Lung transplantation (LTx) is the unique treatment for end-stage lung diseases, but it is limited by the shortage of available lung grafts (1). Ex vivo lung perfusion (EVLP) has been developed to increase the pool of transplantable lungs, by allowing prolonged graft physiological evaluation and for the possible therapeutic reconditioning and repair of damaged donor lungs before LTx (2). As such, EVLP serves the main purpose to alleviate ischemia reperfusion (I/R) injury after LTx, hereby reducing the risk of primary graft dysfunction (PGD) (3). The pulmonary endothelium is a major target of I/R injury during LTx (4). At the onset of reperfusion, pulmonary endothelial cells produce highly reactive oxygen (ROS) and nitrogen species (RNS) which may precipitate endothelial damage and dysfunction, resulting in increased vascular permeability, extravasation of immune cells and lung injury (4–6). Different strategies for pulmonary endothelial protection in the context of lung I/R injury have been evaluated, such as the targeted endothelial delivery of antioxidants (7) or the inhibition of adhesion molecules activities (8), to name a few. Although promising in the experimental setting, their translational application for clinical LTx remains to be established (9).

Previous investigators reported that endothelial cells (ECs) from non-pulmonary origin could be protected against I/R injury through the transient exposure to a mild heat stress (HS) and the induction of a heat-shock response (10, 11). The latter is a highly conserved adaptive mechanism conferring tolerance to a diversity of stressful stimuli, through the induced expression of various heat shock proteins (HSPs), acting as molecular chaperones for the maintenance of proteostasis (12). Proposed mechanisms of HS-dependent endothelial protection in I/R conditions comprise the activation of K_ATP channel (10) and the stimulation of NO-mediated vasodilation (11). Furthermore, moderate heat exposure may improve endothelial integrity and function through the expression of various protective genes and the reduction of oxidative stress and inflammation (13, 14). These effects on the systemic endothelium support the hypothesis that heat exposure could also benefit the lung endothelium, particularly in the conditions of endothelial damage and dysfunction associated with I/R injury.

The EVLP platform is ideally suited to address this hypothesis, as it permits to selectively expose the lungs to a precisely calibrated temperature with well-defined time limits. We recently provided evidence that transient heat exposure during EVLP conferred protection against I/R injury of damaged lung grafts, by attenuating lung inflammation, oxidative stress, epithelial damage and physiological deterioration after LTx (15). In the present study, we sought to expand further on the effects of ex-vivo therapeutic heat application, by addressing more specifically its effects on the lung endothelium. Furthermore, since prolonged reperfusion can emphasize the damaging effects of I/R injury on the endothelium (16), the effects of heat exposure were evaluated in an EVLP model with an extended reperfusion time of 6 hours.

Lung I/R injury following lung transplantation may precipitate primary graft dysfunction, with significant short- and long-term consequences (23). A hallmark of I/R injury is endothelial damage and dysfunction, resulting in the loss of endothelial barrier function, high permeability edema, vasoactive imbalance, procoagulant activity and the recruitment and activation of inflammatory leukocytes (4–6). The search for novel strategies to protect the endothelium appears therefore crucial to improve LTx outcomes (4). In this respect, the development of EVLP has represented a major breakthrough, by permitting the evaluation of endothelial dysfunction prior to LTx (24, 25), as well as for potential therapeutic interventions (16). In a recent study, we provided evidence that the transient application of a mild heat stress during a relatively short EVLP protocol (3h) reduced innate immune signaling, oxidative stress and epithelial damage in rat lung grafts damaged by warm ischemia (15). Our present results significantly extend these observations, by showing that therapeutic heat application provides significant protection against the endothelial damage and dysfunction developing over an extended period (6h) of EVLP.

Transient (30 min) heating at 41.5° after 1h EVLP promoted the expression of two key inducible heat shock proteins, namely HSP70 and HSP27, starting about 2h after heat exposure and maintained until the end of the 6h reperfusion (Figure 2). By contrast, heat stress did not enhance the expression of constitutive Hsc70, but prevented its downregulation, as noted in non-heated lungs after 6h. It is noteworthy that the activity of both HSP70 and Hsc70 is regulated by ATP binding and hydrolysis (26). As we did not measure lung tissue ATP, we cannot determine whether possible changes in ATP content during EVLP might have influenced their activities. Previous investigations revealed that both HSP70 and HSP27 could reduce endothelial damage and permeability in non-pulmonary vessels by maintaining the integrity of cell junction proteins (27) and by preventing actin polymerization (28). Here, we found that heat-treated lungs displayed a marked reduction of lung edema after 6h reperfusion, an effect translating into preserved oxygenation capacity and static lung compliance (Figure 3). Since the formation of lung edema during reperfusion primarily reflects the increase of endothelial permeability (29, 30), this observation points to significant protection of the endothelial barrier function by heat therapy, an assertion further supported by the decrease in perivascular edema at histological evaluation (Figure 4). In addition, we may also consider the role of reduced capillary hydrostatic forces (31), as suggested by the attenuated increase of PA pressure during reperfusion of heat-treated lungs. It is here worth to mention that, owing to the constant flow conditions in our model, the increase of PA pressure during EVLP was entirely related to an increase in PVR, the latter being significantly blunted by heat therapy (Figure 5). It has been shown that an increase in PVR directly correlates with endothelial dysfunction during experimental LTx (32), and that the degree of PVR increase during EVLP is associated with the severity of PGD after LTx (33, 34). Collectively, these arguments support that heat therapy significantly attenuated endothelial dysfunction and the increased endothelial permeability developing over the 6h of EVLP.

The endothelial barrier function depends on the integrity of inter-endothelial cell connections, comprising adherens (primarily VE-cadherin) and tight (primarily occludins and claudins) junction proteins, connected to the actin cytoskeleton through interactions with catenins, zonula occludens proteins and cingulin (22). Opening of these intercellular junctions precipitates high permeability edema, as occurs during inflammation or ischemia-reperfusion. Factors increasing endothelial permeability, such as inflammatory mediators, VEGF and thrombin, trigger signaling cascades leading to the remodeling of endothelial junctions and the inhibition of VE-cadherin adhesive function, largely related to phosphorylation of VE-cadherin at Tyr⁶⁸⁵ by Src protein tyrosine kinase (22). Src itself is activated through Tyr⁴¹⁹ phosphorylation, whereas phosphorylation at Tyr⁵²⁹ (Tyr⁵³⁰ in humans) leads to decreased activity (35). Our findings indicate that such signaling events were, indeed, triggered after 6h EVLP, as indicated by intense Src Tyr⁴¹⁹ phosphorylation, concomitant to Tyr⁵²⁹ dephosphorylation, associated with the phosphorylation of VE-cadherin at Tyr⁶⁸⁵ (Figure 6). Most importantly, these events were abrogated when lungs were subjected to transient heating, confirming the significant potential of this intervention to prevent the deterioration of endothelial barrier function during prolonged reperfusion.

A possible mechanism underlying these observations may rely in a reduction of oxidant-dependent Src kinase activation (36), in view of the protective actions of heat stress against oxidant-dependent processes (15, 37, 38). Notably, Src is very sensitive to activation by peroxinitrite, a potent oxidant species formed from nitric oxide and superoxide (39), which has been associated with the development of endothelial dysfunction in various conditions, including I/R injury (39). We evaluated endothelial peroxynitrite formation in lung sections after 6h EVLP, using immunofluorescent detection of 3-nitrotyrosine, a stable footprint of peroxinitrite, and of the endothelial marker CD31 (Figure 7). In the EVLP_37°C group, the endothelial lining of alveolar vessels was thickened and intensely positive for 3-NT, contrasting with an attenuated 3-NT staining in the thinner alveolar vessels of heat-treated lungs. While these findings indicate that heat stress reduced vascular peroxinitrite generation, the latter was, nevertheless, not abrogated. Thus, additional mechanisms must be considered to explain the profound reduction of Src kinase activation and VE-cadherin phosphorylation following heat application.

One such mechanism could be the direct inhibition of Src activity by HSP70, as reported in cancer cells (40). Furthermore, the inflammatory cytokines TNFα and IL-1β are potent initiators of increased endothelial permeability (22), an effect mediated, in part, by the activation of Src-family protein kinases (41–43). Copious amounts of both TNFα and IL-1β were released in the perfusate of EVLP_37°C lungs over the 6h of observation (Figure 8), which is consistent with data reported by us (17, 20) and other investigators (44, 45). The release of cytokines was largely reduced by heat stress, confirming our recent studies showing that heat therapy lessened innate immune activation during EVLP and LTx through the inhibition of NF-κB and the inflammasome (15). Taken together, these results suggest that heat stress decreased Src kinase activation leading to VE-cadherin phosphorylation through at least two distinct mechanisms, namely the reduction of endothelial peroxinitrite formation, and the blunted expression of inflammatory cytokines.

We then assessed the influence of heat stress on the release of the endothelial biomarkers PECAM-1, vWF, E-selectin and ICAM-1 during EVLP. PECAM-1 (CD31), is a transmembrane protein critical for the maintenance of the normal endothelial barrier, detectable in plasma as a soluble truncated protein (sPECAM) upon cleavage from damaged endothelial cells (46). VWF, constitutively expressed and stored in endothelial cells, plays important roles in endothelial adhesion and hemostasis, and is released extracellularly under conditions of endothelial damage and activation (47). E-selectin and ICAM-1 are inducible adhesion molecules expressed by the endothelium upon activation by inflammatory cytokines (48). We found that sPECAM increased substantially in the perfusate of non-heated lungs, an effect associated with the downregulation of lung PECAM-1 expression, implying that severe endothelial cell injury resulted in the shedding of PECAM-1 into the extracellular milieu. Similarly, the perfusate levels of vWF, E-selectin and ICAM-1 steadily increased over time, further confirming significant endothelial damage and activation during reperfusion. At variance with these findings, heat-treated lungs released considerably less vWF and sPECAM, while maintaining tissue PECAM-1 expression, indicating better preservation of their endothelial integrity. Furthermore, the accumulation of E-selectin and ICAM-1 in the perfusate was also significantly attenuated by heat-stress, albeit in a lesser proportion than sPECAM and vWF, which could reflect a greater functional endothelium in heat-treated lungs, as suggested by their greater PECAM-1 expression.

Our study has several limitations. The first one is that we did not evaluate endothelial function after transplantation. This was related to the experimental design, which aimed to address the effects of heat stress during an extended period of EVLP. Indeed, the progressive amplification of endothelial damage during prolonged reperfusion may limit the duration during which lung grafts can be maintained in an EVLP system. Our work was therefore designed to determine if therapeutic heat stress allowed to safely extend the duration of EVLP. In addition, due to the severe edema noted after 6h in lungs from the EVLP_37°C group, we could not envision their transplantation into recipient animals. A second limitation is that we performed several measurements in whole lung tissue extracts but not directly in endothelial cells. Future studies will therefore be needed using endothelial cells isolated from the lungs at different stages of EVLP, to address their specific response to heat stress in terms of heat shock protein expression and modulation of various molecular pathways. Also, physiological studies in isolated pulmonary vessels will permit to address the influence of heat stress on vascular reactivity and notably on endothelial-dependent vasodilation. Notwithstanding this limitation, our findings that heat stress reduced the phosphorylation of VE-cadherin, a junction protein exclusively expressed by endothelial cells, and prevented the downregulation of the endothelial antigen CD31 (PECAM-1), point to the direct influence of heat stress on endothelial cells. Finally, we acknowledge that the number of observations was relatively small, with inherent limitations to the statistical analyses.

In conclusion, we demonstrate that a transient heat stress, applied on rat lungs damaged by warm ischemia, protects from endothelial damage and dysfunction during prolonged reperfusion. This conclusion is supported by the significant attenuation of lung edema and physiological deterioration, the suppression of Src kinase activation and VE-cadherin phosphorylation, the reduction of endothelial peroxinitrite generation, the prevention of endothelial PECAM-1 shedding and release of von Willebrand factor, as well as by the downregulated expression of E-selectin and ICAM-1. Collectively, these findings support the concept that ex-vivo heat stress may represent a promising strategy to protect the endothelium from the damaging effect of reperfusion, thereby allowing to safely extend the duration of EVLP.

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

The animal study was approved by Direction générale de l’agriculture, de la viticulture et des affaires vétérinaires (DGAV) de l’Etat de Vaud. The study was conducted in accordance with the local legislation and institutional requirements.

RP: Conceptualization, Data curation, Formal Analysis, Investigation, Writing – original draft, Writing – review & editing. AD: Conceptualization, Data curation, Formal Analysis, Investigation, Writing – original draft, Writing – review & editing. MA: Data curation, Formal Analysis, Investigation, Writing – review & editing. JL: Data curation, Formal Analysis, Investigation, Writing – review & editing. HR-C: Data curation, Formal Analysis, Investigation, Writing – review & editing. LL: Conceptualization, Data curation, Formal Analysis, Funding acquisition, Project administration, Writing – original draft, Writing – review & editing. TK: Conceptualization, Data curation, Formal Analysis, Funding acquisition, Project administration, Writing – original draft, Writing – review & editing.