Controlling the Burden of COVID-19 by Manipulating Host Metabolism

Abstract

The coronavirus disease 2019 (COVID-19) pandemic caused by the coronavirus—severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) continues to cause global health problems, but its impact would be minimized if the many effective vaccines that have been developed were available and in widespread use by all societies. This ideal situation is not occurring so other means of controlling COVID-19 are needed. In this short review, we make the case that manipulating host metabolic pathways could be a therapeutic approach worth exploring. The rationale for such an approach comes from the fact that viruses cause metabolic changes in cells they infect, effective host defense mechanisms against viruses requires the activity of one or more metabolic pathways, and that hosts with metabolic defects such as diabetes are more susceptible to severe consequences after COVID-19. We describe the types of approaches that could be used to redirect various aspects of host metabolism and the success that some of these maneuvers have had at controlling other virus infections. Manipulating metabolic activities to control the outcome of COVID-19 has to date received minimal attention. Manipulating host metabolism will never replace vaccines to control COVID-19 but could be used as an adjunct therapy to the extent of ongoing infection.

Introduction

Throughout human history, mankind has been exposed to novel infectious agents derived from other residents of our planet. Oftentimes, these became the cause of major diseases, particularly when many humans relinquished the hunting and gathering lifestyle and instead lived in communities that farmed and kept livestock. Measles, for example, was likely acquired from ruminants where it caused the disease rinderpest, now extinct thanks to effective control measures (91).

New disease encounters are frequently devastating to previously unexposed populations, such as happened when Native Americans were exposed to smallpox in the 16th century and Europeans to yellow fever virus when engaging themselves in the African slave trade (62,89). Fortunately, many novel infections are ultimately controlled since infection and survival is followed by immunity, which for smallpox and yellow fever is fully protective and lifelong (15). For other pandemics, immunity is less sustained, as occurs with influenza, the cause of a worldwide pandemic that killed millions in 1918-19 (90).

Today's pandemic caused by the coronavirus severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (coronavirus disease 2019 [COVID-19]) is causing major problems in almost all countries of the world. Fortunately, recovery from COVID-19 usually leaves persons immune, although as yet we do not know for how long and certainly some do suffer clinical reinfection (124). Moreover, in an astonishingly short time, effective vaccines were produced, some of which appear to be even more effective than immunity generated from natural infection (96). Accordingly, the COVID-19 pandemic could be a temporary inconvenience with normal lifestyle just around the corner.

Alas, this may not happen anytime soon, despite overwhelming evidence that vaccines are extremely effective and cause minimal side effects there is a high level of reluctance to use them even in countries where vaccine are readily available. Thus, although vaccinating everyone to produce herd immunity is the most logical and effective solution to the COVID-19 pandemic, approaches to complement the use of vaccines need to be developed to tide us over and blunt the impact of the infection, until the virus runs out of susceptible individuals.

This raises the issue of which alternative approaches are available to combat COVID-19. Some viruses, such as human immunodeficiency virus and hepatitis C virus can be successfully combated with antiviral drugs, which is fortunate since effective vaccines are still unavailable for those two viruses (116). For COVID-19, the FDA has fully approved Remdesivir to combat the infection and other drugs such as molnupiravir are close to being approved. However, to be effective antiviral drugs need to be administered early after infection and are not expected to be useful during the most damaging immunopathological phase of infection. Moreover, antiviral drugs are not fully effective and always have some level of toxicity, which is true for all those developed so far to combat COVID-19 (59,74,125).

Many viruses can be combatted by using interferons, but when these molecules are used in therapy, they have many side effects and may not work well against COVID-19 because of its interferon evasion response (119). It is evident that additional therapies are required to combat the various stages of COVID-19 and in this short review we plead the case for using strategies that change the function of one or more aspects of host metabolism, an approach currently mainly used to control some tumors and autoimmune diseases. The rationale for exploring this approach as a control measure against virus infections is because viruses cause metabolic changes in the cells they infect, multiple host immune components that react to the infection rely on several metabolic pathways to achieve their optimal function, and that disease resulting from COVID-19 can be very severe in persons with metabolic diseases.

