Reformulation of the Necroptosis Pathway in Reperfused Myocardial Infarction

Necroptosis is a regulated necrosis program that mediates cardiac damage during reperfused myocardial infarction (MI/R). Preclinical studies showing limitation of infarct size by necrostatin-1, an inhibitor of RIPK1 (receptor interacting protein kinase 1), have led to the widely held assumption that necroptosis in MI/R is mediated by RIPK1 → RIPK3 → MLKL (mixed lineage kinase domain-like) signaling¹. While undertaking a systematic genetic analysis of necroptosis during MI/R in mice, we made some unexpected observations suggesting that this formulation is not correct in this context.

While present in heart tissue and cardiac non-myocytes, RIPK1 protein is undetectable, or barely detectable, in adult murine and human cardiomyocytes (Fig A–C). Further, analysis of existing mass spectrometric data shows that while RIPK1 is present in adult human hearts, it is not detected in cardiomyocytes isolated from these hearts². Because low, but functionally relevant, levels of RIPK1 might nevertheless reside in cardiomyocytes, we tested whether Ripk1 deletion would reduce infarct size in vivo. Deletion in cardiomyocytes resulting in an 86% decrease in transcripts did not alter infarct size in response to 45 min left coronary artery ischemia and 24 h reperfusion (Fig D). Similarly, infarct size was unaffected by Ripk1 deletion in fibroblasts (41% knockdown of transcripts) (Fig E). The interpretation of these results could be confounded by the fact that conditional deletion is never complete. In addition, since RIPK1 contains both a pro-necroptosis kinase domain and a pro-survival scaffold domain, we considered that simultaneous loss of both domains could balance effects on cell death. Accordingly, we studied mice with homozygous knockin of kinase-dead Ripk1 allele which, being germ line, would be present in all cells. Again, no effect on infarct size was observed (Fig F). In addition, phosphorylation of RIPK3 and MLKL during MI/R was independent of RIPK1 kinase activity (Fig G). These data demonstrate that the kinase function of RIPK1, the function that promotes necroptosis, is not needed in any cell type for cardiac damage during MI/R and raise the possibility that reduction in infarct size resulting from necrostatin-1 is an off-target effect. To investigate the latter, we compared the effect of necrostatin-1 on infarct size in control versus cardiomyocyte-specific Ripk1 knockout mice and found similar magnitudes of cardioprotection (Fig H). Taken together, these data indicate that RIPK1 does not mediate cardiac damage during MI/R in adult mice.

Figure. Necroptosis during reperfused myocardial infarction is mediated by ZBP1 → RIPK3 → MLKL, rather than RIPK1 → RIPK3 → MLKL, signaling.

A-C. Immunoblots showing that RIPK1 is undetectable, or barely detectable, in adult murine and human cardiomyocytes. D, E. Absence of effect of cardiomyocyte-specific (Myh6-Mer-Cre-Mer^tg/_ ; Ripk1^flox/flox) or fibroblast-specific (Tcf21-Mer-Cre-Mer^KI/+; Ripk1^flox/flox) deletion of Ripk1 on infarct size in response to 45 min left coronary artery distribution ischemia/24 h reperfusion in vivo. Area at risk (AAR) and infarct assessed by Evans blue and tetrazolium chloride staining, respectively. F, G. Infarct size and RIPK3 and MLKL phosphorylation were not affected by homozygous germline knockin of kinase-dead Ripk1 allele (K45A). H. Necrostatin-1 (1.65 mg/kg into LV cavity immediately before reperfusion) cardioprotects to similar extents in control and cardiomyocyte-specific Ripk1 knockout mice. I, J. Germ line deletion of Ripk3 and Mlkl decreases infarct size. K, L. ZBP1 is transcriptionally induced in the heart following 45 min ischemia/4 h reperfusion. N=3 biological replicates per condition. M. ZBP1 levels increase in adult murine cardiomyocytes subjected to hypoxia (1h) and reoxygenation for the times indicated (AnaeroPak system). N=5–6 biological replicates per time point for wild type cardiomyocytes and 3 biological replicates for knockout cardiomyocytes, N. RIPK3 co-immunoprecipitates with ZBP1 in the heart. Mice homozygous for 3 x FLAG knockin into Zbp1 used as indicated. Example representative of 5 independent experiments. O, P. Germ line Zbp1 deletion reduces phosphorylation of RIPK3 and MLKL in response to MI/R. N=3 biological replicates per condition. Panel O is a phos-tag gel. Q. Zbp1 deletion reduces infarct size during MI/R. R. Zbp1 deletion reduces cell death in adult mouse cardiomyocytes subjected to 1 h hypoxia/4 h reoxygenation. N=4–5 biological replicates per condition. In vivo experiments employed 12–16-week age male mice. All mice are on a C57BL/6J background except Mlkl KO mice and their controls which were C57BL/6N. The surgeon was blinded to genotype both during the surgery and analysis of infarcts. Infarct size was assessed at the 24 h reperfusion time point in all experiments. Data from the 179/204 (87.7%) mice that exhibited the pre-specified AAR 40–65%, myocardial pallor with coronary occlusion and restoration of myocardial color upon suture release, and that survived to the sacrifice time point are reported. Signaling changes, which are manifested earlier, were assessed at the 4 h reperfusion time point in all experiments. Data displayed as mean ± SEM. Statistics: Panels D, E, F, I, J, Q – two-tailed Student’s t-test. Panels H, L, O, R – two-way ANOVA with Tukey’s test. *, **, ***, **** – P < 0.05, 0.01, 0.001, and 0.0001, respectively. Scale bar – 1 mm. RIPK1 – receptor interacting protein kinase 1. GAPDH – glyceraldehyde 3-phosphoate dehydrogenase. cTnT – cardiac troponin T. CM – cardiomyocyte. CF- cardiac fibroblast. AAR – area at risk. LV – left ventricle. INF – infarct. Nec-1 – necrostatin −1. RIPK3 – receptor interacting protein kinase 3. MLKL- mixed lineage kinase domain-like. ZBP1 – Z-nucleic acid binding protein 1. MI/R – myocardial infarction with reperfusion. H/R – hypoxia/reoxygenation. KO – knockout. K/I – knockin. All primary data, analytical methods, and materials will be made available to other researchers and can be obtained by contacting corresponding author RNK.

Given this unexpected result, we re-examined whether necroptosis itself is critical for MI/R, as previously suggested by three studies showing that germ line Ripk3 knockout reduces infarct size³. We first confirmed those results (Fig I). However, RIPK3 can also kill cells independently of necroptosis through inducing inflammation or apoptosis, outcomes dependent on the RIPK3 scaffold rather than its kinase activity which mediates necroptosis. The core necroptosis pathway is defined by RIPK3 → MLKL, where phosphorylated MLKL is responsible for plasma membrane disruption. To further probe the importance of necroptosis in MI/R, we assessed the effect of Mlkl deletion. Germ line Mlkl knockout reduced infarct size (Fig J) to a similar extent as Ripk3 knockout (Fig I). As individual deletion of both core components of the necroptosis pathway limits infarct size, we conclude that necroptosis is indeed critical in MI/R.

If necroptosis is critical in MI/R but RIPK1-independent, how does RIPK3 get activated to bring about MLKL phosphorylation and cell death? Germ line Ripk1 deletion is lethal in perinatal life because loss of the RIPK1 scaffold unleashes CASP8-dependent apoptosis and RIPK3-dependent necroptosis. The necroptosis component can be rescued not only by deletion of RIPK3 but also by deletion of ZBP1 (Z-nucleic acid binding protein 1)⁴. ZBP1 is an innate immune sensor of Z-conformation DNAs and double stranded RNAs which can activate a type 1 interferon response or necroptosis. We hypothesized that ZBP1 substitutes for RIPK1 to activate necroptosis during MI/R. We observed that ZBP1 is transcriptionally induced during MI/R in vivo and that its levels increase in isolated adult murine cardiomyocytes subjected to hypoxia/reoxygenation (Fig K–M). Further, ZBP1 interacts with RIPK3 in heart tissue both at baseline and during MI/R (Fig N). Importantly, germ line deletion of ZBP1 reduces RIPK3 and MLKL phosphorylation (Fig O, P) and infarct size (Fig Q) during MI/R in vivo as well as killing of isolated adult cardiomyocytes subjected to hypoxia/reoxygenation (Fig R). Our findings show that ZBP1 → RIPK3 → MLKL, rather than RIPK1 → RIPK3 → MLKL, signaling mediates infarct generation in MI/R. Consistent with this model, a recent paper reported that ZBP1 knockdown reverses exacerbation of necroptosis and infarct size resulting from deletion of leucine rich repeat containing G protein-coupled receptor 6⁵. Limitations of this study: Given that ZBP1 and RIPK3 interact at baseline in the heart as has been observed in other systems, further work will be needed to determine the molecular trigger by which this interaction activates RIPK3. While the isolated cardiomyocyte experiments show that the described pathway operates in these cells, the spectrum of cardiac cell types that undergo necroptosis to impact infarct size awaits further investigation and will guide cell type-specific knockout studies. Animal protocols were approved by Albert Einstein College of Medicine and Emory University and human studies by University of Nebraska.

Non-standard Abbreviations and Acronyms

AAR

area at risk

CASP8

caspase-8

CF

cardiac fibroblast

CM

cardiomyocyte

Cre

Cre recombinase

cTnT

cardiac troponin T

flox

flanked by LoxP

GAPDH

glyceraldehyde 3-phosphoate dehydrogenase

H/R

hypoxia/reoxygenation

INF

infarct size

K/I

knockin

KO

knockout

LDH

lactate dehydrogenase

loxP

locus of X-over P1

LV

left ventricle

Mer

mutated estrogen receptor ligand binding domain

MI

myocardial infarction

MI/R

reperfused myocardial infarction

MLKL

mixed lineage kinase domain-like

Myh6

myosin heavy chain 6

RIPK1

receptor interacting protein kinase 1

RIPK3

receptor interacting protein kinase 3

Tcf21

transcription factor 21

tg

transgene

WT

wild type

ZBP1

Z-nucleic acid binding protein 1