Functional Genomics and Computational Approaches Identify Novel Small Molecules Targeting Quiescent Leukemia Stem Cells

Chemotherapy or targeted cancer therapies have greatly improved the treatment outcome of patients with leukemia; however, many will ultimately die because of disease relapse and development of drug resistance. Leukemias are cancers of the blood cells that result from alteration of the normal physiological constraints that regulate hematopoietic stem cells (HSCs). General characteristics of leukemia stem cells (LSCs) such as self-renewal, self-protection and proliferative quiescence represent inherent mechanisms that at least partially explain drug resistance and recurrence in post-therapy leukemia patients.

Acute myeloid leukemia (AML) is a heterogeneous disease, both biologically and clinically, in which a number of distinct genetic abnormalities have been described. Several recent studies suggest that this heterogeneity extends to LSCs and can vary between patient subgroups, and even within individual patients. Moreover, the complexity of AML is further complicated by the existence of functionally diverse leukemic and preleukemic clones. Accordingly, the hierarchical organization of AML suggests that this may be relevant to current therapies that primarily target proliferating progenitors/blast cells, which lack self-renewal capacity, and not LSCs. In the current study, we rationalized that understanding how LSCs differ from normal HSCs at the molecular level, is an essential first step towards developing novel targeted therapies and achieving permanent disease remission.

Despite the identification of novel LSC-specific markers, there is considerable heterogeneity in expression of these markers amongst AML patients. However, in addition to marker-enrichment strategies, LSCs can be identified by virtue of their quiescent and slow-cycling properties. For example, label-retaining cells can be isolated and used in functional assays but significant technical limitations impede broad utility of this approach. To this end, we describe the development and use of novel multi-fluorescent protein markers and DNA bar codes integrated into the cellular genomes by lentivirus, as single-cell tracking devices for monitoring LSCs in vivo. We demonstrate how LSCs can transition between a "proliferation phase" and a "quiescence phase" in vivo. Furthermore, using high-throughput quantitative transcriptome sequencing (Q-RNA-Seq) and RNAi genetic perturbation's focusing on well-defined self-renewal signaling pathways, we develop a differential network-based model to identify LSC-specific genes and subsequently prioritize/rank candidates as potential drug targets.

In the current study, we identify several molecular targets deregulated in quiescent versus proliferating LSCs and a mutual set of signaling pathways that facilitate leukemic transformation downstream of diverse initiating mutations/lesions. Remarkably, both quiescent and dividing LSCs but not HSCs, were 'addicted' to SSRP1 - an essential component of the ubiquitous FACT chromatin remodeling complex. Two orally available quinacrine-related DNA-intercalating compounds inhibiting function of FACT (CBL0100 and CBL0175, respectively) suppressed LSC proliferation in vitro and in vivo, as demonstrated by production of leukemic clonogenic cells (CFU) and long-term engraftment of immunodeficient NSG mice, by simultaneous inhibition of NF-kB (stimulated and basal forms) and activation of p53. Furthermore, in a secondary transplantation experiment, leukemic cells obtained from CBL0175 treated mice (primary) failed to engraft into secondary NSG mice in a serial transplantation model by selectively targeting the LSC compartment. Collectively, we present a novel network-based polypharmacology approach that provides unique opportunities to preferentially ablate LSCs (quiescent and dividing types), with potentially profound clinical implications.