Oral connexin43–peptide–loaded milk extracellular vesicles mitigate lethal radiation injury while preserving tumour radiosensitivity

Abstract

Ionizing radiation is central to cancer therapy, where normal-tissue toxicity limits dose escalation, and is also a major hazard in accidental or mass-casualty exposures, where practical radiomitigators for radiosensitive organs are urgently needed. We investigated orally delivered milk extracellular vesicles (mEVs) loaded with the connexin43 carboxyl-terminal peptide αCT11 (XOlacta) in a murine 14 Gy total-body irradiation model and a syngeneic GL261 glioma radiotherapy model. A single oral XOlacta dose given 1 h after total-body irradiation conferred 42% 30-day survival despite an expected 100% lethality from a 14 Gy challenge, with marked preservation of ileal architecture and femoral bone marrow cellularity, and significant benefit was retained when dosing was delayed to 24 h. Biodistribution studies with fluorescently labeled mEVs showed radiation-enhanced uptake in brain, gut, and bone marrow, consistent with injury-licensed targeting of radiosensitive tissues. In glioma-bearing mice, XOlacta protected normal tissues without diminishing tumour radiosensitivity. These findings identify oral αCT11-loaded mEVs as a potent, non-parenteral radiomitigator that preserves normal-tissue integrity while maintaining tumour response in a therapeutic context, and support connexin-linked mEV-targeting as a promising strategy for both radiotherapy adjuncts and management of high-dose radiation exposures.

1. Introduction

Ionizing radiation is central to curative treatment of many solid tumours but is limited by acute injury to radiosensitive tissues such as the gastrointestinal tract, brain and bone marrow [1,2]. Current radioprotectants, including agents such as amifostine, palifermin and hematopoietic growth factors, can compromise anti-tumour efficacy, require parenteral administration, and mainly protect hematopoietic tissues at lower radiation doses [3]. Milk-derived small extracellular vesicles (mEVs) offer a scalable, cold-chain–independent, orally deliverable platform with emerging utility for peptide delivery and tissue repair [4–6], with potential as an adjunct to clinical radiotherapy and as a mass-casualty countermeasure [7,8].

Connexin43(Cx43) carboxyl-terminal peptides such as αCT1 and αCT11 promote wound-healing, including mitigation of skin radiation injury [9], but their use is limited by topical delivery and rapid plasma degradation [10,11]. To overcome this, an esterified pro-peptide (4OMe-αCT11) permits efficient loading inside mEVs and supports oral delivery in preclinical studies of myocardial ischaemia–reperfusion [6,12]. Here, using a lethal total-body-irradiation model and a syngeneic glioma radiotherapy model, oral αCT11-loaded mEVs(XOlacta) are evaluated as systemic radiomitigators of survival, normal-tissue injury, and tumour response.

2. Materials and methods

All procedures complied with institutional IACUC approvals and applicable federal regulations. Detailed methods are provided in the Supplementary Information.

Briefly, purified bovine mEVs were isolated and characterized (Supplementary Fig. 1) and loaded with 4OMe-αCT11 to generate XOlacta as described previously [4,6]. For Total-Body-Irradiation (TBI) studies, C57BL/6 mice received a single 14-Gy dose, followed by oral gavage with vehicle, or unloaded mEVs, or XOlacta (~10¹¹ mEVs/mouse: ~4.0 ×10¹²/kg body weight) at 1 h or 24 h post-irradiation. Survival, body weight, and histopathology of ileum and femoral bone marrow were assessed over 30 days. For biodistribution, CTDR-labeled mEVs (~10¹¹ mEVs/mouse: 4.0 ×10¹²/kg body weight) were given orally immediately after irradiation, and tissues collected 4 h later.

For tumour studies, flank gliomas (GL261) were established in C57BL/6 mice, as we have described previously [13]. Animals received no treatment, XOlacta alone, radiation alone (two 7 Gy fractions at weeks 2 and 3), or combined radiation plus oral XOlacta. Tumour volume, gut histology, and CTDR-mEV uptake in tumours and normal tissues were evaluated.

3. Results and discussion

3.1. XOlacta mitigates lethal total -body-irradiation (TBI), while preserving gut and bone marrow

A single 14 Gy TBI produced 100% mortality in vehicle- and unloaded mEV-treated mice by day 9, consistent with an LD₁₀₀/₃₀ profile in this model [14]. In contrast, a single oral dose of XOlacta 1 h after irradiation yielded 42% survival at 30 days, with significantly improved Kaplan–Meier survival and recovery of body mass (Fig. 1A–C). Delayed dosing at 24-h post-TBI also conferred a significant survival benefit, though of lesser magnitude. Yet, body mass recovery in the 24 h post-TBI group was similar to that of the 1 h group (Fig. 1C). All non-irradiated control mice survived to study end.

Fig. 1. Effect of XOlacta on mouse survival and tissue integrity following Total-Body-Irradiation (TBI).

A) Experimental schema for testing XOlacta efficacy for total body exposure to a lethal dose (14 Gy) of radiation. B) LD_100/30 Kaplan Meier analysis of group survival over 30 days post-TBI. N = 12 mice/group, using Gehan-Breslow-Wilcoxon test, *p < 0.05, **p < 0.01, ****p < 0.0001. C) Group weight-change over 30 days post-TBI. D) Gut H&E histology - Ileum. E) Histopathological grading of gut tissues post-TBI. N = 3 mice/group, ANOVA with post-hoc testing, mean + S.E.M., ****p < 0.0000001, **p < 0.01. F) H&E histology of femur, including bone marrow (BM) content. G) Histopathological grading of % cellularity of femur BM niches. N = 3 mice/group, ANOVA with post-hoc testing, mean + S.E.M. **p < 0.01. H) Fluorescent CTDR-mEV uptake in irradiated and non-irradiated brain, gut(ileum) and femur bone marrow 4-h post-TBI. I) Quantification of CTDR-mEV particles per nucleus in the same 3 tissues as in (H). n = 6–8 mice/group, Welch’s 2-tailed T-test compared irradiated and non-irradiated tissue pairs, mean + S.E.M, ****p < 0.0001; ***p < 0.0001.

At 30 days, ileal histopathology scores after XOlacta at 1 or 24 h matched non-irradiated controls, with preserved villi, crypts, and mucosal architecture (Fig. 1D and E; Supplementary Fig. 2; Supplementary Table 1). In contrast, vehicle- and unloaded mEV-treated irradiated mice sampled at day 7- prior to universal death by day 9-displayed severe villus blunting, crypt loss, mucosal disruption, and significantly worse composite injury scores. Femoral bone marrow from XOlacta-treated mice remained cellular at levels comparable to non-irradiated controls, whereas irradiated vehicle and mEV groups were profoundly hypocellular (Fig. 1F and G; Supplementary Fig. 2).

These findings identify XOlacta as a potent oral radiomitigator that preserves gut and bone marrow after otherwise lethal exposure. Few candidate countermeasures combine this level of efficacy with oral dosing, delayed administration windows, and compatibility with high-dose radiation settings [15]. The mechanism by which XOlacta mitigates radiation injury remains to be defined, but the ability of αCT11 to increase Cx43 S368 phosphorylation, which is associated with reduced hemichannel opening, represents one potential avenue for future investigation [11].

3.2. Radiation directs oral mEV uptake to radiosensitive tissues

We examined the biodistribution of CTDR-labeled mEVs given orally after 14 Gy TBI. Four hours after irradiation, fluorescent puncta were enriched in the brain, ileum, and bone marrow of irradiated mice relative to sham-irradiated controls (Fig. 1H and I). Quantification showed significantly increased CTDR particle counts per nucleus in brain and marrow, with a similar trend in the gut.

Together with prior work on injury-licensed oral mEV uptake [6] and work on other Cx43 peptides [16], these data are consistent with a model in which radiation injury promotes preferential accumulation of mEVs in damaged radiosensitive tissues, rather than establishing this mechanism directly. In this proposed framework radiation-triggered tissue injury, Cx43 phosphorylation and hemichannel activation, inflammatory signalling, and ATP release may provide plausible hypothetical targets for exploring mEV uptake patterns, as well as the protection of radiation-sensitive tissues observed in this study.

The use of 4OMe-αCT11 peptide as the active component of XOlacta further addresses classic obstacles in peptide radioprotectant development. Multi-site esterification enables efficient loading into mEVs, protects the peptide from gut-luminal and systemic degradation, and promotes sustained local availability after uptake, potentially amplifying Cx43-mediated cytoprotective and pro-regenerative signalling described in cardiac and wound models [6,11].

3.3. XOlacta preserves GL261 glioma tumour radiosensitivity

Systemic XOlacta did not alter GL261 flank tumour growth compared with untreated controls over six weeks, indicating that αCT11-mEVs do not alter baseline tumour growth (Fig. 2A and B). Radiotherapy alone (two 7 Gy fractions) significantly reduced tumour volumes, and combined radiation plus oral XOlacta yielded similarly small tumours with no statistical difference from radiation alone.

Fig. 2. Effect of XOlacta on GL261 tumour growth during radiation therapy.

A) Experimental schema and GL261 tumour growth curves showing that combined radiation therapy (RT = 2 × 7 Gy dose a week apart) plus oral XOlacta treatment suppresses tumour growth compared with XOlacta alone or no treatment over the 6-week study. N = 11–14 mice/group, 2-way ANOVA with adjustment for multiple comparisons using Dunnett’s multiple comparisons test method, mean + S.E.M. *p < 0.05, **p < 0.01, ***p < 0.001. C) Representative H&E histology from gut-ileum of representative mice from groups in 2B. D) Post-mortem dissection of tumours from groups in 2B. E) Imaging of CTDR mEV uptake 4-h after oral gavage in tumour, gut and bone marrow following 7 Gy irradiation. F) Quantification of CTDR mEV uptake 4-h after oral gavage in same tissues as (E) following 7 Gy irradiation versus non-irradiated controls. n = 6–8 mice/group, Welch’s 2-tailed T-test compared irradiated and non-irradiated tissue pairs, mean + S.E.M, ***p < 0.0001; **p < 0.01.

Terminal dissections (Fig. 2D) and measurements also showed no differences in tumour size or metastatic incidence between radiation-alone and radiation + XOlacta groups (Supplementary Fig. 3; Supplementary Table 2). Ileal histology from the same animals showed better preservation of villi and crypts in irradiated mice treated with XOlacta, consistent with maintained normal-tissue protection during fractionated tumour irradiation (Fig. 2C). Four of the 53 mice developed non-healing skin ulcers overlying their tumours; however, no radiation-induced skin lesions occurred in mice receiving radiation plus XOlacta (Supplementary Table 2).

CTDR-mEV uptake studies showed that GL261 tumours were intrinsically mEV-avid, exhibiting significantly higher punctate CTDR signal than unirradiated gut or bone marrow, and that 7 Gy irradiation did not further increase tumour uptake (Fig. 2E and F). Unlike 14 Gy exposure, the 7 Gy protocol did not significantly enhance mEV accumulation in the normal radiosensitive tissues.

3.3.1. Implications and limitations

Our data identify oral Cx43-peptide–loaded mEVs (XOlacta) as radiomitigators that improve survival, restore gut and marrow architecture, and preserve tumour radiosensitivity. The platform offers practical advantages for clinical use: oral dosing, good human tolerability, room-temperature stability of a lyophilized product, and a scalable bovine-milk source compatible with stockpiling and rapid deployment [5,6]. In the radiotherapy setting, these properties support development of XOlacta as a convenient adjunct capable of widening the therapeutic window between tumour control and normal-tissue injury. At the same time, the single-fraction total-body-irradiation regimens used here are supralethal in our mouse strain, whereas in humans single whole-body doses of ~4–5 Gy already approach the LD50, indicating that the models also mimic severe, accident-like exposures and provide a stringent test of orally delivered, tissue-selective radiomitigators.

Key limitations include use of one mouse strain, one glioma line, and specific radiation regimens. Extending these findings to additional tumours, fractionation schemes, and large-animal models to define therapeutic index is necessary. Future work should also include formal toxicology (including hepatic and renal biomarkers), clarify the relative roles of inflammatory versus Cx43-dependent signalling in normal-tissue protection, and test whether enhanced oral mEV targeting is shared across diverse cancers. Although XOlacta neither reduced nor enhanced radiosensitivity in GL261 flank gliomas, it remains of interest to determine whether Cx43-targeted peptides can sensitize breast cancer, melanoma, and glioblastoma/glioma cells to radiation therapy, as reported for chemotherapy combinations [13,17–19].

The present study provides proof-of-principle that a Cx43-peptide–loaded mEV formulation can act as an orally deliverable, tissue-selective radiomitigator, supporting further preclinical development as both a clinical adjunct to radiotherapy and a medical countermeasure for radiological and nuclear emergencies.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.canlet.2026.218686.

Data availability

All data supporting the findings of this study are available within the Article and its Supplementary Information files or from the corresponding author upon reasonable request.