Viral Myocarditis

Viral myocarditis has been recognized as a cause of congestive heart failure for >50 years, but it is still a challenging disease to diagnose and treat.^1,2 The history and clinical features are often nonspecific, and practical serological markers are not available during the acute phase of the disease. Even after proper diagnosis, no clinically proven treatment exists to inhibit the development of subsequent dilated cardiomyopathy (DCM) and, in some cases, death. Accordingly, to facilitate future scientific work into this difficult clinical entity, this review proposes a clinical paradigm that focuses on the phases of viral infection and the molecular insights that are important for these phases of the infectious process with a focus on interactions between the virus and the cardiac myocyte.

Myocarditis is defined as inflammation of the heart muscle. The gold standard for diagnosis has been the Dallas criteria based on histopathology from an endomyocardial biopsy.³ It is now recognized that the Dallas criteria are not sensitive for myocarditis because they do not consider the presence of viral genome in the heart.⁴ Furthermore, an invasive procedure is required to obtain a sample of the myocardium.

Many viruses have been implicated as causes of myocarditis. These most commonly include adenoviruses and enteroviruses such as the coxsackieviruses. Recently, parvovirus B19 has been associated with a significant percentage of patients diagnosed with myocarditis and DCM.⁵ However, a growing body of data indicates that parvovirus is present in a large percentage of patients who do not have myocarditis.^6–8 Of the viruses that cause myocarditis, the cellular and molecular mechanisms associated with coxsackieviral infection of the heart have been most thoroughly investigated with murine models. Therefore, the mechanistic experiments described in this review focus primarily, but not exclusively, on coxsackieviral myocarditis. It is important to note that coxsackieviruses are members of the Picornaviridae family, enterovirus genus.

Evidence of viral infection as a cause of heart failure dates back many years. In a recent multicenter analysis of 624 patients in the United States with histologically proven myocarditis, the presence of various virus genomes was confirmed in 239 of biopsy samples (38%).⁹ Of the virus genomes identified, adenovirus, enterovirus, and cytomegalovirus were the most common groups of viruses. The presence of viral genome in endomyocardial biopsy samples also has been reported in a subset of patients with idiopathic DCM even in the absence of classic histological myocarditis.^5,9 Although it is clear that viral genomes can be identified in a subset of patients with acute myocarditis and DCM, the impact of the presence of viral genomes on cardiac function and clinical outcome is still controversial. One clinical study showed the association between viral persistence in the heart and progressive cardiac dysfunction.¹⁰ In contrast, another clinical study reported that the presence of viral genomes per se could not be a predictor of cardiac death or heart transplantation in patients with clinically suspected myocarditis.¹¹ The latter clinical study further pointed out that the presence of inflammatory T cells and/or macrophages with enhanced expression of HLA class II molecules in the heart can be a promising predictor of the clinical outcome even in the absence of viral genomes and Dallas criteria–positive findings.¹¹ Although both of these clinical studies were well designed, differences are present in baseline characteristics of patients and viral profiles identified. Thus, adjustment of inclusion criteria is necessary for a better comparison of these studies. In addition, because these clinical studies were performed in Germany, additional verification, including worldwide multicenter studies with consistent inclusion criteria, would give us a better understanding of whether the presence of viral genomes can be a predictor of the clinical outcome from the therapeutic point of view.

Phases of Viral Myocarditis: From the Perspective of the Virus

It has recently been reported that infection of the cardiac myocyte is required for the induction of cardiac dysfunction and inflammation when mice were systemically infected with coxsackievirus B3 (CVB3).¹² This demonstrates a crucial role of CVB3 infection of cardiac myocytes in the heart during the development of CVB3-mediated myocarditis. Nevertheless, some viruses found in endomyocardial biopsy samples do not always infect cardiac myocytes (Table 1^13–23). For these and other reasons, this article focuses primarily on interactions between the virus and the cardiac myocyte by dividing the phases of virus-mediated heart disease into a preinfection phase and 3 phases that occur after the virus comes into contact with the cardiac myocyte.

A preinfection phase, or phase 0, is important as a time when susceptibility to viral infection is determined and preventive measures for virus-mediated heart disease could be instituted. Phase 1 is defined as the period of time when active replication of live virus is occurring within the myocardium. Phase 2 is defined by a later time point in which replication of live virus has ceased but viral genome persists within the ventricular myocardium. In addition to identifying myocardium that has been infected, persistence of the viral genome could potentially have a role in the progression of the disease by maintaining an immune response or directly injuring the myocardium. Phase 3, a remodeling phase of viral myocarditis, occurs after the viral genome is no longer present in the heart and the inflammatory response is likely to be similar to that observed in other forms of chronic DCM. If a previously infected heart were examined only at this time point, it is likely that the viral cause of the heart disease would not or could not be identified.

Phase 0 (Preinfection)

Important Considerations That Precede Viral Infection: Prevention and Susceptibility

From a clinical point of view, prevention of and susceptibility to viral myocarditis are 2 important areas of investigation that could have an influence on disease morbidity. One strategy considered for the prevention of viral myocarditis is immunization against enteroviral infection. Feasibility of an immunization strategy using an attenuated coxsackievirus variant has been demonstrated in mouse models of viral myocarditis.²⁴ In addition, it has been demonstrated that expression of immune proteins by a coxsackieviral expression vector can prevent the development of viral myocarditis.^25,26 However, many issues need to be addressed before an immunization strategy would likely be pursued aggressively in the general population. These include a lack of a clear understanding of the incidence of myocarditis caused by each of the several viruses that have been implicated in the disease process and their associated morbidity and mortality. One would need to design a strategy that prevented several viruses and viral serotypes. It is difficult at this point in time to predict the risk-to-benefit ratio for immunization against viruses that can cause cardiomyopathy.

Factors that determine susceptibility to viral infection of the heart and the development of cardiomyopathy are not fully known. Why is it that individuals who live in close proximity to each other may be infected with the same virus, but all do not develop myocarditis? Why do some infected individuals go on to develop mild versus severe myocarditis or cardiomyopathy? Interestingly, no clear evidence to date has demonstrated that individuals who developed myocarditis are generally more susceptible to other infectious diseases compared with individuals who do not develop myocarditis. This implies the presence of genetic and environmental factors that predispose specifically to viral myocarditis.

In murine models of viral myocarditis, a variety of factors such as malnutrition, pregnancy, exercise, sex hormones, and age have been reported to affect the susceptibility to myocarditis. Selenium deficiency has also been implicated in virus-mediated cardiomyopathy in humans in the Keshan province in China, where the diet was deficient in selenium. This is part of a disease complex known as Keshan disease.²⁷ Although the exact mechanisms underlying Keshan disease are still obscure, the effect of selenium deficiency on the susceptibility to CVB3-mediated myocarditis has been examined with murine models. Intriguingly, a normally avirulent (noncardiovirulent) CVB3 (CVB3/0) has been found to acquire virulence in selenium-deficient mice after intraperitoneal infection.²⁸ Subsequently, 6 nucleotide changes have been identified in the now virulent CVB3 genome isolated from the avirulent CVB3/0–infected selenium-deficient mice.²⁹ The mutated 6 nucleotides in the virulent CVB3 have completely corresponded with nucleotides found in known virulent CVB3 (CVB3/M1, CVB3/20) genomes. These studies indicate the possibility that a change in viral virulence can occur within certain individuals on the basis of acquired factors such as selenium deficiency.

In addition to the acquired predisposition to the viral infection of the heart, the differences in susceptibility to CVB3-mediated myocarditis among inbred strains of mice have demonstrated the importance of host genetic factors, including haplotypes of the major histocompatibility complex.^30,31 In a recent study using genetic linkage analysis, 3 genomic loci independent of the major histocompatibility complex haplotypes also were found to be associated with susceptibility to CVB3-mediated myocarditis in mice.³² Importantly, to evaluate the severity of myocarditis, this group examined not only histological evidence of inflammation but also the percent of the myocardium in which Evans blue dye was taken up into the myocardial cells. This is generally a marker for sarcolemma disruption in the heart caused mainly by direct CVB3 infection, thus indicating that the susceptibility genes might be important determinants of early viral infection.³³ Indeed, the decay accelerating factor (DAF; CD55), a coreceptor for CVB3 infection, and the well-known antiviral cytokine type I interferon gene cluster are contained in the disease loci, indicating the importance of this locus in controlling the CVB3 infection.

Human genetic studies of patients with myocarditis are sparse. However, 2 reports show an association between myocarditis and genetic factors such as HLA-DQ locus and CD45 polymorphism.^34,35 Because the number of patients examined in these studies is small, comprehensive human genetic studies in myocarditis patients should be considered in the future. A better understanding of the factors that determine susceptibility will allow identification of nodal pathways that can be targeted for the prevention and therapy of viral myocarditis.

Phase 1 (Active Replication of Live Virus Within the Myocardium)

This phase of viral myocarditis spans from the time of initial virus infection to elimination of viral replication. From a clinical perspective, fulminant myocarditis and acute (nonfulminant) myocarditis often fit within this phase of viral myocarditis.

A number of molecular mechanisms important during phase 1 of viral myocarditis have been identified. They include mechanisms of viral entry into the host myocardial cell, innate immune mechanisms that regulate the early phase of viral replication, and mechanisms by which the virus directly affects the infected host cell and how it exits from the infected cell to infect adjacent cells. Each of these mechanisms is a putative target for novel therapeutics.

Mechanisms of Viral Entry Into the Cell

It is notable that both coxsackievirus and adenovirus use the same receptor, the coxsackievirus and adenovirus receptor (CAR), to infect a cell.³⁶ In the absence of CAR expression in the cardiac myocyte, the virus does not infect the cardiac myocyte, no inflammation is present in the myocardium, and cardiomyopathy does not develop.¹² In addition to CAR, coxsackieviruses are known to use DAF and adenoviruses use integrins (α_vβ3 and α_vβ5) as coreceptors for virus infection.³⁷ In epithelial cells, coxsackievirus initially attaches to DAF on the apical cell surface of the cell. This facilitates interaction of the virus with CAR localized at the tight junction of the epithelial cells, an area not normally accessible to viral particles.³⁸ The binding of coxsackievirus to DAF triggers actin rearrangement through AbI kinase activation and permits movement of the virus particle into the tight junction (Figure 1³⁹). Because CAR preferentially localizes at intercalated disks in adult cardiac myocytes,^40,41 similar or as-yet unknown mechanisms that could mediate virus particle accessibility to CAR might exist in the heart.

• Download figure
• Download PowerPoint

CAR is a transmembrane protein with 2 extracellular immunoglobulin domains (Figure 2⁴²). It belongs to the family of intercellular adhesion molecules that include intercellular adhesion molecule and vascular cell adhesion molecule receptors for rhinovirus and encephalomyocarditis virus, respectively, members of the Picornaviridae family.^43,44 The extracellular region of CAR is thought to bind to another CAR molecule on an adjacent cell as an antiparallel homodimer.⁴⁵ Its expression level is highest in the early postnatal period, and the overall level of expression tends to decrease with age. However, after 1 week of age, the majority of CAR is localized at the intercalated disk.⁴⁰ In addition, it localizes to the cell-cell junctions of the atriventricular node, and knockout of CAR results in loss of connexin-45 in the atriventricular node and complete heart block.^41,46

• Download figure
• Download PowerPoint

It is thought that the level of CAR expression has an important role in determining susceptibility to viral infection and the development of myocarditis. The high level of CAR expression in the heart at young ages may, at least partially, explain the apparently higher susceptibility to myocarditis in children. It is of interest that upregulation of CAR in the heart has been reported in humans with DCM, although it is not yet known whether the upregulation of CAR has a pathogenic role in DCM or whether it is associated with the incidence of viral infection in DCM.⁴⁷ In addition to the expression level, the impact of CAR gene mutations and polymorphisms on the susceptibility to myocarditis and DCM may be important. As far as we know, a significant relationship between CAR polymorphisms and the development of myocarditis and DCM has not been identified.⁴⁸

Innate Immunity

Once the virus successfully interacts with the receptor, a survival battle between the intruding virus and the host immune system begins. This critical time point has a major effect on the clinical outcome because, if the initial immune response is ineffective and the virus is not eliminated, chronic myocarditis can evolve, with the potential development of DCM. The immune system of higher vertebrates is typically divided into 2 broad categories, innate and adaptive immunity. Although adaptive immunity refers to antigen-specific immune responses, innate immunity is defined by antigen-independent defense mechanisms that come into play immediately after the appearance of a pathogen in the body. The innate immune system is evolutionally conserved and is the first line of the defense mechanisms for protecting the host from invading microbial pathogens.⁴⁹ In the first 4 to 5 days after cardiomyopathic virus infection, before the adaptive immunity becomes fully active, innate immunity plays a central role in various organs, including heart, to minimize virus replication and propagation.⁵⁰ Cytokines are important mediators of the innate immune response.

Interferons

Interferons, which are potent antiviral cytokines, and are one of the first and best studied of the innate immune mediators. Type I interferons include interferons of the α and β subtype. Interferon-γ is the only type II interferon. Interferon therapy has been clinically approved for cancers, multiple sclerosis, and hepatitis C infection, thus demonstrating its clinical utility. It has been shown that both type I and II interferons can inhibit CVB replication in cultured cells, and administration of type I interferons can ameliorate CVB-induced myocarditis in mice.^51–53 However, it is not clear how important endogenous interferon signaling is in the control of early viral replication in the heart. To test this, mice lacking either type I or type II interferon receptor were infected with CVB3. CVB3 infection in type I receptor–deficient mice led to a marked increase in viral replication in the liver and a marked increase in mortality. However, no significant increase was found in viral RNA in the heart of the type I interferon receptor–deficient mice, indicating that the increase in mortality was not secondary to an increase in virus infection in the heart.⁵⁴ A related experiment that focused on the effect of disruption of interferon-β on CVB3 infection has been reported. An absence of interferon-β caused a marked increase in mortality after CVB3 infection, but no significant increase in CVB3 titer was found in the heart of the interferon-β–deficient mice.⁵⁵ These findings were somewhat surprising and demonstrate that the presence of endogenous type I interferon receptor signaling is required to prevent high-level viral replication in noncardiac organs such as the liver but that there is no significant effect on early viral replication in the heart of mice infected with CVB3. In contrast, the absence of type II interferon signaling did not have a significant effect on mortality and resulted in only a mild increase in the viral titers in heart and liver.⁵⁴ These results demonstrate that endogenous type I interferons are essential for limiting viral replication in the whole animal but have little effect on viral replication in the heart. However, it should be noted that this does not exclude the possibility that exogenous administration of interferon could have a beneficial effect on the heart and other organs.⁵³

Toll-Like Receptors

The discovery of the Toll receptor in Drosophila that recognizes pathogen-associated molecular patterns⁵⁶ has highlighted a new area of investigation in the innate immune system. A human homolog of Toll (Toll-like receptor [TLR]) was subsequently identified, and it was shown that the receptor can recognize pathogen-associated molecular patterns of microbial pathogens and activate innate immune system as a defense against pathogen invasion.⁵⁷ To date, 10 and 13 members of TLRs have been identified in humans and mice, respectively.⁵⁸ TLR signaling is activated by a variety of ligands that are generally associated with infectious pathogens. In humans, it has reported that a TLR3 P544S mutation leads to the expression of dominant-negative TLR3, which predisposes to herpes simplex virus 1 encephalitis.⁵⁹ In addition, an association between TLR4 or TLR2 mutations and bacterial septic shock has been reported.^60,61 TLR2, TLR3, TLR4, TLR7, TLR8, and TLR9 are reported to mediate antiviral effect by inducing type I interferons (Table 2⁶²). In the human heart, although the mRNA expression level of TLR7, TLR8, and TLR9 is negligible, relatively low but clear mRNA expression of TLR3 and 4 has been reported from 2 independent groups using quantitative polymerase chain reaction.^63,64 Given the data above on the lack of an antiviral effect of endogenous interferon signaling in the heart, it is not clear whether TLR induction of interferon signaling in the cardiac myocyte is likely to have a potent effect on viral replication or if other signaling mechanisms might be important.

TLR3 and TLR7/8 signaling can be activated by double-stranded RNA (dsRNA) and single-stranded RNA, respectively (Table 2). This is relevant to myocarditis in that the enteroviral genomes consist of positive, single-stranded RNA. After entry into the cell, the positive-strand RNA is released from the viral capsid proteins. The viral genome replicates using the positive-strand RNA as its template. This results in formation of dsRNA intermediates. Therefore, both single-strand RNA and dsRNA are present in virally infected cells. Infection of TLR3 knockout mice with encephalomyocarditis virus (EMCV), another positive, single-strand RNA virus of the picornavirus family that causes myocarditis, leads to significantly earlier mortality in association with increased viral replication and myocardial injury in the heart compared with wild-type mice.⁶⁵ Considering the facts that this group estimated the virus titer and myocardial injury at 3 and 5 days after EMCV infection and that histological inflammation of the heart in TLR3 knockout mice was even greater than wild-type mice, TLR3-mediated recognition of EMCV infection and subsequent activation of antiviral mechanisms seem to be quite important innate immune mechanisms to minimize virus replication in the heart. Interestingly, unlike the decrease in inflammatory cytokines such as tumor necrosis factor, interleukin (IL)-1β, and IL-6, a paradoxical increase was found in interferon-β in the TLR3-deficient heart after EMCV infection. This suggested that TLR3-dependent innate immunity may activate as-yet unknown interferon-independent antiviral mechanisms in the heart (Figure 3⁶⁶). TLR4 mainly recognizes lipopolysaccharide of Gram-negative bacteria (Table 2). The importance of TLR4 in CVB3 infection of the heart also has been reported using TLR4 knockout mice.⁶⁷ Although the mechanisms underlying CVB3-mediated TLR4 signaling activation are still unclear, TLR4 deficiency significantly increased CVB3 titer in the heart 2 days after infection. TLR7/8 recognizes synthetic imidazoquinoline-like molecules and single-stranded RNA,⁶⁸ and TLR9 recognizes bacterial and viral CpG DNA motif.⁶⁹ A role of these other TLRs in early viral replication in the heart has not been determined yet; however, deficiency of myeloid differentiation factor-88 (MyD88), an important adaptor molecule for TLR2, TLR4, TLR5, TLR7, and TLR9 signaling, decreased CVB3 titers in the heart at 4, 7, and 10 days after infection with better survival compared with wild-type infected mice.⁷⁰ This rather unexpected result suggests the complexity of TLR signaling regulation in the heart. It might be possible that MyD88-independent antiviral signaling could be enhanced in the absence of MyD88 during CVB3 infection or that a MyD88-dependent signaling pathway could activate CVB3 replication by unknown mechanisms in the heart (Figure 3). Nevertheless, taken together, these data using TLR knockout mice show that TLR-mediated virus-sensing mechanisms in the heart have an important role in the pathogenesis of myocarditis.

• Download figure
• Download PowerPoint

RNA Helicases That Sense Incoming Virus Infection

Unlike other TLRs mainly found on cell surface, TLR3, TLR7/8, and TLR9 preferentially localize in the endosomal compartment.⁵⁸ This suggests that these TLRs recognize viral nucleic acids released inside the endosomes after viral internalization. Given that most viral replication intermediates can be found in the cytosol outside the endosomes, the TLR-mediated virus-sensing mechanisms inside the endosomes may not be sufficient for detecting the viruses within the host cell. Recently, it has been found that intracellular viral dsRNA is recognized by 2 RNA helicases, retinoic acid–induced protein I (RIG-I)⁷¹ and melanoma differentiation–associated gene 5 (MDA-5).⁷² Similarly, intracellular viral DNA can be recognized by DNA-dependent activator of interferon-regulatory factors (DAI; also known as DLM-1/ZBP1).⁷³ Although the in vivo role of DAI against DNA virus infection such as adenovirus has not yet been determined, the in vivo significance of the RIG-I and MDA-5 pathway in RNA virus infection was confirmed by the generation of RIG-I⁷⁴– and MDA-5⁷⁵–deficient mice. In addition, the role of each RNA helicase in the recognition of various viruses has been determined: RIG-I is essential for the recognition of paramyxoviruses, influenza virus, and Japanese encephalitis virus; MDA-5 is critical for detection of EMCV and the synthetic analog of viral dsRNA, polyriboinosinic:polyribocytidylic acid.^75,76 Structurally, both RIG-I and MDA-5 contain 2 important domains: the caspase activation and recruitment domain (CARD) and RNA helicase domain. The RNA helicase domain is responsible for dsRNA recognition and binding, which leads to the dimerization and structural alterations of RIG-I and MDA-5 that enable CARD to interact with downstream adaptor proteins. Recently, a crucial adaptor molecule that connects RIG-I and MDA-5 to downstream antiviral mechanisms has been found by independent groups and has been ascribed 4 different names: MAVS, IPS-1, VISA, and Cardif.^77–80 Here, we use mitochondrial antiviral signaling (MAVS). MAVS has an N-terminal CARD and a C-terminal mitochondrial transmembrane domain. The interaction of dsRNA/RIG-I or MDA-5 complex with MAVS through the CARD mediates the activation of transcriptional factors such as nuclear factor-κB and interferon regulatory factors 3 and 7, which eventually lead to various innate immune reactions, including type I interferon expression (Figure 4⁸¹). The in vivo essential role of MAVS in innate immune response against a variety of RNA viruses has subsequently been confirmed with MAVS knockout mice.^82,83 Surprisingly, it has been found that EMCV titer increases by ≈1000-fold in the heart of MDA-5 or MAVS knockout mice 48 hours after infection.^75,83 This indicates a pivotal role of MDA-5–MAVS pathway in sensing EMCV infection in the heart.

• Download figure
• Download PowerPoint

JAK-STAT Signaling and Suppressors of Cytokines Signaling

It has been demonstrated that a profound innate antiviral defense mechanism exists within the cardiac myocyte that can be inhibited by suppressor of cytokine signaling (SOCS) 1 or SOCS3.^84,85 SOCS family proteins are known to be negative-feedback regulators of janus kinase (JAK) and signal transducers and activators of transcription (STAT) signaling. Cytokines such as interferon-α/β, interferon-γ, and IL-6 exert their effect by binding to specific receptors in the cell membrane that subsequently activate intracellular signaling through JAK and STAT signaling. Activated STAT translocates into nucleus and activates transcription of cytokine-responsive genes, including SOCS. The induced SOCS molecules inhibit JAK-mediated phosphorylation of the cytokine receptor, inhibiting activation of STAT signaling.⁸⁶ This negative-feedback regulation via SOCS tightly regulates the duration and intensity of the cytokine-induced JAK-STAT signaling (Figure 5). CVB3 infection is associated with activation of JAK-STAT signaling in the heart with an induction of SOCS1 and SOCS3 mRNA. To understand the in vivo significance of the SOCS expression in the cardiac myocyte, transgenic mice that express SOCS1 or SOCS3 under the direction of α-myosin heavy chain promoter were infected with CVB3. Cardiac-specific transgenic expression of SOCS1 or SOCS3 markedly increased cardiac myocyte susceptibility to CVB3 infection (Figure 6). Because expression of the transgene was limited to cardiac myocytes without expression in immune cells, the results clearly demonstrate a crucial role for innate immune mechanisms that can be affected by SOCS within the cardiac myocyte. These innate immune mechanisms in the cardiac myocyte identify potentially important targets for diagnostic and therapeutic strategies aimed against viral myocarditis. In addition, as-yet unidentified genetic alterations of these pathways could significantly alter susceptibility to viral myocarditis.

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

Mechanisms of Direct Virus-Mediated Myocardial Injury

Viruses that successfully avoid elimination by the innate immune system begin to replicate, producing viral proteins that can cause direct myocardial injury. For example, CVB3 infection is sufficient to induce myocardial injury in severe combined immune deficiency (SCID) mice in which matured T and B lymphocytes are absent.⁸⁷ The mechanisms underlying direct myocardial injury by picornaviruses have been well studied among cardiomyopathic viruses. Picornavirus protease 2A has been shown to cleave eukaryotic initiation factor-4G, resulting in inhibition of the host cell protein synthesis machinery.⁸⁸ In addition, it has been shown that CVB3 protease 2A specifically cleaves the hinge 3 region of dystrophin and disrupts the integrity of the sarcolemma membrane.³³ Emphasizing the potential importance of protease 2A, it has been shown that cardiac-restricted expression of CVB3 protease 2A alone is sufficient to induce cardiomyopathy.⁸⁹ Intriguingly, dystrophin deficiency itself affects the susceptibility to CVB3 infection of the heart by enhancing viral propagation to adjacent myocytes.⁹⁰ Considering the importance of dystrophin in the development of DCM in patients with Duchenne or Becker muscular dystrophy, the increased susceptibility to viral infection in cardiac myocytes that lack dystrophin suggests the possibility that recurrent viral infections of patients with abnormalities in the dystrophin-glycoprotein complex may contribute to the resulting cardiomyopathy. In addition to the proteolytic cleavage of proteins in the cardiac myocyte, it has reported that both CVB3 proteases 2A and 3C can induce apoptosis through activation of extrinsic caspase-8–mediated pathway and intrinsic mitochondria-mediated apoptosis pathway.⁹¹ Taken together, it is apparent that viral proteases can have a variety of detrimental effects on the cardiac myocyte. Inhibition of the effect of these viral proteases could be a promising approach against virus-mediated cardiomyopathy.

Lymphocyte-Mediated Myocardial Injury

After 6 to 7 days after virus infection in the mouse, activation of adaptive immunity begins. A prominent finding in the latter part of phase 1 is infiltration of T lymphocytes in the heart, which usually peaks at 7 to 14 days after virus infection.⁹² The T-cell infiltration leads to antithetical results: the clearance of virus-infected myocytes (beneficial) and the cell-mediated myocardial injury or necrosis (detrimental). In fact, it is well known that T-cell infiltration coincides well with severe acute pathological damage in the myocardium.⁹² Considering the fact that the cardiac myocyte is a terminally differentiated, nondividing cell, the T-cell–mediated viral elimination accompanied by myocardial damage has been thought to be less favorable compared with that in the infected dividing cells, which can be made up by proliferation of uninfected cells. It has been reported that CD⁴⁺ T-cell (helper T lymphocyte) and CD⁸⁺ T-cell (cytotoxic T lymphocyte) double-knockout mice have less mortality with less myocarditis after CVB3 infection.⁹³ Interestingly, despite the complete ablation of both CD⁴⁺ and CD⁸⁺ T cells, no significant difference has been found in the virus titer of the heart at 4 and 7 days after CVB3 infection compared with control mice, indicating the importance of T-cell–independent virus clearance mechanisms in the heart. Consistent with the CD⁴⁺ and CD⁸⁺ T-cell double-knockout mice data, the pathological role of T lymphocytes in the development fatal myocarditis also has been reported using different mouse models such as p56^lck and CD45 knockout mice.^94,95

Phase 2 (Persistence of Viral Genome Without Detectable Viral Replication)

A number of animal models of coxsackievirus infection indicate that the viral genome can persist for an extended period of time in the myocardium, even though it may not be possible to isolate replication-competent virus from the tissue. To determine whether expression of a replication-incompetent viral genome in the cardiac myocyte could induce a DCM, transgenic mice were generated that expressed a replication-restricted CVB3 cDNA mutant exclusively in the heart. This allowed low-level expression of coxsackieviral genomes in the cardiac myocyte without formation of infectious virions, thus preventing a productive viral replication cycle. This resulted in the synthesis of viral plus- and minus-strand RNA without formation of infectious viral progeny. Histopathological analysis of transgenic hearts revealed typical morphological features of myocardial interstitial fibrosis, hypertrophy, and degeneration of myocytes, thus resembling DCM in humans. This occurred in the absence of virus-neutralizing antibodies.⁹⁶ These findings clearly indicate that the expression of a replication-defective viral genome in the cardiac myocyte can contribute to the development of cardiomyopathy.

In humans, detection of viral genome has been demonstrated on many occasions in patients with myocarditis and occasionally in patients with DCM, but it is unusual that replication-competent virus can be isolated from the myocardium in patients with myocarditis. It should be noted that the ratio of positive- to negative-strand enteroviral RNA has been shown to be greater with active virus replication than with persistent virus infection.⁹⁷ This indicates that the active enterovirus replication is restricted at the level of the viral positive-strand RNA synthesis in the persistent infection state.

Phase 3 (Remodeling in the Absence of Replicating Virus or Viral Genome)

Because patients who present at this phase of viral myocarditis have no evidence of virus or viral genome within the myocardium, they generally are managed much like other patients with DCM. In most cases, it may not even be possible to determine whether the inciting event was a viral infection. This highlights the importance of specific and sensitive diagnostic methods for viral myocarditis that can be used at early stages of viral infection.

Conclusions

Considerable evidence now exists that viral infection of the myocardium has a profound and significant effect on the initiation and progression in viral myocarditis. Attention to the effect of the virus within the myocardium will likely facilitate future diagnostic and therapeutic modalities that may improve our management of this difficult disease.

Footnotes