www.aging-us.com AGING 2020, Vol. 12, No. 6 Research Paper The LSD1 inhibitor iadademstat (ORY-1001) targets SOX2-driven breast cancer stem cells: a potential epigenetic therapy in luminal-B and HER2-positive breast cancer subtypes Elisabet Cuyàs1,2,*, Juan Gumuzio3,*, Sara Verdura1,2, Joan Brunet4,5,6,7, Joaquim Bosch-Barrera4,5, Begoña Martin-Castillo8, Tomás Alarcón9,10,11,12, José Antonio Encinar 13, Ángel G. Martin3, Javier A. Menendez1,2 1 Program Against Cancer Therapeutic Resistance (ProCURE), Metabolism and Cancer Group, Catalan Institute of Oncology, Girona, Spain 2 Girona Biomedical Research Institute (IDIBGI), Girona, Spain 3 StemTek Therapeutics, Bilbao, Spain 4 Medical Oncology, Catalan Institute of Oncology (ICO), Girona, Spain 5 Department of Medical Sciences, Medical School University of Girona, Girona, Spain 6 Hereditary Cancer Program, Catalan Institute of Oncology (ICO), Bellvitge Institute for Biomedical Research (IDIBELL), L'Hospitalet del Llobregat, Barcelona, Spain 7 Hereditary Cancer Program, Catalan Institute of Oncology (ICO), Girona Biomedical Research Institute (IDIBGI), Girona, Spain 8 Unit of Clinical Research, Catalan Institute of Oncology, Girona, Spain 9 ICREA, Barcelona, Spain 10 Centre de Recerca Matemàtica (CRM), Barcelona, Spain 11 Departament de Matemàtiques, Universitat Autònoma de Barcelona, Barcelona, Spain 12 Barcelona Graduate School of Mathematics (BGSMath), Barcelona, Spain 13 Institute of Research, Development and Innovation in Biotechnology of Elche (IDiBE) and Molecular and Cell Biology Institute (IBMC), Miguel Hernández University (UMH), Elche, Spain *Equal contribution Correspondence to: Javier A. Menendez; email: jmenendez@idibgi.org Keywords: epigenetics, reprogramming, cancer stem cells, breast cancer, patient-derived xenografts Received: December 21, 2019 Accepted: February 5, 2020 Published: March 18, 2020 Copyright: Cuyàs et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. ABSTRACT SOX2 is a core pluripotency-associated transcription factor causally related to cancer initiation, aggressiveness, and drug resistance by driving the self-renewal and seeding capacity of cancer stem cells (CSC). Here, we tested the ability of the clinically proven inhibitor of the lysine-specific demethylase 1 (LSD1/KDM1A) iadademstat (ORY-100) to target SOX2-driven CSC in breast cancer. Iadademstat blocked CSC-driven mammosphere formation in breast cancer cell lines that are dependent on SOX2 expression to maintain their CSC phenotype. Iadademstat prevented the activation of an LSD1-targeted stemness-specific SOX2 enhancer in CSC-enriched 3-dimensional spheroids. Using high-throughput transcriptional data available from the METABRIC dataset, high expression of SOX2 was significantly more common in luminal-B and HER2- enriched subtypes according to PAM50 classifier and in IntClust1 (high proliferating luminal-B) and IntClust 5 (luminal-B and HER2-amplified) according to integrative clustering. Iadademstat significantly reduced mammospheres formation by CSC-like cells from a multidrug-resistant luminal-B breast cancer patient-derived www.aging-us.com 4794 AGING xenograft but not of those from a treatment-naïve luminal-A patient. Iadademstat reduced the expression of SOX2 in luminal-B but not in luminal-A mammospheres, likely indicating a selective targeting of SOX2-driven CSC. The therapeutic relevance of targeting SOX2-driven breast CSC suggests the potential clinical use of iadademstat as an epigenetic therapy in luminal-B and HER2-positive subtypes. INTRODUCTION dependent homolog of the amine oxidase family that demethylates monomethyl or dimethyl lysine 4 (K4) of The transcription factor SOX2, a master regulator of histone H3. LSD1 blockade with the small molecule embryonic and induced pluripotent stem cells [1–4], is inhibitor CBB1007 has been shown to enhance causally related to tumor initiation, aggressiveness, and repressive H3K9 methylation at the stemness-specific metastasis likely due to its ability to induce and enhancer of SOX2, thereby validating the notion that maintain the stemness of cancer stem cells (CSC) [5, 6]. LSD1 might serve as a selective epigenetic target for Given that a key mechanism of drug resistance relates therapeutic ablation of SOX2-driven cancer stemness to the incapacity of most standard therapeutics to [15]. Although CBB1007-like competitive LSD1 eradicate the minor subpopulation of CSC with self- inhibitors, which have been developed based on the renewal and seeding capacity, SOX2 has been structure of LSD1 with a peptide inhibitor derived from suggested as an attractive anti-cancer target to prevent the N-terminal tail of histone H3 [16], might be CSC-mediated clinical relapse [7, 8]. Unfortunately, the considered good candidates to selectively target CSC “undruggable” characteristics of transcription factors with SOX2-driven pluripotent stem cell properties [17], such as SOX2 has largely constrained the clinical most of them are in a preclinical stage. potential of SOX2-centered therapeutic strategies in major cancer types. Iadademstat (formerly ORY-1001; Oryzon Genomics, Barcelona, Spain), a clinically proven, highly potent and Pre-clinical approaches such as SOX2-targeting selective covalent small-molecule inhibitor of LSD1 [18– siRNAs, shRNAs, or miRNAs offer little therapeutic 22], is an emerging therapeutic in hematological value because of their poor efficacy and delivery. By malignancies. Iadademstat has been shown to induce blast the same token, zinc-finger-based artificial transcription cell differentiation and reduce the leukemia-propagating factors, which can modify the epi-transcriptional state stem cell compartment in acute myeloid leukemia (AML). of endogenous promoters with single locus specificity Initial results from a Phase I/IIa clinical trial of [9, 10], have been employed to reduce SOX2 mRNA iadademstat demonstrated its safety and good tolerability and protein via targeting of proximal SOX2 promoters together with preliminary signs of anti-leukemic activity in cultured cancer cells and xenografts [7, 11], but their in refractory and relapsed AML [20]. Based on these poor in vivo delivery to solid tumor tissue limits their findings, the Phase IIa ALICE study is currently ongoing usefulness for stable SOX2 down-regulation in a in elderly patients with AML not eligible for intensive clinical context. Targeting of SOX2-related upstream/ chemotherapy to combine iadademstat with standard of downstream signaling pathways has become a more care azacytidine (https://www.clinicaltrialsregister.eu/ctr- plausible approach, and pharmacological blockade of search/trial/2018-000482-36/ES). Beyond hematological either the FBXW2-MSX2 axis with pevonedistat [12], cancers, blocking LSD1 with iadademstat has been the EGFR-STAT3 pathway with the cationic triphenyl- proposed as a valid strategy in some solid tumors such as methane pharmacophore gentian violet [13], or small-cell lung cancer (SCLC) and melanoma [21, 22]. EGFR/SRC/AKT signaling with the EGFR inhibitors Indeed, the Phase II CLEPSIDRA trial is recruiting gefitinib and erlotinib and the Src inhibitor dasatinib relapsed SCLC patients to receive iadademstat in [14], have been proposed as strategies to target human combination with platinum-etoposide chemotherapy cancers with SOX2 overexpression. It is unknown, (https://www.clinicaltrialsregister.eu/ctr-search/trial/2018- however, how much of the anti-cancer activity of these 000469-35/ES). In addition, the capacity of iadademstat- indirect approaches can be attributable to SOX2 driven inhibition of LSD1 activity to activate immune depletion. Moreover, the aforementioned strategies responses has recently been proposed as a new means to mostly target the proximal promoters of the SOX2 gene overcome resistance to immune checkpoint inhibitors in driving SOX2 expression in the differentiated states of melanoma [22]. Iadademstat-driven reversion of tumor- cancer cells, and epigenetic re-activation of stemness- driving undifferentiated cell states in genomically-diverse specific enhancers that cause a subpopulation of tumor malignancies strongly supports the notion that LSD1 cells to shift towards a CSC state is unaffected. might serve as a highly selective epigenetic target for the Mechanistically, such an approach can be achieved by elimination of cancer cells with pluripotent stem cell-like inactivation of lysine-specific demethylase 1 properties [15, 16, 23, 24]. To test this hypothesis, we (LSD1/KDM1A), a flavin adenine dinucleotide (FAD)- here investigated the ability of iadademstat to target www.aging-us.com 4795 AGING SOX2-driven CSC in breast cancer, an unexplored cancer iadademstat might operate via both a type for iadademstat-based therapy. catalytic/enzymatic mechanism involving highly potent, direct targeting of the FAD cofactor (at lower Because the mechanism of action of iadademstat has concentrations), and a scaffolding/structural mechanism been proposed to either impede the removal of the involving inhibition of the chromatin binding activity of methyl group from mono-methylated and di- LSD1/RCOR1 (at higher concentrations). Docking methylated K4 and K9 of histone 3 on LSD1-targeted simulations of iadademstat in a crystal structure of genes via a catalytic/enzymatic mechanism [18], or to human LSD1 (chain A) including RCOR1/CoREST promote enhancer activation of subordinate genes (chain B) – a co-repressor that collaborates to through the displacement of LSD1 from chromatin via demethylate mono- and di-methylated H3-K9 in a scaffolding/structural mechanism [19], we first nucleosomes – and a histone H3 peptide (chain C), computationally investigated the capacity of produced eight clusters of docking poses (Figure 1). iadademstat to target the LSD1-bound FAD cofactor When the docking results were ranked according to the and to disturb the anchorage of LSD1 and its co- ascent of the binding energies for iadademstat (up to - repressor (RCOR1/CoREST) to chromatin. Second, 9.09 kcal/mol; Table 1), those clusters exhibiting the because epigenetic re-activation of SOX2 expression highest affinity (#1, #3, and #4, see inset on the top via a pluripotency-specific enhancer can cause a right of Figure 1) were in the nanomolar range and were subpopulation of tumor cells to dynamically acquire a predicted to occupy the same binding site as FAD in CSC state, we evaluated the capacity of iadademstat LSD1. Cluster #6 was predicted to interfere with the to target the mammosphere-forming capacity -a well- position of the histone H3 peptide, whereas clusters #7 accepted surrogate reporter of CSC activity- in and #8, with affinities in the low micromolar range, established in vitro models bearing distinct mutational were predicted to interact with both the LSD1 enzyme landscapes (i.e., BRCA1-mutated basal-like MDA- and RCOR1/CoREST. MB-436 and HER2 gene-amplified/luminal-B BT-474 cell lines) but sharing a common dependency on To add protein flexibility and provide additional SOX2 expression to maintain their CSC phenotype. information about different intra- and inter-molecular Third, because SOX2 confers sensitivity to LSD1 movements, we performed short molecular dynamics inhibition, we characterized the SOX2 expression (MD) simulations over the course of 10 ns together with pattern using the PAM50 classifier and the integrative binding free energy calculations under the Molecular clustering of transcriptional data available from the Mechanics Poisson-Boltzmann Surface Area (MM- Molecular Taxonomy of Breast Cancer International PBSA) approximation. The results highlighted the consortium (METABRIC). Fourth, we finally extremely high affinity of iadademstat at the FAD- evaluated the clinical relevance of iadademstat as a targeted clusters #3, and #4, which reached -103.909 novel anti-SOX2 epigenetic breast cancer therapy by and -111.598 kcal/mol, respectively (Table 1). assessing its ability to impact both the expression of SOX2 and the tumorsphere-forming capacity of CSC- Using an AlphaScreen™ assay with a biotinylated like cells derived from breast cancer patient-derived histone H3 peptide methylated at lysine 4, purified xenografts (PDX). LSD1, and a highly specific antibody that recognizes demethylated substrate, we confirmed the in vitro RESULTS efficacy of iadademstat to dose-dependently suppress the demethylase activity of LSD1, with a mean IC 50 of Binding mode of iadademstat to the LSD1-CoREST- 12 nmol/L (Supplementary Figure 1). histone H3 complex Iadademstat specifically suppresses the Two different models have been proposed to explain the mammosphere-formation potential of breast cancer mechanism of action of iadademstat. In the first, stem cells iadademstat rapidly and irreversibly binds the LSD1 cofactor FAD in a manner analogous to the monoamine One of the gold standards for evaluating the presence of oxidase inhibitor tranylcypromine [18]. In the second, CSC is their ability to form in vitro mammospheres in iadademstat physically separates LSD1/RCOR1 from low-density non-adherent serum-free medium the SNAG-domain transcription repressor GFI1 and supplemented with growth factors [25, 26–29]. We chromatin in a cell-type-specific manner [19]. Taking assessed the anti-CSC activity of iadademstat in triple- advantage of the solved three-dimensional structure of negative breast cancer (TNBC), a highly aggressive LSD1 in a ternary complex with its histone peptide breast cancer subtype driven by highly enriched CSC, substrate and RCOR1/CoREST, we aimed to which are related to therapy resistance, tumor relapse, computationally explore a working model whereby and metastasis [25, 30]. Accordingly, estrogen receptor www.aging-us.com 4796 AGING Figure 1. Binding mode of iadademstat to LSD1. Top. The backbone of LSD1 (chain A)/REST corepressor 1 (chain B)/histone H3 peptide (chain C) heterotrimeric complex is shown. For each cluster of the docked iadademstat (salmon color), only the molecule (spheres) with better binding energy is shown. The molecular docking was performed using the A and B chains in the absence of FAD and histone H3 peptide; however, the clusters of docked iadademstat are shown superimposed on the position that would occupy both the FAD and the histone H3 peptide. The cluster number is also indicated. The insert on the left shows the peptide histone H3 (chain C, backbone as cartoon and side chains as sticks) and the FAD (represented as spheres and with the green carbons). The insert on the right shows only the best pose of iadademstat docked in each cluster and the situation of the histone H3 peptide. The clusters #1, #3 and #4 of iadademstat would occupy the same position of the FAD and are shown superimposed. Bottom. The detailed map of the molecular interactions of iadademstat in each cluster is detailed (see also Table 1). Each inset shows the detailed interactions of each compound docked to the LSD1 heterodimer, indicating the participating amino acids involved in the interaction and the type of interaction (hydrogen bonds, hydrophilic interactions, salt bridges, Π-stacking, etc). www.aging-us.com 4797 AGING Table 1. Details of the interaction of iadademstat docked to the LSD1 heterodimer (see Figure 1A). Cluster G, Dissoc. MM/PBSA Residues of the receptor that contact number [kcal/mol] constant, [µM] Members solvation binding iadademstat energy [kcal/mol] Ile-284, Gly-285, Ser-286, Gly-287, Leu-307, Glu-308, Ala-309, Arg-310, Gly-315, Thr-588, 1 -9.09 0.21811 5% -10.018 Ala-589, Val-590, Arg-591, Thr-624, Leu-625, Pro-626, Val-629, Gln-632, Ala-636, Val-637, Trp-756, Ala-757 (chain A) Leu-625, Val-629, Leu-630, Lys-631, Gln-632, Gln-633, Pro-635, Ala-636, Val-637, Gln-638, 2 -8.84 0.32872 9% -21.945 Phe-639, Val-640, Pro-642, Leu-643, Thr-648 (chain A) Gly-285, Ser-286, Gly-287, Val-288, Ser-289, Gly-290, Glu-308, Ala-309, Arg-310, Gly-314, 3 -8.75 0.38632 5.5% -103.909 Gly-315, Arg-316, Val-317, Thr-624, Leu-625, Pro-626, Trp-756, Gly-800, Glu-801, Thr-810, Val-811, Ala-814 (chain A) Arg-316, Leu-329, Gly-330, Ala-331, Met-332, 4 -8.74 0.39379 17% -111.598 Val-333, Thr-335, Tyr-571, Leu-659, Asn-660, Lys-661, Trp-751, Ser-760, Tyr-761, Ala-809, Thr-810, Val-811 (chain A) Ala-178, Phe-179, Arg-182, Leu-183, Pro-184, 5 -7.95 1.49000 4.5% -68.647 His-185, Gly-338, Gly-339, Asp-557, Phe-558, Glu-559, Phe-560, Thr-561, Tyr-807 (chain A) Thr-335, Ala-539, Asn 540, Trp-552, Asp-555, Phe-558, Glu-559, Phe-560, Tyr-761, Ser-762, 6 -7.92 1.57000 2.5% -29.621 Tyr-763, Val-764, Asn 806, Tyr-807, Pro-808, Ala-809, Thr 810, His-812 (chain A) Lys-481, Ser-482, His-484, Arg-485, Thr-488 7 -7.09 6.40000 16.5% -18.198 (chain A) and Leu-372, Pro-373, Glu-374, Val-375, Ile-376, Gln-377, Asp-407, Val-408, Gly-410 (chain B) Tyr-391 (chain A) and Lys-309, Pro-310, Pro-311, Lys-312, Gly-313, 8 -6.33 22.98000 8% -32.309 Met-314, Phe-315, Leu-316, Ser-317, Gln-318 (chain B) For the best-docked iadademstat molecule of each cluster, the Gibbs free energy, the dissociation constant, the number of molecules members (as %), and the MM/PBSA solvation binding energy are shown. (ER)-negative/progesterone receptor (PR)-negative/ formation by iadademstat was not due to non-specific HER2-negative BRCA1mut/PTENmut MDA-MB-436 toxicity, as MTT-based cell viability assays run in cells, which can form smooth and round spheres in parallel in 10% serum-supplemented adherent suspension culture [25] and are dependent on SOX2 conditions showed no significant cytotoxic activity of expression to maintain their CSC phenotype [31], were iadademstat, even when employing concentrations as used in the Cell2Sphere™ assay to evaluate the impact high as 30 µmol/L – a dose that completely prevented of iadademstat on the ability of CSC to survive and mammosphere formation in MDA-MB-436 cells proliferate as floating microtumors. Specifically, we (Figure 2). examined the effects of iadademstat on the total number, size, and aspect of MDA-MB-436 mammo- Iadademstat suppresses pluripotency enhancer- spheres growing under stem cell-selective conditions. driven activation of SOX2 in breast cancer stem cells Compared with the untreated controls, exposure to graded concentrations of iadademstat resulted in a dose- LSD1-blocking compounds are known to differentially dependent decrease in the total number of target pluripotent cancer cells including teratocarcinoma, mammospheres (IC50 = 3.98 µmol/L; Figure 2). embryonic carcinoma, and seminoma, or embryonic stem Notably, the strong decrease in mammosphere cells that express SOX2, while having minimal growth- www.aging-us.com 4798 AGING inhibitory effects on non-pluripotent cancer or normal Iadademstat fails to target ALDH+ breast cancer somatic cells [15, 16, 23]. We therefore envisioned that stem cells iadademstat might suppress CSC function by repressing the re-activation of SOX2 in breast CSC, a transcriptional We next tested whether the mechanism of action of phenomenon that specifically occurs through activation iadademstat to specifically target mammosphere- of the distal enhancer of the SOX2 promoter that also forming CSC-like cells might be due to a more general controls SOX2 transcription in pluripotent stem cells [32, phenomenon involving drivers of breast cancer 33]. When we transfected tamoxifen-resistant, luminal- stemness other than SOX2. Thus, we evaluated its B/HER2+ BT-474 cells [34] with a luciferase reporter capacity to target cells with high levels of aldehyde vector containing the SOX2 distal enhancer region, we dehydrogenase-1 (ALDH1), a biomarker that has been observed a robust induction (9.2-fold on average) of suggested to label a tumorigenic cell fraction capable of reporter activity in mammosphere cultures when self-renewal [35, 36]. ALDH1+ cell subpopulations compared with the adherent culture control (Figure 3A). enriched for cancer-initiating activity can be readily Of note, the enhancer-driven transcriptional activation of identified by flow cytometry using the Aldefluor ® SOX2 differentially occurring in mammosphere cultures reagent, which quantifies ALDH activity by measuring was dose-dependently suppressed by iadademstat (up to the conversion of the ALDH substrate BODIPY 80% reduction at 10 µmol/L; Figure 3A). Importantly, aminoacetaldehyde to the fluorescent product BODIPY iadademstant-driven SOX2 silencing drastically reduced aminoacetate. Using HER2-overexpressing BT-474 the number of BT-474 mammospheres (data not shown). cells as a breast cancer model naturally enriched with Figure 2. Iadademstat suppresses mammosphere formation in a basal-like established cell line. Figure shows representative microscope representations (×2.5 magnification) of mammospheres formed by MDA-MB-436 cells growing in sphere medium for 6 days in the absence or presence of graded concentrations of iadademstat. The number of mammospheres (>100 µm diameter) is expressed as means (columns)  SD (bars). MTT uptake-based measurement of cell viability is expressed as percentages uptake (OD570) relative to untreated controls (=100% cell viability). The results are expressed as percentages means (columns)  SD (bars). *P < 0.05 and **P < 0.005, statistically significant differences from the untreated (control) group. www.aging-us.com 4799 AGING ALDH1+ cells (>50%), we detected a decrease in the like cell line MDA-MB-436 exhibits similar protein proportion of ALDH1+ cells following 72 h treatment features to those of the luminal-A breast cancer subtype with increasing concentrations of iadademstat (up to [37]. Because SOX2-overexpressing cells are particularly 35% decrease at 25 µmol/L iadademstat) as compared sensitive to LSD1 inhibitors [15, 17], we explored the with vehicle-treated controls (Figure 3B). possibility that SOX2 expression might associate with the luminal entity across breast tumor subtypes. When we SOX2 expression status associates with luminal-B examined the expression status of SOX2 in two sets of and HER2-positive breast carcinomas breast cancer cell lines organized by luminal, basal-A (i.e., basal-like breast cancer intrinsic subtype), and Comprehensive comparison of molecular portraits basal-B (i.e., claudin-low breast cancer intrinsic subtype) between cell lines and breast cancer tumors confirmed sub-classes [38, 39], most of the SOX2-overexpressing the luminal-B intrinsic subtype classification of the breast carcinoma cell lines were found to belong to the ER+/HER2+ BT-474 cell line but revealed that the basal- luminal subclass (Figure 4A). Figure 3. Iadademstat inhibits stemness-associated SOX2 expression in a luminal-B/HER2+ established cell line. (A) Schematic representation of SOX2 promoter structure indicating the proximal core promoter region and the location of the distal enhancer, which is induced exclusively upon CSC-driven mammosphere formation but not in cell-adherent differentiating conditions. Results are expressed as fold-induction of mammosphere culture-associated SOX2 reporter activity above adherent culture control in the absence or presence of graded concentrations of iadademstat. The results are expressed as percentages means (columns)  SD (bars). *P < 0.05 and **P < 0.005, statistically significant differences from the untreated (control) group. (B) Representative Aldefluor® assay to identify BT-474 cells with high ALDH activity (ALDH+) in the absence or presence of graded concentrations of iadademstat for 3 days. The ALDH inhibitor diethylaminobenzaldehyde (DEAB) was used as negative control. Monolayer cultures were fed with iadademstat on day 1. (Note: 1 µmol/L FM19G11, an epigenetic repressor of key genes involved in stemness including SOX2 [98], was employed as a positive control). www.aging-us.com 4800 AGING We then extracted breast cancer date sets from the integrative clusters (IntClust1-10) each associated with Molecular Taxonomy of Breast Cancer International distinct somatic aberrations (CNAs) and gene Consortium (METABRIC) [40, 41] to explore the expression changes [45, 46], the transcript level of association between SOX2 expression and multiple SOX2 was found to be significantly upregulated in the breast cancer subtypes. When the METABRIC breast IntClust1, which is constituted by ER-positive tumors cancer data set was classified into each of the four Gene predominantly classified into the higher proliferation Expression prognostic Index Using Subtypes (GENIUS) luminal-B intrinsic subtype, and in the IntClust5, which subgroups using a 3-gene classifier (i.e., HER2+, ER- mostly encompasses HER2-amplified breast composed /HER-, ER+/HER2- high proliferation, ER+/HER2- low of both HER2-enriched and luminal-B intrinsic proliferation) [42, 43], SOX2 expression was found to subtypes (Figure 4B, right panel). be significantly higher in the HER2+ and ER+/HER2- high proliferation (luminal B-like) subgroups (Figure Iadademstat suppresses the formation of CSC- 4B, left panel). When the METABRIC breast cancer enriched mammospheres derived from a multidrug- data set was classified into each of the five intrinsic resistant luminal-B breast cancer patient in a SOX2- subtypes (i.e., luminal-A, luminal-B, HER2-enriched, related manner basal-like, and normal-like) using the research-based 50-gene prediction analysis of microarray (PAM50) Because patient-derived xenograft (PDX) tumor models classifier [44], SOX2 expression was found to be more faithful recapitulate human tumor biology and significantly higher in the luminal-B and HER2- drug responsiveness than established human cancer cell enriched subtypes (Figure 4B, middle panel). When the lines [47, 48], we employed CSC-enriched mammo- METABRIC data set was classified into each of the 10 spheres generated by growing single-cell suspensions Figure 4. SOX2 expression is enriched in luminal-B and HER2-positive breast cancer subtypes. (A) Relative enrichment of SOX2 expression (213721_at) in breast cancer cell lines organized by luminal, basal A, and basal B sub-classes [38, 39]. (B) Box plots presents the SOX2 gene expression in primary breast tumors from the METABRIC project classified in distinct subtypes using 3-gene (left), PAM50 (middle), and integrative clusters (right) classifiers. The color line presents median, box shows interquartile region and whiskers – the highest (max) and the lowest (min) value. www.aging-us.com 4801 AGING from excised PDX tumors under adherent-free epirubicin, 5-fluorouracil, cyclophosphamide, taxotere, conditions [49] to test the anti-CSC activity of paclitaxel, bevacizumab, and gemcitabine. iadademstat in a clinically-relevant scenario (Figure 5). The BRE-0188 (ER+/PR+/HER2-) PDX model was Iadademstat treatment failed to decrease the number of generated from a clinical sample obtained from a 65- mammospheres originated from CSC-like cells derived year-old female with luminal-A invasive ductal from the luminal-A BRE-0188 PDX (Figure 5A); the size carcinoma with the presence of lymph node metastases. of BRE-0188 mammospheres, however, was The patients had not received any chemotherapy or significantly decreased after treatment with iadademstat. radiotherapy prior to surgery. The BRE-0192 Treatment of BRE-0188 mammospheres cultures with (ER+/PR+/HER2-) PDX model was generated from a iadademstat failed to alter SOX2 expression. Conversely, clinical sample obtained from a 45-year-old female with iadademstat treatment significantly decreased the number luminal-B invasive lobular breast carcinoma with the but not the size of mammospheres originated from the presence of lymph node metastases [50]. The patient multidrug-resistant CSC-like cells derived from the was a poor responder to prior therapies, including luminal-B BRE-0192 PDX (Figure 5B). Moreover, Figure 5. Iadademstat targets SOX2-driven CSC in breast cancer patient-derived xenografts. Cell2Sphere™ assays using BRE-0188 (A) and BRE-0192 (B) PDXs were performed as per the manufacturer’s instructions (http://stemtektherapeutics.com/en/cell2sphere#cell2 sphere_kit). Increasing concentrations of iadademstat were added to sextuplicate sets of wells on day 1. ImageJ was used to quantify the number (left panels) and size (middle panels; central lines indicate mean values) of 9-day-old mammospheres. Right panels. Total RNA from untreated and iadademstat-treated mammosphere cells was evaluated in technical triplicates for the abundance of SOX2 (Hs01053049_s1) relative to housekeeping gene 18S (Hs99999901_s1). The transcript abundance was calculated using the delta Ct method (i.e., the difference of Ct value between the target SOX2 gene and the endogenous 18S control) and presented as relative quantification. www.aging-us.com 4802 AGING treatment with iadademstat notably reduced the irrespective of their mutational landscape (e.g., BRCA1- expression of SOX2 in mammospheres collected from mutated basal-like, HER2 gene-amplified), but also in the BRE-0192 PDX. certain breast cancer intrinsic subtypes that might be intrinsically addicted to the major role of SOX2 in self- DISCUSSION renewal growth and expansion of CSC-like cells (i.e., luminal-B and HER2-positive). Such clinically relevant We provide the first evidence that the clinically proven specificity of the LSD1 inhibitor iadademstat against inhibitor of LSD1 iadademstat can be used to SOX2-driven CSC was confirmed not only in established circumvent the challenge of pharmacologically breast cancer cell lines that are dependent on SOX2 manipulating the epigenetic re-activation of SOX2 that expression to maintain their CSC phenotype [31, 34] but causes a subpopulation of tumor cells to shift towards a also in CSC-enriched mammospheres generated by CSC state, which has constrained the clinical relevance growing single-cell suspensions from excised PDX of SOX2-centered therapeutic strategies in major cancer tumors. types such as breast cancer. The therapeutic potential of targeting SOX2-driven CSC supports the clinical use of The fact that the SOX2 reporter assay detects variations in iadademstat as a novel anti-SOX2 epigenetic breast SOX2 transcription regardless of the proportion of cells cancer therapy, particularly in SOX2-enriched luminal- expressing a CSC marker along with the ability of B and HER2-positive subtypes. iadademstat-induced knockdown of SOX2 to sharply decrease the quantity of CSC-like cells capable of self- Some malignant tumors such as breast carcinoma depend renewal in vitro but not to completely suppress on SOX2 for their tumor-initiating ability [34, 51–53]. subpopulations with high enzymatic activity of ALDH1 Elevated LSD1 levels are associated with enhanced (ALDH1high), which has been shown to mark a breast SOX2 expression, and SOX2-overexpressing cells are cancer population enriched for proliferating CSC, particularly sensitive to LSD1 inhibitors. Accordingly, strongly suggest that the anti-CSC activity of iadademstat LSD1 has been proposed as a selective epigenetic target might be restricted to SOX2-driven CSC bio-behaviors for therapy in SOX2-expressing cancers, particularly in (e.g., 1-3% of all luminal-B/HER2+ BT-474 cells) that those carcinomas in which high SOX2 expression occurs do not necessarily overlap with that of ALDHhigh via 3q26 chromosomal amplification (e.g., lung SCLC, proliferating cells. Using CSC-enriched mammospheres serous ovarian carcinoma, cervical, head and neck, oral, obtained by growing single-cell suspensions from excised and esophageal carcinomas [6]). However, such LSD1 PDX tumor under adherent-free conditions, treatment inhibitor-based therapeutic strategies might become even with iadademstat resulted in significantly reduced more relevant as cancer is entering a new era where the mammosphere formation (but not reduced mammosphere concept of cellular/phenotypic plasticity involving the size) in a multi-drug resistant luminal-B breast carcinoma dynamic interconversion of cells with and without CSC and in significantly reduced mammosphere size (but not states is challenging how we treat and understand tumors reduced mammosphere number) in a treatment-naïve relapse [54–61]. LSD1 controls SOX2 expression by luminal-A breast cancer. The number of mammospheres binding to its distal enhancer, which drives SOX2 reflects the quantity of CSC-like cells capable of self- activation in pluripotent stem-like cells including CSC renewal in vitro, while the size of mammospheres is an [15, 33, 62]. Pharmacological blockade of LSD1 indirect measure of the self-renewal capacity of each selectively promotes methylation of H3K4 and H3K9 mammosphere-generating cell; thus, cell proliferation within the regulatory (enhancer) region of the SOX2 during mammosphere growth determines the size of the promoter, suppressing SOX2 activity and stimulating cell mammospheres [64–66]. Taking together, these results differentiation by augmenting H3K4 methylation on the indicate that iadademstant inhibited CSC-driven promoters of differentiation genes [15]. In a process that mammosphere formation efficiency through suppressing is not mutually exclusive, the differentiation genes that LSD1/SOX2 axis in a luminal-B breast carcinoma, are directly controlled by LSD1 can indirectly inhibit the whereas the size of mammospheres in a luminal-A tumor expression of SOX2 that confers stem cell-like traits to was regulated through other underlying proliferative breast cancer cells [63]. Indeed, the therapeutic mechanism via LSD1. Accordingly, treatment of implications of our current findings with the LSD1 mammospheres from the luminal-B (but not the luminal- iadademstat might involve molecular scenarios in which A) breast cancer patient with iadademstat significantly the epigenetic re-activation of SOX2 within a group of reduced the expression of SOX2, likely indicating a transformed cells in response to microenvironmental selective targeting of SOX2-driven CSC. The therapeutic cues, stochastic genetic and epigenetic alterations, and/or potential of targeting SOX2-driven CSC supports the treatment-imposed selective pressures can cause them to clinical use of iadademstat as a novel anti-SOX2 shift toward a CSC state, LSD1-driven aberrant epigenetic breast cancer therapy, particularly in endocrine activation of SOX2 via stemness-specific enhancers therapy-resistant luminal-B cases -which are known www.aging-us.com 4803 AGING employ SOX2 to increase the proportion of CSC-like thereby suppressing the re-activation of SOX2 cells, rendering them insensitive to tamoxifen [34]- and exclusively occurring in mammosphere-initiating breast in HER2-positive disease –in which SOX2 over- CSC. The ability of iadademstat to efficiently and expression correlates with poor differentiation [67] and specifically target the on/off LSD1-driven SOX2 HER2-targeted therapies such as trastuzumab fail to regulatory process that provides higher tumorigenic eliminate SOX2-overexpressing CSC [68]. We have to potential to cells with an epigenetically acquired CSC acknowledge that the LSD1/SOX2 axis may represent a phenotype [61, 72] might open new therapeutic provocative potential target for CSC elimination not only horizons that incorporate iadademstat in the anti-breast in luminal-B and HER2-positive tumor but also in the cancer armamentarium. We further propose that basal-like subtype, which can also be found in the SOX2- iadademstat might functionally deplete tumor-initiating overexpressing IntClust5 [46]. In this regard, because CSC-like cellular states that sustain tumorigenicity by SOX2 reactivation has been shown to depend on the impacting on fundamental controllers of cell fate recruitment of the tumor suppressor protein BRCA1 to choice, an epigenetic mechanism involving both the the pluripotency-related distal enhancer in the SOX2 downregulation of SOX2 and the re-activation of promoter and concomitant modification of H3K4 and epigenetically suppressed differentiation programs in H3K9 at the same enhancer [62], our findings in basal- SOX2-enriched breast cancer subtypes such as like MDA-MB-436 cells support the idea that BRCA1- Luminal-B and HER2-positive. mediated predisposition to breast and ovarian cancer [69] might involve a controlled loss of SOX2 expression, Furthermore, we are rapidly appreciating that nuclear establishing a direct link between LSD1-regulated reprogramming-like phenomena inducing the expression of SOX2, acquisition of stem-like cell acquisition of epigenetic plasticity and phenotype phenotypes, and BRCA1-related breast/ovarian cancer malleability should be viewed as a fundamental element initiation. Indeed, SOX2 is frequently gained in BRCA1 of a tissue’s capacity to undergo successful repair, aging germline mutated tumors and is preferentially expressed degeneration or malignant transformation [73–77]. in sporadic basal-like phenotypes having similar Thus, chronic or unrestrained cell plasticity would drive phenotypic and clinical characteristics to breast cancer aging phenotypes by impairing the repair or the arising in BRCA1 mutation carriers. Accordingly, SOX2 replacement of damaged cells and such uncontrolled might play a driver role in the development of their less phenomena of in vivo reprogramming might also differentiated/stem cell-like phenotypic traits charac- generate CSC-like cellular states [73–77]. teristics of the basal-like breast cancer phenotype [70]. Pharmacological tools selectively targeting the LSD1- Moreover, SOX2 overexpression occurs not only in SOX2 axis might be appropriate to experimentally high-grade serous ovarian tumors, but also in pre- uncouple the apparently counterintuitive capacity of malignant, fallopian tube epithelial cells from LSD1 blockade to promote reprogramming phenomena BRCA1/BRCA2 mutation carriers who underwent by regulating the balance between pluripotency and prophylactic salpingo-oophorectomy [71] – opening the differentiation [78–83] while preventing SOX2-driven possibility for considering LSD1-targeted epigenetic cancer stemness. This would raise the possibility of approaches such as iadademstat for breast/ovarian pharmacologically managing, in the appropriate cancer prevention. direction and intensity, the physiological versus pathological processes of SOX2-related reparative SOX2 activation has proved instrumental for the plastic cellular reprogramming in aging and cancer. acquisition of aberrant stemness properties in cancer cells. Accordingly, certain CSC bio-behaviors can be MATERIALS AND METHODS defined as rare therapy-resistant, self-renewing cancer cells that aberrantly express SOX2, which might Molecular docking provide specificity for CSC-targeted drug screening. The use of fluorescence protein expression-based The human histone demethylase LSD1 (UniProt code transcriptional reporters for activation of the LSD1- O60341)/REST corepressor 1 (UniProt code regulated enhancer element of the SOX2 gene promoter Q9UKL0)/histone H3 peptide (UniProt code P68431) can specifically identify cells with tumor-initiating ternary complex structure (3 Å resolution, PDB code activity; compounds that would be capable of impeding 2X0L) was obtained from the Research Collaboratory SOX2 activation might be viewed as valuable for Structural Bioinformatics Protein Data Bank (PDB). candidates for drugs aimed to target CSC. In this regard, The molecular structure of iadademstat was obtained our findings unravel for the first time the ability of from PubChem (PubChem_ID: 71543365). The specific iadademstat to inhibit an LSD1-targeted distal enhancer edition of the LSD1 protein structure involving the that specifically controls the expression of the stem cell removal of water, FAD, and histone H3 peptide, was transcription factor SOX2 in pluripotent stem cells, made using PyMol 2.0 software (PyMOL Molecular www.aging-us.com 4804 AGING Graphics System, v2.0 Schrödinger, LLC, at concentration of 0.9% and pH was maintained at 7.4. http://www.pymol.org/) without further optimization. Hydrogen atoms were added to the protein structure at the appropriate ionizable groups according to the Molecular docking analysis of iadademstat against calculated pKa in relation to the simulation pH (i.e., a LSD1 was performed as previously described [84–88]. hydrogen atom will be added if the computed pKa is The selected protein structure was subjected to higher than the pH). The pKa was computed for each geometry optimization using the repair function of the residue according to the Ewald method [95]. All FoldX algorithm [89]. To search for potential binding simulation steps were run by a preinstalled macro sites of iadademstat, a global molecular docking (md_run.mcr) within the YASARA suite. Data were procedure was performed with AutoDock/Vina using collected every 100 ps. YASARA v19.4.27 software [90, 91], where a total of 999 flexible docking runs were set and clustered (7 Å) TMM/PBSA was implemented with the YASARA around the putative binding sites. The YASARA pH macro md_analyzebindenergy.mcr to calculate the command was set to 7.4. The docking software has a binding free energy with solvation of iadademstat, scoring function to give an approximate calculation of complex, and free protein for the LSD1 form the Gibbs free energy variation (ΔG, kcal/mol) between complexes. The binding free energy (kcal/mol) was LSD1 and iadademstat in each binding pose, with more expressed according to the following equation: positive energy values indicating stronger binding [92]. All the values were included in the corresponding table Ebinding = [poterec(i) + solverec(i) + potelig + solvelig] with a negative sign; only the ΔG value for the best - [potecmp(i) + solvecmp(i)] compound docked in each cluster is shown in Table 1. To calculate this parameter, which is used to rank where i is the position number, “pote” is the potential compounds, Autodock Vina uses a force field scoring energy for the complex (potecmp), free protein function that considers the strength of electrostatic (poterec), or free ligand (potelig), and “solve” is the interactions, hydrogen bonding between all atoms of the solvation energy for the complex (solvecmp), free two binding partners in the complex, intermolecular van protein (solverec), or free ligand (solvelig). More der Waals forces, and also solvation and entropy positive binding free values indicate better binding. contributions [93]. Docking results usually cluster around certain hot spot conformations. Two complexed LSD1 enzymatic activity compounds were considered to belong to different clusters if the ligand Root-Mean-Square Deviation of Enzymatic reactions were performed in an AlphaScreen their atomic positions was greater than a minimum of 6 format in duplicate at room temperature for 60 minutes Å. Dissociation constants were recalculated from the in a 10 µL mixture containing assay buffer, histone H3 average binding energy of all compounds of each peptide substrate, LSD1 (BPS#50103, lot#130806-D) cluster. The number of iadademstat-docked molecules enzyme, and iadademstat (RG-6016, Cat. No. S7795, included in each compound cluster is indicated as Selleckchem.com). The 10-µL reactions were carried "members", as a percentage in Table 1. The key out in 384-well Optiplates (Perkin Elmer Life Sciences, residues of LSD1 interacting with iadademstat in each Waltham, MA). A serial dilution of the compounds was cluster were detected using also YASARA v19.4.27 first performed in 3.3% DMSO/assay buffer. From this software [90, 91]. All of the figures were prepared using step, 3 µL of iadademstat was added to 4 µL of enzyme PyMol 2.0 software and all interactions were detected and incubated for 30 minutes at room temperature. using the PLIP algorithm [94]. After this incubation, 3 µL of substrate was added to initiate the reaction. The final DMSO concentration was Molecular dynamics simulations 1%. After the reaction, 5 µL of anti-mouse acceptor beads (Perkin Elmer, diluted 1:500 with 1× detection YASARA dynamics v19.4.27 was also used for all the buffer) or 5 µL of anti-rabbit acceptor beads (Perkin MD simulations with AMBER14 as a force field. The Elmer, diluted 1:500 with 1× detection buffer) and 5 µL simulation cell was allowed to include 20 Å of primary antibody (BPS#52140E,F, diluted 1:200 with surrounding the protein and filled with water at a 1x detection buffer) were added to the reaction mix. density of 0.997 g/mL. Initial energy minimization was After brief shaking, the plate was incubated for 30 carried out under relaxed constraints using steepest minutes. Finally, 10 µL of AlphaScreen streptavidin- descent minimization. Simulations were performed in conjugated donor beads (Perkin Elmer, diluted 1:125 water at constant pressure-constant temperature (25°C) with 1× detection buffer) were added. After 30 minutes, conditions. To mimic physiological conditions, counter the samples were measured in the AlphaScreen ions were added to neutralize the system; Na + or Cl- microplate reader (EnSpire Alpha 2390 Multilabel were added in replacement of water to give a total NaCl reader, Perkin Elmer). www.aging-us.com 4805 AGING The AlphaScreen intensity data were analyzed and BRE-0192 PDX) mammospheres was recorded as a compared using Graphpad Prism software (GraphPad measurement of CSC content. Images were recorded Software Inc., San Diego, CA). In the absence of using a BioTek Cytation 5 image cytometer at 2.5× iadademstat, the AlphaScreen or fluorescence intensity magnification. Prior to image acquisition, spheroid (Ft) was defined as 100% activity. In the absence of cultures were stained with a fluorescent vital dye to enzyme, the intensity (Fb) was defined as 0% activity. The increase the accuracy of spheroid detection and analysis. percent activity in the presence of iadademstat was The system was then set to count number, size, and calculated according to the following equation: %activity aspect ratio of the objects. Thresholds were set to >100 = (F-Fb)/(Ft-Fb), where F=the A-screen intensity in the µm in size and 0.4 as aspect ratio (with 1 being the aspect presence of iadademstat. Once A-screen data were ratio of a perfect circle). Aspect ratio did not vary upon converted to LSD1 activity (%), those values were then iadademstat dosage. plotted against a series of iadademstat concentrations using non-linear regression analysis of sigmoidal dose- Cell viability response curves generated with the equation Y=B+(T- B)/1+10((LogEC50-X)×Hill Slope), where Y=percent activity, Cell viability was determined using a standard B=minimum percent activity, T=maximum percent colorimetric MTT-based reduction assay 72 h after activity, X=logarithm of compound and Hill Slope=slope exposure to graded concentrations of iadademstat. factor or Hill coefficient. The IC50 value was determined as the concentration of iadademstat causing a half- Aldefluor® activity assay maximal inhibition of control activity. The Aldefluor® assay was performed as per the SOX2 profiling in breast cancer datasets manufacturer’s instructions (StemCell Technologies, Vancouver, BC, Canada), with or without the addition We interrogated the publicly available METABRIC of graded concentrations of iadademstat for 72 h. breast cancer dataset in the United Kingdom and Canada Analysis was performed using a MACSQuant® [40], in which mRNA expression was measured using the Analyzer 10 flow cytometer (Miltenyi Biotec, Bergisch Illumina HT-12v13 platform and CNA with the Gladbach, Germany) for data acquisition. Affymetrix SNP 6.0 array. Gene-level expression files from METABRIC were downloaded from the cBioportal SOX2 enhancer reporter assay for Cancer Genomics (https://www.cbioportal.org/). We used the 3-gene, PAM50, and integrative clusters BT-474 cells were transfected with 5 µg of pGL3 Luc subtypes provided in the METABRIC dataset. control (Promega, Madison, WI, USA) or pGL2-Sox2- enhancer-Luc reporter plasmids [32, 33] using Breast cancer xenograft models Lipofectamine Plus (Invitrogen, Carslbad, CA). Twenty- four hours after transfection, the culture was split into For generation of BRE-0188 and BRE-0192 PDX two parts: one part was seeded in two-dimensional models, all clinical samples were collected under adherent culture plates and the other part was cultured in written informed consent (according to the Declaration non-adherent culture conditions to allow mammosphere of Helsinki) and a declaration for commercial use of the formation, in the absence or presence of graded samples from the Consultative Committee for the concentrations of iadademstat. After 48 h, cells were Protection of Persons in Biomedical Research harvested and luciferase activity was measured in (CCPPRB) of Dijon University Hospital under duplicate with the Glomax 20/20 luminometer (Promega) authorization by a French Ministry of Higher Education, and normalized by protein concentration in the extracts. Research and Innovation for human tissue collection, Results were expressed as fold induction of sphere and redistribution (CSP articles L 1243-3, L 1243-4, culture reporter activity above adherent culture control. and L 1245-5). PDX tumors were passaged by serial transplantation in immunocompromised mice [50]. Quantitative real-time PCR Mammosphere formation Total RNA was extracted from mammosphere cells using the Qiagen RNeasy Kit according to the Mammosphere formation was monitored using manufacturer’s instructions. One microgram of total Cell2Sphere™ assays (StemTek Therapeutics, Bilbao, RNA was reverse-transcribed to cDNA using the Spain) as per the manufacturer’s instructions [96, 97]. Reaction Ready™ First Strand cDNA Synthesis Kit Graded concentrations of iadademstat were added to (SABiosciences, Frederick, MD). PCR arrays were triplicate sets of wells on day 1 and the number of either processed according to the SABiosciences RT-PCR 6- (MDA-MB-436, BT-474) or 9-day-old (BRE-0188, manual and analyzed using an Applied Biosystems www.aging-us.com 4806 AGING 7500 Fast Real-Time PCR System with an automated https://doi.org/10.1515/hsz-2014-0317 baseline and threshold cycle detection. The data were PMID:25781683 interpreted using the web-based PCR array analysis tool 2. Rizzino A, Wuebben EL. Sox2/Oct4: A delicately from SABiosciences. balanced partnership in pluripotent stem cells and embryogenesis. Biochim Biophys Acta. 2016; Statistical analysis 1859:780–91. https://doi.org/10.1016/j.bbagrm.2016.03.006 All statistical analyses were performed using GraphPad PMID:26992828 Prism software. Cell-based experimental data are presented as mean ± S.D. Comparisons of means of ≥ 3 3. Takahashi K, Yamanaka S. Induction of pluripotent groups were performed by analysis of variance stem cells from mouse embryonic and adult fibroblast (ANOVA) and the existence of individual differences, cultures by defined factors. Cell. 2006; 126:663–76. in case of significant F values at ANOVA, were https://doi.org/10.1016/j.cell.2006.07.024 assessed by multiple contrasts. The expression of SOX2 PMID:16904174 in the breast cancer subtypes (METABRIC dataset) was 4. Takahashi K, Tanabe K, Ohnuki M, Narita M, Ichisaka T, examined using one-way ANOVA (Kruskal-Wallis) Tomoda K, Yamanaka S. Induction of pluripotent stem with Dunnett’s test. P values < 0.05 were considered to cells from adult human fibroblasts by defined factors. be statistically significant (denoted as *). All statistical Cell. 2007; 131:861–72. tests were two-sided. https://doi.org/10.1016/j.cell.2007.11.019 PMID:18035408 ACKNOWLEDGMENTS 5. Novak D, Hüser L, Elton JJ, Umansky V, Altevogt P, We are grateful to the Cluster of Scientific Computing Utikal J. SOX2 in development and cancer biology. (http://ccc.umh.es/) of the Miguel Hernández University Semin Cancer Biol. 2019. [Epub ahead of print]. (UNH) and the Centro de Supercomputación of the https://doi.org/10.1016/j.semcancer.2019.08.007 University of Granada (ALHAMBRA-CSIRC) for PMID:31412296 providing computing facilities. The authors would like 6. Mamun MA, Mannoor K, Cao J, Qadri F, Song X. SOX2 in to thank Dr. Kenneth McCreath for editorial support. Cancer Stemness: Tumor Malignancy and Therapeutic Potentials. J Mol Cell Biol. 2018. [Epub ahead of print]. CONFLICTS OF INTEREST https://doi.org/10.1093/jmcb/mjy080 PMID:30517668 Stock ownership: Á.G.M., StemTek Therapeutics (CEO). All other authors have no competing interests to 7. Hüser L, Novak D, Umansky V, Altevogt P, Utikal J. declare. The authors declare that the research was Targeting SOX2 in anticancer therapy. Expert Opin Ther conducted in the absence of any commercial or financial Targets. 2018; 22:983–91. relationships that could be construed as a potential https://doi.org/10.1080/14728222.2018.1538359 conflict of interest. Ethics approval was not required for PMID:30366514 this study as per the local legislation. 8. Zhang S, Sun Y. Targeting oncogenic SOX2 in human cancer cells: therapeutic application. Protein Cell. FUNDING 2020; 11:82–84.. https://doi.org/10.1007/s13238-019-00673-x Work in the Menendez laboratory is supported by the PMID:31748974 Spanish Ministry of Science and Innovation (Grant 9. Beltran AS, Blancafort P. Remodeling genomes with SAF2016-80639-P, Plan Nacional de l+D+I, founded by artificial transcription factors (ATFs). Methods Mol Biol. the European Regional Development Fund, Spain) and by an unrestricted research grant from the Fundació 2010; 649:163–82. Oncolliga Girona (Lliga catalana d’ajuda al malalt de https://doi.org/10.1007/978-1-60761-753-2_10 càncer, Girona). The Spanish Ministry of Economy and PMID:20680834 Competitiveness (MINECO, Project AGL2015-67995- 10. Garcia-Bloj B, Moses C, Sgro A, Plani-Lam J, Arooj M, C3-1-R) and the Generalitat Valenciana (PROMETEO/ Duffy C, Thiruvengadam S, Sorolla A, Rashwan R, 2016/006) supports work in the Encinar laboratory. Mancera RL, Leisewitz A, Swift-Scanlan T, Corvalan AH, Blancafort P. Waking up dormant tumor suppressor REFERENCES genes with zinc fingers, TALEs and the CRISPR/dCas9 system. Oncotarget. 2016; 7:60535–54. 1. Feng R, Wen J. Overview of the roles of Sox2 in stem https://doi.org/10.18632/oncotarget.11142 cell and development. Biol Chem. 2015; 396:883–91. PMID:27528034 www.aging-us.com 4807 AGING 11. Stolzenburg S, Rots MG, Beltran AS, Rivenbark AG, https://doi.org/10.1016/j.ccell.2018.02.002 Yuan X, Qian H, Strahl BD, Blancafort P. Targeted PMID:29502954 silencing of the oncogenic transcription factor SOX2 in 19. Maiques-Diaz A, Spencer GJ, Lynch JT, Ciceri F, Williams breast cancer. Nucleic Acids Res. 2012; 40:6725–40. EL, Amaral FM, Wiseman DH, Harris WJ, Li Y, Sahoo S, https://doi.org/10.1093/nar/gks360 Hitchin JR, Mould DP, Fairweather EE, et al. Enhancer PMID:22561374 Activation by Pharmacologic Displacement of LSD1 12. Yin Y, Xie CM, Li H, Tan M, Chen G, Schiff R, Xiong X, from GFI1 Induces Differentiation in Acute Myeloid Sun Y. The FBXW2-MSX2-SOX2 axis regulates stem cell Leukemia. Cell Rep. 2018; 22:3641–59. property and drug resistance of cancer cells. Proc Natl https://doi.org/10.1016/j.celrep.2018.03.012 Acad Sci USA. 2019; 116:20528–38. PMID:29590629 https://doi.org/10.1073/pnas.1905973116 20. Somervaille T, Salamero O, Montesinos P, Willekens C, PMID:31548378 Perez Simon JA, Pigneux A, Recher C, Popat R, 13. Pietrobono S, Morandi A, Gagliardi S, Gerlini G, Molinero C, Mascaro C, Maes T, Bosch F. Safety, Borgognoni L, Chiarugi P, Arbiser JL, Stecca B. Down- phamacokinetics (PK), pharmacodynamics (PD) and Regulation of SOX2 Underlies the Inhibitory Effects of preliminary activity in acute leukemia of ORY-1001, a the Triphenylmethane Gentian Violet on Melanoma first-in-class inhibitor of lysine-specific histone Cell Self-Renewal and Survival. J Invest Dermatol. 2016; demethylase 1A (LSD1/KDM1A): initial results from a 136:2059–69. first-in-human phase 1 study. Blood. 2016; 128:4060. https://doi.org/10.1016/j.jid.2016.06.610 https://doi.org/10.1182/blood.V128.22.4060.4060 PMID:27373978 21. Augert A, Eastwood E, Ibrahim AH, Wu N, Grunblatt E, 14. Singh S, Trevino J, Bora-Singhal N, Coppola D, Haura E, Basom R, Liggitt D, Eaton KD, Martins R, Poirier JT, Altiok S, Chellappan SP. EGFR/Src/Akt signaling Rudin CM, Milletti F, Cheng WY, et al. Targeting NOTCH modulates Sox2 expression and self-renewal of stem- activation in small cell lung cancer through LSD1 like side-population cells in non-small cell lung cancer. inhibition. Sci Signal. 2019; 12:eaau2922. Mol Cancer. 2012; 11:73. https://doi.org/10.1126/scisignal.aau2922 https://doi.org/10.1186/1476-4598-11-73 PMID:30723171 PMID:23009336 22. Sheng W, LaFleur MW, Nguyen TH, Chen S, 15. Zhang X, Lu F, Wang J, Yin F, Xu Z, Qi D, Wu X, Cao Y, Chakravarthy A, Conway JR, Li Y, Chen H, Yang H, Hsu Liang W, Liu Y, Sun H, Ye T, Zhang H. Pluripotent stem PH, Van Allen EM, Freeman GJ, De Carvalho DD, et al. cell protein Sox2 confers sensitivity to LSD1 inhibition LSD1 Ablation Stimulates Anti-tumor Immunity and in cancer cells. Cell Rep. 2013; 5:445–57. Enables Checkpoint Blockade. Cell. 2018; 174:549– https://doi.org/10.1016/j.celrep.2013.09.018 563.e19. PMID:24139802 https://doi.org/10.1016/j.cell.2018.05.052 16. Mimasu S, Umezawa N, Sato S, Higuchi T, Umehara T, PMID:29937226 Yokoyama S. Structurally designed trans-2- 23. Yin F, Lan R, Zhang X, Zhu L, Chen F, Xu Z, Liu Y, Ye T, phenylcyclopropylamine derivatives potently inhibit Sun H, Lu F, Zhang H. LSD1 regulates pluripotency of histone demethylase LSD1/KDM1. Biochemistry. 2010; embryonic stem/carcinoma cells through histone 49:6494–503. deacetylase 1-mediated deacetylation of histone H4 at https://doi.org/10.1021/bi100299r lysine 16. Mol Cell Biol. 2014; 34:158–79. PMID:20568732 https://doi.org/10.1128/MCB.00631-13 17. Wang J, Lu F, Ren Q, Sun H, Xu Z, Lan R, Liu Y, Ward D, PMID:24190971 Quan J, Ye T, Zhang H. Novel histone demethylase LSD1 24. Hino S, Kohrogi K, Nakao M. Histone demethylase LSD1 inhibitors selectively target cancer cells with controls the phenotypic plasticity of cancer cells. pluripotent stem cell properties. Cancer Res. 2011; Cancer Sci. 2016; 107:1187–92. 71:7238–49. https://doi.org/10.1111/cas.13004 https://doi.org/10.1158/0008-5472.CAN-11-0896 PMID:27375009 PMID:21975933 25. Manuel Iglesias J, Beloqui I, Garcia-Garcia F, Leis O, 18. Maes T, Mascaró C, Tirapu I, Estiarte A, Ciceri F, Vazquez-Martin A, Eguiara A, Cufi S, Pavon A, Lunardi S, Guibourt N, Perdones A, Lufino MM, Menendez JA, Dopazo J, Martin AG. Mammosphere Somervaille TC, Wiseman DH, Duy C, Melnick A, et al. formation in breast carcinoma cell lines depends upon ORY-1001, a Potent and Selective Covalent KDM1A expression of E-cadherin. PLoS One. 2013; 8:e77281. Inhibitor, for the Treatment of Acute Leukemia. Cancer https://doi.org/10.1371/journal.pone.0077281 Cell. 2018; 33:495–511.e12. PMID:24124614 www.aging-us.com 4808 AGING 26. Magee JA, Piskounova E, Morrison SJ. Cancer stem 35. Ginestier C, Hur MH, Charafe-Jauffret E, Monville F, cells: impact, heterogeneity, and uncertainty. Cancer Dutcher J, Brown M, Jacquemier J, Viens P, Kleer CG, Cell. 2012; 21:283–96. Liu S, Schott A, Hayes D, Birnbaum D, et al. ALDH1 is a https://doi.org/10.1016/j.ccr.2012.03.003 marker of normal and malignant human mammary PMID:22439924 stem cells and a predictor of poor clinical outcome. Cell Stem Cell. 2007; 1:555–67. 27. Visvader JE, Lindeman GJ. Cancer stem cells: current status and evolving complexities. Cell Stem Cell. 2012; https://doi.org/10.1016/j.stem.2007.08.014 10:717–28. PMID:18371393 https://doi.org/10.1016/j.stem.2012.05.007 36. Korkaya H, Paulson A, Iovino F, Wicha MS. HER2 PMID:22704512 regulates the mammary stem/progenitor cell population driving tumorigenesis and invasion. 28. Weiswald LB, Bellet D, Dangles-Marie V. Spherical Oncogene. 2008; 27:6120–30. cancer models in tumor biology. Neoplasia. 2015; https://doi.org/10.1038/onc.2008.207 17:1–15. PMID:18591932 https://doi.org/10.1016/j.neo.2014.12.004 PMID:25622895 37. Jiang G, Zhang S, Yazdanparast A, Li M, Pawar AV, Liu Y, Inavolu SM, Cheng L. Comprehensive comparison of 29. Brooks MD, Burness ML, Wicha MS. Therapeutic molecular portraits between cell lines and tumors in Implications of Cellular Heterogeneity and Plasticity in breast cancer. BMC Genomics. 2016 (Suppl 7); 17:525. Breast Cancer. Cell Stem Cell. 2015; 17:260–71. https://doi.org/10.1186/s12864-016-2911-z https://doi.org/10.1016/j.stem.2015.08.014 PMID:27556158 PMID:26340526 38. Neve RM, Chin K, Fridlyand J, Yeh J, Baehner FL, Fevr T, 30. Martin-Castillo B, Lopez-Bonet E, Cuyàs E, Viñas G, Clark L, Bayani N, Coppe JP, Tong F, Speed T, Spellman Pernas S, Dorca J, Menendez JA. Cancer stem cell- PT, DeVries S, et al. A collection of breast cancer cell driven efficacy of trastuzumab (Herceptin): towards a lines for the study of functionally distinct cancer reclassification of clinically HER2-positive breast subtypes. Cancer Cell. 2006; 10:515–27. carcinomas. Oncotarget. 2015; 6:32317–38. https://doi.org/10.1016/j.ccr.2006.10.008 https://doi.org/10.18632/oncotarget.6094 PMID:17157791 PMID:26474458 39. Kao J, Salari K, Bocanegra M, Choi YL, Girard L, Gandhi 31. Bhat-Nakshatri P, Goswami CP, Badve S, Sledge GW Jr, J, Kwei KA, Hernandez-Boussard T, Wang P, Gazdar AF, Nakshatri H. Identification of FDA-approved drugs Minna JD, Pollack JR. Molecular profiling of breast targeting breast cancer stem cells along with cancer cell lines defines relevant tumor models and biomarkers of sensitivity. Sci Rep. 2013; 3:2530. provides a resource for cancer gene discovery. PLoS https://doi.org/10.1038/srep02530 One. 2009; 4:e6146. PMID:23982413 https://doi.org/10.1371/journal.pone.0006146 32. Leis O, Eguiara A, Lopez-Arribillaga E, Alberdi MJ, PMID:19582160 Hernandez-Garcia S, Elorriaga K, Pandiella A, Rezola R, 40. Curtis C, Shah SP, Chin SF, Turashvili G, Rueda OM, Martin AG. Sox2 expression in breast tumours and Dunning MJ, Speed D, Lynch AG, Samarajiwa S, Yuan Y, activation in breast cancer stem cells. Oncogene. 2012; Gräf S, Ha G, Haffari G, et al, and METABRIC Group. 31:1354–65. The genomic and transcriptomic architecture of 2,000 https://doi.org/10.1038/onc.2011.338 breast tumours reveals novel subgroups. Nature. 2012; PMID:21822303 486:346–52. 33. Takanaga H, Tsuchida-Straeten N, Nishide K, Watanabe https://doi.org/10.1038/nature10983 A, Aburatani H, Kondo T. Gli2 is a novel regulator of PMID:22522925 sox2 expression in telencephalic neuroepithelial cells. 41. Pereira B, Chin SF, Rueda OM, Vollan HK, Provenzano Stem Cells. 2009; 27:165–74. E, Bardwell HA, Pugh M, Jones L, Russell R, Sammut SJ, https://doi.org/10.1634/stemcells.2008-0580 Tsui DW, Liu B, Dawson SJ, et al. The somatic mutation PMID:18927476 profiles of 2,433 breast cancers refines their genomic 34. Piva M, Domenici G, Iriondo O, Rábano M, Simões BM, and transcriptomic landscapes. Nat Commun. 2016; Comaills V, Barredo I, López-Ruiz JA, Zabalza I, Kypta R, 7:11479. Vivanco M. Sox2 promotes tamoxifen resistance in https://doi.org/10.1038/ncomms11479 breast cancer cells. EMBO Mol Med. 2014; 6:66–79. PMID:27161491 https://doi.org/10.1002/emmm.201303411 42. Desmedt C, Haibe-Kains B, Wirapati P, Buyse M, PMID:24178749 Larsimont D, Bontempi G, Delorenzi M, Piccart M, www.aging-us.com 4809 AGING Sotiriou C. Biological processes associated with breast https://doi.org/10.21873/anticanres.12604 cancer clinical outcome depend on the molecular PMID:29848686 subtypes. Clin Cancer Res. 2008; 14:5158–65. 51. Ben-Porath I, Thomson MW, Carey VJ, Ge R, Bell GW, https://doi.org/10.1158/1078-0432.CCR-07-4756 Regev A, Weinberg RA. An embryonic stem cell-like PMID:18698033 gene expression signature in poorly differentiated 43. Haibe-Kains B, Desmedt C, Rothé F, Piccart M, Sotiriou aggressive human tumors. Nat Genet. 2008; 40:499– C, Bontempi G. A fuzzy gene expression-based 507. computational approach improves breast cancer https://doi.org/10.1038/ng.127 prognostication. Genome Biol. 2010; 11:R18. PMID:18443585 https://doi.org/10.1186/gb-2010-11-2-r18 52. Gangemi RM, Griffero F, Marubbi D, Perera M, Capra PMID:20156340 MC, Malatesta P, Ravetti GL, Zona GL, Daga A, Corte G. 44. Tibshirani R, Hastie T, Narasimhan B, Chu G. Diagnosis SOX2 silencing in glioblastoma tumor-initiating cells of multiple cancer types by shrunken centroids of gene causes stop of proliferation and loss of tumorigenicity. expression. Proc Natl Acad Sci USA. 2002; 99:6567–72. Stem Cells. 2009; 27:40–48. https://doi.org/10.1073/pnas.082099299 https://doi.org/10.1634/stemcells.2008-0493 PMID:12011421 PMID:18948646 45. Dawson SJ, Rueda OM, Aparicio S, Caldas C. A new 53. Wen Y, Hou Y, Huang Z, Cai J, Wang Z. SOX2 is required genome-driven integrated classification of breast to maintain cancer stem cells in ovarian cancer. Cancer cancer and its implications. EMBO J. 2013; 32:617–28. Sci. 2017; 108:719–31. https://doi.org/10.1038/emboj.2013.19 https://doi.org/10.1111/cas.13186 PMID:23395906 PMID:28165651 46. Russnes HG, Lingjærde OC, Børresen-Dale AL, Caldas C. 54. Yuan S, Norgard RJ, Stanger BZ. Cellular Plasticity in Breast Cancer Molecular Stratification: From Intrinsic Cancer. Cancer Discov. 2019; 9:837–51. Subtypes to Integrative Clusters. Am J Pathol. 2017; https://doi.org/10.1158/2159-8290.CD-19-0015 187:2152–62. PMID:30992279 https://doi.org/10.1016/j.ajpath.2017.04.022 55. Gupta PB, Pastushenko I, Skibinski A, Blanpain C, PMID:28733194 Kuperwasser C. Phenotypic Plasticity: Driver of Cancer 47. Siolas D, Hannon GJ. Patient-derived tumor xenografts: Initiation, Progression, and Therapy Resistance. Cell transforming clinical samples into mouse models. Stem Cell. 2019; 24:65–78. Cancer Res. 2013; 73:5315–19. https://doi.org/10.1016/j.stem.2018.11.011 https://doi.org/10.1158/0008-5472.CAN-13-1069 PMID:30554963 PMID:23733750 56. Jolly MK, Celià-Terrassa T. Dynamics of Phenotypic 48. Byrne AT, Alférez DG, Amant F, Annibali D, Arribas J, Heterogeneity Associated with EMT and Stemness Biankin AV, Bruna A, Budinská E, Caldas C, Chang DK, during Cancer Progression. J Clin Med. 2019; 8:E1542. Clarke RB, Clevers H, Coukos G, et al. Interrogating https://doi.org/10.3390/jcm8101542 open issues in cancer precision medicine with patient- PMID:31557977 derived xenografts. Nat Rev Cancer. 2017; 17:254–68. 57. Wainwright EN, Scaffidi P. Epigenetics and Cancer Stem https://doi.org/10.1038/nrc.2016.140 Cells: Unleashing, Hijacking, and Restricting Cellular PMID:28104906 Plasticity. Trends Cancer. 2017; 3:372–86. 49. Ehmsen S, Pedersen MH, Wang G, Terp MG, Arslanagic https://doi.org/10.1016/j.trecan.2017.04.004 A, Hood BL, Conrads TP, Leth-Larsen R, Ditzel HJ. PMID:28718414 Increased Cholesterol Biosynthesis Is a Key 58. Scaffidi P, Misteli T. Cancer epigenetics: from Characteristic of Breast Cancer Stem Cells Influencing disruption of differentiation programs to the Patient Outcome. Cell Rep. 2019; 27:3927–3938.e6. emergence of cancer stem cells. Cold Spring Harb https://doi.org/10.1016/j.celrep.2019.05.104 Symp Quant Biol. 2010; 75:251–58. PMID:31242424 https://doi.org/10.1101/sqb.2010.75.007 50. Asano M, Matsui J, Towle MJ, Wu J, McGonigle S, DE PMID:21047903 Boisferon MH, Uenaka T, Nomoto K, Littlefield BA. 59. Corominas-Faja B, Cufí S, Oliveras-Ferraros C, Cuyàs E, Broad-spectrum Preclinical Antitumor Activity of López-Bonet E, Lupu R, Alarcón T, Vellon L, Iglesias JM, Eribulin (Halaven®): Combination with Anticancer Leis O, Martín ÁG, Vazquez-Martin A, Menendez JA. Agents of Differing Mechanisms. Anticancer Res. 2018; Nuclear reprogramming of luminal-like breast cancer 38:3375–85. cells generates Sox2-overexpressing cancer stem-like www.aging-us.com 4810 AGING cellular states harboring transcriptional activation of 67. Yang F, Zhang J, Yang H. OCT4, SOX2, and NANOG the mTOR pathway. Cell Cycle. 2013; 12:3109–24. positive expression correlates with poor https://doi.org/10.4161/cc.26173 differentiation, advanced disease stages, and worse PMID:23974095 overall survival in HER2+ breast cancer patients. 60. Vazquez-Martin A, Cufí S, López-Bonet E, Corominas- Oncotargets Ther. 2018; 11:7873–81. https://doi.org/10.2147/OTT.S173522 Faja B, Cuyàs E, Vellon L, Iglesias JM, Leis O, Martín AG, Menendez JA. Reprogramming of non-genomic PMID:30464534 estrogen signaling by the stemness factor SOX2 68. Oak PS, Kopp F, Thakur C, Ellwart JW, Rapp UR, Ullrich enhances the tumor-initiating capacity of breast cancer A, Wagner E, Knyazev P, Roidl A. Combinatorial cells. Cell Cycle. 2013; 12:3471–77. treatment of mammospheres with trastuzumab and https://doi.org/10.4161/cc.26692 salinomycin efficiently targets HER2-positive cancer PMID:24107627 cells and cancer stem cells. Int J Cancer. 2012; 131:2808–19. 61. Iglesias JM, Leis O, Pérez Ruiz E, Gumuzio Barrie J, Garcia-Garcia F, Aduriz A, Beloqui I, Hernandez-Garcia https://doi.org/10.1002/ijc.27595 S, Lopez-Mato MP, Dopazo J, Pandiella A, Menendez PMID:22511343 JA, Martin AG. The Activation of the Sox2 RR2 69. Miki Y, Swensen J, Shattuck-Eidens D, Futreal PA, Pluripotency Transcriptional Reporter in Human Breast Harshman K, Tavtigian S, Liu Q, Cochran C, Bennett LM, Cancer Cell Lines is Dynamic and Labels Cells with Ding W, et al. A strong candidate for the breast and Higher Tumorigenic Potential. Front Oncol. 