Persistent polyomavirus infection is characterized by three major states of infection: the latent or attenuated state, a state of limited activation, and activated virus growth accompanied by tissue destruction. Determination of which factors are responsible for virus activation includes not only the role of the immune system but also the possible influence of other viruses that often co-infect immunoimpaired patients.
Cytomegalovirus infection is common in BMT and RT patients, and HIV infection is one of the major risks for PML. Recently, HHV-6 was found to co-localize with JCV in PML lesions. In addition, co-infection with a second polyomavirus may also have an impact on virus expression. The most important candidates for heterologous transactivation are BKV, herpesviruses, and retro-viruses, which are known to have viral transactivating capacities. Heterologous transactivation of viruses can happen at different stages of virus-host interaction and virus growth. In the case of polyomaviruses, the most prominent transactivators are regulatory early proteins acting as either replicating enzymes or as transcription activators. Alternatively, the interaction of cellular proteins on the heterologous promoter can be influenced indirectly by changing the pattern of transcription factors within the cell.
Concomitant infections of BKV and JCV in the urogenital tract affect the same target cells as verified by the detection of viruses in kidney tissue and by viruria. It is a common event, frequently established in individuals after renal and bone marrow transplantation, in HIV-infected patients (Markowitz et al., 1993), in pregnant women (Arthur et al., 1989; Markowitz et al., 1991), and in immunocompetent individuals (Arthur et al., 1989; Chesters et al., 1983; Dorries and Elsner, 1991; Flaegstad et al., 1991; Grinnell et al., 1983b;
McCance, 1983; Sundsfjord et al., 1994b). Nonetheless, the influence of double infections on renal symptomatology or extent of viruria has not been reported (Arthur et al., 1989; Sundsfjord et al., 1994a). Even in BMT patients, JCV in close association with BKV viruria (Azzi et al., 1999; Chan et al., 1994) exhibited no particular differences among single excreters and co-excreting patients. Hence co-expression had no noticeable influence on either JCV or BKV viruria and associated disease (Chan et al., 1994).
The presence of virus DNA in the peripheral blood cells of both healthy and immunocompromised persons was established by PCR analyses and ISH with a radioactive virus-specific probe of genomic length. Reduced sensitivity of the in situ technique detected merely single virus-infected cells; however, the cells carried almost identical amounts of BKV and JCV DNA (Dorries et al., 1994). Because both viruses are able to replicate under restrictions in lymphocytes in vitro (Atwood et al., 1992; Portolani et al., 1985), it is conceivable that they are both periodically activated to virus growth in peripheral blood cells. Whether this has consequences or may influence each other's activity is not yet known.
BKV was not expected to invade the CNS at a high rate, and there have been only rare reports of the presence of both virus genomes in brain tissue in either the asymptomatic brain (Elsner and Dorries, 1992; Vago et al., 1996) or the PML patient (Ferrante et al., 1995; Vago et al., 1996). Cell types promoting polyomavirus persistence in the CNS are not yet defined. Nevertheless, the detection rate of genomic virus DNA in study groups from different laboratories revealed that the amount of BKV DNA in tissue specimens was considerably lower than that of JCV DNA (Elsner and Dorries, 1992; Ferrante et al., 1995; Vago et al., 1996; White et al., 1992). This corresponds to PCR analyses of the CSF demonstrating that BKV is rarely detectable in a large number of samples (Gibson et al., 1993; Hammarin et al., 1996; Perrons et al., 1996; Vago et al., 1996). Although concomitant infection with BKV and JCV can frequently be detected in tissue specimens and in all groups of polyomavirus-infected patients, a transactivating mechanism and the resulting effects remain rather unlikely.
Transactivation may also involve herpesviruses, which can act on polyoma-virus DNA replication. Cytomegalovirus (CMV) is highly prevalent in the human population and can infect virtually any organ of its host (Sinzger and Jahn, 1996; Tevethia and Spector, 1989). Co-infection can occur in the kidney, lung, CNS, and lymphoid organs. Specifically, epithelial cells, fibroblasts, and endothelial cells are potential common host cells for BKV and CMV. Stromal cells and CD34-positive bone marrow progenitor cells might be cell types that can be co-targeted by JCV and CMV (Mendelson et al., 1996; Sinclair and Sissons, 1996). CMV infection is often activated in AIDS patients after RT, and there is a high incidence of CMV infection in patients with hemorrhagic cystitis (HC) after BMT (Childs et al., 1998). In AIDS patients no co-detection and no correlation between polyomavirus and CMV viruria was observed (Sundsfjord et al., 1994a). Similarly, the high incidence of CMV after kidney transplantation (Tolkoff-Rubin and Rubin, 1997) is not matched by an enhanced activity of polyomavirus infection.
Molecular interaction of CMV with JCV is believed to affect the level of DNA replication (Heilbronn et al., 1993) or possibly transcriptional activity. In contrast, co-infection of CMV and BKV in tissue culture does not result in activation of BKV infection (Goldstein et al., 1984). Lack of interaction among CMV and the human polyomaviruses was confirmed by treatment with acyclo-vir being able to reduce CMV activation in HC patients, but having no influence on BKV-associated HC. Although BKV T antigen is able to induce the expression of CMV immediate early and early gene expression (Kristoffersen et al., 1997), co-infection of both viruses in the same cell in vivo has not yet been reported and analyses of polyomavirus load in lymphoid subpopulations do not point to an interaction in vivo (personal observations).
At present, the most interesting virus detected in close association with JCV is human herpesvirus type six (HHV6) in oligodendrocytes within PML lesions (Blumberg et al., 2000). Co-localization was detected by in situ PCR, and correlation of polyomavirus infection with that of HHV6 was astonishingly high. HHV6 is ubiquitous, with a high prevalence in the adult population. It establishes life-long infection in the brain, the urogenital tract, the lung, the liver, and peripheral blood cells. It is conceivable that polyomaviruses and HHV6 have a common host cell not only in the CNS but also in peripheral organs. HHV6 activation occurs frequently after transplantation and is often associated with CMV infection. In the adult, it is usually asymptomatic and not associated with severe illness unless accompanied by CMV (Stoeckle, 2000). Due to the transactivation mechanisms that come into effect by infections with other herpesviruses, a comparable interaction of HHV6 with the human polyomaviruses is conceivable. However, it is a single study exclusively using in situ PCR, which demonstrated co-localization of JCV and HHV6 in PML lesions. Moreover, in situ PCR is one of the most sensitive and most difficult methods at present available and as such is inclined to non-specific signals. Although fascinating, the findings need to be confirmed by other methods and have to be proven by further studies on possible molecular mechanisms involved.
The question of whether the retrovirus HIV-1 may transactivate JCV is important because it became clear that PML is one of the life threatening opportunistic infections in AIDS patients. Molecular studies revealed that transacti-vation occurs in vitro at the level of transcription by HIV-Tat induction of the JCV promoter, and BKV T antigen is able to transactivate the HIV long terminal repeat (LTR). One site that has recently been described for BKV infection is the skin in AIDS-related Kaposi's sarcoma. The detection rate of BKV in the skin appeared to be higher than at other sites (Monini et al., 1996). However, transactivation events of a persistent virus in combination with progression to tumor development could explain the exceedingly high level of BKV (Cavallaro et al., 1996; Corallini et al., 1996). CNS infection of HIV-1 in oligodendrocytes is regularly low (Bagasra et al., 1996); however, Tat protein appears to be secreted by HIV-infected cells and might then be taken up by the JCV-infected oligodendrocyte and induce transcription on Tat-responsive genes. This thesis was examined by Valle et al. (2000), who demonstrated a localization of accumulated Tat protein to the nuclei of JCV-infected oligodendrocytes.
Recently, PML cases were reported without immunosuppression in the setting of co-infection of another retrovirus, HTLV-1 (Okada et al., 2000; Shimizu et al., 1999). This raised the question of whether HTLV-1 could activate JCV expression by interaction of the transactivating protein, Tax. Tax has been shown to interact with other viruses, such as HIV, CMV, and SV40. In vitro studies revealed a glial cell-specific interaction of Tax with the JCV promoter. Therefore, a comparable interaction to that of HIV-1 Tat protein is conceivable (Okada et al., 2000).
In summary, the detection of amplification products belonging to both poly-omavirus species is strong evidence for concomitant infection in all tissues found positive for polyomavirus DNA. The number of individuals with simultaneous JCV and BKV infection is high, probably reflecting the true incidence of polyomavirus infection in the population. However, changes in virus-associated expression or the histologic picture due to co-infection have never been reported; therefore, an influence on expression activity by transactivation events appears to be rather unlikely. Similarly, co-infection with other viruses has been reported in a large number of patients and healthy individuals. Although heterologous transactivation may come into effect in individual cases, a common interaction among heterologous viral transactivators with human polyomaviruses cannot be stated. Consequently, transactivating mechanisms probably do not play a general role in the control of the polyomavirus life cycle and pathogenesis.
Was this article helpful?