In contrast to well-established hierarchical concepts of tumor stem cells, leukemia-initiating cells in B-cell precursor acute lymphoblastic leukemia have not yet been phenotypically identified. Different subpopulations, as defined by surface markers, have shown equal abilities to reconstitute leukemia upon transplantation into immunodeficient mice. Using a non-obese diabetes/severe combined immunodeficiency human acute lymphoblastic leukemia mouse model and cell cycle analysis annotating cells to distinct cycle phases, we functionally characterized leukemia-initiating cells and found that cells in all stages of the cell cycle are able to reconstitute leukemia in vivo, with early cycling cells (G1blow population) exhibiting the highest leukemia-initiating potential. Interestingly, cells of the G2/M compartment, i.e. dividing cells, were less effective in leukemia reconstitution. Moreover, G1blow cells were more resistant to spontaneous or drug-induced cell death in vitro, were enriched for stem cell signatures and were less metabolically active, as determined by lower levels of reactive oxygen species, compared to G2/M stage cells. Our data provide new information on the biological properties of leukemia-initiating cells in B-cell precursor acute lymphoblastic leukemia and underline the concept of a stochastic model of leukemogenesis.

The FOXO1 transcription factor plays an essential role in the regulation of proliferation and survival programs at early stages of B-cell differentiation. Here, we show that tightly regulated FOXO1 activity is essential for maintenance of B-cell precursor acute lymphoblastic leukemia (BCP-ALL). Genetic and pharmacological inactivation of FOXO1 in BCP-ALL cell lines produced a strong antileukemic effect associated with CCND3 downregulation. Moreover, we demonstrated that CCND3 expression is critical for BCP-ALL survival and that overexpression of CCND3 protected BCP-ALL cell lines from growth arrest and apoptosis induced by FOXO1 inactivation. Most importantly, pharmacological inhibition of FOXO1 showed antileukemia activity on several primary, patient-derived, pediatric ALL xenografts with effective leukemia reduction in the hematopoietic, lymphoid, and central nervous system organ compartments, ultimately leading to prolonged survival without leukemia reoccurrence in a preclinical in vivo model of BCP-ALL. These results suggest that repression of FOXO1 might be a feasible approach for the treatment of BCP-ALL.

SMAC-mimetics represent a targeted therapy approach to overcome apoptosis resistance in many tumors. Here, we investigated the efficacy of the SMAC-mimetic BV6 in B-cell precursor acute lymphoblastic leukemia (BCP-ALL). In ALL cell lines, intrinsic apoptosis sensitivity was associated with rapid cIAP degradation, NF-κB activation, TNF-α secretion and induction of an autocrine TNF-α-dependent cell death loop. This pattern of responsiveness was also observed upon ex vivo analysis of 40 primograft BCP-ALL samples. Treatment with BV6 induced cell death in the majority of ALL primografts including leukemias with high-risk and poor-prognosis features. Inhibition of cell death by the TNF receptor fusion protein etanercept demonstrated that BV6 activity is dependent on TNF-α. In a preclinical NOD/SCID/huALL model of high-risk ALL, marked anti-leukemia effectivity and significantly prolonged survival were observed upon BV6 treatment. Interestingly, also in vivo, intrinsic SMAC-mimetic activity was mediated by TNF-α. Importantly, BV6 increased the effectivity of conventional induction therapy including vincristine, dexamethasone and asparaginase leading to prolonged remission induction. These data suggest SMAC-mimetics as an important addendum to efficient therapy of pediatric BCP-ALL.

Despite increasingly successful treatment of pediatric ALL, up to 20% of patients encounter relapse. By current biomarkers, the majority of relapse patients is initially not identified indicating the need for prognostic and therapeutic targets reflecting leukemia biology. We previously described that rapid engraftment of patient ALL cells transplanted onto NOD/SCID mice (short time to leukemia, TTLshort) is indicative of early patient relapse. Gene expression profiling identified genes coding for molecules involved in mTOR signaling to be associated with TTLshort/early relapse leukemia.Here, we now functionally address mTOR signaling activity in primograft ALL samples and evaluate mTOR pathway inhibition as novel treatment strategy for high-risk ALL ex vivo and in vivo. By analysis of S6-phosphorylation downstream of mTOR, increased mTOR activation was found in TTLshort/high-risk ALL, which was effectively abrogated by mTOR inhibitors resulting in decreased leukemia proliferation and growth. In a preclinical setting treating individual patient-derived ALL in vivo, mTOR inhibition alone, and even more pronounced together with conventional remission induction therapy, significantly delayed post-treatment leukemia reoccurrence in TTLshort/high-risk ALL.Thus, the TTLshort phenotype is functionally characterized by hyperactivated mTOR signaling and can effectively be targeted ex vivo and in vivo providing a novel therapeutic strategy for high-risk ALL.

Intensified treatment of pediatric acute lymphoblastic leukemia (ALL) has lead to increased survival rates of about 80%, however therapy fails in the remaining patients leading to relapse of the disease associated with inferior prognosis. Because treatment failure is, at least in part, due to defects in apoptosis programs, novel therapeutic strategies that counter apoptosis resistance are needed. “Inhibitor of Apoptosis” (IAP) proteins block the apoptosis machinery at a central point and are highly expressed in acute leukemias, thereby providing a target structure for therapeutic intervention. Molecules antagonizing these apoptosis inhibitors, so called SMAC-mimetics, therefore provide a promising strategy to overcome apoptosis deficiency and effectively treat high risk ALL. In this study, we investigated the effects of the small molecule SMAC-mimetic BV6 (kindly provided by Genentech) in B cell precursor- (BCP-) ALL.

As novel therapeutics are urgently needed for treating acute lymphoblastic leukemia (ALL) we established a poly-agent chemotherapy regimen in our NOD/SCID/huALL xenograft model investigating its effectivity on different patient-derived xenograft (pdx) ALL samples, which further can be used to evaluate novel substances in combination modalities.

In acute lymphoblastic leukemia (ALL) leukemia initiating cells (LICs) have been considered to be organized in an hierarchical fashion, however recent findings demonstrated LIC-activity also in more committed cells supporting a stochastic stem cell concept. Thus, the nature of leukemia initiating cells in ALL still remains elusive. As an alternative approach to define LICs by expression of cellular markers as commonly employed, we addressed LIC activities in ALL by functional investigation of cellular subfractions of distinct cell cycle phases. Patient-derived xenograft BCP-ALL cells were sorted according to cell cycle stages (i.e. G0/G1 and G2/M) and subsequently transplanted onto NOD/SCID mice. All cell fractions led to engraftment indicating LIC activity of all leukemia cells. However, cells isolated from G0/G1 cell cycle phases led to early leukemia onset in contrast to cells from late cell cycle (G2/M) constantly showing lowest LIC activity. Strikingly, this difference in LIC activity was maintained upon secondary transplantation. In an alternative approach, we investigated metabolic activities in cellular leukemia subfractions and identified low metabolic activity in cells of early G0/G1 cell cycle phases compared to increased cellular metabolism in cells of late G2/M. To address LIC-capacities of ALL cells with distinct metabolic activities on a functional level, cellular fractions were sorted according to low or high ROS levels and subsequently transplanted onto NOD/SCID mice. Interestingly, a prolonged engraftment was observed upon transplantation of ROShigh cells in contrast to faster leukemia repopulation in recipients transplanted with ROSlow cells, showing that the metabolic activity is indicative for its leukemia initiating activity. In summary, we identified LIC-activity in all leukemia subpopulations. Importantly, our findings indicate that leukemia initiating cells in ALL are enriched in early cell cycle and characterized by low metabolic activity.

Previously, we found that rapid leukemia engraftment (short time to leukemia, TTL short ) in the NOD/SCID/huALL (non-obese diabetic/severe combined immuno-deficiency/human acute lymphoblastic leukemia) xenograft model is indicative of early patient relapse. As earlier intact apoptosis sensitivity was predictive for good prognosis in patients, we investigated the importance of apoptosis signaling on NOD/SCID/huALL engraftment. Intact apoptosome function as reflected by cytochrome c-related activation of caspase-3 (CRAC-positivity) was strongly associated with prolonged NOD/SCID engraftment (long time to leukemia, TTL long ) of primary leukemia cells, good treatment response and superior patient survival. Conversely, deficient apoptosome function (CRAC-negativity) was associated with rapid engraftment (TTL short ) and early relapse. Moreover, an intact apoptosis signaling was associated with high transcript and protein levels of the pro-apoptotic death-associated protein kinase1 (DAPK1). Our data strongly emphasize the impact of intrinsic apoptosis sensitivity of ALL cells on the engraftment phenotype in the NOD/SCID/huALL model, and most importantly also on patient outcome.

Abstract 2985 Recently, we identified in a NOD/SCID human (hu) B-cell precursor (BCP) acute lymphoblastic leukemia (ALL) mouse model two different engraftment phenotypes that we named time to leukemia (TTL) short (TTLshort) and TTL long (TTLlong) and that reflected the rapid or late onset of leukemia in the mice. We showed that the rapid and late leukemia engraftments in the mice highly correlated with the relapse free survival of patients initially stratified as standard risk at diagnosis. In order to analyze if the two distinct phenotypes are characterized by different frequency of leukemia initiating cells (LICs), we investigated the NOD/SCID repopulating activity of 4 individual BCP ALL samples. We performed a limiting dilution transplantation assay using freshly isolated spleen cells form 2 TTLshort and 2 TTLlong xenografted mice with full-blown leukemia. We created 5 groups of 8 mice each and transplanted intravenously 105, 104, 103, 102, 101 cells per mouse/group. Leukemia engraftment in the peripheral blood (PB) was routinely evaluated by flowcytometry and the mice were sacrificed at the onset of disease manifestation. We defined a mouse as engrafted when we could detect ≥ 1% CD19 positive cells in the PB. In 3 out of 4 samples, 100% of the mice transplanted with 105 and 104 cells engrafted; only in one sample (TTLlong) we had a 75% and 12,5% successful engraftment using 105 and 104 cells respectively. The percentage of engrafted mice using 103 cells was 87,5% and 75% for the 2 TTLshort samples and 37,5% and 0% for the 2 TTLlong specimens. No mice transplanted with the 102 cells isolated from the 2 TTLlong samples engrafted whereas 50% and 12,5% of the mice transplanted with the 102 cells isolated from the 2 TTLshort engrafted successfully. No one of the mice transplanted with 101 cells engrafted. The LIC frequency was calculated using the Poisson single-hit model; a higher LIC frequency was calculated in the 2 TTLshort (1/329 and 1/739) compared to the 2 TTLlong (1/2159 and 1/74028). As cells in the G0 phase of the cell cycle are considered as the reservoir for the new cycling cells and putative LICs, we analyzed by flowcytometry the percentage of G0 cells, defined as cells staining negative for the Ki-67 antigen, within TTLshort (N=8) and TTLlong (N=7). No statistical difference was found between the analyzed samples. Nevertheless, we observed that in both TTLshort and TTLlong almost all cells were actively progressing through the cell cycle. Aiming to further characterize distinctive biological features related to the different LIC frequency, we analyzed by flowcytometry the cell cycle profiles of 9 TTLshort and 8 TTLlong mice. We used a cell cycle analysis based on the simultaneous labeling of DNA (7-AAD) and RNA (Pyronin Y) that allows to distinguish between cells that are in G1 but have not yet actively entered the S phase of the cell cycle (named G1a cells) from cells that are actively progressing form G1 to S (named G1b cells). A higher percentage of G1a cells was detected in the TTLshort compared to TTLlong (p = 0.004 Mann-Whitney Test) suggesting the presence of a stand-by-cell fraction in the TTLshort almost ready to proceed through the cell cycle. On the other hand, a higher percentage of G1b cells was found in the TTLlong compared to TTLshort (p = 0.001, Mann-Whitney Test). No significant difference was found in the proportion of cells in the “S/G2/M subgroup”. We also analyzed by Western Blot CYCLIN B1 expression in TTLshort (N=5) and TTLlong (N=5): a statistically different CYCLIN B1 expression was found between the two groups (p = 0.009 Mann-Whitney Test). As this molecule is involved in the progression of cells from the G2 to M phase of the cell cycle, we analyzed by flowcytometry the mitotic cells fraction defined as the cells positive for the phosphorilation of the Ser10 on the histone H3 in 6 TTLshort and 5 TTLlong specimens. A higher proportion of P-H3(Ser10) positive cells was detected in the TTLshort compared to the TTLlong (p = 0.045, Mann-Whitney Test). All together these data indicate that the two NOD/SCID engraftment phenotypes are characterized by different frequency of LICs. Furthermore, our functional analyses reveal a distinctive progression of the leukemic cells through the cell cycle showing a higher number of cells (G1a cells) ready to progress through the G1-S phases in the TTLshort/poor prognosis leukemia. Disclosures: No relevant conflicts of interest to declare.

We recently showed that rapid engraftment of patient ALL in NOD/SCID mice is indicative for poor patient survival. Moreover, gene expression analysis identified differenzial expression of molecules regulating the mTOR pathway. We now functionally address mTOR activation assessing P-S6 levels in xenograft ALL and evaluate mTOR inhibition as novel treatment strategy ex vivo and in vivo.

Abstract 2518 B cell precursor acute lymphoblastic leukemia (BCP-ALL) is a heterogeneous disease and frequently associated with genetic alterations. However about one quarter of ALL patients lack characteristic chromosomal rearrangements representing a subset of leukemia not well understood. Expression of cytokine receptor-like factor 2 (CRLF2) has recently been shown to be up-regulated as well as mutated in BCP-ALL patients, including Down syndrome (DS-ALL) patients, lacking recurring chromosomal translocations. Several alterations in CRLF2 resulting in its overexpression have been described: a focal ∼320 kilobase interstitial deletion of the pseudo-autosomal region of the sex chromosomes (Xp22.23 or Yp11.32) that creates a fusion of CRLF2 to the G-protein-coupled purinergic receptor P2RY8 gene (P2RY8-CRLF2), translocation of CRLF2 with the IGH@ locus of chromosome 14, and a point mutation in CRLF2 resulting in a phenylalanine-to-cysteine substitution at amino acid 232 (F232C). Recently it has been shown that patients with CRLF2-P2RY8 fusion gene are associated with poor prognosis although they were initially stratified into intermediate risk (IR) group by minimal residual disease (MRD) criteria. This emphasizes the need to identify the mechanism of CRLF2 disregulation in leukemia thereby providing new therapeutic strategies. We have characterized leukemia samples for the presence of CRLF2 expression at the protein level by FACS, at transcript level by RT-PCR, mutations in CRLF2 and JAK2 by next generation amplicon sequencing and CRLF2-P2RY8 fusion gene and IGH@ translocation by RT-PCR. Interestingly, in our recent study we have observed activated mTOR pathway in a subset of B-ALL xenograft leukemia without recurring genetic alterations that were sensitive to mTOR inhibition ex vivo and in vivo. In view of the recently suggested interconnection between TSLP/CRLF2 and PI3K/mTOR signaling networks we analyzed the mTOR pathway activation in CRLF2 rearranged ALL. In this study we functionally address mTOR pathway activation in xenograft ALL samples with CRLF2 rearrangements and without any known chromosomal rearrangements ex vivo. We have adopted the NOD/SCID/huALL xenotransplant mouse model for this study. Xenograft leukemia cells with CRLF2 rearrangements and without recurring genetic alterations (CRLF2 rearrangements, n=4; CRLF2 wild type, n=11) were re-transplanted onto NOD/SCID mice and ALL cells were harvested at disease onset from leukemia bearing mice. The mTOR pathway activation was analyzed by flow cytometry assessing phosphorylation of ribosomal protein S6 (P-S6), a molecule downstream of the mTOR pathway. No significant difference in basal P-S6 level was found between CRLF2 rearranged and CRLF2 wild type leukemia. Moreover, reduction of the mTOR pathway activation upon mTOR inhibition by rapamycin and dual PI3K-mTOR inhibitor BEZ235 was analyzed. No significant reduction of the mTOR activity was observed in CRLF2 rearranged leukemia compared to CRLF2 wild type leukemia. Since JAK2 mutations have been associated with CRLF2 rearrangements, we also analyzed the JAK-STAT5 pathway assessing P-STAT5 ex vivo. Only one patient derived xenograft ALL sample carried a JAK2 mutation however no differential activation of JAK-STAT5 pathway was observed between JAK2wt/CRLF2r and JAK2mut/CRLF2r samples ex vivo. Also no significant difference in basal P-STAT5 was observed between CRLF2 rearranged and CRLF2 wild type leukemia. Additionally we checked the effect of mTOR inhibition on P-STAT5. No significant reduction in P-STAT5 level upon mTOR inhibition was observed pointing P-STAT5 activation independent of mTOR activation. In this study we have shown that CRLF2 rearranged leukemia is not associated with activated mTOR pathway and also not sensitive to mTOR inhibition either by rapamycin or dual PI3K-mTOR inhibitor NVP-BEZ235. In CRLF2 rearranged and CRLF2 wild type leukemia JAK-STAT5 pathway did not show differential activation and was unaffected by mTOR inhibition. It remains to be clarified whether both pathways could be activated with appropriate stimuli as was demonstrated for diagnostic specimens. To fully understand the exact mechanism of CRLF2 rearranged leukemia development and providing new therapeutic strategies, study the interlinks with other cell-signaling pathways is mandatory. Disclosures: No relevant conflicts of interest to declare.