Nitric oxide for the prevention and treatment of viral, bacterial, protozoal and fungal infections [version 2; peer review: 1 approved with reservations]

Although the antimicrobial potential of nitric oxide (NO) is widely published, it is little used clinically. NO is a key signalling molecule modulating vascular, neuronal, inflammatory and immune responses. Endogenous antimicrobial activity is largely mediated by high local NO concentrations produced by cellular inducible nitric oxide synthase, and by derivative reactive nitrogen oxide species including peroxynitrite and S-nitrosothiols. NO may be taken as dietary substrate (inorganic nitrate, L-arginine), and therapeutically as gaseous NO, and transdermal, sublingual, oral, intranasal and intravenous nitrite or nitrate. Numerous preclinical studies have demonstrated that NO has generic static and cidal activities against viruses (including β-coronaviruses such as SARS-CoV-2), bacteria, protozoa and fungi/yeasts in vitro. Therapeutic effects have been seen in animal models in vivo, and phase II trials have demonstrated that NO donors can reduce microbial infection. Nevertheless, excess NO, as occurs in septic shock, is associated with increased morbidity and mortality. In view of the dose-dependent positive and negative effects of NO, safety and efficacy trials of NO and its donors are needed for assessing their role in the prevention and treatment of infections. Trials should test dietary inorganic nitrate for preor postexposure prophylaxis and gaseous NO or oral, topical or intravenous nitrite and nitrate for treatment of mild-to-severe infections, including due to SARS-CoV-2 (COVID-19). This review summarises the evidence base from in vitro, in vivo and early phase clinical studies of NO activity Open Peer Review Reviewer Status Invited Reviewers 1 version 2 (revision) 22 Oct 2021 version 1 05 Jul 2021 report Khosrow Kashfi , City University of New York School of Medicine, New York, USA 1. Any reports and responses or comments on the article can be found at the end of the article. Page 1 of 38 F1000Research 2021, 10:536 Last updated: 22 OCT 2021


Introduction
Nitric oxide (NO), an inorganic molecule, is generated endogenously by prokaryotes and eukaryotes from L-arginine and L-citrulline by a family of NO synthase enzymes (NOS;  [1][2][3] In higher animals, it is also generated by reduction of dietary and endogenous nitrate (NO 3 Chemistry and biology of nitric oxide NO is a small diatomic hydrophobic colourless gas that diffuses easily and has a short half-life. With an unpaired electron (NO • ), it is a free radical and so is chemically reactive and unstable in the presence of oxygen and superoxide producing reactive nitrogen oxide species (RNOS). In reality, the chemistry of NO is more complex and it exists in several redox forms: nitrosonium cation (NO + ), NO (NO • ) and nitroxyl anion NO À ). 23 As an inorganic molecule, its central role in biology as a signalling molecule was only discovered in the 1980s, 24 the identification of which led to the Nobel Prize for Physiology and Medicine in 1998. 25 Ironically, the medicinal use of NO in the form of glyceryl trinitrate (GTN) for angina prophylaxis antedates the modern understanding of the biological synthesis and role of NO by more than a century. 26 Experimental studies demonstrating nitric oxide inhibition of … Viruses Numerous preclinical in vitro studies have demonstrated that NO sources (stimulated endogenous NO, inorganic and organic nitrates, L-arginine) and PDE5-inhibitors can reduce replication in all seven types of virus as defined in the Baltimore classification (Table 2); 27 this includes Class IV viruses (positive-sense RNA viruses) incorporating several coronaviruses 28-33 including SARS-CoV-2. [34][35][36] Most studies showed efficacy although neutral studies were reported for porcine reproductive and respiratory virus (an arterivirus, which is closely related to coronaviruses) and rhinovirus. 37,38 Bacteria Multiple studies have assessed the effect of NO on bacteria and inhibitory effects have been seen across a wide range of gram negative, gram positive and acid-fast bacteria (Table 3). NO sources included L-arginine, NO, nitrite, organic nitrates, and endogenously-generated NO. Multiple mechanisms for efficacy have been reported, as discussed below.    Protozoa NO sources have been tested on both intracellular and extracellular protozoa (Table 4) with sources involving activated macrophages, sodium nitrite, glyceryl trinitrate, sodium nitroprusside (SNP) and S-nitroso-L-acetylpenicillamine (SNAP).

Fungi and yeasts
The effects of NO on several fungi and yeasts have been studied (Table 5). NO was donated exogenously through stimulating macrophages or as acidified nitrite. In vitro experiments demonstrated reduced replication whilst in vivo experiments in mice showed reduced infection.

Derivatives of nitric oxide
Whilst endogenous NO derived from eNOS and nNOS is physiologically active via its second messenger (cGMP), the antimicrobial effects of NO relate to its toxic effects when present at higher concentrations. Although it is technically challenging to measure free NO concentrations, studies suggest that NO concentrations derived from iNOS are 10-100Â higher than those resulting from eNOS/nNOS (Table 6). NO concentrations resulting from exogenous administration lie between those from eNOS/nNOS and iNOS but approximate more closely to those from iNOS than eNOS. Importantly, much antimicrobial NO activity is likely to reflect the effects of derivative molecules rather than NO itself: • Nitric oxide (NO • ). In general, bacteria deficient in low molecular weight thiols such as glutathione (e.g. Staphylococci spp.) are sensitive to attack by NO whereas those with high thiol concentrations are resistant to NO.
Since these molecules differ in their stability, reactivity, location and cellular diffusivity, the overall effect of NO will depend on the molecular species involved and its location.

