Abstract Leukaemia stem cells (LSCs) not only drive and sustain acute myeloid leukemia (AML), but also frequently survive chemotherapy and cause relapse. Specific targeting of LSCs has great potential to cure AML. However, there are no known surface markers that are specific for LSCs in all AML patients. Most of the markers present on LSCs are also found on haematopoietic stem cells (HSCs). Rather than focusing on the identification of LSC-specific cell surface markers, differences in signaling pathways between LSCs and HSCs could be exploited to specifically target LSCs. The aryl hydrocarbon receptor (AHR) pathway was shown to be critical for the self-renewal properties of HSCs. Here we investigate the effect of AHR chemical modulation on LSCs in two murine bone marrow transplantation models (MBMTM) of AML. We used a PICALM-MLLT10-driven MBMTM AML model with a LSC frequency (LSC-F) of about 1:4 (95% confidence interval (CI): 1:13-1:2). Leukemic bone marrow (LBM) cells from this model were treated ex vivo for 96 hours with the AHR agonist ITE (10 µM), the AHR antagonist CH-223191 (10 µM) or cultured with vehicle only (dimethyl sulfoxide (DMSO). We then performed limiting dilution assays (LDA) to measure LSC-F of the treated cells by transplanting 50, 250 and 500 cells into groups of three mice each. The LSC-F was reduced 11-fold to 1:44 (95% CI: 1:167 – 1:12) in the vehicle cultured cells, to 1:163 (95% CI: 1:429 – 1:62) in the AHR agonist (ITE) -treated cells and, very unexpectedly, became unmeasurable in the AHR antagonist (CH223191) -treated cells. None of the nine mice transplanted with AHR antagonist-treated cells developed leukemia. To investigate whether this LSC eliminating effect of the AHR antagonist CH223191 can also be observed in other murine AML models, we tested the effect of AHR antagonist treatment in a KMT2A/MLLT3-driven murine leukemia. Bone marrow cells from a KMT2A/MLLT3 leukemic mouse were treated with the AHR antagonist CH-223191 (10 µM) or cultured with the vehicle (DMSO) as described above. Three groups of four mice were transplanted with 4,000, 40,000 and 400,000 cells each of the AHR antagonist-treated cells and three groups of four mice were transplanted similarly with DMSO-treated cells (i.e., 12 mice in each arm). All mice transplanted with 400,000 cells developed leukemia with similar latencies in both the CH-223191 and the DMSO arm. However, all mice (n=4) transplanted with 40,000 CH-223191-treated cells survived, while only one of the four mice survived in the group transplanted with 40,000 cells that had been exposed to DMSO. None of the eight mice transplanted with 4000 cells (treated or untreated) developed leukemia. These results suggest that AHR antagonist treatment was also negatively affecting the LSCs in the KMT2A/MLLT3 leukemia model. We then evaluated the effect of CH-223191 treatment on healthy HSCs using a long-term competitive repopulation assay (CRA). The proportion of blood cells derived from HSCs transplanted after treatment with the AHR inhibitor CH-223191 (10 µM; for 96-hours) steadily increased over a period of 52 weeks in the recipients. The contribution to hematopoiesis of inhibitor-treated HSCs to long term hematopoiesis was slightly stronger than what we observed from untreated HSCs in the control arm of our CRA. These results indicate that healthy long term repopulating HSCs were not harmed by AHR inhibition. Taken together, our data show that inhibition of AHR signaling using CH-223191 is able to eliminate LSCs in two distinct murine AML models, and that treatment with the same inhibitor has no or even protective effects on normal HSCs. These data suggest that inhibition of AHR signaling could be exploited to develop novel treatment strategies for AML that are both more effective and have fewer side effect than traditional chemotherapy approaches.
Desai et al. (Mon,) studied this question.