Epstein–Barr virus-acquired immunodeficiency in myalgic encephalomyelitis—Is it present in long COVID?

Myalgic encephalomyelitis or chronic fatigue syndrome (ME/CFS) is a chronic, multisystem disease characterized by unexplained, disabling fatigue and a combination of nonspecific symptoms lasting at least 6 months [1, 2]. The main symptoms are fatigue, post-exertional malaise, cognitive dysfunction, sleep disturbances, generalized pain, and gastrointestinal discomfort [3‐5]. The prevalence of patients with ME/CFS in the USA is estimated to be between 1.7 million and 3.38 million [6], corresponding to a range from 0.1 to 2.5% in the general population, depending on the diagnostic criteria applied [7]. The main causes are latent herpesvirus infection [2, 8‐34], a chronic inflammatory state [1, 35‐46], hypocortisolism [35, 47‐60], autoimmune antibodies [13, 20, 24, 29, 35, 61‐77], genetic predisposition [8, 35], or microclot formation [33, 78, 79].

SARS-CoV-2 infection, which causes a multisystem disease known as COVID-19 [80], is responsible for a worldwide pandemic with substantial mortality and morbidity [81]. Of all COVID-19 cases, approximately 81% have mild disease, 14% severe disease, and 5% develop critical illness [82]. It has recently been observed that approximately half of recovered COVID-19 patients continue to experience at least one long-term symptom [83, 84]. Within this group, long-lasting fatigue is particularly common, affecting 28.3% of patients [83]. The most common persistent symptoms are asthenia, dyspnea, pain, exercise intolerance, sleep disturbances, smell disturbance, taste disturbance, cognitive dysfunction, and neurological and digestive disturbances [5, 85‐89]. Several terms, such as long COVID [89, 90], long-haul COVID-19 [91, 92], post-COVID syndrome [93], and post-acute sequelae of SARS-CoV-2 infection (PASC) [94], are used to refer to the persistence of these symptoms more than 12 weeks after SARS-CoV-2 infection [95, 96]. Although its origin is still unknown, among the main causes related to long COVID (LC) are the persistence of chronic SARS-CoV-2 infection [33, 97‐105], autoimmunity [106‐116], chronic inflammatory state with immune alteration [110, 117‐127], Epstein–Barr virus reactivation [128‐140], hypocortisolism [128, 133, 141‐143], or microclot formation [33, 144‐149].

The similarity in both symptoms and the different causes of ME/CFS and long COVID (LC) have led to searches for some connection between these two diseases [5, 33, 124, 139, 150‐159]. The results of these studies suggest a high degree of similarity between LC and ME/CFS. Of the 29 symptoms observed in patients with ME/CFS, 25 were reported in patients with LC [156]. In addition, there are several articles that have linked EBV reactivations to the development of LC [138, 160‐162]. Therefore, given the relationship of EBV with both diseases [2, 8‐18, 20‐28, 30‐33, 128‐130, 133‐140], we propose that latency and recurrent EBV reactivation could generate both an acquired immunodeficiency syndrome in genetically predisposed individuals ME/CFS patients and, second, immunosuppression due to SARS-CoV-2 in the case of LC patients.

It should be mentioned that not only EBV infection has been linked to ME/CFS, but also human herpesvirus (HHV)-6, cytomegalovirus (CMV), human parvovirus B19, and enteroviruses [22, 23, 28, 163‐168], as well as bacterial and parasitic infections [169]. However, as in the case of LC, in these other post-infectious subgroups of ME/CFS, EBV could also play a possible role in immunological alterations and symptomatology, since the virus is present in 90% of the population [170] and therefore could be reactivated in the presence of immunosuppression [128‐130, 133‐140].

Epstein–Barr virus (EBV) is known to be involved in the development of some autoimmune diseases and certain cancer and is suspected to underlie the development of chronic fatigue syndrome/myalgic encephalomyelitis and LC [2, 8, 23, 26, 33, 129, 133, 134, 136‐138, 171‐182]. However, the pathways it exploits leading to the development of these diseases remain unclear, although it is thought to be mainly related to the genetic susceptibility of an individual [171].

HLA-II genes are highly polymorphic due to coexistence with different pathogens as humans moved to different regions, leading to the emergence of new alleles providing resistance to these pathogens [183, 184]. DR2-DQ6, DR3-DQ2, and DR4-DQ8 are among the oldest extant haplotypes [185]. These ancestral haplotypes have survived over time by being able to recognize a greater number of antigens than other haplotypes, enabling them to detect and eliminate a greater number of pathogens [183]. For example, HLA-DR4 alleles are associated with increased clearance of hepatitis B virus infection [186], and HLA-DRB1*0401 or DRB1*1501 alleles are associated with increased clearance of hepatitis C virus infection [187]. Surviving a greater number of pathogens has allowed these ancestral haplotypes to be inherited from generation to generation, but with the disadvantage that those older pathogens with which they have coexisted have also acquired resistance against them via co-evolution of their evasion mechanisms [171, 183]. One of these pathogens is EBV, which is the only member of the genus Lymphocryptovirus adapted to humans and was transferred to a hominid ancestor millions of years ago [188]. Its gp42 protein is an evasion mechanism that has evolved against these ancestral haplotypes, as the virus infects cells through the interaction of its viral protein gp42 with the β1 domain of HLA-II in the cell [189‐191]. In this way, its membrane fuses with that of the cell, forming a new gp42–HLA-II complex that alters antigenic presentation to CD4 T cells (Fig. 1) in individuals carrying these alleles [189, 190, 192], managing to suppress the activation (HLA-DR) of CD4 T cells and the presentation of EBV nuclear antigen (EBNA)-1 on HLA-II molecules [171]. Not presenting EBNA-1 to CD4 T cells via HLA-II allows these cells with latency I to forgo elimination, since CD4 T cells are the only ones capable of recognizing this latency [171, 193]. This allows any inflammatory stimulus (such as another infection) in any tissue to recruit leukocytes (among which are cells with EBV latency), ultimately leading to the formation of EBV-infected ectopic lymphoid structures (Fig. 2) [171‐176, 194‐196]. When unrecognized, they proliferate in these ectopic lymphoid structures, triggering the activation and release of IFN-γ by NK cells to restrict the transformation of B cells by the virus and as a response to the first inflammatory stimulus [197, 198]. In addition, the presence of foreign antigens from the first inflammatory stimulus leads to terminal differentiation and activation of EBV latent B cells, allowing the transition from the latent to the lytic phase of the virus, originating foci of viral reactivation [171, 172, 194]. Therefore, exposure to foreign antigens from the first stimulus or to EBV viral antigens leads to activation of CD4 T cells and IFN-γ release, provided that the cells presenting these antigens to CD4 T cells are not previously infected with EBV, since the gp42–HLA-II complex in these cells would prevent their activation [171, 194]. Both IFN-γ released by NKs and CD4 T cells upregulates HLA-II expression in adjacent cells (such as epithelial cells), which favors the acquisition of a nonprofessional antigen-presenting cell phenotype [199, 200]. The increased expression of HLA-II in these cells allows the newly generated viral particles to infect more cells through the gp42–MHC-II interaction, which ultimately leads to an increased number of latent cells that perpetuate the inflammatory state in that tissue by releasing proinflammatory substances (Fig. 1), such as Epstein–Barr virus-encoded small RNAs (EBERs) [201‐203], producing abortive reactivation, in which the resulting proteins can cause inflammation [204, 205] or induce the release of new virions that provoke an immune response [171]. This chronic inflammation provokes a chemokine response, leading to an increase in the recruitment of a greater number of B cells, allowing the perpetuation of viral infection and, ultimately, the formation of EBV reservoirs in these tissues [171, 194]. Increasing the number of EBV-dormant cells would also increase the evasion mechanisms of these cells by decreasing the activation and cytotoxic capacity of EBNA-1-specific CD4 T cells through the release of IL-10 (increased humoral Th2 response), chemotaxis of regulatory T lymphocytes (Treg) due to increased CCL20 expression in EBNA-1-mediated EBV latency cells, and release of EBV miRNA contained in exosomes, which could suppress the expression of target genes in the viral or host genome to maintain latent EBV infection [206‐208]. Consequently, the greater the number of antigen-presenting cells infected by this virus, the lower the activation of CD4 T cells, either by inhibition of the TCR–HLA-II interaction by the gp42–HLA-II complex or due to the evasion mechanisms of these latently infected cells. Thus, an acquired immunodeficiency in CD4 T-cell function develops, leading to increased proliferation of latent EBV cells, which increases the risk of neoplastic transformation or autoimmune disease in these tissues [171].

The characteristics of this acquired immunodeficiency could vary according to biological sex (Fig. 3). In fact, estrogens decrease the CD4/CD8 T-lymphocyte ratio and increase B-cell survival, antibody release, and expression of the class II major histocompatibility complex [209‐211]. In women, this increase in antibody levels would increase their resistance to viral infections under normal conditions, but under pathological conditions there is increased B cell survival, decreased CD4 T cells, and increased estrogen-induced HLA-II expression, which confer certain disadvantages against EBV, as they increase the survival of EBV-transformed B cells, and the increase in HLA-II expression allows more cells to become infected [171, 209]. In addition, increased HLA-II expression can lead to increased presentation of native cellular autoantigens or viral antigens, which can undergo posttranslational modifications, such as citrullination, and form neoantigens, through the endogenous antigen processing pathway, triggering the activation of autoreactive cells [171, 212‐215]. Therefore, both the increase in the survival of transformed B cells and the increase in antigen presentation generated by the increase in HLA-II expression by estradiol may favor an increase in the presentation of both self and foreign antigens during an infectious process, and those “abnormal” plasma cells producing autoantibodies survive longer, which consequently increases the probability of women developing autoimmune diseases or even cancer [171]. In the case of androgens, testosterone regulates the immune system by increasing the Th1 response and CD8 T cell activation while downregulating the NK cell response and HLA-II expression [209] . As antigen-presenting cells are important for the differentiation of CD4 or helper T cells into Th1 or Th2 cells, depending on the cytokines they release, sex hormones could determine whether CD4 T cells differentiate more into Th1 or into Th2 [209]. Thus, women have a higher incidence of sex-related diseases but generate a more effective response against extracellular infections by having more antibodies (Th2 response) [209]. The female immune system might however be more ineffective against intracellular pathogens than in the male immune system, as more cellular (Th1) response is needed to eliminate virus-infected cells [209].

In addition to these considerations, the menstrual cycle in women may also affect acquired immunodeficiency [216]. During the cycle, elevated estrogen and progesterone levels induce Th2-type responses, whereas low estrogen concentrations induce Th1-type responses [216]. In addition, progesterone suppresses Th1-type responses by stimulating regulatory T cells and increasing Th2-type responses [216]. Therefore, there is a predominantly Th1 response during the beginning of the follicular phase, the end of the luteal phase, and menstruation, since estrogen and progesterone levels are not elevated [216]. In contrast, during the late follicular phase, ovulation, and in the luteal phase, there is a predominantly Th2 environment, which causes some immunosuppression [216]. For this reason, during the phases with a predominant Th2 response, greater immunosuppression can appear, causing outbreaks of viral reactivation as occurs, for example, with the herpes simplex virus type 1, which in these phases usually causes vesicles to appear on the lips [217]. Therefore, if there is a previous immunodeficiency, it would worsen during these phases.

Among the immunological alterations presented by ME/CFS patients, there is reduced cytotoxic response by T lymphocytes and natural killer cells (NKs) [1, 35, 44, 357], reduced T-cell responses to mitogens and other specific antigens with decreased activation [1, 35, 358], deficient EBV-specific B- and T-cell response [21], increased Th2 response (especially IL-10) [35, 36, 38, 41, 46, 359], increased regulatory T cells (CD4+CD25+FoxP3+) [359], and increased IFN-γ levels [359]. In individuals with ME/CFS, ancestral HLA-II haplotypes, and latent EBV, these immunological alterations can be explained by the EBV-acquired immunodeficiency model, where the impairment of CD4 T-cell function and activation caused by these ancestral HLA-II haplotypes could generate T-cell depletion and thus an increase in EBV-infected ectopic lymphoid structures [8, 171]. These infected tissues, when not controlled by the adaptive response, could end up with chronic inflammation due to the activation of TLR3 receptors or RIG-I, which results in continuous detection of viral genetic material (EBERs) [231‐233], causing a chronic innate response and inflammation through the continuous release of proinflammatory cytokines, generating an increase in IL-1, IL-6, TNFα and IFN-γ in individuals with ME/CFS [37]. Chronic release of proinflammatory cytokines of innate immunity, increased Treg cell activity, aerobic glycolysis of infected cells, and activation of neutrophils by ANCA lead to excessive ROS production, and this could explain the increased oxidative and nitrosative stress, depletion of antioxidant substances, increased aerobic glycolysis in circulating cells, and increased damage to fatty acids, DNA, and proteins that occur in patients with ME/CFS [296‐298, 360]. This increased oxidative and nitrosative stress could be exploited by EBV-infected cells for transformation by methylation of viral DNA, preventing its replication and favoring the establishment of different types of latency that allow evading the immune system [299]. Furthermore, increased IL-10 levels in ME/CFS patients support the presence of a persistent infectious state and decreased activity of T cells and NKs [359].

Increased levels of TNF-α and IFN-γ in ME/CFS also explain the increased intestinal inflammation and permeability, development of irritable bowel, and increased bacterial translocation in these patients [35, 292, 359].

This acquired immunodeficiency could also justify the lytic or abortive reactivation of this and other latent pathogens, such as other herpesviruses and Parvovirus B19, that occurs in patients with ME/CFS [9, 10, 15, 17, 34, 361‐363] and could generate, on the one hand, the production of transient autoantibodies such as ANCAs, ANAs, or antiphospholipid antibodies (such as anticardiolipin) [71, 360], causing exacerbation of symptoms, and on the other hand, the increased release of viral dUTPases that increase neurological symptoms [10, 11, 26], generating more inflammation and disease flares.

Patients with ME/CFS may present with insulin resistance, increased blood insulin levels, high waist circumference, high triglycerides, and hypocortisolism [47‐60, 364]. Chronic elevated levels of insulin and proinflammatory cytokines could explain the negative regulation of cortisol secretion, where increased insulin causes a deficient cortisol response to hypoglycemia, as occurs in patients with ME/CFS assessed using the insulin-induced hypoglycemia test [54, 58, 253, 262, 263, 365]. Thus, it appears that patients with ME/CFS attempt to maintain a state of transient hypoglycemia caused by increased proinflammatory cytokines, hyperinsulinemia, and hypocortisolism to maintain a chronic innate immune response, decreased viral replication, and enhanced CD8 T-cell effector response [247‐249], though without being able to resolve the infection by presenting an alteration in CD4 T cell function and, therefore, a deficient cytotoxic response of CD8 T cells in individuals with “weak” HLA-II haplotypes against EBV. The increase of EBV latent cells in different tissues could lead to a more severe state of insulin resistance and, thus, to an increased risk of developing metabolic syndrome, type 2 diabetes mellitus, cardiovascular disease and Alzheimer’s disease in these patients [58].

Within the serotonergic alterations presented by patients with ME/CFS, it has been suggested that viral infection may cause damage to the blood–brain barrier, leading to the entry of viral genetic material that activates TLR3 receptors in microglia, leading to increased release of IL-1β and TNF-α, which increases SERT expression and activity in astrocytes and, therefore, increases 5-HT reuptake and its degradation to the metabolite 5-HIAA, thereby reducing extracellular levels of 5-HT [271, 285‐288]. This would result in reduced activation of neuronal 5-HT1A receptors, possibly inducing the sensation of fatigue and the development of depression [285‐287, 289]. Furthermore, during infectious processes in ME/CFS patients, both increased proinflammatory cytokines and increased oxidative and nitrosative stress due to activation of microglia via TLR3 could lead to increased IDO activity in microglia, resulting in reduced tryptophan levels (increased its catabolism), increased kynurenine catabolites, and decreased 5-HT and melatonin synthesis [366, 367]. Depletion of tryptophan together with increased kynurenine catabolites could lead to an impaired Th1 cell response and increased promotion of Th2 cells, favoring immune tolerance to the disease [325, 366]. Furthermore, within kynurenine metabolites, quinolinic acid has been shown to be an NAD+ precursor that has neurotoxic properties through its binding to the N-methyl-d-aspartate (NMDA) receptor, which leads to sustained Ca²⁺ influx and thus increased nitrosative stress [366]. This increased nitrosative and oxidative stress results in activation of poly(ADP-ribose) polymerase-1 (PARP-1) to prevent DNA damage [366]. Continued overactivation of PARP-1, however, results in depletion of intracellular NAD+ and ATP stores with consequent impairment in energy production and mitochondrial function, which have been reported in ME/CFS patients [366]. In addition, quinolinic acid stimulates synaptosomal glutamate release by acting as an NMDA receptor agonist and, in astrocytes, decreases the expression of glutamate transporters and increases their release, thereby increasing extracellular glutamate levels [367, 368]. Elevated glutamate, apart from generating neurotoxicity, can cause mitochondrial dysfunction and lower energy [366]. Therefore, increases in kynurenine and its metabolites resulting from increased IDO activity could contribute to neuroinflammation, increased glutamate, altered immune system, depression, altered gut microbiota, and fatigue in ME/CFS patients [366, 367, 369]. However, it is not only increased quinolinic acid that can explain the overstimulation of NMDA receptors. In brain areas of increased Cu use, ceruloplasmin deficiency and activation of microglia in response to the presence of viral genetic material could also contribute to the overstimulation of NMDA receptors, increased levels of nitric oxide/peroxynitrite levels, and the cognitive impairment present in ME/CFS patients [346, 369, 370].

Decreased activity of Cu-dependent enzymes, such as SOD, and decreased levels of ceruloplasmin in patients with ME/CFS [371] can be explained by a deficiency in Cu uptake due to overexpression of MTs in enterocytes in response to mucosal infection. This Cu deficiency could cause a decrease in ceruloplasmin synthesis and a decrease in the activity of Cu-dependent enzymes, contributing to increased oxidative stress.

On the other hand, the hypozincemia observed in patients with ME/CFS could occur due to increased proinflammatory cytokines in response to infection, as they increase the expression of zinc transporters and zinc-binding peptides, leading in the long term to a reduction in their availability in blood and their uptake by immune cells [328, 329, 372]. This results in a decrease in their activation in the presence of antigens and altered immune responses with downregulation of intracellular antioxidant activity [236, 329, 335, 372]. In addition, this deficiency could lead to increased MHC-II expression allowing infection of a greater number of cells by EBV in these patients [329].

Another abnormality present in patients with ME/CFS is the presence of fibrinolysis-resistant amyloid fibrin microbleeds and platelet hyperactivation [78], which could also be a consequence of the overproduction of proinflammatory cytokines or due to the binding of EBERs to TLR3 [231‐233, 308, 309].

In the case of LC, there is increased levels of depleted T cells (PD-1+/Tim-3+) [128], impaired T-cell function [119, 128], decreased recruitment of activated T cells [125], impaired cytotoxic response of NK cells [373], increased Th2 response [128], increased Treg cells [374‐376], increased proinflammatory cytokines (IFN-γ, TNFα, IL-1, and IL-6) and IL-10 [118, 125, 377, 378], increased oxidative stress [377, 379‐381], decreased antioxidant defenses [380], hypozincemia [380, 382], hypocortisolism [128, 133, 141‐143], presence of amyloid fibrin microbleeds [33, 146‐149], presence of transient autoantibodies (ANAs, ANCA and anti-cardiolipin) [106, 113, 383, 384], insulin resistance [385, 386], and increased IDO activity and kynurenine levels [387].

Just as the binding affinity of EBV peptides to MHC class II molecules can influence the generation of an antiviral response [191, 218], it could also occur with SARS-CoV-2 [388]. It is known that single amino acid alterations in immunogenic peptides can enhance or decrease the binding of the peptide to the MHC class II molecule, causing the differentiation of CD4 T cells into Th1 cells if the binding is strong or into Th2 cells if the binding is weak [389]. Therefore, having weak HLA-II haplotypes against SARS-CoV-2 could influence the loss of binding affinity between MHC class II molecules and viral peptides, leading to decreased immune protection and viral recognition [388]. For example, mutations in the Spike protein in Omicron cause the HLA-DRB1*03:01 allele to lose affinity for this protein [390]. Therefore, individuals with LC could have “weak” HLA-II haplotypes against SARS-CoV-2 different from those against EBV or they could be the same weak haplotypes against EBV. More research is needed in this area. In both cases, there would be impaired control of SARS-CoV-2-infected cells by CD4 T cells, as these are essential for their immunosurveillance and for maintaining CD8 T cell cytotoxic responses [230, 391]. Therefore, chronic antigenic stimulation by SARS-CoV-2 in the absence of CD4 T-cell support results in a loss of immunosurveillance of infected cells, generating chronic SARS-CoV-2 infection with increased viral reservoirs in tissues and leading to immune “exhaustion” by chronic exposure to these viral antigens, as occurs in other chronic infections [227, 230]. This immune exhaustion would lead to a decrease in the formation of memory T cells and an increase in the generation of short-lived effector cells and exhausted T cells, where these exhausted antiviral T cells, unable to become activated, are functionally ineffective and unable to exert their cytotoxic functions in response to antigenic stimulation [227, 229], resulting in impaired immunosurveillance of SARS-CoV-2-infected cells and other cells with latency to another virus, such as EBV [128, 139]. This could result in the recruitment of EBV-latent leukocytes to inflamed tissues, leading to the formation of EBV-infected ectopic lymphoid structures and the development of EBV-acquired immunodeficiency as in post-EBV-infected ME/CFS. This cellular impairment of EBV control can be observed by increased antibody responses to EBV and increased reactivation in patients with LC [8, 128‐130, 133‐140]. It has also been observed that symptoms of fatigue and neurocognitive dysfunction in patients with LC are associated with EBV reactivation [138].

Another commonality of both diseases is that women are at higher risk of developing them than men [35, 392]. This may be because estrogens decrease the CD4/CD8 T lymphocyte ratio, increase B-cell survival, antibody release, and HLA-II expression, and decrease the Th1 antiviral response [209‐211]. In addition, TLR3 expression on platelets is higher in women, thus increasing the risk of increased platelet hyperactivation and hypercoagulability [279].

In summary, according to the model presented here, ME/CFS develops in three steps: first, an acquired EBV immunodeficiency develops in individuals with “weak” HLA-II haplotypes against EBV, which prevents control of latency I cells. Second, ectopic EBV-latent lymphoid structures form in different tissues (including the CNS) that promote inflammatory responses and further impairment of cell-mediated immunity. Third, immune exhaustion occurs due to chronic exposure to viral antigens and the disease consolidates. In the case of LC, before the first step there is prior infection with SARS-CoV-2 in individuals with “weak” HLA-II haplotypes against this virus and/or EBV.

Based on this model of development of both diseases, one might think that treatment with hydrocortisone would be beneficial to resolve hypocortisolism, but it has been observed to have limited and even adverse effects in patients with ME/CFS, since hypofunction of the HPA axis is the consequence, and not the cause, of immunological alteration [54]. Moreover, given the increased Th2 response and impaired Th1 antiviral response, the use of hydrocortisone could further potentiate this immunodeficiency and EBV reactivation [393, 394].

Given that there could be an increase in 5-HT levels in inflamed tissues and peripheral blood but a decrease in levels in the CNS, it could be thought that the use of selective serotonin reuptake inhibitors (SSRIs) could help to restore serotonergic impairment. At the CNS level, long-term use of SSRIs could increase extracellular levels of 5-HT by inhibiting SERT, preventing it from degrading to 5-HIAA and thus making more 5-HT available to bind to neuronal 5-HT1A receptors and improve fatigue [395]. At the peripheral level, however, it could be detrimental because elevated peripheral blood 5-HT levels could be involved in increasing insulin and 5-HT release by pancreatic β-cells via the 5-HT3 receptor and inhibition of glucagon secretion by α-cells, resulting in exacerbation of hyperinsulinemia and hypoglycemia [282, 283]. This, together with the fact that SSRIs also cause increased ROS and oxidative damage in β-cells, could result in an increased risk of developing type 2 diabetes in the long term [396]. This could occur for example with fluvoxamine treatment. Khani and Entezari-Maleki suggested that fluvoxamine, an S1R agonist, could alleviate the symptoms of LC associated with EBV reactivation [131, 132]. It would do so by reducing the activation of XBP1 and the endoplasmic reticulum stress response [131, 132]. As an SSRI, it could also enhance the availability of 5-HT in the central nervous system [397]. However, it might exacerbate intestinal symptoms and inflammation due to an increase in the availability of 5-HT at the intestinal level. Therefore, this aspect should be taken into account when considering the use of fluvoxamine in this context.

Given the increased insulin resistance due to infection and increased blood insulin, treatment with metformin could be beneficial, as it decreases lipogenesis and gluconeogenesis, improves insulin sensitivity, inhibits complex I of the respiratory chain, reduces ROS production, improves antioxidant activity, and suppresses the production of proinflammatory cytokines by macrophages [398, 399]. Of particular interest is that it is able to downregulate NADPH oxidase [399], thus preventing the transformation of EBV-infected cells by inhibiting the intracellular ROS production and DNA damage necessary for generating latency. It would also reduce microclot formation due to endothelial damage caused by the positive regulation of NOX2 by EBNA-1 in EBV-transformed endothelial cells or by the activation of TLR3 by EBERs [256, 302, 316, 317]. Furthermore, metformin does not stimulate endogenous insulin production, so it is not associated with a risk of hypoglycemia [399]. The beneficial effect of metformin has been observed, for example, in patients with fibromyalgia, in cultured skeletal muscle cells from patients with ME/CFS, and even in reducing the severity of acute SARS-CoV-2 infection [400‐402].

The use of antioxidant supplements such as N-acetylcysteine (NAC), could also help to decrease oxidative stress, as they prevent the proapoptotic effect due to oxidative and nitrosative damage generated by increased tryptophan catabolites or increased NOX2 activity [327]. Another advantage of the use of antioxidants is the reduced risk of developing EBV-associated cancer [301]. NAC can inhibit NF-κB activation, thereby decreasing the resistance and survival of EBV latency cells [403, 404]. In addition, NAC suppresses the induction of Th17 cells and Tregs and promotes Th1 responses [327]. It even reduces glutamate levels in the CNS, thus reducing its neurotoxicity [405]. It has also been shown to be effective in reducing chronic inflammation and leukocyte recruitment to inflamed tissue caused by persistent EBV infection [404]. Other antioxidant supplements such as ubiquinol, Zn, vitamin E, vitamin C, and melatonin would also increase antioxidant defenses, preventing their depletion and reduce chronic oxidative stress due to infection [296, 297, 299‐301]. Specifically, Zn could also suppress the expression of proinflammatory cytokines and MHC-II on the surface of antigen-presenting cells, preventing them from being infected by EBV [329]. It would even increase blood zinc values in individuals with hypozincemia, restoring its uptake by immune cells and improving their activity [236]. It has been observed that zinc supplementation improves glucose metabolism and insulin sensitivity in patients with diabetes and could thereby also prevent its development [406].

This observation suggests that the use of zinc ionophores such as hydroxychloroquine could be useful, since they increase intracellular Zn, decrease the expression of MHC-II, and inhibit the replication of SARS-CoV-2. On the other hand, they promote the reactivation of EBV and other herpesviruses, which could contribute in the long term to the development of lymphomas [407, 408]. Another benefit of hydroxychloroquine is that it inhibits TLR3 activation and therefore also reduces the inflammatory response and development of coagulation disorders [308]. Among its major disadvantages is that it limits antigen-specific T-cell responses, reduces Th1 responses, and inhibits antioxidant production, thereby increasing oxidative stress [409, 410]. Therefore, the disadvantages of hydroxychloroquine use outweigh its advantages, preventing its use for relieving EBV-acquired immunodeficiency and potentially worsening the condition of patients with LC or ME/CFS [410].

Treatment with antivirals such as valganciclovir or valacyclovir could reduce lytic and abortive reactivation, inflammation, the appearance of transient autoantibodies, disease flares, and insulin resistance [244, 247].

There are contradictions in the use of hyperbaric oxygen therapy, since on the one hand the increased oxidative stress it generates allows for increased pathogen clearance, synthesis of various growth factors, and angiogenesis [411]. On the other hand, increased oxidative stress can generate higher levels of ROS and reactive nitrogen species, leading to increased oxidative and nitrosative damage, mitochondrial aging, DNA damage, and maintenance of chronic inflammation [411]. The benefit of this therapy may lie in the fact that generated hyperoxia stimulates the release of IκBα, which inhibits and reduces levels of NF-κB, thereby resulting in lower transcription rates of genes for proinflammatory cytokines (IFN-γ, TNF-α, IL-1β, IL-6, and IL-8), ultimately decreasing inflammation [411]. In addition, the lack of oxygen transport due to vascular damage and microclot obstruction that these patients present could be resolved by improving tissue oxygenation through a combination of hyperoxia and hyperbaric pressure [412, 413]. This could also prove beneficial by decreasing replication in those tissues with lower oxygen tension, such as the brain and some mucous membranes (oral and digestive), where conditions of relative hypoxia allow certain viruses, such as EBV, to enter the lytic phase and replicate and thus generate reservoirs and greater inflammation in these tissues [414‐416]. Therefore, this therapy could be useful for those viruses that do not generate latency, such as SARS-CoV-2, but could be detrimental for viruses that do, since it promotes the increase of latent cells by increasing oxidative stress.

Therefore, the symptoms of individuals with EBV-acquired immunodeficiency could be improved through the combined chronic use of antioxidant supplements, antivirals, and metformin, at least until a treatment is found that completely eliminates EBV latency cells. The use of anticoagulants to reduce thrombotic sequelae and improve the quality of life of patients with LC and ME/CFS could also be considered [78, 417].

EBV infection in individuals with “weak” EBV HLA-II haplotypes prevents the control of latency I cells (Fig. 1). The escape of these cells from immunosurveillance results in the formation of ectopic EBV-latent lymphoid structures in different tissues that promote inflammatory responses by releasing EBERs, producing abortive reactivation, and releasing new virions. Increased viral reactivation may increase the production of transient autoantibodies and the release of viral dUTPases. The increase in EBV latent cells also leads to an increase in evasion mechanisms by decreasing the activation and cytotoxic response of EBNA-1-specific CD4 T cells through IL-10 release, Treg recruitment, and the release of EBV miRNA contained in exosomes. Activation of TLR3 by EBERs or of TLR2 by EBV dUTPases results in the release of proinflammatory cytokines (IL-1β, IL-6, IL-8, IL-12, TNF-α, and IFN-γ) and IL-10. Hypozincemia could occur due to increased proinflammatory cytokines, as they increase metallothionein expression, causing a reduction in their availability in blood and their uptake by immune cells. The increase of metallothioneins in enterocytes reduces Cu uptake and ceruloplasmin synthesis. Decreased Cu transport results in decreased activity of Cu-dependent enzymes (CU/Zn-SOD, DAO, MAO, and cytochrome C oxidase). In addition, increases in proinflammatory cytokines leads to the formation of amyloid fibrin microclots resistant to fibrinolysis, either by increased production of serum amyloid A (SAA) or by the release of β-amyloid (Aβ) peptide by platelets, causing capillary blockage. High levels of IFN-γ produced by NK cells in response to persistent infection cause insulin resistance by downregulating insulin receptor transcription in myocytes. To compensate for insulin resistance, the pancreas increases insulin production, resulting in compensatory hyperinsulinemia. Hyperinsulinemia reduces glycogenolysis in the liver, leading to transient hypoglycemia and decreased metabolism in peripheral tissues, resulting in exercise intolerance. Hyperinsulinemia can decrease ACTH secretion and, together with chronic exposure to IL-10, TGF-β1 and TNFα could also decrease cortisol secretion. Activation of TRL2 and TLR3 also causes inhibition and decreased expression of intestinal SERTs, resulting in extracellular accumulation of 5-HT and activation of 5-HT receptors, generating increased motility, nausea, and intestinal secretions. Intestinal serotonergic alteration together with the increase in proinflammatory cytokines also generates chronic inflammation, resulting in decreased acid secretion, breakdown of the intestinal barrier, decreased expression of enzymes necessary to digest carbohydrates, and alteration of the intestinal adaptive immune system, leading to the appearance of food intolerances and bacterial overgrowth in the small intestine. The occurrence of SIBO results in malabsorption of fat-soluble vitamins and vitamin B12. When activated via 5-HT2A receptors by increased 5-HT in peripheral blood or by binding of EBERs to TLR3, platelets bind fibrinogen, leading to increased aggregation and coagulation processes. Activation of TLR3 receptors in microglia causes increased release of IL-1β and TNF-α, which increases SERT expression in astrocytes and, therefore, 5-HT reuptake and its degradation to the metabolite 5-HIAA, thus reducing extracellular 5-HT levels. The decrease in 5-HT reduces the activation of neuronal 5-HT1A receptors, possibly inducing the sensation of fatigue. Increases in both proinflammatory cytokines and oxidative and nitrosative stress by activation of microglia via TLR3 could lead to increased IDO activity in microglia, resulting in reduced tryptophan levels, increased kynurenine catabolites, and decreased 5-HT and melatonin synthesis. Quinolinic acid (kynurenine metabolites) is an NAD+ precursor that has neurotoxic properties through binding to the N-methyl-d-aspartate (NMDA) receptor, followed by sustained Ca²⁺ influx leading to increased nitrosative stress. The increased oxidative and nitrosative stress generated by all these processes could be exploited by EBV-infected cells for their transformation through viral DNA methylation, preventing their replication and favoring the establishment of different types of latency that allow evasion of the immune system.