Proliferation Rate of Rabbit Corneal Epithelium during Overnight Rigid Contact Lens Wear

Abstract

purpose. To examine cell proliferation of the normal corneal epithelium and during extended rigid gas-permeable (RGP) lens wear.

methods. Twenty-three New Zealand White rabbits were fitted unilaterally with either a low oxygen transmissible (Dk/t) or hyper-Dk/t RGP lens, with the other eye serving as a control. The rabbits were injected with 5-bromo-2-deoxyuridine (BrdU) 24-hours later and killed at three time points: 1, 3, and 7 days after injection. Corneas were processed for immunocytochemistry, and sequential digital images were taken from the superior limbus to the central epithelium with an epifluorescence microscope. The total number of BrdU-labeled cell pairs was quantified.

results. The limbus in normal corneas was significantly less populated with BrdU-labeled cells than the central and peripheral epithelium (P < 0.05). The peripheral epithelium adjacent to the limbus was marked by a peak of labeled cells (P < 0.05). Both types of RGP lenses produced an increase in BrdU labeling in the limbus and a dramatic decrease in the central epithelium (80% for low Dk/t, 37% for hyper Dk/t). At day 3 and 7 after BrdU injection, the low-Dk/t lens continued to show decreased BrdU labeling centrally, whereas the limbus remained increased. Hyper-Dk/t lens wear however, showed persistent limbal elevation but equivalent numbers of BrdU-labeled cells centrally at days 3 and 7, compared with control corneas. Keratocytes unexpectedly showed BrdU labeling during RGP lens wear.

conclusions. Limbus, peripheral, and central epithelium were characterized by different proliferation rates in the normal rabbit cornea. RGP lens wear significantly altered the homeostatic proliferation pattern of the epithelium with the low-Dk/t lens having the most dramatic effect. RGP contact lens wear appears to stimulate proliferation of keratocytes.

The outermost surface of the human cornea is formed by a 50-μm thick epithelium, composed of five to seven cell layers. It is a resilient, self-renewing cellular barrier that serves as the first line of defense against infection. A homeostatic equilibrium among basal cell division, cell differentiation, and surface cell exfoliation is important in the maintenance of overall corneal epithelial integrity. It is now widely believed that normal replenishment of the corneal epithelium is a continuous dynamic process originating in the limbus, where epithelial stem cells are located. ^{1 2} These slow-cycling cells have a tremendous capacity to divide, producing transient amplifying (TA) epithelial cells. ³ TA cells migrate toward the center of the cornea ^{4 5} and divide a few times over a relatively short time before leaving the basal cell layer and undergoing epithelial cell differentiation, which terminates in surface cell exfoliation. 

Daily and overnight contact lens wear is known to induce changes in the corneal epithelium of animals and humans. Such changes include superficial punctate keratitis, epithelial edema, epithelial thinning, ^{6 7} decreased cell exfoliation as collected with an irrigation chamber, ^{8 9 10} enlargement of superficial epithelial cells on the surface, ^{9 11 12} decreased sensitivity, ¹³ epithelial microcysts, ¹⁴ reduced epithelial adhesion, ¹⁵ epithelial glycogen depletion during anoxia, ¹⁶ a decline of the corneal epithelial barrier function, ^{17 18} and increased bacterial binding. ^{9 19} These changes in the epithelium may predispose contact lens wearers to increased risk of infectious corneal ulceration with possible devastating visual outcomes. ^{20 21 22} Taken together, these data suggest that contact lens wear disturbs this homeostatic balance of epithelial cell birth, centripetal migration, and superficial cell loss. Surprisingly, only two reports have been published on the effect of contact lens wear on epithelial cell proliferation in the cornea. 

Hamano and Hori ²³ found that short-term (24–48 hours) extended soft contact lens wear severely decreases the number of mitotic figures observed in the central epithelium. Recently, Ren et al. ²⁴ were the first to demonstrate with BrdU labeling that short-term (24 hours) rigid contact lens wear also inhibits mitosis in the central corneal epithelium of the rabbit. The effects of long-term overnight contact lens wear on proliferating corneal epithelial cells and their fate over time, however, are unknown. 

The purpose of this study was to determine in the rabbit model the long-term effects of lens oxygen transmissibility on the corneal epithelial proliferation rate after prolonged overnight lens wear and to observe labeled proliferating cells over time. The proliferation rate was studied by labeling the dividing epithelial cells with 5-bromo-2-deoxyuridine (BrdU). ^{24 25 26} Labeled BrdU-cells were monitored over time and across corneal spatial locations as animals were killed at three different intervals after BrdU injection. The BrdU-labeled cells were counted in a wholemount preparation by epifluorescence microscopy, to cover a large corneal area. 

Methods

Twenty-seven New Zealand White rabbits weighing 3.0 to 3.5 kg and of either sex were used in the study. All rabbits were treated according to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The animals were screened for ocular disease with a handheld biomicroscope, housed in individual cages at constant room temperature (19–23°C) and relative humidity of 30% to 50%, were maintained in a constant 12-hour light–dark cycle. Food and water were provided ad libitum. To facilitate lens wear, a standard partial nictitating membranectomy was performed on each eye with rabbits under anesthesia with 30 to 50 mg/kg ketamine (Ketaset; Fort Dodge, Fort Dodge, IA) and 3 to 5 mg/kg xylazine (Rompun; Bayer, Shawnee Mission, KS) and after the application of two topical anesthetic drops (0.5% tetracaine hydrochloride ophthalmic solution; B&L, Tampa, FL). Prophylactic antibiotic ointment (polymyxin B, AK-Poly-Bac; Akorn, Decatur, IL) and drops of 0.3% gentamicin sulfate ophthalmic solution (B&L) were applied to the rabbits’ eyes after the surgery. ²⁴ The rabbits were allowed to recover for a minimum of 1 week. 

After recovery from the partial removal of the nictitating membrane, rabbits were randomly assigned to two experimental contact lens–wearing groups (n = 23). One eye of each rabbit was randomly selected and fitted with either rigid gas-permeable (RGP) lens; the contralateral eye served as a control. Four non–lens-wearing rabbits were used to examine the effects of removal of the nictitating membrane on BrdU labeling with the other eye serving as a control. Table 1 shows the characteristic material properties of each test RGP lens. All lenses were spherical lenses with a monocurve lens design, a diameter of 14.0 mm, and a uniform thickness of 0.15 mm. The best-fitting base curve was selected after trial fitting with radii of 7.60, 7.80, and 8.00 mm, by using fluorescein and cobalt blue light. All contact lenses were placed on the eye at 9.00 AM and left on continuously until the animal was killed. 

After 24 hours of overnight wear, anesthetized rabbits were injected intravenously (marginal ear vein) with 5-bromo-2-deoxyuridine (BrdU; 200 mg/kg) in phosphate-buffered saline (PBS; 30.3 mg/ml; pH 7.4) using a 23-gauge needle. BrdU substitutes for thymidine during the S phase of the DNA replication and has been used previously to label proliferating corneal epithelial cells in the rat ^{26 27} and rabbit. ^{24 25} The BrdU-labeling technique in this study was introduced by Beebe et al. ²⁶ in the rat and modified by Ren et al. ²⁴ in the rabbit. Given that a circadian rhythm considerably affects the proliferation rate in different species, ^{28 29 30} all rabbits were injected with BrdU between 9:00 and 9:15 AM. Animals were then killed between 9:00 and 9:15 AM with an overdose of pentobarbital sodium (Nembutal; Abbott Laboratories, North Chicago, IL) at 1, 3, and 7 days after BrdU injection. 

Immunocytochemistry

Corneas were fixed in situ with 1% paraformaldehyde in PBS for 3 minutes, excised along 3 mm of scleral rim and cut in a vertical stripe from the superior to the inferior rectus muscle. Subsequently, the tissues were processed through a series of staining and washing as follows: The tissues were washed 3 minutes in TD buffer (PBS with 1% dimethyl sulfoxide [DMSO] and1% Triton X-100), placed in acetone (−20°C) for 3 minutes, washed in TD buffer for 3 minutes, placed in 1% HCl for 3 minutes, washed in TD buffer for 3 minutes, incubated in whole goat serum (1:10) for 30 minutes at 37°C, and stained overnight in diluted (1:30) monoclonal mouse anti-BrdU antibody in washing buffer (Boehringer Mannheim, Indianapolis, IN) at room temperature with agitation (100 turns per minute). The second day, the tissues were washed with TD buffer three times for 30 minutes and stained with FITC-conjugated goat anti-mouse secondary antibody (ICN, Costa Mesa, CA) overnight at room temperature with agitation (100 turns per minute). On the final day, the tissues were washed three times in TD buffer. 

Fluorescence Microscopy and Digitizing

Whole corneas were mounted epithelial side up on a glass slide and assessed by epifluorescence microscopy. Digital fluorescent microscopic images (588 μm vertically × 984 μm horizontally) were collected sequentially from the superior limbus to the central cornea (×10 objective). The number of nuclei appearing as pairs were counted per unit area (0.578 mm² using the UTHSCSA Image Tool program developed at the University of Texas Health Science Center at San Antonio. Beebe et al. ²⁶ have also shown that 97.8% to 98.3% of BrdU-labeled cells occur as pairs in the rat epithelium. When a single cell was detected, it was matched with another single cell in the same image as a pair. 

The normal limbus was easily identified as a zone of very little BrdU labeling compared with the highly labeled peripheral corneal epithelium and the adjacent conjunctiva. ²⁴ In addition, the limbal region could be located by the termination of scleral vessels viewed under transmitted light and the beginning of the avascular cornea. 

Statistics

The statistical analysis was performed by computer (Sigmastat ver. 1.0; SPSS, Chicago, IL). If the data were normally distributed and equal variance was obtained, then parametric repeated one-way and two-way analyses of variance (ANOVA) were applied, depending on the number of factors. The nonparametric Friedman repeated-measures ANOVA on ranks or the two-way ANOVA on ranked data were performed in all other cases. In this study, because of the large intra-animal variability in BrdU labeling, we compared the contact lens–exposed corneas with the contralateral control corneas of the same rabbit. To compare the effects of the low-Dk/t directly with the hyper-Dk/t-test lens, a separate study with a larger sample size is needed. The level of significance was set at P < 0.05, and if the power of the appropriate statistical test was not more than 0.8, the statistical analysis was omitted. If the data demonstrated significance after the ANOVA test, a pair-wise multiple comparison procedure was performed with the Student-Newman-Keuls (SNK) test. 

Results

Normal Rabbit Corneal Epithelium

Labeling of corneal epithelial cells in the S phase was accomplished with one BrdU injection. A control experiment (n= 4) did not show any significant differences in BrdU labeling between normal rabbits with the nictitating membrane on and those with it off (data not shown). To observe the fate of these BrdU-labeled cells over time, rabbits were killed at different intervals after injection. Figure 1 shows the results of normal control eyes at 1 day after injection. BrdU-labeled pairs were counted in images from the limbus (Fig. 1 ; 1 on the x-axis) to the central epithelium (Fig. 1 ; 11–15 on the x-axis). As previously reported in the rat corneal epithelium, most BrdU-labeled cells were organized in pairs. ^{24 26} There was a significant difference between the corneal regions (P < 0.001, Friedman repeated-measures ANOVA on ranks). A multiple-comparison procedure SNK test was used to compare each region individually. The limbus, with a mean of 67.4 ± 10.1 (SEM), was significantly less populated with BrdU-labeled cell pairs than the peripheral and central epithelium (P < 0.05, SNK test). The adjacent peripheral epithelium clearly demonstrated the highest degree of labeling. At day 1, a peak in the peripheral epithelium reached a mean of 204.3 ± 19.8 (SEM) BrdU-labeled pairs per standardized area, whereas the central epithelium showed an average of 132.9 ± 24.5 labeled pairs (significant, P < 0.05, SNK test). 

Low-Dk/t Contact Lens

Figure 2A illustrates an example of BrdU labeling in the central epithelium after low-Dk/t lens wear and in the control (Fig. 2C) . Figure 3 summarizes the data quantified after low-Dk/t contact lens wear compared with the contralateral control eyes 1 day after BrdU injection. The pattern of BrdU labeling between low-Dk/t lens and control corneas is significantly different (P < 0.001, two-way ANOVA on ranked data). The limbus of the contact lens group demonstrated an increase in labeling, whereas the cell proliferation in the central epithelium was dramatically lower than in the control group. The total inhibition of proliferation in the central epithelium of the low-Dk/t contact lens is approximately 80%, whereas by contrast the limbus is activated by approximately 150%. The number of BrdU-labeled cells in the limbus remained consistently elevated over time as shown at days 3 (Fig. 4) and 7 (Fig. 5) . There was almost no change over the first 1 week during low-Dk/t RGP lens wear across the central epithelium. The total number of BrdU-labeled cells remained between 16 to 20 per unit area of 0.578 mm² (NS, P > 0.05; one-way ANOVA). Despite the initial high increase in cell proliferation rate in the limbus, no wave of labeled cells moving toward the center was detectable over the 7 days after BrdU injection. However, in the peripheral epithelium, the number of labeled cells increased substantially from day 1 to day 3 and then decreased at day 7 to levels comparable to those on day 1. 

Hyper-Dk/t Contact Lens

Similar to but less dramatic than the low-Dk/t contact lens, hyper-Dk/t contact lens wear at day 1 decreased the number of labeled cells centrally (Fig. 2B) , whereas labeled cells increased in the limbus. Central epithelial cell labeling decreased by approximately 37% and the limbal epithelial labeling increased by approximately 80% (Fig. 6) . Of note, by day 3 the number of labeled epithelial cells appeared similar to that in control eyes, with the central cornea having a slightly higher number of labeled cells (Fig. 7 ). At day 7, a decrease in labeled cells was observed in the periphery, whereas the number of labeled cells in the central epithelium was identical with that in the control group (Fig. 8) ; however, the number of BrdU-labeled cells remained elevated in the limbus at days 3 and 7. 

Keratocytes

Unexpectedly, BrdU-labeled keratocytes were detected as shown in Figure 9 . Whereas, control corneas only rarely demonstrated BrdU-labeled keratocytes in the periphery, both the low- and hyper-Dk/t contact lens wear groups showed noticeable keratocyte labeling. Labeling was most distinct in the peripheral and midperipheral corneal stroma, whereas the central stroma showed no labeling. Although not quantified in this study, it appeared that the 3- and 7-day groups showed more BrdU-labeled keratocytes than the 1-day group. 

Discussion

The prolonged presence of an RGP contact lens on the ocular surface markedly altered the normal proliferation pattern of the limbal and corneal epithelium, independent of the oxygen transmissibility of the test contact lens. BrdU was incorporated only into epithelial cells preparing to divide at the time of the injection, giving rise to two labeled daughter cells that remained paired. Wearing of both the low- and hyper-Dk/t contact lens caused increases in BrdU labeling in the limbus. By contrast, the central epithelium demonstrated a lower proliferation rate than did control eyes. The central inhibition of proliferation was most pronounced in the low-Dk/t group, which showed an 80% reduction of BrdU labeling compared with control, whereas the hyper-Dk/t group showed a 37% inhibition. 

What factor(s) might cause the significant decrease in corneal epithelial cell mitosis? Oxygen plays a role, as indicated by the difference between the low- and hyper-Dk/t contact lens groups; however, it is probably not the only factor. The hyper-Dk/t contact lens should provide the ocular surface with sufficient oxygen with an equivalent oxygen percentage (EOP) of 19.13%—slightly less than the normal 21% for the open eye at sea level. The physical presence of the contact lens may also influence the proliferation rate of the corneal epithelium. As noted by Millodot, ¹³ 8 hours of low-Dk/t rigid contact lens wear results in decreases in corneal sensitivity of as much as 94%. Polse ³¹ attributed the decreased sensitivity to sensory adaptation caused by the mechanical stimulation of the rigid contact lens. In vitro experiments have further established that sensory innervation to corneal epithelial cells stimulates proliferation. ³² Thus, there may be a link between contact lens–mediated decreased corneal sensitivity and decreased proliferation of the corneal epithelium. 

The unexpected increase in BrdU labeling in the limbus in prolonged overnight wear appears to represent lens-related mitotic upregulation during RGP lens wear, possibly related to the dramatic decline in basal cell division centrally. Indeed, it has been shown in vivo that the normally slow-cycling limbus can increase the cellular division rate after injury in the central epithelium ^{3 33} : Cotsarelis et al. ³ found the proliferative rate of limbal epithelial cells in the rat increased eight- to ninefold, 12 hours after removing 1 mm of the central epithelium. Recently, Chung et al. ³⁴ found a 4.5-fold increase of BrdU-labeled cells in the rat limbus after central wounding. Based on the expression of cell cyclins D and E, they hypothesized that the limbal epithelium enters the S phase of the cell cycle more rapidly than do the corneal epithelial cells after wounding of the central epithelium. Taken together, these results suggest that the limbus is very sensitive to changes occurring in the central epithelium and that it is capable of rapidly upregulating epithelial cell proliferation if challenged. Although, the mitotic upregulation in the limbus induced by contact lens wear seems less dramatic than during wound healing, it may indicate that the lens-exposed corneal epithelium is under constant stress, particularly during prolonged overnight wear. Alternatively, the constant mechanical rubbing of the RGP contact lens edge may also be the cause of a localized increase in limbal BrdU labeling. 

The 3- and 7-day follow-up after the single-pulse BrdU injection enabled us to monitor the overall migration and secondary divisions of the epithelial cells labeled at day 1. The low-Dk/t contact lens test group did not show any large wave of labeled cells moving from the periphery to the central epithelium within the first week after BrdU injection. The total number of BrdU-labeled cells in the central epithelium remained the same at days 1, 3, and 7, suggesting a persistent suppression of central proliferation. The hyper-Dk/t contact lens test group showed an initial inhibition of central cell division at day 1; however, it seemed to recover fully centrally at day 3 (increased secondary divisions and/or centripetal migration?) as if test lens wear produced only a transient effect that was followed by rapid adaptive compensation. Holden et al. ⁷ demonstrated in a retrospective human study that long-term overnight contact lens wear reduces the oxygen uptake of the corneal epithelium when the lens is removed. These results indicate physiological adaptation of the corneal epithelium to prolonged hypoxia. A more recent prospective double-masked clinical study on long-term (1 year) overnight lens wear showed corneal epithelial adaptation in human patients. ³⁵ After an initial significant decrease in corneal epithelial thickness and surface cell exfoliation, partial recovery was observed during the subsequent year of extended lens wear. Correspondingly, the binding of Pseudomonas aeruginosa bacteria to collected exfoliated surface epithelial cells was significantly increased during the first 1 to 3 months of extended wear but recovered fully to baseline values at the conclusion of the study (12 months of extended wear). Additional studies are needed to examine the long-term effects of extended contact lens wear on epithelial proliferation and to identify whether the corneal epithelium shows significant physiological adaptation in maintaining a homeostatic environment. 

In this study, the proliferation rate of the epithelium as measured with BrdU labeling was not equally distributed across the normal cornea of the rabbit, confirming and extending the results of Ren et al. ²⁴ If the proliferation rate were identical in the central, peripheral, and limbal epithelial zones, an equal number of BrdU-labeled cells would be expected 24 hours after the injection. However, in each rabbit there was a distinct difference visible across the epithelium. The limbus was marked by a zone of very low BrdU labeling, whereas the immediately adjacent peripheral epithelium had a marked increased in labeling of epithelial cells, covering a narrow area approximately 1 to 1.5 mm wide. BrdU labeling decreased gradually toward the center of the normal corneal epithelium to a baseline level higher than in the limbus. Lavker et al. ³⁶ have shown in the rat that the labeling index of the limbal epithelium is lower than that of the central corneal epithelium. Similarly, Håskjold et al. ³⁰ found, by using tritiated thymidine labeling, a lower labeling index in the limbus over a 24-hour time span, and Ren et al. ²⁴ also found a lower limbal proliferation rate using BrdU labeling. Our results support these observations. The low BrdU labeling in the limbus appears to indicate a population of slow-cycling epithelial stem cells and is consistent with the current hypothesis that corneal stem cells reside in the limbus. ^{1 2}  

Our findings are in disagreement, however, with those studies reporting that the proliferation rate of the central and peripheral corneal epithelium in the rabbit and mouse are equal. ^{25 36 37} These prior studies did not identify the small band of increased labeling in the peripheral epithelium, immediately adjacent to the limbus. This apparent discrepancy with our present study results may be explained in part by species differences and inadequate sampling. In the present study, corneas were evaluated in whole flatmounted preparations, covering a much wider volume and containing significantly more epithelial cells than in histologic cross-section preparations. Furthermore, the time of labeling during the circadian cycle may have an impact on the labeling of the central and peripheral epithelial cells. Håskjold et al. ³⁰ showed that more tritiated thymidine-labeled cells were found in the periphery between 9.00 AM and 1.00 PM than in the central epithelium, whereas both were equal between 9.00 PM and 5.00 AM. They also demonstrated that the mitotic rate ³⁰ and percentage of tritiated labeled cells ³⁸ over 24 hours was higher in the periphery adjacent to the limbus than in the central corneal epithelium. 

Alternatively, it is possible that the proliferation rate of the superior peripheral epithelium, as assessed in this study, differs from the lateral and medial peripheral epithelium. Wiley et al. ³⁹ detected regional heterogeneity in the human cornea with AE1-positive– and AE5-negative–stained basal epithelial cells extending from the limbus into the superior and inferior peripheral epithelium but less in the lateral–medial meridian. Lauweryns et al. ⁴⁰ also found predominately basal epithelial cells in the superior peripheral meridian with phenotypes similar to the limbal basal cells. By contrast, almost none were found in the medial and temporal peripheral epithelium, indicating regional differences across the meridians of the peripheral corneal epithelium. Using BrdU and triated thymidine double labeling, Lehrer et al. ³³ demonstrated in the mouse cornea that 8% of the peripheral epithelial cells adjacent to the limbus were capable of undergoing at least two divisions within 72 hours, whereas double-labeled cells were rarely detected in the central epithelium up to 96 hours. In addition, in vitro cell culture experiments showed that the human peripheral epithelial cells are capable of undergoing more population doubling before becoming terminally differentiated than central epithelial cells. ⁴¹ Considering these results together, it is therefore reasonable to propose that the peak of BrdU-labeled cells in the periphery of the rabbit cornea reflects a higher overall division rate in the peripheral epithelium than in the central epithelium, at least in the superior region. 

Lavker et al. ³⁶ questioned the “driving-force” hypothesis for centripetal migration caused by a (dividing) population pressure in the limbus. Although the cell proliferation rate in the limbal epithelium is lower than in the central epithelium, our finding of increased labeling in the peripheral corneal epithelium, immediately adjacent to the limbus, may represent a driving force for migration of epithelial cells toward the center of the corneal epithelium. It has also been suggested that a strong adhesion of epithelial basal cells to the basement membrane in the limbus resists lateral population pressure in the conjunctival direction. ⁴² In this model of population pressure, the loss of superficial epithelial cells may remain an important factor affecting centripetal movement and upward differentiation of corneal epithelial cells as proposed by Lavker et al. ³⁶ In the epidermis for example, an increased loss of superficial keratinocytes resulted in an increased proliferation rate and turnover time, ⁴³ and even though no direct link between epithelial exfoliation and proliferation rate in the cornea has been demonstrated thus far, both are simultaneously suppressed during overnight contact lens wear. ^{8 9 10 23 24} The question remains whether decreased exfoliation is a consequence of decreased proliferation or whether the roles are reversed: Does the decreased exfoliation negatively affect basal cell proliferation? Further experiments on the relationships between epithelial cell exfoliation and proliferation in the cornea are needed to determine how epithelial renewal in the cornea is controlled. 

There are limitations in extrapolating our results to humans wearing RGP contact lenses. First, the rabbits were fitted with RGP lenses with large diameter (14.0 mm), whereas humans wear RGP lenses with an average diameter between 8.5 and 10.0 mm. Therefore, a substantially larger corneal surface area in the rabbit was exposed to the test contact lens. Furthermore, a smaller lens also results in more overall oxygen supplied to the corneal epithelium. Secondly, rabbits blink approximately once every 6 minutes, ⁴⁴ much less than humans with a blinking frequency of nearly 12 per minute. ⁴⁵ The flushing of the tear film between the corneal epithelium and the contact lens is thus much more frequent (blinking) and more efficient (smaller diameter contact lens) in humans. The effect of tear film stagnation (debris build-up) on the proliferation rate is not known. Third, the established difference in lactate dehydrogenase (LDH) isozymes expression between the rabbit (Heart-4) and human cornea (Muscle-4) suggests that glycolysis in the rabbit cornea is more dependent on an aerobic metabolism, and the rabbit cornea may therefore react more rapidly to hypoxia than the human cornea. ⁴⁶  

A previous study reported that keratocytes in the normal cornea show no labeling with BrdU unless the epithelium is mechanically removed, ⁴⁷ confirming widespread conventional belief that keratocytes do not actively divide. Although, it was not the purpose of this study to examine the proliferation rates of other corneal cells, it was interesting to note the obvious keratocyte labeling with BrdU after RGP contact lens wear, demonstrating active keratocyte mitosis. Based on these new findings, we propose that keratocyte mitosis after RGP lens wear may be required to replenish lost keratocytes. Indeed, Jalbert and Stapleton ⁴⁸ have reported an estimated decrease in keratocyte density with long-term extended contact lens wear in humans, and Yamamoto et al. ⁴⁹ observed TUNEL-positive (apoptotic) keratocytes in the rabbit cornea after RGP lens wear. 

 Supported in part by EY10738 (HDC), The Pearle Vision and Chilton Foundations, Dallas, Texas; Senior Scientist Awards (JVJ, HDC); the Olga Keith Weiss Scholar Award, and an unrestricted grant from Research to Prevent Blindness. PML was a recipient of a William C. Ezell Fellowship from the American Optometric Foundation, Rockville, Maryland, during part of the study.

 Submitted for publication March 20, 2001; revised June 18, 2001; accepted July 11, 2001.

 Commercial relationships policy: E (KY); C (HDC); N (all others).

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked“ advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

RGP Test Lenses

Lens Type	Materials	Dk*	Dk/t, †	EOP, ‡	Diameter (mm)	Thickness (mm)	Base Curves (mm)
Low-Dk test lens	SMA/MMA	15	10	5.76	14.0	0.15	7.60–8.00
Hyper-Dk test lens	SiSt/FMA	146	97	19.13	14.0	0.15	7.60–7.80

 FMA, fluoromethacrylate; MMA, methylmethacrylate; SiSt, siloxanylstyrene; SMA, siloxanylpropylmethacrylate.

*  Oxygen permeability coefficient (cm/sec)(mL O₂/mL mm Hg) × 10⁻¹¹.

^†  Dk/t thickness of the lens.

^‡  Equivalent oxygen percentage; 21% is the normal value at sea level under open-eye conditions.

Figure 1.

 

The number of BrdU-labeled cells in the normal cornea 24-hours after BrdU injection. Filled squares represent 15 images of the cornea, from the limbus to the central corneal epithelium. Slight BrdU labeling was observed in the limbus (image 1) and peak labeling at the adjacent peripheral epithelium (image 3). Bar, 50μ m.

Figure 2.

 

BrdU-labeled epithelial cells at day 1 in the central cornea of (A) a low-Dk/t lens–wearing eye, (B) a hyper-Dk/t lens–wearing eye, and (C) a control eye. Bar, 50μ m.

Figure 3.

 

BrdU labeling of the low-Dk/t group and its paired control at day 1 after injection. Symbols are situated to correspond with 15 images of the cornea, from the limbus to the central corneal epithelium.

Figure 4.

 

Low-Dk/t group and control at day 3. Symbols are situated to correspond with 15 images of the cornea, from the limbus to the central corneal epithelium.

Figure 5.

 

Low-Dk/t group and control at day 7. Symbols are situated to correspond with 15 images of the cornea, from the limbus to the central corneal epithelium.

Figure 6.

 

Hyper-Dk/t group and control at day 1. Symbols are situated to correspond with 15 images of the cornea, from the limbus to the central corneal epithelium.

Figure 7.

 

Hyper-Dk/t group and control at day 3. Symbols are situated to correspond with 16 images of the cornea, from the limbus to the central corneal epithelium.

Figure 8.

 

Hyper-Dk/t group and control at day 7. Symbols are situated to correspond with 16 images of the cornea, from the limbus to the central corneal epithelium.

Figure 9.

 

(A) BrdU labeling of keratocytes at day 3 after injection in the low-Dk/t RGP lens group (small arrows). (B) BrdU-labeled keratocytes were out of focus during examination of the epithelium (small arrows) in the same image as is shown in (A). Bar, 50 μm.