Infection and Consequences of COVID-19

Usually infection with COVID-19 is the result of exposure to aerosols produced by infected persons that usually have a symptomatic respiratory illness and often cough. However, favoring the spread of COVID-19, asymptomatic persons can also be infectious either before their developing symptoms or even from those that remain asymptomatic (93). In rare instances, infection may derive from virus on fomites such as doorknobs and other surfaces and perhaps by handshakes and other interactions. The infection source that is most dangerous are particles between the size of 2.5–10 μm since these can remain suspended in the air for some time and once inhaled can readily avoid removal by the innate protection components in the upper respiratory system (25). Larger virus bearing particles are usually trapped in the upper respiratory tract and may or may not cause local lesions.

At the upper respiratory site, the infection is quickly contained by the innate defense system and the subsequent adaptive immune response may be modest or even unapparent (47,64). Once virus is present in the lower respiratory tract, alveolar epithelial cells are infected and the stage is set for severe lesions that can be fatal. Many of the clinical signs in the lung infection are thought to be more the consequence of host immune responses to the infection rather than the direct damaging effects of viral replication (70,128). The term cytokine storm is often used to describe this so-called stage 3. Patients during stage 3 are usually minimally infectious and many go on to recover and generate strong adaptive immune responses (7,129). However, others do not and succumb especially if unable to receive intensive medical care.

In several persons, the virus may spread to infect other organs beyond the respiratory tract and recovery is a long, drawn-out process. Additional organs involved include the kidneys, gastrointestinal tract, and the brain all of which contain cells that express the viral receptor, ACE2, although damage to such organs may not be the direct effect of viral infection (127). A minority of individuals go onto to develop a syndrome now referred to as Long COVID, a still mysterious syndrome that is further discussed in a later section (105).

As with all infectious diseases, the outcome of infection in different persons is highly variable and is affected by multiple factors as we and others have discussed in other reviews (5,12). With COVID-19, persons who are old or suffer from metabolic problems such as diabetes and obesity, or are immunologically suppressed, are more susceptible. However, a major factor that impacts on the outcome of infection is the dose of virus exposure as was discussed in another publication (99).

Minimal doses can easily be resisted and wearing a mask markedly diminishes the infection dose. Very large doses can likely overcome resistance, even in those persons with potent immunity because of past infection or vaccination (40). Other factors that impact on the outcome include host genetics, infecting viral strain, ongoing and past exposure to other infections, nutritional status, the content of the resident microbiome, and likely the status of host metabolism, the focus of this review.

Viruses Cause Metabolic Changes in Cells They Infect

Viruses themselves are not metabolically active and derive all of their metabolic products from the cells they infect. This usually means that the infected cell is induced to undergo some metabolic changes to sustain and replicate the virus (98). There is a wide spectrum of different virus–cell interactions and each is likely to have different metabolic consequences (109). Some viruses are rapidly cytopathic or cytolytic and the infected cell usually dies after producing viral progeny. In such interactions, all metabolic pathways are eventually shut down, but before this happens, the virus may stimulate particular pathways to provide the virus with structural and regulatory molecules required for new virus production (46,106).

For instance, with cytopathic cytomegalovirus, infected cells initially elevate the glycolysis and fatty acid synthesis pathways and also increase the production of intermediate molecules involved in the tricarboxylic acid (TCA) cycle (115). The cell then makes new enveloped virus before it dies. Other virus–cell interactions are noncytopathic and the infected cells may continue for a long time to make new virions that bud from the cell membrane in the case of enveloped viruses. With some tumor viruses such as Kaposi's Sarcoma-associated herpesvirus, the virus reprograms the cell to undergo dramatic metabolic changes that permit it to bypass normal growth limitations and become a tumor (30,133). COVID-19 is usually considered to be cytolytic with the infected cell dying after virus is replicated (132).

There is a growing library of information about the change in activity in some metabolic pathways that occurs before the total shutdown happens after COVID-19. For example, an infected cell may switch its energy supply from oxidative phosphorylation to glycolysis (3,26). There is also evidence that lipid metabolism is upregulated by several mechanisms in productively infected cells (80).

The severe clinical stage of COVID-19 involves pronounced inflammation that is characterized by excessive macrophage infiltration and their production of multiple inflammatory mediators and reduced production of regulatory cytokines such as interferon (1,49). How this effect is mediated at a molecular mechanistic level still needs to be fully described. However, the outcome is unlikely to be the direct effect of viral infection of mononuclear immune cells, although some reports have described the presence of viral proteins in monocytes, but did not show they were synthesized in the cell (33,63).

