Understanding COVID-19 susceptibility and presentation based on its underlying physiology.

We have entered the third year of the COVID-19 pandemic with a record number of deaths worldwide (∼6 million, ∼1 million in the United States) and a fear of future waves of severe disease spread and mortalities. About half a billion people are affected by COVID-19, with new surges emerging in some parts of the world. Since November 2021, a new fast-spreading but apparently less severe SARS-CoV-2 mutant, Omicron (including its subvariant BA.2 and its more recent subvariant XE), has become dominant. It is anticipated that this highly infectious variant, which seems to cause relatively milder disease, would increase herd immunity and may convert the pandemic to an endemic. But these are speculations, and although vaccinations have provided some relief, hospitals are still overflowing with COVID-19 patients and great uncertainty exists as to whether more severe mutations will occur somewhere and rise rapidly again. Meanwhile, physicians and scientists strive to discover mechanisms and targets to treat the patients and literature is piling on the data from ongoing work. A wealth of knowledge is available on coronavirus collections of the American Physiological Society (APS) journals (https://journals.physiology.org/covid19). Physiological Reviews has provided its readers an important resource in form of COVID-19 collections https://journals.physiology.org/topic/physrev-collections/covid19?seriesKey=physrev&tagCode= that consist of current and previous reviews on the topic. As discussed below, some of these have covered critical issues related to the pandemic, including severe disease-causing mechanisms in susceptible populations as observed in sick, immunosuppressed, elderly, and pregnant individuals. The purpose of this commentary is to discuss and integrate findings from some of the articles published in Physiological Reviews as well as other high-impact APS journals.

The first paper discusses whether cancer predisposes patients to COVID-19 (1). Comorbid patients such as those with cancer receiving immunosuppressing chemotherapy and radiation treatments are at higher risk for pathogens including COVID-19. Here, the authors examined additional pathways by which cancer may predispose patients to COVID-19. Host metabolic dysfunction and disorder are critical determinants of the pathogenesis of COVID-19 and its severity. As put together in the review by Sica et al. (1), metabolic and immunological complications underlie the “homeostatic frailty” and COVID-19 susceptibility observed in elderly patients. These “complications” are mostly caused by imbalances between inflammatory and anti-inflammatory processes [such as increased levels of proinflammatory mediators like interleukin (IL)-6 and tumor necrosis factor-α (TNF-α) and concurrently reduced levels of anti-inflammatory mediators such as IL-10], immune senescence, and compromised resolution of inflammation. Besides balanced inflammation, physiological immune senescence, and resolution, another key feature of intact immune regulation is the speed of emergency hematopoiesis that guarantees an adequate supply of lymphoid and myeloid cells when needed. During stress and in pathological conditions such as cancer, alterations in the magnitude and composition of the hematopoietic output affect the immune response and disease outcome. In particular, the microenvironment of the hematopoietic stem cells affects the plasticity and activation status of immune cells through specific surface receptors. An example of such stem cell plasticity is seen in cancer patients where tumors induce an alternative macrophage activation state known as M2 phenotype that promotes immunosuppression to facilitate tumor growth. Therefore, metabolic alterations associated with comorbidities such as cancer, hypertension, diabetes, and aging may heavily influence the immune system during viral infection by affecting immune cell populations and decreasing antiviral activity. Viral infections such as those with SARS-CoV-2 may, on the other hand, promote classical M1 macrophage activation. Although cancer and viral infection thus promote different functional macrophages (M2 vs. M1), these counterregulatory modulations seem to be central to the pathological alterations of the immune system in COVID-19. Here it is important to note that M1 macrophages are classically activated, typically by IFN-γ or lipopolysaccharide (LPS), and produce proinflammatory cytokines, phagocytize microbes, and initiate an immune response. M1 macrophages produce nitric oxide (NO) and reactive oxygen intermediates (ROIs) to protect against bacteria and viruses. M2 macrophages are alternatively activated by exposure to certain cytokines such as IL-4, IL-10, or IL-13. M2 macrophages produce anti-inflammatory cytokines and also produce either polyamines to induce proliferation or proline to induce collagen production and are therefore associated with wound healing and tissue repair. Therefore, activation of M2 macrophages would counteract the classical proinflammatory status after viral infection and also promote tissue repair. As such, Sica et al. (1) note that macrophage activation pathways (both M1 and M2) may present important therapeutic targets to mitigate severity of COVID-19 in such comorbid populations. For example, nicotinamide phosphoribosyltransferase (NAMPT, an enzyme essential for NAD+ synthesis) was shown to be a critical modulator of mobility, immunometabolism, and the immunosuppressive or antiviral activity of M2 macrophages (FIGURE 1). Moreover, viral pathogens themselves have been shown to alter important metabolite levels such as those of NAD+ in cells involved in both innate and adaptive immunity. Thus, fluctuations in the bioavailability of metabolites like NAD+ during host-pathogen interaction may affect pathogen load, host immune response, and disease onset. This may explain the potential loss of homeostatic robustness and overwhelming inflammation observed in cancer patients affected by COVID-19. Although the studies discussed in this review highlight the susceptibility and vulnerability of cancer patients, the risk of death from COVID-19 varies across different cancer types and is still not conclusive.

FIGURE 1.

Metabolic dysfunction, altered immune response, and thrombosis at the heart of COVID-19 severity in susceptible populations. Molecular effects of SARS-CoV-2 and its many variants (as shown at top) leading to CD38 activation, NAD depletion [its hydrolysis by CD38 leads to production of adenosine diphosphate ribose (ADPR), cyclic ADPR, and nicotinic acid adeninine dinucleotide phosphate (NAADP)], cell death, and release of nucleotides and damage-associated molecular patterns (DAMPs) to cause activation of purinergic receptors, increased cytosolic calcium (Ca2+), reduced nicotinamide phosphoribosyltransferase (NAMPT), reduced antiviral activity of M2 macrophages, increased reactive oxygen species (ROS) production, cytokine storm, increased thrombosis, and metabolic dysfunction. Clinically, these effects result in overwhelming inflammation, hypoxia, hypercapnia, and hypercoagulation. Loss of homeostatic robustness results in increased disease severity in susceptible populations such as that of elderly, comorbid, and pregnant individuals.

Interestingly, Horenstein et al. (2) put forth a compelling case that the enzyme CD38, a cyclic ADP ribose hydrolase that utilizes NAD+ and its precursors [nicotinamide mononucleotide (NMN) and nicotinamide (NAM)] as substrate, is at the center of the SARS-CoV2 infection, leading to intracellular calcium mobilization, an interferon gamma response, and a reactive oxygen species burst in immune cells. CD38 is found on the surface of many immune cells. It has receptor binding activity that activates B and T cells and NAD+ hydrolyzing activity that may increase cytokine and reactive oxygen species production. Besides its NAD+ hydrolyzing activity, CD38 also has receptor binding activity that activates B and T cells and may as such increase cytokine and reactive oxygen species production. The enzymatic NAD+ hydrolysis by CD38 and the downstream metabolites [e.g., adenosine diphosphate ribose (ADPR), cyclic ADPR, and nicotinic acid adenine dinucleotide phosphate (NADP)] that arise because of its action are regulators of cytosolic calcium fluxes and key to induction of signaling pathways involved in the antiviral response and proinflammatory processes. However, viral infections are known to quickly deplete NAD+ levels in the host. In preclinical models of aging, CD38-positive immune cells were shown to accumulate in tissues of older animals, leading to NAD+ depletion. Increased accumulation of senescent cells was also observed in animal models infected with coronavirus, indicating NAD+ depletion. Similarly, senescent cells infected with SARS-CoV-2 spike protein exhibited a hyperinflammatory response that increases CD38 activity and depletes NAD+. These studies seem to support the hypothesis of a critical role for the CD38/NAD+ axis in COVID-19 pathogenesis (2) and further explain the mechanisms of susceptibility of certain comorbid populations as highlighted above by Sica et al. (1). Although CD38-positive immune cells have been shown to be increased upon infection with other viruses (e.g., HIV, RSV) (2) evidence for SARS-CoV-2 infection is still scant, and more data are required to support that CD38 content is sufficiently increased to cause NAD+ depletion in COVID-19. Notably, there are also other NAD+-consuming enzymes [poly ADP ribose polymerases (PARPs)] and receptors [aryl hydrocarbon receptor (AhR)] that are overexpressed in COVID-19 as a consequence of inflammation. AhR targets and activates both PARPs and CD38, leading to cell death primarily due to NAD+ depletion. Cell death causes release of damage associate molecular patterns (DAMPs) such as ATP and adenosine, which may in turn promote thrombosis via activation of purinergic receptors P1 and P2. Thus, another consequence of CD38 upregulation is increased thrombosis, which just happens to be also an important predictor of COVID-19 severity. Thus, multiple CD38-mediated pathways that eventually lead to NAD+ depletion, immune cell metabolic dysfunction, and thrombosis may be central to COVID-19 pathogenesis and severity (2).

