Necroptosis Drives Major Adverse Cardiovascular Events During Severe COVID-19

Background The mechanisms used by SARS-CoV-2 to induce major adverse cardiac events (MACE) are unknown. Thus, we aimed to determine if SARS-CoV-2 can infect the heart to kill cardiomyocytes and induce MACE in patients with severe COVID-19. Methods This observational prospective cohort study includes experiments with hamsters and human samples from patients with severe COVID-19. Cytokines and serum biomarkers were analyzed in human serum. Cardiac transcriptome analyses were performed in hamsters' hearts. Results From a cohort of 70 patients, MACE was documented in 26% (18/70). Those who developed MACE had higher Log copies/mL of SARS-CoV-2, troponin-I, and pro-BNP in serum. Also, the elevation of IP-10 and a major decrease in levels of IL-17 , IL-6, and IL-1r were observed. No differences were found in the ability of serum antibodies to neutralize viral spike proteins in pseudoviruses from variants of concern. In hamster models, we found a stark increase in viral titers in the hearts 4 days post-infection. The cardiac transcriptome evaluation resulted in the differential expression of ~ 9% of the total transcripts. Analysis of transcriptional changes of the effectors of necroptosis (mixed lineage kinase domain-like, MLKL) and pyroptosis (gasdermin D) showed necroptosis, but not pyroptosis, to be elevated. Active form of MLKL (phosphorylated MLKL, pMLKL) was elevated in hamster hearts and, most importantly, in the serum of MACE patients. Conclusion SARS-CoV-2 can reach the heart during severe COVID-19 and induce necroptosis in the heart of patients with MACE. Thus, pMLKL could be used as a biomarker of cardiac damage and a therapeutic target.

Introduction  has been the most devastating infectious disease since the 1918 in uenza pandemic over 100 years ago [1], shaking healthcare systems and rattling the scienti c community. It is projected to cause a cumulative worldwide loss of about USD 12.5 trillion and up to 8.76 million lives through 12/2024 [2,3].
While infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causative agent of COVID-19, is often associated with respiratory and pulmonary-related diseases, it can also lead to detrimental effects on several organs and induce systemic complications. Major adverse cardiac events (MACE) (i.e., heart attacks, arrhythmias, heart failure, and strokes) are some of the most frequently diagnosed complications of COVID-19 [4][5][6][7]. MACE has been shown to occur during acute hospitalization and even once the patient is discharged from the hospital, leading to worse clinical outcomes, including higher mortality [8][9][10]. Despite these observations, the exact mechanism(s) by which infection with SARS-CoV-2 leads to MACE is unknown.
SARS-CoV-2 uses the angiotensin-converting enzyme 2 (ACE2) as the receptor to enter host cells. ACE2 is an enzyme within the renin-angiotensin system, involved in blood pressure regulation and electrolyte homeostasis [11,12]. Baseline ACE2 expression is highest in the thyroid, heart, kidney, testis, and small intestine [13]. Furthermore, the expression of ACE2 is increased in cardiac tissue upon biological stress and chronic cardiovascular diseases such as heart failure [14] and respiratory infection by Streptococcus pneumoniae or in uenza A virus [15]. These data suggest that the myocardium is susceptible to SARS-CoV-2 entry and can potentially lead to the development of MACE via direct viral infection of the cardiac tissue.
Several studies have documented the development of MACE in patients with 17], mainly those with severe infection, comorbid conditions, and older age. Moreover, some researchers have documented that MACE is also frequent in patients that survive the acute infection and are now recognized as part of the long-COVID syndrome [18]. However, the underlying mechanisms of these complications still need to be better understood. Of importance, some studies have shown the presence of SARS-CoV-2 viral particles in the cardiac tissue of patients that succumbed to infection [16, [19][20][21]. Recent reports showed that SARS-CoV-2 could infect human cardiomyocytes via ACE2, replicate, and cause cell death in vitro [22,23]. However, it is unknown whether cardiac infection with SARS-CoV-2 and the associated tissue injuries lead to the development of MACE. In addition, it is uncertain whether serum biomarkers could be used to identify patients at higher risk of developing MACE during acute COVID-19.
Here we attempted to bridge these gaps to identify potential therapeutic targets to prevent MACE in patients infected with SARS-CoV-2.

Material And Methods
This observational prospective cohort study includes experiments with hamsters and human samples gathered of subjects admitted to the Clínica Universidad de La Sabana in Chía, Colombia, with con rmed COVID-19 diagnosed by reverse transcription polymerase chain reaction (RT-PCR). All consecutive patients admitted with severe disease to the participating center between November 2019 and May 2020 were included. Data were collected prospectively by the attending physicians by reviewing medical records, laboratory data, and blood samples within the rst 24 hours of hospital admission were gathered to dissect the underlying mechanisms of MACE in these patients. This study was approved by the Institutional Review Board (IRB) of the Clínica Universidad de La Sabana, and all patients signed informed consent to participate in the study (CUS-LFR-012).

Subjects and data collection
The human cohort includes hospitalized patients older than 18 years. The de ned disease severity was de ned based on the World Health Organization criteria. Severe illness was diagnosed in patients with SpO 2 ≤ 94% on room air, including patients on supplemental oxygen, oxygen through a high-ow device, or no-invasive ventilation. Critical illness was diagnosed in patients requiring invasive mechanical ventilation and/or extracorporeal membrane oxygenation or end-organ dysfunction.
During hospital admission, the following variables were collected: demographic data, comorbidities, symptoms, physiological variables collected during the rst 24 hours of hospital admission, systemic complications, and laboratory reports. A retrospective chart review was conducted at hospital discharge to double-check the registered data.

Study de nitions
MACE is a composite outcome [24,25] that encompasses patients who develop any of the following clinical diagnoses. Cardiac arrhythmia (new or worsening): change from the sequence of electrical impulses in the electrocardiogram (EKG), compared to EKG at hospital admission or in past medical history [26]. Heart Failure (new or worsening): a clinical syndrome with symptoms and/or signs secondary to functional or structural cardiac abnormality, which may occur with or without previous cardiac disease documented through an echocardiogram, evidence of pulmonary or systemic congestion, and/or an increase of serum biomarkers such as Pro-BNP [27]. Myocardial injury: acute cardiac cell injury corroborated by the rise of serum troponin values with at least one value above the 99th percentile of the normal reference value of each local laboratory; development of pathological Q waves and/or new ischemic changes in EKG; evidence of coronary thrombus by angiography and/or new loss of viable myocardium, or regional wall motion abnormality identi ed in the echocardiogram [28,29]. Finally, stroke was de ned as a neurological de cit caused by an acute focal injury of the central nervous system by a vascular cause (i.e., cerebral infarction, intracerebral hemorrhage, or subarachnoid hemorrhage) [30].

Virus strain and hamster infection
SARS-CoV-2 isolate USA-WA-1/2020 was used for these studies. The reagent was deposited by the Centers for Disease Control and Prevention and obtained through BEI Resources, NIAID, NIH: SARS-Related Coronavirus 2, Isolate USA-WA-1/2020, NR-52281. Male golden Syrian hamsters of 3-4 weeks old were purchased from Charles River Laboratory and used at 5-6 weeks of age. All procedures were per approved IACUC (2020040AR and 2020048AR) and Institutional Biosafety Committee protocols (08496). Hamsters were infected with the indicated dose of virus intratracheally (i.t.) as previously described (31); animals were sedated with ketamine (100 mg/kg) and xylazine (5 mg/kg), and the tongue was pulled forward to visualize the trachea. Three hundred microliters of virus at a dose of 9 × 10 5 PFU of SARS-CoV-2, USA-WA-1/2020 suspension was applied to the trachea, and then the nose was covered to stimulate respiration.

Testing
All the speci c performed testing, such as the quanti cation of cytokines and chemokines, RNA sequencing, neutralization assays, and western blots are explained in the Additional le 1.

Statistical analysis
Categorical variables are presented in counts (percentages) and were evaluated through the Chi-square test or Fisher's exact test. Continuous variables with normal distribution are expressed as means (standard deviation); variables with no normal distribution are expressed as median (interquartile ranges). For continuous variables with normal distribution, the t-Student test was performed, and for variables with no normal distribution Wilcoxon-Mann-Whitney test was used. Descriptive and bivariate analysis of the information was performed to determine the association between in ammatory pro le and clinical outcomes, such as in-hospital mortality, the requirement of invasive mechanical ventilation, ICU admission, and hospital length of stay.

Results
A total of 70 patients with con rmed SARS-CoV-2 infection admitted to the intensive care unit (ICU) were included in the study (Fig. 1A). The median (IQR) age was 61.5 (50. 5 (Fig. 2C). Notably, D-dimer and presepsin were not increased in MACE patients at hospital admission (Fig. 2D, Fig. 2E).

Neutralization of SARS-CoV-2 variants and the development of MACE
Spike-speci c neutralizing antibodies are generally considered correlates of protection against COVID-19. We want to determine whether there were any signi cant differences in the neutralization titers between the MACE and no-MACE groups, which could explain the disease trajectory following ICU admission. No signi cant differences were observed in the ability to neutralize pseudoviruses from the variant of concerns: SARS-CoV-2 (D614) or Beta, Gamma, Delta, and Omicron subvariants (both BA.1 and BA.2). While a partial increase in the neutralization of D614 was observed in the no-MACE group and an increase of Delta neutralization in the MACE group, these were not statistically signi cant. Our results suggest that at the time of ICU admission, the robustness of neutralizing antibody activity does not appear to be involved in preventing MACE (Fig. 3). Moreover, these data also suggest that MACE was not associated with a particular viral variant of concern.
Hearts of hamsters infected with SARS-COV-2 show upregulation of major pathways associated with injury, cell death, antiviral immune responses, and metabolic changes.
To understand the underlying effects of severe SARS-CoV-2 infection in the heart, we used a recently established Golden Syrian Hamster model of severe infection [31]. Hamsters were infected intratracheally with SARS-CoV-2 strain USA-WA-1/2020 at a 9 X 10^5 PFU dose. Four days later, mice were sacri ced, and hearts were excised for assessment of viral titers and RNA sequencing (Fig. 4A). We observed a stark increase in viral titers (plaque forming units, PFU) in the hearts of hamsters 4 days post-infection when compared to mock-infected animals (Fig. 4B). At this time point, viral titers in the lungs were shown to start decreasing in a previous report by our group, using the same model [31]. Then, the cardiac transcriptome was evaluated by transcriptome sequencing (RNA-seq) of uninfected hearts and hearts infected with SARS-CoV-2 on day 4 post-infection. The infection resulted in the differential expression of 1,084 transcripts with a p-value of ≤ 0.05 out of 12,556 transcripts for a change of 8.63% of the total cardiac transcriptome. Gene ontology (biological processes) analysis for the signi cantly changed transcripts showed major changes in terms associated with programmed cell death (Fig. 4C), regulation of reactive oxygen species (ROS) (Fig. 4D), defense response to the virus (Fig. 4E), and carbohydrate metabolic processes (Fig. 4F). Of note, cell death and ROS activity in the heart have been linked to active pathogenesis of cardiovascular diseases [32] and cardiac damage in models of severe pneumococcal and pandemic in uenza infections [15,[33][34][35]. The stark response to viral infection shown in the transcriptional increase of factors such as I tm2, Irf7, Stat1, and Eif2a, among others (Fig. 4E), indicates a heart actively mounting a host response to the invading SARS-CoV-2 (Fig. 4B).
Hamsters and humans show the activity of programmed necrosis, i.e., necroptosis in the heart and serum, respectively.
Analysis of transcriptional changes of the effectors of the two major programmed necrosis pathways, necroptosis (mixed lineage kinase domain-like; MLKL) and pyroptosis (gasdermin D; GSDMD), showed necroptosis to be starkly elevated in SARS-CoV-2 infected hamsters compared to the hearts of the uninfected group (Fig. 5A). We used immunoblots to con rm whether these transcription changes re ected protein and activity levels. Blots for the phosphorylated MLKL (pMLKL), the active form of MLKL, in hamster hearts showed signi cant activity of necroptosis in the hearts of SARS-CoV-2-infected hamsters (Fig. 5B).
While hamsters provide a pivotal model to study SARS-CoV-2 cardiac pathogenesis, we aimed to de ne if such a strong indicator of tissue injury was present in our MACE-experiencing human cohort. Immunoblots for pMLKL showed a distinct increase in the presence of this molecular marker of necroptotic cell death in the serum of MACE patients compared to those without MACE (Fig. 5C). These results suggest necroptosis activation in patients with MACE, which correlates with the ndings described in the animal model. The complete uncropped gel and blot images can be found in the Additional File ( Figure S1) Markers of cardiac injury, in ammation, necroptosis, and circulating SARS-CoV-2 strongly correlate with the development of MACE Using an annotated heatmap, we evaluated the association of MACE with all analyzed serum biomarkers (Fig. 6A). We observed that Troponin-I and pMLKL were the main variables associated with the development of MACE in humans. Then, single linear regressions were performed to analyze these relations further. We found a strong correlation between serum pMLKL and serum Troponin-I levels with a regression coe cient of 0.3589, p = 0.0086 (Fig. 6B) in MACE patients but not in the no-MACE (Fig. 6E).
Notably, the serum viral burden evaluated by SARS-CoV-2 copies was associated with Troponin-I release in patients with MACE (Fig. 6C), with a regression coe cient of 0.2576, p = 0.03, but not in no-MACE patients (Fig. 6F). SARS-CoV-2 copies and pMLKL levels in serum were found not to correlate in both the MACE (Fig. 6D) and no-MACE (Fig. 6G) groups. Of note, the top four biomarkers for the development of MACE were observed using a mean accuracy plot and de ned to be pMLKL, Troponin-I, serum presence of SARS-CoV-2, and Pro-BNP (Additional File Figure S2). These ndings illustrate the relation between necroptosis, cardiac injury, and viral burden with the development of MACE.

Discussion
In this comprehensive assessment of the underlying mechanisms of MACE, we found that SARS-CoV-2 can infect hamster heart tissue, propagate, and induce cardiac damage by activating necroptosis. We also observed that human participants with COVID-19 who developed MACE had a higher viral burden, necroptosis activity, and markers of cardiac injury (troponin and pro-BNP) in the serum. In this cohort, we did not detect any signi cant difference in the neutralizing antibody breadth against SARS-CoV-2 variants. This observation suggests that antibody-neutralizing activity is not associated with protection from MACE progression at admission. These ndings are novel and constitute an important advance in understanding MACE in patients with COVID-19 and propose some potential therapeutic targets.
Several epidemiological studies have extensively characterized MACE during and after COVID-19 [7,18,36,37]. A recent systematic review of 150 studies with 33,805 patients showed that COVID-19 increased the risk of developing MACE. These complications were more frequently documented in patients with preexisting cardiovascular comorbidities and disease severity [38]. However, a recent report by Xie et al. showed that even mild cases of COVID-19 have a high risk of developing MACE, such as heart failure and stroke, 1-year after recovering from the disease. Cardiac dysfunction has substantial implications for the treatment of COVID-19 but also demands deep mechanistic research to aid in reducing acute and longlasting disease sequela (i.e., long-COVID) [10]. Of note, using a rodent model of infection Dhanyalayam et al. observed that SARS-CoV-2 could persist in the cardiac tissue of male and female humanized ACE2 of infected mice (USA-WA1/2020) at 10 days post-infection. The SARS-CoV-2 infection led to sex-dependent morphological changes in the hearts of mice. Males were found more susceptible to pathological changes such as immune cell in ltration, and a slight increase in brosis was observed in both sexes [39]. However, the molecular and immune determinants that modulate such pathology are unknown. The results from the presented study propose direct SARS-CoV-2-driven necroptosis and pyroptosis as a mechanism for cardiac injury and cardiovascular complications associated with COVID-19. An increase in circulatory SARS-CoV-2 signal (DNA copies/mL) is associated with increased markers of cardiac injury (Troponin-I and pro-BNP) and the presence of necroptosis effector pMLKL in serum. To corroborate this observation, we infected golden Syrian hamsters with SARS-CoV-2 and detected an increase in virus titers in the heart and the activity of necroptosis via transcriptomics and immunoblots. Taken together, our results propose a new mechanism for SARS-CoV-2-driven cardiac injury via direct cardiac infection and activation of necroptosis. It also provides a new therapeutic target to reduce the acute and long-term cardiac damage associated with COVID-19.
Most epidemiological and systematic reviews regarding MACE and COVID-19 have suggested Troponin-I increase as a major biomarker for MACE and prevalent disease outcomes, which was expected [5,7,22,40,41]. In this study, we found that viral burden and high concentrations of the necroptosis effector pMLKL in the serum were more accurate in predicting the development of MACE and possibly long-term cardiac events in convalescent COVID-19 patients. This is important as none of the MACE patients in this study had major changes in circulating cytokines and chemokines associated with cell death and injury or in biomarkers of cardiac damage driven by in ammation (i.e., presepsin). In ammation has been proposed as the primary driver of severe COVID-19 [42][43][44] and, in some cases, of MACE [42][43][44].
However, here we found that the importance of systemic in ammation in the development of MACE is relatively marginal.
Necroptosis is a highly in ammatory form of cell death regulated by the receptor-interacting serine/threonine-protein kinases (RIPK)1 and RIPK3 and MLKL [45,46]. Our study suggests that necroptosis can be a therapeutic target to reduce the acute and long-term effects of COVID-19 on the heart. Supporting this notion, our group has shown that inhibiting necroptosis or oxidative stress reduces cardiac injury observed during and after pneumococcal pneumonia [34]. Treatment with FDA-approved tyrosine kinase and necroptosis inhibitor "Ponatinib" (by inhibition of RIPK1 and RIPK3 [47]) was found to reduce collagen deposition, troponin release, and promote cardiac function up to 3 months after pneumococcal pneumonia when used as an adjunct therapy to antibiotics in a mouse model [34]. Notably, while bene cial during bacterial infection, Ponatinib blocks upstream effectors of necroptosis required in the cell defense against viruses [47,48] ; thus, future studies should aim to use selective MLKL inhibitors to reduce cardiac injury driven by SARS-CoV-2.
While this study represents a comprehensive approach to dissecting some mechanisms of MACE, it has several limitations. First, we could not study the heart tissue of patients that developed MACE as our group lacked access to these tissues. However, we performed a robust clinical and paraclinical characterization of these patients that allowed us to identify necroptosis and viral burden as crucial factors for developing MACE. Second, the number of patients included in this study was small, which limits our capacity to develop high-accuracy predictive models using the serum biomarkers identi ed in our study. However, future studies will de ne how these markers directly correlate with MACE and if additional effectors can be linked to these phenotypes. Third, the kinetics of the study is not longitudinal; however, the data presented on Log copies/mL of SARS-CoV-2 were from samples collected upon admission to ICU. Thus, MACE patients showed more viral load persistence in serum. In

Consent for publication: Not Applicable
Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.