Mir-155 Is Downregulated in Familial Adenomatous Polyposis And

Published OnlineFirst August 2, 2018; DOI: 10.1158/1541-7786.MCR-18-0115

Oncogenes and Tumor Suppressors Molecular Cancer Research miR-155 Is Downregulated in Familial Adenomatous Polyposis and Modulates WNT Signaling by Targeting AXIN1 and TCF4 Anna Prossomariti1,2, Giulia Piazzi1,2, Leonarda D'Angelo2, Sara Miccoli3,4, Daniela Turchetti3,4, Chiara Alquati1,2, Claudio Montagna1,2, Franco Bazzoli1, and Luigi Ricciardiello1,2,4

Abstract

Adenomatous Polyposis Coli (APC) gene mutations are somatic second hit in the APC gene was found in adenomatous responsible for the onset of familial adenomatous polyposis polyps from 6 of 9 FAP patients. Heterozygous APC gene (FAP) and sporadic colorectal cancer and have been associated mutations in FAP patients were associated with altered expres- with miRNAs dysregulation. The capacity of miR-155, a cancer- sion of candidate miRNAs and increased levels of AXIN1 and related miRNA, to target components of the WNT/b-CATENIN AXIN2 mRNAs. miR-155-5p was downregulated in FAP pathway suggests that APC gene mutations, controlling patients and in the APC and b-CATENIN–mutant colorectal miRNAs expression, may critically regulate WNT/b-CATENIN cancer cell lines, and critically regulates WNT/b-CATENIN signaling. To this end, APC gene target sequencing was per- cascade by targeting both AXIN1 and TCF4. Importantly, formed on colonic adenomatous polyps and paired normal miR-155-5p may sustain long-term WNT/b-CATENIN activa- mucosa clinical specimens from FAP patients (n ¼ 9) to elu- tion in colorectal cancer cells upon WNT3A stimulation. cidate the role of miR-155-5p in APC-mutant setting. The expression of selected miRNAs and WNT/b-CATENIN signaling Implications: This study supports a key role of miR-155-5p components was characterized in FAP patients and non-FAP in modulating WNT/b-CATENIN signaling in colorectal cancer control subjects (n ¼ 5). miR-155-5p expression and functional and unravels a new mechanism for AXIN1 regulation which effects on WNT cascade, cell survival, growth, and apoptosis represents a potential therapeutic target in speciﬁc tumor were investigated in different colorectal cancer cell lines. A subtypes. Mol Cancer Res; 16(12); 1965–76. 2018 AACR.

Introduction detectable in more than 90% of sporadic colorectal cancer cases, represent a critical hit for colorectal cancer initiation in particular Familial adenomatous polyposis (FAP) is a rare autosomal in microsatellite stable cancers (3–5). dominant inherited syndrome caused by germline mutations in The APC gene encodes for a scaffolding multi-domain protein the Adenomatous Polyposis Coli (APC) gene. FAP patients develop crucially involved in the WNT/b-CATENIN signaling pathway (6). hundreds to thousands of colorectal adenomas, carrying a high Indeed, APC, together with AXIN1, GLYCOGEN SYNTHASE lifetime risk of developing colorectal cancer (1). According to the KINASE 3b (GSK3b), CASEIN KINASE I, and E3-UBIQUITIN Knudson's two-hit hypothesis, FAP patients have heterozygous LIGASE b-TRCP, constitute the "b-CATENIN destruction com- APC mutations in their normal tissues, but are biallelically altered plex" (7). In the absence of WNT signals, this complex mediates in neoplasms as a consequence of a somatic second hit in the b-CATENIN phosphorylation, ubiquitination, and proteasomal APC gene (2). Moreover, somatic mutations in the APC gene, degradation. Upon WNT stimulation, LOW-DENSITY LIPOPRO- TEIN RECEPTOR-RELATED PROTEINS 5/6 (LRP5/6), together with FRIZZLED and DISHEVELLED (DVL), mediate AXIN1 1Department of Medical and Surgical Sciences, University of Bologna, Bologna, 2 recruitment to the plasma membrane leading the release of Italy. Center for Applied Biomedical Research (CRBA), S.Orsola-Malpighi Hos- b pital, University of Bologna, Bologna, Italy. 3Medical Genetic Unit, S.Orsola- nonphosphorylated (active) -CATENIN from the destruction APC Malpighi Hospital, University of Bologna, Bologna, Italy. 4Center for the Studies complex and its cytosolic accumulation (8, 9). gene muta- of Hereditary Cancers, Department of Medical and Surgical Sciences, University tions, leading to b-CATENIN nuclear translocation and its inter- of Bologna, Bologna, Italy. action with TCF/LEF transcription factors, result in the partial Note: Supplementary data for this article are available at Molecular Cancer induction of WNT downstream targets (i.e., AXIN2, CYCLIND1, Research Online (http://mcr.aacrjournals.org/). and C-MYC; ref. 10). Indeed, for complete WNT signaling activa- APC- b-CATENIN– Corresponding Authors: Luigi Ricciardiello, University of Bologna, Via Massar- tion, or mutant colorectal cancer cell lines enti 9, PAD 11, Bologna 40138, Italy. Phone/Fax: 39-051-2143381; E-mail: require WNT ligands, in particular the canonical ligand WNT3A [email protected]; and Giulia Piazzi, Center for Applied Biomedical (11–13). Importantly, WNT3A sustains signaling activation Research, S.Orsola-Malpighi Hospital, University of Bologna, Via Massarenti 9, through AXIN1 degradation in the absence of functional APC (14). Bologna 40138, Italy. Phone/Fax: 39-051-2143902; E-mail: Despite the critical role of APC gene mutations in the onset [email protected] of colorectal cancer is widely recognized, colorectal cancer is a doi: 10.1158/1541-7786.MCR-18-0115 heterogeneous and complex disease characterized by different 2018 American Association for Cancer Research. molecular subtypes, involving alterations of multiple oncogenic

www.aacrjournals.org 1965

Prossomariti et al.

pathways and mechanisms, that would be relevant targets for the Experimental procedures on DNA isolation, the APC gene development of both preventive and therapeutic strategies (15). target sequencing, filtering of variants, and pathogenic predictions Accordingly, the same molecular subtypes have also been recently are detailed in Supplementary Information. Variants are reported identified in colorectal cancer cell lines (16). according to Human Genome Variation Society guidelines (22). miRNAs are functional small noncoding RNAs able to bind the 30-untranslated (30UTR) region of their target mRNAs, partially Cell line transfections and treatments inhibiting their expression by inducing mRNA degradation or The human colorectal cancer cell lines RKO, DLD-1, SW480, impairing its translation (17, 18). Noteworthy, because each and HCT116 were obtained from the ATCC. CACO-2 cells were miRNAs has several different cellular targets, also in the same purchased by ECACC. DLD-1 were cultured in RPMI-1640, where- biological pathway, their activity can be very complex and pivotal as RKO, SW480, CACO-2, and HCT116 in Iscove's Modified for development and progression of diseases (19). Recently, an Dulbecco Media, respectively (EuroClone). Culture media were aberrant expression profile of several miRNAs has been found in supplemented with 10% FBS, 100 U/mL penicillin, 100 mg/mL colonic adenomatous polyps and colorectal cancer specimens streptomycin, and 2 mmol/L glutamine (Euroclone), and cells (20, 21). However, because colorectal cancer tissues harbor mul- were maintained at 37 C and 5% CO2. For miR-155 induction, tiple mutations in different genes, establishing whether a specific colorectal cancer cell lines were grown in antibiotic-free genetic alteration is responsible for miRNAs deregulation could Opti-MEM media (Gibco; Thermo Fisher Scientific) and be difficult to ascertain. Thus, FAP patients, carrying germline transiently transfected with 50 nmol/L of miR-155-5p Pre-miR mutations in the APC gene, are ideal candidates to identify miRNA Precursor (PM12601; Ambion, Thermo Fisher Scientific; miRNAs deregulated in response to WNT signaling activation. pre-miR-155-5p) or with Pre-miR miRNA Precursor Negative We hypothesize that APC gene mutations might be responsible for Control (NC, AM17110; Ambion, Thermo Fisher Scientific; aberrant miRNAs expression which in turn pivotally controls pre-miR-NC) for 24 hours using Lipofectamine 2000 (Invitrogen; WNT/b-CATENIN signaling. Thermo Fisher Scientific) according to the manufacturer's In this study, we identified a set of APC-regulated miRNAs in instructions. FAP patients. Importantly, miR-155-5p was downregulated both For WNT3A stimulation, DLD-1 and SW480 cells were stimu- in colonic tissues from FAP patients and APC- or b-CATENIN– lated with human recombinant WNT3A (50 ng/mL; R&D mutant colorectal cancer cell lines, and acted as a critical regulator Systems) for 1 hour. Cell lines have been authenticated through of WNT/b-CATENIN signaling, targeting both AXIN1 and TCF4 short tandem repeat (STR) profiling by using the AmpF‘ (also named TCF7L2) genes. In addition, we demonstrated that STR Identifiler PCR Amplification Kit according to the manufac- miR-155-5p in combination with WNT3A stimulation resulted in turer's instructions and the GeneMapper ID 3.5 software (Thermo a rapid AXIN1 degradation leading to a boosted WNT activation. Fisher Scientific). Our data clarify the role of miR-155-5p in the modulation of WNT/b-CATENIN signaling, providing new relevant insight into RNA extraction and quantitative real-time PCR AXIN1 regulation in APC-mutant settings. Total RNA was extracted from human colonic fresh tissues and colorectal cancer cell lines using the TRIzol reagent (Invitrogen; Thermo Fisher Scientific) according to the manufacturer's proto- Materials and Methods col. RNA concentration was measured using the Nanodrop 1000 Human tissues collection spectrophotometer (Thermo Fisher Scientific). For miRNA anal- Nine patients with clinical and genetic diagnosis of FAP were ysis, total RNA was reverse transcribed using the TaqMan Micro- recruited at the S.Orsola-Malpighi Hospital (Bologna, Italy). RNA Reverse Transcription Kit or TaqMan Advanced miRNA Small colonic adenomatous polyps (<5 mm) and paired mor- cDNA Synthesis Kit (Applied Biosystems; Thermo Fisher phologically appearing normal mucosa (NM) samples (at least Scientific) according to the manufacturer's instructions. The levels 3 cm from adenomatous polyps) were collected from recruited of mature hsa-miR-15a-5p (Assay ID: 477858_miR), hsa-miR- FAP patients during endoscopy. Fresh tissues were immediately 17-5p (Assay ID: 478447_miR), hsa-miR-21-5p (Assay ID: frozen in liquid nitrogen and stored at 80C until use. If 477975_miR), hsa-miR-135a-5p (Assay ID: 478581_miR), hsa- necessary, colonic adenomatous polyps were fixed in formalin miR-16-5p (Assay ID: 477860_miR), hsa-miR-155-5p (Assay ID: for histopathologic evaluation. 002623), hsa-miR-423-5p (Assay ID: 478090_miR), and RNU6B Fresh NM colonic tissues from five fecal immunochemical test– (Assay ID: 001093) were assessed using TaqMan MicroRNA positive non-FAP control subjects (ages 50–70 years) were used as Assays or TaqMan Advanced miRNA Assays (Applied Biosystems; reference samples for gene expression analyses. Thermo Fisher Scientific). The study was conducted in accordance with the Declaration of For mRNA analysis, total RNA was converted to cDNA using the Helsinki and approved by the Ethics Committee of the S.Orsola- High-Capacity RNA-to-cDNA Kit (Applied Biosystems; Thermo Malpighi Hospital (Bologna, Italy). Written-informed consent Fisher Scientific) according to the manufacturer's instructions. was obtained from each patient. Quantitative real-time PCR (q-PCR) reactions were performed in triplicate on a Mx3000P QPCR thermal cycler (Stratagene; APC gene target sequencing and in silico analysis Thermo Fisher Scientific) using the SYBR Select Master Mix for APC gene target sequencing was performed on 10 colonic CFX or with TaqMan Gene Expression Master Mix for mRNAs and adenomatous polyps and 9 adjacent morphologically NM sam- TaqMan Universal Master Mix or TaqMan Fast Advanced Master ples from enrolled FAP patients using the Ion Torrent Personal Mix for miRNAs (Applied Biosystems; Thermo Fisher Scientific). Genome Machine System (Thermo Fisher Scientific). All exons Primer sequences used for qPCR are AXIN2 (forward primer: and the 30UTR region of the APC gene (RefSeq NM_000038.5) 50-GGA CAA ATG CGT GGA TAC CT-30; reverse primer: 50-TGC were sequenced. TTG GAG ACA ATG CTG TT-30), CYCLIND1 (forward primer:

1966 Mol Cancer Res; 16(12) December 2018 Molecular Cancer Research

miR-155-5p Regulates Wnt Signaling in APC-Mutant Settings

50-AGG AGC TGC TGC AAA TGG-30; reverse primer: 50-GGC ATT activity was used to normalize Firefly luciferase activity for each TTG GAG AGG AAG TG-30), C-MYC (forward primer: 50-CGT AGT sample. TGT GCT GAT GTG TGG-30; reverse primer: 50-CTC GGA TTC TCT GCT CTC CTC-30), and TCF1 (forward primer: 50-CAG ACG GAT Apoptosis evaluation TGG TGT GGT C-30; reverse primer: 50-GCA CTG TCA TCG GAA DLD-1 and SW480 cells (104 cells/well) were seeded in 96-well GGA AC-30). mRNA expression of AXIN1 (Assay ID: plates in quadruplicate and transfected with pre-miR-155-5p or Hs00394718-m1), TCF4/TCF7L2 (Assay ID: Hs01009044-m1), pre-miR-NC (50 nmol/L) for 24 hours. CASPASE-3 and -7 activ- and GAPDH (Assay ID: Hs03929097-g1) was analyzed using ities were measured using the Caspase-Glo 3/7 Assay (Promega) TaqMan probes (Applied Biosystems; Thermo Fisher Scientific). following the manufacturer's instructions. Cleaved and total GAPDH was used as reference gene for mRNA normalization. CASPASE-3 protein expression was also analyzed through RNU6B or miR-423-5p were used as endogenous controls for Western blot as indicated above. miRNA normalization. DD Fold induction levels were obtained using the 2 Ct method, Cell counting, viability, and colony-forming assays by normalizing against the appropriate reference gene or endog- Cell counting was performed using trypan blue at 24 hours enous control. after transfection. For cell viability, 104 cells/well were seeded in 96-well plates in quadruplicate, transfected with 50 nmol/L of Western blot analysis pre-miR-155-5p or pre-miR-NC for 24 hours, and analyzed using Total proteins were extracted from colorectal cancer cells using CellTiter-Glo Luminescent Cell Viability Assay (Promega) accord- RIPA buffer. Nuclear and cytosolic protein extracts were obtained ing to the manufacturer's recommendations. For colony-forming using the NE-PER nuclear extraction Kit according to the manu- assays, 24 hours after transfection with pre-miR-155-5p or facturer's instructions (Pierce; Thermo Fisher Scientific). Proteins pre-miR-NC, cells (2 103 cells/well for DLD-1 and 500 cells/ were separated on a 4%–12% NuPAGE Novex Bis-Tris Gels well for SW480) were seeded in 6-well plates in triplicate and (Invitrogen; Thermo Fisher Scientific) in MOPS buffer (Novex; cultured until colony formation. Then, colonies were washed Thermo Fisher Scientific) and transferred onto nitrocellulose twice in PBS, fixed in paraformaldehyde 4%, stained with 0.5% membrane. After blocking, membranes were incubated overnight crystal violet, and counted using ImageJ software (NIH; https:// at þ4C with the following primary antibodies: AXIN1, phospho– imagej.nih.gov/ij/). b-CATENIN (Ser33/37/Thr41), total b-CATENIN, phospho- GSK3b (Ser9), total GSK3b, TCF4/TCF7L2, cleaved CASPASE-3 Bioinformatic and statistical analysis (Asp175), and total CASPASE-3. GAPDH was used as reference Interaction of miR-155-5p with the AXIN1 target gene was protein for total and cytosolic fractions. LAMINB was used as predicted using the miRNA target prediction algorithm PITA (23). reference protein for nuclear fraction. Primary antibodies' spec- Statistical analysis was performed using Graphpad 5.0 Software ification and condition are shown in Supplementary Table S1. (GraphPad Software Inc.). One-way ANOVA was used to compare After incubation with appropriate secondary horseradish perox- mean differences among three or more groups, whereas unpaired idase–conjugated antibodies (Amersham Protran; GE Health- and paired t tests were applied to calculate the mean differences care), the signal was detected with a luminol enhancer solution between two unmatched or matched groups, respectively. One (WESTAR EtaC; Cyanagen), and images were acquired using sample t test was used to measure differences between a sample the Chemidoc XRSþ (Bio-Rad). Densitometry analyses were mean and a reference value. Whenever necessary, the values were performed using Image Lab software (Bio-Rad). square root transformed to stabilize the variance. For q-PCR and Western blot analyses, data were presented upon square root Dual-luciferase assays transformation. Data are shown as mean SEM or as box and TOPFlash reporter plasmid (containing wild-type TCF-binding whiskers plot (min to max). P values < 0.05 were regarded as sites) was used to evaluate WNT signaling activity. The wild-type- statistically significant. 30UTR clone of AXIN1 (WT-30UTR-AXIN1; SC209177; OriGene), the mutant-30UTR clone of AXIN1 (MUT-30UTR-AXIN1; CW302969; custom mutated by OriGene modifying the wild- Results type SC209177 clone in the miR-155-5p seed region as follow: APC gene target sequencing AGCA to UCGU), or the wild-type-30UTR clone of TCF4/TCF7L2 To identify FAP patients harboring a somatic second hit in the (WT-30UTR-TCF4; OriGene) was used to assess the inhibitory APC gene, we performed APC gene target sequencing on colonic effect of miR-155-5p on the selected target genes. adenomatous polyps and paired morphologically NM. Clinical For luciferase assays, DLD-1, SW480 (105 cells/well), and characteristics of enrolled FAP patients are reported in Table 1. All CACO-2 (80,000 cells/well) cells were seeded in 24-well germline APC gene mutations were located outside of the muta- plates in antibiotic-free Opti-MEM media (Gibco; Thermo tion cluster region of the APC gene (codons 1,286–1,514; Fisher Scientific), in triplicate. Pre-miR-155-5p or pre-miR-NC ref. 24; Table 2) and were confirmed at target sequencing in both (50 nmol/L), Renilla (40 ng), and the reporter plasmids, NM and adenomatous polyp samples. WT-30UTR-AXIN1 (50 ng), MUT-30UTR-AXIN1 (50 ng), No additional disease-causing variants were found in morpho- WT-30UTR-TCF4 (50 ng), or TOPFlash (200 ng), were cotrans- logically NM samples. fected in all cell lines using Lipofectamine 2000 (Invitrogen; We were able to identify somatic APC gene second hits in Thermo Fisher Scientific). colonic adenomatous polyps from 6 of 9 patients (Table 3). For Twenty-four hours after transfection, luciferase activity was patients P7 and P8, no disease-causing second hits in the APC determined using the Dual-Luciferase Reporter Assay (Promega) gene were found. No sequencing data on colonic adenomatous according to the manufacturer's instructions. Renilla luciferase polyps from patient P5 were available.

www.aacrjournals.org Mol Cancer Res; 16(12) December 2018 1967

Prossomariti et al.

Table 1. Clinical characteristics of enrolled FAP patients Patient ID Gender Age (years) Prophylactic surgery Location of collected samples FAP_01 M 56 Colectomy þ IRA Rectum FAP_02 F 30 — Ascending and Transverse FAP_03a M29 — Ascending FAP_04 M 28 — NA FAP_05b F20 — Ascending FAP_06b M 53 Colectomy þ IRA Rectum FAP_07a F20 — NA FAP_08 M 53 Proctocolectomy Ileal pouch FAP_09 M 42 — Ascending and Transverse Abbreviations: F, female; IRA, ileorectal anastomosis; M, male; NA, not available. arelated patients. brelated patients.

All the adenomas undergoing to the histologic analysis were cell line, which is wild-type for both APC and b-CATENIN genes. classified as tubular adenomas with low-grade dysplasia. Interestingly, we found significantly lower miR-155-5p levels in the APC and b-CATENIN–mutant cell lines compared with the APC gene mutations lead to miRNA deregulation and wild-type cell line RKO (Supplementary Fig. S1A). To investigate AXIN1/2 mRNA induction in FAP patients the effects of miR-155-5p on WNT/b-CATENIN cascade, we To gain insight into whether APC loss of function leads to a transiently transfected APC- and b-CATENIN–mutant colorectal specific miRNA signature, we evaluated the expression of selected cancer cell lines with an miR-155-5p precursor (pre-miR-155-5p) miRNAs (miR-15a-5p, miR-16-5p, miR-17-5p, miR-21-5p, obtaining significantly higher levels of miR-155-5p in all cell lines miR-135a-5p, and miR-155-5p) involved in WNT/b-CATENIN (Supplementary Fig. S1B). Upon miR-155-5p induction, AXIN1 signaling regulation (25, 26), in colonic tissues from 9 FAP protein levels were reduced in DLD-1, SW480, and CACO-2 cell patients and 5 non-FAP control subjects. We found significantly lines, with a stronger effect on DLD-1 cells (Fig. 2A; Supplemen- higher levels of miR-21-5p (Fig. 1A), miR-135a-5p (Fig. 1B), and tary Fig. S2A, S2B, and S2C). No relevant changes in AXIN1 miR-17-5p (Fig. 1C), and significantly lower levels of miR-155-5p protein were observed in HCT116 (Supplementary Fig. S3A and (Fig. 1D) in both colonic NM and adenomatous polyps from FAP S3B). Interestingly, upon miR-155-5p overexpression, phospho– patients compared with non-FAP controls, whereas no significant b-CATENIN was decreased in DLD-1 and increased in SW480, changes in the expression level of miR-15a-5p (Fig. 1E) and miR- CACO-2 (Fig. 2A; Supplementary Fig. S2A, S2B, and S2C), and 16-5p (Fig. 1F) were observed. Importantly, no significant HCT116 cell lines (Supplementary Fig. S3A and S3B). changes in miRNA levels were detected between NM and matched The analysis of cytosolic/nuclear b-CATENIN distribution adenomatous polyps in FAP patients. showed no significant changes in any of the tested cell lines, Subsequently, we evaluated the mRNA expression levels of although a slight increase of cytosolic b-CATENIN was found in AXIN1 and AXIN2. Compared with non-FAP controls, we found SW480, CACO-2, and HCT116 cell lines (Fig. 2A; Supplementary significantly increased levels of AXIN1 mRNA in both NM and Figs. S2E and S2F; S3A and S3C). Moreover, a concomitant minor adenomatous polyps of FAP patients (Fig. 1G) and AXIN2 mRNA reduction in nuclear b-CATENIN levels was observed in CACO-2 in NM of FAP patients (Fig. 1H). cells during miR-155-5p upregulation (Fig. 2A; Supplementary These results suggest that miRNA deregulation may contribute Fig. S2F). to the disease development in FAP patients. Because GSK3b mediates b-CATENIN phosphorylation (29), we also assessed GSK3b phosphorylation at Ser9 which inhibits its miR-155-5p regulates WNT/b-CATENIN signaling in kinase activity (30). GSK3b phosphorylation was unchanged in APC-mutant colorectal cancer cell lines DLD-1, CACO-2 (Fig. 2A; Supplementary Fig. S2A and S2C), and To evaluate whether miR-155-5p levels were decreased HCT116 cells (Supplementary Fig. S3A and S3B), whereas a in colorectal cancer cell lines with constitutively active reduction of phosphorylated GSK3b (p-GSK3b) was observed in WNT/b-CATENIN signaling as well, we measured miR-155-5p SW480 cells upon miR-155-5p overexpression (Fig. 2A; Supple- levels in DLD-1, SW480, and CACO-2 cell lines which harbor mentary Fig. S2B). Importantly, miR-155-5p induction led to different APC-truncating mutations (27), in HCT116 cells that TCF4 protein downregulation in DLD-1, SW480, and CACO-2 carry a heterozygous mutation in b-CATENIN (28), and in RKO cell lines (Fig. 2A; Supplementary Fig. S2A, S2B, and S2C),

Table 2. Germline APC gene mutations in enrolled FAP patients Patient ID Position (hg19) Exon Variant c.DNA (NM_000038.5) Coverage (X) Protein effect Variant type FAP_01 5:112174625 16 c.3336_3340delAAATC 1,989 p.(Asn1113Serfs4) Frameshift FAP_02 5:112154723 10 c.994C>T 1,802 p.(Arg332) Nonsense FAP_03 5:112164616 14 c.1690C>T 1,999 p.(Arg564) Nonsense FAP_04 5:112174123 16 c.2833delA 1,995 p.(Arg945Glyfs10) Frameshift FAP_05 5:112174489 16 c.3202_3205delTCAA 1,982 p.(Ser1068Glyfs57) Frameshift FAP_06 5:112174489 16 c.3202_3205delTCAA 1,986 p.(Ser1068Glyfs57) Frameshift FAP_07 5:112164616 14 c.1690C>T 2,000 p.(Arg564) Nonsense FAP_08 5:112151204 9 c.847C>T 1,999 p.(Arg283) Nonsense FAP_09 5:112074269 — c.532-2A>T 1,996 — Splice variant

1968 Mol Cancer Res; 16(12) December 2018 Molecular Cancer Research

miR-155-5p Regulates Wnt Signaling in APC-Mutant Settings

Table 3. Somatic APC gene second hits in adenomatous polyps from FAP patients Patient ID Position (hg19) Exon Variant c.DNA (NM_000038.5) Coverage (X) VAF (%) Mutant cells (%) Protein effect Variant type FAP_01P 5:112175348 16 c.4057G>T 1,999 5.7 11.3 p.(Glu1353) Nonsense FAP_02P 5:112173830 16 c.2544_2545insA 1,976 7.5 15.1 p.(Asp849Argfs2) Frameshift FAP_03P1 5:112175513 16 c.4222G>T 1,991 10.7 21.5 p.(Glu1408) Nonsense FAP_04P1 5:112175751 16 c.4460C>T 1,999 2.9 5.7 p.(Thr1487Ile) Missense FAP_04P2 5:112175576 16 c.4285C>T 2,000 8.8 17.6 p.(Gln1429) Nonsense FAP_06P 5:112175639 16 c.4348C>T 1,994 17.8 35.5 p.(Arg1450) Nonsense FAP_09P 5:112175639 16 c.4348C>T 1,658 6.6 13.3 p.(Arg1450) Nonsense Abbreviations: P, polyps; VAF, variant allele frequency (number of variant reads/number of total reads).

Figure 1. miRNAs and AXIN1/2 expression levels in FAP patients and non-FAP control subjects. Relative expression of (A) miR-21-5p, (B) miR-135a-5p, (C) miR-17-5p, (D) miR-155-5p, (E) miR-15a-5p, (F) miR-16-5p, (G) AXIN1,and(H) AXIN2 in FAP patients (n ¼ 9) and non-FAP control subjects (n ¼ 5). Data are shown as box and whiskers plot (min to max). Comparisons between FAP patients and non-FAP control subjects were performed using the unpaired t test, whereas differences between matched NM and adenomatous polyps from FAP patients were analyzed using the paired t test. , P < 0.05; , P < 0.01; , P < 0.001 compared with non-FAP control subjects.

www.aacrjournals.org Mol Cancer Res; 16(12) December 2018 1969

Prossomariti et al.

Figure 2. Effects of miR-155-5p induction on WNT/b-CATENIN signaling pathway in colorectal cancer cells. A, Western blot representative images for WNT signaling components in DLD-1, SW480, and CACO-2 cell lines transfected with miR-155-5p or NC precursor. WNT downstream targets mRNA analysis in (B) DLD-1, (C) SW480, and (D) CACO-2 transfected with miR-155 or NC precursor. Graphs show results from at least three independent experiments. Data are shown as mean of square root–transformed values SEM. Comparisons were performed by the unpaired t test. , P < 0.05; , P < 0.01; and , P < 0.001 compared with NC. E, TOPFlash reporter assay on DLD-1, SW480, and CACO-2 cotransfected with miR-155 or NC, Renilla, and TOPFlash reporter plasmid. Data represent the normalized luciferase activity obtained calculating the Fireﬂy/Renilla luciferase ratio for each sample. Graphs report results from three independent experiments. Statistical signiﬁcance was calculated using the one sample t test. , P < 0.05 and , P < 0.01 compared with NC.

whereas no variations in TCF4 protein levels were observed in A modest increase in AXIN2 mRNA levels was appreciated in HCT116 (Supplementary Fig. S3A and S3B). all the analyzed cell lines upon miR-155 induction (Fig. 2B–D; We further characterized the effects of miR-155-5p on Supplementary Fig. S3D). b-CATENIN transcriptional activity evaluating the mRNA expres- These data were confirmed by the TOPFlash reporter sion profile of selected WNT signaling downstream targets. In activity analysis which showed no changes in luciferase accordance with the effects of miR-155-5p on b-CATENIN and activity in DLD-1 and a significant reduction in both TCF4 proteins, we found no changes in CYCLIND1 and a modest SW480 and CACO-2 cell lines upon miR-155-5p induction increase of C-MYC and TCF1 mRNAs in DLD-1 cells (Fig. 2B), (Fig. 2E). and a slight increasing trend in all the same targets in HCT116 Altogether, our results demonstrate that miR-155-5p, by cells (Supplementary Fig. S3D). Otherwise, CYCLIND1, C-MYC, modulating AXIN1 and TCF4 protein expression, has different and TCF1 mRNAs were reduced in both SW480 (Fig. 2C) and cell-line–dependent effects on WNT cascade in APC-mutant colo- CACO-2 cells (Fig. 2D). rectal cancer cell lines.

1970 Mol Cancer Res; 16(12) December 2018 Molecular Cancer Research

miR-155-5p Regulates Wnt Signaling in APC-Mutant Settings

Figure 3. miR-155-5p binding to the AXIN1 30UTR. A, Schematic representation of base pairing between miR-155-5p and AXIN1-mRNA-30UTR. The mutant seed sequence of the AXIN1-30UTR is indicated in gray. Dual-luciferase reporter assays of (B) DLD-1, (C) SW480, and (D) CACO-2 transfected with miR-155-5p or NC precursors, Renilla, and either the wild-type or mutant 30UTR clone of AXIN1. Data represent the normalized luciferase activity obtained calculating the Fireﬂy/Renilla luciferase ratio for each sample. Graphs report results from three independent experiments. Statistical signiﬁcance was calculated using one sample t test. , P < 0.05 and , P < 0.01 compared with NC.

AXIN1 and TCF4 are targets of miR-155-5p miR-155-5p sustains WNT signaling activation in APC-mutant We showed that miR-155-5p induction was associated with colorectal cancer cells by promoting AXIN1 degradation upon AXIN1 and TCF4 protein destabilization in APC-mutant cell WNT stimulation lines. Thus, we hypothesized that both AXIN1 and TCF4 AXIN1 protein degradation, upon WNT3A stimulation, has might be direct targets of miR-155-5p. Noteworthy, although been described as a critical mechanism for regulating AXIN1/ TCF4 has been previously described as direct target of miR- b-CATENIN interaction during WNT signaling activation (11, 12). 155-5p in cervical and breast cancer cell lines (31, 32), the In particular, a relevant decrease in the AXIN1 protein has been relationship between miR-155-5p and AXIN1 remained to reported 4 hours upon WNT3A stimulation (15). be further defined. Hence, using the miRNA target prediction To establish the effects of a short-time treatment with WNT3A algorithm PITA (23), we first verified the presence of an alone or in combination with miR-155-5p, we treated both miR-155-5p–responsive element in the AXIN1 30UTR mRNA untransfected and transfected colorectal cancer cells with WNT3A finding one putative binding site for miR-155-5p with for 1 hour. Importantly, although the WNT3A stimulation caused seed sequence placed at positions 221 to 228 bp (with ddG a slight increase in AXIN1 protein and an accumulation of of 0.92). phosphorylated b-CATENIN in both DLD-1 and SW480 untrans- Then, we conducted dual-luciferase assays cotransfecting fected cell lines (Fig. 4A; Supplementary Fig. S6A and S6B), the DLD-1, SW480, and CACO-2 cells with pre-miR-155-5p or exposure of miR-155-5p–transfected colorectal cancer cells to pre-miR-NCinthepresenceofaluciferase vector containing WNT3A led to a marked reduction of AXIN1 protein and an either the wild-type (WT-AXIN1-30UTR) or the mutant 30UTR impairment of b-CATENIN phosphorylation in both cell lines sequence of AXIN1 (MUT-AXIN1-30UTR; Fig. 3A). As expected, (Fig. 4A; Supplementary Fig. S6C and S6D). No significant miR-155 induction significantly reduced the luciferase activity changes in cytosolic and nuclear b-CATENIN distribution were of wild-type WT-AXIN1-30UTR in all cell lines, whereas trans- observed in both cell lines treated with WNT3A alone (Fig. 4A; fectionwithMUT-AXIN1-30UTR had no effect on luciferase Supplementary Fig. S6E and S6F). Interestingly, although in miR- activity in DLD-1 and a slight effect on SW480 and CACO-2 155–transfected DLD-1 cells the treatment with WNT3A increased cells (Fig. 3B–D).Furthermore,wealsoconfirmed the ability nuclear b-CATENIN and significantly reduced its cytosolic frac- of miR-155-5p to target the wild-type TCF4-30UTR region in tion (Fig. 4A; Supplementary Fig. S6G), a modest increase in DLD-1, SW480, and CACO-2 cell lines (Supplementary cytosolic b-CATENIN was observed in SW480 (Fig. 4A; Supple- Fig. S4). These results confirm that both AXIN1 and TCF4 mentary Fig. S6H). Moreover, total and phosphorylated GSK3b are targets of miR-155 in APC-mutant cell lines. Noteworthy, were not affected by WNT3A stimulation either when used alone miR-155-5p induction was associated with AXIN1 and TCF4 or when used in combination with miR-155-5p induction in both protein degradation in DLD-1, SW480, and CACO-2 cell lines, cell lines (Fig. 4A; Supplementary Fig. S6A–S6D). whereas no changes in AXIN1 and TCF4 mRNAs expression Noteworthy, although treatment with WNT3A alone led to an levels were observed with the exception of CACO-2 in which increase in TCF4 protein in SW480 cells (Fig. 4A; Supplementary TCF4 mRNA was also significantly reduced (Supplementary Fig. S6B), TCF4 protein downregulation, observed upon miR-155 Fig. S5A–S5C). Based on these data, we concluded that miR- induction, was maintained in cell lines transfected with pre-miR- 155 acts on these targets mainly by impairing their translation 155-5p and stimulated with WNT3A (Fig. 4A; Supplementary rather than affecting their mRNAs stability. Fig. S6C and S6D). As regard the effects on WNT downstream

www.aacrjournals.org Mol Cancer Res; 16(12) December 2018 1971

Prossomariti et al.

Figure 4. Effects of WNT3A and miR-155 on canonical WNT cascade in DLD-1 and SW480 cells. A, Western blot analysis for WNT signaling components in DLD-1 and SW480 stimulated with recombinant WNT3A or transfected with miR-155 or NC and then treated with WNT3A for 1 hour. WNT downstream target mRNA levels in (B) DLD-1 and (C) SW480 treated with WNT3A alone or in (D) DLD-1 and (E) SW480 transfected with miR-155 or NC and treated with WNT3A. Graphs show results from at least two independent experiments. Data are shown as mean of square root–transformed values SEM. Comparisons were performed by two-tailed unpaired t test. , P < 0.05 compared with NC þ WNT3A.

1972 Mol Cancer Res; 16(12) December 2018 Molecular Cancer Research

miR-155-5p Regulates Wnt Signaling in APC-Mutant Settings

targets' mRNAs expression, no changes were observed in DLD-1 consequent strong hyperproliferative stimulus, early lesions, and SW480 cells upon WNT3A stimulation (Fig. 4B and C). while harboring a second inactivating hit in the APC gene, had Otherwise, in DLD-1 cells transfected with pre-miR-155-5p and an incomplete WNT signaling activation while showing miRNA's treated with WNT3A, a significant increase of C-MYC, TCF1, and alteration yet. AXIN2 mRNA levels was observed (Fig. 4D). Differently, in A panel of APC-regulated miRNAs has been recently described SW480 cells, the downregulation of the WNT downstream targets in a colorectal cancer cell line and in a mouse model of FAP observed in presence of miR-155-5p alone persisted also upon (25, 36). Noteworthy, to the best of our knowledge, this is the first WNT3A treatment despite not reaching statistical significance study evaluating miRNA's expression level in both NM and (Fig. 4E). adenomatous polyp tissues from FAP patients. Our data suggested that miR-155-5p may contribute to desta- In particular, in this study, we focused our attention on bilize the b-CATENIN destruction complex in APC-mutant colo- miR-155-5p, a relevant cancer-related miRNA. Previous studies rectal cancer cells upon WNT stimulation by promoting AXIN1 reported increased levels of miR-155 in patients with sporadic degradation. colorectal cancer suggesting a potential oncogenic role of this miRNA (37, 38). However, we found a significant downregula- miR-155-5p affects cell survival, growth, and apoptosis in tion of miR-155-5p in FAP patients, as well as in the APC or APC-mutant colorectal cancer cell lines b-CATENIN–mutant colorectal cancer cell lines. Previously pub- The role of WNT signaling on cell proliferation and apoptosis lished data reported an epigenetic silencing of miR-155-5p as a is widely recognized (33, 34). Based on our data showing that causal role of miR-155-5p downregulation in these colorectal miR-155-5p regulates WNT cascade in APC-mutant settings, we cancer cell lines (39). Moreover, miRNAs are frequently involved hypothesized that miR-155-5p could also affect cell growth, in mutual feedback regulatory loops which critically control the proliferation, and apoptosis. activation of oncogenic pathways, including WNT signaling (40). We found that miR-155-5p significantly reduced colony Thus, it is possible that effectors of WNT pathway, which is formation (Fig. 5A and B) and cell number in miR-155-5p– activated in FAP patients as a consequence of germ-line APC gene transfected DLD-1 and SW480 cell lines (Fig. 5C). However, cell mutations, may lead to miR-155 transcriptional repression. viability assay conducted on DLD-1 and SW480 showed no Intriguingly, we showed that FAP patients with low miR-155- differences upon miR-155-5p upregulation on both DLD-1 and 5p levels had significantly higher levels of AXIN1 and AXIN2 SW480 cell lines (Fig. 5D). mRNAs. An upregulation of AXIN2 was previously reported in In order to clarify whether increased apoptosis could be respon- adenomas from FAP patients and sporadic colorectal cancer sible for the reduced cell survival and growth observed upon specimens, whereas, to the best of our knowledge, AXIN1 expres- miR-155-5p induction in DLD-1 and SW480 cells, we also mea- sion has not yet been investigated in FAP. sured CASPASE 3/7 activities in these cells. In DLD-1, but not in Among WNT signaling core components, AXIN1 scaffold pro- SW480, miR-155-5p led to a significant increase of caspase activity tein is critical for controlling b-CATENIN levels stabilizing the (Fig. 5E). Moreover, CASPASE-3 activation in DLD-1 was also destruction complex (14). Otherwise, the interaction between confirmed by the increase of cleaved CASPASE-3 protein (Fig. 5F; AXIN1 and phospho-LRP6, occurring early after WNT stimula- Supplementary Fig. S7A). Accordingly, no differences in cleaved tion, is a crucial step to trigger WNT signaling activation, thus CASPASE-3 were found in SW480 cells (Supplementary Fig. S7B). supporting a dual role of AXIN1 in WNT signaling regulation Our data suggested that miR-155-5p, by modulating canonical (41, 42). Importantly, an increase of AXIN1 in the initial phases of WNT signaling, controls cell survival and apoptosis in WNT signaling activation has recently been demonstrated (43), APC-mutant colorectal cancer cells. and new evidence suggests that AXIN1 degradation does not have a causal role for WNT signaling initiation, rather representing a Discussion later event aimed at sustaining the prolonged signaling activation (44). Thus, based on our results and literature data, we speculate Mutations in the APC gene represent a "gatekeeping" event in that in FAP patients WNT signaling initiation is enhanced as a colorectal cancer initiation (35). In this study, we first aimed to consequence of miR-155-5p downregulation and AXIN1 induc- identify APC-regulated miRNAs in 9 FAP patients finding a panel tion, whereas the ectopic miR-155-5p expression would lead to of four WNT-related miRNAs (miR-21-5p, miR-135a-5p, miR-17- AXIN1 degradation and long-term WNT signaling activation. 5p, and miR-155-5p) differentially expressed between NM sam- APC gene targeting has been previously proposed as one of ples from FAP patients and non-FAP controls. Importantly, these the main mechanisms for WNT signaling activation by miR-155 results showed that alterations in the expression of specific miR- (19, 45, 46). In this work, we first demonstrated that AXIN1 is a NAs may have a role in disease progression in FAP patients. direct target of miR-155-5p in APC-mutant colorectal cancer cells Unexpectedly, we found no significant differences in miRNA and confirmed evidence obtained in cervical and breast cancer levels between NM and matched adenomatous polyps in FAP cells on its ability to target TCF4 (31, 32). Although we demon- patients despite most of the analyzed adenomatous lesions (6 of strated a direct effect of miR-155-5p on AXIN1-30UTR, we also 9) harbor a causative somatic second hit in the APC gene. believe that miR-155 might promote AXIN1 destabilization by However, we cannot rule out an APC LOH event or promoter additional indirect mechanisms. Conversely, no direct effects on hypermethylation as possible second hit in the remaining 3 AXIN1 and TCF4 proteins were observed in the b-CATENIN– patients. mutant cell line HCT116 upon miR-155-5p induction. These Even though this result could be counterintuitive, we believe results suggest that b-CATENIN–mutant cell lines might be resis- that it might be due to the small size of the analyzed adenomatous tant to miR-155-5p effects on WNT signaling. polyps (<5 mm). Indeed, differently from advanced lesions Interestingly, we found that miR-155 induction in SW480 and characterized by a hyperactivation of the WNT signaling and CACO-2 led to a reduction of WNT/b-CATENIN downstream

www.aacrjournals.org Mol Cancer Res; 16(12) December 2018 1973

Prossomariti et al.

Figure 5. miR-155-5p effects on cell growth, survival, and apoptosis in DLD-1 and SW480 cells. (A) Representative images and (B) quantiﬁcation of clonogenic assay on DLD-1 and SW480 cells transfected with miR-155 or NC. (C) Cell count, (D) cell viability, (E) CASPASE 3/7 activity, and (F) Western blot representative images of cleaved and total CASPASE-3 on DLD-1 and SW480 cells transfected with miR-155 or NC. Graphs represent results from three independent experiments. Statistical signiﬁcance was calculated using one sample t test. , P < 0.05 and , P < 0.01 compared with NC.

targets, whereas a slightly increased expression of these targets ﬁndings observed regarding the miR-155-5p regulation of WNT was observed in DLD-1. Thus, we hypothesized that when signaling might be strongly dependent on the molecular subtype miR-155-5p exerts a major effect on its target AXIN1, as it of the tested colorectal cancer cell lines which partially reﬂect occurs in DLD-1, even in the presence of TCF4 downregulation, colorectal cancer heterogeneity in patients. For instance, miR-155 b-CATENIN is able to activate WNT downstream targets. On the was found to be overexpressed in microsatellite instability (MSI) other hand, when the effect of miR-155-5p on TCF4 is predom- tumors and able to increase MSI by targeting the core mismatch inant, as happens in SW480 and CACO-2 cells, it prevents the repair proteins MLHI and MSH2 (48). Moreover, somatic AXIN1 b-CATENIN/TCF4-driven transcriptional activity. Multiple mutations have been found in a high percentage of MSI colorectal mechanisms could explain the differences observed. First of all, cancer cases (49). Interestingly, the MSI DLD-1 cell line carries an it is possible that the functional effects of miR-155-5p on its target AXIN1 (L396M) mutation that abrogates its ability to bind GSK3b genes AXIN1 and TCF4 in colorectal cancer cells might represent a (50). Thus, it is possible that the WNT/b-CATENIN signaling result of a competition among mRNA target genes. Indeed, it is activation observed in miR-155–transfected DLD-1 cells could known that an miRNA can be sequestered by its targets through be explained by a further destabilization of the b-CATENIN a mechanism which represents a critical issue for posttranscrip- destruction complex due to AXIN1 mutation. In addition, tional regulation processes (47). Secondly, it is possible that the we proved that miR-155 overexpression could sustain the

1974 Mol Cancer Res; 16(12) December 2018 Molecular Cancer Research

miR-155-5p Regulates Wnt Signaling in APC-Mutant Settings

WNT/b-CATENIN signaling in the presence of WNT3A by pro- Disclosure of Potential Conflicts of Interest moting AXIN1 degradation. Importantly, it is known that differ- No potential conflicts of interest were disclosed. ent levels of WNT signaling activation may result in distinct cell fates: in particular, its hyperactivation leads to enhanced apopto- Authors' Contributions sis (51, 52). According to this hypothesis, our data showed that Conception and design: A. Prossomariti, L. Ricciardiello miR-155-5p, targeting both AXIN1 and TCF4, was able to finely Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A. Prossomariti, S. Miccoli, D. Turchetti regulate WNT/b-CATENIN signaling activation, thus resulting in Analysis and interpretation of data (e.g., statistical analysis, biostatistics, apoptosis induction or cell proliferation inhibition. Indeed, upon computational analysis): A. Prossomariti, G. Piazzi, L. Ricciardiello miR-155 overexpression, we observed CASPASE-3 activation in Writing, review, and/or revision of the manuscript: A. Prossomariti, G. Piazzi, DLD-1, but not in SW480 cells, suggesting that the activation of F. Bazzoli, L. Ricciardiello cleaved CASPASE-3 found in DLD-1 could be a consequence of Administrative, technical, or material support (i.e., reporting or organizing AXIN1 loss. CASPASE-3 activation at low AXIN1 levels has been data, constructing databases): L. D'Angelo, C. Alquati, C. Montagna Study supervision: L. Ricciardiello described in human melanoma cells, suggesting that loss of Other (critical revision of the manuscript): F. Bazzoli AXIN1 sensitizes cells to apoptosis (53). Although the biological role of WNT-mediated apoptosis has not yet been clarified, our Acknowledgments data support the hypothesis that high levels of WNT activity The authors thank Dr. Massimiliano Bonafe for providing Renilla and may lead to apoptosis of colonic cells (54, 55). On the other TOPFlash plasmids. hand, in SW480 characterized by a stronger miR-155–mediated This work was funded by Italian Association for Cancer Research (Investi- TCF4 degradation (as opposed to DLD-1), we observed a reduc- gator Grant IG14281 to L. Ricciardiello), Italian Foundation for Cancer Research (Fellowship "David Raffaelli" 13837 to A. Prossomariti), European Commu- tion of cell growth due to the inhibition of WNT-driven tran- nity's Seventh Framework Program (Pathway-27, under grant agreement n. scriptional activity. 311876 to L. Ricciardiello), and Programma di Ricerca Regione-Universita In conclusion, our data suggest a critical role of miR-155 in 2010-2012 Regione Emilia Romagna-Bando Giovani Ricercatori "Alessandro APC-mutant settings and define new relevant mechanisms Liberati PRUa1GR-2012-007" to G. Piazzi. mediated by miR-155 on WNT/b-CATENIN regulation in colorectal cancer. Importantly, we suggest that the level of The costs of publication of this article were defrayed in part by the payment of advertisement miR-155, the target concentrations in APC-mutant cells, and page charges. This article must therefore be hereby marked in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. the molecular subtype of tumors might lead to different fi outcome making our ndings pertinent to subsets of colorectal Received February 2, 2018; revised May 31, 2018; accepted July 23, 2018; cancer tumors. published first August 2, 2018.

References 1. Kinzler KW, Nilbert MC, Su LK, Vogelstein B, Bryan TM, Levy DB, et al. 14. Lee E, Salic A, Kruger€ R, Heinrich R, Kirschner MW. The roles of APC and Identification of FAP locus genes from chromosome 5q21. Science Axin derived from experimental and theoretical analysis of the Wnt 1991;253:661–5. pathway. PLoS Biol 2003;1:E10. 2. Knudson AG. Hereditary cancer, oncogenes, and antioncogenes. 15. Guinney J, Dienstmann R, Wang X, de Reynies A, Schlicker A, Soneson C, Cancer Res 1985;45:1437–43. et al. The consensus molecular subtypes of colorectal cancer. Nat Med 3. Markowitz SD, Bertagnolli MM. Molecular basis of colorectal cancer. 2015;21:1350–6. N Engl J Med 2009;361:2449–60. 16. Sveen A, Bruun J, Eide PW, Eilertsen IA, Ramirez L, Murum€agi A, et al. 4. Morin PJ, Sparks AB, Korinek V, Barker N, Clevers H, Vogelstein B, et al. Colorectal cancer consensus molecular subtypes translated to preclinical Activation of beta-catenin-Tcf signaling in colon cancer by mutations in models uncover potentially targetable cancer cell dependencies. beta-catenin or APC. Science 1997;275:1787–90. Clin Cancer Res 2018;24:794–806. 5. Novellasdemunt L, Antas P, Li VS. Targeting Wnt signaling in colorectal 17. Bartel DP. MicroRNAs: genomics, biogenesis, mechanism, and function. cancer. A Review in the theme: cell signaling: proteins, pathways and Cell 2004;116:281–97. mechanisms. Am J Physiol Cell Physiol 2015;309:C511–21. 18. Valencia-Sanchez MA, Liu J, Hannon GJ, Parker R. Control of translation 6. Fearnhead NS, Britton MP, Bodmer WF. The ABC of APC. Hum Mol Genet and mRNA degradation by miRNAs and siRNAs. Genes Dev 2006;20: 2001;10:721–33. 515–24. 7. Stamos JL, Weis WI. The b-catenin destruction complex. Cold Spring Harb 19. Ben-Hamo R, Efroni S. MicroRNA regulation of molecular pathways as a Perspect Biol 2013;5:a007898. generic mechanism and as a core disease phenotype. Oncotarget 8. Bilic J, Huang YL, Davidson G, Zimmermann T, Cruciat CM, Bienz M, et al. 2015;6:1594–604. Wnt Induces LRP6 signalosomes and promotes dishevelled-dependent 20. Pellatt DF, Stevens JR, Wolff RK, Mullany LE, Herrick JS, Samowitz W, et al. LRP6 phosphorylation. Science 2007;316:1619–22. Expression profiles of miRNA subsets distinguish human colorectal carci- 9. Clevers H, Nusse R. Wnt/b-catenin signaling and disease. Cell 2012;149: noma and normal colonic mucosa. Clin Transl Gastroenterol 2016;7:e152. 1192–205. 21. Slattery ML, Herrick JS, Pellatt DF, Stevens JR, Mullany LE, Wolff E, et al. 10. Molenaar M, van de Wetering M, Oosterwegel M, Peterson-Maduro J, MicroRNA profiles in colorectal carcinomas, adenomas and normal colon- Godsave S, Korinek V, et al. XTcf-3 transcription factor mediates beta- ic mucosa: variations in miRNA expression and disease progression. catenin-induced axis formation in Xenopus embryos. Cell 1996;86:391–9. Carcinogenesis 2016;37:245–61. 11. Willert K, Shibamoto S, Nusse R. Wnt-induced dephosphorylation of axin 22. den Dunnen JT, Dalgleish R, Maglott DR, Hart RK, Greenblatt MS, releases beta-catenin from the axin complex. Genes Dev 1999;13:1768–73. McGowan-Jordan J, et al. HGVS recommendations for the description of 12. Yamamoto H, Kishida S, Kishida M, Ikeda S, Takada S, Kikuchi A. Phos- sequence variants: 2016 update. Hum Mutat 2016;37:564–9. phorylation of axin, a Wnt signal negative regulator, by glycogen synthase 23. Kertesz M, Iovino N, Unnerstall U, Gaul U, Segal E. The role of site kinase-3beta regulates its stability. J Biol Chem 1999;274:10681–4. accessibility in microRNA target recognition. Nat Genet 2007;39:1278–84. 13. Voloshanenko O, Erdmann G, Dubash TD, Augustin I, Metzig M, Moffa G, 24. Miyoshi Y, Nagase H, Ando H, Horii A, Ichii S, Nakatsuru S, et al. Somatic et al. Wnt secretion is required to maintain high levels of Wnt activity in mutations of the APC gene in colorectal tumors: mutation cluster region in colon cancer cells. Nat Commun 2013;4:2610. the APC gene. Hum Mol Genet 1992;1:229–33.

www.aacrjournals.org Mol Cancer Res; 16(12) December 2018 1975

Prossomariti et al.

25. Li Y, Lauriola M, Kim D, Francesconi M, D'Uva G, Shibata D, et al. 41. Zeng X, Huang H, Tamai K, Zhang X, Harada Y, Yokota C, et al. Initiation of Adenomatous polyposis coli (APC) regulates miR-17-92 cluster through Wnt signaling: control of Wnt coreceptor Lrp6 phosphorylation/activation b-catenin pathway in colorectal cancer. Oncogene 2016;35:4558–68. via frizzled, dishevelled and axin functions. Development 2008;135: 26. Yan Z, Che S, Wang J, Jiao Y, Wang C, Meng Q. miR-155 contributes to the 367–75. progression of glioma by enhancing Wnt/b-catenin pathway. Tumour Biol 42. Baig-Lewis S, Peterson-Nedry W, Wehrli M. Wingless/Wnt signal transduc- 2015;36:5323–31. tion requires distinct initiation and amplification steps that both depend 27. Yang J, Zhang W, Evans PM, Chen X, He X, Liu C. Adenomatous polyposis on Arrow/LRP. Dev Biol 2007;306:94–111. coli (APC) differentially regulates beta-catenin phosphorylation and ubi- 43. Yang E, Tacchelly-Benites O, Wang Z, Randall MP, Tian A, Benchabane H, quitination in colon cancer cells. J Biol Chem 2006;281:17751–7. et al. Wnt pathway activation by ADP-ribosylation. Nat Commun 2016;7: 28. Ilyas M, Tomlinson IP, Rowan A, Pignatelli M, Bodmer WF. Beta-catenin 11430. mutations in cell lines established from human colorectal cancers. Proc 44. Li VS, Ng SS, Boersema PJ, Low TY, Karthaus WR, Gerlach JP, et al. Wnt Natl Acad Sci USA 1997;94:10330–4. signaling through inhibition of b-catenin degradation in an intact Axin1 29. Liu C, Li Y, Semenov M, Han C, Baeg GH, Tan Y, et al. Control of beta- complex. Cell 2012;149:1245–56. catenin phosphorylation/degradation by a dual-kinase mechanism. Cell 45. Zhang X, Li M, Zuo K, Li D, Ye M, Ding L, et al. Upregulated miR-155 in 2002;108:837–47. papillary thyroid carcinoma promotes tumor growth by targeting APC and 30. Frame S, Cohen P, Biondi RM. A common phosphate binding site explains activating Wnt/b-catenin signaling. J Clin Endocrinol Metab 2013;98: the unique substrate specificity of GSK3 and its inactivation by phosphor- E1305–13. ylation. Mol Cell 2001;7:1321–7. 46. Zhang Y, Wei W, Cheng N, Wang K, Li B, Jiang X, et al. Hepatitis C virus- 31. Lei C, Wang Y, Huang Y, Yu H, Huang Y, Wu L, et al. Up-regulated miR155 induced up-regulation of microRNA-155 promotes hepatocarcinogenesis reverses the epithelial-mesenchymal transition induced by EGF and by activating Wnt signaling. Hepatology 2012;56:1631–40. increases chemo-sensitivity to cisplatin in human caski cervical cancer 47. Jens M, Rajewsky N. Competition between target sites of regulators shapes cells. Ahmad A, editor. PLoS One 2012;7:e52310. post-transcriptional gene regulation. Nat Rev Genet 2014;16:113–26. 32. Xiang X, Zhuang X, Ju S, Zhang S, Jiang H, Mu J, et al. miR-155 promotes 48. Valeri N, Gasparini P, Fabbri M, Braconi C, Veronese A, Lovat F, et al. macroscopic tumor formation yet inhibits tumor dissemination from Modulation of mismatch repair and genomic stability by miR-155. Proc mammary fat pads to the lung by preventing EMT. Oncogene 2011;30: Natl Acad Sci U S A 2010;107:6982–7. 3440–53. 49. Mazzoni SM, Fearon ER. AXIN1 and AXIN2 variants in gastrointestinal 33. Gregorieff A, Clevers H. Wnt signaling in the intestinal epithelium: from cancers. Cancer Lett 2014;355:1–8. endoderm to cancer. Genes Dev 2005;19:877–90. 50. Webster MT, Rozycka M, Sara E, Davis E, Smalley M, Young N, et al. 34. Benchabane H, Ahmed Y. The adenomatous polyposis coli tumor sup- Sequence variants of the axin gene in breast, colon, and other cancers: an pressor and Wnt signaling in the regulation of apoptosis. Adv Exp Med Biol analysis of mutations that interfere with GSK3 binding. Genes Chromo- 2009;656:75–84. somes Cancer 2000;28:443–53. 35. Kinzler KW, Vogelstein B. Gatekeepers and caretakers. Nature 1997; 51. Ming M, Wang S, Wu W, Senyuk V, Le Beau MM, Nucifora G, et al. 386:761–3. Activation of Wnt/b-Catenin protein signaling induces mitochondria- 36. Necela BM, Carr JM, Asmann YW, Thompson EA, Ambros V, Asli N, et al. mediated apoptosis in hematopoietic progenitor cells. J Biol Chem Differential expression of MicroRNAs in tumors from chronically 2012;287:22683–90. inflamed or genetic (APCMin/þ) models of colon cancer. PLoS One 52. Kim K, Pang KM, Evans M, Hay ED. Overexpression of beta-catenin induces 2011;6:e18501. apoptosis independent of its transactivation function with LEF-1 or the 37. Volinia S, Calin GA, Liu CG, Ambs S, Cimmino A, Petrocca F, et al. A involvement of major G1 cell cycle regulators. Mol Biol Cell 2000; microRNA expression signature of human solid tumors defines cancer gene 11:3509–23. targets. Proc Natl Acad Sci 2006;103:2257–61. 53. Biechele TL, Kulikauskas RM, Toroni RA, Lucero OM, Swift RD, James RG, 38. Qu YL, Wang HF, Sun ZQ, Tang Y, Han XN, Yu XB, et al. Up-regulated et al. Wnt/b-catenin signaling and AXIN1 regulate apoptosis triggered by miR-155-5p promotes cell proliferation, invasion and metastasis in colo- inhibition of the mutant kinase BRAFV600E in human melanoma. Sci rectal carcinoma. Int J Clin Exp Pathol 2015;8:6988–94. Signal 2012;5:ra3. 39. Suzuki H, Takatsuka S, Akashi H, Yamamoto E, Nojima M, Maruyama R, 54. Wong MH, Huelsken J, Birchmeier W, Gordon JI. Selection of multipotent et al. Genome-wide profiling of chromatin signatures reveals epigenetic stem cells during morphogenesis of small intestinal crypts of lieberkuhn is regulation of MicroRNA genes in colorectal cancer. Cancer Res 2011; perturbed by stimulation of Lef-1/beta -catenin signaling. J Biol Chem 71:5646–58. 2002;277:15843–50. 40. Peng Y, Zhang X, Feng X, Fan X, Jin Z. The crosstalk between microRNAs 55. Lazarova DL, Bordonaro M, Carbone R, Sartorelli AC. Linear relationship and the Wnt/b-catenin signaling pathway in cancer. Oncotarget 2017; between Wnt activity levels and apoptosis in colorectal carcinoma cells 8:14089–106. exposed to butyrate. Int J Cancer 2004;110:523–31.

1976 Mol Cancer Res; 16(12) December 2018 Molecular Cancer Research

miR-155 Is Downregulated in Familial Adenomatous Polyposis and Modulates WNT Signaling by Targeting AXIN1 and TCF4

Anna Prossomariti, Giulia Piazzi, Leonarda D'Angelo, et al.

Mol Cancer Res 2018;16:1965-1976. Published OnlineFirst August 2, 2018.

Updated version Access the most recent version of this article at: doi:10.1158/1541-7786.MCR-18-0115

Supplementary Access the most recent supplemental material at: Material http://mcr.aacrjournals.org/content/suppl/2018/08/02/1541-7786.MCR-18-0115.DC1

Cited articles This article cites 55 articles, 20 of which you can access for free at: http://mcr.aacrjournals.org/content/16/12/1965.full#ref-list-1

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/16/12/1965. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.