Host Restriction Factors Modulating HIV Latency and Replication in Myeloid Cells

In addition to CD4+ T lymphocytes, myeloid cells, and, particularly, differentiated macrophages, are targets of the human immunodeficiency virus type-1 (HIV-1) infection via interaction of gp120Env with CD4 and CCR5 or CXCR4. Both T cells and macrophages support virus replication although with substantial differences. In contrast to activated CD4+ T lymphocytes, HIV-1 replication in macrophages occurs in nondividing cells and it is characterized by virtual absence of cytopathicity both in vitro and in vivo. These general features should be considered in evaluating the role of cell-associated restriction factors aiming at preventing of curtailing virus replication in macrophages and T cells particularly in the context of designing strategies to tackle the viral reservoir in infected individuals receiving combination antiretroviral therapy. In this regard, we will here also discuss a model of reversible HIV-1 latency in primary human macrophages and the role of host factor determining restriction or reactivation of virus replication in myeloid cells.


Figure 1. Restriction and transcription factors acting on HIV-1 infection in myeloid cells. See
for details.

Restriction factors and counteracting viral proteins active in myeloid cells
Among other accessory genes encoded by HIV, VpR, a virion-associated protein, has been early characterized as a relevant factor to allow efficient virus replication in macrophages [56,57]. Its mechanism(s) of action has/have been debated for several years and likely does not rely on a single modality. Experimental evidence supports an active role of VpR as transcriptional booster of provirus expression potentially involving its interaction with the intracellular glucocorticoid receptor [58] followed by its translocation from cytoplasm into the cell nucleus and leading to cell cycle arrest in G2/M phase [59][60][61]. The importance of VpR in HIV pathogenesis is supported by in vivo studies in NHP [62] and in HIV-1+ long-term nonprogressors (LTNP) infected with VpR-defective viruses [63] as well as by a rare case of human infection in a laboratory setting [64], as discussed in detail in [26]. It has been suggested that VpR could cooperate with Vif in the interaction with A3G leading to its proteasomal degradation [65].
In addition to A3G, other members of the APOBEC family have shown restrictive effects on HIV-1 infection. This aspect of their biology has been recently revisited for its relevance to HIV infection and the role of restriction factors, with particular regard to SAMHD1 (SAM domain and HD domain-containing protein 1), a hydrolase processing deoxynucleotides triphosphates (dNTPs) physiologically involved in DNA repair mechanisms [13]. SAMHD1 acts by depleting the pool of dNTPs that are the "building blocks" necessary to the reverse transcription process in order to synthesize viral DNA before its integration into the host cell genome. Vpx is an accessory protein of the HIV-2, but not of the HIV-1 genome, that targets SAMHD1 for proteasomal degradation thus allowing efficient virus replication [75,76]. As HIV-1 is devoid of Vpx it has been highly debated whether a similar mechanism of SAMHD1 inactivation would be expressed by other accessory viral proteins. in this regard, Ferreira and colleagues have reported that, although in the absence of cell division [13], macrophages in G0 phase express p21/Waf1, previously shown to represent a negative regulator of virus replication in macrophages [41], together with high levels of SAMHD1 thereby resulting in a highly restricted state in terms of virus replication. Their switch to G1 has been associated with downregulation of p21/Waf1, increased expression of cyclin dependent kinase 1 (CDK1) and inactivation of SAMHD1 by means of its phosphorylation leading to increased levels of dNTPs and unleashing of virus replication. This pathway was reverted by conditions such as genotoxic stress and response to "danger signals" that favored a back-transition from G1 to G0, as reviewed in [13].
The IFN-inducible human myxovirus resistance protein B (MxB), related, but distinct from MxA, has been recently described as restriction factor for HIV-1 infection by acting after the completion of the reverse transcription process but before proviral integration likely trough interaction with capsid-cyclophilin A [77][78][79][80].
A well-described feature of HIV-1 replication in macrophages is the lack of induction of a robust type 1 IFN response upon infection, likely explained by the activity of the TREX1 exonuclease [81,82], although a modest induction of IFN induced genes (ISG) can be observed upon infection [83,84]. In the absence of a specific cytosolic sensor, it has been hypothesized that the perturbation of the plasma membrane caused by the interaction with HIV-1 virions would be sufficient to trigger a canonical activation of a STimulator of INterferon Genes (STING)-dependent pathway involving the IFN-responsive factor 3 (IRF-3) [85,86]. In addition, a second wave of ISG expression has been reported to occur a few days after infection following proviral integration and synthesis of the viral regulatory protein Tat, as Once integrated as proviral DNA, the expression of the HIV genome falls under the control of both viral and host factors influencing its transcription, RNA splicing and export from the nucleus to the cytoplasm, translation into viral proteins and their assembly with full-length viral RNA at the plasma membrane to generate new progeny virions. The role of negative regulators of proviral transcription (described in Table 2) will be discussed later.
The last steps of the viral life cycle, namely the budding and release of new virions, is a target of modulation by several RF, as earlier described in the case of Tetherin/BST-2, an IFN-inducible tetraspanin that keeps the virions stuck at the cell surface and not released; its action is counteracted by the viral accessory protein Vpu that promotes its degradation [87,88]. In addition to IFN, also the presence of the viral accessory protein Nef has been associated with increasing levels of Tetherin expression in macrophages [89]. In addition, proteins of the IFITM family (in particular,

IFITM1, 2 and 3) interfere with the release of new virions by inserting themselves into the Env of nascent virions thereby
impairing cell fusion in models of cell-to-cell viral spreading [90,91]. Of note is the fact that their restriction seems particularly effective in macrophages [92]. Furthermore, two recently identified restriction factors (membrane- SERINC3 and SERINC5 are host molecule that significantly decreases virion infectivity by interfering with the fusogenic properties of HIV-1 Env glycoproteins whereas the accessory viral protein Nef prevents their incorporation into nascent virions at the plasma membrane [53,54]. Their effect has been mostly studied in cell lines, but although they seem to be dispensable for virus replication in primary activated CD4+ T cells, recent studies have described its relevance particularly for primary macrophages, although with significant inter-donor variability [94]. Of interest, SERINC5 expression is upregulated during differentiation of monocytes into MDM [95] and its incorporation into virions has been linked to the upregulated production of pro-inflammatory cytokines, including tumor necrosis factor- (TNF-), IL-6 and others, an effect that was also prevented by Nef [94]. SerinC5-induced release of pro-inflammatory cytokines by MDM occurred in synergy with the CCR5 antagonist Maraviroc that blocks virions bound to CD4 on the plasma membrane [96]. In this regard, SERINC5-expressing virions showed a greater susceptibility to inhibition by either Maraviroc or neutralizing antibodies [97]. Thus, in addition to its direct antiviral effect, SERINC5 incorporation into virions may serve as "danger signal" to the infected cells to trigger or enhance a pro-inflammatory response finalized to counteract the infection.
A restriction factor lately emerged as capable of influencing the capacity of HIV to infect cells is the IFN-inducible cholesterol 25-hydroxylase that produced oxysterols, mediators of several process including inflammation and immune activation [98,99]. In addition to HIV, this ER-associated enzyme can limit the infectivity of several enveloped TRIM22, also known as Staf50, is an ISG whose expression is profoundly upregulated by IFN stimulation of different cell types [38]. Our group described it as the key factors differentiating U937 cell clones with a restrictive phenotype in terms of supporting HIV-1 replication ("Minus clones") in comparison to those fully permissive (Plus clones) [105]. TRIM22 does not possess a DNA binding domain and acts, at least in part, by preventing the binding of Sp1, a positive transcription factor constitutively expressed by many cell types, to the core promoter of the HIV-1 provirus [106]. CIITA was originally described to be the key transcription factor for MHC Class II antigen expression under the control of IFN-, as reviewed [44]. It was then demonstrated to play a significant role in the inhibition of both HTLV-1/HTLV-2 transcription and by competing with the regulatory protein Tat for binding to P-Tefb (a complex formed by Cyclin t1 and CDK9) in order to enhance HIV-1 proviral transcription also in myeloid cells [107].
Therefore, restimulation of HIV-1 infected M1-MDM by M1 cytokines induced both positive and negative restriction factors with the latter dominating over the former (Figure 2). Therefore, M1 2 -MDM could represent a potential model to investigate the hypothesis that, together with CD4+ T cells, also myeloid cells including macrophages could contribute to the establishment and maintenance of the viral reservoir resistant to cART. Indeed, cocultivation of

Conclusions
HIV-1 infection of CD4+ cells involves primarily a prominent subset of T lymphocytes and myeloid cells. These latter encompass very different cell types, such as myeloid dendritic cells, circulating monocytes, eventually extravasating to become MDM in inflammatory conditions, and TRM that acquire very distinctive features according to the anatomical site (from Kupffer cells in the liver to the microglia in the CNS). Although the prominent role of latently infected CD4+ T cells in establishing and diversifying the HIV reservoir of cells carrying replication-competent proviruses has been well established, a growing number of evidence indicates that myeloid cells could also contribute significantly to this unsolved issue preventing the eradication of HIV-1 infection from cART-treated individuals. The identification of specific extracellular and intracellular factors influencing the susceptibility of myeloid cells to become targets of either latent or productive infection could be a crucial goal in order to define effective strategies aiming at the curtailment of the HIV-1 reservoir or to its definitive silencing.
Author Contributions: GP designed the overall structure of the review; IP, SG and EV contributed to the literature analysis concerning restriction factors, PD contributed to the optimization of the M1-MDM model of HIV restriction. All authors reviewed and concurred with the hypotheses and interpretations of this article.