Skip to main content
NCBI home page
The Journal of Pharmacology and Experimental Therapeutics logoLink to The Journal of Pharmacology and Experimental Therapeutics
. 2024 Feb;388(2):518–525. doi: 10.1124/jpet.123.001814

The Involvement of Unfolded Protein Response in the Mechanism of Nitrogen Mustard–Induced Ocular Toxicity

Assylbek Zhylkibayev 1, Trong Thuan Ung 1, James Mobley 1, Mohammad Athar 1, Marina Gorbatyuk 1,✉
PMCID: PMC10801749  PMID: 37914413

Abstract

Nitrogen mustard (NM) is a known surrogate of sulfur mustard, a chemical-warfare agent that causes a wide range of ocular symptoms, from a permanent reduction in visual acuity to blindness upon exposure. Although it has been proposed that the two blistering agents have a similar mechanism of toxicity, the mode of NM-induced cell death in ocular tissue has not been fully explored. Therefore, we hypothesized that direct ocular exposure to NM in mice leads to retinal tissue injury through chronic activation of the unfolded protein response (UPR) PERK arm in corneal cells and VEGF secretion, eventually causing cell death. We topically applied NM directly to mice to analyze ocular and retinal tissues at 2 weeks postexposure. A dramatic decline in retinal function, measured by scotopic and photopic electroretinogram responses, was detected in the mice. This decline was associated with enhanced TUNEL staining in both corneal and retinal tissues. In addition, exposure of corneal cells to NM revealed 228 differentially and exclusively expressed proteins primarily associated with the UPR, ferroptosis, and necroptosis. Moreover, these cells exhibited activation of the UPR PERK arm and an increase in VEGF secretion. Enhancement of VEGF staining was later observed in the corneas of the exposed mice. Therefore, our data indicated that the mechanism of NM-induced ocular toxicity should be carefully examined and that future research should identify a signaling molecule transmitted via a prodeath pathway from the cornea to the retina.

SIGNIFICANCE STATEMENT

This study demonstrated that NM topical exposure in mice results in dramatic decline in retinal function associated with enhanced TUNEL staining in both corneal and retinal tissues. We also found that the NM treatment of corneal cells resulted in 228 differentially and exclusively expressed proteins primarily associated with ferroptosis. Moreover, these cells manifest the UPR PERK activation and an increase in VEGF secretion. The latter was also found in the corneas of the cexposed mice.

Introduction

Sulfur mustard (SM) was extensively used in the military operations of the Iran-Iraq War between 1980 and 1988, affecting over 64,000 individuals (Ghazanfari et al., 2009; Amini et al., 2020). SM exposure causes severe ocular complications, ranging from a permanent reduction in visual acuity to blindness, which can develop in exposed individuals even 40 years postexposure (Feizi, 2019). Thus, a retrospective study involving survivors of an SM terror attack reported a reduction of retinal function in severely affected individuals (Shoeibi et al., 2017), supporting earlier research that detected central retinal vein occlusion, macular edema, diminished visual acuity (Shoeibi et al., 2016), and an overexpression of vascular endothelial growth factor (VEGF)-A in the tears (Abbaszadeh et al., 2014) of the survivors of the SM attack.

The mechanism of toxicity and structure of nitrogen mustard (NM), another chemical weapon, are similar to those of SM, and NM serves as an SM surrogate. Both SM and NM are potent vesicants that act as strong alkylating agents that cause skin blistering. These agents induce corneal injury and cause neovascularization (Goswami et al., 2016, 2019; McNutt et al., 2021). However, unlike SM, NM is commercially available, and studies of NM can be conducted without a containment facility. This has prompted NM researchers to examine the mechanism of vesicant-mediated ocular damage because there is a critical need to develop a countermeasure program to fight the injury. Thus, topical NM exposure in rabbits has been reported to cause severe injury of the anterior ocular segment, leading to infected conjunctiva, extensive corneal epithelial damage, and corneal neovascularization (Morad et al., 2005). A recent study of NM-exposed mice revealed retinal damage via gliosis (Mahaling et al., 2023). The authors demonstrated that the mechanism of toxicity in Müller cells is associated with the activation of the NLR family pyrin domain containing 3 (NLRP3) inflammasome in a caspase-1–dependent manner (Mahaling et al., 2023). The study did not assess the retinal function of the NM-exposed mice or further investigate whether glycolysis provokes the retinal cell death despite the importance of these issues. Another study of topical NM exposure investigated corneal and retinal injuries associated with protein citrullination, a post-translational modification in which arginine is converted into citrulline (Umejiego et al., 2023). Moreover, according to this study, both the corneal and retinal tissues of the exposed mice exhibited acute cellular damage with different kinetics, and hypercitrullination is common in deteriorated ocular tissues. The authors discovered that the acute (1 hour postexposure) and immediate (1 day postexposure) retinal responses were associated with gliosis in Müller cells and citrullination in retinal ganglion cells (Umejiego et al., 2023). Based on this key finding, this study suggested that the molecular signal initiated by topical corneal exposure propagates deeply into the retina and that the retinal tissue damage may be chronic.

Our recent studies of the vesicants phenylarsine oxide and lewisite (LEW) revealed chronic unfolded protein response (UPR) activation in the exposed corneal cells and retinal tissues (Zhylkibayev et al., 2023). Moreover, significant cytokine upregulation associated with chronically activated UPR was detected in both phenylarsine oxide- and LEW-exposed ocular tissues, suggesting the presence of crosstalk between these pathways (Zhylkibayev et al., 2023). In SM-exposed rabbit ocular tissue, VEGF expression reached its highest point at 28 days postexposure and led to corneal neovascularization.

Interestingly, another vesicant, LEW, was found to induce the highest VEGF expression at 14 days postexposure and sustain neovascularization for up to 56 days postexposure (Tewari-Singh et al., 2017). This observation is particularly significant because a study involving victims of SM attacks also demonstrated a strong correlation between elevated VEGF-A165 levels in the subjects’ tears and corneal neovascularization (Tewari-Singh et al., 2016). Furthermore, anti-VEGF therapy applied to SM-exposed rabbits significantly reduces neovascularization, underscoring the pivotal role of VEGF in transmitting molecular signals across the eye (Abbaszadeh et al., 2014) and in propagating molecular signals across the eye (Kadar et al., 2014; Gore et al., 2018).

It is well established that VEGF expression is susceptible to regulation by various external factors; a notable contributor is activating transcription factor 4 (ATF4), which acts as a mediator in the UPR protein kinase RNA-like ER kinase (PERK) pathway (Roybal et al., 2004; Oskolkova et al., 2008; Afonyushkin et al., 2010). Additionally, it has been reported that oxidative stress plays a role in these processes (Kim and Byzova, 2014). It has been proposed that both SM and NM act by depleting glutathione (GSH) and causing a significant downregulation of GSH reductase, ultimately leading to oxidative stress (Layali et al., 2018). For instance, GSH is known to regulate the proper formation of disulfide bonds in proteins. Moreover, in biopsies of SM-exposed lung tissue, researchers have detected hypoxia-induced and reactive oxygen species–reactivated gene expression, which further underscores the role of oxidative stress in these pathways (Tahmasbpour et al., 2016). All of these findings suggest that the mechanism of mustard’s ocular toxicity warrants careful examination, particularly in the context of ER homeostasis imbalance.

The primary objective of the current study was to explore whether direct ocular exposure to NM leads to chronic retinal injury characterized by retinal cell death and impaired retinal function. Our study also delved into the molecular mechanisms underpinning damage to corneal and retinal tissues and to determine whether these mechanisms are linked to the activation of the UPR and heightened secretion of VEGF. Our findings revealed a significant loss of retinal cells and a decline in retinal function in mice following topical exposure to NM. Additionally, in primary corneal cells, we observed concurrent chronic activation of the PERK pathway in the UPR and an increased secretion of VEGF. Taken together, our findings underscore the importance of further investigating the interplay between UPR activation and VEGF elevation in the context of ocular damage induced by NM exposure.

Materials and Methods

Cell Culture and Treatment

NM analog mechlorethamine HCl (Sigma Aldrich, 122564) was used for in vivo and in vitro studies. The primary human retinal endothelial cell line was purchased from Cell Systems (ACBRI 181). The culturing flasks were precoated 30 minutes before splitting with an endothelial attachment factor solution (Cell Applications, 123-500), and cells were grown in a basal medium (Lonza, CC-3156) supplemented with growth factors (Lonza, CC-3162). Human corneal keratocytes (HCKs) were purchased from Cell Applications (632-05a) along with corneal keratocytes growth medium (6111K-500). Primary corneal epithelial (HCE) cells were purchased from ATCC (PCS-700-010) along with growth medium (PCS-700-030) supplemented using a growth kit (PCS-700-040). Cells were subjected to 4–8 passages and treated when they reached 70%–80% confluence. NM at a concentration of 100 µM was used for exposure of cells over 24 hours, and sterile distilled water served as the vehicle control. The dose was determined based on the published data. (Goswami et al., 2016). To that end, 100 mM stock solution of NM dissolved in water, following the safety guidelines from the Occupational Health Safety Department of the University of Alabama at Birmingham, was prepared. For mass spectrometry, cells were collected after 24 hours of treatment. Four control and four NM-treated cell samples were used for the liquid chromatography–mass spectrometry study. Cells were lysed in M-PER lysis buffer (Fisher, PI78501) supplemented with protease and phosphatase inhibitors (Thermo Fisher Scientific, 78425). Protein concentrations were determined using the BCA Protein Assay Kit (Thermo Fisher Scientific, 23227). Ten micrograms per sample was loaded into Novex NuPAGE 10% Bis-Tris Protein gels (Invitrogen, NP0315BOX) stained with Colloidal Coomassie Blue Staining Kit (Invitrogen, LC6025). Each gel lane was digested with trypsin (Promega, V5280) in three fractions overnight.

Mass Spectrometry

This method was broadly described in our previous study (Ludwig et al., 2016). Briefly, samples digested with trypsin were loaded onto a 1260 Infinity high-performance liquid chromatography stack (Agilent Technologies) and separated using a 75-micron i.d. × 15-cm pulled-tip C-18 column (Jupiter C18 300 Å, 5 microns, Phenomenex). The Thermo Q Exactive HF-X mass spectrometer, equipped with a Nanospray Flex ion source (Thermo Fisher Scientific), was used in the study to perform proteomic analysis. Data were analyzed using Shiny Go web resource (http://bioinformatics.sdstate.edu/go/). A Venn diagram was created using Excel program.

Immunoblotting

Cells were scraped, collected after treatment, and lysed with radioimmunoprecipitation assay buffer (Cell Signaling Technology, 9806) supplemented with 1% Halt Protease Inhibitor and a phosphatase inhibitor cocktail (Thermo Fisher Scientific, 87786). Forty- to sixty-microgram protein samples were loaded onto 4%–20% Mini-PROTEAN TGX Precast Gels (Bio-Rad, 4561093EDU) and transferred to a polyvinylidene difluoride membrane from Bio-Rad (1704272). Membranes were then incubated for 1 hour in 5% skim milk (Bio-Rad, 1706404) prepared with 1× Tris-buffered saline (Bio-Rad, 1706435) with 0.1% Tween 20 (Sigma-Aldrich, P1379). Primary antibodies were diluted in 5% bovine serum albumin (Fisher, BP9703-100) dissolved in 1× Tris-buffered saline. The following antibodies were used in the study with 1:2000 dilutions: anti–phospho-eukaryotic translation initiation factor 2A (eIF2α) (Cell Signaling Technology, 3398), anti-eIF2α (Cell Signaling Technology, 9722), and anti-ATF4 (Cell Signaling Technology, 11815F). Anti–tribbles pseudokinase 3 (TRIB3) with a 1:1000 dilution (Santa Cruz Biotechnology, sc-390242), anti–growth arrest and DNA damage-inducible protein 34 (GADD34) with a 1:2000 dilution (Proteintech, 50-556-111), and anti–β-actin with a 1:10,000 dilution (Sigma-Aldrich, A2066) antibodies were used in the experiments. Secondary antibodies (1:10,000) were purchased from Li-Cor: horseradish peroxidase goat anti-Rabbit IgG (926-80011) and horseradish peroxidase goat anti-Mouse IgG (926-80010). Images of membranes were captured and analyzed using the Odyssey XF system (Li-Cor).

Animal Studies

In the present study, C57Bl/6J wild-type mice of both sexes were purchased from the Jackson Laboratory (Bar Harbor, ME) at the age of 8 weeks. All procedures were confirmed by the Institutional Animal Care and Use Committee of the University of Alabama at Birmingham with protocol 22537. The animals were maintained on a standard light and dark cycle (12/12 hours) with ambient room temperature and light illumination. The animals were kept in plastic cages equipped with safe environmental pads and standard bedding. They had unrestricted access to water and chow throughout the duration of the study. During experimental procedures, all animals were housed in special single-use prebedded disposable cages from Innovive (MSX2 set, San Diego, CA) in a quarantine room with special care and surveillance. Animals were euthanized using carbon dioxide gas following cervical dislocation.

NM Topical Application

NM was dissolved in pure distillated water with all required precautions. A full-face respirator (Fisher, 18-999-4556) equipped with a 3M 6000 cartridge (Fisher, 18-999-4550) was strictly used during preparation and exposure. Animals were anesthetized with a ketamine (50 mg/kg) and xylazine (10 mg/kg) cocktail intraperitoneally; pain was managed using buprenorphine extended release (1 mg/kg) approximately 30 minutes before exposure by subcutaneous injection. Whatman paper eye patches (2 × 2 mm) were applied on the ocular surfaces of sedated mice for 3 minutes to treat them with 5 μl/eye NM at a dose of 80 µg/eye. Control eyes received the vehicle (dH2O, 5 μl) only. Both eyes were carefully washed with pure water. After exposure to NM, the animals were kept in a continuously operated chemical hood for 3 hours for intensive observation and transferred to a quarantine room for daily care during experiments. The mice were euthanized after 2 weeks for the electroretinography (ERG) procedure, followed by histologic and immunochemical (IHC) analyses (Supplemental Fig. 1).

Electroretinography

After the animals were anesthetized using the ketamine/xylazine cocktail, phenylephrine 2.5% (Paragon BioTeck, 42702-102-15) drops were applied into the eye to induce mydriasis. Upon completion of dilation, the eye was coated with Gonak 2.5% (Akorn, Lake Forest, IL). ERG amplitudes were measured simultaneously on both eyes. Contact electrodes were placed on the cornea, whereas reference and ground electrodes were placed subcutaneously on the head area and tail. Between animals, the electrodes were sterilized with 70% ethanol. Flashes of varying intensity were directed toward the eyes in a Ganzfeld chamber, with the responses recorded by UTAS BigShot (LKC Technologies, Gaithersburg, MD). After the ERG measurements, the animals were euthanized and their eyes enucleated for histology. LKC EM software was used for the analysis of waveforms.

Histology and Immunochemistry

Hematoxylin and eosin staining was performed by fixing enucleated eyeballs in 4% paraformaldehyde as described earlier (Starr et al., 2019; Saltykova et al., 2021). Cryopreserved eyes were sectioned with 12-μm thickness using the Leica CM1510S cryostat (Leica, Buffalo Grove, IL). Corneal thickness was counted in 20× magnification with a 200-µm step distance between points. The vehicle-treated left eyes and NM-exposed right eyes were used for corneal morphometry.

The IHC analysis was conducted on the corneal sections using an anti-VEGF antibody (dilution 1:100; Santa Cruz Biotechnology, sc-365578). Donkey anti-Mouse IgG and Alexa Fluor 555 were employed as secondary antibodies (Invitrogen, PIA32773). Imaging analysis was carried out using the BZ-X800 fluorescence microscope (Keyence, Itasca, IL). To quantify the intensity of VEGF staining, two randomly selected fields at the center of the corneas were used. ImageJ software was employed to calculate the ratios of integrated density of VEGF signal over the area of interest (n = 6). A terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay was described in our previous study (Zhylkibayev et al., 2023). TUNEL-positive nuclei were counted manually by masked investigators (n = 7).

ELISA

Cells were treated with 100 µM NM for 24 hours, and conditioned media were collected for the detection of VEGF concentrations. Human VEGF was detected using an ELISA (Novus Biologicals, NBP1-91272). The procedure was performed according to the manufacturing company’s protocol.

Statistical Analysis

A Student two-tailed paired t test was used to compare two groups using Graphpad Prism 9 software. All statistical data were expressed as mean ± S.D. P < 0.05 was considered as significant.

Results

The primary objective of this study was to examine the impact of topical ocular exposure to NM on ocular pathobiology, focusing on the extension of its effects from the cornea to the retinal tissue. Our hypothesis posited that molecular signals originating from damaged corneal tissue may transmit a prodeath stimulus to the retina. To investigate the prominent biologic processes and potential molecular pathways affected by NM exposure, we subjected corneal keratocytes to NM treatment of proteomic analysis.

NM Exposure Alters the Proteomics of Corneal Keratocytes

The HCK cells were treated with 100 μM of NM for 24 hours and then harvested to prepare a protein extract for the liquid chromatography–mass spectrometry study. Overall, the study identified 1292 proteins. The Venn diagram in Fig. 1A presents the study results, highlighting the identification of 228 proteins in total to be significantly different between the treated and untreated groups (P < 0.05). Among all identified proteins, 111 were upregulated in control and 117 were upregulated in NM groups. These numbers include 13 identified proteins that were present exclusively in NM-treated cells (Fig. 1A, right panel) and 14 proteins that were absent in NM-treated cells and unique for the untreated control (Fig. 1A, left panel). Therefore, 228 differentially and exclusively expressed proteins were detected in the treated cells. A further analysis of 228 proteins demonstrated the increased and decreased proteins in NM-treated cells as compared with the control. To identify altered biologic processes using the Gene Ontology database, we then plotted all the increased and decreased proteins separately in the ShinyGO program (Fig. 1B). We have identified that the response to unfolded proteins, topologically incorrect proteins, and RNA and miRNA processing were the top five most enriched biologic processes found to be increased in NM-treated proteomics (up to 10-fold). Regarding the biologic processes that exhibited a decrease in activity, we observed a reduction in collagen-fibril and extracellular-matrix organization as well as in cell-substrate junction organization. Using the Kyoto Encyclopedia of Genes and Genomes PATHWAY database, we then plotted the upregulated proteins from NM-treated cells to search for enrichment in cellular signaling (Fig. 1C). Protein export, ferroptosis, protein processing in the endoplasmic reticulum (ER), N-glycan biosynthesis, necroptosis, and apoptosis were among the pathways that exhibited the most enrichment (up to 10-fold) after the treatment. We then uploaded the decreased proteins and found that they were responsible for the extracellular matrix–receptor interaction, DNA replication, proteosome, metabolism (glycolysis/gluconeogenesis), and protein processing in the ER. The upregulated proteins in NM-treated cells belonged predominantly to cellular components—including intracellular ferritin, oligosaccharyltransferase, and ER-containing complexes—and cytoplasmic stress granules (Fig. 1C). The proteins that exhibited a decrease in NM-treated cells, as visualized in the ShinyGo program, included collagen type 1 trimer, CMG and DNA replication complexes, and ficolin-1-rich granules (Fig. 1D).

Fig. 1.

Fig. 1.

Open in a new tab

The proteomics study was conducted with human corneal keratocytes exposed to nitrogen mustard for 24 hours. (A) The Venn diagram shows 228 differentially expressed proteins as a result of the treatment. In addition, 14 and 13 proteins were unique to cells treated with NM and control cells, respectively. A list of the unique proteins resulting from the treatment with NM is shown on the right. Proteins that are exclusively present in control and absent in NM-treated cells are shown on the left. (B) Upregulated and downregulated proteins associated with biologic processes in NM-treated cells based on the Gene Ontology (GO) knowledgebase are shown. (C) Upregulated and downregulated cellular pathways in NM-treated cells using the Kyoto Encyclopedia of Genes and Genomes (KEGG) PATHWAY database are depicted. (D) Upregulated cellular components and downregulated molecular function in NM-treated cells using the Gene Ontology database are presented.

The results of the proteomic analysis of NM-treated cells showed a significant enrichment of proteins associated with the UPR and of biologic processes and molecular pathways occurring in the ER. This discovery led us to explore whether UPR signaling is indeed activated in these treated cells and whether this activation is correlated with the generation of a prodeath signaling molecule in the treated corneal cells. We hypothesized that the mode of transmission for such a signaling molecule involves extracellular secretion.

NM Exposure Results Regarding Activation of the UPR PERK Pathway in Corneal and Retinal Cells

The UPR is a cellular adaptive response that launches prosurvival programs in cells to restore cellular homeostasis by activating PERK, activating transcription factor 6 (ATF6), and inositol-requiring enzyme type 1α (IRE1α). Under sustained cellular stress, such as that induced by vesicant exposure, we observed that persistent UPR activation shifted the cellular survival program toward death (Zhylkibayev et al., 2023). Although all three UPR arms may play significant roles in ocular tissue damage, chronically activated PERK signaling is well accepted to mostly contribute to apoptotic cell death. Our published work with mice has demonstrated that persistent PERK activation induces retinal cell death through apoptosis (Rana et al., 2014; Bhootada et al., 2016; Starr et al., 2018; Gorbatyuk et al., 2020). The results of our western blot analysis, presented in Fig. 2, indicated that the PERK arm mediators p-eIF2α, ATF4, and GADD34 were significantly upregulated in HCK cells after 24 hours NM treatment. Although the treated and untreated cells exhibited no difference in TRIB3 protein levels, TRIB3 mRNA expression was significantly increased (by 60%) after the treatment (Fig. 2A). Like HCK cells, human retinal endothelial cells responded to the NM treatment and exhibited an increase in p-eIF2α and GADD34. Overall, these data indicated that the UPR PERK arm was activated after NM exposure (Fig. 2B).

Fig. 2.

Fig. 2.

Open in a new tab

UPR PERK signaling activated in the cells treated with nitrogen mustard for 24 hours. (A) The human corneal keratocytes demonstrated an increase in PERK downstream mediators p-eIF2α, ATF4, and GADD34. (B) The treatment of human retinal endothelial cells with nitrogen mustard resulted in the upregulation of p-eIF2α and GADD34 (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; n = 4 for each group). HREC, human retinal endothelial cell; ns, not significant.

Considering that sustained activation of the UPR PERK arm leads to apoptotic cell death and recognizing that treatment of HCK cells with NM induces various modes of cell death, we proceeded to investigate the response of wild-type mice to NM exposure. To do so, we topically applied 5 µl of NM solution containing 80 µg of vesicant for 3 minutes onto the corneal surface using a Whatman paper eye patch. Two weeks later, the mice were analyzed with electroretinography and TUNEL staining and subjected to a histologic assessment.

Topical NM Exposure in Mice Results in Both Corneal and Retinal Degeneration

Our analysis of corneal thickness in mice exposed to NM showed an increased corneal thickness of the treated eyes and suggested corneal edema (Fig. 3A). Later, we found a dramatic increase in the number of TUNEL-positive cells in the treated corneas, which served as evidence of cell death (Fig. 3B). Notably, the retinas of the treated mice also had an increased number of TUNEL-positive cells, suggesting that a topically applied vesicant can injure the retina as well as the cornea.

Fig. 3.

Fig. 3.

Open in a new tab

Nitrogen mustard topical exposure in mice leading to corneal and retinal damage. (A) Treated corneas manifest an increase in thickness. This increase was associated with cell death determined by the TUNEL staining analysis (****P < 0.0001; n = 5). Scale bar, 50 μm. (B) The number of TUNEL-positive cells increased in the treated corneas (*P < 0.05; ***P < 0.001) and retinas (n = 7). Scale bar, 50 μm. (C) Results of the scotopic and photopic electroretinogram analyses recorded with 25 cd*s/m2 flash intensities are shown (upper panel; n = 5). The images of the representative scotopic and photopic electroretinogram amplitudes registered in treated versus untreated eyes are depicted (bottom panel). n = 5 for each group. **P < 0.01. ns, not significant; PhNR, photopic negative response.

Indeed, significant reductions in the scotopic A- and B-wave amplitudes and the photopic B-wave amplitudes were registered in treated eyes as compared with the vehicle treatment (Fig. 3C). Interestingly, the photopic negative response characterizing the retinal ganglion cell function had not diminished at 2 weeks postexposure. These data suggested that retinal degeneration in the mice resulted from either the distribution of the NM agent itself through the corneal tissue or the propagation across the eyeball of a molecular signal generated by the corneal cells. Therefore, we searched for a molecular signal transmitted by the corneal cells that may initiate retinal pathogenesis. Our rationale for this search was grounded in our previous studies, which described the role of ATF4 and CHOP proteins in the overexpression of VEGF and cytokines (Wang et al., 2013; Rana et al., 2014; Bhootada et al., 2016).

NM Exposure Leads to an Increase in VEGF Secretion from Human Corneal Epithelial Cells and Keratocytes

Twenty-four hours after NM treatment, we collected the cell-growing media to analyze the VEGF concentrations (Fig. 4A). We found that whereas the control HCE cells had approximately 400 pg/ml of VEGF, the media of NM-treated HCE cells had over 700 pg/ml of VEGF (P < 0.0001). Like the HCE cells, the control HCK cells had over 700 pg/ml of VEGF, whereas NM-treated cells exhibited a twofold increase in secreted VEGF (P < 0.001). These data demonstrated that the cells exposed to NM activated the UPR PERK arm, enhancing VEGF secretion. We then analyzed the corneas of the treated eyes using IHC to identify the anti-VEGF signal (Fig. 4B). A dramatic increase in VEGF signaling in treated versus untreated corneas was detected.

Fig. 4.

Fig. 4.

Open in a new tab

Exposure to nitrogen mustard leading to increased VEGF secretion. (A) The human corneal keratocytes and epithelial cells exposed to 100 µM NM for 24 hours manifested an increase in VEGF secretion (n = 4). (B) The nitrogen mustard–treated cornea demonstrates an increase in the VEGF signal (n = 6). Representative images are shown. **P < 0.01; ***P < 0.001; ****P < 0.0001. Scale bar, 50 μm. HCEC, human corneal epithelial cell.

Discussion

Although NM, a surrogate for SM, has been accessible to the research community for over 80 years and has been validated for use in animals, the exact molecular mechanisms underlying its action remain the subject of ongoing investigation. Furthermore, despite the growing interest in this vesicant, there is a notable gap in expanded research on NM exposure and its relation to retinal degeneration. Consequently, to the best of our knowledge, the current study is the first to comprehensively analyze the impact of NM on retinal function and cell loss in treated animals. The primary findings of this study revealed that direct ocular exposure to NM has detrimental effects not only on corneal tissue but also on the retina. We observed significant cell loss in both of these exposed tissues in mice. Furthermore, our investigation showed that the loss of retinal cells is correlated with a decline in scotopic and photopic electroretinogram responses, indicating a loss of retinal function. Our results also demonstrated that corneal cell death can occur through activation of the UPR and that VEGF overexpression results in various cell-death pathways. Understanding and mitigating the prodeath signal transmitted from treated corneal cells to the retina is essential to preserving patient vision and enhancing the safety and effectiveness of corneal procedures.

We found that NM exposure leads to upregulation of the UPR in corneal keratocytes and cell death through ferroptosis, necroptosis, and apoptosis. Overall, ferroptosis, a recently identified mode of programmed cell death, differs from apoptosis by its dependency on iron and the accumulation of lipid peroxides. Whereas ferroptosis has never been reported for SM- or NM-exposed tissue or cells, necrosis and apoptosis have been detected after SM treatment (Dabrowska et al., 1996; Heinrich et al., 2009). Recent research has suggested a potential crosstalk relationship between the UPR PERK arm and ferroptosis. Lee et al. (2018) proposed that the ferroptotic agent artesunate, a glutathione S-transferase inhibitor, promotes ATF4-mediated CHOP and TRIB3. Furthermore, previous studies conducted by our team and other investigators have demonstrated that sustained ATF4 overexpression leads to retinal degeneration (Bhootada et al., 2016) and that the upregulation of CHOP induces apoptosis (Oyadomari and Mori, 2004). Indeed, ferroptosis may induce p-eIF2α→ATF4 axis activation (McGrath et al., 2021). It is important to note that NM with genotoxic properties can induce a wide array of mutations at G-C and A-T base pairs that may, in turn, cause DNA damage and premature transcriptional termination (Pieper and Erickson, 1990; Povirk and Shuker, 1994). In light of these genotoxic effects, it is plausible to conclude that NM treatment leads to the accumulation of misfolded proteins and subsequently activates the UPR. Nevertheless, the exact mechanisms through which NM exposure regulates the UPR in treated corneal cells remain to be fully elucidated. To address this knowledge gap, future experiments will need to overcome the current study’s limitations.

The activated UPR, ferroptosis, or necroptosis in keratocytes may upregulate VEGF and other proinflammatory cytokines (Wang et al., 2013; Rana et al., 2014; Hänggi et al., 2017; Li et al., 2022). We found that both HCE and HCK cells treated with NM exhibited an increase in VEGF secretion. Moreover, in treated HCK cells, the observed increase in VEGF secretion was concomitant with UPR activation, suggesting that VEGF may serve as a prodeath signaling molecule transmitted to the retina. In a previous study of lewisite, significant upregulation of the cytokines Il-1β, Il-6, and Cox2 was found in the retina of treated mice 24 hours after exposure (Zhylkibayev et al., 2023), suggesting that acute cellular stress was transmitted from the cornea. Moreover, secreted VEGF may trigger corneal neovascularization, which was observed in NM-exposed animals (Goswami et al., 2019), suggesting that strategies aimed at targeting UPR→VEGF may slow down the magnitude of corneal neovascularization in exposed individuals. Therefore, to overcome the limitations of our study, our future research will attempt to determine the upstream regulatory mechanisms of VEGF regulation in NM-treated cells. Additionally, to analyze VEGF expression in affected retinal cells, it is crucial to extend the time points beyond 50 days.

After 2 weeks of exposure, the retina exhibited a loss of photoreceptor function, and both rods and cones were affected. The functional loss was consistent with the retinal cell loss detected with TUNEL staining. Interestingly, TUNEL staining usually fails to discriminate between apoptosis, necrosis, necroptosis, and ferroptosis (Grasl-Kraupp et al., 1995; Shindo et al., 2019; Tadokoro et al., 2020). Therefore, given the mixed mode of the cell death detected in NM-treated keratocytes, we cannot be certain about the apoptotic mechanism in affected ocular tissues. In addition, the mode of cell death may differ between corneal cells (epithelial cells, keratocytes, and endothelial cells) and neuronal cells. Therefore, future research should identify the exact mechanism of retinal cell deterioration after direct NM exposure, thus overcoming the current study’s limitations.

In summary, the current study delineated links between NM exposure and ocular pathogenesis in mice, specifically tied to the activation of the UPR PERK arm and the subsequent increase in VEGF secretion in corneal cells, ultimately leading to cell death. To address the limitations of our study, further investigations are necessary. These future experiments should aim to determine whether UPR activation in corneal cells generates a molecular signal that is transmitted to the posterior segment of the eye, focusing on the potential role of VEGF in this signaling pathway. Furthermore, to enhance our understanding of the molecular mechanisms underlying NM-induced ocular tissue damage, it is necessary to explore the potential crosstalk between the UPR and ferroptosis or between the UPR and necrosis. Such research efforts will be critical to advancing our knowledge of ocular NM toxicity and facilitating the development of effective medical countermeasures.

Data Availability

The authors declare that all the data supporting the findings of this study are available.

Abbreviations

ATF4

activating transcription factor 4

eIF2α

eukaryotic translation initiation factor 2A

ER

endoplasmic reticulum

ERG

electroretinography

GADD34

growth arrest and DNA damage-inducible protein 34

GSH

glutathione

HCE

human corneal epithelial

HCK

human corneal keratocyte

IHC

immunohistochemistry

LEW

lewisite

NM

nitrogen mustard

PERK

protein kinase RNA-like ER kinase

SM

sulfur mustard

TRIB3

tribbles pseudokinase 3

TUNEL

terminal deoxynucleotidyl transferase dUTP nick end labeling

UPR

unfolded protein response

VEGF

vascular endothelial growth factor

Authorship Contributions

Participated in research design: Zhylkibayev, Ung.

Conducted experiments: Zhylkibayev, Ung.

Performed data analysis: Zhylkibayev, Ung, Mobley, Athar, Gorbatyuk.

Wrote or contributed to the writing of the manuscript: Mobley, Athar, Gorbatyuk.

Footnotes

This study was supported by National Institutes of Health National Eye Institute [Grant 5R01EY034110] (to M.G.).

The authors do not have a conflict of interest to disclose.

Primary laboratory of origin: University of Alabama at Birmingham (Birmingham, AL).

Inline graphicThis article has supplemental material available at jpet.aspetjournals.org.

References

  1. Abbaszadeh M, Aidenloo NS, Nematollahi MK, Motarjemizadeh Q (2014) Investigating the Association between Angiogenic Cytokines and Corneal Neovascularization in Sulfur Mustard Intoxicated Subjects 26 Years after Exposure. Toxicol Int 21:300–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Afonyushkin T, Oskolkova OV, Philippova M, Resink TJ, Erne P, Binder BR, Bochkov VN (2010) Oxidized phospholipids regulate expression of ATF4 and VEGF in endothelial cells via NRF2-dependent mechanism: novel point of convergence between electrophilic and unfolded protein stress pathways. Arterioscler Thromb Vasc Biol 30:1007–1013. [DOI] [PubMed] [Google Scholar]
  3. Amini HSolaymani-Dodaran MMousavi BAlam Beladi SNSoroush MRAbolghasemi JVahedian-Azimi ASalesi MGuest PCSahebkar A, et al. (2020) Long-term Health Outcomes Among Survivors Exposed to Sulfur Mustard in Iran. JAMA Netw Open 3:e2028894. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bhootada Y, Kotla P, Zolotukhin S, Gorbatyuk O, Bebok Z, Athar M, Gorbatyuk M (2016) Limited ATF4 Expression in Degenerating Retinas with Ongoing ER Stress Promotes Photoreceptor Survival in a Mouse Model of Autosomal Dominant Retinitis Pigmentosa. PLoS One 11:e0154779. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Dabrowska MI, Becks LL, Lelli JL Jr, Levee MG, Hinshaw DB (1996) Sulfur mustard induces apoptosis and necrosis in endothelial cells. Toxicol Appl Pharmacol 141:568–583. [DOI] [PubMed] [Google Scholar]
  6. Feizi S (2019) Reply. Cornea 38:e24–e25. [DOI] [PubMed] [Google Scholar]
  7. Ghazanfari TFaghihzadeh SAragizadeh HSoroush MRYaraee RMohammad Hassan ZForoutan AVaez-Mahdavi MRJavadi MAMoaiedmohseni S, et al. ; Sardasht-Iran Cohort Study Research Group (2009) Sardasht-Iran cohort study of chemical warfare victims: design and methods. Arch Iran Med 12:5–14. [PubMed] [Google Scholar]
  8. Gorbatyuk MS, Starr CR, Gorbatyuk OS (2020) Endoplasmic reticulum stress: New insights into the pathogenesis and treatment of retinal degenerative diseases. Prog Retin Eye Res 79:100860. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Gore A, Horwitz V, Cohen M, Gutman H, Cohen L, Gez R, Kadar T, Dachir S (2018) Successful single treatment with ziv-aflibercept for existing corneal neovascularization following ocular chemical insult in the rabbit model. Exp Eye Res 171:183–191. [DOI] [PubMed] [Google Scholar]
  10. Goswami DG, Kant R, Ammar DA, Kumar D, Enzenauer RW, Petrash JM, Tewari-Singh N, Agarwal R (2019) Acute corneal injury in rabbits following nitrogen mustard ocular exposure. Exp Mol Pathol 110:104275. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Goswami DG, Tewari-Singh N, Dhar D, Kumar D, Agarwal C, Ammar DA, Kant R, Enzenauer RW, Petrash JM, Agarwal R (2016) Nitrogen Mustard-Induced Corneal Injury Involves DNA Damage and Pathways Related to Inflammation, Epithelial-Stromal Separation, and Neovascularization. Cornea 35:257–266. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Grasl-Kraupp B, Ruttkay-Nedecky B, Koudelka H, Bukowska K, Bursch W, Schulte-Hermann R (1995) In situ detection of fragmented DNA (TUNEL assay) fails to discriminate among apoptosis, necrosis, and autolytic cell death: a cautionary note. Hepatology 21:1465–1468. [DOI] [PubMed] [Google Scholar]
  13. Hänggi KVasilikos LValls AFYerbes RKnop JSpilgies LMRieck KMisra TBertin JGough PJ, et al. (2017) RIPK1/RIPK3 promotes vascular permeability to allow tumor cell extravasation independent of its necroptotic function. Cell Death Dis 8:e2588. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Heinrich A, Balszuweit F, Thiermann H, Kehe K (2009) Rapid simultaneous determination of apoptosis, necrosis, and viability in sulfur mustard exposed HaCaT cell cultures. Toxicol Lett 191:260–267. [DOI] [PubMed] [Google Scholar]
  15. Kadar T, Amir A, Cohen L, Cohen M, Sahar R, Gutman H, Horwitz V, Dachir S (2014) Anti-VEGF therapy (bevacizumab) for sulfur mustard-induced corneal neovascularization associated with delayed limbal stem cell deficiency in rabbits. Curr Eye Res 39:439–450. [DOI] [PubMed] [Google Scholar]
  16. Kim YW, Byzova TV (2014) Oxidative stress in angiogenesis and vascular disease. Blood 123:625–631. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Layali I, Shahriary A, Rahmani Talatappe N, Tahmasbpour E, Rostami H, Beigi Harchegani A (2018) Sulfur mustard triggers oxidative stress through glutathione depletion and altered expression of glutathione-related enzymes in human airways. Immunopharmacol Immunotoxicol 40:290–296. [DOI] [PubMed] [Google Scholar]
  18. Lee YS, Lee DH, Choudry HA, Bartlett DL, Lee YJ (2018) Ferroptosis-Induced Endoplasmic Reticulum Stress: Cross-talk between Ferroptosis and Apoptosis. Mol Cancer Res 16:1073–1076. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Li GLin YZhang YGu NYang BShan SLiu NOuyang JYang YSun F, et al. (2022) Endometrial stromal cell ferroptosis promotes angiogenesis in endometriosis. Cell Death Discov 8:29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Ludwig MR, Kojima K, Bowersock GJ, Chen D, Jhala NC, Buchsbaum DJ, Grizzle WE, Klug CA, Mobley JA (2016) Surveying the serologic proteome in a tissue-specific kras(G12D) knockin mouse model of pancreatic cancer. Proteomics 16:516–531. [DOI] [PubMed] [Google Scholar]
  21. Mahaling B, Sinha NR, Sokupa S, Addi UR, Mohan RR, Chaurasia SS (2023) Mustard gas exposure instigates retinal Müller cell gliosis. Exp Eye Res 230:109461. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. McGrath EP, Centonze FG, Chevet E, Avril T, Lafont E (2021) Death sentence: The tale of a fallen endoplasmic reticulum. Biochim Biophys Acta Mol Cell Res 1868:119001. [DOI] [PubMed] [Google Scholar]
