CXCR4, the receptor for bone marrow stroma derived SDF-1, has recently been studied in normal hematopoiesis and hematologic malignancies. Increased expression of CXCR4 by leukemic blasts has been reported by us and others (Konoplev S. et al, Cancer 2007) to be associated with poor prognosis in acute myeloid leukemia (AML). However, all in-vitro studies are usually carried out under unphysiological, i.e. normoxic (21% O2) conditions. We hypothesized that the pO2 in vitro has major impact on the expression of CXCR4, a key receptor for cell migration and intracellular signalling. Thus, pO2 of bone marrow aspirates was measured using the i-STAT Portable Clinical Analyzer and a hypoxic workstation was used providing constant low oxygen content. Surface and total CXCR4 expression was examined in leukemic cell lines and patient samples by flow cytometry, confocal microscopy and Western blotting (WB). In 19 patients, the median pO2 of the bone marrow was determined as 46.1±12.8 mmHg (6.1±1.7%) with no significant difference between patients with AML (n=7, pO2 41.3±11.2 mmHg) and patients in CR (n=12, pO2 48.3±15.9 mmHg). This level of hypoxia significantly increases surface and total expression of CXCR4 in the leukemic cell lines U937 and OCI-AML3 as well as in samples from patients with AML, as compared to normoxic conditions (~2.8fold increase). This increase happened mainly within the first 2–8 hours of hypoxia and was unrelated to increased CXCR4 transcription, as shown by PCR. Re-oxygenation of leukemic cells resulted in a statistical significant degradation of CXCR4 (~3fold decrease) in all examined cell lines and patient samples (n=10). This loss of CXCR4 is very rapid (within 5 minutes of re-oxygenation) and was detected by flow cytometry, confocal microscopy and WB. This phenomenon was independent of proteasome activity and ATP. Detailed analysis of membraneous lipid rafts by sucrose density separation, cholesterol depletion and flow cytometry analysis for GM1 gangliosides showed structural (distinct re-distribution of Lck in lipid rafts) and quantitative changes (loss of cholesterol and CXCR4) during re-oxygenation. Moreover, part of the loss of CXCR4 can be attributed to sequestration of microparticles into the extracellular environment as shown by WB of supernatant of re-oxygenated cells and by a significant increase (~1.5fold) in the amount of microparticles released by cells (cell lines U937 and OCI-AML3 and additional patient samples) during the process of re-oxygenation, as measured by flow cytometry. In summary, this study determined the oxygen content of CR and leukemic bone marrow samples as 6.1±1.7%. This pO2 is associated with an increase in CXCR4 expression on AML cells, while re-oxygenation leads to a rapid decrease of CXCR4, perhaps in part by shedding of CXCR4- containing microparticles. These studies point to the importance of studying leukemic blasts under physiologic, i.e. hypoxic conditions.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

Sign in via your Institution