inhibition of alkali-induced corneal ulceration and perforation by a

Investigative Ophthalmology & Visual Science, Vol. 31, No. 1, January 1990Copyright © Association for Research in Vision and Ophthalmology

Inhibition of Alkali-Induced Corneal Ulcerotionand Perforation by a Thiol Peptide

Frank PV Burns,* Robert D. Gray,t and Christopher A. Paterson*

Corneal ulceration and perforation following a severe alkali burn occur as a consequence of collagendestruction by locally released enzymes. A thiol peptide, which recently was shown to be a potentinhibitor of corneal collagenase in vitro, was tested in alkali-burned rabbit corneas to determine itseffectiveness in inhibiting corneal ulceration. Following a standard alkali burn to one eye of eachrabbit, ten animals were treated topically six times daily and subconjunctivally one time daily with a 1mM solution of the peptide for a period of 3 weeks. A control group of ten rabbits was administeredvehicle only using the same regimen as the experimental group. Corneal ulceration occurred in ten outof ten of the control eyes and seven out of ten progressed to perforation. The experimental groupdemonstrated ulcerations in four out of nine animals, only one of which was deep (one of nine), and noperforations. There was no significant difference when comparing the onset of ulceration between thetwo groups, but the difference was significant when comparing the total number of ulcerations (0.02< P < 0.05), deep ulcerations (0.01 < P < 0.02) and perforations (0.001 < P < 0.01) between the twogroups. Histologic examination of the corneas after 3 weeks of treatment revealed that the experimen-tal, thiol-treated corneas that did not ulcerate contained relatively few PMNs, whereas the controlcorneas demonstrated a marked inflammatory infiltrate in the form of PMNs, most notably at sites ofcorneal ulceration. These findings demonstrate that a synthetic thiol peptide inhibits alkali-inducedcorneal ulceration and perforation in vivo. Invest Ophthalmol Vis Sci 31:107-114,1990

The alkali-burned eye presents a major therapeuticchallenge to the ophthalmologist.1 Depending on theclinical condition of the cornea post-burn, numeroustherapeutic approaches are used in treating the al-kali-injured eye, including steroids, heparin, colla-genase inhibitors, contact lenses, fibronectin, con-junctival flaps and corneal transplantation.1'2 Re-cently, several experimental therapeutic approaches,including sodium citrate and sodium ascorbate, havebeen advocated.3"5 Numerous studies in the past havedemonstrated that the corneal damage following analkali burn is due to the destruction of corneal colla-gen by locally released enzymes.6"8 Corneal collage-nase has been implicated as one of the principle en-

From the Departments of ""Ophthalmology and Visual Sciencesand tBiochemistry, University of Louisville School of Medicine,Louisville, Kentucky.

Presented in part at the International Congress of Eye Researchmeeting, San Francisco, California, September, 1988.

Supported by NIH Grants EY-06918 (CP), 1F32-EY-06048-01A1 (FB) and AM-31364 (RDG and A. F. Spatola), the KentuckyLions Eye Research Foundation and an unrestricted grant fromResearch to Prevent Blindness, Inc.

Submitted for publication: November 28, 1988; accepted June12, 1989.

Reprint requests: Christopher A. Paterson, PhD, DSc, Professorand Director of Research, Kentucky Lions Eye Research Institute,301 E. Muhammad AH Boulevard, Louisville, KY 40202.

zymes involved in the destruction of the cornea.910

Therefore, a well justified approach to therapy hasbeen to inhibit collagenase, thus preventing this de-struction.

Inhibitors of collagenase have been used for about20 years in the treatment of alkali-induced cornealulceration. Many compounds have been tested fortheir efficacy in preventing corneal ulceration andperforation in this setting but, in general, they havebeen found to be relatively ineffective in vivo.1 Com-pounds that have been used as collagenase inhibitorsinclude acetylcysteine," cysteine,1112 Na2EDTA13

and penicillamine.14 Acetylcysteine (Mucomyst) isapproved for use as a mucolytic agent only, but isused topically as a collagenase inhibitor in humancorneal alkali burns.1'2 Tetracycline compounds havebeen recently shown to inhibit alkali-induced cornealulceration, presumably by inhibiting collagenase.15

Recently we have shown that a newly developed/3-mercaptomethyl tripeptide (Fig. 1) is highly activein inhibiting purified collagenase from alkali-burnedrabbit corneas, in vitro.16 In comparison to othercompounds used in the treatment of alkali-inducedcorneal ulceration, the thiol compound was demon-strated to be far greater in potency in inhibiting cor-neal collagenase.

In view of our in vitro findings, the present studywas undertaken to evaluate the efficacy of the thiol

107

Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933150/ on 04/12/2018

108 INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / January 1990 Vol. 31

HSCH2CHCO-NHCH-CONH-CH-CONH2

CH2 CH2 CH3

CH

(CH3)2

HSCH2-Leu-Phe-Ala-NH2 (Thiol Peptide)

Fig. 1. Structure of the thiol peptide used in this study. Thecompound is a |8-mercaptomethyl tripeptide developed specificallyas an inhibitor of collagenase. The thiol peptide has previously beenshown to be a potent inhibitor of corneal and synovial collagenaseand has been demonstrated to be far more potent than its carbox-yalkyl analogue.

peptide in preventing alkali-induced corneal ulcer-ation and perforation in rabbits.

Materials and Methods

General Considerations

Twenty New Zealand Dutch strain albino rabbitsof both sexes weighing between 2.0 and 2.5 kg wereanesthetized by intramuscular injection of 10 mg/kgxylazine and 37.5 mg/kg ketamine HC1. One eye ofeach animal received topical tetracaine and wasproptosed for alkali burning. A sharply defined 12.7mm corneal burn was produced by pipetting 0.5 mlof 2N sodium hydroxide into a plastic well heldfirmly against the cornea for 60 sec.3 The interior ofthe well and the surface of the eye were then thor-oughly irrigated with saline for 5 sec. Irrigation wasthen continued for 5 sec after removal of the well.Erythromycin ophthalmic ointment (Pharmafair,Inc., Hauppage, NY; 0.5%) was immediately appliedto the eye and once daily thereafter. Rabbits wererandomly assigned into two groups often, one groupbeing treated with the peptide and the other beingtreated with vehicle only. The methods of this studywere in accord with the ARVO Resolution on the Useof Animals in Research.

Solution Preparation

Preparation of the thiol peptide, shown in Figure 1,has been described previously.17 The peptide was dis-solved in 95% ethanol containing 1 raM acetic acid.The solution was then dissolved in Adsorbotear with-out EDTA or thimerosol (Alcon, Inc., Fort Worth,TX) to give a final peptide concentration of 1 mM.This solution was then stored at 4°C until ready foruse. Ethanol concentration varied between 2 and 5%in the final solution depending upon the concentra-tion of the stock solution of the peptide. Fresh solu-

tions of the peptide were made every 24 to 48 hr.Concentration of the peptide in Adsorbotear wasconfirmed by HPLC monitoring of the solution forrepresentative peaks. Equal volumes of ethanol con-taining 1 mM acetic acid were added to a secondsolution of Adsorbotear for use as the control vehiclewhich was also kept at 4°C.

Peptide preparations used for subconjunctival in-jections were made immediately prior to the time oftreatment. The thiol peptide was dissolved in 95%ethanol containing 1 mM acetic acid. The solutionwas added to normal saline to form a 1 mM concen-tration of peptide in a total volume of 0.5 ml. Ethanolconcentration varied between 2 and 5% in the finalsolution, depending upon the concentration of thestock solution of the peptide. Equal volumes of 95%ethanol containing 1 mM acetic acid were dissolvedin normal saline for use as the control vehicle. Solu-tions were used immediately after preparation.

Treatment Regimen

Animals were treated topically with one drop ofeither the peptide or the vehicle only in the alkali-burned eye six times daily, every 2 hr from 8 AM to 6PM. Animals were returned to their cages immedi-ately after treatment. At 8 PM daily, rabbits weretreated with subconjunctival administration of pep-tide or the vehicle only. Following topical anesthesiawith tetracaine, injections of 0.5 ml of the solutionswere made subconjunctivally at the twelve o'clockposition of each alkali-burned eye using a sterile tu-berculin syringe and a 30-guage needle. All animalswere noted to have subconjunctival blebs in the supe-rior fornix following injection. At 9 PM daily, topicalerythromycin ointment was applied to the alkali-burned eye of each rabbit. This treatment regimenwas continued for 3 weeks following the alkali burn.

Clinical Observations and Tissue Analysis

External examinations of each eye were done oncedaily during the entire study. Detailed, double-masked, slit-lamp examinations of each rabbit wereperformed every other day initially for the first 10days and then every day following the onset of cor-neal ulceration. Corneas were examined for the pres-ence of corneal defects, ulceration, perforation, vas-cularization or infection. Each cornea was assigned aclinical score according to the severity of corneal ul-ceration by classification into the following groups:(1) no ulceration: score = 0; (2) superficial ulceration(ulcers limited to the anterior one-third of the cor-nea): score = 1; (3) moderate ulceration (ulcers ex-tending to the middle one-third of the cornea): score= 2; (4) deep ulceration (ulcers extending to the pos-


No. 1 INHIBITION OF ALKALI-INDUCED CORNEAL ULCEfWION / Burns er ol 109

terior one-third of the cornea): score = 3; (5) desce-metocele: score = 4; (6) perforation: score = 5. Cor-neal features were documented by macrophotogra-phy using an Olympus Zuiko Medical Macro camera.Photography was done twice weekly initially for thefirst 10 days and then as needed to document anysignificant change in pathology. Topical and subcon-junctival treatments were continued until the ani-mals were sacrificed by pentobarbital overdose viaear vein injection on day 21 of the experiment. Ani-mals which developed corneal perforations were sac-rificed at the time the perforation was discovered.

After sacrifice, the anterior chamber of the alkali-burned eye was entered with a scalpel blade, and theentire cornea was excised from the eye with cornealscissors. Corneas were immediately placed in 10%formalin. Following formalin fixation, corneas wereembedded in paraffin and stained with hematoxylinand eosin in preparation for routine histologic exami-nation. One animal in the experimental group wassacrificed on the fourth day of the experiment aftersuffering a spinal injury; the data were excluded fromthe final analysis of the results.

Statistics

The results were analyzed for significance using theChi-square test with Yate's correction and the stu-dent t-test. The clinical severity of ulcers was com-pared for significance using the Mann-Whitney Utest.

Results

The severe alkali burns employed in this study re-sulted in corneal stromal opacities extending to thelimbus of each burned eye. None of the eyes becameinfected and none of the corneas became significantlyvascularized. Clinical examination of the subcon-junctival injection sites of both the experimental andcontrol eyes showed no signs of necrosis or tissuedamage and there was no difference in clinical ap-pearance of the injection sites between the controland inhibitor groups.

Table 1. Analysis of clinical observations for cornealalkali burns at end of experiment

Table 2. Analysis of degree of corneal ulcerationand perforation occurring after experimentalalkali burns

Experimental Controls(n = 10)

Nonulcerated corneasTotal vascularizationsSuperficial stromal ulcersModerate stromal ulcersDeep stromal ulcersDescemetocelesPerforations

5021100

0020107

Treatmentgroup

Control(n = 10)

Thiol(n = 9)

* 0.02 < P < 0.05.to.oi <p<om.±0.001 </><0.01

Totalulcers

10

4*

Totaldeep ulcers

8

It

Totalperforations

1

o*

A striking difference was noted between the experi-mental and control groups in the development ofcorneal ulcerations and perforations. The controlgroup demonstrated ulcerations in ten (100%) oftencorneas compared with four (44%) of nine corneas inthe experimental group (0.02 < P < 0.05). A total ofeight (80%) of the ten control corneas progressed todeep ulceration or descemetocele formation followedby seven (70%) often progressing to perforation. Theexperimental group showed one (11%) of nine pro-gressing to deep ulceration (0.01 < P < 0.02) andnone progressing to descemetocele formation or per-foration (0.001 < P < 0.01). Table 1 summarizes theoverall results from the two groups and Table 2 illus-trates the difference between the, two groups in termsof incidence of overall ulceration, deep ulcerationand perforation.

The mean length of time for appearance of cornealulceration was 14 ± 4 days (range 8-19) in the controlgroup and 16 ± 1 days (range 15-17) in the experi-mental group. There was no significant differencewhen comparing the time of onset of ulceration. Thetendency was for the control group to develop ulcerssooner and, once established, to progress rapidly todeep ulcers and descemetoceles. The mean length oftime for the development of corneal perforation was17 ± 2 days (range 15-20) in the control group. Asmentioned previously, the experimental group dem-onstrated no perforations.

Figure 2 illustrates data obtained from clinical ob-servations during the course of the study. Exam 0represents the day of the alkali burn. The final exami-nation (exam 12) occurred on day 20 post-burn. Thenumber of control eyes decreased as they were re-moved secondary to perforation. The depth of ulcerswas represented by a numerical score, as explained inMethods, above. The depth of ulceration progres-sively worsened in the control group from exam 5through exam 12, while the thiol treatment signifi-cantly reduced the progression of ulceration. The twogroups were significantly different in average clinical


110 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / January 1990 Vol. 31

£oo

CO

"5o

'c

oa)oo>

4--0 - No Ulcer1 - Superficial2 - Moderate3 - Deep4 - Descemetocele5 - Perforation

.A

n=5

n=6An=4

2 -

1 • •

At Exom 0 -

Control (n=

Thiol (n=9)

A — AO—O

ControlThiol

Fig. 2. The progression ofclinical changes of the al-kali-burned corneas overthe course of the study.Exams were done every 2days initially and then dailyfrom exam 6 through exam12. The X-axis correspondsto the exam number. Theday of the alkali burns wasdesignated as exam 0whereas exam 12 corre-sponded to day 20 post-burn, which was the finalday of the study. The "n"values decreased in the con-trol group towards the endof the experiment as ani-mals were removed becauseof corneal perforations. TheY-axis corresponds to theaverage clinical score of thetwo groups. This score wascalculated by averaging thescore of all the animals ineach group based upon the

following clinical scoring system: no ulcers, score = 0; superficial ulcers (depth to the anterior one-third), score = 1; moderate ulcers (depth tothe middle one-third), score = 2; deep ulcers (depth to the posterior one-third), score = 3; descemetoceles, score = 4; perforations, score = 5.Corneal ulcers that perforated (score = 5) before the end of the experiment continued to be included in the analysis of the ulcer severity untilthe end of the experiment. The two groups were significantly different when comparing the severity of ulceration statistically using theMann-Whitney U test (P < 0.01).

-o. .o- •O'-h -

0 2 4 6 8 10

Exams Following Corneal Alkali Burns

12

severity when comparing them statistically using theMann-Whitney U test (P < 0.01).

Light Microscopy

All of the corneas were examined by light micros-copy following the clinical aspect of the study. It wasnoted that none of the corneas in either group showedsignificant reepithelialization. The degree of vascular-ization varied between corneas, but it was limited tothe periphery, and there was no notable difference invascularization between the two groups.

The anterior stroma and mid-stroma surroundingthe ulcerations and perforations, in general, were in-filtrated with inflammatory cells. In comparing ul-cerating and nonulcerating corneas from the twogroups, it was of interest that corneas in the experi-mental group which were not ulcerating (five of nine)were essentially devoid of polymorphonuclear leuko-cytes (PMNs) in the central area of the cornea. Theseexperimental corneas also showed a total loss of ker-atocytes but essentially no loss of stromal substance.The other corneas in the experimental group hadvarying degrees of PMN infiltration which seemed tocorrelate with the depth of ulceration present. A rep-resentative nonulcerating cornea from the experi-mental group which demonstrates a lack of PMNinfiltration is seen in Figure 3A.

The control corneas demonstrated a markedamount of PMN infiltration throughout six of the tencorneas. It was noted that all four control corneas thatdid not have generalized PMN infiltration progressedto perforation, and three of the four perforated by day15 and the fourth on day 17 of the study. The moststriking infiltration in these corneas took place at sitesof corneal ulceration. A representative area of PMNinfiltration in a nonperforated control cornea isshown in Figure 3B. It also was noted that the threeeyes in the control group that did not perforate hadPMN infiltration throughout the cornea, varyingfrom mild in one to marked in two.

Discussion

Ocular alkali burns create a complex therapeuticchallenge for the ophthalmologist.1 Alkali injuriescan progress to corneal ulceration and perforation,thus causing loss of vision and permanent visual dis-ability. The control and prevention of this pathologicprogression have been the focus of numerous studies.The treatment approaches that have been used intreating alkali injuries have been thoroughly re-viewed.12 Treatment modalities that have been suc-cessful in reducing the incidence of corneal ulcerationand perforation in the alkali-burned cornea fall intofive categories: (1) agents which inhibit the locomo-


No. 1 INHIBITION OF ALKALI-INDUCED CORNEAL ULCEIWION / Burns er ol 111

Fig. 3. Representativehistologic sections from al-kali-burned corneas. (A)Nonulcerated cornea fromthe experimental group(rabbit #776) following 3weeks of topical and sub-conjunctival treatment with1 mM thiol peptide. Thecornea is essentially devoidof cells and there are nosigns of ulceration (hema-toxylin-eosin, original mag-nification X100). (B) Ulcer-ated cornea from the con-trol group (rabbit #772)following 3 weeks of topicaland subconjunctival treat-ment with vehicle only. Theanterior and mid-stroma areinfiltrated with PMNs andother inflammatory cells atthe site of an ulcer (hema-toxylin-eosin, original mag-nification XI00).

A

B

tion and respiratory burst of PMNs, such as sodiumcitrate51819; (2) agents that prevent reepithelializationand the infiltration of PMNs into the cornea, such asglued-on contact lenses and cyanoacrylate tissue ad-hesives20"22; (3) agents that inhibit collagenase, suchas acetylcysteine1112 and tetracycline15; (4) anti-in-flammatory agents, such as medroxyprogesterone23;and (5) agents that increase the production of colla-gen within the cornea, such as ascorbic acid.3'418

Collagenase has been identified as a major contrib-utor to the breakdown of the cornea following suchan injury.910 Compounds that have been used pre-viously as collagenase inhibitors in treating alkali

burns in humans have been relatively ineffective invivo1 and, thus, work has continued to produce moreeffective treatment approaches. Our findings in thepresent study indicate that a recently developed pep-tide derivative designed to inhibit collagenase has asignificant impact upon reducing alkali-induced cor-neal ulceration and perforation at markedly lowerconcentrations than previous compounds. This effectof the peptide almost certainly can be ascribed to itsability to interfere with the collagenase-mediated de-struction of the corneal stroma.

Synthetic inhibitors of collagenase have been de-signed specifically to bind with collagenase and, thus,


112 INVESTIGATIVE OPHTHALMOLOGY G VISUAL SCIENCE / January 1990 Vol. 31

block the enzyme's collagen-cleaving action.1724'25

The mechanism by which the peptide presumablyfunctions is by binding to the substrate recognitionsite of the enzyme and coordinating with catalyticallyessential Zn2+ at the active site.26 Development ofthese peptides is based upon strategies used to de-velop potent inhibitors of angiotensin-converting en-zyme.26 Previous studies have shown that these com-pounds are highly potent inhibitors of collagenasesfrom several sources including: (1) pig synovial colla-genase17'25; (2) rabbit V-2 tumor collagenase17'25; and,most recently, (3) purified corneal collagenase fromalkali-burned rabbit corneas.16 Thiol peptides, suchas the one used in this study (Fig. 1), have been dem-onstrated to be far greater in potency than their N-carboxyalkyl counterparts.16'25 In comparison toother compounds that have been demonstrated toreduce the incidence of alkali-induced corneal ulcer-ation, the thiol peptide was shown to be far greater inpotency in inhibiting corneal collagenase in vitro.This included comparison to acetylcysteine, cysteine,sodium citrate, sodium ascorbate and tetracyclinecompounds.16 For this reason, the thiol peptide wasevaluated in vivo for its efficacy in reducing alkali-in-duced corneal ulceration.

In the present study we have shown one particularthiol peptide to be a potent inhibitor of corneal ulcer-ation and perforation following a severe alkali burnin rabbits. The burn was very severe; 70% of the con-trol eyes perforated within 3 weeks. The area of theburn extended to the limbus, which prevented signifi-cant vascularization or reepithelialization. Collagendestruction occurred rapidly once ulceration ap-peared. In the experimental (thiol-treated) groupthere was a significant reduction in the progression ofulceration and perforation. Overall, the majority ofcorneas in the experimental group did not ulcerateand only one progressed to deep ulceration; this was amarked difference in comparison to the extreme pa-thology noted in the control group. This observationstrongly supports the contention that corneal ulcer-ation following the release of collagenases is effec-tively inhibited by a synthetic collagenase inhibitor.

Qualitative examination of the histologic sectionsof the ulcerated eyes showed a marked PMN infiltra-tion moving from the limbal region inward (Fig. 3B).Infiltration of the cornea with acute and chronic in-flammatory cells is a well described phenomenon oc-curring during corneal ulceration, with the PMNbeing the predominant cell type following alkaliburns.20'27'28 It was of some interest to note that ex-amination of nonulcerated experimental eyes re-vealed relatively few inflammatory cells in the corneafollowing treatment with the thiol peptide (Fig. 3A).

In six of the ten control corneas, histological exami-nation revealed PMNs throughout the stroma. Theremaining four corneas, all of which perforated, wereinfiltrated by a large number of PMNs peripherallyaround the sites of perforation, but PMNs were notseen centrally. It is likely that ulceration and perfora-tion in these corneas took place at the corneal periph-ery as a result of the accumulation of PMNs andrelease of their lysosomal enzymes at that site. Per-haps PMN infiltration into the central stroma did notoccur because early perforation did not afford thephysiological environment or time frame for chemo-taxis to take place. It is also possible that the lack ofadequate vascularization reduced PMN infiltration,although the finding that six of ten control corneaswere diffusely infiltrated even without significant vas-cularization argues against this theory.

Although the primary impact of the thiol peptide isto directly block the destruction of stromal collagenby collagenase, there also might be an indirect effectupon PMN chemotaxis. Infiltration of the PMNs intoa connective tissue matrix may require collagenaseand other proteases to cleave stromal collagen fibrilssurrounding the PMN, thus allowing more easy pen-etration and chemotaxis of the PMNs to occur. Thus,inhibition of collagenase might result in reducedPMN infiltration, thereby enhancing the preventionof ulceration and perforation in these corneas. Whilethis speculation is based only upon qualitative histo-logic observation, the concept that PMN infiltrationwas reduced in the alkali-burned cornea by treatmentwith a potent inhibitor of collagenase is interesting; asimilar observation was made previously when rab-bits with alkali-burned corneas were treated with tet-racycline systemically.15

Furthermore, it has been proposed that alkali-burned collagen fragments may be a mediator forPMN influx.29 It has been demonstrated that the su-pernatant from alkali-burned commercial collagensis chemotactic for PMNs and stimulates PMN respi-ratory burst; it was suggested that following alkali-burning of the cornea, peptide fragments of cornealcollagen could diffuse away from the cornea acrossthe surfaces or through the stroma to the limbal areawhere interaction with serum albumin could facili-tate PMN locomotion. After invading the corneathrough the damaged collagen matrix, the PMNsmight then release their proteolytic enzymes, whichwould lead to further stormal destruction. Perhapsthe thiol peptide has interfered with this feedbackmechanism by inhibiting the proteolytic enzymes re-leased by the PMN, thus preventing further genera-tion of chemotactic collagen fragments.

Collagenolytic activity has been found to be pro-


No. 1 INHIBITION OF ALKALI-INDUCED CORNEAL ULCERATION / Burns er ol 113

duced by non-PMN sources, such as from epithelialcells3031 and fibroblasts.3233 Initially, lysis of epithe-lial cells and fibroblasts would cause release of theirenzymes and it is likely that the peptide inhibitedmetalloproteases, which were released by these cells.However, because these cells were destroyed by thealkali burn, it is unlikely that they were a continuoussource of proteolytic enzymes following initial re-lease. Thus, the peptide's effect most likely was notdue solely to inhibition of enzymes derived fromsources other than PMNs. Although the source ofcollagenase in these corneas is not clear, it is likelythat collagenase and proteases released by PMNswere inhibited by the peptide and that the overallprotective effect of the peptide was due to this inhibi-tion. The corneas treated with the peptide alsoshowed a total loss of keratocytes but essentially noloss of stromal substance. This finding indicates thatstromal collagen was protected from degradation bythe thiol peptide.

In summary, this study has shown that a thiol pep-tide, designed to inhibit collagenase, reduces the inci-dence of corneal ulceration and perforation in alkali-burned rabbit corneas. The treatment of the alkali-burned cornea should be a combined approach usingagents that have different mechanisms of action. Be-cause the PMN plays a major role by releasing de-structive enzymes, including collagenase, the inhibi-tion of its locomotion and activation with an agentsuch as citric acid is a fundamental therapeutic com-ponent. Second, ascorbic acid treatment, which re-places the depleted levels of the co-factor necessaryfor collagen production, is also regarded as a crucialcomponent. Finally, a potent metalloprotease inhibi-tor should be a component of therapy because of theneed to inhibit collagenase and other proteases re-leased into the cornea. The thiol peptide used in thisstudy could be a candidate for this aspect of treat-ment.

Key words: cornea, alkali burn, corneal ulceration, thiolpeptide, collagenase inhibitor

Acknowledgments

The authors thank Ms. Robyn Mclntire for assistance inthe preparation of this manuscript. We also wish to thankDr. A. F. Spatola for advice in the chemical synthesis aspectof this study, and Ms. Sandy Hensley and Mr. Greg Shirkfor assistance in the care of animals.

References

1. McCulley JP: Chemical injuries. In The Cornea: ScientificFoundations and Clinical Practice, Smolin G and Thoft RA,editors. Boston/Toronto, Little, Brown, and Co., 1983, pp.422-435.

2. Pfister RR: Chemical corneal burns. Int Ophthalmol Clin24:157, 1984.

3. Levinson R, Paterson CA, and Pfister RR: Ascorbic acid pre-vents corneal ulceration and perforation following experimen-tal alkali burns. Invest Ophthalmol 15:986, 1976.

4. Pfister RR, Paterson CA, and Hayes SA: Topical ascorbatedecreases the incidence of corneal ulceration after experimen-tal alkali burns. Invest Ophthalmol Vis Sci 17:1019, 1978.

5. Pfister RR, Nicolaro ML, and Paterson CA: Sodium citratereduces the incidence of corneal ulcerations and perforationsin extreme alkali-burned eyes: Acetylcysteine and ascorbatehave no favorable effect. Invest Ophthalmol Vis Sci 21:486,1981.

6. Brown SI, Weller CA, and Wasserman HE: Collagenolytic ac-tivity of alkali-burned corneas. Arch Ophthalmol 81:370,1969.

7. Itoi M, Gnadinger MC, Slansky HH, Freeman MI, and Dohl-man CH: Collagenase in the cornea. Exp Eye Res 8:369, 1969.

8. Brown SI, Weller CA, and Akiya S: Pathogenesis of ulcers ofthe alkali-burned cornea. Arch Ophthalmol 83:205, 1970.

9. Berman M, Dohlman CH, Gnadinger M, and Davison P:Characterization of collagenolytic activity in the ulceratingcornea. Exp Eye Res 11:255, 1971.

10. Gordon JM, Bauer EA, and Eisen AZ: Collagenase in humancornea: Immunologic localization. Arch Ophthalmol 98:341,1980.

11. Slansky HH, Berman MB, Dohlman CH, and Rose J: Cysteineand acetylcysteine in the prevention of corneal ulcerations.Ann Ophthalmol 2:488, 1970.

12. Brown SI, Akiya S, and Weller CA: Prevention of the ulcers ofthe alkali-burned cornea: Preliminary studies with collagenaseinhibitors. Arch Ophthalmol 82:95, 1969.

13. Brown SI and Weller CA: Collagenase inhibitors in preventionof ulcers of alkali-burned cornea. Arch Ophthalmol 83:352,1970.

14. Francois J, Cambie E, Feher J, and Van Den Eeckhout E:Collagenase inhibitors (penicillamine). Ann Ophthalmol5:391, 1973.

15. Seedor JA, Perry HD, McNamara TF, Golub LM, Buxton DF,and Guthrie DS: Systemic tetracycline treatment of alkali-in-duced corneal ulceration in rabbits. Arch Ophthalmol105:268, 1987.

16. Burns FR, Stack MS, Gray RD, and Paterson CA: Inhibitionof purified collagenase from alkali-burned rabbit corneas. In-vest Ophthalmol Vis Sci 30:16, 1989.

17. Gray RD, Miller RB, and Spatola AF: Inhibition of mamma-lian collagenases by thiol-containing peptides. J Cell Biochem32:71, 1986.

18. Pfister RR, Haddox JL, and Lamb KM: Citrate or ascorbate/citrate treatment of established corneal ulcers in the alkali-in-jured rabbit eye. Invest Ophthalmol Vis Sci 29:1110, 1988.

19. Pfister RR, Haddox JL, Dodson RW, and Deshazo WF: Poly-morphonuclear leukocytic inhibition by citrate, other metalchelators and trifluoperazine: Evidence to support calciumbinding protein involvement. Invest Ophthalmol Vis Sci25:955, 1984.

20. Kenyon K, Berman M, Rose J, and Gage J: Prevention ofstromal ulceration in the alkali-burned rabbit cornea by glued-on contact lens: Evidence for the role of polymorphonuclearleukocytes in collagen degradation. Invest Ophthalmol Vis Sci18:570, 1979.

21. Dohlman CH, Slansky HH, Laibson PR, Gnadinger MC, andRose J: Artificial corneal epithelium in acute-alkali burns. AnnOphthalmol 3:357, 1969.


114 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / January 1990 Vol. 31

22. Dohlman CH, Carroll JM, Richard J, and Refojo MF: Furtherexperience with glued-on contact lens (artificial epithelium).Arch Ophthalmol 83:10, 1970.

23. Newsome DA and Gross J: Prevention by medroxyprogester-one of perforation in the alkali-burned rabbit cornea: Inhibi-tion of collagenolytic activity. Invest Ophthalmol 16:21, 1977.

24. Gray RD, Saneii HH, and Spatola AF: Metal binding peptideinhibitors of vertebrate collagenase. Biochem Biophys ResCommun 101:1251, 1981.

25. Gray RD, Darlak K, Miller RB, Stack MS, and Spatola AF:Inhibition of mammalian collagenase by thiol and N-carboxy-alkyl peptide derivatives. FASEB J 2:A345 (Abstract), 1988.

26. Cushman DW, Cheung HS, Sabo EF, and Ondetti MA: Designof potent competitive inhibitors of angiotensin-converting en-zyme: Carboxyalkanoyl and mercaptoalkanoyl amino acids.Biochemistry 16:5484, 1977.

27. Brown SI and Weller CA: The pathogenesis and treatment ofcollagenase-induced diseases of the cornea. Trans Am AcadOphthalmol Otolaryngol 74:375, 1970.

28. Foster CS, Zelt RP, Mai-Phan T, and Kenyon KR: Immuno-suppression and selective inflammatory cell depletion: Studiesof a guinea pig model of corneal ulceration after ocular alkaliburning. Arch Ophthalmol 100:1820, 1982.

29. Pfister RR, Haddox JL, Dodson RW, and Harkins LE: Alkali-burned collagen produces a locomotory and metabolic stimu-lant to neutrophils. Invest Ophthalmol Vis Sci 28:295, 1987.

30. Brown SI and Weller CA: Cell origin of collagenase in normaland wounded corneas. Arch Ophthalmol 83:74, 1970.

31. Pfister RR, McCulley JP, Friend J, and Dohlman CH: Colla-genase activity of intact corneal epithelium in peripheral alkaliburns. Arch Ophthalmol 86:308, 1971.

32. Hook RM, Hook CW, and Brown SI: Fibroblast collagenase:Partial purification and characterization. Invest Ophthalmol12:771, 1973.

33. Newsome DA and Gross J: Regulation of corneal collagenaseproduction: Stimulation of serially passaged stromal cells byblood mononuclear cells. Cell 16:895, 1979.


inhibition of alkali-induced corneal ulceration and perforation by a

Documents