In addition, SARS-CoV-2 infection of human primary monocytes in vitro results in the upregulation of pathways involved in lipid uptake pathways and of the major transcriptional factors peroxisome proliferator-activated receptor gamma and sterol regulatory element-binding protein 1 involved in lipogenesis and diacylglycerol O-acyltransferase 1 in triacylglycerol synthesis (33).

The unresolved issue is to determine how the changes in activation state and metabolism are caused by the infection and, perhaps of more relevance, how understanding the effects at a mechanistic level can translate into practical therapies to reverse the tissue damaging process. Our own bias is that the pathological inflammation that occurs in the respiratory tract and some other organs in severe COVID-19 is likely to be mainly an indirect effect, a notion supported by the fact that the lesions are often well controlled by anti-inflammatory therapy such as dexamethazone or using mAb therapy that targets certain cytokines or their receptors (24,81). We anticipate that correcting the activity of some metabolic pathways that are dysregulated in reactive cells might also be useful to control infection and facilitate recovery.

COVID-19 Can Cause Metabolic Changes in Infected Persons

As with many virus infections, patients infected with COVID-19 may undergo changes in several aspects of their metabolism and in some cases, these could dictate the outcome of infection. Moreover, as discussed in a subsequent section, COVID-19 in those with existing metabolic disorders such as diabetes can suffer more severe consequences (36,45,95). One aspect of metabolism that is often changed as a consequence of COVID-19 is glucose metabolism. Thus, many so-called metabolically healthy patients develop hyperglycemia (>180 mg/dL or >10 mmol/L) even early after infection and those that are admitted to hospitals often have a poor outcome (19,121).

How the infection causes hyperglycemia is not yet fully understood but, in some cases, it may result from virus infection of the pancreatic beta cells resulting in lower insulin production (38,104). Another idea is that the inflammatory response induced leads to insulin resistance by an as-yet-unexplained mechanism (31). This so-called stress related hyperglycemia can occur with several acute and chronic infections such as tuberculosis (67) and may be the consequence of producing excess cytokines that act in the liver (69). It has also been suggested that cells in some organs respond to high blood glucose levels by upregulating ACE2 receptors, making them susceptible to COVID-19 (104). Currently, we are unaware of what effects attempts to control hyperglycemia with drugs such as metformin may have on the outcome of COVID-19.

COVID-19 patients may also develop changes in lipid and amino acid metabolism (21). Moreover, clinical studies have indicated that patients with dyslipidemia suffer more severe disease and when hospitalized have a poor prognosis. For example, a retrospective study in Wuhan that examined the lipid profile of persons with different disease outcomes showed that persons that succumbed often had a severe decrease in lipid levels as the disease progressed (122). It also seemed that the dyslipidemia and poor prognosis was associated with an elevated inflammatory response (2,16,42), but how this caused the dyslipidemia needs to be further investigated as do the value of therapies to modulate the metabolic changes.

Longitudinal studies using metabolomics have been done to compare any changes that occur in COVID-19 patients compared to controls. Such studies can uncover metabolic changes that may relate to disease severity and potentially guide responses to therapy (101). One such study revealed major effects on tryptophan metabolism and the associated kynurenine pathway (111). Critical patients often have a marked increase of kynurenic acid in their sera, indicative of elevated tryptophan metabolism (61). It is also well known that the kynurenine pathway is necessary for the normal functioning of several immune cells so its suppression could result in poor control of COVID-19 (79).

Activation of the kynurenine pathway is also correlated with an increase in the proinflammatory cytokine interleukin (IL)-6 that contributes to lesion severity during the lung damaging stage 3 lesions (111). One valuable therapy is the inhibit IL-6 and/or its receptor binding using mAb therapy (24,54).

In addition to tryptophan metabolism being changed during COVID-19 reports that several other amino acid levels are also changed. These include aspartate, arginine, tyrosine, glutamine, and lysine, but further studies are needed to show how these changes influence the outcome of infection (111). In the case of glutamine, which is metabolized to become a neurotransmitter, conceivably, changes in its metabolism could be linked in some way to the onset of “brain fog,” a common sign in patients suffering the so-called Long COVID syndrome (87) that is discussed in a later section.

COVID-19 in Persons with Metabolic Diseases

It soon became evident after reports of COVID-19 first appeared that disease was often far more severe with higher rates of mortality in those who had pre-existent metabolic diseases or were immunocompromised (8). Most reports dealt with type 1 diabetes or obesity where many patients often develop type 2 diabetes (48). For example, in a United Kingdom study people with either form of diabetes where 50% more likely to develop severe complications with many sent to the ICU and diabetics accounting for a third of all those dying from the viral infection (58,78). Indeed, patients with severe signs during COVID-19 often were discovered later to be suffering from undiagnosed diabetes.

Curiously, the pancreatic beta cells in some infected patients had an above average number of ACE2 receptors, making them potentially more susceptible to the disease via direct viral infection, but this mechanism is yet to be proven to occur (37). Additionally, some of the treatments used to control severe COVID-19 lesions, such as dexamethasone can result in signs of hyperglycemia and metabolic acidosis, as occurs in untreated diabetes (117). In some instances, dexamethasone treatment can cause apoptosis of pancreatic beta cells so explaining the onset of diabetes (44). In addition, naturally produced cortisol in response to infections and immune responses may act on the liver by way of the hypothalamic-pituitary-adrenal axis to cause hyperglycemia (69).

Patients with diabetes, and to a lesser extent, obesity, may be more susceptible to other virus infections such as influenza (102). This is thought to occur because aspects of both innate and adaptive immunity function less effectively to control infections. For instance, diabetics may have reduced neutrophil functions that include neutrophil extracellular trap function, decreased production of elastase, beta defensins and myeloperoxidase, reduced superoxide responses, and diminished phagocytic activity (53,57,92). NK cell and macrophage functions may also be suppressed and diabetics make diminished T cell-derived proinflammatory cytokine responses (55).

Diabetics may also be less able to make anti-inflammatory responses such as Treg responses and the production IL-10 and transforming growth factor-beta that would normally function to diminish the extent of the cytokine storm that occurs during severe COVID (50). Many of the problems caused by diabetes can be controlled by using insulin to reduce hyperglycemia, but using insulin to control hyperglycemia in nondiabetics can be problematic (126). However, it appears likely that controlling glucose metabolism during COVID-19 could be managed with other therapies that include the use of 2-Deoxy-D-glucose (2DG) and metformin. At least one study has evaluated the use of 2DG to control COVID-19 (14,65).

Obesity is the other common metabolic problem that predisposes COVID-19 infected patients to develop more severe and often lethal consequences. In fact, a study in France showed that out of 85 persons infected with COVID-19, 87.7% of those patients who required invasive mechanical ventilation were obese (100). In other studies, 25% of cases with severe COVID-19 were obese (22). In China obese people had on average 3.4-fold more severe consequences of COVID-19 (17) and were significantly more likely to develop the acute respiratory distress syndrome (34). In summation, a meta-analysis on 75 studies involving 399, 461 COVID-19 patients revealed that obese patients had a 113% higher chance for hospitalization, 74% higher rate of ICU admission, and 48% higher rate of mortality compared to those without obesity (94).

Explanations for the higher susceptibility of obese persons to COVID-19 are multiple and include a change in balance of both innate and adaptive aspects of immunity (66,86). One key factor that is changed during obesity is the hormone leptin. Leptin acts on adipose and innate immune cells to induce them to secrete proinflammatory cytokines via manipulation of the JAK-STAT and NF-kB dependent pathways (73). Enhanced production of leptin in adipose tissue induces activation of STAT-3 and thereby enhances the production of IL-18, IL-6, TNF-a, and IL-6 from proinflammatory M1 macrophages and a change toward the hyper-inflammatory state. This leads to a cytokine storm that decreases the chance of recovery from COVID-19 (32).

Murine studies revealed that leptin-deficient mast cells form more anti-inflammatory M2 macrophages leading to cessation of inflammation and the beginning of tissue recovery (131). One common outcome of obesity is insulin-resistant diabetes and the gamut of immune consequences that make diabetics more susceptible to COVID-19. It will be important to determine whether there are any metabolic changes that could be corrected in obese persons infected with COVID-19 that would lessen the impact of their infection.

Long COVID

There is an increasing awareness that even after apparent recovery from acute clinical disease at least 10% of patients experience a variable range of symptoms that can be long lasting and are referred to as Long COVID (82,105). The syndrome is more common in middle-aged persons, but can affect all age groups. There is a highly variable range of clinical signs and these persist for 12 weeks or more. The symptoms are multisystemic and fluctuating. They include fatigue, dry cough, shortness of breath, gastrointestinal distress, sleep disorders, headaches and muscle aches, anxiety, and depression (29). Additionally, some patients experience one or more neurological disturbances and the term “brain fog” is often used. As with some previous undefined chronic conditions, such as the chronic fatigue syndrome, initially Long COVID was not taken seriously, but that is no longer the case (18,110).

However, the cause and pathogenesis of Long COVID remains poorly understood and is likely to vary between individuals. Two general ideas are that it is the direct consequence of persisting virus that disseminates to organs beyond the respiratory tract and damaging their function (130). The second proposal is that the syndrome represents a disbalance of the immune system set off by the infection.

In support of this proposal many patients have changes in cytokine levels and T cell subset and myeloid cell representation. Some individuals develop autoimmune-like symptoms (39,118). The most difficult concept to explain in Long COVID is the mechanism by which the many neurological signs that are reported occur (85). Long COVID remains a mysterious syndrome and given the bias of our own research interests, we propose that the disfunction of one or more metabolic pathways in affected organs may partially explain Long COVID pathogenesis. Thus, there is some evidence that tryptophan and glutamine metabolism may be altered in some Long COVID patients (35,72). Conceivable, the impact of Long COVID could be minimized by therapies that correct any such metabolic changes and this concept merits further investigation.

Would Manipulating Metabolism Represent a Practical Approach to Control COVID-19 Lesions?

The idea of targeting one or more metabolic pathways to change the outcome of a virus infection has received very minimal attention and appears unexplored as an approach to control COVID-19. Some potentially useful approaches that would minimize lesions by reshaping metabolism might be effective only when used before infection. Such an example is diet adjustment that can limit the extent of viral damaging lesions acting in many cases by changing the composition of the gut microbiome. For example, diets that are high in fiber and enriched in certain types of fatty acids result in changes in the gut microbiome composition (13,88).

In such persons the result is the induction of responses to virus infections that damage tissue to a lesser degree and favor lesion resolution (6,113). Thus, in experiments that compared the outcome of influenza infection in mice fed different levels of fat in their diet, those receiving a high-fat diet experienced twofold higher mortality along with enhanced respiratory lesions and the higher production of inflammatory cytokines (76). The type of fat in the diet can also influence viral pathogenesis. For example, a diet rich in omega-3 fatty acids favors the production of anti-inflammatory T cell-derived mediators (56), which can limit tissue damaging reactions and might be useful to counteract the lung damaging stage 3 reactions during COVID-19.

Another approach that can act to limit viral-induced inflammatory reactions that we have shown is effective in a herpesvirus model is to supplement the diet with the short chain fatty acids propionate or butyrate (108). Similarly, we and others have shown that suppressing the utilization the amino l-glutamine can reduce the extent of viral-induced inflammatory lesions (68,107), as can supplements with some vitamins and minerals such as zinc (41,51,123). Additionally, blocking fatty acid metabolism via use of the Acetyl-CoA carboxylase isoform ACC1 and carnitine-palmitoyl transferase I inhibitor (43), results in reduced lesions in West Nile virus-infected mice (52).

It might be useful to use one or more of the previously mentioned approaches that are expected to diminish the severity of lesions associated with a virus infection such as SARS-CoV-2 if used before infection, what is really needed are metabolic manipulations that improve the outcome of COVID-19 when begun during ongoing infection. This topic has received minimal investigation in any virus infection, but we contend it is a highly relevant control measure particularly if combined with diagnostic approaches such as metabolomics to detect and quantify any metabolic dysfunctions present in COVID-19 patients. Some approaches that target metabolism to improve the outcome of infection are listed in Table 1.

Table 1.

Some Potential Approaches to Manipulate Metabolism to Suppress Coronavirus Disease 2019

Pre-exposure	Acute phase	Severe inflammatory phase	Long COVID phase
Dietary alterations (Fiber, Fat type and Level) (13,56,88)	Antivirals (Remdesivir, Molnupiravir, Lopinavir/Ritonavir) (59,74,125)	Metabolic treatment (2DG, Metformin, Etomoxir, 25-hydroxycholesterol) (11,14,65,68,77,134)	Dietary tryptophan supplementation (35,61,111)
Dietary vitamin supplements (Vitamin A, B complex, C & D) (9,71,97,112)	Metabolic treatment (2DG, Metformin, Etomoxir, 25-hydroxycholesterol) (11,14,65,68,77,134)	Corticosteroid therapy (81)	Dietary glutamine supplementation (23,72)
Dietary mineral supplements (Zinc) (41,123)	Monoclonal antibody (Bamlanivimab plus Etesevimab, Casirivimab plus Imdevimab, Sotrovimab) (27)	Anticytokine monoclonal antibody therapy (Anti-IL-6, Anti-IL-1β receptor) (54,60)
Microbial regulation (10,20)	Interferons (α&β) (4)	Itaconate/Dimethyl fumarate/4-Octyl-Itaconate (75,84,83)
	Itaconate/Dimethyl fumarate/4-Octyl-Itaconate (75,84,83)

Most of these have not been tested against COVID-19.

2DG, 2-Deoxy-D-glucose; COVID-19, coronavirus disease 2019; IL, interleukin.

In a herpes virus system in which lesions are attributed to an immune inflammatory reaction to the virus, as occurs also in stage 3 of COVID-19, we have shown that administering 2DG during the early stages of lesion development can markedly suppress their severity (11,114). The beneficial outcome correlated with diminishing the response of proinflammatory T cells yet preserving the responsiveness of anti-inflammatory regulatory T cells (11). A similar outcome was achieved if animals with ongoing lesions received a drug that suppressed glutamine metabolism, an amino acid that is required to support the proliferation and cytokine-producing capacity of inflammatory T cells (107). In addition, in ongoing studies with a model of herpes simplex encephalitis, also a lesion that involves inflammatory T cells, treatment with the glucose metabolism inhibiting drug metformin resulted in notable lesion control (unpublished studies).

Other approaches worthy of investigation derive from their success in managing ongoing autoimmune lesions. For example, it would be relevant to investigate the use of drugs that affect lipid metabolism. Thus, in a model of autoimmune encephalitis mediated by proinflammatory T cells that reacted with central nervous system autoantigens use of etomoxir, a drug that stimulates fatty acid oxidation suppressed the development of lesions likely by way of its inhibitory effects on proinflammatory T cells (77), a mechanism that occurs in stage 3 COVID-19.

Another approach found useful to inhibit exuberant inflammatory reactions is to elevate levels of the metabolite itaconate, a molecule with immunomodulatory activity that also has antiviral activity against several viruses that include SARS-CoV-2 (28,83). It is known that itaconate levels may correlate with the severity of COVID-19 (103). Itaconate is derived from citrate and acts to inhibit some components in the mitochondrial TCA cycle (84). When itaconate, or better still some stabilized cell-permeable derivatives, are administered systemically inflammatory reactions are diminished largely because of changes in the cellular composition of the reactions (75). This metabolic reprogramming caused by itaconate may also be achieved using ligands such as LPS that cause itaconate upregulation in animal hosts (84).

There are several additional approaches that may cause metabolic pathway changes that have yet to be explored as an approach to diminish the consequences of COVID-19. These include the administration of some vitamins such as vitamin D whose level appears diminished in some patients with COVID-19 (9,71). Others have suggested that vitamin A might be useful because of its known anti-inflammatory and antioxidant effects (112). Other miscellaneous approaches advocated as useful include drugs that change cholesterol metabolism such as the derivative 25-hydroxycholesterol that has antiviral effects against coronaviruses such as SARS-CoV-2 and MERS in vitro or in animal model systems (120,134). One expects that this area of investigation will receive far more attention as a means of controlling COVID-19 particularly if vaccine hesitancy and limited availability in some societies remain an issue.

Author Disclosure Statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding Information

This work was supported by Foundation for the National Institutes of Health Grants; R21AI142862 and R01EY005093.