|Home | About | Journals | Submit | Contact Us | Français|
Here we show potent inhibition of HIV-1 replication in a human T cell line and primary human CD4+ cells by expressing a single antiviral protein. Nullbasic is a mutant form of the HIV-1 Tat protein that was previously shown to strongly inhibit HIV-1 replication in nonhematopoietic cell lines by targeting three steps of HIV-1 replication: reverse transcription, transport of viral mRNA, and trans-activation of HIV-1 gene expression. Here we investigated gene delivery of Nullbasic, using lentiviral and retroviral vectors. Although Nullbasic could be delivered by lentiviral vectors to target cells, transduction efficiencies were sharply reduced primarily because of negative effects on reverse transcription mediated by Nullbasic. However, Nullbasic did not inhibit transduction of HEK293T cells by a murine leukemia virus (MLV)-based retroviral vector. Therefore, MLV-based virus-like particles were used to transduce and express Nullbasic-EGFP or EGFP in Jurkat cells, a human leukemia T cell line, and Nullbasic-ZsGreen1 or ZsGreen1 in primary human CD4+ cells. HIV-1 replication kinetics were similar in parental Jurkat and Jurkat-EGFP cells, but were strongly attenuated in Jurkat-Nullbasic-EGFP cells. Similarly, virus replication in primary CD4+ cells expressing a Nullbasic-ZsGreen1 fusion protein was inhibited by approximately 8- to 10-fold. These experiments demonstrate the potential of Nullbasic, which has unique inhibitory activity, as an antiviral agent against HIV-1 infection.
Since the beginning of the epidemic, HIV-1 has infected more than 60 million people and killed more than 25 million (UNAIDS, 2009). The primary therapy used to treat HIV-1 infection and AIDS is combinatorial antiviral chemotherapy, using drugs that act by inhibiting at least two steps of virus replication. Although this therapy has been incredibly effective at limiting virus replication and disease progression, challenges remain when treating chronic infections, including the inability to eliminate virus infection. In the absence of a potent vaccine to prevent HIV-1 infection, strategies to eliminate infection are under investigation. Alternatively, the cure of HIV-1 infection in the “Berlin patient” has raised interest in alternative genetic therapies capable of introducing an HIV-1-resistant immune system to people living with HIV-1 and AIDS (Hutter et al., 2009).
Among the approaches tested to render human immune cells resistant to HIV-1, protein-based antiviral inhibitors that act by targeting viral proteins as dominant (or trans-dominant) negative proteins or engineered innate cellular inhibitors have been described. A dominant negative protein is typically an altered form of a protein that can inhibit the normal function of its wild-type counterpart. Dominant negatives for many viral proteins have been described including Tat (Green et al., 1989; Pearson et al., 1990; Modesti et al., 1991), Rev (Bevec et al., 1992; Malim et al., 1992), Vif (Walker et al., 2010), Nef (Fackler et al., 2001), and Gag (Trono et al., 1989; Furuta et al., 1997; Lee et al., 2009; Checkley et al., 2010). For example, engineered Tat proteins with altered basic domains possess trans-dominant negative phenotypes against wild-type Tat function in transcription (Pearson et al., 1990), and a trans-dominant-negative Rev mutant called RevM10 has been described that was shown to inhibit wild-type Rev function (Malim et al., 1991). The RevM10 mutant protein is dominant negative because it retains the ability to bind the HIV-1 Rev response element (RRE) but is unable to promote CRM1 (chromosome region maintenance 1)-mediated export of the viral mRNA from the nucleus, thereby inhibiting HIV-1 replication (Malim et al., 1991; Stauber et al., 1998). RevM10 mutant, the first trans-dominant negative protein to undergo a pilot clinical trial in HIV-1-infected patients (Woffendin et al., 1996; Ranga et al., 1998), was beneficial to cell survival in the context of HIV infection (Podsakoff et al., 2005). A human engineered TRIM5 (tripartite motif 5)–cyclophilin protein introduced into primary CD4+ T cells and macrophages provided extraordinary protection from HIV-1 infection in vitro and in immune-compromised mice (Neagu et al., 2009). TRIM5–cyclophilin inhibits HIV-1 infection in the newly infected cell by binding to capsid protein (CA), leading to downregulated viral reverse transcription. In summary, various antiviral proteins have been described, each of which targets a single step of the virus life cycle, often by targeting a single viral or cellular protein.
We described a mutant of the two-exon HIV-1 Tat protein, termed Nullbasic, that inhibited multiple steps of the HIV-1 replication cycle (Meredith et al., 2009), a unique feature compared with other antiviral proteins. Nullbasic was created by replacing the entire arginine-rich basic domain of wild-type Tat with glycine/alanine residues. Like similarly mutated one-exon Tat mutants, Nullbasic exhibited dominant negative effects on Tat-dependent HIV-1 gene expression. However, unlike previously reported mutants (Green et al., 1989; Pearson et al., 1990; Modesti et al., 1991; Ulich et al., 1996), Nullbasic inhibited reverse transcription and also effectively suppressed the steady state levels of unspliced and singly spliced viral mRNA, an activity caused by inhibition of HIV-1 Rev activity (Meredith et al., 2009). HeLa CD4+ cells constitutively expressing Nullbasic were strongly protected from a spreading infection by HIV-1 (Meredith et al., 2009). This study investigated whether Nullbasic could be used to effectively protect human T cells from HIV-1 infection. We show that Nullbasic can provide strong and sustained protection from HIV-1 infection.
HEK293T, Phoenix-A (also known as ΦNX-A) (Swift et al., 2001), and PG13-myc-αerb-CD28ζ (von Kalle et al., 1994; Teng et al., 2004) cells were grown in Dulbecco's modified Eagle's medium (DMEM; Invitrogen, Carlsbad, CA) supplemented with penicillin (100U/ml), streptomycin (100μg/ml), and 10% (v/v) fetal bovine serum (FBS) (referred to as DF10 medium). HeLa and Jurkat cells were grown as described previously but using RPMI 1640 medium supplemented as described previously (RF10 medium). All cells were grown at 37°C in humidified incubators, using 5% CO2.
Peripheral blood mononuclear cells (PBMCs) were isolated from buffy coat supplied by the Australian Red Cross Blood Service. CD4-positive cell selection was performed with human CD4 MicroBeads and magnetic separation columns according to the instructions of the manufacturer (Miltenyi Biotec, Bergisch Gladbach, Germany). Isolated cells were determined to be >98% CD4+ by flow cytometric analysis. CD4+ cells were incubated in RPMI 1640 containing 20% FBS (RF20) and human interleukin-2 (IL-2, 20 units/ml; Roche, Indianapolis, IN) at 1 million cells per milliliter in flasks precoated with purified anti-human CD3 (clone OKT3) and purified anti-human CD28 (clone CD28.2) antibodies (BioLegend, San Diego, CA). After 3 days, cells were transferred to an uncoated flask and grown in RF20 and IL-2 as previously described.
HIV-1pNL4-3 and HIV-189.6 viral stocks were made by transfection of proviral DNA plasmids in HEK293T cells (Adachi et al., 1986; Doranz et al., 1996). All viral stocks were stored in small aliquots at −80°C.
All oligonucleotide primers used are shown in Supplementary Table S1 (supplementary data are available online at www.liebertonline.com/hum). pLOX-CW-Nullbasic-EGFP, pLOX-CW-EGFP, pCMVΔR8.91, and pCMV-VSV-G plasmids used for production of lentiviral VLPs (virus-like particles) have been described elsewhere (Zufferey et al., 1997; Salmon et al., 2000). The pCDN3.1-Nullbasic-FLAG plasmid was previously described (Meredith et al., 2009). The plasmid pGCsamEN has been described (Treisman et al., 1995). The pGCsamEN-Nullbasic-EGFP construct was made by PCR of Nullbasic-EGFP from pcDNA3.1-Nullbasic-EGFP (Meredith et al., 2009), using primers P1 and P2, which introduce unique NotI and SnaBI restriction sites compatible with the multiple cloning site of pGCsamEN (Supplementary Table S1). Similarly, pGCsamEN-EGFP was made by PCR of an EGFP cassette from pCDN3.1-EGFP, using primers P2 and P3, and pGCsamEN-ZsGreen1 was made by PCR of ZsGreen1 from pLVX-IRES-ZsGreen1 (Clontech, Mountain View, CA), using primers P4 and P5. A precursor plasmid, pcDNA3.1-Nullbasic-ZsGreen1 was constructed by PCR of pLVX-ZsGreen, using primers P6 and P7, which introduce unique XhoI and XbaI restriction to replace EGFP in pCDN3.1-NB-EGFP with ZsGreen1. Next, primers P1 and P5 were used to generate the Nullbasic-ZsGreen1 DNA, which was then ligated into the NotI/SnaBI sites of pGCsamEN to make pGCsamEN-Nullbasic-ZsGreen1.
HEK293T cells grown in 10-cm dishes were cotransfected with 15μg of pCMVΔR8.91 plasmid, 5μg of pCMV-VSV-G, and 5μg of pLOX-CW-Nullbasic-EGFP or pLOX-CW-EGFP, using a calcium phosphate method as previously described (Fish and Kruithof, 2006). After 16hr, the transfected cells were incubated in 10ml of DF10 medium and 48hr later lentiviral VLP stocks were harvested and stored in aliquots at −80°C. The HIV-1 CAp24 concentration in lentiviral VLP supernatants was measured with an HIV-1 p24 ELISA kit (Zeptometrix, Buffalo, NY).
PG13-myc-αerb-CD28ζ cells produce retroviral VLPs as previously described (von Kalle et al., 1994; Teng et al., 2004). This cell line was transduced secondarily with CAp24 (100 or 500ng) containing lentiviral VLPs conveying EGFP or Nullbasic-EGFP, respectively, in hexadimethrine bromide-coated 24-well plates. Transduced cells expressing high levels of EGFP were isolated with a FACSAria (BD Biosciences, San Jose, CA), and the purified cells were expanded. Flow cytometry was used to monitor EGFP or Nullbasic EGFP expression. To make ΦNX-A VLPs, cells were transfected with 10μg of pGCsamEN-EGFP, pGCsamEN-Nullbasic-EGFP, GCsamEN-ZsGreen1, or GCsamEN-Nallbasic-ZsGreen1. Where indicated, 5μg of pCDNA3.1-Nullbasic-FLAG or the parental plasmid was cotransfected. VLPs were collected after 48hr and stored in aliquots at −80°C. VLP titers were determined according to a previously described method (Fassati, 2009), and reverse transcriptase (RT) levels for all VLP and HIV-1 stocks were determined by RT colorimetric assay (Roche) according to the manufacturer's instructions.
Lentiviral VLP stock containing 1μg of CAp24 was treated with DNase I (Worthington Biochemicals, Lakewood, NUJ) and pelleted through 1×phosphate-buffered saline (PBS) containing 20% (v/v) sucrose at 100,000×g for 2hr, using an SW 41 Ti rotor (Beckman Coulter, Brea, CA). The VLP pellet was washed with 1×PBS (137mM NaCl, 2.7mM KCl, 4.3mM Na2HPO4, 1.47mM KH2PO4; pH 7.4) and resuspended in 500μl of 1×PBS overnight at 0°C. The concentrated VLP stock was used for RNA packaging and endogenous reverse transcription (ERT) assays (described below). To measure VLP RNA, VLPs containing 200ng of CAp24 were solubilized with TRIzol (Life Technologies, Carlsbad, CA) according to the manufacturer's instructions and a synthetic kanamycin RNA (10ng; Promega, Madison, WI) was added to the mixture before isolation of RNA. RT-PCR was performed with SuperScript III RT and random hexamer oligonucleotides. The synthesized DNA was analyzed by qPCR in triplicate, using primers P8 and P9. A serially diluted pLOX-CW plasmid was used to determine VLP cDNA levels, using primers P8 and P9. Primers P10 and P11 were used to determine kanamycin cassette RNA recovery and RT-PCR by the comparative CT method (Schmittgen and Livak, 2008).
The ERT assay is described in detail elsewhere (Warrilow et al., 2008). ERT reactions were performed in triplicate with concentrated VLP stocks that were normalized to RT. A no-dNTP ERT reaction was always included. Purified viral cDNA levels were measured by real-time qPCR using primers P8 and P9.
Expression of EGFP and Nullbasic-EGFP was confirmed by Western blot using lysates from cells prepared and electrophoresed as previously described (Meredith et al., 2009). Lysates were equalized for total protein concentration. Nullbasic was detected with a rabbit anti-Tat antibody (Diatheva, Fano, Italy). EGFP was detected with a rabbit anti-GFP antibody (Cell Signaling Technology, Danvers, MA). Endogenous β-tubulin was detected with a mouse anti-β-tubulin antibody (Sigma-Aldrich, St. Louis, MO). Primary antibodies were detected with horseradish peroxidase (HRP)-conjugated goat anti-mouse or anti-rabbit antibodies (Life Technologies).
Lentiviral and retroviral VLP stocks (1ml) normalized to RT activity were added to hexadimethrine bromide (Sigma-Aldrich)-coated 24-well plates and centrifuged at 1000×g. For ΦNX-A VLP transductions, VLP stocks (2ml) normalized to RT activity were applied to plates coated with RetroNectin (TaKaRa Bio, Otsu, Japan) according to the manufacturer's instructions. Cells were added (1.5×105 HEK293T, HeLa, or Jurkat cells; 3.5×105 activated primary CD4+ cells) in 2×medium as indicated. ZsGreen1 and EGFP expression was measured 48hr posttransduction with a FACSCalibur flow cytometer (BD Biosciences). Expression of myc-αerb-CD28ζ was determined by staining with a mouse anti-Myc–Alexa Fluor 488 monoclonal antibody (Life Technologies) and flow cytometry. Where indicated, transduced cells were collected after 72hr with a FACSAria (BD Biosciences). EGFP and ZsGreen1 expression in cells was monitored by flow cytometry with a FACSCalibur (BD Biosciences). Cells collected by fluorescence-activated cell sorting (FACS) were incubated for 48hr before infection with HIV-1.
All infections were performed in triplicate, using 4×105 Jurkat cells or 1×106 primary CD4+ cells. Cells were infected with HIV-1pNL4-3 or HIV-189.6 containing 20ng of CAp24 in 24-well plates in a volume of 500μl of Opti-MEM I with 5% FBS at 37°C for 2hr. After infection, the medium was replaced with RF10 or RF20 with IL-2, respectively. Culture supernatant and cells were collected as indicated and assayed with an HIV-1 CAp24 ELISA kit (Zeptometrix); EGFP and ZsGreen1 expression was measured by flow cytometry.
The CellTiter 96 AQueous One Solution cell proliferation assay was used following the manufacturer's instructions. CellTiter 96 AQueous One Solution reagent was directly to a sample of culture supernatant incubating for 2hr and then recording absorbance at 490nm with a 96-well plate reader.
VLPs equivalent to 0.6μg of CAp24 were treated with DNase I at 37°C for 1hr before addition to a 60-mm dish of 239T cells. Where indicated, VLPs were inactivated by heating at 70°C for 10min. If the HIV RT inhibitor nevirapine was used, cells were incubated with a final concentration of 200μM nevirapine for 2hr at 37°C before transduction. Transduction was carried out at 4°C for 1hr and then at 37°C for 3hr. The cells were washed three times with PBS containing 0.5mM EDTA and lysed in 0.5ml of hypotonic buffer (10mM HEPES [pH 7.9], 1.5mM MgCl2, 10mM KCl, and 1mM dithiothreitol [DTT]) for DNA PCR, or lysed in TRIzol reagent for RNA extraction. To isolate DNA from cytoplasmic lysate, the washed cells were incubated in hypotonic buffer for 10min at 4°C and then lysed with a 2-ml homogenizer (20 strokes). The lysate was centrifuged at 7500×g for 20min at 4°C to remove nuclei and the supernatant was assayed by qPCR using primers targeting the EGFP region. Nucleic acid concentrations of lysates were normalized by PCR amplification of the mitochondrial DNA-encoded cytochrome c oxidase subunit II gene, using oligonucleotide primers P12 and P13. To extract RNA, the washed cell pellet was lysed in 0.5ml of TRIzol reagent and the RNA was obtained by passage through a Direct-zol RNA MiniPrep column (ZYMO Research, Irvine, CA). Kanamycin RNA (100ng) was added to each lysate to monitor the purification process and analysis. In-column DNase I treatment was carried out before a final wash to remove DNA contaminations. The RNA was eluted in 35μl of water and used to make cDNA, using random primers, followed by qPCR. A cDNA reaction lacking reverse transcriptase was always included to monitor for DNA contamination.
