Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Feb 1.
Published in final edited form as: Exp Eye Res. 2024 Dec 18;251:110213. doi: 10.1016/j.exer.2024.110213

PI3K Signaling and Lysyl Oxidase is Critical to Corneal Stroma Fibrosis Following Mustard Gas Injury

Nishant R Sinha 1,2,4, Alexandria C Hofmann 1,2,4, Laila A Suleiman 2, Riley Laub 2, Ratnakar Tripathi 1,2, Shyam S Chaurasia 3, Rajiv R Mohan 1,2,4,*
PMCID: PMC11798705  NIHMSID: NIHMS2044667  PMID: 39706242

Abstract

Sulfur mustard gas (SM), an alkylating and vesicating agent, has been used frequently in many wars and conflicts. SM exposure to the eye results in several corneal abnormalities including scar/fibrosis formation. However, molecular mechanism for SM induced corneal fibrosis development is poorly understood. After SM insult to the eye, excessive synthesis/secretion of extracellular matrix components (ECM) including collagen (COL) I, COL III, and lysyl oxidase (LOX) by corneal myofibroblasts causes corneal fibrosis, however, precise mechanism remains elusive. This study tested the hypothesis that Phosphoinositide 3-kinase (PI3K) signaling alterations post SM in cornea enhances stromal ECM synthesis and corneal fibrosis. New Zealand White Rabbits were used. The right eyes were exposed to SM (200mg-min/m3) and left eye to the air for 8min at MRI Global. Rabbit corneas were collected on day-3, day-7, and day-14 for molecular analysis. SM exposure caused a significant increase in mRNA expression of PI3K, AKT, COL I, COL III, and LOX and protein levels of LOX in a time-dependent manner in rabbit corneas. The in vitro studies were performed with human corneal stromal fibroblasts (hCSFs) by growing cultures in −/+ nitrogen mustard (NM) and LY294002, a PI3K specific inhibitor, for 30min, 8h, 24h, 48h, and 72h. NM significantly increased mRNA and protein levels of PI3K, AKT, COL I, COL III, and LOX. On the contrary, LY294002 in NM hCSFs significantly reduced PI3K, AKT, COL I, COL III, and LOX protein expression. We concluded that PI3K signaling mediates stromal collagen synthesis and LOX production following SM injury.

Keywords: Sulfur mustard, mustard gas keratopathy, cornea, myofibroblast, LOX, collagens

1. Introduction

Sulfur mustard gas (SM) is highly toxic, volatile, lipophilic, and blister-inducing alkylating chemical. SM exposure to the eye causes severe multiple corneal injuries including fibrosis and irreversible blindness referred as mustard gas keratopathy (Balali-Mood and Hefazi, 2006; Fuchs et al., 2021). In healthy eye, the cornea is transparent and avascular, but SM exposure disables corneal clarity and function (Sinha et al., 2023, 2022; Tripathi et al., 2020). The precise molecular pathways mediating SM induced corneal dysfunction are largely unknown (Fuchs et al., 2021; McNutt and Mohan, 2020; Mishra and Agarwal, 2022). The cornea is mainly comprised of three layers, epithelium, stroma and endothelium. The stroma, constituting 90% of the cornea, plays a major role in maintaining corneal transparency and homogenous light refraction It is contains keratocytes, lamellae of collagen bundles, extracellular matrix, and proteoglycans (Kumar et al., 2023; Mohan et al., 2022; Sinha et al., 2021).

Stromal collagen fibrils unique organization and arrangement is essential in maintaining corneal shape, size, and homogenous light refraction (Meek, 2009). The stroma predominantly contains interweaved bundles of collagen type I (COL 1) (Meek and Knupp, 2015). Conversely, collagen type III (COL III) is a repair collagen only found at abundant levels post-injury (Saika et al., 1996). Current literature on the molecular mechanism activated to preserve stromal collagen after mustard gas exposure is limited. This study is a continuation of our previous work that showed SM exposure to the cornea compromises the unique collagen architecture in the stroma. Previously, we showed SM exposure caused a decrease in collagen fibril size and increased maximum and minimum inter-fibril distance (IFD) (Sinha et al., 2022).

Lysyl oxidase (LOX) is a secreted copper-dependent amine oxidase expressed in various cell types, including keratinocytes, fibroblasts, adipocytes, osteoblasts, smooth muscle cells, and endothelial cells (Harris, 1976; Shetty et al., 2015; Wang et al., 2017). The most well-known function of LOX is the initiation of collagen synthesis and elastin cross-linking (Siegel et al., 1970; Wang et al., 2017). Currently, the study of LOX in the cornea is at a very juvenile stage due to the limited ability to culture corneal fibroblasts with an extracellular matrix (ECM) microenvironment (Chen and Raghunath, 2009; Priyadarsini et al., 2020; Shetty et al., 2015). However, LOX has been studied in patient samples with keratoconus (KC), a corneal thinning disease. Studies have shown that increased LOX mRNA and protein levels correlate with the progression of KC (Dudakova and Jirsova, 2013; Pahuja et al., 2016; Shetty et al., 2015).

Phosphoinositide 3-kinase (PI3K) is a known regulatory protein for LOX and COL III gene transcription through transforming growth factor beta (TGFβ1) in non-ocular cells/tissues. TGFβ1 activates the phosphoinositide 3-kinase (PI3K) signaling cascade to transcribe LOX mRNA (Atsawasuwan et al., 2008; Taylor et al., 2011; Voloshenyuk et al., 2011). This study determined a new mechanism of action for the regulation of extracellular matrix protein expression following mustard gas exposure.

2. Materials and Methods

2.1. Animal usage and care

The Institutional Animal Care and Use Committees of the University of Missouri, Columbia, MO, and the MRI Global, Kansas City, MO have approved the study. Rabbits were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmology and Vision Research. Fifty-four healthy male New Zealand White rabbits (2.5–3.0kg) with no clinical ocular symptoms purchased from the Charles Rivers Laboratories were used. Rabbit eyes were divided into three groups: Group 1 (naïve) eyes had no treatment, Group 2 (vehicle) eyes received air vapor on the eye only, and Group 3 (SM) eyes received SM vapor at a target concentration of 200 mg-min/m3 for 8 minutes on the eye (Sinha et al., 2022; Tripathi et al., 2020). Six rabbits were used per group for each timepoint. At endpoints, rabbits were humanely euthanized, and corneal tissues were collected and used for preparing serial tissue sections for immunofluorescence, protein lysates for western blotting, and cDNA for qRT-PCR studies following reported methods (Sinha et al., 2023; Tripathi et al., 2020).

2.2. Cell Culture

Primary human corneal stromal fibroblasts (hCSFs) were generated in the lab from human donor corneas purchased from the Saving Sight Foundation using established protocols (Balne et al., 2020; Lim et al., 2016; Mohan et al., 2016; Sharma et al., 2015). In brief, epithelium and endothelium were removed from donor human corneas using gentle scraping. Afterwards, corneal stroma was cut into eight partitions (buttons) and plated on a 60 mm tissue culture dish. Corneal buttons were given 10% serum media till fibroblasts sprouted and plate became 70% confluent. Upon reaching confluency, fibroblasts were passaged as needed (Sinha et al., 2023). All in vitro experiments were done in six tissue culture well plates using two wells for each treatment. Each experiment was repeated three times, thus, n=6 was used to ensure high scientific rigor and statistical significance. First, hCSF cells were cultured with or without NM (100 ng/mL) in serum free media for the entire duration of the experiment (Tripathi et al., 2020). Secondly, hCSF cells were treated with or without 50 μM LY294002, a PI3K specific inhibitor, then exposed to 100ng/mL NM in media. Vehicle treatment (VT) group made of equal volume of DMSO used to prepare PI3K inhibitor media was included. VT were collected at the final timepoint studied.

2.3. Real-Time RT-PCR

Corneal tissues were minced in a Tissuelyser LT (Cat # 85600, Qiagen, Valencia, CA, USA) in RLT buffer (cat # 79216, Qiagen). hCSFs were collected in RLT buffer. Total RNA was isolated using the RNeasy kit (Cat # 73404, Qiagen) following the manufacturer’s instructions as we previously published (Tripathi et al., 2020). Afterwards, mRNA was converted to cDNA using GoScript™ Reverse Transcription System (Cat # A50001, Promega, Madison WI, USA) according to the manufacture’s protocol (Gupta et al., 2017). The relative mRNA expression was calculated using the 2-ΔΔCt method and reported as a relative fold change over the corresponding control values. qRT-PCR was performed in triplicate for each sample, and a minimum of three independent experiments were conducted.

2.4. Regular PCR

Untreated hCSFs and HeLa cells were collected in RLT buffer. After mRNA extraction and cDNA conversion was performed using GenScript Taq DNA Polymerase kit (Cat # E00007, GenScript, Piscataway, NJ, USA) followed by PCR in BT100 Thermal cycler (Bio Rad, Hercules, CA, USA) according to the manufacture’s protocol. The PCR products were separated on a 2% agarose gel, stained with ethidium bromide, and imaged with iBright (Cat # A44114, Thermo Fisher, Waltham, MA, USA) as described previously (Balne et al., 2023).

2.5. Western Blotting

Cell lysates were collected in RIPA buffer and protein lysate concentrations were quantified using the Bradford quantification method (Cat # 500006, Bio Rad). Western blots were run using previously published protocols (Sinha et al., 2023). In brief, 50ng of proteins were loaded into a precast 4–12% SDS page gel and protein samples were separated via electrophoresis. Afterwards, the gel was transferred to a nitrocellulose membrane and stained with primary and secondary antibody. Primary antibody was diluted in (5% or 2.5%) FFM or (5% or 2.5%) BSA and Secondary antibody was diluted in (5% or 2.5%) FFM or (5% or 2.5%) BSA. The membranes were developed using 1 mL of a solution containing 1:1 luminol to peroxide buffer SuperSignal™ West Dura Extended Duration Substrate (Cat # 34075, Thermo). The following antibodies were used: ɑSMA (Cat # M0851, DAKO), PI3K (Cat # 6G10, Cell Signaling Technology, Danvers, MA, USA), and AKT (Cat # 40D4, Cell Signaling). The membrane was then analyzed using a C-Digit digital Western Blot analyzer (Li-COR Inc, Lincoln, NE, USA).

2.6. ELISA

Cell lysates were collected for protein extraction in RIPA buffer followed by sonification (3× 10s at 50hz with 30s gaps). Protein lysate concentrations were quantified following Bicinchoninic acid (BCA) assay quantification method (Cat # 23227, Invitrogen). ELISA kits for COL I (Cat # ab210966, Abcam), COL III (Cat # NBP2–75855 Novus Biologicals, Centennial CO, USA) and LOX (Cat # EHOLR1, Thermo) were performed following vendors protocols. In brief, 100μL of protein lysates were added to pre-coated ELISA plates and left overnight at 4C. Afterwards, plates were washed using wash buffer and incubated in detection antibody. Then plates were washed and streptavidin-HRP was added, followed by a wash and developed using substate solution till color change. Upon color change, stop buffer was added and plates were read using Bio Tek Synergen H1 (Agilent, Santa Clara, CA, USA).

2.7. Immunofluorescence

Frozen eight-micron thick corneal sections were placed at room temperature for 10 min, blocked with 5% donkey serum for 1 hour at ambient temperature, incubated at 1:50 dilution in anti-LOX primary antibody (Cat # Sc-373995, Santa Cruz Biotechnology, Dallas, TX) overnight at 4°C, and 4-hour in goat anti-mouse Alexa Fluro 488 secondary antibody at 1:500 dilution (Cat # A11001, Thermo). Tissue sections were mounted using DAPI antifade medium (Cat # H1200, Vector Laboratories, Newark, CA, USA) and premier coverslips (Thermo). 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). ImageJ with FiJi addon was used to quantify immunofluorescence intensity.

2.8. Statistical Analysis

Each experiment was conducted independently, each sample was tested in triplicate, and the values were expressed as mean ± SEM. For statistical analysis, one-way or two-way analysis of variance (ANOVA) with Bonferroni post hoc test was used. The value of p<0.05 was considered as the level of significance. GraphPad Prism 10.1 (GraphPad Software, La Jolla, CA, USA) software was used for statistical analysis.

3. Results

3.1. Time-dependent alterations in LOX and ECM post SM/NM

3.1.1. Time-dependent changes in LOX protein in rabbit cornea post SM in vivo

A time-dependent qualitative and quantitative alterations in the expression of LOX protein were evaluated in rabbit corneal sections using immunofluorescence and ELISA, respectively. As evident from Figure 1A, SM ocular exposure resulted in a significant increase in LOX expression in corneal epithelium at day-3 (p<0.05), day-7 (p<0.0001), and day-14 (p<0.0001) and in the stroma at day-3 (p<0.0001), day-7 (p<0.0001), and day-14 (p<0.0001) compared to the corresponding negative controls (air-exposed or naïve corneas). Figure 1B shows the digital quantification of qualitative LOX protein in the epithelium which was significantly high at all tested time points compared to negative controls. The comparison of LOX expression within the SM-exposed corneas found highest LOX levels at day-14 followed by day-7 and day-3 (day-7 vs day-3 (p<0.0001), day-14 vs day-7 (p<0.0001), and day-14 vs day-3 (p<0.0001)). Figure 1C shows the digital quantification of qualitative LOX expression in corneal stroma which was significantly increased at day-3 (p<0.001), day-7 (p<0.0001), and day-14 (p<0.0001) compared to corresponding negative controls. A time-dependent significant increase in LOX levels were noted (day-7 vs day-3 (p<0.0001), day-14 vs day-7 (p<0.0001) and day-14 vs day-3 (p<0.0001).

Figure 1: Expression of LOX protein in rabbit corneas −/+ SM in vivo.

Figure 1:

Open in a new tab

Immunofluorescence was used to detect LOX expression in rabbit corneas at 3, 7, and 14 days after SM exposure (A). LOX expression was qualifiedly measured using imageJ to determine signal intensity. LOX significantly increased in a time-dependent manner in the epithelium (B) and stroma (C) of rabbit corneas exposed to SM compared to corresponding vehicle exposure group. At day-3, day-7, and day-14 LOX expression was significantly increased in the corneal epithelium and corneal stroma compared to the naïve group. At day-14 LOX expression was increased compared to day-7 and day-3 in the corneal stroma and epithelium. (n=3; scale bar = 100 μm)

Figure 2 presents a time-dependent quantitative LOX expression measured with ELISA in whole corneal lysates. The LOX levels were significantly increased at all three tested time points compared to the corresponding negative controls (p<0.0001). The comparisons of LOX expression within SM-exposed cornea were also significant for day-7 vs day-3 (p<0.05) and day-14 vs day-7 (p<0.0001).

Figure 2: Total LOX protein levels in rabbit corneas −/+ SM in vivo.

Figure 2:

Open in a new tab

ELISA was used to detect total LOX in a rabbit cornea 3, 7, and 14 days after SM exposure. This data confirmed the increase in LOX expression detected in immunofluorescence. A significant increase in LOX was detected at day-7 (p<0.01) and day-14 (p<0.0001) with a trending increase at day-3 (*=p<0.05, **=p<0.01, ****=p<0.0001, n=3).

3.1.2. Time-dependent expression of PI3K/AKT and ECM in rabbit cornea post SM in vivo

A time-dependent mRNA expression of PI3K, AKT, and ECM components (LOX, COL I, and COL III) in SM-exposed and control corneas was analyzed with qRT-PCR. SM exposure caused a significant increase in PI3K (Fig. 3A) at day-7 (p<0.01) and day-14 (p<0.001) and AKT (Fig. 3B) at day-7 (p<0.01) and day-14 (p<0.0001) compared to VT (air-exposed rabbit corneas). SM-exposed corneas had a time-dependent increase in LOX (Fig. 3C) at day-3, day-7 (p<0.05) and day-14 (p<0.01) compared to VT. SM exposed corneas had significantly increased COL I (Fig. 3D) at day-3 (p<0.01), day-7 (p<0.0001), and day-14 (p<0.0001) levels compared to VT. Also, the COL III (Fig. 3E) was significantly increased at day-7 (p<0.001) and day-14 (p<0.0001) in SM exposed corneas compared to VT corneas. None of the SM-exposed corneas of day-3 showed significant differences from VT except the COL I (p<0.01).

Figure 3: mRNA expression of PI3K/AKT and ECM components in rabbit corneas −/+ SM in vivo.

Figure 3:

Open in a new tab

mRNA expression of PI3K, AKT, and ECM components (LOX, COL I, and COL III) of rabbit corneas exposed to SM was analyzed using qRT-PCR at day-3, day-7, and day-14. SM exposure caused a significant increase in PI3K (A) at day-7 (p<0.01) and day-14 (p<0.001) compared to VT. AKT (B) significantly increased at day-7 (p<0.01) and day-14 (p<0.0001) compared to VT. SM exposure caused a trending increase in LOX (C) at day-3 and a significant increase at day-7 (p<0.05) and day-14 (p<0.01) compared to VT. COL I (D) significantly increased at day-3 (p<0.01), day-7 (p<0.0001), and day-14 (p<0.0001) in SM exposed corneas compared to VT. COL III (E) significantly increased at day-7 (p<0.001) and day-14 (p<0.0001) in SM exposed corneas compared to VT corneas with a trending increase at day-3 (*=p<0.05, **=p<0.01, ****=p<0.0001, n=3).

3.1.3. LOX and LOX-Like expression in human corneal fibroblasts in vitro

The expression of LOX and LOX-Like (LOXL) proteins 1–4 was detected with PCR in human corneal stromal fibroblasts. Figure 4A shows LOX, LOXL1, LOXL2, LOXL3, and LOXL4 expression in healthy hCSF. Figure 4B shows the expression of LOX and LOXL1–4 in HeLa cells, which was essentially used as positive control.

Figure 4: Characterization of LOX and LOX-like (Loxl) mRNA expression in hCSFs.

Figure 4:

Open in a new tab

Reverse transcriptase PCR showed human corneal stromal fibroblasts expressed LOX, LOX1, LOX2, LOX3, and LOX4. β-actin and HeLa cells were used as positive controls for primers and cells expressing LOX, respectively. (n=3)

3.1.4. Time-dependent mRNA expression of ECM genes in hCSF treated with NM in vitro

qRT-PCR detected a significantly increased mRNA expression of LOX, COL I, and COL III at 24h (p<0.001), 48h (p<0.0001), and 72h (p<0.01) in NM-exposed hCSFs compared to VT (Fig. 5A–C). Interestingly, the mRNA expression of LOX in NM-exposed hCSFs at 72h was lower than the 48h timepoint (Fig. 5A). The NM treatment to hCSF significantly increased mRNA of COL I at 24h (p<0.0001), 48h (p<0.001), and 72h (p<0.0001 compared to VT (Fig. 5B). Also, a significant increase in COL III mRNA expression at 48h (p<0.001), and 72h (p<0.0001) observed NM-treated cells compared to the VT group (Fig. 5C) except for 24h (p=ns).

Figure 5: Time-dependent expression of LOX, COL I, and COL III mRNAs in hCSF −/+ NM in vitro.

Figure 5:

Open in a new tab

mRNA levels of LOX (A) were significantly increased at 24h (p<0.001), 48h (p<0.0001), and 72h (p<0.01) compared to VT. However, the mRNA levels were lower at 72h compared to 48h. COL I (B) had a significant increase in mRNA at 24h (p<0.0001), 48h (p<0.001), and 72h (p<0.0001) compared to VT. COL III (C) had a significant increase in mRNA at 48h (p<0.001), and 72h (p<0.0001) compared to VT. Representative VT shown collected at 72h (**=p<0.01, ***=p<0.001, ****=p<0.0001;)

3.1.4. Time-dependent expression of ECM proteins in hCSF treated with NM in vitro

Protein expression of LOX, COL I, and COL III were measured using ELISA. NM treatment to hCSFs significantly increased LOX protein levels at 8h (p<0.001), 24h (p<0.001), 48h (p<0.0001), and 72h (p<0.0001) compared to VT (Fig. 6A). The COL I protein expression at 8h (p<0.01), 24h (p<0.0001), 48h (p<0.0001), and 72h (p<0.0001) in NM treated hCSF was also significantly high compared to VT (Fig. 6B). Similarly, expression of COL III protein at 24h (p<0.0001), 48h (p<0.0001), and 72h (p<0.0001) in NM treated hCSF was significantly higher compared to VT (Fig 6C). No significant increase was seen at 30 min for LOX, COL I, and COL III. Unlike others, no significant increase was seen for COL III at 8h.

Figure 6: Time-dependent expression of LOX, COL I, and COL III proteins in hCSF −/+ NM in vitro.

Figure 6:

Open in a new tab

Protein expression of LOX (A) had a significant increase at 8h (p<0.05), 24h (p<0.05), 48h (p<0.01), and 72h (p<0.01) compared to VT. Similarly, COL I (B) had a significant increase in protein expression at 8h (p<0.05), 24h (p<0.05), 48h (p<0.01), and 72h (p<0.01) compared to VT. On the other hand, COL III (C) protein expression was delayed and showed a significant increase starting at 24h (p<0.01), 48h (p<0.0001), and 72h (p<0.0001) compared VT. Representative VT shown collected at 72h (**=p<0.01, ***=p<0.001, ****=p<0.0001; timepoint compared to VT)

3.2. NM induced ECM modulation involves PI3K/AKT signaling in vitro

The mRNA and protein levels in NM-exposed and VT cultures were measured by qRT-PCR (Fig 7) and western blotting (Fig 8), respectively, to study involvement of PI3K/AKT signaling. The qRT-PCR analysis demonstrated significantly increased mRNA levels of αSMA, PI3K, and AKT in NM-exposed hCSF cultures at 24h, 48h, and 72h (Fig. 7). The αSMA mRNA expression was significantly high at 24h (p<0.01), 48h (p<0.0001), and 72h (p<0.0001) in NM treated hCSFs compared to the VT. Similarly, the expression of PI3K and AKT transcripts was significantly high in NM treated hCSFs at 24h (p<0.0001), 48h (p<0.0001), and 72h (p<0.01) compared to the VT.

Figure 7: NM induced LOX, ɑSMA, PI3K, and AKT mRNA signaling expression in hCSF −/+ NM in vitro.

Figure 7:

Open in a new tab

hCSF cells exposed to NM for 24h, 48h, and 72h had an increase in mRNA and protein expression of αSMA, LOX, PI3K, and AKT in a time dependent manner. mRNA levels of αSMA had a significant increase at 24h (p<0.01), 48h (p<0.0001), and 72h (p<0.0001) compared to the NT. mRNA levels of PI3K had a significant increase at 24h (p<0.0001), 48h (p<0.0001), and 72h (p<0.01). However, the 72h timepoint had less mRNA expression compared to the 48h timepoint. Similarly, mRNA levels of AKT had a significant increase at 24h (p<0.0001), 48h (p<0.0001), and 72h (p<0.01) although less significant at 72h compared to the 48h timepoint.

Figure 8: NM induced PI3K and AKT protein expression in hCSF −/+ NM in vitro.

Figure 8:

Open in a new tab

PI3K and AKT proteins were detected using western blot (representative blots in A). Densitometry analysis of westerns is in B. hCSF treated with NM had a significant increase in protein levels of PI3K at 8h (p<0.05), 24h (p<0.0001), 48h (p<0.0001), and 72h (p<0.0001) compared to VT. AKT protein levels were significantly increased at 8h (p<0.0001), 24h (p<0.0001), 48h (p<0.0001), and 72h (p<0.0001) compared to VT (B). VT were collected at 72h. (*=p<0.05, **=p<0.01, ***=p<0.001, ****=p<0.0001; timepoint compared to VT)

Western blot measured levels PI3K and AKT proteins in hCSF in the presence and absence of NM (Fig. 8A) and densitometry was performed to quantify expression levels these proteins (Fig. 8BC). NM treatment to hCSFs significantly increased PI3K protein levels at 8h (p<0.05), 24h (p<0.0001), 48h (p<0.0001), and 72h (p<0.0001) compared to VT (Fig. 8B). Likewise, NM-treated hCSFs showed significantly increased AKT protein levels at 8h (p<0.0001), 24h (p<0.0001), 48h (p<0.0001), and 72h (p<0.0001) compared to VT (Fig. 8C). No significant increase was seen at 30 min for PI3K and AKT and at 8h for PI3K.

3.3. Modulation of NM-induced ECM proteins via PI3K signaling in hCSF in vitro

To study modulation of NM-induced ECM proteins via PI3K signaling in hCSF, LY294002, a PI3K specific inhibitor, was used. The pretreatment of LY294002 to hCSF to NM showed a time-dependent decrease in the expression of PI3K and AKT (Fig. 9) and ECM (LOX, COL I and COL III) proteins (Fig. 10). The effects of PI3K signaling inhibition on the ECM protein levels were calculated in percent change. The hCSF+LY294002 significantly reduced PI3K (FIG. 9B) levels at 8h (32%±2.4%; p<0.05), 24h (38%±3%; p<0.0001), 48h (34%±0.48; p<0.0001), and 72h (30%±2%; p<0.0001) hCSFs without LY294002 treated with NM. Likewise, LY294002 in NM-exposed hCSF significantly reduced AKT (Fig. 9C) protein expression at 8h (37%±4.8%; p<0.0001), 24h (33%±2%; p<0.0001), 48h (32%±2%; p<0.0001), and 72h (31%±3%; p<0.0001) compared to NM-exposed hCSFs without LY294002. No significant increase was seen at 30 min for PI3K and AKT.

Figure 9: Time-dependent change in PI3K and AKT protein levels in hCSF in −/+ NM and LY294002 in vitro.

Figure 9:

Open in a new tab

PI3K and AKT proteins were detected using western blot (representative blots in A) and densitometry analysis of westerns is in B. hCSF only (red) and hCSF+LY294002, a PI3K specific inhibitor, (blue) were exposed to NM in a time-dependent manner. hCSF+LY294002 demonstrated significantly reduced PI3K expression at 8h (p<0.05), 24h (p<0.0001), 48h (p<0.0001), and 72h (p<0.0001) compared to NM only group. The PI3K specific inhibitor significantly reduced AKT expression at 8h (p<0.0001), 24h (p<0.0001), 48h (p<0.0001), and 72h (p<0.0001) compared to the NM only group. (*=p<0.05, ****=p<0.0001; timepoint compared to VT)

Figure 10: Time-dependent change in extracellular matrix protein expression in −/+ NM and LY294002 in hCSF in vitro.

Figure 10:

Open in a new tab

COL I, COL III and LOX proteins were detected using ELISA. hCSF only (red) and hCSF+LY294002, a PI3K specific inhibitor, (blue) were exposed to NM in a time-dependent manner. hCSF+LY294002 treated with NM had a significant reduction in LOX protein (A) expression at 8h (p<0.0001), 24h (p<0.0001), 48h (p<0.0001), and 72h (p<0.0001) compared to NM only. hCSF+LY294002 treated with NM had a significant decrease in COL I protein (B) at 8h (p<0.0001), 24h (p<0.0001), 48h (p<0.0001), 72h (p<0.0001) compared to NM only group. hCSF+LY294002 treated with NM had a significant decrease in COL III protein (C) expression at 24h (p<0.0001), 48h (p<0.0001), and 72h (p<0.0001) compared to NM only group. VC were collected at 24h. (# = p<0.0001 NM + inhibitor compared to NM only) (* = p<0.05, ** = p<0.01, *** = p<0.001, **** = p<0.0001; compared to VT). VT were collected at 72h. (*=p<0.05, **=p<0.01, ***=p<0.001, ****=p<0.0001; timepoint compared to VT)

The levels of LOX, COL I, and COL III proteins were measured using ELISA in hCSF treated with NM with or without LY294002 pretreatment (Fig. 10). A significant reduced LOX levels (Fig. 10A) at 8h (33%±1.4%; p<0.0001), 24h (33%±3.4%; p<0.0001), 48h (32%±1.8%; p<0.0001), and 72h (37%±2.3%; p<0.0001) were detected in hCSF+LY294002 treated with NM than NM-treated hCSFs without LY294002. Likewise, NM-treated hCSF+LY294002 exhibited a significantly decrease in COL I protein (Fig. 10B) levels at 8h (16%±0.18%; p<0.0001), 24h (12%±0.12%; p<0.0001), 48h (13%±0.21%; p<0.0001), and 72h (14%±0.21%; p<0.0001) compared to the NM-treated hCSFs without LY294002 (Fig. 10C). A similar inhibitory pattern for COL III protein was observed at 24h (16%±5%; p<0.0001), 48h (42%±4.1%; p<0.0001), and 72h (73%±3%; p<0.0001) in NM-treated hCSF+LY294002 when compared to NM-treated hCSFs without LY294002. No significant increase was seen at 30 min for total LOX, COL I and COL III. The COL III expression was also non-significant at 8h.

4. Discussion

Sulfur mustard is a potent alkylating agent that causes corneal haze and vision loss on contact with the eye (Sinha et al., 2022). Currently, the molecular mechanisms of SM induced vision loss are unknown. A distinct intricate collagen fibril framework in stroma is required for the corneal transparency and sharp vision (Meek and Knupp, 2015). We have previously reported that SM causes an increase in collagen type III, distortion in inter-fibril distance, and decrease in collagen lamellae size (Sinha et al., 2022). However, the role of lysyl oxidase (LOX) in SM induced injury was not studied. LOX is a copper-dependent enzyme and is known to play an important role in the maintenance of corneal ECM and its thickness/shape. This study tested a hypothesis that PI3K/AKT and LOX function are involved in alterations of stromal ECM levels and production in cornea.

Irregularity in stromal collagen architecture leads to increased light scattering, which causes blurry vision because light cannot be focused on the retina (Meek, 2009). Keratoconus is a well-studied corneal thinning disease that decreases stromal ECM, causing vision loss (Zhou et al., 2017). Conversely, recent studies from our and other labs have investigated the role of collagens on stromal transparency using normal and transgenic injured/uninjured mouse corneas and diabetic/non-diabetic porcine corneas (Mutoji et al., 2021; Sinha et al., 2021). These studies show that increased total collagen, changes in collagen size, and variations in collagen distribution led to a loss of transparency and a change in the wound healing response.

The primary function of LOX is the initiation of collagen and elastin crosslinking (Harris, 1976; Siegel et al., 1970; Wang et al., 2014, 2017). LOX functionality is essential to normal corneal function, as seen by Dudakova et al. 2015 (Dudakova and Jirsova, 2013). Therefore, upstream regulators of LOX were used as anti-fibrotic targets in the cornea. This will most likely give more promising results in vivo as severe downregulation of LOX is associated with KC (D’Souza et al., 2021). Low LOX expression is known to cause corneal thinning and deviations in light scattering resulting in astigmatism or blurry vision as seen in patients (McKay et al., 2019).

In addition to corneal stroma pathologies, LOX has been studied in epithelial recovery. A recent study in pulmonary airway disease has shown LOXL2 attenuates epithelial proliferation and remodeling by activating the AKT pathway (Zeng et al., 2024). Another study has shown upregulation of LOX in the gut enhances BMP2 to strengthen the epithelial barrier. (Sasaki et al., 2024). While at low levels LOX promotes epithelial healing in non-ocular tissue, the overproduction of LOX has shown to promote fibrosis at the epithelial lumen barrier. Rabbit corneas have shown increased LOX expression in the corneal epithelium following SM injury. This may contribute to irregular and delayed remodeling of the epithelial stromal barrier (Sinha et al., 2022).

Immunofluorescence examination of LOX in rabbit eyes exposed to SM vapor in vivo showed a significant time-dependent increase in LOX protein expression. Furthermore, the increase in LOX can be seen in the epithelial layer and the stromal layers of the cornea. Thus, further mechanistic analysis of LOX regulation in the stroma was conducted in hCSF in vitro. hCSF exposed to nitrogen mustard (NM), an analogous agent to SM, showed a significant time-dependent increase in LOX, COL I, and COL III mRNA and protein. However, the increase in LOX mRNA expression at 72h is less than 48h. The decrease in LOX expression may be the result of the mRNA primer being specific for steady-state LOX mRNA and the primer avoids detection of LOX-L isotypes. Therefore, LOX mRNA at 72h may have undergone further processing. Further mRNA analysis showed PI3K and AKT followed the same significant time-dependent expression as LOX. The decrease in PI3K and AKT may signify a connection to LOX production. However, the possibility of increased repressor activity or mRNA degradation occurring between 48h and 72h may be present. Thus, protein levels of PI3K and AKT were assessed in a time-dependent manner with inhibitors.

PI3K activity was analyzed by testing AKT protein expression in a time-dependent manner. AKT expression was significantly increased from 8h to 72h and PI3K followed the same time-dependent expression observed in AKT. Pretreatment with a PI3K specific inhibitor, LY294002, was used to validate PI3K and AKT signaling as a possible mechanism for LOX activity following mustard gas injury. LY294002 showed the inhibition of PI3K had a direct effect on LOX, COL I, and COL III at later timepoints.

A limitation of this study is use of chemical inhibitor LY294002, which strongly inhibited PI3K, however it was not as effective in AKT inhibition. This may be because AKT is targeted by other pathways like mTORC or autophagy (Deng et al., 2022; Kempuraj and Mohan, 2022). This may be why significant reduction of LOX is only prevalent at the late time point. A future study using siRNA for PI3K or specifically inhibiting AKT would strength the results of the current study.

5. Conclusion

In conclusion, the result of this study supports a notion that PI3K is a potential mechanism for regulating LOX protein expression in the cornea after mustard gas injury. Further studies are essential to gain mechanistic understanding of PI3K and LOX functionality on stromal collagen architecture in vivo.

Highlights:

  1. LOX, COL I, and COL III are regulated in cornea post alkylating injury

  2. LOX and all LOX like isotypes are expressed in corneal stromal fibroblasts

  3. PI3K influences stromal ECM in cornea post mustard gas ocular exposure

Acknowledgements

This work was primarily supported by the NEI/NIH U01EY031650 grant (RRM) from the National Eye Institute, NIH, Bethesda, Maryland, USA, and Ruth M. Kraeuchi Missouri Endowed Chair Fund (R.R.M) of the University of Missouri, Columbia, Missouri, USA. A partial support came from the RO1EY0343319 and R01EY030774 grants (R.R.M.) from the National Eye Institute, NIH, Bethesda, Maryland, USA; and Merit 1I01BX000357 and RCS IK6BX005646 awards (R.R.M.) from the United States Department of Veterans Affairs BLR&D, Washington DC, USA.

