Cytidine Deaminase Deficiency Reveals New Therapeutic Opportunities

Home , Deoxyuridine

Author Manuscript Published OnlineFirst on September 6, 2016; DOI: 10.1158/1078-0432.CCR-16-0626 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

1 Cytidine deaminase deficiency reveals new therapeutic opportunities

2 against cancer

3 Hamza Mameri1,2,3, Ivan Bièche4, Didier Meseure5, Elisabetta Marangoni6, Géraldine

4 Buhagiar-Labarchède1,2,3, André Nicolas5, Sophie Vacher4, Rosine Onclercq-Delic1,2,3,

5 Vinodh Rajapakse7, Sudhir Varma7, William Reinhold7, Yves Pommier7 and Mounira Amor-

6 Guéret1,2,3

8 Authors’ affiliations:

9 1Curie Institute, PSL Research University, CNRS, UMR 3348, F-91405, Orsay, France.

10 2CNRS UMR 3348, Centre Universitaire, Bât. 110. 91405, Orsay, France

11 3Université Paris Sud, Université Paris-Saclay, CNRS, UMR 3348, F-91405 Orsay, France.

12 4Curie Institute – Genetic Department, 75005 Paris, France.

13 5Curie Institute, PSL Research University, 75005 Paris, France; Platform of Investigative

14 Pathology, 75005 Paris, France.

15 6Curie Institute, PSL Research University, Translational Research Department 75005 Paris,

16 France.

17 7Developmental Therapeutics Branch and Laboratory of Molecular Pharmacology, Center for

18 Cancer Research, NCI, NIH, Bethesda, MD, 20892, USA.

20 Corresponding author: Mounira Amor-Guéret, Curie Institute, CNRS UMR 3348, Centre

21 Universitaire, Bât. 110. 91405, Orsay, France. Phone: 33 1 69 86 30 53; Fax: 33 1 69 86 94

22 29; E-mail: [email protected]

23 Disclosure of Potential Conflicts of Interest

24 No potential conflicts of interest were disclosed

25 Running Title: CDA deficiency in Cancer

Downloaded from clincancerres.aacrjournals.org on September 28, 2021. © 2016 American Association for Cancer Research. Author Manuscript Published OnlineFirst on September 6, 2016; DOI: 10.1158/1078-0432.CCR-16-0626 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

1 Translational relevance

2 This study revealed for the first time that CDA expression is lost in a large proportion of

3 tumors, mostly due to DNA methylation, and that tumors from the same classically defined

4 groups may display differences in CDA expression status resulting in contrasting cellular

5 properties, such as levels of sister chromatids exchanges. Thus, the use of CDA expression

6 status in tumor cells defines two new subgroups: CDA-deficient tumors and CDA-proficient

7 tumors. Our results indicate that immunohistochemistry assessments of CDA levels could be

8 used to determine the CDA status of tumors, with potential implications for treatment. In

9 particular, we identified aminoflavone, which reached phase II clinical trials, as a proof-of-

10 principle candidate for the targeting of CDA-deficient tumor cells, with no effect on CDA-

11 proficient cells. Thus, CDA expression status could be used as a new marker to guide

12 anticancer therapy.

1 Abstract

2 Purpose: One of the main challenges in cancer therapy is the identification of molecular

3 mechanisms mediating resistance or sensitivity to treatment. Cytidine deaminase (CDA) was

4 reported to be downregulated in cells derived from patients with Bloom syndrome, a genetic

5 disease associated with a strong predisposition to a wide range of cancers. The purpose of this

6 study was to determine whether CDA deficiency could be associated with tumors from the

7 general population and could constitute a predictive marker of susceptibility to anti-tumor

8 drugs.

9 Experimental design: We analyzed CDA expression in silico, in large datasets for cancer

10 cell lines and tumors, and in various cancer cell lines and primary tumor tissues using

11 immunohistochemistry, PDXs, RT-qPCR, and western blotting. We also studied the

12 mechanism underlying CDA silencing and searched for molecules that might target

13 specifically CDA deficient tumor cells using in silico analysis coupled to classical cellular

14 experimental approaches.

15 Results: We found that CDA expression is downregulated in about 60% of cancer cells and

16 tissues. We demonstrate that DNA methylation is a prevalent mechanism of CDA silencing in

17 tumors. Finally, we show that CDA-deficient tumor cells can be specifically targeted with

18 epigenetic treatments and with the anticancer drug aminoflavone.

19 Conclusions: CDA expression status identifies new subgroups of cancers, and CDA

20 deficiency appears to be a novel and relevant predictive marker of susceptibility to antitumor

21 drugs, opening up new possibilities for treating cancer.

23 Key words: Cytidine deaminase, DNA methylation, Cancer, Biomarker, Drug sensitivity,

24 Cancer therapy

1 Introduction

2 Despite major advances in the development of chemotherapy, many cancers continue to have

3 a poor prognosis, due to the resistance of cancer cells to antineoplastic drugs through intrinsic

4 or acquired mechanisms (1). It is, thus, highly important to identify markers predicting the

5 response to anticancer treatment, and new molecular targets for novel anticancer treatments.

6 Two families of cytidine deaminases exhibit different biological functions, deaminating either

7 the free cytidines, as performed by cytidine deaminase (CDA or CDD; EC3.5.4.5), or

8 deaminating the cytidines incorporated within the DNA or RNA polymers, as performed by

9 the AID/APOBECs (Activation-induced deaminase /Apolipoprotein B mRNA editing

10 catalytic polypeptide-like) proteins (2,3). We previously reported that CDA deficiency leads

11 to DNA damage and to genetic instability (4,5), known to be associated with cancer

12 development, suggesting a possible relationship between CDA underexpression and cancer.

13 CDA is an enzyme of the pyrimidine salvage pathway catalyzing the hydrolytic deamination

14 of cytidine and deoxycytidine to uridine and deoxyuridine, respectively (6). In addition to

15 native nucleotides, CDA also deaminates and inactivates nucleoside analogs, such as

16 gemcitabine and cytosine arabinoside (Ara-C), agents widely used to treat cancer (7). CDA

17 thus plays an important role in the sensitivity/resistance of cancer cells to treatment with

18 cytidine analogs (8,9). Indeed, early severe toxicity has been reported in cancer patients with

19 low levels of CDA activity treated with gemcitabine (10,11). CDA overexpression might be

20 thus a marker for resistance to chemotherapy based on cytidine analogs.

21 CDA gene polymorphism has been widely explored (11–17) , but little is known about its

22 transcriptional and posttranscriptional regulation. One study recently reported that miR-484

23 downregulates CDA gene expression by targeting its 3’-UTR, sensitizing breast cancer cells

24 overexpressing CDA to gemcitabine (18). CDA overexpression is thus a good marker for

25 resistance to chemotherapy based on cytidine analogs.

1 CDA overexpression has recently been identified as a potential target for anticancer treatment.

2 Indeed, CDA has been shown to be involved in the deamination of oxidized and

3 epigenetically modified cytidine nucleosides (19,20). These cytidine analogs are selectively

4 converted to their uracil counterparts and then incorporated into DNA in cancer cells

5 overexpressing CDA, resulting in DNA damage and cell death (19,20).

6 In this study, we focused on CDA deficiency rather than CDA overexpression. We found that

7 CDA was downregulated in about 60% of cancer cells and tissues. The CDA gene was mostly

8 inactivated by DNA methylation. We identified a new subgroup of cancers not expressing

9 CDA that were susceptible to the specific toxic effects of drugs such as aminoflavone. Thus,

10 CDA deficiency in cancer cells is a new biomarker of cellular response to anticancer drugs.

11 Our findings open up promising new avenues for the treatment of cancer.

1 Materials and Methods

2 Cell culture and treatments

3 We used 33 cancer cell lines in this study (Supplementary Material 1): 19 breast cancer cell

4 lines from the Translational Research Department of the Curie Institute (ZR75-1, T47D,

5 HCC-1428, BT-474, MCF-7, MDA-MB-361, MDA-MB-468, MDA-MB-231, MDA-MB-

6 436, HCC-38, HCC-70, HCC-1187, HCC-1937, HCC-1143, BT-20, BT-549, HCC-1954,

7 SKBR-3, HS578T) and two nonmalignant breast cell lines (MCF-12A and 184B5), four lung

8 cancer cell lines (H522, H23, HOP-92, HOP-62), three ovarian cancer cell lines (IGROV-1

9 SKOV-3 and OVCAR-8) from the NCI (21,22), one melanoma cell line (A2058) from Dr.

10 Stephan Vagner’s laboratory (UMR3348 CNRS, Curie Institute), and two cervical cancer cell

11 lines (HeLa-Ctrl and HeLa-shCDA) and two Bloom syndrome cell lines (BS-Ctrl and BS-

12 CDA), previously described (5). All the breast cell lines were authenticated at the

13 Translational Research Department of the Curie Institute using standard DNA microsatellite

14 short tandem repeat (STR) method. Lung and ovarian cancer cell lines were authenticated at

15 the NCI using standard DNA microsatellite STR method (21,22). No specific authentication

16 of the A2058 melanoma cell line was performed. The isogenic HeLa-Ctrl/HeLa-shCDA and

17 BS/BS-CDA cell lines were established in our laboratory and are checked routinely for

18 several phenotypic characteristics such as sister chromatid exchange and ultrafine anaphase

19 bridge frequencies (4,5).

20 All cells were routinely checked for the absence of mycoplasma and were maintained in the

21 recommended media (see Supplementary Material 1) before the extraction of DNA, RNA and

22 proteins.

23 For evaluation of the induction of CDA expression, RNA was isolated from cell lines

24 continuously treated with 1 or 2.5 µM of 5-Aza-2′-deoxycytidine (5-Aza-dC - Sigma Aldrich)

25 for 96 hours.

1 Cell viability was carried out with 3-(4,5-dimethyl-2-thiazolyl)-2,5 diphenyl-2H-tetrazolium

2 bromide (MTT- Life Technologies) in 96-well microplates. We assessed the functionality of

3 CDA by plating HCC-1954 and IGROV-1 cells at densities of 2000 and 3000 cells/well,

4 respectively, on the day before pretreatment and at a density of 800 cells/well for control

5 conditions. Cells were then left untreated or subjected to pretreatment with 1 µM 5-Aza-dC

6 for 96 h, The cells were washed twice with PBS buffer, placed in fresh medium and incubated

7 for 72 hours in the presence of various concentrations of gemcitabine, from 0.001 to 1 µM

8 (Sigma Aldrich). The data were normalized to corresponding controls, for each condition. We

9 evaluated aminoflavone cytotoxicity after 72 h of treatment, by plating MCF-7, MDA-MB-

10 468, MDA-MB-231, SKOV-3, and OVCAR-8 cells at a density of 3000 cells/well and

11 IGROV-1 cells at a density of 4000 cells/well. Aminoflavone (NSC 686288) was provided by

12 Dr Yves Pommier (Developmental Therapeutics Branch - NCI).

14 DNA sequencing, quantitative PCR and western blotting

15 The 950 base pairs downstream from the translation initiation codon in the CDA promoter

16 region and the four exons were amplified by PCR, with the Phusion Polymerase enzyme

17 (Promega). The reaction was performed with 50 ng of genomic DNA isolated from 12 breast

18 cancer and two normal-like cell lines. The specific primers used for amplification and

19 nucleotide sequencing to base-pair resolution (Eurofins Genomics) are presented in

20 Supplementary Material 2.

21 The procedure for real-time PCR (RT-qPCR) was as described by Gemble et al.(5). In brief,

22 total RNA was extracted from PDX tissues and from cell lines with the RNeasy Mini Kit

23 (Qiagen). Reverse transcription was performed on 1 µg of RNA with the GoScript enzyme

24 (Promega). The cDNA obtained was used at a dilution of 1/10 for real-time PCR with the

25 SYBER Green supermix reagent (Biorad) in a Biorad CFX96 machine. Each sample was run

ΔΔ 1 in triplicate. Relative expression was determined by the 2- Ct method. GAPDH and TBP were

2 used as internal controls. The specific primers used for RT-qPCR analysis are presented in

3 Supplementary Material 2.

4 For western blotting, cells were harvested by centrifugation and lysed in 8 M urea, 50 mM

5 Tris-HCl, pH 7.5 and 150 mM β-mercaptoethanol buffer supplemented with protease inhibitor

6 (ThermoScientific). They were then sonicated and heated. Protein concentration was

7 estimated with the BCA kit (Pierce) and the equivalent of 20 µg or protein per cell lysate was

8 run on a 4–12% Bis-Tris pre-cast gel (Life Technologies). The proteins were then transferred

9 to PVDF membranes, which were probed with the appropriate antibody. Protein bands were

10 visualized with a CCD camera (BioRad). Details of the primary and secondary antibodies

11 used are provided in Supplementary Material 2.

13 Immunohistochemistry

14 Immunohistochemistry was carried out as described by Baldeyron et al.,(23). Briefly,

15 paraffin-embedded tissue blocks obtained at the initial diagnosis were retrieved from the

16 archives of the Biopathology Department of Curie Institute Hospital. Sections (3 µm thick)

17 were cut with a microtome from the paraffin-embedded tissue blocks. Tissue sections were

18 dewaxed and rehydrated through a series of xylene and ethanol washes. A primary anti-CDA

19 antibody (Ab) was used (Supplementary Material 2). The sections were processed with a

20 Dako machine for immunostaining. The specificity of the CDA Ab was confirmed by

21 applying the same protocol to paraffin-embedded human tissue sections and cell block

22 sections. The sections were rehydrated by incubation in PBS for 5 minutes and then incubated

23 with anti-CDA antibody for 1 hour. Antibody binding was detected by incubation with a

24 secondary antibody coupled to a peroxidase-conjugated polymer (Dako Envision +) after

25 treatment with DAB solution (Dako K3468) for 5 minutes, and Mayer’s hematoxylin for 1

1 minute. The sections were then mounted in resin. We evaluated CDA immunostaining on

2 histological sections from 19 normal human tissues (20 samples per tissue), and from 6

3 primary tumor tissues (50 samples per cancer type). For each section we evaluated two

4 immunohistological scores:

5 Intensity score: Score 0: no staining, Score 1+: weak staining, Score 2+: moderate staining,

6 Score 3+: intense staining.

7 Frequency score: Score 0: no staining, Score 1+: 1% - 33% stained cells, Score 2+: 34% -

8 67% stained cells, Score 3+: 68% - 100% stained cells.

9 Then we defined a final score (H score = frequency score x intensity score)

10 This H score was equal to 1 in normal colon tissue, and 1.5 in lung, breast, melanoma, ovary

11 and endometrium normal tissues. It means that the expression of CDA in normal tissues is

12 between > 1 and < 2.

13 Thus, the cut-off of CDA expression in tumor tissues was defined as: CDA under-expression

14 by H score between 0 and 1 (CDA low), and CDA overexpression by H score between 2 and

15 3 (CDA high). Thus, the data are presented as a combination of the percentage of CDA-

16 positive cells and intensity scores. The analysis was carried out by two independent

17 pathologists.

19 Breast cancer patient-derived xenografts

20 The PDX models used here were established as described by Marangoni et al. (24). Briefly,

21 breast cancer fragments were obtained from patients at the time of surgery, with the prior

22 written informed consent of the patients. Fragments (30 to 60 mm3) were grafted

23 subcutaneously into the interscapular fat pad of 8- to 12-week-old female Swiss nude mice,

24 under avertin anesthesia. Mice were maintained in specific pathogen-free animal housing

25 (Curie Institute) and received estrogen (17 mg/mL) in their drinking water. Xenografts

1 appeared at the graft site two to eight months after grafting. They were subsequently

2 transplanted from mouse to mouse and stored frozen in DMSO-fetal calf serum (FCS)

3 solution or dry-frozen in liquid nitrogen for RNA isolation. The experimental protocol was

4 performed in accordance with French regulations.

6 Sister chromatid exchange (SCE) assay

7 This assay was performed as described by Gemble et al.(5). In brief, cells were plated on

8 glass slides in the presence of 10 μM 5-bromodeoxyuridine (BrdU) (Sigma Aldrich). After

9 two divisions, colchicine (Sigma Aldrich) was added (0.1 μg/ml) and the cells were incubated

10 for 1 h. Cells were then incubated in a hypotonic solution (1:5 (vol/vol) FCS-distilled water)

11 and fixed with a 3:1 (vol/vol) mixture of methanol and acetic acid. They were then stained by

12 incubation with 10 μg/ml Hoechst 33258 (Sigma Aldrich) in distilled water for 20 minutes.

13 The slides were rinsed with 2×SSC (Euromedex) and exposed to ultraviolet light at a

14 wavelength of 365 nm and a distance of 10 cm for 105 minutes. The slides were then rinsed in

15 water, stained with 2% Giemsa (VWR) for 16 minutes, rinsed in water, dried and mounted in

16 EUKITT (Sigma Aldrich). Metaphases were captured and chromosomes were visualized

17 under a Leica DMRB microscope at a magnification of ×100. We determined the number

18 SCEs per chromosome.

20 DNA methylation data

21 We analyzed 482,422 CpGs in the NCI-60 cell lines with Illumina Infinium Human

22 Methylation 450 Beadchips. The DNA methylation datasets are available under accession

23 number GSE66872. The methylation values are presented from 0 to 1. The data were

24 normalized and analyzed as described by Nagales et al. (25).

25 The negative correlations between CDA promoter methylation and CDA expression on The

1 Cancer Genome Atlas (TCGA) samples (26) were generated through the Broad Institute

2 FireBrowse portal (27,28) and the cBioPortal for Cancer Genomics database (29–31) , all the

3 cBioPortal data (expression, mutation, copy number, significance analyses) being loaded

4 directly from FireBrowse. The only promoter CpG site presenting a high significant negative

5 correlation with CDA expression in both NCI-60 cell lines and TCGA samples was selected.

7 Transcriptomic data

8 A collection of 40 human breast tumor cell lines (mostly from ATCC) was established in the

9 Translational Research Department of the Curie Institute. Gene expression profiles were

10 generated with the Affymetrix Exon array and Genosplice algorithms to summarize

11 multiprobe measurements as single mRNA levels.

12 CDA expression levels were extracted from various transcriptomic datasets: breast tumor cell

13 lines of the Curie Institute collection (32), NCI-60 (CellMiner tools) (33), Cancer Cell Lines

14 Encyclopedia (CCLE) (34), Gene Expression Across Normal and Tumor Tissue database

15 (GENT) (35,36), the TCGA portal , and the Gene Expression Omnibus database (GEO) (37).

16 All these data are publicly accessible.

18 Statistics

19 All data analysis and processing were performed with GraphPad Prism 6 software. Pearson’s

20 correlation analysis was used to assess the association between two variables. P values for

21 sister SCEs were calculated by Mann-Whitney tests. CDA mRNA levels in normal and

22 cancerous tissues were compared in two-tailed unpaired t-tests. Differences in the induction of

23 CDA expression by 5-Aza-dC, as assessed by RT-qPCR, were evaluated in two-tailed paired

24 t-tests. Survival curves were compared in paired t–tests for HeLa-shCDA versus HeLa-Ctrl

25 cells treated with aminoflavone and HCC-1954 and IGROV-1 cells with and without 5-Aza-

1 dC pretreatment. Unpaired t-tests were used for the other cell lines. Differences were

2 considered statistically significant if P<0.05.

1 Results

2 CDA expression is downregulated in a large panel of cancer cell lines and tissues

3 We first analyzed CDA expression in cancer cell lines by combining in silico analyses on

4 cancer cell lines from the Curie Institute, the NCI, and the CCLE (Broad-Novartis Cancer

5 Cell Line Encyclopedia), and experimental approaches using qRT-PCR and western blotting

6 on a set of 26 representative cancer cell lines from the Curie Institute and the NCI. We found

7 that CDA was expressed weakly or not at all in 25 of 34 (73%) breast cancer cell lines from

8 the Curie Institute and 44 of 60 (73%) cancer cell lines derived from nine different organs and

9 tissues from the NCI (Fig. 1A, left and right panels). Similarly, about 60% (700) of the 1036

10 cancer cell lines from 24 different cancer tissues from the CCLE database did not express

11 CDA (Supplementary Fig. S1A). We validated these results, by RT-qPCR and western

12 blotting on a set of 26 representative cancer cell lines from the Curie Institute and the NCI

13 cancer cell line collections (19 breast cancer cell lines, 4 lung cancer cell lines, and 3 ovarian

14 cancer cell lines) (Fig. 1B).

15 We investigated whether the absence of detectable CDA expression observed in the majority

16 of cancer cell lines also applied to primary tumor tissues, by performing RT-qPCR to analyze

17 CDA mRNA levels in human primary breast tumors xenografted into nude mice (patient-

18 derived xenografts, PDXs). As control, we analyzed Cda mouse analog in parallel, to ensure

19 that mouse stroma was not interfering with human tumor data. This approach made it possible

20 to avoid the contamination of primary tumor tissues with normal cells from the stroma

21 (usually up to 30%). We found that 56 of the 66 (~84%) human primary breast tumors studied

22 had no significant CDA expression, and did not find any correlation between the levels of

23 mouse Cda expression and human CDA expression, indicating that mouse stroma did not

24 contribute to the xenograft data (Fig. 1C). We also analyzed CDA protein levels in six types

25 of primary cancer tissues (50 per type), and in 19 normal human tissues (20 per tissue) by

1 immunohistochemistry (IHC) with an anti-CDA antibody validated by IHC on isogenic

2 Bloom syndrome-derived cells not expressing CDA (BS-Ctrl) or expressing exogenous CDA

3 (BS-CDA) (5) (see the materials and methods section and the Supplementary Fig. S1B).

4 According to CDA staining intensity in normal tissues (Supplementary Fig.S1C, upper panel),

5 we stratified CDA expression into two groups on the basis of staining intensity scores: CDA

6 low (scores of 0 and 1) and CDA high (scores of 2 and 3) (Fig. 1D). About 50%

7 (endometrium) to 88% (breast triple-negative, ovary and colon) of cancer tissues displayed

8 very low levels of CDA expression.

9 We then compared CDA mRNA levels between healthy and cancerous tissues of different

10 origins, by replotting the CDA mRNA data downloaded from Gene Expression Omnibus

11 (GEO) found in different genomic data sources (Nextbio, Oncomine). Tumor tissues are often

12 contaminated with normal tissues that might express CDA, leading to inappropriate

13 interpretations of CDA expression in some tumor tissues. Nevertheless, CDA expression

14 levels were significantly lower in several tumors than in healthy tissues (Fig. 1E,

15 Supplementary Fig. S1D-E). These results (summarized in Supplementary Table 1) reveal that

16 CDA is overexpressed in some tumor tissues, such as those of pancreas, stomach, thyroid and

17 bladder cancers, as previously reported (19,35), but underexpressed in other tumor tissues,

18 such as those of liver, cervix, colon and esophagus cancers. We confirmed these results by

19 RT-qPCR on a small in-house cohort of colon tissues. We found that CDA expression levels

20 were significantly lower (P=0.0128) in tumor tissues (n=10) than in healthy tissues (n=5)

21 (Fig. 1F).

22 Finally, analysis of a recently published gene expression dataset used to determine the

23 molecular mechanism of cervical cancer progression (38) revealed that CDA expression

24 decreased considerably with cervical cancer progression (Supplementary Fig. S1F). This

25 result is consistent with the data presented in Fig. 1E, showing lower levels of CDA

1 expression in cervical cancer tissues than in non-cancerous tissues. Overall, these results

2 suggest that CDA expression tends to be lost during carcinogenesis, at least in some tissues,

3 such as the cervix.

5 CDA is downregulated by DNA methylation

6 We investigated the mechanism underlying the downregulation of the CDA gene in tumor

7 cells, by first analyzing CDA copy number in the DNA of the CCLE and NCI60 cell lines. No

8 significant correlation was found between CDA mRNA levels and CDA gene copy number

9 (Supplementary Fig. S2A). The downregulation of CDA levels cannot therefore be attributed

10 to genetic deletions in tumor cells.

11 We then carried out sequencing analysis, to determine whether CDA (promoter and exons)

12 was mutated in 11 breast cancer cell lines that did not express CDA, through comparison with

13 two breast cancer cell lines expressing high levels of CDA (HCC-1143 and MDA-MB-231,

14 see Fig. 1B) and breast cell lines derived from healthy tissues with strong or weak CDA

15 expression (MCF- 12A and 184B5, respectively, Supplementary Fig. S2B). No genetic

16 mutation likely to lead to CDA inactivation was identified (Supplementary Table 2A).

17 However, we found several SNPs that had previously been identified and listed in the Single

18 Nucleotide Polymorphism Database (dbSNP) (15,39–41) . These results are consistent with

19 the CDA gene sequencing results for the NCI-60 cell lines (Supplementary Table 2B) and

20 demonstrate that CDA is not inactivated through genetic alterations in cancer cells.

21 We then explored the possible role of epigenetic regulation of CDA gene expression. We

22 mapped the CpG methylation sites in the CDA gene. Using the NCI-60 methylation datasets,

23 we mapped the CpG methylation sites in the CDA gene, and identified 8 CpG methylation

24 sites, three in the body of the gene, two within 1500 bp of the transcription start site (TSS),

25 one in the TSS200, one in the 5’-UTR and one is the 3’-UTR (Supplementary Fig. S2C). We

1 then analyzed the levels of methylation of the CDA promoter, using the dataset for the

2 methylation of NCI60 cell lines (42). We calculated Pearson’s correlation coefficients for the

3 relationships between methylation at the various CpG methylation sites and CDA expression.

4 We found a highly significant negative correlation between CDA transcript and methylation

5 levels at CpG two sites, cg04087271 (TSS200) and cg00784581 (5’-UTR) (Pearson r= -

6 0.4184, P=0.009) (Fig. 2A), the six other exhibiting no significant negative correlation with

7 CDA expression. High levels of CDA methylation were found in 42% of the CDA-deficient

8 cell lines (19 of 45), such as MCF-7, and IGROV-1, and no methylation was detected in cell

9 lines overexpressing CDA, such as MDA-MB-231 and HOP-92, except for the LOXIMVI

10 melanoma cell line.

11 For the validation of these methylation data, we treated a set of cancer cell lines derived from

12 breast, lung, ovarian and melanoma tumors not expressing CDA (Fig. 1B and Supplementary

13 Fig. S2D) with the DNA methyltransferase activity inhibitor 5-Aza-2’-deoxcytidine (5-Aza-

14 dC), resulting in DNA demethylation (43). We found that 5-Aza-dC induced a strong increase

15 (up to 1000-fold induction) in CDA mRNA levels (Fig. 2B) without major toxicity

16 (Supplementary Fig. S2E) in the 7 CDA-deficient cell lines analyzed. By contrast, it had little

17 or no effect on CDA transcript levels in the MDA-MB-231, HOP-92, HCC-1143 and HCC-

18 1937 control cell lines, which have constitutively high levels of CDA (~two-fold induction).

19 It has been reported that the selection of CDA overexpression in response to prolonged drug

20 exposure is responsible for resistance to gemcitabine (18,44) and that the ectopic expression

21 of CDA in CDA-deficient cancer cells leads to a significant increase in resistance to

22 gemcitabine (18,45). We thus evaluated the functionality of the CDA protein produced after

23 5-Aza-dC treatment, by breast and ovarian cancer cells, HCC-1954 and IGROV-1,

24 respectively. The cells were left untreated or were subjected to pretreatment with 5-Aza-dC

25 for 96 hours and then to treatment with various concentrations of gemcitabine over a period of

1 72 hours. The induction of CDA protein production by 5-Aza-dC led to a significant increase

2 in gemcitabine resistance (Fig. 2C). Our data are consistent with 5-Aza-dC inducing the

3 expression of functional CDA protein in CDA-deficient cancer cells, leading to resistance to

4 gemcitabine.

