Remote ischemic preconditioning attenuates intestinal mucosal damage: insight from a rat model of ischemia–reperfusion injury

Remote ischemic preconditioning (RIPC) is a phenomenon, whereby repeated, non-lethal episodes of ischemia to an organ or limb exert protection against ischemia–reperfusion (I/R) injury in distant organs [1]. This effective, simple and low-risk procedure can be easily induced by transient occlusion of blood flow to an arm with a blood pressure cuff [2]. Despite intensive research there is still an apparent lack of knowledge about the RIPC-mediated mechanisms. Different studies indicated that besides neurogenic pathways and a systemic anti-inflammatory response, humoral mediators released into the systemic circulation by the RIPC-stimulus might play a key role in transferring the protective signal to the remote target tissues [1, 3].

Depending on the RIPC-mediated mechanisms, various physiological and pathophysiological processes can be influenced and regulated within the target tissue/organ. Several authors have shown that RIPC is able to attenuate apoptotic events induced by I/R injury in different organs [4‐6]. Besides the regulation of apoptosis, preventing oxidation of cellular macromolecules by increasing antioxidant capacity within the target tissue seems to be an equally important mechanism of RIPC-mediated organ protection [7‐9]. RIPC effects may also be mediated by hypoxia-inducible factor-1α (HIF-1α), a key transcription factor regulating cellular adaptation to hypoxia [10, 11], although the underlying mechanisms are only poorly understood so far [12, 13]. We have recently shown that RIPC is also able to reduce the activity of matrix metalloproteinases (MMPs) in serum and cardiac tissue of patients undergoing cardiac surgery and that these events are associated with decreased cellular injury in the target tissue, suggesting MMP activity as another potential factor in RIPC mediated organ protection [14, 15].

Experimental and clinical studies have mainly focused on the protective effects of RIPC on myocardial, cerebral and renal I/R injury. However, some studies also suggested beneficial effects of RIPC in the intestine [16, 17], in which I/R injury is associated with high mortality rates [18]. In the present study, we have established a rat model of intestinal I/R injury to evaluate possible protective effects of RIPC on intestinal I/R injury. Employing intestinal tissue and blood we furthermore analyzed the cellular and molecular mechanisms of intestinal I/R injury as well as the impact of RIPC on these events.

Remote ischemic preconditioning (RIPC) can be simply induced by repeated cycles of transient occlusion of blood flow to a limb by inflating a blood pressure cuff and exerts potent protection from I/R injury in different organs. The majority of RIPC studies mainly focused on the potential effects of RIPC on myocardial, cerebral and renal I/R injury [23, 24]. Although intestinal I/R-injury is associated with extremely high mortality rates and RIPC could represent a promising treatment option, the RIPC-mediated effects and the underlying mechanism in the intestine are only poorly investigated.

Dickson et al. showed for the first time in a rat model that RIPC is able to increase the hypoxic tolerance within the intestine after I/R injury. Effluents from preconditioned hearts restored the intestinal motility after I/R-injury in the isolated rat jejunum, suggesting that humoral mediators are responsible for the protective effects [25]. Saeki et al. revealed, that 3 cycles of 15 min RIPC reduce the intestinal injury in a rat model of small bowel transplantation [26], while other studies using only one cycle of hindlimb ischemia did not show protective effects on intestinal I/R injury [27]. Based on the observation that repeated cycles of I/R are required for RIPC to be efficient, in the rat model employed in this study, repeated bilateral hindlimb RIPC (3 × 5 min) was applied directly before induction of intestinal I/R-injury (30 min of ischemia and 60 min of reperfusion). Our decision to use bilateral hindlimb ischemia instead of conditioning only one location originated from observations of animal studies that suggested that bilateral RIPC might be more effective than unilateral RIPC [28, 29].

Histomorphological analyses of the intestinal tissues revealed fewer signs of mucosal injury in RIPC treated animals compared to the control group (I/R injury only). Moreover, we showed a statistically significant reduction of the intestinal injury score (Chiu-Score) in the RIPC group, indicating cell and tissue protective effects of RIPC in our model. Our results are concordant with the findings from Saeki et al. implicating the existence of an “early window” of RIPC-mediated protection in the intestine [26], similar to the described windows for cardioprotection, where a biphasic course of RIPC-mediated protection with an early and late window was proposed [30]. In the heart, and possibly also in our intestinal I/R model, the early window of protection is believed to be caused by a fast activation of intracellular signalling pathways involving already preformed molecules, while the late window of organ protection is characterised by (slower) transcriptional changes and de-novo synthesis of protective proteins [30, 31]. LDH measurements in sera of all animals revealed a significant increase in LDH release after I/R. RIPC reduced the levels of LDH to values that were statistically not different from the sham group. Although serum LDH values of the I/R and IR + RIPC group did not differ significantly, the results are in line with the histological findings and Chiu-scoring. The lack of statistically significant differences between the I/R and I/R + RIPC group and the only moderate increase of LDH release after I/R might be due to the fact that blood and serum were obtained 60 min after the ischemic insult and that this time point might have been too late to detect a significant systemic increase of LDH activity. This hypothesis is supported by the results of others who have shown that levels of serum markers for intestinal cell damage are increased directly at reperfusion and not at later time points [32].

Several authors suggested that anesthesia is a confounder of organprotection by RIPC especially in the heart and that propofol anesthesia impairs the cardioprotective effect of RIPC [33, 34]. According to a meta-analysis by Zhou et al. and a recently published study by Cho et al. use of volatile anesthetics also attenuates the cardioprotection afforded by RIPC [35, 36]. Although, no studies have so far investigated whether or not anesthesia is also a confounder of intestinal protection by RIPC, we cannot rule out this possibility. In our study, in all animals and all treatment groups (including controls) interventions including the RIPC procedure were performed on the “background” of sevoflurane anesthesia and under these conditions RIPC was able induce protection from intestinal I/R injury. Therefore, RIPC is effective, even under sevoflurane anesthesia. Nevertheless, it remains unclear whether RIPC mediated intestinal protection is to some extent influenced by sevoflurane or if sevoflurane does alter secondary parameters of I/R and RIPC (e.g. protein expression, hydrogen peroxide generation, MMP activity).

Whereas intensive research over the last years has identified several proteins that are responsible for the RIPC-mediated mechanism of cardioprotection, there is an apparent lack of knowledge concerning the respective mediators in the intestine. To screen for possible RIPC-regulated proteins, antibody-based cell stress protein arrays were employed. These arrays simultaneously detect the expression levels of 26 proteins involved in cell stress and cellular defence mechanisms. Our results showed that compared to sham operated animals Cytochrome C, Cited2 and ADAMTS1 were strongly expressed in intestinal tissues from rats submitted to I/R. RIPC applied directly before the induction of I/R injury resulted in higher expression levels of Cytochrome C and ADAMTS1, while Cited2 was downregulated by RIPC.

Interestingly, Cytochrome C, a component of the mitochondrial electron transport chain, which plays also an important role in the initiation of apoptosis [37], was upregulated by RIPC. This observation is unexpected as we detected protective effects of RIPC on the intestinal I/R injury and would assume reduced apoptosis. It is commonly accepted that translocation of mitochondrial Cytochrome C into the cytosol leads to complex formation of Cytochrome C, Caspase-9 and Apaf-1, the so-called Apoptosome, which initiates cellular death by the activation of effector Caspases like Caspase 3 and 7 [37, 38]. Besides its role in electron transfer in the respiratory chain and the initiation of apoptosis, Cytochrome C is also known to possess antioxidative functions [37, 39]. Some studies reported an early increase of mitochondrial Cytochrome C protein levels after drug-induced apoptosis and these results were interpreted as an early cellular defence mechanism to avoid apoptosis [40‐42]. In our experimental setting total protein of intestinal tissue samples was isolated and fractional analyses were not performed. Therefore, we cannot discriminate between effects caused by cytosolic or mitochondrial Cytochrome C and can only speculate about the role of the molecule in RIPC mediated mucosal protection. However, Caspase 3 and 7 activities in the intestinal tissues were significantly increased after I/R, suggesting that at least some of the histomorphologically visible mucosal injury is related to apoptosis. Interestingly, RIPC did not alter the Caspase activity in any of our study groups although RIPC was able to reduce mucosal injury. These findings suggest that RIPC does not reduce I/R injury via an attenuation of Caspase 3 and 7 mediated apoptosis and that the upregulation of Cytochrome C by RIPC in our animal model is most likely not related to apoptotic events via of Caspase 3 and 7. To further analyse possible antioxidative effects of RIPC, H₂O₂ formation and malondialdehyde (MDA) serum levels were measured. Neither intestinal I/R injury nor RIPC were able to change MDA serum levels. However, there was a slight increase of H₂O₂ serum concentrations after I/R injury and a trend of reduced H₂O₂ levels in all RIPC groups.

Proteome profiling revealed a RIPC-mediated increased expression of ADAMTS1 within the intestine after I/R injury. ADAMTS1 (a disintegrin and metalloproteinase with thrombospondin motifs 1) is a member of the MMP family involved in several physiological and pathophysiological processes like cellular differentiation, angiogenesis, inflammation, cancer development and arteriosclerosis [43]. Only little is known about the exact role of ADAMTS1 during hypoxia and I/R-injury. However, several authors suggested a HIF1-α induced upregulation of ADAMTS1 under hypoxia in endothelial cells and an early expression after myocardial infarction [44, 45]. Recently, studies from our group indicated an involvement of MMPs in RIPC-mediated organ protection [14, 15]. As ADAMTS1 also belongs to the class of MMPs, we extended our study and further evaluated the activities of different MMPs after I/R injury using gelatin zymography. Our results did not show any regulation of MMP-2 activity in intestinal tissue or serum, although several studies suggested that in heart, brain, kidney and liver a decreased MMP-2 activity after I/R injury is associated with reduced tissue injury and organ dysfunction [14, 46, 47]. Instead, intestinal I/R injury lead to a slightly increased gelatinase activity at 130 kDa in the serum samples and RIPC almost completely abolished this activity in the I/R and sham group. Other authors have described a gelatinase activity containing molecule of 130 kDa as Lipocalin-2/MMP-9 complex and contributed its function to inflammatory processes and tumour invasion [48]. Lipocalin-2 increases the proteolytic activity of MMP-9 and also protects the enzyme from degradation [48]. Whilst a reduced activity of MMP-9 is known to be associated with protection against I/R injury [49, 50], the possible effects of Lipocalin-2 and Lipocalin-2/MMP-9 complex on I/R injury is still not clear and controversially discussed [51‐54].

As mentioned above, protein profiling of intestinal tissues revealed an increased expression of Cited2 after intestinal I/R injury, whereas RIPC was associated with decreased expression levels of Cited2. Cited2 is reported to be a negative regulator of HIF-1α and is activated under hypoxic conditions [55]. HIF-1α, a key transcription factor regulating cellular adaptation to hypoxia by activating several genes involved in antioxidative defence, apoptosis, glucose metabolism and angiogenesis, is promoting cellular survival during hypoxia [10, 31]. Whereas several studies indicated a pivotal role of HIF-1α in the process of local preconditioning, only little is known about the effects of HIF-1α during RIPC-mediated organ protection [12, 13]. Previous studies from our group demonstrated, that RIPC is able to increase HIF-1α protein levels in cardiac tissues from patients undergoing cardiac surgery and that this effect is associated with decreased Troponin-T levels as marker for cardiac injury [56]. Moreover, Weber et al. demonstrated that RIPC plasma protects human endothelial cells cultured in vitro from hypoxia-induced cell damage and that HIF-1α could play a key role in these events [20]. These results are also in accordance with newer studies demonstrating, that a RIPC-induced upregulation of HIF-1α is linked to decreased tissue injury after cerebral [57] and myocardial I/R [58, 59]. Our western blot analyses of intestinal tissues derived after I/R injury showed that RIPC leads to significantly increased HIF-1α protein levels, supporting the hypothesis, that RIPC is able to increase HIF-1α levels within the target tissue. Although current knowledge assigns HIF-1α a central role in RIPC mediated protection, further studies (i.e. using HIF-1α knockout models) are necessary to clearly show the direct involvement of HIF-1α in organ protection, especially in the intestine. Proteome profiling arrays detecting the phosphorylation status of 45 kinases showed an increased phosphorylation of 13/45 proteins in the I/R group (compared to the sham group). Interestingly, RIPC did not lead to an additional increase in phosphorylation of signaling kinases. These results are somewhat surprising, as numerous authors have postulated that the protective effects of RIPC in the heart are associated with an increased phosphorylation of pro-survival kinases (e.g. Erk1/2 and Akt) [12, 60, 61]. The increased phosphorylation of several signaling molecules by I/R may be interpreted as an intrinsic cellular defense mechanisms to control I/R-induced damage. Our observation, that RIPC does not further increase the phosphorylation status of these molecules suggests that the kinases investigated in our study are not responsible for the described protective effects of RIPC on intestinal I/R-injury and that RIPC-mediated protective mechanisms in the intestine differ from the mechanisms described for the heart. However, compared to the heart which is predominantly composed of cardiomyocytes, the composition of intestinal tissue is more complex, including various different cell types such as endothelial cells, epithelial cells, fibroblasts, immune cells etc. As the signaling arrays were performed using intestinal tissue homogenates, RIPC-mediated effects on one cell type (e.g. endothelial cells) might be masked by other cell types which did not respond to RIPC. We can also not exclude the possibility, that the time period from induction of RIPC (T 10 min) to sample preparation (T 150 min) might have been too long so that RIPC induced phosphorylation had already decreased.

There are several potential limitations of our study that need to be discussed: (i) As a result of the experimental setup, intestinal tissue samples could only be obtained at one time point at the end of the reperfusion period. Therefore, we cannot provide information about the temporal dynamics of I/R and RIPC associated cellular and molecular mechanisms. Regarding the ROS production, Caspase activity and LDH release, earlier time points may provide better insights into possible RIPC mediated antioxidative, antiapoptotic and cell protective effects, while later time points may be more suitable in revealing differences in MMP expression/activity. (ii) Human cell stress proteome profiling arrays were employed in the study. Although the manufacturer confirmed the suitability and cross reactivity of the arrays with rat tissue, we cannot exclude slightly different binding capacities of the array specific antibodies. However, arrays were performed with pooled tissue samples and only used as screening tool for the detection of cell stress proteins and pathways that could be involved in I/R-injury and RIPC. The respective proteins or downstream targets were in the following steps further analysed using unpooled samples and different biochemical methods. (iii) Although there is convincing evidence that the 130 kDa factor with gelatinase activity resembles a Lipocalin-2/MMP-9 complex, further biochemical analyses are necessary to confirm this hypothesis beyond doubt. (iv) Our study suggests that RIPC underlying mechanisms involve HIF-1α, ADAMTS1, Cited2, Cytochrome C and a decreased serum activity of a 130 kDa factor with gelatinase activity. However, functional interactions or a causal connection of these molecules were not demonstrated and further experiments (e.g. expression knock-down or function blocking studies) are needed to substantiate the findings of our work.

Taken together, our results show RIPC mediated protection against intestinal I/R injury in rats and suggest that the underlying mechanisms may involve HIF-1α and a decreased serum activity of a 130 kDa factor with gelatinase activity. Further work has to clarify, whether RIPC also has the potential to preserve intestinal barrier function in patients suffering from intestinal I/R injury and is able to reduce the devastating consequences of intestinal I/R-injury in the clinic.