Microvascular Injury, Thrombosis, Inflammation, and Apoptosis in the Pathogenesis of Heatstroke
A Study in Baboon Model
Objective— Severe heatstroke is a leading cause of morbidity and mortality during heat waves. The pathogenesis of tissue injury, organ failure, and death in heatstroke is not well understood.
Methods and Result— We investigated the pathways of heatstroke-induced tissue injury and cell death in anesthetized baboons (Papio hamadyras) subjected to environmental heat stress until core temperature attained 42.5°C (moderate heatstroke; n=3) or onset of severe heatstroke (n=4) signaled by a fall in systolic blood pressure to <90 mm Hg and rise in core temperature to 43.1±0.1°C. Three sham-heated animals served as controls. Light and electron microscopy revealed widespread hemorrhage and thrombosis, transmural migration of leukocytes, and microvascular endothelium injury in severe heatstroke. Immunohistology and ultrastructural analysis demonstrated increased staining of endothelial von Willebrand factor (vWF), tissue factor (TF), and endothelial leukocyte-platelet interaction. Extensive apoptosis was noted in spleen, gut, and lung, and in hematopoeitic cells populating these organs. Double-labeling studies colocalized active caspase-3 and TF with apoptotic cells. Findings in sham-heated animals were unremarkable.
Conclusion— These data suggested that microvascular injury, thrombosis, inflammation, and apoptosis may play an important role in the pathogenesis of heatstroke injury.
Exposure to high ambient temperature can result in heatstroke, a clinical condition characterized by a core temperature rising rapidly above 40°C associated with central nervous system (CNS) alterations such as delirium, convulsions, or coma.1 Although prevalent in hot climates, heatstroke also occurs sporadically in epidemic form in temperate zones during heat waves.2,3 The heat wave that affected Europe in the summer of 2003 led to an unprecedented 45 000 excess deaths, of which one-third was attributed to heatstroke.2,3 It is predicted that global warming will cause an increase in the frequency and severity of heat waves with an associated rise in mortality, unless proactive measures are taken.3
The pathogenesis of tissue injury and cell death in heatstroke is not well understood, which may explain the high mortality and morbidity as no specific mechanisms can be targeted for treatment.1 Studies in cell lines and animal models suggest that heat directly induces tissue injury and cell death.4–7 Extreme temperatures (49 to 50°C) cause damage to most cellular structures and their function, resulting in cell death by necrosis in less than 5 minutes.5 Rat models subjected to moderate whole body hyperthermia showed that accelerated apoptosis also contributes to cell death, but whether apoptosis is an important cause of cell death in patients with heatstroke is not known.6
Recent evidence associated closely the host inflammatory and hemostatic responses to heat stress with multiple organ system dysfunction/injury (MOSD) and death.1,8–10 Increased circulating proinflammatory cytokines was documented in human and experimental heatstroke and their levels correlated with outcome.7,8,11,12 In laboratory rats, administration of interleukin (IL)-1 receptor antagonist or activated protein C (APC) at onset of heatstroke prevented severe arterial hypotension, organ injury, and improved survival.11,12 Disseminated intravascular coagulation (DIC), a syndrome characterized by widespread intravascular thrombosis that can compromise adequate blood supply to various organs, has been implicated in the MOSD and high mortality of heatstroke.9,10,13 In a large series of 132 patients with heatstroke, the DIC with clinically mild to severe bleeding that was observed in 30% of the patients contributed directly to death in half of them.9 A strong association between acute respiratory distress syndrome and DIC has been observed in patients with heatstroke, suggesting that intravascular deposition of fibrin may account for the organ failure.13 Postmortem findings in patients with fatal heatstroke include diffuse bleeding at several sites, hemorrhagic necrosis, and widespread microthrombi in most organs of the body.10,14 These observations suggest that excessive inflammation and DIC may be major pathological mechanisms.
Although helpful, necropsy studies in humans are retrospective and a time delay can result in postmortem changes.10,14 Primate models of heatstroke offer an alternative method for prospective exploration of the pathogenic mechanisms of tissue injury, organ failure, and death.7,15 We have recently established that baboons subjected to environmental heat stress reproduce the clinical spectrum from moderate to severe heatstroke in human heatstroke.7 Baboons with moderate heatstroke display hyperthermia, CNS alteration, and self-limited MOSD, which subside after 72 hours with full recovery. Comparable with human heatstroke, the occurrence of hemodynamic alteration in baboons subjected to heat stress signals the onset of systemic signs of severe heatstroke that can culminate in death.7,15 Using these 2 experimental baboon models of moderate and severe heatstroke, the objective of this study was to explore the pathways that contribute to tissue injury in heatstroke, as well as to determine whether apoptosis is a major mechanism of cell death.
After approval of the study protocol by the Institutional Review Board, anesthetized baboons (Papio hamadyras) were subjected to environmental heat stress until core temperature attained 42.5°C (moderate heatstroke; n=3) or onset of severe heatstroke (n=4) signaled by a fall in systolic blood pressure to <90 mm Hg and rise in core temperature to 43.1±0.1°C, as described previously.7 Three sham-heated animals served as controls (please see supplemental materials, available online at http://atvb.ahajournals.org).
Blood and Tissue Sampling
Blood samples for hematologic and biochemical assays were collected before heat exposure (baseline), at the end of heat exposure (T+0), and 3 hours later (T+3). Tissue samples from lung, liver, heart, kidney, spleen, jejunum, and adrenal were obtained at immediate autopsy in all animals that died spontaneously or after euthanasia at 72 hours from the onset of heatstroke.
Complete blood count, coagulation, liver, and renal profiles were performed on automated devices. Plasma IL-6, thrombomodulin, and protein C antigen levels were measured in plasma using specific ELISA (Quantikine, R&D Systems and Diagnostica Stago, respectively) and according to the instructions of the manufacturers.
Paraffin embedded tissues were sectioned at 3-μm thickness and stained by hematoxylin and eosin (H&E). Electron microscopy (EM) was performed on small tissue blocks, approximately 5 mm3. Tissue for transmission electron microscopy (TEM) was sectioned at 0.7 μm after postfixation in osmium tetroxide (OsO4), stained with lead citrate and uranyl acetate, mounted on copper grids, and examined under a Phillips 301 electron microscope source.
Terminal dUTP Nick-End Labeling Assay
TUNEL assay was performed using in situ cell death detection kit (Roche Diagnostics), according to the manufacturer’s directions.
Immunohistochemistry and Immunofluorescence
The detection of factor VIII-related antigen (vWF) was performed on fixed tissue sections using Ventana Medical System primary antibody according to the manufacturer’s instructions. Immunostaining for active caspase-3 was performed in embedded sections using anticaspase-3 antibody (CPP32 clone, Cayman) according the manufacturer’s directions followed by HRP labeled Envision polymer (Dako). For caspase-3 and TUNEL double labeling, the sections were first stained with TUNEL as described above, followed by anticaspase-3 primary antibody and Alexa Fluor 647 labeled goat antirabbit secondary antibody (Invitrogen). TF and TUNEL double staining was performed on cryosections where staining with anti-TF primary antibody (clone TF9-10H10, Santa Cruz) and PE-labeled goat anti-mouse secondary antibody (Dako) were performed first, followed by TUNEL staining as described above. Sections were mounted with Vectamount (VectorLabs) and visualized by confocal microscope (Perkin Elmer Ultraview).
This was done using SAS version 9.1.3 (SAS Institute Inc). The data are presented as mean±SE for each group. Differences were considered significant at P<0.05 as determined using either repeated measures analysis of variance (ANOVA) or the Student t test.
Multiple Organ Damage and Outcome
Severe heatstroke baboons displayed MOSD that manifested as renal failure, rhabdomyolysis, and liver injury (supplemental Figure I). Baboons with moderate heatstroke displayed similar changes but with less severity. Sham-heated animals had an uneventful course.
