The long-term interests of our group have been the interactions between invading retroviruses and the regulatory machinery of infected host cells. Heterotransmission of retroviruses, i.e. transfer from their natural host into the foreign species, has been a potent tool for such studies thanks to a broad scale of suppressive effects raised by non-permissive host cells against the non-adapted virus. The barriers against retroviral heterotransmission as well as the mechanisms by which retroviruses can overcome them and adapt to the new host are not well understood yet. The world-wide human immunodeficiency virus type 1 (HIV-1) epidemic is a clear example of previous heterotransmission from non-human primates to human. In addition to the experimenatal approach, a large amount of knowledge of virus-host co-evolution comes from whole-genome studies of human endogenous retroviruses (HERVs) and retrotransposons.

 DNA Methylation of Integrated Retroviruses   

          An extreme example of interclass heterotransmission of Rous sarcoma virus (RSV) from chicken into non-permissive mammalian cells has been a key model for the postulation and evidence of the proviral state and oncogene transduction. Furthermore, RSV displays a handful of striking phenomena in the mammalian host cell, especially efficient transcriptional suppression of integrated RSV proviruses by de novo CpG methylation. Therefore, our experimental system is also interesting from the point of view of epigenetic silencing of integrated retrotransposable elements. Escape from this transcriptional control might, on the other hand, be useful in improving the long-term transcription of retrovirus-based vectors used for the gene transfer. We, therefore, look for the pattern of DNA methylation within proviral long terminal repeats (LTRs), leader sequences and cellular flanking regions, and we try to protect RSV proviruses using antimethylation signals.

          We have shown previously that the RSV LTR is sensitive, at least in mammalian cells, to the transcriptional repression by DNA methylation. Along the same line, junction of a CpG island with RSV LTR resulted in partial resistance to the post-integrative inactivation and anti-methylation protection of the proviral reporter in mammalian cells. We suggest that the DNA methylation may be part of the natural barriers against retroviral transmission between heterologous host species.

          To understand in more detail the mechanisms of de novo DNA methylation and transcriptional suppression of integrated retroviruses, we started to study CpG methylation patterns in proviral LTRs by the bisulphite technique. Two interesting observations resulted from this work: first, after exogenous infection of chicken cells, RSV proviruses lack almost any methylation, whereas in mammalian cells they are heavily methylated and suppressed (manuscript in preparation). This corresponds with a low level of methylation observed within HIV-1 LTR in in vitro infected cells or AIDS patients. Second, in one interesting case of the simplified LTR, v-src, LTR provirus integrated into a hypermethylated genomic region, the provirus overcomes the suppressive constraints of the integration site, keeps its non-methylated status and demethylates the flanking sequences. This is the first example of such target site demethylation by retroviral integration. In addition to the exogenous retroviruses, we are also interested in CpG methylation of HERV-Ws, especially the syncytin, a captive HERV whose functional env is used in the process of syncytiotrophoblast differentiation.

          We are also interested in the pathogenic consequences of persistent infection of avian leukaemia viruses (ALVs), because there are few data for comparison between oncoviral and lentiviral pathogenesis (17). Previously, we have established our inbred flock of domestic ducks as a valuable model for studies on ALV pathogenesis in a semi-permissive host. Recently, we have started a new project of identification of the ASLV-C retrovirus receptor. For the aforementioned projects, we have broadly used our traditional models – RSV in vitro infection of avian vs mammalian cells or in vivo sarcoma induction in chickens. A substantial part of our research was based on our panel of inbred chicken lines, some of which are congenic in the major histocompatibility complex (MHC) (9, 22).

          Whole human genome surveys of HERVs, Alus and long interspersed nuclear elements (LINEs) have been proven as efficient tools in testing some of hypotheses raised by our previous findings on retroviral integration specificity and the effect of integration sites on retrovirus expression. Especially, the analysis of HERV-Ws turned to be fruitful in the current search for the role of HERV-Ws in human development.    

          The last but not least research topic in our group is the characterization of the novel tumour-associated CA IX protein with carbonic anhydrase activity. This is a cell surface protein expressed in a high percentage of certain human carcinomas (renal, cervical, lung, colorectal, etc.) displaying a cell-to-cell adhesive function.  Our interest in CA IX is reasoned by the expectation that it can serve both as a diagnostic marker and a target molecule for therapy.