The patient was a 61-year-old Japanese male who was admitted because of poor performance status. He had systemic lymphadenopathy, hepatosplenomegaly and generalized erythematous skin. Laboratory data indicated hypercalcemia, renal failure and positivity of anti-HTLV-I antibody. The peripheral blood white cell count was elevated and contained 61% of flower-like cells whose immunological phenotype was CD2(+), CD3(+), CD4(+), CD8(+), CD25(+), CD7(-) and CD19(-). From these clinical data, he was diagnosed with an acute type of ATLL. He received various combinations of chemotherapies, but had no remission and died. The peripheral mononuclear cells were frozen and preserved for the further analyses and the establishment of HUATAK.

HUVEC was purchased from Lonza Walkersville, Inc. (Walkersville, MD, USA) and was maintained in EGM-2 medium with a Bullet Kit (CC-3162), which contains a growth supplement kit with hydrocortisone, hFGF-b, VEGF, R3 IGF-1, ascorbic acid, heparin, FBS, hEGF and GA-1000 (Lonza Walkersville, Inc.).

ATLL cells were co-cultivated with HUVEC using Iscove’s modified Dulbecco’s medium (IMDM) containing 20% human plasma, 10 ng/ml of IL-2 (PeproTech), and 20 nM damnacanthal (Calbiochem, Merck, Tokyo, Japan).

The established cell line, HU-ATTAK will be deposited to Riken Bioresource Center for academic use. (http://www.brc.riken.go.jp/lab/cell/english/).

Surface immunophenotyping of the lymphoma, HU-ATTAK and HUVEC was performed using the following panel of lymphoid-associated monoclonal antibodies: HLA-DR (CR3/43), CD4 (RPA-T4), (DAKO Denmark), CD2 (RPA-2.10), CD3 (UCHT1), CD5 (L17F12), CD8 (SK1), CD10 (WM15), CD13 (WM15), CD14 (M5E2), CD19 (SJ25C1), CD20 (L27), CD25(M-A251), CD33 (P67.6), CD45 (2D1), CD56 (NKH-1), CD57 (NK-1), TCRαβ(WT31), TCRγδ (11F2), (BD Pharmingen, Japan BD, Tokyo, Japan), CD1a (BL6), CD7 (8H8.1), (Beckman Coulter, Inc. CA, USA), CD134(W4-3) and CD252(TAG-34) (Medical & Biological Laboratories, Nagoya, Japan). Multiparameter analysis of gated cell populations during cell suspension studies was used to yield more definitive immunophenotypic information (CELLQuest software, Becton-Dickinson, CA, USA).

Chromosomal analysis of the original leukemic cells and HU-ATTAK was performed at the Special Reference Laboratory (Tokyo, Japan). The analysis of HU-ATTAK was performed with conventional methods and karyotyped according to ISCN.

For the extraction of high molecular weight DNA, the original bone marrow tumor and HU-ATTAK cells were treated with proteinase K and extracted with phenol and chloroform as described elsewhere [17Kagami Y, Jung J, Choi YS, et al. Establishment of a follicular lymphoma cell line (FLK-1) dependent on follicular dendritic cell-like cell line HK Leukemia 2001; 15: 148-56.]. Data of Southern blot analysis for the HTLV-I integration site were acquired from the Special Reference Laboratory.

DNA labelling, array fabrication, hybridization, normalization and analysis were performed as described previously [18Ota A, Tagawa H, Karnan S, et al. Identification and characterization of a novel gene, C13orf25, as a target for 13q31-q32 amplification in malignant lymphoma Cancer Res 2004; 64: 3087-95.]. The array CGH glasses used in the present study were ACC Version 5.0 [19Nakagawa M, Nakagawa-Oshiro A, Karnan S, et al. Array comparative genomic hybridization analysis of PTCL-U reveals a distinct subgroup with genetic alterations similar to lymphoma-type adult T-cell leukemia/lymphoma Clin Cancer Res 2009; 15: 30-8.]. The array consisted of 2304 BAC/PAC clones spotted in duplicate. These clones were from libraries RP11 and RP13 for BAC clones, and RP1, RP3, RP4 and RP5 for PAC clones. These clones were obtained from the BAC/PAC Resource Centre at the Children’s Hospital, Oakland Research Institute, Oakland, CA, USA (http://bacpac.chori.org/). The BAC/PAC clones were aligned with each chromosome on the basis of data from Ensembl Genome Data Resources (release 40; http://www.ensembl.org/) or the National Centre for Biotechnology Information (Build 36; http://www.ncbi.nlm.nih.gov/). The locations of all the clones used for array CGH were confirmed by fluorescent in situ hybridization (FISH). Clones that showed unreliable data in 17 unrelated normal individuals were excluded from the data analysis. Clones of the sex chromosomes were also excluded. A total of 2125 clones (covering 2867 Mb, 1.34-Mb resolution) were further analyzed.

Thresholds for copy number gain and loss were determined so that the estimated false discovery rate (FDR) was approximately 0.05. The thresholds for gain and loss were +0.186868 and −0.144762, respectively. We defined the region of gains and losses as: (a) three contiguous clones showing gain (log2 ratio +0.19 to +1.0) or loss (log2 ratio −1.0 to −0.15); or (b) clones, if not contiguous, showing high copy number gain (log2 ratio> + 1.0) or homozygous loss (log2 ratio <−1.0).

HUVEC were seeded in a 12-well cluster dish (Costar, Cambridge, MA, USA) on day -1. On day 0, after removal of the medium, 2 ml of a HU-ATTAK cell suspension at a concentration of 5 X 10⁴/ml was plated onto each well. Every 3 days, floating viable cells after pipetting were concentrated to 1 ml and the number of viable cells was counted. Then 10⁵ HU-ATTAK cells were resuspended in 2 ml of fresh medium and seeded into new wells at a concentration of 5 X 10⁴/ml. For the IL-2-dependent growth assay, 10⁵ HU-ATTAK cells were cultured in a 12-well cluster in the presence or absence of 10 ng/ml IL-2. The viable cells were counted after three days.

Dose dependency of IL-2 was examined as follows: 5 X 10³ HU-ATTAK cells in the presence of HUVEC were cultured in 2 ml of culture medium containing IL-2 in a 12-well cluster dish. The IL-2 was serially diluted from 270 ng/ml by one-thirds. After three days, 100 µl of cell suspension from each well was transferred to a 96-well cluster dish. The number of viable cells was measured using a colorimetric assay (CellTiter 96TM; Aqueous Non-Radioactive Cell Proliferation Assay Kit, Promega, Madison, WI, USA). The mean values of the absorbance at 490 nm in the quadruplicate wells were plotted as a growth curve.

10⁵ HU-ATTAK cells were cultured on HUVEC with or without IL-2 for 2 days. These cells were washed and stained with FITC-labeled Annexin V (AN) and propidium iodide (PI) (Trevigen Inc., Gaithersburg, MD, USA). Early apoptotic cells which express phosphatidylserine on the outer cell membranes bind AN, and dead cells with a compromised cell membrane allow PI to bind to the cellular DNA. Using flow cytometric analysis, the degrees of apoptosis were measured with two populations of AN+/PI- (early apoptotic cells) and AN+/PI+ (late apoptotic cells and necrotic cells).

A Transwell culture was established using a 0.4-µm pored bottom cup (Transwell® Costar, Cambridge, MA, USA). On day -1, HUVEC were seeded into a 6-well cluster dish (Costar). On day 0, 1.5 ml of the culture medium was replenished, the pored bottom cups were settled, and then 10⁵ HU-ATTAK cell suspensions were inoculated. The number of viable cells was counted on days 3, 5 and 7.

10³ HUVEC/well were cultured in a 96-well cluster dish (Costar) on day -1. On day 0, after removal of the medium, 100 µl of HU-ATTAK suspension at a concentration of 5 X 10⁴/ml was plated onto each well which contained various concentrations of monoclonal anti-OX40 or anti-OX40 ligand antibody in quadruplicates. Viable cell numbers of each cell suspension were counted every three days after pipetting. The cell density was maintained at 5 X 10⁴/ml, and 5 X 10³ HU-ATTAK cells were seeded into a new well of a 96-well cluster dish containing precultured HUVEC. If the cell density was less than 5X10⁴/ml, the whole cell suspension was centrifuged and the medium was changed to a new medium, then seeded into a new well.

