Progress towards a standardized model of ocular sulfur mustard injury for therapeutic testing

Abstract

Sulfur mustard (SM) remains a highly dangerous chemical weapon capable of producing mass casualties through liquid or vapor exposure. The cornea is highly sensitive to SM toxicity and exposure to low vapor doses can cause incapacitating acute injuries. At higher doses, corneas fail to fully heal and subsequently develop a constellation of symptoms known as mustard gas keratopathy (MGK) that causes reduced quality of life and impaired or lost vision. Despite a century of research, there are no specific treatments for acute or persistent ocular SM injuries. Here I summarize toxicological, clinical and pathophysiological mechanisms of SM vapor injury in the cornea, describe a preclinical model of ocular SM vapor exposure for reproducible therapeutic studies, and propose new approaches to improve evaluation of therapeutic effects. I also describe recent findings illustrating the delayed development of a transient but severe recurrent corneal lesion that, in turn, triggers the emergence of secondary keratopathies characteristic of the chronic form of MGK. Development of this recurrent lesion is SM dose-dependent, although the severity of the recurrent lesion appears SM dose-independent. Similar recurrent lesions have been reported in multiple species, including humans. Given the mechanistic relationship between the recurrent lesion and chronic, secondary keratopathies, I hypothesize that preventing the development of the recurrent lesion represents a novel and potentially valuable therapeutic approach for treatment of severe corneal SM injuries. Although ocular exposure to SM vapor continues to be a challenging therapeutic target, establishing consistent and reproducible models of corneal injury that enhance mechanistic and pathophysiological understanding will help satisfy regulatory requirements and accelerate the development of effective therapies.

Keywords: Chemical warfare agents; Corneal toxicity; Mustard gas keratopathy; Preclinical models; Reactive organic vapor; Sulfur mustard; Therapeutics.

Figure 1:

Summary of SM physiochemical properties and pathophysiological effects on the eye. (A) List of physical properties related to physical state and reactivity. Physical properties are from Pubchem (CID 10461). (B) Relationship between temperature and SM vapor concentration calculated using Antoine’s equation (Perry et al., 2015). P = vapor pressure, T = temperature in Kelvin. Antoine’s coefficients A (7.4749753), B (1940.711) and C (204.671) are from (Penski, 1993). (C) Representative images of human eyes demonstrating acute SM lesions (top) and MGK (bottom; modified from (Rajavi et al., 2017)). The spectrum of acute severity is shown, ranging from mild (left panel, conjunctival hyperemia, vascular dilation and no corneal involvement), moderate (center panel, mild signs plus corneal involvement and generalized edema) and severe (right panel, moderate signs plus corneal inflammation, corneal edema and corneal defects). The MGK images illustrate the development of irreversible pathologies such as neovascularization and corneal opacity.

Figure 2:

Representative covalent reactions of SM with water, DNA and protein targets. Original SM atoms are cyan; reactant atoms are black. The boxed panel depicts the cyclization of SM to the episulfonium ion. Conversion of episulfonium ions to hydroxyl groups is shown for simplicity and does not imply that proteins and other biological molecules may not be cross-linked. R is amino acid side chain; Pr is protein adduct. Structures were assembled using Chemdraw v19.1.

Figure 3:

(A) Location of cornea in the eye and role in light refraction. (B) Anatomical structure of the cornea with key tissues labeled. Both images are modified from (McNutt et al., 2016).

Figure 4:

The corneal SM vapor cap exposure method. (A) Assembly of the standard 14 mm corneal vapor cap. Whatman #1 filter paper is cut slightly larger than the screw-cap lid to ensure a tight fit. The o-ring diameter is measured at the top of the torus, which forms a seal with the cornea and thus defines the total corneal exposure area. (B) A saturating volume of neat SM liquid (10 uL) is deposited on the filter paper in the bottom of vapor cap. The syringe was created with BioRender.com. (C) The vapor cap is inverted on a glass slide for 1 min, then placed directly on the rabbit cornea and held in place with a pair of forceps to ensure a tight seal (<20 grams of pressure). Although not optimal, the cap can be manually held in place if necessary. (D) Representative images of acute SM injury after a 150 second exposure (estimated dose of 1500 min•mg/m³), showing increased corneal thickness in OCT images, increased opacity and frank epithelial loss without substantial injury to the eye lids (modified from (McNutt et al., 2012b)).

Figure 5:

Sample data from rabbit eyes exposed using the SM vapor cap for 60 seconds or 90 seconds. (A, B) Representative traces demonstrating REL emergence and progression after 60 second (A) and 90 second (B) exposures. Each color represents longitudinal changes in corneal thickness in the exposed cornea of an individual rabbit. Corneas were chosen to illustrate the full range of REL latencies in this study. (C-E) Violin plots demonstrating the effects of SM vapor dose on latency to REL onset (C; defined as corneal thickness exceeding 200% of baseline at ≥ 3 weeks after exposure), latency to peak edema during REL (D) and peak corneal thickness during REL (E). **, p < 0.01; ***, p < 0.001, ****, p < 0.0001. All data are modified from (McNutt et al., 2021)