HIV-1 transcription and latency: an update

Abstract

Combination antiretroviral therapy, despite being potent and life-prolonging, is not curative and does not eradicate HIV-1 infection since interruption of treatment inevitably results in a rapid rebound of viremia. Reactivation of latently infected cells harboring transcriptionally silent but replication-competent proviruses is a potential source of persistent residual viremia in cART-treated patients. Although multiple reservoirs may exist, the persistence of resting CD4+ T cells carrying a latent infection represents a major barrier to eradication. In this review, we will discuss the latest reports on the molecular mechanisms that may regulate HIV-1 latency at the transcriptional level, including transcriptional interference, the role of cellular factors, chromatin organization and epigenetic modifications, the viral Tat trans-activator and its cellular cofactors. Since latency mechanisms may also operate at the post-transcriptional level, we will consider inhibition of nuclear RNA export and inhibition of translation by microRNAs as potential barriers to HIV-1 gene expression. Finally, we will review the therapeutic approaches and clinical studies aimed at achieving either a sterilizing cure or a functional cure of HIV-1 infection, with a special emphasis on the most recent pharmacological strategies to reactivate the latent viruses and decrease the pool of viral reservoirs.

Figure 1

Dynamics of plasma virus levels in a cART-treated HIV + individuals. After initiation of cART, viremia undergoes three phases reflecting the decay rates of different populations of HIV-1 latently infected cells. The first phase represents the rapid decay of productively infected CD4+ T cells (activated CD4+ T cells having a half-life of ~1 day). The cells responsible for the second phase, which have a half-life of about 14 days, are not definitively identified (possibly partially activated CD4+ T cells or other cell types such as macropages or dendritic cells). The third phase is a constant phase in which viremia reaches levels below the limit of detection of clinical assays (50 copies viral RNA per ml of plasma). During this plateau phase, occasional viremic episodes (called blips) are detected despite cART. Reservoirs of HIV-1 are responsible for the low but stable level of residual viremia observed during the third phase. This residual viremia is partly derived from the activation of latently infected resting (memory) CD4+ T cells (or subsets of these cells) and partly from another unknown cell source (such as long-lived HIV-infected cells). A rapid rebound of viremia is observed if cART therapy is stopped. Therapeutic strategies achieving control of viremia below detection level after cART cessation could lead to a functional cure. Strategies achieving elimination of HIV-1 from the human body could lead to a sterilizing cure.

Figure 2

Reactivation of HIV-1 transcriptional latency. During latency, nuc-1 blocks transcriptional initiation and/or elongation, Tat is absent and only short mRNAs corresponding to TAR are produced. Nuc-1 is maintained hypoacethylated by HDACs recruited to the 5’LTR via several transcription factor (YY1, CTIP-2, p50-p50 homodimer, CBF-1). The corepressor CTIP-2 interacts with the Sp1 transcription factor at three sites in the HIV-1 5’ LTR and recruits HDACs and the HMT Suv39h1, which trimethylates H3K9 leading to the recruitment of HP1. Other histone methylation repressive marks such as H3K9Me2 or H3K27Me3 catalyzed by the HMT G9a and EZH2, respectively, are also implicated in HIV-1 latency. In addition, during latency, the HIV-1 promoter is hypermethylated at two CpG islands surrounding the HIV-1 transcriptional start site. The dotted arrows indicate that DNMTs are most likely recruited to the HIV-1 promoter but this recruitment has not been demonstrated so far. In latent conditions, the active form of NF-kappaB (p50-p65 heterodimers) is sequestered in the cytoplasm by the inhibitor of nuclear factor kappaB (IκB), while NF-kappaB p50-p50 homodimers occupies the kappaB sites at the viral LTR region. The kappaB sites can also be occupied by CBF-1 and by STAT5Δ/p50 heterodimer in monocytic cells. The phosphorylated form of NFAT is also in the cytoplasm in latency conditions. Moreover, in resting CD4+ T cells, P-TEFb, composed of CDK9 and human cyclin T1, is sequestered in an inactive form by the HEXIM-1/7SK snRNA regulatory complex. In this context, several compounds have been proposed for trasncriptionalreactiation of HIV-1 including HDACIs (SAHA, VPA) to target the hypoacetylated state of nuc-1, HMTIs (chaetocin, BIX-01294, DZNEP) to target HMTs, DNMTIs (5-Aza-CdR) to target 5’LTR DNA methylation, PKC or Akt agonists (sprostratin, bryostatin) to activate the NF-kappaB signaling pathway, cytokines (IL-2, IL-7, GM-CSF) to activate STAT5 and inducers of P-TEFb release (HMBA, JQ1).

Figure 3

Modulation of HIV-1 replication by miRNA. HIV-1 impacts the cellular miRNA pathways in several ways. The figure shows a schematic representation of the HIV-1 provirus with the indicated viral genes and LTRs, including the transcription start site in the 3’LTR. PCAF and Cyclin T1, which are both essential cofactors for viral transcription, are targeted by the indicated miRNAs in resting T cells and monocytes. A cluster of miRNAs is targeted directly towards the 5'end of the HIV-1 genome and their inhibition rescues infectious virus from resting T cells. Finally two sets of miRNA that are derived from the HIV-1 genome are shown. Their location within the LTR (TAR and nef) is duplicated (arrows).