Adjuvant radiotherapy is a standard therapy for breast cancer treatment after conservative surgery and mastectomy (Yang and Ho, 2013). However, exposure of healthy tissue, in particular, the heart, to ionizing radiation (IR) often increases the risk for the development of cardiovascular diseases (CVD), especially atherosclerosis (Darby et al., 2003; Marks et al., 2005; Baker et al., 2011; Cutter et al., 2011; Yusuf et al., 2011; Darby et al., 2013; Aleman et al., 2014; Van Nimwegen et al., 2015; Baselet et al., 2016). Although modern radiotherapy techniques reduce the volume of the heart and major coronary vessels exposed to high doses of IR, some exposure is often unavoidable, especially in the case of left-sided breast cancer, in which case the accumulative dose received by the heart area is in the order of ∼6.6 Gy (compared to 2.9 Gy for right-sided breast cancer) (Darby et al., 2013; Eftekhari et al., 2015; Al-Kindi and Oliveira, 2016; Cho et al., 2018). A population-based case-control study in women who underwent radiotherapy for breast cancer indicated a significant increase of 7.4% in the rates of major coronary events (i.e. myocardial infarction, coronary revascularization, or death from ischemic heart disease) per increase of 1 Gy in the cardiac exposure dose, without apparent threshold (Darby et al., 2013). In addition to breast cancer, exposure of the cardiovascular system to IR may occur during thoracic cancer radiotherapy such as head-and-neck cancer, Hodgkin’s lymphoma, and esophageal cancer (Borghini et al., 2013). Indeed, growing evidence indicates a link between IR exposure at high and medium doses (>0.5 Gy) and atherosclerosis development (Stewart et al., 2006; Hoving et al., 2008; Shimizu et al., 2010; Icrp et al., 2012; Darby et al., 2013; Kreuzer et al., 2015; Little, 2016; Cheng et al., 2017). However, the underlying molecular mechanisms remain not entirely understood. In addition, there is growing epidemiological evidence nowadays that indicates an excess risk for CVD at much lower IR doses than previously thought (<0.5 Gy) (Yamada et al., 2005; Kreuzer et al., 2015; Baselet et al., 2016; Little, 2016). However, these data are suggestive rather than persuasive due to a lack of statistical power and the fairly limited knowledge on the underlying molecular mechanisms. Hence, effects of radiation exposure on the cardiovascular system, especially at low dose exposure, are still not fully characterized, possibly resulting in improper radiation protection.

Exposure of endothelial cells to IR induces DNA damage and ROS which in turn induces oxidative stress, cellular Ca²⁺ overload, inflammation, senescence and apoptosis which may cause endothelial cell dysfunction, an early marker for atherosclerosis (Baker et al., 2011; Yusuf et al., 2011; Borghini et al., 2013; Madan et al., 2015; Baselet et al., 2016; Tapio, 2016). Intercellular communication mediated by gap junctions and hemichannels, may modulate endothelial response to IR, and therefore, the atherosclerotic process (Azzam et al., 2001; Little, 2006; Ohshima et al., 2012). Gap junctions permit passive diffusion of small hydrophilic molecules and ions of below ∼1.5 kDa molecular weight between adjacent cells. Gap junctions are usually open under physiological conditions, and they are important both for slow physiological processes (e.g. cell growth and proliferation) and fast activities such as electrical impulse conduction in the heart or communication between vascular endothelial cells to transmit upstream vasodilatory messages (De Wit and Griffith, 2010). In contrast to gap junctions, plasma membrane hemichannels are typically closed under physiological conditions (Orellana et al., 2011), and can open in response to several stress signals (e.g. mechanical stimulation, stress-associated stimuli such as ischemia, oxidative stress, a modest increase in intracellular Ca²⁺ or pro-inflammatory conditions) (Ramachandran et al., 2007; De Vuyst et al., 2009; Orellana et al., 2009; Johansen et al., 2011; Meunier et al., 2017). When open, hemichannels facilitate bidirectional passage over the plasma membrane of ions and signaling or metabolic molecules of below ∼1.5 kDa molecular weight (e.g. Ca²⁺, prostaglandin E2, glutathione, glutamate, ATP, NAD⁺, and others) (Wang et al., 2013a). Uncontrolled opening of hemichannels can result in loss of cell-essential metabolites, excessive entry of Na⁺ and Ca²⁺ and ATP leakage which in turn can lead to activation of apoptotic caspases, nitric oxide production, and inflammation (Decrock et al., 2009a, b; Saez et al., 2010; Ohshima et al., 2012; Decrock et al., 2017; Leybaert et al., 2017). These events may potentially trigger significant changes in cell homeostasis and cause cell dysfunction.

Gap junctions and hemichannels are composed of transmembrane connexin (Cx) proteins. Endothelial cells, the interior lining of blood vessels and cardiac valves, express three main Cx isotypes, namely Cx37, Cx40, and Cx43. Increasing evidence indicates that Cxs play a role in atherosclerosis progression (27, 29–31). Cx37 and Cx40 are suggested to have atheroprotective effects as it was reported that healthy endothelial cells have a widely distributed Cx37 and Cx40 expression pattern, while both Cxs are lost in the endothelium covering the advanced atherosclerotic plaques (Kwak et al., 2002; Morel, 2014). In contrast, Cx43 has been associated with atherogenesis as it is highly expressed in atherosclerotic plaques compared to control (Kwak et al., 2002; Morel, 2014). Cx43 also has atherogenic properties as decreasing Cx43 expression reduces the formation of atherosclerotic lesions in vivo (Kwak et al., 2003; Wong et al., 2003; Wang et al., 2013b) while in vivo Cx43 upregulation increased the expression of cell adhesion proteins such as VCAM-1, thereby enhancing monocyte-endothelial adhesion, a key event in initiating atherosclerosis (Yuan et al., 2015). Furthermore, Cx43 has been implicated in endothelial cellular stiffness that is associated with cardiovascular disease and atherosclerosis (Okamoto et al., 2017).

Besides the role of Cx in the development of atherosclerosis, it was previously reported that Cx expression is sensitive to ionizing radiation. Cx43 expression was increased in human skin fibroblast after exposure to 10 mGy of α-particles (Azzam et al., 2003). In addition, Cx43 expression is elevated in response to low dose gamma-ray exposure (¹³⁷Cs source) in human neonatal foreskin fibroblast (Glover et al., 2003), in response to 5 Gy X-rays in mouse brain endothelial cell line bend3 (Banaz-Yasar et al., 2005) and in cardiac myocytes in an in vivo animal model upon exposure to high-LET radiation (Amino et al., 2010). Finally, it was reported that 0.5 Gy of gamma-rays exposure induced Cx43 hemichannels opening in B16 melanoma cells (Ohshima et al., 2012).

Although Cxs and their channels have been reported to be involved in the pathogenesis of atherosclerosis and to be sensitive to IR exposure, their role in radiation-induced endothelial cell response were never investigated (Azzam et al., 2003; Pfenniger et al., 2013). We previously demonstrated that single and fractionated IR induces acute and persistent upregulation of Cx43 gene and protein expression in the coronary artery and microvascular endothelial cells (Ramadan et al., 2019). In addition, we demonstrated that IR induces acute and long-lived Cx43 hemichannel opening in a dose-dependent manner (Ramadan et al., 2019). Here, we aimed at investigating the involvement of Cx43 hemichannels in endothelial cell responses induced by IR exposure, by making use of the Cx43-targeting peptide TAT-Gap19 (Abudara et al., 2014). TAT-Gap19 was previously reported to specifically block Cx43 hemichannels in different in vitro experimental models, as observed by ATP release and dye uptake assays (Abudara et al., 2014; Willebrords et al., 2017; Maatouk et al., 2018; Saez et al., 2018; Walrave et al., 2018). These findings were supported by electrophysiological measurements of Cx43 hemichannel unitary currents demonstrating that Gap19 inhibits Cx43 hemichannels in HeLa cells overexpressing Cx43 (Wang et al., 2013b; Gadicherla et al., 2017), in acutely isolated pig ventricular cardiomyocytes (Wang et al., 2013b) and in primary astrocytes (Freitas-Andrade et al., 2019). In depth mechanistic investigations based on surface plasmon resonance studies revealed a direct binding of Gap19 to the C-terminal tail of Cx43, thereby preventing the CL loop/CT tail interaction which is essential for Cx43 hemichannel activity, whereas closure of gap junctions is prevented (Ponsaerts et al., 2010; Iyyathurai et al., 2018). Linking Gap19 to the HIV-derived TAT internalization sequence promotes its membrane permeability and reduces the concentration needed for half-maximal Cx43 hemichannel inhibition by 5-folds. We found that TAT-Gap19 reduced intracellular ROS generation, cell death, inflammation, and premature cell senescence induced by IR in the immortalized human coronary artery and microvascular endothelial cells. Collectively, these results indicate a functional linkage between hemichannel opening and post-irradiation ROS, inflammation, and cell death/senescence responses.

Patients treated with thoracic radiotherapy have an increased risk of cardiovascular disease, but the underlying pathophysiology is complex and not fully understood (Yamada et al., 2005; Stewart et al., 2006; Hoving et al., 2008; Shimizu et al., 2010; Icrp et al., 2012; Darby et al., 2013; Kreuzer et al., 2015; Baselet et al., 2016). Here, we aimed to investigate the role of Cx43-based hemichannels in radiation-induced endothelial cell damage, an early marker of atherosclerosis, making use of the peptide inhibitor TAT-Gap19 (Wang et al., 2013b; Delvaeye et al., 2018). We found that this peptide significantly reduced oxidative stress, cell death, key inflammatory cytokines and premature cell senescence induced by X-ray exposure of endothelial cell lines derived from coronary artery and microvascular endothelial cells (Figure 6). Below we discuss these findings in more detail.

A predominant effect of IR exposure is ROS production and oxidative stress (Robbins and Zhao, 2004; Azzam et al., 2012; Yamamori et al., 2012). Accordingly, in TICAE (coronary artery-derived) and TIME cells (microvascular endothelial cells TIME), we found a dose-dependent increase in oxidative stress after 0.1 and 5 Gy of X-ray exposure as measured by CM-H2DCFDA-based live cell imaging approache. The highest oxidative stress response was observed at 45 min p.i. in both cell types, followed by a decline to the 2 and 3 h time points; the latter were, however, still significantly above non-irradiated controls. The decline at 2–3 h may be linked to antioxidant defense reactions of enzymes such as superoxide dismutases (SOD), glutathione, glutathione peroxidases, catalase and others (Robbins and Zhao, 2004; Azzam et al., 2012; Szumiel, 2015). A similar timeline has been observed in primary microvascular endothelial cells exposed to gamma-rays (Laurent et al., 2006). The observed differences between TICAE and TIME cells in the extent of ROS production is possibly related to different radiosensitivity and distinct endothelial properties at different sites of the vascular tree (Aird, 2012; Park et al., 2012). Interestingly, we found that blocking Cx43 hemichannels with TAT-Gap19 significantly reduced ROS production in both cell types at 45 min p.i. Cx43 hemichannel opening has been demonstrated previously in osteocytes in response to oxidative stress, which involved changes of intracellular Ca²⁺ concentration (Kar et al., 2013; Riquelme and Jiang, 2013). Oxidative stress was also reported to induce Cx43 hemichannels opening in fibroblastoid rat mammary tumor cell line (Marshall Cells) which express Cx43 only (Ramachandran et al., 2007). Further downstream, hemichannel opening has been linked to an extracellular depletion of the major antioxidant glutathione, thereby exaggerating oxidative stress in renal tubular epithelial cells (Chi et al., 2016). Along the same line, Cx43 hemichannel inhibition with TAT-Gap 19 increased SOD activity in mice liver treated with thioacetamide to induce liver fibrosis (Crespo Yanguas et al., 2018). We hypothesize that IR exposure induces ROS production in endothelial cells, which opens Cx43 hemichannels and leads to either loss of hemichannel-permeable antioxidants such as glutathione or activates other signaling pathways that lead to the loss of larger antioxidants such as SOD. TAT-Gap19 attenuated this oxidative stress response, but given the important bidirectional relation between [Ca²⁺]i elevation and ROS generation (Yan et al., 2006; Gorlach et al., 2015), it is also possible that hemichannel inhibition directly reduces ROS production by inhibiting Ca²⁺ entry through hemichannels, though further investigations are needed to clarify the involvement of Cx43 hemichannels in intracellular oxidative stress induced by ionizing radiation exposure.

It is well known that oxidative stress is associated with the activation of various cascades including the DNA damage response, apoptosis and inflammatory pathways, thereby amplifying endothelial dysfunction (Elahi et al., 2009; Frey et al., 2009; Soucy et al., 2011; Schramm et al., 2012; Tapio, 2016; Kattoor et al., 2017; Incalza et al., 2018). We observed radiation-induced DNA damage in the two cell types at 1 h p.i., as evidenced by gamma-H2AX and TP53BP1 foci analysis. These findings are consistent with previous studies where a dose-dependent increase in DNA damage was observed in EA.hy926 endothelial hybrid cells, in umbilical vein endothelial cells, as well as in coronary artery endothelial cells in response to X-ray doses in the range of 0.05 to 2 Gy (Rombouts et al., 2013; Baselet et al., 2017), and in microvascular endothelial cells after 3 and 10 Gy of gamma-rays (Laurent et al., 2006). Our findings showed that TAT-Gap19 did not alter this DNA damage response in the irradiated TICAE and TIME cells.

We found that cell death in the two cell types mainly occurred after 5 Gy irradiation, with TICAE cells responding with Annexin V and Caspase 3/7 elevation while TIME cells responded with Annexin V elevation and membrane leakage as indicated by cellular dextran fluorescein positivity. The absence of Caspase 3/7 elevation in TIME cells confirms the findings in primary human microvascular endothelial cells exposed to X-rays in the 2 to 12 Gy dose range (Muschter et al., 2018). Overall, the cell death observed in our study is in line with previous findings in umbilical vein endothelial cells (Rombouts et al., 2013) and microvascular endothelial cells (Langley et al., 1997; Rousseau et al., 2011). Interestingly, we show that TAT-Gap19 significantly reduced cell death in irradiated TICAE and TIME cells. Cx43 hemichannels are known to mediate cell death in various model systems [reviewed in Decrock et al. (2009b)] as well as in cardiomyocytes exposed to hypoxia-reoxygenation or ischemia-reperfusion (Wang et al., 2013b; Gadicherla et al., 2017), by facilitating the passage of soluble factors and molecules such as ATP, ROS and Ca²⁺ (Leist et al., 1997; Li et al., 2001; Hur et al., 2003; Kalvelyte et al., 2003; Rodriguez-Sinovas et al., 2007; Decrock et al., 2009a, b; Saez et al., 2010; Belousov et al., 2017). We (in a previous study) and others reported that IR induces hemichannel-related ATP release (Kang et al., 2008; Ohshima et al., 2012; Ramadan et al., 2019) which may cause intracellular ATP depletion, thereby activating cell death process through necrosis or apoptosis (Leist et al., 1997; Kalvelyte et al., 2003; Decrock et al., 2009a). Hemichannel opening has furthermore been suggested to facilitate ROS entry into the cells eventually leading to cell death (Ramachandran et al., 2007). Moreover, Cx43 hemichannel inhibition with TAT-Gap19 was previously reported to reduce cell death in various model systems (Wang et al., 2013b; Gadicherla et al., 2017), and to have anti-apoptotic activity by increasing the expression of Bcl-2 and decreasing that of Bax (Chen et al., 2019). Our experiment further showed that TAT-Gap19 decreased cell death in the control non-irradiated condition, as observed by an increase in Annexin V and Caspase 3/7 detection in TICAE cells and by an increase in Annexin V and Dextran fluorescein detection at 72 h in TIME cells. This could be explained by Cx43 hemichannels opening in response to several stress stimuli (Ramachandran et al., 2007; De Vuyst et al., 2009; Orellana et al., 2009; Johansen et al., 2011; Meunier et al., 2017), which could be associated with manipulation of cells during experiments or confluent cells in culture, resulting in a basel level of ATP leakage and cell death, counteracted by TAT-Gap19 protection.

Inflammation plays a key role in atherosclerosis development and progression (Libby, 2012). In this study, we showed that radiation induced an increase in atherogenic inflammatory markers (IL-6, MCP-1, PECAM-1, IL-1β, TNF-α, CRP, VCAM-1, E-Selectin, Endothelin-1, IL-8, and PAI-1) mainly at 5 Gy, in both TICAE and TIME cells. Observations in various endothelial cells confirm an elevation of these inflammatory markers mainly at higher irradiation doses (>2 Gy) (Hallahan et al., 1995; Van Der Meeren et al., 1999; Lanza et al., 2007; Scharpfenecker et al., 2009; Kiyohara et al., 2011; Abderrahmani et al., 2012; Haubner et al., 2013; Sievert et al., 2015; Himburg et al., 2016; Baselet et al., 2017). We found that TAT-Gap19 significantly reduced the elevated IL-1β, IL-8, VCAM-1, MCP-1, and Endothelin-1 levels in the two cell lines, at late time points (72 h and 7 days), while it also had early (24 h) inhibitory effects in TIME cells mainly at 5 Gy. In addition, TAT-Gap19 displayed early (24 h) inhibition of TNF-α, E-Selectin, PECAM-1, CRP, and ICAM-1 mainly at 5 Gy but only in TIME cells, indicating that Cx43 hemichannel involvement in the inflammatory response is time and cell line dependent. All cytokines listed are known to be strongly involved in the pathogenesis of radiation-induced atherosclerosis. The cytokine response is triggered by IR/oxidative stress-induced NF-κb signaling and results in endothelial cell activation with subsequent expression of adhesion molecules (Kim et al., 2001; Roedel et al., 2002; Elahi et al., 2009; Liao, 2013; Sievert et al., 2015). IL-6 has been shown to be responsible for propagating downstream inflammatory responses, and expression of adhesion molecules, such as VCAM-1, ICAM-1, PECAM-1, and E-Selectin, on the vascular wall which is known to initiate the recruitment of macrophages and leukocyte from the vasculature, that acts together with MCP-1 and IL-8 to induce infiltration of macrophages into the subendothelial cell layer, leading to the formation of atherosclerotic lesions (Harrington, 2000; Libby et al., 2002; Schuett et al., 2009; Kiyohara et al., 2011; Libby, 2012). Moreover, increased the production of the potent vasoconstrictor peptide Endothelin-1 was linked to endothelial cell dysfunction, as it may decrease endothelial nitric oxide synthase (eNOS) expression, thereby reducing nitric oxide (NO) vasodilatory signaling and it was also reported to activate macrophages (Kinlay et al., 2001; Bousette and Giaid, 2003; Loomis et al., 2005; Napoli et al., 2006; Bohm and Pernow, 2007). Importantly, it is thought that Cx43 hemichannel opening may continue and propagate the inflammatory scene (Retamal et al., 2007; Guo et al., 2016; Li et al., 2018; Saez et al., 2018). In addition, upregulated Cx43 expression often correlates with increased inflammatory responses, while reduced Cx43 expression inhibits inflammation (Mori et al., 2006; Retamal et al., 2007; Cronin et al., 2008; Danesh-Meyer et al., 2008; Tsuchida et al., 2013; Mugisho et al., 2018). TAT-Gap19 has been demonstrated to significantly decrease IL-1β and TNF-α in non-alcoholic fatty liver mouse model (Willebrords et al., 2017). Peptide5, another mimetic peptide-based hemichannel blocker, was reported to reduce TNF-α and IL-1β in rat model of spinal cord injury (O’Carroll et al., 2013). This peptide also reduced inflammation in a rat model of light-induced retinal damage, as assessed by glial fibrillary acidic protein (GFAP) and leukocyte common antigen (CD45) immunohistochemistry (Guo et al., 2016; Mat Nor et al., 2018). A possible mechanism by which Cx43 hemichannels contribute to inflammation is through their regulatory effect on high-mobility group box 1 (HMGB1), which serves as a damage-associated molecular pattern molecule (DAMP) (Li et al., 2018). HMGB1, released by apoptotic cells, induces IL-6, IL-8, and MCP-1 secretion and increases the expression of ICAM-1 and VCAM-1 in endothelial cells (Sun et al., 2013). Furthermore, by releasing ATP, Cx43 hemichannels enhance inflammation through downstream activation of the NLRP3 inflammasome (Mugisho et al., 2018; Mugisho et al., 2019). Cx43 hemichannels may thus act to amplify and perpetuate an initial inflammatory condition. Interestingly, IR induces a rapid and persisting increase in Cx43 gene and protein expression in TICAE and TIME cells (Ramadan et al., 2019), thereby creating pro-inflammatory substrate.

In contrast to the inflammation mitigating effect of TAT-Gap19 on irradiated endothelial cells, the peptide transiently increased some other inflammatory markers at various time points (Supplementary Table 1). TAT-Gap19 is composed of the nonapeptide Gap19, linked to a HIV-derived TAT sequence to promote its membrane permeability (Abudara et al., 2014). It is possible that this HIV-derived membrane translocation sequence may induce pro-inflammatory side effects in the cells. In line with this, alpha-carboxy terminus 1-peptide (αCT1), another Cx43 based peptide containing the antennapedia translocation motif, triggered a transient increase in TNF−α upon topical corneal application (Moore et al., 2014). Here, we report that the TAT translocation motif by itself induces a transient increase in IL-6, MCP-1, IL-1β and CRP in TICAE cells at 24 h p.i. (the time point at which TAT-Gap19 significantly increased all assessed cytokines in TICAE cells). In addition, TAT peptide increased MCP-1 and VCAM-1 only in TIME cells at 72 h post exposure (of note; TAT-Gap19 significantly increased the 0 Gy control and some of the irradiated conditions in these cytokines at 72 h p.i. in TIME cells). Taken this into account, our results indicate that TAT-Gap19 exerts a late inflammation mitigating effect, by reducing the atherogenic inflammatory markers IL-1β, IL-8, VCAM-1, MCP-1, and Endothelin-1, in both TICAE and TIME cells after IR exposure.

Interestingly, inflammation has been reported to contribute to cellular senescence (Freund et al., 2010; Tian and Li, 2014). Premature cell senescence is known to contribute to endothelial dysfunction by stimulation a complex pro-inflammatory response, increasing ROS levelsand it is known to trigger the initiation of cell death by reducing cell repair ability (Collado et al., 2007; Xue et al., 2007; Donato et al., 2015). Our experiments showed an increase in SA-β gal activity in TICAE and TIME cells mainly at 7 and 9 days p.i. This increase in SA-β gal activity was persistent until 14 days p.i. only at 0.1 Gy. For 5 Gy, an unexpected decrease in SA-β gal activity was observed, which may be caused by the frequent medium change for this time point, resulting in washout of senescent and/or apoptotic cells, which in our experiment mainly occurred at the high dose, and replacement by proliferating cells. Premature senescence in TICAE and TIME cells was also apparent from the increase in IGFBP-7 (at 5 Gy) and GDF-15 (at 0.1 and 5 Gy) 7 days p.i. Our results are in line with previous studies, where cellular senescence was observed after low and high doses of IR exposure in different types of endothelial cells, including human coronary artery (Oh et al., 2001; Ungvari et al., 2013; Yentrapalli et al., 2013a, b; Kim et al., 2014; Lowe and Raj, 2014; Baselet et al., 2017). Interestingly, TAT-Gap19 significantly suppressed the radiation-induced increases in SA-β gal activity, mainly at 5 Gy at 7 and 9 days and from 0.1 Gy at 14 days, and GDF-15 levels from 0.1 Gy in TICAE and TIME cells suggesting a role for Cx43 hemichannels. Chronic Cx43 hemichannel opening, as reported in our previous study at 72 h after irradiation in TICAE and TIME cells (Ramadan et al., 2019), may result in ATP leakage that in turn can activate downstream cellular processes including propagation of intercellular Ca²⁺ waves, ROS production, activation of the NLRP3 inflammasome pathway and inflammation (De Bock et al., 2012; Mugisho et al., 2018; Tonkin et al., 2018), which are all known to stimulate premature cellular senescence (Kurz et al., 2004; Acosta et al., 2013; Tchkonia et al., 2013). In addition, senescence can be communicated to neighboring cells in a paracrine manner involving secretion of the senescence-associated secretory phenotype (SASP) and inflammasome activation (Acosta et al., 2013). Presumably, Cx43 hemichannels may perhaps facilitate paracrine senescence communication between cells via its effects on pro-inflammatory parameters like IL-8, MCP-1, and IL-6 which are inhibited by TAT-Gap19 in our study. We further showed that TAT-Gap19 decreased senescence in some of the control non-irradiated conditions. Several stress stimuli might be associated with confluent cells in culture for longer times (7 to 14 days), which may result in a basal level of senescence that can be protected by TAT-Gap19, as shown in our experiment.

In summary, this study demonstrates that TAT-Gap19 mitigates radiation-induced endothelial cell damage by reducing oxidative stress, cell death, pro-inflammatory and pathological factors like IL-1β, IL-8, VCAM-1, MCP-1, and Endothelin-1 and premature senescence. This effect is observed mainly for the 5 Gy dose. Further research is needed to investigate the underlying pathways and molecular mechanisms leading to radiation-induced Cx43 hemichannel opening, including changes in membrane potential and intracellular Ca²⁺, redox status, the surface amount of Cx43 hemichannels and the channel opening probability. Investigating the amount of functional connexin hemichannels at the cell surface may also help in understanding the variable response of TAT-Gap19 in different cell types. In addition, it will be interesting to gather information regarding the downstream effects of Cx43 hemichannel opening that trigger major changes in endothelial function and survival such as unveiling the role of purinergic signaling, intracellular Ca²⁺, ionic balance and intracellular redox status. Other open question that needs to be resolved is what is the effect of other Cx43 hemichannel targeting approaches, such as Gap19 without HIV-derived TAT, siRNA/shRNA or other mimetic peptides, on endothelial cell response after irradiation. This may open perspectives for the establishment of novel therapeutic strategies to protect from secondary radiotherapy side effects in thoracic cancer patients.

In conclusion, to the best of our knowledge, this is the first study to show the possible involvement of endothelial Cx43 hemichannels in contributing to various radiation-induced processes, such as oxidative stress, cell death, inflammation and premature cell senescence that lead to endothelial activation and dysfunction, which is known as an early marker for atherosclerosis. Therefore, targeting Cx43 hemichannels holds potential to therapeutically protect against radiation-induced endothelial cell damage.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation, to any qualified researcher.

RR conducted all the experiments and wrote the manuscript text. EV did a MSc project contributing to inflammatory marker measurement at 48 h and Dextran fluorescein experiments. DA did a MSc project contributing to ROS experiment using Flow Cytometry, inflammatory markers measurements at 24 and 72 h, and apoptosis live cell imaging experiments. IG performed a MSc project contributing to live cell imaging of ROS and DNA damage experiment. DH participated in Calcium measurement experiment which directed us in the discussion. ED, SB, LL, and AA contributed in designing the experiments and supervision of the work. All authors, except the MSc students EV, DA, and IG contributed equally in reviewing the manuscript. All authors read and approved the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.