The peripheral mononuclear cells frozen prior to the start of chemotherapy, were thawed and cultured in IDMEM with 20% human plasma and 10 ng/ml of human IL-2 in the presence or absence of HUVEC. However, the cells did not grow for long under either condition. During the search for agents to stimulate the growth of T-cell lymphoma in vitro, we found that the addition of 20 nM damnacanthal to the medium stimulated the growth of leukemic cells for some ATLL cases in the presence of IL-2 (Fig. 1A). We therefore added 20 nM damnacanthal to the IDMEM culture together with 20% human plasma and 10 ng/ml IL-2 without HUVEC. However, this condition did not support growth and the tumor cells ceased to grow. When HUVEC was added to this condition on September 2007, mononuclear cells kept growing vigorously for more than 4 months. Interestingly, after 4 months, the growth rate was identical whether damnacanthal was present or absent, although tumor cell growth was strictly dependent on the presence of IL-2 and HUVEC (Fig. 1B). This cell line which continued to grow for 9 months from the start of culture was designated as HU-ATTAK.

Fig. (1) (A) Representative growth curve of an ATLL case in the presence (●) or absence (■) of 20 nM damnacanthal without HUVEC. A cell line was not established in this case. (B) Representative growth curve of HU-ATTAK on HUVEC cells in the presence (■) or absence (▲) of 20 nM damnacanthal. A similar growth curve was observed in an independent experiment. Cell numbers are plotted using a log scale.

The phenotypic analysis of the original leukemia and HU-ATTAK showed expression of CD3, CD4 and CD25, which are characteristic of ATLL. CD8 was positive in the original leukemic cells, but it became negative in the new cell line. The analyses of other phenotype of HU-ATTAK indicated that CD1(-), CD2(+), CD5(+), CD7(-), CD10(-), CD13(+/-), CD19(-), CD56(-), CD57(+/-), HLA-DR(+), TCR-αβ(-), and TCR-γδ(-), which showed some aberrations from normal helper T-cells.

The proviral integration pattern of the original peripheral mononuclear cells and HU-ATTAK identified by means of Southern blot analysis showed that the clonal integration pattern of HTLV-I and the two flanking sequence bands were identical. However, the bands of the internal sequence of the gag-env gene were partially deleted in HU-ATTAK, which suggested that microdeletion occurred during establishment of the HU-ATTAK cell line (Fig. 2A).

Fig. (2) (A) Southern blot analysis of HTLV-I integration pattern for mononuclear cells and HU-ATTAK. Left panel shows patient PBMNC and the right panel shows the HU-ATTAK cell line derived from the same patient. E, EcoRI digestion and P, Pst1. The bands of the flanking sequence (thick arrows) are identical to each other. The three bands of internal sequences of HTLV-I (dotted arrows) were deleted, while a new band (open arrow) was detected in HU-ATTAK. (B) CGH analysis of original PBMNC and HU-ATTAK. The loci of amplification and deletion are shown.

Chromosomal analysis showed that the karyotype of the original leukemic cells was idic(18)(p11.2), i(19)(p10) and +mar1, which was maintained in HU-ATTAK. Array CGH analysis was used for the critical analysis of constitutive DNA abnormalities. The major findings detected in the original leukemic cells were regions of gain in 1q21-25, 1q32-44, 14q32.2, 18q11-23 and 19p13.11, and a region of loss in 18p11 (Fig. 2B). The majority of these aberrations were conserved in HU-ATTAK. The results of these analyses proved that HU-ATTAK was derived from the original leukemic cells.

After establishment of the HU-ATTAK cell line on HUVEC in the presence of IL-2, we examined to what extent this cell line depended for its growth on each of the factors. First, IL-2 was removed from the culture and as a consequence HU-ATTAK cells stopped growing, and almost all the cells died within three days even in the presence of HUVEC cells (Fig. 3A). The cell death mechanism in the absence of IL-2 was examined with the Annexin V/PI staining method (Fig. 3B). The cultured cells were stained with FITC-labeled annexin V and propidium iodide (PI) and analyzed using a FACS scan 24, 48 and 72 hours later. Annexin V positive apoptotic cells increased from 24 hours, which demonstrated the anti-apoptotic effect of IL-2.

Fig. (3) Promotion and inhibition of HU-ATTAK cell growth by IL-2. (A) HU-ATTAK cell death without IL-2. HUVEC was cultured in a 12-well cluster dish on day -1. 105 cells of HUATTAK were cultured with or without HUVEC or IL-2. The numbers of viable cells were counted on day 3. This is a representative of two independent experiments. (B) Apoptotic cell death of HU-ATTAK in the absence of IL-2. Flow cytometric analysis of apoptosis after 48 hours from the deprivation of IL-2 is shown. The fluorescence intensities of AN (abscissa) and PI (vertical) are plotted. A representative of three experiments is shown. (C) Dose response curve of various concentrations of IL-2 in the presence of HUVEC. HUVECs were cultured in a 12-well cluster dish on day -1. 105 cells of HU-ATTAK were cultured in 2 ml of medium containing serially diluted IL-2. The mean value of quadruplicate assays is shown. This is a representative of two independent experiments.

The growth promotion by IL-2 was examined in the presence of serially diluted concentrations of IL-2, and dose dependence was observed for concentrations from 370 pg/ml to 30 ng/ml. The growth rate was less than twice that obtained with the minimum concentration, indicating that the growth promotion activity of IL-2 was weak (Fig. 3C). These findings indicated that IL-2 had a potent anti-apoptotic effect and a weak growth promotion action.

Next, HUVEC cells were removed from the culture. Fig. (4) shows the growth curve of HU-ATTAK cells in the presence of IL-2 but without HUVEC. In the culture without HUVEC, the growth rate was similar to that of the control culture for three days, after which proliferation stopped and the cells survived for 2 weeks. These findings indicated that HUVEC promoted the proliferation of HU-ATTAK in the presence of IL-2.

Fig. (4) Growth curve of HU-ATTAK cells in the presence or absence of HUVEC. 105 cells of HU-ATTAK with IL-2 were cultured with or without HUVEC. The numbers of viable cells were counted every 3 days. This is a representative of two independent experiments. Cell numbers are plotted using a log scale.

We examined whether direct contact with HUVEC cells is required for HU-ATTAK to grow. HU-ATTAK was cultured in an inner cup separated from HUVEC cells by a 0.4-µm microporous membrane to prevent direct cell-to-cell contact. In the absence of direct contact, HU-ATTAK stopped growing a few days later and finally died. These findings suggested that the molecule(s) expressed on the surface of HUVEC could sustain HU-ATTAK growth (Fig. 5).

Fig. (5) Separation of HUVEC and HU-ATTAK prevents the growth of HU-ATTAK. HU-ATTAK and HUVEC were separated with a micropore membrane and cultured. As a control, HUATTAK cells were cultured with HUVEC without a membrane. The numbers of viable cells in the three wells were counted on days 4, 7, 10, 13 and 16. This is a representative of two independent experiments. Cell numbers are plotted using a log scale.

HUVEC expresses the OX40 ligand, which is a molecule that acts against ATLL cells by means of adhesion and anti-apoptosis [20Imura A, Hori T, Imada K, et al. OX40 expressed on fresh leukemic cells from adult T-cell leukemia patients mediates cell adhesion to vascular endothelial cells: implication for the possible involvement of OX40 in leukemic cell infiltration Blood 1997; 89: 2951-8., 21Kunitomi A, Hori T, Maeda M, Uchiyama T. OX40 signaling renders adult T-cell leukemia cells resistant to Fas-induced apoptosis Int J Hematol 2002; 76: 260-6.]. OX40 ligand expression was limited to HUVEC while OX40 expression was limited to HU-ATTAK (Fig. 6A). We examined the effect of the OX40 ligand on HU-ATTAK growth by adding an antibody to the culture system. In the presence of the anti-OX40 ligand antibody, the growth of HU-ATTAK was inhibited to the same level as that in the culture without HUVEC (Fig. 6B). On the other hand, the inhibitory effect was weak in the presence of the anti-OX40 antibody (Fig. 6C). These findings indicated that growth promotion by HUVEC is dependent on its expression of the OX40 ligand.

Fig. (6) Growth inhibition by anti-OX40 ligand and anti-OX40 antibody. (A) Expression of OX40 ligand or OX40 on HUVEC or HUATTAK as determined by flow cytometry. (B) Anti-OX40 ligand antibody completely inhibits the growth of HU-ATTAK. HU-ATTAK cells in quadruplicate culture were cultured in the presence of 0 µg/ml (□), 1 µg/ml (▲), 10 µg/ml(●) and 30 µg/ml(■) of OX40 ligand antibody. The negative control was the absence of HUVEC (○). Cells were counted on days 3, 6 and 8. This is a representative of two independent experiments. Cell numbers are plotted using a log scale. (C) Anti-OX40 antibody partially inhibits the growth of HU-ATTAK. HU-ATTAK cells in quadruplicate culture were cultured in the presence of 0 µg/ml (□), 1 µg/ml (▲), 10 µg/ml (●) and 30 µg/ml (■) of OX40 antibody. The negative control was without HUVEC (○). Cells were counted on days 3, 6 and 8. This is a representative of two independent experiments. Cell numbers are plotted using a log scale.