Dexamethasone targets actin cytoskeleton signaling and inflammatory mediators to reverse sulfur mustard-induced toxicity in rabbit corneas

Purpose:

Sulfur mustard (SM), a bi-functional alkylating agent, was used during World War I and the Iran-Iraq war. SM toxicity is ten times higher in eyes than in other tissues. Cornea is exceptionally susceptible to SM-injuries due to its anterior positioning and mucous-aqueous interphase. Ocular SM exposure induces blepharitis, photosensitivity, dry eye, epithelial defects, limbal ischemia and stem cell deficiency, and mustard gas keratopathy leading to temporary or permanent vision impairments. We demonstrated that dexamethasone (Dex) is a potent therapeutic intervention against SM-induced corneal injuries; however, its mechanism of action is not well known. Investigations employing proteomic profiling (LC-MS/MS) to understand molecular mechanisms behind SM-induced corneal injury and Dex efficacy were performed in the rabbit cornea exposed to SM and then received Dex treatment. PEAKS studio was used to extract, search, and summarize peptide identity. Ingenuity Pathway Analysis was used for pathway identification. Validation was performed using immunofluorescence. One-Way ANOVA (FDR<0.05; p<0.005) and Student’s t-test (p<0.05) were utilized for analyzing proteomics and IF data, respectively. Proteomic analysis revealed that SM-exposure upregulated tissue repair pathways, particularly actin cytoskeleton signaling and inflammation. Prominently dysregulated proteins included lipocalin2, coronin1A, actin-related protein2, actin-related protein2/3 complex subunit2, actin-related protein2/3 complex subunit4, cell division cycle42, ezrin, bradykinin/kininogen1, moesin, and profilin. Upregulated actin cytoskeleton signaling increases F-actin formation, dysregulating cell shape and motility. Dex reversed SM-induced increases in the aforementioned proteins levels to near control expression profiles. Dex aids corneal wound healing and improves corneal integrity via actin cytoskeletal signaling and anti-inflammatory effects following SM-induced injuries.

1. Introduction

Sulfur mustard (SM), bis(2-chloroethyl) sulfide, is a bifunctional, cytotoxic, alkylating chemical warfare agent (CWA) (Geraci 2008). Albert Niemann first reported its blistering properties (1860) (Guthrie 1860, Niemann 1860, Ghabili, Agutter et al. 2011, Panahi, Fattahi et al. 2018). It was first deployed in warfare during World War I and more recently in the Iran-Iraq War (1980–1988) (Geraci 2008, Balali-Mood, Afshari et al. 2011). SM is lipophilic, solubilizes physiological lipids (Kehe and Szinicz 2005, McNutt, Kelly et al. 2021), reacts with biomolecules (Kehe and Szinicz 2005), targets nucleic acids (Kumar, Tewari-Singh et al. 2015) causing genotoxicity (Ghabili, Agutter et al. 2011, Goswami, Tewari-Singh et al. 2016, McNutt, Kelly et al. 2021), and causes oxidative stress (Ghabili, Agutter et al. 2011).

More than 90% SM-exposed Iranian survivors reported ocular injuries (McNutt, Hamilton et al. 2012). Eyes are ~ten times more sensitive than other organs to CWAs (Balali-Mood, Afshari et al. 2011, McNutt, Hamilton et al. 2012, McNutt, Kelly et al. 2021) and cornea is most vulnerable because of its external location and aqueous-mucous interface (McNutt 2023). SM-injuries can be acute or chronic, depending upon the dose and duration of exposure, and may include irritation, lacrimation, inflammation, photophobia, blepharospasm, corneal neovascularization, opacity, and ulceration, epithelial defects, dry eye, conjunctival and corneal scarring, and limbal stem cell deficiency and mustard gas keratopathy (Balali-Mood, Afshari et al. 2011, McNutt, Hamilton et al. 2012, Goswami, Mishra et al. 2021, McNutt, Kelly et al. 2021).

SM-induced corneal damage in rabbit models includes ulceration, opacity, inflammation, and neovascularization at the clinical level (Goswami, Mishra et al. 2021). Histopathological effects include corneal thickening, epithelial degradation, epithelial-stromal separation, keratocyte death, inflammatory cell infiltration, and increase in blood vessel numbers (Goswami, Tewari-Singh et al. 2016). It also increases inflammatory mediator expressions, namely cyclooxygenase (COX)-2, matrix metalloproteases (MMP)-9, and vascular endothelial growth factor (Goswami, Tewari-Singh et al. 2016, Goswami, Mishra et al. 2021). However, over a century since SM was deployed as a CWA and despite decades of research, approved countermeasures are still lacking. Studies in our lab, using ex vivo cultured rabbits (Goswami, Kant et al. 2018) and humans (Neha, Rama et al. 2023) corneas and in vivo rabbit eyes (Goswami, Mishra et al. 2022, Mishra, Kant et al. 2023), and from other groups (Kadar, Dachir et al. 2009, Goswami, Mishra et al. 2022, Mishra, Kant et al. 2023) have shown that dexamethasone (Dex), an FDA-approved anti-inflammatory drug, is an effective therapeutic against mustard vesicant-induced corneal injuries (Goswami, Kant et al. 2018). Although literature has bolstered Dex efficacy against SM-induced corneal injuries, mechanistic details of Dex action remain obscure. Using a more advanced high throughput omics analysis technique will help us identify the proteins that imperatively participate in tissue injury post SM exposure and in tissue repair mechanism after Dex treatment of SM exposed tissue. Proteomic profiling serves as a complement to the genomic outcomes derived from previous studies conducted by different groups, and confirms the translational implications of the gene-level efficacy of Dex in countering dysregulations that occur post SM exposure. Therefore, we performed proteomic analysis of SM-exposed and Dex-treated corneas post SM-exposure to get deeper insights into mechanisms underlying Dex efficacy against SM-induced corneal injuries.

2. Method and material

2.1. Animals and study design

For this study, New Zealand white male rabbits, weighing 5.5 lbs to 8 lbs were purchased from Charles River Laboratories and housed in MRIGlobal conventional animal facility. All animal studies were conducted after the approval of the respective Institutional Animal Care and Use Committees. In brief, one week after acclimatization animals were divided into two groups; group 1: both the eyes were exposed to SM and then treated with Dex (n=6/timepoint), Group 2: only right eye was exposed to SM (once, for 90 sec.), left eye received no SM exposure or Dex treatment, hence used as experimental control (n=5/timepoint) (Mishra, Kant et al. 2023). SM exposure was performed using already approved “Vapor Cup Exposure System” (Mishra, Kant et al. 2023). In our previous study, we assessed the efficacy of Dex administration, initiated 2 h after SM exposure and then twice daily (every 12 h) or thrice daily (every 8 h) for 28 days (Mishra, Kant et al. 2023). Results showed that Dex administration every 8 h was more effective against SM-induced corneal injuries at clinical, pathophysiological, and molecular levels (Mishra, Kant et al. 2023). Based on these findings, for in-depth analysis of Dex mechanism of action in the present study, out of total four treatment groups (control, SM exposed, SM+Dex every 8 h, and SM+Dex every 12h) we performed proteomics analysis on corneas from three groups: control (no SM or Dex), SM (SM exposed), and SM+Dex (SM exposed+Dex every 8 h), at day 28 post SM-exposure. For proteomic profiling, snap-frozen tissues were utilized (n=4–5). For histological analysis, 10% formalin fixed and OCT preserved tissues were used, as described previously (Mishra, Kant et al. 2023). A schematic study outline is provided in Fig. 1A.

Fig. 1:

Study design, principal component analysis (PCA), and heatmap of corneal tissue proteomics. (A) Study design: 28-day time point corneal samples from control, SM only, and SM+Dex groups were used for proteomic analysis. (B) PCA plot shows the clustering of study samples within the groups and their overlapping. (C) Heatmap for proteins of interest identified after statistical analysis (p<0.05, using FDR<0.05, and One Way ANOVA posthoc analysis) from peptide database in proteomic analysis. (n=4–5)

2.2. Sample preparation: solubilization and digestion of rabbit corneas for LC-MS/MS

Samples were solubilized, reduced, alkylated, digested with trypsin and cleaned up using the PreOmics iST kit (cat#P.O.00001, Preomics, Planegg, Germany) for LC-MS/MS analysis. Isolated corneal tissue (4–5 mg) was lysed in 100 μl of preomics lysis buffer and sonicated with 10 cycles of probe on/off for protein solubilization and then heated at 95°C (10 min) for sample reduction and alkylation. Samples were digested with trypsin (3 h) and cleaned using the kit resin and buffers. All the instructions were followed as in the PreOmic iST kit protocol. Samples were re-suspended in the LC load buffer provided with the kit and brought to a concentration of 0.8 mg/mL as measured by BCA assay.

2.3. LC–MS/MS

Digested corneal peptides (8 μg) were chromatographically resolved on-line using a 2.1×250 mm 2.7 μm AdvanceBio peptide mapping column (Agilent Technologies, Santa Clara, California, USA) using a 1290 Infinity II LC system (Agilent, Santa Clara, CA). Mobile phases comprised of 0.1% formic acid (A) and 90% aq. acetonitrile +0.1% formic acid (B). Samples were separated chromatographically at a flow rate of 0.25 ml/min using a gradient of 5–30% B over 89 min and 30–40% B over 11 min for a total 100 min gradient at 50°C. Gradient method was followed by a column wash at 90% B (5 min). Data were collected on a 6550 Q-TOF (quadrupole time of flight) equipped with a dual Agilent Jet Stream source (Agilent) operated using intensity-dependent CID MS/MS (collision-induced dissociation-MS/MS) to generate peptide identifications from pooled group samples (Control, SM, SM+Dex). Samples were run individually in MS-only mode for general protein quantitation. MS/MS data were collected in positive ion polarity over mass ranges 260–1700 m/z at a scan rate of 8 spectra/s for MS scans and mass ranges 50–1700 m/z at a scan rate of 3 spectra/s for MS/MS scans. All charge states, except singly charged species, were allowed during MS/MS acquisition. MS-only data were collected in positive ion polarity over mass ranges 260–1700 m/z at a scan rate of 1.5 spectra/s.

2.4. Database search for peptides identification

PEAKS studio proteomics version 10.5 was used to extract, search, and summarize peptide identity results from pooled samples of each group (PEAKS, Waterloo, ON, Canada). Default settings were used to extract spectra without merging scans and detecting correct precursor with mass only. Peptide identifications were performed using the PEAKS DB search engine combined with PEAKS de novo sequencing. Extracted spectra were first searched against the SwissProt Oryctolagus cuniculus database and then the Unreviewed TrEMBL Rabbit Database. Samples were searched using semi-specific trypsin enzyme, allowing up to two missed tryptic cleavages with variable carbamidomethyl (C), deamidated (NQ), and oxidation (M) modifications. Monoisotopic peptide mass tolerance was set to ±20.0 ppm and the MS/MS tolerance to ±0.05 da. Peptides were validated by setting the false discovery rate (FDR) to 0.5%, corresponding to a −log 10 p-value score greater than or equal to 20.7 in the data set. The resulting peptide hits were used for the generation of an accurate mass and retention time (AMRT) library.

