SummaryThe human retrovirus HTLV-1 causes a hematological malignancy or neuroinflammatory disease in ∼10% of infected individuals. HTLV-1 primarily infects CD4+ T lymphocytes and persists as a provirus integrated in their genome. HTLV-1 appears transcriptionally latent in freshly isolated cells; however, the chronically active anti-HTLV-1 cytotoxic T cell response observed in infected individuals indicates frequent proviral expression in vivo. The kinetics and regulation of HTLV-1 proviral expression in vivo are poorly understood. By using hypoxia, small-molecule hypoxia mimics, and inhibitors of specific metabolic pathways, we show that physiologically relevant levels of hypoxia, as routinely encountered by circulating T cells in the lymphoid organs and bone marrow, significantly enhance HTLV-1 reactivation from latency. Furthermore, culturing naturally infected CD4+ T cells in glucose-free medium or chemical inhibition of glycolysis or the mitochondrial electron transport chain strongly suppresses HTLV-1 plus-strand transcription. We conclude that glucose metabolism and oxygen tension regulate HTLV-1 proviral latency and reactivation in vivo.

No abstract

HTLV-1 plus-strand transcription begins with the production of doubly-spliced tax/rex transcripts, the levels of which are usually undetectable in freshly isolated peripheral blood mononuclear cells (PBMCs) from HTLV-1-infected individuals. However, the presence of a sustained chronically active cytotoxic T-cell response to HTLV-1 antigens in virtually all HTLV-1-infected individuals, regardless of their proviral load, argues against complete latency of the virus in vivo. There is an immediate burst of plus-strand transcription when blood from infected individuals is cultured ex vivo. How is the HTLV-1 plus strand silenced in PBMCs? Is it silenced in other anatomical compartments within the host? What reactivates the latent provirus in fresh PBMCs? While plus-strand transcription of the provirus appears to be intermittent, the minus-strand hbz transcripts are present in a majority of cells, albeit at low levels. What regulates the difference between the 5′- and 3′-LTR promoter activities and thereby the tax-hbz interplay? Finally, T lymphocytes are a migratory population of cells that encounter variable environments in different compartments of the body. Could these micro-environment changes influence the reactivation kinetics of the provirus? In this review we discuss the questions raised above, focusing on the early events leading to HTLV-1 reactivation from latency, and suggest future research directions.

The human T cell leukemia virus persists in vivo in 103to 106clones of T lymphocytes that appear to survive for the lifetime of the host. The plus strand of the provirus is typically transcriptionally silent in freshly isolated lymphocytes, but the strong, persistently activated cytotoxic T lymphocyte (CTL) response to the viral antigens indicates that the virus is not constantly latent in vivo. There is now evidence that the plus strand is transcribed in intense intermittent bursts that are triggered by cellular stress, modulated by hypoxia and glycolysis, and inhibited by polycomb repressive complex 1 (PRC1). The minus-strand gene hbz is transcribed at a lower, more constant level but is silent in a proportion of infected cells at a given time. Viral genes in the sense and antisense strands of the provirus play different respective roles in latency and de novo infection: Expression of the plus-strand gene tax is essential for de novo infection, whereas hbz appears to facilitate survival of the infected T cell clone in vivo.

Kulkarni et al.: The reciprocal antagonistic interplay between HIV-1 Tat and SIN3A specifies the epigenetic switch between de-repression versus maintenance of HIV-1 gene silencing. Retrovirology 2013 10(Suppl 1):O12.