Studies on the Ocular Pharmacology of Prostaglandin D2

Home , Prostaglandin analogue, Prostaglandin D2

Studies on the Ocular Pharmacology

D. F. Woodward, 5. 3. Hawley, L. S. Williams, T. R. Ralston, C. E. Protzman, C. S. Spada, and A. L. Nieves

Prostaglandin D2 (PGD2) exerts a variety of biologic actions in the eye; these include ocular hypotension and inflammatory effects on the conjunctiva. The profile of activity of PGD2 in ocular tissues was compared to that of BW 245C, a selective agonist for the PGD2-sensitive (DP) receptor, and to that of the biologically active metabolites of PGD2, 9a,ll/?-prostaglandin F2 (9a,ll/?-PGF2) and prostaglandin J2 (PGJ2). PGD2 produced a dose-dependent decrease in intraocular pressure and in the conjunctiva it caused increased conjunctival microvascular permeability, eosinophil infiltration and goblet cell depletion. Although BW 245C was equipotent to PGD2 as an ocular hypotensive agent, it did not cause pathological effects in the conjunctiva. Thus, the ocular hypotensive effect of PGD2 may be separated from inflammatory effects on the conjunctiva by employing a selective DP-receptor agonist such as BW 245C. 9a,ll/?-PGF2 was a weak ocular hypotensive and did not cause conjunctival inflammation. PGJ2 produced no significant effect on intraocular pressure. PGJ2 did not elicit a microvascular permeability response in the conjunctiva, but was inflammatory in other respects and caused eosinophil infiltration and goblet cell depletion similar to PGD2. Thus, both the ocular hypotensive actions and the conjunctival pathology of PGD2 may be replicated individually by employing PGD2 analogues and metabolites. Invest Ophthalmol Vis Sci 31:138-146,1990

Prostaglandin D2 (PGD2) is a major cyclooxygen- mation. Recent advances in PGD2 biology, however, ase metabolite of arachidonic acid with a diverse have made characterization of ocular PGD2 pharma- spectrum of biologic activities. These include inhibi- cology feasible. These advances are: (1) the identifi- tion of platelet aggregation,' airway smooth muscle cation of a highly selective agonist (BW 245C) for the 23 34 1213 contraction, vasomotor activity, antineoplastic PGD2-sensitive receptor ; and (2) the elucidation effects,5 neuromodulation6 and inflammatory activi- of new spontaneous and enzymatic degradation 7 ties. The possibility that PGD2 may exert some of its pathways for PGD2 and studies on the pharmacology effects through a distinct PGD2-sensitive receptor has of these PGD2 derivatives, ie, 9a, 110-prostanglandin 1415 been suggested from comparison of rank order of po- F2 (9«,lli8-PGF2), prostaglandin J2 (PGJ2) and 12 13 16 17 tency for natural prostanoids in selected prepara- A PGJ2. - ' 89 tions and from radioligand binding studies in plate- Studies in a variety of PGD2-sensitive systems have 10 12 lets. " Many tissues, however, contain more than revealed that BW 245C, PGJ2 and A PGJ2 can dif- one prostanoid receptor subtype that may elicit iden- ferentially mimic PGD2 responses. BW 245C is a po- tical or opposing responses. The presence of a mixed tent and selective agonist in PGD2-sensitive systems prostanoid receptor population, together with the such as relaxation of human uterus,18 relaxation of typically modest degree of selectivity of natural pros- rabbit stomach strip13 and inhibition of platelet ag- tanoids for a particular receptor subtype, can make gregation.1213 Thus, BW 245C appears to be a useful rank ordering of natural prostanoid potency as an agent for describing DP-receptors according to the approach to pharmacological definition an indeter- current working classification.19 BW 245C does not minate exercise. This is likely to be particularly prob- produce PGD2-like contraction of the guinea pig tra- lematical in the eye for multifactorial responses, such chea and the dog cerebral artery, and does not cause 13 as changes in intraocular pressure (IOP) and inflam- the inhibition of cell growth associated with PGD2. The 11-ketoreductase metabolite of PGD2, 9«,11/8- PGF2, reproduces the tracheal myotropic actions of From Allergan, Inc., Department of Pharmacology, Irvine, Cali- PGD2, and is an inhibitor of platelet aggregation but, fornia. in contrast to PGD2, 9a, 1 l/3-PGF2 increases blood Submitted for publication: September 12, 1988; accepted May 15 20 pressure. ' The antiproliferative actions of PGD2 26, 1989. 12 Reprint requests: Dr. D. F. Woodward, Allergan, Inc., Depart- are replicated by PGJ2 and A PGJ2, which exhibit ment of Pharmacology (LS-OA), 2525 Dupont Drive, Irvine, CA reduced myotropic and antiaggregation actions rela- 131617 92715. tive to the parent prostanoid. Because of the

138

Downloaded from iovs.arvojournals.org on 09/24/2021 No. 1 PGD2 OCULAR PHARMACOLOGY / Woodward er ol 139

differential activity profile of these agents in PGD2- and the plasma was removed and diluted in saline. sensitive systems, the effects of PGD2, PGJ2, BW The absorbency of plasma and aqueous humor sam- 245C and 9a, 1 l/3-PGF2 were compared in the eye in ples was then determined at 620 nm using a spectro- an attempt to elucidate the pharmacological basis of photometer (Beckman DU7). By reference to a stan- the ocular hypotensive21 and inflammatory actions22 dard curve and the dilutions involved, plasma leak- ofPGD2. age could be calculated in terms of milliliters albumin per milliliter aqueous humor. These and other in- Materials and Methods flammatory effects were statistically analyzed by comparison of treated and control eyes according to Intraocular pressure (IOP) studies were conducted the student paired t-test. Responses in each treatment in New Zealand, white/Dutch Belted, cross-bred rab- group are numerically expressed as the difference be- bits of either sex that weighed between 1.5-2.5 kg. tween test and control values. These animals had not previously received topical Increases in conjunctival microvascular permeabil- drugs of any type. IOP was measured with a pneuma- ity and histological studies were performed in male tonometer (Digilab, Cambridge, MA) calibrated albino guinea pigs of the Hartley strain, weighing against the eyes of anesthetized rabbits by closed 350-450 g. This species was preferred for inflamma- stopcock manometry. The correlation coefficient tion studies since the guinea pig and the human con- 22 over the range 10-30 mm Hg was 0.98. The animals junctiva appear to respond similarly to PGD2. were acclimated to pneumatonometry by taking Conjunctival extravascular albumin accumulation unrecorded measurements before experimental de- and blood content were quantified by a previously terminations of IOP. Corneal anesthesia for tonome- described technique involving 51Cr-erythrocytes and 125 23 51 125 try was provided by topical application of one drop of I-bovine serum albumin. Na2 CrO4 and I-bo- 0.05% proparacaine (Allergan, Irvine, CA). vine serum albumin were purchased from New En- The effects of topically applied prostanoids on IOP gland Nuclear (Boston, MA). Radiolabelled erythro- were determined by applying 25 ^1 of drug solution to cytes and albumin were injected into the dorsal foot one eye and 25 /xl of vehicle to the contralateral eye as vein 15 min before prostanoid administration to the 9 a control. Prostaglandin D2, 9-deoxy-A prostaglan- eye. Topical ocular application volumes were 20 {A, din D2 (prostaglandin J2), 9a, 110-prostaglandin F2, contralateral eyes received an equal volume of vehi- (Cayman, Kalamazoo, MI) and 5-(6-carboxyhexyl)- cle as a control. The animals were sacrificed 15 min l-(3-cyclohexyl-hydroxypropyl) hydantoin (BW. after prostanoid administration by intracardiac injec- 245C, a gift from Burroughs Wellcome, Beckenham, tion of T-61 euthanasia solution (Hoechst, Somer- UK) were dissolved in 2% Na2CO3 and immediately ville, NJ). The eyes were then enucleated and the neutralized with HC1. All solutions were prepared bulbar conjunctiva was dissected free of the globe at fresh on a daily basis and were used within 15 min the limbus and trimmed free of extraocular muscle. after preparation. IOP was recorded at 30 min and Tissue samples from each individual animal were immediately before topical drug administration and counted in a 7 scintillation counter, together with a at 30, 60, 120, 180 and 240 min thereafter. Statistical 1.0 ml blood sample from the same animal. Tissue analyses employed the student paired t-test. and blood samples were finally oven-dried at The integrity of the rabbit blood-aqueous barrier 50-55 °C to constant weight. Extravascular albumin was determined by employing Evans blue as a colori- content and blood content were calculated as ml/g metic marker of albumin leakage, a technique exten- dry weight of bloodless tissue as previously de- sively employed in inflammation research. Ten mil- scribed,24 and values for each animal were finally ex- liliters of 2.5% Evans blue23 was injected into the pressed as the difference between test and control marginal ear vein. After a 15 min period, PGD2, BW tissues. 245C, PGJ2, 9«,1 l/3-PGF2a, prostaglandin E2 (PGE2: Leukocyte infiltration and goblet cell numbers Cayman, Kalamazoo, MI) or prostaglandin F2a were determined by microscopic studies at 6 hr post- (PGF2a: Cayman, Kalamazoo, MI) solutions were prostanoid administration. Immediately after sacri- then administered topically to one eye in a 25 /ul fice by intracardiac T-61, 10% neutral-buffered for- volume, and the contralateral eye received 25 ix\ ve- malin was administered to the ocular surface. The hicle as a control. The animals were sacrificed 30 min globe and attached eyelid ring were then surgically later by i.v. injection of 1 ml sodium pentobarbital excised intact and fixed in 10% neutral-buffered for- (Eutha-6: Western Medical, Aracadia, CA). Aqueous malin for 24 hr at room temperature. Tissues were humor samples were then extracted from each eye embedded in paraffin and two 6 ^im sections were and a 1 ml blood sample was obtained from the heart. obtained per eye in such a manner as to prevent the The blood was centrifuged at 4000 rpm for 10 min same cell populations from being counted in both

Downloaded from iovs.arvojournals.org on 09/24/2021 140 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / January 1990 Vol. 31

sections. The sections were stained by Luna's tech- Results nique for eosinophil granules.25 Leukocytes and goblet cell numbers were counted as previously de- The effects of PGD2, BW 245C, PGJ2, and 26 scribed and graphically expressed as the difference 9a, 1 Ij8-PGF2 on rabbit IOP are compared in Figure between treated and control tissues per high power 1. PGD2 and BW 245C, over an identical range of field (h.p.f.). concentrations (0.01%, 0.1%, 1%), produced a dis- The investigations described conform to the tinct ocular hypotensive response (Fig. 1A-F). A ARVO Resolution on the Use of Animals in Re- modest ocular hypertensive response was apparent search. for the 1% concentration of PGD2 (Fig. 1C), but not

Fig. 1. Effect of graded doses of PGD2, BW 245C, PGJ2 and 9a,l Ij8-PGF2 on rabbit intraocular pressure. Eyes that received PGD2 (A) 0.01%, (B) 0.1%, (C) 1%; BW 245C (D) 0.01%, (E) 0.1%, (F) 1%; PGJ2 (G) 1 I I H1 i . . 0.01%, (H) 0.1%, (I) 1%; 9a,lli3-PGF2(J)0.01%,(K) 30-i 0.1%, (L) 1% are represented by (•). Contralateral, 25- control eyes are represented by (•). Points are mean values ± SEM. *P < 0.05 and **P < 0.01 represent significant decreases in in- 2 15- traocular pressure relative to baseline (0 hr). n = 8.

10 i i 1 I i I I i J K 30n

E J20- Q. 2 15-

10 • III i i i "~TIII 0 12 3 4 12 3 0 12 3 4 TIME(hr) TIME (hr) TIME(hr)

Downloaded from iovs.arvojournals.org on 09/24/2021 No. 1 PGD2 OCULAR PHARMACOLOGY / Woodward er ol 141

Table 1. Comparison of the effect of PGD2, BW hibit ocular hypertensive activity. Gross observations 245C, PGJ2, 9a,l 10-PGF2, PGE2 and PGF2a of the ocular surface indicated that only the 1% doses on blood-aqueous barrier integrity as of PGD2 consistently caused obvious ocular surface indicated by leakage of Evans blue-labelled redness. albumin into aqueous humor The effects of PGD2, BW 245C, PGJ2 and 9a, 11/3- PGF on the rabbit ocular blood-aqueous barrier are ml Albumin/ml aqueous humor 2 compared to those of PGE2 and PGF2a in Table 1.

PGD2 (1%) 0.0003 ± 0.0003 PGD2, BW 245C and 9a,\l(3-PGF2 were virtually 0.0006 ± 0.0003 BW245C (1%) devoid of activity. PGF2a and PGJ2 elicited a modest PGJ2 (1%) 0.0044 ±0.0031 response but only PGF2a produced a statistically sig- 9a,U/3-PGF2(l%) 0.0003 ± 0.0003 PGE2 (1%) O.O75f± 0.0014 nificant effect. PGE2, as expected, caused a much PGF2a (1%) 0.0063* ± 0.0024 more pronounced disruption of the blood-aqueous

Values are mean ± SEM for test-control eyes. barrier. * P < 0.05, ^P < 0.01 represent a significant difference between test and The microvascular permeability responses to control eyes. n values are PGD2 = 7, BW 245C = 6, PGJ2 = 8, 9a, 1 l/3-PGF2 = 4, PGE2 PGD2, BW 245C, PGJ2 and 9a, 11,8-PGF2 in the bul- = 5, PGF2o = 8. bar conjunctiva of the guinea pig are compared in Figure 2. PGD2 caused a sharp rise in extra vascular albumin content over a 0.005-0.5% concentration for 1% BW 245C (Fig. IF). In contrast, PGJ2 did not range with the 0.5% concentration, causing a re- produce a decrease in IOP over the 0.01-1% concen- sponse that was only slightly greater than that evoked tration range (Fig. 1G-I). 9a, 1 l/?-PGF2 displayed oc- by 0.05% PGD2 (Fig. 2A). BW 245C (Fig. 2B), PGJ2 ular hypotensive activity only at the 1% concentra- (Fig. 2C) and 9a, 1 l/3-PGF2 (Fig. 2D) did not elicit a tion (Fig. 1J-L). PGJ2 and 9a, 1 l/3-PGF2 did not ex- similar microvascular permeability response. Blood

A. 1 .2 - B. 1 .2 - z 1.0- PGD, s 1 .0- BW245C m 0.8- 0.8- 0.6- i: 0.6 - -I CC 0.4- => Q 0.4- O 0.2- < i 0.2- 0.0- 5 0.0- 0.2 I 0.2 0005 6-05 0-5 0 005 0 05 CONCENTRATION (%) CONCENTRATION (%)

c. 1 .2- D. 1 .2 - z £ 1 .0- PGJ, 1 .0 - 9«,11/3PGF2 3 GQ 0.8- 0.8 - 0.6- $> o.aH ! _l OC 3 Q 0.4- go 0.4- O (0 O) (/> CD 0.2 - < := o : 0.0- o.o -

-0.2 •0.2 0005 005 0-5 0« CONCENTRATION (%) CONCENTRATION (%)

Fig. 2. Effect of (A) PGD2, (B) BW 245C, (C) PGJ2) (D) 9a, 1 l/3-PGF2 on guinea pig conjunctival microvascular permeability. Points are the mean values ± SEM of test-control tissues. **P < 0.01 represents a significant difference between test and control tissues, n = 6.

Downloaded from iovs.arvojournals.org on 09/24/2021 142 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / January 1990 Vol. 31

Table 2. Comparison of the blood content response The effect of prostanoids on eosinophil infiltration to PGD2, BW 245C, PGJ2 and 9a, 110-PGF2 in into the guinea pig conjunctiva is depicted graphi- guinea pig bulbar conjunctiva cally in Figure 3. The cellular responses obtained by Dose prostanoid quantitative histology are statistically analyzed between treatment groups in Table 3. The responses for 0.005% 0.05% 0.5% each dose of BW 245C were compared to those for

PGD2 0.078 ± 0.065 0.16 ±0.037 0.16 ±0.027 PGD2, PGJ2 and 9a, 1 l/?-PGF2 at equal dose levels by BW 245C 0.047 ±0.016 0.027 ±0.015 0.045 ± 0.028 the student non-paired t-test. PGD2 caused a dose- PGJ2 0.012 ±0.013 0.054 ± 0.02 0.039 ±0.01 dependent, diffuse eosinophil infiltrate as reflected by 9a,ll]8-PGF2 0.07 ±0.036 significantly increased eosinophil numbers in both Values are mean ± SEM for test-control tissues (ml blood/g dry wt. tissue); epithelial and subepithelial areas (Fig. 3A). BW 245C, n = 6. over an identical concentration range (0.005-5%), did not cause an eosinophil infiltrate (Fig. 3B). PGJ2 content values obtained during the microvascular produced a dose-dependent eosinophil infiltrate that permeability studies are summarized in Table 2. In- resembled that associated with PGD2 (Fig. 3C). creases in blood content for PGD2 showed, approxi- 9a, 1 l/3-PGF2 was inactive as an eosinophil chemoat- mately, the same dose-dependency that occurred tractant (Fig. 3D). At a 5% dose, PGD2, PGJ2 and with respect to microvascular permeability. BW 9a, 11(8-PGF2 all produced a modest (approximately 245C and PGJ2 did not produce an increase in blood two cells/h.p.f.) neutrophil infiltrate in the conjuncti- content of similar magnitude to that associated with val epithelium, whereas no neutrophils were apparent 0.05% and 0.5% PGD2. for BW 245C (data not illustrated).

A. 8 - B. 8 - PGDo BW245C 7 - 7 - 6 - 6 - 5 - 5 - 4 - 4 - 3 - 3 - 2 - 2 - 1 - 1 - 0 - 0 - - 1 - - 1 -

- 2 - 2 I I i 0005 005 0 5 0005 005 05 CONCENTRATION (%) CONCENTRATION (%)

D. c. 8 - 8 - 9«,11/?PGF 7 - PGJ 7 - 2 6 - 6 - 5 - 5 - 4 - 4 - 3 - 3 - 2 - 2 - 1 - 1 - 0 - 0 - - 1 - • 1 - - 2 - 2 O-tfOS 0 05 0 5 0005 0-5 CONCENTRATION (%) CONCENTRATION (%)

Fig. 3. Eosinophil infiltration into the guinea pig conjunctiva in response to (A) PGD2, (B) BW 245C, (C) PGJ2, (D) 9a, 11 j8-PGF2. Eosinophil infiltrates into the epithelial and subepithelial areas are represented by (•) and (A), respectively. Points are the mean of test-control eye values ± SEM. *P < 0.05, **P < 0.01 represent a significant difference between test and control eyes, n = 6.

Downloaded from iovs.arvojournals.org on 09/24/2021 No. 1 PGD2 OCULAR PHARMACOLOGY / Woodward er al 143

Table 3. Statistical comparison (student non-paired t-test) of BW 245 C versus PGD2, PGJ2, and 9a, 11/3- PGF2 treatments at equal dose levels on eosinophil infiltration and goblet cell numbers

Eosinophils Eosinophils Goblet (epithelium) (subepithelium) cells * 0.005% PGD2 vs. 0.005% BW 245C NS NS 0.005% PGD2 vs. 0.005% BW 245C NS NS NS 0.005% 9a, Ilj3-PGF2 vs. 0.005% BW 245C — — — 0.05% PGD2 vs. 0.05% BW 245C NS NS NS 0.05% PGJ2 vs. 0.05% BW 245C NS NS NS O.O5%9a,110-PGF2 vs. 0.05% BW 245C NS NS NS • 0.5% PGD2 vs. 0.5% BW 245C * 0.5% PGJ2 vs. 0.5% BW 245C t t t 0.5%9a,ll/3-PGF2 vs. 0.5% BW 245C NS NS NS 5% PGD2 vs. 5% BW 245C t ' NS t 5% PGJ2 vs. 5% BW 245C NS NS * 5%9a,ll)3-PGF2 vs. 5% BW 245C t t t *P< 0.05.

PGD2 produced a marked depletion of goblet cell maximal response appeared less than that produced numbers at the 0.5 and 5% concentrations (Fig. 4A by PGD2. In contrast, BW 245C (Fig. 4B) did not and Table 3 for between group statistical analyses). alter goblet cell numbers even when a 5% concentra- PGJ2 also caused a depletion in conjunctival goblet tion was topically applied to the guinea pig eye. A 5% cell numbers (Fig. 4C), although the potency and concentration of 9«,ll/3-PGF2 caused a small, but

A. B. 6 - 4 - BW245C 2 - 0 -

c. 6 - D. 6 -

4 - PGJ 4 - 9af1l/3PGF2 2 - 2 - j 0 - 0 -

CO S/HPF - 2- _i - 2- - 4- LU - 4- o • • 6- - 6- - 8- - 8-

-10- -10-

-1 2 •12- 0 005 0 05 0 5 5 0-5 5 CONCENTRATION (%) CONCENTRATION (%)

Fig. 4. Effect of (A) PGD2, (B) BW 245C, (C) PGJ2, (D) 9a, 1 l/3-PGF2 on the guinea pig conjunctival goblet cell population. Points are the mean of test-control eye values ± SEM. *P < 0.05, **P < 0.01 represent a sigi.ificant difference between test and control eyes, n = 6.

Downloaded from iovs.arvojournals.org on 09/24/2021 144 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / January 1990 Vol. 31

significant effect on goblet cells compared to BW has previously been shown to produce conjunctival 245C (Fig. 4D). erythema, chemosis and eosinophil infiltration in 22 The pathological effects of PGD2 on the guinea pig both guinea pigs and human volunteers. The pres- conjunctiva are compared with those of BW 245C ent studies confirm and extend previous findings and PGJ2 in Figure 5. Figure 5A illustrates the effect performed in the guinea pig conjunctiva where re- of 0.5% PGD2; note the epithelial and subepithelial sponses to PGD2 resemble the human conjunctival eosinophil infiltrate and the reduced goblet cell num- response to PGD2. In addition to inflammatory bers compared to vehicle-treated control eyes (Fig. events, PGD2 also produced dose-dependent goblet 5D). PGJ2 (0.5%) elicited a similar pathological re- cell depletion. In contrast to PGD2, BW 245C did not sponse (Fig. 5C) to that produced by PGD2. In con- increase conjunctival microvascular permeability trast, no leukocyte infiltration or depletion of the and no pathological effects on the conjunctiva were goblet cell population was observed for the 0.5% con- apparent even at a 5% dose. However, BW 245C re- centration of BW 245C (Fig. 5B), and the conjunctiva tained the ocular hypotensive activity associated with appeared indistinguishable from a vehicle-treated, PGD2. It appears that a DP-receptor agonist such as control conjunctiva (Fig. 5D). BW 245C may provide a valuable approach for sepa- rating PGD2-like ocular hypotensive activity from Discussion PGD2-like inflammatory effects on the conjunctiva. Among the analogues and metabolites examined,

Recent reports have demonstrated that PGD2 pro- only PGD2 and BW 245C potently lowered IOP. Ac- duces pronounced ocular hypotension in rabbits27 cording to the current working classification for pros- 19 and that PGD2-methyl ester lowers IOP in a human tanoid receptors, these results indicate that DP-re- 28 volunteer. In terms of ocular surface effects, PGD2 ceptor stimulation causes a decrease in IOP. It is un-

<* „'

• -C:

Fig. 5. Representative examples of the effects of PGD2, BW 245C, and PGJ2 on the guinea pig conjunctiva (Luna's stain, X100). PGD2 (A, top left) and PGJ2 (C, bottom left) produced a distinct eosinophil infiltration (E) and reduced the number of gobtet cells (G) relative to control tissues (D, bottom right). In contrast, conjunctivae treated with BW 245C (B, top right) appeared normal and indistinguishable from control (D). Prostanoid effects are compared at a 0.5% concentration.

Downloaded from iovs.arvojournals.org on 09/24/2021 No. 1 PGD2 OCULAR PHARMACOLOGY / Woodward er ol 145

likely that other major prostanoids, such as PGE2, sidered as not entirely satisfactory as a model for PGF2a and prostacyclin (PGI2) cause ocular hypoten- prostanoid effects on the human ocular surface. The sion by DP-receptor stimulation. PGE2 and PGF2a do ability of PGD2 to produce similar ocular pathologi- not inhibit human platelet aggregation,27'28 a re- cal and hyperemic effects over a similar dose range in sponse typically associated with BW 245C.1213'29 BW guinea pigs and man reaffirms the guinea pig as a 12 245C does not have high affinity for PGI2 receptors more suitable model for studying PGD2 pharmacol- and its pharmacological effects may be entirely delin- ogy of the ocular surface. Comparison of blood con- 18 eated from PGI2. Unlike PGD2, PGE2 and tent values suggests that the ocular surface hyperemia 30 PGF2a, BW 245C does not contract thromboxane- typically evoked by PGD2 would be diminished by a sensitive smooth muscle preparations.13 Moreover, selective DP-receptor agonist. 31 the potent and selective TxA2-mimetic U-46619 32 In summary, the pathological actions of PGD2 on does not lower IOP, which indicates that the TP-re- the conjunctiva may be separated from ocular hypo- ceptor does not participate in the ocular hypotensive tensive effects by employing a selective DP-receptor response to any prostanoid. It appears likely that DP- agonist, according to these studies in laboratory ani- receptor-induced ocular hypotension is distinct and mal models. By virtue of its selectivity, BW 245C unrelated to the actions of other prostanoids on IOP. provides a valuable "tool" for further studies on the DP-receptor stimulation does not appear involved DP-receptor in laboratory animal species and pri- in the inflammatory effects of PGD2 on the conjunc- mates with respect to mechanism and site of action tiva. Only PGD2 increased conjunctival micro vascu- and efficacy associated with chronic dosing. lar permeability; BW 245C, PGJ and 9a,l 1,3-PGF 2 2 Key words: prostaglandin D , prostaglandin J , 9«, 11/3- were inactive despite their marked pharmacological 2 2 prostaglandin F2, BW 245C, intraocular pressure, conjunc- activity in other systems. BW 245C did not cause tiva, inflammation eosinophil infiltration into the conjunctiva or goblet cell depletion. Interestingly, PGJ2 appeared virtually Acknowledgment equipotent to PGD2 as an eosinophil chemoattrac- tant and also caused goblet cell depletion. The pres- The authors express their gratitude to J. Boag for excel- lent secretarial assistance. ent studies seem to constitute the first report that PGJ2 can potently mimic PGD2 effects other than inhibition of cell proliferation. It is not clear from References these studies whether PGD2 and PGJ2 are causing 1. Whittle BJR, Moncada S, and Vane JR: Comparison of the pathological effects on the conjunctiva by stimulating effects of prostaglandin E, and D2 on platelet aggregation in a single receptor population; studies on isolated eo- different species. Prostaglandins 16:373, 1978. 2. Schneider WE and Drazen JM: Comparative in vitro effects of sinophils are required to resolve this issue. A further arachidonic acid metabolites on tracheal spirals and parenchy- consideration is that either or both PGD2 and PGJ2 mal strips. Am Rev Resp Dis 121:835, 1980. 12 may be enzymatically converted to A PGJ2 by reac- 3. Wasserman MA, Ducharme DW, Griffin RL, Degraaf GL, tions that involve serum albumin.1733 and Robinson FG: Bronchopulmonary and cardiovascular ef- A previous report has described PGD as an ocular fects of prostaglandin D2 in the dog. Prostaglandins 13:255, 2 1977. hypotensive that is devoid of ocular side effects, such 27 4. Toda N: Different responsiveness of a variety of isolated dog as flare and hyperemia, in the rabbit. The present arteries to prostaglandin D2. Prostaglandins 23:99, 1982. studies confirm that PGD2 does not disrupt the rabbit 5. Sakai T, Yamaguchi N, Shiroko Y, Sekiguchi M, Fujii G, and blood-aqueous barrier. BW 245C and 9«,1 l/3-PGF2 Nishino H: Prostaglandin D2 inhibits the proliferation of at a 1 % dose were also without meaningful effect on human malignant tumor cells. Prostaglandins 27:17, 1984. 6. Hayaishi O: Prostaglandin D2: A neuromodulator. In Ad- the blood-aqueous barrier. In contrast, 1% PGE2 vances in Prostaglandin, Thromboxane and Leukotriene Re- produced substantial plasma leakage into the search 12, Samuelsson B, Paoletti R, and Ramwell P, editors. aqueous humor, which is consistent with previous New York, Raven Press, 1983, pp. 333-337. 27 34 studies. ' PGF2a also produced significant disrup- 7. Flower RJ, Harvey EA, and Kingston WP: Inflammatory ef- tion of the blood-aqueous barrier of an intermediate fects of prostaglandin D2 in rat and human skin. Br J Pharma- col 56:229, 1976. magnitude. No conjunct val hyperemic effect was 27 8. Jones RL: Cardiovascular actions of prostaglandins D and E in previously reported for 50 /xg PGD2 in rabbits, the sheep: Evidence for two distinct receptors. In Advances in which may be viewed as surprising since this exceeds Prostaglandin and Thromboxane Research 1, Samuelsson B, the dose necessary to cause ocular surface vasodilata- Paoletti R, and Ramwell P, editors. New York, Raven Press, tion in human subjects.22 The present studies also 1976, pp.221-230. 9. Whittle BJR, Mugridge KG, and Moncada S: Use of the rabbit suggest that the rabbit ocular surface is not particu- transverse stomach-strip to identify and assay prostacyclin, larly sensitive to PGD2, which is generally regarded PGA2, PGD2 and other prostaglandins. Eur J Pharmacol as a potent vasodilator. Thus, the rabbit may be con- 53:167, 1979.

Downloaded from iovs.arvojournals.org on 09/24/2021 146 INVESTIGATIVE OPHTHALMOLOGY G VISUAL SCIENCE / January 1990 Vol. 31

10. Schafer AI, Cooper B, O'Hara D, and Handin RI: Identifica- val vasopermeability and blood-aqueous barrier integrity. In- tion of platelet receptors for prostaglandin I2 and D2. J Biol vest Ophthalmol Vis Sci 26:481, 1985. Chem 254:2914, 1979. 24. Woodward DF, Wasserman MA, and Weichman BM: Investi- 11. Siegl AM, Smith JB, and Silver MJ: Specific binding sites for gation of leukotriene involvement in the vasopermeability re- prostaglandin D2 on human platelets. Biochem Biophys Res sponse associated with guinea pig tracheal anaphylaxis: Com- Comm 90:291, 1979. parison with cutaneous anaphylaxis. Eur J Pharmacol 93:9, 12. Town M-H, Casals-Stenzel J, and Schillinger E: Pharmacologi- 1983. cal and cardiovascular properties of a hydantoin derivative, 25. Luna LG: Manual of Histologic Staining Methods of the BW 245C, with high affinity and selectivity for PGD2 recep- Armed Forces Institute of Pathology, 3rd ed. New York, To- tors. Prostaglandins 25:13, 1983. ronto, London, Sydney, McGraw-Hill, 1968, pp. 111-112. 13. Narumiya S and Toda N: Different responsiveness of prosta- 26. Woodward DF, Spada CS, Hawley SB, and Nieves AL: Con- glandin D2-sensitive systems to prostaglandin D2 and its ana- junctival eosinophil infiltration evoked by histamine and im- logues. Br J Pharmacol 85:367, 1985. mediate hypersensitivity: Modification by Hr and H2-receptor 14. Liston TE and Roberts LJ: Transformation of prostaglandin blockade. Invest Ophthalmol Vis Sci 27:1495, 1986. D2 to 9a, 1 l/3-(15S)-trihydroxyprosta-(5z,13E)-dien-l-oic acid 27. Goh Y, Nakajima M, Azuma I, and Hayaishi O: Prostaglandin (9a, 11/3-prostaglandin F2): A unique biologically active prosta- D2 reduces intraocular pressure. Br J Ophthalmol 72:461, glandin produced enzymatically in vivo in humans. Proc Natl 1988. Acad Sci USA 82:6030, 1985. 28. Goh Y. European patent application 87107049. 6. 15. Beasley CRW, Robinson C, Featherstone RL, Varley JG, 29. Samuelsson B, Granstrom E, Green K, Hamberg M, and Hardy CC, Church MK, and Holgate ST: 9a, 11/3 prostaglan- Hammarstrom S: Prostaglandins. Ann Rev Biochem 44:669, din F2, a novel metabolite of prostaglandin D2 is a potent 1975. contractile agonist of human and guinea pig airways. J Clin 30. Jones RL and Wilson NH: Partial agonism of prostaglandin H Invest 79:978, 1987. 2 analogs and 11-deoxy-prostaglandin F2a to thromboxane-sen- 16. Fukushima M, Kato T, Ota K, Arai Y, and Narumiya S: 9- 9 sitive preparations. In Advances in Prostaglandin and Throm- deoxy-A -prostaglandin D2: A prostaglandin D2 derivative boxane Research 6, Samuelsson B, Ramwell PW, and Paoletti with potent anti-neoplastic and weak smooth muscle contract- R, editors. New York, Raven Press, 1980, pp. 467-475. ing activities. Biochem Biophys Res Comm 109:626, 1982. 31. Whittle BJR, Moncada S, Mullane K, and Vane JR: Platelet 17. Mahmud I, Smith DL, Whyte MA, Nelson JT, Cho D, Tokes and cardiovascular activity of the hydantoin BW 245C, a po- LG, Alvarez R, and Willis AL: On the identification and bio- tent prostaglandin analogue. Prostaglandins 25:205, 1983. logical properties of prostaglandin J2. Prostaglandins Leuko- trienesMed 16:131, 1984. 32. Coleman RA, Humphrey PPA, Kennedy I, and Lumley P: Prostanoid receptors: The development of a working classifica- 18. Sanger GJ, Jackson A, and Bennett A: A prostaglandin analog tion. Trends in Pharmacological Science 5:303, 1984. which potently relaxes human uterus, but not gut muscle. Eur J Pharmacol 81:141, 1982. 33. Coleman RA, Humphrey PPA, Kennedy I, Levy GP, and 19. Coleman RA, Humphrey PPA, and Kennedy I: Prostanoid Lumley P: Comparison of the actions of U-46619, a prosta- receptors in smooth muscle: Further evidence for a proposed glandin H2-analogue, with those of prostaglandin H2 and classification. Trends in Autonomic Pharmacology 3:35, 1982. thromboxane A2 on some isolated smooth muscle preparations. BrJ Pharmacol 73:773, 1981. 20. Giles H and Leff P: The biology and pharmacology of PGD2. Prostaglandins 35:277, 1988. 34. Kulkarni PS and Srinivasan BD: The effect of intravitreal and 21. Bito LZ: Comparison of the ocular hypotensive efficacy of topical prostaglandins on intraocular inflammation. Invest eicosanoids and related compounds. Exp Eye Res 38:181, Ophthalmol Vis Sci 23:383, 1982. 1984. 35. Kikawa Y, Narumiya S, Fukushima M, Wakatsuka H, and 9 l2 22. Abelson MB, Madiwale NA, and Weston JH: The role of pros- Hyaishi O: 9-deoxy-A ,A -13,14-dihydro prostaglandin D2, a taglandin D2 in allergic ocular disease. In Advances in Immu- metabolite of PGD2 formed in human plasma. Proc Natl Acad nology and Immunopathology of the Eye, O'Conner GR and Sci USA 81:1317, 1984. Chandler JW, editors. Chicago, Year Book Medical Publishers, 36. Neufeld AH and Sears ML: The site of action of prostaglandin 1982, pp. 163-166. E2 on the disruption of the blood-aqueous barrier in the rabbit 23. Woodward DF and Ledgard SE: Effect of LTD4 on conjuncti- eye. Exp Eye Res 17:445, 1973.

Downloaded from iovs.arvojournals.org on 09/24/2021