A mutant Tat protein inhibits infection of human cells by strains from diverse HIV-1 subtypes

HEK 293T (ATCC), TZM-bl [24, 25] and Phoenix-Ampho [26] cell lines were grown in Dubelcco’s modified Eagle’s medium (DMEM; Life Technologies) supplemented with 10% (^v/_v) fetal bovine serum (FBS), penicillin (100 IU/ml) and streptomycin (100 μg/ml) (referred to as DF10 medium). TZM-bl expressing NB-mCh or mCh cell lines were established by transduction of NB-mCh or mCh virus-like particles (VLPs) and then selected by fluorescent activated cell sorter (FACS) for the top 10% of mCherry positive cells by mean fluorescent intensity (MFI).

Peripheral blood mononuclear cells (PBMCs) were isolated from healthy donor’s buffy coat supplied by Australian Red Cross Blood service using Ficoll density gradient centrifugation. CD4⁺ cells were isolated from the PBMCs by using a magnetic-activated cell sorting human CD4⁺ cell isolation kit (Miltenyi Biotec) as per the manufacturer’s instruction. The selected cells were grown in 6 cm tissue culture dishes and stimulated using plates pre-coated with purified anti-human CD3 (clone HIT3a) and anti-human CD28 (clone CD28.2) antibodies (BioLegend) in RPMI medium supplemented with 20% (^v/_v) FBS and 5 ng/ml interleukin-2 (IL-2) (hereafter called RF20 IL-2) for 2 days. All cells were grown at 37 °C in humidified incubators with 5% CO₂.

pSRS11-SF-γC-EGFP was a gift from Axel Schambach and Christopher Baum [27]. pSRS11-SF-γC-NB-mCh or pSRS11-SF-γC-mCh or pSRS11-SF-γC-NB-ZSG1 or pSRS11- SF-γC-ZSG1 construct was made by replacing the enhanced green fluorescent protein gene in pSRS11-SF-γC-EGFP with NB-mCh or mCh or NB-ZSG1 or ZSG1. A proviral plasmid pGCH making HIV-1_NL43 (GenBank accession number AF324493) was previously described [13]. The proviral plasmid pZAC (GenBank accession number JN188292.1) was obtained from Jochen Bodem [28]. The proviral plasmids pELI and pMAL (Los Alamos accession number A07108 and A07116 respectively) were provided by Damian Purcell [29]. The exon tat genes with hemagglutinin epitope were synthesized by GenScript and ligated into pcDNA3.1⁺ plasmid (Thermofisher Scientific).

HIV-1 subtype B, C, D and A/D were produced from pGCH, pZAC, pELI and pMAL proviral plasmids respectively. HEK 293 T cells were grown on a 10 cm plate at ~80% confluency and transfected with 10 μg of each proviral plasmid then incubated for 24 h at 37 °C. On the next day, the transfected cells were washed with 1 x phosphate buffered saline (PBS) and the DF10 media was replaced. The supernatant containing HIV-1 VLPs was collected 48 and 72 h post transfection and the amount of HIV-1 capsid (CA) protein in each supernatant was measured by enzyme-linked immunosorbent assay (ELISA) (Zeptometrix) as recommended by the manufacturer.

NB-mCh or mCh or NB-ZSG1 or ZSG1 VLPs were produced in Phoenix-amphotropic retroviral packaging producer cell line by co-transfection of 7.5 μg of pSRS11-SF-γC vector expressing NB-mCh or mCh or NB-ZSG1 or ZSG1 and 1.5 μg of Gag-Pol expressing plasmid using X-tremeGENE^TM DNA transfection reagent (Roche) in a 10 cm plate. Six hours post transfection, the cells were washed with PBS and the media was replaced. The VLPs were collected 48 and 72 h post transfection and filtered through a 0.45 μm filter.

Cell lysates were made from 5 × 10⁶ NB-mCh or mCh-TZM-bl cells, or from 3 × 10⁶ CD4-NB-ZSG1, CD4-ZSG1 or non-transduced CD4 cells in cell lysis buffer (50 mM Tris HCl pH 7.4, 150 mM NaCl, 1 mM ethylenediaminetetraacetic acid and 1% (^v/_v) Triton X-100). The total protein concentration was measured by a Bradford assay using Bio Rad protein assay (Bio Rad) and equivalent amounts of protein were used for analysis. The blots were stained with a anti-mCherry rabbit antibody (BioVision), a rabbit anti-Tat antibody (Diatheva), a mouse anti-ZsGreen1 (Origene), a rabbit anti-β-tubulin antibody (Sigma Aldrich), or a goat anti-actin antibody (Santa Cruz) as indicated. Appropriate species-specific secondary antibodies conjugated to horse radish peroxidase (HRP) (Cell Signaling Technology) and followed detection by chemiluminescence (BioRad).

Tissue culture dishes (6 cm) were seeded with 5 × 10⁵ TZM-bl cells expressing NB-mCh or mCh and then co-transfected with 1 μg of each subtype Tat plasmid or pCDNA3.1⁺ without an insert and 150 ng of Gaussia luciferase expression plasmid. After 48 h, the cells were washed with PBS and cell lysates were made using Glo Lysis buffer (Promega). Luciferase assays were performed in 96 well white polystyrene microplates as per the manufacturer’s instructions using 10 μl of the cell lysates and Dual-Glo® luciferase substrate (Promega). Luciferase activity in each sample was measured within 20 min by using a luminescence microplate reader and relative values were normalized to Gaussia luminescence in the sample.

Next, 3 × 10⁵ TZM-bl cells expressing NB-mCh or mCh or non-transduced (NT) TZM-bl cells were seeded in 6 well plates. The next day, the cells were infected with HIV-1_NL4–3 (subtype B), HIV-1_ZAC (subtype C) [28], HIV-1_ELI (subtype D) and HIV-1_MAL (A/D recombinant subtype) virus supernatant containing 20 ng of CA, or a mock supernatant for 48 h. The cells were washed with PBS and then cell lysates were made using Glo Lysis buffer (Promega). Luciferase activity was measured as described above.

NB-ZSG1 or ZSG1 VLPs were concentrated using the precipitation method through the addition of 20% (^v/_v) of 34% polyethylene glycol 8000 (Sigma Aldrich) and 10% (^v/_v) of 0.3 M NaCl solution. The solution mixture was incubated at 4 °C for 1.5 h, mixed every 30 min and then centrifuged at 1500 × g for 1 h at 10 °C. The supernatant was discarded and the precipitate was resuspended in 600 μl RF20 IL-2 medium. The concentrated VLP (150 μl) was added to Retronectin (Takara) coated 24 well plate and incubated at 37 °C for 30 min. 5 × 10⁵ stimulated CD4+ cells were added to each well and incubated for 3 days. Transduced cells were processed by FACS to collect ZSG1 positive cells which were grown for 3 days further. The RF20-IL2 media was replaced every day.

TZM-bl cells expressing NB-mCh or mCh or NT (3 × 10⁵ cells/well) cultured in a 6 well plate were infected with a virus stock containing 20 ng CA of HIV-1_NL4–3, HIV-1_ELI and HIV-1_MAL or 40 ng CA of HIV-1_ZAC, or a mock supernatant for 2 h at 37 °C. A larger amount of HIV-1_ZAC was required to yield measurable infections. The virus was then removed by washing the cells 3 times with PBS and the infected cells were cultured at 37 °C with 5% of CO₂. The culture supernatants were sampled on day 3 and 5 post infection. The amount of HIV CA present was measured using a CA ELISA kit (Zeptometrix) according to the manufacturer’s instruction.

Primary CD4⁺ T cells (5 × 10⁵ NB-ZSG1 or ZSG1 or NT) were infected with virus stocks containing 2 ng CA of each HIV-1 subtype for 2 h at 37 °C. After infection, the cells were washed with PBS and cultured in RF20 IL-2 medium. Cell and supernatant samples were collected on days 0, 3, 7, 10 and 14 by centrifugation at 500 × g for 5 min. The amount of viral CA in the supernatant was measured by ELISA. The cells were fixed with 1% paraformaldehyde in PBS solution and NB-ZSG1 or ZSG1 expression was measured by flow cytometry.

Cell metabolic activity was measured by MTS assay using a CellTiter 96® aqueous one solution cell proliferation reagent (Promega) according to the manufacturer’s instructions. Cell proliferation was quantified using a Violet Proliferation Dye 450 (BD Horizon^TM) assay in accordance with the manufacturer’s instructions and violet fluorescence was measured using a violet laser-equipped BD LSRFortessa^TM IV flow cytometer. Apoptosis events were quantified using a PE Annexin V apoptosis detection kits (BD Pharmingen^TM) as per the manufacturer’s instructions. Camptothecin, which induces apoptosis in CD4+ T cells, was used as a positive control.

A schematic diagram of Nullbasic protein and the basic domain mutations (amino acids 49–57) to glycine and an alanine are shown in Fig. 1a. A NB-mCh fusion protein, used previously to investigate Nullbasic inhibition on HIV-1 Rev activity [14], was inserted into a SIN γ-retroviral vector pSRS11-SF-γC (Fig. 1b) [27]. VLPs produced in Phoenix amphotropic packaging cells [30] were used to transduce TZM-bl cells and those stably expressing NB-mCh or mCh (hereafter referred to as TZM-bl-NB-mCh or TZM-bl-mCh, respectively) were collected by FACS. Expression of NB-mCh and mCh was confirmed by flow cytometry analysis of the purified cells (Fig. 1c) and western blot analysis using an anti-mCherry antibody (Fig. 1d).

It is worth reviewing the amino acid sequence heterogeneity of Tat from these different strains by comparing the alignment of Tat proteins from HIV-1_NL4–3 (subtype B), HIV-1_ZAC (subtype C) [28], HIV-1_ELI (subtype D) and HIV-1_MAL (A/D recombinant subtype). The four different amino acid sequences have amino acid residue substitutions in all domains except for amino acids 43 to 56 (Fig. 2, boxed), which includes the basic domain, completely conserved. The first 20 residues of each Tat protein include 17 positions that are conserved or functionally similar. The carboxyl terminal amino acids 90 to 100 of subtype C and D Tat proteins have conserved amino acid sequences KKKVE and ETDP (Fig. 2, underlined), which are also present in Nullbasic (derived from HIV-1_BH10). All four Tat proteins maintained the cysteine residues in the cysteine-rich domain with one exception; HIV-1_ZAC (subtype C) has a C31S substitution.

Tat mediates HIV-1 transactivation by binding to trans-activation response (TAR) RNA in the R region of HIV-1 LTR and recruiting P-TEFb [31] that then binds a super elongation complex (SEC) [32, 33]. P-TEFb consists of cyclin T1 and cyclin dependent kinase 9 (CDK9) [31]. In Tat, K41 is important for intramolecular hydrogen bonding and structural integrity of the Tat core [34] and is present in all Tat proteins shown except HIV-1_ZAC which has T41. A crystal structure of the Tat-P-TEFb complex showed that the surface of 37% amino acids 1–49 are complementary to the kinase complex, and this model indicates that the interactions between Tat and P-TEFb can accommodate substitutions commonly present in different Tat genes [34]. However, Tat interacts with other cellular proteins many of which are important for HIV transcription [35]. Therefore, it is possible that these subtle differences could affect the ability of Nullbasic to inhibit the transactivation by Tat from the HIV-1 strains shown.

To test if transcription by Tat from different HIV-1 strains can be inhibited by Nullbasic, TZM-bl-NB-mCh and TZM-bl-mCh cell lines stably carrying a HIV-1-LTR firefly luciferase reporter were co-transfected with eukaryotic expression plasmids that express either Tat_NL4–3, Tat_ZAC, Tat_ELI, Tat_MAL or an empty expression plasmid and a Gaussia luciferase expression plasmid to control for transfection efficiency. Tat can activate the HIV-1-LTR firefly luciferase reporter in TZM-bl cell lines, while Gaussia luciferase reporter enables linear quantification of relative transfection efficiencies between samples [36]. The amount of firefly luciferase activity present in lysates from the transfected TZM-bl cells was measured and the relative luminescence unit (RLU) values were normalized to Gaussia luciferase RLUs measured in the culture supernatant. The results show that expression of NB-mCh in TZM-bl-NB-mCh cells reduced transactivation of the HIV-1 LTR luciferase reporter in all subtypes tested from ~70 to ~90% (Fig. 3). Although Tat_ZAC showed a trend towards less inhibition than others, this difference did not achieve statistical significance. Transfection of an empty expression plasmid did not affect the level of luciferase made by TZM-bl-NB-mCh cells compared to TZM-bl-mCh cells. Overall, the level of Nullbasic inhibition of the Tat proteins tested here was similar to a previous report [4].

Post infection, the integrated provirus produced Tat can activate the HIV-1-LTR luciferase reporter in TZM-bl cells, which is an indicator of virus infection and replication. To evaluate Nullbasic inhibition of Tat-mediated transactivation of different HIV-1 strains, we infected TZM-bl-NB-mCh and TZM-bl-mCh cells with HIV-1_NL4–3 (subtype B), HIV-1_ZAC (subtype C), HIV-1_ELI (subtype D) and HIV-1_MAL (recombinant A/D subtype) produced in HEK293T cells or with mock supernatant. Lysates prepared from HIV-1 infected TZM-bl cells 48 h post infection were assayed for firefly luciferase activity and the results were normalized to total protein concentration in the cell lysates. NB-mCh significantly reduced the amount of RLUs produced compared to control lysates made from infected TZM-bl-mCh cells. HIV-1_NL4–3 viral Tat transactivation of the LTR-luciferase reporter was inhibited by ~97%; slightly lower inhibitions of transactivation were measured for all other HIV-1 strains; HIV-1_ZAC by ~90%, HIV-1_ELI by ~89% and HIV-1_MAL by ~91% (Fig. 4). However, the difference between transactivation inhibition in HIV-1_NL4–3 and other strains did not achieve statistical significance. The level of luciferase made by mock-infected TZM-bl-NB-mCh cells compared to TZM-bl-mCh cells was unchanged. The combined experiments demonstrate that NB-mCh can inhibit transactivation of the TZM-bl LTR reporter by each Tat tested albeit at a slightly reduced level compared to HIV-1_NL4–3.

Finally, we tested if the presence or absence of NB-mCh could inhibit viral replication (rather than assessing effects on the integrated HIV-1 LTR-luciferase reporter in the cell line) of each HIV-1 strain in the TZM-bl cell lines. In this 5 day experiment, detection of CA by ELISA requires virus replication. Therefore, TZM-bl-NBmCh and TZM-bl-mCh cells were infected, supernatants were collected on day 3 and 5 post infection, and the amount of CA in each sample was by measured by ELISA. As shown in Fig. 5a-d, the data shows that each HIV-1 strain replicated in NT TZM-bl and TZM-bl-mCh cells with an increasing level of CA from day 3 to 5. However, HIV-1 replication of all strains was inhibited in TZM-bl-NBmCh cells. Comparing day 5 CA levels by TZM-bl-mCh to TZM-bl-mCh cells showed the production of HIV-1 CA levels dropped by >99% for HIV-1_NL4–3, ~97% for HIV-1_ZAC, ~98% for HIV-1_ELI, and ~97% for HIV-1_MAL. As previously observed, the trend was that NB-mCH inhibited HIV-1_NL4–3 better than other strains tested. Nevertheless, the combined results support the hypothesis that Nullbasic can inhibit virus replication of the HIV-1 strains tested in TZM-bl cells.

We previously used an MLV-based γ-retroviral vector, pGCsamEN [11], containing NB-ZSG1 or ZSG1 to transduce primary CD4⁺ cells. In those experiments, cells that expressed NB-ZSG1 significantly delayed HIV-1_NL4–3 replication compared to cells expressing ZSG1. However, pGCsamEN is not a SIN vector and expression of a transgene is via the MLV-LTR promoter. Non-SIN γ-retroviral vector can be transcriptionally repressed in cells [37], which can be lessened by SIN-based γ-retroviral vectors that used strong constitutively expressed internal promoters. Therefore, SRS11-SF-γC-NB-ZSG1 VLPs were used to transduce CD4⁺ T cells in this study. The transduced CD4⁺ T lymphocytes were selected by FACS using parameters previously described and a western blot was performed to confirm NB-ZSG1 and ZSG1 expressions (hereafter referred to as CD4-NB-ZSG1 and CD4-ZSG1, respectively) in the sorted cells (see Additional file 1).

CD4-NB-ZSG1, CD4-ZSG1 and NT CD4 cells were infected with each HIV-1 strain and virus replication was monitored for 14 days (Fig. 6). All HIV-1 strains were inhibited but some strain specific differences were noted. For example, HIV-1_NL4–3, HIV-1_ZAC, and HIV-1_ELI were strongly inhibited. At day 14 post infection, in CD4-ZSG1 to CD4-NB-ZSG1 cells, CA levels reduced by 90% for HIV-1_NL4–3, and 88% for HIV-1_ELI and HIV-1_MAL. However, no CA was detected after infection with HIV-1_ZAC. It is worth noting that based on CA expression levels, this HIV-1_ZAC replicated poorly in TZM-bl (Fig. 5b) and CD4⁺ T cells (Fig. 6b). Compared to HIV-1_NL43, HIV-1_ZAC replication in CD4+ T cells on day 14 post infection was ~100-fold lower. A reduced level of HIV-1_ZAC replication compared to HIV-1_NL4–3 was reported previously [28]. Therefore, this may account for a lack of detectable CA by HIV-1_ZAC in CD4-NB-ZSG1 cells.

NB-ZSG1 and ZSG1 expression was monitored in uninfected and HIV-1-infected CD4⁺ T cells by flow cytometry. CD4⁺ T cell populations were initially collected by FACS so that ZSG1 positive cell population was >97%. The percentage of ZSG1 positive cells in the population by 14 days ranged from 90% (for HIV-1_ELI) to 99% (uninfected CD4⁺ T cells) in all experiments, whereas NB-ZSG1 levels were approximately 82% in uninfected and infected CD4+ T cells (Fig. 7). However, NB-ZSG1 expression reported here was much higher than in previous experiments using pGCsamEN vectors where the percentage of CD4⁺ T cells expressing NB-ZSG1 ranged from 40 to 50% of cells [11].

We investigated if high level expression of NB-ZSG1 in CD4⁺ T cells had a measurable detrimental effect that could explain why levels of NB-ZSG1 positive cells declined. First, cell viability was evaluated by an MTS colorimetric assay. This assay measures cellular metabolism based on reduction of the MTS tetrazolium compound by NAD(P)H-dependent oxydoreductase enzymes largely in the cytosolic compartment of dividing cells [38, 39]. No significant difference between CD4-NBZSG1 and CD4-ZSG1 cells were found, although both NB-ZSG1 and ZSG1 cells had a slightly lower metabolic activity than NT cells, suggesting the transduction, cell purification processes or ZSG1 were responsible (Fig. 8a). Second, we measured cell proliferation using a cell permeable VPD450 dye for monitoring of cell division by flow cytometry. No difference in proliferation was observed between CD4-NB-ZSG1 and CD4-ZSG1 (Fig. 8b). Given that HIV-1 wild type Tat is reported to have pro- and anti-apoptotic activities in primary CD4⁺ T cells (reviewed in [40]), an annexin V assay was used to determine if NB-ZSG1 expression affected cellular apoptosis. All of the cells tested had very little or no apoptosis (Fig. 8c), therefore we cannot account for a drop in NB-ZSG1 levels in CD4⁺ T cells due detrimental effects of NB-ZSG1 on the cellular pathways tested. However, Nullbasic may affect other cellular pathways other than those tested leading to reduced levels of Nullbasic expression in the cells.

In summary, here we show that Nullbasic can inhibit replication of HIV-1 strains from different subtypes. The outcome indicates that the replication pathways affected by Nullbasic are most likely shared by these HIV-1 strains.