FOXO1 promotes the expression of canonical WNT target genes in examined basal‐like breast and glioblastoma multiforme cancer cells

Basal‐like breast cancer (BBC) and glioblastoma multiforme (GBM) are aggressive cancers associated with poor prognosis. BBC and GBM have stem cell‐like gene expression signatures, which are in part driven by forkhead box O (FOXO) transcription factors. To gain further insight into the impact of FOXO1 in BBC, we treated BT549 cells with AS1842856 and performed RNA sequencing. AS1842856 binds to unphosphorylated FOXO1 and inhibits its ability to directly bind to DNA. Gene Set Enrichment Analysis indicated that a set of WNT pathway target genes, including lymphoid enhancer‐binding factor 1 (LEF1) and transcription factor 7 (TCF7), were robustly induced after AS1842856 treatment. These same genes were also induced in GBM cell lines U87MG, LN18, LN229, A172, and DBTRG upon AS1842856 treatment. By contrast, follow‐up RNA interference (RNAi) targeting of FOXO1 led to reduced LEF1 and TCF7 gene expression in BT549 and U87MG cells. In agreement with RNAi experiments, CRISPR Cas9‐mediated FOXO1 disruption reduced the expression of canonical WNT genes LEF1 and TCF7 in U87MG cells. The loss of TCF7 gene expression in FOXO1 disruption mutants was restored by exogenous expression of the DNA‐binding‐deficient FOXO1‐H215R. Therefore, FOXO1 induces TCF7 in a DNA‐binding‐independent manner, similar to other published FOXO1‐activated genes such as TCF4 and hes family bHLH transcription factor 1. Our work demonstrates that FOXO1 promotes canonical WNT gene expression in examined BBC and GBM cells, similar to results found in Drosophila melanogaster, T‐cell development, and murine acute myeloid leukemia models.