MS1 Level data were extracted and aligned using Profinder V.10.0 software (Agilent). Retention times, neutral masses, and chemical formulas from the AMRT library were used to perform a batch-targeted feature extraction. Samples were extracted with an ion count threshold set to two or more ions and 10000 counts and a score threshold of 70 in two or more samples. The score was based on the quality of the mass, isotope abundances, and isotope spacing of compounds found in each sample to a targeted chemical formula within a specified retention time window generated from the AMRT library. Charge states 2–6 were allowed with H+ adducts using the peptide isotope model. Retention time window and mass window alignment setting tolerances were set to ±0.35 min and ±10 ppm, respectively. Final extraction, alignment, and peak area results for identified peptides from the batch-targeted feature extraction were rolled up into total protein area sums and used for quantitation between samples.

2.5. Network analysis and pathway enrichment

Network and pathway enrichment analyses were performed for proteins found to be significantly dysregulated post-SM exposure and Dex treatment (biological function, pathways, and protein-protein networking) using web-based bioinformatics application Ingenuity Pathway Analysis (IPA; Qiagen, Valencia, CA; http://www.ingenuity.com). Proteins with fold-change values were uploaded onto IPA for Core analysis. Fisher’s Exact test was used to determine the statistical significance of pathway enrichment (p<0.05). Activation/inhibition of identified pathways was determined via z-score (≥2 and ≤−2) generated by IPA. Further literature review was also performed for deriving networks.

2.6. Immunofluorescence (IF)

IF was performed to validate protein expressions observed in proteomics analyses, using protocols previously described (Mishra, Kant et al. 2023). Briefly, formalin-fixed paraffin-embedded tissues were sectioned (5 μm; sagittal sections). Sections were cleared, rehydrated, incubated in sodium citrate (pH. 6; 90°C; 15 min), and permeabilized (0.3% triton X [cat#T8787, Sigma-Aldrich, St. Louis, MO, USA] in PBST [PBS, cat#BP399, Fisher Scientific, Hampton, NH, USA]; 2 h). Non-specific binding was blocked (5% BSA; cat#A9418, Sigma-Aldrich, St. Louis, Missouri, USA), and slides were incubated with respective primary antibodies (LCN2: cat#515876, CDC42: cat#SC-8401, CORO1A: cat#SC-100925; Santa Cruz Biotechnology, Dallas, Taxes, USA; dilution 1:50) overnight at 4°C. Further, tissues were incubated with Alexa 647-tagged fluorescence secondary antibody (cat# A-21236; Invitrogen Life Technologies, Waltham, Massachusetts, USA; dilution 1:500), washed with 0.1% triton X high-salt PBS, and mounted (cat#H-1800, Vector Laboratories, Newark, California, USA).

2.7. Phalloidin IF (P-IF)

P-IF was performed for the identification of filament actin (F-Actin). OCT embedded corneal tissues were sectioned (8 μm thick; cryostat: cat#CM1850, Leica Biosystems, Deer Park, IL, USA) and preserved at −80°C until staining. For P-IF, slides were brought to room temperature and fixed (4% paraformaldehyde, 25 min). Slides were then washed (thrice, 5 min/wash in PBS), incubated in 0.1% triton X (5 min in PBS), followed by washing. Slides were further incubated, in dark (90 min) with 1X Phalloidin CruzFluor 488 Conjugate (cat#SC-363791, Santa Cruz Biotechnology, Dallas, Taxes, USA) diluted in 0.1% triton X in PBS. After incubation, tissues were washed (twice, PBS) and mounted (cat#H-1800, Vector Laboratories, Newark, California, USA). Once dried, slides were analyzed for F-Actin localization.

2.8. Imaging and analysis

All IF/P-IF imaging was performed using Nikon Eclipse Ti2 inverted confocal microscope (NIS elements AR version 5.20). For analysis, 8–10 regions were selected from each sample randomly; images were captured at 200X magnification and further magnified two-times for signal intensity analysis, using QuPath software (version 0.4.3). Average score for each sample/group was calculated to yield the final fluorescence intensity score.

2.9. Statistical analyses

R (version 4.2.3) was used for statistical analysis. One-Way ANOVA was used to measure the omnibus effect, i.e., any difference between groups (FDR <0.05). Pairwise post-hoc testing was used to compare specific treatment groups. Principal component analysis (PCA) was executed and heatmap was generated using the prcomp function (for PCA) and heatmap function (using centered and scaled data). Student’s t-test was used for analyzing IF and P-IF data (p<0.05; n=4–5/group). Data presented as mean (±standard error of mean).

3. Results

3.1. Overview of proteomic profiling and protein clustering

Proteomic analysis was performed on the corneal tissues (groups: Control, SM, and SM+Dex) obtained at 28 day time point (Mishra, Kant et al. 2023). Time point of investigation was based on clinical, histopathology, and protein expression analysis in our previous study (Mishra, Kant et al. 2023). LC-MS/MS obtained peaks were analyzed for peptide identification. Further, searching through SwissProt Oryctolagus Cuniculus Database (containing 1405 proteins entries) and Unreviewed TrEMBL Rabbit Database (containing 49,707 proteins entries), we identified 857 unique proteins in all samples from the groups (control, SM, and SM+Dex). Visualization of proteomic similarities and differences across the three study groups, PCA (Fig. 1B) and heatmap (Fig. 1C) were employed.

In the PCA analysis, study groups were segregated by the first (PC1: 43.5%) and second (PC2: 16.0%) components of the total variation. SM group samples were loosely clustered (samples were spread across the plot), while the control group formed a tight and sequestered cluster, further from the SM group cluster. Sample clustering for the SM+Dex group was in the vicinity of and overlapped with the control group cluster, indicating that Dex treatment reversed many of the SM-induced effects on protein levels. A heatmap visualization displayed the level of alterations in the expression of significant proteins (Fig. 1C).

3.2. DEX treatment alleviated injury-associated stress post-SM exposure

SM exposure modulated proteins associated with corneal injury as well as repair mechanisms, and Dex treatment markedly reversed these expressions. Thirty proteins had statistically significant differential expression amongst the study groups (Table 1). These 30 proteins, the proteins of interest (POI), were selected for further analysis, group comparisons, and IPA-generated pathways and ontologies. Group-wise comparisons indicated that 29 proteins were significantly dysregulated by SM exposure (Fig. 2A). Twenty-six proteins were significantly dysregulated after Dex treatment (Fig. 2A). These proteins may contribute to the mechanistic aspects of Dex efficacy against SM-induced injuries. A comparison of SM+Dex and control groups revealed the involvement of 13 proteins that showed significant differential expression. This cluster comprised of nine proteins that were affected by SM exposure and reversed by Dex treatment; however, could not achieve expression profiles close to those of the control corneas (Fig. 2B & D). Of the 24 proteins significantly upregulated by SM exposure, 22 were downregulated after Dex treatment (Fig. 2C). Additionally, of the five proteins downregulated post-SM exposure, three were upregulated after Dex treatment (Fig. 2C). Thus, Dex treatment reversed 86% of the POIs modulated post-SM exposure.

Table 1:

Differentially expressed and statistically significant proteins identified from proteomics analysis

			SM vs. Control		SM+Dex vs. SM		SM+Dex vs. Control
Description	Gene ID	Any Treatment Effect (FDR)	Log10 (FC)	p-value	Log10 (FC)	p-value	Log10 (FC)	p-value
Fibrillar collagen NC1 domain-containing protein	UNK#	0.0375	−0.85	0.00028***	0.19	0.22615	−0.66	0.00120**
Alpha-1 type V collagen (Fragment)	COL5A1	0.0390	−0.76	0.00038***	0.17	0.24987	−0.59	0.00156**
Elongation factor Tu; Elongation factor Tu mitochondrial	TUFM	0.0254	−0.34	0.00043***	0.41	0.000054***	0.08	0.23450
Complement factor I	CFI	0.0270	−0.23	0.00647**	0.39	0.00011***	0.16	0.03067*
Dolichyl-diphosphooligosaccharide--protein glycosyltransferase subunit 1	RPN1	0.0254	−0.20	0.00007***	0.13	0.00164**	−0.08	0.02716*
Apolipoprotein H	CEP112	0.0254	−0.13	0.06528	0.36	0.00009***	0.23	0.00229**
Ezrin	EZR	0.0254	0.13	0.03390*	−0.32	0.00008***	−0.19	0.00377**
FERM domain-containing protein	FEMD	0.0254	0.17	0.00485**	−0.30	0.00004***	−0.14	0.01189*
Arp2/3 complex 34 kDa subunit	ARPC2	0.0390	0.20	0.00038***	−0.16	0.00115**	0.04	0.33180
Adenylyl cyclase-associated protein	CAP1	0.0254	0.23	0.00011***	−0.16	0.00115**	0.07	0.07722
Actin-related protein 2/3 complex subunit 4	ARPC4	0.0390	0.24	0.00035***	−0.18	0.00209**	0.06	0.17651
Actin related protein 2	ARP2	0.0417	0.24	0.00037***	−0.17	0.00356**	0.08	0.11129
Profilin	PFN	0.0254	0.25	0.00017***	−0.23	0.00019***	0.02	0.71173
Lymphocyte cytosolic protein 1	LCP1	0.0423	0.27	0.00091***	−0.24	0.001299**	0.03	0.64545
78 kDa glucose-regulated protein	GRP78	0.0423	0.27	0.00166**	−0.29	0.00068***	−0.02	0.73732
Histone H1.3	H1.3	0.0254	0.34	0.00008***	−0.26	0.00041***	0.08	0.15112
SH3 domain-containing protein	SORBS	0.0441	0.35	0.00056***	−0.24	0.00440**	0.10	0.14389
Afamin	AFM	0.0450	0.41	0.03718*	−0.82	0.00050***	−0.41	0.03046*
Kininogen 1	KNG1	0.0423	0.42	0.00130**	0.00	0.97397	0.41	0.00094***
Carbonic anhydrase	CA	0.0305	0.49	0.00030***	−0.44	0.00036***	0.05	0.61472
Deadenylyl-alpha-globin (Fragment)	UNK#	0.0423	0.56	0.00068***	−0.47	0.00175**	0.10	0.40202
Alpha-globin 1	HBA	0.0429	0.57	0.00072***	−0.47	0.00198**	0.09	0.44347
Phosphatidylethanolamine-binding protein 1	PEBP1	0.0254	0.58	0.00030***	−0.58	0.00021***	0.00	0.99134
Globin A1; Hemoglobin subunit gamma	HBA1	0.0390	0.59	0.00034***	−0.47	0.00117**	0.12	0.28756
Paraoxonase	PON1	0.0423	0.67	0.00042***	−0.40	0.00893**	0.27	0.05067
Cell division control protein 42 homolog	CDC42	0.0192	0.70	0.00000***	−0.33	0.00179**	0.38	0.00061***
Thrombospondin 1	THBS1	0.0450	0.84	0.00081***	−0.68	0.00241**	0.16	0.36182
Coronin 1A (Fragment); Coronin	CORO1A	0.0264	0.95	0.00010***	−0.42	0.0161**	0.53	0.00453**
Heparin cofactor 2	SERPIND1	0.0390	1.02	0.00126**	0.09	0.68663	1.11	0.00047***
Lipocalin 2	LCN2	0.0450	1.94	0.00052***	−0.97	0.02420*	0.97	0.02459*

SM; sulfur muatard, Dex; dexamethasone, FC; fold change,

p<0.05; *,

p<0.01; **,

p<0.001; ***

^#unknown

Fig. 2:

Effect of sulfur mustard toxicity and dexamethasone treatment on proteomic profile in different groups. (A) Significantly dysregulated proteins identified in the proteomic analysis and proteins modulated post SM exposure or Dex treatment (FDR<0.05; One Way ANOVA posthoc analysis) (B) Distribution of significant proteins in SM vs C, SM+Dex vs SM, and SM+Dex vs C group comparisons. (C) Dex treatment effects on proteome profile of SM exposed corneal tissue. (D) List of common proteins to the respective group comparisons of section B (n=4–5) C, control; Dex; dexamethasone; SM, sulfur mustard; # unknown

3.3. Protein ontology and pathways identification

POIs (with fold change values) were uploaded onto IPA, and dysregulated pathways were predicted using the Core Analysis. The general summary predicted the involvement of 24 POIs in diseases and associated functioning, particularly organismal injuries and abnormalities. Of the 24 proteins, 15 were associated with inflammatory and immune responses post-injury or infection, 15 had regulatory cellular functions like cell shape and cell motility, 15 played essential roles in cell survival-related signaling, and seven were associated with fibrosis post-injury, trauma, or after severe infection (Fig. 3A).

Fig. 3:

Function of differentially expressed proteins, and top signaling pathways modulated post sulfur mustard exposure and dexamethasone treatment in corneal tissue. (A) Differentially expressed proteins categorization based on functioning at cellular level and diseases involvement. Purple circle, organismal injury and abnormalities; Green circle, cell movement/migration and cell morphology; Blue circle, cell death/survival mechanisms; Orange circle, inflammation and immune response; Red circle, tissue fibrosis. (B) Top 10 signaling pathways altered in corneal tissue post SM exposure or Dex treatment. Significance determined on the basis of (p<0.05). (C) Network of pathways dysregulated post SM exposure or Dex treatment in corneal tissue. Red arrow, pathways activated/inhibited by SM exposure; Green arrow, pathways activated/inhibited after Dex treatment in SM exposure corneal tissue. Dex; dexamethasone; SM, sulfur mustard.

Pathway analysis showed that of the top ten most significant pathways, actin cytoskeleton signaling (ACS), FCɣR-mediated phagocytosis in macrophages and monocytes, RHOA and Rho family-GTPases signaling, clathrin-mediated endocytosis, integrin signaling, actin nucleation by ARP-WASP complex, regulation of actin-based motility by Rho, N-formyl-methionine-leucine-phenylalanine (fMLP) signaling in neutrophils, and RAC signaling pathways were predicted to be activated by SM and inhibited by Dex (Fig. 3B). Only RHOGDI signaling was predicted to be inhibited by SM exposure and activated by Dex treatment (Fig. 3B.).

Common proteins associated with the top pathways dysregulated by SM exposure and Dex treatment included ACTR2, ARPC2, ARPC4, CDC42, EZR, MSN, KNG1, and PON1. Several proteins were involved in multiple pathways, giving rise to an interaction network; alteration in the activity of one pathway would affect/modulate the activity of all other pathways (Fig. 3C). Further, protein interaction networks were generated for each group to predict the effect of each protein on other proteins either appearing in proteomic analysis or predicted by IPA using its web base database (Fig. S1).

3.4. Dex treatment mitigated SM-induced increase in LCN2, CDC42, and CORO1A expressions

3.4.1. Proteomic analysis

LCN2, CDC42, and CORO1A were determined to have grave importance, due to their intensity of modulation and/or their imminent role in the significant pathways that appeared in IPA analysis. SM increased LCN2, CDC42, and CORO1A expressions compared to the controls (~87-, ~5-, and 9-fold, respectively; p<0.001) (Table 1). Dex significantly reversed the LCN2 (89% reversal), CDC42 (53% reversal), and CORO1A (62% reversal) expressions to 0.11-, 0.47-, and 0.38-folds, respectively, compared to the SM group (p<0.05) (Table 1).

3.4.2. IF

IF staining for LCN2, CDC42, and CORO1A was performed to validate the proteomic profiles. Interestingly, corneal epithelial cells and mixed cell populations in the corneal stroma had higher LCN2 expression upon SM exposure as compared to the controls (Fig. 4A–D; Fig. S2A–B). In corneal epithelial cells, Dex treatment significantly reversed the SM induced increase in LCN2 expression (control: 1.22±0.13, SM: 9.33±0.92, SM+Dex: 5.3±0.78; p<0.001), corresponding to a 53% reversal (Fig. 4A–B). Dex treatment demonstrating a 94% reversal in SM induces LCN2 expression in the stroma (control: 0.05±0.0076, SM: 0.995±0.211, SM+Dex: 0.108±0.021; p<0.001) (Fig. 4C–D; Fig. S1B). CORO1A expression also increased in the cornea post-SM exposure and a ~90% reduction reversal was observed upon Dex treatment (control: 0.00414±0.00132; SM: 1.034±0.166, SM+Dex: 0.1097±0.0309; p<10⁻⁶) (Fig. 4E–F; Fig. S2C). SM exposure caused an 8.6-fold increase in CDC42 expression in the corneal epithelium compared to the controls and Dex treatment caused a 67% reversal (control: 2.09±0.34, SM: 18.1±1.18, SM+Dex: 7.43±0.75; p<10⁻¹⁰) (Fig. 5A–B; Fig. S3A). A similar pattern of CDC42 expression was observed with SM exposure and Dex treatment in the stroma as well (control: 0.18±0.048, SM: 0.70±0.085, SM+Dex: 0.090±0.026; p<10⁻⁶) (Fig. 5C–D; Fig. S3B).

Fig. 4:

SM induced increase in LCN2 and CORO1A expression and reversal post Dex treatment. (A) Bar graph of LCN2 expression assessed through immunofluorescence (IF) post SM exposure and its reversal after Dex treatment in corneal epithelium. (B) Representative IF images of LCN2 expression in corneal epithelium. (C) Bar graph of LCN2 expression assessed using IF post SM exposure and its reversal after Dex treatment in corneal stroma. (D) Representative IF images of LCN2 expression in corneal stroma. (E) Bar graph presenting the increased CORO1A protein expression assessed using IF post SM exposure and its reversal after Dex treatment in corneal stroma. (F) Representative IF images of CORO1A expression in corneal stroma. Fluorescence intensity was measure using QuPath analysis software (version-0.4.3). Statistical significance calculated using Student’s T-test. Green arrows, LCN2 expression in epithelium; White arrows, LCN2 expression in stroma; Yellow arrows, CORO1A expression in stroma; Pink arrows, DAPI stained nucleus; scale, 50 μm; ***p<0.001; n=3–5. CORO1A, coronin 1A; Dex, dexamethasone; LCN2, lipocalin 2; SM, sulfur mustard.

Fig. 5:

Elevation in CDC42 expression post SM exposure and reversal upon Dex treatment in corneal epithelium and stroma. (A) Bar graph presenting CDC42 expression as assessed using immunofluorescence (IF) post SM exposure and its reversal after Dex treatment in corneal epithelium. (B) Representative IF images of CDC42 expression in corneal epithelium. (C) Bar graph presenting CDC42 expression as assessed using IF post SM exposure and its reversal after Dex treatment in corneal stroma. (D) Representative IF images of CDC42 expression in corneal stroma. Fluorescence intensity was measured using QuPath analysis software (version-0.4.3). Statistical significance was calculated using Student’s T-test. Green arrows, CDC42 expression in epithelium; Yellow arrows, CDC42 expression in stroma; Pink arrows, DAPI stained nucleus; scale, 50 μm; ***p<0.001; n=3–5. CDC, cell division cycle 42; Dex, dexamethasone; SM, sulfur mustard.

3.5. DEX mitigates SM-induced increase in F-Actin expression in corneal epithelium and stroma

Further, as 50% of the POIs were associated with cell motility and morphology, P-IF for F-actin was performed as it is vital for these processes (Peng, Bera et al., Kozma, Ahmed et al. 1995, Chen, Hobbie et al. 1996, Jiang, Enomoto et al. 2009, Bamburg and Bernstein 2010, Galkin, Orlova et al. 2011, Zheng, Qin et al. 2023). F-actin expression and actin dynamics were altered at the cellular level. In control tissues, F-actin expression was observed in the superficial (topmost) epithelial membrane layers. SM drastically increased the F-actin expression and Dex treatment led to a ~93% reversal in the corneal epithelium (control: 22.66±1.707, SM: 104.15±12.72, SM+Dex: 28.46±4.04; p<10⁻⁶) (Fig. 6A–B; Fig. S3C). In SM exposed tissues, F-actin expression was distributed throughout the cytoplasm. In the stroma, Dex treatment almost completely reversed SM-induced increase in F-actin expression (control: 0.079±0.014, SM: 4.43±0.531, SM+Dex: 0.118±0.025; p<10⁻¹⁰) (Fig. 6C–D; Fig. S3D).

Fig. 6:

Actin polymerization post SM exposure and Dex treatment in corneal tissues. (A) Bar graph presenting F-actin expression as assessed using immunofluorescence (IF) post SM exposure and its reversal after Dex treatment in corneal epithelium. (B) Representative IF images of F-actin expression in corneal epithelium. (C) Bar graph presenting F-actin expression as assessed using IF post SM exposure and its reversal after Dex treatment in corneal stroma. (D) Representative IF images of F-actin expression in corneal stroma. Fluorescence intensity was measured using QuPath analysis software (version-0.4.3). Statistical significance was calculated using Student’s T-test. Red arrows, F-actin expression in epithelium; Yellow arrows, F-actin expression in stroma; Pink arrows, DAPI stained nucleus; Scale, 50 μm; ***p<0.001; n=3–5. Dex, dexamethasone; F-actin, filamentous actin; SM, sulfur mustard.

4. Discussion

In this study, a thorough examination of mechanisms underlying SM-induced corneal injury and Dex treatment was conducted using proteomic analysis. POIs were primarily associated with injury, cell motility, phagocytosis/endocytosis, inflammation, fibrosis, and cell proliferation/viability (Fig. 7A). Top signaling pathways regulated/affected by SM and Dex were also associated with cell migration/motility and tissue inflammation (Fig. 7B).

Fig. 7:

Interaction networks of proteins of interest (POI) and actin cytoskeletal signaling, and schematic cell migration and actin filament formation at the cellular level. (A) Network for POIs. (B) Actin cytoskeletal signaling (ACS) depicting stepwise activation of proteins that participate in actin polymerization and F-actin stabilization. Pink color shapes, POIs from proteomic analysis; Orange color shape, activated proteins and mechanisms in ACS; Blue color shape, inhibited mechanism ACS. (C) Schematic of actin polymerization at cellular level and cell migration. ACS, actin cytoskeleton signaling.

In our analysis, ACS signaling was the most crucial pathway implicated with Dex action in countering SM toxicity in corneas. Other pathways were found to either support ACS, like RHOA and Rho family GTPases signaling, actin nucleation by ARP-WASP Complex, and regulation of actin-based motility by Rho, RAC signaling, and RHOGDI signaling (Fig. 7B) (Cotteret and Chernoff 2002, Maddala, Reddy et al. 2003, DerMardirossian, Rocklin et al. 2006, Haga and Ridley 2016, Zheng, Qin et al. 2023); or these pathways had recourse to ACS, like FCγR-mediated phagocytosis in macrophages and monocytes signaling, clathrin-mediated endocytosis signaling, and fMLP signaling in neutrophils (Hackam, Rotstein et al. 1997, Joshi, Butchar et al. 2006, Phillipson and Kubes 2011, Yang, Colosi et al. 2022). ACS regulates the dynamic network of actin filaments and is vital for proper cell shape/structure, cell adhesion, cell motility, endocytosis/exocytosis, cell division, intracellular transport, and cell signaling (Chen, Hobbie et al. 1996, Melendez, Grogg et al. 2011, Mattila, Batista et al. 2016, Lechuga and Ivanov 2021, Fu, Liu et al. 2022). ACS plays a crucial role in tissue repair mechanism as it increases the cell migration and structural remodeling at the site of injury and accelerates the wound closure. Dex treatment reversed SM-induced increases in the expression of ACTR2, ARPC2, ARPC4, CDC42, EZR, KNG1, MSN, and PFN proteins; hence, opposes the activation of ACS. Putatively, early administration of Dex (2 h post-SM exposure) prevented severe damage to corneas and/or Dex administration every 8 h for 28 days boosted tissue repair mechanisms; hence, further actin cytoskeleton and tissue repair signaling are not required in Dex treated tissues. In the process of actin polymerization PFN favors the exchange of ADP with ATP on actin monomers and promotes activated actin monomer binding at the barded (+ end) end of the actin filament (Pernier, Shekhar et al. 2016, Tang and Gerlach 2017). Whereas actin depolymerization factors ADF/cofilin disassemble actin monomers at the minus (− end) end and make them available at the barbed end (Bamburg and Bernstein 2010, Galkin, Orlova et al. 2011).

Rho-GTPases are critical molecular switches that regulate ACS dynamics via CDC42, RhoA, and Rac (Wilk-Blaszczak, Singer et al. 1997), also evident from our ontology and functional analyses. CDC42 is a key regulator of ACS-dependent processes (Kozma, Ahmed et al. 1995, Wilk-Blaszczak, Singer et al. 1997, Melendez, Grogg et al. 2011, Haga and Ridley 2016, Tang and Gerlach 2017). Extracellular accumulation of KNG1 causes activation of CDC42 via binding with on G protein-coupled receptors (GPCR), B1 and B2 at the cell membrane (Hall 1997, Knox, Corbett et al. 2001, Mizuno and Itoh 2009, Melendez, Grogg et al. 2011, Haga and Ridley 2016). Upon activation, CDC42 GTPase binds with Neural Wiskott Aldrich Syndrome protein (N-WASP), disrupting an auto inhibitory domain of N-WASP that makes it available to interact with ARP2/3 complex (Fu, Liu et al. 2022). Further, N-WASP-ARP2/3 complex binds with the existing actin filament and starts the branched elongation of actin filament that increases the actin filament density at the leading edge via promoting actin nucleation (Wilk-Blaszczak, Singer et al. 1997, Cotteret and Chernoff 2002, Melendez, Grogg et al. 2011, Haga and Ridley 2016, Zheng, Qin et al. 2023). Proteins, like the Ezrin, Radixin and Moesin (ERM) bundle, bind with the F-actin to stabilize the actin filament structure (Arpin, Chirivino et al. 2011). Other proteins of this pathway are also associated with injury mechanisms. KNG1 activation upon injury and inflammation, also stimulates MMPs, (MMP-2 and MMP-9) causing extracellular matrix degradation (Erices, Corthorn et al. 2011), IL-6, and COX-2 expressions (Feng, Hsiung et al. 2010). Moreover, it increases vascular permeability and promotes vasodilation helping immune cell infiltration (Dray and Perkins 1993, Hall 1997, Maurer, Bader et al. 2011). Dex treatment caused a decrease in SM-induced MMP-9 and COX-2 elevations, (Goswami, Mishra et al. 2021) preventing structural damage and inflammation, as reported previously (Goswami, Mishra et al. 2021, Goswami, Mishra et al. 2022, Mishra, Kant et al. 2023).

Uncontrolled ACS may also cause structural damage due to breakdown of adhesion-junctions (AJ) (Lechuga and Ivanov 2021) and tight junctions (TJ) (Gehren, Rocha et al.) that are crucial for cell-to-cell adhesion, tissue integrity, and selective permeability of epithelial membrane (Gehren, Rocha et al., Lechuga and Ivanov 2021). Structural damage to the corneal epithelium and inflammation in the stroma have been well-documented (McNutt, Hamilton et al. 2012, Goswami, Mishra et al. 2021).