2014; ovarian cancer susceptibility gene BRCA1. Science. 4:308. 1994; 266:66–71. https://doi.org/10.3389/fonc.2014.00308 https://doi.org/10.1126/science.7545954 PMID:25414831 PMID:7545954 62. Kondo T, Raff M. Chromatin remodeling and histone 70. Rodriguez-Pinilla SM, Sarrio D, Moreno-Bueno G, modification in the conversion of oligodendrocyte Rodriguez-Gil Y, Martinez MA, Hernandez L, Hardisson precursors to neural stem cells. Genes Dev. 2004; D, Reis-Filho JS, Palacios J. Sox2: a possible driver of the 18:2963–72. basal-like phenotype in sporadic breast cancer. Mod https://doi.org/10.1101/gad.309404 Pathol. 2007; 20:474–81. PMID:15574597 https://doi.org/10.1038/modpathol.3800760 63. Wu Y, Wang Y, Yang XH, Kang T, Zhao Y, Wang C, Evers PMID:17334350 BM, Zhou BP. The deubiquitinase USP28 stabilizes LSD1 71. Hellner K, Miranda F, Fotso Chedom D, Herrero- and confers stem-cell-like traits to breast cancer cells. Gonzalez S, Hayden DM, Tearle R, Artibani M, Cell Rep. 2013; 5:224–36. KaramiNejadRanjbar M, Williams R, Gaitskell K, https://doi.org/10.1016/j.celrep.2013.08.030 Elorbany S, Xu R, Laios A, et al. Premalignant SOX2 PMID:24075993 overexpression in the fallopian tubes of ovarian cancer 64. Dontu G, Abdallah WM, Foley JM, Jackson KW, Clarke patients: discovery and validation studies. MF, Kawamura MJ, Wicha MS. In vitro propagation and EBioMedicine. 2016; 10:137–49. transcriptional profiling of human mammary https://doi.org/10.1016/j.ebiom.2016.06.048 stem/progenitor cells. Genes Dev. 2003; 17:1253–70. PMID:27492892 https://doi.org/10.1101/gad.1061803 72. Iglesias JM, Gumuzio J, Martin AG. Linking Pluripotency PMID:12756227 Reprogramming and Cancer. Stem Cells Transl Med. 65. Yu F, Yao H, Zhu P, Zhang X, Pan Q, Gong C, Huang Y, 2017; 6:335–39. Hu X, Su F, Lieberman J, Song E. let-7 regulates self https://doi.org/10.5966/sctm.2015-0225 renewal and tumorigenicity of breast cancer cells. Cell. PMID:28191771 2007; 131:1109–23. 73. Mosteiro L, Pantoja C, Alcazar N, Marión RM, https://doi.org/10.1016/j.cell.2007.10.054 Chondronasiou D, Rovira M, Fernandez-Marcos PJ, PMID:18083101 Muñoz-Martin M, Blanco-Aparicio C, Pastor J, Gómez- López G, De Martino A, Blasco MA, et al. Tissue damage 66. Xu C, Sun X, Qin S, Wang H, Zheng Z, Xu S, Luo G, Liu P, and senescence provide critical signals for cellular Liu J, Du N, Zhang Y, Liu D, Ren H. Let-7a regulates reprogramming in vivo. Science. 2016; 354:aaf4445. mammosphere formation capacity through Ras/NF-κB https://doi.org/10.1126/science.aaf4445 and Ras/MAPK/ERK pathway in breast cancer stem PMID:27884981 cells. Cell Cycle. 2015; 14:1686–97. https://doi.org/10.1080/15384101.2015.1030547 74. Ocampo A, Reddy P, Martinez-Redondo P, Platero- PMID:25955298 Luengo A, Hatanaka F, Hishida T, Li M, Lam D, Kurita M, www.aging-us.com 4811 AGING Beyret E, Araoka T, Vazquez-Ferrer E, Donoso D, et al. stem cells induced from mouse somatic cells by small- In Vivo Amelioration of Age-Associated Hallmarks by molecule compounds. Science. 2013; 341:651–54. Partial Reprogramming. Cell. 2016; 167:1719– https://doi.org/10.1126/science.1239278 1733.e12. PMID:23868920 https://doi.org/10.1016/j.cell.2016.11.052 83. Sun H, Liang L, Li Y, Feng C, Li L, Zhang Y, He S, Pei D, PMID:27984723 Guo Y, Zheng H. Lysine-specific histone demethylase 1 75. Menendez JA, Alarcón T. Senescence-Inflammatory inhibition promotes reprogramming by facilitating the Regulation of Reparative Cellular Reprogramming in expression of exogenous transcriptional factors and Aging and Cancer. Front Cell Dev Biol. 2017; 5:49. metabolic switch. Sci Rep. 2016; 6:30903. https://doi.org/10.3389/fcell.2017.00049 https://doi.org/10.1038/srep30903 PMID:28529938 PMID:27481483 76. Folguera-Blasco N, Cuyàs E, Menéndez JA, Alarcón T. 84. Encinar JA, Fernández-Ballester G, Galiano-Ibarra V, Epigenetic regulation of cell fate reprogramming in Micol V. In silico approach for the discovery of new aging and disease: A predictive computational model. PPARγ modulators among plant-derived polyphenols. PLOS Comput Biol. 2018; 14:e1006052. Drug Des Devel Ther. 2015; 9:5877–95. https://doi.org/10.1371/journal.pcbi.1006052 https://doi.org/10.2147/DDDT.S93449 PMID:29543808 PMID:26604687 77. Folguera-Blasco N, Pérez-Carrasco R, Cuyàs E, 85. Galiano V, Garcia-Valtanen P, Micol V, Encinar JA. Menendez JA, Alarcón T. A multiscale model of Looking for inhibitors of the dengue virus NS5 RNA- epigenetic heterogeneity-driven cell fate decision- dependent RNA-polymerase using a molecular docking making. PLOS Comput Biol. 2019; 15:e1006592. approach. Drug Des Devel Ther. 2016; 10:3163–81. https://doi.org/10.1371/journal.pcbi.1006592 https://doi.org/10.2147/DDDT.S117369 PMID:31039148 PMID:27784988 78. Adamo A, Sesé B, Boue S, Castaño J, Paramonov I, 86. Cuyàs E, Verdura S, Lozano-Sánchez J, Viciano I, Barrero MJ, Izpisua Belmonte JC. LSD1 regulates the Llorach-Parés L, Nonell-Canals A, Bosch-Barrera J, balance between self-renewal and differentiation in Brunet J, Segura-Carretero A, Sanchez-Martinez M, human embryonic stem cells. Nat Cell Biol. 2011; Encinar JA, Menendez JA. The extra virgin olive oil 13:652–59. phenolic oleacein is a dual substrate-inhibitor of https://doi.org/10.1038/ncb2246 catechol-O-methyltransferase. Food Chem Toxicol. PMID:21602794 2019; 128:35–45. 79. Whyte WA, Bilodeau S, Orlando DA, Hoke HA, https://doi.org/10.1016/j.fct.2019.03.049 Frampton GM, Foster CT, Cowley SM, Young RA. PMID:30935952 Enhancer decommissioning by LSD1 during embryonic 87. Cuyàs E, Verdura S, Micol V, Joven J, Bosch-Barrera J, stem cell differentiation. Nature. 2012; 482:221–25. Encinar JA, Menendez JA. Revisiting silibinin as a https://doi.org/10.1038/nature10805 novobiocin-like Hsp90 C-terminal inhibitor: PMID:22297846 computational modeling and experimental validation. 80. Yang P, Wang Y, Chen J, Li H, Kang L, Zhang Y, Chen S, Food Chem Toxicol. 2019; 132:110645. Zhu B, Gao S. RCOR2 is a subunit of the LSD1 complex https://doi.org/10.1016/j.fct.2019.110645 that regulates ESC property and substitutes for SOX2 in PMID:31254591 reprogramming somatic cells to pluripotency. Stem 88. Cuyàs E, Gumuzio J, Lozano-Sánchez J, Carreras D, Cells. 2011; 29:791–801. Verdura S, Llorach-Parés L, Sanchez-Martinez M, Selga https://doi.org/10.1002/stem.634 E, Pérez GJ, Scornik FS, Brugada R, Bosch-Barrera J, PMID:21433225 Segura-Carretero A, et al. Extra Virgin Olive Oil 81. Wang Q, Xu X, Li J, Liu J, Gu H, Zhang R, Chen J, Kuang Contains a Phenolic Inhibitor of the Histone Y, Fei J, Jiang C, Wang P, Pei D, Ding S, Xie X. Lithium, Demethylase LSD1/KDM1A. Nutrients. 2019; 11:E1656. an anti-psychotic drug, greatly enhances the https://doi.org/10.3390/nu11071656 generation of induced pluripotent stem cells. Cell Res. PMID:31331073 2011; 21:1424–35. 89. Schymkowitz J, Borg J, Stricher F, Nys R, Rousseau F, https://doi.org/10.1038/cr.2011.108 Serrano L. The FoldX web server: an online force field. PMID:21727907 Nucleic Acids Res. 2005; 33:W382–8. 82. Hou P, Li Y, Zhang X, Liu C, Guan J, Li H, Zhao T, Ye J, https://doi.org/10.1093/nar/gki387 Yang W, Liu K, Ge J, Xu J, Zhang Q, et al. Pluripotent PMID:15980494 www.aging-us.com 4812 AGING 90. Krieger E, Vriend G. YASARA View - molecular graphics 95. Krieger E, Nielsen JE, Spronk CA, Vriend G. Fast for all devices - from smartphones to workstations. empirical pKa prediction by Ewald summation. J Mol Bioinformatics. 2014; 30:2981–82. Graph Model. 2006; 25:481–86. https://doi.org/10.1093/bioinformatics/btu426 https://doi.org/10.1016/j.jmgm.2006.02.009 PMID:24996895 PMID:16644253 91. Krieger E, Darden T, Nabuurs SB, Finkelstein A, Vriend 96. Corominas-Faja B, Cuyàs E, Lozano-Sánchez J, Cufí S, G. Making optimal use of empirical energy functions: Verdura S, Fernández-Arroyo S, Borrás-Linares I, force-field parameterization in crystal space. Proteins. Martin-Castillo B, Martin ÁG, Lupu R, Nonell-Canals A, 2004; 57:678–83. Sanchez-Martinez M, Micol V, et al. Extra-virgin olive https://doi.org/10.1002/prot.20251 oil contains a metabolo-epigenetic inhibitor of cancer PMID:15390263 stem cells. Carcinogenesis. 2018; 39:601–13. 92. Lionta E, Spyrou G, Vassilatis DK, Cournia Z. Structure- https://doi.org/10.1093/carcin/bgy023 PMID:29452350 based virtual screening for drug discovery: principles, applications and recent advances. Curr Top Med Chem. 97. Cuyàs E, Verdura S, Folguera-Blasco N, Bastidas-Velez 2014; 14:1923–38. C, Martin ÁG, Alarcón T, Menendez JA. Mitostemness. https://doi.org/10.2174/1568026614666140929124445 Cell Cycle. 2018; 17:918–26. PMID:25262799 https://doi.org/10.1080/15384101.2018.1467679 PMID:29886796 93. Morris GM, Huey R, Olson AJ. Using AutoDock for ligand-receptor docking. Curr Protoc Bioinformatics. 98. Moreno-Manzano V, Rodríguez-Jiménez FJ, Aceña- 2008; Chapter 8:Unit 8.14. Bonilla JL, Fustero-Lardíes S, Erceg S, Dopazo J, https://doi.org/10.1002/0471250953.bi0814s24 Montaner D, Stojkovic M, Sánchez-Puelles JM. PMID:19085980 FM19G11, a new hypoxia-inducible factor (HIF) modulator, affects stem cell differentiation status. J 94. Salentin S, Schreiber S, Haupt VJ, Adasme MF, Schroeder M. PLIP: fully automated protein-ligand interaction Biol Chem. 2010; 285:1333–42. profiler. Nucleic Acids Res. 2015; 43:W443–7. https://doi.org/10.1074/jbc.M109.008326 https://doi.org/10.1093/nar/gkv315 PMID:19897487 PMID:25873628 www.aging-us.com 4813 AGING SUPPLEMENTARY MATERIALS Supplementary Figure Supplementary Figure 1. Iadademstat inhibits LSD1 activity. Dose-response curves of LSD1 demethylation activity were created by plotting AlphaScreen signals as the function of iadademstat concentration. Circles and error bars represent mean values and S.D., respectively. Data are representative of two independent experiments (*P < 0.05). www.aging-us.com 4814 AGING