their content of urokinase-type plasminogen activator (uPA), tissue-type plasminogen activator (tPA), plasminogen activators inhibitors (PAIs), plasminogen, and ...
Investigative Ophthalmology & Visual Science, Vol. 33, No. 9, August 1992 Copyright © Association for Research in Vision and Ophthalmology
Urokinase-Type Plasminogen Activator in Human Aqueous Humor Stephanie F. Bernatchez,*§ Cyrus Tabatabay,t and Dominique Deling In the anterior segment of the eye, fibrin clots must be rapidly resorbed to prevent further fibrosis and scarring. The aqueous humor of patients undergoing cataract surgery was analyzed for the presence of components of the fibrinolytic cascade. In 30 patients, aqueous humor and plasma were compared for their content of urokinase-type plasminogen activator (uPA), tissue-type plasminogen activator (tPA), plasminogen activators inhibitors (PAIs), plasminogen, and total proteins. With gel electrophoresis and zymographic assays of serial dilutions of plasma and aqueous humor, all these components were found to be present at lower concentrations in aqueous humor than in plasma. For total proteins, the aqueous/plasma ratio was approximately 0.003, and for plasminogen it was 0.001. Interestingly, the aqueous/plasma ratio for uPA was not as low and varied from 0.01 to 0.03. A significant proportion of the uPA in aqueous humor was present in the two-chain active form. In addition to uPA, aqueous humor contained lower levels of tPA, but no detectable levels of reactive plasminogen activators inhibitors (PAIs). The presence of a relatively high concentration of active uPA shows that the proteolytic balance of the aqueous humor in the anterior chamber of the eye is shifted toward fibrinolysis. Invest Ophthalmol Vis Sci 33:2687-2692,1992
Aqueous outflow through the trabecular meshwork
fied, which are the products of separate but related
is important for the maintenance of normal intraocular pressure. The protein composition of aqueous humor has been compared with that of plasma in human1 and calf2 using two-dimensional gel electrophoresis. The most abundant protein in aqueous humor is albumin.2 Aqueous humor contains alpha-2-macroglobulin (820 kD), whereas some smaller plasma proteins are absent, indicating that aqueous humor is not a simple filtrate of plasma.1 Fibrinolytic activity has been detected in the anterior segment of the eye and has been proposed to contribute to the maintenance of normal aqueous outflow.3'4 This proteolytic activity was attributed to the presence of plasminogen activators (PAs)—serine proteases that convert plasminogen, an inactive zymogen, into plasmin, a neutral protease of broad specificity responsible for fibrinolysis.56 In mammals, two types of PAs have been identiFrom ""Centre Hospitalier de l'Universite Laval, Quebec, Canada, fClinique d'Ophtalmologie, Hopital Cantonal Universitaire, Geneve, and fDepartement de Pathologie, Centre Medical Universitaire, Geneve, Suisse. § Present address: Departement de Pharmacie galenique, Universite de Geneve, Geneve, Suisse. This work was supported by a grant from the Swiss National Science Foundation (3.59-0.87) and by funds from the State of Geneva. Submitted for publication: July 10, 1991; accepted March 11, 1992. Reprint requests: Dr. D. Belin, Dept. Pathologie, CMU, 1, rue Michel-Servet, CH-1211 Geneva, Switzerland.