To further elucidate the mechanisms of increased mortality and morbidity in COVID-19 patients with comorbidities, Ji et al. (3) reviewed the literature to conclude a clear pitch for increased plasmin activity as a common risk factor for COVID-19. Severe COVID-19 comorbidities present with increased levels of plasmin, a protease that degrades fibrin formed by the activity of thrombin during the process of thrombosis (3, 4). Ji et al. argued that the S (spike) protein of the coronavirus may be cleaved by plasmin, trypsin, cathepsins, elastase, and members of the transmembrane proteaseserine subfamily (TMPRSS). Cleavage of the S protein facilitates binding of the virus’s S1 region with angiotensin-converting enzyme 2 (ACE2) receptors on host cells, and thus probably enables virus entry and fusion. Accordingly, the authors suggest that antiproteases targeting the hyperproteolysis caused by plasminogen may present a promising approach to fight COVID-19 (3). Reduced SARS-CoV-2 clearance causing increased viral load and elevated fibrin degradation products and D-dimers in plasma are important factors related to increased mortality in older COVID-19 patients. Moreover, increased plasmin in patients with various comorbidities could increase COVID-19 pathogenesis by multiple ways. For example, plasmin can cleave up to 16 sites including the cleavage sites for trypsin, chymotrypsin, prostasin, and elastases of the human epithelial sodium channel (ENaC) γ-subunit, resulting in ENaC activation and increased Na+ reabsorption. Elevated renal plasmin results in hypertension by cleaving ENaC in the collecting tubule, which increases salt retention, causing expansion of blood volume. On the other hand, cleavage of ENaC in distal lung epithelial cells may enhance the removal of alveolar edema resulting from damage to the alveolar epithelium following SARS-CoV-2 infection. This effect may, however, be counteracted by a parallel downregulation of ENaC activity by plasmin-independent mechanisms. For example, SARS-CoV-1 has been shown to reduce ENaC activity by activating PKC (5), and similar inhibition of ENaC in response to SARS-CoV-2 infection has been proposed as a consequence of increased signaling via PKC, paracrine nucleotides, and/or cystic fibrosis transmembrane conductance regulator (CFTR) activation (6). In COVID-19, the natural defense by antiproteases such as α2-antiplasmin may not suffice to adequately limit plasmin-mediated hyperfibrinolysis. Importantly, this concept may lead one to consider tranexamic acid (TXA), a synthetic analog of lysine that binds lysine receptor sites on plasminogen and prevents its conversion to plasmin, as a promising therapeutic strategy for COVID-19 (7). On the opposite end of the spectrum, as mentioned above, plasmin is important in cleaving fibrin clots, reducing clot burden, and improving pulmonary blood flow. For that reason, a number of investigators suggested tissue plasminogen activator (tPA) as a therapy to dissolve fibrin clots in patients with ARDS, and two clinical trials have been registered to test this concept (NCT04357730 and NCT04356833) (7). However, acute bleeding is a known risk of tPA therapy, and therefore additional better alternates need to be sought. In their Letter to the Editor, Barker and Wagener (7) suggested that TXA may be a useful adjuvant therapy in the outpatient setting before onset of overt COVID-19 symptoms to prevent hospitalization and morbidity, whereas TXA may be less effective or even detrimental at later stages for the reasons discussed above. Thus, proteases such as plasmin seem to play an important role in COVID-19 pathology and may present potential, albeit double-edged, and presumable time-dependent targets for therapy.