Lentiviral vectors are commonly used to deliver genes to human cells. As many lentiviral vectors are derived from HIV-1, we sought to determine whether Nullbasic negatively affected lentiviral virus-like particle (VLP) production or infectivity as with HIV-1 (Meredith et al., 2009). A Nullbasic-EGFP or EGFP gene cassette was inserted in pLOX-CW and cotransfected with pCMVΔ8.91 and pCMV-VSV-G in HEK293T cells to make infectious lentiviral VLPs (Fig. 1a). Supernatant were recovered after 48hr and assayed for capsid protein, reverse transcriptase (RT), and packaged mRNA made by pLOX-CW vectors (Fig. 1b and c). We observed no significant difference in the level of VLP production, indicating that Nullbasic-EGFP expression did not interfere with viral protein expression or particle assembly. An equivalent amount of Nullbasic-EGFP or EGFP viral supernatant, normalized to total capsid protein, was used to infect three human cell lines: Jurkat, HEK293T, and HeLa cells (Fig. 2). Sharp decreases in the levels of EGFP-positive cells were observed for all cell lines transduced with Nullbasic-EGFP compared with EGFP VLPs irrespective of the target cell line. This outcome was consistent with our observation that Nullbasic reduced HIV-1 infectivity when viral particles were made by cells expressing Nullbasic or Nullbasic-EGFP, primarily by having a pronounced negative effect on HIV-1 reverse transcription (Meredith et al., 2009). To determine whether Nullbasic had a similar effect on lentiviral particles, endogenous reverse transcription (ERT) reactions were performed with equivalent amounts of viral supernatant normalized to RT levels as measured in a standard homopolymeric primer:template RT assay. As shown in Fig. 3a, pLOX-NB-EGFP VLPs produced approximately 4-fold less negative-strand strong stop DNA than pLOX-EGFP VLPs, suggesting that Nullbasic decreased the infectivity of VLPs largely by inhibiting reverse transcription. Considerably less viral DNA was detected if nucleotides were excluded from the ERT reactions, indicating DNA synthesis was de novo. To confirm that the reduced amount of ERT DNA was not somehow influenced by the Nullbasic-EGFP gene cassette, lentiviral VLPs were made using pLOX-CW-EGFP that was cotransfected with an expression vector containing Nullbasic-FLAG or the empty vector pCDNA3.1 as a control (Fig. 3b). ERT assays were performed with viral supernatant collected 48hr posttransfection and normalized to RT concentration. As shown in Fig. 3b, lentiviral VLPs conveying EGFP that were made in cells expressing Nullbasic-FLAG also had a significantly reduced ability to undergo reverse transcription, supporting the observation that Nullbasic-EGFP could inhibit reverse transcription by VLPs. As shown previously, much less viral DNA was detected if nucleotides were excluded from the reactions, indicating DNA synthesis was de novo, and Western blot analysis confirmed that Nullbasic-FLAG was expressed by the transfected cells (Fig. 3c). We also tested the comparative transduction efficiency and reverse transcription capacity of VLPs conveying Nullbasic-EGFP or EGFP in HEK293T cells that were transduced in parallel so that reverse transcription products and total lentiviral RNA in cells could be isolated and quantified. As shown in Fig. 4a, a 4- to 5-fold defect in reverse transcription was observed after transduction with HIV-1-based lentiviral VLPs conveying Nullbasic-EGFP compared with EGFP. Reverse transcription was strongly reduced if the cells were treated with nevirapine or if VLPs were heat-treated, indicating that cDNA synthesis was de novo. However, we observed no significant difference in lentiviral RNA levels in cells, indicating that transduction efficiencies were similar (Fig. 4b). Interestingly, total RT activity was inhibited by a minor but statistically significant amount as measured with a homopolymer primer:template [a poly(A) and oligo(dT)-based] system. However, a much larger reverse transcription defect was observed in ERT or cell transduction experiments, suggesting that Nullbasic affects the viral reverse transcription complex rather than RT per se. We conclude that Nullbasic can inhibit reverse transcription of lentiviral particles, resulting in reduced infectivity by lentiviral VLPs derived from HIV-1. Therefore standard lentiviral vectors conveying Nullbasic are not suitable for gene delivery to human cells.
To investigate whether Nullbasic could inhibit MLV-based retroviral vectors, a PG13-myc-αerb-CD28ζ cell line was transduced with lentiviral VLPs conveying Nullbasic-EGFP or EGFP (Fig. 5a). PG13-myc-αerb-CD28ζ cells constitutively produce MLV-based VLPs that, after transduction, convey a Myc-tagged αerb-CD28ζ cell surface protein to target cells. As shown in Fig. 5b, EGFP-positive cells were collected by FACS and new PG13-myc-αerb-CD28ζ cell lines stably expressing EGFP were established. We should point out that to compensate for reduced infectivity of lentiviral VLPs with NB-EGFP, a larger volume of pLOX-CW-NB-EGFP VLPs was used to obtain the high levels of transductant cells shown in Fig. 5b (bottom). Flow cytometric analysis of the purified cells indicated that >95% of the cells were positive for EGFP or NB-EGFP (data not shown). Western blot analysis of lysates made from each cell line confirmed expression of Nullbasic-EGFP or EGFP in the cells (Fig. 5c). We noted that the two protein bands detected by an anti-Tat antibody were most likely a result of posttranslational modification known to occur on Tat such as ubiquitination, phosphorylation, and acetylation. Using supernatants containing equivalent amounts of total RT, Jurkat cells were transduced and analyzed by flow cytometry for the presence of an extracellular Myc-epitope tag. No significant difference in the number of transduced cells was observed (Fig. 5d), indicating that expression of Nullbasic in PG13-myc-αerb-CD28ζ cells did not inhibit the infectivity of MLV-based VLPs that were produced.
The Nullbasic-EGFP or EGFP gene cassette was inserted into the pGCsamEN (Chuah et al., 1995) retroviral vector and transfected into the MLV-A VLP-producing cell line ΦNX-amphotropic (Fig. 6a) (Swift et al., 2001). VLP-containing supernatant was obtained and relative titers were determined by serial dilution transduction of HeLa cells. We routinely obtained supernatants containing from 1×105 to 4×105 EGFP CFU/ml irrespective of the pGCsamEN construct used. The transduced Jurkat cells underwent FACS, with the EGFP-positive cell population collected as indicated (Fig. 6b). Subsequent flow cytometric analysis showed the cell populations were >97% positive for EGFP or NB-EGFP. As expected, Western blot analysis of lysates made from the cells showed that Nullbasic-EGFP or EGFP was expressed (Fig. 6c). For each infection, 106 Jurkat-Nullbasic-EGFP, Jurkat-EGFP, and parental cells were each incubated with HIV-1pNL4-3 (equivalent to 20ng of CAp24) for 2hr. The virus was removed and the infected cells were incubated for 4 weeks, with twice-weekly sampling of the culture medium; the samples were assayed for HIV-1 capsid protein by ELISA. As shown in Fig. 7, similar virus replication kinetics were observed in parental Jurkat and Jurkat-EGFP cells, with peaks on days 21 and 17, respectively. Virus replication was consistently inhibited in Jurkat-NB-EGFP cells, in which only low levels of CAp24 (<500pg/ml) were detected after 3 weeks. This result was consistently observed in three independent experiments. We also noted steady expression of NB-EGFP in Jurkat cells, suggesting that the protein had little or no deleterious effects on the cells. This is in sharp contrast to HIV-1 Tat, which can accelerate HIV-1 replication and induce cytotoxic effects in human T cells (Caputo et al., 1990; Benjouad et al., 1993). We conclude that expression of NB-EGFP is well tolerated and has potent antiviral effects that can sharply delay virus replication in Jurkat cells.
Transduction experiments were performed to determine whether Nullbasic could inhibit HIV-1 replication in primary human CD4+ cells. However, MTS assays of our initial attempts to make transductant cells with NB-EGFP or EGFP indicated a considerable level of cellular toxicity, probably caused by EGFP expression in activated CD4+ primary cells (Liu et al., 1999; and data not shown). Therefore, the EGFP fusion protein partner was replaced with the better-tolerated ZsGreen1 (Wouters et al., 2005) protein for these experiments (Figs. 8a and and9)9) after confirming that NB-ZsGreen1 had a similar ability to inhibit HIV-1 infectivity as previously described (data not shown; and Meredith et al., 2009). Highly enriched human CD4+ cells isolated from buffy coat cells were transduced with MLV-based retroviral vectors conveying NB-ZsGreen1 or ZsGreen1. After 72hr, ZsGreen1-positive cells were collected by FACS as shown (Fig. 8b) and Western blot analysis confirmed expression of NB-ZsGreen1 (Fig. 8c) with an anti-Tat antibody. We noted that the overall fluorescence intensity of cells expressing ZsGreen1 or NB-ZsGreen1 was lower than that observed for comparable EGFP proteins in Jurkat cells (Fig. 6b), although ZsGreen1 has a comparable fluorescence intensity profile (Day and Davidson, 2009). We transduced 106 CD4+ cells with HIV-189.6 (a dual-tropic R5X4 laboratory HIV-1 strain) supernatant containing 20ng of CAp24 for 2hr. The virus was removed by washings and samples of supernatant and cells, collected on day 0 and every 7 days thereafter, were assayed for HIV-1 CAp24 by ELISA and for expression of Zs-Green1 or NB-ZsGreen1 at each time point by flow cytometry. We also compared cell viability with a commercial MTS assay (Fig. 9c). As shown in Fig. 9a (left), where cells were infected with HIV-1, the percentage of NB-ZsGreen1-positive CD4+ cells decreased from 80% on day 0 to 35% on day 7 and increased by day 21 to 54%. A similar pattern of NB-ZsGreen1 expression was observed in uninfected cells although at slightly lower levels (Fig. 9b, left). Higher percentages of CD4+ cells expressed ZsGreen1; 90% on day 0 and thereafter decreasing to 65–80% of CD4+ cells (Fig. 9a, right), irrespective of the presence of HIV-1 (Fig. 9b, right). In three identical experiments, HIV-1 levels were reduced by approximately 8- to 10-fold when NB-Zs-Green1 was expressed in CD4+ cells as compared with ZsGreen1. MTS assays of uninfected cells and HIV-1-infected NB-ZsGreen1 cells showed no significant change in cell viability compared with untreated control CD4+ cells. However, HIV-1-infected CD4+ ZsGreen1 cells demonstrated reduced viability, most likely a result of HIV-1 infection. These experiments support the conclusion that NB-ZsGreen1 and ZsGreen1 are tolerated by activated human CD4+ primary cells and that NB-ZsGreen1 strongly inhibited HIV-1 replication.
Here we demonstrated that Nullbasic, a mutant Tat protein, was effectively delivered by MLV-based retroviral vectors and its expression in primary human CD4+ cells and Jurkat T cells provided strong protection from HIV-1 replication. Mutant forms of Tat that inhibit HIV-1 gene expression in T cell lines have been described but these proteins were truncated within the basic domain (amino acids 49–57), expressed only the first exon, or were truncated in the second exon (86-amino acid Tat) and had fewer substitutions in the basic domain (Pearson et al., 1990; Modesti et al., 1991). Some of these mutant Tat proteins were shown to provide modest protection from HIV-1 replication when expressed in human T cell lines (Ulich et al., 1996). Whereas this study confirmed that downregulation of virus infectivity and reverse transcription by Nullbasic was demonstrably important, previous analysis of the antiviral response induced by these shorter Tat mutant proteins was, generally, limited to effects on HIV-1 gene expression (Green et al., 1989; Pearson et al., 1990; Modesti et al., 1991), and although not demonstrated it seems possible that other dominant negative Tat proteins could negatively affect reverse transcription. In addition, the negative effect on reverse transcription by Nullbasic was limited to HIV-1 and HIV-1-derived lentiviral vectors; there was no significant effect on infectivity by MLV-based retroviral vectors made by PG13 cells, suggesting that reverse transcription was unaffected. Overall, the extended level of protection from virus replication provided by Nullbasic suggests it may be useful as an agent to inhibit HIV-1 replication.
Nullbasic was reported to inhibit HIV-1 transcription by Tat, Rev-mediated mRNA transport, and reverse transcription (Meredith et al., 2009). Meredith and colleagues showed that when Nullbasic was cotransfected with a HIV-1 proviral vector in sufficient quantity, HIV-1 virus production could be strongly reduced (Meredith et al., 2009). However, lentiviral genomic RNA incorporation and particle production were not significantly affected when Nullbasic was expressed by pLOX-CW-NB-EGFP, suggesting that levels of Nullbasic expressed by pLOX vector were insufficient to inhibit Rev. Although these analyses hinted that Nullbasic might be successfully delivered to target cells by lentiviral vectors, overall transduction ability of the lentiviral particles conveying Nullbasic were markedly reduced compared with controls, an outcome that could be largely attributed to a defect in reverse transcription. However, Nullbasic did not affect infectivity of MLV-A virus-like particles, indicating that Nullbasic most likely targets the HIV-1 reverse transcription complex (RTC), cellular components important for HIV-1 reverse transcription, or may act as a trans-dominant negative inhibitor given evidence that Tat can support improved HIV-1 reverse transcription (Harrich et al., 1997). The precise mechanism by which Nullbasic inhibits reverse transcription remains to be determined.
In all experiments, HIV-1 replication in Jurkat-NB-EGFP cells was remarkably inefficient over a 4-week period, whereas sharply reduced HIV-1 replication was observed in activated primary CD4+ cells expressing NB-ZsGreen1 compared with control cells. We suspect that one significant factor for reduced inhibition in primary cells was due to reduced expression of NB-ZsGreen1 compared with NB-EGFP in Jurkat cells. Flow cytometric analysis of the overall fluorescence intensity profiles, where the relative fluorescence intensity of ZsGreen1 is reported to be 117% of EGFP (Day and Davidson, 2009), indicated an approximate 0.5-log reduction in ZsGreen1 expression in primary CD4+ cells compared with Jurkat-NB-EGFP cells. Second, a drop in the penetrance of NB-ZsGreen1 expression was observed from day 0 to 7. Although the exact cause for the decrease in the number of ZsGreen1+ cells is unclear, we did not observe significant change is the cell viability compared with control cells. It is possible that NB-ZsGreen1 was downregulated by a gene-silencing mechanism in some CD4+ cells (Cherry et al., 2000). Nevertheless, HIV-1 replication was reduced by approximately 8- to 10-fold in CD4+-NB-ZsGreen1+ cell populations in three independent experiments. The reduced virus production, as indicated by decreased capsid protein secretion, most likely resulted from greatly suppressed virus replication after infection on day 0 when approximately 80% of cells were NB-ZsGreen1+. Expression of Nullbasic in cells can decrease virus production in infected cells, resulting in reduced virus replication kinetics (Meredith et al., 2009). To address this possibility we attempted intracellular staining of the primary cells for HIV-1 capsid protein to determine which cell population was producing viral protein. However, this analysis was unsuccessful. Although we cannot exclude HIV-1 replication in NB-ZsGreen1+ cells, delayed virus replication in unprotected cells is more likely. In future experiments, vectors capable of strong and stable expression of NB-ZsGreen1 in activated CD4+ primary cells should help address this issue. Whereas expression of HIV-1 Tat can accelerate HIV-1 replication and induce cytotoxic effects in human T cells (Caputo et al., 1990; Benjouad et al., 1993), Nullbasic-EGFP and Nullbasic-ZsGreen1 were well tolerated HIV-1 inhibitors in Jurkat and primary CD4+ cells, respectively. Further analysis of how different domains of Tat contribute to Nullbasic antiviral effects may illuminate new candidate protein inhibitors with improved expression while maintaining strong anti-HIV-1 properties.
In conclusion, here we show that expression of Nullbasic fusion proteins could provide excellent protection against HIV-1 replication in Jurkat and primary CD4+ cells. Two previous studies showed that combined expression of Tat mutant proteins and RevM10 (Ulich et al., 1996) or their fusion (Aguilar-Cordova et al., 1995) provided synergistic inhibition of HIV-1 replication, a strategy that may be relevant to Nullbasic. Therefore, the unusual properties conveyed by Nullbasic make it a candidate antiviral agent. At the very least, analysis of its mechanism of HIV-1 inhibition could advance discovery of novel antiviral strategies.
This work was funded by a grant from the Australian Centre for HIV and Hepatitis Research and by an Australian Research Council Future Fellowship to D.H.
The authors declare that they have no competing interests.