NF-κB Inhibition Facilitates the Establishment of Cell Lines That Chronically Produce Human T-Lymphotropic Virus Type 1 Viral Particles

ABSTRACT

Most human T-lymphotropic virus type 1 (HTLV-1)-infected HeLa and SupT1 cells cease proliferation and become senescent immediately after infection by HTLV-1 or transduction of the HTLV-1 tax gene. The cellular senescence response triggered by Tax is caused by hyperactivated NF-κB and mediated by cyclin-dependent kinase inhibitors, p21^CIP1/WAF1 and p27^KIP1. When NF-κB activity is blocked by a degradation-resistant form of IκBα, ΔN-IκBα, Tax-induced senescence is averted. Here, we show that NF-κB inhibition through the expression of ΔN-IκBα allows cells of a human osteosarcoma (HOS) cell line to be chronically infected by HTLV-1. Stable HTLV-1-producing HOS cell clones can be readily established and isolated. These clones continue to proliferate in culture; express Tax, Rex, Gag, and Env proteins persistently; and transmit HTLV-1 to naive HOS, SupT1, and Jurkat T reporter cell lines readily after cocultivation. As HOS cells are adherent to culture plates, infected T cells in suspension can be easily collected and characterized. The ease with which chronic and productive HTLV-1 infection can be established in cell culture through inhibition of NF-κB affords a useful means to examine in depth the molecular events of HTLV-1 replication and the mechanisms of action of viral genes.

IMPORTANCE This paper describes a system for establishing cell lines that can be productively infected by human T-lymphotropic virus type 1 (HTLV-1) and can spread HTLV-1 to susceptible cells. Such a system can facilitate the study of HTLV-1 replication in cell culture.

INTRODUCTION

Human T-lymphotropic virus type 1 (HTLV-1) is a complex human retrovirus that infects approximately 10 to 20 million people worldwide. It is the causative agent of adult T-cell leukemia/lymphoma (ATL), HTLV-1-associated myelopathy/tropical spastic paraparesis (HAM/TSP), HTLV-1 uveitis, and other inflammatory diseases (1, 2). HTLV-1 infects a wide variety of cells, including T lymphocytes, B lymphocytes, monocytes, endothelial cells, and fibroblasts. This is due in part to its use of a ubiquitous cell surface molecule, glucose transporter 1, as the receptor for virus entry (3). Other molecules, such as neuropilin 1 and heparan sulfate proteoglycans, also contribute to viral infection (4, 5). The broad tropism of HTLV-1 notwithstanding, its transmission requires cell-to-cell contact (6). Cell-free HTLV-1 particles are poorly or not directly infectious (6). Interestingly, it has been shown recently that dendritic cells exposed to free HTLV-1 particles can rapidly transmit the virus to CD4⁺ T cells (7). Cell-to-cell transmission of HTLV-1 occurs through “virological synapses” formed in part through LFA1 and ICAM1 (8, 9). A recent study has found that HTLV-1 particles are stored as carbohydrate-rich, biofilm-like extracellular assemblies that rapidly attach to target cells for virus transmission (10).

HTLV-1 infection in cell culture is usually achieved by cocultivating naive cells with mitotically inactivated HTLV-1-producing cells or by cell-free infection using vesicular stomatitis virus (VSV) G-pseudotyped viral particles (11–13). To track cellular changes that occur after HTLV-1 infection, we generated several reporter cell lines using an expression cassette that contains 18 copies of the Tax-inducible HTLV-1 21-bp repeat, the viral TATA element, the complete R region, and a part of the U5 sequence fused to the enhanced green fluorescent protein (EGFP) gene (14). This reporter cassette can be stably integrated into cells of interest by using a self-inactivating lentivirus vector known as SMPU. With reporter cell lines derived in this way, we were able to show that HeLa cells cease proliferation within one or two division cycles after infection by HTLV-1 or transduction of the HTLV-1 tax gene (15, 16). HTLV-1-infected HeLa cells, like their tax-transduced counterparts, expressed high levels of cyclin-dependent kinase inhibitors—p21^CIP1/WAF1 and p27^KIP1—developed mitotic abnormalities, and became arrested in senescence (15, 17). Similarly, SupT1 T cells also became irreversibly arrested upon HTLV-1 infection (15). Most cells infected by HTLV-1 are unable to spread the infection to susceptible cells. This is likely due to the former's inability to form the virological synapse for virus transmission. Although previous studies have shown that human osteosarcoma (HOS) cells infected by HTLV-1 could spread/transmit the virus to other susceptible cells (18), clonal HOS cell lines chronically and stably producing HTLV-1 were difficult to establish (18).

We have shown recently that the senescence checkpoint response triggered by Tax represents a cellular safeguard against persistent and potentially oncogenic activation of IκB kinase/NF-κB (16, 19). When NF-κB activity is constitutively repressed by ΔN-IκBα, a degradation-resistant form of IκBα that lacks the NH₂-terminal 36 amino acid residues of the wild-type protein, Tax-induced senescence is prevented (16). Here, we demonstrate that when NF-κB activation is blocked, HOS cell lines that are chronically and productively infected by HTLV-1 can be established. They readily spread HTLV-1 to naive HOS cells and to SupT1 and Jurkat T cells.

MATERIALS AND METHODS

Cell lines and antibodies.

HOS cells were grown in Dulbecco's modified Eagle's medium (DMEM). The HOS-G/ΔN-IκBα cell line was derived by transducing HOS-G cells with LV-ΔN-IκBα-SV-Puro (16) and selecting for puromycin after 48 h. HOS-G/ΔN-IκBα clones were isolated in 96-well plates by limiting dilution and confirmed for ΔN-IκBα expression by immunoblotting. A HOS-RFP/ΔN-IκBα cell line was derived similarly. 729-B cells were grown in Gibco Iscove's modified Dulbecco's medium (IMDM). MT2, Jurkat, and SupT1 cells were grown in RPMI medium. All media were supplemented with 10% fetal bovine serum, 2 mM l-glutamine, and 100 U/ml of penicillin and streptomycin. Tax 4C5, actin, and IκBα antibodies have been described previously (16). Anti-HTLV1 gp46 and anti-HTLV-1 p19 were from Zeptometrix. Anti-HTLV-1 p24 was from ABL. Rabbit anti-HTLV-1 Rex was a kind gift of Gisela Heidecker (National Cancer Institute, Frederick, MD).

Plasmid construction.

The piggyBac transposon plasmid and a piggyBac plasmid carrying a transposase gene, PB-Tase, were obtained from Pentao Liu (Sanger Institute, United Kingdom) (20). The red fluorescent protein (RFP)-encoding sequence was excised from pTag-RFP-N with BamHI and NotI and ligated into the BamHI and NotI sites of the piggyBac-CAG-Oligo vector. A Tax-inducible enhancer/promoter cassette that contains 18 copies of the Tax-responsive 21-bp repeat element was excised from p18x21-DSred-Neo with XbaI and BamHI and ligated upstream of the RFP reporter gene. The plasmid was designated PB-18x21-RFP. Similarly, the firefly luciferase-encoding sequence was PCR amplified from the pLenti-CMV-Puro-LUC plasmid using oligonucleotides with the sequences CGCGGATCCATGGAAGACGCCAAAAACATAA and GAAGCGGCCGCTTACACGGCGATCTTTCCGCCC as forward and reverse primers, respectively, and then cloned into the BamHI and NotI sites of the PB-18x21-RFP plasmid, essentially by replacing the RFP region with the firefly luciferase gene, and designated PB-18x21-Luc. To construct a piggyBac transposon carrying a puromycin resistance gene under the control of the simian virus 40 (SV40) promoter, the SV40-Puro-long terminal repeat (LTR) region from HR-CMV-Tax-SV40-Puro (14) was digested with SalI and NheI and cloned into the XhoI and NheI sites of the piggyBac-CAG-Oligo plasmid. The plasmid was designated PB-SV40-Puro.

DNA transfection.

HOS-RFP, Jurkat-RFP/Luc, and SupT1-RFP/Luc cell lines were generated by introducing the piggyBac transposon plasmids through electroporation. Briefly, 3 × 10⁶ HOS cells were electroporated with 4.5 μg of PB-18x21-RFP and 1.5 μg of the piggyBac transposase expression plasmid, PB-Tase, and cultured in a 100-mm dish. After 48 h, limiting dilution was performed in a 96-well plate, and individual clones were expanded and confirmed for the expression of RFP after transduction with an adenovirus vector for the tetracyclin trans-activator (Ad-tTa) or for Tax (Ad-Tax). For the generation of Jurkat-RFP/Luc and SupT1-RFP/Luc cell lines, 10 × 10⁶ Jurkat or SupT1 cells were electroporated with 3.5 μg of PB-18x21-RFP, 3.5 μg of PB-18x21-Luc, 1.5 μg of PB-SV40-Puro, and 1.5 μg of PB-Tase and cultured in a 100-mm dish. Forty-eight hours after electroporation, Jurkat or SupT1 reporter cells were selected with puromycin (1 μg/ml) and cloned by limiting dilution in a 96-well plate. Individual cell clones of interest were expanded and observed for the expression of the RFP or luciferase gene after transduction with either Ad-tTa or Ad-Tax by immunofluorescence or luciferase assay.

HTLV-1 infection following coculture with 729B cells.

