The effect of salicylates on aqueous fibrinolysin

Judson P. Smith, ///,* and Gerald R. Christensen

Assuming that fibrinolytic activity may play a role in cases of recurrent bleeding in traumatic hyphema, a microassay method for determining fibrinolytic activity in the aqueous humor was developed. Results indicate that the rabbit aqueous contains relatively low levels of fibrinolysin and. high levels of fibrinolysin activators. Systemic administration of acetylsalicylic acid in therapeutic doses is shown to reduce the fibrinolytic activity in the aqueous of the rabbit substantially.

Key words: aqueous humor, fibrinolysin, fibrinolysin activators, acetylsalicylic acid, aspirin, salicylates, profibrinolysin, rabbit, hyphema.

hile the majority of traumatic hyphe- to seal off the ruptured vessels, and a mas resolve without sequelae, as many as firm fibrin clot develops. The factors re- 26 per cent of cases have been shown sponsible for the breakdown of these hemo- to develop secondary bleeding.1 This sec- static controls leading to secondary bleed- ondary hemorrhage often leads to severe ing concern certain fibrinolytic processes glaucoma due to mechanical obstruction of occurring in the plasma and aqueous hu- the trabecular meshwork and pupil. It can mor. In particular, the enzyme fibrinolysin be inferred from hemostatic processes else- is assumed to play a significant role in where in the body that with the primary the development of the secondary bleed. hemorrhage the ruptured vessels contract At present it is uncertain whether clot ly- and recess, a platelet plug rapidly develops sis occurs primarily in response to plasma fibrinolytic activity or in response to the fibrinolytic activity present in the aqueous From the Department of Ophthalmology, Uni- humor. That fibrinolysin is a major factor versity of Texas Medical Branch, Galveston, in a secondary bleed is also an assumption, Texas. and it is quite possible that other variables This study was supported in part by National in the process of hemostasis play a larger Eye Institute Grant 03105030 DHEW T501 role. EY00058 05. In 1952, Ungar, Damgaard, and Hum- Manuscript submitted Dec. 1, 1970; revised mel2 reported that salicylates inhibited the manuscript accepted Feb. 8, 1971. protease fibrinolysin (as well as many Reprint requests: Dr. Gerald R. Christensen, Department of Ophthalmology, University of other enzymes). They attempted to corre- Texas Medical Branch, Galveston, Texas 77550. late the anti-inflammatory action of the "Junior Student, University of Texas Medical salicylates with their ability to inhibit Branch. fibrinolysin, a protease which is believed 266

Downloaded from iovs.arvojournals.org on 10/05/2021 Volume 10 Salicylates on aqueous fibrinolysin 267 Number 4

Urokinase, tissue activators, staphylokinase

Key: ^ Antikinases »-= potentiation ^ i •= inhibition Profibrinolysin r*- Fibrinolysin -"*- I i Proactivator— •Activator Streptokinase Fig. 1.

