Adult T-cell leukemia cells overexpress Wnt5a and promote osteoclast differentiation

Abstract

Adult T-cell leukemia/lymphoma (ATL) is etiologically linked to infection with the human T-cell leukemia/lymphoma virus type 1 (HTLV-I). ATL is classified into 4 distinct clinical diseases: acute, lymphoma, chronic, and smoldering. Acute ATL is the most aggressive form, representing 60% of cases and has a 4-year survival of < 5%. A frequent complication and cause of death in acute ATL patients is the presence of lytic bone lesions and hypercalcemia. We analyzed the Wnt/β-catenin pathway because of its common role in cancer and bone remodeling. Our study demonstrated that ATL cells do not express high levels of β-catenin but displayed high levels of LEF-1/TCF genes along with elevated levels of β-catenin (LEF-1/TCF target genes) responsive genes. By profiling Wnt gene expression, we discovered that ATL patient leukemia cells shifted expression toward the noncanonical Wnt pathway. Interestingly, ATL cells overexpressed the osteolytic-associated genes-Wnt5a, PTHLH, and RANKL. We further show that Wnt5a secreted by ATL cells favors osteoclast differentiation and expression of RANK. Our results suggest that Wnt5a is a major contributing factor to the increase in osteolytic bone lesions and hypercalcemia found in ATL patients. Anti-Wnt5a therapy may prevent or reduce osteolytic lesions found in ATL patients and improve therapy outcome.

Figure 1

HTLV-I cell lines express varying levels of β-catenin but remain responsive to Wnt stimulation. (A) Western blot analysis of β-catenin expression from total cellular extracts of PBMCs, Jurkat, and HTLV-I cell lines. Extracts were normalized to actin expression. (B) Real-time PCR was performed on β-catenin cotranscription factors, TCF-1, TCF-4, and LEF-1 from cDNA derived from HTLV-I cell lines. Real-time PCR was performed in duplicate and samples were normalized to GAPDH expression. Fold change was calculated by comparing values with resting PBMCs (R.PBMCs). (C) Real-time PCR expression of Wnt/β-catenin downstream target genes, c-myc, survivin, VEGF, c-jun, COX-2, and cyclin D1 from total cDNA from HTLV-I cell lines. Real-time PCR was performed in duplicate. Fold change was calculated by comparing values with Jurkat normalized gene expression. (D-F) PCR amplification of typical canonical Wnts (D), noncanonical Wnts (E), and their coreceptors (F) from cDNA derived from HTLV-I–transformed (MT2 and MT4) and immortalized (1185 and LAF) cell lines. Resting PBMCs (R.PBMCs) and the non–HTLV-I T-cell line, Jurkat, were used as controls. GAPDH amplification in nonsaturating conditions served as a control for the quality and quantity of the samples.

Figure 2

Tax and p30 increase β-catenin transcriptional activity through AKT and/or p38 pathways. (A) Western blot analysis of phosphorylated GSK3-β (Ser9) expression from total cellular extracts of R.PBMCs, Jurkat, and HTLV-I cell lines. Extracts were normalized to actin expression. (B) Western blot analysis of total cellular extract for phosphorylated GSK3-β (Ser9) from 293T cells transfected with pc-Tax (0.5 µg) or PMH-p30 (2.0 µg). Extracts were analyzed 48 hours after transfection and normalized to actin expression. Western blots were stripped and reprobed for GSK3-β expression. Tax and HA-p30 expression were demonstrated in the transfected lysates. (C) PCR analysis of Tax and p30 expression in HTLV-I cell lines. Jurkat cDNA was used as a negative control. (D) pc-Tax (0.1, 0.2, and 0.5 µg) and PMH-p30 (0.5, 1.0, and 2.0 µg) were transfected into 293T cells along with the TOPflash luciferase reporter plasmid. Cellular extracts were used in luciferase assays to measure the level of β-catenin activity. All experiments were performed in duplicate, and values represent the average reading normalized to Renilla luciferase or protein concentration. Error bars represent the population standard deviation for each sample. (E) 293T cells were transfected with or without 1.0 µg PMH-p30, along with 0.1 µg of wild-type or mutant Tax and the TOPflash reporter plasmid, and analyzed as in (D). (F) 293T cells were transfected as in (B) and analyzed for β-catenin expression by real-time PCR. Real-time PCR was performed in duplicate, and samples were normalized to GAPDH expression. Fold change was calculated by comparing values with PMH-transfected cell–normalized gene expression. (G) 293T cells were transfected as in (D). Total extracts were analyzed 48 hours after transfection and normalized to actin expression. Tax and HA-p30 expression were demonstrated in the transfected lysates. (H-I) 293T cells were transfected in duplicate with 0.1 µg pc-Tax (H) or 1.0 µg PMH-p30 (I), along with 1.0 µg of dominant-negative AKT (DN AKT) or dominant-negative p38 (DN p38), and the TOPflash reporter plasmid. Values represent the average reading normalized to protein concentration. Error bars represent the population standard deviation for each sample.

Figure 3

ATL patient samples express high levels of β-catenin cotranscriptional genes, TCF-1 and LEF-1. (A) Western blot analysis of β-catenin, phosphorylated GSK3-β (Ser9), and STAT3 expression from total cellular extracts of 6 ATL patient samples. Extracts were normalized to actin expression. (B) Real-time PCR analysis of β-catenin expression from cDNA from ATL patients. Samples were normalized to GAPDH expression, and the fold change was calculated by comparing values with R.PBMCs’ normalized gene expression. (C) ATL cell lines were treated with MG132 (2.5 µM) or dimethyl sulfoxide control for 8 hours, followed by Western blot analysis on total cellular extract with β-catenin antibody. The upper band corresponds to mono-ubiquitinated β-catenin, the middle band corresponds to wild-type β-catenin, and the lower band corresponds to degraded β-catenin after MG132 treatment. (D) ATL cell lines were treated with 10 mM LiCl for 24 hours. Total cell lysates were normalized to actin expression before Western blot analysis with β-catenin or p-GSK3β (Ser9). (E) Real-time PCR expression of Wnt/β-catenin downstream target genes, c-myc, survivin, VEGF, c-jun, COX-2, and cyclin D1from total cDNA from ATL patients. Fold change was calculated by comparing values with Jurkat-normalized gene expression. The ATL patient cDNAs were additionally analyzed for Tax and p30 expression by qualitative PCR. Jurkat and LAF cDNA were used as negative and positive controls, respectively. (F) Real-time PCR for TCF-1, TCF-4, and LEF-1 from cDNA derived from ATL patients. Real-time PCR was performed in duplicate and samples were normalized to GAPDH expression. Fold change was calculated by comparing values with R.PBMC-normalized gene expression. (G) Cell cycle analysis of 1185 cells treated with or without IL-2, serum, and hydroxyurea after 48 hours. (H) Cells were treated as in (G) and analyzed by real-time PCR analysis. Real-time PCR was performed in duplicate and samples were normalized to GAPDH expression. Fold change was calculated by comparing values with nontreated 1185 cells’ normalized gene expression.

Figure 4

Wnt gene profiling demonstrates elevated expression of Wnt5a in ATL patient samples. (A-C) PCR amplification of typical canonical Wnts (A), noncanonical Wnts (B), and their coreceptors (C) from cDNA derived from ATL patient samples. R.PBMCs were used as controls. GAPDH amplification in nonsaturating conditions served as a control for the quality and quantity of the samples. Open boxes highlight genes differentially expressed between R.PBMCs and ATL patients. (D-F) Real-time PCR analysis of Wnt5a, PTHLH, and RANKL expression from cDNA derived from ATL patients. Samples were normalized to GAPDH expression and the fold change was calculated by comparing values to R.PBMCs normalized gene expression.

Figure 5

Wnt5a is involved in osteoblast differentiation in ATL cells. (A) Real-time PCR analysis of Wnt5a expression from cDNA derived from ATL cell lines. Samples were normalized to GAPDH expression and the fold change was calculated by comparing values with R.PBMCs’ normalized gene expression. Jurkat expression served as a negative control. (B) C2C12 and doxycycline-treated MT1-DKK1 or MT1-pTRIPZ cells were cultured in 0.4 µM Transwell plates with or without 100 ng/mL rBMP-2. After 5 days, C2C12 cells were lysed and used in alkaline phosphatase assays. ALP activity was measured as the amount of pNP generated in µmol per volume of the sample per reaction time and normalized to protein concentrations. Results are representative of 2 independent experiments. (C) MT1 cells stably expressing DKK-1 were induced with or without 2 µg/mL doxycycline (Dox) for 72 hours. Real-time PCR was performed on cDNA for DKK-1 expression and normalized to GAPDH expression. Results are representative of 2 independent inductions. (D) C2C12 cells were cocultured with MT1-pTRIPZ cells on 0.4 µM Transwell plates. Three hours after culturing, 50 ng/mL rWnt3a was added and cells were grown at subconfluence for an additional 48 hours. RNA was extracted from C2C12 cells and used for real-time PCR analysis. Results are representative of 2 independent experiments. Samples were normalized to GAPDH expression and the fold change was calculated by comparing values with C2C12 cells without rWnt3a-normalized gene expression. (E) Model demonstrating the C2C12 Transwell system used in the study. MT1 cells were seeded in a Transwell insert and C2C12 cells were seeded at subconfluency on the bottom of the plate. Secreted Wnt5a was blocked by the addition of anti-Wnt5a antibody. (F) C2C12 cells were cultured in a Transwell plate with and without MT1-TRIPZ cells and with or without Wnt5a antibody. Three hours after the addition of MT1-pTRIPZ cells and Wnt5a antibody, 30 ng/mL of rRANKL or control was added. Cells were cultured for a further 48 hours, after which RNA was extracted and used for real-time PCR. Results are representative of 2 independent experiments. Samples were normalized to GAPDH expression and the fold change was calculated to control normalized gene expression.