First Department of Internal Medicine, University of Occupational and Environmental Health, Japan, School of Medicine, Kitakyushu, Japan
Address reprint requests to: Yoshiya Tanaka, MD, First Department of Internal Medicine, University of Occupational and Environmental Health, Japan, School of Medicine, 1-1 Iseigaoka Yahatanishi-ku, Kitakyushu, 807-8555 Japan
Hypercalcemia is observed in >80% of ATL. Serum MIP-1α levels were elevated in all 24 ATL with hypercalcemia but undetectable in all 10 patients with humoral hypercalcemia of malignancy with solid tumors and in 34 of 37 ATL without hypercalcemia. We propose that serum MIP-1α is a clinical hallmark for hypercalcemia in ATL.
Introduction: High serum cytokines levels are not always associated with hypercalcemia in patients with adult T-cell leukemia (ATL), suggesting that other factors are involved in the pathogenesis of ATL patients with hypercalcemia. This study was designed to determine the role of macrophage inflammatory protein-1α (MIP-1α), a chemokine recently described as an osteoclast stimulatory factor, in ATL-associated hypercalcemia.
Materials and Methods: We measured serum interleukin (IL)-1β, IL-6, TNF-α, parathyroid hormone-related protein (PTHrP), and MIP-1α levels in ATL patients by enzyme-linked immunosorbent assays. FACScan was used to measure the expression of RANKL on ATL cells. Osteoclast formation in cocultures of ATL cells and peripheral blood mononuclear cells (PBMCs) was evaluated by TRACP staining.
Results: High serum MIP-1α levels were noted in all 24 ATL patients with hypercalcemia and in 3 of 37 ATL patients without hypercalcemia. The elevated levels of MIP-1α and calcium in ATL patients decreased after effective chemotherapy, emphasizing the role of MIP-1α in ATL hypercalcemia. ATL cells spontaneously produced MIP-1α. MIP-1α significantly enhanced human monocyte (precursor cells of osteoclasts) migration and induced RANKL expression on ATL cells. ATL cell-induced osteoclast formation from PBMCs was inhibited by anti-MIP-1α antibody and osteoprotegerin.
Conclusion: Our results suggest that MIP-1α can induce RANKL on ATL cells in autocrine fashion and that RANKL seems to mediate the hypercalcemic effect of MIP-1α in ATL. We propose that MIP-1α is the clinical hallmark of hypercalcemia in ATL and could be a potentially useful therapeutic target.
ADULT T-CELL LEUKEMIA (ATL) is a mature CD4+ T-cell malignancy caused by infection with human T-lymphotrophic virus (HTLV)-I and is associated with a marked increase of peripheral leukemic cells and monoclonal growth during the acute phase. Mechanisms of hypercalcemia in malignancy have been explored recently. The high frequency of hypercalcemia is the most striking feature of ATL; >70% of ATL patients have hypercalcemia during the clinical course of the disease, particularly during the aggressive stage of ATL.(1) Such a frequency is the highest among malignancies, and hypercalcemia is more severe in patients with ATL than in those with other malignancies.(2)
Several pathological studies of ATL patients with hypercalcemia have indicated that the latter is caused by increased number of osteoclasts and accelerated bone resorption.(1) Several cytokines, such as interleukin-1 (IL-1)(3) and parathyroid hormone-related protein (PTHrP),(4) have been implicated in ATL-associated hypercalcemia. We previously reported that ATL cells proliferate and produce IL-1 in a calcium-dependent manner in vitro.(5–9) Among these factors, PTHrP is considered to play an important role in stimulating osteoclasts, resulting in increased bone resorption. Immunodeficient mice implanted with leukemic cells from patients with ATL exhibited hypercalcemia and overexpressed PTHrP.(10) However, PTHrP cannot directly induce the differentiation of hematopoietic precursor cells to osteoclasts.(11) Furthermore, high levels of PTHrP in the serum are not always associated with hypercalcemia in patients with ATL, suggesting that another factor is involved in the pathogenesis of hypercalcemia.(12)
ATL is characterized by a rapid infiltration of circulating ATL cells into a variety of tissues, a process often associated with poor prognosis. In this regard, as well as T-cells, we reported that macrophage inflammatory protein (MIP)-1α induces integrin-mediated adhesion of ATL cells to the endothelium and their subsequent migration, which results in severe infiltration of ATL cells into various tissues.(13,14) On the other hand, it is reported that MIP-1α is a potent osteoclastogenic factor that acts directly on osteoclast precursors and osteoclast formation.(15–18) Therefore, we hypothesized that MIP-1α is a marker for ATL-associated hypercalcemia. In this study, we show that serum MIP-1α levels were elevated in all 24 ATL patients with hypercalcemia but not detected in 34 of 37 ATL patients without hypercalcemia or in all 10 HHM patients with solid tumors. Furthermore, we found that MIP-1α was spontaneously produced by ATL cells and that MIP-1α stimulated monocyte migration and bone resorption. We postulate that autocrine stimulation of MIP-1α is essential for osteoclastogenic activation of bone metabolism, resulting in marked bone resorption and subsequent hypercalcemia.
MATERIALS AND METHODS
ELISA of MIP-1 in supernatant or cytosol of ATL cells
Freshly isolated normal T-cells and ATL cells (1 × 106) were washed in PBS and lysed with 250 μl of PBS containing 2% N-octyl-β-D-glucopyraide (OGP; Sigma Chemical, St Louis, MO, USA). The culture supernatants were collected from normal T-cells (1 × 106) and ATL cells (1 × 106) after 24 h of incubation in RPMI1640 with 5% fetal calf serum (FCS) at 37°C without any stimulation. The level of MIP-1α protein in each sample was measured by MIP-1α ELISA system (R&D Systems, Minneapolis, MN, USA). The sensitivity of the assay was 4.0 pg/ml of MIP-1α. Results were expressed in nanograms per milliliter per 1 × 105 cells.
Preparation of T-cells and ATL cells
Highly purified CD4+ T- and ATL cells were prepared by exhaustive negative selection(19) from peripheral blood monoclonal cells (PBMCs) of consented normal donors and ATL patients using magnetic beads (Dynal, Oslo, Norway) and a cocktail of various antibodies, including CD19 monoclonal antibody (mAb) FMC63, CD16 mAb 3G8, CD11 mAb NIH11b-1, CD14 mAb 63D3, and CD8 mAb B9.8.4.
Monocyte migration assay
Transendothelial migration of monocytes was assessed using 3-μm-pore 24-well microchemotaxis chambers (Transwell; Costar, Cambridge, MA, USA). The inner wells were seeded with HMEC-1 (1 × 105 cells/inner well) in 200 μl endothelial cell medium. The endothelial cell monolayers were cultured for 24 h at 37°C and rinsed with assay medium before performing the assays. The inner wells were placed in microchemotaxis chambers, each containing 500 μl of endothelial cell medium (without endothelial cell growth supplement) with MIP-1α (0.1-100 ng/ml).51Chromate-labeled monocytes (1 × 106/well) were added to the endothelial cell monolayers, which were incubated at 37°C in 5% CO2 for 1.5 h. Cells that migrated into the lower wells were lysed with 0.1% Triton X-100 (Sigma), and γ-emissions of well contents were determined.
Fluorescence-activated cell-sorting analysis
Staining and flow cytometric analysis of osteoblasts and synovial cells were carried out by standard procedures as described previously(20) using a fluorescence-activated cell sorter (FACScan; Becton Dickinson, Mountain View, CA, USA). Briefly, cells (1 × 105) were incubated with specific mAbs and subsequently with fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse IgG Ab or rabbit anti-goat IgG Ab at saturating concentrations in FACS medium consisting of HBSS (Nissui, Tokyo, Japan), 0.5% human serum albumin (Green-cross, Osaka, Japan), and 0.2% NaN3 (Sigma Aldrich) for 30 minutes at 4°C. After three washes in FACS medium, the cells were analyzed with FACScan. Amplification of the mAb binding was provided by a three-decade logarithmic amplifier. Quantification of the cell surface antigens on one cell was performed using beads (QIFKIT; Dako Japan, Kyoto, Japan).
Osteoclast formation in coculture system of ATL cells and PBMCs
Isolated PBMCs (4 × 105 cells/well) were resuspended in α-MEM containing 10% FCS and 50 ng/ml of macrophage colony-stimulating factor (M-CSF) and seeded in 48-well tissue culture plates (Costar 3548; Corning, New York, NY, USA). Three days later, adherent cells were used for subsequent cocultures with ATL cells. ATL cells were added to 48-well plates (1 × 104 cells/well) and cocultured for 9 days in α-MEM containing 10% FCS, 50 ng/ml of M-CSF, and 10−7 M 1,25(OH)2D3. Anti-MIP-1α (10-100 ng/ml) or OPG (100 ng/ml) were added to the coculture system on culture day 4. After 9 days of culture, some dishes were stained for TRACP as described previously.(21) The number of TRACP+ multinucleated cells that contained more than three nuclei were identified as osteoclasts and counted by light microscopy.
45Ca release assay
Bone resorbing activity (BRA) was assayed according to the Ca release assay with some modification. Bones were prelabeled with45Ca in vivo by injecting pregnant ICR mice 1 day before delivery with 1.85 Mbq45CaCl2. Shafts of the radii dissected from 16-day-old fetal mice were cultured in 0.5 ml of serum-free α-MEM medium supplemented with 2 mg/ml of bovine serum albumin (BSA; Sigma). The long bones were cultured on stainless steel minimesh grids at the interphase between the medium and 5% CO2 atm at 37°C. All long bones were preincubated in the medium for 30 h to remove exchangeable45Ca. The long bones were transferred to fresh medium with or without various concentrations of MIP-1α. After 3.5 days of culture with these agents and/or antibodies,45Ca released into the medium or remaining in the 5% TCA extracts of the bones was measured by a liquid scintillation counter. Percent45Ca release was calculated as [medium cpm/(medium cpm + bone cpm)] × 100 (%).
Data are expressed as mean ± SD. Differences between groups were tested for statistical significance using the Student's t-test. A p value <0.05 denoted the presence of a statistically significant difference.
Serum cytokine levels in ATL with hypercalcemia and humoral hypercalcemia of malignancy in patients with solid cancers
To identify the molecules involved in ATL-associated hypercalcemia and humoral hypercalcemia of malignancy (HHM) with solid cancer, we measured the serum concentrations of several cytokines in ATL patients with hypercalcemia and HHM with solid cancer by ELISA. Among several cytokines associated with hypercalcemia, PTHrP has emerged as an important contributing factor for high serum calcium.(4) However, in this study, serum PTHrP levels were elevated in 8 of 10 HHM patients (80%) and in 6 of 24 ATL patients (25%) with hypercalcemia (Fig. 1). Furthermore, the serum concentrations of IL-1β (60% in HHM and 17% in ATL), IL-6 (90% of HHM and 55% of ATL), and TNF-α (50% of HHM and 12.5% of ATL) had similar results.
Elevated levels of serum MIP-1α in ATL patients with hypercalcemia
We measured serum MIP-1α levels in ATL patients with hypercalcemia and normocalcemia, ATL carriers, HHM with solid cancer, primary hyperparathyroidism with hypercalcemia, and healthy volunteers using ELISA (Fig. 2). Serum concentrations of MIP-1α were elevated in all 24 ATL patients with hypercalcemia (100%) and in 3 of 37 ATL patients with normal calcium levels (8%). All 24 ATL patients with hypercalcemia and 37 ATL patients without hypercalcemia had elevated ATL cell counts and soluble IL-2 receptor levels in peripheral blood. Interestingly, the three ATL patients with high serum MIP-1α developed hypercalcemia later, in association with transformation to active ATL. Furthermore, the elevated MIP-1α in ATL patients decreased after effective chemotherapy. On the other hand, 0 of the 15 healthy volunteers, 8 HTLV-I carriers, 10 primary hyperparathyroidism patients with hypercalcemia, and 10 HHM patients with solid tumors (e.g., lung carcinoma, hepatocellular carcinoma and esophageal carcinoma) had high levels of serum MIP-1α. These results suggest that MIP-1α is a specific factor responsible for hypercalcemia in ATL.
Spontaneous secretion of MIP-1α by ATL cells
We and others have proposed that chemokines such as MIP-1α functionally trigger T-lymphocyte integrins.(13,22,23) ATL cells produced significant amounts of MIP-1α protein in the culture supernatant as well as in the cytosol without any stimulation, whereas normal resting T-cells did not produce any of these chemokines (Fig. 3). These results clearly show that MIP-1α is spontaneously produced by ATL cells.
MIP-1α induces transendothelial migration of monocytes
Because monocytes are thought to be the origin for bone resorbing cells (mature osteoclasts), we next assessed the chemotactic activity of MIP-1α toward monocytes. In the invasion chamber, MIP-1α induced transendothelial migration of monocytes, and the effect was dose-dependent (Fig. 4). The results indicated that MIP-1α enhances the migration of monocytes.
MIP-1α stimulates RANKL expression on ATL cells
We initially assessed the effects of MIP-1α on the expression of RANKL on ATL cells using a flow cytometer. MIP-1α (1 ng/ml) was added to the cells, and the cells were harvested after 6-h incubation. The obtained cells were stained by mAbs and subjected to flow cytometric analysis with subsequent quantification of the cell surface antigens on one cell using standard beads. Approximately 500 molecules of RANKL were spontaneously expressed on ATL cells (Fig. 5). However, MIP-1α (1 ng/ml) induced three times excess of RANKL expression on the cells within the 6-h incubation. These results indicate that MIP-1α enhances the expression of RANKL on ATL cells.
ATL cells induce osteoclast formation
Formation of TRACP+ multinucleated cells (MNCs) from the coculture system of ATL cells and PBMC was assessed in α-MEM containing 10% FCS, 50 ng/ml of M-CSF, and 10−7 M 1,25(OH)2D3. Anti-MIP-1α (10-100 ng/ml) or OPG (100 ng/ml) was added to the coculture system at day 4. After a 9-day culture, marked TRACP+ MNC formation was noted from cultured ATL cells, whereas no TRACP+ MNC formation was induced in cocultures of PBMC alone without ATL cells (Fig. 6). The formation of TRACP+ MNC in cocultures of PBMC and ATL cells was completely inhibited by the addition of anti-MIP-1α Ab (100 ng/ml) or OPG, suggesting that MIP-1α plays a pivotal role in osteoclastogenesis through the upregulation of RANKL on ATL cells. These results imply that the high-affinity adhesion of PBMC and ATL cells is a prerequisite for the efficient signaling of RANKL on ATL cells during osteoclast maturation, suggesting that MIP-1α plays an important role in the osteoclastogenesis through RANKL on ATL cells.
MIP-1α stimulates calcium release in mice long bone cultures
It is reported that MIP-1α is a potent osteoclastogenic factor that acts directly on osteoclast precursors and osteoclast formation.(15–18) To elucidate further whether MIP-1α is involved in bone resorption, we examined the effects of MIP-1α on bone resorption by using mice long bone system.45Ca release from fetal mouse long bones in cultures was stimulated by various concentrations of MIP-1α (10-100 ng/ml) and 10 nM of 1α,25(OH)2D3 (Fig. 7). The results indicated that MIP-1α stimulated osteoclast formation and bone resorption.
Effective treatment of ATL results in reduction of serum calcium and MIP-1a
To further characterize the correlation between serum calcium and MIP-1α in ATL patients, we measured the effects of chemotherapy on serum calcium and MIP-1α levels in eight ATL patients (Fig. 8). Interestingly, patients with inactive ATL (cases 1 and 2) and high serum MIP-1α developed hypercalcemia later. On the other hand, calcium and MIP-1α levels increased simultaneously after conversion to acute phase in cases 3 and 4. Furthermore, in cases 3-8, serum calcium concentration decreased proportionately with serum MIP-1α level by effective chemotherapy. These results further emphasize the clinical importance of MIP-1α in ATL patients with hypercalcemia.
It has been proposed that osteoclastic bone resorption mediated by PTHrP and IL-1 is involved in HHM.(24) The frequency of hypercalcemia in patients with ATL is markedly high compared with HHM.(2) Its frequency in patients with ATL is reported at about 70% during the whole clinical course, although it tends to be more frequent in those patients with clinically aggressive ATL,(1) suggesting that molecules expressed or secreted by ATL cells play an important role in the induction of hypercalcemia. In this study, we showed that serum concentrations of PTHrP, IL-1β, IL-6, and TNF-α did not correlate with serum calcium concentrations in ATL patients. Furthermore, although serum MIP-1α concentrations were elevated in all 24 ATL patients with hypercalcemia, they were not elevated in all 10 HHM patients. These results indicate that MIP-1α plays an important role in hypercalcemia in patients with ATL. However, because PTHrP has emerged as an important factor in the pathogenesis of hypercalcemia in ATL patients,(4,10) our findings in this study are contradictory to those previous observations. Previous studies reported that PTHrP cannot directly induce differentiation of hematopoietic precursor cells to osteoclasts and that high levels of PTHrP in the serum are not always associated with hypercalcemia in patients with ATL.(11,12) Furthermore, Nosaka et al.(25) reported a subgroup of ATL patients with high serum PTHrP concentrations but without hypercalcemia, suggesting that elevated serum PTHrP alone does not always explain the mechanism of hypercalcemia in at least some patients with ATL and that RANKL expression on ATL cells could play a critical role in the pathogenesis of ATL-associated hypercalcemia. These reports support the present findings, and we suspect that PTHrP might be an additional factor that induces the expression of RANKL and exacerbates hypercalcemia.
ATL is characterized by rapid infiltration of circulating ATL cells into various tissues, a process often associated with poor prognosis. We previously proposed that chemokines, MIP-1α and MIP-1β, induce integrin-mediated adhesion of circulating T-cells to endothelial cells.(13) We report here that ATL cells produce large amounts of MIP-1α. In another study, we also showed that MIP-1α induced integrin-mediated adhesion of ATL cells to the endothelium and their subsequent migration, which resulted in extensive infiltration of ATL cells into multiple tissues.(14) Other studies reported that MIP-1α increases osteoclast formation and osteoclast recruitment in rodent systems and is produced by lymphoblastic cell lines(26) and that MIP-1α also stimulates osteoclast formation in human bone marrow cultures.(15) Therefore, we hypothesized that the MIP-1α is a marker for ATL-associated hypercalcemia.
Among the various cytokines produced by ATL cells, including IL-1, IL-6, and TNF-α, the production of MIP-1α and MIP-1β is a characteristic feature of the cells, because the HTLV-I tax induces production of these chemokines.(27,28) Pretreatment of ATL cells with anti-MIP-1α antibody or transfection of antisense oligonucleotide of MIP-1α into ATL cells reduced integrin-mediated adhesion of ATL cells to endothelial cells and subsequent transendothelial migration,(14) indicating that spontaneous production of MIP-1α from ATL cells results in marked transmigration of ATL cells into various tissues including bone and bone marrow. We also observed that MIP-1α induced integrin-mediated adhesion of ATL cells to osteoblasts and bone marrow stromal cells, which leads to stimulation of the cells, and that MIP-1α induced release of45Ca in bone organ cultures. Furthermore, we and others reported that MIP-1α induces integrin-mediated adhesion and chemotaxis of monocytes, including osteoclast progenitor cells, and enhances the production of osteoclastogenic factors such as IL-6, PTHrP, and RANKL from osteoblasts or stromal cells, and hence is involved in osteoclast maturation.(9,29) Because ATL cells are derived from activated helper T-lymphocytes, secreted cytokines can modify the clinical features. Monocytosis is one of the clinical features of ATL in which increased M-CSF should be implicated in the pathogenesis.(30) M-CSF increased precursor cells and supports their differentiation into monocytes and osteoclasts. Without RANKL, these precursors cells differentiate into monocytes, but once the expression of RANKL is induced in ATL cells and such leukemic cells infiltrate the bone marrow, these precursor cells differentiate into osteoclasts. Activated T-lymphocytes are known to induce differentiation of cells into osteoclasts in vitro possibly through RANKL expression.(31) This study indicates that MIP-1α induces RANKL expression on ATL cells and that osteoclast formation is induced by ATL cells and is inhibited by the addition of anti-MIP-1α Ab or osteoprotegerin (OPG). These results indicate that MIP-1α-induced upregulation of RANKL on ATL cells is involved in osteoclast maturation in the coculture system. These findings also indicate that MIP-1α can induce RANKL on ATL cells in an autocrine fashion and suggest that RANKL seems to mediate the hypercalcemic effect of MIP-1α in ATL.
Taken together, our findings suggest that serum MIP-1α level could be considered as a specific marker for hypercalcemia in ATL patients and that the hypercalcemia in ATL is caused by a mechanism different from that of HHM associated with solid cancers. This study indicates that autocrine stimulation of MIP-1α is essential for osteoclastogenic activation of bone metabolism, resulting in marked bone resorption and subsequent hypercalcemia. Thus, MIP-1α seems to be a clinical hallmark for hypercalcemia in ATL and could be a potential therapeutic target. Recently, Oyajobi et al.(32) reported that MIP-1α could induce osteoclastic resorption in vivo and postulated that strategies based on functional blockade of MIP-1α bioactivity might be clinically effective in myeloma. Our findings may also allow the development of new strategies for the treatment of hypercalcemia in ATL and this aggressive disease.
We thank Dr FG Issa for the careful reading and editing of the manuscript. This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan.