, where they were subsequently stored in a similar solution at -4°C overnight. The following morning, eyes were aerated with 100% 0, for approximately 30 min before the long posterior ciliary artery (LPCA) was subsequently dissected free of surrounding fat and connective tissue under a dissecting microscope. External iliac arteries were isolated and removed from New Zealand White rabbits weighing 1.5-2.2 kg that were euthanized following an injection of sodium pentobarbital (50 mg/kg) via an ear vein and subsequent exsanguination after severing both carotid arteries. A segment S-10 mm in length was isolated from the LPCA (300-400 km, ID) just proximal to where the artery branches into upper and lower branches. Segments lo-12 mm in length were isolated from the external iliac artery (500-600 km, ID) in a similar manner. Segments were then cut perpendicular to the long axis of each vessel segment into ring segments 1.5-2.0 mm in width which were fixed vertically between two stainless steel triangles.” One triangle was then attached to a stainless steel rod, which held the tissue in place in a 20 ml water-jacketed, isolated organ bath filled with PSS. The upper triangle was then attached to a FT-03 Grass force displacement transducer for measurements of isometric force, which was displayed on a Grass Model 79A polygraph. Each vessel segment was then allowed to equilibrate in PSS under an optimal preload of 3.5 g for the LPCA and 1.5 g for the external iliac artery for 120 min prior to specific experimental protocols. The solution in the organ bath was continuously aerated with 100% O,, changed every 10-l 5 min, and maintained at 37°C for the duration of the experiment. The PSS contains in mM/L: NaCl, 154; KCl, 5.4; dextrose, 11 .O; tris(hydroxymethyl)aminoethane, 6.0; CaCl,, 2.0 and MgCl,, 1.2. The pH of the solution was adjusted to 7.40 & 0.03 with 6 N HCl at 36-37°C. A similar solution without added Ca++ and containing an additional 40 mM KC1 substituted equimolar for NaCl (45.4 mM total) was utilized to characterize the Ca+ +

VASCULAR RELAXING ACTION OF BETA BLOCKERS Ciliary

Artery

/ /

It +

10 nM gallopamll 100 “M gallopamll

KCI Concentration(mM) I“tg. 1, C:oncentration-dependent inhibitory effects of the Ca+ + channel blocker, gallopamil, on the concentration response relationship for KC1 in porcine I.PCA. Maximum developed force to 80 mM KCI (control) was 1.25 k 0.10 grams tension in the absence of gallopamil, 0.60 ? 0. IO grams tension in the presence of 10 nM gallopamil; and only 0.24 2 0.06 grams tension in the presence of 100 nM gallopamil (N = 8).

concentration response relationship in the LPCA. All solutions included 3 x lo-” M phentolamine to eliminate any potential effects of released catecholamines due to KC1 or any direct interactions of beta adrenoceptor antagonists with alpha adrenoceptors.‘,‘” Individual agents used in this study included KCl, U-46619 (Cayman Chemical Co., Ann Arbor, MI), nifedipine (Fujisawa, Osaka, Japan), nitroprusside (Sigma Chemical Co., St. Louis, MO), phentolamine hydrochloride (Ciba, Summit, NJ), papaverine hydrochloride (Sigma), gallopamil (Knoll AG, Ludwigshafen, Germany), betaxolol (Synthelabo, Paris, France), carteolol (Otsuka, Osaka, Japan), timolol (Mediolast S.p.A, Milano, Italy) and nimodipine (Bayer AG, Leverkusen, Germany). Nifedipine was initially dissolved in lO%%ethylene glycol, 20%’ alcohol and 70% demineralized, deionized water, and then further diluted in PSS.“’ Final concentrations of solvents in the tissue baths were never >0.2%. lndividual concentration response relationships for vascular relaxants were completed in a cumulative manner (without any intervening washout of bath chambers) utilizing LPCA vessel segments precontracted with either the potent thromboxane analog, U-46619 (3 X 10m7 M), or increased KC1 (total 45.4 mM) substituted equimolar for N&l and the external iliac artery with 30 mM additional KC1 (total 35.4 mM). After the addition of a threshold concentration of each re-

s127

laxant, vessel segments were allowed to reach a new, lower steady-state force before the addition of the next subsequent higher concentration. This procedure was continued until the maximal response for each agent was delineated. After the maximally effective concentration of each vascular relaxant was determined, lo-’ M papaverine was then added to define the maximal relaxation attainable in each tissue. In the porcine LPCA precontracted with either KC1 or U-46619 and the rabbit external iliac artery precontracted with KCl, the relaxation following the addition of papaverine after these relaxants is nearly always greater than the total force developed in response to these constrictors. l’hus, total relaxation relative to initial developed force was usually greater than 100%. Mean concentrations resulting in 50% of the maximum relaxation attainable (EC-,,, values) following the addition of papaverine were quantified for each individual concentration response relationship and are presented as geometric means with associated 95% confidence intervals. Calcium concentration response relationships in LPCA were completed in a (:a’+ free, high KC1 (plus 40 mM substituted equimolar for N&l) depolarizing solution in both the absence and presence of selected vascular relaxants. After the initial 120 min equilibration period in PSS. vessel segments were exposed to a (;a’ + free, high KC1 depolarizing solution for 30 min prior to the cumulative addition of Ca’ + (0.03-10.0 mM). Vascular relaxants were added 10 min pri(0.03 mM). or to the initial Ca++ concentration All data are presented as means t standard errors (SE) in grams tension (g) or percentages (%). Statistical analyses were accomplished by using Student’s t-test for unpaired samples and one-way analysis of variance followed by Dunnett’s test. Dif‘ferences were considered significant if p < 0.05.

Results Initial studies were directed towarcl insuring that responses to KC1 in porcine LPCA, like those in most other arteries, were primarily the result entry through typical “L” type voltageof Ca’ + channels. As occurs in many other gated Ca’ ’ vessels, the KCl-induced concentration response relationship (25-80 mM) is shifted downward and to the right in a concentration-dependent manner by the Ca + + channel blocker, gallopamil (Fig. 1). Qualitatively, similar results were noted channel the dihydropyridine Ca’ ’ with blockers, nimodipine and nifedipine (data not

S128

Surv Ophthalmol

38 (Suppl)

May 1994

shown). These results indicate that KCl-induced responses in LPCAapparently reflect Ca++ entry through typical depolarization-activated, dihydropyridine-sensitive Ca+ + channels. Subsequent studies utilized LPCAs precontracted with KC1 in order to characterize the concentration-dependent relaxant effects of betaxo101, carteolol and timolol. This concentration of KC1 produced approximately 80% of the maximum response attainable to KC1 in this vessel. The relaxant effects of these beta adrenoceptor antagonists were concomitantly compared to those of the potent, nitric oxide-dependent vascular relaxant, nitroprusside (Fig. 2). Potassiuminduced tone is relaxed in a concentration-dependent manner by all four vascular relaxants. However, the sensitivity of the precontracted LPCA to these relaxants is betaxolol = nitroprusside > carteolol = timolol. The EC,, values (Table 1) indicate that, whereas the sensitivity to nitroprusside and betaxolol is essentially equal, the sensitivity of the porcine LPCA is six-fold greater for betaxolol (and nitroprusside) than carteolol, and ten-fold greater than timolol. At equimolar concentrations (1 0m4 M), betaxolol produced 48.1 + 3.8% of the maximum relaxation attained following the addition of lo-” M papaverine; whereas carteolol produced 16.7 f 4.2%; timolol 16.6 + 2.1% and nitroprusside 46.3 k 4.5%. Thus, the KC1 precontracted porcine LPCA was equally as sensitive to betaxolol as nitroprusside, and thus both agents were not only equipotent, but equieffective at 10m4 M, whereas the sensitivity to carteolol and timolol at the same concentration was comparatively low. In this vessel the maximum relaxation attainable, relative to the initial level of KCl-induced force, following the addition of 10m4 M papaverine subsequent to betaxolol, carteolol, timolol and nitroprusside was correspondingly, 157.1 + 10.3% 153.3 k 10.8%, 153.1 k 5.9% and 160.2 k 10.1%. Thus, not only was the KCl-induced response completely relaxed, but the initial unstimulated, basal force was also reduced. This suggests that there is a high degree of intrinsic, possibly myogenic tone in the porcine LPCA. In order to determine if there were any vascular selectivity to the relaxant actions characterized above, similar experiments were repeated in a representative peripheral vessel from another species, the rabbit external iliac artery. When this vessel was precontracted with KCl, it also relaxed in a concentration-dependent fashion to betaxolol, carteolol, timolol and nitroprusside

HESTER Ciliary

-%--

Artery

betaxolol carte0101 timolol nitroprusside

* -Ue

Y

ET AL

-5Oj .OOl

.Ol

.l

1

10

100

1000

Concentration(pM) Fig. 2. Relaxant effects of the beta adrenoceptor an-

tagonists, betaxolol, carteolol and timolol, in porcine LPCA precontracted with a 40 mM substituted KC1 solution compared to the relaxant effects of nitroprusside. The maximum forces developed in response to KC1 for these four vascular relaxants were, respectively, 1.44 k 0.20, 1.42 k 0.19, 1.45 + 0.10 and 1.39 k 0.15 grams tension (N = 10-24). The maximum amounts of relaxation were, correspondingly, 1.90 k 0.21, 1.10 2 0.16, 0.80 2 0.07 and 0.85 2 0.10 grams tension. The respective maximum attainable relaxations following the addition of 10e4 M papaverine were 2.11 k 0.22, 2.02 + 0.22, 2.10 + 0.11 and 2.03 k 0.19 grams tension.

(Fig. 3). In comparison to the porcine LPCA, the KC1 precontracted rabbit external iliac artery is ten-fold more sensitive to nitroprusside (p < 0.005), two-fold more sensitive to betaxolol (p < O.OOl), equally sensitive to timolol, and 4.5fold less sensitive (p < 0.001) to carteolol (Table 1). Futhermore, in contrast to the porcine LPCA, the KCl-induced tone in this vessel is slightly more sensitive to nitroprusside than betaxolol (2.5-fold), and much less sensitive to timolol(25fold) and carteolol (63-fold). At equimolar concentrations (1 Oe4 M) and similar to the results in the LPCA, nitroprusside and betaxolol were again equally potent and equally effective vascular relaxants eliciting 72.0 2 3.1% and 72.6 k 3.0%, respectively, of the maximum relaxation attainable following the addition of 0.1 mM papaverine. Conversely, at this same concentration ( 10e4 M), timolol was only minimally effective (13.6 _t 3.2%) and carteolol elicited little, if any, relaxation (1.4 k 0.6%). In the rabbit external iliac artery precontracted with KCl, the maximal relaxation attainable following the addition of

VASCULAR RELAXING ACTION OF BETA BLOCKERS

s129

TABLE 1 EC50

Values and Associated 95% Confidence Intervals for Betaxolol arteolol, Timolol and Nitroprusside in Porcine LPCA + M U-46619 and Rabbit External Iliac Artery Precontracted with 40 mM Substituted KC1 Solution or 3 X IOPrecontracted with 30 mM Substituted KCL

Precontracting Agent _

Porcine LPCA

Rabbit External Iliac Artery KC1

KC1

U-46619

Betaxolol

1.05 x 1O-4M (0.74-1.49) [N= 141

2.69 x 1O-4M (1.774.09) [N=23]

4.48 x 10-j M (3.72-5.40) [N= 131

Carte0101

6.22 x 1O-4M* (3.62-10.67) [N= 121

5.29 x 1O-3M* ( 1.13-24.6) [N=9]

2.81 x 10.’ M* (1.41-5.61) [N= 121

Tim&l

1.00x 1O-3M* (0.45-2.26) [N=21]

5.19x 1O-4M (2.92-9.22) [N = 301

1.13x lo-” M* (0.39-3.24) [N=ll]

Nitroprusside

1.14 x 1O-4 M (0.49-2.64) [N = 201

4.42 x lo.5 M* (2.24-8.72) [N = 231

1.81 x 10.” M** (0.89-3.68) [N = 141

*Significantly different (p < 0.001) from respective betaxolol **Significantly different (p < 0.02) from respective betaxolol

IO-’ M papaverine after the highest concentration of betaxolol, carteolol, timolol or nitroprusside used was correspondingly, 111.5 + 2.3%, 105.9 -+ 1.9%, 112.0 * 1.7% or 108.8 ? 1.0% of the KCl-induced contraction. This indicates a much lower degree of intrinsic tone in this vessel as compared to the porcine LPCA. The next series of experiments were designed to further clarify the potential Ca’ + entry blocking ability of these beta adrenoceptor antagonists suggested by the previous experiments. Thus, the effects of approximately equipotent concentrations of betaxolol (most potent) and timolol (least potent) on the Ca++ concentration relationship, which represents Ca++ entry through voltage-gated Ca’ ’ channels, were characterized and again compared to those of an equieffective concentration of nitroprusside (Fig. 4). The control EC,,, value for Ca++ in the presence of high KC1 was 5.10 X lo-’ M (3.66-7.11). In the presence of lo-’ betaxolol, the Ca’+ concentration response relationship was shifted to the right and consequently the EC,, value was increased significantly (p < 0.001) to 2.86 X lo-” M (2.16-3.78). This represents a 5.6-fold reduction in sensitivity without significantly altering the maximum force developed in the presence of 10 mM Ca’ + An equipotent (relative to relaxing the KCl-induced contractile response) and equimolar concentration (lo-’ M) of nitroprusside response realso shifted the Ca ’ + concentration

EC50 value. EC50 value.

lationship to the right, indicative of a reduction in sensitivity. However, nitroprusside, in contrast to betaxolol, significantly (p < 0.001) reduced the maximum response. The EC,(, value for Ca++ m of nitroprusside was . the presence increased to 1.50 x lo-:’ M (1.02-2.19) reflecting a three-fold depression in sensitivity, whereas the maximum response was concomitantly attenuated by 56.6%. Conversely, the highest concentration of timolol (3 x lo-” M) used in this study had only a marginal effect on sensitivity [EC,,), 1.05 x lo-’ M (0.6-1.8)] and no effect on the maximum. These results suggest potential differences in the mechanism of relaxing action of the beta adrenoceptor antagonists and nitroprusside. The rightward shift of the concentration response relationship without a concomitant reduction in the maximum is indicative of competitive type inhibition of voltage-gated ca++ entry. Conversely, nitroprusside not only shifts the relationship to the right, but also reduces the maximum by more than half, suggesting a noncompetitive type of voltage-gated Ca + + entry inhibition. Finally, the relaxant effects of betaxolol, carteolol, timolol and nitroprusside were characterized and compared in the porcine LPCA precontracted with the thromboxane analog, U-46619. This provided an avenue for contrasting the effectiveness of these vascular relaxants on specific receptor-mediated mechanisms of activation rel-

~130

Surv Ophthalmol

HESTER ET AL

38 (Suppl) May 1994

Iliac

Artery

Ciliary

Artery

control betaxolol timolol nitroprusside

---3A U

q

Y

betaxolol carteolol timolol nitroprusside

-2oL----7.Ol .l

,

I

1

10

100

1000

Concentration(pM) Fig. 3. Comparative relaxant effects of betaxolol, carteolol, timolol and nitroprusside on KCI-induced responses in rabbit external iliac artery. The maximum developed forces in response to a 30 mM substituted KC1 solution were, respectively, 4.29 _C 0.34, 5.20 2 0.45,4.64 k 0.34 and 4.80 + 0.38 grams tension (N = 11-14). Maximum relaxations to each relaxant were, correspondingly, 4.62 f 0.31, 2.10 ?I 0.26, 1.68 + 0.26 and 4.16 k 0.17 grams tension. Maximum attainable relaxations to each vascular relaxant plus 10m4 M papaverine were 4.72 t 0.30, 5.47 k 0.45,5.15 2 0.35 and 5.19 * 0.37 grams tension, respectively.

ative to the above studies on general membrane depolarization and consequent voltage-gated Ca++ entry. Vessels were precontracted with a concentration ofU-46619 (3 X IO-‘M) that produced a sustained level of force that was approximately equal to that produced by the additional 40 mM KCl. All four vascular relaxants reduced U-4661 g-induced tone in a concentration-dependent manner (Fig. 5). Not unexpectedly, the U-466 1g-induced contraction was 2.6-fold more sensitive to the relaxant actions of nitroprusside than the equipotent KCl-induced contraction (Table 1). Conversely, this contraction was 2.6fold less sensitive (p < 0.005) to betaxolol than the corresponding KCl-induced contraction. Sensitivity of the U-466 19 precontracted vessel to betaxolol was 6.1-fold less than to nitroprusside but still markedly greater than carteolol (19.7-fold). Betaxolol (10e4 M) produced 35.4 + 3.1% of the maximum relaxation attainable following the addition of papaverine at the end of each experiment. Similarly, at equimolar concentrations nitroprusside was equieffective and produced 34.1

.1

.Ol

Calcium

1

10

Concentration(mM)

Fig. 4. The

inhibitory effects of betaxolol (1O-4 M, timolol (3 x 10m4 M) and nitroprusside (10e4 M) on the Cat + concentration response relationship in porcine LPCA. Vessels were exposed to a Cat+ free, substituted KC1 (plus 40 mM substituted equimolar for NaCl) solution for 30 min prior to the addition of the initial concentration of CaCl,. Vascular relaxants were added 10 min prior to Caf ’ (20 min after the Ca+ + free, high KC1 solution). Initial developed forces in response to a 40 mM substituted KC1 solution containing 2.0 mM CaCl were 1.65 ? 0.12 (N = 26), 1.67 k 0.27 (N = 9), 1.66 k 0.13 (N = 8), and 1.62 k 0.19 grams tension (N = 7) for control, betax0101, timolol and nitroprusside groups, respectively. Corresponding maximum responses to 10 mM Ca+ + KC1 solution were added to a Ca+ + free, substituted 1.98 -C 0.13, 1.63 k 0.31, 1.91 k 0.16 and 0.86 2 0.19 grams tension.

+ 3.1%, whereas timolol produced only 21.8 4 2.7%, and carteolol only 12.7 * 3.9%. The maximum concentrations of betaxolol (lo-’ M), carteolol (1O-3 M), timolol (3 X 1O-4 M) and nitroprusside (1 O-4 M) used elicited relaxations that were correspondingly, 68.8 + 2.7%, 36.7 * 5.0%,39.7 _’ 2.8%and62.1 -+ 3.6%ofthemaximum papaverine-induced response. Thus, even though the U-46619 precontracted vessel is much more sensitive to nitroprusside than betaxolol (relative to EC,, values), these drugs were approximately equieffective at 10m4 and lo-” M (68.8 vs 62.1%), respectively. The maximal relaxations attainable following the addition of papaverine as a function of the U-46619-induced response were respectively, 163.2 + 11.2%, 155.5 f 9.7%, 135.1 2 5.0%, and 150.6 2 8.7%. These results again suggest a high degree of intrinsic,

VASCULAR RELAXING ACTION OF BETA BLOCKERS Ciliary

Artery

75 -

50

25

’

+

belaxolol

*

carte0101

-+

timolol

e

nitroprusside

1 -25) .Ol

.l

1

10

100

1000

Concentration(yM) /‘lg. 5. The concentration-dependent relaxant effects of betaxolol, carteolol and timolol in porcine LPC.4 precontracted with 3 x 1 O-’ M U-46619compared to nitroprusside. Ring segments were first constricted with Ll-466 19 and subsequently relaxed in a cumulative and concentration-dependent fashion. Finally. these vessels were totally relaxed with lo-’ M papa~erine. The maximum forces developed in response to 1!-46619 for these four vascular relaxants were, rcspectivcly, 1.31 2 O.lJ,O.88 rt 0.14, 1.10 -t 0.08and 1.3 I t- 0.13 grams tension (N = 9-17). Maximum of aniounts relaxation were, correspondingly, I.36 & 0.12, 0.45 ? 0.07, 0.60 ? 0.07 and 0.96 2 0.07 grams tension. The respective maximum attainable relaxations following the addition of lO_ M papaverine were 2.01 ? 0.17, 1.29 2 0.16, 1.47 +_ 0.12 and 1.60 III 0.13 grams tension.

possibly

myogenic

tone

in the porcine

LPCA.

Discussion ‘This study has clearly illustrated the direct vascular relaxing effects of three of the principle beta adrenoceptor antagonists used currently in the treatment of glaucoma - betaxolol, carteolol and timolol. This was readily apparent in an ocular vessel (porcine LPCA) and was also confirmed in a more peripheral artery of another species (rabhit external iliac artery). The results are consistent with previous studies using the satne or different beta adrenoceptor antagonists in other vessels, such as rabbit ear artery,’ canine coronary artery,““.“’ rat aorta,’ rat femoral, mesenteric and renal arteries,‘.” and bovine retirtal arAlso consistent within this same terioles.” literature is that all beta adrenoceptor antagonists are not equally potent with regard to their ability to relax vascular smooth muscle. In the

s1:31

current study, arteries from the porcine eye and the hindlimb of the rabbit were much more sensitive to the vascular relaxing properties of betaxolol than either timolol or carte&l. Thus, betaxolol elicited more relaxation than these other two beta adrenoceptor antagonists at equimolar concentrations, which is similar to previous studies comparing betaxolol to atenolol and bunitrolol in rat arteries2 or comparing propranolol to timolol in bovine retinal arterioles:’ The data in the current study further suggest that a primary component of the vascular t-elaxant effects of these beta adrenoceptor antagonists results from inhibition of (;a’ + entering through typical L type, voltage-gated Ca’ + channels. Additionally, it has been concluded that this action is not due to any direct interactions with either alpha or beta adrenoceptors and probably does not represent any significant degree of membrane stabilizing or local anesthetic-like effects. Finally, the potency of betaxolo1 relative to carteolol and timolol for this direct vascular relaxing action suggests that if’ this property in any way might contribute to the beneficial effects seen on visual field improvement and retinal sensitivity in the chronic treatthat it is more likely to be ment of glaucoma“~” associated with the actions of betaxolol than the other two beta adrenoceptor antagonists. The vascular relaxing actions characterized in this study are probably not directly related to arty selective interactions with beta adrenoceptors. like bovine retinal arterThe porcine LP(IA,” ioles,’ relaxes little, if at all following beta adrenoceptor stimulation. Furthermore, we have found that neither the epirtephrine nor the norepinephrine concentration response relationship in the LPC;A is affected tor blockade unpublished relaxing

(Hester, observations).

capacity

indicates

that

at either

beta,

by beta adrenocrp(Ihen and Becker; Similarly

the limited

noted

in this study

of carteolol

intrinsic

sympathomimetic

or 1’adrenoceptors

does

activity not

cott-

significantly to the vascular relaxant effects of these beta adrenoceptor antagonists. Additionally, beta adrenoceptor selectivity does not seem to play a role in this vascular t-ektxant action. Even though betaxolol, a cardioselective beta adrenoceptor antagonist, is a much more potent vascular relaxant than either- timo101 or carteolol, both nonselective beta adrenoceptor antagonists, propranotol, another wmselective antagonist, seems to be equally as efrective in relaxing bovine retinal arterioles” as betaxolol is in relaxing the porcine L.l’( :A. Furthermore. tribute

S132

Surv Ophthalmol

38 (Suppl) May 1994

other studies using the beta, selective or cardioselective agent, atenolol,26 have clearly demonstrated that atenolol had minimal effects on Kfinduced vascular tone. The results of this study also imply that any membrane stabilizing ability or local anestheticlike properties do not contribute significantly to the vascular relaxing action of beta adrenoceptor antagonists, since none of the three beta adrenoceptor antagonists currently used has any such readily apparent action. The results of other reported studies3,‘5,25,26 and our own unpublished results clearly suggest that propranolol, a beta adrenoceptor antagonist with significant membrane stabilizing ability, appears to be equal to betaxolol in terms of the vascular relaxant properties characterized in this study and in that of Bessho et al.’ However, at least one of these studies15 has clearly shown that the vascular relaxing profile of the local anesthetic, lidocaine, is quite different from that of propranolol. We also have found that the relaxant profile of both procaine and lidocaine in porcine LPCA more clearly resembles that of timolol than betaxolol (Hester, Chen and Becker; unpublished observations). The most likely mechanism of the vascular relaxing action of the beta adrenoceptor antagonists characterized in this study appears to be inhibition of voltage-gated Caf ’ entry. The sensitivity of the KCl-induced response in the LPCA to the concentration-dependent inhibitory effects of the Ca+ + channel blockers, gallopamil, nifedipine and nimodipine, certainly suggests that the increase in force due to increased K+ in this vessel is primarily due to Ca’+ entry resulting from membrane depolarization and consequent activation of L type, voltage-gated Ca++ channels. Thus, the betaxolol-induced relaxation of K+-induced responses implies inhibition of potential-dependent Ca++ entry via these typical “L” type Ca+ + channels. Previous studies with propranolol in isolated cardiac and vascular smooth muscle have suggested that beta adrenoceptor antagonists may directly interact in an inhibitory fashion with this dihydropyridine-sensitive Ca’ + channel.28 Likewise, the inhibitory profile of betaxolol on the Ca+ + concentration response relationship in porcine LPCA indicates that the inhibitory effect may be competitive in nature. Betaxolol clearly shifts the relationship to the right in a parallel manner without altering the maximum response to the highest concentration of Ca’ + (10 mM). Alternatively, these data suggest that at a constant level of membrane depolarization (due to

HESTER ET AL increased extracellular KC]), and presumably a constant and maintained increase in Ca++ conductance (depolarization activated, L type Caf + channels), that the inhibitory effects of betaxolol are reversed by increasing the extracellular Ca+ + concentration. Previous studies in rat aorta have demonstrated similar parallel shifts in the Cat+ concentration response relationship for betaxolol and the Ca’ + channel blocker, diltiazem’; and the relaxing ability of propranolol in canine coronary artery was readily reversed by increasing the extracellular concentration of Ca + +.‘5.26Additionally, more recent studies in rat aorta,* using the Ca ++ indicator dye, fura-2, have shown that increasing the extracellular Ca++ concentration in the presence of betaxolol clearly results in an increase in intracellular Ca++ concentration and simultaneously a reversal of the betaxolol-induced relaxation. Extensive studies over the years have consistently demonstrated that the inhibitory effects of the classical “L” type Ca’ + channel blockers are usually reversed by increasing the extracellular concentration of Ca+ +.5,7 Thus, a major component of the vascular relaxant action of betaxolol appears to be inhibition of potential-dependent Ca++ entry. Similar conclusions were gleaned from comparisons of the differential responsiveness of the Ca++ concentration response relationship to betaxolol and nitroprusside. Nitroprusside not only shifted the relationship to the right, but additionally depressed the maximum response, indicative of a noncompetitive type of inhibition. This certainly suggests that the mechanisms of action of betaxolol and nitroprusside are different. Likewise, the marked differential sensitivity of the U-46619-induced contraction to nitroprusside and betaxolol (6.1-fold greater for nitroprusside than betaxolol) again implies different primary mechanisms of action. These implications are consistent with earlier comparative studies using K+-induced depolarization and specific receptor-mediated mechanisms (e.g., norepinephrine, serotonin, prostaglandin F,,) for increasing force to clearly delineate the selectivity of Ca++ channel blockers for potential-dependent Ca+ + entry and nitroprusside and other nitric oxide producing vascular relaxants for specific receptor-linked Ca++ entry processes’“.‘3,14 that may include, but are not limited to voltage-gated Ca++ channels. Whereas increased K+ depolarizes the cell membrane activating voltage-gated Ca++ channels and consequently inducing Ca+ + entry, receptor-medi-

VASCULAR RELAXING ACTION OF BETA BLOCKERS ated processes stimulate Ca+ + release from internal stores and Ca+ + entry through both receptor-linked as well as voltage-gated Ca+ + channels. One can only speculate on the clinical relevance of these findings. The peak concentration of‘betaxolol and timolol determined in the aqueous humor within an hour after topical administration of either timolol”‘.“-24 or betaxolol?t’,‘“.“’ was approximately lo-” M, whereas that for carteolol was even higher.‘” This concentration is certainly sufficiently high”‘~‘“” to account for the inhibitory effects of these agents on any beta adrenoceptors involved in aqueous humor production and thus could be involved in their known capacity for reducing IOP.‘7,” This concentration also represents the threshold concentration for the direct vascular relaxing action of betaxolol, as reported by Bessho et al’ and demonstrated in our current study. However, it is much lower than the threshold concentrations for atenolol,“’ bunit.rolol,’ or timolol”; we had similar findings for timolol and carteolol in 0111 current study. This supports the concept that additional factors, other than beta adrenoceptor blockade, might contribute to the therapeutic benefits of this agent. In this regard, recent studies4,‘x have indicated that the extent of IOP lowering capacity for betaxolol is somewhat less than that of timolol, yet the longterm effects on visual field and retinal sensitivity are better for betaxolol than timolol. Betaxolol has been clearly shown to reduce total peripheral resistance, whereas atenolol, another cardioselective beta adrenoceptor antagonist, or propi-anolol, a nonselective agent, increases resistance.‘)7 It is conceivable that the vascular relaxing action manifested peripherally in 7li710 might also contribute to the beneficial effects noted in the eye. Thus. this direct vascular relaxing action 7iu inhibition of voltage-gated Ca+ + entry might contribute to any potential capacity for maintaining optic nerve head blood flow either through localized vasodilation” or inhibition of vasospasm.” In fact, a recent study” has shown that nicardipine, a dihydropyridine Ca’ + channel blocker, increased optic nerve head blood tlow and maintained retinal blood flow despite XI associated reduction in mean systemic blood pressure.

Conclusion In conclusion, the beta adrenoceptor antagonists, betaxolol, carteolol and timolol, have direct vascular relaxing properties in addition to their

s133 capacities for blocking beta adrenoceptors. The primary mechanism of this action appears to result from inhibition of Ca’+ entry through depolarization-activated membrane Ca’ + channels. The potency of betaxolol relative to carteolol and timolol for eliciting vascular relaxation implies that this action is much more likely to be seen with betaxolol than with the other two beta adrenoceptor antagonists. It is further suggested that this direct vascular relaxing action and any potential associated ef‘fects on ocular blood flow might contribute to the overall therapeutic profile seen with long-term use of betaxoiO1.

References 1. Ashhrook

DW. Purdy RE, Hurlbut DE, et al: A novel response to propranoloi: contractile response in the isolated rabbit ear artery. L//e Scz 26: 155-163. 1980 9_. Bessho H, Suzuki J. Tobe A: Vascular effects ofhetaxolol, a cardioselective p-adrenoceptor antagonist. in isolated rat arteries. Jpn J Pharnuxol 55:351-358. 1991 HAJ, De Mey JGR: Ef3. Boonen HCM, Struyker-Boudier fects of tertatolol on the responsiveness of isolated femoral. mesenteric, and renal resistance arteries to adrenergic stimuli. J Cnrdiomasr Phnnnnc-ol 15: 124- 129, 1990 J: Long-term effect of ophthalmic P4. Collignon-Brach adrenoceptor antagonists on intraocular pressure and retinal sensitivity in primary open-angle glaucoma. C:w? Eve K?\ 11: l-3, 1992 A: Specific pharmacology ot calcium in 5. Fhxkenstein myocardium, cardiac pacemakers, and vascular smooth muscle. ,dnn Ktw Pharmncol Tox~cnl 17:149-166. 1977 L. Mahler F: Do vaso6. Gasser P. Flammer J, Guthauser spams provoke ocular- disease? .lq$11:21S-220, 1990 I‘, Miller R. Wiho M: Calcium antagonism 7. Godfraind and calcium entry blockade. Pl~cc~mc~v/ Rn)\ 3X::\2 l-4 16. 1986 JE, Delehanty J: Efftct of topical carteolol on 8. Grunwald the normal human retinal circularicm. Iwu\l O~~~~~~~?~~~ IL ,%I 33:18X-1856, 1992 nicardipinr 9. Harino S, Riva CE. Petrig BL: Intravenous in cats increases optic ne;ve head hut not r-etinal h&d flow Irrwsl Obhlhnlmol I’;.\ Sci 3 3:2X85-2890. 1992 nitrate on 10. Hester RK: The effects of 2-rricotinamidoethyl agonist-sensitive (;a+ ’ release and Ca+ ’ entry in rabbit acjrta. J Phnrmnd EX/J Thrr 2 3 ?: 1OO- I 1 I, 19X5 11. Hester RK. Chen 2. Becker E,], et al: The vascular rrlaxant effects of hetaxolol. carteolol and timolol cm potassium-induced tone in the porcine long posterior ciliary artery (abstract). 11nw’l Oph/ha/mo/ l’1.r SC/ (Supfil)3?:8()9. 1992 in ‘I‘vson CA. 12. Hester RK. Ramos KS: Vessel Cylinders. FraLiel- IM teds): M&o~c 1)~7hxrf o/oe~. San Diego. AGdemic &ess; 19k3, pp 169-l X 1 “. and D600 IS. Hester RK, Weiss GB: Effects ofnitroprusslde on norepinephrineand KCI-stimulated (:a+ ’ activation and contraction systems in canine renal veins as compared to canine renal artery. ,/ (:crrtllorw\r Phrumnd h:762-571. 1984 I-4. Hester RK. Weiss GB, Frye N’J: Diftering actions of nitroprusside and D-600 on tension and “;‘<:a tluxcs in canine renal arteries.,] Phrtmmcol E.ul, 7‘lrrr 170X:155-160, 1979 LJ, et al: EtFects ot beta15. Haste AM, Boels PJ. Andries antagonists on contraction of hovine retinal microarteries IT)xr~tw. fnrv\l O~hthnlmol I’/.\ SC{ 31: 1231-1237, 1990 AP: .4utoregulation of choroldal 16. Kirl JL$‘, Shepherd

S134

17. 18.

19.

20.

21.

22.

23.

24.

25.

Surv Ophthalmol

38 (Suppl)

May

1994

blood flow in the rabbit. Invest Ophthnlmol Vis Sci 33: 2399-2410, 1992 Lesar TS: Comparison of ophthalmic B-blocking agents, Clin Pharmacol 6:45 1-463, 1987 Messmer C, Flammer J, Stiimplig D: Influence of betax0101 and timolol on the visual fields of patients with glaucoma. Am J Ophthalmol 112:678-681, 1991 Mostaghim R, Maddox YT, Ramwell PW: Endothelial potentiation of relaxation response to beta adrenoceptor blocking agents. JPharmacolExp Ther239;797-801,1986 Phan TM, Nguyen KPV, Giacomini JC, Lee DA: Ophthalmic beta blockers: determination of plasma and aqueous humor levels by a radioreceptor assay following multiple doses. J Ocular Pharmacol 7:243-252, 199 I Pillunat L, Stodtmeister R: Effect ofdifferent antiglaucomatous drugs on ocular perfusion pressure. J Ocular Pharmacol4:231-242, 1988 Polansky JR: Side effects of topical ophthalmic therapy with anti-inflammatory steroids and B-blockers. Curr Opin Ophthalmol3:259-272, 1992 Polansky JR, Alvarado JA: Isolation and evaluation of target cells in glaucoma research: hormone receptors and drug responses. Curr Eye Res 4:267-279, 1985 Polansky J, Friedman Z, Fauss D, et al: Effects of betaxolol/timolol on epinephrine stimulated cyclic-AMP levels in human trabecular meshwork cells. Int Ophthalmol 13: 95-98, 1989 Rokutanda M, Araki S, Sakanashi M: A pharmacological

HESTER

ET AL

investigation on a possible calcium antagonistic action of propranolol. Arch Int Pharmacodyn Ther 262:99-108. 1983 M, Takeo S: Characterization of propranolol26. Sakanashi induced relaxation of coronary artery. Jpn J Pharmucol 33:603-6 10, 1983 on 27. Satoh N, Suzuki J, Bessho H, et al: Effects ofbetaxolol cardiohemodynamics and coronary circulation in anesthetized dogs: comparison with atenolol and proprano101. ~pn J Piarmacoi54:113-119, 1990 ’ A 28. Weishaar RE. Ouade MM. Pueslev TA. et al: Interaction between propr&olol and ca&um channel blockers in cardiac and vascular smooth muscle. J Mol Cell Cardiol 20:897-903, 1988

This study was partially supported by a grant from Alcon Laboratories, Inc. (Fort Worth, TX). Thanks are extended to Patty Sampson for her patient assistance with the preparation and processing ofthe manuscript. Preliminary reports of this study were presented at the Annual Meeting of the Association for Research in Vision and Opthalmology, Sarasota, Florida, May 3-8, 1992. Reprint address: Dr. Kelly Hester, Department of Medical Pharmacology and Toxicology, College of Medicine, Texas A&M University Health Science Center, College Station, Texas, U.S.A. 77843-l 114.