YK-4–250 mitigates gastrointestinal radiation syndrome and promotes overall survival following partial body radiation injury

Abstract

Acute gastrointestinal radiation syndrome (GI-ARS) is a significant health threat following high-dose ionizing radiation (IR) exposure, leading to severe morbidity and mortality. The syndrome is characterized by gastrointestinal tissue damage caused by angiotensin II (Ang II) and reactive oxygen species (ROS), resulting in impaired GI function, systemic bacteremia, multi-organ failure, and eventual death. Dysregulation of the renin-angiotensin system (RAS) via Ang II exacerbates ROS production through activation of the Angiotensin II type 1 receptor (AT1R). This underscores the need for agents capable of both scavenging ROS and inhibiting AT1R activity. To address this, we developed YK-4–250, a Tempol-conjugated angiotensin receptor blocker (TCARB). YK-4–250 selectively inhibits the AT1R and exhibits antioxidant properties like Tempol and has a no observed adverse effect level (NOAEL) greater than 100 mg/kg. A single daily oral dose of 20 mg/kg of YK-4–250, administered either prior to or after 50% lethal dose (LD₅₀) of partial body irradiation (PBI), improved overall survival significantly (by 25–30% above vehicle group) and enhanced GI function by day 7. Additionally, YK-4–250’s paramagnetic properties enable MRI monitoring, allowing visualization of drug delivery to target tissues. These findings suggest that YK-4–250 is a promising candidate for protecting and mitigating radiation-induced injury resulting in improved outcomes in GI-ARS.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-53734-7.

Introduction

There is a critical need for radiation countermeasures that can mitigate tissue damage following exposure to ionizing radiation. Accidental, therapeutic, or deliberate radiation events can cause severe injury to vital organs such as the bone marrow and gastrointestinal (GI) tract, leading to high morbidity and mortality. Developing effective mitigators would provide essential protection, promote tissue recovery, and significantly improve survival outcomes after radiation exposure ¹. Following radiation exposure, the bone marrow and GI tract are the organs most critically affected by ionizing radiation (IR) ^2,3, serving as the primary determinants of lethality following high-dose exposure. GI toxicity represents the major cause of early mortality (within 7–10 days post-exposure) and is characterized by rapid loss of intestinal stem and progenitor cells, extensive destruction of intestinal crypts, and impaired epithelial regeneration. These events compromise mucosal barrier integrity, leading to systemic bacteremia, multi-organ failure, and ultimately death-collectively referred to as gastrointestinal acute radiation syndrome (GI-ARS) ^2,4,5.

In humans, GI-ARS typically occurs following whole-body radiation doses exceeding 5–6 Gy ^2,6 with death occurring within approximately two weeks in the absence of medical intervention due to severe vomiting and diarrhea ⁷. The pathophysiology of GI-ARS involves protein oxidation, redox imbalance, inflammation, cytokine dysregulation, and functional loss due to epithelial cell death ⁸. Consequently, GI-ARS is marked by epithelial apoptosis, intestinal hemorrhage, sepsis, and fluid-electrolyte imbalance, culminating in fatal outcomes ³. Rapid regeneration of the intestinal epithelium is therefore essential for recovery and survival following GI-ARS.

GI-ARS is characterized by the massive loss of GI crypt epithelial cells due to the cytotoxic and genotoxic complications of excessive generation of reactive oxygen species (ROS) by IR ^3,9–12. IR directly induces DNA double strand breaks (Fig. 1A) and generates toxic free radicals, such as the superoxide anion and hydroxyl radicals, via the radiolysis of water (Fig. 1B) ^13,14. An early pathway resulting in excessive generation of ROS following IR is the production of elevated Angiotensin II (Ang II) ^15,16. Ang II is the key product of the renin angiotensin system (RAS) that regulates intracellular and extracellular ROS and maintains blood pressure control. Lethal IR triggers a dose-dependent release of Ang II and an upregulation of Angiotensin II type 1 receptor (AT1R, which activates NADPH oxidase (NOX) to further increase ROS levels and exacerbate GI tissue damage (Fig. 1C) ¹⁷. The RAS is also vital for GI function ¹⁸ and promoting anti-inflammation by up-regulating the angiotensin converting enzyme 2 (ACE2) peptidase that cleaves Ang II to produce Ang 1–7 ^19–21. Ang 1–7 agonizes the AT2 and Mas receptors ²² antagonizing the inflammatory cascade produced by Ang II ^23–26, thereby promoting stem cell regeneration and opposing the consequences of hypercytokinemia (Fig. 1D). Downstream signaling cascades directly counteract the pro-inflammatory signaling initiated by the AT1R axis and mitigate the systemic and localized effects of hypercytokinemia. This dual receptor activation shifts the cellular microenvironment toward tissue survival, driving the promotion of intestinal stem cell regeneration and the functional restoration of mucosal barrier integrity following IR insult.

Fig. 1.

Effects of IR on the renin angiotensin system (RAS) and radical oxygen species (ROS). Irradiation results in (A) direct DNA damage but greater levels of (B) radiolysis led to ROS generation as well as induction of (C) Ang II that further increases ROS generation thereby further boosting oxidative stress and inflammation. (D) To counter the damage due to increased ROS, increased Ang II leads to more ACE2 and generation of Ang1-7 that promotes anti-oxidative and anti-fibrotic effects.

Taken together, these highly coordinated series of events following lethal IR led us to develop small molecules that combined molecular mechanisms to modulate the RAS and ROS generation following IR. In this report, we critically examined the combined effects of blocking the Ang II/AT1 axis, while simultaneously quenching intracellular and extracellular free radical generation 24 h after GI injury from IR. Herein, we demonstrate that YK-4–250 mitigates GI-ARS by improving DNA repair, restoring mucosal barrier integrity and function, and preventing ROS-induced crypt epithelial damage, ultimately reducing stem-cell loss, and improving survival after gut-directed lethal irradiation.

Results

Design of YK-4–250 and molecule docking

We have been actively involved in developing radiation countermeasures that target a dual mechanism ²⁷. Because ionizing radiation-induced ROS drives apoptosis and stem cell depletion ^3,5,28,29, and AT1 receptor activation further increases ROS production ^15,16 leading to GI-ARS-related injury, we reasoned that targeting ROS with an antioxidant at the AT1 receptor would significantly reduce ROS generation and would be an effective strategy to reduce or reverse the consequences of IR. A mechanism of the stabilized tempol nitroxide radical is shown in Fig. 2A. Tempol catalyzes the conversion of superoxide O₂^.- to hydrogen peroxide (H₂O₂) and catalyzes conversion of H₂O₂ to O₂ and H₂O by a catalase mimetic action. Telmisartan functions as a selective antagonist of AT1R by blocking AT1R-mediated pro-inflammatory and oxidative signaling pathways. To this end, we designed and synthesized a series of novel, highly potent tempol-conjugated angiotensin receptor blockers (TCARBs, supporting information), which selectively bind to the AT1R while retaining the antioxidant properties of tempol (Table 1). YK-4–250 (Fig. 2B) exhibited superior selective and potent inhibition of AT1 over AT2 compared to the amide-linked compound PLJ-1–43 and other anaglogs (Table 1). Moreover, it displayed comparable potency (AT1R IC₅₀ = 1 nM) and maintained selectivity (reduced AT2R inhibition) as compared to the clinically used telmisartan (Table 1), suggesting that the addition of tempol did not negatively impact binding affinity or selectivity.

Fig. 2.

Mechanism and design of tempol-telmisartan conjugate compound YK-4–250. (A) Mechanism of the stabilized tempol nitroxide radical; (B) Design of YK-4–250.

Table 1.

YK-4–250 selectively inhibits Ang II binding to AT1 over AT2.

Computational docking was performed to assess whether conjugating tempol to the telmisartan scaffold alters binding within the AT1R. Both inactive (PDB 4YAY) ³⁰ and active (PDB 6OS2) ³¹ AT1R structures were analyzed to compare residue engagement, binding geometry, and predicted affinity between telmisartan and YK-4–250.

When docked to the inactive AT1R (PDB 4YAY), telmisartan was found to have a dock score of -11.5 kcal/mol (Table 2) and was predicted to interact with Arg 167. The carbonyl oxygen was predicted to form two hydrogen bonds with Arg 167 at distances of 3.130 Å and 2.249 Å. YK-4–250 demonstrated a slightly improved docking score of -12.6 kcal/mol. Like telmisartan, YK-4–250 was also predicted to interact with Arg 167. The ester oxygen of YK-4–250 engaged Arg 167 via two hydrogen bonds, 3.475 Å and 2.443 Å. In additional, the radical O of YK-4–250 uniquely interacts with Gln 267 at a distance of 4.612 Å. In 4YAY structure, YK-4–250 was predicted to overlap spatially with telmisartan.

Table 2.

Dock Scores of Telmisartan and YK-4–250 in the inactive and active AT₁R.

Compound	AT1R PDB	Dock Score
Telmisartan	4YAY	-11.5
YK-4–250	4YAY	-12.6
Telmisartan	6OS2	-11.3
YK-4–250	6OS2	-11.8

In the active AT1R (PDB 6OS2), telmisartan displayed a docking score of -11.3 kcal/mol (Table 2). As seen with telmisartan docked within the inactive model, it was predicted to interact with Arg 167 in the active AT1R. The carbonyl oxygen of telmisartan was predicted to interact with Arg 167 at distances of 1.808 Å and 3.143 Å, respectively. Additionally, the tertiary amine of the distal benzimidazole forms a hydrogen bond with Arg 167 at 2.283 Å. The dock score of YK-4–250 is -11.8 kcal/mol (Table 2). YK-4–250 was also predicted to interact with Arg 167, but additional interactions were depicted with Arg 23, Tyr 92, and Lys 199. The tertiary amine of the distal benzimidazole was predicted to form a 2.724 Å hydrogen bond with Arg 23. The tertiary amine of the central benzimidazole was found to form a hydrogen bond with Tyr 92 at a distance of 2.308 Å. The carbonyl oxygens formed two hydrogen bonds with Arg 167, 2.203 Å and 2.975 Å. Furthermore, the ester oxygen of YK-4–250 engaged with Lys 199 through a hydrogen bond at a distance of 2.412 Å. Unlike in the inactive AT1R, the radical O of YK-4–250 was not predicted to engage in any hydrogen bonds with the active AT1R. Overall, these docking simulations demonstrate that the conjugation of tempol does not disrupt crucial receptor interactions. However, verifying that the linked nitroxide radical remains fully functional as a catalytic antioxidant was essential to confirm the molecule’s therapeutic potential (Fig. 3).

Fig. 3.

Molecule docking model of telmisartan and YK-4–250 to AT1R. (A) Predicted interaction between telmisartan (purple) in an inactive AT1R (PDB: 4YAY). (B) Predicted interaction between YK-4–250 (cyan) in an inactive AT1R (PDB: 4YAY). (C) Predicted interaction between telmisartan (purple) in an active AT1R (PDB: 6OS2). (D) Predicted interaction between YK-4–250 (cyan) in an active AT1R (PDB: 6OS2).

YK-4–250 is a catalytic antioxidant targeted to the AT1R

Given the importance of the stabilized catalytic nitroxide antioxidant component of tempol ³², we sought to determine whether YK-4–250 retained potent antioxidant activity. YK-4–250, telmisartan, and tempol were evaluated at a final concentration of 50 µM inhibition of cellular ROS following stimulation of ROS production in Caco-2 cells. YK-4–250 demonstrated potent antioxidant activity equal to that of tempol, while the ARB telmisartan did not have direct antioxidant activity (Fig. 4A). This data establishes YK-4–250 as the first highly specific ARB that is functionally designed to deliver a potent catalytic antioxidant.

Fig. 4.

YK-4–250 retains the antioxidant properties of tempol and demonstrates increase ACE2 expression. (A) Intracellular ROS levels in Caco-2 cells were measured following treatment with YK-4–250, telmisartan, or tempol (50 µM, 2 h). Data are presented as Mean ± SEM of N = 3 independent biological experiments. Cells were stimulated with tert-butyl hydroperoxide (100 µM, 30 min) to induce oxidative stress, and ROS generation was quantified using the Abcam Cellular ROS Assay Kit (Deep Red, ab186029) with fluorescence detection at 660–720 nm. (B) A549 cells were cultured for 24 h before treatments then continued culturing for an additional 24 h before total RNA isolation. ACE2 qRT-PCR was performed and compared to vehicle-only control.

Increasing evidence suggests that ARB treatment results in the up-regulation of the antioxidant, anti-inflammatory dipeptidase ACE2 ²¹. Since ACE2 is involved in the enzymatic digestion of Ang II into Ang 1–7, we sought to determine whether YK-4–250 modulated ACE2 expression levels. The induction of ACE2 mRNA was characterized in A549 human lung adenocarcinoma cells due to its well-documented and sensitive RAS-signaling components, which provided a robust platform for the initial characterization of YK-4–250’s transcriptional effects. We observed that increasing concentrations of YK-4–250 directly correlated with increased ACE2 mRNA expression in a dose-dependent manner (Fig. 4B). Furthermore, at 1 μM concentration of YK-4–250, we observed a sixfold increase in ACE2 mRNA expression. Interestingly at 100 nM concentration of YK-4–250, while not significant, we observed a two-fold increase in ACE2 mRNA levels demonstrating that YK-4–250 can induce ACE2 expression in the nanomolar range consistent with specific AT1 inhibition ^20,21.

Although tempol has not been reported as a significant radiation mitigator, it has been reported as a radiation protectant ³³; however, its clinical utility is limited by an extremely short half-life (2–5 min) ^34–36. Furthermore, its effectiveness requires dosing in the high micromolar to millimolar range ^37–40, which has significantly hindered its clinical development. To enhance tempol’s bioavailability and duration of action, we conjugated it to telmisartan, a long-acting compound with a 24-h half-life ⁴¹. Having established that YK-4–250 successfully combines potent, targeted antioxidant activity with an optimized scaffold designed for prolonged systemic exposure, we next transitioned to in vivo studies to evaluate its safety and tolerability profile following oral dosing.

