Potential neurovirulence of common cold virus.

An elegant recent study by Thomas Dufresne and Matthias Gromeier suggests that a causative agent of the common cold, coxsackievirus A21 (CAV21), is potentially neurovirulent and could, under the right circumstances, cause a poliomyelitis-like illness. CAV21 and poliovirus are members of the enterovirus genus (family Picornaviridae) and show remarkable genetic similarity. However, CAV21 causes upper respiratory-tract infections whereas poliovirus causes neurological disease including poliomyelitis, which clinically manifests as acute flaccid paralysis. The contrast in clinical presentation has been attributed to the different receptors used for cell invasion: intercellular adhesion molecule-1 (ICAM-1) is the receptor for CAV21 and CD155 is the receptor for poliovirus. Not surprisingly, the genomes of these two related viruses show the greatest dissimilarity in the capsid (or viral coat) region. Murine ICAM-1 does not support CAV21 binding and mice are not normally susceptible to infection with this virus. Dufresne and Gromeier stably inserted the gene encoding human ICAM-1 (hICAM-1) into the genome of a mouse (which thus became transgenic), resulting in expression of hICAM-1 by the animal’s cells. They then inoculated one gastrocnemius muscle of the transgenic mouse with CAV21. Acute flaccid paralysis developed in the injected muscle, but not in any other muscle. CAV21 replication, motorneuron destruction, and inflammation were detected in the ipsilateral anterior horn of the spinal cord, but viral replication was not seen in the injected muscle. The abnormalities in the spinal cord did not occur if the sciatic nerve was transected before inoculation (figure). These observations suggest that CAV21 can be pathogenic for motorneurons and that this neurovirulence is dependent on invasion of the central nervous system by retrograde transport along nerve axons. These findings are scientifically important because they suggest that CAV21 has all the machinery required to cause a poliomyelitis-like illness, but is prevented from doing so by virtue of its receptor not being expressed at the neuromuscular junction. Dufresne and Gromeier found low levels of hICAM-1 at the neuromuscular junction of hICAM-1 transgenic mice, but not in wild-type mice. Unfortunately, they did not report whether there is hICAM-1 expression at the human neuromuscular junction, so leaving doubt as to how the observations translate to human beings. They also found hICAM-1 expression on motorneurons in the spinal cord of transgenic mice, but no paralysis occurred when CAV21 was injected directly into the central nervous system. This apparently paradoxical observation could be explained by a requirement for neuronal infection of a co-receptor that is expressed at the neuromuscular junction, but not by neurons in the central nervous system. A role has been found for decay accelerating factor as a coreceptor for CAV21 attachment to human cell lines and, indeed, decay accelerating factor is expressed at the neuromuscular junction, but could not be found on neurons in the central nervous system. 2004 sees the 50 anniversary of Enders, Weller, and Robbins receiving the Nobel Prize in Medicine for culturing poliovirus, and the year coincides with the final stages of the Global Polio Eradication Programme. How clinically significant is the discovery of neurovirulent potential for CAV21 from a public-health perspective and should it concern us? Dufresne and Gromeier chose to refer to the acute flaccid paralysis observed with CAV21 infection in hICAM1-transgenic mice as “poliomyelitis”. In line with WHO definitions, we think this term is best reserved for poliovirus and that “poliomyelitis-like illness” should be used for acute flaccid paralysis caused by other agents. Thus poliomyelitis eradication is not directly affected by Dufresne and Gromeier’s findings, but their results increase our concern that the global public-health burden of acute flaccid paralysis will not disappear with the eradication of poliovirus. Several reasons suggest that CAV21 itself is unlikely to be a clinically significant cause of poliomyelitis-like illness in human beings. CAV21 infection in the hICAM-1 transgenic mice is site-restricted and much less aggressive than poliomyelitis in human beings. Paralysis only occurred in the hICAM-1 transgenic mice via the intramuscular route and not with intravenous, intranasal, or central nervous system routes, and paralysis remained localised to the injected muscle. This observation is analogous to “provocation poliomyelitis” where skeletal muscle injury predisposes an individual to poliomyelitis from concurrent poliovirus infection. Unlike CAV21 in hICAM-1 transgenic mice, poliovirus can spread to the central nervous system directly in the context of viraemia, as well as via retrograde axonal transport. In addition, poliovirus usually enters human beings via the oral route and not by the direct intramuscular route. Dufresne and Gromeier did not report the result of oral administration of CAV21 in hICAM-1 transgenic mice. What could account for the difference in severity between CAV21-induced poliomyelitis-like illness and poliomyelitis? The most likely answer is differences in binding of viral capsid with either hICAM-1 or CD155 and the tissue distribution of these receptors and any relevant co-receptors, such as decay accelerating factor. CAV21 could present more of a threat to 10 Llovet JM, Wurmbach E. Gene expression profiles in hepatocellular carcinoma: not yet there. J Hepatol 2004; 41: 336–39. 11 Iizuka N, Oka M, Yamada-Okabe H, et al. Comparison of gene expression profiles between hepatitis B virusand hepatitis C virus-infected hepatocellular carcinoma by oligonucleotide microarray data on the basis of a supervised learning method. Cancer Res 2002; 62: 3939–44.