Cellular and other targets of nitric oxide
The targets for NO and associated reactive nitrogen species are multitudinous: • DNA, through deamination of adenine, cytosine and guanine; 41 cross-linking; breakage of strands; inhibition of DNA repair enzymes such as DNA alkyl transferases (and so preventing transfer of the guanine alkyl group to protein); and disruption of DNA replication by inhibition of ribonucleotide reductase; 42 as in S. enterica and vaccinia virus.
• RNA, through disruption of RNA replication by inhibition of viral ribonucleotide reductase.
• Inhibition of mitochondrial function, specifically through inactivation of iron-sulphur complexes within respiratory chain enzymes. 43 • Protein modification at cysteine, methionine, phenylalanine, tryptophan and tyrosine residues, e.g. by RNOS. Such protein effects will reduce enzyme activity, as seen for DNA, proteases 44 and mitochondrial function, as in Coxsackievirus. 44 • Limit late protein synthesis, e.g. through posttranslational modification of viral proteases. (Early protein translation/synthesis is not typically affected.) • Induction of lipid peroxidation.
• Limit glutaminolysis by shuttling glutamine to glutathione synthesis, as in cytomegalovirus. 45 • Interaction with sulfhydryl-containing constituents of the bacterial cell. 46 • Disrupt zinc homeostasis, as in S. enterica. 47 • Limit virion assembly/particle formation.
• Reduce bacterial adhesion to NO-releasing surfaces. 48 Nitric oxide may also play an augmenting role as an antimicrobial agent. Examples include the adjuvant roles of NO when given with type I interferons in the treatment of DNA viruses 49 and L-arginine when given with conventional chemotherapy in smear-positive TB. 50 In addition, NO's vasculo-active effects are likely to be beneficial in preventing infection and its severity, with NO: • Reversing endothelial dysfunction and so potentially reducing endotheliitis, 11 as occurs in COVID-19. 9,10 • Reducing leucocyte function (e.g. adhesion, chemotaxis, phagocytosis); 4 COVID-19 is associated with increased phagocyte counts. 10,51 • Reducing platelet activation and platelet-leucocyte conjugation and so reducing micro-and macro-thrombosis, as seen in COVID-19. 10,52 • Improving organ blood flow and perfusion through smooth muscle relaxation and vasodilation and so likely reversing infection-related vasoconstriction as seen in COVID-19, 10 including in the pulmonary circulation. 53 These actions of NO are all mediated via the second messenger cGMP.
Antimicrobial production of nitric oxide NO is produced by some bacteria, archaea and yeasts via several pathways including denitrification of nitrate to nitrite and then to NO 54 and oxidation of L-arginine to NO and L-citrulline as catalysed by a bacterial nitric oxide synthase (bNOS), a process that can be inhibited by NOS inhibitors. 55 Whilst eukaryotic NOS contains both catalytic and reductase domains, prokaryotic bNOS lacks the latter relying instead on other cellular reductases to generate NO; the one exception to this is the bNOS present in Sorangium cellulosum which does include a reductase module.
In contrast to the signalling role of NO in mammals, NO synthesis in bacteria has multiple functions which vary between antimicrobial species: [56][57][58][59] • Protection against oxidative stress with NO limiting thiol reduction and so the formation of hydroxyl radicals (Bacillus anthracis/subtilis, Staphylococcus aureus 56 ) ( Table 1.10).
• Protection against oxidative stress with NO activating catalase (B. subtilis). Such defence will limit damage from phagocytic respiratory bursts. 56 • Protection against oxidative stress by reducing endogenous NO synthesis and increasing the expression of NO dioxygenase to detoxify NO (Candida albicans). 58 • Biosynthesis of toxins, e.g. thaxtomins (a phytotoxin) interfere with potato plant wall synthesis (Streptomyces turgidiscabies). 56 • Activation of aerobic and nitrate respiration to optimise growth (S. aureus). 59,60 • Protection against antimicrobial agents including amoxycillin, cefuroxime, gentamicin and novobiocin (B. anthracis/cereus/thuringiensis/weihenstephanensis, S. aureus), 57 and azoles (Candida albicans). 58 The production by some microbes of endogenous NO to protect against oxidative stress is ironic since hosts are using NO to try to destroy the microbe! Resistance to nitric oxide Microbial resistance to antibiotics is an increasingly common problem and has left some bacteria with few treatment options, e.g. drug-resistant Neisseria gonorrhoeae. Hence, it is vital to consider whether resistance to NO is innate in some microbes or can be acquired. As already highlighted, some microbes have an intrinsic ability to produce their own NO and so attenuate the effects of oxidative stress (e.g. B. anthracis/subtilis, C. albicans, S. aureus 56,58 ), activate aerobic respiration (S. aureus 59 ) or protect against antimicrobial agents (B. anthracis/cereus/thuringiensis/weihenstephanensis, C. albicans, S. aureus 57,58 ).
Microbes may also have mechanisms for deactivating NO. One mechanism is via a NO reductase which reduces NO to nitrous oxide and then nitrogen, e.g. as occurs in fungi 61 (Table 1.11). Bacteria have different NO reductases but similarly produce nitrous oxide, 61 as seen in Pseudomonas aeruginosa. 54 Loss-of-function mutations in NO reductase may be lethal, possibly because intracellular NO concentrations rise to toxic levels.
A second mechanism for detoxifying NO is via NO dioxygenase oxidation to nitrate (Table 1.12). The pre-eminent NO dioxygenase is flavohaemoglobin, 62 as present in bacteria (e.g. Salmonella enterica, S. aureus, Vibrio cholerae, Yersinia pestis, 22,63,64 ) and yeasts. A related haemoglobin, truncated haemoglobin, detoxifies NO in mycobacteria. Of note, Mycobacterium leprae has undergone reductive genome evolution losing more than 2,000 genes, including some that protect against RNOS; as a result, M. leprae has fewer defences against NO than Mycobacterium tuberculosis. 65 Bacterial lactate dehydrogenase also detoxifies NO, as seen in S. aureus. 22 Importantly, these detoxifying enzymes only cope with low levels of NO and are not protective against high NO levels.
As a result, microbes show differing sensitivities to NO, as seen for common airways pathogens where sensitivity was ranked (sensitivity most to least): P. aeruginosa~C. albicans > S. aureus > Klebsiella pneumoniae~Staphylococcus epidermis. 66 However, there is little evidence that bacteria can acquire de novo resistance to NO, as confirmed in experiments on strains of E. coli, P. aeruginosa, S. aureus and Staphylococcus epidermidis. 67 This property is unsurprising since NO has multiple mechanisms for antimicrobial activity and these are likely to be invoked orders of magnitude faster than any microbe can process metabolically, especially if protein synthesis is required. Equally, the main mechanisms for antibiotic resistance (drug inactivation, altered binding sites or metabolism and reduced drug permeability) are unlikely to be relevant to many NO sources. Whether viruses, protozoa and fungi can develop resistance to NO remains unclear.
Administering nitric oxide, donors and related compounds L-arginine and L-citrulline In the presence of NOS, administration of L-arginine may enhance NO synthesis (Table 7) although intracellular L-arginine levels are not normally rate limiting and so administration may not have physiological effects. 68 Although oral preparations of L-arginine are commercially available, consumption of high doses is associated with profuse diarrhoea (P Bath, personal observation). However, L-citrulline may be a more efficient method for delivering L-arginine since it has more efficient intestinal absorption, lower first pass metabolism and higher renal reabsorption, does not induce arginase and is safe and tolerable at high doses. 4 Inhaled nitric oxide Gaseous NO may be inhaled with the aim of improving pulmonary haemodynamics and killing microbes. Multiple trials are underway for COVID-19 prevention and treatment (Table 9). NO may also be created in real time by combining sodium nitrite and citric acid and administering this either as a nasal spray (for local therapy) or via nebuliser (for combined nasal and bronchial therapy).

Organic nitrates
Organic nitrates such as GTN, isosorbide dinitrate (ISDN) and isosorbide mononitrate (ISMN) are widely used in vascular medicine for the prevention and treatment of angina, treatment of chest pain in unstable angina and myocardial infarction, treatment of severe heart failure, and blood pressure lowering after cardiac surgery and in acute stroke (Table 7). There is increasing concern that chronic use of organic nitrates may cause major adverse cardiac events and death, 69 reduce daily activity, 70 and not improve quality of life or exercise capacity. 70 Potential explanations include the development of tolerance, and induction of endothelial dysfunction and cell damage through oxidative stress, e.g. production of free radicals/peroxynitrite. 71,72 Importantly, organic nitrates only release NO in cells and tissues expressing mitochondrial aldehyde dehydrogenase-2. 73 For example, SNP and SIN-1 inhibit monocyte chemotaxis whilst organic nitrates (ISDN, GTN and molsidomine) do not; 74 this contrasts with smooth muscle cells which vasodilate with all five agents. Since aldehyde dehydrogenase-2 suffers from use-inactivation, nitrate tolerance (tachyphylaxis) and endothelial dysfunction develops 72 and bioconversion only restarts following a nitrate-free period. Several in vitro studies have demonstrated the potential antimicrobial effects of organic nitrates (Table 7). Other non-organic nitrates include pentaerythritol tetranitrate and erythrityl tetranitrate.
Therapeutic inorganic nitrite and nitrate Therapeutic use of inorganic nitrite is limited with intravenous administration used in cyanide poisoning (British National Formulary). Topical acidified sodium nitrite has been shown to reduce cutaneous infections secondary to a variety of viruses and bacteria although its general use is probably limited by skin irritation and erythema (Table 7). A recent study found that acidified nitrite improved wound healing in rats with diabetes. 75 Dietary inorganic nitrate NO may also be produced from dietary inorganic nitrate, as is present in high concentrations in green leafy and some root vegetables, e.g. spinach, lettuce, rocket, beetroot, celery, fennel, radish and Chinese cabbage. 76 Nitrate is absorbed from the proximal gastrointestinal tract, excreted by salivary glands, reduced to nitrite by oral bacteria and then absorbed in the gastrointestinal tract. A number of bacterial species situated on the dorsal surface of the tongue perform this conversion via nitrate reductases. 77 (In the absence of oxygen, nitrate and nitrite are commonly used by bacteria as terminal electron acceptors for respiration. 78 ) Through this symbiotic relationship, the mammalian host provides the nutrients and the environment in return for nitrite production by bacteria. 79 Absorbed and circulating nitrite is then further reduced to NO, a process that is enhanced in hypoxic or acidic conditions and by multiple mechanisms including deoxyhaemoglobin, deoxymyoglobin, xanthine oxidoreductase and endothelial nitric oxide synthase. 80,81 As such, most effects of dietary nitrate will be vascular and perivascular. The beneficial vascular protective effects of vegetable consumption are very clear epidemiologically, as present in the classical Japanese   92 Beyond anti-inflammatory effects on blood cells, nitrite or nitrate reduce soluble pro-inflammatory factors including C-reactive protein, chemokine (C-X-C motif) ligand-1/2, endothelin-1, interleukins-1β/6/10/12p70, interferon-γ, monocyte chemoattractant protein and tissue necrosis factor-α. 99 Dietary nitrate has profound metabolic effects and appears to have the potential for reversing the metabolic syndrome and have anti-diabetic effects through improving insulin sensitivity and lowering blood glucose levels. [100][101][102] Overall, the pharmacological effects of beetroot juice have been demonstrated in younger and older people, and in people with cardiovascular disease, e.g. diabetes, obesity, hypertension, hypercholesterolaemia, heart failure and stroke. 103 Importantly, inorganic nitrate (given as beetroot juice) may be taken by pregnant women. 104 Experimentally, watermelon juice and chard gel may be used as an alternative source of dietary nitrate. 105,106 Phosphodiesterase-5 inhibitors PDE5-inhibitors, such as dipyridamole and sildenafil, enhance the physiological effects of NO as mediated by cGMP. Whether these agents should have antimicrobial effects is unclear since they do not enhance NO levels per se; nevertheless, both drugs have exhibited antimicrobial activity (Table 7) and are being tested in COVID-19 trials (Table 9). 107,108 Stimulation of endogenous nitric oxide-dependent nitric oxide production Endogenous NO production may also be stimulated externally. First, nasal breathing promotes the production of NO from the paranasal sinuses and this has bronchodilatory, vasodilatory and potential antimicrobial activities. 109 This natural defence mechanism may be attenuated with mouth breathing, as occurs with increasing age and obesity. Second, ultraviolet radiation (UVA and UVB) stimulates the release of NO from both keratinocytes and melanocytes; NO has multiple effects including attenuation of free radical damage, melanogenesis, blood pressure lowering 110  • NONOoates (diazeniumdiolates), e.g. with activity against C. albicans, E. coli. 110,115 • NO coating of medical device surfaces and tubing, 116,117 e.g. using S-nitroso-N-acetyl-D-penicillamine to kill Staphylococcus aureus and P. aeruginosa.
• RRx-001, a small molecule nitric oxide donor. 118 NO sources can also be categorised by whether administration is local (e.g. cutaneous nitrite or intranasal preparations), systemic (e.g. dietary nitrate, L-arginine or L-citrulline, oral isosorbide or sildenafil, sublingual GTN, intravenous GTN or SNP) or mixed local and systemic (transdermal GTN). Local administration allows high and potentially cidal concentrations of NO to be achieved without unwanted systemic effects. Intravenous formulations might allow for systemic infections to be treated.

