Effects of Gelatin Hydrogel Containing Chymase Inhibitor on Scarring in a Canine Filtration Surgery Model

Abstract

Purpose.: To investigate the effects of gelatin hydrogel (GH) containing a chymase inhibitor (CI) on intraocular pressure (IOP) and conjunctival scarring in a canine model of glaucoma surgery.

Methods.: Glaucoma surgery models were made in beagles. As the first experiment, GH was implanted. IOP was measured for 4 weeks, followed by histologic evaluation. As the second experiment, GH containing a CI or GH alone was implanted. IOP and bleb features were evaluated for 12 weeks. The densities of proliferative cell nuclear antigen (PCNA)–positive cells, fibroblasts, and mast cells were quantified. The mRNA levels of transforming growth factor–β (TGF-β) and chymase were determined by real-time PCR.

Results.: In the first experiment, IOP was significantly lower in the eyes treated with GH than that in the control eyes. The conjunctival area normalized by the scleral area was reduced in the treated eyes. In the second experiment, IOP reduction was maintained for 12 weeks in the eyes treated with GH containing a CI, but not in the eyes treated with GH alone. In addition, the bleb score was larger, whereas the adhesion score and densities of PCNA-positive cells, fibroblasts, mast cells, and chymase-positive cells were lower in the eyes treated with GH containing a CI. The mRNA levels of TGF-β and chymase were significantly decreased in the eyes treated with GH containing a CI.

Conclusions.: Implanting GH alone maintained IOP reduction, whereas GH containing a CI enhanced the IOP-reducing effect by suppressing cell proliferation. This drug delivery system might be useful for maintaining filtering blebs for a longer duration after glaucoma surgery.

Trabeculectomy, a major filtering surgery, has been widely performed on patients with glaucoma to obtain a sufficient reduction of intraocular pressure (IOP). However, filtration bleb dysfunction often occurs due to wound healing by Tenon's capsule tissue, which leads to subconjunctival fibrosis. ^1,2 The proliferation and migration of Tenon's fibroblasts to the wound site could in part explain why filtration wounds scar and lead to failure of filtration surgery. ³ Application of mitomycin C (MMC) has greatly improved the results of glaucoma surgery, through strongly suppressing proliferation of fibroblasts. However, by making thin blebs with MMC there is an increased risk of complications such as infectious endophthalmitis. ^{4 –6} Therefore, a more physiological approach to suppressing fibroblast proliferation is needed. 

Previous reports suggest that mast cells are involved in inflammation, scarring, and adhesion formation. ^{7 –9} It was demonstrated that intraperitoneal adhesion formation was reduced in mast-cell–deficient mice compared with normal control mice. ⁷ Mast-cell stabilizers were effective in attenuating adhesion formation in rat surgical adhesion models. ^8,9 Oshima et al. ¹⁰ reported that tranilast, a mast-cell stabilizer, directly inhibited cell proliferation of fibroblasts established from rabbit Tenon's capsule. In a clinical study, topical instillation of tranilast significantly alleviated ischemia of the filtering bleb and reduced IOP. ¹¹ Therefore, mast cells also represent a target for development of antiscarring agents that can be used after trabeculectomy. Chymase is a chymotrypsin-like serine protease contained in the secretory granules of mast cells. ¹² Chymase has been known to induce the accumulation of neutrophils, eosinophils, and other inflammatory cells. ¹³ Further, it was shown that human chymase induced the cell growth of human fibroblasts via upregulation of transforming growth factor (TGF)–β. ¹⁴ Using a hamster model of glaucoma surgery, we previously demonstrated that chymase activity and the number of chymase-positive mast cells in conjunctival tissues were significantly increased during the wound-healing process. ¹⁵ We also reported that dog chymase significantly increased cell proliferation of cultured canine Tenon's capsule fibroblasts, and that scores for adhesion degree in the chymase inhibitor (CI)–treated eyes were significantly decreased in the canine surgical model. ¹⁶ In addition, mast cells and chymase-positive cells were reportedly increased in a monkey trabeculectomy model. ¹⁷ These reports suggest that chymase may play an important role in the development of scarring after glaucoma surgery and that its inhibition may directly prevent the scarring with fewer complications than that of the currently used antimetabolites, such as MMC or 5-fluorouracil. 

On the other hand, there have been several investigations into subconjunctivally implanted drug delivery systems (DDSs) that would provide a localized and sustained release of antiproliferative drugs over an extended period after glaucoma surgery. ^{18 –26} The DDSs investigated to date for that purpose include collagen implants, ¹⁸ biodegradable polymers, ^{19 –23} nonbiodegradable polymers, ²⁴ liposomes, ²⁵ microspheres, ²⁶ and drainage devices. ²⁷ However, none of these has yet been developed for clinical use. Gelatin hydrogel (GH) is one of the biodegradable materials useful for the DDS of bioactive proteins in other fields of medicine. ²⁸ Using the GH, the controlled release of bioactive growth factors over a time range of 5 days to 3 months has been possible. ^29,30 Various growth factors, including basic fibroblast factor (bFGF) and transforming growth factor (TGF)–β1, have been incorporated in GH, and their controlled release has been effective for regeneration therapy of various tissues. ^31,32 This DDS has been applied to clinical therapies, such as for severe skin lesions complicating autoimmune vasculitis syndromes, ³³ peripheral arterial diseases, ³⁴ and severe ischemic limb pain. ³⁵ On the other hand, implanting other biodegradable materials (solid hyaluronic acid film and honeycomb-patterned film) under the conjunctiva has been investigated for maintaining a filtering bleb, ^{36 –38} but not applied as the DDS. 

The aim of the present study was to test the effects of GH containing a CI, as a new DDS, on IOP reduction, filtration bleb formation, and cell proliferation in a canine model of glaucoma surgery. 

Materials and Methods

Materials and Drugs

The GH was constituted by chemically cross-linking acidic gelatin with glutaraldehyde. Glutaraldehyde (0.163%) was added to a 3% solution of basic gelatin (acid-treated porcine skin collagen; Nitta Gelatin, Osaka, Japan; isoelectric point, 9.0; average molecular weight, 99,000 kDa; water content, 97.8 wt%). A sheet of chemically bridged gelatin (2 mm thick) was obtained by placing this mixture on a polypropylene weighing plate (138 mm in diameter) and allowing it to stand at 4°C for 12 hours. Discs (diameter, 6 mm) were punched from this bridged hydrogel and treated with 100 mM glycine solution in water at 37°C for 3 hours to inactivate unreacted glutaraldehyde, then washed several times with distilled water and frozen. The freeze-dried GH was sterilized with ethylene oxide gas. 

A CI, Suc-Val-Pro-Phe^P(OPh)₂, was a gift from Jozef Oleksyszyn (Wroclaw University of Technology, Poland). ³⁹ A solution of 10 μM of the CI in PBS was dripped onto the GHs and left overnight at 4°C. 

Animals and IOP Measurement

Twelve beagle dogs (weighing 9.0 to 10.0 kg) were obtained from a commercial supplier (Japan SLC, Shizuoka, Japan). Dogs were fed regular chow, had free access to tap water, and were housed in an air-conditioned room at approximately 23°C and 60% humidity with a 12-hour light–dark cycle. The experimental procedures for animals were conducted in accordance with the ARVO Statement for Use of Animals in Ophthalmic and Vision Research. IOP was measured with a calibrated pneumatonometer (Model 30 Classic; Medtronic Solan, Jacksonville, FL) under general anesthesia with pentobarbital (35 mg/kg, administered intravenously [IV]) in a face-front position. 

Glaucoma Filtration Surgery Model

Dogs were anesthetized with an IV injection of pentobarbital (35 mg/kg body weight). A 10-mm fornix-based flap of conjunctiva and the Tenon's capsule (length, 5 mm) was made. After a 3 × 1-mm scleral portion was removed at the limbus, peripheral iridectomy was performed, followed by closing the conjunctiva with a 10-0 nylon suture. After surgery, 1 cm of 3 mg/g ofloxacin ointment was applied to the eye. 

Experimental Protocol

For the first experiment, 12 eyes of six dogs were used. In the treated eye, 5 × 5 mm of a GH was placed under the conjunctiva before closing the conjunctiva in the surgery. Nothing was placed in the fellow eye to serve as a control. IOP as well as bleb and adhesion scores were assessed weekly for 4 weeks, followed by histologic evaluation of the eye after dogs were euthanized by injecting a lethal dose of pentobarbital sodium. Conjunctival and scleral areas of the lesion were then measured, as described in the following text. 

For the second experiment, 12 eyes of six dogs were used. A 5 × 5-mm piece of GH containing CI was placed under the conjunctiva. A GH alone was placed in the fellow eyes. IOP and morphologic changes in bleb features were evaluated at 2, 4, 8, and 12 weeks postoperatively, followed by euthanizations, then the number of proliferative cell nuclear antigen (PCNA)–positive cells, fibroblasts, mast cells, chymase-positive cells, and angiotensin II–positive cells were quantified. 

For the additional experiment, 12 eyes of six dogs were used. Similarly to the second experiment, a 5 × 5-mm piece of GH containing CI was placed under the conjunctiva. A sponge (Material Quick Absorber [MQA]; Inami & Co., Tokyo, Japan), soaked with 0.04% MMC solution in distilled water, was placed under the conjunctiva for 5 minutes in the fellow eyes. Dogs were euthanized at 4 weeks postoperatively for histologic evaluation, to quantify vessels in the lesion as well as above-mentioned types of cells. 

Bleb Scores and Adhesion Scores

Blebs were examined via a slit lamp and were graded as previously described by Perkins et al., ⁴⁰ according to a qualitative scale of 1 to 4, reflecting increasing bleb height and size as follows: 1, minimal height, conjunctiva thickening, no microcysts; 2, microcysts present; 3, elevated bleb covering 2 to 3 clock hours of the eye; and 4, greatly elevated bleb covering >4 clock hours. A score of 0 indicated no observable bleb. 

The adhesion degree was determined by comparing the percentage of the area of adhesion to the whole flap area (5 × 10 mm²) as an index. ¹⁶ Scores were determined as follows: Score 1, the flap could be detached by lightly lifting it, or the area of adhesion was not >20%; Score 2, the area of adhesion was between 20% and 40%; Score 3, the area of adhesion was between 40% and 60%; Score 4, the area of adhesion was between 60% and 100%. 

Histology and Immunohistochemistry

Conjunctival and scleral tissue specimens were fixed with 10% buffered formalin and embedded in paraffin. Sections (5 μm thick) were cut, mounted on silanized slides (Dako Japan, Kyoto, Japan), and deparaffinized with xylene and a series of graded ethanol. Each section was stained with azan and toluidine blue to identify collagen fibers and mast cells, respectively. The conjunctival and scleral areas of the lesion where the flap was made were measured by using a computerized morphometry system (MacSCOPE, Ver 2.2; Mitani Co., Fukui, Japan) and the ratio of the number of pixels in the conjunctival area to that in the scleral area was determined. 

To retrieve the antigen, sections were pretreated with 10 mM citrate buffer (pH 6.0) and autoclaved for 15 minutes at 120°C before immunohistochemical staining with PCNA. Sections were soaked in absolute methanol containing 3% hydrogen peroxide for 30 minutes at room temperature to remove endogenous peroxidase activity. To suppress nonspecific binding, sections were incubated with 10% nonimmune goat serum for 10 minutes. Sections were then incubated with mouse monoclonal antibody against PCNA (PC10, M0879; Dako Japan) for 15 hours at 4°C. After being washed in PBS, the slides were incubated with biotin-conjugated goat anti-mouse immunoglobulin G (IgG) antibody for 30 minutes. After being washed with PBS, the sections were incubated with avidin–biotin–peroxidase complex (Dako Japan) for 30 minutes and washed once more with PBS. Finally, sections were incubated with 0.05% 3,3-diaminobenzidine. The slides were then washed in running tap water, counterstained with hematoxylin, and mounted in Canadian balsam. Nonimmunized mouse IgG was used as a negative control. No significant immunohistochemical reactions occurred in the control sections. 

To identify fibroblasts and vessels, anti-bovine vimentin (Wako, Osaka, Japan) and anti-human α-smooth muscle actin (α-SMA; Dako, Glostrup, Denmark) antibodies were used, respectively. To detect chymase-positive cells and angiotensin II–positive cells, anti-chymase antibody, which was contributed by Prof. George H. Caughey (University of California San Francisco, San Francisco, CA), and rabbit polyclonal antibody against angiotensin II (IgG Corp., Nashville, TN) were used, respectively. Sections were incubated for 1 hour at room temperature with each antibody, followed by reaction with appropriate reagents from a streptavidin–biotin peroxidase kit (Dako) and 3-amino-9-ethylcarbazole. The sections were lightly counterstained with hematoxylin. 

We counted PCNA-positive cells, fibroblasts, mast cells, chymase-positive cells, and angiotensin II–positive cells at the sites where they accumulated in the conjunctival and scleral lesions using a light microscope (number per ×100 field). The average number of each type of cell in five selected fields was calculated. We also quantified the number of vessels stained with anti–α-SMA antibody, similarly. 

Real-Time Polymerase Chain Reaction

The total RNA (0.5 μg) was transcribed into cDNA with reverse transcriptase (Superscript III) and random hexamers (both from Invitrogen, Carlsbad, CA). The mRNA was measured by RT-PCR with specialized software (LightCycler; Roche Diagnostics, Tokyo, Japan) using fluorogenic probes (TaqMan; Roche Molecular Systems). All primers and probes for RT-PCR of TGF-β, chymase, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were commercially designed (Roche Diagnostics). The primers were 5′-cattaacgggttcagttccag-3′ (forward) and 5′-agcaggaagggtcggttc-3′ (reverse) for TGF-β, 5′-atccctcagacccaagagg-3′ (forward) and 5′-ggaagctggatctttattgagg-3′ (reverse) for chymase, and 5′-aatgtatcagttgtggatctgacc-3′ (forward) and 5′-gcttcactaccttcttgatgtcg-3′ (reverse) for GAPDH. The probes were 5′-ggccacca-3′ for TGF-β, 5′-tccaggtc-3′ for chymase, and 5′-cctggaga-3′ for GAPDH. The mRNA levels of TGF-β and chymase were normalized to those of GAPDH. 

Statistical Analysis

Each measurement was expressed as the mean ± SD or SE. Statistical comparisons of multiple groups used repeated-measures ANOVA followed by other tests. Bleb and adhesion scores were evaluated in a nonparametric test and statistically analyzed by Mann-Whitney U test. Other parameters were evaluated by Student's t-test for paired or unpaired data. Differences were considered statistically significant at P < 0.05. 

Results

Effects of Gelatin Hydrogel Alone in a Canine Filtration Surgery Model

The initial values of IOP (mean ± SD, mm Hg) were 19.2 ± 2.7 in the eyes treated with GH alone and 19. 2 ± 2.4 in the control eyes. IOP was reduced 1 week after surgery in both treated and control eyes (Fig. 1). Two weeks after the surgery, IOP began to increase again in the control eyes, whereas it remained reduced in the treated eyes. IOP in the treated eyes was significantly lower than that in the control eyes 3 (P = 0.011, paired t-test) and 4 weeks (P = 0.021, paired t-test) postoperatively (P = 0.040, repeated-measures ANOVA). 

Figure 1.

 

IOP changes in the eyes treated with GH or the control eyes. Data are shown as the mean ± SD for six dogs. *P < 0.05, †P < 0.1 vs. controls (paired t-test).

Bleb scores were kept high at least until 4 weeks in the eyes treated with GH alone, whereas they were reduced from 3 weeks in the control eyes (Fig. 2A). Repeated-measures ANOVA revealed a significant difference (P < 0.01) between bleb scores of the eyes treated with GH alone and the control eyes. Bleb scores were significantly higher in the eyes treated with GH alone than those in the control eyes at 1 to 4 weeks (P = 0.007, 0.016, 0.004, 0.004, respectively; Mann-Whitney U test). The adhesion score was significantly lower in the eyes treated with GH alone compared with that in the control eyes (P = 0.028, Mann-Whitney U test; Fig. 2B). 

Figure 2.

 

Bleb score changes (A) and adhesion scores (B) at 4 weeks postoperatively in the eyes treated with GH or the control eyes. Data are shown as the mean ± SD for six dogs. **P < 0.01, *P < 0.05 vs. controls (Mann-Whitney U test).

As shown in Figures 3A and 3B, in the eyes treated with GH, the conjunctiva was less thickened compared with the control eyes. The ratio of conjunctival area to scleral area was significantly less in the treated eyes than that in the control eyes (P = 0.011, paired t-test; Fig. 3C). 

Figure 3.

 

Representative photomicrographs of the sections obtained from the eye treated with GH (A) and the control eye (B) 4 weeks after surgery and stained with azan stain. The conjunctiva and the sclera are surrounded by red and blue lines, respectively. Scale bars, 1 mm. The ratio of the conjunctival area to the scleral area in GH-treated and control eyes (C). Data are shown as the mean ± SD for six dogs. *P < 0.05 (paired t-test).

Effect of Gelatin Hydrogel Containing a Chymase Inhibitor in a Canine Filtration Surgery Model

The initial values of IOP (mean ± SD, mm Hg) were 14.5 ± 2.0 in the eyes treated with GH containing a CI and 14.4 ± 2.7 in the eyes treated with GH alone. The IOP reduction was maintained to almost the same level until 4 weeks postoperatively in both eyes (Fig. 4). In the eyes with GH alone, IOP returned to the initial level 12 weeks postoperatively, whereas IOP reduction was maintained in the eyes treated with GH containing a CI (P = 0.040, repeated-measures ANOVA). There were significant differences between the IOPs of the eyes treated with GH alone and those treated with GH containing a CI at 8 and 12 weeks (P = 0.039 and P = 0.002, paired t-test). 

Figure 4.

 

IOP changes by treatment with GH containing A CI or GH alone. Data are shown as the mean ± SD for six dogs. *P < 0.05, **P < 0.01 vs. GH alone (paired t-test).

Bleb scores were kept high until 8 weeks in the eyes treated with GH containing a CI, whereas they were reduced from 8 weeks in the eyes treated with GH alone (Fig. 5A). Repeated-measures ANOVA revealed a significant difference (P = 0.013) between bleb scores of the eyes treated with GH alone and those treated with GH containing a CI. Bleb scores were significantly higher in the eyes treated with GH containing a CI than scores in the eyes treated with GH alone at 8 and 12 weeks (P = 0.045, Mann-Whitney U test). The adhesion score was significantly lower in the eyes treated with GH containing a CI compared with that in the eyes treated with GH alone (P = 0.045, Mann-Whitney U test; Fig. 5B). 

Figure 5.

 

Bleb score changes (A) and adhesion scores at 12 weeks postoperatively (B) by treatment with GH containing a CI or GH alone. Data are shown as the mean ± SD for six dogs. *P < 0.05 vs. GH alone (Mann-Whitney U test).

As shown in Figures 6A and 6B, in the eyes treated with GH alone, the conjunctival surface was rough and thickened, whereas in the eyes treated with the GH containing a CI, the conjunctival surface was smooth and thickening was inhibited. The ratio of conjunctival area to scleral area was significantly lower in the eyes treated with the GH containing a CI than that in the eyes treated with GH alone (P = 0.011, paired t-test; Fig. 6C). 

Figure 6.

 

Representative photomicrographs of the sections obtained from the eyes treated with GH (A) and GH containing a CI (B) 12 weeks after surgery and stained with azan stain. The conjunctiva and the sclera are surrounded by red and blue lines, respectively. Scale bars, 1 mm. The ratios of the conjunctival area to the scleral area in the same eyes are shown in (C). Data are shown as the mean ± SD for six dogs. **P < 0.01 (paired t-test).

In the eyes treated with GH alone, many PCNA-positive cells were found, not all, but many of which were fibroblasts stained with antivimentin antibody (Figs. 7A, 7B). The densities of PCNA-positive cells and fibroblasts in the lesion were significantly decreased or had a tendency to be decreased in the eyes treated with the GH containing a CI, compared with the eyes treated with GH alone at 4 and 12 weeks postoperatively, and those were also decreased at 12 weeks compared with those at 4 weeks in both eyes (Figs. 7E, 7F). 

Figure 7.

 

Representative immunohistochemical staining of serial sections for PCNA (A, C) and vimentin (B, D) in eyes treated with GH alone (A, B) or GH containing a CI (C, D). Scale bars, 50 μm. Changes in the numbers of PCNA-positive cells (E) and fibroblasts (F) by treatment with GH containing a CI and GH alone. Data are shown as the mean ± SE for six dogs. **P < 0.01, *P < 0.05, †P < 0.1 (unpaired t-test).

Distributions of mast cells in the eyes treated with the GH containing a CI and the eyes treated with GH alone at 4 and 12 weeks are shown in Figures 8A–D. Their number seems to be obviously fewer in the eyes treated with the GH containing a CI than that in the eyes treated with GH alone. The number of mast cells was statistically decreased in the eyes treated with the GH containing a CI, compared with the eyes treated with GH alone at 4 and 12 weeks (Fig. 8E). The number of chymase-positive cells was also fewer in the eyes treated with the GH containing a CI than that in the eyes treated with GH alone at 4 weeks (Fig. 8F). 

Figure 8.

 

Distribution of mast cells. Representative photomicrographs of the sections obtained from the eyes treated with GH alone (A, C) and GH containing a CI (B, D) 4 weeks (A, B) and 12 weeks (C, D) after surgery and stained with toluidine blue. Scale bars, 50 μm. Changes in the numbers of mast cells (E) and chymase-positive cells (F) by treatment with GH containing a CI and GH alone. Data are shown as the mean ± SE for six dogs. **P < 0.01, *P < 0.05, †P < 0.1 (unpaired t-test).

The density of angiotensin II–positive cells was lower in the eyes treated with the GH containing a CI and the eyes treated with MMC than that in the control eyes and the eyes treated with GH alone (Fig. 9A), whereas the density of vessels was decreased in the eyes treated with MMC, but not in the eyes treated with the GH containing a CI (Fig. 9B). 

Figure 9.

 

Densities of angiotensin II–positive cells (A) and vessels (B) at 4 weeks in the control eyes, the eyes treated with GH alone (GH), GH containing a CI, and MMC. Data are shown as the mean ± SE for six dogs. ***P < 0.001, **P < 0.01, *P < 0.05 (unpaired t-test).

Expression of mRNA of TGF-β as well as that of chymase was significantly inhibited in the eyes treated with the GH containing a CI, compared with the eyes treated with GH alone, as shown in Figure 10. 

Figure 10.

 

Changes in the mRNA levels of TGF-β (A) and chymase (B) normalized to those of GAPDH by treatment with GH containing a CI and GH alone. Data are shown as the mean ± SE for six dogs. ***P < 0.001, *P < 0.05 (unpaired t-test).

Discussion

The present study demonstrated for the first time that implanting a GH alone maintained IOP reduction in a canine glaucoma surgery model and that the GH containing a CI, as a new DDS, extended the period of filtration bleb formation and IOP reduction by suppressing cell proliferation in the lesion. 

We have previously shown that angiotensin-converting enzyme is involved in scarring in the animal model of glaucoma filtering surgery. ¹⁵ Dogs were used in the present study because the conversion of angiotensin I to angiotensin II in the vascular tissue of humans, monkeys, hamsters, and dogs is partly dependent on chymase, but not that in rats. ⁴¹ Consequently, there are several reports in which canine models were used to show attenuating effects of CI on adhesion formation or fibrosis after surgery. ^16,42,43 Previously we demonstrated that the same CI suppressed the scarring in a canine conjunctival flap model. ¹⁶ This time, scleral trephining with peripheral iridectomy was chosen as a simple filtration surgery in canine eyes. The dose of the chymase inhibitor used in the present study was chosen based on the findings of the previous experiment. ¹⁶ The effects of this DDS on the glaucoma surgery model were evaluated according to the previous reports. ^16,17,40  

In a previous study, the percentage of subcutaneously implanted GH remaining was approximately 30% at 14 days. ⁴⁴ Therefore, GH probably remains in the tissue for at least a few weeks. In addition, the first experiment of the present study revealed that the ratio of conjunctival area to scleral area was lower in the eyes treated with GH alone than that in the control eyes. Taken together, these results suggest that the probable mechanism of the IOP reduction obtained by implantation of GH alone for 4 weeks would be as follows: a space, caused by hydrogel under the conjunctiva, might prevent adhesion between the conjunctiva and sclera, resulting in maintenance of the filtration bleb and IOP reduction. 

In the second experiment of the present study, the bleb score began to decrease approximately 4 weeks postoperatively in the control eyes treated with GH alone, probably leading to the IOP increase after 4 weeks. The fact that most of the implanted GH degrades within a few weeks postoperatively, as we noted earlier, could explain the time-course change of the bleb formation in the control eyes. In contrast, the bleb formation remained for a longer period and the adhesion at 12 weeks was less in the eyes treated with GH containing a CI than that in the eyes treated with GH alone. In addition, it was also found that PCNA-positive cells and fibroblasts were decreased in the eyes treated with GH containing a CI, indicating a reduction in the numbers of proliferating cells including fibroblasts. The treatment also reduced the number of mast cells, some of which were involved in the release of chymase. Actually, the number of chymase-positive cells was also decreased by the treatment. In addition, the density of angiotensin II–positive cells was reasonably reduced by treatment with GH containing a CI as well as by treatment with MMC. Since it was confirmed that the release of this CI from the sheet of GH lasted for at least a few weeks in a similar way as for bFGF (Takai S, et al., unpublished data, 2006), it seems likely that the CI continued to act in the lesion where inflammation occurred until a few weeks after the surgery. It has also been reported that the CI functions irreversibly, which means that the inhibitor, once bound to the chymase, continues to inhibit it for several weeks. ³⁹ We also showed that CI, gradually released from GH, prevented the development of scarring by suppressing the upregulation of TGF-β levels after glaucoma surgery, as has been observed in other reports. ^45,46  

Although the results obtained in previous studies with other biomaterials are promising, ^{18 –23} the DDS we used in the present study provides some advantages. GH has been clinically used in medical applications and proven to be biocompatible. ²⁸ It immobilized protein through the physicochemical interaction between the gelatin and protein. The protein could be released from the hydrogel over a period of >5 days, accompanied with the degradation of carrier GH. This period of release is relatively long when compared with, for example, that of sodium hyaluronate gel, which releases steroid over 2 days. ²² Furthermore, using a CI, instead of antimetabolites, MMC, or 5-fluorouracil, would probably be less toxic to the ocular tissue. As one of the evidences for less toxicity of a CI, we found a significant decrease of vessels in the eyes treated with MMC, but not in the eyes treated with the GH containing a CI. 

Several limitations of the present study bear mention. First, the effect of the CI might be different in human eyes, since there are interspecies differences in the roles of chymase, as mentioned earlier. The effect of the DDS applied in this study, as well as its level of toxicity, should be verified in primates in the future. In addition, the effects of other biodegradable materials (such as Seprafilm adhesion barrier), and other antiscarring chemicals, such as antibodies against TGF-β, ⁴⁷ should be compared with the GH and the chymase inhibitor in a future study. Further, the effects of other kinds of CI should be compared with the effects of the CI used here. Searching for the best biosafety combination for antiscarring might be a next step in the progress of glaucoma surgery. 

In summary, the implantation of GH maintained the IOP reduction and bleb formation for a longer duration compared with that in the control animals. The GH containing a CI prolonged the effect through the gradual release of the CI, which suppressed the cell-proliferative effect of chymase. This DDS might be useful for maintaining the filtering bleb for a longer duration after glaucoma surgery. 

Footnotes

 Disclosure: S. Kojima, None; T. Sugiyama, None; S. Takai, None; D. Jin, None; M. Shibata, None; H. Oku, None; Y. Tabata, None; T. Ikeda, None

The authors thank Takao Okuno from Osaka Medical College for technical support and KN International, Inc. (Chicago, IL) for English language editing.