Mitochondrial Regulation of EC Health via mtROS

At low levels, mtROS regulate homeostatic processes like EC autophagy, immunity, and inflammation through oxidative signaling pathways (Tang et al., 2014). mtROS oxidatively activate Fyn and Ras proteins that phosphorylate the redox-sensitive p90 Ribosomal S6 Kinase (p90RSK), setting off a kinase cascade that leads to ERK5 S496 phosphorylation (Abe et al., 2000; Vu et al., 2018). Phosphorylated ERK5 then upregulates phosphorylative activation of NF-kB, leading to production of proinflammatory cytokines such as tissue necrosis factor (TNF) and interleukin 6 (IL-6) (Liu et al., 2017; Vu et al., 2018). mtROS can further promote inflammation by oxidizing thiols and inducing formation of a disulfide bond in the NLRP3 inflammasome, which, when fully assembled and activated, produces the IL-1β and IL-18 proinflammatory cytokines (Bae and Park, 2011; Zhou et al., 2011; Abais et al., 2015). These inflammatory cytokines increase expression of adhesion molecules that bind to leukocytes and recruit immune cells, such as macrophages, neutrophils, and T cells, to the site of inflammation (Mai et al., 2013; Ragab et al., 2020). mtROS also mediate cellular adaptation to hypoxia by preventing hydroxylation-associated inhibition of hypoxia-inducible factors, which translocate to the nucleus and increase transcription of genes involved in red blood cell production, angiogenesis, and glycolytic enzymes (Bell et al., 2007; Sena and Chandel, 2012; Schieber and Chandel, 2014). While these processes are essential for cellular responses to environmental stresses, it is also necessary for cells to remove excess mtROS, as elevated levels of mtROS can inflict severe oxidative damage on proteins, DNA, and lipids, which induces mitochondrial dysfunction and inflammation (Tang et al., 2014). The primary mechanism by which cells prevent the accumulation of mtROS involves rapidly inducible superoxide dismutases (SODs) that convert O₂^– to H₂O₂, which is then reduced to H₂O by antioxidants such as catalase and glutathione (Garlid et al., 2013; Tang et al., 2014). Inhibition and dysregulation of these antioxidant pathways lead to increased levels of ROS, resulting in oxidative stress (Figure 3).

Open in a separate window

FIGURE 3

Proposed mechanisms of SARS-CoV-2 mediated increase in mtROS production. When SARS-CoV-2 spike proteins bind to the ACE2 receptor to enter ECs, ADAM17 is thought to be activated and cleave the ACE2 ectodomain, leading to downregulation of ACE2. This leads to an accumulation of Ang II that binds to AT1R, a G-protein-coupled receptor. The binding of Ang II to AT1R activates downstream effectors like c-Src, PKC, and Rac-1, which facilitate assembly and activation of NADPH oxidase. NADPH oxidase produces ${\text{O}}_2^{-}$ that reacts with NO to form ONOO^– and is also converted to H₂O₂ by SOD1. ONOO^– and H₂O₂ both inactivate ETC complexes, which results in increased production of ${\text{O}}_2^{-}$. Mitochondrial SOD2 can then convert ${\text{O}}_2^{-}$ to H₂O₂, which is able to diffuse across membranes. TMPRSS2: transmembrane protease, serine 2; ACE: angiotensin-converting enzyme; ETC: electron transport chain; mtROS: mitochondrial reactive oxygen species; SOD: superoxide dismutase; c-Src: proto-oncogene tyrosine-protein kinase Src; PKC: protein kinase C.

Mitochondrial Dysfunction in Endothelial Pathologies

Dysfunctional mitochondria, characterized by dysfunctional ETC activity and lower ATP production, are known to produce higher levels of mtROS (Davidson and Duchen, 2007; Murphy, 2013). This increased oxidative stress and subsequent upregulation of inflammatory cytokine production can facilitate endothelial dysfunction, which increases endothelial permeability and aggregation of pro-thrombotic platelets (Stenvinkel, 2001; Davignon and Ganz, 2004). Excessive recruitment of immune cells in response to increased inflammation is thought to be one of the primary mechanisms responsible for tissue damage and lung injury in the COVID-19 cytokine storm (Ragab et al., 2020; Ye et al., 2020). Furthermore, both ROS and inflammatory cytokines (TNF and IL-6) disrupt the endothelial barrier between blood vessels and tissue by activating the actin-myosin contractile apparatus, which pulls on the cadherin junctions that link adjacent ECs inward and away from each other (Millar et al., 2016). The endothelial barrier can be further weakened by EC apoptosis (Kluge et al., 2013; Chambers et al., 2018). One of the primary ways mtROS initiates EC apoptosis is by oxidizing intermembrane cardiolipin, which normally has a high affinity for the electron shuttle cytochrome c (Garrido et al., 2006). This causes cytochrome c to be released into the intermembrane space and cytosol, where it allosterically activates apoptosis-protease activating factor 1, which recruits and activates caspases that eventually execute apoptosis through mass proteolysis (Garrido et al., 2006; Bratton and Salvesen, 2010). This loss in endothelial barrier stability is considered the driving force behind the increased pulmonary edema observed in lung injury and ARDS (Millar et al., 2016; Fan et al., 2020). Thus, mitochondrial dysfunction and subsequent dysregulation of mtROS production represent a potential contributor to the lethal cytokine storm and inflammatory pathologies associated with SARS-CoV-2 infection, which will be discussed in greater detail in later sections.

Proposed Mechanisms of SARS-CoV-2-Induced Mitochondrial Dysfunction

Given the extensive damage mtROS can cause, it is critical to examine how SARS-CoV-2 infection can induce mitochondrial dysfunction in ECs. The increase in Ang II levels may be one of the mechanisms responsible for elevating mtROS production and subsequent EC pathologies, as Ang II has been shown to increase ROS production via increased NADPH oxidase (Nox) activation (Liang et al., 2018). Past studies have found that a 4-h, 200 nmol/L treatment of Ang II increased mitochondrial H₂O₂ production by 1.5-2 times in ECs, an increase mediated through Nox-generated ROS (Doughan et al., 2008). Ang II binding to AT₁R, which is coupled to a heterotrimeric G-protein, leads to increased production of diacyl glycerol that activates protein kinase C (Lipp and Reither, 2011). Protein kinase C regulates the phosphorylative activation and translocation of Nox subunit p47phox to the cell membrane (Choi et al., 2008; Wynne et al., 2009; Lipp and Reither, 2011; Nguyen Dinh Cat et al., 2013). Importantly, protein kinase C and c-Src can also activate the small GTPase Rac1 through downstream phosphorylation of GDP to GTP (Figure 1; Harrison et al., 2003; Ohtsu et al., 2006; Gianni et al., 2008). Rac1 then further promotes NADPH oxidase activity by binding to Nox activator p67phox, which contains an activation domain necessary for O₂^– production via electron flow in a GTP-dependent manner (Ueyama et al., 2006; Welch, 2008; Nguyen Dinh Cat et al., 2013; Schroder et al., 2017). Research also indicates that increased levels of Ang II are associated with increased transcription and expression of Nox isoforms and subunits (Sachse and Wolf, 2007; Brown and Griendling, 2009; Nguyen Dinh Cat et al., 2013; Panday et al., 2015). Further research is required to understand the precise control and kinetics of Nox activation following AT₁R activation.

O₂^– created by the NADPH oxidase then react with NO to form ONOO^–, which is known to damage and inactivate the mitochondria’s respiratory complexes through tyrosine nitration and cysteine oxidation (Pacher et al., 2007; Doughan et al., 2008; Dikalov and Nazarewicz, 2013). This inactivation of ETC complexes, one of the hallmark initiators of mitochondrial dysfunction, results in increased reduction of oxygen and subsequent mtROS production due to accumulation of electrons that are unable to be used for ATP production (Pacher et al., 2007; Doughan et al., 2008; Dikalov and Nazarewicz, 2013; Murphy, 2013; Korge et al., 2016). NO can also nitrate ETC complexes but does not contribute as much to increased production of mtROS due to its most sensitive target being complex IV, which does not synthesize O₂^– (Tengan and Moraes, 2017). Furthermore, H₂O₂ formed by dismutated O₂^– can permeate the mitochondrial membrane and oxidize the thiol side chain of cysteine, causing further ETC dysregulation and increased mtROS production (Pacher et al., 2007; Mailloux et al., 2014). Ang II and G-protein-mediated activation of protein kinase C epsilon (PKC-ε) has also been shown to result in phosphorylation and opening of mitoK_ATP, thus further increasing mtROS production (Nunez et al., 2017).

