the role of contact lenses, trauma, and langerhans cells in a

The Role of Contact Lenses, Trauma, and LangerhansCells in a Chinese Hamster Model of AcanthamoebaKeratitis

Franciscus van Klink,* Hassan Alizadeh,^ YuGuang He,\ Jessamee A. Mellon,^Robert E. Silvany,-\ James P. McCulley,\ and Jerry Y. Niederkorn^

Purpose. To determine the role of contact lenses, corneal trauma, and Langerhans cells in thedevelopment of keratitis caused by Acanthamoeba organisms in Chinese hamsters.

Methods. Various methods were used to induce corneal infections in Chinese hamsters, includ-ing application of parasite-laden contact lenses. The role of corneal epithelial defects in pro-moting parasite binding was examined in vitro in a microscopic binding assay. The role ofcorneal abrasion in the development of Acanthamoeba keratitis was also examined in Chinesehamsters exposed to parasite-laden contact lenses. Other experiments evaluated the effect ofinfiltrating Langerhans cells on the incidence and severity of Acanthamoeba keratitis.

Results. Corneal epithelial defects promoted extensive parasite binding to abraded corneascompared to intact, nonabraded counterparts. Corneal abrasion was absolutely necessary forthe induction of Acanthamoeba keratitis in hamsters infected with contaminated contact lenses.Infection was never detected unless the corneas were abraded before exposure to parasite-laden contact lenses. The presence of Langerhans cells in corneas prevented the developmentof Acanthamoeba keratitis.

Conclusions. The highest incidence of Acanthamoeba keratitis occurs in corneas expressing epi-thelial defects and exposed to parasite-laden contact lenses. The presence of Langerhans cellsin corneas exposed to parasite-laden contact lenses prevents the development of Acanthamoebakeratitis. Invest Ophthalmol Vis Sci 1993;34:1937-1944.

./lcanthamoeba keratitis is a serious sight-threateningdisease caused by a ubiquitous, pathogenic, free-livingamoeba. Although Acanthamoeba spp. are is widely dis-tributed in the environment and can be isolated from

From the *Def)artmeiil. of Ophthalmology, Uiiden University Hospital, Leiden, TheNetherlands and Department of Ophthalmology,^ The University of TexasSouthwestern Medical Center, Dallas, Texas 75235.This project supported by an unrestricted grant from Research to Prevent Blindness,Inc. New Yorkand the foundation "Met Dondersfonds" and the foundation "DeDrie Lichten" in the Netherlands. Dr. Niederkorn is a Research to PreventBlindness, Inc. Senior Scientific Investigator.Submitted for publication: March 23, 1992; accepted July 22, 1992.Proprietary Interest Category: N.Reprint requests: Jerry Y. Niederkorn, Department of Ophthalmology, University ofTexas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, Texas75235.

swimming pools, soil, dust, reservoirs, under ice andfrom the nasopharyngeal mucosa from healthy hu-mans,1 corneal infections are very rare and mainly re-stricted to people who wear contact lenses.2 It is gener-ally accepted that the predominant risk factor forAcanthamoeba keratitis is contact lens wear.3 Consider-ing the fact that more than 24,000,000 persons in theUnited States wear contact lenses and the ubiquitousdistribution of the Acanthamoeba, one would expect ahigh incidence of Acanthamoeba keratitis. However, theoccurrence of the disease is very low. Therefore, itseems almost certain that other factors exist in thedevelopment of corneal infection. Such additional riskfactors could be identified and examined in an appro-priate animal model. Moreover, an animal model

Investigative Ophthalmology & Visual Science, May 1993, Vol. 34, No. 6Copyright © Association for Research in Vision and Ophthalmology 1937

Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933399/ on 03/20/2018

1938 Investigative Ophthalmology & Visual Science, May 1993, Vol. 34, No. 6

would also make it possible to explore the pathophysi-ology, cell biology, immunology, genetics, and therapyof this disease.

An animal model must conform to several basiccriteria to be a useful tool in long-term studies. Themost important one is that the model conform toKoch's postulates. The disease should be produced bylive, infectious parasites, and viable parasites must beisolated from diseased tissue and grown in culture.