Increased plasminogen levels are observed in the urine of pregnant women, indicating a risk for increased susceptibility to viral infection and subsequent disease severity. Indeed, pregnant women are known to be especially susceptible to viral pathogens and pneumonia. In their review, Wastnedge et al. (8) evaluated the knowns and unknowns of this unique COVID-19-susceptible group, knowledge about which is still emerging. There are several factors that can contribute toward COVID-19 severity or lack thereof in pregnancy. To accommodate the embryo, the maternal immune system is drastically modulated, and the resulting immunosuppression may alter the host response to viral infection, as well as disease onset and progression. On the other hand, alteration of chest shape and volume during pregnancy will decrease total lung capacity and diminish mucociliary clearance, leading to enhanced susceptibility to viral infection. Pregnant women also have an increased risk for coagulopathies and thromboembolic events that may compound the severity of COVID-19. Endothelial dysfunction is critical in the pathogenesis of COVID-19-related ARDS. Pregnant women with preeclampsia (where endothelial dysfunction is common), a condition associated with increased systemic and pulmonary blood pressures as well as lung injury and pulmonary edema, are therefore at further risk of severe COVID-19. This notion is supported by a large United States-based cohort of women in which pregnant women infected with SARS-CoV-2 were found to have an increased risk of hospitalization, ICU admission, and mechanical ventilation compared with infected nonpregnant women. Placental insufficiency, miscarriage, and stillbirths resulting from SARS-CoV-2 placentitis have recently been reported in a study on pregnant mothers suffering from COVID-19 (9). Therefore, special conditions such as pregnancy need to be further evaluated for the effects of COVID-19.

As discussed above, conditions such as pregnancies or comorbidities like eclampsia seem to make women vulnerable to SARS-CoV2 infection. However, severe outcomes of COVID-19 appear to be more common in male patients (10). Acute myocardial injury and long-term cardiovascular dysfunction frequently occur in male COVID-19 patients. It is plausible that sex hormones affect the SARS-CoV-2 infectivity by modulating ACE2 receptor expression and the activity of the transmembrane serine protease TMPRSS2. It was also proposed that higher ACE2 expression in cardiac cells may support its higher virulence in males; however, viremia and active viral replication in systemic organs is an extremely rare event, even in severe COVID-19. Moreover, as discussed below, cardiac effects of COVID-19 may be better attributed to the circulating mediators released as a response to the primary pulmonary infection. A role of the female hormone 17β-estradiol has also been suggested in reducing ACE2 and TMPRSS2 expression and thereby reducing the risk for or magnitude of SARS-CoV-2 infection. Since binding of SARS-CoV-2 to ACE2 may drive thrombin- and purinergic-mediated thromboinflammation and cardiovascular cell/tissue damage observed in severe COVID-19 pathology, it is expected to be less severe in females that have decreased ACE2 and increased 17β-estradiol. These studies are still emerging, and more mechanistic observations are expected to unfold.

Early on in the pandemic, autopsy studies revealed the presence of extensive pulmonary endothelialitis and immunothrombosis (i.e., an exacerbated immune response in combination with a procoagulant profile) as characteristic signs of severe COVID-19, raising the notion that endothelial activation and injury (in both the lungs and the systemic circulation) may constitute critical early events in a severe disease course. Although systemic disease manifestation and multiorgan failure constitute hallmarks of severe COVID-19, systemic analyses of blood and autopsy tissue only detected viremia and active viral replication in distal organs in a small fraction of COVID-19 cases, including those with fatal outcome. As such, it has become increasingly recognized that SARS-CoV-2 infection is commonly limited to the airways and distal lung regions, whereas extrapulmonary involvement is mostly secondary to local immune responses and the release and systemic dissemination of a cocktail of mediators (often referred to as “cytokine storm”) from the primary site of infection (FIGURE 1). Considering the air space epithelium as the primary site of infection, the endothelium thus emerges as the next immediate barrier for such a systemic dissemination in a dual sense, first for the transition of proinflammatory mediators from the air spaces to the vascular space and second for the effect of circulating mediators on the distal organs. This view is supported by analyses of lung autopsy tissue from the Kwapiszewska laboratory showing upregulation of endothelial activation markers including intercellular adhesion molecule-1 (ICAM-1), von Willebrand factor (vWF), and vascular endothelial growth factor receptor-2 (VEGFR2) that was associated with extravasation of immune cells in COVID-19 samples compared with control lung tissue (11). Consistent with the notion of endothelial cell activation, plasma concentrations of various endothelial markers including vascular cell adhesion molecule-1 (VCAM-1), E-selectin, and CD31 were found elevated in plasma of COVID-19 patients, along with increased levels of inflammatory cytokines including IL-6, IL-8, monocyte chemoattractant protein-1 (MCP-1), or TNF-α (7). Notably, upregulation of adhesion molecules such as CD31 and CD38 (see above) may lead to lymphocyte exhaustion in COVID-19 (2). Additionally, the “cocktail” of inflammatory mediators has the ability to injure by itself the endothelial layer, as demonstrated by Michalick and coworkers (12), who showed that plasma from COVID-19 patients can directly disrupt interendothelial junctions and impair lung endothelial cell barrier function in vitro as a function of disease severity. Notably, all plasma samples from COVID-19 patients in this study were negative for SARS-CoV-2, substantiating the view that systemic viremia is a rare event and not required for systemic disease dissemination in severe COVID-19 (12). From these studies, a scenario emerges in which initial endothelial injury in the lung allows for the systemic release and dissemination of inflammatory mediators via the bloodstream, which in turn activate endothelial cells in various organs (and, in a vicious circle, again the lung), thus exerting local and systemic proinflammatory, procoagulatory, edematogenic, and, ultimately, fibrogenic effects. From such an endotheliocentric view, it seems surprising that, so far, efforts to improve or conserve endothelial barrier function in COVID-19 patients (and, as such, to “confine” the disease to the lung) have been relatively sparse, despite an abundance of preclinical data highlighting the potential of barrier-protective strategies such as the administration of adrenomedullin, sphingosine-1-phosphate, transient receptor potential antagonists, angiotensin-(1–7), or angiopoietin-1 analogs, to name a few.

In addition to alveolo-capillary barrier failure, the development of permeability-type pulmonary edema as a clinical hallmark of COVID-19 and cause of systemic hypoxemia and subsequent systemic organ involvement is further aggravated by a parallel impairment of alveolar fluid clearance mechanisms. In the intact lung, alveolar fluid homeostasis is warranted by an intricate regulation of transepithelial ion transport that in turn drives alveolar fluid absorption by osmotic gradients. At the apical surface of the alveolar epithelium, the most prominent ion channels involved in this physiological process are ENaC and the chloride channels chloride intracellular channel protein 5 (CLIC5) and cystic fibrosis transmembrane conductance regulator (CFTR), respectively, with the basolaterally expressed Na-K-ATPase generating the necessary gradients for ion and fluid transport. As it turns out, these molecular building blocks of alveolar fluid clearance are, however, particularly sensitive to infectious and inflammatory diseases, including COVID-19. Downregulation of ENaC and CFTR protein abundance at the alveolar epithelial plasma membrane has long been recognized as a characteristic feature of pneumonia, including infection with SARS-CoV-1. As such, coexpression of hydrophilic coronaviral proteins with ENaC subunits in Xenopus oocytes decreases amiloride-sensitive Na+ currents and γ-ENaC protein levels at their plasma membranes in a protein kinase C-dependent manner (5). Over and above that, SARS-CoV-2 seems to exert additional inhibitory effects on the expression and regulation of Na-K-ATPase, as recently outlined in a comprehensive perspective by Kryvenko and Vadász (13). Notably, SARS-CoV-2 infection may negatively impact on Na-K-ATPase by a combination of different mechanisms, ranging from SARS-CoV-2 hijacking the transcriptional, translational, and posttranslational machinery of the alveolar epithelium, thereby impairing Na-K-ATPase de novo expression, maturation, and trafficking (including targeting of Na-K-ATPase to the apical instead of the basolateral plasma membrane), to increased endocytosis and proteasomal degradation of Na-K-ATPase (again, reducing its abundance at the basolateral plasma membrane), to indirect effects with clinical characteristics of COVID-19 such as hypoxia, hypercapnia, inflammatory mediator release, and increased coagulation activity, all negatively impacting on Na-K-ATPase abundance and activity. In combination, these direct and indirect mechanisms can be expected to shut down alveolar fluid clearance at large, thus eliminating the main physiological rescue mechanism that could (at least partially) protect the lung from permeability-type alveolar edema. Over and above that, Na-K-ATPase dysregulation in COVID-19 can be expected to further impair barrier function and promote alveolar epithelial cell death given its additional role as adherence junction-forming molecule and as regulator of tight junctions via phosphorylation of occludin and its essential relevance for cell survival via the maintenance of physiological ion homeostasis at the plasma membrane (13).

