Supplemental Information. Dual Function of p38α MAPK in Colon Cancer: Suppression of Colitis-Associated Tumor Initiation

Cancer Cell, Volume 25 Supplemental Information Dual Function of p38α MAPK in Colon Cancer: Suppression of Colitis-Associated Tumor Initiation but Requirement for Cancer Cell Survival Jalaj Gupta, Ivan del Barco Barrantes, Ana Igea, Stratigoula Sakellariou, Ioannis S. Pateras, Vassilis G. Gorgoulis, and Angel R. Nebreda

1 Supplemental Data Figure S1 (Related to Figure 1). Downregulation of p38α in IEC increases AOM/DSSinduced colon tumorigenesis

2 (A) Colon tumors were induced by the AOM/DSS protocol. Average tumor numbers were determined from two independent experiments. Statistical analysis was performed by Student s T-test or Mann Whitney test using GraphPad Prism 4 software. p values < 0.05 were considered statistically significant. Data represent means ± SEM. (B) Colon sections were stained for BrdU or cleaved caspase 3 to determine cell proliferation and apoptosis, respectively. Bars = 50 µm. (C) Analysis by qrt-pcr of mrna expression levels for the indicated pro-inflammatory mediators in mouse colons at the end of the AOM/DSS protocol. Data are means ± SEM (n = 4). (D) IL-6 protein levels in blood serum from mice after the AOM/DSS protocol. Data are means ± SEM (n = 4). (E) Colon tumor lysates were analyzed for p38α expression by western blotting (one mouse per lane).

3 Figure S2 (Related to Figure 2). Downregulation of p38α in IEC does not affect AOMinduced early DNA damage but increases inflammatory cell infiltration upon DSS treatment

4 (A) Mice were injected with AOM or PBS and killed 7 hr later. Colon lysates were analyzed by western blotting (one mouse per lane) using the indicated antibodies. (B) Mice were treated as in (A) and colon sections were stained for cleaved caspase 3 to detect apoptosis. Histograms represent the quantification of the average number of apoptotic cells (Cleaved caspase 3 + ) at the base of crypts in AOM-treated mice. Data are means ± SEM (n = 3). (C) CD45 + leukocytes, CD19 + B-cells and Gr1 + neutrophils in colon lamina propria and intraepithelial compartment of mice treated with 2% DSS for 5 days and analyzed at day 6. Histograms show numbers of the indicated cells per 10 3 total cells. Data represent means ± SEM. (D) Colon sections from mice either untreated or treated with DSS for 5 days were stained for F4/80 to identify macrophages. (E) Expression of inflammatory mediators in EpCAM + epithelial cells and CD45 + leukocytes from DSS-treated mice. Relative mrna expression for the indicated genes was determined by qrt-pcr. Data represent means ± SEM (n = 3). Bars = 50 µm.

5 Figure S3 (Related to Figure 4). Analysis of mutations in the tumors induced by treatment only with DSS in p38α- IEC mice (A) Mutations within exon 3 of the β-catenin gene, that encodes for GSK3β phosphorylation sites, was analyzed by direct sequencing on genomic DNA isolated by laser capture microdissected tumors or normal mucosa. (B) Nuclear β-catenin staining in tumors developed in p38α- IEC mice by repetitive treatment with DSS in the absence of the AOM carcinogen. Bar = 50 µm.

6 Figure S4 (Related to Figure 5). Downregulation of p38α in IEC affects intestinal homeostasis and tight junctions but does not affect IL-23 and IL-17 expression

7 (A) Colon sections from untreated mice were stained with H&E, Ki67 antibody for proliferation, Periodic acid-schiff (PAS) reagent for goblet cells and Chromogranin A (ChgA) antibody for enteroendocrine cells. Bars = 50 µm. (B) Relative mrna expression levels for the indicated genes in the distal colon of untreated mice were determined by qrt-pcr. Data are means ± SEM (n = 4)., p<0.05. (C) Electron microscope images showing colon epithelia sections where the tight-junctions depicted in Figure 5C were taken from. Bars = 200 nm. (D) Relative mrna expression levels of the indicated genes in mouse colons were determined by qrt-pcr. Data are means ± SEM (n = 4)., p<0.05. (E) Relative mrna expression levels of IL-23 and IL-17 in the colon from untreated mice or in AOM/DSS-induced colon tumors were determined by qrt-pcr. Data are means ± SEM (n = 4).

8 Figure S5 (Related to Figure 6). Inducible downregulation of p38α in IEC before and after AOM/DSS treatment

9 (A) Colon, liver, lung and spleen lysates were analyzed by western blotting (one mouse per lane) with the indicated antibodies. (B) Schematic representation of the AOM/DSS protocol used to induce colorectal tumors in p38α- IEC-ERT2 mice. (C) Average tumor number, size and load at the end of the AOM/DSS protocol. Data represent means ± SEM (n = 7). (D) Body weight loss induced by 2% DSS administered in the drinking water. Data represent means ± SEM (n 7)., p<0.05. (E) Colon tumors were induced by the AOM/DSS protocol and then 4-OHT was injected (day 66 to 70). Average tumor number and size were determined at 3 days after the last 4-OHT injection. Data represent means ± SEM (n 4). Tumor lysates were analyzed at the indicated time points by western blotting (one mouse per lane). The histograms show densitometric quantification of p38α protein expression normalized to actin. Data represent means ± SEM (n = 3). (F) Colon tumors were induced by the AOM/DSS protocol and then 4-OHT was injected (day 66 to 70). Average tumor number and size were determined at 10 days after the last 4-OHT injection. Data represent means ± SEM (n 5). Tumor lysates were analyzed at the indicated time points by western blotting (one mouse per lane). The histograms show densitometric quantification of p38α protein expression normalized to actin. Data represent means ± SEM (n = 3). (G) Colon tumors were induced by the AOM/DSS protocol and then 4-OHT was injected (day 66 to 70). At 20 days after the last 4-OHT injection, colon sections were stained with p38α antibody. N, normal tissue; T, tumor. Bars = 50 µm. (H) Laser-captured microdissected (LCM) tumor samples were obtained from mice treated as in (G). Genomic DNA was purified and analyzed by qpcr with primers specific for exon 2 and exon 12 (as a control) of the p38α gene. Relative amount of exon 2 versus exon 12 is shown. Data are means ± SEM (n 6), p<0.05. (I) EpCAM + epithelial cells and CD45 + leukocytes were isolated from colon tumors of mice treated as in (G). Genomic DNA was purified and analyzed by qpcr for the deletion of p38α as in (H). Data represent means ± SEM (n 7).

10

11 Figure S6 (Related to Figure 7). Persistent downregulation of p38α in colon tumors reduces tumor burden without affecting histological tumor grading or epithelial barrier function (A and B) Colon tumors were induced by the AOM/DSS protocol and then 4-OHT was injected (day 66 to 70). Average tumor number and size (A) and tumor grades (B) were determined 45 days after the last 4-OHT injection. Data represent means ± SEM (n = 7). (C) Tumor lysates from different mice were analyzed at 45 days after last 4-OHT injection by western blotting (one mouse per lane). The histograms show densitometric quantification of p38α protein expression normalized to actin. Data represent means ± SEM (n 6). (D) Schematic representation of the protocol used for the sustained downregulation of p38α in AOM/DSS-induced colon tumors. (E) Tumor lysates from different mice treated as in (D) were analyzed by western blotting (one mouse per lane) at 20 days after the second 4-OHT cycle of injections. The histograms show densitometric quantification of p38α protein expression normalized to tubulin. Data represent means ± SEM (n 4)., p<0.05. (F) Average tumor number, size, load and tumor size distribution in mice at the end of the experiment indicated in (D). Data represent means ± SEM (n = 6)., p<0.05. (G) Tumors from mice treated as in (D) were classified into low or high grade (n=6) based on histological analysis. All the tumors detected in WT and p38α- IEC-ERT2 mice were of high grade.

12 (H) Intestinal permeability was determined by measuring the concentration in blood serum of FITC-dextran that was orally administered to mice, with or without colon tumors, 20 days after the last 4-OHT injection. Data are means ± SEM (n 4). Tumor permeability was estimated as described in Supplemental Experimental Procedures by subtracting the permeability in untreated mice from the permeability in tumor-bearing mice. The values obtained for tumor permeability were 562 (1329-767) in WT mice and 302 (1276-974) in p38α- IEC-ERT2 mice. Total tumor permeability was normalized to the average tumor numbers or to the tumor burdens indicated in Figure 6D. In both cases, the normalized tumor permeability was similar in the two genotypes: 52.04 (562/10.80) in WT tumors and 50.33 (302/6) in p38α KO tumors, referring to tumor numbers; 18.47 (562/30.4) in WT tumors and 20.68 (302/14.6) in p38α KO tumors referring to tumor burdens. (I) Colon tumors were induced by the AOM/DSS protocol and then 4-OHT was injected (day 66 to 70). At 20 days after the last 4-OHT injection, colon tumor lysates were analyzed by western blotting (one mouse per lane) with the indicated antibodies. (J) Relative mrna expression levels for the indicated genes in the tumors of mice treated as in (I) were determined by qrt-pcr. Data are means ± SEM (n = 4). (K) Colon tumors from mice treated as in (I) were stained with MPO antibodies. Bars = 100 µm. Quantifications are shown in the lower panel. Data represent means ± SEM (n=4)., p<0.05.

13 Supplemental Experimental Procedures Western blotting Distal colon segments were lysed in buffer containing 1% NP40, 150 mm NaCl, 50 mm Tris HCl ph 7.5, 2 mm EDTA, 2 mm EGTA, 20 mm sodium fluoride, 2 mm PMSF, 2 µm microcystin, 2 mm sodium orthovanadate, 1 mm DTT and 1x EDTA-free complete protease inhibitor cocktail (Roche, #11873580001), using Precellys homogenization and lysis instrument (Bertin technologies). Protein content was quantified using the Bradford assay with BSA as standard (Bio-Rad), and 40 µg of total protein lysate in Laemmli buffer were separated on SDS-PAGE and transferred to nitrocellulose membrane (Whatman #10401396). After blocking (5% non-fat milk and 1% BSA in PBS, 1 hr at RT) membranes were incubated at 4ºC overnight with primary antibodies. The following antibodies were used: p38α (#9218; 1:1000), phospho-p38 (#9211;1:1000), phospho-stat3 Tyr705 (#9145; 1:600), STAT3 (#9132; 1:1000), phospho-iκbα Ser32/36(#9246; 1:600), IκBα (#4812; 1:1000), phospho- ERK1/2 (#9101;1:1000), phospho-akt Ser473 (#9271; 1:1000), phospho-p53 Ser15 (#9284; 1:1000), p53 (#2524; 1:1000), phospho-hsp27 Ser82 (#2401; 1:700), Mcl-1(#5453; 1:600), Bcl-2 (#2870, 1:1000), Bak (#3814; 1:1000), Bax (#2772; 1:1000) from Cell Signaling; γ- H2AX (Millipore #05-636; 1:1000), ZO-1 (#40-2200; 1:1000) and Claudin 1 (#374900; 1:500) from Invitrogen; phospho-jnk (#C12541; 1:500) from BD; JNK (#SC-571; 1:1000) and Hsp27 (#SC-1049; 1:700) and Bcl-XL (#SC-8392; 1:1000) from Santa Cruz; p85 PARP (#G734A; 1:500) from Promega. Tubulin (Sigma #T9026) was used as loading control. After washing with PBS, membranes were incubated with Alexa Fluor 680 or 800-conjugated secondary antibodies (Invitrogen; 1:5000) for 1 hr at RT and were visualized using Odyssey Infrared Imaging System (Li-Cor, Biosciences). Histopathological analysis Tumor grades were scored using a two-tiered system based on up-to-date guidelines of the WHO for tumors of the digestive system. Briefly, adenomas with low-grade dysplasia consist of stratified dysplastic epithelium retaining its columnar shape, with the nuclei in the basal part of epithelium, whereas adenomas with high-grade dysplasia have lost its columnar shape and the nuclei are mainly located in the surface of the crypts. The following scoring system was used for epithelial damage: 1- intact crypts, 2- basal one-third damaged, 3- basal two-thirds damaged, 4- damaged surface epithelium.

14 The severity of inflammation was scored as follows: 1- rare cells in mucosa, 2- increased cells in lamina propria, 3- confluence of cells in the submucosa, 4- transmural inflammation. Immunohistochemistry Paraffin-embedded colon sections were stained with antibodies against Ki67 (Novacastra, NCL; 1:500, 1 hr RT), F4/80 (ebiosciences #14-4801; 1:50, 2 hr RT), MPO (Dako #A0398; 1:1000, 30 min RT), β-catenin (BD Biosciences #610153; 1:100, 1 hr RT), Chromogranin-A (Abcam #151601; 1:1000 overnight 4ºC), cleaved caspase-3 (#9661; 1:200, 1 hr RT), p38α (#9218; 1:50, 2 hr RT) and phospho-p38 (#4631; 1:50, overnight 4 C) from Cell Signaling. In some experiments BrdU (Roche #10280879001) was intraperitoneally injected (1 mg /10 g body weight) and 2 hr later the mice were sacrificed. BrdU staining was performed using anti- BrdU antibody (BD Biosciences #347580; 1:100 1 hr RT). The secondary antibodies used were: HRP conjugated anti-rabbit (ImmunoLogic #DPVR110HRP, 45 min at RT), anti-mouse (Dako #P0447; 1:100, 30 min at RT) and anti-rat (Dako #P0450; 1:75, 30 min at RT). RNA extraction and qrt-pcr Total RNA was extracted from distal colon segments or from colon tumors using TRIzol (Invitrogen) or PureLink RNA mini kit (Ambion #12183018A). After DNase I treatment (Roche #04716728001), total RNA (1-2 µg) was reverse transcribed using a Super script II Reverse Transcriptase (Invitrogen #18064-014) and Random primers (Invitrogen #48190-011). qrt-pcr was performed in triplicates using 4 µl of 1/12 diluted cdna and SYBR green (Bio-Rad #1708886) in 20 µl total volume on a Bio-Rad C1000 thermal cycler machine. Relative quantities ( cycle threshold values) were obtained by normalizing against GAPDH. The primers used are listed in the Supplemental Table. Analysis of tight junctions by transmission electron microscopy Pieces of intestine were fixed overnight at 4ºC in 2% paraformaldehyde and 2.5% glutaraldehyde, postfixed in OsO4 (2%) for 2 hr and then dehydrated at 4ºC through a series of acetone concentrations (50, 70, 90, 96, 100%), prior to being progressively (25, 50, 75, 100%) embedded in Epon 812 epoxy resin. Sections with a thickness of 50 nm were cut with an ultramicrotome UCT6 (Leica Microsystems, Vienna) and placed on TEM grids (Formvar carbon-coated Cu grids). The grids were further contrasted with uranyl acetate and lead

15 citrate. Micrographs were obtained with a Jeol JEM 1010 MT electron microscope (Jeol, Japan) operating at 80 kv. Images were recorded with AnalySIS (SIS, Munster, Germany) on a Megaview III CCD camera. Mouse treatments Mouse genotyping was performed by PCR on genomic tail DNA. Primers and conditions are available upon request. To activate Cre-ERT2, mice were injected intraperitoneally for 5 consecutive days with 1 mg/day of 4-OHT (Sigma, H6278) dissolved in corn oil (Sigma, #C8267). The inhibitor of p38α and p38β PH797804 (Goldstein et al., 2010) was obtained from Selleckchem (#S2726) and was dissolved in 0.5% Methyl cellulose (Sigma #M7140) and 0.025% Tween 20 (Sigma #P1379) at a concentration of 1 mg/ml. Mice were administered daily a dose of 10 mg/kg body weight by oral gavage for 12 consecutive days. Control mice were similarly administered with vehicle (0.5% Methyl cellulose and 0.025% Tween 20). Isolation of epithelial cells and leukocytes from colon Acute colitis was induced by treating mice with 2% DSS for 5 days. One day after termination of the DSS treatment, mice were sacrificed and colons were dissected, opened longitudinally and washed extensively in cold PBS. Colons were then cut into 3-4 pieces and incubated in 8 mm EDTA solution in PBS at 37ºC for 15 min. After 15 min, EDTA solution was replaced with cold PBS and shacked vigorously for 45 s. This process was repeated once more. Supernatants from both incubations were combined, centrifuged at 1200 rpm for 5 min at 4ºC, and pelleted cells were digested with Dispase II (0.5 mg/ml, Roche #04942078001) at 37ºC for 25 min. After the incubation, the digested cell suspension was gently syringed with 18 G needle in order to get enriched single cell population and sequentially passed through 100, 70 and 40 µm mesh filters (BD Biosciences). Filtered cells contain epithelial cells and intra-epithelial leukocytes. To obtain lamina propria leukocyte, colon pieces after EDTA incubations were collected, cut into small pieces (2-3 mm) and digested with mix of collagenase A (1.75 mg/ml, Roche #10103586001) and DNase I (0.05 mg/ml, Sigma #D4263) at 37ºC for 45 min. After digestion, the cell suspension was filtered as above to get single cell suspension containing lamina propria leukocytes. To quantify immune cell populations, single cell suspensions from both purifications were co-stained with panleukocyte antigen CD45 and for specific leukocyte population markers (CD11b, Gr1 and

16 CD19) and analyzed on FACS Aria 2.0 (BD Biosciences). FITC-EpCAM (Santa Cruz #sc- 53532) staining was used to discard epithelial cells by negative selection. To get epithelial cells and total leukocytes from the colon, after induction of acute colitis whole colon tissue was digested in a mix of collagenase and DNase I, syringed and sequentially filtered as described above. Then cells were co-stained with FITC-EpCAM and APC-CD45 (BD #559864). Stained cells were analyzed and sorted for EpCAM + epithelial cells and CD45 + leukocytes. DNA or RNA was isolated from sorted cells for further analysis. Detection of tumor mutations Tumor tissues were obtained by laser capture microdissection (LCM) from paraffinembedded colons and were digested with proteinase K. Genomic DNA was purified by standard phenol-chloroform extraction. Exon 3 of the β-catenin gene was amplified by PCR using the following specific primers: 5 -GCTGACCTGATGGAGTTGGA-3 and 5 - GCTACTTGCTCTTGCGTGAA-3 (amplicon size = 227 bp). PCR products were purified and sequenced using forward and reverse primers. Mutations were detected by observing individual chromatograms. Analysis of p38α floxed exon deletion Genomic DNA was isolated using standard phenol-chloroform extraction from LCMprocessed tumor samples or from EpCAM + epithelial cells and CD45 + leukocytes isolated from tumors and then analyzed by qpcr with primers specific for exon 2 (floxed) and exon 12 (as a control) of the p38α gene. Relative amount of exon 2 versus exon 12 was determined. Primers are listed in the Supplemental Table. Determination of IL-6 in blood serum and colon tissue Blood was collected by cardiac puncture and serum samples were obtained using SST tubes (BD #365968). IL-6 protein concentration in serum was determined by the CBA mouse IL-6 flex set (BD #558301) according to manufacturer s instructions. To determine IL-6 protein levels in colonic tissue, 100 µg of freshly lysed colon tissue were analyzed by ELISA using anti-mouse IL-6 purified capture antibody (ebioscience #14-7060-81) and anti-mouse IL-6 biotin detection antibody (ebioscience #13-7062-81). Mouse IL-6 recombinant protein (ebioscience #14-8061-62) was used to make the standard curve.

17 Analysis of apoptosis in human cancer cell lines Human colon cancer cell lines SW-620 and Caco-2 were grown in Dulbecco s modified Eagle s medium supplemented with 10% heat-inactivated fetal bovine serum, 1% l-glutamine, and 1% penicillin-streptomycin. The p38 MAPK inhibitors SB203580 (10 µm), PH797804 (1 µm) and Birb0796 (200 nm) were added every 48 hr to the cell cultures. After 96 hr, all the cells (including floating cells in the medium) were collected and lysed. Cell lysates were analyzed by western blotting using p85 PARP antibody as an apoptosis marker. Estimation of colon tumor permeability To estimate tumor permeability, we first measured intestinal permeability both in untreated mice and in tumor-bearing mice as described in Experimental Procedures. The intestinal permeability should mainly depend on epithelial barrier function in untreated mice, whereas in tumor-bearing mice both epithelial barrier function and tumorigenesis (i.e., changes in intestinal permeability due to tumor formation) should contribute to the total permeability measured. The intestinal permeability values in untreated mice were subtracted from the total permeability values in tumor-bearing mice to estimate the permeability due to tumorigenesis, which was then normalized against the average tumor numbers or the tumor burdens. Primers used for quantitative RT-PCR Gene Forward Primer (5-3 ) Reverse Primer (5-3 ) GAPDH CTTCACCACCATGGAGGAGGC GGCATGGACTGTGGTCATGAG IL-6 AGTTGCCTTCTTGGGACTGA CAGAATTGCCATTGCACAAC IL-1α GAGAGCCGGGTGACAGTATC TGACAAACTTCTGCCTGACG TNF-α CGTCAGCCGATTTGCTATCT CGGACTCCGCAAAGTCTAAG COX2 AAAAGCTGGGAAGCCTTCTC AAGTGCTGGGCAAAGAATGC p38α CTGACCGACGACCACGTTC CTTCGTTCACAGCTAGGTTGC Chg A AAGTGCGTCCTGGAAGTCATCTC GCTTGGCTTTTCTGGCTTGC Ki67 GTGCTGACCCTGATGGGGAAGG GCTCTTGCCCTGCCTGACACC CyclinD1 CTGCAAATGGAACTGCTTCTGGTGA AGCAGGAGAGGAAGTTGTTGGGGCT

18 Muc2 GCCCGTGGAGTCGTACGTGC TTGGGGCAGAGTGAGGCGGT Tff3 TAATGCTGTTGGTGGTCCTG CAGCCACGGTTGTTACACTG TGFβ1 GGAGGTACCGCCCGGCCCGC GACAGCAATGGGGGTTCGGG IL-10 CCAAGCCTTATCGGAAATGA TTTTCACAGGGGAGAAATCG IL-12p40 AGGTCACACTGGACCAAAGG TGGTTTGATGATGTCCCTGA ZO-1 GCCGCTAAGAGCACAGCAA GCCCTCCTTTTAACACATCAGA Occludin TTGAAAGTCCACCTCCTTACAGA CCGGATAAAAAGAGTACGCTGG ZO-2 ACTCCAGTCCCTATTCCTGAG GCTATTTCGATCCTCGCATTC JAM-C CTGCCTGACTTCTTCCTGCT ATGTACCACTGGGTTTCGGT IL-11 TGTTCTCCTAACCCGATCCCT CAGGAAGCTGCAAAGATCCCA CXCL-1 ACTGCACCCAAACCGAAGTC TGGGGACACCTTTTAGCATCTT CXCL-2 CCAACCACCAGGCTACAGG GCGTCACACTCAAGCTCTG IL-23p19 CCAGCGGGACATATGAATCT AGGCTCCCCTTTGAAGATGT IL-17A GCCCTCAGACTACCTCAACC ACACCCACCAGCATCTTCTC p38α (Exon 2) p38α (Exon 12) GCATCGTGTGGCAGTTAAGA GCCCTCCCTCACTTCAGGAG GTCCTTTTGGCGTGAATGAT TGTGCTCGGCACTGGAGACC