Dysfunctional Mitochondria and Excess mtROS Form Self-Feeding Loop That Pose Risk for Chronic EC Dysfunction

The induction of mitochondrial dysfunction poses chronic EC health risks beyond the initial response to viral infection. Notably, elevated levels of mtROS form a dangerous feedback loop with dysfunctional mitochondria, perpetuating a cycle that promotes inflammation in ECs (Zorov et al., 2014). The aforementioned formation of ONOO^– also occurs within the mitochondrial matrix through mitochondrial O₂^–, which further damages the ETC (Pacher et al., 2007). The activation of p90RSK by mtROS and subsequent ERK5 S496 phosphorylation may further exacerbate oxidative stress, as evidenced by increased levels of redox-sensitive p90RSK enzymatic activity, by inhibiting transcription of the antioxidant nuclear factor erythroid 2-related factor 2 (Singh et al., 2019). mtROS-mediated NLRP3 inflammasome activation and redox-sensitive kinase cascades produce inflammatory cytokines like IL-1β and TNF, which can activate NADPH oxidase and further perpetuate oxidative stress (Figure 1; Kaur et al., 2004; Yang et al., 2007). It has also been suggested that H₂O₂ can stimulate NADPH oxidase’s production of O₂^–, representing another potential feedback loop triggered by SARS-CoV-2 infection (Figure 3; Doughan et al., 2008). H₂O₂ also oxidizes thiols in the zinc finger of PKC-ε, which further increases mtROS production by opening mitoK_ATP and allowing for K⁺ influx (Garlid et al., 2013).

Furthermore, O₂^– generated by both the mitochondria and NADPH oxidase can induce oxidative modifications to DNA sugar-phosphate backbones, resulting in mutations and DNA damage (Li et al., 2013; Dominic et al., 2020b). The damage to mitochondrial DNA (mtDNA) caused by ROS is thought to further induce mitochondrial dysfunction due to improper assembly of ETC complexes (Ide et al., 2001; Doughan et al., 2008). However, this is not conclusively known, as it has been reported that accumulation of mtDNA mutations in mice resulted in reduced ATP production but showed no evidence of increased ROS damage (Kujoth et al., 2005; Trifunovic et al., 2005). Notably, high amounts of mtDNA mutations are correlated with greater induction of apoptotic markers, suggesting that the persistent, feedback loop-fueled production of mtROS still poses a great risk for EC apoptosis and subsequent development of ARDS edema and lung injury (Kujoth et al., 2005).

H₂O₂, the mtROS that more easily diffuses across membranes, reacts with intracellular iron to produce highly reactive hydroxyl radicals that induce nuclear DNA strand breaks (Mello Filho et al., 1984; Balasubramanian et al., 1998; Lee et al., 2011). Zinc-dependent Poly ADP-ribose polymerase-1 (PARP-1) is one of the main DNA damage sensors that binds to these single- and double-strand breaks and activates various DNA repair mechanisms through ADP-ribosylation (Hocsak et al., 2017; Ray Chaudhuri and Nussenzweig, 2017; Dominic et al., 2020b). However, it can be over-activated when there is extensive DNA damage resulting from persistent oxidative stress (Dominic et al., 2020a). This increase in PARP-1 activation also increases expression of c-Jun N-terminal kinases (JNK) and p38 mitogen-activated protein kinase in response to cellular stress (Hocsak et al., 2017; Dominic et al., 2020a). Activated JNK phosphorylates c-Jun, which then helps form the AP-1 transcription factor that suppresses SOD gene expression, thus exacerbating the accumulation of O₂^– (Riera et al., 2015). Activation of JNK and p38 mitogen-activated protein kinase pathways is also associated with changes in mitochondrial membrane potential (MMP) and apoptosis (Hocsak et al., 2017). Critically, extreme changes in MMP in either direction compromise the cell’s antioxidant system, which also increases mtROS production (Li et al., 2013). These wide-ranging causes and consequences of elevated mtROS levels demonstrate the importance of investigating potential SARS-CoV-2-activated feedback loops that exist between mtROS production, mitochondrial dysfunction, and inflammation (Figure 1).

