Mustard Gas Exposure Actuates SMAD2/3 Signaling to Promote Myofibroblast Generation in the Cornea

Sulfur mustard gas (SM) is a vesicating and alkylating agent used as a chemical weapon in many mass-casualty incidents since World War I. Ocular injuries were reported in >90% of exposed victims. The mechanisms underlying SM-induced blindness remain elusive. This study tested the hypothesis that SM-induced corneal fibrosis occurs due to the generation of myofibroblasts from resident fibroblasts via the SMAD2/3 signaling pathway in rabbit eyes in vivo and primary human corneal fibroblasts (hCSFs) isolated from donor corneas in vitro. Fifty-four New Zealand White Rabbits were divided into three groups (Naïve, Vehicle, SM-Vapor treated). The SM-Vapor group was exposed to SM at 200 mg-min/m3 for 8 min at the MRI Global facility. Rabbit corneas were collected on day 3, day 7, and day 14 for immunohistochemistry, RNA, and protein lysates. SM caused a significant increase in SMAD2/3, pSMAD, and ɑSMA expression on day 3, day 7, and day 14 in rabbit corneas. For mechanistic studies, hCSFs were treated with nitrogen mustard (NM) or NM + SIS3 (SMAD3-specific inhibitor) and collected at 30 m, 8 h, 24 h, 48 h, and 72 h. NM significantly increased TGFβ, pSMAD3, and SMAD2/3 levels. On the contrary, inhibition of SMAD2/3 signaling by SIS3 treatment significantly reduced SMAD2/3, pSMAD3, and ɑSMA expression in hCSFs. We conclude that SMAD2/3 signaling appears to play a vital role in myofibroblast formation in the cornea following mustard gas exposure.


Introduction
Sulfur mustard gas (SM) is a chemical warfare agent used in many wars and conflicts since its first use in World War I [1][2][3][4][5][6]. Sulfur mustard is the most abundantly stockpiled chemical warfare agent worldwide due to its easy production, easy deployment, and strong vesicant properties. The recent employment of SM in terrorist activities and the Syrian civil war is a significant concern. SM is known to cause severe injuries to the skin, eyes, and lungs, with long-term complications. For instance, about 50% of the 100,000 Iranians exposed to SM during the Iraq-Iran war from 1980-1988 suffered chronic respiratory, skin, and eye complications [1][2][3][4][5][6]. Unfortunately, 90% of SM gas-exposed victims experience acute and chronic ocular symptoms 30+ years after exposure [4,7,8].
SM exposure to the eyes causes severe corneal injury, ocular pain, irritation, and significant vision impairment [9]. In some cases, ocular complications persisted several months to years after the initial exposure [4,7,10,11]. The ocular pathology resulting from SM exposure is often described as mustard gas keratopathy (MGK). Clinically, MGK is characterized by

Animal Tissue Processing
Following euthanasia, corneas with 2 mm of the sclera were removed by enucleating the eye, puncturing the eye using a number 11 scalpel blade approximately 5 mm posterior to the corneal limbus. A circumferential cut was made to remove the anterior ocular tissues, including the ciliary body and iris. The ciliary body and iris were removed from the cornea using surgical forceps and Westcott scissors under a dissecting microscope (Leica Wild M690, Leica Microsystems Inc., Buffalo Grove, IL, USA). Corneas were cut in two halves: one half was used for histology studies, and the other half was used for molecular studies. The histology half was immediately placed into a mold containing optimal cutting temperature (OCT) compound and snap frozen in a container of 2-methyl butane immersed in liquid nitrogen. Frozen tissues were maintained at −80 • C until sectioning and further evaluation. Tissues were sectioned at 8-µm thickness with a cryostat, mounted on microscopic glass slides (SuperFrost Plus, Fisher Scientific, Waltham, MA, USA), and preserved at −80 • C for subsequent analysis. The molecular half was divided in half to make two quarter corneas. One quarter was used for RNA extraction and cDNA synthesis. One quarter was used for protein analysis.

In Vitro NM Studies
Primary human corneal stromal fibroblasts (hCSF) were generated from human donor corneas purchased from the Saving Sight Foundation, Kansas City, Missouri, using our established protocol [52,[55][56][57]. In brief, a #15 surgical blade was used to remove the corneal epithelium and endothelium gently. Afterward, the cornea was cut into eight even pieces and incubated in 60-mm culture dishes with 3 mL Minimum Essential Medium supplemented with 10% fetal bovine serum (FBS) (Thermo Fisher, Grand Island, NY, USA) and grown at 37 • C in humidified 5% CO 2 incubator. Primary cells were seeded in six-well tissue culture plates at passage three for all in vitro experiments. Each treatment was performed in duplicate, and each experiment was repeated three times (n = 6) for high scientific rigor and to evaluate statistical significance. hCSF cells were cultured with or without NM (100 ng/mL) [44]. NM was prepared by reconstituting mechlorethamine hydrochloride (Cat no. 122564, Sigma-Aldrich, St. Louis, MO, USA) in culture media. Additionally, hCSF cells were pretreated with or without SIS3, a SMAD3-specific inhibitor, and exposed to 100 ng/mL of NM. SIS3 (Cat no. 5291, R&D systems, Minneapolis, MN,

Real-Time PCR
Corneal tissues were minced in a tissue lyser (TissueLyser LT, Qiagen, Valencia, CA, USA) in RLT buffer (Qiagen, Valencia, CA, USA), and the total RNA was isolated using the RNeasy kit (Qiagen, Valencia, CA, USA) following the manufacturer's instructions. hCSF cells were collected using RLT butter. The reverse transcriptase enzyme kit was used to synthesize first-strand cDNA (M8291, Promega, Madison, WI, USA). One Step Plus Real-Time PCR system (Cat no. 4309135, Applied Biosystems, Carlsbad, CA, USA) was used for quantitative PCR (qPCR). A 20-µL reaction mixture containing 2 µL cDNA, 2 µL forward and reverse primers (200 nM each), and 10 µL of 2X All-in-One PowerUp SYBR green master mix (Applied Biosystems, Carlsbad, CA, USA) was run at a universal cycle (95 • C for 10 min, 40 cycles at 95 • C for 15 s, and 60 • C for 60 s). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an endogenous reference gene. The threshold cycle (Ct) was used to detect the increase in the signal associated with the exponential growth of PCR products during the log-linear phase. ∆Ct for each sample was calculated by subtracting the Ct of the target gene from that of the Ct of the endogenous reference gene, and ∆∆CT was calculated by subtracting the ∆Ct of the test sample from the ∆Ct of the control sample. The relative mRNA expression was calculated using the 2 −∆∆Ct method and reported as a relative fold change over the corresponding control values. The amplification efficiency for the qRT-PCR was similar for all templates used, and the difference between linear slopes was less than 0.1. The qPCR was performed in triplicate for each sample, and a minimum of three independent experiments were conducted. Accession numbers and sequences are in Table 1. was loaded. The gel was run and transferred to a nitrocellulose membrane using XCell II Blot Module (E19051, Novex ThermoFisher, Waltham, MA, USA). The nitrocellulose membrane was washed with TBS for 3 × 10 m and blocked with 5% fat-free milk for 1 h at room temperature. Immunostaining was performed using a primary antibody (at recommended dilution) and kept overnight at 4 • C, followed by a 4-h incubation with a secondary antibody (at recommended dilution). The membrane was washed 3× in TBST and incubated in SuperSignal West Femto (34094, Fisher, Pittsburgh, PA, USA) for 10 min. The membrane was then analyzed using a C-Digit digital Western Blot analyzer (C-digit Li-Cor, St, Lincoln, NE, USA). Li-Cor ImageStudioDigits software version 5.2.5 was used for densiometric quantification of western blot bands. This analysis software does not allow tweaking to live capturing and highlights oversaturated images in blue, which prevents quantification of blots. %pSMAD3/SMAD2/3 was calculated by diving pSMAD3 (Cat. No. Sc-517575, Santa Cruz, CA, USA) signal by SMAD2/3 (Cat. No. Sc-133098, Santa Cruz, CA, USA) signal for each timepoint. β-actin (Cat. No. 3700, Cell Signaling, Danvers, MA, USA) signals were compared to the no-treatment group for all timepoints. Then pSMAD3/SMAD2/3 ratios were normalized to respective β-actin expression.

Immunocytochemistry
Cultured cells with and without NM were prepared, post-fixed in 4% PFA at room temperature for 30 m, and blocked with 5% donkey serum for 1 h. Immunostaining was performed using a primary antibody (at recommended dilution) and kept overnight at 4 • C followed by a 4h incubation with a secondary antibody (at recommended dilution). A drop of DAPI antifade Vectashield medium (Cat. No. H1200, Vector Laboratories, Newark, CA) was applied, and tissue sections were mounted with premier coverslips (Thermo Fisher, Waltham, MA, USA). The stained sections were viewed and photographed with a fluorescence microscope (Leica DM 4000B, Leica Microsystems Inc., Buffalo Grove, IL, USA) equipped with a digital camera (SpotCam RT KE, Diagnostic Instruments Inc., Sterling Heights, MI, USA). The number of cell nuclei (DAPI staining) and nuclei with αSMA (Cat. No. M0851, DAKO, Santa Clara, CA, USA) expressions (GFP) was quantified by manually counting six random images taken from culture dishes.

Immunohistochemistry
8-µm thick corneal sections were prepared, postfixed at room temperature for 10 min, and blocked with 5% donkey serum for 1 h at room temperature. Immunostaining was performed using αSMA (Cat. No. M0851, DAKO, Santa Clara, CA, USA) primary antibody (at recommended dilution) and kept overnight at 4 • C, followed by a 4 h incubation with a secondary antibody (at recommended dilution). A drop of DAPI antifade Vectashield medium (Cat. No. H1200, Vector Laboratories, Newark, CA) was applied, and sections were mounted with premier coverslips. The stained sections were viewed and photographed with a fluorescence microscope (Leica DM 4000B, Leica Microsystems Inc., Buffalo Grove, IL, USA) equipped with a digital camera (SpotCam RT KE, Diagnostic Instruments Inc., Sterling Heights, MI, USA).

Statistical Analysis
GraphPad Prism Version 9.2 (GraphPad Software, La Jolla, CA, USA) software was used for statistical analysis. Each experiment was conducted independently with n provided in the test, and the values were expressed as mean ± SD. For statistical analysis, the student's t-test and two-way analysis of variance (ANOVA) with Bonferroni post-hoc test was used. The value of p ≤ 0.05 was considered significant. The sample size was determined using the G*Power (3.1.9.4 software, www.psycho.uni-duesseldorf.de/abteilungen/ aap/gpower3, accessed on 19 April 2021) priori power analysis method to achieve α = 0.05; power ≥ 0.9.
3.1.2. Localization of αSMA in Rabbit Corneas Exposed to SM In Vivo SM-exposed corneas had increased immunostaining of αSMA expression in the corneal stroma in a time-dependent manner. Day-3 corneas had minimal αSMA staining ( Figure 2B), which increased significantly on day 7, and day 14. In addition, day-7 had αSMA staining localized in the anterior stroma ( Figure 2C), whereas day-14 had αSMA staining throughout the corneal stroma ( Figure 2D) compared to the naïve group ( Figure 2A).

Time-Dependent Expression of SMAD Signaling Proteins in Rabbit Corneas
Exposed to SM In Vivo SM-exposed corneas had a significant increase in phosphorylated SMAD3 to total SMAD (pSMAD3/SMAD2/3) expression at day 7 (p < 0.05) and day 14 (p < 0.0001) compared to the naïve group ( Figure 3). ɑSMA expression was significantly increased at day 7 (p < 0.0001) and day 14 (p < 0.0001) compared to the naïve group.

Time-Dependent Expression of SMAD Signaling Proteins in Rabbit Corneas
Exposed to SM In Vivo SM-exposed corneas had a significant increase in phosphorylated SMAD3 to total SMAD (pSMAD3/SMAD2/3) expression at day 7 (p < 0.05) and day 14 (p < 0.0001) compared to the naïve group ( Figure 3). ASMA expression was significantly increased at day 7 (p < 0.0001) and day 14 (p < 0.0001) compared to the naïve group. Figure 2. Expression of αSMA protein in rabbit corneas in vivo −/+ SM. SM-exposed corneas had increased staining of αSMA in the corneal stroma in a time-dependent manner compared to naïve (A). Day-3 did not have much αSMA staining (B) compared to day-7 (C) and day-14 (D). Day-7 had αSMA staining at the anterior stroma (C) and day-14 had αSMA staining throughout the corneal stroma (D). (Red = ɑSMA; Scale bar = 100 µM; n = 6).

Discussion
Sulfur mustard gas (SM) is a potent alkylating agent that causes haze formation upon contact with the eye [58]. Currently, the mechanism of action used by SM to generate corneal haze/fibrosis in the stroma is unknown. Haze results from residential keratocytes transdifferentiating into opaque myofibroblasts (EMT) that produce excess extracellular matrix and persist after physiological wound healing has terminated. Myofibroblasts are absent in a healthy functioning cornea [44,55,59]. Various studies have shown SM has a multifactorial cause of damage, including oxidative stress, inflammation, cell death, and dysregulated signaling pathways [44,45]. In the present study, rabbit corneas exposed to sulfur mustard vapor in vivo showed significantly increased mRNA levels of TGFβ and αSMA on day 3, day 7, and day 14. IHC staining for αSMA in tissue sections confirmed increased αSMA expression and myofibroblast formation in the corneal stroma.
Western blot analysis of SM-exposed rabbit corneas in vivo were used to identify the SMAD signaling pathway as a mechanism for SM-induced myofibroblast formation. Protein analysis showed a significant increase in the pSMAD3/SMAD2/3 ratio and αSMA levels at day 7 and day 14 after SM exposure, thus demonstrating mustard gas exposure caused excessive myofibroblast accumulation in rabbit corneas. Further, a correlative time-dependent increase in SMAD signaling was present after SM exposure in vivo. This expression is analogous to our previous findings suggesting sulfur mustard-induced opacity and edema on day 7, followed by haze/fibrosis on day 14 [58]. Therefore, SMAD signaling was further examined as a possible mechanism for SM-induced corneal myofibroblast in vitro.
Human corneal stromal fibroblast cells (hCSFs) were used to establish TGFβ/SMAD signaling as a molecular pathway for myofibroblast formation. hCSFs were treated with nitrogen mustard (NM), a SM analog used in the lab, in a time-dependent manner in vitro. A significant increase in protein and mRNA levels of αSMA at 24 h, 48 h, and 72 h in were observed in hCSFs, which signifies the formation of myofibroblasts [60]. Additionally, a significant increase in mRNA levels of SMAD2, SMAD3, SMAD4, and SMAD7 was observed in hCSFs. SMAD2, SMAD3, and SMAD4 are described as profibrotic components in corneal wound healing [25]. In addition to keratocyte transdifferentiation to myofibroblasts, the phosphorylation of SMAD3 has also been related to increased ECM production at early time points [25]. Thus, a future study examining the expression of collagen and other ECM would further our understanding of mustard gas-induced fibrosis. Conversely, SMAD7 is an inhibitory member of the SMAD family that prevents the SMAD2/3 complex from binding to SMAD4 and entering the nucleus. Therefore, the increase in SMAD7 at 24 h and 48 h may have a regulatory role in preventing myofibroblast formation. However, a decrease in SMAD7 mRNA was seen at 72 h, which may result from a downstream repressor protein blocking new SMAD7 transcription.
Western blot analysis of hCSFs treated with NM showed a significant time-dependent increase in TGFβ and αSMA protein levels from 24 h to 72 h. SMAD activity was analyzed by testing the pSMAD3 to total SMAD (pSMAD3/SMAD2/3) ratio in a time-dependent manner in vitro. The pSMAD/SMAD2/3 ratio was significantly increased at all timepoints following NM exposure. A time-dependent change can be seen from 30 m to 48 h; however, a slight decrease can be seen at 48 h compared to 72 h. This decrease may result from all hCSFs being converted to myofibroblasts, as seen by ICC staining of ASMA by 72 h. In addition, a decrease in pSMAD3 and total SMAD2/3 can be seen at 72 h, which might be due to ubiquitination/degradation.
Further analysis with a SMAD3-specific inhibitor, SIS3, validated the role of SMAD3 and SMAD signaling as a predominant pathway following mustard gas injury. SIS3 caused a significant reduction in pSMAD3/SMAD2/3 from 30 m to 24 h and αSMA protein expression from 8 h to 24 h compared to the NM only group. An interesting finding was the increase in total SMAD and pSMAD3 even after the use of a chemical inhibitor. These results suggest that the synthesis of new SMAD proteins maybe a unique property of mustard gas keratopathy due to the dual nucleophilic chemical property of the mustard gas. For other alkylating agents, such as sodium hydroxide (NaOH), it typically takes 72 h post-exposure for 80-90% of cells to become myofibroblasts [57]. However, NM caused almost all cells to become myofibroblast between 48 h and 72 h. Another possibility for the continued increase in SMAD2/3 and pSMAD3 is the crosstalk between other signaling pathways driven by mustard gas toxicity. Thus, SIS3 showed the profibrotic cascade of the SMAD family being a dominant pathway in myofibroblast formation compared to inhibitory SMAD7 in hCSF treated with NM. Future RNAseq or luciferase assay studies would provide insight into the observed disconnect between mRNA production and protein translation.

Conclusions
We conclude that TGFβ/SMAD signaling is a predominant pathway for myofibroblast formation in the cornea exposed to mustard gas toxicity.