Relevance of in vitro studies to preclinical and clinical studies
Most microbial studies presented above and in Tables 2-5  Nitric oxide for clinical infections Oral health and gastrointestinal infections As already highlighted, oral bacteria (e.g. Corynebacterium pseudodiphtheriticum, Fusobacterium nucleatum, Nocardia spp., Prevotella melaninogenica, S. aureus, S. epidermidis, Veillonella spp.) are vital for the reduction of salivary nitrate to nitrite as part of the entero-salivary circulation; nitrite is further reduced to NO. 119 This represents a symbiotic relationship between bacteria and the mammalian host; the host provides the nutrients and the environment in return for nitrite production, 79 as in the absence of oxygen, nitrate and nitrite are commonly used by such bacteria as terminal electron acceptors for respiration. 78 Oral consumption of nitrate and the resulting increase in nitrite in the oro-pharynx leads to salivary alkalinisation (pH~7.0 to 7.5) 120 and so reduction in detrimental bacteria and caries. 77 Similarly, nitrate supplementation was associated with increased oral Rothia spp. and Neisseria spp, and diminished oral Prevotella spp. and Veillonella spp.; in parallel, plasma nitrite levels rose and systemic blood pressure fell. 121 Salivary nitrite production is related to the abundance of oral-nitrate-reducing bacteria. 122 In contrast, bacteria and yeast, in particular Lactobacillus spp., Streptococcus spp. and C. albicans, are key to the development of dental caries through the production of acid. Equally, antibiotics that kill nitrate-reductase-containing bacteria inhibit oral nitrite production and so increase the risk of oral thrush. 123 Acidified nitrite has antibacterial activity against Helicobacter pylori in vitro, 124 an experiment that likely mimics the scenario seen by these bacteria in the stomach after a nitrate/nitrite-rich meal.

Cutaneous infections
The skin is a potent source of nitric oxide and production is increased with exposure to sunlight (specifically ultraviolet radiation) sufficient to lower blood pressure. 125,126 Hence, skin-derived NO may form a natural dermatological antimicrobial defence. Numerous studies have demonstrated that topical acidified sodium nitrite reduces cutaneous infections due to a variety of viruses and bacteria (Table 7) although prophylaxis had to be continued in some cases since NO suppressed replication without necessarily being viro-toxic. [127][128][129][130][131][132] Inorganic nitrate has been used for the treatment of human papilloma virus. 133 Phase II clinical trials have found that acidified nitrite in cream reduced Leishmania major/ tropica amastigotes and promastigotes with a reduction in cutaneous leishmaniasis 129 and increased cure rates in tinea pedis. 127 Novel NO agents are in development to treat skin conditions (Table 8).

Respiratory infections
In animal and human experiments, NO substrate (L-arginine) and a NO donor (SNP) has been shown to improve the mucociliary activity of the upper respiratory tract 134 suggesting a modulatory role for NO in nasal barrier function and clearance. Novel NO agents building on this observation are in development (Table 8): Endogenous NO has potent pulmonary haemodynamic and bronchodilator effects physiologically. The importance of endogenous NO in preventing infection is apparent experimentally where inhibition of NO results in increased susceptibility to microbes including Leishmania spp., Mycobacterium spp. and Plasmodium spp. 135 Similarly, NO sources are used therapeutically, for example sildenafil in the management of pulmonary hypertension (Table 7).
In respect of airway epithelial cells, nitrite reduced P. aeruginosa biofilm growth. 136 In infection, NO reduces pulmonary vascular resistance and intrapulmonary shunt, and improves oxygen partial pressure in patients with acute severe pneumonia. 137 More specifically, inhaled NO improves arterial oxygenation enabling a reduction in inspired oxygen therapy and airway pressure support, and reduces lung infiltrates, in patients with severe acute respiratory syndrome (SARS). 33 These findings continue after termination of NO therapy suggesting that NO has both pulmonary vasodilator and anti-SARS effects. Small uncontrolled clinical studies have suggested that iNO may be beneficial in COVID-19. [138][139][140][141][142] iNO and novel NO agents are in development, primarily for COVID-19 at present (Tables 8, 9).
Dipyridamole, a phosphodiesterase 5 inhibitor, may also have similar beneficial effects in severe COVID-19. 35 A phase II clinical trial found that L-arginine might have beneficial effects when given on top of conventional therapy for tuberculosis (Table 7). 50

Urinary tract infections
There may also be a role for dietary nitrate/inorganic nitrite in the prevention and treatment of urinary tract infections. Many of the lower urinary tract opportunistic organisms (e.g. E. coli) possess nitrate reductases, this forming the basis of urine dipstick detection of nitrite. In acidic urine conditions, nitrite is reduced to NO with toxicity to bacteria; for example, transferring nitrite-rich urine containing E. coli to a more acidic environment (e.g. pH 5.5) dose-dependently inhibited bacterial growth, 143 an effect potentiated by vitamin C. The antibacterial potency is comparable to conventional antibiotics such as trimethoprim and nitrofurantoin. This approach has been tested by filling urinary catheter retention balloons with nitrite and ascorbic acid, resulting in measurable amounts of NO outside the membrane and effectively killing two strains of E. coli in the surrounding urine. 144 A similar approach found decreased bacterial counts and prevented biofilm formation by P. aeruginosa, K. pneumoniae, and Enterobacter cloace (but not E. coli or S. aureus). 145 Last, instillation of bacillus Calmette-Guerin (BCG, an attenuated strain of Mycobacterium bovis) into the bladder is used for the treatment of superficial/non-muscle invasive bladder cancer and carcinoma in situ. BCG induces long-term increases in NOS activity in urothelial cells 146,147 and the formed NO is toxic to the malignant cells. The use of BCG to provide non-specific protection against SARS-CoV-2 is to be tested 148,149 although vaccination in infancy does not appear to protect against COVID-19 in adults. 150 Other infections Nitrate (usually KNO 3 ) and nitrite (NaNO 2 ) have been used for millennia to preserve food, especially meat and fish. 151 Food preparation leads to reduction of nitrate to nitrite, and nitrite inhibits bacterial growth, especially Clostridium botulinum, a key and severe cause of neurotoxin poisoning. Additionally, nitrite adds colour to food, 60,152 flavour (in part by overcoming rancid tastes) and is an antioxidant.
NO donors have also been investigated for eradicating or dispersing biofilm organisms. For example, GTN synergises with citrate and ethanol in eradicating biofilms (related to S. aureus, MRSE, P. aeruginosa and C. albicans) in an experimental catheter lock model. 153 Similarly, isosorbide mononitrate synergised with antibiotics to disperse then kill S. aureus. 154 An NO-releasing contact lens has been developed to treat microbial keratitis due to P. aeruginosa and S. aureus). 155 GTN may have improved outcome after infection in participants enrolled into the RIGHT-2 trial, a study where paramedics recruited patients with suspected stroke and randomised them to GTN versus sham. Overall, the trial was neutral. 156 However, in a planned subgroup analysis of those participants with a final diagnosis of a non-stroke mimic, functional outcome was better with GTN. 156 In a post hoc analysis of participants in this subgroup, GTN was associated with a better outcome in those with a final diagnosis of infections of the respiratory and urinary tracts which raises the possibility that NO was treating the infectious cause underlying the stroke mimic diagnosis.
The prostaglandin-cyclic adenosine-phosphodiesterase-3 system As with the NO system, the prostaglandin-cyclic adenosine monophosphate-phosphodiesterase-3 (PG-cAMP-PDE3) system has similar vasculo-protective roles with anti-leucocyte, antiplatelet and anti-smooth muscle, and pro-endothelial effects. It is therefore interesting to note that prostaglandins (PGA 1 , PGJ 2 ), including prostacyclin (PGI2 and analogues), have been reported to have antiviral effects. [157][158][159][160] Whether drugs based on these 161 or the PDE3 inhibitor, cilostazol, have efficacy against SARS-CoV-2 remains to be investigated. Further, endogenous NO and PGI 2 work together in the vascular tree, and it is conceivable that their potential antimicrobial effects will similarly synergise. Their combination, in the forms of ISMN and cilostazol, have been tested after stroke 162 but not yet reported for the prevention or treatment of infection.

Interaction between nitric oxide and vaccine efficacy
The interaction between diet, nutrition state and vaccine effectiveness has been assessed in multiple studies, principally in low-middle income countries where vaccination is paramount, especially in children, and yet where malnutrition may be widespread. In a systematic review and meta-analysis of observational studies and randomised controlled trials, there was little suggestion that malnutrition had any effect on vaccine responses 163 ; similarly, supplementation with vitamins and D, and iron and zinc, did not appear to modify responses. In preclinical studies, protein-energy malnutrition had limited influence on vaccine efficacy in mice. 164 The effect of dietary nitrate levels on vaccine efficacy is unstudied.
If nitric oxide derivatives attenuate microbial infections, then the efficacy of vaccines based on live attenuated viruses and bacteria (such as measles, poliovirus, BCG) might be attenuated by treatment with NO. Although there are many factors known to alter vaccine effectiveness (e.g. age), the effect of NO has not been studied.

Post-infection morbidity
Many infections cause long-term morbidity with chronic fatigue syndrome (CFS) and symptoms including fatigue, tiredness, myalgia, cognitive impairment and depression. Example associated microbes include Borrelia burgdorferi (Lyme disease), Chlamydia pneumoniae (community acquired pneumonia), Epstein-Barr virus (infectious mononucleosis), human herpes virus 6 (exanthema subitem), human immunodeficiency virus (AIDS), polio virus, SARS-CoV-1 virus (SARS), SARS-CoV-2 (long-COVID) and West Nile virus (fever). [165][166][167] Although CFS may represent chronic or latent infection, it is more likely to reflect the presence of post-infectious chronic inflammation. Hypothetically, these patients might benefit from inorganic nitrates in view of their positive effects on exercise performance (elite athletes take beetroot juice for this purpose) and cognition, 76,95 and potentially antimicrobial effects, a question that needs addressing (Table 7). A phase II trial of L-citrulline is studying this approach in patients with post-polio syndrome. 168 Excess nitric oxide during infection During severe infection, sepsis (defined as "life-threatening organ dysfunction caused by a dysregulated host response to infection") often develops. Septic shock is a subset of sepsis and is a leading cause of death worldwide. 169 It manifests as hyper-or hypo-pyrexia, altered mental state, hypotension, tachycardia, tachypnoea, hypoxia, anuria and/or lactataemia. This can occur with many infections due to: • Gram negative bacteria: Bacteroides fragilis, C. pneumoniae, Enterobacter spp., E. coli, Haemophilus influenzae, Klebsiella spp., Legionella spp., Neisseria meningitidis, Proteus spp., P. aeruginosa, Yersinia pestis. 63 • Gram positive bacteria: Clostridium spp., Enterococcus spp., Listeria monocytogenes, Staphylococcus spp., Streptococcus agalactiae/pneumoniae/pyogenes Typically, autoamplification of circulating cytokines (so-called cytokine storm) leads to excess NO synthesis, mostly derived from inducible NOS, leading to high circulating NO levels and the development of septic shock. In these circumstances, treatment with exogenous NO might be inappropriate. Trials of inhibiting endogenous NO synthesis with NOS-inhibitors in critically ill patients with sepsis have been reported although, disappointingly, did not improve outcome; indeed, the non-specific NOS-inhibitor, NG-methyl-L-arginine hydrochloride (L-NMMA, 546C88), was associated with increased death. 174 It is not clear why inhibiting NO synthesis was ineffective but non-selective NOS inhibitors were used meaning that both toxic (iNOS) and beneficial (eNOS) sources of NO were inhibited; pharmacologically, such inhibitors will have reduced cardiac output, organ perfusion and tissue oxygenation. In the absence of licensed selective iNOS inhibitors, perhaps the analogous approach used in the management of hyperthyroidism using block (with carbimazole) and replace (thyroxine) might be effective, i.e. block NOS activity and replace with a low dose of a NO donor. That excess NO is dangerous does not mean that pharmacological doses of NO cannot be effective ( Figure 1, Tables 6, 11) since all effective interventions in medicine have an inverted "U" dose response.
Some infections have opposing in vitro and in vivo responses to NO. For example activated macrophage-derived NO or NO donors such as SNAP reduced Trypanosoma brucei proliferation in vitro 175,176 whereas endogenous iNOS-derived NO suppressed protozoa-antigen specific T-cell proliferative responses and so worsened infection, at least in infected mice. 176 Intracellular protozoal infections are unlikely to be affected in this manner since macrophage-derived NO would be able to act directly on pathogens such as Leishmania major. 177 Other infections do not appear, at least in vitro, to induce iNOS. For example, Cryptococcus neoformans failed to induce NOS in primed macrophages, 178 apparently due to a lack of TNF-α secretion, probably because the polysaccharide capsule masked the signal for TNF-α secretion. Interestingly, non-encapsulated mutants of C. neoformans did induce endogenous NOS.
High levels of iNOS activation were antimicrobial in studies of malaria. Based on monocyte-derived mRNA levels in circulating blood, uncomplicated malaria was associated with increased levels of iNOS activation in contrast to patients with severe malaria who had lower levels. 179 The dual effects of NO in malaria, i.e. both low and high levels appear to be hazardous, are further reviewed. 180,181 Epidemics and pandemics Over recorded history, most epidemics and pandemics have resulted from viral infections including Ebola (viral haemorrhagic fever), influenza (H1N1, H2N2, H3N2, H3N8), HIV-1 (AIDS), polio (poliomyelitis), smallpox, yellow fever, zika or corona (OC43, MERS-CoV, SARS-CoV-1/2) viruses. Bacterial pandemics have resulted from Vibrio cholerae (cholera), S. enterica (typhoid fever) and Yersinia pestis (plague). Studies in vitro have reported findings suggesting that NO can reduce infection for some of these pathogens (Tables 2, 3) but information appears to be lacking for smallpox, yellow fever, zika and cholera (Table 10).
With multiple pandemics over the last 100 years, it is only inevitable that further ones will occur and some, like COVID-19, will comprise a "global catastrophic biological risk". 182 Global pandemics will most likely be caused by a respiratoryspread virus that crosses over from animals such that humans have no inherent immunity to it. Likely candidates include orthomyxoviruses (especially influenza A viruses such as H7N9), paramyxoviruses (e.g. measles, mumps, croup), pneumovirus (e.g. human metapneumovirus), coronaviruses and picornaviruses (especially rhinoviruses and enteroviruses). All of these have had strains that have crossed from animals to humans. This emphasis on RNA viruses is because DNA viruses tend to have lower mutation rates and, therefore, evolve more slowly and are less likely to escape the human immune system within the first rounds of infection. Nevertheless, DNA viruses, such as pox or herpes viruses from great apes or monkeys, do have the potential to jump species. Non-viral causes of pandemics are less likely since most bacteria will be treatable with broad-spectrum antibacterial agents, most fungi are thermally restricted, and prions would require massive food contamination (and only spread slowly). 182 Protozoa are usually thermally restricted although global warming may allow malaria to spread more widely in temperate zones.
Unfortunately, pandemics/epidemics may co-exist as seen with SARS-CoV-2 and dengue in Brazil, 183 and both with S. enterica in Pakistan; 184,185 in part, this reflects increasing travel with aircraft providing a portal for numerous microbes. 186 Of theoretical concern was the potential for COVID-19 and epidemic influenza to co-exist during winter in the Northern hemisphere, this possibly leading to a dramatic increase in deaths. 187 Nevertheless, 'flu rates were very low in both southern and northern hemisphere 2020 winters, presumably due to hands, face, space, mask and fresh air measures. All-in-all, the absence of a true broad-spectrum of antiviral agents is a major concern 188 and a potential agent such as NO with antimicrobial effects that extend beyond viruses would be most welcome.

Implications for SARS-CoV-2 and COVID-19
One possible explanation for the observation that COVID-19 outcomes are worse in older people, males, black or Asian ethnicity, and those with co-morbidities such as diabetes, hypertension, stroke and chronic lung disease, 189 is that these groups have lower vascular NO activity 6,7,11 and so mount a sub-optimal host response against infection. Increasing NO availability is therefore a potential therapeutic strategy. Several NO sources have potential relevance to preventing and treating COVID-19. L-arginine, sodium nitrite, GTN, SNP, NO and dipyridamole each have clinical antimicrobial activity and can be administered, variously, orally, intravenously or as NO gas in the intensive care unit. Transdermal  GTN, and oral ISMN, dipyridamole and sildenafil may be administered in the community or hospital. Of these, NO gas, dipyridamole and sildenafil are already being tested for preventing or treating COVID-19 (Table 9). It remains to be determined if increasing dietary nitrate may be a cost effective and safe intervention of widespread health relevance for the prevention of COVID-19 and, indeed, other emerging, pandemic, epidemic or endemic infections. Recent trial evidence provides indirect supporting evidence for the potential anti-SARS-CoV-2 effect of NO. First, dexamethasone and tocilizumab reduced death in patients in intensive care units, 190,191 and these agents and NO share anti-inflammatory effects. And second, interferon-ß reduced the need for intensive care in COVID-19 patients; 192 type I interferons increase iNOS activity and so have antimicrobial effects, as seen with L. major and Burkholderia pseudomallei. 193,194

Discussion and conclusions
Nitric oxide is a fundamental molecule with wide-ranging and potent vascular, anti-platelet, anti-inflammatory and tumoricidal effects. Further, there is a large volume of literature spanning the last 30+ years demonstrating that NO also has potent in vitro antimicrobial effects on a wide variety of viruses, bacteria, protozoa, fungi and yeasts; these are supported by a modest number of in vivo studies. Further, several positive clinical phase II trials of NO have been reported in viral, bacterial, protozoa and fungal infections, these relating particularly to skin and respiratory infections administered by cream and gas respectively. Although not from randomised trials, there is also evidence that dietary nitrate modifies the oral biome and so reduces dental caries.
However, these results cannot be considered persuasive on their own. First, few neutral or negative studies have been reported suggesting that there may be a risk of publication bias. Second, conflicting data in some dual-protocol studies with positive in vitro and neutral/negative in vivo data suggest that although NO is antimicrobial per se, the local tissue environment may overcome or reverse this effect. Third, organic nitrates can suffer from tolerance and may lead to the generation of reactive NO species such as peroxynitrite and S-nitrosothiols which might exacerbate rather than attenuate infection. Fourth, resistance may develop although this seems unlikely to be a generic issue, not least because NO levels can change, and be changed, much faster than any microbe can raise defensive mechanisms. Fifth, some microbes can produce their own NO and use this to resist the oxidative stress induced by external NO and its derivatives. Sixth, excess NO production is associated with the development of septic shock which might suggest that any NO is ineffective. Potentially, unsuccessful trials of non-selective NOS-inhibitors in severe sepsis may have confused the issue, perhaps by suggesting that treatment with NO is not important in infection. Seventh, positive clinical studies have been performed in environments where very high local concentrations of NO can be achieved and without the risk of reactive responses, in particular on the surface of tissues such as cream on the skin, dietary nitrate in the mouth, nitrite in the stomach, NO gas in the lungs and nitrite in the bladder; whether NO is effective as an antimicrobial within tissues and the vascular tree remains to be determined.
There are many sources of NO suitable for studying the prevention and treatment of milder infections in the community and hospital (e.g. topical sodium nitrite, oral NO donors such as ISMN, or oral PDE5 inhibitors such as dipyridamole or sildenafil), and treatment of serious infections in hospital (e.g. intravenous L-arginine, sodium nitrite or NO donors such as GTN or SNP, of NO gas). NO may also be delivered via a high nitrate diet, thus offering a widely available and inexpensive public health approach to potentially reducing and attenuating the severity of infections worldwide. This approach has the added advantage that such diets are already known to reduce vascular disease and some cancers, and possibly other inflammatory diseases and dementia.
In summary, the wealth of in vitro data suggest that NO has generic antimicrobial effects. However, some data suggest that NO may be ineffective or even hazardous and these reinforce our view for the need for large scale clinical trials of NO donors in the community and hospitals to prevent and treat infections. Although such studies need to focus urgently on the COVID-19 pandemic (especially with the lack of broad spectrum antiviral agents 188 ), other pathogens also need to be targeted. However, patients with established septic shock should not be administered NO donors to avoid exacerbating vascular collapse. One utopian vision would be demonstration that high dietary nitrate intake produces pre-or postexposure prophylaxis against infections and their severity in the community whilst NO donors are effective antimicrobial treatments for use by general practitioners and in hospitals.

Data availability
No data are associated with this article. nitrate, L-arginine)….", please also discuss L-citrulline as a source of NO. We recently reviewed this topic and you may want to use this review as a source to expand on this 1 .
In Table 1, please add L-citrulline as a substrate and give the appropriate biochemical equation.
○ Under Therapeutic inorganic nitrite and nitrate, please consider adding the use of nitrite/nitrate in animal models of wound healing. Refer to Afzali et al. 2020 2 for an introduction to this. ○ Under novel nitric oxide agents, please add RRx-001, which is a novel NO modulator 3 .

○
As summarized in the review, dietary intake of foods such as beetroot that are rich as a source of nitrate have many health benefits. And as mentioned, "Dietary nitrate has profound metabolic effects and appears to have the potential for reversing the metabolic syndrome and have anti-diabetic effects.". However, although the beneficial metabolic effects of inorganic nitrate and nitrite in type 2 diabetes mellitus have been documented in animal experiments, this is not the case for humans. Perhaps this should be mentioned on page 18 of the review 4 .
○ 2020, and the reference itself.
As summarized in the review, dietary intake of foods such as beetroot that are rich as a source of nitrate have many health benefits. And as mentioned, "Dietary nitrate has profound metabolic effects and appears to have the potential for reversing the metabolic syndrome and have anti-diabetic effects.". However, although the beneficial metabolic effects of inorganic nitrate and nitrite in type 2 diabetes mellitus have been documented in animal experiments, this is not the case for humans. Perhaps this should be mentioned on page 18 of the review 4 .