Footnotes

Disclaimer

The contents do not represent the views of the U.S. Department of Veterans Affairs or the United States Government. Authors have no conflict of interest

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Atsawasuwan P, Mochida Y, Katafuchi M, Kaku M, Fong KSK, Csiszar K, Yamauchi M, 2008. Lysyl oxidase binds transforming growth factor-beta and regulates its signaling via amine oxidase activity. J Biol Chem 283, 34229–34240. 10.1074/jbc.M803142200 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Balali-Mood M, Hefazi M, 2006. Comparison of early and late toxic effects of sulfur mustard in Iranian veterans. Basic Clin Pharmacol Toxicol 99, 273–282. 10.1111/j.1742-7843.2006.pto_429.x [DOI] [PubMed] [Google Scholar]
  3. Balne PK, Gupta S, Landon KM, Sinha NR, Hofmann AC, Hauser N, Sinha PR, Huang H, Kempuraj D, Mohan RR, 2023. Characterization of C-X-C chemokine receptor type 5 in the cornea and role in the inflammatory response after corneal injury. Exp Eye Res 226. 10.1016/j.exer.2022.109312 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Balne PK, Sinha NR, Hofmann AC, Martin LM, Mohan RR, 2020. Characterization of hydrogen sulfide toxicity to human corneal stromal fibroblasts. Ann N Y Acad Sci 1480, 207–218. 10.1111/nyas.14498 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chen CZC, Raghunath M, 2009. Focus on collagen: In vitro systems to study fibrogenesis and antifibrosis _ state of the art. Fibrogenesis Tissue Repair 2, 1–10. 10.1186/1755-1536-2-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Deng CF, Zhu N, Zhao TJ, Li HF, Gu J, Liao DF, Qin L, 2022. Involvement of LDL and ox-LDL in Cancer Development and Its Therapeutical Potential. Front Oncol 12, 1–17. 10.3389/fonc.2022.803473 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. D’Souza S, Nair AP, Sahu GR, Vaidya T, Shetty R, Khamar P, Mullick R, Gupta S, Dickman MM, Nuijts RMMA, Mohan RR, Ghosh A, Sethu S, 2021. Keratoconus patients exhibit a distinct ocular surface immune cell and inflammatory profile. Sci Rep 11, 1–16. 10.1038/s41598-021-99805-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dudakova L, Jirsova K, 2013. The impairment of lysyl oxidase in keratoconus and in keratoconus- associated disorders. J Neural Transm 120, 977–982. 10.1007/s00702-013-0993-1 [DOI] [PubMed] [Google Scholar]
  9. Fuchs A, Giuliano EA, Sinha NR, Mohan RR, 2021. Ocular toxicity of mustard gas: A concise review. Toxicol Lett 343, 21–27. 10.1016/j.toxlet.2021.02.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Gupta S, Rodier JT, Sharma A, Giuliano EA, Sinha PR, Hesemann NP, Ghosh A, Mohan RR, 2017. Targeted AAV5-Smad7 gene therapy inhibits corneal scarring in vivo. PLoS One 12, e0172928. 10.1371/journal.pone.0172928 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Harris ED, 1976. Copper-induced activation of aortic lysyl oxidase in vivo. Proc Natl Acad Sci U S A 73, 371–374. 10.1073/pnas.73.2.371 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Kempuraj D, Mohan RR, 2022. Autophagy in Extracellular Matrix and Wound Healing Modulation in the Cornea. Biomedicines 10. 10.3390/BIOMEDICINES10020339 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Kumar R, Sinha NR, Mohan RR, 2023. Corneal gene therapy: Structural and mechanistic understanding. Ocular Surface 29, 279–297. 10.1016/j.jtos.2023.05.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Lim RR, Tan A, Liu Y-CC, Barathi VA, Mohan RR, Mehta JS, Chaurasia SS, 2016. ITF2357 transactivates Id3 and regulate TGFβ/BMP7 signaling pathways to attenuate corneal fibrosis. Sci Rep 6, 20841. 10.1038/srep20841 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. McKay TB, Priyadarsini S, Karamichos D, 2019. Mechanisms of Collagen Crosslinking in Diabetes and Keratoconus. Cells 8, 1–28. 10.3390/cells8101239 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. McNutt PM, Mohan RR, 2020. The need for improved therapeutic approaches to protect the cornea against chemotoxic injuries. Transl Vis Sci Technol 9, 1–7. 10.1167/tvst.9.12.2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Meek KM, 2009. Corneal collagen-its role in maintaining corneal shape and transparency. Biophys Rev 1, 83–93. 10.1007/s12551-009-0011-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Meek KM, Knupp C, 2015. Corneal structure and transparency. Prog Retin Eye Res 49, 1–16. 10.1016/j.preteyeres.2015.07.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Mishra N, Agarwal R, 2022. Research models of sulfur mustard- and nitrogen mustard-induced ocular injuries and potential therapeutics. Exp Eye Res 223. 10.1016/J.EXER.2022.109209 [DOI] [PubMed] [Google Scholar]
  20. Mohan RR, Kempuraj D, D’Souza S, Ghosh A, 2022. Corneal stromal repair and regeneration. Prog Retin Eye Res. 10.1016/J.PRETEYERES.2022.101090 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Mohan RR, Morgan BR, Anumanthan G, Sharma A, Chaurasia SS, Rieger FG, 2016. Characterization of Inhibitor of differentiation (Id) proteins in human cornea. Exp Eye Res 146, 145–153. 10.1016/j.exer.2015.12.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Mutoji KN, Sun M, Elliott G, Moreno IY, Hughes C, Gesteira TF, Coulson-Thomas VJ, 2021. Extracellular Matrix Deposition and Remodeling after Corneal Alkali Burn in Mice. Int J Mol Sci 22. 10.3390/ijms22115708 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Pahuja N, Kumar NR, Shroff R, Shetty R, Nuijts RMMA, Ghosh AA, Sinha-Roy A, Chaurasia SS, Mohan RR, Ghosh AA, 2016. Differential molecular expression of extracellular matrix and inflammatory genes at the corneal cone apex drives focal weakening in keratoconus. Invest Ophthalmol Vis Sci 57, 5372–5382. 10.1167/iovs.16-19677 [DOI] [PubMed] [Google Scholar]
  24. Priyadarsini S, Whelchel A, Nicholas S, Sharif R, Riaz K, Karamichos D, 2020. Diabetic keratopathy: Insights and challenges. Surv Ophthalmol. 10.1016/j.survophthal.2020.02.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Saika S, Ooshima A, Shima K, Tanaka S, Ohnishi Y, 1996. Collagen types in healing alkali-burned corneal stroma in rabbits. Jpn J Ophthalmol 40, 303–309. [PubMed] [Google Scholar]
  26. Sasaki M, Hara T, Wang JX, Zhou Y, Kennedy KV, Umeweni CN, Alston MA, Spergel ZC, Ishikawa S, Teranishi R, Nakagawa R, Mcmillan EA, Whelan KA, Karakasheva TA, Hamilton KE, Ruffner MA, Muir AB, 2024. Lysyl Oxidase Regulates Epithelial Differentiation and Barrier Integrity in Eosinophilic Esophagitis. Cell Mol Gastroenterol Hepatol 17, 923–937. 10.1016/J.JCMGH.2024.01.025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Sharma A, Sinha NR, Siddiqui S, Mohan RR, 2015. Role of 5′TG3′-interacting factors (TGIFs) in vorinostat (HDAC inhibitor)-mediated corneal fibrosis inhibition. Mol Vis 21, 974–984. [PMC free article] [PubMed] [Google Scholar]
  28. Shetty R, Sathyanarayanamoorthy A, Ramachandra RA, Arora V, Ghosh AA, Srivatsa PR, Pahuja N, Nuijts RMMA, Sinha-Roy A, Mohan RR, Ghosh AA, Ghosh AA, 2015. Attenuation of lysyl oxidase and collagen gene expression in keratoconus patient corneal epithelium corresponds to disease severity. Mol Vis 21, 12–25. [PMC free article] [PubMed] [Google Scholar]
  29. Siegel RC, Pinnell SR, Martin GR, 1970. Cross-linking of collagen and elastin. Properties of lysyl oxidase. Biochemistry 9, 4486–4492. 10.1021/bi00825a004 [DOI] [PubMed] [Google Scholar]
  30. Sinha NR, Balne PK, Bunyak F, Hofmann AC, Lim RR, Rajiv R, 2021. Collagen matrix perturbations in corneal stroma of Ossabaw mini pigs with type 2 diabetes 666–678. [PMC free article] [PubMed]
  31. Sinha NR, Tripathi R, Balne PK, Green S, Sinha PR, Buyank FE, Giuliano EA, Chaurasia SS, Mohan RR, 2022. Time-dependent in situ structural and cellular aberrations in rabbit cornea in vivo after mustard gas exposure. Exp Eye Res 109247. 10.1016/J.EXER.2022.109247 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Sinha NR, Tripathi R, Balne PK, Suleiman L, Simkins K, Chaurasia SS, Mohan RR, 2023. Mustard Gas Exposure Actuates SMAD2/3 Signaling to Promote Myofibroblast Generation in the Cornea. Cells 12. 10.3390/cells12111533 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Taylor MA, Amin JD, Kirschmann DA, Schiemann WP, 2011. Lysyl oxidase contributes to mechanotransduction-mediated regulation of transforming growth factor-β signaling in breast cancer cells. Neoplasia 13, 406–418. 10.1593/neo.101086 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Tripathi R, Balne PK, Sinha NR, Martin LM, Kamil S, Landreneau JR, Gupta S, Rodier JT, Sinha PR, Hesemann NP, Hofmann AC, Fink MK, Chaurasia SS, Mohan RR, 2020. A novel topical ophthalmic formulation to mitigate acute mustard gas keratopathy in vivo: A pilot study. Transl Vis Sci Technol 9, 1–15. 10.1167/tvst.9.12.6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Tripathi R, Sinha NR, Kempuraj D, Balne PK, Landreneau JR, Juneja A, Webel AD, Mohan RR, 2022. Evaluation of CRISPR/Cas9 mediated TGIF gene editing to inhibit corneal fibrosis in vitro. Exp Eye Res 220, 109113. 10.1016/j.exer.2022.109113 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Voloshenyuk TG, Landesman ES, Khoutorova E, Hart AD, Gardner JD, 2011. Induction of cardiac fibroblast lysyl oxidase by TGF-β1 requires PI3K/Akt, Smad3, and MAPK signaling. Cytokine 55, 90–97. 10.1016/j.cyto.2011.03.024 [DOI] [PubMed] [Google Scholar]
  37. Wang L, Uhlig PC, Eikenberry EF, Robenek H, Bruckner P, Hansen U, 2014. Lateral growth limitation of corneal fibrils and their lamellar stacking depend on covalent collagen cross-linking by transglutaminase-2 and lysyl oxidases, respectively. Journal of Biological Chemistry 289, 921–929. 10.1074/jbc.M113.496364 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Wang TH, Hsia SM, Shieh TM, 2017. Lysyl oxidase and the tumor microenvironment. Int J Mol Sci 18, 1–12. 10.3390/ijms18010062 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Zeng R, Zhang D, Zhang J, Pan Y, Liu X, Qi Q, Xu J, Xu C, Shi S, Wang J, Liu T, Dong L, 2024. Targeting lysyl oxidase like 2 attenuates OVA-induced airway remodeling partly via the AKT signaling pathway. Respir Res 25. 10.1186/S12931-024-02811-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Zhou HY, Cao Y, Wu J, Zhang WS, 2017. Role of corneal collagen fibrils in corneal disorders and related pathological conditions. Int J Ophthalmol 10, 803–811. 10.18240/ijo.2017.05.24 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES