Prostaglandin F2α increases uveoscleral outflow in the cynomolgus monkey

Erp.

Eye Res. (1989)

49, 389-402

Prostaglandin

F,, Increases Cynomolgus

B’ANN lkpartment

TRUE

GABELT

of Ophthalmology,

(Received 19 December

Uveoscleral Monkey* ANL)

University Wisconsin.

Outflow

in the

PAUL L. KACFMAN~

of Wisconsin U.S.A.

Medical

1988 and accepted i?r revl.sed form

School, Madisoll.

22 2farch

19x9)

Cynomolgus monkeys were treated topically in one eye twi(*e daily with prostaglandin F,,-lisopropylester (PGF,;IE) for nine doses. On treatment day 4. 3 hr after the seventh dose. intraocular pressure (IOP) in the treated eye was reduced by 65% compared to the controls. to < 5 mmHg. On treatment day 5, 3 hr after the ninth dose, total outflow facility was determined by two-level constant pressure perfusion of the ant,erior rhamber. Immediately thereafter, uveoscleral outfIow was determined by intracamerally infusing 11p51j- or [‘“‘T/-albumin and fluoresceinated dextran, and calculating the volume of anterior chamber fluid required to have deposited the quantity of tracer recovered from the various ocular and periocular tiuuues. Simultaneously, trabecular outflow was determined by calculating the volume of anteriol chamber fluid required to have deposited the quantity of tracer recovered from the general circulation. Total facility was - 50% higher in treated than in control eyes, but the effect was variable, of marginal statistical significance, and perhaps due to increased pseudofacility or uveoscleral facility. Uveosc!eral outflow was approximately two to three-and-a-half times higher in treated than in control eyes, the magnitude of the effect being dependent upon the timing and pressure at which the perfusion was conducted. Trabecular outflow was reduced by - 75 % in the treated eyes relative to control so that, the proportion of total outflow comprised by trabecular outflow in the treated eyes was only one third that in the controls. Total aqueous flow was slightly (- 20 %) but not significantly reduced in the treated eyes. The IOP lowering effect of PGF,, in the cynomolgus monkey is due largely if not exclusively to an increase in uveoscleral outflow of aqueous humor, with aqueous outflow being redirected from the trabecular to the uveoscleral route. Key words: aqueous humor outflow ; ciliary muscle; intraocular pressure ; LV~caca f”asciculari.~ : prostaglandin F,, . trabecular meshwork; uveoscleral outflow.

1. Introduction The ocular hypotensive effect of low doses of topically applied prostaglandin PGF,, has been demonstrated in normotensive rabbits (Camras, Bito and Eakins, 1977 ; Lee, Podos and Severin, 1984), cats (Bito, Draga, Blanc0 and Camras, 1983a; Bito, Srinivasan, Baroody and Schubert, 1983b; Stern and Bito, 1982), monkeys (Bito et al., 1983a; Camras and Bito, 1981; Camras. Podos, Rosenthal, Lee and Severin. 1987 ; Crawford, Kaufman and Gabelt, 1987 ; Crawford and Kaufman, 1987 ; Lee et, al., 1984; Stern and Bito, 1982) and humans (Alm and Villumsen. 1986; Giuffrd, 1985; Lee, Shao, Xu and Qu, 1988), and in glaucomatous monkeys (Camras and Bito, 1981; Lee, Podos, Howard-Williams and Severin, 1985; Lee, Podos, Serle and Camras, 1987) and humans (Villumsen and Alm, 1987). The mechanism for the intraocular pressure (IOP) reduction was unclear unt,il it was demonstrated that pilocarpine antagonizes PGF,,-induced ocular hypotension in the cynomolgus monkey, suggestlowers IOP by increasing uveoscleral outflow (Crawford and ing that PGF,, Kaufman, 1987). Nilsson, Stjernschantz and Bill (1987) indirectly estimated * Supported by National Eye Institute grant EY02698. t To whom reprint requests should be addressed at the Department Wisconsin, Clinical Science Center, 600 Highland Avenue, Madison, 0014~1835/89/0~389+

14 $03.00/O

of Ophthalmology, Wisconsin 53792, 0 1989 Academic

University U.S.A. Press I.imitrd

of

:I!)0

I(. ‘I‘ (:.\I~lcl“I’

uveos~leral

outflow

of tot,al aqueous intraramerally

in this

flow infused

a single

qualitatively

by

We

here

report.

measurements

species

I’ I. li.\I~E’\I.-\N from

the

t,hrough the ant’erior chamber [ “‘1 I- or I’x1I )-aihumin) and

thr general circulation [ “‘1 I- or [“‘I ]~aIbnmin). following

monkey

.4SI)

(calculaM from They c*al(~ulatrtl

topical

dose

of

autoracliography t,he effect.

of uvcoscaleral

bet)wrrn

(calculat,ed flow of anterior

measurements

from the chambrr

dilution fluid

the accxumulation of intracamrrally a significaant incarease in uveosc~leral

P(+F,, of t’he

of multiple m&flow.

tliff’ermc~r

isopropylest,er ocular

t.opical tot,al

and doses

outflow,

(I E) periocular

ant1

infused outflon

c~ontirmc~d

this

tissues.

of PGP,,

on direct

and

facility

t,otal

of into

quantit.at.ivr in c~~riomolgus

monkeys.