genes. Urokinase-type PA (uPA), a protein of 55 kD, is synthesized and secreted as a single-chain inactive proenzyme7 that is converted into the two-chain active enzyme by limited proteolysis. Tissue-type PA (tPA) has a molecular weight of 70 kD, and its singlechain and two-chain forms are enzymatically active.5 Histochemical localization of PAs in human and monkey eyes has revealed fibrinolytic activity3'4 and the presence of tPA8"10 in several anterior chamber tissues. Fibrinolytic assays have documented PA production by cultured bovine corneal endothelial cells," and cultured human trabecular meshwork cells have been shown to synthesize tPA.12 Recently, the presence of tPA in the aqueous humor of several species, including man, has been documented.910'13'14 However, to our knowledge, the presence of uPA in aqueous humor has not yet been reported, although uPA was detected in cultures of bovine corneal endothelial cells.15 To further characterize the proteolytic potential of the anterior chamber of the eye,16 we have searched for components of the PA/plasmin system in human aqueous humor collected prior to cataract surgery. This analysis documents the presence of uPA in the human aqueous humor. Materials and Methods Materials Blood and aqueous humor were obtained from 30 patients submitted to cataract surgery. The procedure
2687
2688
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / August 1992
was explained and prior consent was obtained from the patients. Blood samples were taken in the operating room before surgery. The aqueous humor was obtained by corneal paracentesis after retrobulbar anesthesia and prior to conjunctival desinsertion to minimize bleeding.17 Plasminogen was purified from human plasma.18 Mr 55,000 urokinase, 100 U/^g,19 was obtained from the Green Cross Corporation (Osaka, Japan); amiloride (A7410) was from Sigma (St. Louis, MO), and diisopropylfluorophosphate (DFP) was from Merck AG (Darmstadt, Germany). Rabbit antibodies to human uPA and tPA were generous gifts of Dr. W.-D. Schleuning and Dr. E. Dowdle, respectively, and were used as described.20 Preparation of Samples
Blood samples were collected in tubes containing citrate or EDTA as anticoagulant and were centrifuged at 800 X g for 5 min at 0°C. The plasma was analyzed directly or stored in aliquots at -20°C after initial freezing in liquid nitrogen. Aqueous humor samples, collected prior to cataract surgery as an initial step of the procedure, were centrifuged at 800 X g for 5 min at 0°C to eliminate cellular contaminants and were analyzed directly or stored as described above. No significant differences were observed in the zymographic pattern of fresh and frozen aqueous humor samples. Gel Electrophoresis and Zymography of PAs and Plasminogen
Samples were electrophoresed in the presence of SDS under nonreducing conditions in 10% polyacrylamide gels (SDS-PAGE) using the buffer system of Laemmli.21 To estimate the relative protein content in aqueous humor and plasma, serial threefold dilutions of plasma were electrophoresed in parallel with the aqueous humor samples. The lowest and highest plasma dilutions analyzed always contained, respectively, more and less protein, PA, or plasminogen than the aqueous humor samples. The plasma dilutions therefore could be used for a semiquantitative analysis of aqueous humor. Total proteins were revealed by staining of the gels with coomassie blue. Albumin, the major protein in plasma and aqueous humor, was used to visually estimate total protein content. PAs were detected by the zymographic method of Granelli-Piperno and Reich,22 with minor modifications.2023 Briefly, after SDS-PAGE of the samples, the gels were washed twice for 10 min in 2.5% Triton X100, twice for 10 min in Tris 0.1 mol/1 Tris/HCl pH 8.1, and layered over an agar gel containing casein (2% weight/volume) and plasminogen ). The PAs diffuse into the agar gel and con-
Vol. 33
vert plasminogen into plasmin, causing local caseinolysis. The transparent lytic bands are dark under indirect illumination, and the opaque background of undigested casein is white. To facilitate the identification of uPA and tPA, and to discriminate uPA/PAI and tPA/PAI complexes, 1 mmol/1 amiloride was included in agar gels. This compound inhibits specifically uPA and its related complexes.24 To detect plasminogen in the samples, plasminogen was omitted and urokinase (50 mU/ml) was included in agar gels. Discrimination between the proenzyme and enzyme forms of uPA was accomplished as described previously,23 by incubating the samples with diisopropylfluorophosphate (DFP), which inhibits the enzyme form, prior to SDS-PAGE and zymography. Photographs of zymograms were taken under dark-ground illumination after incubation at 37°C for 15-24 hr. Immunoprecipitation
Immunodepletion of PAs was performed as described previously.20 Briefly, plasma and aqueous humor samples (20 fx\) were incubated with 2 n\ of antihuman uPA (0.5 mg/ml), antihuman tPA (1 mg/ml), or nonimmune rabbit IgG (1.2 mg/ml). After 2 hr at 4°C, fixed Staphylococcus aureus(\0 /A of a 10% w/v suspension) were added. After 30 min at 20°C, the samples were centrifuged for 5 min in a microfuge to remove immune complexes, and the supernatants were subjected to SDS-PAGE and zymography. Results The aqueous humor and plasma samples were first compared for their total protein content. Coomassie blue staining of the gels showed that the protein concentration in aqueous humor was approximately 0.3% of that found in plasma (results not shown), a result similar to previously reported values.17-25 Aqueous humor and plasma from two patients undergoing cataract surgery were analyzed for PAs by a zymographic assay, using an underlying agar gel containing casein and plasminogen (Fig. 1 A). The pattern observed with aqueous humor was strikingly different from that of plasma. Three major bands of comparable intensity, and with Mr of 55,80, and 110 kD, were detected in plasma (lanes 1-6). In aqueous humor (lanes 7 and 8), one major band with an Mr of 55 kD was detected, as well as traces of 70 kD and 110 kD bands, which became more visible upon prolonged incubation. Similar results were obtained after zymographic analysis of aqueous humor and plasma from 28 additional patients. A comparison of the intensities of the lytic bands generated by the serial dilutions of plasma with those of the aqueous humor samples
UPA IN HUMAN AQUEOUS HUMOR / Dernorchez er ol
No. 9
MrxlO"
80-
Fig. 1. Zymographic analysis of plasminogen activators in human plasma and aqueous humor from two patients. Patient 1: lanes 1, 3, 5, and 7; patient 2: lanes 2, 4, 6, and 8. (A) 0.3 n\ of plasma (lanes 1 and 2), 0.1 fi\ of plasma (lanes 3 and 4), 0.03 fil of plasma (lanes 5 and 6), and 10 p.1 of aqueous humor (lanes 7 and 8) were subjected to SDS-PAGE and zymography. The Mr of the bands were determined by comparison with standards proteins electrophoresed in an adjacent lane and stained with coomassie blue. (B) Zymographic analysis in the absence of plasminogen. Lanes 1 and 2:0.3 ix\ of plasma; lanes 3 and 4: lOjul of aqueous humor. Both gels were incubated for the same period of time.
indicated that the concentration of the 55 kD band in aqueous humor was 1-3% of that found in plasma. To characterize the caseinolytic proteases detected by zymography (Fig. 1 A), the samples were analyzed in parallel on an agar gel containing casein only (Fig. IB). In the absence of plasminogen, no proteolytic activity could be detected in aqueous humor (lanes 3 and 4), whereas a minor band of 80 kD was observed in plasma (lanes 1 and 2). Thus, the 55 kD, 70 kD, and 110 kD lytic bands reflect the presence of PAs in the samples. A 80 kD protease was previously described in zymograms of human plasma and may represent prekallikrein or factor XII.5'26 The PAs detected in aqueous humor and plasma were identified by immunodepletion with uPA- and tPA-specific antibodies. Figure 2A shows the results obtained with a pair of plasma and aqueous humor samples from one patient and with aqueous humor
2689
from two other patients. The 55 kD band, which constitutes the major activity in aqueous humor, was selectively immunodepleted with the anti-uPA antibodies (lanes 2). Free tPA was not detectable in plasma, and the 110 kD band abundant in plasma probably is a complex between tPA and its plasma inhibitor PAIl,27 because it was immunodepleted by the anti-tPA antibodies (lane 3). The low recovery of plasmatic tPA/PAI-1 complexes and of aqueous humor tPA after precipitation with irrelevant (lanes 2) and nonimmune (lanes 4) IgG results from nonspecific adsorption to Staphylococcus aureus2* The 80 kD band was not recognized by the PA-specific antibodies. The identification of PAs was confirmed by zymography in the presence of amiloride, a competitive inhibitor of uPA catalytic activity24 (Fig. 2B). Under these conditions, the plasmatic tPA/PAI-1 complexes and the comparatively smaller amount of tPA in aqueous humor was more readily observable. The apparent lower molecular weight of plasmatic uPA, which is particularly evident in Figure 2B, was a result of compression by albumin, as previously described.26 Thus, the major PA in aqueous humor is uPA. Urokinase-type PA is synthesized and secreted as an inactive proenzyme in most tissues and biological fluids,5-6 except in excreted urine, where the active enzyme predominates.28 To determine whether uPA also is present in aqueous humor as the proenzyme, aqueous humor from three patients was incubated prior to zymography with diisopropylfluorophosphate (DFP), which irreversibly inhibits the active 2chain form of uPA but does not affect the proenzyme (Fig. 3). In all cases, the amount of uPA activity was decreased by DFP treatment, although the extent of inhibition varied. There was a three- to fourfold inhibition with aqueous humor of patient 2, whereas patients 1 and 3 showed only an approximately twofold inhibition of uPA activity. In contrast, the activity of purified 2-chain uPA (Fig. 3, uPA lanes), as well as that of tPA in the aqueous humor samples, were essentially abolished by DFP treatment. The activity of single-chain uPA and of tPA/PAI complexes in plasma was resistant to DFP treatment, as expected.1929 Similar results were obtained with fresh aqueous humor, as well as with samples frozen and thawed. Although the relative amounts of proenzyme and active uPA cannot be easily quantitated by zymography,7'23 these results indicate that the single chain inactive and the 2-chain active forms of uPA are present in aqueous humor. Plasminogen concentration was estimated by zymography in the presence of uPA and casein. Aqueous humor was found to contain approximately 0.1% of the plasminogen found in plasma (results not shown). Attempts to detect PA inhibitors in aqueous
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Augusr 1992
2690
Vol. 33
Fig. 2. Characterization of the plasminogen activaplasma 1 ah 1 ah 2 ah 3 tors (PA) in human plasma and aqueous humor. (A) 0.3 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 fA of plasma (patient 1, plasma 1) and 10 jd of aqueous humor from patient I and two other patients (ah I to ah 3) were treated directly with sample buffer (lanes I) or immunodepleted with antihuman uPA IgG (lanes 2), with antihuman t-PA IgG {lanes 3), or with nonimmune IgG (lanes 4). The immunosupernatants were subjected to SDS-PAGE and zymography. Low recovery of tPA/PAI complexes in plasma and of t-PA in aqueous humor results from nonspecific adsorption to Staphylococcus aweusP (B) 0.3 ^1 of plasma (patient 1, plasma I) and 10^1 of aqueous humor from three other patients (ah 2 to ah 4) were subjected to SDS-PAGE in duplicate. The gels were layered onto casein-agar-plasminogen gels. Amiloride 1 mmol/t was included in one ofthe substrate gels (right panel) . The four patients analyzed in thisfigureare different from the two patients analyzed in Figure 1.
humor by assaying for the formation of high molecular complexes with 125I-labelled uPA2023 and by reverse zymography20-29 were unsuccessful (results not shown). Discussion This study was undertaken to further characterize the proteolytic activity of human aqueous humor.3'4'30 The samples were collected from patients undergoing cataract surgery, and cataract generally is believed not
MrxlO"3 110-
P 1 1 2
ah 1 1 2
ah 2 1 2
ah 3 1 2
uPA 1 2
- +
- +
- +
- +
- +
55 DFP
Fig. 3. Sensitivity of plasminogen activators (PA) in human plasma and aqueous humor to DFP. Added to each sample was a phosphate-buffered solution containing 1 mg/ml bovine serum albumin (PBS-BSA; - , lanes 1), or 20 mmol/1 DFP in PBS-BSA (+, lanes 2). The samples were incubated for 2 hr at 20°C and subjected to SDS-PAGE and zymography. pi: 0.1 ^1 of plasma from patient 1; ah 1 to 3: KM of aqueous humor from the same patient and two other patients; uPA: active human urinary u-PA (3 mU) was used as a positive control for the efficiency of DFP treatment.
to affect the functional properties of aqueous humor. Upon zymographic analysis in the presence of plasminogen, the major PA in aqueous humor was a 55 kD band, which we identified as uPA. This PA comigrated with purified uPA, selectively reacted with an antiserum raised against human uPA, and was inhibited by amiloride. In agreement with previous reports,910'1314 small amounts of tPA also were detected in aqueous humor. A fraction of uPA in aqueous humor was inhibited by DFP, an active site titrant of serine proteases, indicating the presence of the 2chain active enzyme and its proenzyme form. Active uPA and tPA are rapidly inactivated by PAIs, whereas pro-uPA and PAIs can accumulate concomitantly in the extracellular milieu.23 Hence, our failure to detect free PAI in aqueous humor was not unexpected, considering the presence of active uPA and tPA in aqueous humor. Thus, in contrast to plasma, the normal aqueous humor has a high fibrinolytic potential. The site or sites of synthesis ofthe uPA detected in aqueous humor are not yet identified. The enzyme may originate from plasma or may be synthesized and secreted by ocular tissues—for instance, by corneal endothelial cells.15 Large amounts of uPA activity have been detected in the murine cornea,31 although the relative contribution of the endothelium is not known. In this context, it seems important to stress that the PAs present in aqueous humor may represent
No. 9
UPA IN HUMAN AQUEOUS HUMOR / Qernarchez er al
only a fraction of the total fibrinolytic activity in the anterior chamber of the eye. Indeed, uPA and plasminogen can associate with binding sites at the cell surface, while tPA binds to a number of extracellular matrix components.6 Therefore, it will be interesting to determine whether uPA receptors and plasminogen binding sites are present at the surface of corneal endothelial cells or of the lens epithelium as well as the amounts of cell-bound enzymes. In addition, the observation that cultured corneal endothelial cells can internalize uPA/PAI complexes32 suggests that the cells surrounding the aqueous humor may contribute in a variety of ways to the balance of proteolytic activity in this biological fluid. Several roles for the activation of plasminogen by PAs have been proposed.56 Whereas tPA is primarily associated with fibrinolysis, uPA generally was thought to participate in the extracellular proteolytic events that accompany cell migration and tissue remodeling. The recent detection of uPA and tPA mRNAs in distinct portions of renal tubules33 has led to the suggestion that both PAs may contribute to the maintenance of tubular patency. A similar situation may occur in the anterior chamber of the eye, where both enzymes could cooperate to prevent fibrin deposition. There is preliminary evidence that total PA activity is reduced in the aqueous humor of glaucoma patients, suggesting that the resistance to aqueous outflow associated with increased intraocular pressure may be associated with increased protein deposition in the anterior chamber.34 More direct evidence for a role of plasminogen activation in aqueous humor is provided by the observation that exogenously added plasmin facilitates outflow.35 In conclusion, the identification of uPA as a major PA in the aqueous humor suggests that this enzyme participates in the maintenance of a functional aqueous outflow in the normal human eye. Key words: aqueous humor; urokinase; plasminogen activator; protease References 1. Yu TC and Okamura R: Comparative study of native proteins in aqueous humor and serum—Detection of characteristic aqueous humor proteins. Jpn J Ophthalmol 31:235, 1987. 2. Pavao AF, Lee DA, Ethier CR, Johnson MC, Anderson PJ, and Epstein DL: Two-dimensional gel electrophoresis of calf aqueous humor, serum, andfilter-boundproteins. Invest Ophthalmol Vis Sci 30:731, 1989. 3. Kwaan HC and Astrup T: Localization offibrinolyticactivity in the eye. Arch Pathol 76:595, 1963. 4. Pandolfi M and Kwaan HC: Fibrinolysis in the anterior segment of the eye. Arch Ophthalmol 77:99, 1967. 5. Dano K, Andreasen PA, Grondahl-Hansen J, Kristensen P, Nielsen LS, and Skriver L: Plasminogen activators, tissue degradation, and cancer. Adv Cancer Res 44:139, 1985.
2691
6. Vassalli J-D, Sappino A-P, and Belin D: The plasminogen activator/plasmin system. J Clin Invest 88:1067, 1991. 7. Pedersen LC, Lund LR, Nielsen LS, Dano K, and Skriver L: One-chain urokinase-type plasminogen activator from human sarcoma cells is a proenzyme with little or no intrinsic activity. J BiolChem 263:11189, 1988. 8. Tripathi BJ, Geanon JD, and Tripathi RC: Distribution of tissue plasminogen activator in human and monkey eyes. An immunohistochemical study. Ophthalmology 94:1434, 1987. 9. Park JK, Tripathi RC, Tripathi BJ, and Barlow GH: Tissue plasminogen activator in the trabecular endothelium. Invest Ophthalmol Vis Sci 28:1341, 1987. 10. Geanon JD, Tripathi BJ, Tripathi RC, and Barlow GH: Tissue plasminogen activator in avascular tissues of the eye: A quantitative study of its activity in the cornea, lens, and aqueous and vitreous humors of dog, calf, and monkey. Exp Eye Res 44:55, 1987. 11. Fehrenbacher L, Gospodarowicz D, and Shuman MA: Synthesis of plasminogen activator by bovine corneal endothelial cells. Exp Eye Res 29:219, 1979. 12. Shuman MA, Polansky JR, Merkel C, and Alvarado J A: Tissue plasminogen activator in cultured human trabecular meshwork cells. Invest Ophthalmol Vis Sci 29:401, 1988. 13. Hayashi K, Nakashima Y, Sueishi K, Tanaka K, and Inomata H: Fibrinolytic activity and localization of plasminogen activator in bovine vitreous fluid and aqueous humor. Invest Ophthalmol Vis Sci 27(suppl):49, 1986. 14. Tripathi RC, Park JK, Tripathi BJ, and Millard CB: Tissue plasminogen activator in human aqueous humor and its possible therapeutic significance. Am J Ophthalmol 106:719, 1988. 15. Mirshahi M, Mirshahi S, Soria C, Soria J, Thomaidis A, Pouliquen Y, and Faure JP: Production of proteases type plasminogen activator and their inhibitor in cornea. Biochem Biophys ResCommun 160:1021, 1989. 16. O'Rourke J, Wang W-P, Donnelly E, Wang E, and Kreutzer DL: Extravascular plasminogen activator and inhibitor activities detected at the site of a chronic mycobacterial-induced inflammation. Am J Pathol 126:334, 1987. 17. Tripathi RC, Millard CB, Tripathi BJ: Protein composition of human aqueous humor: SDS-PAGE analysis and post-mortem samples. Exp Eye Res 48:117, 1989. 18. Deutsch DG and Mertz ET: Plasminogen purification from human plasma by affinity chromatography. Science 170:1095, 1970. 19. Wun T-C, Schleuning W-D, and Reich E: Isolation and characterization of urokinase from human plasma. J Biol Chem 257:3276-3283, 1982. 20. Busso N, Belin D, Failly-Crepin C, and Vassalli J-D: Plasminogen activators and their inhibitors in a human mammary cell line (HBL-100): Modulation by glucocorticoids. J Biol Chem 261:9309, 1986. 21. Laemmli UK: Cleavage of structural proteins during the assembly of the head bacteriophage T4. Nature 227:680, 1970. 22. Granelli-Piperno A and Reich E: A study of proteases and protease-inhibitor complexes in biological fluids. J Exp Med 148:223, 1978. 23. Vassalli J-D, Dayer J-M, Wohlwend A, and Belin D: Concomitant secretion of prourokinase and of a plasminogen activatorspecific inhibitor by cultured human monocytes-macrophages. J Exp Med 159:1653, 1984. 24. Vassalli J-D and Belin D: Amiloride selectively inhibits the urokinase-type plasminogen activator. FEBS Lett 214:187, 1987. 25. Caprioli J: The ciliary epithelium and aqueous humor. In Adler's Physiology of the Eye: Moses RA and Hart WM, editors. St. Louis, The C.V. Mosby Company, 1987, pp. 204-222.
2692
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Augusr 1992
26. Tissot JD, Schneider P, Hauert J, Ruegg M, Kruithof EKO, and Bachmann F: Isolation from human plasma of a plasminogen activator identical to urinary high molecular weight urokinase. J Clin Invest 70:1320, 1982. 27. Erickson LA, Hekman CM, and Loskutoff DJ: The primary plasminogen-activator inhibitors in endothelial cells, platelets, serum, and plasma are immunologically related. Proc Natl AcadSci USA 82:8710, 1985. 28. Rijken DC, Wijngaards C, and Welbergen J: Immunological characterization of plasminogen activator activities in human tissues and body fluids. J Lab Clin Med 97:477, 1981. 29. Loskutoff DJ, Van Mourik JA, Erickson LA, and Lawrence D: Detection of an unusually stable fibrinolytic inhibitor produced by bovine endothelial cells. Proc Natl Acad Sci USA 80:2956, 1983. 30. Franceschetti A and Eichenberger E: Fibrinolyse in Kammer-
31.
32. 33.
34.
35.
Vol. 33
wasser menschlicher und tierischer Augen. Experientia 15:130, 1959. Miskin R, Axelrod JH, Griep AE, Lee E, Belin D, Vassali J-D, and Westphal H: Human and murine urokinase cDNAs linked to the murine aA-cristallin promoter exhibit lens and non-lens expression in transgenic mice. Eur J Biochem 190:31, 1990. Shuman MA and Merkel CH: Urokinase binding to bovine corneal endothelial cells. Exp Eye Res 41:371, 1985. Sappino A-P, Huarte J, Vassalli J-D, and Belin D: Sites of synthesis of urokinase and tissue-type plasminogen activators in the murine kidney. J Clin Invest 87:962, 1991. Mehra K.S, Dube B, Mikuni I, and Dube RK: Reduced fibrinolytic activity in aqueous humor of chronic simple glaucoma. Tokai J Exp Clin Med 9:33, 1984. Perkins ES and Saiduzzafar H: The effect of plasmin on the facility of outflow in cynomolgus monkeys. Exp Eye Res 8:386, 1969.