Drosophila melanogaster has a single FOXO protein (dFoxO) that can act with armadillo (beta-catenin homolog) to pattern wings, eyes, and induce apoptosis.Overexpression of armadillo with sd-Gal4 led to an ectopic bristle phenotype near the wing margins in adult flies.Loss-of-function dfoxo mutants (dfoxo-IR, which had half the normal expression of dFoxO, dfoxoD94/+, or dfoxo21/+) suppressed the armadillodriven ectopic bristle phenotype [59].Furthermore, ectopic gene expression of wf (wingful, human NOTUM homolog) reporter gene wf-LacZ in third instar wing disks and salivary glands was induced by overexpression of armadillo (via ptc-Gal4); this phenotype was strongly suppressed by dfoxo loss-of-function mutants (dfoxo-IR, dfoxoD94/+, or dfoxo21/+) [59].In the Drosophila eye, overexpression of armadillo or dFoxO (using GMR-Gal4) led to a small eye phenotype due to increased apoptosis.Loss-of-function mutants in armadillo suppressed the small eye phenotype of dFoxO overexpression, whereas loss-offunction dfoxo suppressed the small eye phenotype of armadillo overexpression [59].In addition to numerous genetic interactions observed between dFoxO and armadillo, these proteins are physically associated.Myc-tagged armadillo protein associated with FLAGtagged dFoxO in co-immunoprecipitation analyses [59].FOXO transcription factors can bind to, and act in concert with, beta-catenin in additional contexts, such as HEK 293T cells, to induce the gene expression of GHOST transcripts (genes hidden outside the standard targets) such as GADD45A [60].In preosteoblasts, FOXO1 and beta-catenin induce genes such as GADD45A in response to oxidative stress [22].Betacatenin-FOXO1 was also shown to hinder kidney fibrosis [61,62].
Foxo1 was required for canonical WNT gene expression in osteoblasts and a murine acute myeloid leukemia (AML) model induced by activated betacatenin.Forced expression of Foxo1 led to increased expression of Axin2 and Lef1 in murine osteoblasts [63].Furthermore, constitutively active beta-catenin (Ctnnb1 CAosb ) had increased expression of targets Axin2 and Lef1, which were significantly reduced by osteoblast-specific deletion of Foxo1 [63].In murine osteoblasts, active beta-catenin (Ctnnb1 CAosb ) drives AML development [64].Osteoblast deletion of Foxo1 prevented Ctnnb1 CAosb -driven AML development as evidenced by a lack of blasts and neutrophils with nuclear hypersegmentation in the blood, as well as blast infiltration of liver at periportal sites; this type of liver infiltration is commonly observed in murine AML models driven by active beta-catenin and fusion proteins AML1/MDS1/EVI [63,65].Therefore, evidence points to a role for Foxo1 in driving WNT gene expression in AML development.
There is evidence that FOXO factors impede canonical WNT gene expression in some contexts, including osteosarcoma, pancreatic, and colon cancer cells.FOXO1 tumor suppressor activity inhibited the WNT pathway in osteosarcoma cell lines U2OS and MG63 [66].FOXO1 was overexpressed for 48 h using an integrated FOXO1-ER fusion that was induced to go to the nucleus by 4-OHT treatment; this led to decreased WNT pathway reporter activity using TOPflash (reporter contains TCF/LEF1 sites that regulate luciferase expression) [66].In addition, FOXO1-induced cells had less detected beta-catenin protein based on immunofluorescence.FOXO1 induction using 4-OHT also led to cell cycle arrest and apoptosis, highlighting canonical FOXO roles in cancer.
FOXO1 hindered the WNT pathway in pancreatic cancer [67].FOXO1 was overexpressed using an integrated, selected virally transduced cassette, leading to induction in the gene expression of a long noncoding RNA LINC01197 in a series of pancreatic cancer cell lines AsPC1, BxPC3, and PANC1 [67].Reduction in FOXO1 by siRNA in HPNE (immortalized pancreatic duct cells) led to a loss in LINC01197 expression.LINC01197 RNA attached to biotin beads interacted with beta-catenin protein in RNA pull-down assays.Overexpression of LINC01197 for 48 h decreased WNT target gene expression, as measured by quantitative real-time PCR (qRT-PCR; MYC, CCND1, CD44, and MMP26), and disrupted co-immunoprecipitation between beta-catenin and TCF4 (using endogenous proteins) from pancreatic ductal adenocarcinoma cells.Conversely, LINC01197 siRNA for 48 h increased WNT target gene expression measured by qRT-PCR (MYC, CCND1, CD44, and MMP26, and the interaction between endogenous TCF4 and beta-catenin in HPNE cells.The authors present a model in which FOXO1 inhibits the TCF4/beta-catenin interaction in pancreatic cancer via induction of LINC01197 [67]. FOXO transcription factors have a complex role in the solid tumors BBC and GBM.These factors act as tumor suppressors by activating apoptosis under the right conditions and are pro-oncogenic by directly inducing stem gene expression [19,23,[68][69][70].However, the extent to which FOXO transcription factors influence BBC and GBM biology, especially under basal conditions, remains an open area of investigation.In this work, we treated BBC and GBM cells with AS1842856 to discover novel FOXO-regulated processes.AS1842856 was identified for its ability to specifically bind to the unphosphorylated form of FOXO1, thereby blocking the ability of this factor to directly bind to DNA [71].A set of WNT pathway target genes were robustly induced by AS1842856 not only in BT549 cells but also broadly in a series of GBM cell lines.In contrast to AS1842856 treatment (which blocks FOXO1 DNA binding), FOXO1 gene disruption or RNAi targeting led to reduced LEF1 and TCF7 gene expression in U87MG cells.Expression of the DNA-bindingdefective FOXO1-H215R mutant restored TCF7 gene expression to FOXO1 disruption mutants.Therefore, the DNA-binding activity of FOXO1 is not required to induce TCF7, similar to other published targets TCF4 and hes family bHLH transcription factor 1 (HES1) [72,73].Our work is the first to demonstrate that FOXO1 promotes canonical WNT gene expression in examined solid tumor-derived BBC and GBM cells, similar to results found in Drosophila melanogaster, T-cell development, and beta-catenin-driven AML models.

FOXO1 disruption
The FOXO1 gene was disrupted in U87MG cells using the OriGene CRISPR Cas9 knock-out kit Cat#: KN400477 (Rockville, MD).Mutants were confirmed by western blot and sequencing.The truncated protein is predicted to start in the middle of the DNA-binding domain and contain the C terminus of the protein.

Transfection of FOXO1-H215R mutant
One million log-phase U87MG cells were transfected using a LONZA electroporation kit (VPG1001 for neurons) and program T-020 with 500 nanograms of indicated DNA (control vector or FOXO1-H215R mutant) [73].RNA was prepped from cells 24 h post-transfection.

RNA extraction and sequencing
One lM AS1842856 or (DMSO control) was added to BT549 cells for 48 h.Total RNA was extracted using Qiagen RNAeasy and DNAse reagents (Hilden, Germany).Three biological replicates of control and AS1842856treated samples were used for RNA extraction.The quality of the extracted RNA was analyzed using a NanoDrop spectrophotometer (Thermo Fisher, USA) and Agilent Bioanalyzer (Santa Clara, CA).
RNA-Seq analysis, including mapping to the human genome, sequence assembly, and identification of differentially expressed genes (DEGs), was performed by the Novogene Corporation Inc. (Sacramento, CA).Original image data files from high-throughput sequencing (Illumina) were transformed into sequenced reads (called Raw Data or Raw Reads) by CASAVA base recognition.STAR software was used for mapping.FPKM (Fragments Per Kilobase of transcript sequence per Million base pairs sequenced) was utilized to estimate gene expression levels.Differential expression analysis was performed using DESeq2 R package.
RNA-Seq data were deposited to the GEO database (Accession: GSE179856).RNA-Seq data were analyzed to identify DEGs associated with AS1842856 treatment.Differentially expressed genes between AS1842856 treated and control cell lines were investigated using Gene Set Enrichment Analysis (GSEA) [74].GSEA was performed by setting ENSEMBL Gene ID v7.3 Chip platform, 1000 permutations, Signal2Noise, and weighted enrichment statistics.Gene expression box plots from clinical samples were prepared using the GEPIA 2 online tool [75].

Quantitative real-time PCR
Log-phase cancer cell lines were treated with 1 lM AS1842856 for 48 h (unless otherwise noted).Extracted total RNA was used to prepare cDNA using Superscript Reverse Transcriptase II (Invitrogen, Carlsbad, CA).Primers spanning coding sequences for the genes were designed using Primer3 (v0.4.0) (https://bioinfo.ut.ee/ primer3-0.4.0/) and ordered from Sigma-Aldrich (Saint Louis, MO).Amplification and expression of genes were performed using the Applied Biosystems StepOne Real-Time PCR System (Foster City, CA).TUBB (beta-tubulin gene) was used as the control for relative normalization and quantification performed by the 2 ÀDDCT method (primer list in Table S1) [76].The Student's t-test was employed to assess statistical significance.

Ethics approval
Work was performed with Institutional Biosafety Committee approval from the University of Texas Rio Grande Valley: Registration number: 2016-003-IBC.

RNA-Seq revealed AS1842856 treatment increased WNT gene expression in BBC BT549 cells
FOXO transcription factors reside at least in part in the nucleus in BBC and GBM cells and promote stem gene expression programs [23,70,77].The roles of FOXO factors in BBC and GBM, especially under basal conditions, remained to be fully characterized.To identify FOXO1 target genes (direct and indirect), RNA-Seq analysis was performed with BT549 cells that were treated with 1 lM AS1842856 (which blocks the ability of FOXO1 to directly bind DNA) for 48 h [71].Gene expression levels were estimated by the abundance of transcripts (count of sequencing) that mapped to the genome.Differential expression analysis between DMSO control (three biological replicates) and AS1842856 treated samples (three biological replicates) was performed using the DESeq2 R package.6754 differentially expressed genes (DEGs) were identified.Gene Set Enrichment Analysis was performed to discover processes impacted by AS1842856 treatment.Examining the Molecular Signatures Database Hallmark genesets (h.all.7.4symbols.GMT), we found that two were significantly enriched in AS1842856-treated samples Table 1, whereas 16 were enriched in the controls (Table 2).

Validation of WNT target induction by AS1842856 treatment through qRT-PCR
To validate differentially expressed WNT-related genes obtained from RNA-Seq data, we performed qRT-PCR.Primers were designed using PRIMER3 (v0.4.0) online tool (Table S1).WNT pathway genes LEF1, TCF7, AXIN2, and HEY1 were increased upon AS1842856 treatment in BT549 cells (Fig. 2A).Validation of differential gene expression by qRT-PCR built confidence in the RNA-Seq analysis.We sought to examine whether FOXO1 inhibition would induce canonical WNT targets in another BBC cell line.None of the examined WNT genes were increased in BBC cell line MDA-MB-468 upon AS1842856 treatment (Fig. 2B).We investigated a set of GBM cell lines for changes in WNT pathway targets, given that this cancer also has FOXO-regulated stem gene signatures [1,23,70].Treatment with AS1842856 led to increased expression of LEF1, TCF7, and AXIN2 in U87MG, LN229, LN18, A172, and DBTRG cells (Fig. 2C-F and Fig. S1A).
AXIN2 induction was particularly robust in these settings (26 to 477-fold increased with AS1842856 treatment compared with control samples).PROX1 and  HEY1 were induced by AS1842856 treatment in LN229 and A172 cells (Fig. 2D,F).PROX1 was diminished in U87MG and LN18 cells upon AS1842856 treatment (Fig. 2C,E).Thus, AXIN2, LEF1, and TCF7 were induced by AS1842856 treatment in six out of seven examined BBC and GBM cell lines.Experiments were done using 10-fold less AS1842856 (samples ranging from 100 nM to 1 lM).Induction of LEF1 was observed even with 100 nM of AS1842856 in BT549 and U87MG cells.Prior studies found that > 1 lM of AS1842856 was needed to efficiently inhibit FOXO3 or FOXO4, whereas 100 nM of the drug was sufficient to inhibit FOXO1 (Fig. 3A,B) [71,81].It is likely that FOXO3 and FOXO4 functionally overlap with FOXO1 in regulating WNT genes, as was observed in AML [63].

AS1842856 treatment impacted WNT pathway genes in colon cancer cells
The WNT pathway is of central importance in colon cancer, with 80% having inactivating mutations in the APC gene that encodes a protein that promotes betacatenin destruction [35,82,83].The WNT pathway sustains cancer stem cells and promotes the proliferation of cells as they migrate before differentiation [84,85].To ascertain the impact of FOXO1 on WNT pathway gene expression in colon cancer, we treated HCT116 and SW480 cells with AS1842856.In the HCT116 cell line, AXIN2, TCF7, and HEY1 were significantly induced upon AS1842856 treatment (Fig. S1B,C); WNT7B was also decreased in HCT116 cells (Fig. S1E).
In the SW480 cell line, HEY1 and WNT5A were slightly upregulated, whereas WNT7B was decreased upon AS1842856 treatment (Fig. S1C,F).Therefore, AS1842856 had varied impacts on colon cancer cell lines.
Secreted WNT ligands are implicated in breast cancer development.Studies indicated that WNT5A might have tumor suppressive roles, at least in some contexts, whereas WNT5B activates the WNT pathway in BBC [45,46].WNT7B expression was significantly associated with poor prognosis in breast cancer [86].RNA-Seq data revealed that several WNT-secreted factors were differentially expressed upon AS1842856 treatment in BT549 cells.WNT5A was induced by treatment, whereas WNT5B and WNT7B were diminished, Table 3.These genes were analyzed in a set of BBC and GBM cell lines by qRT-PCR.BT549 and LN18 had increased WNT5A and decreased WNT5B and WNT7B (Fig. S2A,E).MDA-MB-468 only had diminished WNT5B and WNT7B (Fig. S2B).U87MG and DBTRG cells had decreased WNT7B with AS1842856 treatment (Figs S1D and S2C).LN229 and A172 had increased WNT5B with AS1842856 treatment (Fig. S2D,F).Therefore, the impact of AS1842856 treatment on WNT ligands differed depending on the context in examined BBC and GBM cells.

AS1842856 treatment led to increased active and total beta-catenin in BT549 and U87MG cells
We examined beta-catenin subcellular localization and abundance to gain insight into why AS1842856 treatment led to increased target gene expression.Nuclear and cytoplasmic fractions were prepared for BT549 and U87MG cells treated with or without the drug.We found that the nuclear and cytoplasmic distributions of beta-catenin were comparable, but the amount of total and active beta-catenin was increased with AS1842856 treatment (Fig. 3C).GSK3b phosphorylates beta-catenin leading to its degradation.GSK3b protein was reduced in AS1842856 treated samples, likely explaining the increase in beta-catenin (Fig. 3C).We performed microscopy with AS1842856 treated cells to determine whether this drug impacted beta-catenin membrane recruitment (using transfected beta-catenin-GFP).No change in betacatenin membrane recruitment was observed in AS1842856-treated samples (Fig. S3).

FOXO1 RNAi led to decreased canonical WNT gene expression
To rigorously investigate the contribution of FOXO1 to WNT gene regulation, we performed RNAi experiments.We found the reduction in FOXO1 significantly decreased LEF1 and TCF7 gene expression (Fig. 4A,  B).The FOXO1 RNAi data contrast with results obtained with AS1842856 treatment, which induced WNT genes.To examine the role of FOXO1 in WNT gene regulation in another way, we built FOXO1 disruption mutants using CRISPR Cas9 genome editing in U87MG cells (confirmed by western blot and sequencing).We found that LEF1 and TCF7 gene expression was decreased in FOXO1 disruption mutants (Fig. 4C,D and Fig. S4).Exogenous expression of the DNA-binding-defective FOXO1-H215R mutant significantly restored TCF7 expression to FOXO1 disruption mutants in the U87MG background (Fig. 4E).FOXO1 mutant H215R has a histidine that directly contacts DNA replaced with arginine.Numerous studies have validated that this mutant fails to bind tested target DNA sequences, including 3XIRS [72,73].Transfections of FOXO1-H215R mutant were done via electroporation using a protocol for mouse neurons, given that these cells were difficult to transfect, similar to our previous work with FOXO3 mutants in GBM cells [70].Strikingly, it appears that the DNA-binding activity of FOXO1 is dispensable for TCF7 induction, possibly explaining why AS1842856 treatment induced TCF7, as this drug only impedes the ability of FOXO1 to directly bind to DNA.In support of this idea, a previously published FOXO1-H215R-induced target TCF4 was also significantly increased by AS1842856 treatment in our RNA-Seq data (P = 1.25E-38) [73].
We performed RNAi experiments to examine contributions from FOXO3 and FOXO4 to WNT pathway gene expression (24-h timepoint).We found that the reduction in FOXO3 only reduced LEF1 and TCF7 in BT549 cells (not U87MG cells), whereas FOXO4 RNAi led to reduced LEF1 and TCF7 in U87MG cells and reduced LEF1 in BT549 cells (Fig. 5A-D).Therefore, FOXO1 promotes LEF1 and TCF7 gene expression in U87MG and BT549 cells with some functional overlap with FOXO3 and FOXO4.

WNT targets were robustly overexpressed in GBM and colon cancer clinical samples
The WNT pathway has many roles in cancer, including within stem and proliferating cells [46,86].To gain insight into WNT pathway expression in clinical samples, we queried the GEPIA2 database, which has gene expression and matched control clinical data [74].AXIN2, LEF1, TCF7, and WNT5A were robustly increased in GBM compared with control samples (Fig. 6A-D).AXIN2, LEF1, TCF7, and PROX1 were highly expressed in colon cancer compared with control samples (Fig. 7A-D).These results support the notion that canonical WNT gene expression is elevated and relevant in GBM and colon cancer clinical samples.Refining regulatory mechanisms that promote these WNT outputs, such as FOXO1-mediated, will shed light on the biological underpinnings that drive GBM and colon cancer progression.