2. Materials ( ‘hwuicals

and

and

Methods

drugs

PGF,,-IE was obtained from Pharmacia Ophthalmics AB (Uppsala, Sweden) as U-0535% aqueous solution also containing 0.5 % polysorbate 80 and @9 % NaCl. This stock solution was further diluted with @9% XaCl to give a final concentration of @02 % PGF,;IE. Fluorescein isothiocyanate-conjugated dextran (FITC-Dex ; molecular weight 46600) was also supplied by Pharmacia Ophthalmics. lZ5T and 1311 were obtained, respectively. from NEN Products (Boston, MA) and ICN Biomedicals, Inc. (Irvine, CA). Albumin was isolated from cynomolgus monkey (~ucacafascicLLlaris) plasma on Blue Sepharose CL-6B (Pharmacia Fine Chemicals, Piscataway, PU‘J) according to Travis, Bowen. Tewksbury, Johnson and Pannell (1976). lyophilized, and stored dessicated at 4°C until use. Two mg of the purified albumin was enzymatically iodinated with 0.5 mC’i “‘1 or 1311 using Enzymobeads (BioRad, Richmond, CA). Immediately before the experiment an aliquot was separated from free iodine on a Sephadex G-25 column (Pharmacia Fine Chemicals). PQF,,

treatment

protocol

Eleven young adult female eynomolgus monkeys (Macaca fasciculuriu) weighing 2.7 to 3.6 kg were studied. None had received any drug topically or systemically for over 2 months : most had never received any drugs and none had previously undergone ocular perfusion or surgery. The fully conscious animals were trained to enter a net in which they were manually r&rained in the supine position by a gloved assistant; they were never chaired or treatments. Two 5~1 drops of PGF,,-IE (total dose = 2 pg). anesthetized for PGF,, separated by 30 set, were administered t,opically to the rentral cornea of one eye (six right. ryes. five left eyes in alternating sequence) with a micropipette. If necessary, the rye was then blotted with tissue to remove any excess moisture and the animal was returned to its cagr. Dosing was at 0730 hr and 1500 hr for 3 days. and at N 0900 hr on day 4. The c*ontralatrral eye rrceived neither drug nor vehicle. Total

facility,

uwo~&wzl

outflow.

und

trahwular

outjlou~

101’ was checked with a miniiied Goldmann applanation tonometer (Kaufman and Davis, 1980). under intramuscular (i.m.) ketamine anesthesia (10 mg kg-‘), on the fourth treatment day at 4 hr intervals between hrs 2 and 4 after the morning (seventh) PGF,, dose. The ocular hypotensive response of the monkey thus verified, the perfusion experiment was conducted on the following day after the morning (ninth) dose. The monkey was anesthesized initially with i.m. ketamine. the eyes were examined by slit-lamp biomicroscopy (PLK, who did not know which eye was drug-treated). and the IOPs were determined tonometrically. Deep anaesthesia was then induced wit.h i.m. pent,obarbital Na (3&35 mg kg-‘). Each eye was cannulated with three unbranched needles connected, via polyethylene tubing, stopcocks, and micro-T-pieces, t,o a continuously weighed fluid reservoir, a pressure transducer, a small pump for mixing the contents of the anterior chamber, and motorized c,oupled push-pull gas tight infusion-wit,hdrawal syringes (Hamilton Co., Reno. NV) (Bill, 1977: Sperber and Bill, 1984). The perfusand was B&&y’s solution (B&r&y, 1964). Following a 5-10 min stabilization period at spontaneous IOP, total facility was determined by two-level con&ant pressure perfusion for 3&45 min at 15 and 24-4 mmHg (B&r&y, 1964).

I’GF,,

INCREASES

UVEOSCLERAL

OIITFLOW

391

The anterior chamber contents were then exchanged over approximately 10 min with - 2 ml of perfusand containing 2 x lo-* M FITC-Dex and 5 x lo6 cpm ml-’ of either [1”“1] (right eyes)- or [la111 (left eyes)-albumin (the final albumin concentration was adjusted to 91% with unlabeled albumin). The infusion-withdrawal rate was then reduced to 2-3 pl mini by slowing the motorized syringes, and continued for another 50 or 80 min during which time the chamber contents were continuously mixed at N 60 ~1 min-’ via the external pump and mixing circuit. The anterior chamber was then rinsed over - 10 min with 4 ml of perfusand without tracer, after which IOP was elevated to - 30 mmHg for 3-5 min to wash tracer from the trabecular outflow pathway. Blood samples were collected from a saphenous vein prior to the initial cannulation and after the final unlabeled rinse. The animal was then killed by i.v. methohexital overdose. The eyes were enucleatrd within 2 min of death, rinsed with saline, and placed on ice until dissection. Tissue remaining in the socket was removed, rinsed with saline. and designated as distal periocular tissue. Periocular tissue adhering to the globe during enuclration was removed and designated as proximal periocular tissue. The globe was stabilized in a cloth and plastic well, and the cornea was removed. The anterior chamber was rinsed with saline, and the globe then placed in a plastic dish where it was cut around the equator and dissected into limbal corneosclera (a l-2 mm wide ring extending posteriorly from the corneoscleral junction and including the trabecular meshwork and Schlemm’s canal). anterior sclera (a 5 mm wide ring extending from the posterior edge of the limbal corneosclera to the equator), posterior sclera (the remainder of the scleral envelope), iris, anterior uvea (ciliary body, choroid and retinal pigmented epithelium anterior to the equator), posterior uvea (remainder of the choroid and retinal pigmented epithelium), neural retina and fluid (vitreous and any fluid released during dissection of the globe). In some animals a central vitreous sample was obtained prior to dissection of the ocular tissues to distinguish between physiologic diffusional and iatrogenic mechanical entry of tracer into the vitreous. The entire dissection process was completed within 15 min after death: the order (treated vs. control) varied from animal to animal. The tissues and fluids, as well as aliquots of the infusion solution (sampled from the infusion tubing at the end of the experiment) were counted for lz51 or isi1 in a Packard Instruments (‘-5330 Auto-Gamma Spectrometer. Each sample was then homogenized (or vortexed in the case of fluids and solutions) in 10% ZnSO,.GH,O, and the protein precipitated at neutral pH with an equal volume of 0.5 N NaOH (Somogyi. 1930). Experiments were divided into three categories. For most animals facility determination encompassed minutes [- 180 to - 220 (n = 8)] following the ninth PG dose. Reservoirs connected to the eyes were then closed and IOP was allowed bo return to its spontaneous level during the next - 10 min. Infusion of labeled perfusand and determination of trabeeular and uveoscleral outflow at the spontaneous IOP (n = 6) then encompassed minutes - 235 to - 325, the first 10 min of which were occupied by anterior chamber exchange. Since uveoscleral outflow is reduced at very low IOP levels (Bill, 1967), two additional animals were run with the same time course while IOP in both eyes was maintained at 17-18 mmHg via an external reservoir filled with labeled solution. It also seemed plausible that the increase in uveoscleral outflow could peak early, so that measurements later in the period of maximum hypotony might not reflect the maximum effect (Kilsson et al., 1987). Therefore. in another animal the experiment was conducted sooner after the ninth PG dose (facility, 855120 min; labeled infusion, 135195 min), with IOP during the labeled infusion maint’ained at 18 mmHg in both eyes. Cveoscleral outflow (F,) was considered to be the volume (VU) of labeled anterior chamber fluid required to have deposited the amount of tracer recovered from the ocular and periocular tissues, divided by the duration (T) of the labeled infusion @ill. 1966). For each tissue and fluid compartment, quantity of tracer in tissue or fluid (cpm or ng) c; = concentration of tracer in perfusand (cpm or ng ~1~~). and