  23. McNutt PMKelly KEMAltvater ACNelson MRLyman MEO’Brien SConroy MTOndeck CABodt SMLWolfe SE, et al. (2021) Dose-dependent emergence of acute and recurrent corneal lesions in sulfur mustard-exposed rabbit eyes. Toxicol Lett 341:33–42. [DOI] [PubMed] [Google Scholar]
  24. Morad Y, Banin E, Averbukh E, Berenshtein E, Obolensky A, Chevion M (2005) Treatment of ocular tissues exposed to nitrogen mustard: beneficial effect of zinc desferrioxamine combined with steroids. Invest Ophthalmol Vis Sci 46:1640–1646. [DOI] [PubMed] [Google Scholar]
  25. Oskolkova OV, Afonyushkin T, Leitner A, von Schlieffen E, Gargalovic PS, Lusis AJ, Binder BR, Bochkov VN (2008) ATF4-dependent transcription is a key mechanism in VEGF up-regulation by oxidized phospholipids: critical role of oxidized sn-2 residues in activation of unfolded protein response. Blood 112:330–339. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Oyadomari S, Mori M (2004) Roles of CHOP/GADD153 in endoplasmic reticulum stress. Cell Death Differ 11:381–389. [DOI] [PubMed] [Google Scholar]
  27. Pieper RO, Erickson LC (1990) DNA adenine adducts induced by nitrogen mustards and their role in transcription termination in vitro. Carcinogenesis 11:1739–1746. [DOI] [PubMed] [Google Scholar]
  28. Povirk LF, Shuker DE (1994) DNA damage and mutagenesis induced by nitrogen mustards. Mutat Res 318:205–226. [DOI] [PubMed] [Google Scholar]
  29. Rana T, Shinde VM, Starr CR, Kruglov AA, Boitet ER, Kotla P, Zolotukhin S, Gross AK, Gorbatyuk MS (2014) An activated unfolded protein response promotes retinal degeneration and triggers an inflammatory response in the mouse retina. Cell Death Dis 5:e1578. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Roybal CN, Yang S, Sun CW, Hurtado D, Vander Jagt DL, Townes TM, Abcouwer SF (2004) Homocysteine increases the expression of vascular endothelial growth factor by a mechanism involving endoplasmic reticulum stress and transcription factor ATF4. J Biol Chem 279:14844–14852. [DOI] [PubMed] [Google Scholar]
  31. Saltykova IV, Elahi A, Pitale PM, Gorbatyuk OS, Athar M, Gorbatyuk MS (2021) Tribbles homolog 3-mediated targeting the AKT/mTOR axis in mice with retinal degeneration. Cell Death Dis 12:664. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Shindo ROhmuraya MKomazawa-Sakon SMiyake SDeguchi YYamazaki SNishina TYoshimoto TKakuta SKoike M, et al. (2019) Necroptosis of Intestinal Epithelial Cells Induces Type 3 Innate Lymphoid Cell-Dependent Lethal Ileitis. iScience 15:536–551. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Shoeibi N, Balali-Mood M, Abrishami M (2016) Sulfur Mustard Exposure and Non-Ischemic Central Retinal Vein Occlusion. Iran J Med Sci 41:59–63. [PMC free article] [PubMed] [Google Scholar]
  34. Shoeibi N, Mousavi MN, Balali-Mood M, Moshiri M, Darchini-Maragheh E, Mousavi SR, Abrishami M (2017) Long-term complications of sulfur mustard poisoning: retinal electrophysiological assessment in 40 severely intoxicated Iranian veterans. Int J Retina Vitreous 3:7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Starr CR, Nyankerh CNA, Qi X, Hu Y, Gorbatyuk OS, Sonenberg N, Boulton ME, Gorbatyuk MS (2019) Role of Translational Attenuation in Inherited Retinal Degeneration. Invest Ophthalmol Vis Sci 60:4849–4857. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Starr CR, Pitale PM, Gorbatyuk M (2018) Translational attenuation and retinal degeneration in mice with an active integrated stress response. Cell Death Dis 9:484. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Tadokoro TIkeda MIde TDeguchi HIkeda SOkabe KIshikita AMatsushima SKoumura TYamada KI, et al. (2020) Mitochondria-dependent ferroptosis plays a pivotal role in doxorubicin cardiotoxicity. JCI Insight 5:e132747. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Tahmasbpour E, Ghanei M, Qazvini A, Vahedi E, Panahi Y (2016) Gene expression profile of oxidative stress and antioxidant defense in lung tissue of patients exposed to sulfur mustard. Mutat Res Genet Toxicol Environ Mutagen 800–801:12–21. [DOI] [PubMed] [Google Scholar]
  39. Tewari-Singh NCroutch CRTuttle RGoswami DGKant RPeters ECulley TAmmar DAEnzenauer RWPetrash JM, et al. (2016) Clinical progression of ocular injury following arsenical vesicant lewisite exposure. Cutan Ocul Toxicol 35:319–328. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Tewari-Singh N, Goswami DG, Kant R, Ammar DA, Kumar D, Enzenauer RW, Casillas RP, Croutch CR, Petrash JM, Agarwal R (2017) Histopathological and Molecular Changes in the Rabbit Cornea From Arsenical Vesicant Lewisite Exposure. Toxicol Sci 160:420–428. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Umejiego E, Paramo R, Zafiris A, Mullane E, Bargagna-Mohan P, Mohan R (2023) A corneo-retinal hypercitrullination axis underlies ocular injury to nitrogen mustard. Exp Eye Res 231:109485. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Wang X, Wang G, Kunte M, Shinde V, Gorbatyuk M (2013) Modulation of angiogenesis by genetic manipulation of ATF4 in mouse model of oxygen-induced retinopathy [corrected]. Invest Ophthalmol Vis Sci 54:5995–6002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Zhylkibayev A, Srivastava R, Anantharam P, Crotch C, Athar M, Gorbatyuk M (2023) Retinal injury mouse model and pathophysiological assessment of the effect of arsenical vesicants. Exp Eye Res 226:109354. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from The Journal of Pharmacology and Experimental Therapeutics are provided here courtesy of American Society for Pharmacology and Experimental Therapeutics

RESOURCES