HTLV-1-producing 729B cells were mitotically inactivated with mitomycin C (10 μg/ml) at 37°C for 30 min and then washed once with serum-free medium and phosphate-buffered saline (PBS) to completely remove the residual mitomycin C. For coculture experiments, HOS-G or HOS-G/ΔN-IκBα cells in 2 ml DMEM plus 10% fetal bovine serum were seeded in 6-well plates at a density of 2 × 10⁵ cells per well. Mitotically inactivated 729B cells were added to each well at a density of 1 × 10⁶ per well and incubated for 24 h. After 24 h, the supernatant containing the 729B cells or dead cells was replaced with fresh DMEM plus 10% fetal bovine serum. HTLV-1-infected HOS-G or HOS-G/ΔN-IκBα cells were observed for green fluorescent protein (GFP) expression at days 2 and 6 using an epifluorescence microscope. HOS-G/ΔN-IκBα cells infected with HTLV-1 were expanded in 150-cm² flasks and sorted into 96-well plates using GFP as a marker at 488 nm with a FACSAriaII fluorescence cell sorter (BD Bioscience). More than 87% purity of the sorted cells was observed when the cells were reanalyzed after sorting. Clones of interest were identified by fluorescence microscopy and further expanded.

Western blotting.

Cell lysates were prepared using RIPA buffer, and protein concentrations were determined by bicinchoninic acid (BCA) protein assay (Thermo Scientific Pierce, Inc.). Cell supernatants containing virus particles were collected and centrifuged briefly for 5 min at 500 × g at 4°C to remove cell debris. Later, the clear supernatants were filtered through 0.22-μm Millex-GP PES membrane filters and centrifuged. The supernatants were removed, and the virus pellets were dissolved in 2× SDS sample buffer. Proteins were separated by sodium dodecyl sulfate-12% polyacrylamide gel electrophoresis and transferred to polyvinyl difluoride (PVDF) membranes. The PVDF membranes were probed for p19, p24, Tax, Rex, gp46, IκBα, or actin antibodies, followed by the addition of goat anti-mouse horseradish peroxidase (HRP) or goat anti-rabbit HRP (Santa Cruz) and detection by enhanced chemiluminescence (Luminata; Millipore).

Transmission electron microscopy.

729B or HOS-G/ΔN-IκBα-HTLV 1F11 cells were grown in 150-cm² Corning flasks. The supernatants were harvested, centrifuged at 500 × g to remove cell debris, and filtered through a 0.22-μm Millex-GP PES membrane filter. The filtrates were pelleted through a sucrose cushion (20% sucrose in PBS) for 2 h at 25,000 rpm at 4°C. The virus pellets were fixed in 2% glutaraldehyde-2% formaldehyde overnight at 4°C. Finally, the virus particles were negatively stained and visualized under an electron microscope equipped with a digital camera system.

Cell-to-cell transmission of HTLV-1.

HOS-G/ΔN-IκBα/HTLV-1 and HOS-RFP/ΔN-IκBα cells were detached from the flask using PBS containing 0.5 mM EDTA, washed with PBS once, and then cocultured at 3 × 10⁵ cells in total per well in a 6-well plate at a 1:1 ratio. To enhance cell susceptibility to virus infection, MgCl₂ at a final concentration of 15 mM was added to the culture medium. For HTLV-1 transmission to the SupT1 or Jurkat reporter cell line, HOS-G/ΔN-IκBα/HTLV-1 cells (1.5 × 10⁵) were cocultured with SupT1-RFP/Luc or Jurkat-RFP/Luc at a 1:5 ratio in 6-well plates in the presence of 15 mM MgCl₂. The cells were maintained and observed for RFP expression using a fluorescence microscope or processed for luciferase assay. For transmission inhibition studies, SupT1-RFP/Luc cells were seeded at 5 × 10⁵ cells per well in 6-well plates and incubated overnight in the presence or absence of various concentrations of the reverse transcriptase inhibitor azidothymidine (AZT) or the integrase inhibitor raltegravir. The next day, HOS-G/ΔN-IκBα/HTLV-1 cells were added at a density of 1 × 10⁵ cells per well to the drug-treated SupT1-RFP/Luc cells and cultured for another 48 or 72 h. Then, the SupT1-RFP/Luc cells in suspension were collected, lysed, and assayed for luciferase.

Luciferase assay.

A dual-luciferase assay was performed according to standard procedures recommended by the manufacturer (Promega).

Lentivirus and adenovirus vectors.

The recombinant adenovirus vectors Ad-tTA and Ad-Tax and a lentiviral vector carrying an N-terminal mutant form of IκBα have been described previously (16).

RESULTS

Derivation of HOS reporter cell lines for HTLV-1 infection.

To identify and track the fate of HTLV-1-infected cells, we had previously derived reporter cell lines carrying the 18x21-EGFP reporter cassette, which directs the expression of enhanced green fluorescent protein from a transcriptional promoter that consists of 18 copies of the Tax-responsive 21-bp repeat element upstream of the minimal HTLV-1 promoter (15, 16). Using these cell lines, we have demonstrated that HTLV-1 infection causes HeLa and SupT1 cells to express high levels of cyclin-dependent kinase inhibitors, p21^CIP1/WAF1 and p27^KIP1, and to become senescent. Subsequent studies from our laboratory have demonstrated that senescence represents a cellular safeguard against persistent and potentially oncogenic activation of NF-κB by Tax (16). When NF-κB activity is blocked by ΔN-IκBα or by short hairpin RNA (shRNA) knockdown of the IKKα or IKKγ/NEMO subunit, Tax-induced senescence is prevented (16, 19). We therefore wondered if NF-κB inhibition could facilitate establishment of cell lines that are chronically and persistently infected by HTLV-1.

A wide variety of lymphoid and nonlymphoid cells, including T lymphocytes, B lymphocytes, monocytes, endothelial cells, and fibroblasts, can be infected by HTLV-1. However, establishment of cell lines that stably and chronically produce HTLV-1 virions has been difficult, most likely because of the cellular senescence checkpoint response induced by Tax. Furthermore, for reasons that are unclear at present, very few cell lines are capable of spreading HTLV-1 infection. The human osteosarcoma cell line HOS had been shown previously to support not only productive HTLV-1 infection, but virus transmission to other cells (18). For this reason, we constructed a series of HOS reporter cell lines.

A HOS cell line containing a stably integrated Tax-responsive 18x21-EGFP reporter cassette has been described in previous reports (14, 15). To generate the HOS-18x21-RFP cell line for monitoring HOS-to-HOS virus transmission, the EGFP coding sequence in the 18x21-EGFP cassette was replaced with that of RFP, and the 18x21-RFP cassette was inserted into the piggyBac transposon vector to produce the PB-18x21-RFP plasmid. HOS cells were then electroporated with PB-18x21-RFP, together with an expression plasmid for the piggyBac transposase, to enable integration of the 18x21-RFP reporter cassette, and 48 h after DNA transfection, the cells were subjected to limiting dilution in 96-well plates. To identify HOS-18x21-RFP clones that stably harbor the 18x21-RFP reporter, we transduced candidate RFP clones with Ad-Tax, an adenovirus expression vector for Tax, at a multiplicity of infection (MOI) of 5 for 48 h and screened for RFP expression by fluorescence microscopy. The HOS-RFP cell clones that showed strong RFP expression in response to Ad-Tax were isolated and expanded. The Tax responsiveness of the HOS-G and HOS-RFP cell lines is illustrated in Fig. 1A, where HOS-G and HOS-RFP cells, respectively, were transduced with Ad-Tax or Ad-tTa for 48 h and photographed using an Olympus IX81 microscope equipped with a charge-coupled-device (CCD) camera. As indicated, HOS-G and HOS-RFP cells expressed green and red fluorescent proteins, respectively, after Ad-Tax, but not Ad-tTa, infection (Fig. 1A). As expected, a reporterless HOS cell control did not show GFP or RFP signal after Ad-Tax or Ad-tTa infection.

FIG 1.

(A) Derivation of HOS reporter cell lines. HOS cells were either transduced with the LV-SMPU-18x21-EGFP vector or transfected by electroporation with the piggyBac-18x21-RFP transposon, together with an expression plasmid for the piggyBac transposase, PB-Tase, to enable integration of the reporter gene cassette. Infected or transfected cells were cloned by limiting dilution into 96-well plates. Individual cell clones were expanded and screened for the stable integration of the reporter cassette by infection with Ad-Tax or Ad-tTa. HOS reporter cell lines were then visualized by fluorescence (FL) and bright-field phase-contrast (BF) microscopy. (B) Stable expression of degradation-resistant ΔN-IκBα in HOS reporter cell lines. HOS-18x21-EGFP and HOS-18x21-RFP cells were transduced with LV-ΔN-IκBα-SV-puro (LV-ΔN-IκBα-puro) or not (Mock) and selected in medium containing puromycin (1 μg/ml). HOS-G/ΔN-IκBα (or HOS-RFP/ΔN-IκBα) clones were then isolated by limiting dilution. (C) Immunoblot analysis of 4 HOS-G/ΔN-IκBα cell lines (2B8, 1F2, 1B3, and 1D9) and the parental HOS-G cell line using IκBα and control β-actin (Actin) antibodies.

We next asked if NF-κB inhibition in HOS cells can facilitate chronic HTLV-1 infection. To this end, HOS-G cells were transduced with a lentivirus vector, LV-ΔN-IκBα-SV40-Puro, that carries the coding sequence of the degradation-resistant NF-κB superrepressor, ΔN-IκBα, and selected for puromycin resistance (Fig. 1B). HOS-G/ΔN-IκBα clones were then isolated in 96-well plates after limiting dilution, expanded, and confirmed for ΔN-IκBα expression by immunoblotting (Fig. 1C). The NF-κB shutoff by ΔN-IκBα had no adverse effects on the growth of HOS cells, as HOS-G/ΔN-IκBα cells continued to grow like their parental controls. A HOS-G/ΔN-IκBα clone known as 1B3 was chosen for further experiments. By using the same strategy, we also generated a HOS-RFP/ΔN-IκBα cell line.

NF-κB inhibition facilitates productive HTLV-1 infection and establishment of stable HTLV-1-producing HOS cell lines.

HOS-G and HOS-G/ΔN-IκBα cells were then infected by cocultivation at a 1:5 ratio with the HTLV-1-producing 729B cells that had been mitotically inactivated by mitomycin C treatment. As indicated in Fig. 2A, most HTLV-1-infected (GFP-positive) HOS-G cells ceased proliferation immediately, in accordance with the notion that HTLV-1 infection leads to chronic NF-κB activation, which triggers a cellular senescence response. For reasons that are unclear, under these conditions, HTLV-1-infected HOS cells did not spread HTLV-1 to the surrounding cells. HOS-G/ΔN-IκBα cells that were infected by HTLV-1 appeared to continue to proliferate after infection (Fig. 2B). Clusters of about 32 proliferating GFP-positive HOS-G/ΔN-IκBα cells infected with HTLV-1 could be easily seen at day 6 after coculture with 729B cells.

FIG 2.

Infection of HOS-G and HOS-G/ΔN-IκBα cell lines by coculture with the 729B cell line. HOS-G (A) and HOS-G/ΔN-IκBα (B) cells were cocultured with the mitotically inactivated HTLV-1-producing 729B cell line for 24 h at a cell-to-cell ratio of 5 (729B):1 (HOS-G or HOS-G/ΔN-IκBα). After 24 h, 729B cells in suspension were removed, and the HOS cells were visualized by fluorescence (FL) and bright-field phase-contrast (BF) microscopy at days 2 and 6 postinfection.

As HOS-G/ΔN-IκBα cells infected with HTLV-1 seemed able to continue to proliferate, we sought to clone and characterize them in depth. To this end, we sorted GFP-positive cells from HTLV-1-infected HOS-G/ΔN-IκBα cultures into 96-well plates. Individual GFP-positive clones were isolated, expanded, and characterized for HTLV-1 viral antigen expression by immunoblotting. Three HOS-G/ΔN-IκBα/HTLV-1 clones, 1F5, 1F11, and 2C5, were chosen for further analyses. Total cell lysates and culture supernatants were prepared from these clones as described in Materials and Methods. As indicated by immunoblotting, all HOS-G/ΔN-IκBα/HTLV-1 cells were productively infected by HTLV-1 and expressed Rex, Tax, Gag (p19 matrix and p24 capsid), and Env (gp46) (Fig. 3A). HTLV-1 Gag protein appeared as a 55-kDa precursor protein, p55, which was processed into p19 and p24 proteins after virus release. Although p19 and p24 were only weakly detectable in total cell lysates of the productively infected clones 1F5, 1F11, and 2C5 (Fig. 3A, middle), they could be easily detected in the culture supernatants of all clones (Fig. 3A, right), indicating efficient virus budding. We noted that 1F5, 1F11, and 2C5 clones expressed low levels of Env (gp46) compared to control MT2 or 729B cells (Fig. 3A, middle). This may be in part related to the copy number of HTLV-1 DNA in infected HOS cells.

FIG 3.

Productive HTLV-1 infection in HOS-G/ΔN-IκBα cell lines. (A) Immunoblot analyses of HTLV-1-infected HOS-G/ΔN-IκBα clones. HOS-G/ΔN-IκBα cells were cocultured with 729B cells as described in the legend to Fig. 2 and sorted at one cell per well into 96-well plates with the BD FACSAria cell sorter. Clones of interest were identified by fluorescence microscopy and further expanded. The cell lysates, as well as the culture supernatants, of HOS-G/ΔN-IκBα and its HTLV-1-infected progeny cell lines were prepared and subjected to immunoblot analyses using p19, p24, Tax, Rex, gp46, and actin antibodies. (B) Electron microscopy of HTLV-1 viral particles produced by 729B cells and the HOS-G/ΔN-IκBα/HTLV-1 1F11 clone. The culture supernatants of 729B cells and the HOS-G/ΔN-IκBα/HTLV-1 1F11 clone were harvested, spun down at 500 × g to remove cell debris, and filtered through a 0.22-μm Millex-GP PES membrane filter. The filtrates were pelleted through a 20% sucrose cushion in PBS for 2 h at 25,000 rpm at 4°C. The virus pellets were fixed in 2% glutaraldehyde-2% formaldehyde overnight at 4°C, followed by negative staining, and then visualized under an electron microscope.

We next visualized viral particles released by HOS-G/ΔN-IκBα/HTLV-1 cells by electron microscopy. The culture supernatant of clone 1F11 was concentrated and processed for electron microscopy as described in Materials and Methods. Similar experiments were also carried out for 729B cells for comparison. As indicated in Fig. 3B, 90-nm particles with what appeared to be icosahedral capsids could be detected in the concentrated culture supernatants of 729B (Fig. 3B, top) and HOS-G/ΔN-IκBα/HTLV-1 (Fig. 3B, bottom) cells, consistent with the notion that HOS-G/ΔN-IκBα/HTLV-1 clones are productively infected and chronically produce HTLV-1 viral particles. Collectively, these results suggest that cells productively infected by HTLV-1 mostly undergo senescence as a result of chronic NF-κB activation by Tax. However, when NF-κB activity is blocked, the senescence checkpoint response is avoided, and the productively infected cell (PIC) population can grow and divide and can be established as individual virus-producing clones.

Productively infected HOS-G/ΔN-IκBα/HTLV-1 cells transmit HTLV-1 to HOS and T cells.

We next asked if HOS-G/ΔN-IκBα/HTLV-1 could transmit HTLV-1 to susceptible cells. To visualize HTLV-1 infection and to quantitatively measure HTLV-1 transmission, we generated SupT1-RFP/Luc (firefly luciferase) and Jurkat-RFP/Luc cell lines. SupT1 and Jurkat cells were electroporated with three piggyBac transposons harboring SV40-Puro, 18x21-RFP, and 18x21-Luc cells, together with an expression plasmid for the piggyBac transposase to enable integration of all three transposons. After 48 h, both SupT1 and Jurkat reporter cells were selected with puromycin (1 μg/ml). Later, these cells were subjected to limiting dilution in 96-well plates. Individual Jurkat-RFP/Luc and SupT1-RFP/Luc clones were expanded and screened for RFP and firefly luciferase expression after transduction by Ad-Tax or Ad-tTA (at an MOI of 2). As shown in Fig. 4B, a 134-fold and a 41-fold increase in luciferase activity were detected for Ad-Tax-transduced Jurkat-RFP/Luc and SupT1-RFP/Luc cells, respectively. Attempts to generate reporter T cell lines expressing ΔN-IκBα were unsuccessful, most likely because NF-κB is essential for T cell proliferation and survival. As might be expected, the SupT1-RFP/Luc and Jurkat-RFP/Luc reporter cell lines were readily infected in coculture with the HTLV-1-producing 729B cells, as measured by RFP fluorescence and luciferase reporter activities (Fig. 4C and D).

FIG 4.

Transmission of HTLV-1 by 729B/HTLV-1 cells to T reporter cell lines. (A and B) Derivation of SupT-1 and Jurkat reporter cell lines. SupT-1-RFP/Luc and Jurkat-RFP/Luc reporter cell lines were generated using three transposons—piggyBac-18x21-RFP, piggyBac-18x21-Luc, and piggyBac-SV-Puro—as described in Materials and Methods and isolated by limiting dilution. Clones of interest were infected with Ad-Tax or Ad-tTa and identified by fluorescence microscopy (A) and luciferase assay (B). (C and D) SupT1-RFP/Luc and Jurkat-RFP/Luc cells were cocultured at a 1:5 ratio for 48 h with mitotically inactivated HTLV-1-producing 729B cells that had been stably transfected with the ACH molecular clone of HTLV-1. Infection was monitored by RFP fluorescence (C) and luciferase reporter activities (D) as described in Materials and Methods. The values shown are the fold changes normalized to the control. The error bars indicate standard deviations.

To investigate HTLV-1 transmission from HOS cells to target cells, we cocultured HOS-G/ΔN-IκBα/HTLV-1 cells with HOS-RFP/ΔN-IκBα, SupT1-RFP/Luc, or Jurkat-RFP/Luc cells. HOS-G/ΔN-IκBα/HTLV-1 cells readily transmitted HTLV-1 to these target cells, as shown by the appearance of RFP-positive cells (compare Fig. 5A, B, and C). HOS-G/ΔN-IκBα/HTLV-1 and HOS-RFP/ΔN-IκBα cells appeared not to form syncytia, as judged from their apparently normal cell sizes (Fig. 5A). Likewise, no syncytia were observed between HOS-G/ΔN-IκBα/HTLV-1 cells and infected SupT1-RFP/Luc or Jurkat-RFP/Luc cells.

FIG 5.

Transmission of HTLV-1 from HOS-G/ΔN-IκBα cells to HOS-RFP/ΔN-IκBα, SupT1-RFP/Luc, or Jurkat-RFP/Luc cells. (A) HTLV-1 transmission between HOS cells. HOS-G/ΔN-IκBα/HTLV-1 1F11 cells were cocultured with HOS-RFP/ΔN-IκBα at a 1:1 ratio and visualized by fluorescence (FL) and phase-contrast (BF) microscopy after 48 h. (B and C) HTLV-1 transmission between HOS and T cell lines SupT1 and Jurkat. HOS-G/ΔN-IκBα/HTLV-1 1F11 cells were cocultured with SupT1-RFP/Luc (B) and Jurkat-RFP/Luc (C) cells at a ratio of 1:5 for 48 h. The SupT1-RFP/Luc and Jurkat-RFP/Luc cells were visualized by fluorescence and phase-contrast microscopy. (D) Luciferase activities of HTLV-1-infected SupT1-RFP/Luc and Jurkat-RFP/Luc cells. Cells in suspension from panels B and C were harvested and assayed for firefly luciferase activities as described in Materials and Methods. (E) The reverse transcriptase inhibitor AZT and the integrase inhibitor raltegravir inhibit HTLV-1 infection. HOS-G/ΔN-IκBα/HTLV-1-1F11 cells were cocultured with SupT1-RFP/Luc in the presence of AZT at 0, 5, 10, 20, and 50 μM for 48 h or raltegravir at 0, 10, 20, 50, and 100 nM for 72 h. SupT1-RFP/Luc cells in suspension were collected and assayed for luciferase activities. The values shown are the fold changes normalized to the control. The error bars indicate standard deviations.

Because NF-κB activation was normal in both T reporter cell lines, they became growth arrested as single RFP⁺ cells upon HTLV-1 infection (Fig. 5B and C). The 18x21-Luc cassette in the T reporter cell lines allows virus infection to be measured quantitatively. Even with only a small percentage (approximately 0.4%, as determined by flow cytometry [results not shown]) of cells infected, SupT1-RFP/Luc and Jurkat-RFP/Luc cells showed approximately 4-fold and 5-fold increases in luciferase activity after coculture with HOS-G/ΔN-IκBα/HTLV-1 cells (Fig. 5D). These results indicate that HOS cells productively infected with HTLV-1 can transmit the virus to HOS and T cells, as monitored by both RFP expression and luciferase activity. Under our experimental conditions, the efficiency of HOS-G/ΔN-IκBα/HTLV-1 cells in transmitting HTLV-1 to SupT1 or Jurkat cells is approximately 20% that of 729B cells, as judged by the luciferase activities shown in Fig. 4D and 5D (approximately 3.9-fold versus 19-fold and 5.3-fold versus 27-fold basal luciferase activity for SupT1 and Jurkat cells, respectively). This may be due to the lower level of virus production in the HOS-G/ΔN-IκBα/HTLV-1 clone used and/or better cell-to-cell contact established between 729B and SupT1 or Jurkat cells.

We also tested the infectivity of the cell-free virus particles produced by HOS-G/ΔN-IκBα/HTLV-1 cells. To this end, the culture supernatant from HOS-G/ΔN-IκBα/HTLV-1 clone 1F11 was collected, filtered through a 0.22- μm PES membrane filter, and used to infect HOS-RFP/ΔN-IκBα, SupT1-RFP/Luc, and Jurkat-RFP/Luc reporter cell lines. No RFP signal or luciferase activity could be detected, however (unpublished results). This is in agreement with published literature showing that HTLV-1 transmission occurs largely through cell-to-cell contact and that cell-free viral particles are poorly infectious unless dendritic cells are used to present them to target cells (7).

Inhibition of HTLV-1 transmission by reverse transcriptase and integrase inhibitors.

Finally, to demonstrate that HTLV-1 transmission from HOS to T cells occurred through de novo infection that involves virus entry, reverse transcription, and integration, we treated the infection with a reverse transcriptase inhibitor, AZT, and an integrase inhibitor, raltegravir, using luciferase reporter activities as readouts. SupT1-RFP/Luc cells were pretreated with AZT (0, 5, 10, 20, and 50 μM) or raltegravir (0, 10, 20, 50, and 100 nM) for 24 h, cocultured with HOS-G/ΔN-IκBα/HTLV-1 cells in the presence of inhibitors for 48 h (AZT) and 72 h (raltegravir), and then harvested for luciferase activity measurement. As indicated in Fig. 5E, both AZT and raltegravir reduced the luciferase activity in the newly infected reporter cells in a dose-dependent manner. In aggregate, our results indicate that chronic and productive HTLV-1 infection can be established in HOS cells whose NF-κB activity is strongly repressed. These infected HOS cells readily transmit HTLV-1 to susceptible HOS and T cells.

DISCUSSION

We had previously shown that most HTLV-1-infected HeLa or SupT1 cells became senescent and ceased proliferation immediately (15, 21). The cellular senescence caused by HTLV-1 infection is a result of Tax-driven NF-κB hyperactivation (16). It is mediated by the CDK inhibitors p21 and p27 and likely serves as a cellular safeguard against potentially oncogenic chronic NF-κB activation (16). In this report, we have extended the early studies to show that when the senescence-imposed restriction on HTLV-1-infected cells is lifted by suppressing NF-κB through the stable expression of a degradation-resistant form of IκBα, ΔN-IκBα, HOS cell clones productively infected by HTLV-1 can be readily established. The resultant HOS-G/ΔN-IκBα/HTLV-1 cell lines continue to proliferate and produce Tax, Rex, Gag, and Env proteins. We have shown further that these productively infected HOS cells can transmit HTLV-1 when cocultured with target reporter cell lines of HOS or T cell origin. Thus, by blocking the HTLV-1-induced senescence checkpoint response and by taking advantage of the ability of HOS cells to serve as both HTLV-1 transmitter and target cells, we have established a cell culture system that can support all stages of the HTLV-1 replicative cycle. Similar attempts to repress NF-κB in T lymphoid cell lines were not successful. This reflects the importance of NF-κB in T cell proliferation and survival. The HTLV-1 antisense protein HBZ has been shown previously to suppress the classical NF-κB pathway (22). Although theoretically HBZ may facilitate productive viral infection, its inhibition of NF-κB is moderate and is effective in delaying the onset of senescence only when the level of Tax is low (16). Thus, strong NF-κB inhibition by a superrepressor, such as ΔN-IκBα, is necessary for the establishment of cell lines productively infected by HTLV-1. Finally, under the present conditions, the efficiency of HTLV-1 infection via HOS is not as high as desirable (0.4%). This is likely due to the lower levels of viral gene expression in current HTLV-1-producing HOS cell clones.

Because HTLV-1-transmission requires cell-to-cell contact, most infections in cell culture involve cocultivation of umbilical cord or peripheral blood lymphocytes with lethally irradiated virus-producing HTLV-1-transformed cell lines, such as MT2, as virus donors. The endpoint of such studies is often T cell transformation, which occurs weeks to months after infection. How early events of viral infection impact cell fate cannot be examined in such studies. Furthermore, HeLa and T cell lines, such as Jurkat or SupT1, while readily infected during cocultivation, are unable to transmit HTLV-1 to other susceptible cells, thereby limiting the infection to only half of the HTLV-1 replicative cycle.

The HOS cell line is a nonlymphoid human osteosarcoma cell line permissive to HTLV-1 infection and spread (18). The present results show that most HOS cells infected with HTLV-1 after cocultivation with 729B (or MT2) underwent senescence and ultimately disappeared from culture within 2 or 3 weeks. However, for those HOS cells whose NF-κB activity is blocked by ΔN-IκBα, HTLV-1 infection became chronic. Such stably and productively infected HOS cells continued to proliferate and readily transmitted HTLV-1 to target HOS or T cells. In agreement with published results from many laboratories, although HTLV-1 viral particles were readily detected in the culture supernatants of HOS-G/ΔN-IκBα/HTLV-1 cells, they were not infectious. In contrast, and consistent with the need for cell-to-cell contact for HTLV-1 transmission, when chronically HTLV-1-infected HOS cells were cocultured with uninfected HOS, Jurkat, or SupT1 cells, virus transmission was readily detected. Two advantages of the HOS-G/ΔN-IκBα/HTLV-1-mediated HTLV-1 transmission are noted. (i) HTLV-1 producing 729B cells, when cocultured with HOS or T-cell lines, produced syncytia in the form of multinucleated giant cells. This was not observed with HOS-G/ΔN-IκBα/HTLV-1 cells, however. This may be due to the lower levels of virus production in general and envelope protein expression in particular by HOS-G/ΔN-IκBα/HTLV-1 cells, which render the cells less fusogenic. (ii) HTLV-1-producing cell lines capable of virus transmission, such as T and B cell lines, MT2 and 729B, respectively, grow in suspension. They are more difficult to separate from target T lymphocytes of peripheral blood or cord blood origin. As HOS-G/ΔN-IκBα/HTLV-1 cells are adherent in nature, they can be used to infect T cells by coculture, and the infected T cells can be easily separated for further analysis. In summary, our study indicates that cell lines chronically infected by HTLV-1 can be easily established once NF-κB activation is inhibited. The application of this finding to HOS cells yields stably infected HOS cell lines that readily transmit HTLV-1 to target cells. We anticipate that these results and cell lines will facilitate studies of HTLV-1 replication.