Dysfunctional Mitochondrial Metabolism

A phenomenon that has gone relatively understudied is SARS-CoV-2’s effect on mitochondrial metabolism, which may be relevant because many viruses increase glycolytic metabolism and TCA intermediate concentrations in ECs to ensure successful replication (Goodwin et al., 2015; Sanchez and Lagunoff, 2015). Interestingly, lipid metabolic analysis of SARS-CoV survivors found elevated levels of cysteine, aspartic acid, and alanine, which are all intermediates in the tricarboxylic acid (TCA) cycle (Wu et al., 2017). This study is of particular interest because it suggests that SARS-CoV-2, which has proteins that are highly homologous to those of SARS-CoV, may cause similar long-term metabolic changes that further promote chronic oxidative stress and inflammation in ECs (Xu et al., 2020). Increasing TCA cycle metabolism poses a risk for even greater endothelial dysfunction, as high levels of metabolites overburden the ETC, resulting in increased electron leakage and associated mtROS production (Groschner et al., 2012). Additionally, succinate oxidation by Complex II drives production of inflammatory IL-1β in macrophages, while an accumulation of fumarate binds to the antioxidant glutathione and mediates an increase in oxidative stress (Zheng et al., 2015; Martinez-Reyes and Chandel, 2020). Notably, while succinate levels were not found to be increased in the serum of COVID-19 patients, the authors of the study in question hypothesized that succinate levels may have been significantly lowered by the patients’ intense respiratory therapy at the time of blood draw (Thomas et al., 2020). Of further interest is the finding that lipopolysaccharide, which is known to increase succinate levels, has been correlated with more severe COVID-19 outcomes; for example, newer smokers and patients with pneumonia and other lung infections – demographics associated with high lipopolysaccharide levels – are known to be at greater risk of COVID-19 complications and mortalities (Petruk et al., 2020).

Furthermore, low NAD + /NADH ratios are associated with vascular dysfunction and have been observed in senescent ECs (Csiszar et al., 2019; Dominic et al., 2020b). This reduction in NAD + levels is compounded by the shift of infected and senescent cells to more glycolytic metabolism, as glycolysis takes up NAD + to yield NADH for electron transfer in mitochondrial ATP synthesis (Birch and Passos, 2017). These findings are corroborated by the fact that activated macrophages, which also shift toward glycolysis for ATP production, show increased levels of mtROS and inflammation (Mills et al., 2016). Low NAD + /NADH ratios are characteristic of minimal need for ATP and promote passing of electrons from complex I to oxygen, thus producing O₂^– (Widlansky and Gutterman, 2011). The DNA damage repair initiated by PARP-1 further exacerbates the reduced amount of NAD + , as PARP-1 requires NAD + to donate ADP-ribose (Fang et al., 2014; Birch and Passos, 2017). This NAD + depletion leads to decreased expression of NAD + -dependent SIRT1, a deacetylase that enhances activity of the transcriptional coactivator PGC-1α (Gomes et al., 2013; Zhou et al., 2017; Dominic et al., 2020b). PGC-1α increases expression of genes involved in mitophagy, the process by which dysfunctional or injured mitochondria are destroyed by autophagosomes. Thus, PGC-1α reduction resulting from NAD + depletion may further perpetuate mitochondrial dysfunction in ECs following SARS-CoV-2 infection (Figure 4; Fang et al., 2014; Vainshtein et al., 2015).

Open in a separate window

FIGURE 4

Oxidative stress results in NAD+ depletion and perpetuates mitochondrial dysfunction. Oxidative damage of telomeric DNA invokes the PARP-1 DNA damage response, which consumes NAD+ in order to ADP-ribosylate DNA and proteins. This activation of PARP-1 exacerbates accumulation of mtROS by decreasing expression of SOD1 and decreasing MMP. The excess mtROS also induce senescence via oxidative telomere damage, which leads to a metabolic shift toward glycolysis. This increased glycolytic activity results in further NAD+ depletion in ECs, which, in turn, leads to decreased expression of PGC-1α. Without PGC-1α to promote expression of mitophagy-related genes, the dysfunctional mitochondria are not properly removed from the cell and are thus able to continue producing more mtROS, contributing to a chronic state of oxidative stress in ECs. MMP, mitochondrial membrane potential; mtROS, mitochondrial reactive oxygen species; NAD+, nicotinamide adenine dinucleotide; NADH, nicotinamide adenine dinucleotide + hydrogen; SOD1, superoxide dismutase 1; PARP-1, poly ADP-ribose polymerase 1; JNK, c-Jun N-terminal kinase; PGC-1α, peroxisome proliferator-activated receptor gamma coactivator 1-alpha.

These alterations in endothelial metabolism have been associated with vascular dysfunction and atherosclerosis in aging patients (Sabbatinelli et al., 2019). Thus, addressing the glycolysis pathway may be a crucial point of intervention, as indicated by the inhibition of the pro-inflammatory IL-1β in mice treated with the glycolytic inhibitor 2DG (Tannahill et al., 2013). However, there are some conflicting reports that suggest endothelial senescence has no effect on, or may even actually reduce, glycolytic activity in ECs, making the precise role of senescence on EC glycolytic metabolism controversial (Urata et al., 2019). Overall, current evidence suggests SARS-CoV-2 infection engenders dysregulated mitochondrial metabolism in ECs, thus contributing to the proposed pathways of infection-induced oxidative stress and mitochondrial and endothelial dysfunction (Table 1).

Chronic Inflammatory Effects of Senescence

This proposed induction of senescence downstream of viral infection has additional negative implications for EC health. Previous research shows that senescent epithelial cells and fibroblasts are more susceptible to viral infection, as evidenced by higher levels of viral proteins from both influenza and Varicella zoster viruses in the senescent cells (Kim et al., 2016). Researchers hypothesize that senescence causes increased viral susceptibility by downregulating expression of interferons, a class of cytokines that are known to limit viral replication (Fritsch and Weichhart, 2016; Kim et al., 2016). Senescence may also impact the regulatory effects of mitochondrial ACE and ACE2, as aged rats exhibit higher mitochondrial ACE/ACE2 ratios that are associated with increased ROS production and oxidative stress (Escobales et al., 2019). Notably, the p53 transcription factor, which can initiate cell cycle arrest and senescence, can also reduce expression of manganese superoxide dismutase, thus perpetuating oxidative stress and subsequent senescence (Moiseeva et al., 2009; Passos et al., 2010). Thus, senescence engages in a vicious cycle of oxidative stress damage and inflammatory cytokine production, similar to the synergistic interactions between dysfunctional mitochondria and excess mtROS production (Figure 5).

This conclusion suggests that it is critical to prevent mitochondrial dysfunction and senescence following SARS-CoV-2 infection. Importantly, even after the inflammatory cytokine storm has passed and the patient has recovered from COVID-19, senescence remains irreversible, and senescent cells can accumulate long after their causative stress, as is the case in radiation-induced SASP (He and Sharpless, 2017). This is particularly relevant, as we have discussed how SARS-CoV-2 may induce continuous excess ROS production and p90RSK-mediated ERK5 S496 phosphorylation (Vu et al., 2018). Previous research indicates that HIV cART treatment increases long-term risk for cardiovascular disease by inducing p90RSK-mediated SASP, so SARS-CoV-2 infection may have a similar effect in inducing chronic senescence and inflammation in ECs (Singh et al., 2019). This further exacerbates the oxidative stress-mediated induction of SASP, production of inflammatory cytokines, and subsequent dysfunction in ECs.

Senescence also increases risk for EC apoptosis by decreasing phosphorylation and expression of endothelial NOS activity, thus reducing the protection NO offers against apoptosis and, in turn, chronically increasing risk of vascular leakage and lung injury (Hoffmann et al., 2001; Chambers et al., 2018). NO is known to deactivate caspase via S-nitrosylation of the caspase’s catalytic cysteine site, thus inhibiting one of the primary executors of apoptosis (Dimmeler and Zeiher, 1999; Liu and Stamler, 1999). Senescence-induced loss of NO exacerbates the already decreased levels of NO resulting from NO reacting with excess ROS and may explain the results of Varga et al.’s study, which found immunohistochemical staining patterns consistent with EC apoptosis in COVID-19 patients (Varga et al., 2020). Notably, extensive pulmonary EC apoptosis has been observed in the lungs of ARDS patients, further supporting the proposed role of senescence in promoting EC apoptosis and associated COVID-19 pathologies (Abadie et al., 2005). Together, these reports suggest that SARS-CoV-2 can induce SASP and subsequently promote chronic EC inflammation and further damage from oxidative stress.

SASP-Produced TGF-β Perpetuates Redox Imbalance and Senescence

One of the cytokines released in SASP is transforming growth factor beta (TGF-β), which has been viewed as a significant contributor to pulmonary pathologies in severe COVID-19 cases (Tominaga and Suzuki, 2019; Chen, 2020). While TGF-β is synthesized and secreted as an inactive complex bound to latency-associated protein, mtROS can oxidize the latency-associated protein and reduce its affinity for TGF-β, in turn freeing and activating TGF-β (Pociask et al., 2004; Liu and Desai, 2015). Once activated, TGF-β binds to TGF-β receptor 2, which recruits, phosphorylates, and activates TGF-β receptor 1 (Pardali et al., 2017). This TGF-β complex then phosphorylates various Smad proteins that translocate to the nucleus and bind to promoter regions in the DNA, upregulating and downregulating expression of various genes (Pardali et al., 2017). Smad proteins can repress expression of the telomerase reverse transcriptase gene, which results in reduced telomere length and subsequent senescence (Li et al., 2006; Tominaga and Suzuki, 2019). Smad proteins have also been shown to increase expression of the NADPH oxidase Nox 4 subunit, which produces ROS that fuel previously discussed pathways associated with mitochondrial dysfunction and increased mtROS production (Jiang et al., 2014a). TGF-β is thought to further induce redox imbalance by suppressing the expression of glutamate cysteine ligase, the rate-limiting enzyme in synthesis of the antioxidant glutathione, although the mechanism of this effect remains unclear (Liu and Gaston Pravia, 2010; Liu and Desai, 2015). These findings implicate TGF-β, mtROS, and SASP in a chronic feedback loop, wherein the production of TGF-β as a result of SARS-CoV-2-mediated SASP fuels further oxidative stress and inflammation (Figure 6).

Thrombosis and Coagulopathy

Atherosclerosis and Associated Cardiovascular Disease

Anti-inflammatory Therapies

The cytokine storm in COVID-19 plays a significant role in initiating potent and widespread inflammatory EC damage. Thus, much of the current work concerning acute treatment of the cytokine storm via anti-inflammatory drugs can help to mitigate the initial damage that may lead to long-term dysfunctional phenotypes in multisystem ECs. Two of the most discussed drugs in recent literature and popular media are chloroquine and hydroxychloroquine, which are meant to treat SARS-CoV-2 infection by preventing the induction of inflammatory cytokines like TNF and IL-6 (Jean et al., 2020; Schrezenmeier and Dorner, 2020; Ye et al., 2020). Although these drugs were shown to have some positive effects in preliminary cohort studies, newer studies suggest that chloroquine and hydroxychloroquine do not significantly correlate to recovery and provide little to no benefits as treatment options (Arshad et al., 2020; Boulware et al., 2020; Gao et al., 2020; Gautret et al., 2020; Geleris et al., 2020; Mahevas et al., 2020; Mitja et al., 2020; Rosenberg et al., 2020). A 19-year-old female taken off of HCQ found better results and full recovery when given colchicine, an anti-inflammatory inhibitor of NLRP3 and IL-6 that is normally implicated in the treatment of CVDs (Leung et al., 2015; Kajikawa et al., 2019; Tardif et al., 2019; Recalcati et al., 2020). The GRECCO randomized clinical trial (RCT) tested colchicine as a potential therapeutic for COVID-19 and found that it significantly reduced clinical deterioration in treatment groups relative to control groups (Deftereos et al., 2020). Interestingly, colchicine also displayed anti-thrombotic properties in COVID-19, as treatment groups in the study exhibited significantly lower D-dimer levels (Deftereos et al., 2020; Rabbani et al., 2020). More importantly, however, the recent RECOVERY trial found that Dexamethasone, an anti-inflammatory glucocorticoid, significantly reduced mortality in COVID-19 patients who needed respiratory support (RECOVERY Collaborative Group et al., 2020). Although these results are promising, more clinical studies need to be done to confirm the efficacy of Dexamethasone so that it can potentially become a widely used therapeutic. The RECOVERY trial presents an interesting and novel proposal for using anti-inflammatory glucocorticoids in COVID-19 treatment (RECOVERY Collaborative Group et al., 2020). Future studies should evaluate the safety and efficacy of Dexamethasone and other glucocorticoids as a means of mitigating the downstream effects of the initial cytokine storm that may propagate acute and chronic inflammation.

The Prospect of Confronting EC Senescence

The selective targeting of senescence may be another means by which the potential acute and long-term effects of SARS-CoV-2 infection can be mitigated, as stimulated senescent cells produce SASP proteins that can facilitate systemic inflammation. Based on recent evidences, some of the most promising anti-senescent therapeutics include treatments that decrease SASP biomarker levels (McHugh and Gil, 2018; Amaya-Montoya et al., 2020). Most studies that have displayed significant inhibition of SASP biomarkers have focused on the suppression of signaling mechanisms (e.g., NF-kB, mTOR, p38MAPK, BRD4) in fibroblasts rather than ECs (Chien et al., 2011; Freund et al., 2011; Moiseeva et al., 2013; Laberge et al., 2015; Tasdemir et al., 2016; Wang et al., 2017). However, inhibition of the pro-inflammatory JAK-STAT pathway and TNF-α decreases SASP expression in HUVECs (Xu et al., 2015; Prattichizzo et al., 2016). The central idea of reducing SASP biomarker levels through the inhibition of signaling pathways also suggests the potential of blocking p90RSK activity, which activates and maintains senescent EC phenotypes through NF-kB activation and telomere shortening (Figure 5). Future studies should explore the use of p90RSK inhibitors, such as FMK-MEA, as potential therapeutics for EC senescence caused by SARS-CoV-2 infection. Furthermore, senescent cells can be effectively targeted through a class of drugs known as senolytics, which are aimed at inducing apoptotic pathways in senescent cells (McHugh and Gil, 2018; Amaya-Montoya et al., 2020). For example, Navitoclax, a potent senolytic drug, works by impeding the actions of anti-apoptotic Bcl proteins in senescent cells and has robust apoptotic outcomes in HUVECS (Zhu et al., 2016; McHugh and Gil, 2018). Although more investigation into the efficacy of senolytic drugs is needed, they may help in preventing acute and chronic complications of COVID-19 by reducing senescent cell burden in recovering patients.

Antioxidant Therapies

Antioxidant therapies are another class of potential therapeutics for COVID-19 and its associated pathologies. Theoretically, antioxidants would attenuate the extensive presence of mtROS within ECs and could prevent the chronic oxidative damage proposed by this review (Figure 8). This line of thinking explains the proposition of using supplementary antioxidant substances like vitamin C and melatonin in future COVID-19 treatment-related studies (Carr, 2020; Zhang et al., 2020). In particular, vitamin C has been linked to improved outcomes in septic patients, making it a conceivable candidate for study in hospitalized COVID-19 patients (Fowler et al., 2014; Zabet et al., 2016; Marik et al., 2017). However, although many licensed antioxidants are meant to target COVID-19 related pathologies (i.e., inflammation, atherosclerosis, CVDs), studies on the safety and/or efficacy of these types of treatment options in humans are limited (Zhang et al., 2016). Moving forward, more cohort studies on antioxidant treatments are necessary, as many practitioners continue to treat COVID-19 patients with antioxidants, such as IV vitamin C and zinc, without extensive support from clinical data.

We would like to express our sincere gratitude to Dr. Jun-ichi Abe (Department of Cardiology, MD Anderson) and Dr. Dale J. Hamilton (Center for Bioenergetics, HMRI) for their valuable feedback and critical input throughout the writing and editing process. We would also like to credit biorender.com, which was used to design the majority of our figures.