Download A partial-thickness epithelial defect increases the adherence

Investigative Ophthalmology & Visual Science, Vol. 30, No. 6, June 1989
Copyright © Association for Research in Vision and Ophthalmology
A Partial-Thickness Epithelial Defect Increases the
Adherence of Pseudomonas aeruginosa to the Cornea
Stephen A. Klorz,*f Yue-Kong Au,t and Raghunarh P. Misro-f^
Some patients with infectious keratitis have no clinically demonstrable corneal abrasion predisposing
them to infection. Subtle, undetectable corneal injuries may facilitate bacterial adherence to the
cornea, eventually leading to keratitis. To study this concept, we have developed a rabbit model in
which a partial-thickness corneal epithelial defect was induced by filter paper impression on the
cornea that removed one to two layers of corneal epithelium. Following this injury, the corneas were
incubated with Pseudomonas aeruginosa, washed, and the number of bacteria adhering to the injured
corneas as well as to control corneas was quantitated. Corneas treated with filter paper, either ex vivo
or in vivo, allowed 20 times more bacteria to adhere than did the untreated control corneas (/* < 0.01).
This superficial epithelial defect increased Pseudomonas adherence to the cornea for up to 72 hr after
injury. When corneal injury was extended to the stroma, the adherence of Pseudomonas was further
augmented as compared to adherence to the superficially injured cornea. Thus, we conclude that a
clinically subtle, partial-thickness corneal epithelial injury can markedly facilitate the adherence of
Pseudomonas aeruginosa, which may be an important predisposing factor for infectious keratitis.
Invest Ophthalmol Vis Sci 30:1069-1074,1989
It is well appreciated clinically that corneal abrasion is a predisposing factor to bacterial keratitis.
Heretofore macroscopic injuries have been used in
animal models of the disease. Nevertheless, infectious
keratitis often caused by Pseudomonas aeruginosa
occurs frequently in the absence of a history of corneal trauma, particularly among extended wear soft
contact lens wearers and dry eye patients.1"3 Presumably disruption of the tear film (in dry eye patients),
an unspecified compromise of corneal defenses, or a
subtle corneal injury (in contact lens wearers) predisposes those individuals to infectious keratitis.
In this report we investigated the effect of a partialthickness corneal epithelial injury as well as a fullthickness epithelial injury on the subsequent adherence of Pseudomonas aeruginosa. Using this approach we found that after the removal of one to two
layers of corneal epithelium, the adherence of Pseudomonas to the cornea was enhanced 20-fold or
greater.
Materials and Methods
Animals
Male and female adult New Zealand white rabbits
were used in this study in accordance with the ARVO
Resolution on the Use of Animals in Research. Enucleated eyes were also obtained from cattle and sheep
at a local abbatoir and transported in phosphate buffered saline (PBS) at 4°C until use.
Bacteria
A strain of Pseudomonas aeruginosa isolated from
a patient with soft contact lens-associated infectious
keratitis and characterized in previous work4 was
maintained by weekly transfer on trypticase soy agar.
For adherence assays a loopful of bacteria was placed
in 10 ml of Mueller-Hinton broth (MHB) and grown
at 26°C with shaking for 16 hr. The bacteria were
washed in phosphate-buffered saline (PBS) by centrifugation and diluted to a 1 McFarland standard (approximately 1.5 X 108 bacteria/ml) in PBS.
From the Departments of *Medicine, tOphthalmology and JPathology, Veterans Administration Medical Center and Louisiana
State University, Shreveport, Louisiana.
Supported by the Veterans Administration Research Service, the
Department of Ophthalmology, LSU School of Medicine, Shreveport, and by a donation from the Alcon Corp., Fort Worth, Texas.
Submitted for publication: August 23, 1988; accepted December
12, 1988.
Reprint requests: Stephen A. Klotz, MD, Infectious Diseases
Section (111 ID), VA Medical Center, 510 East Stoner Avenue,
Shreveport, LA 71101 -4295.
Corneal Trauma
Ex vivo experiments: Eyes were removed aseptically from rabbits, cattle and sheep. The right eye of
each animal served as a normal control. A subtle epithelial injury was induced by pressing a strip of Number 42 filter paper (Whatman, Hillsboro, OR) to the
corneal surface until adherent and then gently removing it. This removed one to two layers of epithe-
1069
Downloaded From: http://iovs.arvojournals.org/ on 05/14/2017
1070
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / June 1989
Vol. 30
Fig. 1. (A) Photomicrograph of cross-section of the
comeal surface demonstrating denuded area of epithelium (to the right) and
loosely adherent epithelium
(original XlOO). (B) Magnification of (A) to show the
margin of injury with area
of normal epithelium to the
left and partially denuded
surface to theright(original
X400). (C) Cross-section of
the stromal injury. The endothelial surface is to the
bottom and there is a complete absence of epithelium
on the top surface (hematoxylin and eosin stain, original X400).
lial cells from the comeal epithelium (Fig. 1A, B).
