Non-cytolytic enterovirus

Non-cytolytic enterovirus is an aberrant form of enterovirus able to cause chronic infection. Normally enteroviruses only produce acute infections, but mutations acquired in the host can convert enterovirus B serotypes such as coxsackievirus B into non-cytolytic viruses able to create persistent intracellular viral RNA infections.

This viral RNA infection self-replicates and propagates very slowly, and does not destroy the cells it inhabits, but nevertheless it may cause pathological effects. Persistent non-cytolytic enterovirus is found in a number of chronic illnesses, from myalgic encephalomyelitis to diabetes, and some researchers postulate this non-cytolytic form of the virus may play a causal role in these diseases.

Also known as non-cytopathic enterovirus, and defective enterovirus.

Introduction
Non-cytolytic enterovirus is an aberrant form of viral infection enterovirus B serotypes (such as coxsackievirus B and echovirus) can transmute into within the host, a transformation that allows a virus normally only capable of acute infections to create persistent ones. Evidence for the existence of this form of infection dates back to at least 1990, but an understanding of the molecular mechanism behind the conversion of lytic to non-cytolytic enterovirus was only obtained in 2005, when Professor Nora Chapman figured out the mechanism.

The transformation of normal lytic (wild-type) enterovirus into the non-cytolytic form results from mutations (specifically, deletions in the genome) the virus acquires in the host during the first weeks of acute infection. This altered form of enterovirus may then cause persistent intracellular infections which replicate very slowly without killing the cells the virus resides in. Non-cytolytic enterovirus infections may lead to cellular dysfunction, affect immune signaling, induce autoimmunity, or elicit a pro-inflammatory immune response. Non-cytolytic enterovirus is difficult for the immune system to eradicate, so it can persist for very long periods, and is associated with a range of chronic diseases.

Persistent enteroviral infections have been found in: The persistent enterovirus infections found in the above diseases are in several cases (those shown in italics) explicitly demonstrated to be non-cytolytic. Several researchers including Prof Nora Chapman, Prof Steven Tracy and Dr John Chia theorize that persistent non-cytolytic enterovirus infection may be a cause of ME/CFS. Persistent enterovirus is being investigated as a possible cause for the other above-listed diseases as well.
 * The brain, muscle and stomach tissues of myalgic encephalomyelitis / chronic fatigue syndrome (ME/CFS) patients; see enterovirus infection studies.
 * The heart muscle in murine and human coxsackievirus B (CVB) chronic myocarditis.
 * The heart muscle in dilated cardiomyopathy (a sequelae of myocarditis).
 * The cerebrospinal fluid and spinal cord tissue of amyotrophic lateral sclerosis (a motor neuron disease).
 * The brainstem in Parkinson's disease.
 * The salivary gland epithelial cells and lymphocyte infiltrates in primary Sjögren's syndrome.
 * The cerebrospinal fluid in post-polio syndrome.
 * The murine pancreas and human pancreas, which has implications for type 1 diabetes.
 * The intestines in ileocecal Crohn's disease.
 * The heart valve tissues in valvular heart disease.
 * The calf muscle in peripheral arterial disease.
 * The stomach tissues in functional dyspepsia and chronic gastritis.

Non-cytolytic enterovirus is also known as non-cytopathic enterovirus. Cytolytic refers to destruction of a cell by rupture of its membrane, and cytopathic signifies damage to a cell, typically by rupture. Thus the terms non-cytolytic and non-cytopathic indicate this mutated enterovirus does not destroy the cells in which it resides. Non-cytolytic enterovirus may also be referred to as defective enterovirus and terminally deleted enterovirus, terms which refer to the mutations (deletions) in its genome which render this virus defective and aberrant.

The nature of non-cytolytic infection
When enterovirus is initially contracted by a host, it creates an acute lytic infection. Lytic infection involves the virus entering into host cells and replicating rapidly, producing thousands of new viral particles (virions) in each infected cell, then rupturing and killing the cell through lysis, allowing the new virions to escape and infect more cells. The immune response will clear such acute infections from the host within weeks. Until relatively recently, enterovirus was thought only capable of acute lytic infection; it was not generally believed to cause chronic infection.

However, it is now clear that during the first weeks of acute infection, enterovirus B serotypes can sometimes transmute into the non-cytolytic form — a form which can then go on to create persistent infections in the host. This form is so named because cell destruction by lysis is rarely seen in the cells it infects. The conversion from lytic to non-cytolytic enterovirus results from certain mutations (deletions at the end of the viral genome) which may manifest during the acute phase. Once these mutations occur, it changes the capabilities of the virus, producing a modified pathogen that is now able to evade the immune system, and form persistent infections.

