Cancer cells can undergo a senescence-like phenotype in response to genotoxic stress from chemotherapy. Relapse of acute myeloid leukemia (AML) frequently occurs months or years after initial chemotherapy and the functional consequence of chemotherapy-induced senescence (CIS) has not been elucidated. We propose that CIS acts as a stress-survival mechanism in AML, allowing cells to persist in a dormant state with potential to repopulate leukemia post-treatment.

To quantify senescence-associated-β-galactosidase (SA-β-gal) activity, we used a fluorogenic substrate (C12-FDG) that enabled sensitive quantification of the senescence marker in viable cells by flow cytometry. We found increased SA-β-Gal activity after exposure to the chemotherapeutic agent cytarabine (AraC) in p53-deficient myeloid leukemia cell lines (K-562 and KG-1), indicating that p53 is not essential for CIS in AML. Using an ex vivo culturing model, we found that patient-derived AML cells demonstrate a greater SA-β-gal response to increasing dosage or duration of AraC treatment until cells undergo apoptosis and exhibit diminished C12-FDG levels. This suggested that CIS is not a distinct condition, but rather a continuous response to the degree of genotoxic stress. ATR kinase activation mediates senescence by enforcing cell cycle arrest in response to genotoxic stress during replication. Treatment of AML specimens with an inhibitor of ATR (VE-821) reduced the initiation of senescence and markedly impaired cell survival after exposure to a moderate dose of AraC. This implied that ATR mediates CIS, independent of p53, to increase stress survival following AraC treatment.

In order to characterize the response to genotoxic stress, we treated patient-derived AML cells with AraC, analyzed RNA expression, and performed gene set enrichment analysis (GSEA) with senescence-associated genes compiled from literature. Inflammatory mediators and extracellular matrix proteins are strongly induced after AraC treatment; these are partly related to the senescence-associated secretory phenotype (SASP) and senescence-messaging secretome (SMS) (p<0.001). Thus, primary AML cells also display a senescent gene expression signature after AraC exposure.

To model the latent dormant state after chemotherapy, we treated patient-derived AML cells with high doses of AraC (1000 and 10 000 nM for 3 days) that killed the majority of the leukemia population in vitro. Cells surviving initial AraC treatment with 1000 nM persisted in a senescent-like dormant state for 3-4 weeks before initiating leukemia repopulation, while cells treated with 10 000 nM AraC continued to be in a dormant state. To recapitulate the relapse in vivo, we established an AML relapse model using primary human specimens engrafted in NSG mice. Administration of AraC, using a physiologically relevant dose and schedule (60 mg/kg/day x 5 days), to AML-engrafted NSG mice reduced peripheral blood leukemic cells and total body leukemic burden 8 days after initiation of therapy to a small residual leukemia fraction (nadir). This reduction in leukemic burden was reversed after 4 weeks. Gene expression analysis of purified human AML cells at nadir (day 8) were significantly enriched for senescence signatures (p<0.001). This enrichment was reduced by day 29 post-treatment as expression was partially reversed to untreated levels. Notably, GSEA revealed reduction of hematopoietic and leukemia stem cell signatures at nadir, which partially recovered in some AML cases at day 29.

Finally, to demonstrate that senescent-like cells maintain leukemia-repopulating potential, we sorted for low (untreated control cells), moderate, and high levels of SA-β-gal activity in C12-FDG+ AML cells after AraC treatment. Transplantation of the sorted cells into NSG mice demonstrated repopulation of leukemia from senescent-like cells. Notably, mice transplanted with equal numbers of high C12-FDG+ senescent-like cells had shorter overall survival compared to mice with moderate or low C12-FDG+ cells.

Altogether, our results show that AML cells after chemotherapy can persist in a reversible senescent-like dormant state with leukemia repopulation capacity and reveal a novel mechanism of chemotherapy resistance with therapeutic potential in AML.

Disclosures

Duy:GlaxoSmithKline: Research Funding. Melnick:Janssen: Research Funding.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution