Pathogen infection and autoimmune disease

For many readers, the title ‘pathogen infection and autoimmune disease’ might suggest that pathogens are the suspected drivers of autoimmune disease. This, of course, is a perfectly reasonable assumption, as pathogens of all kinds, including viruses, bacteria and parasites, can cause direct damage to organs and tissues and induce a strong immune response that aims at a fast elimination of the foreign intruder. Pathogen infections induce several inflammatory cascades that result in the attraction, differentiation and expansion of cells of the innate as well as the adaptive immune system. Further, mechanisms such as bystander activation, pathogen‐induced necroptosis, superantigen cross‐linking and molecular mimicry have been reported to be involved in the breakdown of self‐tolerance (Fig. 1) 1, 2, 3, 4, 5, 6.

Figure 1.

Impact of pathogens on autoimmunity. Several mechanisms have been suggested how pathogens might be involved in the initiation or acceleration of autoimmune diseases (left) or how they might decelerate an ongoing autoimmune process or even protect from autoimmune disease (right). This figure provides a schematic overview of some of them without claiming to be complete. Initiation/acceleration: pathogens can cause direct damage to infected cells (1) leading to necrosis or initiate a necroptosis programme in infected cells (2). The infection‐associated release of a plethora of inflammatory mediators may trigger bystander activation (3), including the up‐regulation of major histocompatibility complex (MHC) molecules, resulting in an enhanced presentation of pathogen and host peptides. Pathogen‐derived superantigens, such as staphylococcal enterotoxins, may cross‐link MHC molecules directly with the T cell receptor (TCR) without requiring a specific peptide (4), resulting in a polyclonal T cell activation. Molecular mimicry between pathogen and host (5) might activate pathogen‐specific lymphocytes that also cross‐react to self‐antigens. Both direct damage and determinant epitope spreading might result in the exposure of cryptic or previously sequestered epitopes (6) to which no tolerance had been established. Protection/deceleration: pathogens, particularly parasites, can induce an immune balance shift towards a type 2 (Th2/Tc2) response (7). Immunomodulatory cytokines and other suppressive factors may induce tolerogenic dendritic cells and/or bystander regulatory T cells (8). Pathogen infection at a remote location might cause a deviation (9) for aggressive cells that might follow a stronger chemokine gradient leading away from the site of autoimmune destruction. Finally, pathogens might induce target autoantigen‐specific regulatory T cells. Note that additional factors, including genetic predisposition, microbiome and others, impact the outcome of the autoimmune process.

Conversely, however, pathogens might also exhibit a protective role. The hygiene hypothesis states that infection with pathogens might be one of the reasons why regions with a low hygienic standard have lower incidences of autoimmune diseases and allergies 7, 8. Countries with a lower gross domestic product per capita often display lower hygienic standards and individuals living in such countries are more likely to suffer from tuberculosis or hepatitis A. In addition, the risk of contracting traveller’s diarrhoea is high in low‐income countries 8. Several mechanisms have been suggested to explain how such a protective role might be conveyed (Fig. 1). In particular, infection with parasites has formed the basis of the hygiene hypothesis, as parasites predominantly induce a type 2 (Th2/Tc2) immune response that suppresses the type 1 (Th1/Tc1) immune response that, together with a type 17 (Th17/Tc17) response, is considered a driving force in the development of autoimmunity 9.

Many of these suggested mechanisms have been identified in animal models, in which mostly transgenic mice that are prone to develop a certain autoimmune disease are infected by pathogens 8. In this way, pathogens have been demonstrated to suppress the autoaggressive immune response by cytokine‐mediated bystander suppression, expansion of regulatory dendritic cells and induction of antigen‐specific regulatory T cells 10, 11. Another mechanism is immune deviation, in which a pathogen infection at a remote site might lure autoaggressive lymphocytes away from the site of autoimmune destruction 12.

However, one has to consider the additional factors which contribute to the observed regional differences in the prevalence of autoimmune diseases. Many areas with a lower hygienic standard are located close to the equator and therefore have a higher exposure to sunlight. As a result, individuals living in these regions generate higher levels of vitamin D, which is believed to exhibit a protective effect 13. Conversely, there are examples of regions located on the same latitude, such as Finland and neighbouring Russian Karela, that display a dramatically different prevalence of type 1 diabetes (T1D) 14. Another example is the much higher prevalence of T1D and multiple sclerosis (MS) in Sardinia compared to the surrounding southern European populations, such as mainland Italy or Corsica 15. As these regions exhibit a similar hygienic standard and are on a similar latitude, other factors such as genetic predisposition seem to have a tremendous effect. Interestingly, migration studies have revealed that the prevalence for MS is decided before the age of 15 years, as individuals who move from high‐ to low‐risk regions after the age of 15 remain at high risk of developing MS. In contrast, when they move before the age of 15, they acquire the low risk of the surrounding region 5.

Indeed, for almost all autoimmune‐related diseases, genetic risk factors have been identified and thoroughly investigated, indicating that the human leucocyte antigen (HLA)‐haplotype has by far the highest influence on the development of autoimmunity. However, the inception of genomewide association studies (GWAS) has facilitated the identification and unbiased analysis of many additional risk factors 16. Metagenomics is another recent advance in technology, and has allowed for a detailed analysis of the composition of the microbiome. This has revealed an association of several autoimmune diseases with a certain degree of dysbiosis 8. Such a reduction in the diversity of the microbiome has been further followed‐up in a very intriguing longitudinal study with infants who were genetically predisposed to T1D 17. During the study, stool samples were collected from a cohort of 33 newborns from Finland and Estonia who had been selected upon an HLA risk genotyping during a period of 3 years. Interestingly, a dysbiosis has been observed in those infants who had started to generate antibodies to β cell antigens. Importantly, this shift to a decreased microbiome diversity occurred before the progression to disease but after seroconversion, and was restricted to those seroconverters that did, indeed, progress to T1D later 17.

In this special issue of Clinical & Experimental Immunology, we have assembled review articles that reflect on the current evidence for the involvement of pathogens in the aetiology and/or pathogenesis of autoimmune diseases. Although many associations between pathogen infections and the occurrence of autoimmune diseases have been reported, firm proof of a detrimental or a protective effect is rare. In order to provide such proof, the following aspects would have to be fulfilled: (1) statistical evidence for an association between pathogen infection and disease; (2) identification of the pathogen in patients with the disease; and (3) presence of pathogen‐specific antibodies and/or T cells. In case of a possible molecular mimicry, such antibodies and/or T cells should be cross‐reactive with endogenous components. (4) Reproduction of the events in (an) animal model(s), in which an infection of model animals should result in a similar autoimmune disease. As if fulfilling these aspects would not be difficult enough to begin with, the situation is further complicated by the possibility that more than one pathogen might be involved. We all have been and will be infected by a multitude of pathogens throughout our lifetime, some of which might be involved in the initiation and/or acceleration of a disease, others which might display protective properties, and others again which might have no impact at all. Here, we will present articles that report on the pursuit of possible suspects.

First, Teresa Rodriguez‐Calvo opens the investigation for T1D 18, with a focus on how enteroviruses (EV) might be involved in the pathogenesis of T1D. Epidemiologically, EV have been associated with this disease for a long time 19. However, the consequences of such infections for T1D have since been debated. Recently, more and more studies report the presence of EV components in the islets of Langerhans of patients with T1D. In particular, the analysis of patient tissue samples from the Network for Pancreatic Organ Donors with Diabetes (nPOD) contributed to the current opinion that EV might indeed play a role in the pathogenesis of T1D, albeit perhaps not alone.

Primary biliary cholangitis (PBC) is an autoimmune liver disease that has been associated with pathogens for quite some time. Tanaka, Leung and Gershwin review how molecular mimicry between bacteria such as Escherichia coli and Novospingonium aromaticavorans, and the major group of mitochondrial autoantigens (2‐oxoacid dehydrogenase complexes), is involved in the development of PBC 20. They further report on possible reasons for the recently observed shift in the male to female ratio in PBC patients. Compared to a few decades ago, the male to female ratio has increased from 1 : 10 to 1 : 6, and is as high as 1 : 3 in Lombardy, Italy. Because the xenobiotic 2‐octynoic acid, which is present in cosmetics and some food additives, confers molecular mimicry to the 2‐oxoacid dehydrogenase complexes, similar to components/proteins of E. coli and N. aromaticavorans, an increase in xenobiotics consumption might explain the changes in gender distribution.

Much less is known about the aetiology of another autoimmune liver disease; namely, autoimmune hepatitis (AIH). Although the liver is the target of many prominent pathogen infections, including several hepatitis viruses, there has been no firm proof that such viruses might also lead to autoimmune‐mediated liver damage. Christen and Hintermann provide insight into the current evidence for the involvement of pathogens in the pathogenesis of AIH and discuss the problems associated with obtaining firm proof 21. Further, they hypothesize about the consequences of the very successful modern therapies for hepatitis C on the development of AIH. If hepatitis C virus (HCV) were involved in the aetiology of AIH, for example by breaking tolerance to hepatic structures via molecular mimicry, the chronic presence of HCV components might prevent a reaction to similar endogenous structures in the liver. However, upon total elimination of HCV such an immune deviation might stop, and in response AIH could possibly slowly develop in the cured HCV patient.

Addison’s disease (ADD) is an autoimmune disease that affects the adrenal cortex resulting in the lack of steroid hormone production. Hellesen and Bratland 22 review the current evidence for environmental factors to be involved in ADD. Although a major focus of ADD research has been the identification of genetic risk factors, there are also some indications that viruses, such as Epstein–Barr virus (EBV) or HCV, might be involved in its aetiology. Interestingly, many of the identified risk factors encode for proteins that are involved in a normal anti‐viral response, such as the plasma cell differentiation factor BACH2, nuclear factor of activated T cells, cytoplasmic 1 (NFATC1) and the class II, major histocompatibility complex transactivator (CIITA). In addition, they report on the involvement of the chemokine C‐X‐C motif chemokine 10 (CXCL10), which is prominently up‐regulated during acute and chronic viral infections and has been demonstrated to be an important chemotactic factor attracting autoaggressive T cells to the adrenal cortex in ADD.

Lang et al. look at the role of viruses in the pathogenesis of autoimmunity from a mechanistic view 23. They found that some subtypes of antigen‐presenting cells, in particular CD169⁺ macrophages, exhibit a mechanism of enforced viral replication in order to establish an enhanced anti‐viral defence mode. In CD169⁺ macrophages the ubiquitin‐specific protease 18 (Usp18) has been identified as a key player that inhibits the responsiveness to type I interferons (IFNs) upon virus infection. This results in an enforced viral replication in these cells, a promotion of the CD169⁺ macrophage development and subsequently a superior activation of innate and adaptive immune responses. Importantly, they have found in a virus‐induced mouse model that Usp18 has an influence on the development of T1D.

Finally, the review article by De Luca and Shoenfeld brings the microbiome into the discussion 24. They reflect upon the role of the microbiome on the maintenance homeostasis of the immune system and the alterations in the microbiome that have been observed in patients and experimental animals with autoimmune diseases. Further, they discuss modern approaches to re‐establish a healthy microbiome, such as probiotic diet and faecal transplantation, and their possible impact on the progress of autoimmunity.

The reviews gathered in this special issue of Clinical & Experimental Immunology report current evidence for pathogens to be involved in the pathogenesis of autoimmunity. They draw a picture that is very similar for most autoimmune diseases. First, there is epidemiological evidence for pathogens to be involved in either the initiation and acceleration, or the abrogation and deceleration of autoimmune processes. Secondly, pathogens have been identified at the site of autoimmune destruction. Thirdly, many mechanisms of how such pathogens impact upon autoimmunity have been identified in animal models. In the coming months, we hope to flesh this out further with updates on how viruses, such as EBV, might be involved in the development of MS.

However, for most autoimmune diseases there is still a lack of evidence that infection with a specific pathogen might be decisive in determining the outcome. There are several reasons why there is still no firm proof. On one hand, the involvement of many additional factors such genetic susceptibility, diet, exposure to chemicals, microbiome diversity, stress, sunlight exposure and many more has been demonstrated. On the other hand, it is very likely that not a single pathogen infection, but rather the sum of all pathogen infections acquired throughout a lifetime, determine which and how many autoimmune diseases affect a given individual. Thereby, some pathogens might have an accelerating and some a decelerating effect 25. Thus, large prospective studies such as ‘The Environmental Determinants of Diabetes in the Young’ (TEDDY) and the ‘Diabetes Auto Immunity Study in the Young’ (DAISY) that run for several decades are of growing importance in order to identify and characterize the interplay of environmental factors that are involved in the pathogenesis of autoimmune diseases.