We have been unable to establish corneal infec-tions in immunosuppressed mice and two species ofimmunocompetent rats. It was therefore concludedthat some mammalian species were more susceptibleto infection than others. Accordingly, we evaluatedthe parasite's ability to bind and penetrate full-thick-ness corneal buttons in vitro. The results indicatedthat of the eleven vertebrate species tested, only thecorneas from human, pig, and the Chinese hamsterwere susceptible to in vitro infection with Acantha-moeba castellanii.4

As predicted from the in vitro studies, pigs werefound to be susceptible to in vivo corneal infectionand could serve as a valuable model for investigatingthe pathogenesis of Acanthamoeba keratitis.5 Based onin vitro studies,4 we predicted that the Chinese ham-ster would also be susceptible to in vivo infection.

The present study shows that Acanthamoeba kerati-tis develops in Chinese hamsters exposed to contami-nated contact lenses and that this host can be used as apromising animal model of Acanthamoeba keratitis. Us-ing this model, the role of contact lenses and trauma inpromoting the infection was studied.

The involvement and the role of Langerhans cellsin the pathogenesis of Acanthamoeba keratitis was stud-ied for two reasons. First, Langerhans cells may con-tribute to corneal inflammation as has been suggestedin herpes simplex stromal keratitis.67 Second, Langer-hans cells are highly efficient antigen-presenting cellsand might accelerate the development of protectiveimmunity and thereby mitigate corneal disease. Thesetwo propositions were explored in the present study.

MATERIALS AND METHODS

Animals

Chinese hamsters were purchased from Cytogen Re-search and Development, Inc. (West Roxbury, MA).Animals were used when 4-6 wk old and all corneaswere examined before initiating the experiments toexclude animals with preexisting lesions or anomalies.AH procedures were performed on the right eye. Theleft eyes were not manipulated. All animals were han-dled in accordance with the ARVO Resolution on theUse of Animals in Research and adhered to the tenetsof the Declaration of Helsinki.

Parasite Cultivation

Acanthamoeba castellanii, originally isolated from a dis-eased human cornea, was obtained from AmericanType Culture Collection (ATCC #30868; Rockville,MD) and grown axenically in PYG (peptone-yeast-glu-cose) medium as described elsewhere.8

Modes of Corneal Exposure to Parasites

Several methods were used in an attempt to establishcorneal infections with A. castellanii. These included:(A) intrastromal injection; (B) subconjunctival injec-tion; (C) irrigation of abraded cornea with parasite-rich inocula; (D) deposition of parasite suspensionsinto the conjunctival cul-de-sac after tarsorrhaphy;and (E) application of parasite-laden "contact lens"onto intact or abraded corneal surfaces. Intrastromalinjections were performed by injecting approximately1.0 Ail of a suspension of 4 X 106 parasites/ml using avery fine 5 ix\ glass micropipette. Injections were aidedby an automatic dispenser (Hamilton Co., Inc., Whit-tier, CA) while viewing the cornea under an operatingmicroscope as described elsewhere.910 Subconjuncti-val injections were performed using 30-gauge needlesmounted on tuberculin syringes. Parasites were depos-ited into the conjunctival cul-de-sac also with a 1 mltuberculin syringe and 30-gauge needle without per-forating the conjunctiva. In each case, a PYG suspen-sion (3 X 106 parasites/ml) was used. Parasite-ladencontact lenses were prepared as described later.

Contact Lens Preparation

Contact lenses were prepared from Spectapor dialysismembrane tubing (Spectrum Medical Industries, LosAngeles, CA) using a 3 mm trephine. Sterile dialysistubing lenses (DTL) were incubated at 35°C for 24 hrwith A. castellanii in 200 /u.1 PYG medium at a concen-tration of 3 X 106 organisms/ml (50% trophozoites,50% cysts) in sterile 96-well microtiter plates (Costar,Corp., Cambridge, MA). Attachment of parasites tothe DTL was verified microscopically in a "hangingdrop" technique.4

In Vivo Corneal Infection

Chinese hamsters were anesthetized with sodium pen-tobarbital (80 mg/kg; Abbott Laboratories, NorthChicago, IL) injected intraperitoneally. Topical anes-thesia consisted of proparacaine (Alcon Laboratories,Inc., Ft. Worth, TX). Twenty-five percent of the cor-neal surface was gently abraded under an operatingmicroscope with a sterile cotton applicator before ap-plication of the parasite-laden DTL. After applicationof the DTL, the eyelids were closed by tarsorrhaphywith 10-0 Mersilene® sutures (Ethicon, Inc., Somer-ville, NJ). Control animals were treated the same waybut without abrasion of the cornea. An additional


Hamster Model of Keratitis Caused by Acanthamoeba Organisms 1939

group was exposed to sterile DTL to examine possiblebacterial infections or nonspecific corneal damage in-duced by the DTL. No topical or systemic antibioticswere used.

Clinical Observations

Sutures were removed 7 days after exposure to para-sites and the DTL were removed. The experimentaland the contralateral eyes were observed under a dis-secting microscope. Animals were observed 3-4 timeseach week throughout the course of the study, and theeyes were scored according to the criteria given in Ta-ble 1.

Culture of Parasites

A. castellanii organisms were isolated from the re-moved DTL by cultivating on lawns of Escherichia coli(ATCC #25922) cultured on non-nutrient agar. E. colicultures were observed for the development of char-acteristic "amoeba trails."411 Corneas demonstratingclinical signs of infection were examined for the pres-ence of A. castellanii organisms and bacteria. A. castel-lanii organisms were isolated by culturing cornealspecimens on lawns of E. coli as described earlier. Bac-teria were isolated by culturing corneal specimens onBBL Brucella agar (Becton Dickinson MicrobiologicalSystems, Cockeysville, MD). Bacterial identificationwas made using bacterial identification test panels(Biolog Micro Station System, Biolog Inc., Hayward,CA). At different time points of infection, animalswere killed and their corneas were cut into two equalpieces. One half of each cornea was homogenized insaline using a glass tissue grinder and the homogenatewas layered onto a lawn of E. Coli and cultured for A.castellanii. The other half of the cornea was processedfor histology.

Histology

Corneal specimens were fixed in Carson's formalin,progressively dehydrated to 95% ethanol and embed-ded in glycol methacrylate (Sorvall, Wilmington, DE).

Three /im sections were cut and stained with hematox-ylin and eosin.

Induction of Langerhans Cell Migration intothe Central Cornea

Latex beads were deposited into shallow corneal epi-thelial incisions as a simple and reproducible methodfor inducing the migration of Langerhans cells intothe center of the corneal epithelium.12 Briefly, animalswere anesthetized with Metofane (Pitman-Moore Inc.,Mundelein, IL), and sterile 1.0 /xM polystyrene latexbeads suspended in phosphate-buffered saline (pH= 7.2) were deposited into shallow incisions in thecenter of the corneal epithelium 10 days before infec-tion. Animals were infected with DTL after abrasion ofthe cornea, as described earlier. For comparison, ani-mals without latex bead treatment were similarly in-fected.

Langerhans cell migration was also induced by theintracorneal injection of interleukin 1 (IL-1), as de-scribed in detail previously.9 Approximately 1 /xl (7—12U) of recombinant mouse IL-1 (Genzyme, Cambridge,MA) was injected through a very fine glass micropi-pette of approximately 5 ix\. Injections were per-formed with the aid of an automatic dispenser whileviewing the cornea under an operating microscope.Animals were infected either 2 or 7 days after the in-jection of IL-1 as described before. Control animalswere injected with phosphate-buffered saline.

Adenosine Diphosphatase Staining

Langerhans cells were characterized by adenosine di-phosphatase (ADPase) staining. ADPase staining is aconvenient and sensitive method for demonstratingthe presence of Langerhans cells in skin13 and cor-nea.12 To count and verify migration of Langerhanscells into the central cornea, Chinese hamsters weresacrificed ten days after latex bead treatment, and 2 or7 days after IL-1 treatment. Experimental and controleyes were enucleated and corneal epithelial sheetswere prepared as previously described and stained forADPase activity.1213 Flat whole epithelial sheets were

TABLE l. Clinical Scoring Criteria for Acanthamoeba keratitis in Chinese Hamster

Epithelial Defects Stromal Edema Vascularity Stromal Opacity

0 = no vessels1 = vessels on < 25% of

circumference of cornea

0 = no defects 0 = no edema1 = ulceration, defect, 1 = edema < 25%

< 25% of surfacearea

2 = defect 25 < 50% 2 = edema 25 < 50% 2 = vessels 25 < 50%of surface area

3 = defect 50 < 75% 3 = edema 50 < 75% 3 = vessels 50 < 75%of surface area

4 = defect > 75% ofsurface area

4 = edema > 75% 4 = vessels > 75%

0 = stroma clear1 = opacity < 25% of cornea area,

pupil easy visible

2 = opacity 25 < 50%, pupilhardly visible

3 = opacity 50 < 75%, pupil notvisible

4 =viaiuic

opacity > 75%, pupil andnot visiblenot visible

iris



mounted on glass slides with Gel Mount (BiomedaCorp., Foster City, CA), and ADPase-positive cellswere counted in the limbus and in the central corneaby counting the whole field using 400 X magnificationunder a normal light microscope at six randomly cho-sen fields for each cornea. The mean value was re-corded as number of Langerhans cells/mm2.

Attachment of Acanthamoeba castellanii toChinese Hamster Corneas In Vitro

Animals were killed 10 days after latex bead treatmentof the cornea or immediately after abrasion of the cor-neal epithelium and both eyes were enucleated. Theleft, untreated eyes were used as controls. Eyes weresterilized in an iodine solution for 45 seconds andrinsed thoroughly with saline. The eyes were placedinto tapered tubes 12 mm in length made from theupper part of 200 ix\ pipette-tips. Using this method,the eyes became stuck and a "well" with the cornealsurface of the eye as the bottom was created. A 96-wellmicrotiter plate was used as an holder for the separate"wells." Parasites were labeled by a modified tech-nique of Weisman and Korn.14 Fluorescent latex bead-labeled Acanthamoeba (1 X 106 parasites/ml), wereadded to each well. After incubating at 35°C for 24 hr,the medium was removed, and each well rinsed twicewith fresh PYG medium. Eyes were gently removedfrom the tube and placed into deep chamber slideswith the epithelial side up. The corneas were then cov-ered with a coverslip so that the center of the corneawas flattened and the number of Acanthamoeba at-tached to the cornea counted under a microscope(200 X magnification). Using an ocular grid, four ran-domly chosen fields were counted for each cornealspecimen and the results were expressed as the num-ber of parasites/mm2.

Statistical Analysis

An unpaired, one-tailed Student's / test and the Wil-cox log rank test were used for statistical analysis. Aprobability value < 0.05 was considered statisticallysignificant.

RESULTS

Chinese Hamster ModelAs predicted by previous in vitro studies,4 Chinesehamsters were susceptible to in vivo corneal infectionwith A. castellanii. Of the various techniques for ex-posing corneas to viable parasites, exposure via para-site-laden contact lenses combined with abrasion ofthe corneal epithelium was the only method that con-sistently produced corneal infection in a large numberof the experimental hosts. We were not able to inducekeratitis by intrastromal injections, subconjunctivalinjections, irrigation of abraded corneas with parasite-rich inocula, or by deposition of parasites into the con-junctival cul-de-sac (Table 2).

Using parasite-laden DTL, the specific role of cor-neal trauma was examined. Eyes from 18 Chinesehamsters were abraded with cotton swabs and theninfected with parasite-laden DTL. For comparison, 12additional animals were also infected with parasite-laden DTL, but without abrasion of the cornea. At theremoval of the contact lens on day 7 after infection,the majority of animals (66%) with abraded corneasdeveloped keratitis as described later. By contrast, nosign of keratitis developed in any of the animals withnonabraded corneas. The highest incidence of infec-tion occurred when the parasite-laden DTL remainedin the eye for at least 7 days (Table 2).

Acute signs of infection were visible 7 days afterexposure to parasite-laden DTL. Eyelids were in-flamed and swollen, and the conjunctivae were in-

TABLE 2. Modes of Corneal Exposure to Acanthamoeba

Exposure to Parasites*

Intrastromal injectionSubconjunctival injectionDeposition in cul-de-sacInoculation after abrasion of

the corneaContact lens (sterile)Contact lens (no abrasion)Contact lens (with abrasion)

3 days exposure5 days exposure7 days exposure10 days exposure

Mean Clinical Scoref

No signs (8)No signs (8)No signs (8)

No signs (8)No signs (12)No signs (12)

4+ (1/8)4+ (4/9)4+ (12/18)4+ (6/9)

Histopathology

No inflammationNo inflammationNo inflammation

No inflammationNo inflammationNo inflammation

PMN infiltratePMN infiltratePMN infiltratePMN infiltrate

Isolation of Acanthamoeba

NegativeNegativeNegative

NegativeNegativeNegative

PositivePositivePositivePositive

* Controls for imrasiromal injection, subconjunctival injection, deposition into cul-de-sac, and inoculation onto abraded cornea consistedof hamsters exposed to parasite-free PYG medium in the same manner as the respective experimental group (e.g., intrastromal injectionof slerile PYG medium). Controls for experiments involving contact lenses consisted of sterile contact lenses incubated in parasite-freePYG medium. Controls for experiments testing the effect of abrasion consisted of hamsters exposed to parasite-laden contact lensesplaced onto intact corneas that were not abraded with cotton-tipped swabs.t Clinical scores calculated as described in Table 1. Numbers in parentheses represent the number of hamsters in each group.



jected and hyperemic. Corneal stromas were moder-ately to severely opaque because of infiltration andedema. The epithelium was ulcerated in different de-grees in most (90%) of the infected animals (Figure 1).Neovascularization was visible in the more severely in-fected animals. Histologic examination of acutely in-fected corneas revealed disruption of stromal lamel-lae, edema, neovascularization, and an inflammatoryinfiltrate consisting predominantly of neutrophils (Fig-ure 2). Viable parasites were isolated from cornealsmears, scrapings, and from the removed DTL. Cor-neas started to clear 3 days after removal of the DTL.Epithelial defects healed, stromal infiltration andedema resolved, and the keratitis was healed in 50% ofthe infected animals. By the fourteenth day after in-fection, all signs of keratitis had disappeared, althoughmild scarring of the cornea developed in a small num-ber (10%) of hamsters.

Pathogenic bacteria such as Pseudomonas spp.were not detected at any time. Furthermore, sterileDTL applied in the same manner as the parasite-ladenDTL did not produce any sign of disease or cornealdamage after 7 days of exposure.

Role of Langerhans Cells in AcanthamoebaKeratitisThe role of Langerhans cells in the development ofAcanthamoeba keratitis was evaluated by comparing theincidence and severity of keratitis in animals with nor-mal corneas versus animals with Langerhans cells inthe central cornea. Migration of Langerhans cells wasinduced by either the instillation of sterile latex beadsor intracorneal injection of recombinant mouse IL-1.Parasite-laden DTL were applied to the corneas 10days after latex-bead treatment and 2 and 7 days afterIL-1 injection. The presence of infiltrating Langer-

FIGURE l. Clinical appearance of acute Acanthamoeba kerati-tis in the Chinese hamster (day 7). The epithelium was ulcer-ated, stroma was opaque due to infiltration and edema. Vas-cularization, although present, was not visible in this pic-ture.

FIGURE 2. Histologic appearance of Acanthamoeba keratitis inthe hamster. (A) Cornea during acute infection showing ex-tensive neutrophilic infiltration, disruption of stromal la-mellae, edema, and neovascularization. Bar = 22 fxM. (he-matoxylin and eosin stain). (B) Higher magnification of cor-nea during acute infection showing neutrophilic infiltrate.Bar = 50 nM. (hematoxylin and eosin stain).

hans cells in the central corneas of latex bead-treatedand IL-1-treated corneas was confirmed by ADPasestaining (N = 6 for all groups). Although the numberof Langerhans cells was not significantly different inthe limbus of the control corneas, latex bead-treatedcorneas, or IL-1-treated corneas, the number of Lan-gerhans cells in the center of the corneas treated witheither latex beads or IL-1 was significantly higher thancontrols (Figure 3). Histologic examination of the la-tex bead-treated and IL-1-treated corneas revealedthat no inflammatory cells were present.

The presence of Langerhans cells in the centralportions of the cornea exerted a profound effect onthe development of Acanthamoeba keratitis. Clinicalsigns of Acanthamoeba keratitis developed in only 14%(1/7) of the latex bead-treated animals whereas 60%(6/10) of the untreated control animals had character-istic corneal lesions. Slightly different results occurredin corneas in which Langerhans cells were inducedwith IL-1. Here the time point of infection after the



CD CONTROLCXD LATEX BEADS (7 DAYS)CX] I L - 1 (2 DAYS)• i IL-1 (7 DAYS)

200 T

160--

120--

80--

40--

*

LIMBUS CENTRAL CORNEA

FIGURE 3. Number of Langerhans cells in Chinese hamstercorneas detected by ADPase staining in limbus and centralcornea 10 days after latex bead treatment, and 2 and 7 daysafter intrastromal IL-J injections. Each bar represents themean and SEM. N = 6 for each group. Asterisk indicatesdifference between that group was significantly differentfrom the control in the central cornea (P < 0.05).

injection of IL-1 seemed crucial. In the group infected2 days after IL-1 treatment, the incidence of infectionwas the same as in the control group (50%), but theseverity of infection was significantly reduced in thegroup treated with IL-1 2 days before exposure toparasite-laden DTL (Figure 4). In the group pre-treated with IL-1 7 days before infection, only 28%(2/7) developed keratitis compared to 71% (5/7) in thecontrol group.