5 We then analyzed in silico CDA promoter methylation levels (Supplementary Fig. S2C) on

6 TCGA samples for 16 different tissue cancers, using the CBioPortal for cancer genomics

7 (29,30) and the Broad Institute FireBrowse portal (27). We found a highly significant

8 correlation between CDA transcripts levels and CDA promoter methylation on two CpG sites

9 (cg04087271 and cg24502330). Methylation of the cg04087271 site was the only one

10 correlating with CDA deficiency in both tumor tissues and NCI-60 cell lines (Fig. 2D). In

11 both cancer cell lines and tumor tissues, CDA promoter methylation levels were significantly

12 higher in samples with low CDA transcript levels (Fig. 2D) than in those with high CDA

13 transcript levels. This correlation was not significant in some other cancer types, but we

14 nevertheless identified a subpopulation of these cancers with low CDA expression and high

15 CDA promoter methylation levels (Supplementary Fig. S2F). Thus, the association of

16 methylation status to gene expression is consistent with the hypothesis that promoter

17 methylation is a principal driver of CDA gene expression in cancer.

19 Loss of CDA expression in tumor cells defines a new tumor subgroup that could be

20 specifically targeted by chemotherapy

21 We previously reported that CDA deficiency in BS cells or CDA depletion in HeLa cells

22 leads to an increase in sister chromatid exchange (SCE) frequency, a hallmark of genomic

23 instability (4,5).

24 We therefore investigated whether constitutive CDA deficiency in tumor cells was also

25 associated with an increase in SCE frequency, by analyzing basal SCE levels in several cancer

1 cell lines derived from breast, lung and ovary tumors. SCE frequency was significantly higher

2 in the cancer cell lines not expressing CDA than in those expressing CDA (Fig. 3A).

3 Thus, tumors from the same classically defined groups may display differences in CDA

4 expression status resulting in contrasting cellular properties, such as SCE levels (e.g. CDA-

5 proficient HCC-1143 cells and CDA-deficient BT-20 cells are both classified as triple-

6 negative breast cancer cells). We thus propose the use of CDA expression status in tumor

7 cells to define two new subgroups: CDA-deficient tumors and CDA-proficient tumors. These

8 new subgroups may differ in their sensitivity to antitumor therapies. The targeting of CDA-

9 deficient tumor cells might therefore open up new possibilities for cancer therapy.

10 The CellMiner web tool (46) can be used to assess the correlation between gene expression

11 and drug sensitivity/resistance. We searched for drugs with antiproliferative activity

12 significantly correlated with CDA expression levels. We identified 277 such drugs, 94 of

13 which were more toxic to CDA-deficient cells and 183 of which were more active against

14 CDA-proficient cells (Supplementary Table 3). We tested our hypothesis that drugs that do

15 not affect CDA-proficient cells can specifically target CDA-deficient cells, by focusing on an

16 aminoflavone (AF) derivative (AFP464; NSC 710464) for which we found a highly

17 significant negative correlation (Pearson r=-0.379, P=0.0031) with CDA deficiency

18 (Supplementary Table 3). Twenty CDA-deficient cell lines of the 43 tested (46.5%), including

19 MCF-7 and IGROV-1, were sensitive to AF, whereas 13 of the 16 (81.25%) CDA-proficient

20 cell lines, including MDA-MB-231, were resistant to this drug (Fig. 3B).

21 Among the selected molecules presenting a significant negative correlation with CDA

22 expression, we choose AF because it has reached phase II clinical trials in the US for treating

23 solid tumors including breast cancer (NCT01015521, NCT01233947, NCT00369200,

24 NCT00348699) (47), and because the anti-tumor efficacy of AF had been validated in vivo by

1 three independent groups on MCF-7 and MDA-MB-468 xenografted breast cancer models

2 (48–50) . No effect of AF was observed on MDA-MB-231 xenografted breast cancer cells

3 (50). Together with our results showing that MCF7 and MDA-MB-468 breast cancer cell

4 lines are CDA-deficient, and MDA-MB-231 cells are CDA-proficient (Fig. 1A and 1B), these

5 data are consistent with our hypothesis that AF specifically targets CDA-deficient cancer cells

6 both in vitro and in vivo.

