Topical losartan inhibition of myofibroblast generation in rabbit corneas with acute incisions

Abstract

PURPOSE:

To study whether deep central corneal incisions close during topical losartan treatment and the effect of topical losartan on myofibroblast generation after incisions in rabbit corneas.

METHODS:

Rabbits (12) had a 0.35mm deep radial incision from the center of the cornea into the limbus in one eye that was approximated with a single 10-0 nylon suture 1 mm inside the limbus. The incision was treated with 50 μl of topical 0.8 mg/ml losartan or 50 μl of balanced salt solution (BSS) vehicle six times per day for one month. Standardized slit lamp photos of the central incisions were analyzed for opacity with ImageJ prior to euthanasia. Triplex IHC was performed on cryofixed corneas for myofibroblast marker alpha-smooth muscle actin (α-SMA), mesenchymal cell marker vimentin, and basement membrane marker laminin alpha-5. Stromal α-SMA-positive myofibroblasts surrounding the incisions were quantitated with ImageJ.

RESULTS:

Topical losartan compared to vehicle did not affect closure of the radial incisions or the opacity that developed surrounding the incisions at one month after injury. Topical losartan compared to vehicle did significantly decrease the average density of stromal myofibroblasts surrounding the incisions.

CONCLUSIONS:

Topical losartan, a known inhibitor of TGF beta signaling, did not affect closure of deep corneal incisions. Losartan decreased myofibroblast generation surrounding nearly full-thickness radial corneal incisions compared to vehicle. The opacity at the incisions was not significantly affected by losartan—likely because corneal fibroblasts that develop in the stroma adjacent to the incisions was not changed by the losartan compared to the vehicle.

Introduction

Several recently published animal studies in rabbits^1-3 and case reports in humans^4,5 have supported a role for topical, but not oral, losartan in the prophylactic prevention and treatment of scarring corneal fibrosis caused by myofibroblasts after many different injuries to the cornea, including infections, chemical burns, trauma, and surgical complications. In each of these disorders, topical losartan was thought to inhibit transforming growth factor (TGF) beta-1 and TGF beta-2 signal transduction molecule erk⁶ and thereby trigger apoptosis of myofibroblasts, and possibly their precursor cells derived from corneal fibroblasts and fibrocytes.^5,7-9 In one of these case reports,⁵ vision compromising stromal fibrosis that occurred many years following radial keratotomy was highly responsive to topical 0.8 mg/ml losartan six times per day that triggered a decrease in stromal fibrosis and an improvement in vision.

However, there is concern about treating recent corneal lacerations or incisions with topical losartan because myofibroblasts have contractile functions that may facilitate closure of these injuries.^7-10 The current study in rabbits was performed to investigate the effects of topical losartan vs. vehicle treatment after a radial, nearly full-thickness incisions from the center of the cornea and into the limbus on normal closure of the incisions and the generation of myofibroblasts at the incision.

MATERIALS AND METHODS

Animals and Surgery

All animal procedures were approved in advance by the Institutional Animal Care and Use Committee at the Cleveland Clinic Foundation and the Animal Care and Use Review Office (ACURO), and animals were treated in accordance with the tenets of the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. Twelve female 12- to 15-week-old New Zealand white rabbits weighing 2 to 3 kg each were included in the study. All rabbits received 60 ml of children’s liquid acetaminophen (Johnson & Johnson, Ft. Washington, PA, USA) per liter of drinking water starting 24 hours prior to surgery and continuing 5 days after the incision. If needed because of lack of eating or drinking, or other signs of pain, the rabbit was also administered a supplemental anesthetic injection of Buprenorphine 0.05 mg/kg SQ twice daily.

The rabbit was placed under general anesthesia with 30 mg/kg ketamine hydrochloride and xylazine 5 mg/kg by intramuscular (IM) injection and topical anesthesia with two drops of 1% proparacaine (Alcon, Ft. Worth, TX) administered to one eye of each rabbit randomly selected to have the incision. A single incision extending from the anatomical center of the cornea inferiorly into the limbus was performed with a Micro Feather 0.35 mm guarded ophthalmic blade (Cat. # 7335G, Feather, Oasis, Glendora, CA), as illustrated in Figure 1. The normal central corneal thickness in these rabbits measured approximately 400 μm with ultrasonic pachymetry. This incision was approximated with a single 10-0 nylon suture placed 1 mm inside the limbus with the knot rotated into the stroma (Fig. 1). The contralateral cornea was included as an unwounded, untreated control since no contralateral effects of injury have been noted in prior studies.^1-3

Fig. 1.

Schematic diagram of incision design and analysis strategy. The 0.35 mm guarded blade was inserted in the central rabbit cornea and the incision extended vertically into the limbus. A single interrupted 10-0 nylon suture was placed 1 mm from the inferior limbus and the knot rotated into the stroma. Standardized slit lamp photos were taken of the cornea, with the light reflex as shown, and a 1 mm wide and 2 mm long zone one mm from the center of the cornea was analyzed with ImageJ in each cornea one month after the incision and treatment. After marking the incision with a gentian violet marker, the corneal rim was removed, cryofixed in OCT solution, and frozen. Each block was bisected at the center of the analysis area and transverse sections were cut from each side of the block for immunohistochemical analysis.

Medications

Beginning immediately after the incision and suture were performed, six eyes were treated with 50 μl of balanced salt solution (BSS, 0.64% sodium chloride, 0.075% potassium chloride, 0.048% calcium chloride dihydrate, 0.03% magnesium chloride hexahydrate, 0.39% sodium acetate trihydrate, 0.17% sodium citrate dihydrate, pH 7.0) six times a day (approximately 8 am, 10 am, 12 noon, 2 pm, 4 pm and 6 pm) and six eyes were treated with 50 μl of 0.8 mg/ml losartan (Merck & Co., Inc., Kenilworth, NJ, USA) in BSS six times a day. Treatment with the vehicle or losartan was continued for one month after the incision surgery. Eyes that had an incision were also treated with one drop of topical ciprofloxacin (Ciloxan, Alcon Laboratories, Inc. Fort Worth, TX) three times a day and at least five minutes from the other topical medication until the epithelium closed (the epithelium in all eyes closed by 5 days after surgery). No corticosteroids were administered in either group.

Standardized slit-lamp photographs and ImageJ measurement of corneal opacity

At 1 month after the incision and topical treatment, each rabbit was placed under ketamine-xylazine general anesthesia and the pupils were dilated with two drops of 1% tropicamide (Akorn Co., Lake Forest, IL, USA) for 30 minutes to facilitate corneal imaging. The study eye in each rabbit had slit-lamp photographs with standardized illumination level on 40% maximal intensity, the angle of illumination approximately 60 degrees from the normal line perpendicular to the central corneal surface, and angle of photography perpendicular to the central corneal surface at 20X magnification so that the light reflex is displaced laterally from the center of the cornea by approximately 3 mm (Fig. 1) with a Topcon (Oakland, NJ, USA) SL-D7 slit-lamp photography system. For each study cornea, the incision image was converted to grayscale with Adobe (San Jose, CA, USA) Photoshop 23.5.2 and the mean opacity in pixels in a 1 mm wide by 2 mm length rectangular area including the central corneal portion of the incision (Fig. 1) was determined by using ImageJ 1.53a analysis software (National Institutes of Health, Bethesda, MD, USA).

Corneal cryo-fixation and sectioning

After imaging, each rabbit was euthanized while under ketamine-xylazine general anesthesia with 100 mg/kg Beuthanasia (Shering-Plough, Kenilworth, NJ, USA) by intravenous injection followed by rapid exsanguination. The epithelial surface of the incision was marked with a gentian violet pen to allow the specimen to be oriented for sectioning. The corneo-scleral rim was removed with sharp Westcott scissors (Fairfield, CT, USA) and 0.12 forceps (Storz, St Louis, MO, USA). The corneo-scleral rim was centered in a 24-mm X 24-mm X 5-mm mold (Fisher Scientific, Pittsburgh, PA, USA) that was filled with optimal cutting temperature (OCT) compound (Sakura Finetek, Torrance, CA, USA) and quick frozen on dry ice. Each block was stored at −80°C until sectioning.

Blocks were bisected at the center of the analysis box with the cut transecting the original incision (Fig. 1) using the suture and gentian violet surface marks for orientation. Ten μm-thick sections were cut from the cornea perpendicular to the incision with a cryostat (HM 505M; Micron GmbH, Walldorf, Germany). Three sections from each cornea were placed on each 25-mm X 75-mm X 1-mm Superfrost Plus microscope slide (Fisher Scientific). Slides with sections were maintained at −20°C prior to immunohistochemistry (IHC).

Immunohistochemistry (IHC) and quantitation of stromal alpha-smooth muscle actin using ImageJ

Triplex IHC for α-SMA (myofibroblast marker), vimentin (mesenchymal cell marker), laminin alpha-5 (basement membrane marker), and 4',6-diamidino-2-phenylindole (DAPI) was performed using previously described methods⁹ with primary antibodies confirmed by Western blotting and IHC to recognize rabbit antigens or isotypic non-specific control antibodies (ThermoFisher Scientific, Waltham, MA, USA) and secondary fluorescent tagged antibodies (Table).^1-3,9 At the concentration of anti-vimentin antibody used in this study, all corneal fibroblasts and myofibroblasts are expected to be vimentin-positive, while most keratocytes are vimentin-negative at this anti-vimentin antibody concentration due to their relatively low level of expression.⁹ Peter Yurchenco, MD, PhD, graciously provided the laminin alpha-5 antibody. All corneas were evaluated with the triplex immunohistochemistry at least four times. Images were obtained at 100X total magnification on a Leica DM6B upright microscope equipped with an automated stage and Leica 7000 T camera using the LASX software (Leica Microsystems, GmbH, Wetzlar, Germany).

Table.

Antibodies used in immunohistochemistry

Antigen	Source	Species source	Ab Isotype	Catalog number	Dilution
PRIMARY ANTIBODIES
α-smooth muscle actin	DAKO	Mouse	IgG2A	M0851	1:400
Vimentin	Abcam	Rat	IgG2A	MAB2105	1:200
Laminin alpha-5	Peter Yurchenco, MD, PhD	Chicken	IgY	-	1:400
SECONDARY ANTIBODIES
Alexa Fluor 568 anti-mouse	Thermo Fisher Scientific	donkey	IgG	A10037	1:200
Alexa Fluor 488 anti-rat	Thermo Fisher Scientific	donkey	IgG	A21208	1:200
Alexa Fluor 647 anti-Chicken	Thermo Fisher Scientific	donkey	IgG	A32787	1:200

All IHC images were converted to 900-pixel width X 672-pixel height 300 DPI images with Photoshop 22.1.1 (Adobe, San Jose, CA). The mean pixels of stromal α-SMA were determined in a 900-pixel wide X 235-pixel high rectangle with ImageJ in three sections for each cornea using the image panels showing only the α-SMA antigen. The mean from three corneal sections was used as the value for stromal α-SMA for each individual cornea. Epithelial thickness overlying the incision was determined from DAPI IHC images of each cornea by determining the mean of five adjacent measurements from the epithelial surface to the epithelial basement membrane (EBM) with a reticle.

Statistics

Comparisons between groups were performed using the Kruskal Wallis Test followed by post-hoc Dunn’s-Bonferroni test. P < 0.05 was considered statistically significant.

Results

All incisions in both the vehicle- and losartan-treated groups closed without gaping or persistent epithelial defects by three to five days after the incision. There were no corneal infections during this study.

Figure 2A shows the standardized slit lamp photo converted to grayscale in the vehicle and losartan groups within the 1 mm X 2 mm analysis area at the incision of each cornea (see Fig. 1). Figure 2B shows the mean pixels of opacity within this rectangular analysis area of the incision for each cornea, as shown in Fig. 1. The difference between vehicle group (mean±SEM, 76±2 pixels) and the losartan group (80±2) was not statistically significant (p = 0.39).

Fig. 2.

Standardized slit lamp opacity analysis. A. A 1 mm X 2 mm area of the incision in each cornea in grayscale was analyzed with ImageJ. Arrows indicate opacity that was diffusely present surrounding the incisions that bisected the shown area from top to bottom. B. The mean ImageJ opacity within the analysis area for each cornea is shown, along with the mean±SEM for each group. The difference between the vehicle-treated group and the losartan-treated group was not statistically significant (p=0.39).

Figure 3A shows representative triplex immunohistochemistry for α-SMA, vimentin, and laminin alpha-5 of an unwounded, untreated control cornea and of the incision analysis area of each vehicle-treated cornea. Figure 3B shows the triplex immunohistochemistry with isotypic control antibodies and of the incision analysis area of each losartan-treated cornea. Figure 3C shows a graph of the total α-SMA pixel units in the 2.0W X 0.5H analysis area for each cornea shown in Fig. 3A and Fig. 3B. The difference in α-SMA between the vehicle-treated (475±253) and the losartan-treated corneas (36±22) was statistically significant (p=0.03). The difference in α-SMA between the vehicle-treated corneas (0±0) and the control, unwounded corneas was statistically significant (p=0.0003). The difference in α-SMA between the losartan-treated corneas and the control, unwounded corneas was not statistically significant (p=0.14).

Fig. 3.

Triplex immunohistochemistry analysis for α-SMA, vimentin (VIM) and laminin alpha-5 (LAMA5). A. A representative section from an unwounded and untreated control cornea and each vehicle-treated cornea is shown. A composite image of all three antigens with DAPI is provided in the top panel, with individual panels for α-SMA (arrowheads indicate α-SMA-positive myofibroblasts), VIM (arrowheads indicate VIM-positive mesenchymal cells) and LAMA5 (arrowheads indicate EBM associated LAMA5). ImageJ 2.00W X 0.5 H analysis rectangles with the upper length approximately at the posterior edge of the epithelium is shown for each cornea. e is epithelium in each panel. * indicates epithelial artifacts of sectioning noted commonly in corneas with incisions at one month after surgery that occurs in the investigators’ experience while the regenerated epithelium is immature. B. Representative sections from each losartan-treated cornea with labelling as in A. A vehicle-treated incision cornea is also shown that had triplex IHC with isotypic control antibodies. Blue is DAPI staining of nuclei in each panel. C. A graph of the mean α-SMA pixel units in each cornea in the vehicle-treated, losartan-treated, and unwounded, untreated control corneas. One vehicle-vehicle treated cornea (vehicle 2 in Fig. 3A) was noted to have a very high level of α-SMA and the value was noted superior to the double lines that indicated it was beyond the scale of the y-axis. The difference between the vehicle-treated group and the losartan-treated group was statistically significant (P=0.03), as was the difference between the vehicle-treated group and the unwounded, untreated group (p=0.0003). The difference between the losartan-treated group and the unwounded, untreated control group was not significant (p=0.14).

Fig. 4. shows the mean epithelial thicknesses overlying the incision in the vehicle and losartan treated groups and in the same region of the unwounded, untreated controls. There was a trend towards significance between the vehicle-treated group (mean 113±14 μm) and the losartan group (mean 73±13) that didn’t achieve statistical significance (p=0.09). The difference between the vehicle-treated incision group and the unwounded, untreated control group (46±2) was statistically sign (p=0.007). The difference between the losartan-treated group and the unwounded, untreated control group was not statistically significant (p=0.30).

