Published on August 27, 2010
Author: Didier Hober
Institution: Laboratory of Virology/EA3610, University Lille 2, Faculty of Medecine, CHRU Lille Institut Hippocrate
Address: Loos-Lez-Lille, 59120, France
Abstract: Type 1 diabetes (T1D) is an autoimmune disease characterized by the disturbance of pancreatic insulin-producing cells, which results in hyperglycemia. The disease is associated with severe complications that impair the quality of life of individuals. The cause of T1D is unknown. Development of the disease is the result of interactions between immunological, genetic, and environmental factors. Viruses are thought to play an important role in the initiation or acceleration of the disease. This is an important issue since it opens the possibility to develop new preventive and therapeutic strategies to fight the disease. The role of enteroviruses in the development of T1D, in particular type B coxsackieviruses, is supported by epidemiological observations. It has been demonstrated that enterovirus infections were significantly more common in recently diagnosed diabetic patients, compared with control subjects. Enteroviral RNA and/or proteins can be detected in blood samples and intestine biopsies of patients with T1D. The hypothesis of a relationship between enteroviruses and the disease has been strengthened by the presence of enteroviral components or infectious particles in the pancreas of patients with T1D. In this review, arguments in favor of a relationship between enterovirus infections and T1D and the mechanisms of the enteroviral pathogenesis of the disease are presented.
Type 1 diabetes (T1D), also termed “insulin-dependent diabetes mellitus” (IDDM) or “juvenile diabetes,” is an autoimmune disease that represents 5-10% of all cases of diabetes, a chronic disease with severe complications that impair the quality of life of individuals and have huge repercussions in term of health expenditure.
In T1D, the pancreatic islet β-cells are selectively and massively destroyed or impaired, which results in a defect in insulin production that is characteristic of the disease. Before the onset of overt clinical manifestations of diabetes, autoantibodies circulate in the peripheral blood of individuals. These autoantibodies target islet β-cell autoantigens [insulin autoantibodies (IAA)], glutamate decarboxylase (GAD65 antibodies), and tyrosine phosphatase [insulinoma-associated (IA-2) autoantibodies].
T1D incidence peaks at 2, 4-6, and 10-14 years and it generally occurs before the age of 40 years. An increasing incidence rate has been observed for the last few decades especially in young individuals (less than five years old). The global annual incidence varies from 1 per 100,000 in Asia to 14 per 100,000 in USA and more than 30 per 100,000 in Scandinavia (Jaidane and Hober, 2008).
The cause of T1D is still unknown. Several factors interact and lead to the development of the disease (Figure 1).
Insulitis, a pancreatic inflammation, is present at the symptomatic onset of T1D, and reflects the immune response to β-cells (Eizirik et al., 2009). T1D is an autoimmune disease, resulting from the failure of peripheral tolerance due to abnormalities of regulatory T lymphocytes (Treg) in the periphery and/or of central tolerance involving the thymus (Rosmalen et al., 2002). Interestingly, anomalies of T lymphocyte subpopulations were described in diabetic and prediabetic patients, suggesting a role for alterations in the T lymphocyte repertoire in T1D genesis (Jaidane et al., 2010). In the thymus, GAD and proteins of the insulin family, such as type 2 insulin-like growth factor (IGF-2), are presented to thymocytes. A low level of thymus expression of insulin, which is influenced by a polymorphism in the insulin gene, is associated with a susceptibility to the disease. Among animal models of diabetes, non-obese diabetic (NOD) mice and bio-breeding diabetes-prone rats do not express insulin precursor (proinsulin 2) or IGF-2 in their thymus (Geenen et al., 2010).
Susceptibility to T1D is influenced by genetic factors. Genes located in the human leukocyte antigen (HLA) class II locus (especially DQ and DR) on chromosome 6, and more than 20 other loci, contribute to T1D. A potent locus for T1D is the insulin gene. A low level of insulin gene expression in the thymus and peripheral lymphoid tissues may disturb the tolerance to this self antigen. In addition, some of the loci are associated with antiviral activities (Hober and Sauter, 2010). Moreover, epigenetic modifications (such as DNA methylation) and parent-of-origin effects may play a role (Beyan et al., 2010).
Within the human genome there are endogenous retrovirus genes that can be activated by environmental factors or hormones. Interestingly a human endogenous retrovirus, termed IDDMK1,2-22, is expressed and released from leukocytes in T1D patients (Conrad et al., 1997). Whether such an endogenous retrovirus can play a role in the pathogenesis of the disease is an intriguing question.
The rapidly increased incidence of T1D in most populations cannot be explained only by genetic modifications in the population.
Differences in incidence of the disease in different parts of the world have been observed, even though the genetic profiles of neighboring populations are similar (Kondrashova et al., 2005). In addition the incidence varies from one season to another and there is a relationship between immigration and disease development.
Therefore it is strongly suspected that the disease is a result of a complex interaction of genetic and environmental factors, such as drugs, toxins, nutriments (e.g., cow’s milk), and, more recently, viruses (Jaidane and Hober, 2008).
The role of several viruses such as rotaviruses, adenoviruses, retroviruses, reoviruses, cytomegalovirus, Epstein-Barr virus, mumps virus, or rubella virus in the pathogenesis of T1D was invoked (Jaidane et al., 2009). However, enteroviruses, especially coxsackievirus B, are among the viruses most able to be involved in the pathogenesis of the disease (Hober and Sauter, 2010).
Type 1 Diabetes and Enterovirus
What is an enterovirus?
Coxsackievirus B (CV-B), coxsackievirus A (CV-A), poliovirus (PV), and other viruses such as enterovirus 69 (EV69), EV70, EV71, etc. belong to the Enterovirus genus of the Picornaviridae family. Based on genome sequence homologies, there are five distinguished species of Enteroviruses: PV, human enterovirus A (HEV-A), HEV-B including the six CV-B serotypes and the Echoviruses, HEV-C, and HEV-D (Stanway et al., 2005). The Enterovirus genus currently encompasses more than 100 types, they are small viruses (around 30 nm diameter), non-enveloped, shaped as an icosahedron, and they have a genome that is a single-stranded, linear, non-segmented positive RNA (infectious and serving as a messenger RNA) made of about 7,500 bases.
This RNA encodes a polyprotein (2,200 amino acids), which, through successive cleavages, gives 11 mature proteins: four are structural (VP1, VP2, VP3, and VP4) and form the capsid, and seven are functional (involved in the viral replication in the host cell). The coding region is flanked by two non-coding regions needed for translation initiation of the enteroviral RNA and for viral RNA synthesis (Jaidane et al., 2010). The enteroviruses enter into cells that harbor specific receptors, they subsequently replicate, which results in the production and release of new viral particles able to infect other cells (Hober and Sauter, 2010) (Figure 2).
Enteroviruses have a large distribution in the world. They are mainly transmitted by the fecal-oral route via the ingestion of contaminated water or food.
Most enterovirus infections are asymptomatic whereas some enterovirus infections, especially the one due to CV-B, have been associated with acute manifestations — non-specific febrile disease combined with gastrointestinal, respiratory, or cutaneous symptoms, meningitis, encephalitis, pericarditis, etc. In addition they are involved in chronic diseases, meningoencephalitis (in individuals with an impaired production of gamma globulins), chronic myocarditis, and dilated cardiomyopathy, and their role in T1D is strongly suspected (Jaidane and Hober, 2008).
Is there a relationship between enteroviruses and type 1 diabetes?
• Early results suggest a possible link
It was observed in several studies that there was a temporal relationship between diabetes onset (late fall/early winter) and the peaks of enterovirus infections (late summer/early fall) (Green and Patterson, 2001). For the first time in 1969 researchers noted that anti-coxsackievirus antibodies were found more frequently in patients with T1D than in control subjects (Gamble et al., 1969). In 1979, a CV-B4 was isolated from the pancreas of a patient who had died of diabetic ketoacidosis (Yoon et al., 1979). That viral strain, when transferred to susceptible mice, led to hyperglycemia with inflammation of the pancreas Langerhans islets and β-cell necrosis, and provoked the appearance of autoantibodies (anti-GAD antibodies) and hyperglycemia six weeks post-infection (Yoon et al., 1979; Gerling et al., 1991). Based on these early results a possible link between enteroviruses, CV-B4 in particular, and T1D was suggested. More recent investigations explored this hypothesis.
• Retrospective and prospective recent investigations
In retrospective studies, anti-enterovirus antibodies are found more frequently in recently diagnosed diabetic patients than in healthy controls (Helfand et al., 1995). A strong argument came from the detection of enterovirus RNA in the peripheral blood of recently diagnosed diabetics, which was more frequent than in healthy controls (Andreoletti et al., 1997; Hyoty 2002; Yin et al., 2002; Craig et al., 2003; Sarmiento et al., 2007; Schulte et al., 2010). The most frequently implicated enteroviruses were CV-B, especially CV-B4 (Nairn et al., 1999; Chehadeh et al., 2000; Yin et al., 2002; Kawashima et al., 2004). Our group showed that interferon-alpha (IFN-α), a marker of viral infection, was circulating in peripheral blood of 75% of T1D patients in association with CV-B infections (Chehadeh et al., 2000).
Enteroviruses, especially their RNA, are present in blood of patients and in other tissues; indeed their proteins can be detected in small intestine biopsies (Oikarinen et al., 2008).
Interestingly, enteroviruses and/or their components have been detected in the target organ of the disease (Taurienen et al., 2010). Indeed, enterovirus RNA was found in autopsy pancreas samples from T1D patients (Ylipaasto et al., 2005); enteroviral particles, especially CV-B4 and enterovirus capsid protein (VP1), were present in pancreatic β-cells from patients, but not from control donors (Dotta et al., 2007). In a study encompassing 72 young recent-onset type 1 diabetic patients, the enterovirus capsid protein (VP1) was detected in autopsy pancreases of most of them (44 out of 72) (Richardson et al., 2009). The presence of a virus in pancreatic islet β-cells might be responsible for the expression of IFN-α observed in β-cells of patients with T1D (Foulis et al., 1987; Huang et al., 1995) (Figure 3).
Prenatal and perinatal exposure to enterovirus may have a role in the development of T1D. Exposure to enteroviruses, especially CV-B, during intrauterine life is associated with increased risk of offspring developing T1D during childhood (Dahlquist et al., 1995b; Dahlquist et al., 1999; Elfving et al., 2008). A temporal relationship between viral infections and the initiation of autoimmune processes corresponding to the appearance of autoantibodies has been studied (Hyoty et al., 2002) (DiMe Study). The follow-up of genetically predisposed children with T1D (according to their genetic profile) or of diabetic children’s siblings, showed that enterovirus infections were more prevalent in children who became positive for β-cell autoantibodies than in healthy controls (Lönnrot et al., 2000; Sadeharju et al., 2003). Mothers whose children developed T1D before age 15 showed an elevated number of enterovirus infections during pregnancy, compared to controls (Dahlquist et al., 1999).
Are Enteroviruses Able to Play a Role in the Pathogenesis of Type 1 Diabetes?
Enteroviruses infect the pancreas of patients; in addition, the infection of other tissues in human beings has been reported as well. The detection of enteroviruses in various tissues, blood, gut, and pancreas of patients with T1D suggests an association between these viruses and the disease (Taurienen et al., 2010).
How can enteroviruses play a role in the pathogenesis of type 1 diabetes? Various mechanisms, not mutually exclusive, through which enteroviruses can play a role in the development of T1D will be presented below (for detailed description see Hober and Sauter, 2010; Jaidane et al., 2010; Taurienen et al., 2010) (Figure 4).
Enteroviruses have a tropism for pancreas and β-cells
CV-B and other enterovirus serotypes can infect human islets and can replicate in β-cells in vitro, which stimulates the production of cytokines (soluble mediators of the “communication” between immune cells) and membrane proteins at the surface of β-cells that are able to intervene in pathogenic processes. CV-B4 inoculated to mice resulted in diabetes with viral replication in β-cells.
In our experiments, CV-B replicated in human β-cells in vitro. An expression of IFN-α by β-cells was observed (Chehadeh et al., 2000). IFN-α can play a pathogenic role through induction of class I HLA molecules and ICAM-1 at the surface of β-cells, which is a characteristic of β-cells of patients with T1D. It is interesting to note that transgenic mice whose β-cells express IFN-α develop diabetes as a result of insulitis and β-cell destruction, due to activation of autoimmune effector cells against islets (Stewart et al., 1993). Thus IFN-α, produced by enterovirus-infected cells, can initiate autoimmunity towards β-cells.
Persistence of enteroviruses
CV-B, such as CV-B4, can establish persistent infections in vitro, especially of human pancreatic β-cells, as demonstrated in our experiments (Chehadeh et al., 2000). In vivo, in a mouse model, the persistence of CV-B4 RNA up to 6 months post injection in islets was associated with β-cell destruction and diabetes onset.
How can a persistent infection of islets induce β-cell impairment? There are several possible mechanisms supported by experimental data. The infection with CV-B4 can disturb the release of insulin in response to high glucose levels. A persistent infection can result in the induction of an autoimmune response directed towards β-cells through presentation of viral antigens and self antigens to T lymphocytes and activation of the immune system (Hyoty, 2002).
Enterovirus infections, and re-infections can initiate or accelerate β-cell impairment through the activation of anti-enteroviral T lymphocytes which, by cross-reactivity, participate in aggravation of the impairment of β-cells persistently infected by another enterovirus serotype (Hyoty, 2002).
The local inflammatory reaction which accompanies the persistent viral replication can also take part in recruitment of autoreactive cytotoxic T lymphocytes (CTL) in the islets.
The infection of β-cells causes an inflammation which provokes cell damages and release of sequestered antigens. Through a bystander activation, the local inflammation results in recruitment of cells, encompassing autoreactive T lymphocytes directed against these self-antigens, and activation of them (Horwitz et al., 2002).
Enteroviruses and molecular mimicry
A partial sequence homology (i.e., molecular mimicry) exists between enteroviral proteins and β-cell self-antigens: the 2C viral protease (one of seven non-structural proteins of enteroviruses) and the GAD65; the VP1 viral capsid protein (and the VP0 precursor) and IAR/IA-2 protein on one hand and the HSP60 protein on the other hand. In a mouse model designed for testing the molecular mimicry hypothesis between P2C and GAD65, the infection with CV-B4 did not accelerate diabetes. However serum collected from patients with an enterovirus infection can react with IA-2/IAR (Harkonen et al., 2003).
Antibody-mediated enhancement of enterovirus infection
Our group discovered the antibody-dependent enhancement of CV-B (especially CV-B4) infection in the human system, which is mediated by anti-CV-B4 antibodies devoid of neutralizing activity (Chehadeh et al., 2001). These antibodies (IgG) are able to increase the replication of CV-B4 in monocytes/macrophages, and hence the CV-B4-induced IFN-α synthesis by these cells (Hober et al., 2001).
The enhancing effect was higher with plasma or IgG from patients with T1D than from healthy individuals and it was higher with peripheral blood cells from patients, due to IgG bound to their cells (Hober et al., 2002).
The target protein of antibodies that increased the CV-B4-induced IFN-α production by PBMCs and CV-B4 infection of PBMCs was identified as the capsid protein VP4, which indicates that this protein is accessible to antibodies at physiological temperatures (Chehadeh et al., 2005; Sauter et al., 2007).
A higher prevalence and level of anti-VP4 antibodies was found in patients with T1D than in healthy subjects (Sauter et al., 2008).
The role of antibody-dependent enhancement in the pathogenesis of enterovirus-caused disease has been shown in animal models where such antibodies can multiply enterovirus viremia in blood and the viral load in the target organs thereby accentuating the histopathologic lesions (Sauter and Hober, 2009). The antibody-dependent enhancement of CV-B4 infection of circulating blood cells could play a role in virus dissemination in the host, and consequently in the pathogenesis of diseases, such as T1D, induced by this virus.
Enteroviruses and thymus
T1D is an autoimmune disease that is the result of a defect in tolerance towards β-cell antigens at the peripheral level through anomalies of regulatory T lymphocytes (Treg) and/or at the central level through disturbance of thymus.
It has been hypothesized that the infection of thymus by an enterovirus could disturb thymic function and play a role in the pathogenesis of T1D. In the human system in vitro, CV-B4 replicates and persists in human thymic epithelial cells and infect human fetal thymus fragments, resulting in disturbance of T lymphocyte maturation (Brilot et al., 2004). In the mouse system, CV-B4 can infect the thymus in vivo and, in vitro, it replicates in thymic cell primary cultures and in fetal thymus organ cultures, resulting in a disturbance of the maturation/differentiation of T lymphocytes (Jaidane et al., 2006; Brilot et al., 2008).
The possibility that virus-induced disturbances of thymus play a role in the pathogenesis of T1D cannot be excluded.
Interactions Between Enteroviruses and the Host Play a Role in the Development of Type 1 Diabetes
Experimental investigations performed in vitro on one hand, and those based on the use of animal models on the other hand (Jaidane et al., 2009), allowed to explore non-mutually exclusive mechanisms of the viral pathogenesis of T1D. Enteroviruses, especially coxsackievirus B, can promote or accelerate autoimmunity through infection and damage of pancreatic β-cells.
Epidemiological and experimental data together suggest that the enteroviral pathogenesis of T1D would be the result of an interplay between infection, β-cells, the innate and adaptive immune system, and host genes. Figure 5 gives an overview of the interactions between enteroviruses and the host that can be involved in the pathogenesis of T1D.
Recent research findings on the relationship between enterovirus infection and T1D have been revealing. Epidemiological studies strongly suggest an association between enterovirus infections and the occurrence of T1D in genetically predisposed individuals. Such a possible association is supported by the detection of these viruses in various tissues and especially the pancreas of patients with T1D. Overall, the results of retrospective and prospective epidemiological studies are in agreement with a relationship between enterovirus infections and T1D.
However, even though enteroviruses have been isolated from tissues of patients with T1D, a definitive causative correlation is yet to be established. Therefore, studies on additional patients with T1D and controls, such as nPOD (network for pancreatic organ donors with diabetes), aimed to screen and isolate enteroviruses are needed to bring forth evidence for a causative role of these infectious agents (The Juvenile Diabetes Research Foundation: www. jdrfnpod.org).
In addition to epidemiological data, experimental results are in favor of the role of enteroviruses in the development of T1D. In vitro and in vivo investigations carried out in animals showed that several pathogenic mechanisms of enterovirus infection may participate in the impairment of β-cells. Interestingly, recent findings strongly suggest that enteroviral infections and type 1 interferons could play a role in the development of T1D (von Herrath, 2009a).
Collectively, the results of experimental studies suggest that interactions between enteroviruses, the innate and adaptive immune system, and the genetic background are involved in the pathogenic processes of T1D.
Further studies are needed to elucidate the relationship between enteroviruses and T1D pathogenesis in order to develop strategies aimed to prevent or to treat the disease (von Herrath, 2009b).