A cardiopulmonary bypass with deep hypothermic circulatory arrest rat model for the investigation of the systemic inflammation response and induced organ damage

Cardiopulmonary bypass (CPB) is a required technique for major cardiovascular surgery. As a core component of “on pump” surgery, extracorporeal circulation and oxygenation of blood is applied. Both processes take place in a heart lung machine (HLM). The exposure of the blood to artificial surfaces activates a variety of signalling cascades which induce an inflammatory response, first described as whole body inflammation syndrome [1]. Evolutionary needed for wound healing, under unfavourable haemodynamic conditions as it can occur during CPB, this may lead in 2-10% of all cases to a systemic inflammatory response syndrome (SIRS), which may further aggravate to multiple organ dysfunction syndrome (MODS) [2],[3]. SIRS is mediated primarily by the cells of the innate immune system. Later anti-inflammatory compensatory effects are promoted by the adaptive immune response [4]. The “one hit model” proposes that a severe SIRS alone is able to induce MODS [5]. Induction of leucocytosis and secretion of the cytokines TNF-α and IL-1β by activated monocytes and macrophages are the first signs for SIRS followed by a raise in IL-6 plasma level and a switch in T_h1/T_h2 cell balance [4],[6]. The activation of the immune system is at least partially responsible for collateral tissue damage observed after CPB [7], but it has to be unlinked from the pure ischemia/reperfusion process. Ischaemia/reperfusion injuries are caused to major tissues, primarily cardiovascular and visceral organs and the central nervous system [8]. Those injuries are mediated by Ca²⁺-overload and reactive oxygen species (ROS), which amongst others are generated by infiltrating macrophages [9],[10] and mainly contribute to morbidity and mortality after successful surgery [11]. The extent of I/R-induced tissue damage is not only restricted to the cardiovascular system but also affects the kidneys, the respiratory system, the liver, the central nervous system and the intestine [12]. Until now, treatment of I/R damage on clinical scale is limited to an increase of fibrinolysis which might indirectly decrease the postoperative inflammatory response [13], whereas therapies that directly suppress I/R damage are lacking. One approach would be to counteract the induction of SIRS following the “one hit model”. For this purpose, we established a rat model which relies on preceding experiments of Jungwirth et al. [14].

Following the van't Hoff equation, lowering the temperature by 10°C decreases the metabolic rate of the myocardium by 50%. In accordance with this concept known since the 19^th century, hypothermia was successfully introduced into cardiac surgery for myocardial protection by Lewis and Taufic in 1953 [15]. Deep hypothermic circulatory Arrest (DHCA) has proven to be an effective mean of ischemia protection not only for the cardiovascular system but even more for the cerebrospinal and renal system [16].

Extending the aforementioned models, we elucidated biochemical events leading to the systemic inflammatory response associated with CPB and DHCA in multiple organs in a clinically relevant approach. We hypothesized that SIRS is not induced by DHCA but it is mainly affected by the following reperfusion, in which organ damage becomes apparent. The here presented model enabled us to determine common hemodynamic parameters and to assess a variety of circulating surrogate markers for the inflammatory response [17] as well as early alterations in protein levels and/or phosphorylation of MAPKs (Mitogen activated protein kinase), STAT3 (Signal transducer and activator of transcription) and Heat-Shock-Proteins, e.g. heme oxygenase 1 (HO-1) and heat-shock-protein 70 (HSP70), on the organ level. Elevated biosynthesis and/or activation of these proteins are triggered by I/R-induced inflammatory signals in the heart and other organs. They mediate key signalling events following I/R and the extent of their induction/activation determines the outcome of tissue adaption and inflammation after CPB and DHCA [10],[18],[19]. MAPK, STAT3, HO-1 and HSP70 are mediators of the I/R- and cytokine-induced organ damage and also potential targets for selective inhibitors or activators which may supress SIRS [2],[19]-[21]. Therefore we considered it as our primary goal to determine the organ-specific signalling status in target organs possibly affected by MODS. Based on information on hemodynamic and metabolic parameters combined with molecular I/R-induced alterations in various organs, the presented rat model appears to be a suitable experimental platform for the in-depth investigation of SIRS and associated signalling events. This may contribute to improve the outcome of patients undergoing CPB and DHCA in cardiac surgery.

All reagents had analytical grade purity and were acquired from Sigma-Aldrich if not stated otherwise.

Ischaemia/reperfusion (I/R) injury contributes to the development of SIRS which enhances morbidity and mortality after surgery requiring CPB and DHCA. The involved mechanisms and molecular pathways are not completely understood, yet. Thus, it is important to provide a suitable animal model which is capable of mimicking signalling events of I/R and inflammation in humans. Based on previously published animal models it therefore was the aim of this study to establish an appropriate animal model, giving special attention to SIRS associated with I/R in multiple organs.

The observed alterations of most of the analysed blood parameters showed, that they underlie an influence by CPB. The above mentioned increase in plasma AST activity is expected to occur after reperfusion, as it represents a marker for liver, skeletal and cardiac muscle damage. The observed decrease in AST activity during the cooling period might be due to haemodilution associated with CPB.

Although the increase of creatinine remained within the reference range, the course of this parameter indicates an impaired renal function developing throughout the experiment, possibly caused by I/R-induced renal tissue damage as reported [14].

The increased levels of cardiac troponin, CK-MB and LDH were in accordance with findings of other groups and have been reported in humans after CPB and I/R, respectively as compared to the levels before surgery [15].

The increase of IL-6 and TNF-α during reperfusion is associated with SIRS and may induce JAK/STAT signalling during CPB. The dramatic increase of IL-6 and TNF-α after the reperfusion is correlated with a strong leucocytosis. At the same time points CRP levels remained low, matching very well the conditions of a beginning SIRS for the intra-operative time frame we decided to investigate. CRP as a marker of the complement system activation is elevated only after one or two days after surgery [31],[32]. The present study could demonstrate that I/R injury as applied in the described model leads to an increase of the pro-inflammatory cytokines IL-6 and TNF-α, which can activate intracellular signalling [17]. For the interpretation of the above data it must be considered that we observed haemolysis in the reperfusion blood samples and that haemolysis can cause an increase of LDH, AST, ALT, potassium and CK-levels [16].

As a further result of SIRS and I/R organ-specific phosphorylation and expression patterns of stress proteins could be detected. As assessed by STAT3 phosphorylation, an inflammatory response was observed in all organs as expected [21]. Those findings are in agreement with the increased number of leucocytes and the higher IL-6 plasma levels in I/R-animals after reperfusion. Previous to the presented experiments and based on literature a number of I/R-induced alternations of the protein expression level and protein phosphorylation level were anticipated, particularly involving MAPK activation as well as heat shock protein induction. However, following our `cardiocentric’ and clinically derived approach those expected changes were not entirely confirmed by the presented experiments. The anticipated alterations were not present for all of the detected proteins in all organs. However, an organ-specific pattern of intracellular response to I/R has already been suggested, e.g. demonstrating divergent results for the heart as opposed to other organs [24],[33]. Especially JNK phosphorylation pattern were dissimilar for most organs, but contradictory results have been reported, indicating that JNK activation may differ in I/R injury [27],[34]. One of the major reasons for divergence in I/R-induced signalling events may be the extent of I/R that actually acts on the individual organs, but also the organ-inherent tolerance to transient ischemic periods. In case of the heart, the level of induced cardioplegia as applied in different models may represent an explanation for the differences between our results and those of other studies. In contrast, abnormal calcium levels can be excluded in our set-up as a trigger for kinase activation and heat shock protein induction, because no difference was noticed in calcium serum levels (see Table 2).

In the presented work we have chosen to undertake a final measurement of protein expression and phosphorylation at the end of the complete I/R procedure. Although this approach has proven valid to demonstrate various aspects of an ideal SIRS-I/R model, it yet may have led to a simplified picture of events occurring over the time period of the entire experiment. Likewise, the one-point detection of the read-out measures may have caused a systematic masking of kinase phosphorylation kinetics that are known to represent a highly time dependent transient effect [18],[24]. Furthermore, the truly SIRS-dependent molecular effects have to be dissected from other I/R variables by ongoing experiments. Thus, in following studies the influence of hypothermia, reperfusion and haemolysis on I/R and SIRS triggered signalling events shall be further analysed.

The following limitations may be applied to our study. Cardiac arrest was achieved by deep hypothermia, no cardioplegic solution was applied. This was done on purpose to exclude signalling induced by excessive application of potassium. Since the focus of the study centers on early signalling events which may protect from or induce organ damage, we did not investigate angiopathic and apoptotic changes induced by I/R. Moreover the transition from SIRS to MODS was not aim of this study. These points will be considered in ongoing studies.

We established a CPB rat model that can reproduce common pathophysiological and molecular alterations that are associated with the induction of SIRS and the activation of specific signalling cascades. This standardised model may serve as a tool to evaluate the extent of the inflammatory reactions and organ damage associated with I/R and SIRS and to investigate the potential of novel therapeutics in a preclinical model. It might be suitable to test the efficacy of immunosuppressive therapeutics applied in major heart surgery using CPB with and without DHCA. The contribution of the different aspects of CPB might be investigated in detail, as the role of oxidative stress and inflammation might be further discriminated by analysing the involved molecular pathways.

ME, EB, and LW carried out the in vivo experiments together with EV and also the biochemical analyses together with AP. ME and AP drafted the manuscript and contributed to the preparation of the figures. PA, HS, AL, EB and UB designed the concept of the study, contributed to the analysis and interpretation of the data, and reviewed as well as revised the manuscript. PA, AL and UB coordinated the in vivo experiments. AP and HS coordinated the work on the biochemical analyses. KK contributed significant intellectual input regarding concerning the establishment and subsequent improvement of the in vivo model and the involved physiology, was also involved in interpretation of the in vivo data, and critically reviewed and revised the manuscript. All authors read and approved the final manuscript.