Downregulation of the Deiminase PADI2 Is an Early Event In

Published OnlineFirst June 8, 2016; DOI: 10.1158/1541-7786.MCR-16-0034

Oncogenes and Tumor Suppressors Molecular Cancer Research Downregulation of the Deiminase PADI2 Is an Early Event in Colorectal Carcinogenesis and Indicates Poor Prognosis Neus Cantarino~ 1, Eva Musulen 2,Vanesa Valero1,3, Miquel Angel Peinado1, Manuel Perucho1, Victor Moreno4,5,Sonia-Vanina Forcales1, Julien Douet1,3, and Marcus Buschbeck1,3

Abstract

Peptidyl arginine deiminases (PADI) are a family of enzymes PADI2 is the dominantly expressed PADI enzyme in colon mucosa that catalyze the poorly understood posttranslational modiﬁcation and is upregulated during differentiation. PADI2 expression is low converting arginine residues into citrullines. In this study, the role or absent in colorectal cancer. Frequently, this occurs already at the of PADIs in the pathogenesis of colorectal cancer was investigated. stage of low-grade dysplasia. Mucosal PADI2 expression is also low Speciﬁcally, RNA expression was analyzed and its association with in ulcerative colitis. The expression level of PADI2 in tumor and survival in a cohort of 98 colorectal cancer patient specimens with adjacent mucosa correlates with differential survival: low levels matched adjacent mucosa and 50 controls from donors without associate with poor prognosis. cancer. Key results were validated in an independent collection of tumors with matched adjacent mucosa and by mining of a publicly Implications: Downregulation of PADI2 is an early event in available expression data set. Protein expression was analyzed by the pathogenesis of colorectal cancer associated with poor prog- immunoblotting for cell lines or IHC for patient specimens that nosis and points toward a possible role of citrullination in further included 24 cases of adenocarcinoma with adjacent dys- modulating tumor cells and their microenvironment. Mol Cancer plasia and 11 cases of active ulcerative colitis. The data indicate that Res; 14(9); 841–8. 2016 AACR.

Introduction has a profound influence on protein–protein interactions when occurring on relevant interaction surfaces. Methylation is another Peptidyl arginine deiminases (PADI) are a family of enzymes posttranslational modification of arginine residues. It has been that catalyze the conversion of arginine residues to citrullines shown that monomethylated but not dimethylated arginines can (reviewed in ref. 1). Five genes located in a well-organized cluster be citrullinated (5). Citrullination prevents posterior methylation on human chromosome 1 encode the five enzymes PADI1, (6). Important functions for PADI enzymes have been reported in PADI2, PADI3, PADI4, and PADI6. Of these, PADI2 is the ances- both the cytosol and the nucleus. PADIs citrullinate components tral homologue and widely expressed in different tissues, while of cytoskeleton in the cytosol such as vimentin (7) but also the other PADI show restricted tissue-specific expression (sum- proteins that have structural extracellular functions such as fibrin marized in ref. 2). Although PADI enzymes share a high degree of (8). A major function of PADI activity in the nucleus is decon- homology, they have some substrate specificity in vitro (3, 4). The densation of chromatin by citrullination of histone proteins (9). citrullination reaction catalyzed by PADI enzymes is strictly This function has been suggested to contribute to the na€ve calcium-dependent and effectively removes the positive charge chromatin state of pluripotent embryonic stem cells (10). of arginine residues. As such, this posttranslational modification Through citrullination of histones and other chromatin regula- tors, PADIs have further been implicated in gene regulation (summarized in ref. 11). 1Institute of Predictive and Personalized Medicine of Cancer (IMPPC), Campus Can Ruti, Badalona, Spain. 2Department of Pathology, Hos- In the innate immune response, chromatin decondensation pital Universitari Germans Trias i Pujol (HGTP), Campus Can Ruti, upon histone hypercitrullination allows the formation of neu- 3 Badalona, Spain. Josep Carreras Leukaemia Research Institute (IJC), trophil extracellular trap (NET) formation that neutrophils use to Campus ICO–Germans Trias i Pujol, Campus Can Ruti, Badalona, Spain. 4Cancer Prevention and Control Program, Catalan Institute of Oncol- capture bacteria (9). PADI enzymes are further involved in a range ogy (ICO), Bellvitge Biomedical Research Institute (IDIBELL) and of auto-inflammatory diseases, including rheumatoid arthritis CIBERESP, Barcelona, Spain. 5Department of Clinical Sciences, Fac- and multiple sclerosis due to their activation during NETosis, ulty of Medicine, University of Barcelona (UB), Barcelona, Spain. apoptosis, and autophagy (12). A common denominator in these Note: Supplementary data for this article are available at Molecular Cancer diseases is an increase in enzymatic PADI activity and increased Research Online (http://mcr.aacrjournals.org/). levels of antibodies against citrullinated antigens (1). These auto- Corresponding Author: Marcus Buschbeck, IJC, Campus Can Ruti, Ctra de Can antogenic antibodies are used as biomarkers for diagnosis and Ruti, Cami de les Escoles w/o number; 08916 Badalona, Spain. Phone: 34- prognosis for rheumatoid arthritis (13). 934651472; Fax: 34-934651472; E-mail: [email protected] More recently, PADI enzymes, in particular PADI2 and PADI4, doi: 10.1158/1541-7786.MCR-16-0034 have been further implicated in cancer (11). PADI4 was found to 2016 American Association for Cancer Research. be overexpressed in epithelial cells from various adenocarcinoma

www.aacrjournals.org 841

Cantarino~ et al.

compared with normal tissue (14). PADI2 was found upregulated used antibodies directed against PADI2 (Proteintech UK; in human HER2-positive breast cancers (15). In mice, overexpres- 12110-1-AP) and C-terminal histone H3 (Abcam; ab204964). sion of a PADI2 transgene was sufficient to cause skin tumors (16). RNA levels were quantified by RT-qPCR and normalized to two It has been proposed that PADIs could contribute to cancer by reference genes (HPRT1 and GAPDH). All oligo sequences are acting as coactivator of oncogenic transcription factors eliciting available on request. tumor-promoting expression programs. To illustrate this, PADI4 was shown to enhance transcriptional regulation of Tal1 in Immunohistochemistry hematopoietic stem cells (17). However, the outcome of such a Immunohistochemistry and hematoxilin–eosin staining on coactivator function is strongly dependent on the cellular context. formalin-fixed paraffin-embedded tissue sections were performed In the estrogen receptor–positive subgroup of breast cancer as described (23). For immunohistochemistry, we have opti- patients, PADI2 expression correlated with good prognosis and mized the signal background for anti-PADI2 (Proteintech UK; in breast cancer cells was shown to facilitate the binding of the reference number 12110-1-AP) using positive and negative con- estrogen receptor to its target genes (18). trol samples. A pathologist scored the expression of PADI2 as Colorectal cancer is the second most common cancer in the positive, focally positive, or negative. western world (19). The role of PADIs in the pathogenesis of colorectal cancer has not been investigated. In this article, we Data analysis report that PADI2 is the primarily expressed PADI in colon To analyze TCGA data, we have used the Wanderer application mucosa and that its expression is low or absent in tumor tissue that we had developed for this purpose and that is described in at an early stage of the colorectal cancer development. Moreover, detail elsewhere (24). The analysis of expression microarray data higher level of PADI2 in the tumor and adjacent mucosa is a for specific genes of interest was performed essentially as previ- predictor of good prognosis. ously described (20). The comparison of expression between tumor samples (n ¼ 98) and normal mucosa (n ¼ 50) was Materials and Methods performed with a linear model, adjusted for age and gender. The sample size for this discovery phase allowed to identify and effect Patient samples and data sets size of 0.5 with 80% power. For the validation, the TCGA dataset The following sample series and data sets were analyzed: RNA had a similar power, and the series of matched paired samples from 16 colorectal tumors and matched normal mucosa; immu- (n ¼ 16) could detect an effect size of 1.0 or larger. For the analysis nohistochemistry of tissue blocks from 24 patients with infiltrat- of prognosis, disease-free survival curves were estimated using the ing adenocarcinoma containing areas of both low- and high-grade Kaplan–Meier method for a categorization of PADI2 expression. dysplasia and normal mucosa; immunohistochemistry of colon Also, a multivariate Cox model was used to estimate hazard ratios, tissue blocks from 11 patients with active ulcerative colitis; 95% confidence intervals and calculate P values from likelihood microarray-based expression data from tumor and adjacent ratio tests. Analyses were adjusted for age and gender, but not for mucosa samples of a cohort of 98 stage II, microsatellite stable stage, because all cases were stage II colon cancer. colon cancer patients and 50 colon mucosa samples from donors without cancer (ref. 20; www.colonomics.org). The project was approved by the Hospital's Ethics Committee with registration Results number PR074/11. Samples were obtained after written, PADI2 expression is downregulated in colorectal cancer informed consent following the guideline of the Declaration of To investigate the expression of PADI family members we have – Helsinki. Also RNA-seq based expression data from The Cancer analyzed RNA expression data from tumor and adjacent mucosa Genome Atlas (TCGA) were used for validation studies (21). samples of 98 colon cancer patient cohort and 50 colon mucosa samples from donors without cancer (20). As shown in Fig. 1A, of Cell culture all PADI genes, PADI2 is the only gene prominently expressed in Colorectal cancer cell lines HCT116, HT-29, SW480, DLD-1, colon mucosa. We found a pronounced and highly significant and Colo205 were grown in Dubelcco's Modified Eagle's Medium downregulation of PADI2 in tumor samples in comparison with (DMEM, Life Technologies; 11960085) supplemented with 10% both matched adjacent normal mucosa and mucosa from donors FBS (Life Technologies; 10270106), 2 mmol/L glutamine, 1 without cancer. This downregulation was equally significant (P < mmol/L pyruvate, and 1% penicillin–streptomycin (Life Tech- 2e16) no matter whether the multiple probes for PADI2 on the nologies; 25030024; 11360039; 15140122). CaCo2 cells were array were analyzed together or separately. In contrast, the dif- grown in DMEM-F12 (Life Technologies; 21331020) supplemen- ference between tumor-associated mucosa compared with ted with 20% FBS, 2 mmol/L glutamine, 1 mmol/L pyruvate, and healthy mucosa was little and only reached significance for some 1% penicillin–streptomycin. All cells were propagated with a split probes (not shown). To validate the downregulation of PADI2 ratio of 1:10 every 2 to 3 days by trypsinization (0.05% Trypsin expression we used two additional sample sets: a publicly avail- with EDTA 4Na 1x, Life Technologies, 25300062) and plating able data set from the TCGA (21) and our own collection contain- 2 onto 10-cm dishes. To differentiate Caco-2 cells, these were ing 16 matched pairs of tumor and adjacent mucosa from the 2 grown until they reached confluency onto 10-cm dishes. Upon same patients. We found that PADI2 was downregulated in 15 of confluency, medium was changed every 2 days and cells were these 16 samples (Fig. 1B). In most samples the downregulation harvested after 20 days. was more than 8-fold. Next, we analyzed available RNA-seq data from the TCGA Consortium. For this we used the Wanderer RT-qPCR and immunoblotting application that we had developed to facilitate the analysis of Cell lysis, Western blot, and RNA extraction were performed the rather bulky TCGA data for individual genes or loci (24). The as previously described (22). For immunoblotting, we have data set based on 41 normal mucosa samples and 262 colorectal

842 Mol Cancer Res; 14(9) September 2016 Molecular Cancer Research

PADI2 Is Lost in Colorectal Cancer

A Figure 1. 12 PADI1 PADI2 PADI3 PADI4/5 PADI6 Normal mucosa The arginine deiminase PADI2 is downregulated in colorectal cancer. Adjacent mucosa 9 A, the expression range of PADI family Tumor members has been extracted from a microarray analysis performed on a 6 cohort of 98 colorectal cancer patients (both tumor and adjacent mucosae) 3 and 50 age-matched cancer-free Expression range donors of normal mucosa. The so- termed "colonomics" cohort has been previously described (20). The B C indicated P value refers to the PADI2 TCGA 1 comparison tumor and normal mucosa 14 n = 41 0 using the Wilcoxon test. For genes n = 262 represented by more than one probe, a −1 12 − summary (PCA) is shown. B, PADI2 2 10 ﬁ − mRNA has been quanti ed by RT-qPCR 3 ed rsem + 1) 8 performed on RNA extracted from 16 −4 colorectal tumor and matched mucosa −5 m aliz 6 samples from the same patients. C, the −6 PADI2 Expression expression of PADI2 has been log2(tumor/normal) 4 mRNA Level change −7 log2(nor calculated from RNA-seq data of 262 P = 8.0e−23 − 2 colorectal cancers and 41 control 8 normal mucosa samples generated Patient: abcdef gh i j k lmnop mal by TCGA. umor Nor T

adenocarcinomas further confirmed the downregulation of confirmed that expression of PADI2 is characteristic of mucosal PADI2 with a Wilcoxon P value ¼ 8.0e23 (Fig. 1C). As promoter enterocytes in the colon crypt (Fig. 2D). A similar expression of DNA methylation frequently correlates with the reduced gene PADI2 was previously described in other epithelial cells from expression (25), we have analyzed DNA methylation data from kidney, endometrium, and mammary gland (28). TCGA. However, we found no correlation of DNA methylation and PADI2 expression (data not shown). Taken together, these Downregulation of PADI2 expression is an early event in results show that downregulation of PADI2 is a frequent feature of colonic tumourigenesis colorectal cancer. Colorectal cancer is a disease with a step-wise progression from early dysplasia to advanced adenocarcinoma with infiltrating PADI2 expression correlates with enterocytic differentiation features (29). To get an understanding at which step the expres- Next, we analyzed the protein levels of PADI2 in a panel of sion of PADI2 is lost, we have performed anti-PADI2 immuno- established colorectal cancer cells lines, including many that are histochemistry in a set of samples from 24 colorectal cancer regularly used for molecular research, such as HCT116, HT-29, patients that contained normal mucosa, areas of dysplasia, and and SW480. PADI2 was virtually undetectable by immunoblot- infiltrating adenocarcinoma in the same tissue sample (Fig. 3A). ting in most cell lines (Fig. 2A). The only colorectal cancer cell line While PADI2 was expressed in normal mucosa, expression was found to express low to moderate levels of PADI2 was HT-29. The frequently already lost in low-grade dysplasia. In 12 of 24 samples MBU-TS4 cell line derived from transformed keratinocytes PADI2 such dysplastic areas were negative for PADI2 and in another 7 has been included as a reference for high endogenous PADI2 samples expression was reduced and only focally retained (Fig. expression (26). Our preliminary results suggest that overexpres- 3B). In line with the mRNA expression results of Fig. 1, PADI2 sion of PADI2 in colorectal cancer cell lines at a comparable or protein was not detected in the large majority (21 of 24) of even higher level than in MBU-TS4 cells does not affect the infiltrating adenocarcinomas (Fig. 3A and B). These results argue adhesion to collagen I (Supplementary Fig. S1A and S1B), but in favor of PADI2 downregulation as an early event in colorectal might render HT-29 cells sensitive to calcium influx-induced tumourigenesis. reduction in proliferation and viability correlating with the transcriptional activation of the cell cycle inhibitory CDKN1A/p21 PADI2 expression is also low in active ulcerative colitis (Supplementary Fig. S1C and S1D). In some cases, it is challenging to distinguish early low-grade Next, we addressed the question whether the low or absent dysplasia of colorectal cancer from reactive but noncancerous expression of PADI2 could be a consequence of the dedifferen- epithelia. This situation occurs in the setting of active ulcerative tiated state of these colorectal cancer cell lines. For this we used colitis, which is a form of inflammatory bowel disease. Ulcerative CaCo2 cells that can be induced to differentiate in cells with mature colitis is characterized by chronic and episodic colonic inflam- enterocytic features (27). While undifferentiated Caco2 cells were mation that affects adolescents and young adults with an etiology negative for PADI by immunoblotting, in differentiated cells the and pathogenesis that have not been fully elucidated (30). His- protein was readily detected (Fig. 2B). Similarly, mRNA levels tologically the process involves the mucosa at the anal portion were strongly upregulated in differentiated Caco2 cells (Fig. 2C). and extends proximately in a contiguous fashion. The affected Immunohistochemical staining of normal colon mucosa further mucosa shows a lymphoplasmacytic inflammation in the lamina

www.aacrjournals.org Mol Cancer Res; 14(9) September 2016 843

Cantarino~ et al.

A CRC Cell lines B Caco-2

HCT116HT-29SW480DLD-1Colo205Caco-2MBU-TS4 0 20 Days of Figure 2. differentiation PADI2 expression is mostly absent in PADI2 PADI2 colorectal cancer cell lines but induced during enterocytic differentiation. Histone H3 Histone H3 A, crude cell lysates from colorectal cancer cell lines were analyzed by immunoblotting using anti-PADI2 and anti-Histone H3 antibodies. A lysate from MBU-TS4 squamous cell C D carcinoma cells known to express Caco-2 Healthy colon mucosa PADI2 were included as a positive control. B, crude cell lysates from 35 proliferating and differentiated Caco-2 30 Differentiation cells were analyzed by immunoblotting 0 Days as in A. C, the level of PADI2 mRNA 25 was quantiﬁed by RT-qPCR in 20 20 Days proliferating and differentiated Caco-2 cells (n ¼ 3; error bars ¼þSD). D, 15 immunohistological analysis of healthy colon mucosa using anti-PADI2 10 antibody. 5 Relative mRNA level Relative mRNA 0 PADI2 PADI4

propia with cryptitis and crypt abscesses by neutrophilic infiltra- cancerous epithelial alterations from dysplasia are not available tion (31). The repeated episodes of active inflammation increase yet. the risk of colorectal cancer usually preceded by dysplasia (32). For this reason, we have analyzed the expression of PADI2 in 11 The identification of dysplasia in these patients is crucial to cases of active ulcerative colitis (Fig. 4). In all 11 cases, the warrant the best management at a curable stage of their disease. expression of PADI2 was not detectable in the epithelial crypt However, biomarkers that allow distinguishing reactive but non- cells arguing that the downregulation of PADI2 is not a specific

A B Samples (n) 01218246 H&E Normal mucosa

Dysplasia Figure 3. Downregulation of PADI2 is an early Adenocarcinoma event in the development of colorectal cancer. A, tissue samples containing areas of normal mucosa, low-grade Positive dysplasia, and inﬁltrating PADI2 adenocarcinoma were analyzed by Focally positive anti-PADI2 immunohistochemistry Negative and hematoxylin and eosin staining (n ¼ 24). A representative example is shown. B, the expression of PADI2 was scored as positive, focally positive, or negative. A quantiﬁcation of all 24 analyzed samples is shown.

844 Mol Cancer Res; 14(9) September 2016 Molecular Cancer Research

PADI2 Is Lost in Colorectal Cancer

Ulcerative 0.12–0.76, P ¼ 0.011). Patients with lower PADI2 levels, H&E PADI2 PADI2 (zoomed) colitis either in tumor or adjacent mucosa had the worse prognosis (Fig. 5). It is well accepted that microenvironment can affect Patient A tumor growth and progression (33). Our results suggest that in addition to a possible tumor-intrinsic function, the absence of PADI2 in adjacent mucosa in colon cancer patients could be part of an altered tumor-promoting environment. Whether this aberrant loss is also present in some cancer-free indivi- Patient B duals (Fig. 1A) conferring them higher cancer risk remains to be further studied. Because PADI2 was also downregulated in noncancerous ulcerative colitis (Fig. 4), we wondered whether the association of PADI2 expression in adjacent mucosa with reduced survival could Patient C be explained tumor-surrounding inflammation. To address this question, we have estimated an inflammatory index using the expression data for 200 genes (listed in Table S1) extracted from fl Figure 4. the in ammatory hallmark gene set from database MSigDB (34). Enterocytic PADI2 expression is also lost in active ulcerative colitis. Samples This signature has been derived from diverse studies of genes from 11 patients with active ulcerative colitis were analyzed by anti-PADI related to inflammatory response. Also, as indirect evidence of immunohistochemistry. Arrow heads indicate example colon crypts with immune cell infiltration, we have used the expression of CD8A reduced PADI2 expression in enterocytes. Zoomed sections of the same crypts that has been described as a specific marker of T lymphocyte fi are shown on the right. Arrows indicate in ltrating neutrophils that stained infiltration (35). The analysis of these two signatures correlated strongly positive for PADI2. with PADI2 expression in adjacent normal mucosa, but in oppo- site directions. The inflammatory index showed an inverse cor- event of cancer initiation. It can be noted that infiltrating neu- relation with PADI2, while CD8A was positively correlated (Sup- trophils stained strongly positive for PADI2 (Fig. 4). As a conse- plementary Fig. S2A). In any case, the analysis of the prognostic quence, the analysis of mRNA extracted from the bulk tissue value of PADI2 in adjacent normal mucosa remained significant sample would be likely to provide a false-positive result. As PADI2 (HR ¼ 0.22; 95% CI, 0.06–0.81, Cox adjusted P ¼ 0.023) after is downregulated in both dysplasia and reactive changes associ- adjustment for age, gender, tumor site, hallmark inflammatory ated with active ulcerative colitis, we conclude that it has limited index, and CD8A expression. We have further used the R-package potential to be used as a discriminating diagnostic biomarker. "estimate" (36) to infer an immune score that seemed unrelated to PADI2 expression (Supplementary Fig. S2A). Again the prog- Lower levels of PADI2 in tumor and adjacent mucosa correlate nostic value of PADI2 expression remained significant when with poor prognosis adjusting for immune score in addition to age, gender, tumor Next, we took advantage of the availability of up to 10 years site, and inflammatory index (Supplementary Fig. S2B). Our of follow-up information on the cohort of 98 colorectal cancer results further suggest that PADI2 expression in adjacent mucosa patients. We have used the expression range calculated for Fig. has prognostic value independent of inflammation and immune 1A to group patients according to their expression of PADI2. cell infiltration. Adjacent mucosa and tumor samples required different cutoff Taken together, our results show that the absence or loss of points because the expression values had little overlap. For the PADI2 expression is a frequent and early event in colorectal tumor samples, the difference between the two groups was cancerogenesis. Assessing the expression level of PADI2 in tumor statistical significant at cutoff 5.4 (HR ¼ 0.29; 95% CI, 0.09– or adjacent mucosa has a prognostic value for colorectal cancer. 0.98, P ¼ 0.046). The tumor-adjacent mucosa required a Our results further suggest that for tumor-adjacent mucosa this higher expression level cutoff at 7.7, which was significantly prognostic value is independent of inflammation and immune associated with disease-free survival (HR ¼ 0.82; 95% CI, cell infiltration.

Adjacent mucosa Tumor 1.0 1.0 High n = 33 Figure 5. 0.8 0.8 n High expression of PADI2 in tumor-adjacent mucosa High = 86 n correlates with good prognosis. A Kaplan–Meier curve of 0.6 0.6 Low = 65 disease-free survival is shown. Using colonomics data from a cohort of 98 colorectal cancer patients (same as Low n = 12 in Fig. 1A) two expression levels were discerned using low 0.4 0.4 [0–7.82] and high (7.8–10.1] for adjacent mucosa and [0– 5.14] and high (5.14–8.05] for tumor. Cox model has been 0.2 0.2 used to calculate P values. Disease-free survival P = 0.011 P = 0.046 0.0 0.0 0246810 0246810

www.aacrjournals.org Mol Cancer Res; 14(9) September 2016 845

Cantarino~ et al.

Discussion PADI2 has an ambivalent role in cancer Adding to the complexity, the role of PADI in tumors seems to PADI2 expression correlates with differentiation be context dependent. In contrast to colorectal cancer, an onco- Here, we report the downregulation of the peptidyl deimi- genic role for PADI2 has been reported in breast cancer and nase family member PADI2 at an early stage of colorectal cancer squamous carcinomas (15, 16). In a cell culture model for breast formation. The function of PADIs is best understood in skin. cancer progression, PADI2 expression correlated with the trans- Keratinocytes express PADI isoforms 1–3 which citrullinate formed phenotype and cells and xenografted tumors were sensi- keratin 1 and filaggrin during the terminal differentiation of tive to Cl-amidine treatment (15). Mice overexpressing PADI2 in a keratinocytes (37). The consequential accumulation of citrul- panel of tissues including skin and mammary epithelia sponta- linated proteins at the outer layers of the epidermis is thought neously developed squamous cell carcinomas in the skin in 37% to be a key event to maintain the cutaneous barrier function. of the cases but no other tumors (16). In vitro data suggested that Similar to epidermis, the colon mucosa is another example of overexpression of PADI2 in skin carcinoma cells could favor an epithelium with a high cell turnover. Here, we show that epithelial-to-mesenchymal transitions and thus increase the cells' PADI2 protein is present in enterocytes and, analogous to the migratory capacity (16). In contrast, in colorectal cancer, PADI2 is epidermis, is predominantly expressed in the more differenti- downregulated and reduced expression in both tumor tissue and ated cells at the tips of the villi and less in the crypts where the adjacent mucosa correlated with reduced survival. Reexpression stem cell compartment is located (Fig. 2D). In this line, the of PADI2 in colorectal cancer cell lines had no influence on mRNA encoding PADI2 was detected upregulated in a study adhesion. Our preliminary experiments however suggested that comparing the global gene expression pattern of colon villi tops PADI2 overexpression could render HT-29 cells sensitive to cal- with crypts (38). While many colorectal cancer cells express cium treatment. Calcium treatment led to a reduction of viability very little PADI2, its expression is induced when forcing tumor and proliferation that correlated with the upregulation of the cell- cells to differentiate in vitro.Hence,wehypothesizethatthe cycle inhibitory gene CDKN1A/p21 (Supplementary Fig. S1C and apparent loss of PADI2 expression in colorectal cancer may be a S1D). This link between PADI2 and p21 is particularly interesting consequence of reflecting the undifferentiated or dedifferen- given that HT-29 cells harbor a change-of-function mutation in tiated status of tumor cells. p53 and thus lack the classic p21-activating pathway. These observations warrant further exploration as any alternative path- Does PADI2 serve as early diagnosis biomarker? way that leads to the activation of p21 might have therapeutic At present, there is no biomarker to help pathologists dis- potential for the many p53 mutant cancers. Taken together, in tinguish between dysplastic changes and noncancerous lesions colon cancer PADI2 seems to act as tumor suppressor, which such as inflammation, thus choosing the most appropriate contrasts with reported functions in skin and mammary cancer. treatment scheme is challenging. Because reduction of PADI2 The dual role of cancer genes as oncogenes and tumor suppressors expression is an early event in colorectal cancer, PADI2 was is common, if not universal, in cancer pathogenesis and depends tested as a biomarker. However, when analyzing PADI2 expres- on the tissue context (44–46). sion in active ulcerative colitis samples, we found that PADI2 is also downregulated in this noncancerous pathology. At first Why does PADI2 in adjacent mucosa predict survival? sight, this finding was somewhat surprising taking into account Patients with higher PADI2 expression in the tumor or the previous studies which have highlighted the role of PADI adjacent mucosa have best prognosis and, conversely, down- activity, in particular PADI4 in inflammatory processes regulation of PADI2 correlates with poor survival. It is intriguing (reviewedinref.39).Inthecontext of ulcerative colitis, the that the expression level of PADI2 in the peritumoral mucosa is pan-PADIinhibitorCI-amidinewasfoundtobeeffectiveina associated with survival. As a possible explanation, it can be mouse model where it reduced inflammation (40). As an speculated that in addition to a tumor-intrinsic function, PADI2 explanation, the authors found that the drug acted primarily might modulate the microenvironment in a way that when lost on inflammatory immune cells and induced their apoptosis favors tumor progression. The modification of secreted compo- while epithelial cells remained unaffected (40). Another study nents by PADI2 has the potential to provide a molecular mech- reported increased levels of citrullination in inflammatory anism. PADIs catalyze the conversion of charged arginine residues bowel disease (41). From their immunohistochemistry, it is into neutral citrullines. The net removal of a positive charge has apparent that a fraction of the citrullinated proteins are extra- great potential to change the substrate's protein–protein interac- cellular. As such, it would be interesting to determine to which tion capacity and function. In fact, Struyf and colleagues already extent infiltrating immune cells cause the observed citrullina- demonstrated in vitro that citrullination of the cytokine CXCL12, tion and to what level epithelial cells contribute to it. Similarly, expressed by colonic epithelial cells, reduces its binding affinity to we lack any knowledge about the identity of the PADI enzyme CXCR4 and CXCR7 receptors located on monocyte and lympho- responsible for the citrullination. It seems likely that PADI2 cyte cell surfaces (47). As a result, PADI may play an anti-inflam- may have distinct roles in inflammatory cells and in differen- matory role, attenuating the recruitment of CXCR4-positive leu- tiated epithelia. While the function of PADI2 and PADI4 in kocytes to sites of inflammation. Our finding that the predictive immune cells has been well described (7, 42, 43), the role of potential of PADI2 was independent of gene signatures related to PADI2 in colonic epithelium remains unexplored. We now inflammation and immune cell infiltration, however, argues contribute with a novel result showing that the apparent loss against this. From studies in skin, PADIs are known to modify of PADI2 expression occurs early in tumorigenesis. It remains components of the extracellular matrix (37). Changes in the to be determined if it can be used as a biomarker for colorectal extracellular matrix have clear potential to affect adhesion and cancer risk as a similar loss in epithelial expression also occurs migration of embedded cancer cells. In a first exploratory set of in ulcerative colitis, which is a cancer predisposing condition. experiments, however, we were unable to detect any influence of

