CD44 standard and CD44v10 isoform expression on leukemia cells distinctly influences niche embedding of hematopoietic stem cells

BACKGROUND A blockade of CD44 is considered a therapeutic option for the elimination of leukemia initiating cells. However, anti-panCD44 can interfere with hematopoiesis. Therefore we explored, whether a CD44 variant isoform (CD44v)-specific antibody can inhibit leukemia growth without attacking hematopoiesis. As a model we used CD44v10 transfected EL4 thymoma cells (EL4-v10). METHODS The therapeutic efficacy of anti-panCD44 and anti-CD44v10 was evaluated after intravenous application of EL4/EL4-v10. Ex vivo and in vitro studies evaluated the impact of anti-panCD44 and anti-CD44v10 as well as of EL4 and EL4-v10 on hematopoietic stem cells (HSC) in cocultures with bone marrow stroma cells with a focus on adhesion, migration, cell cycle progression and apoptosis resistance. RESULTS Intravenously injected EL4-v10 grow in bone marrow and spleen. Anti-panCD44 and, more pronounced anti-CD44v10 prolong the survival time. The higher efficacy of anti-CD44v10 compared to anti-panCD44 does not rely on stronger antibody-dependent cellular cytotoxicity or on promoting EL4-v10 apoptosis. Instead, EL4 compete with HSC niche embedding. This has consequences on quiescence and apoptosis-protecting signals provided by the stroma. Anti-panCD44, too, more efficiently affected embedding of HSC than of EL4 in the bone marrow stroma. EL4-v10, by catching osteopontin, migrated on bone marrow stroma and did not or weakly interfere with HSC adhesion. Anti-CD44v10, too, did not affect the HSC--bone marrow stroma crosstalk. CONCLUSION The therapeutic effect of anti-panCD44 and anti-CD44v10 is based on stimulation of antibody-dependent cellular cytotoxicity. The superiority of anti-CD44v10 is partly due to blocking CD44v10-stimulated osteopontin expression that could drive HSC out of the niche. However, the main reason for the superiority of anti-CD44v10 relies on neither EL4-v10 nor anti-CD44v10 severely interfering with HSC--stroma cell interactions that, on the other hand, are affected by EL4 and anti-panCD44. Anti-panCD44 disturbing HSC embedding in the osteogenic niche weakens its therapeutic effect towards EL4. Thus, as far as leukemic cells express CD44v isoforms, the therapeutic use of anti-panCD44 should be avoided in favor of CD44v-specific antibodies.