Is Epstein Barr virus latency protective against type 1 diabetes?

Adaptation, fitness, and survival are very sensitive to between-individual variability in immune response, which is a cumulative result of both genetic and environmental factors. Host-pathogen interactions during acute or chronic infections reset the levels and activation state of the immune system and differences in immune response have been observed in immune-related disorders including infection, inflammation, autoimmunity, cancer and immune response to immunotherapy. The pattern of immune response correlate with risk to autoimmunity, which could partially explain the etiopathology of the disease. However, the importance of the correlation in clinical practice has remained unclear.

A detrimental impact of viral infection on autoimmune disease has now been documented for multiple sclerosis. Strong evidence implicates EBV in its etiopathology [1]. In particular, a cross-reactivity between antibodies for Epstein-Barr nuclear antigen 1 and glial cell adhesion molecule has been found in patients with multiple sclerosis [2]. The molecular mimicry of the latent viral protein EBV nuclear antigen-1 has also been described as a driver of autoimmunity in lupus erythematosus [3]. In the overall context of autoimmunity, it has been shown that EBNA2 is associated with transcription factors that occupy substantial part of autoimmune risk loci. prominently including the risk of type 1 diabetes and suggesting indirectly EBV implication as well as other disorders [4].

It is well known that the infection of latent, non cell-lytic viruses (EBV, CMV, HIV) have a long lasting effect on the setting of immune system response. Herpesviriadae infection (CMV, HIV, EBV) change the level of immune activity in tumor microenvironment, organ transplantation and as well as autoimmunity. At the same point, the sub-clinical inflammation is characteristic of chronic inflammatory and autoimmune diseases, and the increase in exhausted cytotoxic CD8 + CD28- T cells that has been reported in chronic inflammatory disease and senescence.

Concerning the level of CD8 + T cells in greater detail, when they are exposed to antigens such as viral or tumor components, they, proliferate and expand but are then constrained to the pool of memory cells. However, certain infections cause long persistent expansion of the CD8 + T cell pool, contributing to different basal thresholds of activation of immune system. For instance, after infection with Herpesviriadae, through a mechanism known as CD8 + T cell inflation, expanded T cell populations, that expanded in response for some epitopes do not contract but accumulate as effector memory cells (CD8dimCD28- and long lived CD8bright EM T cells) [5]. Thus, when antigen persists such as in chronic viral infection, CD8 + T cells undergo loss of functions and develop an “exhausted” phenotype as an evolutionary mechanism to prevent immunopathology.

Conceivably, the inter-individual differences observed in response to the anti-CD3 monoclonal antibody Teplizumab (approved for delay of type 1 diabetes). Teplizumab positive effects have been linked to the development of partially-exhausted CD8 + T cells expressing the transcription factor EOMES, KLRG1 and TIGIT [6,7]. The effect of EBV seropositivity on disease status has been established in a clinical cohort of children with type 1 diabetes. In a clinical trial (TN10, AbATE) the most robust response to the Teplizumab was observed in EBV - sero-positive compared to seronegative children. The result has been attributed to reduced antigen receptor signaling pathways in T and B cells. Teplizumab also induces partial exhaustion of CD8 + T cells, which has been confirmed in expression analysis showing that genes related to exhaustion pathways and inhibition of the immune system are seen in the EBV seropositive individuals. Additionally, the genes related to the differentiation of regulatory T cells were upregulated whereas B cells in EBV negative patients have shown increased differentiation into B memory cells and plasmablasts 18 months after therapy [8].

In another recent study, activated islet specific CD8 + memory T cells were prevalent in type 1 diabetes patients who progress rapidly. By contrast, slow disease progression was associated with an exhaustion-like pattern of T cells, with expression of diverse inhibitory receptors, lower cytokine production and reduced proliferation [9].

In accord with those results, we have discovered that similar to the effect of latent viruses on human immune system, the PRF1:p.A91V mutation contributes to the regulation of distinct immune subtypes, notably augmenting the memory cytotoxic T cell compartment: CD8dim, CD8dimCD28-, TCRgammadelta and effector-memory CD8bright T cell subpopulations expansion. Intriguingly, similar the effect of some pathogens, this mutation resets basal immune status, possibly predisposing to persistent subclinical, asymptomatic inflammation observed at a population level, which in turn contributes to multiple sclerosis but conversely protects against type 1 diabetes [10].

In further agreement with these findings, anti-PD-1 PD-L1 antibodies used in cancer therapies, potentiate exhausted CD8 + T cells in nonobese diabetic (NOD) mice (autoimmune diabetes model) and favor rapid onset of autoimmune diabetes in human cancer studies. Moreover, pancreatic inflammation induced by certain viral infections is also protective in mouse models of type 1 diabetes, via expansion of antiviral cytotoxic T lymphocytes involved in insulitis [11,12]. Also, consistent with these findings is the protection conferred by human cytomegalo-virus infection against type 1 diabetes in a small pediatric cohort [13].

Intriguingly, the lower levels of antibodies anti EBV in type 1 diabetes children have been noted in the past, although the interpretation of the finding was not clear [14].

In general, it has also been also demonstrated that a transcriptional signature mirroring CD8 + T cell exhaustion is associated with poor clearance of chronic viral infection but predict better prognosis in various autoimmune and inflammatory disorders (type 1 diabetes, vasculitis, systemic lupus erythematosus, idiopathic pulmonary fibrosis, dengue haemorrhagic fever). A favorable clinical outcome and fewer relapses have been observed for various autoimmune disorders [15]. Therefore, the induction of exhaustion has been proposed as treatment for immune-mediated diseases.

Thus, even if there are many questions to answer, the accumulating evidence suggests that the latent EBV – distinct from its establish impact on tumorigenesis - could have an impact on clinical outcome, in inflammation, autoimmunity and treatment with immune-modulating therapies. The mechanism could be extrapolated to therapeutic approaches in immune-mediated disorders, though its impact could have opposite direction in different autoimmune disorders.

The impact of EBV could have significant information in type 1 diabetes considering the age of onset in that disease. The effect in response to the EBV would however, also be affected for individuals by how long the response persists and how may be changed by infection with other latent-non cell-lytic viruses and possible reactivation status. The interactions are likely to be quite complex, given that fulminant diabetes has been thought to be associated between others also with EBV as a trigger contributing to the disease!

Notably, T cell exhaustion can be induced not only by EBV but also by chronic or latent infections of fungal, bacterial, protozoan and viral pathogens. This suggests that the effects of the outcome of the therapy with teplizumab cannot be explained entirely by EBV status. In fact, a recent study showed that anticommensal antibodies to particular gut microbiota could delay the type 1 diabetes [16].

These findings point to the importance acknowledging pathogen effects on immune system education and exposures during the development of type 1 diabetes. Such considerations rationalize the observations that NOD mice have a higher incidence of autoimmune diabetes when housed in sterile condition. On the other hand, for multiple sclerosis, a possible mechanism differs, because germfree EAE mice develop attenuated disease and exposure during life has been pathognomonic.

Taking together, the literature indicates that the effect of EBV on the host could be positive or negative associated depending on latency, which remains to be investigated further.