Albeit alveolar edema is a cardinal feature of severe COVID-19, clinicians from early on in the pandemic have frequently also observed so-called “happy hypoxemics,” i.e., patients who present with severe hypoxemia in the absence of dyspnea and seemingly rather mild-to-moderate disease severity in lung imaging by chest X-ray or computer tomography. Although the seemingly paradoxical lack of dyspnea in these patients has been alternatively attributed to viral invasion into the central nervous system or basic principles of respiratory physiology not unique to COVID-19, the emergence of severe hypoxemia in seemingly only mildly affected lungs has initially baffled physicians and physiologists alike. Once again, autopsy studies proved instrumental to fuel important functional insights into underlying mechanisms. Specifically, analyses aimed to define the architecture and microanatomy of blood vessels in the distal lungs from serial hematoxylin and eosin (H&E) sections and subsequent computed three-dimensional (3-D) image reconstruction revealed patency of prominent intrapulmonary bronchopulmonary anastomoses (IBAs) in lungs of COVID-19 patients (14). These IBAs connect pulmonary arteries and bronchial arteries and therefore allow the deoxygenated blood to bypass the alveolocapillary compartment by creating a right-to-left shunt with profound hypoxemia. IBAs play a prominent functional role during fetal life and then close at birth but may be recruited again under specific disease conditions including inflammatory and infectious airway disease or thromboembolic events, all of which are characteristic for COVID-19. These anatomical findings have been corroborated at the functional level by a recent study that retrospectively analyzed bedside pulse oximetry measurements at different inspired oxygen fractions to calculate right-to-left shunt from oxygen-hemoglobin dissociation curves constructed by use of a two-compartment mathematical model (15). In a total of 87 admitted COVID-19 patients of whom 31 (36%) died, intrapulmonary shunt was on average 45% higher in patients who died than in those who survived, yet ventilation-to-perfusion ratio (V̇a/Q̇) did not differ. Mortality rate increased with shunt severity and reached 100% in patients with a shunt > 30%. These data support the notion of extensive intrapulmonary right-to-left shunts, possibly via recruited IBAs, as pathological explanation for the profound hypoxemia seen in COVID-19 patients. Although computation of intrapulmonary shunts from pulse oximetry data may provide for a “simple” bedside test with predictive value for outcome, calculations may potentially be affected by COVID-19-related, yet shunt-independent, changes of the oxygen dissociation curve due to systemic acidosis, increased body temperature, or levels of methemoglobin or 2,3-bisphosphoglycerate in blood, as extensively discussed elsewhere (16). More importantly, the reported findings on IBA potency and intrapulmonary shunts in COVID-19 stress the need for a better mechanistic understanding on the regulation of IBA as an important and potentially therapeutically targetable mechanism of systemic hypoxemia.

GRANTS

This work was supported by the CounterACT Program, National Institutes of Health Office of the Director (NIH OD), the National Institute of Neurological Disorders and Stroke (NINDS), and the National Institute of Environmental Health Sciences (NIEHS), Grant Numbers U01ES028182 (to S.A.), R21ES030525 (to S.A.), 5UO1ES027697 (to S.M.) and R21ES032956 (to S.M.).

DISCLOSURES

S. Matalon is an editor of Physiological Reviews and was not involved and did not have access to information regarding the peer-review process or final disposition of this article. An alternate editor oversaw the peer-review and decision-making process for this article. None of the other authors has any conflicts of interest, financial or otherwise, to disclose.

AUTHOR CONTRIBUTIONS

S.A. prepared figures; S.A., S.M., and W.M.K. drafted manuscript; S.A., S.M., and W.M.K. edited and revised manuscript; S.A., S.M., and W.M.K. approved final version of manuscript.