Background: While multi-agent chemotherapy has led to remarkable improvements in survival for children with T-cell acute lymphoblastic leukemia (T-ALL), outcomes remain dismal for the ~ 20% of patients that fail therapy. The goal of this study was to elucidate biological mechanisms contributing to chemoresistance in T-ALL. We hypothesized that chemoresistance owes not only to the aberrant signaling pathways intrinsic to the leukemia cells but also to contributions from an abnormal microenvironment in which they reside. The chemokine receptor ligand pair CXCR4/CXCL12 is important for normal leukocyte trafficking and deregulation is frequently observed in several hematologic malignancies. However, the role this axis plays in T-ALL is largely unknown. Here, we test the hypothesis that CXCR4/CXCL12 creates a sanctuary microenvironment that promotes T-ALL survival and provides protection from chemotherapy.

Methods: A panel of T-ALL cell lines and primary patient samples expanded in NOD/SCID/c null (NSG) mice were cultured in the presence or absence of the CXCR4 antagonist, AMD3100, with or without chemotherapy. The human bone marrow stromal cell line, HS27a, which constitutively expresses the CXCR4 ligand CXCL12, was used to recapitulate the tumor microenvironment ex vivo. Transwell migration and modified pseudo-emperipolesis assays were used to examine CXCL12-mediated chemotaxis. Multi-parameter flow cytometry was used to evaluate the impact of activation of the CXCR4/CXCL12 axis on signaling networks, cell survival, and chemotherapy resistance.

Results: We found CXCR4 membrane expression on all T-ALL cell lines and xenografts samples tested. Incubation with the CXCR4 ligand CXCL12 resulted in activation of survival signaling cascades (PI3K/AKT and MAPK), an effect blocked with AMD3100. Using a transwell system, we found dose dependent chemotaxis of T-ALL lymphoblasts to CXCL12 that was prevented by AMD3100. T-ALL cells also migrated into HS27a stromal cells in a CXCR4/CXCL12 dependent fashion. In addition, co-culturing T-ALL xenografts cells with HS27a imparted a survival advantage that was promptly eliminated by AMD3100 exposure. To test whether the CXCR4/CXCL12 microenvironment contributed to chemotherapy resistance, T-ALL xenografts cells were co-cultured ex vivo in the presence or absence of HS27a stromal cells and bortezomib or dexamethasone. Stromal cells conferred a marked chemoprotective effect that was specifically blocked by AMD3100. This highlights that stroma-mediated chemoresistance in xenograft samples is in part due to stromal-cell mediated activation of the CXCR4/CXCL12 axis.

Conclusions: This study provides evidence for a T-ALL microenvironment that exploits the CXCR4/CXCL12 axis for leukemic cell recruitment, enhanced cell survival, and chemotherapy protection. Our findings also implicate the stroma as a major contributor to chemotherapy resistance in primary expanded patient samples partially due to the activation of the CXCR4/CXCL12 axis. These results further suggest that targeting the stroma through inhibition of this axis may be of therapeutic benefit in patients with chemotherapy resistant T-ALL.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.