846 Mol Cancer Res; 14(9) September 2016 Molecular Cancer Research

PADI2 Is Lost in Colorectal Cancer

PADI2 overexpression on the capacity of colorectal cancer cells to Writing, review, and/or revision of the manuscript: N. Cantarino,~ E. Musulen, adhere to collagen. In a theory first put forward for oral carcinoma, M. Perucho, V. Moreno, S.-V. Forcales, J. Douet, M. Buschbeck Administrative, technical, or material support (i.e., reporting or organizing it has further been suggested that changes in normal epithelium ~ fi data, constructing databases): N. Cantarino can facilitate tumor initiation by providing a permissive eld for Study supervision: N. Cantarino,~ M.A. Peinado, S.-V. Forcales, J. Douet, cancerization (48). In this context, hypermethylation and silenc- M. Buschbeck ing of ADAMTS14, a member of the extracellular matrix metallo- proteinases, in normal colon mucosa associates with a field effect Acknowledgments for cancerization (49). Whether PADI2 could be part of such a We thank Scott Coonrod for the PADI2 expression construct. field effect remains speculative at this stage. Taken together, the lack of PADI2 expression is an early event in the development of colorectal cancer and its expression level Grant Support correlates with differential survival. Future work will be required This work was funded by grants from MINECO (SAF2012-39749 and BFU2015-66559-P) to M. Buschbeck and the Olga Torres Foundation to to determine the substrates and citrullination events that are S.-V. Forcales. Research in the Buschbeck lab is further supported by the relevant for this association and whether these alterations could Deutsche Jose Carreras Leukaemie Stiftung (DJCLS R 14/16), AFM Telethon have an application as biomarkers or points for therapeutic (AFM 18738), the Marie Skøodowska Curie Training network "ChroMe" intervention. (H2020-MSCA-ITN-2015-675610), AGAUR (2014-SGR-35), Fundacio Inter- nacional Josep Carreras, Foundation "Obra Social la caixa," and the Asociacion Espanola~ Contra el Cancer (AECC–Junta de Barcelona). V. Moreno was sup- Disclosure of Potential Conflicts of Interest ported by Catalan Government DURSI (2014SGR647) and Instituto de Salud M.A. Peinado is Chief ScientificOfficer at Aniling. No potential conflicts Carlos III, co-funded by FEDER funds "a way to build Europe" (PS09-1037, of interest were disclosed by the other authors. PI11-01439 and PIE13/00022). N. Cantarino~ was supported by a FPI PhD fellowship (BES-2010-031876), J. Douet by a Juan de la Cierva fellowship (JCI- Authors' Contributions 2011-10831) and M. Buschbeck was a Ramon y Cajal fellow (RYC2010-07337), all funded by MINECO. ~ Conception and design: N. Cantarino, J. Douet, M. Buschbeck The costs of publication of this article were defrayed in part by the ~ Development of methodology: N. Cantarino, V. Valero, J. Douet payment of page charges. This article must therefore be hereby marked Acquisition of data (provided animals, acquired and managed patients, advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate ~ provided facilities, etc.): N. Cantarino, E. Musulen, V. Valero, M. Perucho, this fact. V. Moreno, M. Buschbeck Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): N. Cantarino,~ E. Musulen, V. Moreno, S.-V. Forcales, Received January 28, 2016; revised April 25, 2016; accepted May 23, 2016; M. Buschbeck published OnlineFirst June 8, 2016.

References 1. Bicker KL, Thompson PR. The protein arginine deiminases: Structure, protein citrullination in cancer pathogenesis. Biochem Res Int 2012; function, inhibition, and disease. Biopolymers 2013;99:155–63. 2012:1–11. 2. Vossenaar ER, Zendman AJW, van Venrooij WJ, Pruijn GJM. PAD, a growing 12. Valesini G, Gerardi MC, Iannuccelli C, Pacucci VA, Pendolino M, family of citrullinating enzymes: genes, features and involvement in Shoenfeld Y. Citrullination and autoimmunity. Autoimmun Rev disease. Bioessays 2003;25:1106–18. 2015;14:490–7. 3. Darrah E, Rosen A, Giles JT, Andrade F. Peptidylarginine deiminase 2, 3 and 13. Wiik AS, van Venrooij WJ, Pruijn GJM. All you wanted to know about anti- 4 have distinct specificities against cellular substrates: novel insights into CCP but were afraid to ask. Autoimmun Rev 2010;10:90–3. autoantigen selection in rheumatoid arthritis. Ann Rheum Dis 2011; 14. Chang X, Han J. Expression of peptidylarginine deiminase type 4 (PAD4) in 71:92–8. various tumors. Mol Carcinog 2006;45:183–96. 4. Knuckley B, Causey CP, Jones JE, Bhatia M, Dreyton CJ, Osborne TC, et al. 15. McElwee JL, Mohanan S, Griffith OL, Breuer HC, Anguish LJ, Cherrington Substrate specificity and kinetic studies of PADs 1, 3, and 4 identify potent BD, et al. Identification of PADI2 as a potential breast cancer biomarker and and selective inhibitors of protein arginine deiminase 3. Biochemistry therapeutic target. BMC Cancer 2012;12:1–1. 2010;49:4852–63. 16. McElwee JL, Mohanan S, Horibata S, Sams KL, Anguish LJ, McLean D, et al. 5. Cuthbert GL, Daujat S, Snowden AW, Erdjument-Bromage H, Hagiwara T, PAD2 overexpression in transgenic mice promotes spontaneous skin Yamada M, et al. Histone deimination antagonizes arginine methylation. neoplasia. Cancer Res 2014;74:6306–17. Cell 2004;118:545–53. 17. Kolodziej S, Kuvardina ON, Oellerich T, Herglotz J, Backert I, Kohrs N, et al. 6. Wang Y, Wysocka J, Sayegh J, Lee Y-H, Perlin JR, Leonelli L, et al. Human PADI4 acts as a coactivator of Tal1 by counteracting repressive histone PAD4 regulates histone arginine methylation levels via demethylimina- arginine methylation. Nat Commun 2014;5:1–12. tion. Science 2004;306:279–83. 18. Guertin MJ, Zhang X, Anguish L, Kim S, Varticovski L, Lis JT, et al. 7. Vossenaar ER.Expression and activity of citrullinating peptidylarginine Targeted H3R26 deimination specifically facilitates estrogen receptor deiminase enzymes in monocytes and macrophages. Ann Rheum Dis binding by modifying nucleosome structure. PLoS Genet 2014;10: 2004;63:373–81. e1004613. 8. Masson-Bessiere C, Sebbag M, Girbal-Neuhauser E, Nogueira L, Vincent C, 19. Haggar F, Boushey R. Colorectal cancer epidemiology: incidence, Senshu T, et al. The major synovial targets of the rheumatoid arthritis- mortality, survival, and risk factors. Clin Colon Rectal Surg 2009;22: specificantifilaggrin autoantibodies are deiminated forms of the alpha- 191–7. and beta-chains of fibrin. J Immunol 2001;166:4177–84. 20. Sanz-Pamplona R, Berenguer A, Cordero D, Mollev DG, Crous-Bou M, Sole 9. Brinkmann V, Reichard U, Goosmann C, Fauler B, Uhlemann Y, Weiss DS, X, et al. 1476-4598-13-46. Mol Cancer 2014;13:1–19. et al. Neutrophil extracellular traps kill bacteria. Science 2004;303:1532–5. 21. Weinstein JN, Collisson EA, Mills GB, Shaw KRM, Ozenberger BA, Ellrott K, 10. Christophorou MA, Castelo-Branco G, Halley-Stott RP, Oliveira CS, Loos R, et al. The cancer genome atlas pan-cancer analysis project. Nat Genet Radzisheuskaya A, et al. Citrullination regulates pluripotency and histone 2013;45:1113–20. H1 binding to chromatin. Nature 2014;507:104–8. 22. Creppe C, Palau A, Malinverni R, Valero V, Buschbeck M. A Cbx8-Contain- 11. Mohanan S, Cherrington BD, Horibata S, McElwee JL, Thompson PR, ing polycomb complex facilitates the transition to gene activation during Coonrod SA. Potential role of peptidylarginine deiminase enzymes and ES cell differentiation. PLoS Genet 2014;10:e1004851.

www.aacrjournals.org Mol Cancer Res; 14(9) September 2016 847

Cantarino~ et al.

23. Boix R, Sanz C, Mora M, Quer A, Beyer K, Musulen E, et al. Primary renal cell 37. Senshu T, Kan S, Ogawa H, Manabe M, Asaga H. Preferential carcinoma in a transplanted kidney: genetic evidence of recipient origin. deimination of keratin K1 and filaggrin during the terminal differ- Transplantation 2009;87:1057–61. entiation of human epidermis. Biochem Biophys Res Commun 1996; 24. Diez-Villanueva A, Mallona I, Peinado MA. Wanderer, an interactive viewer 225:712–9. to explore DNA methylation and gene expression data in human cancer. 38. Kosinski C, Li VSW, Chan ASY, Zhang J, Ho C, Tsui WY, et al. Gene Epigenetics Chromatin 2015 Jun 23;8:22. expression patterns of human colon tops and basal crypts and BMP 25. Baylin SB, Jones PA. A decade of exploring the cancer epigenome—biological antagonists as intestinal stem cell niche factors. Proc Natl Acad Sci USA and translational implications. Nat Rev Cancer 2011;11:726–34. 2007;104:15418–23. 26. Cantarino~ N, Fernandez-Figueras MT, Valero V, Musulen E, Malinverni R, 39. Gudmann NS, Hansen NUB, Jensen ACB, Karsdal MA, Siebuhr AS. Bio- Granada I, et al. A cellular model reflecting the phenotypic heterogeneity of logical relevance of citrullinations: diagnostic, prognostic and therapeutic mutant HRASdriven squamous cell carcinoma. Int J Cancer 2016 Apr 13. options. Autoimmunity 2015;48:73–9. doi:10.1002/ijc.30139. [Epub ahead of print]. 40. Chumanevich AA, Causey CP, Knuckley BA, Jones JE, Poudyal D, Chu- 27. Chantret I, Barbat A, Dussaulx E, Brattain MG, Zweibaum A. Epithelial polarity, manevich AP, et al. Suppression of colitis in mice by Cl-amidine: a novel villin expression, and enterocytic differentiation of cultured human colon peptidylarginine deiminase inhibitor. Gastrointest Liver Physiol 2011;300: carcinoma cells: a survey of twenty cell lines. Cancer Res 1988;48:1936–42. G929–38. 28. Cherrington BD, Zhang X, McElwee JL, Morency E, Anguish LJ, Coonrod 41. Makrygiannakis D, af Klint E, Lundberg IE, Lofberg R, Ulfgren A-K, SA. Potential role for PAD2 in gene regulation in breast cancer cells. PLoS Klareskog L, et al. Citrullination is an inflammation-dependent process. ONE 2012;7:e41242. Ann Rheum Dis 2006;65:1219–22. 29. Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell 42. Wang Y, Li M, Stadler S, Correll S, Li P, Wang D, et al. Histone 1990;61:759–67. hypercitrullination mediates chromatin decondensation and neu- 30. Knights D, Lassen KG, Xavier RJ. Advances in inflammatory bowel disease trophil extracellular trap formation. The J Cell Biol 2009;184: pathogenesis: linking host genetics and the microbiome. Gut 2013; 205–13. 62:1505–10. 43. Nakashima K, Arai S, Suzuki A, Nariai Y, Urano T, Nakayama M, et al. PAD4 31. Magro F, Langner C, Driessen A, Ensari A, Geboes K, Mantzaris GJ, et al. regulates proliferation of multipotent haematopoietic cells by controlling European consensus on the histopathology of inflammatory bowel dis- c-myc expression. Nat Commun 2013;4:1836. ease. J Crohn's Colitis 2013;7:827–51. 44. Bosch-Presegue L, Vaquero A. The dual role of sirtuins in cancer. Gen 32. Ullman T, Odze R, Farraye FA. Diagnosis and management of dysplasia in Cancer 2011;2:648–62. patients with ulcerative colitis and Crohn's disease of the colon. Inflamm 45. Warfel NA, El-Deiry WS. p21WAF1 and tumourigenesis. Curr Opin Oncol Bowel Dis 2009;15:630–8. 2013;25:52–8. 33. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 46. Radtke F, Raj K. The role of Notch in tumorigenesis: oncogene or tumour 2011;144:646–74. suppressor? Nat Rev Cancer 2003;3:756–67. 34. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, 47. Struyf S, Noppen S, Loos T, Mortier A, Gouwy M, Verbeke H, et al. et al. Gene set enrichment analysis: a knowledge-based approach for Citrullination of CXCL12 differentially reduces CXCR4 and CXCR7 bind- interpreting genome-wide expression profiles. Proc Natl Acad Sci USA ing with loss of inflammatory and anti-HIV-1 activity via CXCR4. 2005;102:15545–50. J Immunol 2009;182:666–74. 35. Lugli A, Karamitopoulou E, Panayiotides I, Karakitsos P, Rallis G, Peros G, 48. Slaughter DP, Southwick HW, Smejkal W. Field cancerization in oral et al. CD8þ lymphocytes/tumour-budding index: an independent prog- stratified squamous epithelium; clinical implications of multicentric ori- nostic factor representing a "pro-/anti-tumour" approach to tumour host gin. Cancer 1953;6:963–8. interaction in colorectal cancer. Br J Cancer 2009;101:1382–92. 49. Alonso S, Dai Y, Yamashita K, Horiuchi S, Dai T, Matsunaga A, et al. 36. Yoshihara K, Shahmoradgoli M, nez EMI, Vegesna R, Kim H, Torres-Garcia Methylation of MGMT and ADAMTS14 in normal colon MUCOSA: W, et al. Inferring tumour purity and stromal and immune cell admixture biomarkers of a field defect for cancerization preferentially targeting elder from expression data. Nat Commun 2013;4:1–11. African-Americans. Oncotarget 2015;6:3420–31.

848 Mol Cancer Res; 14(9) September 2016 Molecular Cancer Research

Downregulation of the Deiminase PADI2 Is an Early Event in Colorectal Carcinogenesis and Indicates Poor Prognosis

Neus Cantariño, Eva Musulén, Vanesa Valero, et al.

Mol Cancer Res 2016;14:841-848. Published OnlineFirst June 8, 2016.

Updated version Access the most recent version of this article at: doi:10.1158/1541-7786.MCR-16-0034

Supplementary Access the most recent supplemental material at: Material http://mcr.aacrjournals.org/content/suppl/2016/06/25/1541-7786.MCR-16-0034.DC1

Cited articles This article cites 48 articles, 12 of which you can access for free at: http://mcr.aacrjournals.org/content/14/9/841.full#ref-list-1

Citing articles This article has been cited by 7 HighWire-hosted articles. Access the articles at: http://mcr.aacrjournals.org/content/14/9/841.full#related-urls

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/14/9/841. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.