Oxidants and Antioxidants in Viral Diseases: Disease Mechanisms and Metabolic Regulation
The Journal of Nutrition, Volume 127, Issue 5, May 1997, Pages 962S–965S, https://doi.org/10.1093/jn/127.5.962S
Published: 01 May 1997
Reactive oxygen and nitrogen metabolites play a complex role in many diseases and in metabolic regulation. Because viruses replicate in living cells, such metabolites influence the growth of viruses in addition to serving as a host defense mechanism. Low levels of reactive oxygen species (ROS) play a role in mitogenic activation, and the early phase of lytic and nonlytic virus infection indeed resembles that of mitogenic cell activation. In addition to these subtle cell-activating effects shared by many viruses, influenza and paramyxoviruses activate a respiratory burst in phagocytic cells. These viruses are toxic when injected in animals. Cells lavaged from the lungs of mice infected with influenza virus are primed for enhanced superoxide generation. Moreover, xanthine oxidase is enhanced and the buffering capacity of small molecular antioxidants is decreased in the lungs, suggesting that infection leads to oxidative stress. The wide array of cytokines produced in the lungs during influenza could contribute to the systemic effects of influenza. Oxidative stress has also been shown in human immunodeficiency virus (HIV) infection in humans. Via activation of NFKB, ROS may activate viral replication, but oxidants are believed to contribute also to the loss of CD4 T cells by apoptosis. Antioxidants, together with agents interfering with the harmful effects of cytokines and lipid mediators, may have a role in the treatment of viral diseases. Such agents could not only alleviate disease symptoms but also decrease the long-term effects of chronic oxidative stress, which have been linked to the development of cancer in some viral infections.
virus, influenza, oxidants, antioxidants, nitric oxide
Reactive oxygen species (ROS)3
and reactive nitrogen species (RNI) are key elements in antimicrobial and antitumoral defense, but they also contribute to the pathogenesis of a wide array of diseases, including degenerative conditions such as Alzheimer’s disease (for reviews, see Bredt and Snyder 1994, Halliwell et al. 1992). The fact that ROS and RNI are intimately involved in metabolic regulation and physiology is of particular relevance, because viruses depend on the biosynthetic mechanisms of their host cells. By their role in cell activation (reviewed in Burdon 1995), ROS may facilitate or even promote replication of these parasites, depending on the cell and virus involved (Albrecht et al. 1992, Pace and Leaf 1995).
REACTIVE OXYGEN SPECIES AND VIRAL INFECTION
A role of oxidants in the inactivation of viruses was detected as early as 1970 (Belding et al. 1970), but the metabolic role of oxidants in viral infection became apparent only later. In a study on the effects of Semliki Forest virus infection on chick embryo cells, we found that the early biochemical effects of infection were similar to those which appear early in mitogenic cell activation. In addition, it had been reported that many viruses grow better in lymphocytes when the cells are treated with mitogenic lectins and that influenza viruses of the H2N2 subtype were mitogenic in lymphocytes. Looking for evidence of activation in other cell types, we observed that Sendai virus and influenza virus were capable of activating a respiratory burst in phagocytic cells in the absence of antiviral antibody (for a review, see Peterhans 1994). The level of ROS production during the respiratory burst exceeds that required for cell activation (Burdon 1995), which nicely illustrates the dual role of ROS: metabolic regulation of cells and a phagocytic cell defense mechanism. Subsequent work showed that herpes viruses metabolically activate their host cells in the early phase of infection (Albrecht et al. 1992). In addition, confluent, quiescent cell monolayers poorly support viral growth whereas viruses grow well in semiconfluent, proliferating cells.
The observations made with Sendai and influenza viruses in vitro were of interest with respect to early reports on the toxic effects of these viruses in rabbits. Like the activation of the respiratory burst in phagocytic cells, virus replication was not required for induction of toxic effects. In fact, the similarity extended even further, because filamentous particles were more toxic in vivo and more potent in burst activation in vitro (reviewed in Peterhans 1994). On the basis of the close similarity between the two phenomena, we proposed a role for ROS in the pathogenesis of influenza and other viral infections (Peterhans et al. 1987).
INFLUENZA VIRUS INFECTION AND OXIDATIVE STRESS
In this model, mice are infected intranasally with influenza virus. Infection remains restricted to the airways and lungs, but the systemic effects of infection are dramatic and include weight loss, decreased body temperature, and decreased pO2
and increased pCO2
in the blood. Mice die 5 or 6 d after infection. Cells lavaged out of the lung showed a marked increase in O⋅−2
production when stimulated with phorbol myristate acetate (PMA), and xanthine oxidase, an enzyme synthesizing O⋅−2
, was increased in lung homogenates, indicating enhanced ROS production during the course of influenza infection in mice. Analysis of major antioxidants (α-tocopherol, ascorbate and glutathione) revealed no changes in the redox status, but the overall concentrations of these antioxidants decreased during the course of influenza. Taken together, these observations suggested that infection was associated with oxidative stress. In a model somewhat different from ours, Oda and co-workers (1989) showed that intravenously injected pyran copolymer–conjugated superoxide dismutase protected mice from the lethal effect of influenza. Though compelling, this observation is difficult to interpret because pyran copolymers are well-known antiviral agents (Kunder 1993). In addition to ROS, NO⋅
seems to play a role in the pathogenesis of influenza (Akaike et al. 1996, Bykova et al. 1991).
The role of oxidants in influenza is complicated by several factors. As opposed to situations of systemic exposure to oxidants, such as in whole-body irradiation or intoxication with radical-generating agents, influenza infection is restricted to the airways and lungs. In addition, within the infected lungs, areas with severe inflammation may be found next to relatively unaffected areas. The compartmental nature of influenza virus infection also poses problems in the analysis of the tissue redox status, because methods available for this purpose are based on studies of whole-tissue homogenates. In addition, many other mechanisms can contribute to disease symptoms and lesions at locations where no virus is present. For example, humans infected with influenza virus exhibit severe systemic symptoms such as fever, headache, myalgia and anorexia (Nicholson 1992). We demonstrated in the lung lavage fluid of influenza-infected mice a wide array of cytokines and lipid mediators that could mediate systemic effects of the local infection (Hennet et al. 1992b). It is noteworthy in this respect that injection of certain cytokines, in particular interferons, in humans causes symptoms closely resembling influenza (Dvoretzky 1990).
The influenza model also provides another level of complexity in the role of oxidants and antioxidants in viral infection. Although ROS have been shown to be virucidal (Belding et al. 1970), they may also contribute to an increase in the viral titer of influenza virus. The reason for this is in a peculiarity of influenza virus virulence, combined with an effect of ROS on protease inhibitor present in the lung surfactant. Influenza virus hemagglutinin, the surface glycoprotein responsible for receptor binding and entry into the host cell, is synthesized as a precursor protein, HA0. This protein is cleaved intracellularly into the dipeptides HA1 and HA2. Strains that possess a hemagglutinin with an amino acid sequence that fits optimally with the sequence specificity of the intracellular protease is released into the extracellular space in the HA1/HA2 form, whereas virus with less optimal sequence is released mostly in the HA0 form. Only virus of the HA1/HA2 form is infectious. The overall speed at which virus infection spreads in the airways and lungs depends on the relative proportion of infectious virus made in each round of replication. Consequently, cleavage of HA0 into HA1 and HA2 is an important determinant of virulence (for a review, see Rott et al. 1995).
If the noninfectious HA0 form of the virus is released from cells without being cleaved, extracellular proteases present in pulmonary surfactant can proteolytically cleave this protein (Kido et al. 1993). As a protective mechanism, anti-proteases are present on the surface of alveoli. However, the anti-proteases can be inactivated by ROS. In this regard, it is important to note that during lung inflammation phagocytes increase in number and produce ROS (McCusker 1992). We have shown in vitro that oxidant-treated anti-protease is unable to prevent trypsin from cleaving HA0 to HA1/HA2, resulting in a 10,000-fold increase in infectious virus (Hennet et al. 1992a).
HUMAN IMMUNODEFICIENCY VIRUS AND OXIDATIVE STRESS
Humans infected with human immunodeficiency virus (HIV) have been shown to be under chronic oxidative stress. Perturbations of the antioxidant defense system in HIV-infected humans include changes in ascorbic acid, tocopherol, carotenoids, selenium, superoxide dismutase and glutathione. In addition, elevated levels of hydroperoxides and malondialdehyde are found in plasma of HIV-infected individuals. In vitro, manipulations that result in enhanced oxidative stress increase the replication of HIV, possibly via activation of NFKB, a transcription factor that stimulates the replication of HIV and of certain cytokines, among them tumor necrosis factor-α (for a review, see Dröge et al. 1994). How increased oxidant production initially triggered is in HIV-infected humans has not been determined to date. Possibilities include stimulatory effects of gp125 (Pietraforte et al. 1994) and Tat, the viral-transactivating protein secreted from virus-infected cells (Westendorp et al. 1995).
Mycoplasmas that are known to enhance the replication of HIV could also act by increasing oxidative stress. As with paramyxoviruses and influenza viruses, these agents can directly activate the generation of ROS in phagocytic cells (Köppel et al. 1984). Moreover, certain mycoplasmas produce H2
(Arai et al. 1983), and co-infection with mycoplasmas and HIV may result in the release of H2
from T cells (Chochola et al. 1995).
Collectively, oxidant production could enhance HIV replication via activation of NFKB and indirectly through activation of genes that further promote oxidative stress and hence HIV replication (e.g., tumor necrosis factor-α). The altered redox status seems to contribute to AIDS in several ways, including by apoptosis of CD4 T cells and immune dysfunction (for a review, see Dröge et al. 1994).
HEPATITIS-CAUSING VIRUSES AND OXIDATIVE STRESS
Via oxidant stress, viral infections may also contribute in particular to cancer of the liver. Hepatitis-causing viruses are taxonomically quite diverse and include Picorna-, Flavi-, Hepadna- and Caliciviruses as well as delta agent associated with hepatitis B virus (Anonymous 1990). Only in the case of hepatitis B virus is there convincing evidence for a “classical” mechanism of viral carcinogenesis involving viral effects on the host genome (for a review, see Milich et al. 1994). Significantly, infection with hepatitis viruses may last for years and is accompanied by chronic inflammation, a feature shared by infection with certain bacteria and parasites and associated with enhanced production of ROS and RNI (Ohshima and Bartsch 1994). It has recently been argued that in addition to genotoxic effects, ROS and RNI may also contribute to cancer because tissue destruction leads to increased cell proliferation, a condition typical of tissue repair and associated with fixation of mutations (Ames et al. 1995).
NO⋅AND VIRUS INFECTION
Formation of NO⋅
may be a feature of several viral diseases. Of particular interest are infections of the central nervous system, because NO⋅
is a second messenger in neurons (Bredt and Snyder 1994). Formation of NO⋅
could explain some mental alterations as well as changes related to the loss of neurons due to RNI and ROS (Dawson and Dawson 1996). Of interest in this respect is Borna virus, which infects horses and sheep and possibly also humans (Rott et al. 1985). Evidence of a pathogenic role of NO⋅
was obtained also in rabies, experimental allergic encephalomyelitis (Hooper et al. 1995) and lymphocytic choriomeningitis in mice (Campbell et al. 1994).
Lymphocytic choriomeningitis virus (Butz et al. 1994) and bovine viral diarrhea viruses (Adler et al. 1994) were shown to prime mouse and bovine macrophages, respectively, for enhanced NO⋅
production. This effect may be related to the immunosuppression observed during infection in vivo. Human immunodeficiency virus infection (Bukrinsky et al. 1995) and HIV gp120 (Dawson and Dawson 1996) were shown to enhance the production of NO⋅
in human monocytes and exert neurotoxic effects in vitro via NO⋅
NO-mediated toxicity of HIV and Maedi Visnavirus (a lentivirus of sheep) Tat was demonstrated in rodents after intracerebral injection (Hayman et al. 1993, Philippon et al. 1994), and Liu and Hotchkiss (1995) demonstrated that the glycoprotein of woodchuck hepatitis virus enhances the production of NO⋅
in hepatocytes. Although these examples would suggest that NO⋅
plays a negative role in quite diverse viral infections, it is important to note that NO⋅
can also have antiviral effects. For example, treatment of mouse macrophages with interferon-γ led to increased NO⋅
production concomitant with the inhibition of ectromelia, vaccinia and herpes simplex viruses. In the presence of inhibitors of NO⋅
synthase, the protective effect of interferon-γ was abrogated. Inhibition of viral replication by NO⋅
was also reported for vesicular stomatitis virus (reviewed in Mannick 1995).
OXIDANTS IN VIRAL DISEASES: ANTIOXIDANTS AGAINST VIRAL DISEASES?
As discussed above, the role of oxidants in viral diseases is more complex than the often cited duality between antimicrobial and autotoxic effects because it includes metabolic regulation both of host metabolism and viral replication. Antioxidants can therefore be expected to act at many different levels, calling for more investigations on the effect of antioxidants in viral diseases. When antioxidants are given in a blend mimicking that present in fruits and vegetables, toxic or undesired effects are particularly unlikely. This may not necessarily be the case with synthetic antioxidants. Brugh (1977) observed that butylated hydroxytoluene (BHT), a synthetic antioxidant used to prevent deterioration of foodstuffs, protected chickens against the lethal effect of virulent Newcastle disease virus. This antioxidant also prevented or delayed seroconversion induced by vaccination with avirulent Newcastle disease virus. Although direct effects of butylated hydroxytoluene on virus infectivity were suspected to be the cause of these effects, it cannot be excluded that protection against the lethal effect as well as inhibition of the immune response was achieved via influences on host physiology.
As outlined above, a number of additional host mechanisms have been shown or are suspected to contribute to the pathogenesis of viral infections, including excessive cytokine and lipid mediator release and complement activation (Peterhans 1994). It seems logical therefore to seek ways to interfere with the activation or undesired effects of these pathways. The limitations of interfering with such mechanisms of viral diseases are similar to those when interfering with oxidant generation, because these pathways are associated with the normal host physiology as well as with host pathology. It is clear that this approach could be criticized because it involves a wide variety of drugs, rather than the magic one or two putatively specific ones used in modern pharmacotherapy. However, because the symptoms and pathology of viral diseases are ultimately the result of complex host reactions in addition to direct viral effects, there is a scientific basis for this strategy of viral disease therapy. Clearly, this does not obviate the further search for drugs that specifically interfere with viral replication, particularly with viruses that cause chronic infections.
I thank Giuseppe Bertoni and Thomas W. Jungi for critically reading the manuscript.
1” Presented as part of the symposium “Newly Emerging Viral Diseases: What Role for Nutrition?” given at Experimental Biology 96, April 17, 1996, Washington, DC. This symposium was sponsored by the American Society for Nutritional Sciences and supported in part by grants from Henkel Corp., Hoffman-LaRoche, Inc., International Life Sciences Institute, Lederle Consumer Health, Nestle, and the Selenium-Tellurium Development Assoc., Inc. The guest editors for the symposium publication were Orville A. Levander, U.S. Department of Agriculture, Agricultural Research Service, Beltsville, MD, and Melinda A. Beck, University of North Carolina, Chapel Hill, NC.
2 “Supported by Swiss National Science Foundation grant 31-39733.93.
© 1997 American Society for Nutritional Sciences