Polymeric CXCR4 Inhibitors Exhibit Novel Anti-Leukemic Effects

Acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL) are common leukemias with a mortality rate estimated to be greater than 50%. This dismal outcome is in part due to current therapeutics. Intensive remission therapy is difficult for patients to tolerate and despite good complete remission rates, is not curative for all patients. In relapse/refractory status, there are even fewer effective therapeutics. There is a strong need for more effective, less toxic therapeutics.

The surface receptor CXCR4 has become a recent target of interest, as it is highly expressed on leukemia cells and, via its canonical ligand CXCL12, enables cells to evade chemotherapeutics by localizing them to the protective bone marrow niche. The small molecule inhibitor AMD3100 (plerixafor) disrupts this localization by blocking the CXCL12:CXCR4 interaction and is being investigated for its ability to mobilize leukemia cells to the peripheral blood. Thus, AMD3100 acts as a chemosensitizer, but by itself, AMD3100 does not have anti-leukemic effects. It has been recently shown that CXCR4 has additional non-canonical activities outside of CXCL12 that are vital for leukemia survival and maintenance. Genetic knock out of Cxcr4 in an AML model decreases survival and increases differentiation in a CXCL12 independent manner. This implies that CXCR4 has previously unrecognized activities in leukemia and that inhibitors capable of blocking both CXCR4's canonical and non-canonical activities would more effective therapeutics.

We developed a class of polymeric CXCR4 inhibitors (PAMDs) via Michael-type addition of AMD3100 and focused on one analog PAMD-Ch17, which contains additional cholesterol modifications. PAMD-Ch17 was originally designed to delivery siRNAs, but we surprisingly found it had anti-leukemic effects on its own. PAMD-Ch17, but not AMD3100, induces cytotoxicity and promotes differentiation of mouse AML cells in vitro. In colony assays, treatment with PAMD-Ch17 causes a statistically significant decrease in leukemia colony growth while treatment with AMD3100 was indistinguishable from the control. Across a panel of seven human AML and ALL cell lines, PAMD-Ch17 inhibited metabolic activity, a proxy for viability, at the nanomolar doses whereas AMD3100 had no effect even at the highest micromolar dose.

As we predict that PAMD-Ch17 is acting through CXCR4, we compared cell surface expression of the receptor to PAMD-Ch17 sensitivity for each cell line. We found that PAMD-Ch17 IC50 was moderately negatively correlated with CXCR4 expression (r= -0.69), suggesting PAMD-CH17 mediates at least some of its effects through CXCR4. To further address the mechanisms of PAMD-Ch17's activity, we cultured Jurkat T-ALL cells with AMD3100 or PAMD-Ch17 at the same dose for 24 hours and stained using the 12G5 CXCR4 monoclonal antibody, which is known to compete with AMD3100 and CXC12. We found that AMD3100 caused a statistically significant decrease in 12G5 staining as compared to control treated cells, but PAMD-Ch17 induced a complete loss of 12G5 staining that was statistically significant as compared to both control and AMD3100 treated cells. PAMD-Ch17, but not AMD3100, also caused a moderate, but significantly decrease in staining with another antibody against the N-terminus of CXCR4, a region not predicted to be affected by AMD3100 or CXCL12. These results indicate that PAMD-Ch17 more effectively blocks the epitope recognized by the 12G5 antibody but may also interact other domains of CXCR4 or decrease its cell surface expression. We also developed a line of PAMD-CH17 resistant Jurkat cells. The resistant cells have an IC50 of at least 2-fold greater than naïve Jurkat cells. Interestingly, the resistant cells are not recognized by the 12G5 monoclonal antibody but are stained by the N-terminus antibody. These results imply that the CXCR4 expressed on resistant cells has an altered conformation that may less effectively bind PAMD-Ch17.

Collectively, these data suggest that PAMD-Ch17 more effectively inhibits CXCR4 leading to its novel anti-leukemic effects and support the development of PAMDs for the treatment of leukemia.

No relevant conflicts of interest to declare.