Patients with myelogenous leukemias can present with symptoms of bone pain and pathologic fractures, however little is known about the interactions between malignant cells and the bone marrow microenvironment. Additionally leukemia is known to severely interfere with normal hematopoiesis. To further characterize interactions between leukemic cells and the microenvironment, we used a model of blast crisis CML (bcCML) in which immature murine hematopoietic cells are engineered to express the BCR/ABL and Nup98/HoxA9 translocation products. Injection of these cells into naïve mice results in rapid accumulation of leukemic cells in the bone marrow (Dash et al. PNAS, 2002). We investigated the effect of leukemia on the bone marrow microenvironment by first performing immunohistochemical analyses. Leukemic cells were observed to preferentially localize in close contact with bony trabeculae. In addition, leukemic mice also exhibited a dramatic loss of trabecular bone as measured by micro-CT scanning (22.8 ± 1.5% vs 13.6 ± 1.5%, BV/TV n=5 in each group p=0.0048), prompting us to examine bone resorption and the abundance of osteoclasts. Histologic sections from leukemic mice showed an increase in mature osteoclasts (TRAP+, multinucleated cells) at the endosteal surface of the long bones (51 ± 4 OC/section vs 64 ± 3 OC/section p=0.0229). Additionally leukemic mice had a 50% increase in serum C-telopeptide (CTX), a well-established marker of global bone resorption (15.5 ± 0.3 ng/ml vs 21.8 ± 1.0 ng/ml, p=0.0003). Therefore, the presence of leukemic cells appears to strongly stimulate osteoclastogenesis and bone resorption. In addition to increased osteoclasts, a rapid and severe reduction in bone formation was identified in leukemic mice by decreased serum osteocalcin, a well-established marker of bone formation (70.8 ± 6.9 ng/ml vs 39.9 ± 3.2 ng/ml, p=0.0036). To determine effects of leukemia on bone-forming cells, we analyzed osteoblastic cells from the long bones of leukemic animals. Marrow was flushed from the bone and minced bone fragments were digested in collagenase. Cells isolated in this fashion have strong osteoblastic activity and can support hematopoietic stem cells (HSCs) (Chitteti et al, Blood 2010). Leukemic cells were present in the isolated fraction confirming that leukemic cells were closely associated with the bone and not entirely removed when the marrow was flushed. Total cells isolated by collagenase digestion from the long bones of leukemic mice were cultured with leukemic cells and evaluated for osteoblastic colony-forming ability. These cultures demonstrated a reduced ability to form osteoblastic colonies compared to controls (26 ± 2 colonies vs 13 ± 2 colonies p=0.0014). Freshly isolated cells were CD45 depleted to remove leukemic cells and again evaluated for colony-forming ability. CD45 negative cells from leukemic mice also demonstrated reduced ability to form mineralizing osteoblastic colonies in vitro compared to controls when an identical number of cells were cultured (10 ± 2 colonies vs 0 colonies) suggesting that their previous exposure to leukemic cells in vivo was sufficient to decrease their osteoblastic activity. Further in vivo osteoblastic evaluation in leukemic mice showed reduced immunohistochemical staining for the osteoblastic marker osteopontin at the endosteal surface, supporting a leukemic-induced reduction of the mature osteoblastic population. These data demonstrate a severe reduction in both number and function of osteoblastic cells in a leukemic environment. Together these data show severe effects of leukemia on both osteoblastic and osteoclastic cells, which could contribute to the bone-specific problems associated with leukemic disease. Moreover, we propose that osteoblastic defects observed in this model may contribute to the leukemia-induced inhibitory effects on normal hematopoiesis. Studies are ongoing to assess the microenvironmental support for normal hematopoiesis in the leukemic setting and to identify the leukemic signals that modify the bone marrow microenvironment.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.