Acute Myeloid Leukemia Specimens Research Articles

LncRNA LBX2-AS1 inhibits acute myeloid leukemia progression through miR-455-5p/MYLIP axis

Acute myeloid leukemia (AML) is a common blood cancer primarily affecting the bone marrow and blood cells, which is prevalent among adults. Long non-coding RNAs (lncRNAs) have been shown to play a crucial role in the development and progression of AML. LBX2-AS1 is a recently discovered lncRNA that has been linked to the pathogenesis and progression of several types of cancer. This study aimed to investigate the role and possible mechanisms of LBX2-AS1 in AML. Expression levels of LBX2-AS1, miR-455-5p, and their target genes were detected in AML samples and cells by RT-qPCR. Cell proliferation and apoptosis were determined by Cell Counting Kit-8 and 5-ethynyl-2′-deoxyuridine assays, and flow cytometry, respectively. LBX2-AS1 was downregulated in AML specimens and cells, and overexpression of LBX2-AS1 significantly inhibited cell proliferation and enhanced apoptosis in vitro. We also determined the effects of LBX2-AS1 overexpression in an AML mouse model by in vivo bioluminescence imaging. Mechanistically, LBX2-AS1 acts as a competitive endogenous RNA, which promotes myosin regulatory light chain interacting protein (MYLIP) expression by sponging miR-455-5p. Knockdown of MYLIP or upregulation of miR-455-5p antagonized the effect of LBX2-AS1 overexpression on the progression of AML. LBX2-AS1 may thus be a valuable therapeutic target for AML.

Polyamines Are Required for AML LSC Function through Their Role in Regulating eIF5A Dependent Protein Synthesis

Outcomes for patients with acute myeloid leukemia (AML) remain poor due to the inability of current therapies to fully eradicate disease initiating leukemia stem cells (LSCs) (Shlush et al., 2017). Many studies have demonstrated that LSCs have unique metabolic properties compared to AML blasts and normal hematopoietic stem and progenitor cells (HSPCs) (Jones et al., 2021). In this study we sought to quantify metabolite levels in LSCs compared to HSPCs using an unbiased metabolites screen. To examine metabolite levels in enriched LSCs and HSPCs we used mass spectrometry-based metabolomics analysis on enriched LSCs from 18 AML patients and HSPCs from 5 normal bone marrow (BM) specimens. Pathway analysis showed that arginine metabolism and biosynthesis were the most significantly enriched pathways in LSCs compared to HSPCs. Clinical studies have shown limited efficacy of targeting arginine levels using arginine degrading enzymes; therefore, we sought to target pathways downstream of arginine. To identify the pathways that arginine metabolism supports in LSCs, we used stable isotope labeled tracing analysis which showed arginine was metabolized through the urea cycle into polyamines in LSCs. Interestingly, the polyamine spermidine was the most significantly enriched metabolite in LSCs compared to HSPCs in our metabolomics analysis. Moreover, expression of SAT1, the enzyme catalyzing the export of polyamines from the cells, was decreased in AML compared to normal BM. Based on these observations, we sought to determine if the polyamine pathway represents a metabolic vulnerability that can be exploited to target LSCs. Polyamines are cationic compounds which have been previously shown to be important for the growth of various cancers but their role in AML LSCs has yet to be determined. To investigate if polyamines are essential for LSC survival and function, we treated LSCs and HSPCs with N 1, N 11-diethylnorspermine (DENSpm), a polyamine analog causing polyamine depletion through increased SAT1 expression and subsequent cellular export. DENSpm treatment decreased the viability of LSCs in 10 out of 12 of patients tested but had no impact on HSPCs viability. Further, treatment with DENSpm resulted in decreased engraftment potential of 5 AML specimens but did not alter the engraftment of HSPCs from 3 normal BM specimens, suggesting that targeting polyamines represents a promising approach to kill LSCs while sparing HSPCs. We then investigated the mechanisms by which polyamines are essential in LSCs. Using RNA-sequencing analysis on enriched LSCs from 5 AML patients and HSPCs from 3 normal BM specimens we observed an enrichment in myeloid differentiation signatures in DENSpm-treated LSCs but not in DENSpm-treated HSPCs. In line with this, expression of myeloid markers CD11b and CD15 were increased in patient-derived xenograft models engrafted with DENSpm-treated LSCs. As we did not observe such dysregulation in mice engrafted with DENSpm-treated HSPCs, these data suggest polyamine depletion promotes differentiation of LSCs but not HSPCs. Our transcriptomic analysis also unveiled decreased expression of genes involved in protein synthesis in DENSpm-treated LSCs but not in normal HSPCs. Protein synthesis and its regulation are critical for AML survival (Chen et al., 2019; Messling et al., 2022) but the role of polyamines in protein synthesis in LSCs has not been explored. Polyamines stimulate translation notably by serving as precursors for the hypusination of eukaryotic translation initiation factor 5A (eIF5A) (Park et al., 2010). Accordingly, DENSpm led to decreased hypusination of eIF5A in AML cells. This was associated with reduced protein synthesis level in AML cell lines and LSCs but not in normal HSPCs. Strikingly, eIF5A hypusination, protein synthesis and viability were rescued upon co-treatment with spermidine. Moreover, knockdown of eIF5A using siRNA resulted in impaired colony forming potential of AML and altered engraftment in patient-derived xenograft models. Altogether, these results suggest that DENSpm alters AML cell and LSC viability and function at least in part through decreased eIF5A-mediated protein translation. Altogether, our data demonstrates that depletion of polyamines decreased LSC function while sparing normal HSPCs. Mechanistically, our data suggest that LSC targeting is accomplished through decreased eIF5A-mediated synthesis of proteins crucial for LSCs.

NUP98-Rearranged Acute Myeloid Leukemia Stem Cells Superimpose Type I Interferon Signaling with Canonical Self-Renewal Programs

The clinical outcomes of pediatric acute myeloid leukemia (AML) have shown improvement with combination therapies; however, relapse remains a significant concern due to the resistance of leukemia cells to conventional treatments. In some cases, epigenetic heterogeneity within leukemias can contribute to therapy resistance. For example, leukemia stem cells (LSCs) have been shown maintain AML, both before and after therapy. This study aims to identify LSCs within a mouse model of high-risk, NUP98-rearranged pediatric AML, and to identify mechanisms that sustain this population. To identify and isolate a LSC subpopulation, we characterized murine NUP98-HOXD13/Flt3-ITD-driven AML in mice by performing cellular indexing of transcriptomes and epitopes (CITE-seq) analysis at two stages: pre-AML and fully transformed AML. Uniform Manifold Approximation and Projection (UMAP) revealed a highly reproducible pattern of heterogeneity within each AML specimen. Each AML had a small percentage of cells that resembled normal and pre-leukemic multipotent progenitors (MPPs), as well as additional populations with B-cell or myeloid gene expression. Pseudotime analysis indicated a trajectory from MPPs to mature cell-like lineages, implicating the MPP-like population as a potential LSC. This population expressed type I interferon (IFN-1) target genes more highly than non-LSC AML cells and pre-AML cells. This finding is significant because we previously identified IFN-1 as an effector of Flt3-ITD/NUP98-HOXD13 synergy during AML initiation, and in human NUP98-rearranged AML (PMID: 36395068). Differential gene expression analysis also identified several established self-renewal genes, including Hlf, Myct1, Mpl, Slc7a8, and Gata2, as highly expressed in putative LSCs relative to non-LSCs (figure B). This suggests that AML LSCs superimpose inflammatory signaling programs, particularly IFN-1, and core stem cell self-renewal programs to maintain a malignant state. The data raise the question of how IFN-1 shapes LSC identity. We next sought to isolate and validate the putative LSC population. CITE-seq feature barcode data indicated high expression of CD117 and CD48, medium expression of B220, and low expression of Gr1 and CD11B in putative LSCs. We isolated LSCs based on these markers (Gr1-CD11b-B220 MidCD117 highCD48+). Several non-LSC populations were isolated for comparison (e.g., Gr1+CD11B+ and Gr1-CD11B-B220 highCD117 Mid populations. Cytospin observation revealed no morphological differences among the three populations, but cell culture assays showed that only the Gr1-CD11b-B220 MidCD117 highCD48+ cells exhibited proliferative capacity after multiple passages. Limit dilution transplant assays showed that the Gr1-CD11b-B220 MidCD117 highCD48+ population was greatly enriched for LSC function relative to whole bone marrow (figure A). Altogether, we have shown that we can prospectively isolate a LSC population from NUP98-HOXD13/Flt3-ITD AML. The LSC population is highly reproducible across specimens and it demonstrates high degrees of IFN-1 activity. Further work will establish the role of IFN-1/self-renewal gene interactions in LSC maintenance.

PDZD2 Is Required for Normal Erythropoiesis and Its 37-KDa Secreted Product, sPDZD2, Functions As a Soluble Tumor Suppressor in AML

PDZ domain-containing protein 2 (PDZD2) is expressed in normal hematopoietic stem and progenitor cells (HSPC), specifically HSC, MPP, and erythroid progenitors, however, it is almost universally lost through hypermethylation in acute myeloid leukemia (AML). PDZD2 encodes for a 300-kDa protein containing 6 PDZ domains. The full-length protein is cleaved in the cytoplasm in a caspase 3-dependent manner, and the C-terminal fragment containing PDZ domains 5 and 6 is secreted to the tissue microenvironment as a 37-kDa protein known as secreted PDZD2 (sPDZD2). sPDZD2 has been reported to function as a tumor suppressor in some solid tumors, including breast, prostate, and liver cancer. Thus, we hypothesized that PDZD2 is required for normal hematopoiesis and that sPDZD2 functions as a soluble tumor suppressor in the hematopoietic system. We first sought to determine if sPDZD2 is secreted to the bone marrow (BM) microenvironment. Using an ELISA to quantify sPDZD2 levels in BM plasma isolated from healthy donors (n=6) and primary AMLs (n=20), we confirmed its secretion by normal CD34+ HSPCs (NBM) and downregulation in AMLs (NBM: 169 ± 17.53 vs AML: 46.12 ± 6.70 ng/ml; p<0.0001). In addition, AML cell lines (n=7) secreted significantly lower levels of sPDZD2 into the culture medium than normal HSPCs (n=2) (NBM: 28.71 ± 1.76 vs AML: 4.62 ± 1.19 ng/mL; p<0.0001). Next, to study the role of PDZD2 in normal hematopoiesis we performed in vitro myeloid and erythroid differentiation and colony-forming unit assays in CRISPR-Cas9-edited human HSPCs. CD34+ cells from healthy donors were electroporated with recombinant Cas9-sgRNA ribonucleoprotein complexes targeting PDZD2 and editing confirmed by T7I endonuclease assay and Sanger sequencing. PDZD2-edited cells (PDZD2 KO) were plated either on methylcellulose or liquid differentiation cultures (n=6). PDZD2 KO resulted in decreased colony number of all types (Control: 172.5 ± 34.37 vs PDZD2 KO: 137.0 ± 18.25, p<0.05), indicating reduced colony-forming potential. In addition, loss of PDZD2 resulted in a delay in terminal erythroid differentiation with an accumulation of pro-erythroblasts (CD71+CD235a lowCD105 high, p<0.05) and late basophilic erythroblasts (CD71+CD235a highCD105 int, p<0.01) as well as decreased orthochromatic erythroblasts (CD71+CD235a highCD105 neg, p<0.01) in a tri-phasic liquid culture system. By contrast, PDZD2 KO had no impact on myeloid differentiation. To determine the role of Pdzd2 in vivo, we took advantage of a Pdzd2 Gt/Gt gene-trap (GT) mouse model and performed peripheral blood (PB) and BM analysis compared to wild-type (WT) controls. PB counts at 2 months of age (n=6-9 per genotype) showed anemia, with a significant decrease in both red blood cell counts (WT: 9.04 ± 0.26 vs GT: 7.58 ± 0.47 x10 6/mL; p=0.026) and hemoglobin levels (WT: 14.88 ± 0.46 vs GT: 12.30 ± 0.81 g/dL; p=0.036), an observation compatible with the impaired erythropoiesis seen in vitro after PDZD2 KO in human HSPCs. In addition, we observed a trend to decrease megakaryocyte-erythroid progenitors (MEPs) and short-term HSCs in the BM of 3 months-old mice (n=3), though the small size of the cohort prevented robust statistical analysis. Next, we sought to evaluate the tumor suppressor role of sPDZD2 in the context of AML. We treated a panel of AML cell lines (n=9) and primary human AML specimens (n=4) with recombinant sPDZD2 ( r-sPDZD2; dose: 100nM-300nM/day) for a period of 8 days. AML cells showed dose-dependent growth inhibition (CellTiter-Glo, p<0.05 for all) and increased expression of either CD11b or CD15 differentiation markers. Cell cycle analysis on day 8 revealed that AML cells treated with r-sPDZD2 were arrested at G0/G1 phase compared to control. In summary, our study is the first to shed light on a previously unrecognized role for PDZD2 in hematopoiesis. Loss of PDZD2 in HSPC resulted in impaired erythroid differentiation both in vitro and in vivo. Moreover, our findings validate a soluble tumor suppressor role for sPDZD2 in AML, similar to that seen in solid tumors. Importantly, we further demonstrate a potential role for r-sPDZD2 as a novel therapeutic approach for AML.

Identifying Stress Granules As Determinants of Leukemia Stem Cell Maintenance and Stress Adaptation

The development of novel leukemic stem cell (LSC)-targeted therapeutics, which spare hematopoietic stem cells (HSC), is an urgent goal towards reducing the high, LSC-driven, relapse rates in acute myeloid leukemia (AML) patients. Towards elucidating key post-transcriptional networks underlying LSC maintenance we previously performed an in vivo pooled CRISPR dropout screen targeting 128 RNA binding proteins (RBPs) highly expressed in AML LSCs while having low expression in HSCs. This screen highlighted 12 RBPs crucial for murine AML LSC-mediated leukemic growth, of which there was a significant enrichment of stress granule (SG) RBPs. Further, gene set enrichment analysis shows the SG proteome is enriched in human AML LSCs vs non-LSCs while normal HSCs have reduced SG RBP expression relative to multipotent progenitor cells. SGs are dynamic biomolecular condensates, composed of ribonucleoprotein complexes, which may form as a protective mechanism against a variety of stress conditions. SGs are gaining increasing attention as contributors to cancer metastasis and chemoresistance, but their role in LSCs is virtually unexplored. The intriguing enrichment of SG RBPs in our screen hints at a larger role for SGs in LSC maintenance and/or stress resistance. To address this, we focused on the core SG nucleation RBP, G3BP1, a molecular switch that guides SG assembly. Fluorescence microscopy was first performed to describe SG dynamics in leukemia vs primitive normal cells where interestingly, a subset of patient AML cells contained G3BP1 SGs in the absence of added stress. In contrast, primitive healthy murine stem cells had reduced heat shock-induced SG formation in comparison to downstream normal progenitors. A series of shG3BP1 knockdown experiments performed in human leukemia cell lines and patient AML specimens demonstrated that shG3BP1-transduced cells showed increased apoptosis and differentiation and reduced colony forming capacity and competition in vitro compared to control cells. The same set of assays showed minimal impact of SG impairment to healthy umbilical cord blood cells. Importantly, in vivo xenotransplantation demonstrated that LSCs with G3BP1 knocked down were significantly compromised in their capacity to generate leukemic grafts whereas normal HSCs assessed in a similar manner display much less G3BP1 dependency. Towards determination of the SG mechanisms essential for bulk AML and LSC regulation, we performed matched RNA-sequencing and proteomics in G3BP1 knocked down AML cells as well as BioID in AML cell lines, primary patient AML cell subsets, and healthy hematopoietic cells to allow for the uncovering of an AML specific SG proteome. These analyses highlighted numerous potential SG mechanisms of cancer cell stress resistance and/or stem cell regulatory processes including sequestration of apoptotic factors and oncogenic signalling pathways. Of note, RNA-seq and proteomics in SG impaired shG3BP1 leukemia cells showed both a negative enrichment of the SG proteome and LSC signature as well as a positive enrichment of the innate immune response. Interestingly, our BioID uncovered that the AML specific SG proteome is also enriched for positive regulation of RIG-I signaling pathway (GO: Biological Process). Together this data suggests SGs play a protective role through the downregulation of overactive innate immune signalling in AML. Work using optimized small input approaches to map SG RNA and protein interactomes in LSCs vs HSCs to confirm similar regulatory relationships as well as identify LSC-context specific ones is ongoing. Overall, these results elucidate SGs as key AML LSC stress adaptive regulatory networks and highlights them as novel targets for consideration in AML therapeutics.

