Induction of the IL-9 gene by HTLV-I Tax stimulates the spontaneous proliferation of primary adult T-cell leukemia cells by a paracrine mechanism

Abstract

The etiologic agent of adult T-cell leukemia (ATL) is human T cell lymphotropic virus type I (HTLV-I). The HTLV-I protein Tax alters gene expression, including those of cytokines and their receptors, which plays an important role in early stages of ATL. Here we demonstrate that expression of interleukin-9 (IL-9) is activated by Tax via an NF-kappaB motif in its proximal promoter, whereas IL-9 receptor-alpha (IL-9Ralpha) expression is not induced by Tax. However, supporting a role for IL-9/IL-9Ralpha in ATL, a neutralizing monoclonal antibody directed toward IL-9Ralpha inhibited ex vivo spontaneous proliferation of primary ATL cells from several patients. Fluorescence-activated cell sorter analysis of freshly isolated peripheral blood mononuclear cells from these patients revealed high level expression of IL-9Ralpha on their CD14-expressing monocytes. Furthermore, purified T cells or monocytes alone from these patients did not proliferate ex vivo, whereas mixtures of these cell types manifested significant proliferation through a contact-dependent manner. Taken together, our data suggest that primary ATL cells, via IL-9, support the action of IL-9Ralpha/CD14-expressing monocytes, which subsequently support the ex vivo spontaneous proliferation of malignant T cells. In summary, these data support a role for IL-9 and its receptor in ATL by a paracrine mechanism.

Figure 1

IL-9 was secreted into the supernatants of ATL PBMCs and HTLV-I–infected cell lines. (A) NK-92 cell line assay of 6-day culture supernatants of PBMCs from smoldering and chronic ATL patients were performed. Monoclonal antibodies to IL-2 or IL-9 or to both were added to the assay to detect the presence of IL-2 or/and IL-9 in the 6-day culture supernatants. The normal control was typical of that observed from 10 normal donors. (B) IL-9 ELISA was performed to define the IL-9 levels in the 6-day culture supernatants of PBMCs from 11 patients with ATL. The growth medium and the supernatants from normal donor PBMCs culture were used as a control (n = 10, patient vs control, P < .001). (C) An IL-9 ELISA was performed to determine the IL-9 levels in the culture supernatants of HTLV-I–infected cell lines Hut102, MT-1, MT-2, and MJ as well as HTLV-I–negative cell lines Jurkat, CEM, and Hut78.

Figure 2

HTLV-I Tax-transactivated IL-9 expression in Jurkat T cells. (A) IL-9 mRNA levels after Tax induction were detected by Taqman real-time PCR. The copy number of IL-9 mRNA was normalized by the copy number of hypoxanthine guanine phosphoribosyl transferase 1 (HPRT1) mRNA. The data are representative of 3 independent experiments. (B) Western blot analysis using a monoclonal antibody to Tax for relative expression of wild-type HTLV-I Tax (JPX-9) and nonfunctional mutant Tax (JPX-m) in JPX-9 and JPX-m before and after addition of 20 μM CdCl₂. (C) Taqman real-time analysis of HTLV-I Tax and IL-9 message in ATL PBMCs ex vivo cells.

Figure 3

Activation of the IL-9 promoter was mediated by NF-κB and the IL-9 NF-κB motif ATGTCAGGGTTTTTCCGTGTTTG bond to NK-κB in the gel shift assay. (A) Schematic representation of the IL-9 luciferase reporter constructs P1, P2, P3, P4, and P5 (ΔNF-κB). (B) IL-9 luciferase reporter construct P1 (10 μg) and CMV-Renilla (1 μg) were transfected into JPX-9 and JPX-m cells. The Tax expression was induced by addition of CdCl₂, and the promoter activities were assayed at different time points after CdCl₂ addition. The promoter activities were normalized by the Renilla value. The results are representative of 3 independent experiments. (C) 10 μg of the IL-9 luciferase construct and 1 μg CMV-Renilla were transfected into Tax expressing Hut102, MT-1, MT-2, and Tax nonexpressing Jurkat, CEM cells by electroporation. Dual-luciferase assays were performed 48 hours later. Normalized results using Renilla values are representative of 3 independent experiments. Experimental variations are indicated by SE bars. (D) The IL-9 luciferase constructs were transfected into Jurkart and CEM T cells in the absence (pMT2T, empty vector) or in the presence of 10 μg of the Tax expression construct pMT2T-Tax by electroporation. Dual-luciferase assays were performed following the manufacturer's recommendations. The data are representative of 3 independent experiments. (E) Gel shift assay. Extracts obtained from 293T cells transfected with p65 and p50 expression constructs were used for the binding of cNF-κB AGTTTGAGGGGACTTTCCCAGGC and IL-9 NF-κB ATGTCAGGGTTTTTCCGTGTTTG (the underlined sequences are the NF-κB binding sites). The typical p50/p65 heterodimer and p50/p50 homodimer can be readily seen with cNF-κB, whereas IL-9 NF-κB forms a complex that comigrates with the p50/p50 homodimer of the cNF-κB, as shown by an arrow (lanes 2, 6). The binding of cNF-κB to p50/p65 and p50/p50 complexes can be competed off specifically by addition of a 50-fold molar excess of unlabeled IL-9 NF-κB probe (lane 3) but was not affected by addition of a 50-fold molar excess of unlabeled nonspecific SP1 probe (lane 4). Similarly, the binding of IL-9 NF-κB to the p50/p50 complex was specifically competed off by addition of a 50-fold molar excess of unlabeled cNF-κB probe (lane 7) and was not affected by addition of a 50-fold molar excess of unlabeled nonspecific SP1 probe (lane 8). Lanes 1 and 5 represent the negative controls. (F) Gel shift assay. Nuclear extracts from the HTLV-I–positive cell line Hut102 were used for the binding of cNF-κB and IL-9 NF-κB. The patterns of cNF-κB and IL-9 NF-κB binding are very similar to panel E. A vertical line has been inserted to indicate a reposition of gel lanes.

Figure 4

HTLV-I Tax did not transactivate IL-9Rα. (A) Schematic representation of IL-9Rα promoter constructs RP1, RP2, and RP3. (B) IL-9Rα promoter assay in HTLV-I–positive Hut102, MT-1, MT-2 cells and HTLV-I–negative Jurkat, CEM cells. The promoter activities were normalized by comparison to the Renilla value. The results are representative of 3 independent experiments. (C) FACS analysis of surface expression of IL-9Rα on HTLV-I-positive Hut102, MT-1, MT-2 cells and HTLV-I–negative Jurkat and CEM cells. PE-conjugated mouse IgG_2b was used as the control.

Figure 5

Antibody directed to IL9Rα inhibited the spontaneous proliferation of ATL PBMCs ex vivo. (A) Six-day spontaneous proliferation of ATL PBMCs with and without antibody to IL9Rα in the ex vivo culture. A nonspecific antibody, UPC10, was used as a control. The monoclonal antibody to IL9Rα or UPC10 was added to the 96-well plates at day 0. ³H thymidine was added to the culture during the last 6 hours of culture. Cells were then harvested and analyzed for the ³H thymidine incorporation. (B) CD4⁺CD25⁺ T cells from ATL patients proliferated in ex vivo culture. FACS analysis of CD4/CD25 expression was performed at different time points (day 0, day 3, and day 6). (C) CFSE staining of CD3⁺ lymphocytes to monitor their cell division. CFSE was labeled at day 0, and the cells were then put in culture without any stimulation. FACS analysis of CFSE-positive or CD25 CFSE double-positive cells was done at day 6. Unlabeled cells were used as controls. (D) CFSE staining of monocytes (CD14-expressing cells) to monitor cell division.

Figure 6

IL-9Rα–expressing monocytes were required for the spontaneous proliferation of ATL PBMCs through a contact-dependent manner. (A) IL-9Rα was expressed on the monocyte population of ATL PBMCs, which also expressed CD14. IL-9Rα was not expressed on CD3 (T), CD20 (B), or CD16 (NK) positive cells. PBMCs from normal donors were used as a control. (B) Six-day spontaneous proliferation of purified T cells, monocytes, and mixture of purified T cells and monocytes (T cells: monocytes = 1:1). ³H thymidine was added to the culture during the last 6 hours of culture. Anti–IL-9Rα or control antibody UPC10 was added to the culture at T₀. Cells were then harvested and analyzed for their ³H thymidine incorporation. (C) Separated T cells and monocytes were cultured in the same chamber or different chambers of the transwell (0.4 μm) for 6 days. ³H thymidine was added to the culture during the last 6 hours of culture.

Figure 7

Genes encoding adhesion molecules, MHC class I and class II, and cytoskeleton binding proteins were up-regulated in monocytes derived from ATL patients. The gene expression of monocytes from 3 ATL patients were compared with the gene expression of monocytes derived from 3 normal healthy donors. The monocytes were purified using negative selection. cDNA from ATL monocytes or normal monocytes was labeled with Cy5 and hybridized with human universal reference cDNA labeled with Cy3. Red indicates Cy5/Cy3 ratio more than 1, green indicates Cy5/Cy3 ratios less than 1, black indicates no significant change in gene expression, and gray indicates the spot did not meet data selection criteria. These ratios were depicted according to the color scale shown at the bottom. Three sets of functional related genes, including genes encoding adhesion molecules, MHC class I and MHC class II gene, and genes encoding cytoskeleton binding proteins, which had at least 2-fold increases in the ATL monocytes.