Fig. 4.

Epithelial thickness measurements in the vehicle-treated, losartan-treated, and unwounded, untreated control corneal sections in immunohistochemistry analyses such as those shown in Fig. 3. Means ± SEMs are shown for each group. The difference between the vehicle-treated group and the unwounded, untreated control group was highly significant (p=0.007). The trend towards a difference between the vehicle-treated group and the losartan treated group did not reach statistical significance (P=0.09). The difference between the losartan-treated group and the unwounded, untreated control group was not significant (p=0.30).

Discussion

Topical losartan has been shown in several recent studies in rabbits^1-3 and in case reports in humans^4,5 to decrease myofibroblast-mediated scarring fibrosis after corneal injury. This treatment, however, has not been studied in corneas with acute incisions or lacerations where closure of the corneal surface could be hindered, leading to unpredictable refractive changes, paradoxical scarring due to lack of epithelial basement membrane closure and regeneration, or other unforeseen complications unique to lacerations or incisions.

The most important observation in this study is that topical losartan treatment did not interfere with closure of a nearly full-thickness corneal incision that extended from the central cornea through the limbus approximated by a single suture one mm central to the limbus. This, however, needs to be verified in human trials of corneal incisions and lacerations. Although the tightness of the incision closure was not tested in this study, it was reassuring that acute incisions did not gape or otherwise remain unclosed during topical losartan for one month after injury. Therefore, if these observations hold up in human corneas, it appears that topical losartan can be used to decrease scarring opacity in corneas with acute incisions or lacerations.

Since myofibroblasts were inhibited by topical losartan, this suggests that corneal fibroblasts generated around the stromal incision are competent to close simple incisions and lacerations, especially since no myofibroblasts were detected in some losartan-treated incisions that had four our more runs of immunohistochemistry. These corneal fibroblasts are more opaque than keratocytes and contribute to the opacity around the incision.¹¹ However, corneal fibroblasts are transient and would likely either disappear by apoptosis or transition back to keratocytes over several more months of follow-up after the injury.^12,13 Thus, the opacity itself surrounding corneal incisions and lacerations is likely to diminish over time after injury, especially if topical losartan was used to diminish myofibroblast generation and to trigger apoptosis of myofibroblasts that are produced in response to the injury. This treatment is not needed in most cases to decrease the opacity of incisions or lacerations that spare the cornea overlying the entrance pupil. However, in some of these cases, opacity will increase if myofibroblasts are generated in large numbers and extend beyond the incision. A case such as this was recently reported in which stromal fibrosis affecting vision developed years after radial keratotomy (Fig. 5).⁵ In that case, topical losartan 0.8 mg/ml applied six times per day was highly effective in decreasing opacity due to fibrosis surrounding the incision and improved vision within 15 days of beginning treatment. That rapid of response is not likely to occur with all etiologies of corneal fibrosis, depending on the source and extent of the injury, and that likely includes corneal incisions and lacerations. Many cases of corneal fibrosis reported to date in humans take up to six months to respond significantly to topical losartan. Good overviews of the corneal wound healing response were provided in recent reviews.^14-16

Fig. 5.

Losartan treatment of visually significant stromal fibrosis in a radial keratotomy incision years after the original surgery. A. More than a decade after radial keratotomy, decreasing uncorrected and best-corrected vision was noted associated with developing fibrosis (arrows) surrounding an inferior RK incision. B. After topical 0.8 mg/ml losartan in BSS was applied six times per day for 15 days, there was a marked decrease in scarring fibrosis associated with the incision (arrows) and an improvement in uncorrected and best spectacle-corrected vision in the eye. The arrowhead points to an artifact related to the light reflex during photography. The patient was lost to follow-up after the 15-day examination. This case was provided by the treating physicians Rodrigo Carlos de Oliveira, MD and Luiz Antônio Vieira, MD, Sao Paulo, Brazil. 15X mag. Reprinted with permission from the J Ocular Pharm and Ther. 2023;39:191-206.

As with previous studies of topical losartan in rabbits^1-3 and human cases,^4,5 no complications were noted in the present incisional study with the treatment. However, controlled clinical trials with larger numbers of human cases are needed to confirm the efficacy and safety of topical losartan treatment for scarring corneal fibrosis associated with corneal incisions and lacerations.

This study of topical losartan treatment was continued only for one month due to the expense of maintaining and treating rabbits for extended periods of time. Even with only one month of treatment, however, myofibroblasts surrounding the incision were significantly decreased by losartan compared to vehicle treatment. Further decreases would likely be noted with ongoing topical losartan treatment. No difference in vimentin levels (Fig. 3B and Fig. 3C) were noted surrounding the incision in losartan or vehicle treated corneas. The vimentin-positive cells in these corneas that were α-SMA-negative likely included corneal fibroblasts and bone marrow-derived fibrocytes.⁷ Although these are the major precursors to myofibroblasts that have been identified in the stroma of injured corneas, the topical losartan effectively inhibited the development of these myofibroblasts.⁷ Similarly, no differences in deposition of laminin alpha-5 was noted between the losartan- and vehicle-treated corneas (Fig. 3A and Fig. 3B). Laminin-alpha-5 is one of the first EBM components laid down, primarily by epithelial cells, at the junction between the epithelium and the stroma.^17-19 Other components, such as perlecan, nidogens, and collagen type IV are added to the nascent EBM to form the mature EBM by both corneal epithelial cells and corneal fibroblasts,^20,21 even around epithelial plugs that extend into the stroma, as can be noted in Fig. 3A and Fig. 3B. Once the mature EBM is regenerated, levels of TGF beta-1 and TGF beta-2 entering the stroma from the tears and epithelium are diminished to the point that persisting myofibroblasts undergo apoptosis,^7,21-23 whether or not the cornea was treated with topical losartan. The mature EBM can regenerate in a week or less in some corneas but can take months, or may even never fully-regenerate, in some corneas with incisions or lacerations. This is likely why myofibroblasts can be detected in some corneas even years after corneal incisions or lacerations. Unfortunately, there is currently no way to assess when the EBM reaches this final stage of development in the cornea in situ and topical losartan is no longer needed if it’s being used to decrease stromal myofibroblasts and fibrosis. Thus, observations at the slit lamp are necessary after successful treatment when the losartan is ended. If scarring fibrosis begins to recur, then longer topical losartan treatment can be resumed. Clinical trials in humans are needed to determine whether prolonged topical losartan therapy is safe and effective.

As can be noted in Fig. 3A and Fig. 3B, the epithelium tended, on average, to show greater hyperplasia and hypertrophy overlying corneal incisions in vehicle-treated corneas compared to losartan treated corneas in this study. As suggested by prior in vitro studies in corneal epithelium and conjunctival goblet cell epithelium,^24,25 this could indicate that TGF beta increases the proliferation and inhibits the terminal differentiation of some epithelial cells. Although the difference between losartan-treated and vehicle-treated corneas did not reach statistical significance (Fig. 4), likely due to the relatively small numbers of rabbits included in this study, this observation is worthy of additional investigation in situ.

An interesting corollary to this study, is topical losartan could possibly be useful to diminish visually significant astigmatism that develops after injuries or surgeries to the cornea—for example, after some cataract surgery cases with or without sutures.^26,27 This could be especially useful for persistent astigmatism after the sutures were removed, for example suturing at the superior limbus creating with-the-rule astigmatism. Controlled trials, however, will be needed to study and confirm this potential topical losartan application.

In conclusion, topical losartan therapy may provide effective treatment to prevent the development of myofibroblasts and decrease scarring myofibroblasts once they develop in corneal incisions and lacerations where vision is compromised by the development of these fibroblastic cells. Further clinical trials, however, are needed to confirm the efficacy and safety of this treatment in humans.