Cladribine Combined with Homoharringtonine and Cytarabine Achieves a High Remission Rate in Adult Patients with De Novo Acute Myeloid Leukemia, Especially for Adverse-Risk Group: A Prospective, Single Center, Single-Arm, Phase 2 Study

Background: For adult de novo acute myeloid leukemia (AML), the standard 3+7 induction regimen has approximately 60% complete remission (CR) rate. Recently, BCL-2 inhibitor plus intensive chemotherapies as novel induction regimens have shown an above 90% composite complete remission (CRc) rate, however still have limited response effect in an adverse risk group, especially for patients with specific cytogenetic abnormalities such as KMT2A rearrangements, FLT3, RAS, and TP53 mutations. Multiple studies have shown that homoharringtonine (HHT) can target leukemia diseases that have switched their anti-apoptotic dependence from BCL-2 to MCL-1. We recently demonstrated that cladribine (CdA) could eradicate BCL-2 inhibitor-resistant monocytic leukemia stem cells driven by RAS mutants and KMT2A rearrangements. (Pei S, et al. Cancer Discovery. 2023) Together, these clinical and basic findings suggest a rationale for combining CdA with HHT to treat AML patients. In this study, we report the efficacy and safety of a newly designed “CHA” regimen composed of CdA, HHT, and cytarabine (AraC) in adult patients with de novo AML. Methods: This is a prospective, single-center, single-arm phase 2 trial. Newly diagnosed AML patients aged 18-59 years were enrolled. The diagnosis, risk classification, and response evaluation were determined according to ELN 2022 criteria. All enrolled patients received the CHA regimen: CdA (5mg/m2, days 1-3), HHT (2mg/m2, days 1-5), and AraC (100mg/ m2, days 1-7) administered intravenously. The response was evaluated on day 28 via bone marrow routine and minimal residual disease (MRD) measurement by flow cytometry. The primary endpoint was CRc (CR plus CRi). Secondary endpoints were MRD, overall survival (OS), disease-free survival (DFS), event-free survival (EFS), and adverse events. Bone marrow (BM) samples on day 1 and 7 and peripheral blood (PB) samples on days 1-7 were collected for laboratory flow analysis to monitor the kinetics of leukemia reduction. We also treated primary AML specimens with single, dual, and triple combinations of CdA, HHT, and Ara-C to evaluate their anti-leukemic activity in vitro. Statistical analyses were carried out using SPSS 24.0. Survival curves were prepared using the Kaplan-Meier method. A two-tailed P value <0.05 was considered statistically significant. This trial is registered on ClinicalTrials.gov (NCT05906914). Results: A total of 33 patients were enrolled from August 1, 2021, to April 15, 2023. The median age was 49 years. 33% (11/33) of patients were at favorable risk, 42% (14/33) were at intermediate risk, and 24% (8/24) were at adverse risk. After one course of CHA induction therapy, the CRc rate was 87% (27 of 31 patients), and the CR rate was 84% (26 of 31 patients). The CR rates in the favorable, intermediate, and adverse-risk groups were 82% (9 of 11 patients), 75% (9 of 12 patients), and 100% (8 of 8 patients), respectively. Notably, six out of six patients with KMT2A rearrangement and ten out of ten patients with RAS mutations achieved MRD-negative CR. The overall MRD negativity was 89% (24 of 27 patients with CRc). In 31 evaluable patients, four patients did not reach CR; three of them held FLT3 mutations. 30-days treatment-related mortality (TRM) was 6% (2 of 33 patients), and both died of sepsis. The median neutrophil recovery time (from the end of chemotherapy to the neutrophil counts ≥0.5×109/L) and platelets recovery time (from the end of chemotherapy to platelets counts ≥20×109/L) were both 11 days. With a median follow-up time of 13 months (IQR: 6-17 months), the one-year OS was 86%, 1-year DFS was 89%, and the 1-year EFS was 76%. Flow cytometry analysis of serially collected PB and BM samples from nine randomly selected patients revealed that an average of 91.0% and 94.7% of leukemia burden was cleared in the PB and BM by day 7, respectively, demonstrating a rapid and effective response induced by CHA. Finally, our in vitro assay demonstrated a synergistic effect of CdA, HHT, and Ara-C on primary AML specimens including three patients with adverse-risk. Conclusion: We propose the CHA regimen as a novel, effective, and safe chemotherapy regimen for newly diagnosed adult AML patients with remarkable potency in treating adverse-risk group patients, especially those with KMT2A rearrangements and RAS mutations that are otherwise resistant to BCL-2 inhibitor-based therapies, but a limited effect on FLT3mutated AML patients.

Circ_0001946 Promotes the Development of Acute Myeloid Leukemia by Upregulating PDL1

Circ_0001946 has been identified as an oncogenic factor, and the aim of this study was to explore the detailed roles and putative targets of circ_0001946 in acute myeloid leukemia (AML). Levels of circ_0001946 were examined in AML tissues and cells. Furthermore, the regulatory functions of circ_0001946 in AML were explored. The expression of circ_0001946 was evaluated in AML samples and a matched para-carcinoma control, as well as in AML cell lines and a human bone marrow stromal cell line using reverse transcription-quantitative polymerase chain reaction. Cell proliferation was examined using a CCK-8 kit, and migration/invasion was measured by transwell assay. Furthermore, interactions between associated molecules were assessed using RNA pulldown, and the mRNA stability of the relevant gene was examined by mRNA stability assay. Our data indicated that circ_0001946 was upregulated in AML specimens/cells. Additionally, overexpression of circ_0001946 promoted the proliferation, migration, and invasion of AML cells and, vice versa, these biological processes were suppressed by knockdown of circ_0001946. Furthermore, PDL1 is a potential downstream molecule of circ_0001946 in AML and its stability was improved by circ_0001946. The expression of PDL1 was increased in AML specimens and positively correlated with circ_0001946 expression. Moreover, biological behavioral alterations in AML cells induced by oe-circ_0001946 were abrogated by sh-PDL1 and the effects of sh-circ_0001946 were enhanced by treatment with sh-PDL1. Taken together, these data suggest that levels of circ_0001946 are elevated in AML and that circ_0001946 could promote the growth of AML cells. Furthermore, PDL1 is a novel downstream molecule of circ_0001946 in AML. Circ_0001946/PDL1 signaling may play crucial roles in tumor progression in AML and could be a novel candidate for targeted treatments for AML patients.

Abstract A54: Polyamines are required for AML LSC function through their role in regulating eIF5A dependent protein synthesis

Abstract Outcomes for patients with acute myeloid leukemia (AML) remain poor due to the inability of current therapies to fully eradicate disease initiating leukemia stem cells (LSCs) (Shlush et al., 2017). Many studies have demonstrated that LSCs have unique metabolic properties compared to AML blasts (Jones et al., 2021). However, the metabolite profiles of human LSCs and hematopoietic stem and progenitor cells (HSPCs) have not yet been directly compared. To examine metabolite levels in enriched LSCs and HSPCs we used mass spectrometry-based metabolomics analysis on enriched LSCs from 18 AML patients and HSPCs from 9 normal bone marrow (BM) specimens. Pathway analysis showed that arginine metabolism and biosynthesis were the most significantly enriched pathways in LSCs compared to HSPCs. Clinical studies have shown limited efficacy of targeting arginine levels using arginine degrading enzymes; therefore, we sought to target pathways downstream of arginine. Stable isotope labeled tracing analysis showed arginine was metabolized through the urea cycle into polyamines in human LSCs. Polyamines are cationic compounds which have been previously shown to be important for the growth of various cancers but their role in AML LSCs has yet to be determined. To investigate if polyamines are essential for LSC survival and function, we treated LSCs and HSPCs with N1,N11-diethylnorspermine (DENSpm), a polyamine analog causing polyamine depletion through increased cellular export. DENSpm treatment decreased the viability of LSCs and AML blasts in 10 out of 12 of patients tested but had no impact on HSPC viability. Further, treatment with DENSpm resulted in decreased engraftment potential of the 4 AML specimens examined in immune deficient mice but did not alter the engraftment of normal HSPCs, suggesting that targeting polyamines represents a promising approach to kill LSCs while sparing HSPCs. We then investigated the mechanisms regulated by polyamines in LSCs. Using RNA-sequencing analysis on enriched LSCs from 5 AML patients and HSPCs from 3 normal BM specimens we observed decreased expression of genes involved in protein synthesis in DENSpm-treated LSCs but not normal HSPCs. Protein synthesis and its regulation are critical for AML survival (Chen et al., 2019; Messling et al., 2022) but the role of polyamines in protein translation in LSCs has not been explored. Polyamines stimulate translation notably by serving as precursors for the hypusination of eukaryotic translation initiation factor 5A (eIF5A) (Park et al., 2010). Accordingly, DENSpm led to reduced protein synthesis level in MOLM13 AML cell line and primary AML samples. Strikingly, protein synthesis and viability were rescued upon co-treatment with spermidine, suggesting that DENSpm induces AML cells death at least in part through decreased eIF5A-mediated protein translation. Altogether, our data demonstrates that depletion of polyamines decreased LSC function while sparing normal HSPCs. Mechanistically, our data suggest that LSC targeting is accomplished through polyamine regulated protein synthesis. Citation Format: Vincent Rondeau, Cristiana O’Brien, Jacob M Berman, Duhan Yendi, Tianyi Ling, Aleksandra Bourdine, Aarushi Gupta, Jonathan St-Germain, Rachel Culp-Hill, Julie A Reisz, Soheil Jahangiri-Tazehkand, Anastasia Tikhonova, Andrea Arruda, Mark D Minden, Angelo D’Alessandro, Brian Raught, Courtney L Jones. Polyamines are required for AML LSC function through their role in regulating eIF5A dependent protein synthesis [abstract]. In: Proceedings of the AACR Special Conference: Acute Myeloid Leukemia and Myelodysplastic Syndrome; 2023 Jan 23-25; Austin, TX. Philadelphia (PA): AACR; Blood Cancer Discov 2023;4(3_Suppl):Abstract nr A54.

Abstract 3155: Gene expression profiles reveal distinct regulatory activities of transcription factors GATA1 and TAL1 upon AML relapse

Abstract The purpose of this study is to identify key regulatory pathways that potentially drive abnormal gene expression program in relapsed Acute Myeloid Leukemia (AML) patients, by integrative computational analyses on multi-omics molecular profiles. Relapsed AML remains a clinical challenge. Epigenetic heterogeneity may contribute to transcriptional dysregulation and disease progression in AML. However, what specific transcriptional programs and potential regulatory mechanisms contribute to disease relapse are not yet well understood. To characterize the global transcriptional landscapes in relapsed AML, we integrated genomics data from two cohorts of matched diagnosis and relapse patient specimens. We identified 5,416 differentially expressed genes (DEGs) between diagnosis and relapse in Cohort I. Unsupervised clustering yielded three distinct DEG groups: group A, B and C genes that were predominantly (88%) down-regulated, divergently regulated, or predominantly (65%) up-regulated, respectively, upon relapse. The expression pattern of all DEGs separated the patients into two clusters, most robustly by Group B genes. Interestingly, the majority of DEGs did not associate with changes in gene promoter methylation. Similar patterns were observed in Cohort II. We used Binding Analysis for Regulation of Transcription (BART) to identify transcriptional regulators (TRs) that potentially regulated the DEGs not associated with DNA methylation changes, and assessed the differential expression of identified TRs during disease progression. PU.1 was identified as a potential TR for Group A genes and was down-regulated upon relapse. GATA1 and TAL1 were identified as regulating Group B genes and were up-regulated in patient cluster1 and down-regulated in cluster2, consistent with the expression pattern of Group B genes. RBBP5 was a top predicted TR for Group C genes and was up-regulated upon relapse. We next validated the potential functionality of those predicted factors. In NSG mice transplanted with a human AML specimen, TAL1 and GATA1 were downregulated in AML cells collected four weeks after chemotherapy treatment, and were inferred as TRs for the down-regulated genes, similar to the patient data. PU.1 was inferred as regulating the up-regulated genes. Furthermore, we found that the level of differential expression of TAL1, GATA1, and PU.1 in each patient specimen associated with the correlation of DEG profiles between the patient specimen and TR perturbation in human-derived hematopoietic cell lines. Our results support the possibility that in some AML patients, TRs with roles in hematopoiesis and leukemia might contribute to disease relapse. Further mechanistic studies deciphering the molecular and phenotypic events facilitated by these TRs will yield significant insight into disease biology and possible therapeutic targeting approaches in relapsed AML. Citation Format: Zhenjia Wang, Yaseswini Neelamraju, Cem Meydan, Nicholas Dunham, Jorge Gandara, Tak Lee, Subhash Prajapati, Franck Rapaport, Caroline Sheridan, Paul Zumbo, Michael Becker, Lars Bullinger, Martin Carroll, Richard D’Andrea, Richard Dillon, Ross Levine, Christopher E. Mason, Ari Melnick, Donna Neuberg, Stefan Bekiranov, Chongzhi Zang, Francine E. Garrett-Bakelman. Gene expression profiles reveal distinct regulatory activities of transcription factors GATA1 and TAL1 upon AML relapse [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3155.

Long Noncoding RNA Expression Independently Predicts Outcome in Pediatric Acute Myeloid Leukemia.

Optimized strategies for risk classification are essential to tailor therapy for patients with biologically distinctive disease. Risk classification in pediatric acute myeloid leukemia (pAML) relies on detection of translocations and gene mutations. Long noncoding RNA (lncRNA) transcripts have been shown to associate with and mediate malignant phenotypes in acute myeloid leukemia (AML) but have not been comprehensively evaluated in pAML. To identify lncRNA transcripts associated with outcomes, we evaluated the annotated lncRNA landscape by transcript sequencing of 1,298 pediatric and 96 adult AML specimens. Upregulated lncRNAs identified in the pAML training set were used to establish a regularized Cox regression model of event-free survival (EFS), yielding a 37 lncRNA signature (lncScore). Discretized lncScores were correlated with initial and postinduction treatment outcomes using Cox proportional hazards models in validation sets. Predictive model performance was compared with standard stratification methods by concordance analysis. Training set cases with positive lncScores had 5-year EFS and overall survival rates of 26.7% and 42.7%, respectively, compared with 56.9% and 76.3% with negative lncScores (hazard ratio, 2.48 and 3.16; P < .001). Pediatric validation cohorts and an adult AML group yielded comparable results in magnitude and significance. lncScore remained independently prognostic in multivariable models, including key factors used in preinduction and postinduction risk stratification. Subgroup analysis suggested that lncScores provide additional outcome information in heterogeneous subgroups currently classified as indeterminate risk. Concordance analysis showed that lncScore adds to overall classification accuracy with at least comparable predictive performance to current stratification methods that rely on multiple assays. Inclusion of the lncScore enhances predictive power of traditional cytogenetic and mutation-defined stratification in pAML with potential, as a single assay, to replace these complex stratification schemes with comparable predictive accuracy.

Identification of Novel Antigens for Normal Karyotype Triple Mutated Acute Myeloid Leukemia

Introduction Acute myeloid leukemia (AML) is the most common form of acute leukemia in adults. AML is divided in several subgroups based on genetic alterations and prognosis varies among AML subgroups. Normal karyotype AML with NPM1, DNMT3A and FLT3(ITD) mutations (NK triple mutated AML) is associated with a long-term overall survival of ~20% and represents ~6% of AML cases (Papaemmanuil et al., NEJM, 2016). No curative chemotherapy for this subgroup of AML patients currently exists, highlighting the need to develop novel therapeutic strategies for this group of AML patients. Methods To identify novel cell surface antigens for NK triple mutated AML that could ultimately be targeted by immunotherapeutic approaches, we developed a method to map the surface proteome of primary human AML specimens based on biotinylation of cell surface proteins, affinity purification and mass spectrometry. To further enrich mass spectrometry datasets with surface proteins, we developed an algorithm that assesses the probability of a protein to be located on the cell surface based on annotations from various databases. Using this approach, the surface proteome of 100 primary human AML specimens was analysed, which included 12 NK triple mutated AML samples. Differential surfaceome analyses identified several surface proteins specifically expressed by NK triple mutated AML specimens. Expression analyses were performed using publicly available datasets to select proteins with limited expression in normal tissues and hematopoietic cells. Surface expression and AML subgroup specificity for these potential antigens was evaluated by flow cytometry using primary human AML specimens from various AML subgroups (n=76). Moreover, to assess the intra-specimen expression profile of these antigens, single cell RNA-sequencing of 15 primary human NK triple mutated AML specimens was performed. Finally, we evaluated the prognostic impact of antigen expression on overall survival in a subgroup of patients from the Leucegene cohort composed of patients diagnosed with de novo AML with intermediate risk cytogenetics who have been treated with intensive induction chemotherapy (n = 316). Results Surface proteome analysis of NK triple mutated AML specimens and hit filtering according to selection pipeline described above identified several promising NK triple mutated AML antigens. Surface expression analysis of these antigens by flow cytometry showed specificity of antigen expression for NK triple mutated AML specimens. Transcriptomic analyses using RNA sequencing data from primary human AML specimens of the Leucegene cohort (n=691) showed significantly higher antigen transcript levels in NK triple mutated AML specimens compared to samples from other AML subgroups. Moreover, multivariate transcriptome analysis comparing the impact of most frequent mutations in AML on antigen expression revealed that FLT3-ITD mutation is the only significant factor associated with antigen overexpression in AML samples. Single cell RNA sequencing data from NK triple mutated AML specimens revealed that these antigens are uniformly expressed across the vast majority of blast populations, including blast cells with an expression profile reminiscent of leukemic stem cells. Interestingly, in patients with intermediate cytogenetic risk AML, high antigen expression was significantly associated with worse overall survival. Patients with high antigen expression who proceeded to allogeneic hematopoietic stem cell transplantation (HSCT) in first complete remission also had a significantly lower 5-year overall survival rate post-HSCT compared to patients with low antigen expression. Conclusion We identified novel antigens for FLT3-ITD mutated and NK triple mutated AML, which show homogeneous expression in blast populations of AML samples including most primitive ones, making it ideal antigens for immunotherapeutic targeting. Patients with intermediate cytogenetic risk AML and high antigen expression show poor prognosis even after HSCT in first complete remission, highlighting the need to develop novel therapeutic strategies for these patients. Different immunotherapeutic approaches targeting these antigens are currently being developed by our group, with the hope of identifying novel therapeutic strategies for the treatment of NK triple mutated AML.

Identification of Novel Cell Surface Therapeutic Targets for KMT2A-Rearranged Acute Myeloid Leukemia

INTRODUCTION. Acute myeloid leukemia (AML) is a highly heterogeneous disease that represents 80% of adult leukemias and is characterized by chromosomal rearrangements such as translocations and inversions, as well as genetic mutations. These genomic alterations affect the normal maturation process of myeloid precursor cells and lead to their clonal expansion in the bone marrow and peripheral blood. Genetic aberrations define AML subgroups, which are associated with risk stratification (favorable, intermediate or adverse risk) and prognosis. One AML subgroup, KMT2A-rearranged AML (KMT2Ar), is driven by chromosomal translocations involving the lysine methyltransferase 2A gene (KMT2A) located on chromosome band 11q23.3, leading to gene fusions with over 94 distinct partners. KMT2A rearrangements are present in 5 to 10% of adult AML. Patients with t(9;11)(p21.3;q23.3) translocation generating the MLLT3::KMT2A fusion are classified in the intermediate cytogenetic risk group according to the 2022 ELN genetic risk classification, but KMT2A fusions with other translocation partners are included in the adverse risk group with current therapies. Ongoing clinical trials for KMT2Ar AML focus mainly on chemical inhibitors (DOT1L and Menin inhibitors) that target and disrupt KMT2A fusion protein complexes, with encouraging results. Precision medicine, such as antibody-based immunotherapy, also represents a promising approach for the treatment of KMT2Ar AML, but is currently underexploited due to the limited availability of targetable surface antigens. To overcome this issue, we analyzed the cell surface proteome (surfaceome) of primary human KMT2Ar AML samples to identify novel cell surface antigens with high therapeutic potential. METHODS AND RESULTS. We used a proteomics-based approach to analyze the surfaceome of 100 primary human AML specimens from 13 different AML subgroups, including 13 KMT2Ar AML samples. Briefly, we combined cell surface biotinylation with streptavidin affinity purification to enrich cell extracts with cell surface proteins, which were then analyzed by LC-MS/MS to obtain an overview of primary AML sample surfaceomes. Differential surfaceome analyses were performed to identify surface proteins selectively expressed by KMT2Ar AML specimens, and proteins with low expression in normal hematopoietic cells and normal tissues, as determined by expression analyses using publicly available datasets, were further selected. Using this approach, we identified several potential KMT2Ar AML surface antigens, including CD93, reported to be expressed by KMT2Ar AML stem cells. We then analyzed surface expression of these antigens by flow cytometry using a collection of 112 primary human AML specimens including 10 KMT2Ar AML samples. Results indicate that in the majority of KMT2Ar AML samples analyzed, more than 90% of blasts express selected antigens, suggesting that targeting these antigens with immunotherapeutic approaches could potentially lead to the elimination of most AML cells. Furthermore, flow cytometry data showed specificity of antigen expression for KMT2Ar AML samples, opening new therapeutic opportunities for this AML subgroup. Finally, single cell RNA sequencing of 8 primary KMT2Ar AML specimens was performed to further analyze the expression profile of these antigens in patient samples. AML specimens harboring the MLLT3::KMT2A fusion expressed these antigens in the vast majority of blast cells, confirming the homogeneous expression profile of identified antigens and the high interest to target these antigens for KMT2Ar AML. CONCLUSIONS Using a cell surface proteomics-based approach for the analysis of KMT2Ar AML surfaceome, we identified several antigens for this AML subgroup with high therapeutic potential. These antigens are very specific for KMT2Ar AML and are expressed homogeneously in primary human KMT2Ar AML samples, suggesting that targeting these antigens represents a promising therapeutic strategy for these patients. Several immunotherapeutic approaches to target these antigens are currently being developed and tested by our group, with the hope of identifying novel therapeutic strategies for the treatment of KMT2Ar AML.

