Human A549, a type II alveolar epithelial cell line, and NCI-H292, a tracheal epithelial cell line, were maintained in RPMI 1640 containing 10% fetal bovine serum (FBS). MT-2 cells, obtained by coculture of peripheral leukemic cells from an ATL patient with normal umbilical cord leucocytes 20, were used as the HTLV-I-infected T-cell line. MT-2 cells contained proviral HTLV-I DNA and produced viral particles. CCRF-CEM cells were used as the uninfected T-cell line. These T cells were treated with 100 μg/ml of mitomycin C (MMC) for 1 h at 37°C. After washing three times with phosphate buffered saline (PBS), they were cultured with an equal number of epithelial cells in RPMI 1640 containing 10% FBS. The culture medium was changed on the third day after coculture. A549 and NCI-H292 cells were harvested at 3, 5, 8 and 14 days, followed by DNA and RNA extraction, as described below. Samples of the culture supernatant were collected at 3 and 5 days after infection and used to measure the p19 antigen of HTLV-I (ZeptoMetrix, Buffalo, NY), IL-8 (BioSource International, Camarillo, CA) and CCL20 (R&D Systems, Minneapolis, MN) by enzyme-linked immunosorbent assay (ELISA).

Total RNA was extracted with Trizol (Invitrogen, Carlsbad, CA) according to the protocol provided by the manufacturer. First-strand cDNA was synthesized from 5 μg total cellular RNA using an RNA PCR kit (Takara Bio Inc., Otsu, Japan) with random primers. Thereafter, cDNA was amplified. The sequences of the primers were described previously 1821222324252627282930. PCR products were fractionated on 2% agarose gels and visualized by ethidium bromide staining.

DNA was prepared from each sample and stored at -80°C until use. The concentration of extracted DNA was adjusted to 10 ng/μl of the working solution. A quantitative real-time PCR assay was developed to measure the proviral load of HTLV-I in cells, as described previously 18.

We examined lung biopsy specimens from three patients with HTLV-I-related pulmonary diseases or normal lung biopsies, and lung biopsy specimens from transgenic mice bearing Tax or control littermate mice 9. All subjects provided informed consent before samples were obtained. The tissue samples were subjected to immunohistochemical staining using the mouse monoclonal antibody (Ab) to Tax, Lt-4 31. Serial sections were deparaffinized. Antigenic sites bound by the Ab were identified by reacting the sections with a mixture of 0.05% 3,3'-diaminobenzidine tetrahydrochloride in 50 mM Tris-HCl buffer and 0.01% hydrogen peroxide. Sections were counterstained with methyl green.

Cells were lysed in a buffer containing 62.5 mM Tris-HCl (pH 6.8), 2% sodium dodecyl sulfate, 10% glycerol, 6% 2-mercaptoethanol and 0.01% bromophenol blue. Equal amounts of protein (20 μg) were subjected to electrophoresis on sodium dodecyl sulfate-polyacrylamide gels, followed by transfer to a polyvinylidene difluoride membrane and sequential probing with the specific antibodies. The bands were visualized with an enhanced chemiluminescence kit (Amersham Biosciences, Piscataway, NJ). Mouse monoclonal Ab to actin was purchased from NeoMarkers (Fremont, CA). Mouse monoclonal Ab to Tax, Lt-4, was used.

To measure the expression of ICAM-1 and LFA-1 on the surface of epithelial cells after HTLV-I infection, FITC-labeled mouse monoclonal Ab against ICAM-1, LFA-1 α chain or control mouse IgG1 (Coulter Immunotech Co., Marseille, France) was used. Cells were analyzed on an Epics XL flow cytometer (Beckman Coulter, Fullerton, CA) after gating on forward and side scatter to exclude debris and clumps.

A549 cells were transfected with a luciferase reporter construct for the HTLV-I long terminal repeat (LTR), and NF-κB and AP-1 reporter constructs 222830 using Lipofectamine (Invitrogen). After 24 h, the transfected A549 cells were cocultured in the presence or absence of MMC-treated MT-2 or CCRF-CEM cells for 24 h before luciferase assay. Luciferase activities were measured using the dual luciferase assay system (Promega, Madison, WI) and normalized by the Renilla luciferase activity from phRL-TK.

EMSA was performed as described previously 2230. Briefly, 5 μg of nuclear extract was incubated with ³²P-labeled probes. The DNA-protein complex was separated from the free oligonucleotides on a 4% polyacrylamide gel. For competition experiments, the cold oligonucleotide probe or competitors were used, and supershift analysis was performed using Abs against NF-κB subunits p50, p65, c-Rel, p52 and RelB, and AP-1 subunits c-Fos, FosB, Fra-1, Fra-2, c-Jun, JunB and JunD (Santa Cruz Biotechnology, Santa Cruz, CA).

A549 and NCI-H292 cells were cocultured with either MT-2 or CCRF-CEM cells. After cocultivation for 3 days, A549 and NCI-H292 cells were washed extensively and recultured in a fresh medium for another 2 days, followed by thorough washing. At 3 and 5 days post-cocultivation, A549 and NCI-H292 cells were harvested for assessment by RT-PCR for expression of HTLV-I viral antigen. Since T-cell lines were pretreated extensively with MMC, these MMC-treated T cells could not proliferate, as determined by cell proliferation assay. These specimens of A549 and NCI-H292 cells at 3 and 5 days of culture contained no viable MT-2 cells. As shown in Figure 1, A549 and NCI-H292 cells cocultured with MT-2 cells showed strong expression of Tax mRNA. In contrast, A549 and NCI-H292 cells cocultured with CCRF-CEM cells did not express Tax mRNA. Using RNA samples prepared from A549 cells cocultured with non-permissible HTLV-I-infected T cell line, TL-OmI 32, RT-PCR was carried out, but Tax mRNA was not detected (data not shown).

We next performed Western blot analysis to assess the expression of Tax protein in A549 cells cocultured with MT-2 or CCRF-CEM cells. As shown in Figure 2B, A549 cells cocultured with MT-2 cells for 3 days expressed Tax protein. In contrast, A549 cells cocultured with CCRF-CEM cells did not express Tax protein. These results suggest that HTLV-I can be transmitted into lung epithelial cells from HTLV-I producing MT-2 cells.

To confirm the production of viral protein, A549 and NCI-H292 cells were first cocultured for 3 days either alone (control) or with MMC-treated MT-2 cells, then washed extensively and recultured in a fresh medium for 2 days. At the end of this period, the level of HTLV-I p19 core protein was measured in culture supernatants. Production of HTLV-I p19 was evident after 3 day of infection; the levels of HTLV-I p19 in the supernatants of A549 and NCI-H292 cells infected with HTLV-I were 1337 and 1023 pg/ml, respectively. The level of p19 in the supernatant of the MMC-treated MT-2 cells, the number of which corresponds to that used for coculturing, was less than 25 pg/ml. These results argue against the possibility that the p19 in the supernatant was produced by residual MT-2 cells used for infection, and support our conclusion that lung epithelial cells are infected by HTLV-I.

Using DNA samples extracted from cocultured lung epithelial cells, the pX region sequence of HTLV-I proviral DNA was amplified by real-time PCR. In A549 cells, the proviral copy numbers per 100 cells were 100, 100 and 64 at 3, 5 and 14 days, respectively. In NCI-H292 cells, the proviral copy numbers were 100, 84 and 40 at 3, 5 and 8 days, respectively. Taken together, our observations suggest that coculturing of lung epithelial cells with MT-2 resulted in infection with HTLV-I.

Tax activates not only the transcription of the viral genome but also the expression of various cellular genes 33. Therefore, we investigated the expression of inflammatory cytokines, chemokines and cell adhesion molecules in A549 cells cocultured with MT-2 or CCRF-CEM cells by RT-PCR. As shown in Figure 2A, the expression levels of IL-1α, IL-1β, IL-6, IL-8, TNF-α, CCL2 (MCP-1), CCL5 (RANTES), and ICAM-1 were increased in A549 cells cocultured with MT-2 cells, but not in A549 cells cocultured with CCRF-CEM cells at 3 and 5 days. The expression levels of most genes were decreased at 5 days after infection. The expression levels of TGF-β1 and CCL20 (MIP-3α) were increased in A549 cells cocultured with MT-2 cells at 5 and 3 days, respectively. Transcripts of IFN-γ and IL-10 were not detected in any of the samples. Transcripts of inducible nitric oxide synthase (iNOS) and IL-12 p40 were detected in control A549 cells, and expression levels were not different between samples (Figure 2C). Cytokine gene expression in NCI-H292 cells was also studied by RT-PCR, using cDNA samples prepared from cocultured cells. As shown in Figure 2D, high expression levels of IL-1α, IL-1β and TGF-β1 mRNA were detected in control NCI-H292 cells. However, the expression level of TNF-α was increased in NCI-H292 cells cocultured with MT-2 cells at 3 and 5 days.

The expression level of Tax mRNA was equivalent in A549 cells cocultured with MT-2 cells at 3 and 5 days, but the expression level of Tax protein was suppressed at 5 days (Figure 2B). Therefore, the expression of several cellular genes correlated with that of Tax. MT-2 cells expressed IFN-γ and IL-10 mRNA, but not IL-1β and IL-8 mRNA. However, A549 cells cocultured with MT-2 cells expressed IL-1β and IL-8 mRNA, but not IFN-γ and IL-10 mRNA, which suggests that residual MT-2 cells in these samples were not amplified.

We also investigated the production of IL-8 and CCL20 by A549 cells cocultured with MT-2 or CCRF-CEM cells. A549 cells cocultured with MMC-treated MT-2 cells released considerable amounts of IL-8 and CCL20 (Figure 3A). IL-8 was not detected in the media of MMC-treated MT-2 cells. We also measured the surface expression of ICAM-1 and LFA-1 on cocultured A549 cells by flow cytometry. Figure 3B shows that the fraction of cells expressing ICAM-1 but not LFA-1 was higher in A549 cells cocultured with MT-2 cells. Thus, consistent with the ability of HTLV-I to induce transcription of several cellular genes, infection of lung epithelial cells with HTLV-I increased the production of cytokines and chemokines and induced the surface expression of cell adhesion molecule.

Tax activates the expression of various cellular genes through the NF-κB and AP-1 pathways 33. Therefore, we investigated the transcriptional activity of NF-κB and AP-1. A549 cells cocultured with MT-2 cells exhibited high transcriptional activity of NF-κB and AP-1, compared with control A549 cells and A549 cells cocultured with CCRF-CEM cells (Figure 4A). Furthermore, viral promoter LTR was activated in A549 cells cocultured with MT-2 cells, suggesting that A549 cells were infected with HTLV-I. Next, we examined DNA-binding of NF-κB and AP-1. A549 cells were cocultured with MMC-treated MT-2 or CCRF-CEM cells, and the DNA-binding activity of NF-κB and AP-1 was assessed by EMSA. As evidenced from Figure 4B, coculture of A549 cells with MT-2 induced the DNA-binding of NF-κB and AP-1. These complexes were due to specific binding of nuclear proteins to each sequence because the binding activities were reduced by the addition of cold probe but not by an irrespective sequence. This gel shift assay detected an NF-κB complex that was supershifted by anti-p50, anti-p65 and anti-c-Rel Abs, and an AP-1 complex that was supershifted by anti-JunD Ab as was noted in HTLV-I-infected T-cell lines 3435 (Figure 4C).

Finally, we immunostained lung tissues of transgenic mice to assess the expression of viral antigen Tax. Lymphocytes accumulated in alveolar septa of the lungs of transgenic mice (Figure 5A; right lower panel), but not in littermate mice (data not shown). We examined the distribution of Tax protein in the lungs of transgenic mice. Marked expression of Tax was observed in epithelial cells (Figure 5A) as well as lymphocytes (Figure 5B) and macrophages (Figure 5C) in the lungs of transgenic mice. We next immunostained lung tissues obtained from patients with HTLV-I-related pulmonary diseases. Compatible with the lungs of HTLV-I-related pulmonary diseases, accumulation of lymphocytes was noted in alveolar septa (Figure 5D; right lower panel). Tax expression was noted in epithelial cells (Figure 5D), lymphocytes (Figure 5E) and macrophages (Figure 5F) of patients with HTLV-I-related pulmonary diseases, but not those of normal lungs (data not shown).