Lysine-SpecificDemethylase1MediatesAKTActivityand ......Ultra DNA Library Prep Kit for Illumina (NEB) as per the manu-facturer's protocol. Chromatin-bound fraction and whole-cell

MOLECULAR CANCER RESEARCH | CANCER GENES AND NETWORKS

Lysine-Specific Demethylase 1 Mediates AKTActivity andPromotes Epithelial-to-Mesenchymal Transition inPIK3CA-Mutant Colorectal CancerSamuel A. Miller1,2, Robert A. Policastro1, Sudha S. Savant2, Shruthi Sriramkumar2, Ning Ding2,Xiaoyu Lu3,4, Helai P. Mohammad5, Sha Cao4, Jay H. Kalin6,7, Philip A. Cole6,7, Gabriel E. Zentner1,8, andHeather M. O’Hagan2,8

ABSTRACT◥

Activation of the epithelial-to-mesenchymal transition (EMT)program is a critical mechanism for initiating cancer progressionand migration. Colorectal cancers contain many genetic and epi-genetic alterations that can contribute to EMT.Mutations activatingthe PI3K/AKT signaling pathway are observed in >40% of patientswith colorectal cancer contributing to increased invasion andmetastasis. Little is known about how oncogenic signaling pathwayssuch as PI3K/AKT synergize with chromatin modifiers to activatethe EMT program. Lysine-specific demethylase 1 (LSD1) is achromatin-modifying enzyme that is overexpressed in colorectalcancer and enhances cell migration. In this study, we determine thatLSD1 expression is significantly elevated in patients with colorectalcancer with mutation of the catalytic subunit of PI3K, PIK3CA,compared with patients with colorectal cancer with WT PIK3CA.LSD1 enhances activation of the AKT kinase in colorectal

cancer cells through a noncatalytic mechanism, acting as ascaffolding protein for the transcription-repressing CoRESTcomplex. In addition, growth of PIK3CA-mutant colorectal cancercells is uniquely dependent on LSD1. Knockdown or CRISPRknockout of LSD1 blocks AKT-mediated stabilization of theEMT-promoting transcription factor Snail and effectivelyblocks AKT-mediated EMT and migration. Overall, we uniquelydemonstrate that LSD1 mediates AKT activation in response togrowth factors and oxidative stress, and LSD1-regulated AKTactivity promotes EMT-like characteristics in a subset ofPIK3CA-mutant cells.

Implications: Our data support the hypothesis that inhibitorstargeting theCoREST complexmay be clinically effective in patientswith colorectal cancer harboring PIK3CA mutations.

IntroductionCancer cells have numerous genetic and epigenetic alterations that

contribute to tumor formation, progression, and therapy resistance.One role of chromatin-modifying complexes is to maintain cellularidentity by chemically modifying amino acid residues on histone andnonhistone substrates. Lineage-specific transcriptional networksmaintained by these complexes provide context for specific mutationsthat determines the phenotypic outcome of the mutation (1). Thisinformation can be leveraged to identify how chromatin modifiers

synergize with specific driver mutations in tissue-specific tumormodels and discover synthetic–lethal relationships.

Mutations in the PI3K/AKT pathway are critical for invasiveproperties and malignant transformation in colorectal cancer (2). Thetwo most common genetic PI3K/AKT pathway alterations in colo-rectal cancer are activating mutations in the PI3K catalytic subunitgene PIK3CA or loss of the pathway suppressor PTEN (3). Mutationsin PIK3CA occur in roughly 25% of patients with colorectal cancer (4)and have been functionally implicated in epithelial-to-mesenchymaltransition (EMT), migration, and chemoresistance (5).While aberrantactivation of the PI3K/AKT pathway has been implicated in colorectalcancer progression, single-nucleotide PIK3CAmutations that activatethe PI3K/AKT pathway are not significantly associated with altera-tions in patient survival (6). These findings indicate that PI3K path-way–activating mutations may require additional factors for fullactivation of the pathway. Recently, the lysine demethylase JMJD2Awas found to be critical for steps involved in activation of AKT,including the recruitment of AKT to the cell membrane and phos-phorylation of AKT at threonine 308 (7, 8). These studies suggest thataberrant overexpression of chromatin-modifying proteins can furtheractivate the PI3K/AKT pathway and therefore may work synergisti-cally with PIK3CA mutations. Little is known with regard to howchromatin modifiers function in the context of PIK3CA mutation tomediate tumorigenic processes in the gut.

The chromatin modifier lysine-specific demethylase 1 (LSD1) isoverexpressed in colorectal cancer and positively correlates withadvanced tumor staging (9). LSD1 is functionally linked to EMT-likechanges and invasion in colorectal cancer (10–12). LSD1 is a memberof the RE1 silencing transcription factor corepressor (CoREST) com-plex (13), which also contains the scaffolding protein RCOR1 andother chromatin-modifying subunits, including histone deacetylase 1

1Genome, Cell, and Developmental Biology, Department of Biology, IndianaUniversity Bloomington, Bloomington, Indiana. 2Medical Sciences Program,Indiana University School of Medicine, Bloomington, Indiana. 3Center forComputational Biology and Bioinformatics, Department of Biostatistics, IndianaUniversity School of Medicine, Indianapolis, Indiana. 4Department of BiohealthInformatics, Indiana University�Purdue University, Indianapolis, Indiana. 5Epi-genetics Research Unit, Oncology, GlaxoSmithKline, Collegeville, Pennsylvania.6Division of Genetics, Department of Medicine, Brigham andWomen's Hospital,Harvard Medical School, Boston, Massachusetts. 7Department of BiologicalChemistry and Molecular Pharmacology, Harvard Medical School, Boston,Massachusetts. 8Indiana University Melvin and Bren Simon Cancer Center,Indianapolis, Indiana.

Note: Supplementary data for this article are available at Molecular CancerResearch Online (http://mcr.aacrjournals.org/).

Corresponding Author: Heather M. O’Hagan, Indiana University School ofMedicine Bloomington, 1001 East 3rd Street, Bloomington, IN47405. Phone: 812-855-3035; Fax: 812-855-4436; E-mail: [email protected]

Mol Cancer Res 2020;18:264–77

doi: 10.1158/1541-7786.MCR-19-0748

�2019 American Association for Cancer Research.

AACRJournals.org | 264

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

and 2 (HDAC1/2; refs. 14, 15). LSD1 and HDAC1/2 within CoRESTdemethylate and deacetylate active chromatin, respectively, to main-tain a repressive chromatin state. In some cellular contexts, LSD1, as amember of CoREST, demethylates di-methyl Histone H3 Lysine 4(H3K4me2) at the promoter of epithelial genes to drive colorectalcancer (10–12). Recent studies, however, have highlighted catalysis-independent functions for LSD1, where it instead acts as a scaffold forthe CoREST complex to maintain transcriptional repression of line-age-specific genes (16, 17). For example, CoREST can confine expres-sion of neuronal genes to neuronal cells bymediating their silencing innonneuronal cell types through the recruitment ofCoREST(14, 15, 18).Furthermore, mechanistic studies of LSD1 catalytic inhibitors inSCLC (19), AML (20, 21), and erythroleukemia (22) demonstrate thatthese inhibitors reactivate gene expression and alter processes such assurvival, proliferation, and differentiation by disrupting the recruit-ment of CoREST to chromatin by SNAG domain transcription factorsas opposed to inhibiting LSD1 demethylase activity. These studiesfurther support the notion that noncatalytic LSD1 functions are criticalfor tumorigenesis.

We hypothesize that LSD1 overexpression synergizes with PIK3CAmutation to enhance invasive phenotypes in colorectal cancer. In thisstudy, we demonstrate that LSD1 is significantly overexpressed inpatients harboring PIK3CA mutations in the gut, but not in cancersarising from other tissues. This observation is functionally significantas we demonstrate that PIK3CA-mutant colorectal and stomachcancer cells exhibit reduced growth after perturbation of LSD1. Wefurther find that LSD1 regulates activation of AKT at the level ofphosphorylation at serine 473 and EMT characteristics downstream ofactive AKT through a noncatalytic scaffolding role in the CoRESTcomplex. Altogether, we illustrate a paradigm wherein LSD1 syner-gizes with a specific PIK3CAmutation to enhance EMT characteristicsand migration.

Materials and MethodsCell culture and treatments

All cell lines were maintained in a humidified atmosphere with 5%CO2. Our study included five colon cell lines (HT29, SW480, HCT116,LoVo, and RKO) and one stomach cell line (AGS). HT29, SW480,HCT116, and LoVo cells were cultured inMcCoy 5Amedia (Corning)and RKO and AGS were cultured in RPMI1640 media (Corning)supplemented with 10% FBS (Gibco). All cell lines were purchasedfrom the ATCC and authenticated and tested for Mycoplasma byIDEXX on June 20, 2019. All cells used in experiments were passagedfewer than 15 times withmost being passaged fewer than 10 times. Forhydrogen peroxide (H2O2) treatments, 30%H2O2 (Sigma) was dilutedin PBS immediately prior to treatment at 250 mmol/L for 1 hour at37�C. For EGF treatments, cells were starved in media lacking serumfor 48 hours prior to treatment. Cells were then treatedwith 100 ng/mLrecombinant EGF (R&D Systems, 236-EG) for 48 hours. GSK-LSD1(Sigma, SML1072), GSK690693 (Sigma, SML0428), and corin(generously provided by P. Cole and J. Kalin) were solubilized inDMSO (Sigma) prior to treatment. Treatment dosages and durationsare defined in the figure legends.

Knockdown, knockout, and transient transfectionsLSD1 (KDM1A; TRCN0000327856), RCOR1 (TRCN0000128570),

and HDAC1 (TRCN0000195467, TRCN0000195103) knockdown(KD) constructs were purchased from Sigma-Aldrichmission shRNA;empty plasmid was used as a vector control. Lentiviral-mediated KDswere performed as described previously (23). CRISPR/Cas9 LSD1 KO

plasmid (sc-430289) and LSD1 HDR plasmid (sc-430289-HDR) werepurchased from Santa Cruz Biotechnology and knockout was per-formed according to manufacturer's protocol. Individual LSD1 KOclones were isolated for experiments, or mixed population KOs wereused, as defined in the figure legends. Cells were transfected withLipofectamine 3000 (Invitrogen) per the manufacturer's protocol.N-terminus HA-tagged LSD1 plasmid was purchased from SinoBiologicals (HG13721-NY) with pCMV3-N-HA used as a negativecontrol vector.

Site-directed mutagenesisMutagenesis was performed according to manufacturer's protocol

(NEB, E0554) using HA-LSD1 plasmid as template, and confirmed viaSanger sequencing by Eurofins Scientific. LSD1 K661A substitutionprimers were generated using NEBaseChanger.

Primer sequencesThe following primer sequences were used: forward, CAACCT-

TAACgcGGTGGTGTTGTG; reverse, CCAAATCCCATCCTTTGG.Annealing temperature was 58�C.

Chromatin immunoprecipitation sequencingChromatin immunoprecipitation (ChIP) was performed using

Diagenode iDeal ChIP-seq Kit (C01010055, for transcription factors;C01010057, for histone modifications) as per the manufacturer'sprotocol. Libraries were generated for sequencing using the NEBNextUltra DNA Library Prep Kit for Illumina (NEB) as per the manu-facturer's protocol.

Chromatin-bound fraction and whole-cell isolationChromatin-bound (or tight chromatin) fractions were isolated as

described previously (23). For protein isolation, cell pellets were lysedin 4% SDS buffer using a Qiashredder (Qiagen). Relative densitometryfor Western blots was determined using ImageJ software and nor-malized to density of loading controls Lamin-B, b-actin, or HistoneH3. All antibodies used in this study are included in the SupplementaryMaterials and Methods.

Immunofluorescence and imagingA total of 2 � 105 HT29 cells were grown on coverslips at 37�C.

After 48 hours, they were treated with 250 mmol/L H2O2 for 1 hour.The cells were then fixed with 4% paraformaldehyde in PBS for15 minutes at room temperature. Cells were permeabilized with0.5% Triton-X in PBS for 10 minutes at room temperature and thenfixed with 1% BSA in PBST (PBS þ 0.2% Tween 20) for 30 minutes.They were then incubated with anti-LSD1 (1:100) and anti-pAKT(1:100) in 1% BSA in PBST for 1 hour at room temperature. This wasfollowed by incubation with Alexa Fluor anti-rabbit 594 (1:500) andAlexa Fluor anti-mouse 488 (1:1,000) for 1 hour at room temperature.Coverslips were mounted using Prolong Gold Antifade with DAPImolecular probes (Cell Signaling Technology #8961). Images wereacquired using Leica SP8 confocal microscope at a magnificationof 63�. The NA of 63� objective used is 1.4. Images were processedusing ImageJ.

Proliferation assaysAssays were performed using the CellTiter-Glo Luminescent Cell

Viability Assay (Promega #G7572) as per manufacturer's protocol.Briefly, 1 � 103 cells were plated in 96-well plates and allowed toincubate under standard growth conditions. Luminescent signals weredetected on a SYNERGY H1 microplate reader (BioTek) using Gen 5

LSD1 Promotes EMT via AKT in PIK3CA-Mutant CRC

AACRJournals.org Mol Cancer Res; 18(2) February 2020 265

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

software (v2.09). All plate readings were normalized to a control plateto account for variance in plating density.

Clonogenic growth and migration assaysFive-hundred (HT29) or 1,500 (SW480) single cells were plated and

allowed to culture at 37�C. After 10 days, cells were fixed with 10%formalin and stained with crystal violet. Crystal violet–stained cellswere imaged using a SYNGENE G:BOX and quantified using theGeneSys and GeneTools programs. For migration, 7.5 � 104 cells inserum-freemedia were plated into transwell in 24-well plates (Corning#40578) for 48 hours with media containing 10% FBS at the bottom.Transwell inserts were stained usingHema 3 Stat Pack (Thermo FisherScientific #123-869). Migration inserts were randomized prior tomanual quantification and the outer 5% of the inserts were notincluded during quantification to reduce edge-effect bias. All imageswere taken on an EVOS FL Auto microscope.

RNA isolation for RNA sequencing and quantitative PCRTotal RNAwas isolated using the RNeasyMini Kit (Qiagen, 74104).

For RNA sequencing of empty vector and LSD1KDcells, libraries weregenerated using Illumina TruSeq Stranded mRNA (Illumina,20020594), or as previously described for RNA sequencing of DMSOand GSK-LSD1–treated cells (24). For qPCR, RNA was used togenerate cDNA via reverse transcription (Thermo Fisher Scientific,K1642). cDNA was amplified using gene-specific primers andFastStart Essential DNA Green Master (Roche, 06402712001). Cq

values of nonhousekeeping genes were normalized to GAPDH expres-sion. qPCR primer sequences listed below:

LSD1, forward, GGTGAGCTCTTCCTCTTCTGG;LSD1, reverse, TCGGCCAACAATCACATCGT;SNAI1, forward, CTAGGCCCTGGCTGCTACA;SNAI1, reverse, TGGCACTGGTACTTCTTGACA;GAPDH, forward, GAAGGTCGGAGTCAACGGATTT;GAPDH, reverse, ATGGGTGGAATCATATTGGAC.

Sequencing, The Cancer Genome Atlas, and statistical analysesDetailed description included in the Supplementary Materials and

Methods. All experiments were performed in biological triplicateunless otherwise specified, with representative results displayed. Allquantitative plots depict mean � SD unless otherwise stated.

Data availabilityRNA-sequencing and ChIP-sequencing (ChIP-seq) data have been

deposited in NCBI's Gene Expression Omnibus and are accessiblethrough GEO Series accession number GSE139927.

ResultsKnockdown of LSD1 reduces phosphorylation of AKT S473

To initially determine whether LSD1 works synergistically withPIK3CA, we analyzed The Cancer Genome Atlas (TCGA) patient datafrom seven different cancer types where PIK3CA mutations arecommon. LSD1 expression was significantly higher in PIK3CA-mutant versus wild-type (WT) tumors for gastrointestinal cancerscolon adenocarcinoma (COAD) and stomach adenocarcinoma(STAD; Fig. 1A). There was a trend toward increased LSD1 expressionin PIK3CA-mutant rectal adenocarcinomas (READ) in the modestnumber of cases analyzed. All other tumor types either had nosignificant change (BLCA, LUSC) or a significant decrease (BRCA,HNSC) in LSD1 expression in the presence of a PIK3CA mutationcompared with PIK3CA WT tumors.

The PI3K pathway has been extensively shown tomediate signalingthrough the activation of AKT. AKT is synergistically activated (25) byphosphorylation of threonine 308 by PDK1 (26), and serine 473mediated by mTORC2 as part of the mTORC/Rictor complex (27).Chromatin modifiers, specifically histone-demethylating enzymes,have recently been linked to the regulation of AKT phosphoryla-tion (7, 8). Therefore, we wanted to test the hypothesis that LSD1 maybe a mediator of complete AKT activation. shRNA-mediated KD ofLSD1 resulted in a significant reduction in pS473-AKT relative toempty vector (EV) cells in bothHT29 and SW480 colorectal cancer celllines (Fig. 1B and C). There was no significant change in pT308-AKTin LSD1 KD SW480 cells and we were unable to detect pT308-AKT inHT29 cells, potentially due to low sensitivity. To confirm the effect ofLSD1 on pS473-AKT, we used CRISPR to generate two LSD1 knock-out (KO) clones each in the SW480 and HT29 cell lines. KO of LSD1also resulted in reduction of pS473-AKT (Fig. 1D andE). Importantly,reintroduction of LSD1 was sufficient to rescue pS473-AKT levels inboth cell lines (Fig. 1D and E). To test the function of LSD1 in de novoactivation of AKT, we stimulated pS473-AKT using H2O2 (28). LSD1KD blocked H2O2-dependent pS473-AKT and, conversely, overex-pression of LSD1 enhanced H2O2-dependent pS473-AKT (Fig. 1F).Overexpression of LSD1 in combination with H2O2 reduced totalAKT, potentially due to a feedback mechanism associated withhyperphosphorylation of AKT, as has been shown with some AKTinhibitors (29). LSD1-mediated regulation of pS473-AKT observed byWestern blot was confirmed by immunofluorescence under both basaland H2O2-treated conditions (Fig. 1G). pS473-AKT was reduced inLSD1 KD cells compared with EV. LSD1-deficient cells exhibited nochange in pS473-AKT after H2O2 treatment, while LSD1-proficientcells exhibited H2O2-induced pS473-AKT. To test the alternativehypothesis that mutant PIK3CA was directly causing increased LSD1expression, we inhibited PI3K in PIK3CA-mutant colorectal cancercells, but saw no change in LSD1 levels (Supplementary Fig. S1).Together, these results demonstrate that LSD1 is more highlyexpressed in PIK3CA-mutant colorectal cancer compared with WTPIK3CA colorectal cancer and that LSD1 levels positively correlatewith pS473-AKT, highlighting a molecular connection between LSD1levels and activation of the PI3K/AKT pathway.

LSD1 catalytic activity is not required for its regulation ofpS473-AKT

LSD1 catalyzes demethylation of mono- and dimethylated lysineresidues on histone (13) and nonhistone substrates (30). The onco-genic function of LSD1 in colorectal cancer has previously beenattributed to its ability to demethylate H3K4me2 at the promoter ofepithelial genes, causing their repression and promoting more aggres-sive cellular phenotypes via enhanced EMT (11, 12). In contrast, newlyemerging research has documented noncatalytic oncogenic functionsof LSD1 (16, 17, 19–22) and suggests we need to consider LSD1enzymatic activity in the regulation of AKT activation. To address this,we first performed ChIP-seq of LSD1 in SW480 cells to identify directbinding targets genome-wide, generating a catalog of potential cata-lytic target loci. Consistent with other studies, LSD1 enrichment atgenes was primarily detected near transcription start sites (TSS;ref. 31; Fig. 2A). To generate a list of potential LSD1 targets de novoand confirm the specificity of our antibody, LSD1 peaks were called bynormalizing LSD1 enrichment in parental cells to our LSD1 CRISPRKO. We detected 6,221 LSD1 peaks shared between replicates. UsingDNase sequencing data from the SW480 cell line, we observeddepletion of DNase cleavage centered at LSD1 peaks genome-wide(Fig. 2B), confirming occupancy of these regions. This DNase

Miller et al.

Mol Cancer Res; 18(2) February 2020 MOLECULAR CANCER RESEARCH266

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

footprint is not an artifact of LSD1 enrichment at TSSs, as we do notobserve an overall depletion of DNase cleavages over TSSs (Supple-mentary Fig. S2A). To determine whether LSD1 KD alters genome-

wide histone methylation status, we performed ChIP-seq for the LSD1substrate H3K4me2 in control and LSD1KD SW480 cells. There was aslight decrease in H3K4me2 enrichment genome-wide after LSD1 KD,

Figure 1.

LSD1 regulates phosphorylation of AKT in colorectal cancer. A, Box and whisker plot of fragments per kilobase of transcript per million mapped reads (FPKM)expression values for LSD1 across different TCGAdatasets. Box limits are set at the third andfirst quartile rangewith central line at themedian,withwhiskers depicting1.5 times the interquartile range. Data points outside this range are represented at outliers (black dots). Black and blue outline indicates data for WT and PIK3CA-mutant tumors, respectively. Red and purple fill represent significant increase and decrease in LSD1 expression, respectively, between PIK3CA-mutant and PIK3CAWT tumors. The numbers under the box plots are the number of samples used to generate the plots. Western blot analysis of empty vector (shEV) or LSD1 KD inSW480 (B) or HT29 (C) cells. Arrowhead indicates correct position of pT308-AKT band.Western blot quantified by densitometric analysis and normalized to b-actinand shEV. Results are represented as mean� SD (n¼ 3). Significance was determined by two-tailed Student t test. LSD1 CRISPR KO clones with or without LSD1 OEplasmid (HA-LSD1) in SW480 (D) or HT29 (E) cells analyzed by Western blot analysis. F, EV, LSD1 KD, or LSD1 OE cells treated with 250 mmol/L H2O2 for 1 hour. G,Brightfield and immunofluorescence images of EV or LSD1 KDHT29 cells under untreated or H2O2-treated conditions. A fieldwas selected in theH2O2-treated shLSD1cells to facilitate direct comparison of LSD1-deficient and LSD1-proficient cells. White arrow indicates cells with LSD1 expression and orange arrow indicates cellsdeficient in LSD1 (��� , P < 0.001; ���� , P < 0.0001; ns, not significant).

LSD1 Promotes EMT via AKT in PIK3CA-Mutant CRC

AACRJournals.org Mol Cancer Res; 18(2) February 2020 267

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

in contrast to the increase thatwould be expected on the basis of LSD1'sknownhistone demethylase activity (Fig. 2C). Therewas no significantchange in bulk H3K4me2 after LSD1 KD suggesting the slight changein ChIP-seq may not be biologically relevant (SupplementaryFig. S2B). To identify potential direct transcriptional targets of LSD1,

we performed RNA-seq after LSD1 KD in SW480 cells. We combinedthese transcriptional data with our ChIP-seq datasets to assessH3K4me2 levels at genes with LSD1 enrichment near their promotersand significantly increased expression after LSD1 KD. The promoterregions of genes FBN3, TRPM6, and RASD2 are depicted as

Figure 2.

LSD1 catalytic activity is dispensable for regulation of gene expression and activation of AKT.A,Metageneplot and heatmap depicting ChIP-seq of LSD1 inWT (n¼ 2)or LSD1 KO (n¼ 1) SW480 cells at gene enrichment sites genome-wide. Average plots and heatmaps depicting: LSD1 enrichment peak overlapwith DNase-seq peaksin SW480 (B) and H3K4me2 ChIP-seq signal at TSS enrichment sites genome-wide in shEV and shLSD1 SW480 cells (N¼ 3; C). A–C, Values are derived from CPM(counts per million) normalized reads. E, Differentially expressed genes (DEG) from RNA-seq (log2FC� 1 and FDR� 0.05¼ purple) after 40 nmol/L GSK-LSD1 for48 hours versus DMSO or shLSD1 versus shEV in SW480 cells (N ¼ 3). D, ChIP-seq gene tracks of representative DEGs in LSD1 versus EV KD SW480 cells with orwithout LSD1 promoter enrichment. F, Cells pretreated with DMSO or 40 nmol/L GSK-LSD1 for 48 hours then treated with 250 mmol/L H2O2 for 1 hour.Western blotswere quantified by densitometric analysis and normalized to loading control and DMSO. Graph represents mean � SD; ns, not significant. Significance determinedusing one-way ANOVAwith Tukey multiple comparisons test (N¼ 3). G,Mixed population LSD1 KO cells were transfected with vector control, HA-LSD1, or HA-LSD1(K661A) for 48 hours. Whole-cell extract from untreated and cells treated with 250 mmol/L H2O2 for 1 hour were analyzed by Western blot analysis.

Miller et al.

Mol Cancer Res; 18(2) February 2020 MOLECULAR CANCER RESEARCH268

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

representative loci for this analysis. There were no significant changesin H3K4me2 enrichment at the promoter of FBN3, TRPM6, or RASD2(Fig. 2D), nor at the promoter of any gene following LSD1 KD (datanot shown). CDH1, a published target of LSD1, showed significantlyincreased gene expression (log2FC¼ 1.070, FDR ¼ 0.035) after LSD1KD. However, there was no change in H3K4me2 at the CDH1promoter after LSD1 KD, nor did we detect enrichment of LSD1. Arecent study in PC3 prostate cancer cells demonstrated similar results,wherein levels of H3K4me2 following LSD1 KD only changed at thepromoter of 2 genes (16). Together, these data suggest that LSD1impacts gene expression in SW480 colon cells without demethylatingH3K4me2.

To further characterize LSD1's role in SW480 cells, we treatedthese cells with a highly potent and selective inhibitor,GSK-LSD1 (31) and performed RNA-sequencing. Treatment withthis small molecule had a minimal effect on gene expression (0 geneswith expression significantly altered >Log2FC ¼ 1), whereas LSD1KD caused robust activation of gene expression, consistent with therole of LSD1 as a transcriptional repressor (Fig. 2E). In addition,while H2O2 treatment induced the expected increases inpS473-AKT, inhibiting LSD1 did not significantly alter basal orH2O2-induced pS473-AKT levels (Fig. 2F), with similar results inHT29 cells (Supplementary Fig. S2C). Moreover, catalytically defec-tive K661A-mutant LSD1 was able to rescue pS473-AKT levels inLSD1 KO HT29 cells similarly to WT LSD1, both basally and inresponse to H2O2 (Fig. 2G). Together, these studies stronglysupport the concept that LSD1-mediated transcriptional repressionand regulation of AKT activation are independent of LSD1 catalyticactivity in these colorectal cancer cell lines.

LSD1 regulates pS473-AKT by scaffolding the CoREST complexon chromatin

Our RNA- and ChIP-seq data suggest that LSD1-mediated regu-lation of AKT activation is independent of demethylase activity; wetherefore hypothesized that this regulation may instead occur throughLSD1's stabilizing role in the repressive CoREST complex. To addressthis, we performed a chromatin fractionation assay. KD of LSD1 inSW480 cells led to a significant decrease in whole-cell and chromatin-bound levels of the CoREST scaffold protein RCOR1 (Fig. 3A).HDAC1 and HDAC2 can be recruited to assist in the repressivefunctions of CoREST complexes (32). While whole-cell levels ofHDAC1/HDAC2 remain unchanged after LSD1 KD, there is a sig-nificant reduction in the chromatin binding of HDAC1 (Fig. 3A). Thisresult supports the notion that loss of LSD1 reduces formation of theCoREST complex and therefore reduces HDAC1 recruitment tochromatin. This is consistent with a recent study, where loss of LSD1decreased HDAC1 binding at LSD1-enriched enhancers in leukemiacells (20). There was a slight but insignificant decrease in the recruit-ment of HDAC2 to chromatin (Fig. 3A). It is possible that a smallerproportion of cellular HDAC2 is recruited to CoREST complexes incomparison with HDAC1. RCOR1 is one of three CoREST isoformsthat interact with LSD1 (33). RCOR contains a SANT domain thatenhances the complex's interaction with histones and therefore playsan important scaffolding role for the CoREST complex (34). Todetermine whether the CoREST complex functions in the regulationof AKT activation, we performed KD of RCOR1 using shRNA andinduced pS473-AKTusingH2O2. KDof RCOR1 inHT29 cells reducedbasal levels of pS473-AKT and blocked H2O2-induced pS473-AKT(Fig. 3B). In addition, there was a decrease in whole-cell levels of LSD1protein, consistent with the scaffolding function ascribed to RCOR1(Fig. 3B).

To our knowledge, HDAC1 does not function as a scaffold in theCoREST complex, but instead contributes to the formation of arepressive chromatin environment by catalyzing the removal of acetylgroups from histone lysine residues. To determine whether loss ofHDAC1 alone is sufficient to block AKT activation, we knocked downHDAC1 using two independent shRNAs and stimulated AKT usingH2O2. HDAC1 KD had no effect on H2O2-induced pS473-AKT orwhole-cell levels of LSD1 in HT29 (Fig. 3C) or SW480 cells (Supple-mentary Fig. S3A). Overall, KDs of proteins that stabilize the CoRESTcomplex (LSD1 or RCOR1) were sufficient to perturb AKT activationwhile KD of a protein (HDAC1) that interacts with the core CoRESTcomplex had no effect on AKT activation, indicating that an intact,chromatin-associated core CoREST complex is required for fullactivation of AKT.

Recently, a single-molecule hybrid inhibitor targeting multiplechromatin modifiers within the CoREST complex was synthesized.The compound corin contains an LSD1 inhibitor (tranylcypromineanalogue) fused to a class 1 HDAC inhibitor (MS-275) and exhibitsnear irreversible inhibition of the CoREST complex (ref. 35; Fig. 3D).In vitro and cellular studies demonstrated this compound is selectivefor complexes containing HDAC1 and LSD1 over complexes that onlycontain HDAC1. Corin treatment led to a dose-dependent reductionin pS473-AKT, with the effect increasing over time (Fig. 3E). Con-sistent with HDAC1 inhibition, global H3K9Ac levels increased aftertreatment with corin. At similar concentrations, MS-275 alone doesnot reduce pS473-AKT (36).

To test whether these observations mirror the patient data availablethrough the TCGA, we investigated RCOR1 expression in PIK3CAWT and mutant tumors. Like LSD1, RCOR1 showed significantlyhigher levels of expression in PIK3CA mutant compared with WT ingastrointestinal cancer types COAD, READ, and STAD (Fig. 3F).Generally, there was no significant association between RCOR1expression and PIK3CAmutation in nongastrointestinal cancer types(Supplementary Fig. S3B). However, there was no significant corre-lation observed between LSD1 and RCOR1 expression in COAD andREAD datasets, but there was a significant correlation in the STADdataset (Supplementary Fig. S3C). Consistent with the idea that onlyLSD1 or RCOR1 need to be overexpressed to stabilize the other, highexpression of LSD1 or RCOR1 in COADREAD patient data wassignificantly associated with PIK3CA mutation compared withpatients with low expression of both LSD1 and RCOR1, and themagnitude of change was significantly greater than expected byrandom chance from permutation testing (Fig. 3G). Altogether thesedata suggest that LSD1 regulates AKT activation through a scaffoldingfunction for the CoREST complex. An intact chromatin-bound CoR-EST complex is required to mediate positive regulation of AKTactivation via regulation of gene expression. Furthermore, high expres-sion of CoREST core members LSD1 or RCOR1 is significantlyassociated with the presence of PIK3CA mutation.

LSD1 regulates EMT-associated gene programs in PIK3CA-mutant cells

On the basis of our observations from clinical datasets (Fig. 1A), wetested a phenotypic link between PIK3CA mutational status andcellular levels of LSD1. We selected six cell lines for this analysis: fivecolon cancer lines (SW480, HT29, LoVo, HCT116, and RKO) and onestomach cancer line (AGS). SW480 and LoVo are PIK3CA WT,whereas HT29, AGS, HCT116, and RKO all carry PI3K-activatingPIK3CA mutation(s). The cell lines selected have several other com-mon colorectal cancer mutations, but none are unique to the PIK3CA-mutant or WT cells (Supplementary Table S1; ref. 37). LSD1 KD had

LSD1 Promotes EMT via AKT in PIK3CA-Mutant CRC

AACRJournals.org Mol Cancer Res; 18(2) February 2020 269

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

no effect on the growth of PIK3CA WT cell lines (Fig. 4A). Interest-ingly, LSD1 KD caused a significant reduction in growth in allPIK3CA-mutant cell lines. Similar results were obtained with LSD1CRISPR knockout, where LSD1 KO significantly reduced growth inPIK3CA-mutant HT29 cells, but had no effect on the growth ofPIK3CAWT SW480 cells (Supplementary Fig. S4A–S4D). These data

suggest that PIK3CA-mutant cancers may be sensitive to targeting ofLSD1, and raise the unique possibility that LSD1may be important forthe tumorigenic activity of PIK3CA mutations.

Because perturbing LSD1 abrogates activation ofAKT in bothHT29(PIK3CA mutant) and SW480 (PIK3CA WT) cells, but only reducesthe viability of HT29 cells, we hypothesized that AKTmay be uniquely

Figure 3.

LSD1 regulates AKT activation via scaffolding of the CoREST complex on chromatin.A, Chromatin affinity assay performed in shEV or shLSD1 withwhole-cell extract(WCE) or chromatin-bound fraction. Western blots were quantified by densitometric analysis and normalized to loading control and shEV fraction. Significancedeterminedby two-wayANOVAwith Sidakmultiple comparisons test (n¼ 3). Graphs depictmean�SD (�� ,Padj<0.01; ���� ,Padj<0.0001; ns, not significant).B, shEVor shRCOR1 cells treated with 250 mmol/L H2O2 for 1 hour and analyzed byWestern blot analysis. C, shEV or shHDAC1 cells treated as inB.D,Model for corin inhibitormode of action. E,Cells treatedwith DMSOor 3, 5, or 7mmol/L corin over time course and analyzed byWestern blot analysis. F,Box andwhisker plot of fragments perkilobase of transcript per million mapped reads (FPKM) expression values for RCOR1 across different TCGA datasets. Box limits are set at the third and first quartilerangewith central line at themedian, withwhiskers depicting 1.5 times the interquartile range. Data points outside this range are represented as outliers (black dots).Black andblue outline indicates data forWTandPIK3CA-mutant tumors, respectively. Redfill represent significant increase in LSD1 expression, respectively, betweenPIK3CA-mutant and PIK3CAWT tumors. G, Fraction of patients with PIK3CAmutation separated by high expression of LSD1 or RCOR1 versus low expression of bothLSD1 and RCOR1. Plotted below is bootstrapped 90% confidence interval for the mean difference in fraction for patients with the PIK3CAmutation between the twogroups. Significance determined by permutation test (��� , P < 0.001).

Miller et al.

Mol Cancer Res; 18(2) February 2020 MOLECULAR CANCER RESEARCH270

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

Figure 4.

Gastrointestinal cell lines with mutant PIK3CA are sensitive to LSD1 KD. A, SW480, LoVo, HT29, AGS, HCT116, and RKO cell growth over a 5-day time coursedetermined by the CellTiter-Glo Luminescent Cell Viability Assay (n¼ 4). Graph depicts meanþ SD. Statistical analyses are performed using two-way ANOVA andSidak multiple comparisons test with all statistically significant comparisons shown. �� , Padj < 0.01; ���� , Padj < 0.0001. B, Correlation plot of RNA-seq data from LSD1versus EV KD in SW480 and HT29 cells. Significant data points are defined as abs(Log2FC) � 1 and FDR � 0.05 for LSD1 compared with EV KD for each cell line,including those unique to HT29 (purple), unique to SW480 (red), and shared between HT29 and SW480 (black; n ¼ 3). C, Ridge plot depicts LSD1 versus EV KDexpression changes of genes contributing tomaxenrichment score ofHallmark and curatedgene sets accessed from themolecular SignaturesDatabase (v6.2). GSEARatio shows the ratio of genes contributing tomaxenrichment score, to the total number of genes in thegene set. Heatmapon right shows Log2FCvalue for eachgeneenriched in HALLMARK_PI3K_AKT_MTOR_SIGNALING in either HT29 or SW480 cells sorted by HT29 Log2FC.D,Network analysis of uniquely upregulated genes inHT29 after LSD1 KDwith significantly enriched processes from the Reactome database. Similar termsweremanually grouped and size and color of circles were set toindicate number of genes and the P value, respectively. E, Clonogenic growth assay. Significance determined by two-way ANOVA with Sidak multiple comparisonstest (n ¼ 3). Results are represented as mean � SD (�� , Padj < 0.01; ns, not significant).

LSD1 Promotes EMT via AKT in PIK3CA-Mutant CRC

AACRJournals.org Mol Cancer Res; 18(2) February 2020 271

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

targeting pathways associated with cell survival in HT29 but notSW480 cells. To test this hypothesis, we performed RNA-seq in theEV and LSD1 KD HT29 cells and generated datasets that consisted ofgenes with significantly altered expression in LSD1 KD versus EV foreach cell line.When comparing these two datasets, there was very littleoverlap between genes upregulated by LSD1KD, exemplified by aweakpositive correlation (R2 ¼ 0.121), suggesting that the majority ofsignificant changes were unique between the two cell lines(Fig. 4B). The largest set of genes with significantly altered expressionafter LSD KD compared with EV by >3-fold was genes uniquelyupregulated in HT29 cells. The only commonly downregulated genebetween the datasets was LSD1, further validating the specificity of ourKD. To identify conserved and divergent characteristics between thetwo datasets, we performed gene-set enrichment analysis (Fig. 4C). Asvalidation, genes upregulated after LSD1KD in an independent datasetwere enriched and upregulated by LSD1 KD in bothHT29 and SW480cells as determined by RNA-seq. In agreement with our finding thatLSD1 KD reduces recruitment of HDACs to chromatin, genes down-regulated after HDAC1 and HDAC2 overexpression were also signif-icantly enriched and upregulated in our data. In support of our findingthat LSD1 is critical for AKT activation, genes upregulated duringactive PI3K/AKT signaling were significantly enriched in our HT29dataset, and trended toward downregulated after LSD1 KD in bothSW480 and HT29 RNA-seq datasets, indicating reduced active AKTsignaling. Sixty-five PI3K/AKT signaling genes were downregulatedbetween the two cell lines.While 51%were commonly downregulated,69% of the remaining genes were uniquely downregulated in HT29cells indicating LSD1 KD may have a more significant effect on geneexpression in this pathway in HT29 cells. Furthermore, by geneontology analysis of genes uniquely upregulated after LSD1 KD inSW480 cells, no enriched genesets were associated with prolifera-tion indicating that there is likely not a compensatory pathwayactivated in SW480 cells to buffer loss of AKT activation (Supple-mentary Fig. S4E). In addition, genes downregulated by TGFb1were enriched and upregulated in HT29 cells after LSD1 KD(Fig. 4C). This is interesting because cross-talk between thePI3K/AKT and TGFb signaling networks plays a critical role intumor progression (38).

We next narrowed our focus to genes that were uniquely upregu-lated in the HT29 cells after LSD1 KD (N ¼ 523) and performedhypergeometric enrichment analysis using the Reactome database. Inagreement with another study in colorectal cancer cells, ontologiesassociated with immune response were enriched (39). We additionallyidentified pathways associated with extracellular matrix organizationand cellular junction organization, one example being the formation ofhemidesmosomes, which play a role in epithelial cell adherence toextracellular matrices and are lost during EMT (ref. 40; Fig. 4D).Among genes uniquely downregulated after LSD1 KD in HT29 cells,were mediators of EMT-associated elastic fibre formation, TGFB3and LOXL1 (data not shown). Together, these data suggest thatLSD1 regulates EMT characteristics in HT29 cells. Clonogenicassays test the ability of a tumor cell to survive and grow inisolation, an important characteristic of EMT and stem-like cells,which may be a critical attribute for metastasis (41, 42). HT29 cellsexhibited >7-fold reduction in clonogenic survival after LSD1 KD,while there was no significant loss of clonogenic survival inPIK3CA WT SW480 cells (Fig. 4E). Together, these data demon-strate that LSD1 KD causes a significant reduction in growth andsurvival as well as alterations in gene expression programs relatedto EMT in PIK3CA-mutant, but not PIK3CA WT colorectal cancercell lines.

LSD1 regulates protein stability of Snail through regulationof AKT

AKT has been extensively implicated in the EMT process throughprotein stabilization of the EMT-promoting transcription factor Snailfamily Transcriptional Repressor 1 (SNAI1 or Snail). Snail can bephosphorylated by GSK3b leading to its ubiquitination and subse-quent proteasomal degradation (43). Active AKT can phosphorylateGSK3b at serine 9 (44), inhibiting its kinase activity toward Snail,and increasing the protein half-life of Snail. However, the cellularcontexts that support the AKT–GSK3b–Snail axis are not wellunderstood. While the downstream functional outcomes of PIK3CAmutations in different domains are poorly characterized, one studysuggests they may alter different downstream pathways, and requireunique cofactors to exert their oncogenic effects (45). Furthermore,PIK3CA mutations in the p85, C2, helical, and kinase domains allincrease PIK3CA lipid kinase activity, but only mutations in C2,helical, and kinase domains are sufficient to robustly increasepS473-AKT and transform cells (46). The exact functions of C2domain mutations in driving tumorigenesis have remained largelyuncharacterized, as researchers have focused primarily on morecommonly occurring kinase and helical domain mutations. Wehypothesize that the activity of the AKT–GSK3b–Snail axis incolorectal cancer cells is dependent on the PIK3CA mutationalstatus.

To test for the effect ofPIK3CAmutational status onAKT-mediatedstabilization of Snail protein, we inhibited AKT across our cell linepanel using a potent and selective competitive pan-AKT inhibitor,GSK690693 (29). Treating cells with this inhibitor generates a feedbackthat causes hyperphosphorylation of S473-AKT, reduction of AKTprotein levels, and loss of pS9-GSK3b. Inhibiting AKT did not changeor caused an increase in Snail protein levels in PIK3CA WT (SW480/LoVo) or PIK3CA kinase–mutant (HCT116/RKO) cells, respectively(Supplementary Fig. S5A–S5D). Inhibiting AKT in PIK3CA C2(HT29) or C2 and helical (AGS) mutant cells led to a markeddecrease in the protein levels of Snail (Fig. 5A and B) independentof significant changes in the mRNA levels of Snail (Fig. 5C). Using theTCGA pancancer datasets, we stratified seven different cancertypes based on the domain frequency of PIK3CA mutations(Fig. 5D). While mutations in the C2 domain represented onaverage 10.6% of PIK3CAmutations across the different cancer types,mutations in gastrointestinal cancers of the stomach and colon/rectumwere disproportionately higher at 22.2% and 14%, respectively.Together, these results suggest that activation of the AKT–GSK3b–Snail axis may depend on PIK3CA mutational status and correlateswith mutations in the C2 domain which are more prevalent in somegastrointestinal cancers.

We next sought to determine whether LSD1 enhances EMT-likechanges in HT29 cells through positive regulation of the AKT–GSK3b–Snail axis. LSD1 KD resulted in a strong reduction in Snailprotein levels (Fig. 5E) that was independent of significant changes inSNAI1 mRNA (Fig. 5F). As expected, changes in Snail protein levelwere accompanied by decreases in pS473-AKT and pS9-GSK3b.Treatment of cells with proteasome inhibitor MG-132 was sufficientto rescue Snail protein levels after LSD1 KD (Fig. 5E). While reexpres-sion of LSD1 in LSD1 KD cells was sufficient to rescue protein levels ofSnail, reexpressing LSD1 in the context of AKT inhibition was notsufficient to rescue Snail protein levels, further supporting our hypoth-esis that LSD1 is acting upstream of AKT in the stabilization of Snail(Fig. 5G). LSD1did not significantly effect AKTactivation inAGS cellsnor did LSD1 KD significantly reduce Snail protein level, supportingthe notion that LSD1-mediated changes in Snail protein rely on LSD1

Miller et al.

Mol Cancer Res; 18(2) February 2020 MOLECULAR CANCER RESEARCH272

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

regulation of AKT (Supplementary Fig. S5E).While LSD1KD reducedpS473-AKT in SW480 cells, LSD1KDhadno significant effect on Snailprotein level (Supplementary Fig. S5F). This result is consistent withour finding that inhibiting AKT has no effect on Snail protein in thiscell line (Supplementary Fig. S5A). In additional cell lines with WT(LoVo) or kinase domain (RKO, HCT116) mutant PIK3CA whereAKT inhibition did not alter Snail levels (Supplementary Fig. S5B–S5D), LSD1KDdid not alter pS473-AKT levels, further suggesting thisaxis is mutation-dependent (Supplementary Fig. S5G–S5I). Together,these data indicate PIK3CAC2 domain mutations may function in theactivation of the AKT–GSK3b–Snail axis. Furthermore, ourmolecularstudies establish LSD1 as a context-dependent upstream regulator ofthis pathway.

The CRC Subtyping Consortium has previously developed a clas-sifier that can be used to subtype cancers into four groups CMS1 (MSI/immune), CMS2 (canonical), CMS3 (metabolic), and CMS4 (mesen-chymal; ref. 47). While there was no significant difference in LSD1expression between CMS1-CMS4 subtyped COADREAD datasets,LSD1 was uniquely upregulated (P < 0.05) in PIK3CA-mutant CMS4

mesenchymal subtype tumors compared with WT (SupplementaryFig. S6A and S6B). This finding suggests our observation that LSD1 isoverexpressed in PIK3CA-mutant versusWT colorectal cancer tumors(Fig. 1A) is at least in part explained by significant differences inCMS4subtyped tumors. Furthermore, we found that PIK3CA C2 domainmutations were most common in CMS4 subtyped tumors (Supple-mentary Fig. S6C). This observation, in addition to our AKT inhibitorstudies, established a connection between PIK3CA C2 domain muta-tions and mesenchymal phenotypes.