392

13.T. (:.-\UEL’I’

ASI)

1’. 1,. KAl’FRlAN

Trabecular outflow (E:,,,) was considered to be the volume (l;,,,) of labeled ant’erior chamber fluid required to have deposited the amount of [izsT]- or [‘311]-albumin found in the general circulation divided by the duration (T) of the labeled infusion (Bill. 1966). Tn several animals, the blood-equivalent. albumin spa.ce (REAS) was det.erminrd. and averagrd - 7.55% of the body weight at 90 min, and - 7.34% at 60 min (B. True Gabelt and I’. 1,. Kaufman, unpubl. res.): these values were assumed valid for all experiments of the corresponding duration. Then, BEAS x specific activity of blood Ll = specific activity of perfusand

For experiments at spontaneous IOP, with no net infusion of fluid from an external system, the decline in anterior chamber tracer concentration due to newly formed nonlabeled aqueous humor was calculated to be < 10 % by the end of the infusion period, given the N 600 pl volume of the mixing circuit, the - 100 yl volume of the anterior chamber, and an aqueous formation rate of - 1~1 min-r (Erickson-Lamy, Kaufman, McDermott and France, 1984). For experiments at an externally stabilized IOP of 1618 mmHg, with a net infusion of labeled fluid from the reservoir, the decline would have been still lower. This decline was neglected in the calculations of F, and F,,,,. One additional animal underwent the PG treatment and perfusion protocol, but both eyes received [*251]-albumin, the labeled infusion encompassed minutes 135-195 after the ninth PG dose, and IOP was maintained at 16 mmHg. Following enucleation the eyes were frozen by immersion in liquid nitrogen-cooled hexane. Cryostat sections, 30 and 50 pm thick. were then cut, freeze dried, and exposed to X-ray film (Kodak XAR 5) for 48-72 hr. Radioactivity in the sections was compared quantitatively using a Zeinek Soft Laser Scanning Densitometer (Biomed Instruments, Inc., Fullerton, CA).

3. Results IOP The 2yg dose of PGF,,-IE was submaximal (Crawford, Kaufnlan and Majors, 1989), but consistently produced a dramatic and stable IOP decrease averaging N 6 mmHg between hours 2 and 6 after the seventh dose on treatment day 4 (Fig. 1). All animals slated to undergo uveoscleral outflow determinations exhibited a substantial IOP reduction in their PG-treated eye compared to the contralateral control under ketamine anesthesia 3 hr after the seventh PG dose on treatment day 4; the difference was 7*4f@9 mmHg (mean ~s.E.M. ; n = 1 l), representing a 63+8% IOP reduction (P < CMMl) (Table I). On treatment day 5, under pentobarbital anesthesia at an average of - 2 hr 45 min after the ninth PG dose, the manometrically determined difference prior to the facility determination was 3.6&@7 mmHg (n = lo), representing a 31 f6% IOP reduction (P < 0901). Following the facility determination and reequilibration to the spontaneous IOP, at an average of - 3 hr 35 min after the ninth PG dose, the manometrically determined difference in IOP was 29*@7 mmHg (n = lo), representing a 32+6% IOP reduction (P < O-001). No anterior chamber cells or flare were observed biomicroscopically in any eye of any animal, but the pupil in the treated eye was invariably 05-1.0 mm smaller than in the control. Uveoscleral outjlow Considering all the experiments in the three time-IOP groups together, uveoscleral outtlow in the PG-treated eyes was twice as high as in the controls when measured with albumin, and 2.3 times as high as control using dextran [Table II(d)]. The effect

-I

5

3 Time

(hr)

FIQ. 1. Intraocular pressure (IOP) measured tonomrtricall~ in six cynomolgus monkeys undrr k&amine anesthesia at various times following the seventh urnlateral 1 jig P(:F,,-IK dose givrn at 0900 hr on treatment day 4. Data in upper panel are mean &S.E.M. in treated (-~ l l ) and control (--O-O~) eyes. Data in lower panel are mean +s.E.M. IOI’ difference between treated and control eyes. Significant difference between treated and control eyes hy the two-tailed two samplr (uppv~ panel) or paired (lower panel) t-test : a F CCWO5, ‘P < 0.01.

was highly statistically significant for both tracers. The actual uveoscleral flow values determined with the two tracers averaged within 12 f 16°K of one another in the treated eyes and within 0 f 22 O/o in the control eyes. There appeared to be quantitative differences between the three time-IOP groupings. When measurements were conducted at spontaneous IOP (treated, 5%f@9 mmHg ; control, 98+ 1.1 mmHg) 235-325 min after the ninth PC: dose, to 20 (dextran) times uveoscleral outflow in the treated eyes was N 1.7 (albumin) higher than in the controls [Table II(a)]. For measurements conducted in a similar time window but at physiologic IOP (17718 mmHg in both eyes), the treated/control ratio was greater, averaging w 24 (albumin) to 2% (dextran) [Table II(b)]. In the single experiment conducted at physiologic IOP (18 mmHg) but in an earlier time window, the effect was greater than in any other animal, with a treated/control ratio of N 3.2 for albumin and 3.5 for dextran [Table II(c)]. No significant correlation between IOP reduction and uveoscleral outflow increase was apparent. One animal received the full treatment course. but died unexpectedly N 90 min

Day 4 PGF,, *n = 10 n= 11

4.3 + 1.2 46_+ 1.2

Day

5 I’refarility

control

PGF,,

12~0+04 12.0f05

74 *o-9

I)ay

Control 1*9+

I.0

5 Postfavi1it.y

PGF,,

(‘ontrol

6I&OT

9.0 * 04

Day 4: Spontaneous IOP (mmHg, mean &s.E.M. for ?z eyes) measured tonometrically anesthesia 180 min after the morning (seventh) PGF,, dose in all animals. Day Spontaneous IDP measured manometrically under pentobarbital anesthesia 163 + 83-195 min) after the morning (ninth) PGF,, dose. Day 5 Postfacility: Spontaneous manometrically under pentobarbital anesthesia after facility determinations 214 k 132-261 min) after the morning (ninth) PGF,, dose. * One animal, which died before anterior chamber cannulation on day 5. excluded.

TABLE

Uveoscleral

II

outflow on day ii of twice daily Treated

under ketamine 5 Prefacility : 10 min (range : IOP measured 14 min (range :

unilateral

Control

treatment

with

PGF,,

Treated/Control

Albumin Dextran

(a) Spontaneous 078+0.12([1 089+0.13~

IOP; y 235325 046 + 0.03 048rtO.11

Albumin Dextran

(b) IOP = 17-18 mmHg; m 24&335 min (n = 2) 1.45+0.01§ 962+0.11 2.41 f042 1.42*015r 0.52 + 0.07 2.75fO.OXt

Albumin Dextran

(c) IOP = 17-18 mmHg: 203 0.63 1.74 0.49

Albumin Dextran

(d) All IOP-time groups combined (n = 9) 1~07~0~17~ 052 ko.04 290,@24$ l~lO~O~i4~ 0.49 * 0.07 2.33 2 0.24 11

w 135-195

min (n = 6) 1.66 * w20t 1.99*0.221:

min (II = 1) 321 3.53

Data are mean ~s.E.M. uveoscleral outflow (~1 x mini) for w animals, each contributing one treated and one control eye, following the ninth unilateral dose of PGF,, on day 5; min indicates time window following PGF,, encompassed by the measurement. Significantly different from 1.0 by the two-tailed paired t-test : tP < 605 : $ P < 001: 11P < 0.001. Significantly different from contralateral controls by the two-tailed two-sample t-test: T P < WO.5: $P < 901.

after the final dose. The perfusion experiment was conducted nonetheless, with the facility and uveoscleral outflow measurements encompassing - minutes I%&145 and 155-205, respectively, after PG (IOP held at 17 mmHg in both eyes for the latter, and no anesthetic agent). Although the actual values for uveoscleral outflow in both eyes were well above the range seen in the living animals (treated, - 45 ~1 min-’ ; control, - 2.1 ,~l mm’), the treated : control ratio was comparable. The volume of anterior chamber fluid passing into the various parts of the eye was compared in treated and control eyes for all time-IOP categories combined (Fig. Z),

AU

AS

L

PU

PS

R

F

P Tissue

I

AU

AS

L

PU

PS

R

F

P

Fro. 2. Distribution of tracer in the ocular and periowlar tissues of living P(:F,,-trmtrtl and wntrol ryvs prrfused with of anterior chamber flnid wcorer~d albumin and dextran. Data for each tissue are expressed as moan &s.R.M. ,ul mini’ I = iris. AU = anterior uvea, AS = anterior sclera, L = limbus. PL’ = posterior uvra. PS = posterior sclera. K = retina. F = fluids, P = periocular tissue. Symbols, top TOW. both panels: significant difference between results obtained with albumin and dextran for a given tissue in treated or conkol eyes by the two-tailed paired t-test (ditferencr vs. (Ml). Symbols. middle and bottom rows. cont,rol panel : significant, difference between treated and control eyes for albumin (middle row) or dexkan (hottom row) by the two-tailed paired l-test, (difkrruw vs. WI).