Discussion
The WNT pathway facilitates stem cell renewal, cellular differentiation, and proliferation in cancer [32,38,45,46,85,86] B and 2A,C-F).It appears that AS1842856 acts as a FOXO1 modulator instead of solely as an inhibitor.Differential impacts of AS1842856 on FOXO1 target genes are possible, especially given its ability to specifically bind to only the unphosphorylated form of this protein to hinder DNA binding [71].We posit that AS1842856 only inhibited the gene expression of a subset of FOXO1 targets that likely required DNA binding, such as SOD2, INSR, and GADD45A (P-values 1.37E-158, 3.54E-08, and 1.27E-21 respectively).The DNAbinding function of FOXO1 was not required for it to induce TCF7, as evidenced by the ability of FOXO1-H215R to restore its gene expression to FOXO1 disruption mutants (Fig. 4E).These results might explain why AS1842856 (which blocks DNA binding) increased TCF7.Perhaps FOXO1 is released from the DNA by AS1842856 and is repurposed to targets that do not require direct DNA binding, such as TCF7.Future studies that employ chromatin immunoprecipitation experiments will shed light on this possibility.
There is precedence for DNA-binding-independent gene regulation by FOXO1.For example, FOXO1-   H215R induced the related WNT pathway component TCF4 in renal carcinoma cells; this gene was also increased in our RNA-Seq data by AS1842856 treatment (P = 1.25E-38) [73].A related Foxo1 DNA-binding-defective mutant with H212 mutated to R (murine model) induced Hes1 in myoblasts where Foxo1 acted as a co-activator that was brought to targets by the protein Csl (CBF1, suppressor of hairless, Lag-1) [72].Therefore, FOXO1-H215R (and related mutant in mice) induces the expression of numerous transcription factors, including TCF7, TCF4, and HES1.
FOXO1 inhibition impacted AXIN2, LEF1, and TCF7 in some, but not all, examined BBC and colon cancer cell lines.One possibility for this divergence could be that the WNT pathway is poised to activate gene expression in some contexts but not others.Perhaps, the distinct BBC subtype contributes to WNT pathway activity.BT549 cells are of the Basal B subtype, and MDA-MB-468 are Basal A. Moreover, beta-catenin was observed in both the nucleus and cytoplasm of BT549 cells and in the cytoplasm of MDA-MB-468 cells, highlighting a possible mechanism for different WNT gene expression dynamics in these settings [87,88].Identifying the conditions that dictate WNT pathway regulation by FOXO1 in cancer will allow for a greater understanding of divergent responses to AS1842856 treatment in these settings.
FOXO factors were previously reported to negatively regulate WNT pathway targets such as CCND1 in pancreatic and prostate cancers as well as osteosarcoma [66,67,89].In osteosarcoma cells, FOXO1 activation led to decreased CCND1 (encoding cyclin D1) expression [66].Therefore, in some settings, FOXO factors hinder WNT and promote this pathway in others.One possible explanation for these striking differences is that FOXO output is highly contextdependent [90,91].Perhaps, co-factors and/or posttranslational modifications direct the ability of FOXO to promote WNT gene expression in AML and GBM cells, whereas another set of circumstances, such as under stress, direct FOXO to disrupt the WNT pathway in DLD1 and pancreatic cancer cells.The localization of FOXO factors may also impact their ability to regulate the WNT pathway.For example, FOXO3 did not affect LEF1 and TCF7 in U87MG cells and was primarily cytoplasmic in this setting (Figs 3C and  5A).Unraveling the contextual frameworks that dictate the impact of FOXO factors on the WNT pathway will allow for a greater understanding of the mechanisms that drive cancer progression.

Conclusion
FOXO transcription factors are emerging as key regulators of the WNT pathway in settings ranging from osteoblasts to cancer.This work demonstrates for the first time that FOXO1 promotes the expression of WNT pathway target genes LEF1 and TCF7 in BT549 BBC and U87MG GBM cells.FOXO1 RNAi targeting or disruption by CRISPR Cas9 genome editing led to reduced expression of these genes.Our work highlights caveats for utilizing AS1842856, as this drug modulates FOXO1 activity instead of merely inhibiting it.Based on our work, some FOXO1 targets were increased with AS1842856 treatment, such as TCF7 and TCF4, and some were significantly reduced, such as SOD2, INSR, and GADD45A.Over 70 peerreviewed publications employed AS1842856 to identify FOXO1 targets and may need reassessment.There are increasing cases in which FOXO1 regulates transcription in a DNA-binding-independent manner, including published targets TCF4 and HES1, as well as in our work regarding TCF7 (Fig. 4E) [72,73].

Fig. 1 .
Fig. 1.FOXO1 inhibition with AS1842856 induced a set of Hallmark WNT beta-catenin genes based on GSEA.(A) GSEA was performed with RNA-Seq data upon 1 lM AS1842856 treatment for 48 h in BT549 cells.Enrichment plot of Hallmark WNT and beta-catenin.The profile of the running enrichment score (ES) and rank-ordered list are shown for GSEA enrichment.(B) Heatmap of 27 core-enriched genes in HALLMARK_WNT_BETA_CATENIN signaling gene set.

Fig. 2 .
Fig. 2. AS1842856 treatment induced canonical WNT target genes in a series of BBC and GBM cancer cell lines.Indicated cell lines were treated with 1 lM AS1842856 for 48 h and examined for changes in gene expression by qRT-PCR using TUBB as the reference gene.AS1842856 treatment induced canonical WNT target genes in (A) BBC BT549 but not MDA-MB-468 BBC cells (B).(C-F) GBM cell lines had evidence of WNT target gene induction upon AS1842856 treatment; * denotes significantly different by Student's t-test compared with the control (P < 0.05) with SD error bars.Each experiment had three biologically independent replicates.

Fig. 3 .
Fig.3.AS1842856 treatment increased beta-catenin and decreased GSK3b protein expression.(A, B) Indicated BBC and GBM cell lines were treated with 0.1-1 lM AS1842856 for 48 h and examined for changes in gene expression by qRT-PCR using TUBB as the reference gene.AS1842856 treatment increased LEF1 gene expression.(C) BT549 and U87MG cells were treated with AS1842856 for 48 h, and then fractions were prepared; GAPDH and Histone H3 were cytoplasmic and nuclear controls respectively.Active and total beta-catenin were increased, whereas GSK3b was decreased.* denotes significantly different by Student's t-test compared with the control (P < 0.05) with SD error bars.Each experiment had three biologically independent replicates.

Fig. 4 .
Fig. 4. FOXO1 RNAi-targeting or disruption impacted WNT target gene expression.(A) FOXO1 and/or beta-catenin (CTNNB1) RNAi led to reduced LEF1, TCF7, and GADD45A gene expression 24 h post-transfection in U87MG cells by qRT-PCR using TUBB as the control.(B) FOXO1 RNAi led to reduced LEF, and GADD45A gene expression 18 h post-transfection in BT549 cells by qRT-PCR using TUBB as the control.(C, D) FOXO1 disruption by CRISPR Cas9 genome editing in U87MG cells led to reduced LEF1 and TCF7 gene expression measured by qRT-PCR using TUBB as the control; disruption of FOXO1 produced a truncated protein as assessed by western blot analysis.The disruption mutant is predicted to express the C-terminal portion of FOXO1 starting in the middle of the DNA-binding domain.(E) Exogenous FOXO1-H215R (DNA-binding-defective) significantly restored TCF7 gene expression to FOXO1 disruption mutants.* denotes significantly different by Student's t-Test compared with the control (P < 0.05) with SD error bars.Each experiment had three biologically independent replicates.

Fig. 5 .
Fig. 5. FOXO3 and FOXO4 RNAi-targeting reduced WNT target gene expression.(A) FOXO3 RNAi 24 h post-transfection in U87MG cells was assessed by qRT-PCR using TUBB as the control.FOXO3 RNAi had no significant impact on LEF1 or TCF7.(B) FOXO4 RNAi was assessed 24 h post-transfection U87MG cells and reduced LEF1, TCF7, and AXIN2 by qRT-PCR using TUBB as the control.(C, D) FOXO3 or FOXO4 RNAi was assessed 24 h post-transfection by RT-PCR using TUBB as the control in BT549 cells; FOXO3 RNAi samples impacted TCF7, whereas FOXO4 reduction samples had reduced LEF1 and TCF7.* denotes significantly different by Student's t-test compared with the control (P < 0.05) with SD error bars.Each experiment had three biologically independent replicates.