A role for NLRP3 inflammasome in acute myocardial ischaemia-reperfusion injury? Reply

We thank Mezzaroma et al. and Jong and Zuurbier for their interest in our article,1 and also for their valuable comments and concerns regarding our results and interpretations. As stated by Mezzaroma et al., several studies have now addressed the role of inflammasomes in the heart. We fully agree with the authors that in order to understand the role of NLRP3 in the myocardium, identification of the cell types involved in NLRP3-mediated responses is of great importance. We also share the authors' claim that cardiomyocytes are the most prominent cell type in the heart. Likewise, loss of contractile tissue is also the most important consequence of a myocardial infarction. However, this does not necessarily imply that the NLRP3 inflammasome needs to be functional in the cardiomyocytes. In fact, while we clearly do not exclude a functional role of NLRP3 inflammasome in cardiomyocytes, we, in line with the study of Kawaguchi et al.,2 suggest that myocardial fibroblasts, being a potent inflammatory cell, could contribute to NLRP3-mediated inflammation during ischaemia-reperfusion (I/R), and thereby also affecting cardiomyocytes. In the study by Mezzaroma et al., immunofluorescence images indicate that inflammasome activity in cardiomyocytes in vivo in post-MI and in vitro studies in HL-1 cardiomyocytes supports this finding.3 However, the study of Kawaguchi et al.2 does not find inflammasome activity using neonatal mouse cardiomyocytes. Moreover, investigations in our lab have failed to reveal inflammasome activity in adult mouse cardiomyocytes (unpublished data). Thus, a role for the NLRP3 inflammasome in cardiomyocytes is in our view at present unclear, but an important topic in forthcoming studies. We acknowledge that the report by Zuurbier et al.,4 in PLoS One last year should have been considered and referred to in our study. This article did not find any difference in cardiac function and cell death comparing Wt and NLRP3−/− hearts during ex vivo I/R, but revealed highly interesting results on ischaemic preconditioning.4 Unfortunately, we find no apparent explanations for the discrepancy between the Zuurbier study and our study.1,4 To respond to the concerns of Jong and Zuurbier, we admit that the pre-ischaemic cardiac function of the hearts in our study was unstable and that this is not ideal. However, we are confident that this has not had a major impact on our results. Importantly, when calculating left ventricular developed pressure during reperfusion relative to pre-ischaemic values, there were still significant differences between Wt and NLRP3−/− hearts. Moreover, pre-ischaemic coronary flow was similar and within normal range in all hearts, and thus, this cannot explain our reported differences in ischaemic damage and cardiac function. Finally, we would like to stress that all genotypes were strictly randomized and studied using the same system with the same detectors and the same balloons. Nonetheless, the discrepancies between our investigations underscore the need for further studies to elucidate the role for NLRP3 inflammasome within the heart. The ischaemic damage that occurs in the Langendorff model of I/R tends to vary between different experimental setups, mostly depending on unknown factors. We refer to the study of Martinov et al.,5 where the ischaemic damage was relatively low (30% with 40 min ischaemia). Our results are more in line with the recently published study by Lei et al.6 The ex vivo I/R work in this study was performed at the same time as the NLRP3 study, using the same equipment and achieving a similar degree of ischaemic damage (60% with 40 min ischaemia). Zuurbier et al.4 do not present direct data on ischaemic damage in their study, and we cannot directly compare the degree of damage in our studies. Basal expression of NLRP3 in the heart is low. Still, Kawaguchi et al.2 found reduced infarct sizes in ASC−/− mice 48 h after I/R in vivo, partly dependent on resident myocardial cells and supporting an ‘early’ effect of inflammasomes. As pointed out, we did not find significant differences between Wt and NLRP3−/− hearts with regard to IL-1β and IL-18 production upon ex vivo I/R, but there were strong trends and significant reduction in ASC−/− hearts.1 The latter may reflect that ASC is central in several other inflammasomes of potential importance in the heart, and that it is not beneficial to inhibit all myocardial inflammasome activity in the ex vivo I/R model. However, it may also be argued that the studies by Zuurbier et al. and Sandanger et al. point towards important inflammasome-independent effects of myocardial NLRP3, as both studies demonstrate a clear difference in phenotypes comparing NLRP3 and ASC-deficient hearts. This is in line with studies performed in renal I/R injury which show NLRP3 effects independent of inflammatory cytokine production.7 A central issue in future exploration of NLRP3 in the heart will be characterization of its inflammasome-dependent and inflammasome-independent properties and their relative importance in different forms of cardiac stress and injury. Given recent reports on the association of NLRP3 with the mitochondria,8 studies of NLRP3 in the very mitochondria-dense cardiomyocytes will be of particular relevance. Conflict of interest: none declared.