Removal of epithelium was confirmed in each instance by staining the filter paper strips with 1% toluidine blue and verifying the presence of epithelium by
light microscopy (Fig. 2). A severe corneal injury resulting in exposure of the stroma was induced by
soaking globes in 4% cocaine hydrochloride for 30
min followed by removal of the epithelium with a
spatula. This method removed all microscopic trace
of epithelium (Fig. 1C).
In vivo experiments: Rabbits were anesthetized to
the level of absence of pinch reflex with xylazine and
ketamine intramuscularly and corneas were then
treated with filter paper or cocaine as described
Fig. 2. (A) Photomicrograph of toluidine blue-stained corneal epithelial cells on thefilterpaper surface. (Fibrils are the matrix of the filter
paper; original XlOO). (B) Magnification of (A) demonstrating the epithelial cell harvest which is uniformly spaced and about one cell layer
thick (original X400).
Downloaded From: http://iovs.arvojournals.org/ on 05/14/2017
No. 6
5UDTLE EPITHELIAL DEFECT ENHANCES BACTERIAL ADHERENCE / Klorz er ol
above. Further anesthetic was given as needed during
the experiment to maintain the appropriate level of
anesthesia. No ophthalmic drops were used.
Adherence Assay
Conduct of adherence assay using ex vivo eyes. Ex
vivo globes with or without the superficial epithelial
injury or denudation to the stroma were submerged
in 3 ml of PBS containing 1.5 X 108 Pseudomonas/'ml
for 90 min at 37°C without shaking. Following this,
each globe was washed by immersion four times in
sterile PBS and then a 7.2 mm diameter button of the
central cornea was removed using a trephine. Each
button was sectioned into halves which were homogenized in 1 ml MHB with a motor-driven Teflon
grinder (Tri-R Instruments, Inc., Rockville Center,
NY). Serial 10-fold dilutions of the homogenate or
1:1 dilutions were prepared and duplicate 0.1 ml
samples of each dilution were plated onto M-H agar
after the dilutions had been prepared and colonyforming units (cfu) determined the following day.
Uninfected control eyes possessed no detectable cfu
of bacteria.
Conduct of the adherence assay using in vivo eyes.
In vivo experiments were conducted in the same
manner except that Pseudomonas was placed in each
eye by dropper every minute for 90 min. Animals
were sacrificed, the eyes removed, washed, trephined,
homogenized and aliquoted as above.
1071
vivo rabbit eyes subjected to filter paper impression
and viewed with the slit lamp showed no evident corneal injury even after the application of fluoroscein.
Application of rose bengal demonstrated a clearly delineated central staining. Therefore, this method provided a subtle partial thickness epithelial defect.
Adherence of Pseudomonas to the Uninjured Cornea
In preliminary ex vivo experiments it was established that the number of Pseudomonas required in
the inoculum in order for detectable cfu to occur in
the control cornea was large: incubation of the eye for
90 min with: 1.5 X 109 bacteria/ml yielded 305 ± 91
cfu/mm2 of corneal surface; 1.5 X 108 bacteria/ml
yielded 198 ± 99 cfu/mm2; and 1.5 X 107 bacteria/ml
yielded 46 ± 24 cfu/mm2. Incubation with 106 or
lower bacteria/ml did not yield detectable cfu. Therefore, the remainder of the experiments were conducted with inocula of 1.5 X 108 bacteria/ml, an inoculum size which resulted in detectable cfu and was
easily standardized. Adherence of Pseudomonas to
the control ex vivo eye was a rapid process: after 15
min of incubation there were 189 ± 52 cfu/mm2 of
corneal surface; 30 min, 258 + 30 cfu/mm2; and at 90
min, 344 ± 69 cfu/mm2. This represented about a
100% increase in adherence from 15 to 90 min but it
is noteworthy that there was considerable adherence
even after 15 min of incubation.
Adherence of Pseudomonas to Filter Paper-Injured
Ex Vivo Corneas
Statistics
Mean values were compared by two-tailed student
t-test. In some experiments, no assumption was made
about the distribution of data and Chi-square analysis
was performed.5 There were at least three animals per
experiment and each experiment repeated a minimum of three times.
Results
Characterization of Injury Produced by
Filter Paper Impression
We first determined the depth of the superficial
corneal injury induced by the application of filter
paper to ex vivo rabbit and bovine eyes. Cross sections of ex vivo rabbit globes preserved after filter
paper impression showed that epithelium was removed from the corneal surface and there was no
extension of the injury to the basal layers of the epithelium (Fig. 1B). The injury was such that there were
fragments of epithelial cells and intact epithelial cells
present in areas of injury. Stained preparations of the
filter paper showed intact corneal epithelium that appeared to be only one to two cells thick (Fig. 2). Ex
Downloaded From: http://iovs.arvojournals.org/ on 05/14/2017
Control eyes and filter paper-treated eyes were then
incubated with 1.5 X 108/ml Pseudomonas aeruginosa for 90 min, washed and coronal sections of the
cornea removed, stained with toluidine blue and
viewed with the light microscope. It was clearly evident that Pseudomonas aeruginosa adhered in abundance to defects created in the corneal epithelium by
the filter paper treatment (Fig. 3), whereas in control
eyes there were only a few adherent bacteria in random locations not recognizably associated with injured or abnormal epithelium. This was true of eyes
from cows, sheep and rabbits.
Quantitation of the adherence of Pseudomonas to
ex vivo filter paper-injured corneas compared to control or uninjured eyes was then performed. The results of a typical experiment are shown in Table 1. It
was clear that the subtle defect greatly enhanced adherence. The control or uninjured eye had 50 ± 35
cfu/mm2 whereas the filter paper-injured eye had
1398 ± 538 cfu/mm2, an increase of greater than 20
times that of the uninjured eyes (P < 0.01). The epithelial injury was apparently repaired by 72 to 96 hr
since at that time there was no significant difference
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / June 1989
1072
Vol. 30
Fig. 3. (A) En face preparation of bovine cornea stained with a rapid Wright stain (Difco, Detroit, MI) demonstrating epithelial cell nuclei
within several layers of the cornea! epithelium (original X400). (B) Filter paper-injured eye with Pseudomonas aeruginosa (arrows) adhering
to the bottom of comeal defect (original X1000). (C) The edge of the comeal defect with the epithelial cell folded back upon inself and still
joined to the cornea between the two solid arrows. P. aeruginosa (open arrows) adhering to the surface of the defect as well as undersurface of
the epithelial cell (original XI000).
in Pseudomonas adherence to injured or uninjured
eyes (Fig. 4).
Adherence of Pseudomonas to Ex Vivo Stroma
The level of trauma was then extended to include
the removal of corneal epithelium altogether. Eyes
soaked in cocaine for 30 min and then denuded to
stroma with a spatula also demonstrated enhanced
adherence of Pseudomonas (Table 1). The increase of
adherence was on the order of 14 times greater than
that to uninjured eyes (P < 0.01).
Adherence of Pseudomonas to Injured Corneas
In Vivo
logical setting. Injured corneas (whether injuries were
partial or full-thickness epithelial injuries) showed
greater adherence of bacteria than control or uninjured eyes similar to that obtained with ex vivo eyes.
Enhanced adherence of Pseudomonas was seen when
measured by colony forming units and by dilution
determination. Multiple comparisons of the partialthickness epithelial injury versus full-thickness epithelial injury were conducted in effort to determine
which defect allowed a greater adherence of Pseudomonas. In every animal the cornea with stromal exposure from a full-thickness epithelial injury had
greater adherence of bacteria than did the one with
the partial-thickness epithelial injury. The results of a
representative experiment are shown in Table 2. The
Having established that both partial thickness epithelial injury and full thickness epithelial injury or
exposure of stroma clearly increases the adherence of
P. aeruginosa to ex vivo eyes, we repeated these experiments in vivo to determine whether similar increases of adherence would occur in a more physioTable 1. Adherence of Pseudomonas aeruginosa
to rabbit ex vivo corneal surfaces
Treatment
Normal (control)
Partial-thickness
epithelial injury
Normal (control)
Full-thickness epithelial
injury
cfu/mnr
3
50 ±
35
3
3
1398 ± 538
285 ± 106
<0.01
3
4081 + 528
<0.01
The injuries were created byremovalofl to 2 layers of corneal epithelium
with filter paper (partial-thickness epithelial injury) or by exposure of the
stroma by scraping (full-thickness epithelial injury). The corneas were then
incubated with I.5XI0" Pseudomonas/m\ for 90 min. P value determined
from paired data using the student t-test.
• n = number of rabbits.
Downloaded From: http://iovs.arvojournals.org/ on 05/14/2017
Hours After Corneal Injury
Fig. 4. Increase in Pseudomonas adherence to the cornea followingfilterpaper injury. Increase in adherence was determined by the
formula: mean cfu/mm2 of corneal surface of filter paper-injured
cornea MINUS mean cfu of control cornea DIVIDED BY mean
cfu of control cornea. There were two rabbits for each time point.
No. 6
1073
SUBTLE EPITHELIAL DEFECT ENHANCES BACTERIAL ADHERENCE / Klorz er al
mean dilution at which the filter paper-treated cornea
injury was positive (ie, visibly turbid) was 1:64,
whereas eyes with the stromal exposure had a mean
dilution of 1:1024. Furthermore, when colony-forming units/ml in positive dilution tubes were compared, the cornea with stromal exposure from the
full-thickness epithelial injury ranged from 10-100
times greater than the eye with a partial-thickness
epithelial injury.
Discussion
In these experiments we have established that a
partial-thickness epithelial injury undetectable by a
slit-lamp examination markedly enhances the adherence of Pseudomonas aeruginosa to the corneal surface. This was demonstrated with corneas from three
mammalian species.
Previous animal models of Pseudomonas keratitis
have used macroscopic injury in measuring adherence as well as infection: for example, by trephining
to the stroma6; injecting microorganisms into the
cornea7-8 or by scratching the cornea910 (for a review
of these models see Hyndiuk"). In some animals no
manipulation is required to induce infection,12 which
is not considered to be the case in humans. While
these manipulations of the animal corneas are successful in inducing Pseudomonas infections, it is
questionable whether such gross trauma simulates
the pathogenesis of infectious keratitis, especially as it
is seen in patients wearing contact lenses or with dry
eyes. Infectious keratitis in patients wearing soft contact lenses, with dry eyes or in comatose patients
occurs often in the absence of a history of predisposing corneal injury. Therefore, we sought a model in
which the induced corneal injury would be subtle and
reversible. Having induced such an injury, we could
then measure its effects upon the adherence of Pseudomonas aeruginosa to the corneal surface since adherence of a microorganism is a necessary first step in
the pathogenesis of infectious keratitis.9 The model
described here allows us to perform these measurements. The injury induced by impression is entirely
superficial, removing only one to two layers of corneal epithelial cells (Figs. 1, 2). The defect is also
reversible, with return to normal cornea after about
72 to 96 hr as determined by the ability of the bacterium to adhere to the corneal surface.
We have determined that bacteria adhere almost
exclusively to areas where an epithelial defect exists
(Fig. 3). Whether the bacteria are adhering to exposed
extracellular matrix or surfaces of injured cells was
not determined. However, work with Candida yeasts
has demonstrated the clear preference of yeasts to
adhere to extracellular matrix rather than intact en-
Downloaded From: http://iovs.arvojournals.org/ on 05/14/2017
Table 2. Partial-thickness versus full-thickness
epithelial injury in vivo in rabbits
Dilution results (cfu/ml)
Dilution
1024
512
256
128
64
32
16
8
4
2
1
0
Partial-thickness
epithelial injury
Full-thickness
epithelial injury
-(ND)»
-(ND)
-(ND)
-(ND)
+(ND)
+(10)
+(15)
+(70)
+(140)
+(250)
+(660)
+(1135)
+(40)
+(30)
+(70)
+(120)
+(225)
+(920)
+(2170)
+(4585)
+(TNTC)
+(TNTC)
+(TNTC)
+(TNTC)
A comparison of dilutions positive (ie, turbid broth) for Pseudomonas and
the corresponding colony-forming units in the tubes. Eyes had been incubated with 1.5 X 108 Pseudomonas/mi for 90 min.
* ND = none detected; TNTC = too numerous to count.
dothelial cells.1314 Extracellular matrix or epithelial
cell surface glycoconjugates may be playing a similar
role in the process of Pseudomonas adherence to the
cornea as well. In the uninjured infant mouse, corneal sialic acid may serve as a receptor15 whereas in
vitro mannose moieties may be subserving this purpose in the adult rabbit.l6
Surprisingly, we found that stroma was more adherent than corneal epithelium, a finding opposite to
that of Stern et al.17 However, in their experiments
adherence was determined by scanning electron microscopy and the epithelial injury was made by
scraping the cornea with a needle, hence inducing a
defect through epithelium to the stroma. Spurr-Michaud et al found an enhancement of Pseudomonas
adherence on denuded corneal basal lamina over adjacent corneal epithelium in an organ culture system.16 They found, however, that Pseudomonas adherence increased with healing whereas we found the
opposite (Fig. 4).
If adherence of the microorganism to target tissue
is a necessary antecedent to infection, as it is believed
to be in many infectious processes, then this work has
demonstrated that a clinically subtle corneal lesion
leads to a many-fold increase in bacterial adherence
to the injured cornea. Furthermore, extending the
injury to the stroma further increases bacterial adherence. Thus, if a similar partial-thickness epithelial injury occurs in soft contact lens wearers, comatose
patients, or dry eye patients then this observation of
enhanced bacterial adherence may help explain the
pathogenesis of this enigmatic infectious disease.
Key words: bacterial adherence, corneal wound healing,
cornea, infectious keratitis, Pseudomonas aeruginosa
1074
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / June 1989
Acknowledgments
The authors thank Denise Dorsey for help with the manuscript and Charles E. Self for technical assistance.
10.
References
1. Ormerod LD and Smith RE: Contact lens-associated microbial
keratitis. Arch Ophthalmol 104:79, 1986.
2. Donnernfeld ER, Cohen EJ, Arentsen JJ, Genvert GI, and
Laibson PR: Changing trends in contact lens associated corneal ulcers: An overview of 116 cases. CLAO J 12:145, 1986.
3. Chalupa E, Swarbrick HA, Holden BA, and Sjostrand J: Severe
corneal infections associated with contact lens wear. Ophthalmology 94:17, 1987.
4. Butrus SI, Klotz SA, and Misra RP: The adherence otPseudomonas aeruginosa to soft contact lenses. Ophthalmology
94:1310, 1987.
5. Gustafson TL: True Epistat. Richardson, TX, Epistat Services,
1987.
6. Wyman M, Swanson C, Kowalski JJ, Powers JD, and Boraski
EA: Experimental Pseudomonas aeruginosa ulcerative keratitis model in the dog. Am J Vet Res 44:1135, 1983.
7. Gray LD and Kreger AS: Rabbit corneal damage produced by
Pseudomonas aeruginosa infection. Infect Immun 12:419,
1975.
8. Davis SD and Chandler JW: Experimental keratitis due to
Pseudomonas aeruginosa: Model for evaluation of antimicrobial drugs. Antimicrob Agents Chemother 8:350, 1975.
9. Ramphal R, McNiece MT, and Polack FM: Adherence of
Downloaded From: http://iovs.arvojournals.org/ on 05/14/2017
11.
12.
13.
14.
15.
16.
17.
Vol. 30
Pseudomonas aeruginosa to the injured cornea: A step in the
pathogenesis of corneal infections. Ann Ophthalmol 13:421,
1981.
Hyndiuk RA, Davis SP, Hatchell DL, Burd EM, Pederson HJ,
Divine KS, and Nassif KF: Experimental Pseudomonas keratitis: Clinical and pathological observations. Cornea 2:103,
1983.
Hyndiuk RA: Experimental Pseudomonas keratitis. Trans Am
Ophthalmol Soc 79:591, 1981.
Hazlett LD, Wells P, and Berk RS: The mouse cornea as a
model for Pseudomonas infection: Further studies on the penetration and lysis of the unwounded eye. In The Structure of
the Eye. Hollyfield JG, editor. New York, Elsevier North Holland, Inc., 1983, pp. 279-296.
Klotz SA: The adherence of Candida yeasts to human and
bovine vascular endothelium and subendothelial extracellular
matrix. FEMS Microbiol Lett 48:201, 1987.
Klotz SA and Maca RD: Endothelial cell contraction increases
Candida adherence to exposed extracellular matrix. Infect
Immun 56:2495, 1988.
Hazlett LD, Moon M, and Berk RS: In vivo identification of
sialic acid as the ocular receptor for Pseudomonas aeruginosa.
Infect Immun 57:687, 1986.
Spurr-Michaud SJ, Barza M, and Gipson IK: An organ culture
system for study of adherence of Pseudomonas aeruginosa to
normal and wounded corneas. Invest Ophthalmol Vis Sci
29:379, 1988.
Stern G, Weitzenkom D, and Valenti J: Adherence of Pseudomonas aeruginosa to the mouse cornea: Epithelial v stromal
adherence. Arch Ophthalmol 100:1956, 1982.