Molecular Vision 2024; 30:448-464
<http://www.molvis.org/molvis/v30/448>
Received 01 August 2024 | Accepted 28 December 2024 | Published 30 December 2024
Molecular Vision 2024; 30:448-464
<http://www.molvis.org/molvis/v30/448>
Received 01 August 2024 |
Accepted 28 December 2024 |
Published 30 December 2024
Suneel Gupta,1,2 Eric Zhang,1,3 Sampann Sinha,1,2 Lynn M. Martin,1,2 Thomas S. Varghese,1,3 Nathan G. Forck,1,2 Prashant R. Sinha,1,2 Aaron C. Ericsson,4 Nathan P. Hesemann,1,3 Rajiv R. Mohan1,2,3
1Harry S. Truman Memorial Veterans’ Hospital, Columbia, MO; 2Departments of Veterinary Medicine & Surgery and Biomedical Sciences, College of Veterinary Medicine, University of Missouri, Columbia, MO; 3Mason Eye Institute, School of Medicine, University of Missouri, Columbia, MO; 4Departments of Veterinary Pathobiology, College of Veterinary Medicine, University of Missouri, Columbia, MO
Correspondence to: Rajiv R. Mohan, Department of Ophthalmology and Molecular Medicine, University of Missouri, 1600 E. Rollins St., Columbia, MO 65211; email: mohanr@health.missouri.edu
Purpose: During ocular trauma, excessive proliferation and transdifferentiation of corneal stromal fibroblasts cause haze/fibrosis in the cornea. Transforming growth factor β (TGFβ) plays a key role in corneal fibrosis through the Smad signaling pathway. The aberrant activity of TGFβ signaling during ocular trauma (viz. mechanical, infectious, chemical, or surgically altered TGFβ/Smad signaling) leads to regulating the predominant expression of myogenic proteins and the extracellular matrix (ECM). We sought to investigate the functional role of Smad3 in corneal wound repair and stromal ECM assembly using Smad3+/+ wild-type and Smad3−/− deficient mice.
Methods: Corneal injury was introduced with the topical application of an alkali-soaked 2-mm filter disc on the central cornea in the Smad3+/+ (C57BL/6J) and Smad3−/− (129-Smad3tm1Par/J) mouse strains. Slit-lamp and stereo microscopy were used for clinical assessment and corneal haze grading in live animals. Hematoxylin and eosin and Masson’s trichrome staining were used to study comparative morphology and collagen level alterations between the groups. Real-time qRT-PCR, western blot, and immunohistochemistry were used to measure changes in profibrotic genes at the mRNA and protein levels.
Results: Slit-lamp clinical exams and stereo microscopy detected notably less opaque cornea in the eyes of Smad3−/− compared with Smad3+/+ mice at 3 weeks (p<0.01) in live animals. Corneal tissue sections of Smad3−/− mice showed significantly fewer α-smooth muscle actin-positive cells compared with those of the Smad3+/+ animals (p<0.05). The corneas of the Smad3−/− mice showed significantly lower mRNA levels of pro-fibrotic genes, α-smooth muscle actin, fibronectin, and collagen I (p<0.05, p<0.01, and p<0.001). In addition, the matrix metalloproteinase and tissue inhibitors of metalloproteinase levels were significantly increased (p<0.001) in the corneal tissue during alkali injury in both Smad3+/+ wild-type and Smad3−/− deficient mice.
Conclusions: The significant changes in profibrotic genes and stromal ECM proteins revealed a direct role of Smad3 in stromal ECM proteins and TGFβ/Smad-driven wound healing. Smad3 appears to be an attractive molecular target for limiting abnormal stroma wound healing to treat corneal fibrosis in vivo.
The cornea is the clear, dome-shaped outermost layer of the eye. The avascular and transparent nature of the corneal tissue makes it a good model for understanding wound healing events [1-5]. Ocular trauma resulting from various corneal insults (such as mechanical, surgical, and chemical injuries) is a serious clinical problem worldwide [6-8]. Any of these traumatic impairments to the ocular tissue can compromise the normal transparency and curvature of the cornea [3-8]. Transdifferentiation of corneal stromal fibroblasts to myofibroblasts is an important event during trauma and plays a key role in extracellular matrix (ECM) protein secretion [9-11]. Exert contraction across the ECM results in tissue architecture alteration and impaired function, leading to corneal fibrosis pathology and corneal scarring [9-14]. Corneal scarring after trauma/infection impairs vision in 4% of Americans and accounts for 7% of the world’s blindness. Over 23 million people globally, including many children, have corneal fibrosis in one eye and 4.9 million in both eyes [15-19]. Ocular trauma and corneal ulceration cause approximately 1.5–2.0 million new cases of monocular corneal blindness each year. In addition, approximately 1.4 million children have blindness worldwide, associated with lower socioeconomic class and suffering from socioeconomic deprivation [16-20]. The transdifferentiation of fibroblasts into contractile and secretory myofibroblasts and their persistence in the corneal tissue are the common sequelae of corneal fibrosis [11-14]. The transdifferentiation of corneal stromal fibroblasts into myofibroblasts and deposition of ECM proteins in the stroma lead to the formation of a collagen-based scar [12-14,21-23].
During wound healing, transforming growth factor β (TGFβ) signals, through canonical and/or noncanonical signaling pathways [24-28], are known to promote conversion of corneal stromal fibroblasts to myofibroblasts, enhance ECM protein synthesis, and promote translocation of suppressors mothers against decapentaplegic (Smads) [29-32]. In addition, they can alter the activity of proteases, including matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs) [33-39]. MMPs, proteolytic enzymes, are essential in wound healing and have roles in fibrosis, ECM remodeling, and active wound healing [33-39]. The activities of MMPs are tightly regulated by gene expression, activation of zymogens (cysteine switch mechanism), and inactivation of enzymes with inhibitors such as endogenous TIMPs [33-39]. A proper balance of MMP and TIMP activities is critical for ECM homeostasis [35-39]. Smad proteins are unique transducers of the TGFβ family receptor serine-threonine kinases and phosphorylation of regulatory Smads (R-Smads; Smad2 and Smad3), with common mediator Smad (co-Smad; Smad4) and translocation of the Smad complex to the nucleus playing a predominant role in the activation of TGFβ-dependent signaling and their target molecules during fibrosis [25-31]. R-Smad expression and activity/function are tightly and precisely controlled by posttranslational modifications, including (de)phosphorylation, (de)ubiquitylation, sumoylation, deacetylation, methylation, and ADP-ribosylation [32]. Our group demonstrated that targeted Smad7 (inhibitory Smad) gene transfer in stroma via adeno-associated vectors reduced corneal fibrosis in an alkali-injured rabbit cornea in vivo [7]. Based on the Smad7 overexpression gene therapy findings, we hypothesized that Smads are the key regulators in corneal wound healing and directly influence the TGFβ-dependent pathway to regulate the stromal ECM proteins during wound healing events. In the present study, we used Smad3+/+ wild-type (C57BL/6J) and Smad3−/− deficient mice (129-Smad3tm1Par/J) to establish the direct role of Smad3 in corneal wound healing and stromal ECM proteins. The objective of this study was to uncover the direct role of Smad3 in corneal wound healing and stromal ECM protein regulation using the loss-of-function approach with a Smad3−/− deficient mouse model.
Standard research-grade chemicals and reagents were used to perform this study. Sodium hydroxide (NaOH), paraformaldehyde solution, Triton X100, and bovine serum albumin were purchased from Sigma‒Aldrich, St. Louis, MO. Two-methyl butane and antifade mounting 4′,6-diamidino-2-phenylindole dihydrochloride medium (cat # H1200, Vector Laboratories) were obtained from Thermo Fisher, Grand Island, NY. Topical artificial tears were obtained from Rugby Laboratories, Livonia, MI. Hematoxylin and eosin (H&E) solutions were procured from StatLab Medical Products, McKinney, TX. Sterile Weck-Cel ophthalmic spears were purchased from Beaver-Visitec International Inc., Waltham, MA. Surgical forceps and sharp Westcott scissors were purchased from World Precision Instruments Inc., Sarasota, FL. Ketamine hydrochloride (JHP Pharmaceuticals, LLC, Rochester, MI), xylazine hydrochloride (XylaMed, Bimeda Inc., Le Sueur, MN), topical 0.5% proparacaine hydrochloride (Alcon, Ft. Worth, TX), and pentobarbital (Diamondback Drugs, Scottsdale, AZ) were obtained from Harry S. Truman VA Medical Center, Columbia, MO.
All mice experimental procedures were approved by the Institutional Animal Care and Use committees of the University of Missouri (MU) and the Harry S. Truman Memorial Veterans’ Hospital. The study was conducted in accordance with the ARVO recommendations for animal experimentation, and data were presented in accordance with the ARRIVE guidelines.
In this study, Smad3+/+ (C57BL/6J) and Smad3−/− (129-Smad3tm1Par/J) deficient mice 8–12 weeks in age in groups of both male and female sex were used, despite our previous work showing that sex does not play an important role during corneal wound healing [40]. Smad3−/− (129-Smad3tm1Par/J) mice were generated from a breeding pair procured from Dr. Aaron C. Ericsson, Department of Veterinary Pathobiology, MU (Columbia, MO) [41]. Wild-type C57BL/6J (Smad3+/+) mice were purchased from the Jackson Laboratory, Bar Harbor, ME. Both, Smad3+/+ and Smad3−/− mice were used to elucidate the essential and functional role of Smad3 in corneal wound healing and stromal remodeling in vivo. Smad3-deficient mice (Smad3−/−; n=12) and wild-type C57BL/6J mice (Smad3+/+; n=12) (genetic background C57BL/J6) were bred by homozygote breeding. Smad3−/− deficiency was confirmed by a polymerase chain reaction using specific primers for Smad3−/− mice (as listed in Table 1) and horizontal 2.0% agarose gel [41].
Mice were anesthetized with an intraperitoneal injection of a ketamine hydrochloride (100 mg/kg) and xylazine hydrochloride (10 mg/kg) cocktail. A single drop of topical anesthetic, proparacaine hydrochloride (0.5%), was instilled into the eye before alkali wounding to reduce pain and discomfort to the animal before the scar induction, clinical assessment, and clinical biomicroscopy. Corneal manipulations and scar induction procedures were performed under a surgical dissecting microscope as reported previously [42-44]. Mice were thermally supported throughout the procedure and during the anesthetic recovery period. For scar induction, 20 µl of NaOH solution (0.5 M) was applied with a 2-mm-diameter filter paper disc (P8, Filter Paper, Fisher Brand, Fisher Scientific, Pittsburgh, PA) and placed onto the central cornea. The alkali-soaked filter disc contact with the corneal epithelium was maintained for 30 s as reported previously [42-44]. Thereafter, the eyes were copiously washed with a balanced salt solution to remove all NaOH traces. All in vivo experimental manipulations and imaging procedures were performed by keeping the experimental conditions constant to avoid variation and increase reproducibility. Both eyes were kept lubricated with artificial tears during the entire procedure to prevent corneal desiccation.
After alkali scar induction, all mice underwent clinical eye examinations and corneal imaging at regular intervals (up to 3 weeks) for the assessment of scaring and for biomicroscopy images under general anesthesia. A wide-beam slit-lamp microscope (Leica DM 4000B, Leica Microsystems Inc., Buffalo Grove, IL) equipped with a digital camera (SpotCam RT KE, Diagnostic Instruments Inc., Sterling Heights, MI) was used to record wound progression in mouse corneal tissue [42-44]. At the time of haze evaluation, more than five images per animal at each time point were taken from both eyes to capture the fine corneal details. The scarring was graded according to Fantes haze score in an independent blinded manner as reported previously [6,7,45,46]. Briefly, naïve or completely clear corneas were graded 0; during the slit-lamp biomicroscopic evaluations, the presence of fine traces of haze during the tangential illumination of corneal tissue was graded 0.5; more prominent haze but not interfering with the visibility of fine iris details was graded 1; sufficient haze causing mild obscuration of iris detail was graded 2; sufficient haze causing moderate obscuration of the iris and lens was graded 3; and prominent haze with complete opacification of the corneal tissues was graded 4 in the traumatized injury zone.
The corneal epithelium defect/erosion was assessed using the fluorescein eye test under a cobalt blue light and recorded under a green fluorescent protein filter-equipped fluorescence microscope (Leica DM 4000B, Leica Microsystems Inc.) equipped with a digital camera (SpotCam RT KE, Diagnostic Instruments Inc.) as reported previously [47-49]. In this test, fluorescein uptake occurs in damaged corneal epithelium, and normal epithelium shows no dye under the green fluorescent protein filter/cobalt blue light.
To determine the impact of Smad3 gene deficiency on aqueous humor regulation, intraocular pressure (IOP) measurement was evaluated using a handheld tonometer (iCare TONOLAB tonometer; TV02). All IOP measurements were performed between 9 AM and 11 AM to minimize normal diurnal variations and recorded by one operator to avoid operator-based variations [46-49].
Mice were humanely euthanized with an intraperitoneal injection of pentobarbital (150 mg/kg) while under general anesthesia at the termination point. Eyes were enucleated with surgical forceps and Westcott scissors under a dissecting microscope. Corneas were used for immunofluorescence, mRNA, and protein studies. Eyes were immediately placed into molds containing optimal cutting temperature compound and snap-frozen in a container of 2-methyl butane immersed in liquid nitrogen as reported previously [42-44]. Frozen tissues were maintained at −80 °C until sectioning and further evaluation. Tissues were sectioned at 8-µm thickness with a cryo-microtome, mounted on glass microscopic slides, and preserved at −80 °C for subsequent analysis.
The mouse corneal tissue sections of both naïve and alkali-injury Smad3+/+ and Smad3−/− were subjected to H&E staining to examine the gross morphology, cellular infiltration, and ECM alterations. H&E staining was performed in our laboratory using the protocol reported previously [42-44,47-51]. Additionally, corneal sections were sent to the MU Veterinary Medical Diagnostic Laboratory for H&E histology.
The mouse corneal tissue sections of both naïve and alkali-injury Smad3+/+ and Smad3−/− were subjected to Masson’s trichrome staining to evaluate alterations in collagen, a primary component of the ECM. Masson’s trichrome staining was performed independently within the laboratory and by the MU Veterinary Medical Diagnostic Laboratory using a previously reported protocol [42-44,47-51].
The corneal sections were used for localization and characterization of proteins under a fluorescence microscope as reported previously [42-44]. Briefly, the corneal tissue sections were washed three times with 1× phosphate-buffered saline for 5 min and incubated and blocked with 5.0% bovine serum albumin for 1 h. Tissue sections were then placed in a 1:100 dilution of a specific primary antibody (α-smooth muscle actin [α-SMA]: Cat# PA5–32587; Smad3: Cat# SC101154; and CD11b: Cat#550282) overnight at 4 °C and then washed with 1× phosphate-buffered saline for 5 min three times. Tissue sections were further incubated in secondary antibody Alexa Fluor™ 488 (Cat# A21206), Alexa Fluor™ 594 (Cat# A21135), or Alexa Fluor™ 594 (Cat# A21209) at 1:1000 dilution and mounted with 4′,6-diamidino-2-phenylindole dihydrochloride antifade medium. The stained corneal tissue sections were viewed and photographed with a fluorescence microscope equipped with a digital camera as reported previously [42-44,47-51].
The corneal protein lysates were subjected to characterization of specific proteins. Protein lysates were prepared by homogenizing corneas in Pierce RIPA buffer containing a protease inhibitor cocktail (Thermo Fischer Scientific) and quantified by the Bradford assay as described previously [42-51]. Samples were resolved on 4%–12% sodium dodecyl sulfate (SDS) polyacrylamide gel, transferred on a polyvinylidene fluoride membrane, incubated with primary antibodies (α-SMA: Cat # M0851; TIMP-1: Cat # AF980-SP; TIMP-2: Cat # D18B7; MMP-2: Cat # SC13594; MMP-9: Cat # SC13520; and GADPDH: Cat # 47,724) followed by corresponding horseradish peroxidase-conjugated secondary antibodies (HRP-conjugate; Cat # SC2005; Cat # 31,460), and developed with SuperSignal ELISA Femto maximum sensitivity substrate (Cat # 37,075) reagents. The blots were imaged and quantified using the iBright CL1500 imaging system, iBright analysis software (Invitrogen, Thermo Fischer Scientific), and NIH ImageJ (ImageJ 1.50i).
The expression of pro- and active isoforms of matrix metalloproteinases (MMP) protein was detected using gelatin zymography in the presence of 0.1% gelatin under nonreducing conditions as reported previously [25]. Briefly, a 20-μL sample was prepared by mixing total protein lysates (30 μg), Tris-glycine SDS sample buffer (10 μL), and deionized water. Samples were loaded on the 10% SDS polyacrylamide gel (Life Technologies, Novex, Carlsbad, CA) in the presence of 0.1% gelatin and were used under nonreducing conditions at a constant voltage (150 V; for 120 min) in a cold room. The gel was washed and developed using the zymography reagents. The developed gel was stained with SimplyBlue SafeStain diluted as per the vendor’s instructions. Digital quantification of zymograms was performed using NIH ImageJ software (ImageJ 1.50i).
Statistical analysis was performed using GraphPad Prism 9.2 software (GraphPad Software, La Jolla, CA). A Student t test, Wilcoxon rank sum test, or Mann–Whitney tests were used to determine the statistical significance during data analysis of different experiments for biologic validation. The data are presented as the average mean ± standard error (SEM). In statistical analysis, p≤0.05 was considered significant.
Smad3−/− deficiency in mice was confirmed by a standard polymerase chain reaction using specific primers for Smad3−/− mice as listed in Table 1. The agarose gel (2.0%) showed a band at 364 bp, confirming the Smad3−/− strain (Smad deficient; Figure 1A), whereas the Smad3+/+ strain (wild type; Figure 1B) showed a band at 227 bp.
Age-matched Smad3+/+ (C57BL/6J, wild type) and Smad3−/− (Smad3) deficient mice (both male and female) were used to study the functional role of Smad3 during alkali-induced trauma for corneal scarring in vivo. The corneal scarring was assessed throughout the study period using slit-lamp microscopy observations in live animals and graded on Fantes scale (Figure 2). The subjective clinical slit-lamp and stereo microscopy evaluations exhibited intensified prominent scarring in the corneal tissues of Smad3+/+ mice (Figure 2C,D) compared with Smad3−/− mice (Figure 2G,H) at day 7 and day 21, respectively. The Smad3+/+ mouse corneas showed significantly increased scarring (1.6±0.48 fold; p<0.001) compared with the Smad3−/− mouse corneas on the Fantes haze scale (Figure 2I). The decreased corneal scarring in Smad3−/− deficient mouse corneas indicates the direct role of Smad3 in corneal scarring. The age-matched Smad3+/+ (C57BL/6J mouse strain) and Smad3−/− (129-Smad3tm1Par/J mouse strain) corneal tissue of both male and female mice during our clinical evaluation did not show any signs of opacity or other ocular alterations before alkali injury (naïve controls of both strains), as shown by representative slit-lamp microscopic images (Figure 2A,E). Tonometry analyses of mice did not show any statistically significant differences in the IOP of both strains in naïve and post-alkali injury conditions (Table 2). As expected, significant alteration in the corneal IOP was observed on day 1 after alkali injury due to initial alkali trauma to the corneal tissue. No significant changes were recorded in IOP at later time points from day 3 to day 21. The corneal re-epithelization was not impacted in Smad3+/+ and Smad3−/− mouse corneal tissue, as indicated by fluorescein eye imaging in naïve and injured mouse corneas in a time-dependent epithelial assessment (Figure 3).
The effect of Smad3 gene deficiency was evaluated by using the inflammatory marker CD11b protein expression (Figure 4). The post-injury group of animal corneal tissue sections of strains Smad3+/+ and Smad3−/− showed notably enhanced CD11b protein expression (Figure 4C,D) compared with the naïve group of animal corneal tissue sections of strains Smad3+/+ and Smad3−/− (Figure 4A,B) at day 21. However, Smad3 deficiency does not seem to have any significant impact on inflammatory response.
To test the postulate that Smad3 deficiency decreases the expression of pro-fibrotic markers in the cornea during active wound healing, the mRNA levels of fibronectin (FN), collagen 1, collagen 3, and collagen 4 were quantified in Smad3−/− mouse corneal tissue and compared with those of Smad3+/+ mouse corneal tissue (Figure 5). The results of the mRNA expression study revealed a significant reduction in FN (Figure 5A; p<0.001), collagen 1 (Figure 5B; p<0.01), collagen 3 (Figure 5C; p<0.05), and collagen 4 (Figure 5D; p<0.01) and evidenced the direct role of Smad3 in TGFβ1-mediated stromal modulation during fibrosis. The mRNA expression profile of pro-fibrotic gene expression reveals that Smad3 gene deficiency directly regulates the downstream targets of TGFβ/Smad signaling.
To further support our lead hypothesis—a direct role of Smad3 in cellular differentiation during corneal wound healing—the corneal fibrosis marker α-SMA was used to show the protein expression between the Smad3+/+ and Smad3−/− mice in naïve and injured corneas (Figure 6). The mRNA expression of α-SMA was significantly increased (p<0.0001) in the Smad3+/+ mouse injured corneas compared with Smad3−/− mouse injured corneas (Figure 6A). The protein expression of α-SMA was also increased significantly, as shown by immunofluorescence staining (Figure 6B) and western blot analysis (Figure 6C,D), in Smad3+/+ mouse injured corneas compared with Smad3−/− mouse injured corneas collected after euthanasia after alkali trauma on day 21. The western blot analysis data were corroborated with the mRNA and immunofluorescence staining and indicated that Smad3−/− group corneal tissue lysates showed a 2.16±0.56-fold reduction in α-SMA expression compared with Smad3+/+ mouse group corneal tissue (Figure 6D).
To study the morphological and structural changes in corneal tissues, Smad3−/− and Smad3+/+ naïve and post-alkali injured mouse corneal tissue sections were subjected to H&E (Figure 7A) and Masson’s trichrome staining (Figure 7B). No morphological alterations were observed in Smad3−/− and Smad3+/+ naïve mouse corneal tissue sections (Figure 7A a,b). However, Smad3−/− mouse post-alkali injury corneal tissue sections showed less stromal cellular infiltration and distorted collagen lamellae compared with Smad3+/+ mouse day 21 post-alkali injury corneal tissue sections (Figure 7A c,d).
The Masson’s trichrome staining data were corroborated with H&E staining. No noticeable changes were perceived in the collagen level of Smad3−/− and Smad3+/+ naïve mouse corneal tissue sections (Figure 7B a,b). Additionally, Smad3−/− mouse post-alkali injury corneal tissue sections showed less intensity of blue color, indicating better stromal reorganization and collagen lamellae compared with Smad3+/+ mouse day 21 post-alkali injury corneal tissue sections (Figure 7B c,d).
To test the postulate that Smad3 gene deficiency impacts TGFβ/Smad signaling, the mRNA expression levels of TGFβ, Smad2, Smad4, and Smad7 were quantified in Smad3−/− mouse corneal tissue and compared with Smad3+/+ mouse corneal tissue during active wound healing (Figure 8). The results of the mRNA expression study revealed a significant alteration in TGFβ (Figure 8A; p<0.01), Smad2 (Figure 8B; p<0.01), Smad4 (Figure 8C; p<0.05), and Smad7 (Figure 8D; p<0.01) and evidenced the role of Smad3 in TGFβ/Smad signaling. The mRNA expression of TGFβ and Smad reveals that Smad3 gene deficiency directly regulates profibrotic cascades during active wound healing.
MMPs and TIMPs critically regulate the active wound-healing event. To consider this, we examined the effect of Smad3 gene deficiency on MMP and TIMP expression during corneal wound healing in both Smad3−/− and Smad3+/+ mouse corneal tissue (Figure 9). The Smad3−/− mouse corneal tissue showed lower MMP-1, MMP-2, and MMP-9 mRNA expressions (p<0.001) compared with the Smad3+/+ mouse corneal tissue (Figure 9A). Additionally, the protein level western blot data analysis showed that MMPs and TIMPs were expressed at higher levels during the injury as compared with the naïve control in both groups, but in Smad3−/− mouse corneal tissue lysate, the expression was lower than in the Smad3+/+ mouse corneal tissue lysate (Figure 9B). In addition, gelatin zymography showed both the pro and active forms of MMPs, and data were well corroborated with the mRNA and protein expression data (Figure 9C).
To validate and support our lead hypothesis—a direct role of Smad3 in cellular differentiation during corneal wound healing—donor human normal and traumatic corneal tissue sections were used for the assessment of Smad3 expression (Figure 10). The protein expression of Smad3 was increased notably in the traumatic donor human corneal tissue sections (Figure 10B) as compared with the corresponding control nontraumatic donor human corneal tissue sections (Figure 10A). The increased Smad3 expression in the traumatic donor human corneal tissue aligns with our proposed hypothesis and shows the direct role of Smad3 during corneal wound healing.
Corneal injury involves orchestrated cellular growth factors and cytokine response to maintain physiologic homeostasis during active wound healing [1-7]. During active wound healing, many processes are involved, including the activation of fibroblasts and the reorganization and synthesis of the ECM [8-14]. The transparency of the corneal tissue is the key factor for normal vision. Disruption of the corneal transparency leads to vision impairment. Changes in corneal transparency are impacted by various ocular insults (e.g., mechanical, surgical, or chemical). These insults alter the transdifferentiation, production of various growth factors, cytokines, and excess accumulation of ECM proteins, leading to shifts in corneal curvature, myofibroblast formation, and collagen fibrillogenesis [8-14]. During traumatic injury, TGFβ can stimulate both canonical and noncanonical pathways involving TGFβRII, TGFβRI, BMP7, Smad2, Smad3, and Smad4 to affect a variety of cellular responses [23-31,52]. Smad intracellular proteins have the important role of transmitting TGFβ superfamily ligand signals to the nucleus [23-32]. In the TGFβ signaling pathway, the phosphorylation of R-Smads (Smad2 and Smad3) with co-Smad4 and the translocation of this complex regulates the cellular transdifferentiation during the active wound healing that impacts the stromal regeneration and ECM remodeling [7,23-32,53]. Unlike other cytokines in the corneal tissue, TGFβ primarily regulates the ECM protein accumulation, collagens, and fibronectin synthesis and alters the MMP–TIMP balance during the different wound-healing phases [33-39]. Our research group and others showed that overexpression of BMP7 and Smad7 gene transfer via adenoviral vectors topically suppresses the alkali-induced fibrotic response in vivo and showed the translational perspective of Smads as a potential therapeutic target for corneal fibrosis [7,52-54]. In Smad-dependent transdifferentiation, the roles of Smad2 and Smad3 are apparent, as shown by various research groups: Smad2 null mutation leads to embryonic lethality, whereas mice lacking Smad3 survive [53-58]. In addition, corneal researchers have shown that the ablation of the Smad3 gene does not arrest epithelial repair in a mouse cornea [26,53,59]. To consider this, the direct role of Smad3 in corneal stromal wound healing and stromal remodeling required precise investigations. In these studies, we showed that corneal transparency was noticeably superior in the Smad3 gene ablated mouse corneas compared with the wild-type mouse corneas (Figure 2). The absence of the Smad3 gene did not impact the corneal transparency and corneal epithelium’s regular turnover (Figure 3). No significant differences were reported in the IOP of both groups of animals (Table 2), reflecting that ablation of the Smad3 gene does not impact the flow of aqueous humor in ocular tissue. To better understand the functional role of Smad3, naïve and alkali-injured corneal tissue sections were evaluated for inflammatory protein (CD11b) after euthanasia on day 21 in wild-type and Smad3-deficient mice. In general, wild-type (Smad3+/+) alkali-injured corneas showed aggravated inflammatory responses compared with Smad3-deficient (Smad3−/−) alkali-injured corneas (Figure 4). Our mRNA (Figure 5) and protein (Figure 6) data undoubtedly illustrated the role of Smad3 in alkali-driven stromal transdifferentiation and corneal fibrosis. Furthermore, Mason’s trichrome staining confirmed and illustrated that Smad3 plays a critical role in ECM architecture, as shown by excessive collagen synthesis in the corneal tissue (Figure 7). The characteristic organization of stromal collagen fibrils provides mechanical strength and is required for the transparency of corneal tissue to maintain vision. The alterations in collagen levels in Smad3+/+ compared with Smad3−/− revealed the direct role of Smad3 in stromal ECM protein modulation during active wound healing (Figure 7). TGFβ-activating mechanisms depend on receptor-activated Smads [7]. The relative mRNA expression profile of TGFβ and other Smads showed the rapport harmony of Smad3 on TGFβ/Smad signaling in wild-type and Smad-deficient mouse corneal tissue during naïve and alkali trauma conditions (Figure 8). Therefore, in a clinical setting, an approach that incorporates multiple modalities may have greater therapeutic efficiency. Smad3, in particular, has been shown to interact with Smad2 and oligomerize with co-Smads to form a heteromeric complex, thereby interacting with R-Smad-Smad4 oligomerization [7]. Disruption in a heteromeric complex of R-Smads has the potential to control the fibrosis cascade and serve as a potential target to regulate active wound healing. Collectively, our findings suggest that inhibition of Smad3 is an important aspect for regulating TGFβ1-mediated differentiation and ECM protein modulation, as shown by Smad3−/− deficient mice during alkali trauma.
In the past, our research group and others showed the critical role of MMPs in wound healing and ECM matrix remodeling [25,33-39]. This study supports our previous findings that both the pro- and active MMPs are involved during the active phase of wound healing. Increased MMP and TIMP expression levels were also found during the corneal stromal fibroblast to myofibroblast transdifferentiation (Figure 9). The crosstalk of Smad3 with the regulation of MMP and TIMP levels is another important aspect of the degradation and remodeling of the stromal ECM and modulation of stromal opacification [33-39]. These findings agree with other research showing that the ECM components expression, stromal reorganization, and collagen fibrillogenesis depend on Smad3 [20,25,29,33-39]. In the ECM protein modulation, the MMPs and collagens play a decisive role during corneal wound healing, and our finding regarding Smad3 gene deficiency showed the direct relationship between Smad3, collagen, and MMP expression in the cornea (Figure 9). The expression of matrix metalloproteinase-2 is Smad2-dependent. Saika et al. and others showed that in the cornea, extirpation of the Smad3 gene had a similar anti-profibrogenic effect in the healing after an alkali burn in mice [25-27,53,54,59-64]. The results of this work agree with our previous work and other investigators' work showing that the overexpression of Smad7 (inhibitory Smad) blocks the TGFβ/Smad-driven transdifferentiation of keratocytes to myofibroblast transformation and Smad-driven ECM remodeling and collagen fibrillogenesis [7,12,21,26-28,31,53-60]. The literature shows that the TGFβR-II phosphorylates the MH2 domains of Smad3, and translocation of the phosphorylated Smad complex regulates transdifferentiation during corneal fibrosis [60-66]. Our donor human corneal tissue data from the traumatic and nontraumatic samples further support and validate the proposed hypothesis and show the role of Smad3 in corneal wound healing (Figure 10). Taken together, a significant increase in profibrotic genes (α-SMA and FN), changes in collagen levels, MMPs (ECM regulatory enzymes), and changes in TGFβ1 expression depict the direct role of Smad3 in corneal wound healing and stromal remodeling, in agreement with the findings of other corneal researchers showing that phosphorylated Smad3 binds directly to the promoter region of the COL1A2 gene to activate transcription in embryonic mouse dermal fibroblasts [67-69].
In conclusion, this study reveals a direct role of Smad3 in the TGFβ1-mediated cascade that influences transdifferentiation of stromal keratocyte/fibroblasts to opaque myofibroblasts during corneal wound healing. Additionally, it highlights that Smad3 is a promising target for controlling abnormal corneal wound healing and fibrosis in vivo. However, additional research is required for a better mechanistic understanding of Smad3’s role in the treatment of fibrosis in ocular and non-ocular tissues.
This work was supported by the 1I01BX000357 Merit and IK6BX005646 SRCS awards (R.R.M.) from the United States Department of Veterans Affairs BLR&D, Washington DC, USA; R01EY017294 and R01EY030774 grants (R.R.M.) from the National Eye Institute, NIH, Bethesda, Maryland, USA; and University of Missouri Ruth M. Kraeuchi Missouri Endowed Chair Ophthalmology Fund (RRM). Disclaimer: The contents do not represent the views of the USA Department of Veterans Affairs or the United States Government. Declaration of competing interest: None of the authors have any conflict of interest to disclose. Availability of data and materials: The data sets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.