Cardiac Depression in Pigs after Multiple Trauma – Characterization of Posttraumatic Structural and Functional Alterations

The purpose of this study was to define the relationship between cardiac depression and morphological and immunological alterations in cardiac tissue after multiple trauma. However, the mechanistic basis of depressed cardiac function after trauma is still elusive. In a porcine polytrauma model including blunt chest trauma, liver laceration, femur fracture and haemorrhage serial trans-thoracic echocardiography was performed and correlated with cellular cardiac injury as well as with the occurrence of extracellular histones in serum. Postmortem analysis of heart tissue was performed 72 h after trauma. Ejection fraction and shortening fraction of the left ventricle were significantly impaired between 4 and 27 h after trauma. H-FABP, troponin I and extracellular histones were elevated early after trauma and returned to baseline after 24 and 48 h, respectively. Furthermore, increased nitrotyrosine and Il-1β generation and apoptosis were identified in cardiac tissue after trauma. Main structural findings revealed alteration of connexin 43 (Cx43) and co-translocation of Cx43 and zonula occludens 1 to the cytosol, reduction of α-actinin and increase of desmin in cardiomyocytes after trauma. The cellular and subcellular events demonstrated in this report may for the first time explain molecular mechanisms associated with cardiac dysfunction after multiple trauma.

The purpose of this study was to define the relationship between cardiac depression and morphological and immunological alterations in cardiac tissue after multiple trauma. However, the mechanistic basis of depressed cardiac function after trauma is still elusive. In a porcine polytrauma model including blunt chest trauma, liver laceration, femur fracture and haemorrhage serial trans-thoracic echocardiography was performed and correlated with cellular cardiac injury as well as with the occurrence of extracellular histones in serum. Postmortem analysis of heart tissue was performed 72 h after trauma. Ejection fraction and shortening fraction of the left ventricle were significantly impaired between 4 and 27 h after trauma. H-FABP, troponin I and extracellular histones were elevated early after trauma and returned to baseline after 24 and 48 h, respectively. Furthermore, increased nitrotyrosine and Il-1β generation and apoptosis were identified in cardiac tissue after trauma. Main structural findings revealed alteration of connexin 43 (Cx43) and co-translocation of Cx43 and zonula occludens 1 to the cytosol, reduction of α-actinin and increase of desmin in cardiomyocytes after trauma. The cellular and subcellular events demonstrated in this report may for the first time explain molecular mechanisms associated with cardiac dysfunction after multiple trauma.
In the United States approximately 30,000 patients with blunt cardiac trauma were recorded per year 1 . Heart injury was identified as an independent predictor for poor outcome 2 and was associated with prolonged ventilation interval 3 and longer hospital stay 4 . The clinical feature of the so called "commotio cordis" is associated with dysrhythmias, including ventricular fibrillation and sudden cardiac arrest 5,6 . In patients with hypotension disproportionate to the estimated blood loss and inadequate response to fluid resuscitation cardiac injury should be considered 7 . Accordingly, myocardial contusion correlates well with the incidence of perioperative hemodynamic instability 8 and cardiac complications 9 .
Pathophysiological effects of blunt cardiac injury such as firmness of cardiac tissue was found in a pig model of blunt chest trauma 10 . In an isolated Langendorff perfused rat heart model early onset of specific impairment of the left ventricle such as an increase of left ventricular end-diastolic pressure occurred after cardiac contusion and was associated with increased cardiac specific troponin levels immediately after trauma 11 . Furthermore, a direct relationship between serum troponin levels and survival of trauma patients was identified 12 . in sham, multiple trauma followed by damage control orthopedics (DCO) or multiple trauma followed by early total care orthopedics (ETC) at indicated time points. (B) Ejection fraction (EF%) of left ventricle in sham, multiple trauma followed by damage control orthopedics (DCO) or multiple trauma followed by early total care orthopedics (ETC) at indicated time after trauma. n = 6 pigs in each group, # DCO p < 0.05. *ETC p < 0.05. (C) Time course for extracellular histone appearance in serum from pigs after multipe trauma and DCO treatment (grey bars), ETC treatment (black bars) or sham (white bars), *differences to sham procedure were significant, p < 0.05; # differences to baseline concentrations were significant, p < 0.05. For all frames n = 6 for each bar.

Results
Cardiac dysfunction and appearance of extracellular histones in serum after multiple trauma.
( Figure 1) To determine whether functional defects in cardiovascular performance occur after multiple trauma, echocardiographic parameters were assessed before, directly as well as 1, 1.5, 2, 4, 5.5, 12, 24, 27, 30, 36 and 48 h after trauma. Pigs subjected to multiple trauma demonstrated significant abnormalities in systolic parameters such as FS and EF ( Fig. 1) in both treatment groups, DCO and ETC. FS (frame A) and EF (frame B) were reduced between 4 h and 27 h after trauma, respectively, compared to baseline whereas in sham treated animals both parameters remained similar to baseline measurements. Extracellular histone concentration in serum was elevated 1.5, 3 and 5.5 h after trauma compared to baseline in both treatment groups DCO and ETC and 1.5, 5.5 and 24 h compared to sham procedure (frame C).
Elevated H-FABP and troponin I serum concentration and extravasation of erythrocytes in left ventricular tissue after multiple trauma. (Figure 2) To estimate cardiac injury after multiple trauma in pigs, serum markers specific for cardiac cell damage were assessed at different time points after trauma (Fig. 2). H-FABP serum concentration was elevated as early as 1.5 h after trauma and after 3 and 5.5 h compared to sham procedure and 1.5 h and 3 h compared to baseline after trauma in both treatment groups (DCO and ETC) and Time course of cardiac specific troponin I in serum after multiple trauma and damage control treatment (DCO, grey bars) or early total care treatment (ETC, black bars) or sham procedure (white bars). p < 0.05; *differences were significant to sham procedure, # differences were significant to baseline. For all frames n = 6 for each bar. H.E. staining of left ventricle in sham (left), after multiple trauma and DCO (middle) or ETC (right) treatment, respectively (frame C).
SCIENTIFIC REPORTS | (2017) 7:17861 | DOI:10.1038/s41598-017-18088-1 after 24 h in ETC treatment group compared to baseline. Cardiac specific troponin I serum concentrations in pigs were assessed before and at indicated time points after multiple trauma. Troponin I concentrations in serum were elevated after trauma in both treatment groups (DCO and ETC) 3 and 5.5 h after trauma compared to baseline and compared to sham procedure as well as 48 h after trauma in DCO treated animals compared to baseline. Furthermore, H.E. staining of left ventricle tissue revealed no extravasation of erythrocytes in sham animals (0/6) whereas in animals objected to multiple trauma with DCO (2/6) or ETC (5/6) treatment extravasation of erythrocytes was present.
Local inflammation in cardiac tissue after multiple trauma. (Figure 3) Homogenates and tissue sections of the left ventricle were obtained 72 h after multiple trauma followed by DCO or ETC treatment or after sham procedure, respectively. IL-1β concentration of sham treated animals was low whereas after multiple trauma in both treatment groups, DCO and ETC, IL-1β concentrations were significantly elevated (frame A). IL-6 concentration was elevated after multiple trauma/ETC (frame B) compared to sham treated animals. Western Blot analysis revealed reduction of C5aR1 protein in left ventricular tissue after multiple trauma/ETC treatment compared to sham treated animals (frame C). As assessed by immunhistochemistry in left ventricle sections after multiple trauma and either DCO or ETC treatment (frame D) diminished C5aR1 staining was identified compared to sham procedure (frame E) whereas complement receptor C5aR2 protein measurement in left ventricular tissue failed to show any differences between sham treated animals and pigs after multiple trauma (frame F).
Translocation of Cx43 with co-translocation of ZO-1. (Figure 4) To determine whether gap junction proteins in the heart were altered after multiple trauma immunohistochemistry staining of Cx43 gap junction protein in left ventricular tissue sections was performed. In sham treated animals Cx43 was located in intercalated discs whereas after trauma followed by DCO or ETC Cx43 was translocated and scattered into the cytosol (frame A). Subsequently Cx43 and ZO-1 (frame B) were co-translocated into the cytosol compared to sham procedure in both treatment groups DCO and ETC. Western blot analysis revealed increased Cx43 protein concentrations in left ventricular homogenates after multiple trauma and DCO treatment (frame C).

Opposing trend of α-actinin and desmin expression in cardiac tissue after trauma. (Figure 5)
Whereas expression of α-actinin in cardiomyocytes was reduced after multiple trauma followed by DCO and ETC compared to sham procedure (frame A), which was confirmed by fluorescence intensity measurement (frame B), expression of desmin was increased (frame C), likewise confirmed by fluorescence intensity measurement (frame D). Figure 6) Densitry of nitrotyrosine staining revealed a minor but significantly increased expression after multiple trauma in both treatment groups compared to sham procedure (frame A). Staining of caspase 3 showed a significantly increased expression after trauma and ETC treatment (frame B) compared to sham treated animals. In sham treated animals and after multiple trauma and DCO treatment PAS staining was faint whereas after trauma and ETC treatment PAS staining was more intense (frame C), indicating increased amount of glycogen. After trauma followed by ETC and DCO treatment glucose transporter 4 (GLUT 4) was decreased on cardiomycytes compared to sham treated animals (frame D), whereas GLUT 4 mRNA was slightly but not significantly reduced after trauma compared to sham treated animals 72 h after trauma (frame E).

Discussion
A linkage between cardiac depression after trauma and structural and molecular alterations in cardiac tissue is described in this report. During haemorrhage, cardiac oxygen need is enhanced in order to the increased cardiac effort to keep up adequate blood pressure with minimal blood volume, mirrored by increased heart rate 164/min compared to baseline heart rate 73/min (data not shown). In the present context, signs of nitrosative stress in cardiac tissue, were evident after trauma and might be linked to the cardiac impairment. High tissue concentrations of NO have been linked to cardiac depression during sepsis 22 . Moreover, in experimental sepsis and during ischemia increased glucose uptake via GLUT 4 and increased glycogen deposition in the heart were observed together with diminished cardiac performance 23,24 . In the present study, correspondingly increased deposits of glycogen in the heart after trauma and ETC treatment were observed. In contrast to findings during ischemia where GLUT 4 was found to be translocated from intracellular vesicles to the plasma membrane together with its up regulation and switch of their primary substrate from fatty acids to glucose to facilitate anaerobic gycolysis 25 in the present study GLUT 4 on the sarcolemma of cardiomyocytes was decreased. In mice with myocardium restricted deletion of GLUT 4 during ischemia systolic and diastolic contractile dysfunction was observed 26 . Altogether this might be an attempt to explain the reversible character of cardiac dysfunction after trauma.
Besides nitrosative stress and metabolic alteration, the systemic inflammatory response to trauma and the presence of DAMPs might be important protagonists of cardiac dysfunction after trauma. The present data confirm the appearance of histones in serum early after trauma in both treatment groups and further elevation in animals with ETC treatment, indicating increased tissue damage in consequence of femoral nailing compared to external fixation. These results are in accordance with earlier studies demonstrating elevation of circulating histones after trauma 18 and during sepsis 27 in humans. Furthermore, a significant correlation between circulating histone levels and Sequential Organ Failure Assessment (SOFA), organ failure and lethality was observed 18 . The presence of extracellular histones in plasma was associated with accumulation in the heart 28 in vivo and on the plasma membrane and the cytosol of cardiomyocytes in vitro 29 associated with impressive elevation in intracellular calcium and reactive oxygen species which has been linked to defective cardiomyocyte function 29 . In the present study synchronous time course of elevated serum histones and impaired cardiac function were observed. Correspondingly, increased extracellular histone levels after porcine multiple trauma might be linked to cardiac dysfunction. Furthermore, after multiple trauma and ETC caspase 3 in left ventricle sections was increased compared to sham treated animals. These findings are in accordance with earlier studies revealing increased apoptosis in cardiomyocytes in presence of extracellular histones in vitro 29 and cardiomyocyte apoptosis after trauma in rats 30 . Furthermore, incubation of cardiomyocytes with plasma isolated from mice after multiple trauma resulted in cardiomyocyte apoptosis in vitro 31 . However, a limitation of the study is that the impact of either treatment