Inhibition of HIV-1 replication in primary human monocytes by the IκB-αS32/36A repressor of NF-κB

Background The identification of the molecular mechanisms of human immunodeficiency virus type 1, HIV-1, transcriptional regulation is required to develop novel inhibitors of viral replication. NF-κB transacting factors strongly enhance the HIV/SIV expression in both epithelial and lymphoid cells. Controversial results have been reported on the requirement of NF-κB factors in distinct cell reservoirs, such as CD4-positive T lymphocytes and monocytes. We have previously shown that IκB-αS32/36A, a proteolysis-resistant inhibitor of NF-κB, potently inhibits the growth of HIV-1 and SIVmac239 in cell cultures and in the SIV macaque model of AIDS. To further extend these observations, we have generated NL(AD8)IκB-αS32/36A, a macrophage-tropic HIV-1 recombinant strain endowed to express IκB-αS32/36A. Results In this work, we show that infection with NL(AD8)IκB-αS32/36A down-regulated the NF-κB DNA binding activity in cells. NL(AD8)IκB-αS32/36A was also highly attenuated for replication in cultures of human primary monocytes. Conclusions These results point to a major requirement of NF-κB activation for the optimal replication of HIV-1 in monocytes and suggest that agents which interfere with NF-κB activity could counteract HIV-1 infection of monocytes-macrophages in vivo.


Background
HIV-1 infection is characterized by a long period of clinical latency followed by the development of acquired immunodeficiency syndrome, AIDS. During latency and when viral replication is being controlled in patients treated with antiretroviral therapy, HIV-1 is present in cellular reservoirs and continues to replicate, with each ensuing round of replication giving rise to escape mutants, which further replenish viral reservoirs [1,2]. This grim picture calls for novel targeted therapies for eradicating virus-infected cells and for preventing new infections.
Initial infection in vivo by HIV-1 is thought to occur in CD4-positive, CCR5-positive lymphocytes and monocytes. Accordingly, when HIV-1 envelope protein in its oligomerized g160 form contacts the cell surface receptor a signalling cascade is triggered that results in transcriptional activation of specific gene arrays, such as the inflammatory cytokines IL-1 β, IL-6, IL-8, TNF-α, TGF-β; these cytokines, in turn, function to enhance the transcriptional activity of the proviral long terminal repeat (LTR) promoter [3,4]. This cytokine-driven inflammatory-like setting is mediated molecularly by the NF-κB family of transcription factors [5,6]; thus, it serves to reason that preventing NF-κB activation would attenuate HIV-1 replication. Indeed, the LTR of HIV-1 does contain two tandem NF-κB sites [7] and three repeated Sp1 sites [8] upstream of the TATAA box with an additional NF-κB site located in the 5' untranslated region of viral genome [9]. Both sets of NF-κB sequences enhance HIV-1 transcription in response to various signals [9]. However, the Sp1 sites and TATAA box can redundantly sustain the Tat-mediated transactivation of the HIV-1 LTR in the absence of NF-κB sites [10]. It is controversial whether NF-κB cellular factors are required for the HIV-1 replication. Mutant HIV-1 carrying deletions or base-pair substitutions in the NF-κB enhancer in the LTR have been shown to be either competent or incompetent for replication [11][12][13]. These divergent observations are likely explained by differing cellular contexts, such as primary cells versus immortalized cell lines, and varying levels of cellular activation.
IκB inhibitors regulate NF-κB activity [14]. In response to activating stimuli, IκB proteins become phosphorylated, ubiquinated and degraded by proteasomes. This releases cytoplasmic-sequestered NF-κB to enter the nucleus to activate the transcription of responsive genes [14]. The mutant IκB-αS32/36A is defective for serine 32-and serine 36-phosphorylation and is resistant to proteolysis. IκB-αS32/36A acts as a potent inhibitor of the NF-κBdependent gene transcription, including those from the HIV-1 genome [15]. To verify the requirement of NF-κB in the replication of HIV-1 in primary cells, we previously designed HIV-1 and SIV molecular clones containing the IκB-αS32/36A cDNA positioned into the nef region of the respective viral genome [16,17]. We found that these recombinant viruses were highly attenuated for replication in T cell lines as well as in human and simian PHAactivated peripheral blood mononuclear cells, PBMCs [16,17]. These findings supported an interpretation that in these cellular contexts NF-κB is required for efficient viral replication. We also showed that a recombinant SIV which expressed IκB-αS32/36A inhibitor was also highly replication attenuated in vivo in rhesus macaque [17]. Here, we have extended our analysis of IκB-αS32/36A function in HIV-1 replication to primary monocytes. We report that a macrophage-tropic derivative of NL4-3 strain that expresses the proteolysis-resistant IκB-αS32/36A inhibitor of NF-κB replicated poorly in cultured primary human monocytes.

Attenuation of pNL(AD8)IκB-αS32/36A in primary monocytes
We next analyzed the replication properties of the recombinant HIV-1 genomes in cultured human monocytes from different individuals. Based on normalized amounts of input virus, we found that NL(AD8)IκB-αS32/36A was highly attenuated for replication when compared to NL(AD8) and NL(AD8)IκB-antisense ( Fig. 3 A-B). Accordingly, virus-induced syncitium formation was also strongly inhibited in monocytes infected with NL(AD8)IκB-aS32/36A (Fig. 4 A, B). Taken together, our results underscore a critical contribution of NF-κB to HIV-1 growth in monocytes.