The jejunum, liver, spleen, lung, and kidney in control animals were unremarkable. In contrast, the damage noted in multiple organs in severe and moderate heatstroke was extensive or relatively mild, respectively (Figure 1A). Damage manifested as vascular congestion, hemorrhage, thrombosis, increased inflammatory cells, and disruption of architecture. Specifically, the disruption of architecture consisted of massive loss of surface epithelium in the jejunum, centrilobular disruption of the liver, alveolar hyaline membrane formation, and increased glomerular mesangial cellularity.
Ultrastuctural observations of jejunum, liver, and kidney confirmed and extended the light microscopy findings (Figure 1B). In severe heatstroke, jejunum microvilli showed terminal bulbous expansions, whereas in moderate heatstroke these changes were minimal. Ultrastructural analysis of hepatocytes showed that microvilli in severe heatstroke were preserved but the space of Disse was obliterated. By contrast, moderate heatstroke animals displayed disorganization, swelling, and loss of hepatocyte microvilli. In renal glomeruli there was evidence of intracapillary thrombus and activated platelets interacting with both neutrophils and endothelium. No significant changes were observed in moderate heatstroke or control animals.
DIC, Inflammation, and Endothelial Cell Activation/Injury
Heatstroke-induced DIC was assessed by plasma markers, namely significantly prolonged PT, activated partial thromboplastin time (aPTT), elevated D-dimer, and decreased platelet count (Table 1). In baboons with moderate heatstroke, DIC was mild to moderate. Histopathologic findings disclosed characteristic features of DIC, namely diffuse hemorrhage at focal sites in tissue from many organs including the liver, kidneys, and adrenals and widespread intravascular thrombi in small and medium-sized arteries and veins (Figure 2). Thrombus formation was widespread in hepatic and renal veins, and pulmonary arterioles in severe heatstroke animals. In some loci (liver and jejunum not shown), thrombi included entrapped white blood cells.
Inflammation and Endothelial Cell Activation/ Injury
Severe heatstroke elicited marked leukopenia attributable to neutropenia together with systemic inflammation, and endothelial activation/injury as assessed by increased plasma IL-6 and thrombomodulin levels, respectively (Table 1). In contrast, moderate heatstroke elicited marked leukocytosis and mild to moderate systemic inflammation and endothelial activation/injury. Both light and electron microscopy revealed accumulation and pooling of leukocytes associated with distinct margination and transmural migration in vascular beds in most of the organs examined (Figure 3A). In the liver, the sinusoids were packed with granulocytes that extended into the central vein and hepatic vein tributaries. Granulocyte margination was also present in the lungs, myocardium, glomerulus capillaries, and jejunum lamina propria. Figure 3A displays representative electron photomicrographs of lung and jejunum from severe heatstroke baboons, which confirm and extend the ultrastructural findings in kidney, as described above (Figure 1B). These comprised the presence of leukocytes in the interstitial tissue interacting with both platelets and endothelium together with cytoplasm ballooning, widening of the gaps between ECs or breach in the microvascular endothelium barrier and increased content of cytoplasmic Weibel-Palade bodies. Figure 3B shows an increased staining of endothelial vWF into capillary endothelium of liver sinusoids, myocardium, lung, and glomeruli of heatstroke compared with control animals.
Both light and electron microscopy demonstrated characteristic changes of apoptosis, namely nuclear pyknosis and fragmentation in enterocytes, inflammatory cells populating the jejunum lamina propria, lymphoid cells in the spleen, endothelial cells, and lung intracapillary leukocytes (Figure 4A). EM revealed also occasional hepatic and jejunal cells succumbing from autophagic cell death (not depicted) but no morphological evidence of necrosis.16
Figure 4B shows absence of or only a few scattered TUNEL-positive cells in control baboons. In severe heatstroke, TUNEL-positive staining indicative of apoptotic cell death was extensive in jejunum, spleen, and to a lesser extent in lungs. There were scanty TUNEL-positive cells in liver, myocardium, and kidney (not depicted). In moderate heatstroke TUNEL-positive cells were demonstrated in spleen, lung, and in jejunum.