group, femoral nailing or external fixation, on cardiac performance might be underestimated in the present model because compared to the difference of both therapy regimes in humans, where reamed femoral nailing is associated with a considerable iatrogenic trauma, in the present study femoral nailing might be less extensive and thus the difference might be less pronounced. Further studies are needed to identify direct relation between the presence of extracellular histones occurring after trauma and the appearance of cardiac dysfunction. It is Moreover, cytokines such as IL-1β and IL-6, elevated in plasma post trauma 32 , have been shown to be cardio-depressive 33,34 . In the present study, IL-1β in left ventricular homogenates was elevated after the multiple trauma impact in both treatment groups, while IL-6 was elevated after trauma and ETC. IL-6 has been shown to correlate with the severity of tissue injury in the heart 35 . Therefore, cardiac dysfunction after trauma might be provoked by increased tissue levels of cardio-depressive cytokines. Furthermore, after trauma C5aR1 protein was reduced in both therapy concepts. This finding is in contrast to earlier studies in rats showing increased C5aR1 expression on cardiomyocytes 24 h after burn injury 36 and after cecal-ligature and puncture-induced sepsis 17 where absence of C5aR1 was associated with improved cardiac function 37 . One possible explanation for the downregulation of C5aR1 after trauma could be an internalization of the C5aR1 triggered by C5a 38 . Of note, C5aR-C5a interaction in isolated perfused hearts caused some dysfunction of rat cardiomyocytes resulting in compromised cardiac function 17 . Cleavage of the receptor by neutrophil serine protease would be a possible explanation for reduced C5aR protein in cardiac tissue after trauma 39 but MPO measurement in left ventricular homogenates did not indicate any increased neutrophil activity in cardiac tissue after trauma (data not shown).
In the present investigation mechanical damage, nitrosative stress and local inflammatory response after multiple trauma were associated with altered cell-to-cell and intracellular integrity. Cx43 was found to be trans-located from the intercalated disc region to the cytosol after trauma which was associated with the co-translocation of ZO-1. Gap-junction endocytosis results in disruption of functional contact between cardiomyocytes and disruption of coordinated spread of any electrical activation, which again is associated with a loss of mechanical and electrical coupling of cardiomyocytes 40 , arrhythmia and cardiac dysfunction 41 . Co-translocation of Cx43 and ZO-1 has been observed after partial or complete enzymatic dissociation of myocytes from intact ventricle associated with gap junction endocytosis 42 . Cell-cell communication through gap junctions such as Cx43 has been shown to partly prevent apoptosis in vitro 43 . In the present multiple trauma study and DCO concept Cx43 was upregulated in left ventricular tissue, which was associated with absent apoptosis. Taken together cardiac alterations of Cx43 distribution and expression as well as increased apoptotic rate were observed, which to our knowledge the first time that posttraumatic cardiac tissue damage and dysfunction was linked to gap junction pathology.
Further structural cardiac proteins located in the Z-lines were investigated. α-Actinin, which is co-localized with L-type Ca 2+ channels and stabilizes the muscle contractile apparatus in cardiomyocytes, was reduced in left ventricles after multiple trauma compared to sham treated animals. Z-disc proteins have been shown to act as responder to stretch or mechanical tension due to hemodynamic demands 44 . In the present context Z-disc proteins may be altered in response to shear forces applied to the heart during blunt chest trauma. Loss of proteins associated with the sarcomeric skeleton such as α-actinin may therefore contribute to ventricular dysfunction after trauma. Desmin, a major component of cardiac intermediate filaments, likewise located in the Z-lines, forming a physical link between nucleus, contractile proteins, mitochondria and sarcoplasmatic reticulum 45 , was up regulated after trauma and its distribution was altered compared to animals after sham procedure. Desminopathies such as desmin aggregates alter heart biomechanics and calcium dynamics 46 . Desmin was found to be increased in guinea pigs with heart failure 47 and in mice with diastolic dysfunction 48 . In the present study desmin disorganisation in pigs after multiple trauma might be a compensatory mechanism after mechanical damage of cardiac tissue in an early reversible stage.
Taken together, our results suggest that multiple trauma in pigs leads to early cardiac dysfunction, which is associated with cardiac cell damage, local inflammation and disturbed cytoskeletal and gap junction architecture.

Methods
This study presents partial results obtained from a large animal porcine multiple trauma model. The model has been previously described in detail by Horst et al. 49 . Multiple trauma in pigs. Pigs underwent either multiple trauma including a combination (n = 12) of blunt chest trauma, penetrating liver injury and femur fracture followed by pressure-controlled haemorrhage (mean arterial pressure of 40 ± 5 mmHg, maximal withdrawal of 45% of calculated total blood volume) for 90 min or sham procedure (n = 6). For introduction of blunt chest trauma a pair of panels (steel 0.8 cm, lead 1.0 cm thickness) was placed on the right dorsal lower chest. A shock wave was induced by a bold shot (Blitz-Kerner, turbocut JOBB GmbH, Germany) which was applied onto the panel using cattle-killing cartridges (9 × 17; Dynamit Nobel AG, Troisdorf, Germany) as previously described 50,51 . Blunt chest trauma was associated with severe signs of lung contusion and rib fractures (2-3 ribs) as accessed by computer tomography 49 . Thereafter pigs were resuscitated by re-transfusion of the withdrawn blood accompanied by additional crystalloids (Sterofundin ISO ® , 2 ml/kg body weight/h) and rewarming to normothermia (38.7-39.8 °C). Sham procedure (n = 6) included instrumentation and anaesthesia without trauma or haemorrhage. The multiple trauma group (n = 12) was randomized in two therapy arms (n = 6); external fixation of the femur fracture corresponding to damage control orthopaedics (DCO) or femoral nailing appropriate to early total care (ETC) principles.

Animals and
Transthoracic echocardiography in pigs. Echocardiograms were performed as previously described 52 .
Imaging was performed according to the recommendations using a standard cardiac ultrasound machine (Vivid I©, GE Healthcare, United Kingdom). After acquisition of M-Mode images in parasternal long axis systolic and diastolic parameters were measured and shortening fraction (SF%) and ejection fraction (EF%) were calculated. The equations used were as follows: Confocal Imaging. For confocal imaging formalin fixed and paraffin embedded heart tissue was used.
For glucose transporter 4 (GLUT 4) staining rabbit anti-pig GLUT 4 (abcam) was used as primary and goat anti-rabbit (AF-488) as secondary antibody (Jackson ImmunoResearch Laboratories, West Grove, PA, USA). Confocal imaging was performed using Leica SP8 (Leica microsystems, Wetzlar, Germany). Evaluation of fluorescence-intensity was conducted by the Software Image J 54 .
Detection of mRNA for Glut 4. Total RNA was isolated from pig heart homogenates by the TRIZOL ® method (Thermo Fischer Scientific) according to manufacturer's instructions. cDNA was generated and amplified (SYBR ® ) using reagents from Life Technologies. Amplification was performed using QuantStudio 3 (Thermo Fischer Scientific, Waltham, MA, USA). Calculation of the relative quantitative results was performed with the 2 −ΔΔCt algorithm. The following pig primers were used: GLUT 4 5′atgttgcggatgctatgggg 3′(fwd), GLUT 4 5′cctcgggtttcaggcacttt 3′(rev). For housekeeping gene Gapdh 5′gagtgaacggatttggccg 3′(fwd), Gapdh 5′aaggggtcattgatggcgac 3′(rev) were used. Statistical procedures. All values were expressed as means ± SEM. Data were analysed by one-way ANOVA followed by Dunnett's or Tukey's multiple comparison test. P ≤ 0.05 was considered statistically significant. GraphPad Prism 7.0 software was used for statistical analysis (GraphPad Software, Incorporated, San Diego, Ca, USA).