Discussion
Substantial numbers of monocytes are preserved in infected individuals even at later clinical stages of AIDS, when T cell numbers are dramatically reduced. Consistently, in animal models of HIV-1 infection, monocytes are the major reservoir after acute depletion of CD4-positive T cells [19,20]. This indicates that these cells are long last-ing infected moieties that shuttle from mucosal sites to lymph nodes and could function as a major HIV-1 reservoir in vivo. In addition, monocytes are programmed to produce a large amount of inflammatory cytokine, includ-Genome structure and expression of recombinant pNL(AD8)IκB-αS32/36A and pNL(AD8)IκB-antisense molecular genomes Figure 1 Genome structure and expression of recombinant pNL(AD8)IκB-αS32/36A and pNL(AD8)IκB-antisense molecular genomes. Panel A shows the structure of pNL(AD8) derivatives that carry the IκB-αS32/36A-FLAG insert into the nef region in sense (pNL(AD8)IκB-αS32/36A) or antisense (pNL(AD8)IκB-antisense) orientations. Panel B shows the immunoblot analysis using hyperimmune AIDS patient serum of total extracts (10 µg) from 293T cells 24 hours after transfection with the indicated viral plasmids (10 µg). Panel C shows the immunoblot analysis using an anti-FLAG monoclonal antibody of total extracts (10 µg) from 293T cells 24 h after transfection with the indicated viral plasmids (10 µg).
ing IL1-β, IL-6, TNF-α, which are strong inducers of HIV-1 replication [5]. Indeed, HIV-1 envelope binding to CCR5 receptor activates an intracellular signalling cascade that promotes high levels of transcription factors, including NF-κB, which sustain the initial rounds of viral replication and induce the production of inflammatory cytokines which activate surrounding cells to become more susceptible to virus infection [3,4].
Based on the published literature, the role of NF-κB in HIV-1 replication has been controversial [13,16,21]. For instance, the deletion of NF-κB binding sites from HIV-1 Reduced NF-κB DNA binding activity in cells transfected with pNL(AD8)IκB-αS32/36A Figure 2 Reduced NF-κB DNA binding activity in cells transfected with pNL(AD8)IκB-αS32/36A. Panel A shows the NF-κB binding activity of nuclear extracts (5 µg) from 293 T cells transfected with the indicated viral plasmids (10 µg) or were mocktransfected. Panel C shows the Sp1 binding activity of the same nuclear extracts as in panel A. Binding competitions were performed with 100-fold molar excess of the respective unlabelled oligonucleotide. and SIV LTRs [22] has suggested that NF-κB activity may not be required for HIV-1 LTR-directed transcription. Moreover, deletion of NF-κB sequences in the LTR has also been reported not to affect HIV-1 replication in defined cellular settings [11,12]. These latter studies relied on short-term infections of immortalized cells that may not express a physiologic concentration of transcription factors. To address this issue, we have developed a novel HIV-1 strain, NL(AD8)IκB-αS32/36A, which was engineered to express a proteolysis-resistant IκBαS32/36A, and is a strong inhibitor of NF-κB activity. This recombinant virus expresses the envelope of the AD8 strain, a macrophage-tropic virus. Our findings show that NL(AD8)IκB-αS32/36A replication profile is different from that of the NL(AD8)IκB-antisense control. NL(AD8)IκB-αS32/36A failed to produce a productive infection in primary monocytic cells over a thirty-days acute infection (Fig. 3). These results were correlated with a strong inhibition NF-κB activity in NL(AD8)IκB-αS32/ 36A-infected cells (Fig. 2), indicating that in the setting of HIV infection of primary monocytes NF-κB plays a nonredundant role. These results are in agreement with the evidence that IκB-αS32/36A negatively affected the replication of HIV and SIV in PBMC cultures and in monkeys [16,17].
Because IκB-αS32/36A constitutively inhibits NF-κB [15], the potent inhibition of HIV/SIV replication could be due to repression of the NF-κB-dependent activation of HIV/ SIV transcription. However, additional mechanisms might explain the potent inhibition of HIV/SIV replication by IκB-αS32/36A. In this regard, IκB-α regulates the transcriptional activity of NF-κB-independent genes by interacting with nuclear co-repressors, histone acetyltransferases and deacetylases [23,24]. Further studies are required to clarify novel activities of IκB-α in the modulation of the transcriptional machinery. Our results underscore a central role for IκB-α as a potent inhibitor of the replication of HIV-1 in both T cells [16] and monocytes (this study), and point to the NF-κB/IκB network as a suitable target for therapeutic intervention of AIDS.

Conclusions
In this study we have addressed the role of NF-κB/IκB proteins in the replication of HIV-1 in primary human monocytes. We show a strong attenuation in the replication of a macrophage-tropic HIV-1 strain expressing the IκB-αS32/ 36A repressor of NF-κB in primary cultures of human monocytes. These results are consistent with previous evidence of HIV/SIV inhibition by IκB-αS32/36A in PBMCs and in macaques [16,17]. In addition, these findings further support a role of NF-κB inhibitors in blocking HIV-1 replication and suggest novel strategies for the development of anti-viral therapy that targets NF-κB factors.

Transfections and Viral stocks
293T cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% v/v heat-inactivated fetal bovine serum and 3 mM glutamine. Viral stocks were produced by transfecting 293T cells (10 6 ) with viral plasmids (10 µg) using calcium phosphate. Forty hours later, the cell culture supernatant was passed through a 0.45-µm filter and measured for RT activity as previously described [16].

Monocytes cultures and infections
Human monocytes were isolated from PBMC by elutriation, cultured in RPMI, 10% FCS and GMCSF (20 ng/ml) for 48 hours. Infections were performed with viral stocks measured by reverse-transcriptase (RT) activity [16]. Usually, cell cultures (10 5 cells) were infected with 10 5 -10 6 cpm of RT activity. The cell culture supernatants were collected every two days and replaced with fresh medium. The viral production was measured as RT activity in the culture supernatants as previously described [16]. The syncitia formation in cell cultures was evaluated by calculating the average number of syncitia in at least six optical fields.