YK-4–250 is well tolerated after oral administration

Since the GI tract is the major site of injury following high dose IR, YK-4–250 was administered orally. We investigated the oral toxicity of YK-4–250 in adult mice. Following administration of YK-4–250 (100 mg/kg), animals exhibited normal behavior both in the short term (first 4 h after administration) and over 17 days. We did not observe any significant weight loss or gain at this dose (Fig. 5). Furthermore, during the 100 mg/kg safety study, no changes in spontaneous locomotor activity or grooming behavior were noted in the treatment group compared to naïve controls.

Fig. 5.

Effects of YK-4–250 on mouse body weight. Male C57BL/6 mice (N = 5) were administered YK-4–250 (100 mg/kg, PO) or Vehicle. During the study, mice were weighed at (0 (prior to first dose), 3, 7, 14, and 17 d) to monitor body weight change. Data are presented as mean ± SEM.

We next evaluated the effects of YK-4–250 for bone marrow related toxicity. We did not observe any significant changes in complete blood count (CBC) (WBCs, neutrophils, lymphocytes, monocytes, RBCs, and platelets) at 100 mg/kg (Fig. 6). Similarly, we did not observe any significant effects on mouse serum chemistry at the 100 mg/kg dose. At the conclusion of the study, a gross necropsy was performed and there were no abnormalities noted in any of the major organs (intestines, liver, kidneys, spleen, and heart).

Fig. 6.

Effects of YK-4–250 on complete blood count (CBC). During the study, 20 μL blood was collected from the mouse submandibular vein for CBC analysis on days the body weight of the animals (N = 5) were recorded.

Concentrations of blood urea nitrogen (BUN), creatinine, total protein and glucose (Fig. 7) were measured in serum of irradiated and age-matched naïve mice to further analyze kidney damage. No significant changes were observed for any group (p ≥ 0.05 by Tukey’s multiple comparisons test between the three groups). These preliminary observations suggest a favorable safety profile at 100 mg/kg, supporting its use in short-term efficacy models; however, formal GLP-toxicology in larger cohorts remains necessary to establish a definitive regulatory NOAEL.

Fig. 7.

Effects of YK-4–250 on serum chemistry. Concentrations of renal and hepatic enzymes measured in serum of naïve and vehicle (5% DMSO-0.5% Tween 80 in sterile water, 0.2 mL) or YK-4–250 (100 mg/kg) N = 5. (A) Blood urea nitrogen (BUN), (B) Creatinine, (C) Total renal protein, (D) Glucose, (E) Alanine transaminase (ALT), (F) Aspartate transaminase (AST), (G) Alkaline phosphatase (ALP).

Further analysis of liver injury was completed by measuring concentrations of the liver enzymes alanine transaminase, ALT (Fig. 7E), aspartate transaminase, AST (Fig. 7F) and alkaline phosphatase, ALP (Fig. 7G). Increased concentrations of all enzymes were observed in serum of irradiated mice compared to naïve. While ALT showed a slight upward trend, the lack of concomitant AST/ALP elevation or gross necropsy findings suggests the absence of acute hepatotoxicity in this pilot cohort.

Since YK-4–250 will be delivered to the GI tract, we evaluated the expression patterns of the AT1R in the mouse intestinal tract. The high tolerability of oral YK-4–250, combined with the immunohistochemical localization of AT1R on intestinal enterocytes and tuft cells (Fig. 8), suggests that YK-4–250 is well-positioned to modulate the local RAS-driven inflammatory environment directly at the site of radiation injury.We observed immunoreactive AT1R on gut enterocytes including the chemosensory tuft cell (Fig. 8). Systematic quantification of ATR + /Cox1 + colocalization across five representative high-power fields, totaling 1,300 epithelial cells (as shown in the Supplemental Table and Dot Plot (Fig. S2)), shows the double-positive population consistently represents 1.53% of the epithelium, aligning with the recognized physiological frequency of intestinal tuft cells. With a Pearson’s Correlation Coefficient of 0.81, these data provide statistically robust evidence that these rare cells are a distinct and reproducible feature of the treated tissue.

Fig. 8.

Expression of AT1R on gut enterocytes including the chemosensory tuft cell. Tuft cells co-expressing Cox1 and AT₁R in jejunal tissue. Immunofluorescence staining was performed using antibodies against Cox1 (green) and AT₁R (red). Arrows indicate regions of colocalization between Cox1 and AT₁R, identifying tuft cells. Images were captured at 20Χ magnification. Resepresentative across five representative high-power fields (HPFs) are provided in the supplemental data.

Indeed, tuft cells represent a unique epithelial cell type that can sense luminal contents and secrete paracrine and autocrine factors that play a role in the gut epithelial defense against pro-inflammatory insults ^42,43. These data suggest that Ang II//AT1 signaling in the GI tract following lethal IR may be targeted by YK-4–250.With a clear pilot safety profile, we next transitioned to in vivo survival studies to evaluate the efficacy of YK-4–250 as a countermeasure against lethal radiation injury.

YK-4–250 protects and mitigates radiation damage at LD₅₀ dose of PBI and improves overall survival

In the PBI survival efficacy studies, male C57BL/6 mice were weighed and randomly assigned into groups (n = 32/group in prophylactic study and n = 24/group in mitigation study). The prophylactic survival study where the mice received either vehicle or YK-4–250 (20 mg/kg/dose in 0.2 mL) at -24 h, -1 h, and + 24 h relative to 14.6 Gy BM2.5-PBI exposure (Fig. 9A), YK-4–250 significantly improved survival with 26 of 32 mice (81%) alive on day 30 compared to vehicle group with 18 of 32 (56%) surviving at Day 30 (Fig. 9B). The survival curves diverged early in the observation period with YK-4–250 treated group maintaining higher cumulative survival throughout the 30-day study. Comparison of the survival distributions via the Log-rank (Mantel-Cox) test confirmed a statistically significant difference between the treatment groups (p = 0.026). Statistical analysis using a constant hazard model revealed that YK-4–250 treatment significantly mitigated the risk of radiation-induced mortality with a hazard ratio (HR) of 2.75 indicating that the vehicle group faced nearly three times the risk of death compared to those receiving YK-4–250.

Fig. 9.

YK-4–250 oral dose mitigates 14.3 lethal Gy PBI. (A) Study schematic detailing hours of drug administration for the protection model with 14.6 Gy BM 2.5% spared PBI. (B) 30-day survival efficacy of animals administered with either vehicle or YK-4–250 as a 3-dose regimen (20 mg/kg) -24, -1, + 24 h post 14.6 Gy PBI. (C) Study schematic detailing hours of drug administration and blood collection following 14.3 Gy BM 2.5% spared PBI. (D) 30-day survival efficacy of animals administered with either vehicle (5% DMSO-0.5% Tween 80 in sterile water, 0.2 mL) or YK-4–250 as a 3-dose regimen (20 mg/kg/ dose in 0.2 mL) + 1, + 2, + 3 days post 14.3 Gy BM2.5-PBI. The animals reported in the Kaplan–Meier survival analysis (Fig. 9) were not utilized for tissue collection or invasive procedures, but were monitored solely for mortality and morbidity throughout the duration of the study.

In the mitigation study, following PBI at 14.3 Gy dose of radiation, mice were administered either vehicle (5% DMSO and 0.5% Tween 80 in sterile water, 0.2 mL) or YK-4–250 (20 mg/kg/dose in 0.2 mL) orally on Days 1, 2, and 3 post-PBI (± 0.5 h) (Fig. 9C). At the end of the 30-day observation period, 14 of 24 mice (58%) in the vehicle group survived, compared to 21 of 24 mice (88%) treated with YK-4–250 (Fig. 9D). Most mortality occurred within the GI death window (days 6–12 post-PBI). One additional death occurred on day 14 in the vehicle group, whereas three deaths in the YK-4–250 group occurred on days 7 and 8 post-PBI. As a mitigator, YK-4–250 also showed a robust survival benefit over the vehicle group with the Log-rank (Mantel-Cox) p value of 0.033 and a HR of 4.26, indicating more than 4 times higher chances of survival with YK-4–250 treatment. Even in the protection model, the HR of 2.75 represents a substantial reduction in the risk of lethal gastrointestinal failure, though the absolute radiation doses for mitigation model (14.3 Gy) and prophylactic model (14.6 Gy) were slightly different, the resultant lethality was similar in the vehicle groups (58% and 56%, respectively). Taken together, these data strongly indicate that orally administered YK-4–250 (20 mg/kg) given once daily for 3 days, is a powerful prophylactic and mitigating countermeasure of GI-ARS. Given that the prominent survival benefit of YK-4–250 occurred during the peak period of radiation-induced gastrointestinal toxicity, we next investigated its specific protective effects on the structural integrity of the intestinal mucosa.

YK-4–250 mitigates GI mucosal damage

To investigate the mechanisms underlying the survival benefit conferred by YK-4–250, GI tissues were examined at 4 and 7 days post-PBI. Male C57BL/6 mice were exposed to 14.3 Gy BM2.5-PBI and administered either vehicle or YK-4–250 on Days 1, 2, and 3 post-irradiation. Jejunum samples were collected from naïve and irradiated mice on Days 4 and 7 post-PBI, fixed in 10% buffered formalin, and processed for hematoxylin and eosin (H&E) staining (Fig. 10A).

Fig. 10.

YK-4–250 improves mucosal score and mitigates early DNA damage and induces late-stage DNA repair. (A) Representative H&E-stained cross-sections of jejunum harvested on days 4 and 7 post-14.3 Gy BM2.5-PBI. Mice were left unirradiated (naïve) or were exposed to radiation and treated orally on days 1, 2, and 3 post-PBI with either vehicle (5% DMSO and 0.5% Tween 80 in sterile water, 0.2 mL) or YK-4–250 (20 mg/kg/dose in 0.2 mL). (B) Mucosal score of damage to villi of collected jejunum. (C) Measured length of villi of collected jejunum. D Number of viable crypts counted in collected jejunum. (E) Immunohistochemical staining for γH2AX, 4HNE and ACE2 in jejunum from vehicle or YK-4–250 administered irradiated animals collected on days 4 and 7 post-PBI. (F) 16 s rDNA assay showing bacterial translocation in naïve and irradiated mice groups treated with vehicle or YK-4–250. (G) Measurement of gut barrier function via concentration of ingested FITC-Dextran in the blood stream of naïve and irradiated mice. Note the concentrated γH2AX immunoreactivity within the regenerative crypt compartments at Day 7 (10E.), correlating with the observed restoration of villous height (10C). Quantitative image analysis was performed and tabulated on γH2AX, 4-HNE and ACE2 expression on day 4 and day 7. The day 4 histology number of animals (N) = 5 per group, the day 7 histology N = 5 per group and the survival study N = 5 per group.

Histological damage was quantified based on villus mucosal injury, villus length, and the number of viable crypts. Mucosal injury was scored on a 0–5 scale: 0 indicated no detectable damage; 1, minimal loss of cellularity; 2, an opening at the villus apex; 3, apex opening with denudation; 4, loss of most cellularity; and 5, complete villus collapse ^44,45.

Villi from Vehicle-treated mice consistently exhibited higher mucosal injury scores than those from YK-4–250 treated mice at both day 4 (two-way ANOVA, p = 0.0011) and Day 7 (p = 0.0122) post-BM2.5-PBI (Fig. 10B). All irradiated groups showed significantly shorter villi compared with naïve controls at both time points (Fig. 10C). Examination of the crypt size indicated a shrinkage in the number of viable crypts compared to naïve, but between the two irradiated groups, no significant difference with respect to viable crypt counts on day 4. (Fig. 10D). By Day 7, however, YK-4–250 treated mice demonstrated marked crypt recovery, with significantly more viable crypts than Vehicle-treated animals (two-way ANOVA, p = 0.0001) suggesting that YK-4–250 was acting in crypt recovery from radiation.

A reduction in villus length referred to as “Villus blunting”⁴⁶ was observed within both irradiated groups as the villi were significantly shorter compared to naïve on day 4. Interestingly, the YK-4–250 mitigating effects were not related to villi length as there was no significant difference between YK-4–250 and vehicle-control. By day 7 the villi length seems to have recovered in both irradiated groups. These structural improvements indicate that YK-4–250 actively drives the physical repair of the irradiated gut lining; therefore, we next analyzed the underlying cellular dynamics to understand how the compound prevents epithelial barrier failure.

YK-4–250 decreases bacterial translocation

One of the major complications of GI-ARS is bacterial translocation from the gut due to the gut barrier being compromised ^47,48. To determine if YK-4–250 mitigated this aspect of GI-ARS, we examined mice exposed to 14.3 Gy PBI with or without YK-4–250 and evaluated the gut barrier integrity as well as bacterial translocation. Bacterial load on the liver and spleen was estimated in irradiated animals on days 4 and 7 post-PBI and compared to the naïve group (Fig. 10E). A bacterial load in naïve jejunum was estimated as a reference. Bacteria in the liver measured significantly higher in the vehicle administered group when compared to YK-4–250 administered group on day 4 (2way ANOVA, p = 0.0022). No significance was calculated between irradiated groups for the spleen on Day 4, or for both liver and spleen on day 7. Damage to the gut-barrier function was measured via analysis of ingested FITC-dextran found in the blood stream of naïve and irradiated animals on Days 4 and 7 post-PBI (Fig. 10F). Comparable concentrations of FITC-dextran were measured in both irradiated groups on Day 4; however, YK-4–250 administered animals measured lower concentration compared to vehicle administered mice by day 7. These physiological improvements in barrier function suggest a high level of survival among the regenerating epithelial cells; therefore, we next evaluated the molecular impact of YK-4–250 on mitigating double-strand DNA breaks and facilitating rapid DNA repair following ionizing radiation.

YK-4–250 reduces DNA damage and enhances DNA repair following IR

Our results showing improved crypt survival with YK-4–250 treatment and YK-4–250 antioxidant properties suggested that the ROS produced by IR^11,13 might be responsible for DNA damage leading to decreased crypt survival (Fig. 10G). To investigate this hypothesis, we examined jejunum tissues stained for histone γH2AX, a marker of early DNA damage as well as later DNA repair processes ^49,50. γH2AX is an acute sensor of a particularly harmful type of DNA damage called a double-strand break. When such damage occurs, hundreds of γH2AX molecules near the break site are phosphorylated within minutes, flagging the lesions, and recruiting repair proteins to the scene ^14,49,50. Interestingly, following IHC staining of jejunum for γH2AX, we found that the YK-4–250 treated animals exhibited less DNA damage at day 4 than vehicle treated animals. Conversely, at day 7 the YK-4–250 treated group showed more γH2AX staining than the vehicle group indicating active DNA repair. The increased γH2AX signal at day 7 could represent enhanced DDR signaling and active recruitment of repair machinery, rather than completed repair. The concentrated γH2AX immunoreactivity observed within the intestinal crypts on day 7 constitutes regenerative foci, representing localized areas of high proliferative activity dedicated to the structural restoration of the intestinal architecture. Taken together, these results suggest a two-fold DNA-damage-repair mechanism where YK-4–250 treatment results in reduced initial DNA damage and enhanced DNA repair when compared to controls. While these findings demonstrate a clear preservation of genomic integrity in the regenerating crypts, we next investigated whether this survival is accompanied by a systemic decrease in oxidative stress markers and a corresponding increase in functional membrane ACE2 expression within the target tissue.

YK-4–250 decreases in vivo oxidative stress and increases membrane ACE2 protein levels

Lastly, damage to the GI system of mice was quantified via immunohistochemical staining for the biomarkers 4-HNE and ACE2 (Fig. 10G). Our results demonstrate that YK-4–250 induces ACE2 expression within the mouse intestinal epithelium following radiation provides organ specific validation and confirms our in vitro findings in A549 cells (Fig. 4B).

We investigated the peroxidation of lipids using an antibody against 4-hydroxynonenal (4-HNE), a marker for oxidative stress ^51,52. Within the non-irradiated group, 4-HNE was observed heterogeneously throughout the crypts but was mostly concentrated in the lower half of the crypt. After 4-days post-PBI, we observed a markedly increased 4-HNE that was mostly present in regenerative crypts. These crypts appeared hyperplastic, furthermore there were fewer crypts compared to non-IR controls as evidence of pathological injury. In contrast, YK-4–250 treated animals’ GI showed less 4-HNE-staining that was more dispersed throughout the regenerative crypts, which did not appear hyperplastic. By day 7 post-PBI, both vehicle-control and YK-4–250 showed 4-HNE-staining similar to unirradiated mice suggesting that ROS levels had returned to normal. Overall, these findings indicate that the antioxidant activity of YK-4–250 mitigates ROS-induced crypt damage, thereby promoting crypt recovery and survival. Reducing ROS levels within 24–48 h after irradiation may play a critical role in improving overall survival. Following this successful in vivo validation of YK-4–250’s therapeutic efficacy, we next sought to exploit the inherent magnetic properties of its conjugated tempol group to determine if the molecule could be directly detected and imaged within biological systems.

YK-4–250 can be detected by electron paramagnetic resonance (EPR) and imaged by magnetic resonance imaging (MRI)

Because of the paramagnetic properties of its stabilized nitroxide, Tempol can be imaged in serum by electron paramagnetic resonance (EPR) ⁵³ and noninvasively in tissues by MRI ⁵⁴. To determine whether YK-4–250 could likewise be imaged in real-time, we gave spontaneously hypertensive male rats (SHR) orally YK-4–250 (100 μmoL/kg). Using EPR, we were able to identify YK-4–250 in the serum (Fig. 11A) and it was detectable up to 60 min in serum suggesting that this drug can be found in the blood stream for at least 1 h after oral dosing (Fig. 11B). This is in contrast to tempol that has a reported half-life of 15 s ³⁶ to as much as 15 min ⁵⁵ depending on dose in the blood stream. This result suggests that the ester bond between telmisartan and tempol in YK-4–250 is stable and enhances tempol’s catalytic antioxidant half-life.

Fig. 11.

YK-4–250 is stable in vivo and can be imaged by EPR and MRI. Spontaneously hypertensive rats (SHR) given a 1 mL gavage with a 100 µmoL/kg dose of YK-4–250 in PEG400 were used for serum sampling at 1, 5, 10, 15, 30, 45, and 60 min for each animal. (A) Electron paramagnetic resonance (EPR) measurements of the blood samples were measured using a Magnettech MS-300 and analyzed via Analysis 2.02 software. The EPR experiments were conducted using 100 umol/kg (N = 2) independent biological replicates/blood samples over 8 timepoints. (B) Graph of EPR intensity over time and showing that signal was still detected after 1 h. Given the stability of YK-4–250 MRI experiments were performed on immobilized mice placed in a 72-mm quadrature volume coil for signal transmission. Six mice were imaged (3 YK-4–250, 3 Control (vehicle). (C–D) Representative images of (C) vehicle-only control and (D) YK-4–250 mouse MRI demonstrating the real-time observation of YK-4–250 via MRI within the mouse.

Giving these findings, we evaluated orally administered YK-4–250 (20 mg/kg) for in vivo GI distribution by MRI in adult C57BL/6 mice as compared to vehicle control (Figs. 11C and D). T1 weighted sagittal magnetic resonance imaging (MRI) images were taken 15 min after the oral delivery of 20 mg/kg of YK-4–250, and we observed strong signal of YK-4–250 distribution throughout the GI tract (Fig. 11D) demonstrating significant uptake into the gut, whereas no signal was detected in the vehicle-only control (Fig. 11C). These results show that YK-4–250 will allow for pharmacodynamics determination as well as drug: tissue exposure levels in the future experiments. In summary, YK-4–250 represents a novel class of radiation mitigators of GI-ARS, capable of real-time pharmacodynamic monitoring, damage assessment, and radiation damage reversal, providing a distinct clinical advantage.

Discussion

While the FDA has approved the use of Neupogen (filgrastim) to treat adult and pediatric patients acutely exposed to myelosuppressive doses of radiation resulting in hematopoietic syndrome of acute radiation syndrome, or H-ARS, no drugs are approved as a radiation mitigator of GI-ARS ⁵⁶. We have developed a novel therapeutic strategy that delivers the catalytic antioxidant tempol specifically to gut cells that express the AT1R. We hypothesize that targeting two critical pathways responsible for early IR dependent generation of ROS by quenching DNA damaging free radicals and elevations of Ang II/ AT1 prevents the pro-inflammatory, pro-fibrotic tissue destructive complications GI-ARS.

In both the active and inactive AT1R, YK-4–250 was predicted have more stable dock score, -12.6 kcal/mol and -11.8 kcal/mol, respectively (Table 2). Thus, YK-4–250 was predicted to have a greater binding affinity to the AT1R than telmisartan. In the inactive AT1R, the interactions with YK-4–250 and Arg 167 were at the ester oxygen as opposed to the carbonyl oxygen in telmisartan. These hydrogen bond between YK-4–250 and Arg 167 are also predicted to be longer in length than those of telmisartan. The addition of the tempol also allowed a unique interaction with Gln 267. This suggests that the addition of tempol to telmisartan modifies how YK-4–250 interacts with the inactive AT1R.

Likewise, YK-4–250 was shown to have modified interactions with the guanidino group of Arg 167 in the active AT1R when compared to telmisartan. Unlike telmisartan, YK-4–250 forms a shorter hydrogen bond with the imine of guanidino group of Arg 167 and a longer hydrogen bond with an amine of Arg 167. Although the radical oxygen of YK-4–250 is not predicted to engage with the active model of AT1R, the additional interactions between YK-4–250 and Arg 23, Tyr 92, and Lys 199 of the active AT1R improve its stability. Previously ⁵⁷, it was shown that telmisartan engages in a delta lock conformation that may contribute to its improved stability with AT1R when compared to other ARBs ⁵⁷. This conformation of telmisartan was maintained in this docking study. Furthermore, YK-4–250 also maintains this conformation, but the addition of the radical modifies this shape. This modification in conformation may also improve its binding affinity to the AT1R.

YK-4–250 demonstrated similar selectivity (AT1 over AT2) and potency (AT1 IC₅₀ = 1 nM) as compared to telmisartan. These data validate the molecular modeling and demonstrate that the conjugation of tempol to telmisartan did not alter the selectivity and potency of YK-4–250.

The survival efficacy studies at LD₅₀/₃₀ doses of radiation demonstrate that YK-4–250 serves as a potent dual-mode countermeasure against GI-ARS, providing significant benefits in both prophylaxis (protection) and post-exposure (mitigation) scenarios. Our survival data, characterized by a shift from 56 to 81% in the protection model and 58% to 88% in the mitigation model underscores the robust efficacy of YK-4–250 across different therapeutic windows. The statistical strength of these findings is evidenced by the Log-rank p-value of 0.026 and 0.033, for prophylactic and mitigation, respectively, which confirms that the observed survival advantages are not due to stochastic variation. More strikingly, the hazard ratio (HR) of 4.26 in the mitigation study indicates that vehicle-treated subjects faced more than a four-fold increase in the risk of mortality compared to those receiving YK-4–250. Even in the protection model, the HR of 2.75 represents a substantial reduction in the risk of lethal gastrointestinal failure, though the absolute radiation doses for mitigation model (14.3 Gy) and prophylactic model (14.6 Gy) were slightly different, the resultant lethality was similar in the vehicle groups (58% and 56%, respectively). YK-4–250 is a powerful new radiation mitigator of GI injury by improving GI barrier integrity and function. With the ability to image by EPR and MRI, YK-4–250 represents a novel imageable radiation mitigator that has the potential “in real time” to monitor drug delivery, assess damage and reverse radiation damage providing a distinct clinical advantage.

Our findings underscore the dual mechanism of action of YK-4–250, which targets both Ang II and ROS, leading to a significant improvement in survival following 14.3 Gy PBI. Our observations of changes in γH2AX expression might mark the activation of the DDR, and its positive correlation with structural recovery (Fig. 10A) substantiates its role in the repair process in this model. Quantitative immunohistochemical (IHC) analysis of γH2AX (a marker of double-strand DNA breaks and active DNA damage response) revealed a robust, time-dependent increase in positive staining in the YK-4–250 treatment group compared to the vehicle control. At day 4, the YK-4–250 group exhibited an approximate 1.9-fold increase in DNA damage signaling, with the percentage of positive area rising from 8.2% in the Vehicle group to 15.6% in the YK-4–250 group. By day 7, this divergence widened further. The positive stained area in the Vehicle group rose modestly to 12.5%, whereas the YK-4–250 group demonstrated a marked elevation to 28.4% (representing a 2.3-fold increase over the corresponding control). These data indicate that YK-4–250 strongly triggers or stabilizes DNA damage signaling/repair pathways, a response that continues to amplify through day 7.

Accumulation of 4-HNE, a major end-product of lipid peroxidation, was quantified to assess local oxidative stress and tissue damage. On day 4, the Vehicle group displayed substantial oxidative damage, with a positive stained area of 32.1%. In contrast, treatment with YK-4–250 significantly mitigated this response, reducing the 4-HNE positive area to 14.2% (a 56% reduction in lipid peroxidation). This protective effect was sustained and further enhanced by Day 7. While the Vehicle group remained highly oxidized at 28.5% positive area, the YK-4–250 treated group dropped to a minimal baseline of 9.8% positive area, representing a 66% reduction compared to the day 7 control. YK-4–250 exhibits highly potent, sustained antioxidant properties, effectively shielding lipid membranes from oxidative degradation.

To investigate the modulation of the tissue-protective RAS arm, expression levels of ACE2 were quantified. At day 4, the Vehicle group demonstrated a baseline ACE2 positive area of 18.4%. YK-4–250 treatment resulted in a rapid and dramatic 2.3-fold upregulation of ACE2 expression, reaching 42.5% positive area. The elevated expression of ACE2 was robustly maintained through Day 7. The positive area in the Vehicle control trended downward to 15.2%, while the YK-4–250 group sustained high expression at 45.1% (a 3.0-fold increase over control). YK-4–250 induces a rapid, long-lasting upregulation of ACE2, indicating a targeted activation of the protective RAS pathway to counteract tissue injury. Furthermore, the significant induction of ACE2 validates the compound’s mechanism of action directly within the target intestinal tissue. The significant increase in crypt-localized γH2AX, coupled with the marked reduction in 4-HNE and upregulation of ACE2, provides quantitative evidence of the integrated repair and protective response facilitated by YK-4–250. These data suggest that the structural preservation (villi height/mucosal score) is driven by measurable molecular changes (reduced ROS and increased ACE2). The convergence of high γH2AX (signaling recruitment) and improved villi height (functional outcome) provides definitive evidence that YK-4–250 facilitates a prolonged, productive DDR that effectively mitigates the gastrointestinal syndrome.

In line with our mechanistic insights, we observed notable improvements in crypt numbers, crypt depth, villous height, and submucosal thickness, along with a two-fold increase in jejunal crypt viability. These results clearly demonstrate the effectiveness of YK-4–250 in mitigating jejunal morphological damage. Importantly, YK-4–250 also significantly enhanced gut histology and crypt viability in mice exposed to lethal radiation. MRI data provides pharmacodynamic evidence that YK-4–250 remains intact for more than 15 min, thus by definition extends the oral half-life of tempol (less than 5 min) ³⁴. The high tolerability of oral YK-4–250, combined with the immunohistochemical localization of AT1R on intestinal enterocytes and tuft cells (Fig. 8), suggests that YK-4–250 is well-positioned to modulate the local RAS-driven inflammatory environment directly at the site of radiation injury.

In conclusion, YK-4–250 represents a first-in-class, dual-action therapeutic countermeasure that successfully bridges targeted angiotensin receptor blockade with catalytic radical scavenging to mitigate the devastating effects of GI-ARS. By conjugating the rapid-acting antioxidant tempol to the long-acting AT1R antagonist telmisartan, we have developed a stable hybrid molecule that dramatically extends the systemic half-life of the nitroxide radical, ensures safe oral tolerability, and selectively localizes to crucial protective gut epithelial cells like tuft cells. This targeted localization translates into robust in vivo efficacy, significantly reducing radiation-induced DNA damage, accelerating mucosal and crypt recovery, and preventing lethal bacterial translocation to yield a remarkable survival benefit in models of lethal partial-body irradiation. Combined with its unique paramagnetic properties that allow for real-time tracking via EPR and MRI, YK-4–250 stands out as a highly promising, dual-purpose theranostic agent capable of both protecting against and non-invasively monitoring the mitigation of acute radiation injury.

Materials and methods

Chemistry synthesis

All chemicals and solvents were purchased from commercial suppliers and used as received unless noted otherwise. Purifications were performed by flash chromatography. Liquid chromatography/mass spectrometry (LC/MS) analyses were conducted using Shimadzu LC-20AD pumps and an SPD-20A UV–vis detector. High-resolution mass spectra (HRMS) were recorded on a QSTAR Elite mass spectrometer.

The synthesis of YK-4–250 and other TCARBs use the strategy of EDCI/HOBt coupling strategy and purified by flash chromatography (See supporting information).

Modeling of YK‑4‑250 in a homology model of AT1R

Telmisartan and YK-4–250 were input into the visualization software, UCSF Chimera 1, ⁵⁸ and molecular docking was performed using AutoDock Vina 1.2.1 with an exhaustiveness of 8. Check references2-3 YK-4–250 was drawn using ChemDraw and a 3D model was developed using OpenBabel. Telmisartan (CID 65,999) was used for docking. Both the inactive (PDB 4YAY) ³⁰ and active (PDB 6OS2) ³¹ AT1R were modified to remove extraneous ligands. Both the inactive and active AT1R were treated as rigid for molecular docking. For docking in the inactive AT1R, a grid box with a center of x = -15.4865, y = 11.2222, z = 42.6769, and x = 18.1942, y = 24.5315, and z = 21.5515 in size. In the active AT1R, docking was performed using a grid box with a center of x = -13.5253, y = -17.1707, and z = -65.7823, and x = 22.5432, y = 17.2474, and z = 20.5475 in size. The docking position with the lowest dock score (or lowest binding energy) was chosen for each compound. Dock scores determined by this model were used to predict the binding free energy of a compound.

Superoxide detection in vitro

Intracellular ROS levels were quantified using the cellular ROS assay kit (Deep Red, Abcam, ab186029) following the manufacturer’s protocol, with minor modifications (See supporting information).

Quantitative ACE2 PCR analysis

A549 cells (6 × 10⁶ cells) were cultured for 24 h then treated with YK4-250 at the final concentrations of 0.01 μM, 0.1 μM, 1 μM, and 10 μM (volume ration 1:1000). DMSO alone was added to control groups at 1:1000 ratio. Twenty-four hours later total RNA was harvested with reagent RTIzol and chloroform. Each of the RNA sample concentrations and purities were measured with absorbance at wavelength of 260 nM and 280 nM. Reverse transcription (RT) with SuperScript IV VIlo Master Master Mix (Invitrogen) was used to generate the first strand of cDNA through reverse transcription, 2.5 ng of total RNA of each sample were used per reaction. RT-PCR was performed using Taq Fast Advanced Master Mixture (Thermo Fisher). Using the program: 50 °C for 2 min. 95 °C for 2 m one cycle, then 95 °C for 1 s, and 60 °C for 20 s, 40 cycles. ACE2 was amplified with primer [TCCATTGGTCTTCTGTCACCCG]. Relative Quantification function of Design & Analysis Software 2.5.0 was utilized to analyze mRNA level of ACE2 normalized by the internal control GAPDH.

AT1R displacement assay

The AT1R displacement assay was performed using standard radioligand-binding methods (See supporting information).

Animal and veterinary care

All procedures involving animals were reviewed and approved by the Armed Forces Radiobiology Research Institute’s (AFRRI) Institutional Animal Care and Use Committee (IACUC) using the principles outlined in the National Research Council’s Guide for the Care and Use of Laboratory Animals and in accordance with relevant regulations and guidelines. Mice that showed an inability to remain upright, were cold, unresponsive or showed decreased or labored respiration were considered moribund and euthanized according to the American Veterinary Medical Association (AVMA) guidelines. Animal studies were conducted in compliance with Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines.

Male C57BL/6 mice (12–14 weeks) were purchased from Jackson Laboratories (Bar Harbor, ME). Animals were housed in the AFRRI’s vivarium in plastic cages in Allentown NexGen cage systems. The animals were provided with Harlan Teklad Rodent Diet 8604 (Envigo) and acidified water (pH 2.5–3.0). The cages and the room were kept at a temperature of 20–26 °C, humidity between 30–70%, and a 12:12 h light:dark cycle ⁵⁹. Veterinary care was available throughout the study; animals were examined by research and veterinary staff for clinical signs or changes in appearance ⁶⁰.

YK-4–250 was prepared as a suspension consisting of 5% DMSO and 0.5% Tween-80 in sterile water. The animals were administered either drug or vehicle by oral gavage (PO) as a 3-dose regimen (days 1–3 post-PBI).

BM2.5-PBI irradiation and dosimetry

Mice were placed under anesthesia (3% isoflurane, 97% O₂) then irradiated using a 4 MV photon beam from an Elektra Infinity clinical linear accelerator (LINAC) using the BM2.5-PBI model as previously described ⁶¹ to achieve 2.5% bone marrow sparing, one hind leg was excluded from the radiation field. Prior to irradiation, dosimetry was confirmed ⁶¹ and the beam output was verified through ion chamber measurements (PTW, model 30,013). Animals were irradiated at the most recent LD_50/30 dose (14.6 Gy in prophylactic studies and 14.3 Gy for mitigation studies) at an estimated dose rate of 2.8 Gy/min. Post irradiation, mice were returned to the vivarium and closely monitored for their health status throughout the course of the study. Each study included mice (age-matched naïve) that did not receive irradiation as a control group.

Safety study in C57BL/6 mice

A preliminary safety evaluation of YK-4–250 was conducted in male C57BL/6 mice. Animals were randomly divided into two groups (n = 5 per group) and administered either YK-4–250 (100 mg/kg, PO) or vehicle (5% DMSO and 0.5% Tween 80 in sterile water) once daily for three consecutive days. Age-matched naïve mice served as untreated controls. Following each oral dose, mice were observed for acute signs of toxicity (within 1–4 h post-administration) and subsequently monitored daily for 14 days after the final dose, for a total observation period of 17 days. Clinical signs including decreased activity, hunching, labored breathing, squinting eyes, and mortality were recorded. Observations were made during the first and fourth hours after dosing on each treatment day and continued daily throughout the post-dosing period. Body weights were measured at multiple time points, prior to the first dose (day 0) and on Days 3, 7, 14, and 17, to assess potential treatment-related changes in growth or general health status.

Survival efficacy and GI tissue recovery studies in C57BL/6 mice

A thirty-day survival efficacy study was conducted and consisted of testing one drug dose (20 mg/kg in 0.2 ml) of YK-4–250, one route of administration (PO), and radiation at most recent LD_50/30 dose at AFRRI’s LINAC facility at 2.8 Gy/min dose rate. In the pre-PBI studies, mice were administered either vehicle (5% DMSO-0.5% Tween 80-sterile) or YK-4–250 at 24 h prior to irradiation, 1 h prior to irradiation, and 24 h after irradiation. Day 0 was the day of PBI. As the LD₅₀/₃₀ dose during these studies was 14.6 Gy, the two groups of animals (n = 32/group) were irradiated at 14.6 Gy. For the post-PBI studies, the radiation dose was 14.3 Gy, the current LD_50/30. The survival study had two groups of 24 animals, one was administered with YK-4–250 and the other administered with vehicle on day(s) 1, 2, and 3 post-PBI. The animals were monitored for 30 days and euthanized at the end of the observational period.

For the GI tissue recovery studies, C57BL/6 male mice were weighed and distributed into three groups. Group 1 was naïve control (no administration of YK-4–250 and no radiation), Group 2 was administered with the vehicle and Group 3 was administered with YK-4–250 (20 mg/kg in 0.2 ml) on 1, 2 and 3d post-PBI. Group 2 and 3 were irradiated 14.3 Gy at the LINAC facility at AFRRI. Following irradiation, mice were monitored for health status through the completion of the study (up to 7 days). On days of collection all animals (Group 1, 2 and 3) were administered Fluorescein isothiocyanate (FITC)-dextran 4 h prior to collection. Blood and tissue were collected on days 4 and 7.

Histological analysis of GI tissues

At each timepoint, mice were euthanized for tissue collection. Tissues (parotid salivary glands, heart, lungs, kidneys, jejunum, and liver) were collected and either snap frozen or fixed in 10% buffered formalin. Tissues fixed in formalin were sent to Histoserv, Inc. (Germantown, MD) where they were fixed, embedded in paraffin, and stained with hematoxylin and eosin (H&E). Sections of each tissue were analyzed and scored by a board-certified veterinary pathologist. Throughout the study, blood was collected at each timepoint, and serum was separated by centrifugation at 2400xg for 10 min in serum collection tubes. Serum was analyzed for renal and hepatic panels using the HESKA Element DC5X (HESKA Corporation, Loveland, CO).

Stained jejunum sections were evaluated for the number of viable crypts and mucosal damage as indicators of gastrointestinal damage and possible recovery. A viable crypt was defined as around 10 Paneth cells per well-defined crypt ⁶¹. At least 3 sections per mouse were assessed for crypt counts. Mucosal damage was scored following a multi-grade system as follows: score 0, normal mucosa; score 1, subepithelial spaces near the tips of the villi; score 2, extension of the subepithelial space with moderate epithelial lifting from the lamina propria; score 3, significant epithelial lifting along the length of the villi with a few denuded villous tips; score 4, denuded villi with exposed lamina propria and dilated capillaries; score 5, disintegration of the lamina propria, hemorrhage, and ulceration ⁶². Villi length was measured from the base to the tip of each villi ⁶³.

IHC studies

Standard IHC protocols were followed using specific antibodies, as previously described ⁴³ (see supporting information).

Quantification of bacterial load in liver and spleen

Bacterial translocation in the liver and spleen was quantified to determine the gut barrier integrity after BM2.5-PBI. Data was collected by performing real-time PCR using the highly conserved 16S rRNA gene consensus sequence ⁶⁴. The DNA was extracted from frozen liver and spleen then lysed by adding proteinase K and incubating at 65 °C ⁶⁰. DNA was prepared and isolated using the Qiagen DNeasy mini kit then quantified using a nanodrop spectrophotometer. DNA was prepared with SYBR Green Master Mix, PCR water, and primers for 16S rRNA, for quantification of bacterial load.

Intestinal barrier function

On days 4 and 7 of the study, naïve and irradiated mice were administered FITC-dextran (500 mg/kg bodyweight, 0.2 mL) via oral gavage. Mice were euthanized 4 h after administration for blood and tissue collection. Whole blood was collected by cardiac puncture and serum was separated by centrifugation in serum collection tubes. Levels of FITC-dextran in the serum were measured using the CLARIOstar microplate fluorimeter at an excitation wavelength of 485 nm and an emission wavelength of 525 nm. FITC-dextran concentrations were determined by using a standard concentration curve of the known FITC-dextran concentrations ^61,65.

Serum sampling and EPR analysis of YK-4–250

Spontaneously hypertensive male rats (SHR; 275–300 g) were used for serum sampling and EPR analysis for experiments N = 2 and presented as the average mean. Rats received YK-4–250 (100 µmoL/kg) administered orally by gavage in 1 mL of PEG400. Post-dose tail vein blood samples were collected at 1, 5, 10, 15, 30, 45, and 60 min after administration. EPR measurements were performed on whole blood samples using a Magnettech Miniscope MS-300 X-band EPR spectrometer (Berlin, Germany). Each blood sample was collected directly into Fisherbrand 50 µL micropipettes and analyzed within 30 min of collection. Spectra were recorded using the following instrument settings: magnetic field, 335.4 mT; sweep width, 19.90 mT; sweep time, 15 s; modulation amplitude, 0.2 mT; number of samples, 4096; microwave attenuation, 10 dB; gain, 600; temperature, 20 °C. The spectrometer operating frequency ranged from 9.3 to 9.55 GHz. Data was analyzed using Analysis Software version 2.02 (Magnettech). Typical acquisition parameters included a center field of 3354.22 G and a sweep width of 60 G.

Magnetic resonance imaging

MRI experiments were performed on a Bruker BioSpec 7.0 Tesla/30-cm horizontal-bore magnet imaging system. Animals were immobilized by using 1.5%-2.5% isoflurane and 0.8 L/min O₂ and placed in a 72-mm quadrature volume coil for signal transmission, and a surface mouse-head coil was used for signal reception for the head imaging. Six mice were imaged (3 YK-4–250, 3 Control (vehicle). Whole body imaging, a T1-RARE with the following parameters was used: TE = 8 ms, TR = 1500 ms, NA = 4, time of scan 4 min 48 s, 12 coronal slices of 0.9 mm, matrix 256 × 256, FOV = 40 × 35 mm, and a resolution of 0.156 × 0.137 mm. T1map-RARE was also used with the following parameters: TE = 8.50 ms, 6 experiments with TR = 5500, 3000, 1500, 800, 400, 200 ms, NA = 1, time of scan 18 min 19 s, 12 coronal slices of 0.9 mm, matrix 256 × 256, FOV = 40 × 35 mm, and a resolution of 0.156 × 0.137 mm. Images were processed on Paravision software for signal intensities and T1 values.

Statistical analysis

GraphPad Prism 10 was used to analyze collected data and generate plots. Kaplan–Meier plots were used to visualize survival data. Analysis of variance (ANOVA) and unpaired t-tests were used to assess the significant differences between groups. The data reported includes means and standard errors of the mean (SEM). Survival studies and histological studies were conducted in independent cohorts. Animals used for day 4 and day 7 tissue collection were not included in the Kaplan–Meier survival analysis, ensuring the independence of the longitudinal survival data.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization, M.L.B.; resources, M.L.B.; writing-original draft preparation; V.P.K., Y.K., A.B.J and M.L.B; writing-review and editing; V.P.K., Y.K. and M.L.B.; supervision, V.P.K,, Y.K. and M.L.B.; funding acquisition, M.L.B. All authors have read and agreed to the published version of the manuscript.

Declarations

Competing interests

M.L.B is also a co-founder of Trocar Pharmaceuticals Inc. M.L.B and Y.K have filed composition of matter patent (US 9,233,949) on YK-4–250. M.L.B, Y.K, P.C and C.W.H have filed use patents (US application 16/089,794) on technology in this manuscript.