LCN2 was another significant protein found to be upregulated by SM exposure and downregulated by Dex. LCN2, or neutrophil gelatinase-associated lipocalin, is expressed in neutrophils, macrophages, epithelial cells, and adipocytes (Abella, Scotece et al. 2015, Guardado, Ojeda-Juarez et al. 2021). It plays a vital role in hematopoietic cell apoptosis, metabolic homeostasis, iron and fatty acid transport, inflammation, immunomodulation, and chemo-attraction of neutrophils (via a positive-feedback loop) in damaged tissue (Aigner, Maier et al. 2007, Rathore, Berard et al. 2011, Schroll, Eller et al. 2012, Abella, Scotece et al. 2015, Guardado, Ojeda-Juarez et al. 2021). Additionally, LCN2 interacts with the MMP-9 and form LCN2-MMP-9 complex, and inhibits the auto degradation of MMP-9 that eventually results in tissue degradation and structural remodeling (Jaberi, Cohen et al. 2021).

Dex treatment also abrogated SM-induced CORO1A increases in our studies. CORO1A is exclusively expressed in lymphocytes and aids lymphocyte motility, phagocytosis, and micropinocytosis (Foger, Rangell et al. 2006, Jayachandran, Sundaramurthy et al. 2007, Pieters 2008). It regulates ACS mainly in the immune cells and facilitates their infiltration in the injured tissue (Ferrari, Langen et al. 1999, Foger, Rangell et al. 2006, Jayachandran, Sundaramurthy et al. 2007). Lymphocyte cytosolic protein-1 (LCP1), is an f-actin binding protein, which participates in the migration of immune cells, especially hematopoietic cellular lineages (leukocytes) in response to chemotaxis (Morley 2012, Mahat, Garg et al. 2022). Increased ACS along with increased expression of CORO1A and LCP1 in the stroma after SM exposure indicates increased immune cells infiltration post SM-exposure that was countered by Dex treatment, which is in line with our previous findings (Goswami, Mishra et al. 2022, Mishra, Kant et al. 2023). Previous publications on SM induced late pathology corroborates the involvement of inflammations and immune cells in ocular tissue damage, either via increased mRNA expression of pro-inflammatory cytokines or via the immune cell infiltration in the late phase injuries (Horwitz, Cohen-Gihon et al., Mishra, Kant et al. 2023).

FRMD, CAP1, and SORBS are associated with cellular motility and were downregulated post Dex treatment; however, their mechanism of action is unknown. TUFM, RPN1, PEBP1, and PON1 expressions were also reversed by Dex treatment, however, their mechanism of action in SM induced corneal injury and Dex treatment needs to be explored. Thus, we observed several novel molecular mediators of Dex efficacy against corneal SM-toxicity.

Thus, SM exposure in corneas causes activation of ACS that regulates structural integrity, cell motility, extracellular matrix remodeling, breakage of AJs and TJs, and inflammation (via immune cells infiltration). Dex treatment mitigated these effects, in our ongoing and previous studies. These are salient features of epithelial-mesenchymal transition (EMT). Though Dex efficacy against SM-induced corneal injuries has been shown (Nagelhout, Gamache et al. 2005, Hoffart, Matonti et al. 2010, Bian, Pelegrino et al. 2016, Bian, Xiao et al. 2017), no studies have shown the involvement of ACS or EMT. Hence, SM toxicity may be mediated via the involvement of EMT process that is mitigated by Dex treatment, in the corneal tissue. Further studies are necessary to confirm these novel findings.

5. Conclusion

SM activates ACS and putatively EMT in corneas, which are novel findings per our knowledge. Dex confers efficacy against SM-induced corneal injury, not only via anti-inflammatory responses, but also via ACS and EMT regulation (Fig. 7A–C).

Highlights:

• SM activates actin cytoskeletal signaling in corneas

• SM induces an inflammatory response in corneas

• DEX reduces cytoskeletal signaling and inflammation in corneas

Abbreviations

AFM

Afamin

ACS

Actin cytoskeleton signaling

AJ

Adhesion-junctions

AMRT

Accurate mass and retention time

ARP2

Actin related protein 2

ARPC2

Arp2/3 complex 34 kDa subunit

ARPC4

Actin-related protein 2/3 complex subunit 4

CA

Carbonic anhydrase

CAP1

Adenylyl cyclase-associated protein

CDC42

Cell division control protein 42 homolog

CEP112

Apolipoprotein H

CFI

Complement factor I

COL5A1

Alpha-1 type V collagen (Fragment)

CORO1A

Coronin 1A

CWA

Chemical warfare agent

Dex

Dexamethasone

EMT

Epithelial-mesenchymal transition

EZR

Ezrin

FDR

False discovery rate

FEMD

FERM domain-containing protein

fMLP

N-formyl-methionine-leucine-phenylalanine

GRP78

Glucose-regulated protein (78 kDa)

HBA

Alpha-globin 1

HBA1

Globin A1; Hemoglobin subunit gamma

H1.3

Histone H1.3

IF

Immunofluorescence

IPA

Ingenuity Pathway Analysis

KNG1

Kininogen 1

LC-MS/MS

Liquid chromatography-mass spectrometry (MS)/MS

LCN2

Lipocalin 2

LCP1

Lymphocyte cytosolic protein 1

MMPs

Matrix metalloproteinases

N-WASP

Neural-Wiskott-Aldrich syndrome protein

O.C.T

Optimal cutting temperature

P

Phalloidin

PBST

Phosphate buffer saline - triton X-100

PCA

Principal component analysis

PEBP1

Phosphatidylethanolamine-binding protein 1

PFN

Profilin

POI

Proteins of interest

PON1

Paraoxonase

Q-TOF

Quadrupole time of flight

RPN1

Dolichyl-diphosphooligosaccharide--protein glycosyltransferase subunit 1

SERPIND1

Heparin cofactor 2

SM

Sulfur mustard

SORBS

SH3 domain-containing protein

THBS1

Thrombospondin 1

TJ

Tight junctions

TUFM

Elongation factor Tu; Elongation factor Tu mitochondrial