We and others have previously reported that leukemia progression is associated with vast expansion of the hypoxic niches and stabilization of hypoxia-inducible factor 1 alpha (HIF-1α) in leukemic cells (Frolova et al. Cancer Biol Ther. 2012, 10:858; Benito et al. PLoS One 2011, 6(8); e23108:1). Interactions of leukemia and the bone marrow (BM) microenvironment are known to play a key role in the survival and growth of leukemic cells, and we have shown that HIF-1α stabilization in stromal cells of the microenvironment facilitates leukemia homing and progression (Chen et al. Blood 2012, 119:4971). In this study, we aimed to characterize the time-dependent progression of BM hypoxia involving both leukemia cells and components of the BM niche, using the multiphoton intravital microscopy (MP-IVM) technique.

We first generated a transplantable, imageable leukemia model by retrovirally transducing C57Bl6-Ai14 murine BM cells that express red fluorescing tdTomato with the p190-Bcr/Abl oncogene. The resulting p190-Bcr/Abl tdTomato cells caused rapid development of acute lymphocytic leukemia (ALL) in un-irradiated C57Bl6 immunocompetent mice, manifested by infiltration of the spleen, liver, BM within long bones, skull, and central nervous system followed by death within 28 days. Leukemia cells collected from the BM (LBC) of these animals were transplantable into secondary recipients and triggered accelerated ALL development (14-16 days). Time-course analysis of skull and femur bones in the secondary recipients by MP-IVM demonstrated LBC lodging on day 1 after ALL cell injection, followed by rapid accumulation of leukemia cells localized predominantly within the sinusoidal spaces, which were visualized by injecting the vascular fluorescent dye BSA-647 (Fig. 1a).

To detect in vivo hypoxia development, we utilized HS680 (HypoxiSense 680), a carbonic anhydrase IX (CAIX)–targeted fluorescent agent that can be used to image overexpression of CAIX, a direct HIF-1α target, in tumors in response to regional hypoxia. C57Bl6 mice were engrafted with 2 x 105 LBC , and HS680 was injected intravenously at serial intervals followed by MP-IVM. In two separate experiments, increased HS680 fluorescence was detected in bone-lining cells in the BM niches of mice harboring ALL on days 8 and 13, but not in their healthy littermates (Fig 1b). To obtain an independent confirmation of hypoxia, additional mice (n=3) at the same stage (day 14) of leukemia development were sacrificed 3 hr after injection of chemical hypoxia probe pimonidazole (Pimo), and hypoxic BM cells that bound the hypoxia probe were detected by immunohistochemistry. Pimo staining demonstrated vastly spread areas of hypoxia that enclosed both leukemia cells and BM niche cells (Fig 1c), consistent with our previously published observations in different leukemia models.

In summary, these findings demonstrate rapid development of intra-BM hypoxia that parallels leukemia progression and involves not only leukemia cells, but also BM niche cells. The HS680 probe can detect hypoxia in vivo within niche cells but not in leukemia cells, likely because of differential expression of CAIX. Our ongoing studies will characterize the cellular origin of hypoxic niche cells by utilizing immunohistochemical techniques and Col2.3-GFPemd transgenic mice to visualize osteoblasts. We postulate that the tumor microenvironment altered with hypoxic niche cells will influence leukemia development or responses to therapy. To this end, we have generated mice with conditionally deleted HIF-1α within BM stromal cells and are investigating the differences in leukemia homing, progression, and chemoresistance between these mice and mice whose BM stromal cells express HIF-1a.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution