Search

While most patients with acute myeloid leukemia (AML) achieve remission with initial therapy, the majority relapse leading to poor overall survival. Relapse is frequently driven by a rare subset of leukemic stem cells (LSC). Understanding biological mechanisms that maintain LSCs will help identify new therapeutic strategies for this disease. To identify such vulnerabilities, we overlaid the results of a genome-wide CRISPR screen with the expression of genes enriched in functionally defined LSCs. Through our CRISPR screen, we identified 570 genes whose knockout reduced the growth and viability of OCI-AML2 cells. Essential genes for growth and viability by our CRIPSR screen were enriched in the LSC+ population. By overlaying the hits from our CRISPR screen with genes upregulated in LSCs, we identified IPO11, as a top hit, with 7.5-fold increase in the LSC+ fraction compared to the LSC- fraction. IPO11 is a member of the importin-β family of proteins and facilitates the import of protein cargo into the nucleus. Further analysis showed that IPO11 was upregulated in LSC+ (engrafting) vs. LSC- (non-engrafting) primary AML samples, CD34+ vs CD34- AML samples, undifferentiated progenitor vs. myeloid cluster AML samples, and relapse vs de novo AML. IPO11 was increased in AML cells compared to normal hematopoietic cells and increased IPO11 expression was associated with decreased overall survival in AML. By immunoblotting, IPO11 protein was increased in primary AML (n=4) compared to normal hematopoietic cells (n=4). To determine whether IPO11 is necessary for AML growth and viability, we knocked down IPO11 in OCI-AML2, TEX and NB4 leukemia cells with shRNA in lentiviral vectors. Knockdown of IPO11 reduced AML growth and viability by 80-90%. In contrast, knockdown of another importin-β family member, IPO5, that was not a hit in our CRIPSR screen, did not reduce AML growth and viability. Knockdown of IPO11 increased differentiation of AML cells as evidenced by the changes in gene expression, decreased chromatin accessibility, increased CD11b expression and increased non-specific esterase staining. Finally, knockdown of IPO11 reduced the engraftment of TEX cells and the low passage primary AML 8227 cells into immune deficient mice by over 90%. Importantly, IPO11 knockdown reduced engraftment of primary AML cells into mouse marrow. To identify novel cargos of IPO11, we performed proximity-dependent biotin labeling (BioID) coupled with mass spectrometry and identified proteins that interacted with IPO11. Among the top hits were BZW1 and BZW 2 (Basic leucine zipper and W2 domains 1 and 2). BZW1 and BZW2 are members of the bZIP super family of transcription factors. Knockdown of IPO11 reduced levels of BZW1 in the nucleus detected by immunoblotting and confocal microscopy. Commercial antibodies could not detect BZW2. To determine if the nuclear import of BZW1 and 2 were functionally important for the effects of IPO11 on AML stem cell function and differentiation, we knocked down BZW1 and BZW2. Dual knockdown of BZW1 and BZW2 (but not individual) mimicked the effects of IPO11 inhibition and decreased the growth and viability of AML cells. Changes in gene expression after BZW1/2 knockdown were similar to IPO11 knockdown with enrichment in myeloid-differentiated genes. By pathway analysis, we identified that IPO11 knockdown, as well as BZW1/2 knockdown decreased expression of MYC target genes, suggesting a mechanism by which these proteins regulate AML stem cell function. Thus, in summary, we identified IPO11 as an essential gene for the viability of AML cells and stem cells. This work highlights a previously unappreciated role of the protein import pathway in regulating AML stem cell function and highlights a potential new therapeutic target for AML. Disclosures Schimmer: Takeda: Honoraria, Research Funding; Novartis: Honoraria; Jazz: Honoraria; Otsuka: Honoraria; Medivir AB: Research Funding; AbbVie Pharmaceuticals: Other: owns stock .

Neurolysin (NLN) is a zinc metallopeptidase whose mitochondrial function is unclear. We found that NLN was overexpressed in almost half of patients with acute myeloid leukemia (AML), and inhibition of NLN was selectively cytotoxic to AML cells and stem cells while sparing normal hematopoietic cells. Mechanistically, NLN interacted with the mitochondrial respiratory chain. Genetic and chemical inhibition of NLN impaired oxidative metabolism and disrupted the formation of respiratory chain supercomplexes (RCS). Furthermore, NLN interacted with the known RCS regulator, LETM1, and inhibition of NLN disrupted LETM1 complex formation. RCS were increased in patients with AML and positively correlated with NLN expression. These findings demonstrate that inhibiting RCS formation selectively targets AML cells and stem cells and highlights the therapeutic potential of pharmacologically targeting NLN in AML.

A number of unique proteases localize to specific sub-compartments of the mitochondria, but the functions of these enzymes are poorly defined. Here, in vivo proximity-dependent biotinylation (BioID) is used to map the interactomes of seven proteases localized to the mitochondrial intermembrane space (IMS). In total, 802 high confidence proximity interactions with 342 unique proteins are identified. While all seven proteases co-localized with the IMS markers OPA1 and CLPB, 230 of the interacting partners are unique to just one or two protease bait proteins, highlighting the ability of BioID to differentiate unique interactomes within the confined space of the IMS. Notably, high-temperature requirement peptidase 2 (HTRA2) interacts with eight of 13 components of the mitochondrial intermembrane space bridging (MIB) complex, a multiprotein assembly essential for the maintenance of mitochondrial cristae structure. Knockdown of HTRA2 disrupts cristae in HEK 293 and OCI-AML2 cells, and leads to increased intracellular levels of the MIB subunit IMMT. Using a cell-free assay it is demonstrated that HTRA2 can degrade recombinant IMMT but not two other core MIB complex subunits, SAMM50 and CHCHD3. The IMS protease interactome thus represents a rich dataset that can be mined to uncover novel IMS protease biology.

Summary Tafazzin (TAZ) is a mitochondrial transacylase that remodels the mitochondrial cardiolipin into its mature form. Through a CRISPR screen, we identified TAZ as necessary for the growth and viability of acute myeloid leukemia (AML) cells. Genetic inhibition of TAZ reduced stemness and increased differentiation of AML cells both in vitro and in vivo. In contrast, knockdown of TAZ did not impair normal hematopoiesis under basal conditions. Mechanistically, inhibition of TAZ decreased levels of cardiolipin but also altered global levels of intracellular phospholipids, including phosphatidylserine, which controlled AML stemness and differentiation by modulating toll-like receptor (TLR) signaling.

Through a clustered regularly insterspaced short palindromic repeats (CRISPR) screen to identify mitochondrial genes necessary for the growth of acute myeloid leukemia (AML) cells, we identified the mitochondrial outer membrane protein mitochondrial carrier homolog 2 (MTCH2). In AML, knockdown of MTCH2 decreased growth, reduced engraftment potential of stem cells, and induced differentiation. Inhibiting MTCH2 in AML cells increased nuclear pyruvate and pyruvate dehydrogenase (PDH), which induced histone acetylation and subsequently promoted the differentiation of AML cells. Thus, we have defined a new mechanism by which mitochondria and metabolism regulate AML stem cells and gene expression.

The transcription factor c-Maf is extensively involved in the pathophysiology of multiple myeloma (MM), a fatal malignancy of plasma cells. In the present study, affinity chromatography and mass spectrometry were used to identify c-Maf ubiquitination-associated proteins, from which the E3 ligase HERC4 was found to interact with c-Maf and catalyzed its polyubiquitination and subsequent proteasome-mediated degradation. HERC4 mediated polyubiquitination at K85 and K297 in c-Maf, and this polyubiquitination could be prevented by the isopeptidase USP5. Further analysis on the NCI-60 cell line collection revealed that RPMI 8226, a MM-derived cell line, expressed the lowest level of HERC4. Primary bone marrow analysis revealed HERC4 expression was high in normal bone marrow, but was steadily decreased during myelomagenesis. These findings suggested HERC4 played an important role in MM progression. Moreover, ectopic HERC4 expression decreased MM proliferation in vitro, and delayed xenograft tumor growth in vivo. Therefore, modulation of c-Maf ubiquitination by targeting HERC4 may represent a new therapeutic modality for MM.

Bruton’s tyrosine kinase (BTK) is a potential therapeutic target in acute myeloid leukemia (AML) and the small molecule BTK inhibitor ibrutinib has been found to potentiate the cytotoxic activity o...

While most patients with AML achieve remission with standard induction chemotherapy, the majority ultimately relapse. Relapsed AML is due, at least in part, to the persistence of chemoresistant leukemia stem cells (LSCs). The mechanisms of chemoresistance in LSCs are not fully understood. Here, we explored DNA damage repair in LSCs. 8227 cells are low passage primary AML cells that maintain a hierarchical organization with functionally defined stem cells in the CD34+CD38- fraction. We FACS sorted 8227 cells into stem and bulk fractions and measured expression of DNA repair genes. LSCs were primed for DNA repair with increased expression of genes associated with homologous recombination (RAD51, XRCC2, XRCC3) and non-homologous end joining (XRCC4, XRCC5, PRKDC). Next, we treated the cell fractions with daunorubicin, an intercalating anthracycline that causes double stranded breaks. DNA damage and repair were evaluated by measuring foci of 53BP1, RAD51 and γH2AX by fluorescent microscopy and quantified using image J. Compared to bulk cells, 8227 stem cells demonstrated enhanced DNA damage repair with increased foci of 53BP1 and RAD51 and decreased γH2AX foci, compared to their basal levels. Similar findings were noted after exposing the stem and bulk cells to radiation. We recently discovered that the metabolic enzyme hexokinase 2 (HK2) localizes to the nucleus to maintain stem cell number and function. Therefore, we selectively over-expressed HK2 in the nucleus of 8227 and NB4 cells by tagging HK2 with a nuclear localizing sequence (PAAKRVKLD). We confirmed selective over-expression of HK2 in the nucleus by immunoblotting and confocal microscopy. Over-expressing HK2 increased stem cell function as shown by clonogenic growth assays and engraftment into mouse marrow. We then treated these cells with daunorubicin and measured DNA damage repair. Over-expression of nuclear HK2 increased 53BP1 and RAD51 foci with decreased γH2AX foci, similar to the phenotype observed in LSCs. In addition, over-expression of nuclear HK2 conferred resistance to daunorubicin as measured by clonogenic growth assays. In summary, LSCs appear to be primed for DNA repair with increased levels of DNA damage repair genes. After exposure to chemotherapy and radiation, LSCs have increased repair of double strand DNA breaks compared to more differentiated blasts. This accelerated DNA damage repair may partly explain the increased chemoresistance seen in LSCs. Disclosures Schimmer: Takeda:Honoraria, Research Funding;Novartis:Honoraria;Jazz:Honoraria;AbbVie Pharmaceuticals:Other: owns stock ;Otsuka:Honoraria;Medivir AB:Research Funding.

Leukemia stem cells are responsible for the initiation and recurrence of AML. Cell differentiation requires the coordination of metabolic state and gene expression programs. It is known that metabolic intermediates can serve as cofactors in epigenetic modifications in the nucleus. Whether metabolic enzymes can directly localize to the nucleus, influence gene expression and regulate stemness in AML is unknown. To identify mitochondrial metabolic enzymes that localize to the nucleus of stem cells we analyzed stem and bulk fractions of 8227 leukemia cells, which are arranged in a hierarchy with functionally defined stem cells in the CD34+CD38- fraction. Hexokinase 2 (HK2) was detected in the nuclear fraction of 8227 cells and nuclear HK2 was increased in 8227 stem cells compared to bulk cells. Conversely, the metabolic enzymes aconitase, citrate synthesase, enolase, fumarase, GAPDH, glucose-6-phosphate isomerase, phosphofructokinase, pyruvate kinase and succinate dehydrogenase A were not detectable in the nuclear fraction of either stem or bulk cells. We confirmed that HK2 was present in the nuclear fraction of OCI-AML2, NB4, HL60 and TEX AML cell lines and 7/9 primary patient samples. To determine whether nuclear HK2 was functionally important for AML stem cells, we overexpressed HK2 tagged with a nuclear localizing sequence (PAAKRVKLD). Over-expression of nuclear localized HK2 enhanced clonogenic growth and engraftment into immune deficient mice. Likewise, over-expression of nuclear localized HK2 blocked differentiation after treatment with all-trans retinoic acid (ATRA). Next, we investigated whether the kinase activity of HK2 was necessary for its nuclear effects on AML stemness and differentiation. We generated a kinase dead mutant of HK2 (D209A, D657A) tagged with a nuclear localizing signal (PAAKRVKLD). Over-expression of nuclear-localized kinase dead HK2 exerted a similar phenotype to overexpressed nuclear HK2 without manipulation of the kinase domain as it enhanced clonogenic growth and blocked cell differentiation after ATRA treatment. To understand the mechanism by which nuclear HK2 influenced AML stemness, we identified proteins that interacted with nuclear HK2 using proximity-dependent biotin labeling (BioID) and mass spectrometry. Proteins involved in DNA damage repair and chromatin remodeling were identified, including exonuclease 3′-5′ domain-containing 2 (EXD2), sirtuin 1 (SIRT1), E3 ubiquitin-protein ligase (URB5) and tyrosyl-DNA phosphodiesterase 2 (TDP2). Thus, our data suggest that nuclear HK2 influences DNA repair, which has been linked to stemness and differentiation. In support of this hypothesis, we identified a PAR binding motif in HK2 and over-expression of nuclear HK2 conferred resistance to the PARP inhibitor, olaparib. In summary, HK2 localizes to the nucleus in AML stem cells and maintains stemness. These effects are independent of its kinase activity. Rather, HK2 may regulate DNA damage repair through a PAR dependent mechanism. Thus, we have identified a new role for metabolic enzymes in the regulation of stemness and differentiation. Citation Format: Grace Egan, Geethu E. Thomas, Parasvi S. Patel, Jordan Chin, Aaron Botham, Rose Hurren, Neil MacLean, Razq Hakem, Aaron D. Schimmer. The metabolic enzyme Hexokinase 2 localizes to the nucleus and regulates stemness in AML through a kinase independent mechanism [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 3784.

TAK-243 (MLN7243) is a first-in-class inhibitor of the ubiquitin-activating enzyme (UBA1) that catalyzes the first step in the ubiquitylation cascade whereby proteins are tagged with mono- or poly-ubiquitin to induce their degradation or modify their functions. Based on its preclinical efficacy and tolerability, TAK-243 has entered phase 1 clinical trials in advanced malignancies. However, the determinants of sensitivity to TAK-243 remain largely unknown. Therefore, we conducted a positive-selection, genome-wide CRISPR/Cas9 knockout screen in OCI-AML2 cells followed by selection with lethal TAK-243 concentrations to identify genes essential for TAK-243 action. We identified BEN domain-containing protein 3 (BEND3), a transcriptional repressor and a regulator of chromatin organization, as the top gene whose knockout conferred resistance to TAK-243 (FDR = 0.0012). BEND3-targeting gRNAs were enriched up to 10,000-fold after selection with the drug. To validate the screen results, we independently knocked out BEND3 in OCI-AML2 cells and confirmed the resistance phenotype. In vivo, tumors of BEND3-knockout cells were resistant to TAK243 (20 mg/kg twice weekly) as opposed to control tumors that showed dramatic reductions in tumor growth rate. As assessed by immunoblotting, BEND3 knockout dampened TAK-243 effects on ubiquitylation, proteotoxic stress and DNA damage response. Mechanistically, BEND3 knockout upregulated the ABC efflux transporter breast cancer resistance protein (BCRP; ABCG2), and decreased intracellular levels of TAK-243. It also conferred partial cross-resistance to pevonedistat and TAK-981–related selective inhibitors of the NEDD8-activating enzyme (NAE) and the SUMO-activating enzyme (SAE), respectively, as well as known substrates of BCRP (mitoxantrone and doxorubicin). Finally, TAK-243 sensitivity strongly correlated with BCRP expression in a panel of 30 cancer cell lines of different origin, and chemical inhibition of BCRP but not P-gp sensitized intrinsically resistant high-BCRP cells to TAK-243. Thus, our data demonstrate that BEND3 regulates the expression of BCRP for which TAK-243 is a substrate. Moreover, BCRP expression could serve as a predictor of TAK-243 sensitivity. Citation Format: Samir H. Barghout, Ahmed Aman, Zachary Blatman, Karen Arevalo, Geethu Thomas, Neil MacLean, Xiaoming Wang, Rose Hurren, Troy Ketela, Moustafa Abohawya, Taira Kiyota, Rima Al-Awar, Aaron D. Schimmer. BEND3 modulates sensitivity to the UBA1 inhibitor TAK-243 by regulating expression of the multidrug transporter BCRP [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 6321.

Leukemic stem cells (LSCs) rely on oxidative metabolism and are differentially sensitive to targeting mitochondrial pathways, which spares normal hematopoietic cells. A subset of mitochondrial proteins is folded in the intermembrane space via the mitochondrial intermembrane assembly (MIA) pathway. We found increased mRNA expression of MIA pathway substrates in acute myeloid leukemia (AML) stem cells. Therefore, we evaluated the effects of inhibiting this pathway in AML. Genetic and chemical inhibition of ALR reduces AML growth and viability, disrupts LSC self-renewal, and induces their differentiation. ALR inhibition preferentially decreases its substrate COX17, a mitochondrial copper chaperone, and knockdown of COX17 phenocopies ALR loss. Inhibiting ALR and COX17 increases mitochondrial copper levels which in turn inhibit S-adenosylhomocysteine hydrolase (SAHH) and lower levels of S-adenosylmethionine (SAM), DNA methylation, and chromatin accessibility to lower LSC viability. These results provide insight into mechanisms through which mitochondrial copper controls epigenetic status and viability of LSCs.

Summary The mitochondrial caseinolytic protease P (ClpP) plays a central role in mitochondrial protein quality control by degrading misfolded proteins. Using genetic and chemical approaches, we showed that hyperactivation of the protease selectively kills cancer cells, independently of p53 status, by selective degradation of its respiratory chain protein substrates and disrupts mitochondrial structure and function, while it does not affect non-malignant cells. We identified imipridones as potent activators of ClpP. Through biochemical studies and crystallography, we show that imipridones bind ClpP non-covalently and induce proteolysis by diverse structural changes. Imipridones are presently in clinical trials. Our findings suggest a general concept of inducing cancer cell lethality through activation of mitochondrial proteolysis.

We evaluated outcomes of 100 patients with high risk AML treated with Ida-FLAG induction as first-line therapy. 72 achieved remission with one cycle; 19 did not. High risk cytogenetics and TP53 mutations were associated with failure to achieve remission. In those reaching remission, allogeneic bone marrow transplantation was associated with better relapse-free and overall survival. Those not achieving remission with induction therapy were extremely unlikely to reach remission with further therapy and had a dismal prognosis. Exploratory molecular analysis confirmed persistence of the dominant genetic mutations identified at diagnosis. Ex vivo chemosensitivity did not demonstrate significant differences between responders and non-responders. Thus, Ida-FLAG induction has a high chance of inducing remission in patients with high risk AML. Those achieving remission require allogeneic transplantation to achieve cure; those not achieving remission rarely respond to salvage chemotherapy and have a dismal outcome. Alternatives to conventional chemotherapy must be considered in this group.

Targeting Bruton's tyrosine kinase (BTK) with the small molecule BTK inhibitor ibrutinib has significantly improved patient outcomes in several B-cell malignancies, with minimal toxicity. Given the reported expression and constitutive activation of BTK in acute myeloid leukemia (AML) cells, there has been recent interest in investigating the anti-AML activity of ibrutinib. We noted that ibrutinib had limited single-agent toxicity in a panel of AML cell lines and primary AML samples, and therefore sought to identify ibrutinib-sensitizing drugs. Using a high-throughput combination chemical screen, we identified that the poly(ADP-ribose) glycohydrolase (PARG) inhibitor ethacridine lactate synergized with ibrutinib in TEX and OCI-AML2 leukemia cell lines. The combination of ibrutinib and ethacridine induced a synergistic increase in reactive oxygen species that was functionally important to explain the observed cell death. Interestingly, synergistic cytotoxicity of ibrutinib and ethacridine was independent of the inhibitory effect of ibrutinib against BTK, as knockdown of BTK did not sensitize TEX and OCI-AML2 cells to ethacridine treatment. Thus, our findings indicate that ibrutinib may have a BTK-independent role in AML and that PARG inhibitors may have utility as part of a combination therapy for this disease.

Acute myeloid leukemia (AML) cells have high oxidative phosphorylation and mitochondrial mass and low respiratory chain spare reserve capacity. We reasoned that targeting the mitochondrial RNA polymerase (POLRMT), which indirectly controls oxidative phosphorylation, represents a therapeutic strategy for AML. POLRMT-knockdown OCI-AML2 cells exhibited decreased mitochondrial gene expression, decreased levels of assembled complex I, decreased levels of mitochondrially-encoded Cox-II and decreased oxidative phosphorylation. POLRMT-knockdown cells exhibited an increase in complex II of the electron transport chain, a complex comprised entirely of subunits encoded by nuclear genes, and POLRMT-knockdown cells were resistant to a complex II inhibitor theonyltrifluoroacetone. POLRMT-knockdown cells showed a prominent increase in cell death. Treatment of OCI-AML2 cells with 10-50 µM 2-C-methyladenosine (2-CM), a chain terminator of mitochondrial transcription, reduced mitochondrial gene expression and oxidative phosphorylation, and increased cell death in a concentration-dependent manner. Treatment of normal human hematopoietic cells with 2-CM at concentrations of up to 100 µMdid not alter clonogenic growth, suggesting a therapeutic window. In an OCI-AML2 xenograft model, treatment with 2-CM (70 mg/kg, i.p., daily) decreased the volume and mass of tumours to half that of vehicle controls. 2-CM did not cause toxicity to major organs. Overall, our results in a preclinical model contribute to the functional validation of the utility of targeting the mitochondrial RNA polymerase as a therapeutic strategy for AML.

To identify new biological vulnerabilities in acute myeloid leukemia, we screened a library of natural products for compounds cytotoxic to TEX leukemia cells. This screen identified the novel small molecule Deoxysappanone B 7,4' dimethyl ether (Deox B 7,4), which possessed nanomolar anti-leukemic activity. To determine the anti-leukemic mechanism of action of Deox B 7,4, we conducted a genome-wide screen in Saccharomyces cerevisiae and identified enrichment of genes related to mitotic cell cycle as well as vacuolar acidification, therefore pointing to microtubules and vacuolar (V)-ATPase as potential drug targets. Further investigations into the mechanisms of action of Deox B 7,4 and a related analogue revealed that these compounds were reversible microtubule inhibitors that bound near the colchicine site. In addition, Deox B 7,4 and its analogue increased lysosomal V-ATPase activity and lysosome acidity. The effects on microtubules and lysosomes were functionally important for the anti-leukemic effects of these drugs. The lysosomal effects were characteristic of select microtubule inhibitors as only the Deox compounds and nocodazole, but not colchicine, vinca alkaloids or paclitaxel, altered lysosome acidity and induced lysosomal disruption. Thus, our data highlight a new mechanism of action of select microtubule inhibitors on lysosomal function.

Mitochondria contain multiple copies of their own 16.6 kb circular genome. To explore the impact of mitochondrial DNA (mtDNA) damage on mitochondrial (mt) function and viability of AML cells, we screened a panel of DNA damaging chemotherapeutic agents to identify drugs that could damage mtDNA. We identified bleomycin as an agent that damaged mtDNA in AML cells at concentrations that induced cell death. Bleomycin also induced mtDNA damage in primary AML samples. Consistent with the observed mtDNA damage, bleomycin reduced mt mass and basal oxygen consumption in AML cells. We also demonstrated that the observed mtDNA damage was functionally important for bleomycin-induced cell death. Finally, bleomycin delayed tumor growth in xenograft mouse models of AML and anti-leukemic concentrations of the drug induced mtDNA damage in AML cells preferentially over normal lung tissue. Taken together, mtDNA-targeted therapy may be an effective strategy to target AML cells and bleomycin could be useful in the treatment of this disease.

The antimycotic ciclopirox olamine is an intracellular iron chelator that has anticancer activity in vitro and in vivo. We developed an oral formulation of ciclopirox olamine and conducted the first-in-human phase I study of this drug in patients with relapsed or refractory hematologic malignancies (Trial registration ID: NCT00990587). Patients were treated with 5-80 mg/m² oral ciclopirox olamine once daily for five days in 21-day treatment cycles. Pharmacokinetic and pharmacodynamic companion studies were performed in a subset of patients. Following definition of the half-life of ciclopirox olamine, an additional cohort was enrolled and treated with 80 mg/m² ciclopirox olamine four times daily. Adverse events and clinical response were monitored throughout the trial. Twenty-three patients received study treatment. Ciclopirox was rapidly absorbed and cleared with a short half-life. Plasma concentrations of an inactive ciclopirox glucuronide metabolite were greater than those of ciclopirox. Repression of survivin expression was observed in peripheral blood cells isolated from patients treated once daily with ciclopirox olamine at doses greater than 10 mg/m², demonstrating biological activity of the drug. Dose-limiting gastrointestinal toxicities were observed in patients receiving 80 mg/m² four times daily, and no dose limiting toxicity was observed at 40 mg/m² once daily. Hematologic improvement was observed in two patients. Once-daily dosing of oral ciclopirox olamine was well tolerated in patients with relapsed or refractory hematologic malignancies, and further optimization of dosing regimens is warranted in this patient population.

Acute myeloid leukemia (AML) cells and stem cells have unique mitochondrial characteristics with an increased reliance on oxidative phosphorylation (OXPHOS). To identify new biological vulnerabilities in the mitochondrial proteome of AML cells, we conducted an shRNA screen and identified neurolysin (NLN), a zinc metalloprotease whose mitochondrial function is not well understood and whose role in AML has not been previously reported. To begin our investigation into the role of NLN in AML, we analyzed NLN gene expression in a database of 536 AML and 73 normal bone marrow samples. NLN was overexpressed in 41% of AML samples. Overexpression of NLN in primary AML cells compared to normal hematopoietic cells was confirmed by immunoblotting. To validate the results of the screen and to determine whether NLN is required for AML growth and viability, we knocked down NLN in the leukemia cell lines OCI-AML2, MV4-11, NB4, and TEX with shRNA. NLN knockdown reduced leukemia growth and viability by 50-70%. Moreover, knockdown of NLN in AML cells reduced the clonogenic growth of leukemic cells in vitro and the engraftment of AML cells into mouse marrow after five weeks by up to 80% and 85%, respectively. The mitochondrial function of NLN is largely unknown, so we identified NLN's mitochondrial protein interactors in T-REx HEK293 cells using proximity-dependent biotin labeling (BioID) coupled with mass spectrometry (MS). This screen identified 73 mitochondrial proteins that preferentially interacted with NLN and were enriched for functions including respiratory chain complex assembly, respiratory electron transport, and mitochondrion organization. Therefore, we assessed the effects of NLN knockdown on OXPHOS. NLN knockdown reduced basal and maximal oxygen consumption, but there were no changes in the levels of individual respiratory chain complex subunits. To understand how NLN influences OXPHOS, we examined the formation of respiratory chain supercomplexes (RCS). Respiratory chain complexes I, III, and IV assemble into higher order quaternary structures called RCS, which promote efficient oxidative metabolism. NLN knockdown significantly impaired RCS formation in T-REx HEK293, OCI-AML2, and NB4 cells, which was rescued by overexpressing wild-type shRNA-resistant NLN. RCS have not been previously studied in leukemia. Therefore, we analyzed their levels in primary AML patient samples and normal hematopoietic cells. RCS assembly was increased in a subset of AML patient samples and positively correlated with NLN protein expression (R2 = 0.83, p < 0.05), suggesting that NLN mediates RCS assembly in AML. To investigate how NLN may be regulating RCS assembly, we analyzed our BioID results to identify NLN interactors that are known regulators of supercomplex formation. Among the top interactors was the known RCS regulator, LETM1. Knockdown of NLN in AML cells impaired LETM1 assembly. Of note, knockdown of LETM1 also reduced growth and oxygen consumption of AML cells. As a chemical approach to evaluate the effects of NLN inhibition on AML cells, we used the allosteric NLN inhibitor R2, (3-[(2S)-1-[(3R)-3-(2-Chlorophenyl)-2-(2-fluorophenyl)pyrazolidin-1-yl]-1-oxopropan-2-yl]-1-(adamantan-2-yl)urea), whose anti-cancer effects have not been previously reported. R2 reduced viability of AML cells, as well as two primary AML culture models, 8227 and 130578. R2 impaired RCS formation in OCI-AML2, NB4, 8227, and primary AML cells. Moreover, R2 reduced the CD34+CD38- stem cell enriched population in 8227 cells, reduced LETM1 complex assembly, and impaired OXPHOS in OCI-AML2 and 8227 cells. Finally, we assessed the effects of inhibiting NLN in mice engrafted with primary AML and normal hematopoietic cells in vivo. Treatment of mice with R2 reduced the leukemic burden in these mice without toxicity. Moreover, inhibiting NLN targeted the AML stem cells as evidenced by reduced engraftment in secondary experiments. In contrast, inhibiting NLN did not reduce the engraftment of normal hematopoietic cells. Collectively, these results demonstrate that inhibition of NLN preferentially targets AML cells and stem cells as compared to normal hematopoietic cells. In summary, we defined a novel role for NLN in RCS formation. We show that RCS are necessary for oxidative metabolism in AML and highlight NLN inhibition as a potential therapeutic strategy. Disclosures Minden: Trillium Therapetuics: Other: licensing agreement. Chan:Agios: Honoraria; AbbVie Pharmaceuticals: Research Funding; Celgene: Honoraria, Research Funding. Schimmer:Medivir Pharmaceuticals: Research Funding; Novartis Pharmaceuticals: Consultancy; Jazz Pharmaceuticals: Consultancy; Otsuka Pharmaceuticals: Consultancy.

Patients with relapsed AML (Acute Myeloid Leukemia) have a poor prognosis and few therapeutic options. Therefore, it remains critical to identify biological vulnerabilities in relapsed AML cells. To identify such vulnerabilities, we analyzed mRNA expression by RNA sequencing of 11 paired diagnosis and relapsed AML samples and identified 34 genes that were upregulated at FDR IPO11 is a member of the importin-β family of proteins functions that facilitate the import of protein cargo into the nucleus. To date, 10 importin-β family members have been identified, but only IPO11 was upregulated at AML relapse, although, importin β1, TNPO3 AND IPO13 were also hits in the CRISPR screen and necessary for AML growth and viability. Therefore, we focused our study on IPO11. Using independent databases of gene expression, we confirmed that IPO11 mRNA was upregulated in LSC+ (engrafting) vs. LSC- (non-engrafting) samples (rank 125, FDR To determine whether IPO11 is necessary for AML growth and viability, we knocked down IPO11 in OCI-AML2, TEX and NB4 leukemia cells with shRNA in lentiviral vectors. Knockdown of IPO11 induced cell cycle arrest with reduced AML growth and viability by 80-90%. In contrast, knockdown of another importin-β family member, IPO5, that was not a hit in our CRIPSR screen did not reduce AML growth and viability. In addition, we demonstrated that knockdown of IPO11 increased differentiation of AML cells as evidenced by increased CD11b expression and staining with non-specific esterase. Finally, knockdown of IPO11 reduced the clonogeneic growth of the above-mentioned AML cell lines and engraftment of TEX cells and the low passage primary AML sample 8227 cells into immune deficient mice by over 90%. To identify cargos of IPO11 whose nuclear import is necessary for AML survival, we compared IPO11 cargo identified through a Bio-ID mass spectrometry protein-protein interaction screen with hits identified from our CRISPR screen. Through this analysis, we identified 7 hits in common to both screens, including RFC5 (replication fork complex unit 5). RFC5 is an essential DNA polymerase accessory protein, involved in telomere maintenance, DNA replication and mismatch repair. Knockdown of IPO11 decreased the nuclear localization of RFC5 in AML cells. In addition, knockdown of RFC5 in AML cells decreased AML growth and decreased clonogeneic growth. Thus, in summary, comparing genes that are upregulated in paired diagnosis and relapse AML samples to our CRISPR screen we identified IPO11 as a single hit. IPO11 knockdown reduced growth and viability, promoted differentiation, reduced CFU generation of several AML cells lines. In a mouse model we show that IPO11 knockdown dramatically reduced engraftment of leukemic cells. We further identify RFC5, a protein involved in replication and DNA damage response as one of the cargos of IPO11. Our study suggests IPO11 supports LSC survival and relapse and provides a new target for novel therapeutic intervention. Disclosures Schimmer: Otsuka Pharmaceuticals: Consultancy; Medivir Pharmaceuticals: Research Funding; Jazz Pharmaceuticals: Consultancy; Novartis Pharmaceuticals: Consultancy.

Targeting cell metabolism has emerged as a viable treatment strategy in acute myeloid leukemia (AML), a malignant hematological disease characterized by poor patient outcomes and limited chemotherapeutic options. Compared to the normal hematopoietic population, leukemia cells exhibit an altered phenotype characterized by increased mitochondrial mass as well as a greater reliance on oxidative phosphorylation and fatty acid oxidation (FAO) for survival. Mitochondrial FAO is a four-reaction process that catabolizes fatty acids to acetyl-CoA, generating reductive equivalents for the electron transport chain (ETC) and anaplerotic intermediates for the TCA cycle. Clinically approved FAO inhibitors, such as trimetazidine and ranolazine, are tissue-specific, targeting FAO in some tissues but affecting additional pathways in others. To better understand the potential clinical utility of targeting FAO, we systematically tested a panel of clinical and pre-clinical FAO inhibitors, reasoning that the most potent FAO inhibitor would lead to a novel anti-AML target. Avocadyne was the most potent anti-AML compound, inducing leukemic cell death (EC50: 2.5 µM) and suppressing clonogenic growth of primary samples, while sparing normal hematopoietic cells. Further, avocadyne (100mg/kg twice weekly for 5 weeks) reduced patient-derived AML cell engraftment in the bone marrow of immune deficient mice. As a component of avocatin-B, a mixture of two fatty alcohols previously determined to accumulate in the mitochondria, we confirmed that avocadyne inhibited long chain FAO using radiolabeled studies and high resolution respirometry. To identify a molecular target, avocadyne treated cells were immunoprecipitated with antibodies against each intramitochondrial enzyme involved in long chain FAO (e.g., very long acyl-CoA dehydrogenase (VLCAD; step 1) and the alpha and beta subunits of the mitochondrial trifunctional protein: HADHA, HADHB; steps 2-4). Immunoblotting and LC/MS analysis confirmed that avocadyne co-eluted with VLCAD, but not with HADHA or HADHB, confirming a direct physical interaction between avocadyne and VLCAD. VLCAD introduces a double bond to a fully saturated long chain fatty acid and reduces electron transfer flavoprotein (ETF)-bound FAD, transferring these electrons to the ETC. Using fluorescence spectrophotometry and respirometry, avocadyne directly inhibited VLCAD activity, resulting in reduced ETF-supported respiration. The activity of MCAD, an acyl-CoA dehydrogenase catalyzing medium chain fats, was not affected, suggesting avocadyne inhibits long chain FAO exclusively. Further, profiling of acyl-carnitines following avocadyne treatment also showed a pattern characteristic of long chain FAO inhibition at VLCAD. To further understand how VLCAD modulated avocadyne sensitivity, avocadyne-resistant and VLCAD knockdown cells were generated. Lentiviral knockdown of VLCAD sensitized leukemic cells to avocadyne-induced FAO inhibition and death. In contrast, cells resistant to avocadyne had increased protein expression of VLCAD but no change in other long chain FAO enzymes. With increased VLCAD-supported ETF respiration, higher concentrations of avocadyne were required to induce cell death, compared to the parental line. These results show that VLCAD expression modulated leukemic sensitivity to avocadyne. Inhibiting FAO at VLCAD triggered an adaptive metabolic switch towards glycolysis, characterized by increased extracellular acidification. This compensatory increase in glycolysis was ultimately insufficient to prevent the depletion of TCA metabolites and ATP, leading to leukemic death. In contrast, following avocadyne treatment, normal umbilical cord blood-derived mononuclear cells increased glycolytic as well as pyruvate dehydrogenase activity and had no decrease in ATP levels, cell viability, or clonogenic growth. Together, these results highlight, for the first time, VLCAD as a novel anti-AML target and further suggest the clinical utility of FAO inhibition as a potential anti-AML strategy. Disclosures Minden: Trillium Therapetuics: Other: licensing agreement. Schimmer:Medivir Pharmaceuticals: Research Funding; Jazz Pharmaceuticals: Consultancy; Novartis Pharmaceuticals: Consultancy; Otsuka Pharmaceuticals: Consultancy.

Hematopoietic cells are arranged in a hierarchy where mature blood cells arise from stem and progenitor precursors. AML is also hierarchical with differentiated blasts arising from leukemic stem/progenitor cells. Recent studies show that metabolites can affect epigenetic marks; however, it is unknown whether metabolic enzymes can directly localize to the nucleus to regulate stemness in AML and normal hematopoietic cells. Here, we discovered that the mitochondrial enzyme, Hexokinase 2, localizes to the nucleus in AML and normal hematopoietic stem cells to maintain stemness. Metabolic enzymes that localize to nucleus of stem cells were identified by evaluating stem and bulk fractions of OCI-AML-8227 leukemia cells, which are arranged in a hierarchy with functionally defined stem cells. We separated OCI-AML-8227 cells into CD34+38- and CD34-38+ populations by FACS and prepared nuclear and cytoplasmic lysates. Immunoblotting of the lysates revealed that the metabolic enzyme Hexokinase 2 (HK2) was increased in the nuclear fraction of 8227 stem cells compared to bulk cells. In contrast, other mitochondrial enzymes such as Enolase1, Aconitase2, and Succinate Dehydrogenase A & B, were not detected in the nuclear lysates. HK2 is an outer mitochondrial membrane protein that phosphorylates glucose to glucose-6-phosphate, initiating glycolysis. We confirmed nuclear HK2 in OCI-AML-8227 stem cells by confocal microscopy and also demonstrated nuclear HK2 in AML cell lines (OCI-AML2, NB4, K563, and MV411) and in 7 of 9 primary AML samples. We FACS sorted normal cord blood into populations of stem/progenitor (HSC, MPP, MLP, CMP, GMP and MEP) and differentiated (Monocytes, Granulocytes, B, T, and NK) cells. The localization of HK2 in these cells was analysed and quantified by immunofluorescence. Nuclear HK2 was detected in the stem/progenitor cells and progressively declined to minimal levels as cells matured. Next, we explored mechanisms that regulate nuclear localization of HK2. AKT-mediated phosphorylation of HK2 promoted localization to mitochondria while inhibition of phosphorylation increased its nuclear levels. Moreover, the nuclear import of HK2 was dependent on IPO5, a member of b-importin family that imports protein to the nucleus; CRM1 was responsible for HK2 nuclear export. We tested whether the nuclear localization of HK2 was functionally important to maintain stemness. We overexpressed HK2 tagged with nuclear localizing signals (PKKKRKV or PAAKRVKLD) in 8227 and NB4 leukemia cells. Selective overexpression of HK2 in the nucleus did not alter the rate of proliferation of the cells, however there was enhanced clonogenic growth and inhibition of retinoic acid-mediated cell differentiation. Conversely, we selectively reduced nuclear HK2 by expressing HK2 with an outer mitochondrial localization signal while knocking down endogenous HK2 with shRNA targeting the 3'UTR of HK2. Selective depletion of nuclear HK2 in AML cells did not alter growth rate, but did reduce clonogenic growth and increased differentiation after treatment with retinoic acid. To determine whether nuclear HK2 maintains stemness through its kinase activity, we over-expressed a kinase dead double mutant of nuclear HK2(D209A D657A). Nuclear kinase dead HK2 increased clonogenic growth and inhibited differentiation after retinoic acid treatment, demonstrating that HK2 maintains stemness independent of kinase function. To understand nuclear functions of HK2, we used proximity-dependent biotin labeling (BioID) and mass spectrometry to identify proteins that interact with nuclear HK2. A top hit in our screen was Exonuclease 3'-5' domain containing 2 (EXD2), involved in DNA repair. Of note, DNA damage induces differentiation of AML cells. In 8227 cells, nuclear EXD2 was higher in the stem cell fraction compared to the bulk fraction. Moreover, knockdown of EXD2 reduced AML growth, clonogenic growth and decreased nuclear HK2 levels. Finally, nuclear HK2 overexpression conferred resistance to the PARP inhibitor, olaparib. In summary, we discovered that unphosphorylated HK2 localizes to the nucleus in malignant and normal hematopoietic stem cells. Through mechanisms independent of its kinase function, nuclear HK2 maintains AML cells in their stem/progenitor state potentially by regulating DNA damage and repair. Thus, we define a new role for a mitochondrial enzyme in the regulation of stemness and differentiation. Disclosures Minden: Trillium Therapetuics: Other: licensing agreement. Schimmer:Medivir Pharmaceuticals: Research Funding; Otsuka Pharmaceuticals: Consultancy; Novartis Pharmaceuticals: Consultancy; Jazz Pharmaceuticals: Consultancy.

Acute myeloid leukemia (AML) is a hematopoietic malignancy characterized by the accumulation of malignant myeloid cells that have arrested maturation. Most therapeutic regimens approved or under development are cytotoxics. An alternate, but less explored therapeutic approach, is to induce terminal differentiation of AML cells. Upon differentiation, AML cells cease to proliferate or die. Phosphatidylserine decarboxylase (PISD) is a mitochondrial enzyme that converts phosphatidylserine (PS) to phosphatidylethanolamine (PE). Here, we explored the effects of inhibiting PISD on AML growth, stemness and differentiation. Knockout of PISD by CRISPR reduced the growth and clonogenic growth of OCI-AML2 cells. The reported chemical PISD inhibitor, 7-chloro-N-(4-ethoxyphenyl)-4-quinolinamine (aka: MMV007285), reduced growth and viability of OCI-AML2 cells (IC50 = 4.741 μM) and TEX cells (IC50 = 4.868 μM). Using the 8227 primary AML cell culture model, we showed that inhibiting PISD induced cell death in the functionally defined stem cell fraction (CD34+CD38-). MMV007285 also preferentially inhibited the clonogenic growth of primary AML cells (n = 7) over normal hematopoietic cells (n= 3). Moreover, MMV007285 induced AML cell differentiation as evidenced by increased CD11b expression and staining for non-specific esterase. Using high-performance thin layer chromatography (HPTLC), we found that inhibition of PISD with MMV007285 increased intracellular PS. To determine whether increased PS was functionally important, OCI-AML2 cells were treated with PS, resulting in reduced growth and clonogenic growth. Furthermore, PS supplementation targeted AML progenitor cells as it decreased engraftment of TEX cells in mice. Mechanistically, inhibiting PISD induced differentiation and decreased stemness in AML by activating Toll-like receptor (TLR) signaling. Specifically, inhibiting PISD upregulated TLR4 and 8 expression and increased expression of cytokines downstream of TLR activation. We also showed that TLR activation was functionally important to induce AML differentiation. Finally, we evaluated the effects of PISD inhibition in AML mouse models. MMV007285 (300 mg/kg/5 of 7 days orally for 10 days) decreased the growth of OCI-AML2 cells in SCID mice. Moreover, MMV007285 (150 mg/kg/5 of 7 days orally for 5 weeks) impeded the leukemic engraftment of primary AML cell in NOD/SCID mice without toxicity. Using secondary transplants, we showed that MMV007285 also targeted the leukemic stem cells. Taken together, inhibition of PISD altered phospholipid metabolism, inhibited growth and stemness, and increased differentiation in AML cells. Our findings reveal a previously undescribed link between mitochondrial phospholipid metabolism and AML stemness and differentiation, highlighting a potential new therapeutic strategy for AML. Citation Format: Mingjing Xu, Ayesh Seneviratne, Val A. Fajardo, Geethu E. Thomas, G. Wei Xu, Rose Hurren, S. Kim, Neil MacLean, Xiaoming Wang, Marcela Gronda, Danny Jeyaraju, Yulia Jitkova, David Sharon, Ahmed Aman, Rima Al-awar, Steven Chan, Mark D. Minden, Paul LeBlanc, Aaron D. Schimmer. Inhibiting the mitochondrial enzyme phosphatidylserine decarboxylase (PISD) reduces stemness and increases differentiation in acute myeloid leukemia (AML) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3003.

Mitochondria possess unique proteases that localize to specific sub-compartments of the organelle. However, the functions of these proteases are largely ill-defined. Here, we used proximity-dependent biotinylation (BioID) to map the interactomes of seven proteases located in the intermembrane space of the mitochondria. The mitochondrial intermembrane space proteases HTRA2, OMA1, YME1L1, LACTB, IMMP1L, IMMP2L and PARL were cloned in-frame with the abortive E. coli biotin ligase BirA*, and expressed in 293 T-REx cells. Cell culture media was spiked with biotin for 24 hrs, the cells lysed, and biotinylated proteins were isolated and identified by mass spectrometry. In total, we identified 342 different proteins as high confidence interactors of the seven mitochondrial proteases. Of these, 272 are assigned a GO mitochondrial annotation, and 230 proteins interacted with only 1 or 2 proteases in our dataset. Validation efforts were focused on high temperature requirement peptidase A 2 (HTRA2). HTRA2 is a serine protease that is released into the cytoplasm during apoptosis where it binds Inhibitor of Apoptosis Proteins (IAPs). However, little is known about the function of HTRA2 in the mitochondria. HTRA2 interacted with 60 mitochondrial, 11 nuclear and 4 cytoplasmic proteins, including its known interactor XIAP, and consistent with its known localization to these cellular compartments. HTRA2 interacted with 8 out of 13 components of the MIB complex, a multiprotein assembly that is essential for proper mitochondrial cristae formation. Knockdown of HTRA2 with shRNA in 293T-REx cells disrupted cristae formation and this phenotype was rescued by expression of an shRNA-resistant HTRA2 cDNA. Compared to normal hematopoietic cells, HTRA2 mRNA expression levels are increased in a subgroup of primary AML cells. HTRA2 knockdown in OCI-AML2 leukemia cells led to a similar disruption of mitochondrial cristae. Knockdown of HTRA2 in OCI-AML2 cells led to increased levels of the MIB subunit IMMT, but not two other MIB complex subunits, SAMM50 and CHCHD3. Finally, in cell-free assays, we demonstrate that recombinant HTRA2 can degrade recombinant IMMT, but not SAMM50 or CHCHD3.Thus, we have mapped the interactomes of the proteases of the mitochondrial intermembrane space. Through this effort, we discovered that HTRA2 regulates protein levels of the MIB complex subunit IMMT and that disruption of this process affects mitochondrial cristae formation. Citation Format: Aaron D. Botham, Etienne Coyaud, Sanjit Nirmalanandhan, Marcela Gronda, Rose Hurren, Neil Maclean, Jonathan St. Germain, Sara Mirali, Estelle Laurent, Brian Raught, Aaron Schimmer. Mapping the protein interactome of mitochondrial intermembrane space proteases identifies a novel function for HTRA2 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 4529.

ClpP is a mitochondrial protease and a major protein quality control mediator that primarily interacts with metabolic enzymes in mitochondria. Here, we demonstrate that activation of this protease results in prominent anti-cancer activity, and propose ClpP activation as a novel therapeutic strategy for cancer and hematologic malignancies. We used genetic and chemical tools to activate ClpP. In a genetic approach, we tested the anti-cancer effects of ClpP activation by expressing a constitutively active ClpP mutant. Indeed, induction of the active ClpP mutant induced apoptosis in vitro and inhibited tumor progression in vivo. To further explore the antineoplastic effects of ClpP activation, we then performed a chemical screen of an in-house library of on-patent and off-patent drugs and identified imipridones (ONC201 and ONC212) as potent ClpP agonists. Imipridones are first-in-class antineoplastic agents and have shown preclinical efficacy in various malignancies in vitro and in vivo and are currently being evaluated in clinical trials in a diverse spectrum of cancers. Importantly, we and others have shown that their activity is agnostic to TP53 mutational status. Of note, molecular targets of imipridones that bind the drugs and are functionally important for their cytotoxicity have never been identified. Through extensive chemical investigations, including analysis of binding mechanism of the compounds to ClpP in cell free (ITC) and cell based assays (CETSA) as well as molecular analysis of the crystal structure, we demonstrate that these molecules bind ClpP non-covalently, and activate the protease by stabilizing the ClpP 14-mer, enlarging the axial pores of the complex, and inducing structural changes in the residues surrounding and including the catalytic triad. In leukemia, lymphoma and colon cancer cells including primary acute myeloid leukemia (AML) cells, both compounds displayed potent ClpP-dependent cytotoxicity with IC50s in low micro- or nanomolar ranges. Importantly, in primary AML samples, pretreatment ClpP levels correlated with response to imipridones. In lymphoma and AML xenograft models, both genetic and chemical activation of ClpP resulted in antitumor effects, while expression of inactive D190A ClpP mutant induced resistance. Mechanistically, ClpP activation leads to increased degradation of substrates of the enzyme including respiratory chain complex subunits and mitochondrial translation system. The resultant impaired mitochondrial structure and reduction in oxygen consumption is selectively cytotoxic to malignant cells that rely highly on mitochondrial energy production for their survival, whereas normal cells are not affected. In conclusion, ClpP activation is an entirely novel therapeutic strategy for malignant tumors. Our findings also suggest a general concept of inducing TP53-independent cancer cell lethality through activation of mitochondrial proteolysis. Citation Format: Jo Ishizawa, Sarah F. Zarabi, R Eric Davis, Ondrej Halgas, Takenobu Nii, Yulia Jitkova, Ran Zhao, Jonathan St-Germain, Lauren E. Heese, Grace Egan, Vivian R. Ruvolo, Samir H. Barghout, Yuki Nishida, Rose Hurren, Wencai Ma, Marcela Gronda, Todd Link, Keith Wong, Mark Mabanglo, Kensuke Kojima, Gautam Borthakur, Neil MacLean, John Man Chun Ma, Andrew B. Leber, Mark D. Minden, Walid Houry, Hagop Kantarjian, Martin Stogniew, Brian Raught, Emil F. Pai, Aaron D. Schimmer, Michael Andreeff. Mitochondrial ClpP-mediated proteolysis induces selective cancer cell lethality [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2720.

Although environmental factors such as cytokines and stroma are recognized to influence the growth and chemosensitivity of acute myeloid leukemia (AML) cells, little is known about the effect of physical forces such as pressure. The pressure in the marrow of AML patients can be 10–20 fold higher compared to patients with solid tumors or non-malignant conditions,1 possibly due to increased cellularity and fibrosis. Therefore, we sought to establish whether the pressure in the marrow is another environmental factor that can influence chemosensitivity. Using AML cell lines and primary patient samples, we show that high pressure promotes a transition to a gel-like plasma membrane that reduces the intracellular accumulation of daunorubicin in AML cells. Therefore, biomechanical stimulus such as pressure is another environmental factor that can influence chemosensitivity. We designed and manufactured pressure chambers to deliver different levels of physiological pressure to AML cells while maintaining continuous gas exchange (Online Supplementary Figure S1). Two AML cell lines, HL60 and TEX, were cultured at pressures up to 310 mmHg above atmospheric pressure (atm). A pressure of 310 mmHg above atm corresponds to less than 0.5% of the maximum pressure acting on the head of the femur of an upright 160 lb person, and therefore, would be physiologically relevant as erosion of the trabeculae from leukemic infiltration increases sensitivity to external forces.1,2 We found that the growth and viability of these cell lines did not change at high pressure (Figure 1A). Next, we explored the impact of increased pressure on the sensitivity of these cells to three anthracyclines used in the treatment of AML: daunorubicin, idarubicin, and mitoxantrone.3,4 HL60 and TEX cells grown at increased static pressure were more resistant to daunorubicin, (Figure 1B), but not idarubicin (Figure 1C) or mitoxantrone (Figure 1D and Online Supplementary Figure S2). Furthermore, the effects of increased pressure on chemosensitivity were reversible. Figure 1. Acute myeloid leukemia (AML) cells at increased pressure display chemoresistance to danorubicin. (A). Cell viability measured by trypan blue exclusion assay of HL60 and TEX leukemia cells cultured over 4 days at 37°C at atmospheric pressure (atm) ... We also tested the impact of increased static pressure on samples from 5 patients with acute leukemia (4=AML, 1=T-ALL) that were sensitive to daunorubicin (patients’ characteristics are shown in Online Supplementary Table S1). Similar to the cell lines, increased pressure did not alter the growth and viability of the primary cells, but 3 of 5 samples (3 of 4 AML samples) had reduced sensitivity to daunorubicin at high pressure (Figure 1E and Online Supplementary Figure S3). To investigate the biological effects of increased pressure that may influence chemosensitivity, we measured the intracellular accumulation of daunorubicin. Compared to cells at atm, cells cultured at increased pressure accumulated less daunorubicin (Figure 2A and Online Supplementary Figure S4). In contrast, we observed no difference in the accumulation of [3H] mitoxantrone in TEX cells treated at increased pressure (Figure 2B). Figure 2. AML cells have reduced daunorubicin accumulation at increased pressure. (A). HL60 and TEX cells were cultured at 37°C at atm and 310 mmHg above atm for 3 days and then treated with 80, 160 and 320 nM of [3H] daunorubicin for 3 hours at the same ... No prior studies have investigated the impact of static pressure on plasma membrane dynamics in cancer cells. However, studies on red blood cells,5,6 bacterial membranes7 and synthetic lipid vesicles8,9 demonstrate that increased pressure alters membrane dynamics and influences drug and particle uptake. In these systems, the acyl chains of the phospholipids in the membrane straighten under increased pressure causing the membrane to become thicker and transition from a fluid liquid-crystalline state to a more solid gel-like state. The thicker gel-like membrane can reduce the permeability of the membrane to small molecules and can also impair the activity of membrane-associated drug uptake channels. Notably, these changes in membrane dynamics are rapidly reversible upon returning to normal pressure. To determine if plasma membrane dynamics are altered in AML cells at increased pressure, HL60, TEX, and primary patient AML cells were cultured at high pressure and stained with the lipophilic probe, laurdan (6-dodecanoyl-2-dimethylaminonaphthalene) (see Online Supplementary Appendix for description of method). The fluorescence spectrum of laurdan is sensitive to the physical state of membrane phospholipids, and has been used to evaluate plasma membrane dynamics in red blood cells exposed to increased pressure6 and in K562 leukemia cells exposed to increased temperatures.10 Using two- photon microscopy at 37°C, atm, and within 10 min of removing from the pressure chamber, we calculated the generalized polarization (GP) factor of laurdan and observed an increase in the gel-like state of the membrane in both HL60 and TEX cells cultured at high pressure (Figure 2C). Moreover, primary AML cells from patients 1, 2, and 3 that demonstrated a change in chemosensitivity to daunorubicin at higher pressure also showed an increase in the gel-like state in their membrane at increased pressure. In contrast, primary cells from patients 4 (T-ALL) and 5 (AML) that did not display a change in chemosensitivity at increased pressure also showed no change in membrane dynamics at increased pressure (Figure 2D and Online Supplementary Figure S5). Thus, in a subset of AML cells, increased pressure promotes a transition to a gel-like membrane that can reduce the intracellular accumulation of select anthracyclines. Cholesterol is an important regulator of plasma membrane structure and consistency, and the amount of cholesterol in the membrane can vary over 10-fold between cells. Prior studies have evaluated the impact of cholesterol content on membrane dynamics after exposure to biomechanical forces. For example, addition of cholesterol to endothelial cells decreased membrane fluidity and dampened the impact of sheer stress on the plasma membrane.11 Similarly, using synthetic lipid vesicles, as the amount of cholesterol in the vesicle increased, the membrane became more gel-like and little further change was observed upon exposure to increased pressure.12 Conversely, membranes with the least amount of cholesterol are the most liquid and show the greatest change in their membrane dynamics upon exposure to increased pressure. However, it is unknown whether cells alter their membrane composition in response to changes in biomechanical pressure. It is worthy of note that the tested anthracyclines vary in their hydrophobicity. Of the three agents, daunorubicin has the highest logP (logarithm of partition coefficient) value of 1.83 indicating that it is the most hydrophobic, while idarubicin and mitoxantrone have lower logP values of 0.2 and -3.1, respectively. Changes in membrane dynamics would, therefore, preferentially affect the more hydrophobic daunorubicin compared to the other agents. Our study is not without limitations. First, membrane phase transitions are generally rapidly reversible and the imaging was done at atm. Although we imaged the cells within 10 min of removal from the pressure chamber, some of the changes in membrane dynamics may have reversed before imaging could be completed. Second, the increased pressure could have decreased daunorubicin solubility and thus resulted in decreased sensitivity to the drug, although this is very unlikely as we did not detect precipitation of daunorubicin. In summary, our study shows that increased static pressure renders AML cells resistant to daunorubicin by reducing intracellular drug levels, but does not affect sensitivity to idarubicin and mitoxantrone. Thus, these results highlight biomechanical stimuli such as pressure as another environmental factor that can influence chemosensitivity. Moreover, it suggests that some AML patients might benefit from chemotherapeutic regimens that contain anthracyclines such as idarubicin or mitoxantrone rather than daunorubicin.

Acute myeloid leukemia (AML) cells meet the higher energy, metabolic, and signaling demands of the cell by increasing mitochondrial biogenesis and mitochondrial protein translation. Blocking mitochondrial protein synthesis through genetic and chemical approaches kills human AML cells at all stages of development in vitro and in vivo. Tigecycline is an antimicrobial that we found inhibits mitochondrial protein synthesis in AML cells. Therefore, we conducted a phase 1 dose‐escalation study of tigecycline administered intravenously daily 5 of 7 days for 2 weeks to patients with AML. A total of 27 adult patients with relapsed and refractory AML were enrolled in this study with 42 cycles being administered over seven dose levels (50–350 mg/day). Two patients experienced DLTs related to tigecycline at the 350 mg/day level resulting in a maximal tolerated dose of tigecycline of 300 mg as a once daily infusion. Pharmacokinetic experiments showed that tigecycline had a markedly shorter half‐life in these patients than reported for noncancer patients. No significant pharmacodynamic changes or clinical responses were observed. Thus, we have defined the safety of once daily tigecycline in patients with refractory AML. Future studies should focus on schedules of the drug that permit more sustained target inhibition.

D-cyclins are universally dysregulated in multiple myeloma and frequently overexpressed in leukemia. To better understand the role and impact of dysregulated D-cyclins in hematologic malignancies, we conducted a high-throughput screen for inhibitors of cyclin D2 transactivation and identified 8-ethoxy-2-(4-fluorophenyl)-3-nitro-2H-chromene (S14161), which inhibited the expression of cyclins D1, D2, and D3 and arrested cells at the G0/G1 phase. After D-cyclin suppression, S14161 induced apoptosis in myeloma and leukemia cell lines and primary patient samples preferentially over normal hematopoietic cells. In mouse models of leukemia, S14161 inhibited tumor growth without evidence of weight loss or gross organ toxicity. Mechanistically, S14161 inhibited the activity of phosphoinositide 3-kinase in intact cells and the activity of the phosphoinositide 3-kinases α, β, δ, and γ in a cell-free enzymatic assay. In contrast, it did not inhibit the enzymatic activities of other related kinases, including the mammalian target of rapamycin, the DNA-dependent protein kinase catalytic subunit, and phosphoinositide-dependent kinase-1. Thus, we identified a novel chemical compound that inhibits D-cyclin transactivation via the phosphoinositide 3-kinase/protein kinase B signaling pathway. Given its potent antileukemia and antimyeloma activity and minimal toxicity, S14161 could be developed as a novel agent for blood cancer therapy.

Amitriptyline is a classic tricyclic antidepressant (TCA) and has been used to treat the depression and anxiety of patients with cancer, but its relevance to cancer cell apoptosis is not known. In the present study, we demonstrated that amitriptyline inhibited cyclin D2 transactivation and displayed potential antimyeloma activity by inhibiting histone deacetylases (HDACs). Amitriptyline markedly decreased cyclin D2 promoter-driven luciferase activity, reduced cyclin D2 expression, and arrested cells at the G(0)/G(1) phase of the cell cycle. Amitriptyline-induced apoptosis was confirmed by Annexin V staining, and cleavage of caspase-3 and poly(ADP-ribose) polymerase-1. D-Cyclin expression is reported to be epigenetically regulated by histone acetylation. Thus, we examined the effects of amitriptyline on histone 3 (H3) acetylation and demonstrated that amitriptyline increased acetylation of H3 and expression of p27 and p21. Further studies indicated that amitriptyline interfered with HDAC function by down-regulation of HDAC3, -6, -7, and -8, but not HDAC2, and by interacting with HDAC7. Molecular docking analysis and molecular dynamics simulations revealed that amitriptyline bound to HDAC7 and formed strong van der Waals interactions with five residues of HDAC7, including Phe162, His192, Phe221, Leu293, and His326, thus inhibiting HDAC activity. Therefore, we found that amitriptyline inhibited cyclin D2 transactivation and HDAC activity and could be a promising treatment for multiple myeloma.

The cytotoxic nucleoside cytarabine (Ara-C) is a cornerstone of AML induction and consolidation therapies, but drug resistance contributes to disease relapse. Among clinically relevant mechanisms of Ara-C resistance is the over-expression of cytidine deaminase (CDA). CDA degrades Ara-C through deamination into inactive metabolites and its increased expression in AML cells results in drug resistance. Therefore, nucleoside analogues that are not degraded by CDA may be new therapeutic agents for this disease. Through our efforts to develop novel nucleoside analogues that overcome mechanisms of resistance to Ara-C, we identified 5-fluorotroxacitabine (5FTRX), a chain-terminating cytidine-based L-nucleoside. First, we tested whether 5FTRX was a substrate of CDA by evaluating its cytotoxicity in HEK-293 cells over-expressing CDA. Compared to wild type cells, Ara-C was 6-fold less active in cells over-expressing CDA (IC50 wild type HEK293: 3.7µM (95% CI: 3.2-4.2uM) vs IC50 CDA over-expression: 25.6uM (95% CI 21.6-30.3uM), consistent with degradation of Ara-C by CDA. In contrast, cells over-expressing CDA were more sensitive to 5FTRX, compared to wild type cells (IC50 wild type HEK293: 5.5uM (95% CI 4.7-6.3uM) vs IC50 CDA over-expression: 0.4uM (95% CI 0.4-0.5uM), potentially due to depletion of endogenous nucleotide pools by increased CDA. Thus, 5FTRX is not a substrate of CDA. We then treated 6 AML cell lines for 72 hours with increasing concentrations of 5FTRX and then measured cell growth and viability using the MTS assay. 5FTRX reduced the growth of 5 of 6 tested AML cell lines, with mean IC50 values (n= > 3) of 92nM (TEX), 130nM (KG1a), 150nM (MV4-11), 410nM (NB4), and 250nM (OCI-AML2). In contrast, K562 cells that have mutant p53 were resistant with an IC50 >50,000nM. The K562 cells were also resistant to Ara-C. 5FTRX reduced the clonogenic growth of primary AML samples (n=2) with a >90% reduction in growth at 50nM, demonstrating that 5FTRX targets leukemia initiating cells. To test whether 5FTRX induced DNA damage in AML cells, we measured changes in phosphorylated H2AX (pH2AX) after 5FTRX treatment. In the tested cell lines (OCI-AML2, TEX, NB4 and MV4-11 cells), 5FTRX increased pH2AX at concentrations associated with loss of viability. Finally, we evaluated the efficacy and toxicity of 5FTRX in mouse models of AML. 5FTRX displayed robust and dose-dependent inhibition of MV4-11 and OCI-AML2 tumors in mouse xenograft models, with complete tumor regressions and long-lasting tumor growth delays (>20 days at 100mg/kg in MV4-11 and >30 days in OCI-AML2) after a single cycle of 5 days of once-daily drug treatment, with no changes in body weight or behavior. The efficacy of 5FTRX was superior to Ara-C dosed for daily for 10 days at its MTD, 60 mg/kg (35%TGI, no regressions). MV4-11 tumors treated with 5FTRX displayed induction of pH2AX, reduction in proliferation (BrdU incorporation) and induction of necrosis, consistent with the mode of action and dramatic tumor regressions. The anti-leukemic effects of 5FTRX were further evaluated in mice engrafted intrafemorally with primary AML cells. 5FTRX (100 mg/kg i.p, x 5 days) reduced primary AML engraftment >95% compared to controls without toxicity (Fig 1). In summary, 5FTRX was identified as a potent inhibitor of AML cell proliferation in vitro and in vivo. In contrast to Ara-C, 5FTRX was not a substrate for CDA. Further development of analogues of 5FTRX is ongoing using protide prodrugs of the 5FTRX monophosphate to further increase potency and evade additional resistance mechanisms. Taken together, our findings support the further development of 5FTRX-based therapies for the treatment of AML, including AML patients with reduced sensitivity to Ara-C through high CDA expression. #AB and TPS contributed equally to this work. #ADS and MA contributed equally to this work. Disclosures Rizoska: Medivir AB: Employment, Equity Ownership. Rydergård:Medivir AB: Employment, Equity Ownership. Kylefjord:Medivir AB: Employment, Equity Ownership. Rraklli:Medivir AB: Employment. Eneroth:Medivir AB: Employment, Equity Ownership. Pinho:Medivir AB: Employment, Equity Ownership. Norin:Medivir AB: Employment, Equity Ownership. Bylund:Medivir AB: Employment, Equity Ownership. Moses:Medivir AB: Employment, Equity Ownership. Bethell:Medivir AB: Employment, Equity Ownership. Targett-Adams:Medivir AB: Employment, Equity Ownership. Schimmer:Jazz Pharmaceuticals: Consultancy; Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees; Otsuka Pharmaceuticals: Consultancy; Medivir AB: Research Funding. Albertella:Medivir AB: Employment, Equity Ownership.

The majority of mitochondrial proteins are encoded in the nucleus, translated in the cytoplasm and imported into the mitochondria. A subset of cysteine-rich proteins destined for the mitochondrial intermembrane space are oxidized and folded by the Mitochondrial IMS Assembly (MIA) pathway. We found that genes encoding substrates of the MIA pathway are over-expressed in leukemic stem cells compared to bulk AML cells. Therefore, we assessed the effects of inhibiting the MIA pathway in AML by targeting the FAD linked sulfhydryl oxidase ALR, an integral part of the MIA machinery. Knockdown of ALR with shRNA reduced the growth and viability of OCI-AML2, TEX and NB4 leukemia cells. In addition, ALR knockdown reduced engraftment of TEX cells into mouse marrow, demonstrating an effect on the leukemia initiating cells. The small molecule selective ALR inhibitor, MitoBloCK-6 killed AML cells (IC50 of 5-10 mM) and preferentially reduced the clonogenic growth of primary AML cells over normal hematopoietic cells. MitoBloCK-6 treatment (80 mg/kg i.p. 5 of 7 days x 2 weeks) of mice engrafted with primary AML cells strongly reduced the leukemic burden without changing mouse body weight, serum chemistries, or organ histology. In contrast, MitoBloCK-6 did not change engraftment of normal cord blood in similar experiments. As evidenced by secondary transplants, MitoBloCK-6 also targeted leukemic stem cells. As expression levels of ALR substrates are increased in AML stem cells we assessed the effects of ALR inhibition on differentiation in AML. Genetic or chemical inhibition of ALR induced the differentiation of AML cells as evidenced by changes in gene expression, increased differentiation associated CD surface marker expression and increased non-specific esterase. Interrogation of the effects of ALR inhibition on its substrates identified the mitochondrial copper chaperone, Cox17 as the primary downstream target in leukemic cells. Genetic or chemical inhibition of ALR selectively reduced levels of Cox17 protein. Validating the functional importance of these findings, knockdown of Cox17 phenocopied ALR inhibition and reduced AML proliferation and induced AML differentiation. Cox17 is a copper metallochaperone that promotes respiratory chain complex IV assembly by loading copper into the respiratory complex. COX17 knockdown slightly reduced the complex IV enzymatic activity, but did not change basal oxygen consumption or ROS production. However, COX17 knockdown increased intracellular levels of free copper 16-fold as measured by atomic mass spectrometry. Copper is a known inhibitor of adenosylhomocysteinase, a key enzyme involved in the preservation of S-adenosylmethionine (SAM):S-adenosylhomocysteine (SAH) ratio in cells. SAM is a global methyl donor and is critical for DNA methylation. Knockdown of COX17 or ALR inhibition with MitoBloCK-6 decreased levels of SAM and reduced DNA methylation in AML cells. Likewise, the enzymatic activity of adenosylhomocysteinase was reduced in OCI-AML2 cells after MitoBloCK-6 treatment. Importantly, co-treatment with the copper chelator, penicillamine, rescued reductions in SAM, DNA methylation, and cell viability after COX17 knockdown or MitoBloCK-6 treatment. Thus, we have discovered a novel copper-dependent mechanism by which mitochondrial pathways regulate epigenetics and stemness in AML. Moreover, inhibitors of ALR or COX17 may be a novel therapeutic strategy to promote the differentiation of AML cells and stem cells. Disclosures Schimmer: Otsuka Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees; Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees; Medivir AB: Research Funding; Jazz Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees.

Hematopoietic cells are arranged in a hierarchy where stem and progenitor cells differentiate into mature blood cells. Likewise, AML (Acute Myeloid Leukemia) is also hierarchical with leukemic stem and progenitor cells giving rise to more mature and differentiated blasts. Recent studies have shown that mitochondrial enzymes such as IDH2 can regulate AML stemness by altering metabolites that affect epigenetic marks. However, it is unknown whether mitochondrial metabolic enzymes can directly localize to the nucleus to regulate stemness in AML and normal hematopoietic cells. Here, we show that the mitochondrial enzyme, Hexokinase 2, localizes to the nucleus in AML and normal hematopoietic stem cells to maintain stemness. We sought to identify mitochondrial metabolic enzymes that localize to the nucleus of stem cells, by evaluating the stem and bulk fractions from 8227 leukemia cells. 8227 leukemia cells are arranged in a hierarchy with functionally defined stem cells present in the CD34+CD38- fraction. We separated 8227 cells into CD34+CD38- and CD34-CD38+ populations by FACS sorting and prepared lysates of the nuclear and cytoplasmic fractions from each population. Using immunoblotting, we measured levels of mitochondrial enzymes in the subcellular fractions of each population. We discovered that the metabolic enzyme Hexokinase 2 (HK2) was increased in the nuclear fraction of 8227 stem cells compared to bulk cells. In contrast, other mitochondrial enzymes such as Aconitase 2 and Succinate Dehydrogenase B were not detected in the nuclear fractions. HK2 is an outer mitochondrial membrane protein that phosphorylates glucose to glucose-6-phosphate, thereby initiating glycolysis and the entry of glucose metabolites into the TCA cycle in the mitochondria. The nuclear localization of HK2 in mammalian cells has not been previous reported. We confirmed that 8227 cells have nuclear HK2 by confocal fluorescent microscopy and also demonstrated nuclear HK2 in AML cell lines (OCI-AML2, NB4, K562, and MV411) and primary AML samples. We also FACS sorted normal cord blood into populations of stem/progenitor (HSC, MPP, MLP, CMP, GMP and MEP) and differentiated (B cells, T cells, NK cells, Monocytes and Granulocytes) cells. The localization of HK2 in these cell fractions was measured by immunofluorescence and quantified by Metamorph and Imaris. Nuclear HK2 was detected in the stem/progenitor cells and progressively declined to minimal levels as the cells matured (Fig 1A). The mitochondrial localization of HK2 is dependent on AKT-mediated phosphorylation of Thr-473 and inhibited by dephosphorylation by the phosphatase PHLPP1. We asked whether phosphorylation of HK2 regulates the nuclear abundance of HK2. Using AML2 cells, we showed that knockdown of PHLPP1 decreased the abundance of nuclear HK2, while inhibition of AKT increased HK2 in the nucleus. Finally, we tested whether the nuclear localization of HK2 was functionally important to maintain stemness. We over-expressed HK2 tagged with nuclear localizing signals (PKKKRKV and PAAKRVKLD) in 8227 and NB4 leukemia cells. We confirmed the selective over-expression of HK2 in the nucleus of these cells by immunoblotting and immunofluorescence. Increasing nuclear HK2 did not alter the proliferation of the cells under basal conditions. However, increasing nuclear HK2 enhanced clonogenic growth and blocked retinoic acid-mediated cell differentiation. In summary, we discovered that the unphosphorylated form of the metabolic enzyme HK2 localizes to the nucleus in malignant and normal hematopoietic stem cells and is functionally important to maintain stem/progenitor state. Thus, we define a new role for mitochondrial enzymes in the regulation of stemness and differentiation. Disclosures Schimmer: Medivir AB: Research Funding; Jazz Pharmaceuticals: Consultancy; Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees; Otsuka Pharmaceuticals: Consultancy.

UBA1 is the major ubiquitin-activating enzyme that initiates the ubiquitylation cascade whereby proteins are tagged with mono- or polyubiquitin to mark them for proteasomal degradation or modify their functions. Despite having equal levels of UBA1 protein, AML cell lines and primary AML cells are more dependent on UBA1 activity compared to normal hematopoietic cells, rendering them more vulnerable to UBA1 inhibition. Recently, we demonstrated that inhibiting UBA1 with the small-molecule inhibitor TAK-243 was selectively cytotoxic to a subset of AML cells and stem cells in vitro and in vivo through a mechanism at least partly dependent on inducing ER stress (Leukemia, 2018). To identify potential determinants of sensitivity/resistance to TAK-243 (Millennium Pharmaceuticals, Takeda) in AML, we conducted a genome-wide CRISPR/Cas9 knockout screen in OCI-AML2 cells followed by selection with cytotoxic TAK-243 concentrations corresponding to the IC90 and IC99. By next-generation sequencing and enrichment analysis, we then identified genes whose knockout renders AML cells resistant to TAK-243. We identified 34 hits in the IC90 and 11 hits in the IC99 arms of the screen (cut off FDR < 0.2). These hits are involved in signaling pathways including transcriptional regulation, histone methylation, ubiquitin conjugation, cell cycle progression, mTOR and NF-κB signaling pathways, consistent with the broad range of pathways regulated by UBA1-mediated ubiquitylation. We focused our investigation on BEN domain-containing protein 3 (BEND3) that ranked as a top hit in both arms (FDR = 0.0012). Compared to control, all 6 BEND3-targeting gRNAs were enriched up to 10,000 times. BEND3 is a transcriptional repressor that regulates heterochromatin organization. To validate the screen results, we independently knocked out BEND3 in OCI-AML2 cells using the 4 top performing gRNAs in the screen. We confirmed target knockout by immunoblotting. BEND3 knockout did not alter the basal proliferation rate of the cells. However, knockout of BEND3 rendered OCI-AML2 cells resistant to TAK-243 with up to a 4-fold increase in the IC50 by the MTS assay. Resistance to TAK-243 was confirmed by Annexin V/PI staining, PARP cleavage, and colony-forming assays. Cells were cross resistant to the NEDD8-activating enzyme inhibitor pevonedistat (2-fold IC50 increase), but not bortezomib or daunorubicin. As assessed by immunoblotting, BEND3 knockout did not change expression of UBA1, or the related enzymes UBA2, UBA3, or UBA6. BEND3 knockout was associated with reduced induction of ER stress as assessed by levels of CHOP and ATF4 after TAK-243 treatment. Conclusions Through a genome-wide CRISPR screen, we identified BEND3 as a determinant of sensitivity to TAK-243 in AML. Mechanistically, lack of BEND3 expression dampens the ER stress response to UBA1 inhibition. Thus, these results may highlight a new mechanism of sensitivity to TAK-243. Disclosures Schimmer: Jazz Pharmaceuticals: Consultancy; Medivir AB: Research Funding; Otsuka Pharmaceuticals: Consultancy; Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees.

On-patent and off-patent drugs with previously unrecognized anticancer activity could be rapidly repurposed for this new indication given their prior toxicity testing. To identify such compounds, we conducted chemical screens and identified the antihelmintic flubendazole. Flubendazole induced cell death in leukemia and myeloma cell lines and primary patient samples at nanomolar concentrations. Moreover, it delayed tumor growth in leukemia and myeloma xenografts without evidence of toxicity. Mechanistically, flubendazole inhibited tubulin polymerization by binding tubulin at a site distinct from vinblastine. In addition, cells resistant to vinblastine because of overexpression of P-glycoprotein remained fully sensitive to flubendazole, indicating that flubendazole can overcome some forms of vinblastine resistance. Given the different mechanisms of action, we evaluated the combination of flubendazole and vinblastine in vitro and in vivo. Flubendazole synergized with vinblastine to reduce the viability of OCI-AML2 cells. In addition, combinations of flubendazole with vinblastine or vincristine in a leukemia xenograft model delayed tumor growth more than either drug alone. Therefore, flubendazole is a novel microtubule inhibitor that displays preclinical activity in leukemia and myeloma.

The vast majority of mitochondrial proteins are encoded in the nucleus, translated in the cytoplasm and then imported into the mitochondria. A subset of these imported proteins are folded into their mature and functional forms in the mitochondrial inter-membrane space (IMS) by the Mitochondrial IMS Assembly (MIA) pathway. We found that genes encoding substrates of the MIA pathway are over-expressed in leukemic stem cells compared to bulk AML cells. Therefore, we assessed the effects of inhibiting the MIA pathway in AML. We knocked down the mitochondrial sulfhydryl oxidase ALR, a key regulator of the MIA pathway. Knockdown of ALR with shRNA reduced the growth and viability of OCI-AML2, TEX and NB4 leukemia cells. In addition, knockdown of ALR reduced the engraftment of TEX cells into mouse marrow, demonstrating an effect on the leukemia initiating cells. The small molecule selective ALR inhibitor, MitoBloCK-6, mimicked the effects of ALR knockdown and killed AML cells with an IC50 of 5-10 μM. MitoBloCK-6 preferentially reduced the clonogenic growth of primary AML cells (n=4/5) over normal hematopoietic cells (n=4). However, only 3/10 bulk AML cells were sensitive to MitoBloCK-6 induced cell death by Annexin V/PI staining. Next, we evaluated the efficacy and toxicity of ALR inhibition in vivo . We injected primary AML cells or normal cord blood into the femurs of mice and then treated mice with MitoBloCK-6 (80 mg/kg i.p. 5 of 7 days x 2 weeks). MitoBloCK-6 strongly reduced the engraftment of primary AML samples but did not affect engraftment of cord blood. In secondary transplants, MitoBloCK-6 also targeted leukemic stem cells. No change in mouse body weight, serum chemistries, or organ histology was seen. As expression levels of ALR substrates are increased in AML stem cells, we assessed the effects of ALR inhibition on differentiation in AML. Genetic or chemical inhibition of ALR induced the differentiation of AML cells as evidenced by increased CD surface marker expression and increased non-specific esterase. In addition, ALR inhibition was preferentially cytotoxic towards undifferentiated cells and stem cells over differentiated bulk AML cells. Interrogation of the effects of ALR inhibition on its substrates identified the mitochondrial copper chaperone, Cox17 as the primary downstream target in leukemic cells. Inhibition of ALR selectively reduced levels of Cox17 protein and altered mitochondrial cristae structure. Validating the functional importance of these findings, knockdown of Cox17 phenocopied ALR inhibition and reduced AML proliferation, induced differentiation of AML cells, and altered mitochondrial cristae structure, without changing respiratory chain activity or oxygen consumption. Of note, cristae remodelling independent of respiratory chain function has been recently implicated in cellular differentiation and in yeast, Cox17 regulates the cristae organizing machinery. Thus, we have identified novel mechanisms by which mitochondrial pathways regulate the fate and differentiation of AML cells and stem cells Moreover, inhibition of ALR may be a novel therapeutic strategy to promote the differentiation of AML cells and stem cells. Disclosures Schimmer: Takeda Pharmaceuticals: Research Funding; Medivir: Research Funding; Novartis Pharmaceuticals: Honoraria.

Introduction: Ubiquitin-like Modifier Activating Enzyme 1 (UBA1; UAE) is the initiating enzyme in the ubiquitylation cascade in which proteins are tagged with ubiquitin moieties to regulate their degradation or function. Compared to normal hematopoietic cells, AML cell lines and primary AML cells have equal levels of UBA1 protein, but increased requirement for this enzyme. TAK-243 is a potent and selective inhibitor of UBA1 and we determined the preclinical activity, biological effects and mechanisms of resistance to the drug in AML. Results: TAK-243 reduced growth and viability of human AML cell lines (OCI-AML2, TEX, U937 and NB4) in a concentration- and time-dependent manner with IC50's ranging from 15-40 nM after treatment for 48 hours. In primary AML samples, most (n=18/21) were sensitive to TAK-243 with an IC50 Binding of TAK-243 to UBA1 and related E1 enzymes was measured in intact AML cells using the cellular thermal shift assay (CETSA). In AML cell lines and primary AML samples, TAK-243 bound UBA1 at concentrations associated with cell death, but bound other E1 enzymes UBA3 and UBA6 only at much higher concentrations. Next, we evaluated the biological effects of UBA1 inhibition by TAK-243. At concentrations associated with cell death, TAK-243 decreased the abundance of poly- and mono-ubiquitylated proteins in OCI-AML2 cells and primary AML samples. In addition, TAK-243 treatment increased PERK phosphorylation, CHOP, XBP1s and ATF4 which are markers of proteotoxic stress and unfolded protein response. TAK-243 inhibited DNA double strand break (DSB) repair as evidenced by reduced recruitment of 53BP1 to DSBs and sustained γH2AX foci after 3 Gy of irradiation. We assessed the preclinical efficacy and toxicity of TAK-243 in mouse models of AML. OCI-AML2 cells were injected subcutaneously (sc) into SCID mice, and when tumors were palpable, mice were treated with TAK-243 (20 mg/kg sc twice weekly). TAK-243 significantly delayed tumor growth in mice (T/C=0.02) with no toxicity as evidenced by no changes in mouse body weight, serum chemistry, or organ histology. In tumors and organs isolated from the above treated mice, TAK-243 preferentially reduced levels of mono- and poly-ubiquitylated proteins in tumors over normal tissues. As an additional model, primary AML cells from 2 patients were injected into the femurs of NOD-SCID mice. Two weeks after injection, mice were treated with TAK-243 (20 mg/kg sc twice weekly). After 3 weeks of treatment, mice were sacrificed, and AML engraftment in the non-injected femur was measured by flow cytometry. TAK-243 reduced primary AML tumor burden in both tested samples without toxicity. Using secondary transplantations, we demonstrated that TAK-243 had targeted the leukemic stem cells. To understand mechanisms of resistance to TAK-243, we selected a population of TAK-243-resistant OCI-AML2 by culturing cells with increasing concentrations of the drug. Persisting cells were 33-fold more resistant to TAK-243 compared to wild-type cells (IC50 757 vs 23 nM), but had a normal rate of proliferation and remained equally sensitive to bortezomib, daunorubicin, mitoxantrone and the NEDD8-activating enzyme inhibitor pevonedistat. Using CETSA, we showed reduced binding of TAK-243 to UBA1 in the resistant cells. We sequenced UBA1 exons 12-16 and 23-24 that span the adenylation domain. Resistant cells had a missense mutation in exon 16 resulting in substitution of tyrosine with cysteine at codon 583 (Y583C). Y583 in human UBA1 corresponds to Y551 in yeast Uba1, which makes a favorable interaction with TAK-243 in its Uba1 binding site. Therefore, Y583C substitution is predicted to interfere with TAK-243 binding to UBA1. Conclusions: TAK-243 is a potent and selective UBA1 inhibitor that displays preferential activity towards AML cells over normal hematopoietic cells. Acquired mutations affect drug binding and may be a clinically relevant mechanism of resistance. These data support conducting a clinical trial of TAK-243 in patients with AML. Disclosures Hyer: Takeda Pharmaceuticals International Co.: Employment. Berger: Takeda Pharmaceuticals International Co.: Employment. Traore: Takeda Pharmaceuticals International Co.: Employment. Sintchak: Takeda Pharmaceuticals International Co.: Employment. Milhollen: Takeda Pharmaceuticals International Co.: Employment. Schimmer: Takeda Pharmaceuticals: Research Funding; Medivir: Research Funding; Novartis Pharmaceuticals: Honoraria.

AML cells have unique mitochondrial characteristics with an increased reliance on mitochondrial metabolism and oxidative phosphorylation. To identify new biological vulnerabilities in the mitochondria of AML, we conducted a CRISPR knockout screen. CAS9-overexpressing human OCI-AML2 leukemia cells were transduced with a library of 91,320 sgRNAs in barcoded lentiviral vectors targeting 17,237 nuclear-encoded genes. Cells were harvested, genomic DNA was isolated, and the relative abundance of sgRNAs were determined by sequencing barcodes 14 days after puromycin selection. We focused on the sgRNAs targeting the 1050 mitochondrial proteins to identify targets in the mitochondrial proteome whose knockout reduced AML growth and viability. The cardiolipin remodeling enzyme tafazzin (TAZ) was among the top 1% of mitochondrial hits. Using individual sgRNA, we confirmed that knockout of TAZ reduced the growth of CAS9-OCI-AML2 cells by >70%, thus validating the findings from our screen. We also knocked down TAZ with 2 independent shRNA and demonstrated reductions in growth and viability of a panel of AML cells: OCI-AML2 (>80%), TEX (>50%), K562 (>50%), and U937 (>40%). Moreover, TAZ knockdown significantly reduced the engraftment of TEX leukemia cells in vivo by 80%, indicating that TAZ-knockdown reduces AML growth in vivo and can target leukemia initiating cells. In contrast, knockdown of TAZ in mouse models did not impair normal hematopoiesis nor reduced the abundance of hematopoietic stem cells, although more subtle defects in the hematopoietic stem cells might explain transient episodes of neutropenia seen in Barth's syndrome, a congenital condition associated with X-linked TAZ mutations. TAZ is responsible for the majority of Cardiolipin (CL) remodeling under physiological conditions. As expected the knockdown of TAZ in both AML and normal mouse hematopoietic cells increased the substrate (monlysocardiolipin) to product (CL) ratio of TAZ. CL is required for the proper localization, and efficient function of, respiratory chain enzymes. However, in AML cells, knockdown of TAZ did not alter respiratory chain complex activity, basal oxygen consumption, or respiratory chain reserve capacity. Recent studies have shown that mitochondrial pathways can regulate cell fate and differentiation independent of their effects on oxidative phosphorylation. Therefore, we examined changes in AML cell differentiation after TAZ knockdown. Knockdown of TAZ promoted the differentiation of AML cells as evidenced by increased non-specific esterase staining and increased CD11b expression on the cell surface. In breast cancer cells decreasing phosphatidylethanolamine (PE) levels, induced the differentiation of these cells. As TAZ regulates phospholipid remodeling, therefore we measured levels of PE and phosphatidylserine (PS) after TAZ-knockdown by spot densitometry. Interestingly, knockdown of TAZ in OCI-AML2 cells decreased PE and increased PS lipid levels. To determine whether alterations in PE and PS phospholipids are functionally important for differentiation of AML cells, we treated AML cells with MMV007285, an inhibitor of the phosphatidylserine decarboxylase (PISD), an enzyme that converts PS to PE. MMV007285 mimicked the effects of TAZ-knockdown and increased differentiation of OCI-AML2 and 8227 AML cells. In summary, the cardiolipin remodeling enzyme TAZ regulates the differentiation of AML cells by controlling levels of PS and PE, thereby highlighting a new mechanism by which phospholipids and mitochondrial enzymes regulate AML cell fate and differentiation. Moreover, PISD inhibition may be a novel therapeutic strategy to selectively promote the differentiation of AML. Disclosures No relevant conflicts of interest to declare.

The oncogene c-maf is frequently overexpressed in multiple myeloma cell lines and patient samples and contributes to increased cellular proliferation in part by inducing cyclin D2 expression. To identify regulators of c-maf, we developed a chemical screen in NIH3T3 cells stably overexpressing c-maf and the cyclin D2 promoter driving luciferase. From a screen of 2400 off-patent drugs and chemicals, we identified glucocorticoids as c-maf–dependent inhibitors of cyclin D2 transactivation. In multiple myeloma cell lines, glucocorticoids reduced levels of c-maf protein without influencing corresponding mRNA levels. Subsequent studies demonstrated that glucocorticoids increased ubiquitination-dependent degradation of c-maf and up-regulated ubiquitin C mRNA. Moreover, ectopic expression of ubiquitin C recapitulated the effects of glucocorticoids, demonstrating regulation of c-maf protein through the abundance of the ubiquitin substrate. Thus, using a chemical biology approach, we identified a novel mechanism of action of glucocorticoids and a novel mechanism by which levels of c-maf protein are regulated by the abundance of the ubiquitin substrate.

Background Normal epithelial cells undergo anoikis, or apoptosis on loss of anchorage to the extracellular matrix, by initiating the death receptor pathway of caspase activation. However, malignant epithelial cells with metastatic potential resist anoikis and can survive in an anchorage-independent fashion. We hypothesized that c-Fas – associated death domain – like interleukin-1 – converting enzyme – like inhibitory protein (FLIP), an endogenous inhibitor of death receptor signaling, may suppress anoikis. Methods We assessed viability and apoptosis of PPC-1 prostate cancer cells cultured in adherent and suspension condi tions using 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium inner salt and Annexin V staining assays. Expression of the death receptor Fas and activation of caspase 8 were measured using flow cytometry. Expression of Fas ligand was measured by reverse transcription – polymerase chain reaction. FLIP protein expression was measured by immunoblotting. Small-molecule inhibitors of FLIP (including the death receptor sensitizer 5809354) and small-interfering (si) RNA directed against FLIP were used to assess the effects of FLIP inhibition on anoikis of prostate cancer cells in vitro and in vivo. All statistical tests were two-sided. Results PPC-1 cells cultured in suspension resisted anoikis, despite increased expression of Fas (0 versus 8 hours, mean relative percent expression = 100% versus 135%, difference = 35%, 95% confidence interval [CI] = 10% to 61%; P = .02) and Fas L (0 versus 24 hours, mean relative percent expression = 100% versus 208%, difference = 108%, 95% CI = 18% to 197%; P = .02). Knockdown of FLIP expression by siRNA or treatment with 5809354 sensitized prostate cancer cells to anoikis (control siRNA versus FLIP siRNA at 10 nM, mean relative percent viability = 95% versus 51%, difference = 44%, 95% CI = 34% to 54%; P

Treatment regimens for acute myeloid leukemia (AML) continue to offer weak clinical outcomes. Through a high-throughput cell-based screen, we identified avocatin B, a lipid derived from avocado fruit, as a novel compound with cytotoxic activity in AML. Avocatin B reduced human primary AML cell viability without effect on normal peripheral blood stem cells. Functional stem cell assays demonstrated selectivity toward AML progenitor and stem cells without effects on normal hematopoietic stem cells. Mechanistic investigations indicated that cytotoxicity relied on mitochondrial localization, as cells lacking functional mitochondria or CPT1, the enzyme that facilitates mitochondria lipid transport, were insensitive to avocatin B. Furthermore, avocatin B inhibited fatty acid oxidation and decreased NADPH levels, resulting in ROS-dependent leukemia cell death characterized by the release of mitochondrial proteins, apoptosis-inducing factor, and cytochrome c. This study reveals a novel strategy for selective leukemia cell eradication based on a specific difference in mitochondrial function. Cancer Res; 75(12); 2478–88. ©2015 AACR.

AML (acute myeloid leukemia) cells have a unique reliance on mitochondrial metabolism and fatty acid oxidation (FAO). Thus, blocking FAO is a potential therapeutic strategy to target these malignant cells. In the current study, we assessed plasma membrane carnitine transporters as novel therapeutic targets for AML. We examined the expression of the known plasma membrane carnitine transporters, OCTN1, OCTN2, and CT2 in AML cell lines and primary AML samples and compared expression to normal hematopoietic cells. Of the three carnitine transporters, CT2 demonstrated the greatest differential expression between AML and normal cells. Using shRNA, we knocked down CT2 and demonstrated that target knockdown impaired the function of the transporter. In addition, knockdown of CT2 reduced the growth and viability of AML cells with high expression of CT2 (OCI-AML2 and HL60), but not low expression. CT2 knockdown reduced basal oxygen consumption without a concomitant increase in glycolysis. Thus, CT2 may be a novel target for a subset of AML.

Approved proteasome inhibitors have advanced the treatment of multiple myeloma but are associated with serious toxicities, poor pharmacokinetics, and most with the inconvenience of intravenous administration. We therefore sought to identify novel orally bioavailable proteasome inhibitors with a continuous daily dosing schedule and improved therapeutic window using a unique drug discovery platform. We employed a fluorine-based medicinal chemistry technology to synthesize 14 novel analogs of epoxyketone-based proteasome inhibitors and screened them for their stability, ability to inhibit the chymotrypsin-like proteasome, and antimyeloma activity in vitro. The tolerability, pharmacokinetics, pharmacodynamic activity, and antimyeloma efficacy of our lead candidate were examined in NOD/SCID mice. We identified a tripeptide epoxyketone, FV-162, as a metabolically stable, potent proteasome inhibitor cytotoxic to human myeloma cell lines and primary myeloma cells. FV-162 had limited toxicity and was well tolerated on a continuous daily dosing schedule. Compared with the benchmark oral irreversible proteasome inhibitor, ONX-0192, FV-162 had a lower peak plasma concentration and longer half-life, resulting in a larger area under the curve (AUC). Oral FV-162 treatment induced rapid, irreversible inhibition of chymotrypsin-like proteasome activity in murine red blood cells and inhibited tumor growth in a myeloma xenograft model. Our data suggest that oral FV-162 with continuous daily dosing schedule displays a favorable safety, efficacy, and pharmacokinetic profile in vivo, identifying it as a promising lead for clinical evaluation in myeloma therapy.

Mitochondrial respiration is a crucial component of cellular metabolism that can become dysregulated in cancer. Compared with normal hematopoietic cells, acute myeloid leukemia (AML) cells and patient samples have higher mitochondrial mass, without a concomitant increase in respiratory chain complex activity. Hence these cells have a lower spare reserve capacity in the respiratory chain and are more susceptible to oxidative stress. We therefore tested the effects of increasing the electron flux through the respiratory chain as a strategy to induce oxidative stress and cell death preferentially in AML cells. Treatment with the fatty acid palmitate induced oxidative stress and cell death in AML cells, and it suppressed tumor burden in leukemic cell lines and primary patient sample xenografts in the absence of overt toxicity to normal cells and organs. These data highlight a unique metabolic vulnerability in AML, and identify a new therapeutic strategy that targets abnormal oxidative metabolism in this malignancy.

BALB/c thymocytes can be divided into three distinct subsets according to the expression of a ligand for the NK activation receptor NKG2D (NKG2D-L) and the expression of MHC class I (MHC-I). The first subset (MHC-Imid/NKG2D-Lhigh or "N+") is predominately CD4+CD8+ double positive (DP), comprises approximately 35% of thymocytes in a 6-8-week-old adult and contains uncommitted cells that have neither undergone selection nor are committed to death by neglect. The second subset (MHC-Ilow/NKG2D-Llow or "M-"), also mostly DP cells, comprises approximately 50% of thymocytes and consists of cells committed to death by apoptosis, likely due to neglect. By contrast, the third subset (MHC-Ihigh/NKG2D-Llow or "M+") is largely single positive (SP), represents approximately 15% of thymocytes and mostly contains more mature cells that have undergone successful positive selection. The major advantage of the analysis is that it splits DP cells into two subpopulations, one committed to death by apoptosis and the other subjected to selection. The analysis also suggests that NKG2D-L may play a role in thymocyte development.

We demonstrate that IL-2-activated NK cells or lymphokine-activated killer cells recognize and kill syngeneic CD4+ and CD8+ T cells that have been activated by APCs. Induction with APC required TCR-specific Ag, and lysis was perforin mediated. Brefeldin A, which disrupts protein transport, inhibited the sensitivity induced by activation. In BALB/c, expression of NKG2D ligands correlated with lysis and could be inhibited by brefeldin A. As well, addition of anti-NKG2D mAb to a killing assay completely abrogated lysis. Transduction of mouse NKG2D into a human NK cell line, YTSeco, conferred upon it the ability to kill activated BALB/c T cells, indicating that NKG2D is necessary for recognition. Our data provide a basis for studying a role for NK cells in T cell regulation.

Tigecycline is a broad-spectrum, first-in-class glycylcycline antibiotic currently used to treat complicated skin and intra-abdominal infections, as well as community-acquired pneumonia. In addition, we have demonstrated that tigecycline also has in vitro and in vivo activity against acute myeloid leukemia (AML) due to its ability to inhibit mitochondrial translation. Tigecycline is relatively unstable after reconstitution, and this instability may limit the use of the drug in ambulatory infusions for the treatment of infection and may prevent the development of optimal dosing schedules for the treatment of AML. This study sought to identify a formulation that improved the stability of the drug after reconstitution and maintained its antimicrobial and antileukemic activity. A panel of chemical additives was tested to identify excipients that enhanced the stability of tigecycline in solution at room temperature for up to one week. We identified a novel formulation containing the oxygen-reducing agents ascorbic acid (3 mg/mL) and pyruvate (60 mg/mL), in saline solution, pH 7.0, in which tigecycline (1 mg/mL) remained intact when protected from light for at least 7 days. This formulation also preserved the drug's antibacterial and antileukemic activity in vitro. Moreover, the novel formulation retained tigecycline's antileukemic activity in vivo. Thus, we identified and characterized a novel formulation for tigecycline that preserves its stability and efficacy after reconstitution.

Mitochondrial carrier homolog 2 (MTCH2) is a mitochondrial outer membrane protein that functions as a receptor-like protein for pro-apoptotic BID. In addition to its role in apoptosis, recent findings show that MTCH2 also regulates cellular metabolism. In murine hematopoietic cells, loss of MTCH2 increases oxidative phosphorylation and reduces the number of hematopoietic stem cells. Here, we sought to understand the role of MTCH2 in leukemogenesis and knocked down MTCH2 in leukemia cell lines using multiple independent shRNAs. Knockdown of MTCH2 reduced growth and viability of AML cells: OCI-AML2 (>90%), TEX (>80%), U937 (>65%), and HL60 (>75%). MTCH2 knockdown also decreased the clonogenic growth of OCI-AML2 (>60%), TEX (>70%), and U937 (>40%) cells compared to controls. However, MTCH2 knockdown did not induce cell death as indicated by annexin V/PI staining. In addition, knockdown of MTCH2 in TEX cells reduced engraftment into the marrow of non-obese diabetic/severe combined immunodeficiency-growth factor (NOD/SCID-GF) mice (control 17±4%, n=10) vs. sh-MTCH2 (4±0.86%, n=10). In mouse models, knockout of MTCH2 decreased the leukomogenic potential of murine hematopoietic stem cells transformed with the MLL-AF9 oncogene and increased the survival of these mice. To understand the mechanism by which MTCH2 knockdown decreased cell growth, we used genome wide transcriptome analysis with RNA-seq and observed an up regulation of genes involved in cellular differentiation. Consistent with increased MTCH2 knockdown promoting differentiation, OCI-AML2 cells with MTCH2 knockdown displayed increased non-specific esterase staining. Increased differentiation (Lin+ve cells) was also observed in MLL-AF9 with MTCH2 knockout. Knockdown of MTCH2 in TEX and OCI-AML2 cells increased levels of H3 and H4 histone acetylation as demonstrated by immunoblotting. Of note, differentiation and increased H3 and H4 acetylation was not observed after inhibiting other mitochondrial processes, such as mitochondrial protein synthesis or mitochondrial DNA replication. Although MTCH2 is a receptor for BID, the increased H3 and H4 acetylation appeared independent of BID as small molecule BID inhibitors did not alter H3 and H4 acetylation. MTCH2 regulates cell metabolism. Therefore, we measured changes in intracellular metabolites in AML cells after MTCH2 knockdown. In AML cells, MTCH2 knockdown increased levels of lactate (2 fold), but did not change the basal rate of oxygen consumption or the activity of mitochondrial respiratory chain complexes. Loss of mitochondrial pyruvate dehydrogenase complex increases lactate levels and a recent study reported that the translocation of pyruvate dehydrogenase complex from the mitochondria to the nucleus under conditions of mitochondrial stress, increases H3 and H4 histone acetylation (Cell. 2014; 158(1):84-97). Therefore, we measured changes in the localization of pyruvate dehydrogenase complex after MTCH2 knockdown. Knockdown of MTCH2 decreased mitochondrial and increased nuclear dehydrogenase complex in OCI-AML2 and TEX cells. Thus, in summary, MTCH2 regulates the differentiation of AML cells and controls the localization of pyruvate dehydrogenase complex and histone acetylation. These results also suggest a mechanism by which loss of MTCH2 leads to reductions of normal hematopoietic stem cells. Disclosures Schimmer: Novartis: Honoraria.

Mitochondrial DNA (mtDNA) replication requires adequate nucleotide pools from the mitochondria and cytoplasm to support DNA biosynthesis. Gene expression profiling of 542 AML patient samples (GSE13159) demonstrated that 55% of AML patients had upregulated mtDNA biosynthesis pathway expression compared to 73 normal hematopoietic cells (mononuclear cells isolated from peripheral blood and bone marrow). We also identified upregulation of pathways which support mitochondrial nucleotide pools, which include mitochondrial nucleotide transporters and a subset of cytoplasmic nucleotide salvage enzymes, which phosphorylate nucleosides to nucleotides. Upregulation of nucleoside kinases in a subset of primary AML samples compared to normal hematopoietic progenitor cells (normal G-CSF (granulocyte-colony stimulating factor) mobilized peripheral blood stem cells (PBSC's)) was confirmed by immunoblotting. These results suggest that AML cells import cytoplasmic nucleotides to support mitochondrial DNA biogenesis. To determine if cytoplasmic nucleoside kinases regulate mtDNA content, we knocked down nucleoside kinases in AML cells. Partial target knockdown of DCK (deoxycytidine kinase) and CMPK1 (cytidine/uridine monophosphate kinase 1) reduced mtDNA content (60+8%, and 62+13%, respectively compared to controls, 5 and 7 days post-shRNA transduction), indicating a role in mtDNA biogenesis. As expected, knockdown of mtDNA replication factors POLG and TFAM reduced mtDNA content in AML cells. The cytidine nucleoside analog, 2'3'-dideoxycytidine (ddC) is activated by DCK and CMPK1 to produce its triphosphate form, ddC-triphosphate (ddC-TP). To assess nucleoside kinase activity, primary AML and normal hematopoietic cells were treated with ddC and total levels of ddC and ddC-TP were measured by mass spectrometry. Levels of ddC did not differ between AML and normal, but ddC-TP levels was increased in AML samples > 7-fold compared to normal (p< 0.05, one-way ANOVA). Previously we and others demonstrated that AML cells and stem cells have increased mitochondrial biogenesis and reliance on oxidative phosphorylation due to decreased spare reserve capacity and an inability to upregulate glycolysis. ddC-TP inhibits the sole mtDNA polymerase POLG, but not nuclear DNA polymerases. Given the increased activity of nucleoside kinases in AML cells over normal, we examined the effects of ddC treatment on mtDNA content and cellular bioenergetics. AML and normal cells were treated with increasing concentrations of ddC. At increasing times after treatment, ddC depleted mtDNA levels > 85% at 0.5 uM, 3 day treatment in OCI-AML2 and TEX cells as assessed by qPCR. ddC decreased protein expression of mtDNA encoded electron transport chain (ETC) subunits COXI and COX II, but not nuclear encoded subunit COXIV) and reduced basal oxygen consumption. ddC also decreased proliferation of AML cell lines (OCI-AML2, TEX, HL-60, K562) (> 95% reduction at 0.5uM, 10 days). Knockdown of DCK abrogated the effects of ddC on AML cell proliferation. We next examined the effects of ddC in primary human leukemia cells (AML = 7, CML blast crisis = 1, CMML-2 = 1) and normal hematopoietic progenitor cells (n=8). ddC preferentially inhibited mtDNA biosynthesis and reduced viability in a subset of primary cells (6 of 9 AML) compared to normal PBSC's (n=8). Sensitivity to ddC positively correlated with mtDNA depletion. Finally, we evaluated the efficacy and toxicity of ddC in mouse models of human AML. ddC (35 mg/kg daily i.p. x 11 days) caused tumor regression in an OCI-AML2 xenograft model without toxicity (changes body weight, behavior, serum chemistries). In OCI-AML2 cells isolated from treated mice, ddC reduced mtDNA by 95% and mtDNA-encoded ETC proteins by 90%. In addition, ddC (75 mg/kg i.p x 6 weeks) significantly reduced human AML bone marrow engraftment in primary AML (n=2, P Thus, AML cells have increased nucleoside kinase activity that is functionally important for mtDNA biogenesis. We leveraged this unique biological vulnerability to preferentially activate ddC, deplete mtDNA and selectively target oxidative phosphorylation in AML. Disclosures Schimmer: Novartis: Honoraria.

A subset of AML and stem cells have increased mitochondrial stress and increased expression of mitochondrial proteases that degrade misfolded mitochondrial proteins. Given the recent findings of the interplay between mitochondrial homeostasis and mitochondrial protein import, we hypothesized that AML cells have an increased reliance on mitochondrial protein import as a compensatory mechanism for increased mitochondrial stress. To test this hypothesis, we measured expression levels of key mitochondrial protein import genes in publicly available datasets (GSE30377, GSE42414 and GSE 24759) and demonstrated their up regulation in a subset of AML cells over normal hematopoietic cells. Increased expression occurred across the spectrum of molecular mutations and cytogenetic abnormalities. Moreover, expression levels of mitochondrial protein import genes were enriched in functionally defined AML stem cells over bulk cells. To assess the impact of inhibiting mitochondrial protein import in AML, we knocked down the outer mitochondrial membrane import channel TOM40, the inner membrane import channel Tim23 and the oxidase ALR, that folds proteins through a disulfide relay system in the mitochondrial inner membrane space and is a rate limiting step for the import of a subset of mitochondrial proteins. Knockdown of these targets in OCI-AML2, TEX and U937 leukemia cells with shRNA reduced growth and viability of AML cells. Knockdown of ALR targeted the leukemia initiating cells as it abrogated engraftment of TEX leukemia cells into immune deficient mice (shRNA ALR = 5.481 +/- 0.9 % engraftment vs shRNA control= 29.44 +/- 5.4 % engraftment; p =0.0004) . Mechanistically, knockdown of mitochondrial import genes reduced levels of nuclear (ATP5A, SDHB and NDUFB8), but not mitochondrial (CoxII) encoded proteins of the OXPHOS chain. This in turn led to decreased basal oxygen consumption in leukemic cells. As a chemical approach to investigate the impact of inhibiting mitochondrial protein import in AML and normal cells, we tested the effects of MitoBloCK-6 and related analogues that selectively inhibit ALR in zebrafish, hESCs and yeast models. MitoBloCK-6 and related analogues killed leukemia cell lines (OCI-AML2, TEX, Jurkat and NB4) with an IC50of 5-10 µM. At these concentrations, MitoBloCK-6 decreased levels of nuclear (ATP5A, SDHB and NDUFB8), but not mitochondrial encoded (CoxII) proteins of the OXPHOS chain. Demonstrating the functional importance of changes in mitochondrial metabolism by these compounds, rho zero 143B rhabdomyosarcoma cells that lack mitochondrial DNA and rely solely on glycolysis were resistant to MitoBloCK-6. Finally, we tested the effects of MitoBloCK-6 on primary AML and normal cells. Treatment with MitoBloCK-6 (2 µM) inhibited clonogenic growth of (4 of 5) primary AML > 64% but produced 50% of control without toxicity. Thus, we have demonstrated that AML cells have a unique reliance on mitochondrial protein import and inhibition of this pathway may be a new therapeutic strategy to selectively target a subset of AML and AML stem cells. Disclosures Schimmer: Novartis: Honoraria.

The antimycotic ciclopirox olamine is an intracellular iron chelator that has anticancer activity in vitro and in vivo. We developed an oral formulation of ciclopirox olamine and conducted the first-in-human phase I study of this drug in patients with relapsed or refractory hematologic malignancies (Trial registration ID: NCT00990587). Patients were treated with 5-80 mg/m² oral ciclopirox olamine once daily for five days in 21-day treatment cycles. Pharmacokinetic and pharmacodynamic companion studies were performed in a subset of patients. Following definition of the half-life of ciclopirox olamine, an additional cohort was enrolled and treated with 80 mg/m² ciclopirox olamine four times daily. Adverse events and clinical response were monitored throughout the trial. Twenty-three patients received study treatment. Ciclopirox was rapidly absorbed and cleared with a short half-life. Plasma concentrations of an inactive ciclopirox glucuronide metabolite were greater than those of ciclopirox. Repression of survivin expression was observed in peripheral blood cells isolated from patients treated once daily with ciclopirox olamine at doses greater than 10 mg/m², demonstrating biological activity of the drug. Dose-limiting gastrointestinal toxicities were observed in patients receiving 80 mg/m² four times daily, and no dose limiting toxicity was observed at 40 mg/m² once daily. Hematologic improvement was observed in two patients. Once-daily dosing of oral ciclopirox olamine was well tolerated in patients with relapsed or refractory hematologic malignancies, and further optimization of dosing regimens is warranted in this patient population.