Non-cytolytic infections are not a different species of enterovirus; they arise from regular enterovirus B serotypes in common circulation, but which as a result of these mutations get transmuted into a distinct quasispecies. Viral quasispecies are viruses which within the host have acquired small mutations in their genome, but which are still closely related to their parent virus. In the case of non-cytolytic enterovirus, these acquired mutations lead to a dramatically altered virus lifecycle.

The lifecycle of a regular lytic enterovirus centers on the virion, but once transformed to a non-cytolytic infection, the virus then exists in a very different configuration: as strands of naked viral RNA that reside within host cells creating a persistent intracellular RNA infection. This intracellular viral RNA is self-replicating and self-sustaining, and is thought can survive independently of any help from the lytic infection. Furthermore, whereas the lytic virus destroys by lysis the cells it infects, non-cytolytic enterovirus does not typically kill the host cells it inhabits, allowing the non-cytolytic infection to reside in these cells on a long-term basis.



The self-sustaining RNA infection of non-cytolytic enterovirus consists of positive and negative single-stranded viral RNA (ssRNA), as well as double-stranded viral RNA (dsRNA). The latter is formed when the positive ssRNA and negative ssRNA in the cell join to create dsRNA.

Non-cytolytic infections very rarely produce lytic virions (infectious virions with a complete genome), neither in chronic myocarditis nor dilated cardiomyopathy nor ME/CFS. Non-cytolytic infection can nevertheless propagate by packing its defective genomes into capsids (viral shells) to create virions; this is one way the defective virus can spread.

Note that the non-cytolytic state of enterovirus is distinct from viral latency: in the latent state, viruses are typically dormant for long periods and do not engage in viral replication or viral propagation; whereas a non-cytolytic infection actively replicates and propagates, albeit very slowly.

RNA viruses such as enterovirus are generally not capable of latency (usually only DNA viruses are able to go into latency). So latency is not a mechanism that enterovirus could employ to create a persistent presence in the host. But it is now clear enterovirus can form a chronic low-level intracellular infection as a non-cytolytic defective virus and can persist in immunocompetent hosts for very long periods.

Non-cytolytic enterovirus is resistant to immune elimination
Non-cytolytic enterovirus can cause chronic long-lasting infections, as its mutated RNA is not readily cleared by the immune system. The reason why this infection can evade the immune response is not clear.

One theory is that the double-stranded viral RNA (dsRNA) component of this infection confers resistance to immune eradication. The immune enzyme RNase L, released inside the cell as part of the type 1 interferon response to viral infection, is able to destroy (cleave) ssRNA, but dsRNA is resistant to destruction by this enzyme.

Dr John Chia likens dsRNA to a seed, which is hardy and allows the non-cytolytic infection to survive through periods of immune attack. Afterwards when the pressure of the immune response abates, he thinks it is probable that the dsRNA can dissociate back into ssRNA and recommence replication. In this way, the dsRNA may constantly reseed the non-cytolytic infection. Dr Chia points out that self-replicating strands of RNA or DNA are well studied, and are known as replicons.

Another theory suggested by Lévêque et al is that the deletions found in the non-cytolytic enterovirus genome might facilitate immune evasion: alphaviruses are known to utilize such deletions within their genome to evade the type 1 interferon-induced immune response, and non-cytolytic enterovirus has deletions in exactly the same area of the genome as alphavirus. The deletions in the alphavirus genome prevent interferon-induced Ifit1 immune proteins from binding to and disabling alphavirus RNA. If the genomic mutations of non-cytolytic enterovirus also confer the same immune resistance, that may help explain how this non-cytolytic virus evades immune clearance.

Non-cytolytic enterovirus may cause pathogenic effects leading to disease
In the cells it infects, non-cytolytic enterovirus synthesizes the full range of viral proteins, and these viral proteins may interfere with cellular function or cause pathology.

For example, enterovirus 2A protein, a protein produced by non-cytolytic enterovirus, has been shown in a murine model to generate dilated cardiomyopathy just on its own. The 2A protein cleaves dystrophin, decreases the contractility of heart cardiomyocyte cells, as well as increasing cell membrane permeability; and 2A can promote virus spread to adjacent cells. Enteroviral protein has been found in human heart tissues in dilated cardiomyopathy, as well as in the skeletal muscles, stomach and brain tissues of ME/CFS. In vitro studies indicate that non-cytolytic enteroviral RNA synthesizes similar amounts of viral protein as wild-type lytic enterovirus (though non-cytolytic viral RNA with the largest deletions generates less protein).

Furthermore, viral dsRNA is a potent inducer of type 1 interferons, so even the low levels of dsRNA found in non-cytolytic infections may be sufficient to chronically induce interferon, which might then lead to adverse effects (although it remains to be seen if these low levels are capable of stimulating type I interferon secretion, and one study suggests they may not, at least in murine heart muscle cells).

Non-cytolytic enterovirus infection in the heart muscle in chronic CVB myocarditis and dilated cardiomyopathy has also been shown in some studies to induce autoantibodies which target mitochondria and thus substantially inhibit energy metabolism. Article here.

How does lytic enterovirus transmute into the non-cytolytic form?
The mechanism by which lytic (wild-type) enterovirus can get transformed into a non-cytolytic virus within the host was identified by Prof Nora Chapman in a landmark 2005 study.

Chapman says that the conversion of acute lytic enterovirus into a non-cytolytic infection can only occur in specific cell types, namely in non-dividing quiescent cells (such as muscle cells), but cannot occur in dividing cells (like liver cells).

A certain amount of non-cytolytic enterovirus RNA is always generated during lytic replication, through random mutations from replication errors; but in dividing cells, because production of lytic virus is efficient, lytic virus populations dominate over non-cytolytic populations, and the cell thus is co-opted for lytic replication. By contrast, in quiescent cells, due to specific cellular conditions, lytic virus manufacture is very inefficient, and this provides an opportunity for non-cytolytic populations in quiescent cells to grow and prevail. Once non-cytolytic virus establishes itself as the dominant species in the cell, it co-opts the cell for non-cytolytic replication. This transmutation of lytic into non-cytolytic enterovirus occurs during acute infection, in a timeframe in the order of weeks after the virus is first caught.

This is the essence of how non-cytolytic enterovirus infection is born, but we now explain in detail how this transmutation from lytic to non-cytolytic virus occurs.

In the mechanics of lytic enterovirus infection, viral negative single-stranded viral RNA (ssRNA) is employed as a template to create numerous positive ssRNA copies (like a photographic negative producing lots of prints). The positive ssRNA copies are then packed into capsids (viral shells) to make new enterovirus virions. Enteroviruses are positive single-stranded RNA viruses, meaning their genome comprises positive ssRNA; so in order to create thousands of new enterovirus virions, thousands of copies of the positive ssRNA are required.



When enterovirus enters a rapidly dividing cell, it efficiently produces thousands of copies of positive ssRNA with the help of an important factor present in the cell known as hnRNP C (heterogeneous nuclear ribonucleoprotein C). This factor works by binding to the negative ssRNA, where it has the effect of accelerating production of positive ssRNA.

With the aid of hnRNP C, each strand of enterovirus negative ssRNA is able to efficiently produce around 100 strands of positive ssRNA. So in rapidly dividing cells, a 100-fold excess of positive strands over negative strands is observed.

However, in a quiescent cell, the factor hnRNP C is not available to the virus, as this factor is confined to the cell nucleus during quiescence. Only when a cell starts dividing (mitosis) does hnRNP C move into the cytoplasm (the region between the nucleus and the cell membrane), where enterovirus replicates. Without the assistance of hnRNP C, positive ssRNA production is severely impacted, and now each negative ssRNA template is only able to manufacture in the order of 1 strand of positive ssRNA. In consequence, roughly equal numbers of positive and negative ssRNA strands are observed in quiescent cells.

Because of the great shortfall of positive ssRNA in quiescent cells, very few lytic virions are created. As a result, cellular lysis — the destruction of the cell as the virions are released — does not occur. Instead the cell survives, and is populated with roughly equal amounts of enteroviral positive and negative ssRNA.

With these conditions found in quiescent cells, a non-cytolytic enterovirus infection can emerge. The emergence pivots on natural defects appearing in the viral genome. In viral replication, during the synthesis of viral RNA, reproduction errors naturally arise; these errors will often be in the form of deletions in the 5′ (pronounced "five primed") region at the terminal end of the genome (5′ is an untranslated region at the end of the genome). These deletions just result from premature termination of the process of genome transcription.

Now it just so happens that these deletions in the 5′ region are located in precisely the part of the genome that hnRNP C binds to on the negative ssRNA template. Thus once the deletions occur in the virus, this genomic defect permanently prevents efficient production of positive ssRNA.

So now, not only do we have a cell which being quiescent lacks cytoplasmic hnRNP C, but this situation is compounded by the deletions in the viral genome, which prevent hnRNP C binding to the negative strand template. Under these conditions, lytic virus creation greatly inhibited, and the replication process starts duplicating the genomes containing deletions, ie, duplicating the non-cytolytic virus. These circumstances favor the evolution and domination of non-cytolytic enterovirus.



Deletions in the 5′ terminal end of the viral genome in persistent low-level coxsackievirus B infections were first detected by Nora Chapman and colleagues in the 2005 study, and have since been demonstrated in subsequent studies. The deletions range from 7 to 49 nucleotides in length (by comparison the full coxsackievirus B genome comprises around 7,400 nucleotides).

Lévêque et al speculate that these genome deletions might also facilitate immune evasion, thereby making non-cytolytic enterovirus largely invulnerable to immune clearance.

This transmutation of lytic into non-cytolytic infection by means of genome deletions cannot occur in a dividing cell, because although viral replication in dividing cells produces RNA with the same genomic deletions, these deleted genomes are outnumbered by the high amounts of lytic virions efficiently generated, thanks to hnRNP C. And furthermore, during lytic virus production in a dividing cell, the process rapidly kills the cell by lysis, so the deleted genome RNA infection has nowhere to live. Thus conversion to the non-cytolytic form is thought possible only in quiescent cells.

So these are the conditions under which non-cytolytic enterovirus infection is born, and once this intracellular infection is created, it is observed that the immune system cannot easily clear it.

Factors that promote the creation of a non-cytolytic infection
The question arises: why do these persistent non-cytolytic infections not always occur after acute coxsackievirus B infection? What circumstances lead to acute lytic enterovirus transforming into a chronic non-cytolytic infection within the host?

Prof Nora Chapman suggests that it probably requires a high level of acute infection to seed quiescent cells with enterovirus, such that a non-cytolytic infection has a chance to emerge.

Interestingly, Dr Chia found that when corticosteroids are inadvertently given to patients during the acute stage of an enterovirus infection, this is a recipe for triggering enterovirus-associated ME/CFS. Corticosteroids suppress the antiviral immune response, and may thus cause a higher level of acute infection.

The spread of non-cytolytic enterovirus infection in the host
Non-cytolytic enterovirus, though defective due to deletions in its genome, appears capable of propagating itself within the host. This is demonstrated when cultured non-cytolytic enterovirus is injected intraperitoneally into mice: the infection travels to the heart, and then produces a persistent non-cytolytic infection in the heart tissues.

Non-cytolytic enterovirus is able to propagate because although the virus is defective, this viral RNA genome is still packaged into capsids to make virions which can transport the non-cytolytic infection to other cells. Evidence for this virion mode of propagation comes from the fact that when cultured non-cytolytic coxsackievirus B3 is introduced to a cell line, infection of those cells can be prevented by CVB3 neutralizing antibody. (Such non-cytolytic enterovirus virions may be transmitted by means of extracellular vesicles, the formation of cell protrusions, and intercellular bridges).

Interestingly, when non-cytolytic enteroviral RNA is packed into virions, as well as virions containing the normal positive ssRNA genome, negative ssRNA is also encapsidated in almost as many virions. Thus in non-cytolytic infection, virions containing the negative ssRNA anti-genome are found, which is very unusual, and this is a signature of non-cytolytic infection.

Detection of non-cytolytic infections
Non-cytolytic infection replicates at too low a level for detection by cytopathic effect, but non-cytolytic enterovirus can be detected in tissues samples by sensitive reverse transcription-PCR (RT-PCR) and by immunohistochemistry.

In their 1999 study Tam and Messner used RT-PCR to detect the non-cytolytic enteroviral RNA in the tissues, with different primers to distinguish positive from negative viral ssRNA. In order to detect the dsRNA, they first used the RNase I enzyme to destroy the viral ssRNA in the tissue, after which a positive result from RT-PCR testing indicates the presence of dsRNA. For the immunohistochemistry detection of non-cytolytic enterovirus, Dr John Chia uses the 5-D8/1 monoclonal antibody stain for highlighting enterovirus VP1 protein, and the J2 monoclonal antibody stain to reveal the dsRNA component of non-cytolytic infection. Chia employs such tissue staining to detect non-cytolytic enterovirus infections in ME/CFS patients' stomach tissue biopsies. In a study, Chia found enterovirus VP1 protein in 82% of ME/CFS patients' stomach biopsies, compared to 20% of the controls.

Indirect evidence of chronic enterovirus infection is suggested by the persistently elevated coxsackievirus B or echovirus antibody titers often observed in ME/CFS patients with these infections. However these raised antibody titers are usually only detected when sensitive techniques such as plaque reduction neutralization testing or micro-neutralization testing are employed. Complement fixation testing (CFT) is not sensitive enough to detect antibody titers in chronic enterovirus infections (CFT is only useful in the first few months of initial infection).

Dr Chia measured the antibody titers of 200 consecutive ME/CFS patients and 150 controls, using the ARUP Lab CVB and echovirus antibody micro-neutralization tests, and found the patients' titers were often substantially higher than the mean titers of the controls. Highly elevated CVB titers may also be seen in a small subset of dilated cardiomyopathy patients.

PCR blood testing is not sufficiently sensitive for the chronic enteroviral infections found in ME/CFS, because enteroviral RNA is only present at low levels in the blood, Dr Chia discovered. Chia determined that with special techniques and repeated testing, enterovirus RNA can be detected in around 30% of whole blood samples taken from chronically infected enterovirus-associated ME/CFS patients. This finding emerged from Chia's lab work in which over a 5 year period he tested more than 2,500 blood samples from over 510 ME/CFS patients by PCR. Chia observed that only 35% of patients tested positive twice or more for enteroviral RNA in their peripheral blood lymphocytes (PBL), using a highly sensitive semi-nested RT-PCR test kit from Chemicon (with an assay sensitivity of 80 to a 800 copies of RNA per ml of blood). In the same RT-PCR test 4% of the controls tested positive. Interestingly, in these RT-PCR blood tests, Chia found enterovirus RNA in 70% of the severe bedridden patients, but only in 12% of the less ill.

Furthermore, when 20 of the sickest patients' PBL cells were retested every 3 to 6 months for 2 years, Chia found that patients who had tested positive might be negative on the next test, and then might be positive again on a subsequent test; he observed it was rare for a patient with a positive PCR result to consistently test positive. Chia concluded it was clear that the enteroviral RNA present in the blood is at very low levels. Dr Chia says any PCR blood tests which have insufficient sensitivity (of above 1000 copies of RNA per ml of blood) will almost always give negative results in ME/CFS patients.

Treatment of non-cytolytic infections
Interferon alpha therapy has been found temporarily effective for enterovirus-associated ME/CFS in several studies,  with treatment often resulting in large improvements in symptoms. In one experiment by Dr John Chia, ribavirin plus interferon alpha therapy were tried: in 7 out of 10 ME/CFS patients this resulted in substantial improvements in symptoms, the elimination of enteroviral RNA from peripheral blood mononuclear cells (PBMC), and a fourfold reduction in coxsackievirus B antibody titers. However, after therapy was complete, relapse typically occurred 4 to 5 months later, along with antibody titers increasing to pretreatment values and the enteroviral RNA returning to the PBMC.

In another experiment by Dr Chia, interferon alpha plus interferon delta were used in combination to treat severe bedbound ME/CFS patients: 6 out of 11 treated patients were able to return to full- or part-time work as a result, but again relapse occurred some months later, the relapse typically triggered by a bout of heavy exertion.Thus interferon therapy does not appear to be a long-term solution for ME/CFS; however, the large improvements many enterovirus-infected ME/CFS patients experience after interferon therapy does provide evidence that enterovirus is playing an ongoing causal role in ME/CFS.

Clearly better treatments are required, and Dr Chia says there is an urgent priority to develop inhibitors for viral RNA replicase, the main mechanism for RNA replication which facilitates the persistence of non-cytolytic enteroviral RNA in infected cells.

Summary characteristics of non-cytolytic enterovirus infection
Non-cytolytic enterovirus infection:
 * Is a self-sustaining intracellular infection of naked viral RNA possessing genomic deletions which are responsible for transmuting the original lytic virus into the non-cytolytic form.


 * Is born in non-dividing (quiescent) cells such as muscle cells, where cellular conditions promote the evolutionary selection of non-cytolytic virus defective genomes in preference to intact lytic virus genomes.


 * Is made of positive and negative single-stranded viral RNA as well as double-stranded viral RNA. This RNA replicates very slowly and at a low level. The dsRNA may contribute to pathogenicity of non-cytolytic infection.


 * Is likely capable of replicating independently of the lytic enterovirus infection; tissues infected with non-cytolytic enterovirus are often largely devoid of lytic virus.
 * Is able to synthesize all the normal viral proteins in the host cell, which may alter cellular functioning and cause pathogenicity.


 * Is resistant to the immune response, and can thus reside inside host cells for very long periods. Its double-stranded RNA form is postulated to confer resistance to immune clearance; other theories suggest the genomic deletions furnish the non-cytolytic virus with the ability to evade the immune response.


 * Is not cytolytic or cytopathic: there is no significant destruction of cells (a marker of lytic infection) either in vivo or when non-cytolytic infections are cultured in vitro (hence the names non-cytolytic and non-cytopathic). The presence of enteroviral RNA in the tissues in the absence of cell destruction is a signature of non-cytolytic enterovirus infection.


 * Is able to encapsidate its defective genome into capsids to create virions, enabling non-cytolytic virus to spread. Just as many negative strand anti-genomes are encapsidated as positive strand genomes, and negative strand virions are a signature of non-cytolytic enterovirus infection.


 * Is observed to have a reduced positive to negative single-stranded RNA ratio: the ratio in acute lytic infection is around 1:100, but in persistent non-cytolytic infection the ratio is closer to 1:1 and such reduced ratios are a signature of non-cytolytic enterovirus infection.

A brief history of non-cytolytic enterovirus
In the lifecycle of an acute lytic enterovirus infection, this virus will enter a host cell, replicate into thousands of new virions, destroying the cell by lysis in the process. However enterovirus, like most RNA viruses, is not capable of assuming a latent state within cells, and enterovirus infections are generally considered to be acute and rapidly cleared by the host immune response. Indeed, John Chia points out that even today, most physicians are taught that enterovirus does not form chronic infections.

Thus it was puzzling why coxsackievirus B appeared capable of forming persistent infections in the heart muscle tissues in chronic CVB myocarditis and dilated cardiomyopathy, and in the skeletal muscle tissues in ME/CFS. Enteroviral RNA would be detected in these tissues, indicating an ongoing enterovirus infection, but infectious lytic virus can almost never be isolated from the adult heart in chronic myocarditis, nor from the skeletal muscles in ME/CFS. (Lytic virus is detected by its cytopathic effect: when infected tissue samples are added to a cell culture in vitro, if lytic enterovirus is present it will lyse the cells in culture and this cytopathic effect can be observed; but a cytopathic effect is very rarely seen with chronic myocarditis, dilated cardiomyopathy and ME/CFS tissue samples).

This was the conundrum for decades: how can enteroviral RNA be persistently present in these tissues, but without any lytic virus present? This paradox was eventually resolved by the discovery and understanding of the non-cytolytic form of enterovirus, which can create chronic intracellular infections of self-replicating enteroviral RNA, but with very little lytic virus generated.

Some studies in non-cytolytic enterovirus research
Within a few years of their discovery in the late 1940s, the Coxsackie B viruses were shown to be involved in myocarditis, but direct evidence that enterovirus was able to maintain a persistent infective presence in the heart muscle tissues would have to wait until molecular testing techniques became available in the late 1980s which could detect the virus in the tissues by its RNA. Once these molecular techniques were employed, they not only demonstrated that enteroviral RNA is chronically present in the tissues, but they led to an intriguing further discovery: that in persistent enterovirus tissue infections, the ratio of positive to negative strand enteroviral RNA was greatly reduced compared to acute enterovirus infection. This prompted the speculation that persistent enterovirus infections might involve a virus with a mutated defective genome.

Fifteen years later, these genomic defects were first detected by Nora Chapman, Steven Tracy, Kyung-Soo Kim, William Tapprich and colleagues in a landmark 2005 paper. The defects were in the form of deletions in the 5′ region of the genome. In this same paper, Chapman identified the precise molecular mechanism by which these acquired genome defects were able to transform lytic enterovirus (which normally only causes acute infections) into a virus capable of very long-term viral persistence in the tissues. Later studies confirmed the presence of these deletions in the 5′ region in other chronic enterovirus human heart infections, thus providing supporting evidence for Chapman's proposed mechanism.
 * Archard et al 1987:  using molecular hybridization testing detected Coxsackie B virus RNA in 56% of the heart tissue biopsies from myocarditis or dilated cardiomyopathy, whereas biopsies from controls with non-viral heart diseases were all negative. The authors also found viral RNA in the skeletal muscles of patients with chronic myositis (juvenile dermatomyositis and adult polymyositis). This may have been one of the first studies to provide solid evidence of the persistent presence of enterovirus in the tissues.


 * Archard et al 1988:  may have been the first to find evidence for persistent enterovirus infection in the skeletal muscles of ME/CFS patients, using molecular hybridization testing to detect enteroviral RNA. They found enteroviral RNA even in ME/CFS patients who had had this illness for 20 years.


 * Righthand et al 1989: demonstrated that an echovirus 6 infection could be maintained for over 6 years in a cell line, and that this infection was and was not cytolytic (showed no cytopathic effect) when transfected into uninfected susceptible cells.


 * Cunningham et al 1990:  was possibly the first indication that there was something odd about persistent enterovirus infections. In normal lytic enterovirus infections, there is around 100 times more positive strand RNA than negative strand RNA in infected cells. But in the enterovirus-infected muscles of ME/CFS patients, Cunningham et al found roughly equal amounts of positive and negative strand RNA. Thus they found instead of the normal approximately 100:1 positive to negative strand ratio, in ME/CFS the ratio is closer to 1:1. The authors concluded that this finding of near equal amounts, together with the failure to isolate and culture infectious lytic virus or detect virus-specific antigens, suggested that a mutant defective enterovirus may be responsible for these persistent muscle infections.


 * Archard et al 1991:  found a reduced positive to negative strand RNA ratio of around 1:1 in the heart muscle in dilated cardiomyopathy, and in myocarditis after the acute inflammation had cleared up. This is just as Cunningham had found in the skeletal muscles of ME/CFS patients. The authors concluded that the persisting enterovirus infection of the myocardium is probably underpinned by defective virus mutants which do not complete a productive replication cycle.


 * Klingel et al 1992:  found the same reduced positive to negative strand ratio of around 1:1 in an immunocompetent mouse model of chronic myocarditis. They also noted a strong correlation, both spatial and temporal, between viral replication and formation of lesions in both the acute and chronic phase of infection.


 * Tam and Messner 1999:  in a chronic CVB myositis mouse model, detected a decreased positive to negative strand RNA ratio of about 1:1 in the muscle tissues, mirroring what Cunningham had found in ME/CFS patients. They also demonstrated that enteroviral persistence in the muscle is characterized by the formation of enteroviral dsRNA. The authors pointed out that intracellular ssRNA decays within hours, so they postulated that the persistence of non-cytolytic enterovirus may instead arise from the dsRNA component of this infection, suggesting that the stability of dsRNA prevents its elimination by the immune system. They also suggested this dsRNA may contribute to pathogenicity of non-cytolytic enterovirus infection. So in the Tam and Messner model of non-cytolytic enterovirus, the viral dsRNA is hypothesized to be central to sustaining this form of infection.


 * Feuer et al 2002:  proposed that persistent coxsackievirus B infection may be based on infection of quiescent cells, and observed that in quiescent cells, CVB appears to be able to persist in a form similar to viral latency.


 * Kim et al 2005:  in this landmark study, Prof Nora Chapman identified the mechanism which underpins enteroviral RNA persistence. They isolated persistent coxsackievirus B3 from the hearts of mice that had been infected with this enterovirus one year earlier, and discovered mutations in this virus: deletions in the 5′ region at the end of the viral genome. They point out that such deletions are observed to inhibit lytic virus replication by decreasing positive strand viral RNA production, and propose that this lytic virus inhibition provides the opportunity for mutated virus populations to dominate the cell, thereby giving birth to persistent non-cytolytic infection. In the Chapman theory of non-cytolytic enterovirus, the deletions in the viral genome are central to creating and sustaining this persistent form of enterovirus infection.


 * Chapman et al 2008:  found for the first time non-cytolytic enterovirus with deletions in the genomic 5′ region in human coxsackievirus B2 myocarditis. Previously non-cytolytic enterovirus with such deletions had only been detected in mouse models.


 * Chia and Chia 2008:  found enteroviral VP1 protein in stomach tissues of 82% of ME/CFS patients, compared to 20% of the controls, and found enteroviral RNA in 37% of patient tissues compared to 5% of controls. They discovered that in the stomach tissues of ME/CFS patients, lytic virus is not always found, despite the presence of enteroviral RNA. This is a signature of non-cytolytic infection.


 * Feuer et al 2009:  showed that coxsackievirus B3 may persist as a chronic low-level non-cytolytic infection in the central nervous system of mice. Three case studies have also found chronic enterovirus infection in the brains of ME/CFS patients; see post-mortem brain studies.


 * Lévêque et al 2012:  found for the first time non-cytolytic enterovirus in the heart tissues in human idiopathic dilated cardiomyopathy: they found enterovirus genomes with deletions in the 5′ region, and they found a reduced positive to negative enteroviral ssRNA ratio.


 * Tracy et al 2014:  found non-cytolytic coxsackievirus B with deletions in the 5′ region can persist in the pancreas of mice. This has implications in the study of type 1 diabetes, which has long been linked to CVB. The fact that non-cytolytic CVB can form long-term infection in the pancreas raises the possibility that this infection might be a cause of TID, at least in a subset of patients. Pre-existing insulitis in murine pancreatic islets makes the islets more susceptible to CVB infection.


 * Chia et al 2015:  using VP1 and dsRNA staining found both enteroviral VP1 protein and enteroviral dsRNA in the stomach tissues of functional dyspepsia and chronic gastritis patients both with and without an ME/CFS diagnosis. The dsRNA is postulated to be a mechanism of non-cytolytic enteroviral persistence.


 * Smithee 2015: demonstrated that artificially engineered non-cytolytic CVB populations in vitro displayed impaired replication, with RNA levels 100,000-fold lower than those of lytic (wild-type) virus.


 * Bouin et al 2016: found a mix of non-cytolytic (terminally deleted) and complete coxsackievirus B3 genomes in the endomyocardial tissues of a patient with dilated cardiomyopathy, along with a low viral RNA load, and a decreased positive to negative strand RNA ratio of 4.8:1. They found the majority of the CVB genomes in the tissues were terminally deleted (with deletions varying from 15 to 48 nucleotides is length), but a small minority (0.9%) of these genomes were found to be the complete genomes of the lytic (wild-type) virus.


 * Lévêque et al 2017: in a series of in vitro experiments on non-cytolytic coxsackievirus B3 with artificially engineered 5′ terminal deletions, found that this terminally deleted CVB3 RNA synthesized viral proteins in similar quantities to lytic CVB3 infection, although non-cytolytic CVB3 RNA with the largest deletions generated less protein (see Fig 4). They were unable to detect RNA synthesis in CVB3 with deletions larger than 21 nucleotides, though point out that other studies have detected very low levels of RNA replication in such terminally deleted enteroviruses in vitro. They suggest this discrepancy is most likely be due to their testing methods being not sufficiently sensitive to detect the very low levels of RNA present. They also suggest another possibility for this discrepancy: that the small amount of lytic virus which Bouin et al found in vivo mixed in with the terminally deleted virus could act as a helper virus, facilitating terminally deleted viruses to replicate. In this model, non-cytolytic (terminally deleted) viruses must coexist with a small amount of lytic virus in order to facilitate their replication (or facilitate more efficient replication).


 * Flynn et al 2017: argue that the persistent enteroviral RNA found in myocarditis and dilated cardiomyopathy are probably not the cause of these diseases. They point out that in the experiments of Nora Chapman and colleagues inoculating mice with cultured non-cytolytic infection, this infection made its way to the heart, where it persisted, but it did not trigger any detectable pathology. However, Nora Chapman says it likely requires a high level of acute infection to seed a sufficient number of quiescent cells with enterovirus before a non-cytolytic infection emerges.

Note on defective interfering viruses
Non-cytolytic enteroviruses are similar to defective interfering (DI) viruses: both have deletions in their genomes. However, one difference is that DI viruses cannot replicate on their own, they are only able to replicate with the help of the lytic virus with a complete genome, whereas non-cytolytic enteroviruses are thought to be able to replicate independently.

Although true independent replication is not entirely certain, as in a non-cytolytic enterovirus infection, Bouin et al found most of the viral genomes are defective due to deletions, but they also found 0.9% of the genomes to be those of the intact lytic virus. Thus the lytic virus genome is still present in small quantities in non-cytolytic infection, and Lévêque suggests that this small amount of lytic virus RNA in the cell might be helping the defective viruses to replicate. However, this low level of lytic RNA found in non-cytolytic infection is not the same as an acute lytic infection, so in that respect, it is clear that non-cytolytic enterovirus can replicate independently to the lytic virus and its lifecycle.

Another difference is that defective interfering viruses have large genome deletions, typically consisting of only 5-10% of the genome of the parental lytic virus (though in the case of picornavirus DI viruses, these usually comprise 85% or more of the parental genome). By contrast, in the case of the genomic deletions of non-cytolytic enterovirus, less than 0.7% of the parental genome is lost (non-cytolytic enterovirus deletions are 7 to 49 nucleotides in size, and the full coxsackievirus B genome comprises around 7,400 nucleotides). So non-cytolytic enterovirus still possesses most of the parental lytic virus genome, thus retaining most of the functionality.

Learn more
How Does a Lytic Enterovirus Persist and Cause Chronic Disease? Presentation by Prof Nora M. Chapman.

Human Enteroviruses and Chronic Infectious Disease. Steven Tracy & Nora Chapman. Also as pdf (p.23-31).

The role of enterovirus in chronic fatigue syndrome. Dr John Chia.

Replication Defective Enterovirus Infections: Implications for Type I Diabetes. Presentation slides by Nora M. Chapman.

Human Enteroviruses and Type 1 Diabetes (p.27-33). Prof Steven Tracy. Also as pdf (p.27-33).

Revealing enterovirus infection in chronic human disorders: An integrated diagnostic approach.

Enterovirus Persistence as a Mechanism in the Pathogenesis of Type 1 Diabetes.