LSD1 is required for EGF-induced migration in HT29 cellsEGF has previously been used to enhance migratory phenotypes

through the AKT–GSK3b–Snail axis (48). HT29 cells exhibitedlittle to no basal migration, but upon stimulation with EGFexhibited a significant >30-fold increase in migration (Fig. 6Aand B). Importantly, loss of LSD1, inhibition of AKT, or inhibitionof the CoREST complex completely blocked EGF-induced migra-tion (Fig. 6A and B). Treatment with EGF altered HT29 cellularmorphology, with cells becoming elongated and scattered;

Figure 5.

LSD1 regulates Snail stability via AKT in PIK3CA C2 domain–mutant cancer cells. Western blot analysis of HT29 (A) or AGS cells treated with DMSO or 10 mmol/LGSK690693 for 48 hours (B). C, Real-time PCR analysis of SNAI1 RNA expression levels after 48-hour DMSO or 10 mmol/L GSK690693 treatment in HT29 and AGScells. Expression was normalized to Gapdh and DMSO. Results are represented as mean � SD. Significance determined by two-way ANOVA with Sidak multiplecomparisons test. ns, not significant. D, Proportion of total PIK3CAmutations occurring in the different domains indicated across various cancer types in the TCGApancancer datasets. E, shEV and shLSD1 HT29 cells treated with DMSO or 10 mmol/L MG-132 for 4 hours and analyzed by Western blot analysis. F, Real-time PCRanalysis of LSD1 and SNAI1RNA expression levels in shEV and shLSD1 HT29 cells as inC. ���� , Padj <0.0001.G,HT29 shLSD1 cells were transfectedwith HA-LSD1 aloneor in combination with 10 mmol/L GSK690693.

LSD1 Promotes EMT via AKT in PIK3CA-Mutant CRC

AACRJournals.org Mol Cancer Res; 18(2) February 2020 273

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

however, LSD1 KO abrogated EGF-mediated changes in cellularmorphology (Supplementary Fig. S7). Treatment of cells with EGFled to an increase in levels of pS473-AKT, pS9-GSK3b, and Snailat 24 and 48 hours (Fig. 6C). LSD1 KD completely blockedEGF-induced activation of pS473-AKT, pS9-GSK3b, and increasein Snail protein. LSD1 KD also increased levels of the epithelialmarker Claudin-1. Together, these data suggest that LSD1 isrequired for EGF-induced migration mediated by the AKT–GSK3b–Snail pathway.

DiscussionIn this study, we implicate LSD1 in the regulation of AKT activity.

We find that PIK3CA-mutant colorectal cancer cell lines are sensitiveto LSD1 KD, whereas WT cells are not. Using TCGA data, we provideevidence that this connection between LSD1 and PIK3CA mutationstatus may be unique to gastrointestinal tumor types. PI3K inhibitors,which have been successful in gynecologic and breast cancers withmutant PIK3CA, have failed in colorectal cancer clinical trials,

Figure 6.

LSD1 is required for EGF-inducedmigration of cells with an active AKT–GSK3b–Snail axis.A, 10� Brightfield images of crystal violet–stained HT29 cells after 48-hourmigration through transwell insert.WT, LSD1 KO, 10mmol/L GSK690693, or 3mmol/L corin cellswere cotreatedwith 100 ng/mL EGF for 48 hours.B,Quantification ofmigration normalized to migration counts for untreated cells. Results are represented as mean � SD. Significance was determined by one-way ANOVA with Tukeymultiple comparisons test. All significant comparisons are shown. ����, Padj <0.0001. C, shEV or shLSD1 cells were treatedwith 100 ng/mL EGF for 24 or 48 hours andanalyzed by Western blot. D, Model depicting CoREST complex in the regulation of AKT.

Miller et al.

Mol Cancer Res; 18(2) February 2020 MOLECULAR CANCER RESEARCH274

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

suggesting that additional factors influence the PI3K pathway incolorectal cancer. We demonstrate for the first time that LSD1regulates basal AKT activation as well as activation in response toboth exogenous oxidative stress and growth factors and, in a subset ofPIK3CA-mutant cells, LSD1-mediated AKT activity promotes EMT-like characteristics, including migration.

Our findings suggest that LSD1may play a role in the full activationof AKT. We highlight, for the first time, an oncogenic function forLSD1 in the regulation of PI3K/AKT signaling via catalytically inde-pendent regulation of gene expression. We hypothesize that theCoREST complex represses gene(s) that normally function to regulateS473-AKT phosphorylation such as regulators of cell receptors, neg-ative regulators of upstream kinases, and/or phosphatases. When thecomplex is disrupted, these gene(s) are expressed and perturb AKTphosphorylation. Our RNA-seq data in HT29 cells confirm that LSD1KD decreases PI3K/AKT pathway activity. However, it is unclear fromthese datasets how S473-AKT phosphorylation is regulated. Futurestudies will determine the mechanisms by which CoREST complexesinfluence pS473-AKT.

The canonical function of LSD1 is to catalyze the demethylation ofhistones to repress gene expression (13). Previously, a study has shownusing endometrial cancer cells that during estrogen signaling,PI3K/AKT upregulated LSD1 protein levels and LSD1 sustained thisprocess through the demethylation of histones to activate gene expres-sion, thereby creating a positive feedback loop (49). Our study presentsa model that is in contrast to a classical enzymatic role for LSD1 as weestablish a catalytically independent oncogenic function for LSD1 bothin the regulation of PI3K/AKT signaling and in the regulation of geneexpression without the requirement of nuclear hormone signaling.Furthermore, according to our gene expression data obtained fromRNA-seq of LSD1 KD cells, we detected significant increase in CDH1gene expression after LSD1 KD. The change in CDH1 expressionafter LSD1 KD is consistent with previous studies, including studiesperformed in HCT116 cells, which demonstrate LSD1 is recruitedby Snail to the promoter of E-cadherin (CDH1) to demethylateH3K4me2 to repress CDH1 gene expression and drive EMT andmigration (10–12). However, there was no change in H3K4me2 atthe CDH1 promoter after LSD1 KD or enrichment of LSD1 in ourSW480 cells. This result suggests that regulation of CDH1 geneexpression occurs through an alternative mechanism that is inde-pendent of LSD1 enzymatic activity in some colorectal cancer celllines. Our findings are unique from studies where LSD1 interactswith Snail to promote migration through its demethylase activity inthat we demonstrate LSD1 promotes survival and migrationthrough stabilization of Snail by a noncatalytic activity. This is acritical distinction because as inhibitors of LSD1 advance in clinicaltrials, it is crucial to establish the importance of the enzymaticversus scaffolding activity of LSD1, as its tumorigenicity may bedynamic between colorectal cancer tumors or even within tumorcell populations. This information will enable us to determine howLSD1 may be acting in a particular tumor and inform patientenrollment in future clinical studies.

Our study demonstrates that the activation of AKT in colorectalcancer cells depends on the scaffolding function of LSD1 in theCoREST complex. In agreement with our findings, a recent studydemonstrated that a demethylase deadmutant LSD1was able to rescueviability defects in lymphoma cells where LSD1 had been knockeddown (17). This study further demonstrated that an LSD1 mutantunable to interact with CoREST was not sufficient to rescue viabilitydefects induced by LSD1 KD. In our study, we also demonstrate thatthe dual LSD1/HDAC1 inhibitor corin was sufficient to ablate pS473-

AKT, mimicking our KD and CRISPR studies. It is tempting tospeculate that corin, rather than acting through inhibiting enzymaticactivity, cobinds to LSD1 andHDAC1, disrupts the complex's histone-interacting surface, and reduces the complex's interaction with chro-matin, as has been shown with other LSD1 inhibitors (19–22). Thefinding that LSD1 and RCOR1 correlate in STAD, but not COAD orREADmay reflect differences in how these genes are transcriptionallyregulated between the large intestine and stomach. LSD1 and RCOR1are stabilized through their protein–protein interactions with oneanother, (50) consistent with our observation that loss of LSD1decreases RCOR1 protein and vice versa (Fig. 3A and B). Thisassociation also suggests that only LSD1 or RCOR1 needs to beoverexpressed to increase protein levels of the other. Our finding thatHDAC1 KD did not phenocopy loss of LSD1 or RCOR1 in theregulation of pS473-AKT supports this conjecture, but the possibilityremains that corin may function through enhanced repression ofHDAC1 in CoREST complexes.

PI3K/AKT signaling is essential for gut development due to its rolein the activation of EMTand extracellularmatrix (ECM)modificationsthat lead to the specification of foregut precursors (51). Interestingly,we demonstrate that LSD1 plays a significant role in the regulation ofECM organization and EMT-associated gene expression programs. Itis possible that LSD1 and PI3K signaling may cooperate duringspecification of foregut precursors. This hypothesis may also explainwhy there is a positive association between CoREST expression inPIK3CA-mutant gastrointestinal cancers but not in other cancer types,suggesting that this interaction may be lineage dependent. It is alsoconsistent with our observation of EMT-associated gene expressionprograms in a colorectal cancer cell line harboring PIK3CA mutationwithin the C2 domain.

In our TCGA analyses of PIK3CAmutation frequencies, mutationsin the helical and kinase domains were most frequent across thedifferent cancer types while less common mutations such as those inthe C2 domain, RBD and ABD appeared more cancer specific. Muta-tions in the C2 domain were disproportionately more common ingastrointestinal cancers of the stomach and colon/rectum. The exactnature of the relationship between PIK3CA mutation and LSD1expression is currently unclear, and further studies are required. Itis possible that LSD1 expression is more closely linked to specificPIK3CAmutations, and that the prevalence of the specific mutation ina given cancer type is driving the significance of the associationbetween LSD1 expression and PIK3CAmutational status in the TCGAdata. Our study also suggests that activation of the AKT–GSK3b–Snailaxis to promote EMT may be dependent on PIK3CA C2 domainmutation. This mechanism for upregulation of Snail is different from apreviously described mechanism by which PIK3CA with the H1047Rkinase mutation upregulates SNAI1 mRNA expression (52). Further-more, our finding of AKT signaling–dependent Snail stabilization byLSD1 is a distinct mechanism from stabilization of Snail throughinteraction with the CoREST complex (10). An important caveat toour PIK3CAmutational studies is that AGS cells contain mutations inboth the C2 and helical domains of PIK3CA. Our study does notdissect whether helical domainmutations alone are associated with theAKT–GSK3b–Snail axis. Future studies are required to untangle thefunctional role of different PIK3CA mutations and to establish amechanistic understanding of their differential roles in mediating theAKT–GSK3b–Snail axis. While we focus on the interaction betweenLSD1 and PIK3CA C2 domain mutation, we also demonstrated thatLSD1 KD reduced proliferation of cell lines with PIK3CA kinasedomain mutations. It will be important to study LSD1 under thecontext of different PIK3CA mutations to potentially identify a

LSD1 Promotes EMT via AKT in PIK3CA-Mutant CRC

AACRJournals.org Mol Cancer Res; 18(2) February 2020 275

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

synthetic lethal relationship and establish prognostic value in gastro-intestinal cancers.

Herein, we used EGF as tool to induce migration of HT29 cells. Thetumor microenvironment plays a critical role in promoting EMT viaautocrine and paracrine signaling by growth factors such as EGF,which stimulates EMT and migratory phenotypes in tumor cells (53).Further in vivo studies are required to understand how LSD1 mayregulate signaling events between tumors and the surroundingmicroenvironment.

Overall, we propose a model where the CoREST complex can actsynergistically with C2 domain PIK3CAmutations and growth factorsto fully active the PI3K/AKT pathway and stabilize Snail protein toenhance cell migration and survival (Fig. 6D). Genetic or therapeuticperturbation of the CoREST complex is sufficient to block cancer cellmigration and reduce survival. This work suggests that PIK3CA-mutant colorectal cancer may be particularly sensitive to LSD1 inhi-bitors that block the interaction of CoREST with transcription factors.

Disclosure of Potential Conflicts of InterestH.P. Mohammad is a senior scientific director at GlaxoSmithKline. No potential

conflicts of interest were disclosed by the other authors.

Authors’ ContributionsConception and design: S.A. Miller, H.M. O'HaganDevelopment of methodology: S.A. Miller

Acquisition of data (provided animals, acquired and managed patients, providedfacilities, etc.): S.A. Miller, S.S. Savant, S. Sriramkumar, N. DingAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): S.A. Miller, R.A. Policastro, S.S. Savant, X. Lu, S. Cao,P.A. Cole, H.M. O'HaganWriting, review, and/or revision of the manuscript: S.A. Miller, S.S. Savant,J.H. Kalin, P.A. Cole, H.M. O'HaganAdministrative, technical, or material support (i.e., reporting or organizing data,constructing databases): S.A. Miller, H.P. Mohammad, P.A. ColeStudy supervision: S.A. Miller, G.E. Zentner, H.M. O'HaganOther (prepared and provided corin, the small-molecule dual inhibitor): J.H. Kalin

AcknowledgmentsWe thank the Indiana University Center for Genomics and Bioinformatics and the

Indiana University Light Microscopy Imaging Center for their assistance. This workwas supported by the National Institute of Environmental Health Sciences Grant(R01ES023183, to H.M. O’Hagan) and the NIH, National Center for AdvancingTranslational Sciences, Clinical and Translational Sciences Award [TL1 TR001107and UL1 TR001108 (principal investigator, A. Shekhar), to S.A. Miller] as well as theNIH Grant (GM62437, to P.A. Cole).

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

Received July 21, 2019; revised September 24, 2019; accepted November 5, 2019;published first November 8, 2019.

References1. Haigis KM, Cichowski K, Elledge SJ. Tissue-specificity in cancer: the rule, not the

exception. Science 2019;363:1150–1.2. Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz LA Jr, Kinzler KW.

Cancer genome landscapes. Science 2013;339:1546–58.3. Cancer Genome Atlas Research Network. Comprehensive molecular character-

ization of human colon and rectal cancer. Nature 2012;487:330–7.4. Samuels Y, Wang Z, Bardelli A, Silliman N, Ptak J, Szabo S, et al. High

frequency of mutations of the PIK3CA gene in human cancers. Science 2004;304:554.

5. Wang Q, Yan-long Shi YL, Zhou K, Wang LL, Yan ZX, Liu YL, et al. PIK3CAmutations confer resistance to first-line chemotherapy in colorectal cancer.Cell Death Dis 2018;9:739.

6. Liao X, Morikawa T, Lochhead P, Imamura Y, Kuchiba A, Yamauchi M, et al.Prognostic role of PIK3CA mutation in colorectal cancer: cohort study andliterature review. Clin Cancer Res 2012;18:2257–68.

7. Wang G, Long J, Gao Y, Zhang W, Han F, Xu C, et al. SETDB1-mediatedmethylation of Akt promotes its K63-linked ubiquitination and activationleading to tumorigenesis. Nat Cell Biol 2019;21:214–25.

8. Guo J, Dai X, Laurent B, Zheng N, Gan W, Zhang J, et al. AKT methylation bySETDB1 promotes AKT kinase activity and oncogenic functions. Nat Cell Biol2019;21:226–37.

9. Hsu HC, Liu YS, Tseng KC, Yang TS, Yeh CY, You JF, et al. CBB1003, a lysine-specific demethylase 1 inhibitor, suppresses colorectal cancer cells growththrough down-regulation of leucine-rich repeat-containing G-protein-coupled receptor 5 expression. J Cancer Res Clin Oncol 2015;141:11–21.

10. Lin Y, Wu Y, Li J, Dong C, Ye X, Chi YI, et al. The SNAG domain of Snail1functions as a molecular hook for recruiting lysine-specific demethylase 1.EMBO J 2010;29:1803–16.

11. Ding J, Zhang ZM, Xia Y, Liao GQ, Pan Y, Liu S, et al. LSD1-mediated epigeneticmodification contributes to proliferation and metastasis of colon cancer. Br JCancer 2013;109:994–1003.

12. Ferrari-Amorotti G, Fragliasso V, Esteki R, Prudente Z, Soliera AR, Cattelani S,et al. Inhibiting interactions of lysine demethylase LSD1 with snail/slug blockscancer cell invasion. Cancer Res 2013;73:235–45.

13. Shi Y, Lan F, Matson C, Mulligan P, Whetstine JR, Cole PA, et al. Histonedemethylationmediated by the nuclear amine oxidase homolog LSD1. Cell 2004;119:941–53.

14. Iwase S, Januma A, Miyamoto K, Shono N, Honda A, Yanagisawa J, et al.Characterization of BHC80 in BRAF-HDAC complex, involved in neuron-specific gene repression. Biochem Biophys Res Commun 2004;322:601–8.

15. Ballas N, Battaglioli E, Atouf F, Andres ME, Chenoweth J, Anderson ME, et al.Regulation of neuronal traits by a novel transcriptional complex. Neuron 2001;31:353–65.

16. Sehrawat A, Gao L,Wang Y, BankheadA III,McWeeney SK, King CJ, et al. LSD1activates a lethal prostate cancer gene network independently of its demethylasefunction. Proc Natl Acad Sci U S A 2018;115:E4179–88.

17. Hatzi K, Geng H, Doane AS, Meydan C, LaRiviere R, Cardenas M, et al. Histonedemethylase LSD1 is required for germinal center formation and BCL6-drivenlymphomagenesis. Nat Immunol 2019;20:86–96.

18. Andres ME, Burger C, Peral-Rubio MJ, Battaglioli E, Anderson ME, Grimes J,et al. CoREST: a functional corepressor required for regulation of neural-specificgene expression. Proc Natl Acad Sci U S A 1999;96:9873–8.

19. Takagi S, IshikawaY,MizutaniA, Iwasaki S,Matsumoto S, KamadaY, et al. LSD1Inhibitor T-3775440 Inhibits SCLC cell proliferation by disrupting LSD1Interactions with SNAG domain proteins INSM1 and GFI1B. Cancer Res2017;77:4652–62.

20. Maiques-DiazA, SpencerGJ, Lynch JT, Ciceri F,Williams EL,Amaral FMR, et al.Enhancer activation by pharmacologic displacement of LSD1 fromGFI1 inducesdifferentiation in acute myeloid leukemia. Cell Rep 2018;22:3641–59.

21. Vinyard ME, Su C, Siegenfeld AP, Waterbury AL, Freedy AM, Gosavi PM, et al.CRISPR-suppressor scanning reveals a nonenzymatic role of LSD1 in AML.Nat Chem Biol 2019;15:529–39.

22. Yamamoto R, Kawahara M, Ito S, Satoh J, Tatsumi G, Hishizawa M, et al.Selective dissociation between LSD1 and GFI1B by a LSD1 inhibitor NCD38induces the activation of ERG super-enhancer in erythroleukemia cells.Oncotarget 2018;9:21007–21.

23. Ding N, Miller SA, Savant SS, O'Hagan HM. JAK2 regulates mismatch repairprotein-mediated epigenetic alterations in response to oxidative damage.Environ Mol Mutagen 2019;60:308–19.

24. Plotnik JP, Hollenhorst PC. Genome-wide analysis of RAS/ERK signalingtargets. Methods Mol Biol 2017;1487:277–88.

25. Chu N, Salguero AL, Liu AZ, Chen Z, Dempsey DR, Ficarro SB, et al. Aktkinase activation mechanisms revealed using protein semisynthesis. Cell 2018;174:897–907.

Miller et al.

Mol Cancer Res; 18(2) February 2020 MOLECULAR CANCER RESEARCH276

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

26. Alessi DR, James SR, Downes CP, Holmes AB, Gaffney PR, Reese CB,et al. Characterization of a 3-phosphoinositide-dependent protein kinasewhich phosphorylates and activates protein kinase Balpha. Curr Biol1997;7:261–9.

27. Sarbassov DD, Guertin DA, Ali SM, Sabatini DM. Phosphorylation andregulation of Akt/PKB by the rictor-mTOR complex. Science 2005;307:1098–101.

28. Ushio-FukaiM,Alexander RW, AkersM, YinQ, Fujio Y,Walsh K, et al. Reactiveoxygen speciesmediate the activation ofAkt/protein kinaseB by angiotensin II invascular smooth muscle cells. J Biol Chem 1999;274:22699–704.

29. Rhodes N, Heerding DA, Duckett DR, Eberwein DJ, Knick VB, Lansing TJ, et al.Characterization of an Akt kinase inhibitor with potent pharmacodynamic andantitumor activity. Cancer Res 2008;68:2366–74.

30. Nicholson TB, Chen T. LSD1 demethylates histone and non-histone proteins.Epigenetics 2009;4:129–32.

31. Mohammad HP, Smitheman KN, Kamat CD, Soong D, Federowicz KE, VanAller GS, et al. A DNA hypomethylation signature predicts antitumor activity ofLSD1 Inhibitors in SCLC. Cancer Cell 2015;28:57–69.

32. You A, Tong JK, Grozinger CM, Schreiber SL. CoREST is an integral componentof the CoREST- human histone deacetylase complex. Proc Natl Acad Sci U S A2001;98:1454–8.

33. Upadhyay G, Chowdhury AH, Vaidyanathan B, Kim D, Saleque S. Antagonisticactions of Rcor proteins regulate LSD1 activity and cellular differentiation.Proc Natl Acad Sci U S A 2014;111:8071–6.

34. Lee MG, Wynder C, Cooch N, Shiekhattar R. An essential role for CoREST innucleosomal histone 3 lysine 4 demethylation. Nature 2005;437:432–5.

35. Kalin JH, Wu M, Gomez AV, Song Y, Das J, Hayward D, et al. Targeting theCoREST complex with dual histone deacetylase and demethylase inhibitors.Nat Commun 2018;9:53.

36. Chen CS, Weng SC, Tseng PH, Lin HP, Chen CS. Histone acetylation-independent effect of histone deacetylase inhibitors on Akt through the reshuf-fling of protein phosphatase 1 complexes. J Biol Chem 2005;280:38879–87.

37. Barretina J, Caponigro G, Stransky N, Venkatesan K, Margolin AA, Kim S, et al.The Cancer Cell Line Encyclopedia enables predictive modelling of anticancerdrug sensitivity. Nature 2012;483:603–7.

38. Zhang L, Zhou F, ten Dijke P. Signaling interplay between transforming growthfactor-beta receptor and PI3K/AKT pathways in cancer. Trends Biochem Sci2013;38:612–20.

39. Jin L, Hanigan CL, Wu Y, Wang W, Park BH, Woster PM, et al. Loss of LSD1(lysine-specific demethylase 1) suppresses growth and alters gene expression of

human colon cancer cells in a p53- and DNMT1(DNA methyltransferase 1)-independent manner. Biochem J 2013;449:459–68.

40. Borradori L, Sonnenberg A. Structure and function of hemidesmosomes: morethan simple adhesion complexes. J Invest Dermatol 1999;112:411–8.

41. Mani SA, GuoW, LiaoMJ, Eaton EN, Ayyanan A, Zhou AY, et al. The epithelial-mesenchymal transition generates cells with properties of stem cells. Cell 2008;133:704–15.

42. Nieto MA, Huang RY, Jackson RA, Thiery JP. Emt: 2016. Cell 2016;166:21–45.43. ZhouBP,Deng J, XiaW,Xu J, Li YM,GunduzM, et al. Dual regulation of Snail by

GSK-3beta-mediated phosphorylation in control of epithelial-mesenchymaltransition. Nat Cell Biol 2004;6:931–40.

44. Cross DA, Alessi DR, Cohen P, Andjelkovich M, Hemmings BA. Inhibition ofglycogen synthase kinase-3 by insulinmediated byprotein kinase B.Nature 1995;378:785–9.

45. Zhao L, Vogt PK. Helical domain and kinase domain mutations in p110alpha ofphosphatidylinositol 3-kinase induce gain of function by different mechanisms.Proc Natl Acad Sci U S A 2008;105:2652–7.

46. Ikenoue T, Kanai F, Hikiba Y, Obata T, Tanaka Y, Imamura J, et al. Functionalanalysis of PIK3CA gene mutations in human colorectal cancer. Cancer Res2005;65:4562–7.

47. Guinney J, Dienstmann R, Wang X, de Reyni�es A, Schlicker A, Soneson C,et al. The consensus molecular subtypes of colorectal cancer. Nat Med 2015;21:1350–6.

48. Gan Y, Shi C, Inge L, Hibner M, Balducci J, Huang Y, et al. . Differential roles ofERK andAkt pathways in regulation of EGFR-mediated signaling andmotility inprostate cancer cells. Oncogene 2010;29:4947–58.

49. Chen C, Wang Y, Wang S, Liu Y, Zhang J, Xu Y, et al. LSD1 sustains estrogen-driven endometrial carcinoma cell proliferation through the PI3K/AKTpathwayvia di-demethylating H3K9 of cyclin D1. Int J Oncol 2017;50:942–52.

50. Shi YJ, Matson C, Lan F, Iwase S, Baba T, Shi Y. Regulation of LSD1 histonedemethylase activity by its associated factors. Mol Cell 2005;19:857–64.

51. Villegas SN, Rothov�a M, Barrios-Llerena ME, Pulina M, Hadjantonakis AK, LeBihan T, et al. PI3K/Akt1 signalling specifies foregut precursors by generatingregionalized extra-cellular matrix. Elife 2013;2:e00806.

52. Wallin JJ, Guan J, Edgar KA, Zhou W, Francis R, Torres AC, et al. Active PI3Kpathway causes an invasive phenotype which can be reversed or promoted byblocking the pathway at divergent nodes. PLoS One 2012;7:e36402.

53. Maheshwari G, Wiley HS, Lauffenburger DA. Autocrine epidermal growthfactor signaling stimulates directionally persistent mammary epithelial cellmigration. J Cell Biol 2001;155:1123–8.

AACRJournals.org Mol Cancer Res; 18(2) February 2020 277

LSD1 Promotes EMT via AKT in PIK3CA-Mutant CRC

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748

2020;18:264-277. Published OnlineFirst November 8, 2019.Mol Cancer Res   Samuel A. Miller, Robert A. Policastro, Sudha S. Savant, et al.   Colorectal Cancer

-MutantPIK3CAPromotes Epithelial-to-Mesenchymal Transition in Lysine-Specific Demethylase 1 Mediates AKT Activity and

  Updated version

  10.1158/1541-7786.MCR-19-0748doi:

Access the most recent version of this article at:

  Material

Supplementary

  http://mcr.aacrjournals.org/content/suppl/2019/11/08/1541-7786.MCR-19-0748.DC1

Access the most recent supplemental material at:

  Cited articles

  http://mcr.aacrjournals.org/content/18/2/264.full#ref-list-1

This article cites 53 articles, 19 of which you can access for free at:

  Citing articles

  http://mcr.aacrjournals.org/content/18/2/264.full#related-urls

This article has been cited by 3 HighWire-hosted articles. Access the articles at:

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  [email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://mcr.aacrjournals.org/content/18/2/264To request permission to re-use all or part of this article, use this link

on August 28, 2021. © 2020 American Association for Cancer Research. mcr.aacrjournals.org Downloaded from

Published OnlineFirst November 8, 2019; DOI: 10.1158/1541-7786.MCR-19-0748