A summary of the various factors influencing thedevelopment of Acanthamoeba keratitis is found in Ta-ble 3.

Attachment of Acanthamoeba castellanii toChinese Hamster Corneas In Vitro

It was previously shown that the extent to whichAcanthamoeba organisms bound to the corneal epithe-lium of various mammalian species correlated closelywith the parasite's ability to produce corneal infec-tions.45 The next experiments evaluated the effect oflatex bead treatment and corneal abrasion on the bind-ing of Acanthamoeba organisms to corneal epithelium.Abrasion of the cornea before exposure to parasiteshad a profound effect on parasite binding resulting inan 18 fold increase in the number of parasites adher-ing to either latex bead-treated corneas or normal cor-neas (Figure 5). By contrast, latex bead treatment didnot influence the degree of binding of parasites toeither nonabraded or abraded corneas.

DISCUSSION

The incidence of Acanthamoeba keratitis has increasedsharply since its first description in 1973.15 However,

the overall incidence is still low, especially consideringthe ubiquity of the parasite, and the large number ofpeople wearing contact lenses. Although the latter isgenerally accepted as the major risk factor for Acanth-amoeba keratitis, the observed rate of infection is only0.01% of the contact lens wearers in the UnitedStates.3 This suggests that there must be other riskfactors involved in the development of this cornealdisease. An animal model of Acanthamoeba keratitiswould be a useful tool to examine possible risk factors,pathophysiology, immunology, and therapy of this dis-ease.

In the past there have been few attempts to estab-lish an animal model of Acanthamoeba keratitis. Font etal16 reported that infections could be induced by intra-corneal inoculations in immunosuppressed rabbits.Badenoch et al10 described that intrastromal injectionsof Acanthamoeba organisms could only cause infectionswhen coinjected with Corynebacteria organisms and notby injecting Acanthamoeba alone. By contrast, Larkin etal17 were able to induce keratitis by intrastromal injec-tions in the rat by using Acanthamoeba polyphaga. Ineach of these studies, keratitis was induced by the di-rect inoculation of parasites into the corneal stroma.However, He et al5 recently described a model ofAcanthamoeba keratitis in pigs infected by parasite-laden contact lenses.

The Chinese hamster model of Acanthamoeba kera-titis has advantages over most previous models. Themost important one is that the mode of infection is thesame as what is believed to be the major source of

IL-1CONTROL

EpithelialDefect

StromalOpacity

StromalEdema

CLINICAL SCORE

FIGURE 4. Clinical scores for Chinese hamsters exposed toAcanthamoeba parasite-laden DTL 2 days after intrastromalinjection of IL-1 (n = 8) compared to control group (n = 6).50% of both groups developed keratitis. Only infected ani-mals are included in this figure. Each bar represents themean and standard error for that criterion in the infectedanimals of that group. All scores for the IL-1 treated animalswere significantly different from the control group (P< 0.05). "Mean Grade of Infection" represents the averageof the four clinical criteria.



TABLE 3. Factors Influencing Acanthamoeba Keratitis in Chinese Hamsters

ParameterIncidence of Nature ofKeratitis (%) Keratitis Conclusion

Method of exposure*Intrastromal injectionSubconjunctival injectionDeposition into cul-de-sacInoculation onto abraded corneaContact lens + abrasionContact lens without abrasion

Time of exposure to parasite-ladencontact lensf

3 clays5 days7 days10 days10 days (normal contact lens)

Effect of Langerhans' cellsLC induced by IL-1LC induced by latex beadsUntreated controls

0000

660

12446666

0

281460

NoneNoneNoneNoneSevereNone

SevereSevereSevereSevereNone

ModerateMildSevere

Exposure via parasite-laden contact lenses and cornealabrasion is necessary to produce disease

Exposure to parasite-laden contact lenses for 7 days orlonger produces highest incidence of disease

Presence of Langerhans' cells at the time of exposureto parasite-laden contact lenses reduces bothincidence and severity of disease

* Positive controls consisted of parasite-laden contact lenses placed onto abraded corneas for 7 days.t Negative controls consisled of normal contact lenses incubated with PYG medium not containing parasites.

human infection, namely by application of parasite-la-den contact lenses. However, certain minimal require-ments are necessary to establish keratitis in Chinesehamsters: parasite-laden contact lenses are necessaryto induce keratitis, the parasite-laden contact lensmust remain in the eye for a minimum of 7 days, andthe cornea must be partially abraded before exposureto parasite-laden contact lenses. Removal of the con-tact lens before day 7 reduces the rate of keratitis.Using these criteria, the rate of keratitis (ie, 60-70%),was consistent and predictable. The reason that infec-tion did not approach 100% might be attributable toimmunologic considerations. It is possible that a signif-icant percentage (30-40%) of the unmanipulated ani-mals had been previously exposed to Acanthamoebathrough contaminated bedding or through other envi-ronmental sources and, as a result, developed protec-tive immunity. Clinically and histopathologically, thekeratitis closely resembled the acute stages of the hu-man counterpart including epithelial ulceration, cor-neal opacity, edema, neutrophilic infiltration, and neo-vascularization. Moreover, the ability to isolate andculture viable parasite from diseased corneas addscredibility to the Chinese hamster model. Another ad-vantage of this model is that the Chinese hamster isinexpensive (compared to pigs), easy to work with in alaboratory setting, and inbred strains are readily avail-able.

The results from the in vitro binding study demon-strated that abrasion of the corneal epithelium greatlyincreased the binding of the parasite to the cornea.Moreover, the in vivo results confirm that abrasion isnecessary for the development of Acanthamoeba kerati-tis and implies that damage of the corneal epithelium

is an additional risk-factor for Acanthamoeba keratitis.This might mean that infections in contact lenswearers could be partly due to poorly fitting contactlenses or minor trauma related to contact lens wear,thereby resulting in damage of the corneal epithelium.This would occur only in combination with contami-nated contact lenses.

It is somewhat surprising that the presence ofLangerhans cells in the central cornea actually pro-tected the cornea against Acanthamoeba keratitis. These

450-

400-

350-

300-

250-

200-

150 •

100 '

50 •

0 -

CD CONTROLE D ABRADED CORNEAIB LATEX BEADS w/o ABRASIONm LATEX BEADS with ABRASION

EXPERIMENTAL GROUP

FIGURE 5. Attachment of Acanthamoeba castellanii to normalChinese hamster corneas and corneas 10 days after latexbeads treatment, both with and without abrasion of the cor-neal epithelium. Each bar represents the mean and standarderror. N = 10 for all groups. Attachment to the abradedcorneas was significantly higher than the nonabraded cor-neas (P < 0.05). There was no significant difference betweenlatex bead-treated and normal corneas (P > 0.05).



findings contrast with previous studies of herpes sim-plex keratitis, in which the presence of Langerhanscells greatly increased the incidence and severity of thekeratitis.6-7

The mechanism whereby Langerhans cells protectagainst Acanthamoeba keratitis might be an indirect im-mune-mediated process because Langerhans cells arehighly efficient antigen-presenting cells. Alternatively,Langerhans cells might exert a direct effect by damag-ing the parasites or inducing their encystment. Both ofthese possibilities are currently under investigation.The results from the in vitro attachment study showedthat neither the latex bead treatment nor the IL-1treatment used to induce Langerhans cell migrationaltered the binding of parasites to the corneal epithe-lium. Thus, the protective effects of latex bead andIL-1 treatment appear to be mediated by Langerhanscells and are not attributable to inhibition of parasitebinding to the corneal epithelium.

The Chinese hamster model provides exciting op-portunities to explore the pathophysiology, cell biol-ogy, immunology, and therapy of Acanthamoeba kerati-tis. However, it has limitations, especially the fact thatAcanthamoeba keratitis in Chinese hamsters is an acute,self-limiting infection. Therefore, this model mightnot be suitable for studying the progressive andchronic stages of Acanthamoeba keratitis. However, animportant advantage is its facility for immunologicstudies that are not feasible in the pig model. Anotherimportant feature of this model is that the mode ofinfection (ie, contact lenses) is the same as that be-lieved to occur in human patients. Thus, the Chinesehamster is not only a model of disease, but it is also amodel of infection. This in turn offers an opportunityfor designing and testing new therapeutic methods forthe prevention of Acanthamoeba keratitis in contactlens wearers.

Key Words

Acanthamoeba sp, hamster, contact lens, cornea, Langerhans'cells

A ckn owledgmen ts

The technical assistance of Mr. Thomas Cunningham isgreatly appreciated. Mr. John Horna provided excellent pho-tographic assistance. The senior author greatly appreciatesthe support and guidance provided by Dr. Martine J. Jagerthroughout the course of this project.

References

1. Auran JD, Starr MB, Jakobiec FA. Acanthamoeba kera-titis. Cornea. 1987;6:2-26.

2. Moore MB. Acanthamoeba keratitis. Arch Ophthalmol.1988; 106:1181-1182.

3. Stehr-Green J K, Bailey TM, Visvesvara GS. The epide-

miology of Acanthamoeba keratitis in the United States.Am] Ophthalmol. 1989; 107:331-336.

4. Niederkorn JY, Ubelaker JE, McCulleyJP, et al. Sus-ceptibility of corneas from various animal species to invitro binding and invasion by Acanthamoeba castellanii.Invest Ophthalmol Vis Sci. 1992;33:104-112.

5. He YG, McCulleyJP, Alizadeh H, et al. A pig model ofAcanthamoeba keratitis: Transmission via contaminatedcontact lenses. Invest Ophthalmol Vis Sci. 1992; 33:126-133.

6. McLeish W, Rubsamen P, Atherton SS, Streilein JW.Immunobiology of Langerhans cells on the ocularsurface. II. Role of central corneal Langerhans cells instromal keratitis following experimental HSV-1 infec-tion in Mice. Reg Immunol. 1989;2:236-243.

7. Jager MJ, Atherton S, Bradley D, Streilein JW. Her-petic stromal keratitis in mice: less reversibility in thepresence of Langerhans cells in the central cornea.Curr Eye Res. 1990; 10:69-73.

8. Silvany RE, Dougherty JM, McCulleyJP, Wood TS,Bowman RW, Moore MB. The effect of currently avail-able contact lens disinfection systems on Acanthamoebacastellanii and Acanthamoeba polyphaga. Ophthalmology1990;97:286-290.

9. Niederkorn JY, Peeler JS, Mellon J. Phagocytosis ofpaniculate antigens by corneal epithelial cells stimu-lates interleukin-1 secretion and migration of Langer-hans cells into the central cornea. Reg Immunol.1989;2:83-90.

10. Badenoch PR, Johnson AM, Christy PE, Coster DJ.Pathogenicity of Acanthamoeba and Corynebacterium inrat cornea. Arch Ophthalmol. 1990; 108:107-112.

11. Visvesvara GS, Mirra SS, Brandt FH, et al. Isolation oftwo strains of Acanthamoeba castellanii from human tis-sue and isoenzyme profiles. J Clin Microbiol.1983;18:1405-1412.

12. Peeler JS, Niederkorn JY: Antigen presentation byLangerhans cells in vivo: donor derived Ia+ Langer-hans cells are required for induction of delayed-typehypersensitivity but not for cytotoxic T lymphocyte re-sponses to alloantigens. J Immunol. 1986; 136:4362-4371.

13. Chaker MB, Tharp MD, Bergstresser PR. Rodent epi-dermal Langerhans cells demonstrate greater histo-chemical specificity for ADP than for ATP and AMP./Invest Dermatol. 1984; 82:496-500.

14. Weisman RA, Korn ED. Phagocytosis of latex beads byAcanthamoeba. I. Biochemical properties. Biochem.1967;6:485-497.

15. Jones DB, Visvesvara GS, Robinson NM. Acantha-moeba polyphaga keratitis and Acanthamoeba uveitis as-sociated with fatal meningoencephalitis. Trans Ophthal-mol Soc UK. 1975;95:221-232.

16. Font RL, Tapert MJ, Robinson NM, Osato MS, JonesDB. An animal model of Acanthamoeba keratitis. InvestOphthalmol Vis Sci. (suppl). 1982; 22:163.

17. Larkin DFP, Easty DL: Experimental Acanthamoebakeratitis: I. Preliminary findings. Br J Ophthalmol.1990;74:551-555.


the role of contact lenses, trauma, and langerhans cells in a

Documents