I

a P<0-05 b P-CO-01 c P
Treated

39ti

13.‘r.(:A1~EI,‘r

Ah-l)

I’. I,. Ki\lTFMAS

since the distribution did not appear to diff’er in the various groups. The majority of anterior chamber fluid recovered in b0t.h treated and control eyes was found in the anterior uvea, anterior sclera, posterior uvea, posterior sclera and periocular tissue. Since the anterior rhamber had been so copiously rinsed with fluid not containing tracer and since tracer in the central vitreous samples obtained prior to tissue dissection was < 1 “A, while tracer recovered from the fluids after dissecation was < 10 % of the aggregate total for all tissues and fluids. tracer found in the fluids had most probably escaped from the anterior and posterior uvea during dissection rather than entering physiologically. Additionally. t,he retina was probably contaminated with posterior uvea. In living animals, both eyes showed a preferential localization of dextran in the anterior sclera, posterior sclera and periocular tissue as compared to albumin, which was found preferentially in the anterior and posterior uvea. With the exception of albumin in the iris and limbus, both tracers were present in higher concentrations in all other tissues of the treated than the control eyes. In the dead animal, the periocular tissue in t,he treated and control eye contained. respectively, 64% and 35 % of the anterior chamber fluid albumin and 74% and 35% of the anterior chamber fluid dextran; in the living animals the corresponding figures in treated and control eyes, respectively, were 33 % and 13 % for albumin, and 40 “/o and 24 % for dextran. In living animals, proximal periocular tissue in treated and control eyes contained respectively 24 and 7 times more albumin than did distal periocular tissue; for dextran, the corresponding ratios were 17 and 6. In the dead animal, both tracers were distributed equally between the proximal and distal periocular tissues of the treated eye, but in the control eye five times more albumin and seven times more dextran were present in proximal than distal periocular tissue. Autoradiography of sections through the globes of one animal revealed tracer farther posteriorly in the treated than in the control eye (Fig. 3). By quantitative densitometry, the average treated/control ratio for 12 paired sections was 1.33kO.12. significantly > i.0 at P = 0017.

FIG. monkey solution extension

3. Autorsdiograms of transpupillary K234 following - 60 min of - 135-195 min after the ninth of tracer in the PGF,,-treated

anterior unilateral eye.

sections from control (top) chamber exchange and PGF,, dose on treatment

and treated (bottom) eye infusion with [‘251]-albumin day 5. Note more posterior

of

PGF,,

INCREASES

UVEOSCLERAL TABLE

Trabecular

and

total outjlow

on day

daily

unilateral

Control

(a) Spontaneous 0.23 * 0.06 1~05*012 023+910// 0.77+907t

397

III

5 of twice

Treated

OI’TFLOM

treatment

with PCX,,

Treated/Control

IOP ; - 235-325 0.97 kO.25 1.43+924 063+008t 037 +omt

min (n = 5) 0.23 k 005 I/ 0~80+0~12 033 + 0.08 /I 2.32 -+ (bwt *

(b) IOP = 17-18 mmHg; - 24lb-335 min (II = 2) 1.21 kO.57 1.83kO.11 968 + 0.35 1.11*033 2.65 + 058 2.45 k 0.22 057+rr15 043kO.12 0.75 * 0.02 2.26 5 0.30 057fO.l” 0.25 f O-02 t

F l&J

(e) IOP 489

692

4.98

4dFtot FuIF,,,

971 0.29

0.87 0.13

Ftot

= 17-18

mmHg: 4.34

- 135-196

min (n = 1) 1.13 1.39 WXl ?31

Data

encompass the times indicated following the ninth unilateral dose PGF,, on day 5 and are mean pl mini for n animals, each contributing one treated and one control eye, with albumin as tracer; perfusions were run at the indicated IOP. Data for one monkey in group (a) were excluded because of very low trabecular and t&al outflow in both eyes, presumably consequent to the very low IOP bilaterally at which the perfusions were run. Significantly different from 1.0 by the two-tailed paired L-test: tP < 905: $P 4 901 : (/P < (POOl. +s.E.M.

Trabecular

outflow,

total

outjlow,

and

total

outjlow

facility

Total outflow was considered to be the sum of trabecular and uveoscleral outflow. In the experiments conducted at spontaneous IOP, trabecular outflow was reduced eyes compared to the contralateral controls [Table by - 75% in the PGF,,-treated III(a), Ftrab], so that it constituted only N %of the total outflow as opposed to - Q in the controls [Table III(a), F,,,,/F,,]. Total outflow averaged - 20% less in the treated than in the control eyes, but the difference was not statistically significant [Table III(a), B’,,]. The fraction of total outflow constituted by uveoscleral outflow was higher in treated vs. control eyes by virtually the same proportion in all three time-IOP groups [Table III(a- c) ; FJF,,,. Treated/Control = 2.311-0.19. P < 0.001. fl = 81. Total facility after the ninth PGF,, dose averaged 49 + 22 % higher in the treated than in the control eyes of the nine living animals (957 + 0.08 vs. 041 k 605 ,A min-’ mmHg-‘), a difference of marginal statistical significance (910 > P > @05). In the dead animal, total facility after PGF,, was twice as high in t.he treated as in the control eye (1.28 vs. 0.63 ~1 min-’ mmHg-‘). 4. Discussion This study directly demonstrates the increase in uveoscleral outflow responsible for the hypotensive effect of PGF,,. Our results are remarkably similar to those of Nilsson et al. (1987) who, using the isotope dilution technique and a single 1 ,ug topical dose

3!N

I(. ‘I’ ( : .4 I: Ii: l,‘l‘ .\ s I) I’. I.. ii A 1 11; AI .I s

Of P(iF,,-I E. Ol>t~~it)?tl I1VPOS(‘if?I~i\l tiO\V \-&lllPS Of’O’!)X ,I,1 ttli?l. ’ for t rrv\tfd mtl ~,%I ,//I minW’ for control eyes over a -t hr period. This compares to 0.78 /tl mini ’ and (1.X ,MI min-’ respec%ively for the I-a.lbumin values in our studies at spontaneous IOP af’trt multiple 2 //g topical tlosc5 ofP(+F,,-IK ITat+ II(a) /. 1 II our hilnfls. ttw 1 //~‘hW. f’vt‘tt with multiple treatments, produc*ed a more variable IOP fall of shorter duration. I’vroscleral outflow consistently (comprised t\vi(*ta as high a prol~ortion of total outflow in the PC+-treat,ed as in the control eyes. despite the varying IOPs. Tht magnitude of the PGF,, -induced inrrease in uveoscleral outflow was comparable to t,hat caused by ayclodialysis (Suguro. Toris and Pederson. 1985) or BSA~induc*etl iridocyclitis (Toris and Pederson. 1987) in cynomolgi. W’hen IOF’ was maintained at. 17-18 mmHg, uveoscleral outflow in PG-treat,ed eyes was nearly double that at spont,aneous IOP. while uveoscleral outflow was much less IOF~dependent in the c.ont,rol eyes. PGF,, thus may alter a basic physiologic characteristic of t,his pathway. namely its relat,ive IOP-independence (Bill. 1966. 1967). However, ot,her rxplanabions are plausible and additional experiments will be required. Interestingly. uveoscleral out,flow becomes more pressure-sensit,ive following cyclodialysis (Toris and Pederson. 1985) and experimental iridocyclitis (Toris and Pederson, 1987). If the tracer moved through t’he uveoscleral pat’hway at higher velocity in the PGtreated than in the control eyes, the underestimation of uveoscleral outflow and overestimation of trabecular outflow consequent t,o unrecognized loss of tracer from the periocular tissues t,o the general circaulation would be greater in t,he PG-treated e-yes. The percent,age of t.he total tracer recovered which was localized t.o the posterior sclera and periocular tissues was higher in the PC&treated than in the control eyes, and t’he difference was most st’riking in t,he shorb-duration experiment. Following a single 1 ,~g dose of P(:F,,-IE in cynomolgi. uveosc+ral outflow. calculated indirectly as the difference between measured values for total aqueous flow t,hrough the anterior chamber and flow of anterior chamber fluid int,o the general circulation. wa5,T maximally increased l-2 hr after drug administration. but t.hen declined t.oward control values during hours 3-4. the time of maximum hypotony (Nilsson, Sperber and Bill. 1989). The actual values reported. however. were averages for the entire 4 hr (Nilsson et al., 1987). Therefore. it’ seems quit,e plausible that the PG effect on uveoscleral outflow was indeed underestimated because of this phenomenon in both Nilsson’s experiments and ours. Additionally, the uveoscleral flow mrasurement,s in some of our animals may have been done after the peak effect had passed. This would be consistent with Nilsson’s data as described above. and also with the higher value in our one animal perfused earlier after drug administration. The experiments reported here demonstrate directly the profound effect of PGF,, on uveoscleral outflow. and identify some t,echnicaal factors which may have caused an underest’imate of the true effect. The protocol variations and relatively small number of experiment.s precluded our isolating the influence of each individual factor. This can be pursued by indirect measurement techniques allowing survival of the animal (Bill. 1970; Sperber and Bill. 1984: Silsson et al.. 1987: Xilsson et al.. 1989). However. since the drug is being considered for anti-glaucoma therapy in humans ((‘amras et al., 1987; C’amras et al., 1988: Bite. 1987) and indeed has now entered clinical trials in the United States. Sweden. Italy and (‘hina. it seemed important t,o directly document. the principal mode of action. Bill (1966) reported that uveoscleral outflow in cynomolgi was t,wice as high after death as during life. In our single animal that died at the start of the experiment, uveoscleral outflow in the control eve was - 3-4 t,imes higher than in control eyes of

PCF,,

INCREASES

UVEOSCLERAL

OL~TFLOW

399

living animals perfused under comparable conditions. However, the value in the PGtreated eye was - 2-25 times higher still, or - 3 times higher than in living PGtreated eyes. If we assume that the post-mortem increase is due to total relaxation of the ciliary muscle, then t,he persistence of essentially the entire PG effect after death indicates a mechanism of drug action beyond simply muscle relaxation. One possible such mechanism is the quantitative or qualitative alteration of the extracellular material between the muscle bundles (Inomata, Bill and Smelser, 1972). The fact that, the full ocular hypotensive effect of PGF,, in cynomolgi develops only after several days of topical treatment is consistent with such a biochemical/metabolir explanation as is the observation of dissolution or washout of the intermuscular connective tissue matrix in such treated eyes (Tamm and Liitjen-Drecoll, 1988: Liitjen-Drecoll and Tamm, 1988). R#aviola (1983) observed that’ intravenously infused horseradish peroxidase leaked from the vessels putatively draining Schlemm’s canal in rhesus monkeys. On t,his basis, Suguro et al. (1985) proposed excluding tracer in the periocular tissues from Raviola’s electron microscopic calculations of uveoscleral outflow. However, t,echnique would detect even miniscule amount,s of t’racer non-quantitatively. It seems most unlikely that, vessels as large as Schlemm’s canal. collector channels, and episcleral veins would leak physiologically meaningful (in terms of accompanying water transfer) amounts of molecules as large as albumin and dextran, especially in the absence of biomicroscopic evidence (i.e. anterior chamber cells and flare) of an enormous alteration in permeability of the anterior segment vasculature in general. Additionally, the substantial amount of tracer found in the proximal periocular tissues of the animal perfused after death seems especially unlikely to have come from vessels draining Schlemm’s canal. If periocular tracer had come from Schlemm’s canal itself, the limbal corneosclera (containing the meshwork and canal) and the anterior sclera should have contained more tracer than the posterior sclera; in fact, precisely the opposite order was found in the treated and control eyes of the living animals and in the treated eye of the dead one. In the control eye of the dead animal, the anterior and posterior sclera contained - equal amounts of tracer. In any event, even if periocular tissue was eliminated from the calculation of uveoscleral outflow, the treated/control ratio in the living eyes (n = 9, all time-IOP groups combined) was still significantly > 1.0 for both albumin (1.50 f 921 ; P < 0.05) and dextran (1.72 ) 0.18, P < 0.004). Furthermore, since the amount and proportion of tracer in the periocular t,issue of the t’reated eye was so large, the posteriorly draining aqueous must have exited the eye trans.sclerally. rather than being reabsorbed by uveal blood vessels consequent to a PG-induced alteration in uveal vascular permeability. Had the latter occurred, one would expect to find less periocular tracer and more tracer in the uvea and nclera. The relatively minor distributional differences of the two t,racers among the ocular tissues probably relates to differences in configuration and electrical charge of the molecules. with albumin being more globular and negatively charged, and dextran being more linear and positively charged. Nonetheless, uveoseleral flow values calculated using FITC-dextran and radioiodinated albumin were very similar, suggesting excellent overall physiologic comparability of the tracers and validity of the values. Total facility tended to be higher and total aqueous flow lower in the PG-treated than in the control eyes, but as in our earlier studies focusing on those parameters (Kaufman, 1986; Crawford et al., 1987), the differences were neither statistically

100

IS. ‘I’. c: A Ii E: f,T

Al s 1) P. I,. Ii,4

1’ FM AN

significant nor sufficiently large to account for t,he iOP-lowering effect. The apparent reduct,ion in calculated aqueous flow might be consequent, t)o underestimation of uveoscleral outflow, as discussed above. Increased total facility could at least partly represent increased pseudofacility (Lee et al., 1984) or uveosrleral facility (Toris and Pederson, 1987). While definitive experiments t,o determine the effect of PGF,, on conventional outflow facility are obviously needed, it seems clear that this parameter is not the major attack point accounting for PGF,, ‘s ocular hypotensive effect. Rather, PGF,, largely redirects aqueous outflow from the trabecular t,o the uveoscleral rout,e, increasing the latter by threefold and perhaps more. Since the uveoscleral outflow route drains against an intraorbital pressure of - 0 mmHg as opposed to the episcleral venous pressure of - 10 mmHg faced by the trabular route (Kaufman, 1985), this could well account for the marked ocular hppotensive effect, and the profoundly low final IOP. ACKNOWLEDGMENTS This study was supported by National Eye Institute grant no. EY02698. Patrick A. Goeckner provided expert technical assistance. Pharmacia Ophthalmics, AB, Uppsala, Sweden, graciously donated PGF,,-IE and FTTC-Dex. REFERENCES Alm,

A. and Villumsen. J. (1986). Intraocular pressure and ocular side effects after prostaglandin F,, eye drops. A single dose-response study in humans. Proc. Znt. Sot. Eye Res. 4, 14, and satellite International Symposium on Prostaglandins and Related Compounds in Ophthalmology, Juntendo University, Tokyo, 1986. Barany, E. H. (1964). Simultaneous measurements of changing intraocular pressure and outflow facility in the vervet monkey by constant pressure infusion. Invest. Ophthalmol. 3, 135-43. Bill, A. (1966). Conventional and uveo-scleral drainage of aqueous humor in the cynomolgus monkey (Macaca irus) at normal and high intraocular pressures. Exp. Eye Res. 5,45-54. Bill, A. (1967). Further studies on the influence of the intraocular pressure on aqueous humor dynamics in cynomolgus monkeys. Invest. Ophthalmol. 6, 364-72. Bill, A. (1970). The effect of changes in arterial blood pressure on the rate of aqueous humour formation in a primate (Cercopithecus ethiops). Ophthalmic. Res. 1, 193-200. Bill, A. (1977). Basic physiology of the drainage of aqueous humor. In The Ocular and Cerebrospinal Fluids. Fogarty International Center Symposium. (Eds. Bito, L. Z., Davson, H. and Fenstermacher, J. D.). Exp. Eye Res. 25 (Suppl.), 291-304. Bito, L. Z. (1987). Prostaglandins. Old concepts and new perspectives. Arch. OphthaEmoZ. 105, 10369. Bito, L. Z.. Draga, A., Blanco, J. and Camras, C. B. (1983a). Long-term maintenance of reduced intraocular pressure by daily or twice daily topical application ofprostaglandins to cat or rhesus monkey eyes. Invest. Ophthalmol. Vis. Sci. 24, 312-9. Bito, L. Z., Srinivasan, B. D., Baroody, R. A. and Schubert, H. (198313). Non-invasive observations on eyes of cats after long-term maintenance of reduced intraocular pressure by topical application of PGE,. Invest. Ophthulmol. Vis. Sci. 24, 37G-80. Camras, C. B. and Bito, L. Z. (1981). Reduction of intraocular pressure in normal and glaucomatous primate (Aotus trivirgatus) eyes by topically applied prostaglandin F,,. Curr. Eye Res. 1, 205-g. Camras, C. B., Bito, L. Z. and Eakins, K. E. (1977). Reduction of intraocular pressure by prostaglandins applied topically to the eyes of conscious rabbits. Invest. Ophthalmol. Vis. Sci. 16. 1125-34. Camras, C. B., Friedman, A. H., Rodrigues, M. M., Tripat.hi, B. J., Tripathi. R. C. and Podos, S. M. (1988). Multiple dosing of prostaglandin F,, or epinephrine on cynomolgus monkey eyes. III. Histopathology. Invest. Ophthalmol. Vis. Sci. 29, 1428-36.

PGF,,

INCREASES

UVEOSCLERAL

OrTFLOW

40 I

Camras, C. B., Podos, S. M., Rosenthal, J. S., Lee, P.-Y. and Severin, C. H. (1987). Multiple dosing of prostaglandin F,, or epinephrine on cynomolgus eyes. I. Aqueous humor dynamics. Invest. Ophthalmol. Vis. Sci. 28, 463-9. Crawford, K. and Kaufman, P. L. (1987). Pilocarpine antagonizes PGF,-induced ocular hypotension in monkeys. Evidence for enhancement of uveoscleral outflow by PGF,,. Arch. Ophthdmd. 105, 1112-6. Crawford, K., Kaufman, P. L. and Gabelt, B. T. (1987). Effects of topical PGF,, on aqueous humor dynamics in cynomolgus monkeys. &rr. &ye Res. 6, 1035-44. (‘rawford, K.. Kaufman. P. I,. and Majors. I,. .l. (1989). Ijose-response relationships between PGF,,-IE and IOP refraction and pupil diamrtrr in cynomolgus monkeys. ITII’PS~. Ophfhalmol. JTis. Sci. 30 (ART’0 Suppl.). 24. Erickson-Lamy, K. A., Kaufman, P. L., McDermott. M. L. and France, N. K. (1984). Comparative anesthetic effects on aqueous humor dynamics in the cynomolgus monkey. Arch. Ophthalmol. 102, 1815-20. won Graefe-s Giuffri, G. (1985). The effects of prostaglandin F,, in the human eye. Albrecht Arch.

C&n.

Exp.

Ophthalmol.

222,

13941.

Inomata, H., Bill, A. and Smelser. G. K. (1972). Unconventional routes of aqueous humor outflow in cynomolgus monkey (Macaca irus). Am. J. Ophthalmol. 73, 893-907. Kaufman, P. L. (1985). Aqueous humor dynamics. In Clinical Ophth’almology. (Ed. Duane. T. D.). Vol. 3. Ch. 45. Pp. l-24. Harper & Row: Philadelphia. Kaufman, P. L. (1986). Effects of intracamerally infused prostaglandins on outflow facility in cynomolgus monkey eyes with intact or retrodisplaced ciliary muscle. Exp. Eye Res. 43, 81927. Kaufman P. L. and Davis. G. E. (1980). “Minified“ Goldmann applanating prism for tonometry in monkeys and humans. Arch. Ophthalmol. 98, 542-6. Lee, P., Podos, 8. M.. Howard-Williams, J. R. and Severin, C. H. (1985). Pharmacological testing in the laser-induced monkey glaucoma model. Curr. Eye Res. 4, 775-81. Lee, I’. Y., Podos, S. M., Serle, J. B. and Camras. C. B. (1987). Intraocular pressure effects of multiple doses of drugs applied t’o glaucomatous monkey eyes. Arch. Opthalmol. 105. 24952. Lee, P. Y., Podos, S. M. and Severin. C. (1984). Effect of prostaglandin F,, on aqueous humor dynamics of rabbit, cat, and monkey. Znvest. Ophthalmol. Vis. Sci. 25, 1087-93. Lee, P. Y., Shao, H., Xu, L. and Qu, C. K. (1988). The effect of prostaglandin F,, on intraocular pressure in normotensive human subjects. Znvest. Ophthalmol. Vis. Sci. 29, 1474-7. Liitjen-Drecoll, E. and Tamm, E. (1988). Morphological study of the anterior segment of cynomolgus monkey eyes following treatment with prostaglandin F,,. Exp. Eye Res. 47. 761-9. Nilsson. 8. F. E., Sperber, G. 0. and Bill. A. (1989). The effect of prostaglandin F,,-1-isopropylester (PGF,,-TE) on uveoscleral outflow. Tn The Ocular Ejfects qf Prostaglandins and Other Eicosanoids. (Eds Rite. L. Z. and Stjernschantz. .J.). Pp. 729-36. Alan R. Liss: Xew York (in press). P;ilsson, S. F. E.. Stjernschantz. J. and Bill, A. (1987). PGF,, increases uveoscleral outflow. Zrwest. Ophthalmol. Vis. Sci. 28 (ARVO Suppl.). 284. Raviola, G. (1983). Conjunctival and episcleral blood vessels are permeable to horseradish peroxidase. Invest. Ophthalmol. Vis. Sci. 24, 725-36. Somogyi, M. (1930). A method for the preparation of blood filtrates for the determination of sugar. J. Biol. Chem. 86, 65583. Sperber, G. 0. and Bill, A. (1984). A method for near-continuous determination of aqueous humor flow: effects of anaesthetics. temperature. and indomethacin. Exp. Eye RPS. 39, 43553. Stern, F. A. and Bito, L. Z. (1982). Comparison of the hypotensive and other ocular effects of prostaglandins E, and F,, on cat and rhesus monkey eyes. Znvest. Ophthalmol. Vis. Sri. 22, 588-98. Suguro, K., Toris, C. B. and Pederson, J. E. (1985). Uveoscleral outflow following cyclodialysis in the monkey eye using a fluorescent tracer. Invest. Opthalmol. Vis. Sri. 26. 8163.

40:!

I3.‘I’.~~;~HEI,‘l’

ASI)

I’. I,. I~.-\I‘P’nI.~s

Tamm. IX. and Liitjen-I~reroll. E. (1988). Morphological changes in the anterior segment ot cat eyes after treatment wit,h prost,aglandin F,, (PGF,,). Proc. lat. SW. E?ye Res. 5, 35. Toris. t’. B. and Pederson. !J. E. (1985). Effect, of intraocular pressure on uveoscleral outflovr following cyclodialy+is in the monkey rye. Invrst. Ophthnlmol. T’is. ,Vci. 26. 1745-!). Toris. C’. H. and I’ederson. .J. E. (1987). Ayueous humor dynamics in experimental iridocyclitis. Jnwst. Ophthdmol. I’is. Sri. 28. 477-81. Travis. ,I.. ISowen. .I.. Tewkshury. I).. .Johnson. I). ant1 I’annell. f2. (1976). Tsolation of albumin from whole human plasma and fract~ionation of albumin-deplet,ed plasma. Riochum. cJ. 157. 301. 6. Villumsen, .J. and Alm. A. (1987). The effect, of prost,aglandin F,, eye drops in open angle glaucoma. Invest. Ophthalmol. J’is. Sci. 28 (ARVO Suppl.), 378.