Session: 661. Malignant Stem and Progenitor Cells: Poster II
Telomerase-deficient AML was generated by retroviral transduction of G3 Terc-/- LKS+ (Lin-Kit+Sca1+, enriched for hematopoietic stem cells) with pMIG-MLL-AF9 and compared to wild type (WT) controls. Transformed Terc-/- LKS+ colony-forming units (CFU) were mildly reduced at early passage (week 1 Terc-/- 13.1 ± 1 vs. WT 32.7 ± 4 per 1000 cells input, p < 0.01) but became progressively extinguished with serial replating (week 6 Terc-/- 3.8 ± 0.4 vs. WT 27.0 ± 2.7 per 1000 input, p < 0.01). Loss of CFU correlated with enforced differentiation (reduced Kit, increased Gr1), cell cycle arrest and preferential apoptosis of Kit+ cells. In vivo, AML developed with delayed latency, but was fully penetrant in recipients of G3 Terc-/- and WT cells (Terc-/- 64 days vs. WT 45 days, p < 0.01). Leukemic burden and leukemia stem cell (LSC, GFP+Lin-Sca1-Kit+FcgR+CD34-) frequency were similar between Terc-/- and WT AML.
To determine the consequences of telomerase loss on AML LSCs, we performed gene expression profiling of purified LSCs. MLL-AF9-Terc-/- LSCs revealed enrichment of pathways controlling DNA damage/repair, cell cycle and apoptosis. Upstream analysis predicted activation of p53, Rbl1 and Cdkn2a, and inhibition of E2f1 in Terc-/- LSCs. Functionally, shRNA-mediated knockdown of p53 in Terc-/- LSCs partially rescued in vitro CFU, differentiation, cell cycle arrest and apoptosis.
The phenotypic changes in Terc-/- AML were amplified by serial passage, suggesting that replicative stress may exacerbate the deleterious effects of telomerase loss on LSC function. To enforce replicative stress in vivo, we performed serial transplantation of Terc-/- AML vs. WT AML. Terc-/- LSCs were unable to generate secondary AML (survival Terc-/- not reached vs. WT 28 days, p < 0.01) and this was confirmed by limiting dilution analysis (Terc-/- LSC frequency 1:224,000 vs. WT 1:184 p < 0.001). Initial engraftment was similar between Terc-/- and WT LSCs. In vivo leukemogenesis was prevented by cell cycle arrest, DNA damage and massive apoptosis (Terc-/- 76.8% ± 3.6% vs. WT 22.49% ± 2.3%, p < 0.0001). Together, these findings demonstrate that in this murine model, telomerase loss eradicates LSC in vivo.
To validate these findings in human AML we performed lentiviral shRNA knockdown of hTERT in the MLL-AF9-containing AML cell line Monomac6, followed by transplantation into NSGS (NOD/SCID/IL2Rgamma-/- transgenic for hSCF/hIL3/hGMCSF) xenograft recipients. Two independent shTERT constructs revealed significantly increased survival compared to non-transduced and non-targeting controls (sh-hTERT 149.5 days and 146 days vs. non-transduced 47 days, non-targeting 53.5 days, p< 0.01 for both sh-hTERT). hTERT knockdown correlated with reduced hCD45 engraftment, induction of DNA damage, cell cycle arrest and apoptosis compared to non-transduced or non-targeting shRNA controls. Pharmacological inhibition of telomerase (Telomerase Inhibitor IX, T-IX, Sigma) reduced growth of multiple human AML cell lines in vitro. Treatment with T-IX followed by transplantation of equal numbers of Monomac6 cells prolonged NSGS xenograft survival (T-IX 96.5 days vs. DMSO 59 days, p < 0.05).
Finally, we examined the prognostic impact of telomerase-regulated genes in a large cohort of patients with AML. The top 140 differentially expressed genes (p < 0.001) between murine Terc-/- and WT LSCs predicted survival in 2 independent AML clinical trial cohorts. Computational modelling using a random forest approach was able to identify 10 key telomerase-regulated human homologues that could cluster AML patients into prognostically relevant groups and was reproducible in multiple independent datasets.
These findings provide new mechanistic understanding into the effects of telomerase inhibition on MLL-rearranged AML and identify the telomerase complex as a novel therapeutic target for AML.
Disclosures: No relevant conflicts of interest to declare.
See more of: Oral and Poster Abstracts
*signifies non-member of ASH