The laboratory began its work on the biology of the group B coxsackieviruses (CVB, serotypes 1-6 or CVB1-6) and upon the role of these viruses in the causation of human myocarditis more than 20 years ago. Myocarditis is the inflammation (itis) of the heart (cardiac) muscle (myo). It can be a very serious or deadly disease, especially in neonates and young children or strenuously exercising athletes. Myocarditis from all causes may occur as frequently as 1,000 per 100,000, but it is diagnosed much less often as symptoms are often not noted by patients until the disease has become severe. This means that the bulk of myocarditis cases are likely to be symptomatically minor.

A related disease, dilated cardiomyopathy (DCM), is often a sequela of myocarditis; DCM involves the enlargement of the heart muscle and is also quite serious, with the outcome being approximately even chances between improvement/maintenance of the disease or the need for heart transplantation to avoid death. DCM occurs in about 40 persons per 100,000 worldwide. Studies in many different laboratories, including this one, suggest that the incidence of enterovirus involvement is in the range of 15-25% of cases of myocarditis and DCM. Other viruses (for example adenoviruses) are also associated with these diseases. A good web page for more detailed information on these and related heart conditions can be found at the National Library of Medicine; just type in myocarditis or whatever subject interests you.

Our laboratory has been instrumental in describing several new aspects of enteroviral myocarditis. The laboratory cloned infectious complementary DNA (cDNA) copies of three different CVB3 RNA genomes to enable manipulation of the viral genomes1-3. The cloning of an RNA virus genome captures a specific genome in DNA form, which is less prone than the virus population itself to change genetically. Thus, by using such infectious cDNA clones, we are able to always refer to a specific nucleotide sequence with a good degree of confidence.

Furthermore, we have studied a series of different CVB3 isolates in a mouse model of disease to determine how the virus' ability to induce disease varied as a function of the virus itself. Because these viruses replicate well in mice, we can inoculate inbred mouse strains to study specific diseases such as myocarditis and pancreatitis (which is inflammation of the pancreas acinar or exocrine tissue, but not the islets of Langerhans). Using CVB3 strains isolated over a period of nearly 40 years, we have demonstrated that CVB can induce pancreatitis, or pancreatitis and myocarditis, but not just myocarditis alone. That diverse CVB3 strains isolated in different years, act in this fashion suggest that pancreatitis causation is common, myocarditis is much rarer (about <10% of strains that have been assayed induce myocarditis in mice; unpublished data), and strains that cause no disease quite rare. That myocarditis in mice never occurred on its own after CVB inoculation, but only in the presence of pancreatitis, suggested that the mechanism by which the two disease states were induced by the virus was linked4.

Viruses enter cells to replicate and produce new viruses. This entry into cells is facilitated by, in the case of viruses like the CVB, a receptor protein. We embarked on a series of experiments to determine whether we might be able to isolate and identify the protein used by the CVB to enter cells. We and others discovered the protein called CAR for coxsackievirus-adenovirus receptor. This protein belongs to the large family of proteins called the immunoglobulin superfamily. The protein protrudes from the cell surface and when it interacts with the CVB virus shell (virion surface), a part of the CAR protein embeds itself into the virion surface in a structure that was named the "canyon" by the Rossmann laboratory at Purdue University. Once bound to the receptor, the virus can enter the cell, where it replicates to spread the infection.

The CAR protein is structurally similar to other proteins called junctional adhesion molecules that participate in complex structures that allow cells to tightly adhere to other cells. Several research groups have established that the CAR protein can be found in junctions between the epithelial cells that line body surfaces (e.g. in the lungs and intestines) and endothelial cells, which cover the inner surfaces of blood vessels (see photo). CVB access to the CAR protein is believed to be necessary for the initial infection (enteroviruses infect via the gastrointestinal route) and spread through the blood to other organs. Since initial purification of the CAR protein and identification of its partial amino sequence, we generated monoclonal antibodies against CAR (used to stain the CAR in the figure above), and used methods of molecular biology to express various forms of CAR to facilitate study of its function in viral infection5 and normal cell biology6-8, and physiology9.

Having shown that some strains of CVB induce disease much more readily than other strains, it was clear that the ability to induce disease in an animal or humans was encoded in the viral RNA. In a series of experiments to determine the genetic sites in the CVB3 genome that determine whether or not a virus induces myocarditis, it was demonstrated that a relatively short sequence of RNA in the 5' non-translated region, called domain II, was key to the determination of a myocarditic phenotype10,11. This work has been further confirmed and extended by demonstrating that replacement of a cardiovirulent CVB3 strain's domain II by that from a CVB3 strain that does not induce myocarditis12, makes the resultant CVB3 strain unable to induce myocarditis even when mice are inoculated with 1 billion virus particles. For comparison in humans, the average, naturally-occurring infectious dose of a CVB (such as in sputum droplets, fecal contamination, contaminated water supplies, etc.) usually ranges from 1 virus particle upward to a few thousand. This work demonstrated that a specific RNA structure is required to induce myocarditis in mice, and most likely in humans as well. How this structure interacts with the heart cell components is a matter of ongoing research.

One of the mysteries of the enterovirus etiology of human inflammatory heart disease is that many groups have reported the presence of enterovirus RNA in diseased heart samples in the apparent absence of infectious virus particles. Recent work in this laboratory has demonstrated that a typical enterovirus, CVB3, can persist in hearts of mice for months by employing a novel mechanism. The persisting virus has a deletion of the 5' terminal end of the viral RNA. This deletion is in a region important for replication of the viral genome explaining the greatly lowered virus yield and the lack of apparent cytopathology. As infectious virus concentration is determined by the amount of cytopathology, virus in this form in hearts would not normally be detected. We cloned several variations of these genomic 5' ends, placed them into a cloned infectious cDNA copy of a CVB3 genome, and demonstrated that each modified genome can replicate without outside virus help but indeed, all of them are extremely slow to replicate13. This slow replication may allow the virus to persist in a form inside the cells so that the immune system effectively ignores the infections, and therefore cannot clear such viruses from the heart. These viruses also packaged negative strand viral RNA (the enterovirus genome is the positive strand), a finding that has not been previously demonstrated in picornaviruses. This work has demonstrated a rational explanation to explain how viral RNA can be detected in the apparent absence of infectious virus: the virus is actually present but the infectivity is not classically measurable. In recent work, we have confirmed and extended this work by showing that an enterovirus genome detected in a myocarditic human heart also shows the 5' end deletion. This finding has important implications for therapy of enteroviral myocarditis and cardiomyopathy: if the persisting virus is involved in chronic disease, antiviral therapy could help to resolve the disease.

  1. Tracy S et al. 1992 Arch Virol 122:399-409
  2. Chapman N et al. 1994 Arch Virol 135:115-130
  3. Lee CK et al. 2005 J General Virology 86:197-210
  4. Tracy S et al 2000 J Med Virol 62:70-81;
  5. Cunningham, K.A. et al. 2003. Virus Res. 92:179-186
  6. Carson, S.D. et al. 1999. J. Virol. 73:7077-7079
  7. Carson, S.D. 2000. FEBS Lett., 484:149-152
  8. Carson, S.D. 2004. Biochemistry 43:8136-8142
  9. Carson, S.D, and N.M. Chapman. 2001. Biochemistry 40:14324-14329
  10. Dunn J et al. 2000 J Virol 74:4787-4794
  11. Dunn J 2003 J Infect Dis187:1552-61
  12. Lee CK et al. 2005 J General Virology 86:197-210
  13. Kim K-S et al 2005 J Virol 79:7024-7041

CAR (stained green) is present where cells cultured from a colon carcinoma are in contact with one another. Note that they are largely absent from the outer border of the cluster where the cell membranes are not in contact with other cells.