Intracellular Calcium Localization Mediates the Activity of Venetoclax in Targeting Acute Myeloid Leukemia Stem Cells

Venetoclax (BCL2 inhibitor) based regimens have significantly improved outcomes for elderly patients with acute myeloid leukemia (AML). While most patients initially respond to venetoclax-based regimens, upfront resistance or relapse on therapy occurs in many. Therefore, elucidating the mechanistic link between BCL2 inhibition and AML cell death is crucial to finding novel therapeutic strategies to combat venetoclax resistance and to discover predictors of venetoclax response. Previous studies have shown BCL2 has a non-canonical function in modulating oxidative phosphorylation (OXPHOS) activity of AML stem cells (LSCs). However, the mechanistic link between BCL2 and OXPHOS has not yet been established. BCL2 has been shown to regulate intracellular calcium localization by modulating the activity of calcium channels at the ER and mitochondria. Optimal mitochondrial calcium concentrations are crucial for cell metabolism as calcium overload or decrease can inhibit OXPHOS activity. In preliminary studies, we established venetoclax increases mitochondrial calcium levels (Fig. 1A) leading to a subsequent decrease in OXPHOS activity and changes in the activity of key calcium dependent TCA cycle enzymes in primary human AML LSCs. Genetic inhibition of BCL2 in LSCs phenocopied the effects of venetoclax treatment. Based on these data, we hypothesized that BCL2 exerts its effect on LSC metabolism by modulating calcium localization between the ER and mitochondria. Furthermore, we sought to explore whether dysregulation of calcium localization by venetoclax is a mechanistic component of drug activity that leads to suppression of OXPHOS. To determine how BCL2 may be influencing mitochondrial calcium levels, we assessed the relationship between BCL2 and SERCA3, a key ion channel responsible for calcium uptake into the ER. Venetoclax disrupted the physical interaction between BCL2 and SERCA3. Mechanistically, venetoclax perturbed this interaction by decreasing SERCA3 protein levels. These findings were recapitulated by genetic inhibition of BCL2 in LSCs. Functional analysis of SERCA3 via siRNA-mediated knock-down in primary AML specimens demonstrated increased mitochondrial calcium levels, decreased OXPHOS activity and decreased engraftment of immune deficient NSG-S mice. Based on these data, we propose a model in which venetoclax disrupts SERCA function leading to less calcium uptake at the ER and more calcium sequestered into the mitochondria leading to overload, decreased OXPHOS activity and decreased LSC function. Next, we evaluated how calcium pathways were impacted in venetoclax resistant AML patient specimens. Intriguingly, venetoclax did not affect mitochondrial calcium levels (Fig. 1B) or the relationship between BCL2 and SERCA3. Therefore, we tested whether inhibiting the activity of MCU, the channel responsible for calcium uptake into the mitochondrial matrix, would target venetoclax resistant LSCs as it was overexpressed in this population. MCU inhibition led to decreased mitochondrial calcium levels, decreased OXPHOS activity and decreased LSC engraftment. These data indicate that venetoclax-resistant AML cells are no longer subject to changes in calcium localization induced by drug treatment but do retain an overall reliance on calcium regulation for LSC function/survival. Finally, we assessed whether our findings could be leveraged to predict venetoclax sensitivity. Analysis of single cell RNA-seq data from venetoclax sensitive versus resistant AML patient specimens (n=19) revealed multiple calcium pathway gene sets were enriched in venetoclax resistant LSCs. Additionally, levels of certain genes including SERCA3 and MCU were differentially expressed in primitive versus monocytic clusters between sensitive and resistant specimens. Taken together, our data demonstrate an unexpected and critical link between BCL2 and calcium signaling in human LSCs. A surprising aspect of our work is that venetoclax, a BH3 mimetic, acts to modulate the interaction with SERCA3. Lastly, our findings suggest that targeting intracellular calcium dynamics may represent a novel therapeutic strategy in the context of venetoclax resistant AML. In addition, we show that there are significant differences in calcium gene signatures between venetoclax sensitive vs resistant AML specimens that may be useful in predicting patient response to venetoclax based regimens. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Relocalization of PU.1 Underlies the On- and Off-Tumor Effects of SWI/SNF Blockade in AML

SWI/SNF ATP-dependent chromatin remodeling activity has been identified as a strong dependency of acute myeloid leukemia (AML). In AML cell lines, genetic knockdown or chemical inhibition of the SWI/SNF ATPases SMARCA4 (BRG1) or SMARCA2 (BRM) by the selective small-molecule inhibitor BRM014 reduces proliferation and induces leukemic differentiation. These effects arise because SWI/SNF inhibition reduces DNA accessibility at binding sites of PU.1, a factor essential for driving high levels of MYC in leukemic cells. However, given that both SWI/SNF and PU.1 are essential in hematopoietic development, important questions regarding the use of SWI/SNF inhibitors against AML remain. First, the chromatin features that influence SWI/SNF-dependent PU.1 regulation within healthy hematologic and leukemic cells have not been established. Furthermore, the ability of SWI/SNF inhibitors to target leukemic cells within the bone marrow has remained uncertain, and the effects of these drugs in an immunocompetent setting have not been reported, despite the essential functions of SWI/SNF and PU.1 in normal hematopoiesis. Finally, the sensitivity of primary human AML samples and normal hematopoietic cells has also not been established. As a result, many questions regarding the nature of the therapeutic window for SWI/SNF inhibition against AML have remained unresolved. Here we evaluate SWI/SNF inhibition using BRM014 in clinically relevant models and examine the consequences of the deregulation of SWI/SNF-dependent PU.1 sites. We demonstrate that inhibition of SWI/SNF shifts PU.1 occupancy from enhancers to promoters in both leukemic and healthy hematopoietic cells. In AML, the shift of PU.1 away from enhancers to the promoters of differentiation-related genes results in loss of MYC expression and differentiation of leukemic cells. In non-leukemic mice, SWI/SNF inhibition suppresses PU.1-dependent B cell development and bone marrow-derived monocyte populations, reflecting off-tumor effects caused by SWI/SNF-dependent PU.1 relocalization. These effects influence peripheral immune cells, as well lineage-committed hematopoietic progenitors within the bone marrow. Moreover, we observe profound therapeutic response to SWI/SNF inhibition in an immunocompetent AML mouse model (Figure 1). In vivo, BRM014 induces differentiation of peripheral leukemic blasts and reduces the leukemic stem cell burden in bone marrow but also induces lymphocytopenia due to its off-tumor effects on PU.1. Nevertheless, the robust regression of leukemic burden seen in vivo over a short two-week treatment period provides a compelling justification for continued study of SWI/SNF inhibitors in the treatment of AML. Further, response of primary human CD34+ hematopoietic stem and progenitor cells (HSPCs) and AML specimens highlight that normal hematopoiesis is affected at concentrations similar to those needed for therapeutic response in AML (Figure 2). Hence, hematopoietic and immune-related side effects, particularly at higher doses and over the duration of treatment, will be important considerations for clinical use of SWI/SNF inhibitors against AML and other tumors. Overall, our results reveal a variable therapeutic window for SWI/SNF blockade in AML and highlight important off-tumor effects of such therapies in immunocompetent settings. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Simultaneous Inhibition of SIRT3 and Cholesterol Homeostasis Targets AML Stem Cells By Perturbing Fatty Acid β-Oxidation and Inducing Lipotoxicity

Outcomes for patients with acute myeloid leukemia (AML) remain poor due to the inability of current therapies to fully eradicate disease initiating leukemia stem cells (LSCs) (Shlush et al., 2017). Oxidative phosphorylation (OXPHOS) is a highly targetable and essential process for LSC energy production and survival (Lagadinou et al., 2013). Sirtuin 3 (SIRT3), a mitochondrial deacetylase with a multi-faceted role in metabolic regulation (Finley et al., 2016), has been specifically shown to regulate OXPHOS in cancer models (Li et al., 2019, Xu et al., 2019). Thus, we sought to identify if SIRT3 is essential for LSC function. To determine if SIRT3 is essential for LSCs, we measured cell viability, colony forming potential, and engraftment potential of primary AML specimens upon perturbation of SIRT3 function using both RNAi and a small molecule inhibitor (YC8-02) (Li et al., 2019). SIRT3 reduction significantly decreased viability, colony forming potential and engraftment of LSCs (Fig. 1A), suggesting that SIRT3 inhibition decreases LSC function. Importantly, neither SIRT3 knockdown nor inhibition impacted the viability, colony forming potential, or engraftment potential of normal hematopoietic stem and progenitor cells (HSPCs). Altogether, these data suggest that SIRT3 is essential for LSCs and not HSPCs. To determine the molecular mechanisms by which SIRT3 is essential for LSC function, we performed proximity-dependent biotin labeling (BioID) in 293 T-Rex cells to map the interactome of SIRT3. Strikingly, our BioID analysis identified several fatty acid oxidation (FAO) enzymes as proximity interactors including CPT2, ACADVL, ACADM, ECHI, HADH, and ACAA2. To corroborate these findings in LSCs, we performed RNA-sequencing upon SIRT3 inhibition in primary LSCs derived from three primary AML specimens. Gene set enrichment analysis revealed a significant decrease in OXPHOS and FAO gene signatures upon SIRT3 inhibition. Importantly, FAO is known to support OXPHOS which suggests that SIRT3 may regulate OXPHOS through mediation of FAO. To determine if protein interactions and gene expression changes result in functional changes in mitochondrial biology, we measured OXPHOS and mitochondrial ATP production in LSCs upon SIRT3 perturbation in LSCs using a seahorse assay. This analysis revealed that SIRT3 perturbation resulted in decreased OXPHOS and ATP production. To determine if SIRT3 regulates OXPHOS via FAO mediation, we performed mass spectrometry-based lipidomic and metabolic interrogation on LSCs from three AML samples treated with YC8-02, which showed an accumulation of fatty acid upon SIRT3 inhibition. To interrogate fatty acid metabolism further, we performed stable isotope tracing analysis using 13C16-palmitate on LSC enriched from three patient specimens upon SIRT3 inhibition. Tracing studies revealed a reduction in 13C enrichment in tricarboxylic acid (TCA) cycle metabolites in the YC8-02-treated LSCs (Fig. 1B), suggesting an impairment to fatty acid utilization upon SIRT3 inhibition. Decreasing availability of TCA products impairs OXPHOS and ultimately causes cell death. Importantly, viability of LSCs was partially rescued through addition of the TCA intermediate α-ketoglutarate, demonstrating reduced FAO to be an essential part of the mechanism by which SIRT3 targets LSCs. In addition to decreasing OXPHOS, the accumulation of fatty acids should induce cell death by causing lipotoxicity. However, phenocopy experiments revealed that LSCs are protected from lipotoxicity unlike AML blasts or AML cell lines. Lipidomic analysis suggested that fatty acids may be converted to cholesterol esters, which can then be stored or excreted from cells, specifically in LSCs and not in AML cell lines. Therefore, we postulated that targeting cholesterol homeostasis using dipyridamole would enhance the effect of SIRT3 inhibition (Esquejo et al., 2021). The combination treatment significantly reduced the viability of LSCs compared to individual agents, indicating that co-inhibition of SIRT3 and cholesterol homeostasis enhances the toxic effects of FAO inhibition on LSCs. In all, we identify SIRT3 as a key regulator of LSC survival, through the mediation of lipid metabolism. We further demonstrate that LSCs are tolerant to FAO inhibition which can be overcome by concurrent inhibition of FAO and cholesterol metabolism, which has the potential to lead to improved treatment response. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

The AP-1/ETS Transcription Factor Network Regulates Stem Cell Quiescence and Chemotherapy Resistance in Acute Myeloid Leukemia

Despite of the continuous improvements in the treatment for acute myeloid leukemia (AML), therapy resistance and disease relapse have been persistent and attributed to leukemia stem cells (LSCs). Recent single cell RNA sequencing (Galen et al, Bernstein 2019) and deconvoluted bulk RNA sequencing analysis (Zeng et al, Dick 2022) show hierarchical cellular organization in individual AML patient specimens, with LSCs, and quiescent LSCs in particular, at its apex. However, the mechanisms controlling LSCs and their quiescence remain elusive, hindering the development of effective targeted therapies. Here, using human patient AML specimens, we show that activator protein 1 (AP-1) and E26 transformation-specific (ETS) family transcription factors (TFs) coordinately regulate LSC quiescence and disease persistence after chemotherapy. A chemical label tracing technique using carboxyfluorescein succinimidyl ester (CFSE) has emerged as a powerful tool to identify quiescent LSCs, and using this technique, we and others have demonstrated that CFSE-label retaining quiescent LSCs isolated from patient leukemia cells exhibit potent leukemia initiating property, resistance to conventional chemotherapy, and epigenetic plasticity (Takao et al, Kentsis 2018 ASH, Ebinger et al, Jeremias 2016 and 2020). Here, we combined this method with functional genomic approaches to elucidate the mechanisms controlling quiescent LSCs and identify novel therapeutic targets to overcome chemotherapy resistance. First, to identify the candidate genes for LSC quiescence regulators, we performed gene expression and chromatin accessibility profiling of label-retaining quiescent LSCs isolated from human patient leukemia cell-engrafted NSG mice in diverse subtypes of AMLs using mRNA sequencing (RNA-seq) and assay for transposase-accessible chromatin with high-throughput sequencing (ATAC-seq). We identified that several transcription factors important for normal hematopoiesis, which included AP-1 TFs, such as JUN and JUNB, and ETS TFs, such as ETS1 and FLI1, were differentially expressed in quiescent LSCs. Chromatin accessibility profiling further revealed the predominant occupancy of AP-1 and ETS TF family binding motifs in differentially accessible chromatin loci in quiescent LSCs. Importantly, gene expression signatures detected in quiescent LSCs showed significant similarity to gene expression profiles of residual leukemia cells observed in chemotherapy-treated AML patients (GSE116256, Galen et al. 2019), supporting the clinical relevance of our models. Next, to develop a functional screen system to identify the regulators of quiescent LSCs in human patient AML-bearing mice, we designed a lentiviral vector encoding mCherry fluorescent protein and doxycycline-inducible cDNAs marked by specific barcode sequences. cDNA library screening using this lentiviral vector identified several TFs as regulators for AML LSC quiescence, which included JUN as a positive regulator, and ETS1 and FLI1 as negative regulators (Figure 1). Finally, given the evidence of transcriptional activity of JUN to regulate LSC quiescence, we hypothesized that JUN activity may be also associated with chemotherapy resistance and disease persistence in patients. To investigate this, we analyzed JUN expression in individual AML cells in diverse patients before and after chemotherapy treatment (GSE116256, Galen et al. 2019). We found JUN to be highly expressed in persistent AML cells upon induction chemotherapy, especially in leukemia cells with features of hematopoietic stem cells (HSC-like), while ETS1 and FLI1 expression was not (Figure 2). These results argue that JUN activity may be required for chemotherapy resistance, LSC persistence, and ultimate AML relapse. In all, human AML LSC quiescence and therapy response can be dynamically regulated by a distinct transcription factor network. This work defines molecular mechanisms controlling AML LSC quiescence. We anticipate that these findings will lead to the elucidation of essential properties of LSC quiescence and the design of therapeutic strategies for their clinical identification and control. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Disruption of Intracellular Metal Homeostasis As a Vulnerability in SF3B1 Mutated AML

Acute Myeloid Leukemia (AML) is a heterogeneous disease in which outcome is highly dependent on cytogenetics and mutational profile. Adverse risk AML, recently defined and updated in the 2022 European LeukemiaNet (ELN) genetic risk classification, is associated with a poor overall survival. To identify novel therapeutic approaches for AML, we screened a library of 10,000 compounds across 56 genetically diverse primary AML specimens. We then looked for active compounds with increased activity towards adverse risk AML samples and with low cytotoxicity to normal hematopoietic cells. Among the hits identified, we were able to validate that S767 molecule is preferentially active on adverse risk AML samples. Additionally, we found a strong association between SF3B1 mutations and sensitivity to S767. To gain insights into S767 mechanism of action, we performed a CRISPR/Cas9 genome-wide functional screen, and identified multiples genes involved in mitochondrial processes both in synthetic rescue and synthetic lethal genes. In particular, knock-down of enzymes involved in heme biogenesis and iron-sulfur cluster biosynthesis appear to interfere with S767 cytotoxic activity. Given the central role of iron in these two synthetic pathways, and the structure of S767, we hypothesized that our molecule binds and disrupts intracellular iron homeostasis. Using UV-spectrometry analysis, we determined that S767 has metal affinity and binds to Fe2+, but also Cu2+ and Zn2+. We then quantified intracellular metal concentrations by ICP-MS (inductively coupled plasma mass spectrometry) and found that S767 exposure significantly decreases intracellular iron concentrations while increasing that of copper. Since metals are critical cellular cofactors involved in multiple enzymatic reactions, disruption of intracellular metal homeostasis can affect many processes. Mitochondrial respiratory chain is particularly sensitive to these perturbations as copper and iron-sulfur clusters are present in complexes I-IV of the electron transport chain and are essential for mitochondrial complexes function. By measuring oxygen consumption rate in leukemic cells exposed to S767, we found that the molecule indeed disrupts mitochondrial respiration. SF3B1 is a splicing factor recurrently mutated in myeloid malignancies including myelodysplastic syndromes (MDS) and AML. Heterozygous mutations typically affect SF3B1 splicing functions and induces consequently mis-splicing of target mRNAs, including the ABCB7 iron-sulfur cluster transporter. Analysis of the Leucegene transcriptomic data of primary AML specimens revealed mis-splicing and downregulation of ABCB7 in SF3B1 mutated patients. Strikingly, ABCB7 was one of the top synthetic lethal gene identified by chemogenomic CRISPR/Cas9 screen following exposure to S767 molecule. Given that ABCB7 exports iron-sulfur clusters out of the mitochondria to be assembled by cytosolic proteins and plays a role in iron homeostasis, we hypothesized that its aberrant expression in SF3B1 mutated AML participates to the high sensitivity of these leukemias to S767. We thus identified a novel molecule with selective activity in AML. Our data suggest that S767 is not a protein-targeting molecule but rather acts through the disruption of intracellular metal homeostasis, and ultimately affects mitochondrial respiration. Its mechanism of action highlighted a vulnerability in SF3B1 mutated AML which present a defect in ABCB7 expression. Importantly, SF3B1 mutations are now recognized in the 2022 ELN risk classification as an adverse prognostic marker if they are associated with intermediate or adverse risk AML, giving a unique opportunity to develop novel therapeutic approaches for this subgroup of patients with poor clinical outcome. In addition, SF3B1 mutations are found in 25% of MDS cases, which are associated with ring sideroblasts and accumulation of mitochondrial iron, and that could also benefit from such therapeutic approach.

LncRNA MAFG-AS1-induced acute myeloid leukemia development via modulating miR-147b/HOXA9.

Recent references discovered that lncRNAs acted roles in malignant cancer development. However, the role of MAFG-AS1 in acute myeloid leukemia (AML) development remains unknown. MAFG-AS1 and miR-147b were determined in AML cells and specimens using qRT-PCR assay. Cell proliferation was detected by CCK-8 analysis and flow cytometry was carried out to measure cell cycle. Luciferase reporter analysis was done to determine the mechanism of MAFG-AS1 and miR-147b. We noted that MAFG-AS1 was overexpressed in AML cells and in serum and bone narrow from AML compared with normal controls specimen. Elevated expression of MAFG-AS1 increased cell growth, cycle and EMT in AML cell HL-60 cell. MAFG-AS1 sponged miR-147b expression in HL-60 cell. Moreover, miR-147b was downregulated in AML cells and in serum and bone narrow from AML compared with normal control specimen. miR-147b was negatively correlated with MAFG-AS1 in the serum and bone narrow of AML cases. We illustrated that HOXA9 was one target of miR-147b and ectopic expression of MAFG-AS1 enhanced HOXA9 expression HL-60 cell. Forced expression of MAFG-AS1 induced cell growth, cycle, and EMT via promoting HOXA9. These data illustrated that MAFG-AS1 acted as one oncogenic gene and accelerated AML progression via modulating miR-147b/HOXA9 axis.

HMGA2 expression defines a subset of human AML with immature transcriptional signature and vulnerability to G2/M inhibition

AbstractHigh-mobility group AT-hook 2 (HMGA2) is a nonhistone chromatin-binding protein that is normally expressed in stem cells of various tissues and aberrantly detected in several tumor types. We recently observed that one-fourth of human acute myeloid leukemia (AML) specimens express HMGA2, which associates with a very poor prognosis. We present results indicating that HMGA2+ AMLs share a distinct transcriptional signature representing an immature phenotype. Using single-cell analyses, we showed that HMGA2 is expressed in CD34+ subsets of stem cells and early progenitors, whether normal or derived from AML specimens. Of interest, we found that one of the strongest gene expression signatures associated with HMGA2 in AML is the upregulation of G2/M checkpoint genes. Whole-genome CRISPR/Cas9 screening in HMGA2 overexpressing cells further revealed a synthetic lethal interaction with several G2/M checkpoint genes. Accordingly, small molecules that target G2/M proteins were preferentially active in vitro and in vivo on HMGA2+ AML specimens. Together, our findings suggest that HMGA2 is a key functional determinant in AML and is associated with stem cell features, G2/M status, and related drug sensitivity.

P482: NOVEL BISPECIFIC INNATE CELL ENGAGER AFM28 FOR THE TREATMENT OF CD123 POSITIVE ACUTE MYELOID LEUKEMIA AND MYELODYSPLASTIC SYNDROME

Background: Novel treatments for patients with acute myeloid leukemia (AML) and high-risk myelodysplastic syndrome (HR-MDS) with relapsed or refractory (R/R), or measurable residual disease are urgently needed. Targeting of the surface antigen CD123 expressed on both leukemic blasts and stem cells (LSCs) of most patients allows for depletion of both compartments, which is considered a prerequisite for deep anti-tumor responses and induction of prolonged remission. AFM28 is a novel bispecific Innate Cell Engager (ICE®) designed for bivalent high-affinity binding of CD16A on natural killer (NK) cells, redirecting effector cell cytotoxicity to CD123+ tumor cells. Allogeneic NK cell therapy is emerging as a next-generation treatment with demonstrated activity in R/R AML and a very benign safety profile. Retargeting of NK cells with AFM28 may be a particularly effective treatment strategy in patients with CD123+ AML and HR-MDS. Aims: 1. Preclinical characterization of AFM28 pharmacology and toxicology in vitro and in vivo to support clinical development. 2. Assessment of binding and activation of NK cells by AFM28 to test suitability for combination development with allogeneic NK cell therapy. Methods: Flow cytometry (FC) was used to assess binding and retention of AFM28 on NK cells, NK cell activation (degranulation, IFNγ expression) and depletion of leukemic cells from AML and MDS patient specimens. Quantification of secreted cytokines in leukocyte cultures employed bead-based luminex methodology. In vivo efficacy data were acquired by BLI measurement from a murine AML model. Cynomolgus data were acquired from non-GLP dose range finding and a GLP toxicology study using standard methods (ELISA, IHC, FC). Results: AFM28 induced NK cell activation at low picomolar concentrations and mediated efficacious depletion of CD123+ tumor cell lines and primary leukemic cells, including cells with low CD123 expression. NK cell activation was induced more potently than by an effector-enhanced IgG1 and was accompanied by more pronounced degranulation and IFNγ production. Similarly, AFM28 induced depletion of leukemic cells in patient bone marrow samples, without lysis of CD34+/CD123− cells, suggesting sparing of healthy hematopoietic progenitors. Potent anti-tumor activity was confirmed in vivo in a murine tumor model. In preclinical toxicology models, AFM28-induced target cell depletion was associated with minimal release of inflammatory cytokines, including IL-6, TNFα and IFNγ from leukocytes in vitro, and only low-level release of IL-6 in cynomolgus monkeys, despite efficacious depletion of CD123+ basophils and plasmacytoid dendritic cells (pDCs). Characterization of NK cell binding revealed high-avidity interaction, long cell surface retention, and potent induction of NK cell-mediated tumor cell lysis with no impact on NK cell viability. Summary/Conclusion: These data demonstrate that AFM28 exhibits potent pharmacological activity in vitro and in vivo and efficaciously induces NK cell-mediated depletion of primary leukemic cells. Further, toxicology data suggest good tolerability with a low risk of cytokine release syndrome. Together, these findings support the clinical investigation of AFM28 as a potential novel treatment of CD123+ AML and HR-MDS. A first-in-human trial investigating safety and activity of single agent AFM28 in patients with R/R AML is currently in preparation. In addition, these data support the combination of AFM28 with allogeneic NK cell therapy, which holds potential as an effective, novel treatment strategy.