Collagen Cross-Links Reduce Corneal Permeability

Abstract

purpose. To investigate the relationship between corneal permeability and nonenzymatic cross-link density.

methods. Corneas were dissected from 90 cadaveric porcine eyes. Samples were incubated for 24 hours with control solution or methylglyoxal at concentrations of 0.01%, 0.10%, and 1.00%. Nonenzymatic cross-link density in treated and control groups was quantified by papain digest and fluorescence spectrophotometry. Control and treated corneas were mounted in a customized Ussing-type chamber connected to vertical tubing, and specific hydraulic conductivity was determined according to the descent of a column of degassed saline at room temperature. Permeability to diffusion of fluorescein in a static chamber was determined for a similar set of corneal samples.

results. Methylglyoxal treatment effectively increased nonenzymatic cross-link content, as indicated by the average fluorescence for each group. Specific hydraulic conductivity (m²) was reduced with increasing cross-link density. Similarly, the permeability coefficient for the fluorescein solute consistently decreased with increasing methylglyoxal concentration, indicating diffusion impedance resulting from the treatment.

conclusions. Nonenzymatic cross-link density in the cornea can be significantly increased by treatment with methylglyoxal. Porcine cornea showed a nonlinear reduction in solute permeability and specific hydraulic conductivity with increasing cross-link density. This model suggests that age-related nonenzymatic cross-link accumulation can have a substantial impact on corneal permeability.

The cornea is a critical component of the eye’s collagenous shell. It contains an organized network of collagen whose structure and arrangement, along with tight regulation of water content, allow the tissue to remain transparent. The cornea relies on diffusion as the primary mechanism of nutrient flow to the epithelium and to keratocytes tasked with maintaining the highly anisotropic collagen fibrils in extracellular matrix. ^{1 2} Fluid and nutrients flow through and across the cornea from the aqueous humor and the limbus. ^{3 4} Epithelial and endothelial cell layers also have important roles in regulating permeability across the cornea. ⁵ Vision depends on corneal clarity, and there is thus great clinical interest in factors affecting corneal health. 

Beyond its role as a transparent barrier and a refractive element in the visual system, the cornea is an important route for permeation of topically administered drugs into the eye, in addition to other pathways such as transconjunctival absorption. ⁶ Topical delivery in the form of eyedrops remains a mainstay of ophthalmic pharmacotherapeutics. Thus, factors affecting the diffusion of molecules through the cornea are of interest when considering drug design, patient selection, and dosing strategies. 

As with other tissues, aging of the cornea results in the accumulation of collagen cross-links, both enzymatic (lysyl-oxidase dependent) and nonenzymatic (glycation). ⁷ Glucose is believed to encourage nonenzymatic cross-links that are formed through Maillard-type reactions. ⁸ Persons with diabetes often experience increased nonenzymatic cross-link density in collagenous tissues such as the cornea because of higher serum glucose levels throughout life. ⁹ In the sclera, collagen cross-links have been found to be a determinant of nutrient diffusion. ^{10 11} It seems likely that the permeability of the corneal stroma would be similarly affected by collagen cross-linking. Elderly persons and those with diabetes constitute a disproportionate percentage of the ophthalmic patient population ^{12 13 14} ; therefore, factors that could affect corneal permeability in this population are highly relevant clinical considerations. The relationship between these natural and pathologic cross-links and corneal permeability has not been adequately investigated. 

We hypothesized that increasing collagen cross-link density in the cornea significantly decreases permeability and solute diffusion. To investigate whether nonenzymatic cross-links are determinants of permeability, we treated porcine corneas with methylglyoxal (MG), a Maillard pathway intermediate, ¹⁵ and quantified specific hydraulic conductivity and permeability to fluorescein diffusion. 

Methods

Tissue Harvest and Treatment

Ninety freshly excised porcine eyes (age range, 6–8 months; Sierra for Medical Science, Whittier, CA) were used in these experiments. This model was chosen because the porcine cornea is anatomically similar to the human cornea though it is slightly larger and thicker (average, 17–19 mm horizontally × 14–16.5 mm vertically and approximately 1 mm thick). ¹⁶ The eyes were delivered in sealed bags on ice. The cornea was excised with a circumferential dissection around the limbus. Some sclera (1–2 mm) was left attached to the cornea to retain the integrity of the tissue. Adherent iris or ciliary body inside the eye was gently coaxed off with a sterile sponge. The corneal epithelium was left intact. All corneas were excised and refrigerated within 48 hours of death. Corneas used in control and treatment groups were used at the same time point after harvest. The central corneal thickness for each sample was measured with a digital caliper (CD-8 PS; Mitutoyo Corp., Kawasaki, Japan) three times and averaged. Only data collected from corneal tissue were used in this study. 

Corneal samples were incubated at 37°C in 5% CO₂ atmosphere for 24 hours in Dulbecco’s phosphate-buffered saline (PBS), containing 1% penicillin-streptomycin and protease inhibitor cocktail tablets (10213200; Roche, Mannheim, Germany) before their use in experiments. Selected samples also received 0.01%, 0.10%, or 1.00% MG (Sigma, St. Louis, MO). Additionally, horizontal corneal strips 15 mm × 7 mm were dissected and subjected to the same treatments, and mechanical residual stress was evaluated visually as confirmation of cross-linking. 

Corneal Hydration

Variations in corneal weight after treatment were measured by comparing the posttreatment weight to the pretreatment weight. Before incubation in the treatment solutions, samples were blotted dry, weighed, and measured (for thickness). After 24 hours in the treatment solution at 37°C, the corneas were removed, blotted, and reweighed, and the percentage difference was measured as the ratio of the difference between posttreatment and pretreatment weight to the pretreatment weight. ¹⁷ A digital scale (AM50; Mettler Instruments Corp., Hightstown, NJ) with an accuracy of 0.1 mg was used to weigh the tissue. Thickness was measured again after treatment. 

The effects of the treatment solutions on corneal hydration were measured in a similar fashion. Tissue samples were blotted dry and weighed before incubation in the treatment solutions. After 24 hours in the treatment solution, the corneas were removed, blotted, and reweighed (posttreatment wet weight). Samples were then dehydrated at 70°C for 24 hours and reweighed (posttreatment dry weight). To determine the effects of cross-linking on the hydration of cornea, corneal hydration was calculated as the ratio of the difference between the posttreatment wet weight and the posttreatment dry weight to the posttreatment wet weight. Variations in corneal hydration were compared between treatment groups to control (0% MG) cornea to calculate the percentage change in hydration. 

Cross-Link Assay

Nonenzymatic glycation-type cross-links form as the result of the attachment of the carbonyl group of glucose, followed by a ketoamine rearrangement. The cross-links formed because of the Maillard-type reactions that followed, then yielded fluorophores characteristic of a reaction of a sugar with a protein. Therefore, fluorescence of a papain digest supernatant in our experiment is an indicator of nonenzymatic cross-link content. ^{15 18 19}  

Tissue samples were cut from cornea, weighed, and placed in sealed plastic bags at −80°C for no longer than 2 weeks. Only samples weighing between 10 and 25 mg were used. A minimum of 10 mg was used to ensure that at least three aliquots of digest could be generated. A maximum of 25 mg was used to maintain a local measurement of cross-link content. Before execution of the assay, the tissue was pulverized to a fine powder to encourage a thorough reaction with the papain. This was accomplished by transferring the tissue to a test tube, submerging that tube in liquid nitrogen for 2 minutes to further cool the tissue to a mechanically brittle state, transferring the tissue to a milling capsule containing a pair of 12-mm diameter stainless steel balls (also previously submerged in liquid nitrogen for 2 minutes), cyclically displacing the milling capsule and its contents in a milling machine (Retsch MM301 ball mill; Retsch Corporation, Haan, Germany) for 3 minutes at 3 0Hz, and weighing a portion of the pulverized tissue for use in the assay. 

To each pulverized tissue sample was added 1 mL of a 10 mM solution of dithiothreitol in a digestion buffer (0.1 M Na acetate/2.4 mM disodium EDTA). To each tissue sample was then added 20 μL of a 1 mg papain/1 mL digestion buffer solution. The tissue was incubated at 60°C for 12 hours, and 20 μL of the papain solution was added again. After 12 more hours of incubation, the tubes were centrifuged at 10,000 rpm for 12 minutes. The supernatant was removed and used to perform fluorescence measurements. 

To measure fluorescence, a 100-μL aliquot of each papain supernatant was removed to a black ELISA plate, and fluorescence was measured in the plate reader. A spectral scan was used to optimize excitation and emission wavelengths for fluorescence assays. The scan was set to wavelengths 50 nm below and above the tentative excitation and emission wavelengths of 370 nm and 440 nm, respectively, obtained from the literature. ^{17 20} The peak maxima were used when expressing levels of fluorescence per milligram of the tissue. 

Diffusion Assay

In vitro diffusion studies were conducted with housing constructed from PVC unions (American Valve, Greensboro, NC). Corneal tissue was blotted dry and placed between two stainless steel annular platens with 6.35-mm central apertures. The tissue was mounted without stretching so as to minimize asymmetric stress. This assembly was mounted between the two reservoir chambers, and the union housing was threaded together to provide the compression necessary to prevent leakage around the specimen (Fig. 1) . The open ends of the union were sealed (Parafilm; Pechiney Plastic Packaging, Chicago, IL) after both chambers were filled with PBS and visually inspected for leakage. The apparatus was incubated at 37°C in a 5% CO₂ atmosphere for 1 hour to restore normal hydration and temperature. 

The medium in the reservoir facing the intraocular side of the cornea (receiver chamber) was replaced by 9 mL fresh PBS at 37°C, whereas the reservoir facing the extraocular side (donor chamber) was filled with an equal volume of PBS containing 1 mg/mL sodium fluorescein (Sigma, St Louis, MO). The apparatus was returned to the tissue incubator and placed under constant agitation using a rotary hybridization oven (Model 2000; Robbins Scientific Corp., Sunnyvale, CA). The chamber was removed from the incubator and shaken with a mixer (Vortex Genie 2; Fisher Scientific Co., Santa Clara, CA) before samples measuring 200 μL were removed from each reservoir at 30-minute intervals for 4 hours and stored in the dark at −80°C. 

Diffusion from the donor chamber to the receiver chamber was characterized by means of a permeability constant, P (cm/s). The apparent permeability coefficient, P, was calculated with the following equation:  

\[P{=}\ \frac{dQ}{dt}(\frac{1}{60AC})\]

where dQ/dt is the steady state rate of appearance of the fluorescein in the receiver chamber (μg/cm²/min), 1/60 is the conversion factor from minutes to seconds, A is the cross-sectional area of the specimen exposed to the solution (cm²), and C is the initial drug donor concentration (μg/mL). ²⁰ Fluorescence was measured at room temperature with a fluorescence spectrophotometer (SpectraMax M5; Molecular Devices, Sunnyvale, CA). Excitation and emission wavelengths were 492 and 520 nm, respectively. Standard curves of fluorescence versus concentration were obtained by serial dilution of fluorescein in PBS. Concentrations of samples were determined by linear regression analysis within the linear portion of the standard curve. The values of steady state flux were estimated from the linear portion of the cumulative amount of drug permeated (μg/cm²) versus time (min) profile in each case. All permeation studies were carried out in duplicate, and the results were expressed as mean ± SD. 

Specific Hydraulic Conductivity Determination

In vitro permeability studies were conducted using a modified version of the two-chamber system described (Fig. 1) . Tissue samples were mounted in the reverse fashion, with the intraocular side of the cornea facing the donor reservoir and the extraocular side facing the receiver reservoir because we were interested in studying fluid flow from the inside of the eye to the outside. Subsequent tissue mounting steps were similar to those described. 

The assembled apparatus was connected to identical fluid columns on both ends and filled with degassed PBS containing 1% penicillin-streptomycin and protease inhibitors. The system was visually inspected for fluid leakage and allowed to equilibrate for 1 hour. Periodically, the apparatus was tapped lightly to remove any accumulated air bubbles. Donor column height was set at 25 cm above the receiver column height, and the experiment was begun. This height was chosen because it created a pressure differential within a physiologic range (18 mm Hg). Typical tubing contained 1 mL/14.9 cm. The descent of the water column on the donor side was allowed to proceed for 24 hours. At the end of the trial, the difference in height of the water columns was recorded. The experiment was conducted at room temperature. 

Specific hydraulic conductivity (κ) of the tissue samples as derived from Darcy’s law ²¹ was calculated using the following equation:  

\[{\kappa}{=}\ \frac{Q{\mu}L}{A{\Delta}P}\]

where Q (m³/s), the flow rate, is the product of the internal cross-sectional area constant (6.71 × 10⁻⁶ m²) of the individual tubes and the change in height (Δh) of the water column to arrive at a flow volume divided by time, μ is the viscosity (Pa · s) of the diffusion medium under experimental conditions, L (m) is the corneal thickness, A (m²) is the area of the aperture, and ΔP (Pa) is the mean pressure head during the experiment. Results were expressed as mean ± SD. ^{21 22}  

Results

Cross-Linking Treatment and Corneal Hydration

Residual stress in treated corneal strips increased with methylglyoxal concentration, providing mechanical confirmation that cross-linking occurred as expected under the treatment conditions (Fig. 2) . Swelling of the cornea was noted in all tissues after treatment (Table 1) .The percentage increase in thickness compared with untreated tissue, however, remained relatively constant among all groups. The mean ± SD of pretreated corneal thickness measurements for the samples was 1.05 ± 0.03 mm. The percentage change in control cornea thickness (41.5% ± 0.5%) was not statistically different from the thickness of the cornea treated with varying MG (39.2% ± 3.1%, 38.0% ± 1.8%, 37.8% ± 1.98% for MG 0.01%, 0.10%, and 1.0%, respectively; P > 0.05). There were no statistically significant differences in the water content of control cornea (83.2% ± 1.3%) or cornea exposed to varying dosages of MG (84.2% ± 1.3%, 84.5% ± 0.5%, 80.3% ± 0.8%, for MG 0.01%, 0.10%, and 1.0%, respectively; P > 0.05). Similarly, the hydration of the cornea was not significantly altered by the addition of MG (Table 2) . The hydration of control cornea (4.96 ± 0.46 g water/g dry weight) was not statistically different from that of corneas treated with MG (5.38 ± 0.56, 5.47 ± 0.22, 4.10 ± 0.21 for MG 0.01%, 0.10%, and 1.0%, respectively; P > 0.05). 

Cross-Link Content Resulting from Methylglyoxal Treatment

The efficacy of the treatment regimen was verified by measuring nonenzymatic cross-link density by fluorescence from five separate cornea samples from each of four groups. Control samples (MG, 0.00%) had an excitation maximum at 275 nm and an emission maximum at 343 nm (Fig. 3) . In contrast, the fluorescent spectrum of samples treated with MG exhibited excitation and emission maximums at longer wavelengths. The excitation/emission maximums for the MG 0.01%, 0.10%, and 1.0% treatment groups were 321 nm/386 nm, 323 nm/391 nm, and 325 nm/394 nm, respectively. 

Results showed a nonlinear trend of increased fluorescence with increased MG concentration (MG 0.00%, 12.7 ± 2.1 fL/mg; MG 0.01%, 84.2 ± 29.0 fL/mg; MG 0.10%, 135.3 ± 19.2 fL/mg; MG 1.00%, 164.5 ± 23.1 fL/mg; P < 0.05). Statistically significant differences were found between the control group and all groups treated with varying concentrations of MG (Fig. 4) . 

Fluorescein Diffusion through Cornea

Twenty-four porcine eyes (n = 6 for each treatment group) were studied for the effects of MG treatment on the permeability to sodium fluorescein across the cornea. Constant flux of the compound began as early as 30 minutes after initial exposure. 

Increasing the concentration of the MG treatment was associated with a reduction of the permeability coefficient (Fig. 5) . At the higher concentrations of 0.10% MG (7.10 ± 1.95 × 10⁻⁷ cm/s) and 1.00% MG (3.50 ± 2.14 × 10⁻⁷ cm/s), this decline was significant (P < 0.05) compared with controls (1.03 ± 0.20 × 10⁻⁶ cm/s). At 0.01% MG (9.30 ± 1.45 × 10⁻⁷ cm/s), this trend of reducing the permeability coefficient did not reach statistical significance (P = 0.35) compared with controls. 

Specific Hydraulic Conductivity

Fifteen samples (n = 4 for each test group except 1.0% MG; n = 3) were used to study the effects of increased MG treatment on specific hydraulic conductivity in porcine cornea. 

As was anticipated given the results of the diffusion experiments, increasing the concentration of the MG treatment was associated with reduced specific hydraulic conductivity (Fig. 5) . This trend was most pronounced at the highest concentration of 1.00% MG (5.34 ± 0.84 × 10⁻¹⁷ m² vs. 8.20 ± 1.59 × 10⁻¹⁷ m², for the control group; P < 0.05). Mean specific hydraulic conductivity for groups treated with 0.01% and 0.10% MG was reduced but did not reach statistically significantly differences with the control group (6.75 ± 1.65 × 10⁻¹⁷ m² and 6.29 ± 1.39 × 10⁻¹⁷ m²; P = 0.25 and P = 0.12, respectively). 

Discussion

The objective of this study was to investigate the effect on corneal permeability of nonenzymatic cross-links, which accumulate in collagenous tissues with age under normal circumstances and to a greater extent with diseases such as diabetes. We have shown that nonenzymatic cross-links influence the permeability characteristics of porcine cornea in a concentration-dependent manner. This was found to be true for solute diffusion and fluid flow across the cornea. The effect of increased cross-link content was statistically independent of changes in hydration, another known determinant of permeability. ²³  

It is likely that cross-links reduce corneal permeability by constraining the interstitial space in the stroma. Considering cornea as a basic porous medium, as described in the model by Bear and Bachmat, ²⁴ permeability of a given solute depends on three fundamental properties: porosity, average medium conductance (channel size), and average tortuosity (complexity of pathways through the matrix). Collagen fibril architecture in the stroma affects all these variables. Bonds resulting from cross-links impose constraints on collagen fibrils. These constraints can be expected to reduce conductance and to increase average tortuosity. Both these effects tend to decrease permeability. 

Our diffusion experiment focused on the effects of cross-links for a constant small fluorescein solute size (376 Da; radius, 0.5 nm) to test the sensitivity of the diffusion coefficient to changes in methylgyoxal treatment concentration. (Of note, the fluorescein diffusion data from our untreated control group was similar in magnitude to previously reported results. ^{25 26} ) If solute size were increased, we would expect that permeability would decrease exponentially. Ambati et al. ²⁷ measured an exponential decline in permeability with the increasing solute molecular radius of soluble dextran. ²⁷ We have no reason to suspect that the effects of cross-linking would invalidate this exponential decline with increasing solute size. Increases in cross-link content simply decrease the overall level of permeability for a given solute size. Given our small fluorescein solute size, we might expect a small pharmacologic agent, such as dexamethasone (393 Da; radius, 0.5 nm), to behave similarly but for larger molecules such as proteins to have permeability levels in cross-linked tissues dramatically lower than those measured in this study. 

Comparison of our cornea data with similar previously reported results ¹⁰ from porcine sclera suggested the possibility that tortuosity and conductance may be affected asymmetrically by cross-links given different levels of order in the respective collagen networks of these tissues. Collagen fibrils in untreated cornea are aligned and tend to have 30% higher average permeability coefficient and 90% higher average specific hydraulic conductivity than in untreated sclera. These results are likely attributed to higher conductance and lower tortuosity inherent in the cornea’s ordered collagen structure. ^{28 29} We speculate that when exogenous cross-links are added to the cornea, a pronounced effect on conductance results whereas tortuosity is not as significantly affected. Expressed in other terms, cross-linking the cornea tends to reduce the average diameter along a through-path but does not have as pronounced an effect on path length or complexity. Given that the diffusion behavior of a small solute is presumably more affected by changes in conductance than changes in tortuosity, ^{30 31} we would expect cross-linking to reduce the permeability constant more in cornea than in sclera. This is indeed the case; we found that treating sclera ¹⁰ and cornea with 1% methylgyoxal for 24 hours reduces the permeability constant on average by 39% and 71%, respectively (P < 0.01 for each). In contrast, cross-linking the disordered collagen fibrils in sclera has a more equal effect on tortuosity and conductance. This makes intuitive sense because the sclera is more isotropic. Given that tortuosity drives specific hydraulic conductivity more than does change in conductance, ³² we would expect cross-linking to reduce specific hydraulic conductivity in the sclera more than in the cornea. Our results seem to confirm this rationale; 1% MG treated sclera and cornea have specific hydraulic conductivity reductions of 78% and 35%, respectively (P < 0.01 for each). Our results indicate that cross-linking ordered structures has more of an effect on diffusion behavior and less of an effect on pressure-driven flow behavior compared with the effects seen from similar cross-linking in disordered structures. 

We anticipate that our model achieves a clinically relevant cross-link density in the treated corneas, but no comparable data exist on the levels of nonenzymatic cross-links in aged human corneas. Sady et al. ⁹ reported fluorescence levels in normal and diabetic corneas in patients 55 to 80 years of age. Unfortunately, their study used collagen extraction measures different from ours and measured fluorescence at incongruent excitation and emission frequencies, rendering quantitative comparisons with our data infeasible. In porcine sclera, we found that treatment with 1% MG achieved a cross-link content on the order of aged human sclera. ^{10 33} Assuming that similar cross-link accumulation occurs in cornea and sclera in vivo and that young porcine cornea and young porcine sclera have a similar capacity to acquire cross-links, it seems likely that the corneal cross-link content achieved by higher MG concentrations in this study is a reasonable approximation of that found in aged human corneas. 

Beyond aging and disease, the effect of collagen cross-linking in the cornea is also clinically relevant in another context: exogenous corneal cross-linking with riboflavin and ultraviolet light is used as a treatment strategy for corneal ectasia and keratoconus, often in young patients. ^{34 35 36} No data exist on the effects of these treatments on corneal permeability, but, based on the findings of this study, it seems likely that permeability may be substantially reduced, yielding patients less able to receive medical treatments topically or potentially interfering with nutrient diffusion. Further work is necessary to confirm this and to identify possible clinical consequences, especially in light of the fact that these patients will accumulate even more cross-links in their corneas later in life. 

A limitation of this model for age-related cross-link accumulation is that it only accounted for one type of age-associated cross-links. Cross-links that accumulate in human collagenous tissues are enzymatic (lysyl-oxidase dependent) and nonenzymatic (glycation) in origin. In this study, we used methylglyoxal treatment to increase nonenzymatic cross-link levels as a means of approximating the effect of increasing all types of cross-links because levels of enzymatic and nonenzymatic cross-links are directly correlated. ³⁷ With age, both types of cross-links increase, whereas in diabetes, nonenzymatic cross-links are elevated. ⁹  

An additional limitation of this study is that it used a model of aging rather than actual aged tissue. The limited availability of aged porcine corneas makes direct comparison of young with old tissue difficult. Nevertheless, the model itself could prove useful in future studies because we found that increasing the methylglyoxal concentration of the treatment solution yielded higher fluorescence levels in the cornea in a predictable manner that correlated with nonenzymatic cross-link content. ¹⁹ This could be an appropriate technique for evaluating permeability or other properties while modeling the effects of age or disease on the cornea. 

Although it seems likely that the cellular- and tight junction-mediated barrier effects of the corneal epithelium and endothelium would be more robust in vivo than in this cadaver eye model, evidence suggests that keeping the epithelium intact in an ex vivo system allows the maintenance of a substantial diffusion barrier. ³⁸ Therefore, we believe that this experimental approach offers a reasonable means of approximating the permeability effects of modulating properties of the collagenous stroma. 

We identified a substantial reduction of corneal permeability to fluid flow and solute diffusion in treated samples in this study. These findings suggest that in elderly or diabetic patients, transcorneal penetration of medications may be significantly impaired compared with penetration in younger patients. In such patients, topical absorption of medications may be reduced or the conjunctiva may be a more important pathway for absorption. ³⁹ The health of anterior corneal structures may also be compromised because of reduced diffusion through the stroma. Further studies are necessary to elucidate the in vivo consequences of cross-link–associated permeability changes in the cornea. 

 Supported by That Man May See, Inc., and Research to Prevent Blindness.

 Submitted for publication August 14, 2008; revised October 25 and November 22, 2008; accepted February 20, 2009.

 Disclosure: J.M. Stewart, None; D.S. Schultz, None; O.-T. Lee, None; M.L. Trinidad, None

 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.

 

Figure 1.

 

Permeameter (left) and diffusion chamber (right) experimental setups: (A) corneal sample; (B) annular platens; (C) O-rings; (D) union housing; (E) syringe; (F) wax film seal.

Figure 2.

 

Corneal strips treated with 0% (A), 0.01% (B), 0.1% (C), or 1% (D) methylglyoxal. The gradation in cross-link content is reflected in the residual stresses that stiffen the tissue when it is cantilevered.

Table 1.

 

Variation of Corneal Weight and Thickness after Treatment

Treatment (% MG)	Pretreatment Weight/Thickness (g, mm)*	Posttreatment Weight/Thickness (g, mm)*	Difference (g, mm)	Difference (%)
Control, 0	0.54 ± 0.05	0.68 ± 0.08	0.14	20.47
	1.02 ± 0.02	1.75 ± 0.01	0.73	41.53
0.01	0.55 ± 0.11	0.76 ± 0.05	0.21	27.90
	1.06 ± 0.04	1.75 ± 0.03	0.69	39.20
0.1	0.59 ± 0.06	0.79 ± 0.08	0.21	26.22
	1.07 ± 0.04	1.73 ± 0.02	0.66	37.96
1.0	0.60 ± 0.04	0.68 ± 0.03	0.08	11.37
	1.06 ± 0.01	1.71 ± 0.05	0.65	37.78

  n = 3 for all treatment groups.

*  Mean ± SD.

Table 2.

 

Water Content Evaluation of Control and Treated Porcine Corneal Specimens

Treatment (% MG)	Water Content (%)*	Hydration (g water/g dry weight)*	Change in Hydration (%)	P (Compared with Normal Hydration)
0.00	83.2 ± 1.3	4.96 ± 0.46	N/A	N/A
0.01	84.2 ± 1.3	5.38 ± 0.56	8.6	0.36
0.10	84.5 ± 0.5	5.47 ± 0.22	10.3	0.25
1.00	80.3 ± 0.8	4.10 ± 0.21	−17.3	0.26

  n = 3 for all treatment groups. N/A, not available.

*  Mean ± SD.

Figure 3.

 

Plot of the effects of methylglyoxal treatment on the excitation (left) and emission (right) spectra of corneal samples (MG 0.01%, n = 3; all other groups, n = 5).

Figure 4.

 

Plot of corneal nonenzymatic cross-link content as indicated by fluorescence spectroscopy versus methylglyoxal treatment concentration (n = 5 for each group; error bars = SD).

Figure 5.

 

(A) Box plot with horizontal lines that denote the upper quartile, lower quartile, and median values for permeability coefficients of treated porcine cornea (n = 6 for each group). (B) Similar box plot for specific hydraulic conductivity of treated porcine cornea (n = 4 for each group). Error bars = SD. *P < 0.05; statistically significant difference from control group.