Active caspase-3 was detected in the lung, jejunum, and spleen tissues of severe heatstroke baboons compared with controls (Figure 4C). Analysis of double stained sections revealed that most caspase-3 expression was in the area of apoptotic cells (Figure 4D).
Tissue Factor and Apoptosis
TF expression in splenic and jejunal tissues was detected in severe heatstroke animals compared with controls and colocalized with TUNEL-positive cells. (Figure 4E).
The purpose of this study was to depict the pathogenic mechanisms that underlie tissue injury and cell death in an experimental baboon model of heatstroke. The results corroborated previous clinical and experimental observations that MOSD and death in heatstroke are associated with systemic inflammation and disseminated intravascular coagulation.7–9 Importantly, the findings showed that death in baboons with severe heatstroke was associated with (1) widespread hemorrhage and thrombosis (a hallmark of DIC), (2) margination and transmural migration of leukocytes, (3) endothelial cell injury in most vascular beds, and (4) extensive apoptosis. In baboons surviving heatstroke, comparable histopathologic manifestations, albeit very subtle, were apparent at 72 hours after recovery. These findings establish a mechanistic link between an exacerbated inflammation and hemostatic host responses to heat stress and the pathogenesis of heatstroke, and suggest that endothelium and apoptosis have an important role.
DIC is a life-threatening complication of heatstroke as confirmed in this study, but the mechanisms underlying its development are not known.9,10 TF is a well-established trigger of the coagulation cascade in vivo in human and baboon model of many pathological conditions.17–19 In a recent study (submitted for publication), we showed that a large number of procoagulant microparticles exposing phosphatidylserine are released into the circulation in baboons with severe heatstroke, and they express functionally active TF. In our present study, we detected high level of TF expression in splenic and jejunal tissues in severe heatstroke animals compared with controls. Using double immunofluorescence labeling, we observed that TF colocalized with apoptotic cells in both organs, raising the possibility of an interaction between TF and membrane phosphatidylserine, which is typically exhibited at the surface of cells undergoing apoptosis.20 This interaction is known to increase markedly TF activity and also to contribute to TF shedding into the circulation carried by membrane microparticles.21,22 Taken together, these data suggested that TF pathway could be responsible for the initiation of coagulation in severe heatstroke, although this hypothesis clearly demands confirmation by studies using pharmacological inhibition or blocking antibodies of TF activity.
Systemic inflammation and endothelial activation are well established mechanisms of TF-initiated DIC in many conditions.17–19 However, in cultured endothelial cells, leukocytes or whole-blood, heat shock prevented cytokine- and lipopolysaccharide (LPS)-induced inflammation and coagulation activation by inhibiting NF-kB and TF expression, suggesting that heat exerts a protective effect.23,24 This differs from the findings in the present study, which revealed an upregulation of TF and marked activation of coagulation concomitant with systemic inflammation and endothelial cell activation as assessed by circulating IL-6 and soluble thrombomodulin, respectively. These differences between in vitro and in vivo effects of heat shock are not entirely clear, but may be in part attributable to the importance of the interaction of endothelial cells with their environment.25 Indeed, activated endothelium attracts platelets, monocytes, and neutrophils, any of which can initiate and amplify coagulation.25 Also, vascular endothelium is continuously exposed to a range of shear-stress generated by the variation of blood flow, and this can have a major impact on both function and structure of vascular endothelium.18,25 A recent study using a baboon model of E coli sepsis demonstrated that rheological factors in regions with disturbed blood flow contribute to procoagulant response by increasing TF expression and activity.18
Using immunohistology and EM, we showed in this study that severe heatstroke induced both structural and functional changes of the endothelium in several sites including the gut, kidneys, and lungs. The ultrastructural changes encompassed increased formation of ECs surface villi, cytoplasm ballooning, widening of the gaps between ECs, and increased content of cytoplasmic Weibel-Palade bodies. There was also morphological evidence of endothelial cells apoptosis. Ultrastructural analysis showed leukocytes and platelets adhering to the endothelium as well as leukocytes present in the interstitial space, probably subsequent to transmigration from the microvasculature, suggesting an active in vivo cross-talk between vascular endothelium, inflammation and coagulation pathways in heatstroke. Moreover, using immunohistochemical analysis, we found a marked increase in staining of endothelial vWF-Ag in most vascular beds. vWF-Ag is known to facilitate the adherence of platelets to subendothelium of injured ECs, particularly under high shear stress, as in heatstroke, and thus it could be instrumental in promoting thrombosis in this condition.26,27 Indeed, the observation by EM of platelets being attached to both ECs and thrombi supported this interpretation. Further studies may help unravel the molecular mechanisms whereby platelets and leukocytes interact with ECs and thereby contribute to the pathogenesis of heatstroke.
A major finding in our study was the predominance of apoptosis as a mechanism of cell death in heatstroke. Using TUNEL assay we observed increased apoptosis in spleen, jejunum, and lung, particularly in inflammatory cells (lymphocytes, plasma cells, and neutrophils) populating the lamina propria and alveolar capillaries. The increase appeared organ-specific as spleen, jejunum, and lung exhibited more apoptosis than liver, heart, and kidney. In baboons surviving heatstroke, we documented sustained apoptosis, albeit with less intensity. These findings were similar to those in rodents subjected to whole body hyperthermia, where there was widespread apoptosis in spleen, lymph node, thymus, small bowel, and cells of lymphoid origin in the lamina propria.6 Also, using in vitro approach, a number of studies have established that apoptosis can be promoted by application of heat shock.28,29 Taken together, these data imply that apoptosis may play an important role in the physiology/pathophysiology of heat stress.
Because of the high prevalence of apoptosis observed in the present investigation, we considered the possibility that TUNEL may have overestimated the actual rate of apoptosis. Indeed, DNA fragmentation similar to that seen in apoptosis can be present in necrosis.30 However, the following obviated the possibility that the high rate of TUNEL-positivity was attributable to a process other than apoptosis: (1) our tissue was obtained and fixed immediately, thus preventing postmortem changes such as are encountered in human studies,31 (2) the tissue specimens were not treated with proteinase K, a step identified as reducing TUNEL specificity, (3) TUNEL was negative in sham-heated controls but positive in heatstroke animals with morphological features of apoptosis by both H&E staining and EM, and (4) there was positive immunostaining for active caspase-3 in the area of TUNEL positive cells, also typically an apoptosis feature.32
The cause of enhanced apoptosis in baboon heatstroke is unknown. Many factors associated with heatstroke, such as hyperthermia and proinflammatory cytokines have proapoptotic potential; however, there are also other factors, such as heat shock proteins and antiinflammatory cytokines, which have antiapoptotic effects.28,32 In vitro heat shock induces caspase-dependent apoptosis, but the mechanisms are still to be elucidated.28 In the present study, double labeling analysis revealed active caspase-3 in spleen, lung, and jejunum tissues localized in the surroundings of apoptotic cells, suggesting that caspase-3 was implicated in cell death and no other immunologic processes. This was an important finding because it was not only an irrefutable confirmatory method for the TUNEL assay but it also provided mechanistic insight, namely that apoptosis in a clinically relevant animal model of heat shock was caspase-dependant. Hence, this finding affords a possibility to unravel the mechanisms of heat shock-induced apoptosis and the potential to test novel therapy. Caspase-3 is a crucial apoptotic protease in the final common pathway of a death receptor–initiated caspase-8 or intrinsic pathway and mitochondria initiated caspase-9 or extrinsic pathway.32 Pharmacological inhibition of caspase-3 has been shown to decrease the extent of injury in stroke and sepsis models.33,34
The beneficial or harmful role of apoptosis in patients with heatstroke could not be addressed in this study; however, based on our results that TF colocalized with apoptotic cells, we speculate that apoptosis contributed to the thrombogenicity of heatstroke. We further speculate that the predilection of apoptosis for lymphoid tissues (particularly the spleen) in heat stressed baboons is comparable with observations made in animal and human sepsis studies, and hence raises the possibility that apoptosis may contribute to the modulation of the immune response in heatstroke.32 In the setting of sepsis, apoptosis was shown to affect specific subsets of cells namely CD+ T-helper, B- and dendritic cells resulting in immunosuppression and thereby contributing to morbidity and mortality in this condition.32 Whether heat stress induced apoptosis of the same immune effector cells and produced the same deleterious effects merits further studies.
In conclusion, the present study identified microvascular injury, thrombosis, inflammation, and apoptosis as potential pathways of tissue injury and cell death in heatstroke. These pathways may constitute the basis for further research and potential therapeutic targets in this emergent threat.
Sources of Funding
This work was supported by Grant 2020 017 from the Office of Research Affairs, King Faisal Specialist Hospital & Research Center, Riyadh, Saudi Arabia.
Original received October 30, 2007; final version accepted March 12, 2008.
Burger FJ, Fuhrman FA. Evidence of injury by heat in mammalian tissues. Am J Physiol. 1964; 206: 1057–1061.
Sakaguchi Y, Stephens LC, Makino M, Kaneko T, Strebel FR, Danhauser LL, Jenkins GN, Bull JM. Apoptosis in tumors and normal tissues induced by whole body hyperthermia in rats. Cancer Res. 1995; 55: 5459–5464.
Bouchama A, Roberts G, Al Mohanna F, El-Sayed R, Lach B, Chollet-Martin S, Ollivier V, Al Baradei R, Loualich A, Nakeeb S, Eldali A, de Prost D. Inflammatory, hemostatic, and clinical changes in a baboon experimental model for heatstroke. J Appl Physiol. 2005; 98: 697–705.
Bouchama A, Parhar RS, el-Yazigi A, Sheth K, al-Sedairy S. Endotoxemia and release of tumor necrosis factor and interleukin 1 alpha in acute heatstroke. J Appl Physiol. 1991; 70: 2640–2644.
Malamud N, Haymaker W, Custer R. Heatstroke: a clinico-pathologic study of 125 fatal cases. Milit Surg. 1946; 99: 397–449.
Geisbert TW, Young HA, Jahrling PB, Davis KJ, Kagan E, Hensley LE. Mechanisms underlying coagulation abnormalities in ebola hemorrhagic fever: overexpression of tissue factor in primate monocytes/macrophages is a key event. J Infect Dis. 2003; 188: 1618–1629.
Mackman N, Tilley RE, Key NS. Role of the extrinsic pathway of blood coagulation in hemostasis and thrombosis. Arterioscler Thromb Vasc Biol. 2007; 27: 1687–1693.
Bombeli T, Karsan A, Tait JF, Harlan JM. Apoptotic vascular endothelial cells become procoagulant. Blood. 1997; 89: 2429–2442.
Morel O, Toti F, Hugel B, Bakouboula B, Camoin-Jau L, Dignat-George F, Freyssinet JM. Procoagulant microparticles: disrupting the vascular homeostasis equation? Arterioscler Thromb Vasc Biol. 2006; 26: 2594–2604.
Basi DL, Ross KF, Hodges JS, Herzberg MC. The modulation of tissue factor by endothelial cells during heat shock. J Biol Chem. 2003; 278: 11065–11071.
Aird WC. The role of the endothelium in severe sepsis and multiple organ dysfunction syndrome. Blood. 2003; 101: 3765–3777.
Keuren JF, Baruch D, Legendre P, Denis CV, Lenting PJ, Girma JP, Lindhout T. von Willebrand factor C1C2 domain is involved in platelet adhesion to polymerized fibrin at high shear rate. Blood. 2004; 103: 1741–1746.
Milleron RS, Bratton SB. Heat shock induces apoptosis independently of any known initiator caspase-activating complex. J Biol Chem. 2006; 281: 16991–17000.