to play a significant role in the inflamma- animals; five similar rabbits served as controls. tory reaction. However, recent evidence The experimental animals were given pure acetyl- salicylic acid (35 mg. per kilogram) suspended indicates that in plasma, aspirin may in- in a 10 per cent acacia mucilage solution every hibit certain inhibitors in the first stage six hours for 48 hours. The drug was administered of fibrinolysin, at least in specific concen- to each animal via a plastic tube inserted through tration ranges (10~3M salicylic acid).3 That the mouth into the esophagus. Two hours after is, aspirin appears to inhibit certain in- the last dose of aspirin, the aqueous humor of both eyes of each rabbit was aspirated with a hibitors (antikinases) responsible for the No. 27 tuberculin syringe and the aqueous from conversion of profibrinolysin to fibrinolysin. both eyes was pooled and placed in a small This would result in an increase in fibrino- test tube. Anesthesia during paracentesis con- lytic activity. The effect of aspirin on the sisted of two drops locally of a 0.25 per cent second stage of fibrinolysis is being studied proparacaine solution (Ophthetic). currently. Recent studies indicate that the Experiment I. To estimate the fibrinolytic ac- second stage is not affected in the same tivity of aqueous humor, a standard lysis time 3 determination was employed. This was conducted range of concentration. Indeed, it has as follows: 1 per cent Armour bovine fibrinogen been well documented recently that sali- preparation (1 ml.) containing 0.9 per cent cylates decrease the euglobulin lysis time NaCl and 5 per cent imidazole buffer (pH dramatically and may very well become 7.2) was added to a 0.58 per cent NaCl solu- one of the most important drugs in the tion (3 ml.) This dilution produced a final fibrinogen concentration of 0.25 per cent and an prophylaxis of thromboembolic phenome- 1 NaCl concentration of 0.66 per cent. This solu- na.' Stage I of fibrinolysis concerns the tion (0.2 ml.) was added to aqueous humor conversion of profibrinolysin to fibrinolysin (0.2 ml.), diluted to one half the original con- (see Fig. 1). Stage II of fibrinolysis con- centration, and placed in a small standard pro- cerns the breakdown of a fibrin clot into thrombin time tube. To this was added a highly purified Seeger's thrombin solution (0.1 ml.) soluble products. In the human, profibrinol- containing 300 U. per milliliter. After 30 seconds ysin is converted to fibrinolysin by the of clotting time had elapsed, a small clear plastic cleavage of a single arginyl-valine peptide piston weighing 343 mg. was slid gently down bond with formation of a two-chain fibri- the side of the tube, coming to rest on top of nolysin molecule linked by a single disul- the clot. The tubes were then placed in a clear 0 plastic water bath at 37° C. so that the lower fide bond. level of the piston and the black mark on the The following experiment was devised tube could be matched easily at various intervals to determine the endpoint of the experiment. The to estimate the fibrinolytic activity of endpoint was arbitrarily chosen to be time it aqueous humor and to investigate the took the plastic piston to descend down to a effect of therapeutic levels of pure acetyl- black line 4 mm. from the bottom of the tube. salicylic acid on the fibrinolytic activity The rate of descent of the plastic piston was of the aqueous humor. thus a measure of the lysis time. The results are shown in Table I. Materials and methods Experiment II. Another set of lysis time ex- periments was conducted as follows: 1.5 per Five albino rabbits weighing an average of cent profibrinolysin-free Pentex fibrinogen (2 ml.) 2.5 kilograms were used as the experimental i.e., 90 per cent clottable Pentex fibrinogen

Downloaded from iovs.arvojournals.org on 10/05/2021 268 Smith and Christensen, G. Inocstigatioe Ophthalmology April 1971

Table I. Results of Experiment I Table II. Results of Experiment II

Animal (No.) Lysis time (min.) Animals (No.) \ Lysis time (min.) Treated with aspirin Treated with aspirin (every 6 hr. for 48 hr.) (every 6 hr. for 48 hr.) 1 19.7 1 44.3 2 31.5 2 57.1 3 32.0 3 55.0 4 13.2 4 47.3 5 17.5 5 52.2 Mean 22.8 Mean 51.2 S.D. 3.8 S.D. 2.4 Control Control 1 13.1 1 43.3 2 19.0 2 42.0 3 9.2 3 41.0 4 9.5 4 40.9 5 9.7 5 39.3 Mean 12.1 Mean 41.1 S.D. 1.9 S.D. 0.7

adsorbed with Darco-charcoal to remove the fibrinogen concentration \ profibrinolysin adsorbed to the negatively charged time = k Be- fibrinogen molecules0 in a solution with 2 per fibrinolysin activity / cent NaCl and 10-'-M imidazole buffer (pH 7.2) 2 was added to 2% NaCl and 10~ M imidazole cause the initial fibrinogen content in the buffer (2 ml.) (pH 7.2). This initial dilution brought experiment was the same, the conclusion the fibrinogen concentration down to 0.75 per cent, but kept the NaCl and imidazole concen- is that the fibrinolytic activity of aqueous trations the same. Distilled water (2 ml.) was humor is decreased in the aspirin-treated added to the above solution (1 ml.), bringing animals. the NaCl concentration to 0.66 per cent and the The fibrinogen used in Experiment I fibrinogen concentration as in Experiment I. High- ly purified Seeger's thrombin solution was added was Armour fibrinogen, which is known (0.2 ml. of a 300 U. per milliliter solution) to to be heavily contaminated with profibrinol- this. This lysis time was determined in Experi- ysin adsorbed to the negatively charged ment I. The results are noted in Table II. fibrinogen molecule. However, the fibrino- Repeated attempts were made to use the fibrin 7 gen used in Experiment II was highly plate method of Astrup and Mullertz to esti- purified and contained almost no adsorbed mate fibrinolytic activity in the aqueous. How- ever, as only small changes in the lysed areas profibrinolysin. The lysin times with the were obtained with large changes in adsorbed "pure" and "impure" fibrogens in the con- profibrinolysin concentration and as the plates trol animals indicate that normal aqueous tended to undergo spontaneous lysis (with the humor has considerable "activator" activity use of Armour fibrinogen), this method could i.e., "activator" being a substance capable not be used to quantitate fibrinolytic activity with any degree of accuracy of repeatability.8' ° of converting profibrinolysin to fibrinoly- sin. Examples are urokinase, streptokinase, Results and staphylokinase and, in small amounts, It can be seen from Table I that fibrinolysin (see Fig. 1). Thus, with the the fibrinolytic activity of aqueous hu- Armour fibrinogen, the lysis times were mor was substantially decreased in the quite short for the aqueous humor of the aspirin-treated animals. The lysis time control animals, indicating that most of is proportional to the fibrinogen content the adsorbed profibrinolysin was converted of the sample divided by the fibrino- to active fibrinolysin in aqueous humor. lytic activity in the sample (i.e., lysis The blood levels of aspirin in the ex-

Downloaded from iovs.arvojournals.org on 10/05/2021 Volume 10 Salicylates on aqueous ftbrinoh/sin 269 Number 4

perimental animals ranged from 9.4 to of fibrinolysis by inhibition or potentiation 30 mg. per cent. Aspirin may affect the depending on the concentration of aspirin. first and/or second stage of fibrinolysis Whatever the mechanism, the net result by inhibition or potentiation, depending is a decrease in the fibrinolytic activity on the concentration of aspirin in the with the plasma concentration range used aqueous humor. As aqueous humor con- in this experiment. The essential question tains predominantly "activator" and little that remains to be answered with regard preformed fibrinolysin, one can conclude to the treatment of traumatic hyphema that the major effect of aspirin must be is whether fibrinolytic activity in the inhibition of this "activator" in order to aqueous humor or in the plasma is the account for the dramatic prolongation of major factor leading to a secondary bleed- the lysis time in the aspirin-treated ani- ing episode, or whether there is some mals. Aspirin also inhibits certain vitamin in clot stability that predisposes to sec- K-dependent clogging factors, i.e., factors ondary bleeding; also, whether the clot II, VII, IX, and X. occluding the wounded vessel is most like- As aqueous humor contains "activator," ly to be affected by plasma factors or this activity may be expressed in terms aqueous factors. It is also not clear whether of a known activator. Lysis time data with it is necessary for salicylates to enter the the use of various known concentrations anterior chamber to exert their effect. We of urokinase was determined experimental- are currently examining these problems in ly. The mean lysis tune for 0.1 ml. of greater detail. aqueous in the control animals was found REFERENCES to be equivalent to 1.7 U. of urokinase. 1. Milstein, B. A.: Traumatic hyphema—a study In the case of aspirin-treated animals, of 83 consecutive cases. In press. the mean lysis time per 0.1 ml. of aqueous 2. Ungar, C, Damgaard, E., and Hummel, F. is equivalent to 0.1 U. of urokinase. The P.: Action of salicylates and related drugs on results in Tables I and II were calculated inflammation, Amer. J. Physiol. 171: 545, to be statistically significant (p <0.05). 1952. 3. Nanninga, L. B.: Personal communication. Discussion 4. Menon, I. S.: Aspirin and blood fibrinolysis, Lancet 1: 364, 1970. Normal aqueous humor contains con- 5. Summaria, L., Hsieh, B., and Robbins, K. siderable "activator" activity and little pre- C: The specific mechanism of activation of formed fibrinolysin. In animals treated human plasminogen to plasmin, J. Biol. Chem. with aspirin, the over-all net fibrinolytic 242: 4279, 1967. 6. Maxwell, R. W., Nickel, V. S., and Lewan- activity of aqueous humor was substantial- dowski, V.: Preparation of plasminogen-de- ly reduced. As aqueous humor contains ficient fibrinogen and thrombin, Biochem. predominantly "activator" activity and lit- Biophys. Res. Comtnun. 7: 50, 1962. tle preformed fibrinolysin, one can conclude 7. Astrup, T., and Mullertz, S.: The fibrin that the major effect of aspirin in the plate method for estimating fibrinolytic ac- tivity, Arch. Biochem. Biophys. 40: 346, 1952. aqueous humor must be inhibition of this 8. Blix, S.: Studies on the fibrinolytic system in as yet unidentified "activator," in order the euglobulin fraction of human plasma, to account for the dramatic prolongation Scand. J. Clin. Lab. Invest. (Suppl.) 58, of the lysis time with aspirin-treated ani- 1961. mals (approximately an 89 per cent in- 9. Nanninga, L. B.: Rapid determination of pro- fibrinolysin (plasminogen) in plasma, Thromb. crease in duration of the lysis time). Aspi- Diath. Haemorr. 17: 9, 1967. rin may affect the first and/or second stage

Downloaded from iovs.arvojournals.org on 10/05/2021