7 We thus evaluated the causality of the relationship between CDA downregulation and the anti-

8 proliferative activity of AF, by shRNA-mediated CDA depletion in HeLa cells

9 (Supplementary Fig. S3). We found that CDA depletion significantly increased sensitivity to

10 AF treatment (Fig. 3C). We then confirmed the cytotoxicity of AF in six breast and ovary

11 cancer cell lines, three of which were CDA-deficient (MCF-7, MDA-MB-468 and IGROV-1),

12 and the other three being CDA-proficient (MDA-MB-231, OVCAR-8 and SKOV-3). The

13 CDA-deficient cell lines were highly sensitive to AF, whereas the CDA-proficient cell lines

14 were resistant (Fig. 3D, left and right panels).

15 Our results show that CDA expression status can be used as a predictor of sensitivity to AF;

16 CDA deficiency is thus a potential new sensitive biomarker or target for anticancer therapies.

1 Discussion

2 We found that CDA expression was lost in a large proportion of cancer cells and tumor

3 tissues, and our findings identify CDA-deficient tumors as a new subgroup of cancers. SCE

4 frequency was significantly higher in cancer cell lines not expressing CDA than in those

5 expressing CDA. This increase in SCE frequency might reflect a nucleotide pool imbalance

6 due to CDA deficiency that slows down replication fork progression, leading to an increase in

7 replication-associated DNA breaks, that could account, at least in part, for SCE formation (4).

8 The loss of CDA expression is mostly due to DNA methylation and the treatment of CDA-

9 deficient cells with 5-Aza-dC was sufficient to restore the expression of a functional CDA.

10 This is the first study, to our knowledge, to reveal the extent of CDA inactivation and its

11 epigenetic control in cancer.

12 DNA methylation may be the predominant mechanism of CDA silencing, but it is clearly not

13 the only one, as some CDA-deficient cell lines present no CDA gene methylation. The other

14 mechanisms involved in regulating CDA gene expression merit further investigation but are

15 beyond the scope of this study.

16 CDA has already been shown to play a crucial role in the response of cancer cells to widely

17 used nucleoside analogs, such as cytosine arabinoside and gemcitabine, and the dose-limiting

18 toxicity of these drugs (8,51–54) . Our results suggest that IHC assessments of CDA levels

19 could be used to determine the CDA status of tumors, with potential implications for

20 treatment.

21 Oxidized and epigenetically modified cytidine nucleosides specifically target tumors

22 overexpressing CDA (19,20). We found that 5-Aza-dC treatment strongly induced the

23 expression of a functional CDA in CDA-deficient tumor cells, with little or no effect on CDA

24 expression in CDA-proficient cells. These findings suggest that DNA-demethylating agents

25 could be assessed as a possible treatment for CDA-deficient tumors, to induce CDA

1 overexpression and then sensitize these tumors to treatment with oxidized and epigenetically

2 modified cytidine nucleosides.

3 Finally, our results suggest that the targeting of CDA deficiency might offer new possibilities

4 for treatment. In silico screening with the NCI CellMiner analysis tool identified AF as a

5 proof-of-principle candidate for the targeting of CDA-deficient tumor cells. AF acts mainly

6 by activating aryl hydrocarbon receptor (AhR), which in turns induces the expression of

7 cytochrome P450 CYP1A1/1A2/1B1 that metabolizes AF to toxic products, inducing

8 replication stress, DNA damage and apoptosis via a p53/p21mediated pathway (48,55–58).

9 We found that AF was specifically effective in CDA-deficient tumor cells, and that this drug

10 had no effect on CDA-proficient cells, in agreement with the literature (48–50).

11 Thus, the subgroup of tumors not expressing CDA could be specifically targeted by such

12 treatment, and CDA expression status could be used as a new marker to guide anticancer

13 therapy. Molecules not yet shown to be active against this tumor subgroup will probably be

14 discovered through the systematic screening of CDA-proficient and –deficient cells.

15 In conclusion, although additional human trials are needed, our results constitute a proof-of-

16 concept that CDA deficiency may turn out to be a new predictive marker of susceptibility to

17 antitumor drugs that could be used as a new target for anticancer therapies, thus opening up

18 new possibilities for the treatment of cancers.

1 Disclosure of Potential Conflicts of Interest

2 No potential conflicts of interest were disclosed.

4 Authors’ contributions

5 Conception and design: H. Mameri, I. Bièche, Y. Pommier, M. Amor-Guéret

6 Development of methodology: H. Mameri, I. Bièche, D. Meseure, E. Marangoni, Y.

7 Pommier, M. Amor-Guéret

8 Acquisition of data: H. Mameri, E. Marangoni, G. Buhagiar-Labarchède, A. Nicolas, S.

9 Vacher, R. Onclercq-Delic, V. Rajapakse, S. Varma, W. Reinhold

10 Analysis and interpretation of data: H. Mameri, I. Bièche, D. Meseure, E. Marangoni, G.,

11 G. Buhagiar-Labarchède, A. Nicolas, S. Vacher, R. Onclercq-Delic, V. Rajapakse, S. Varma,

12 W. Reinhold, Y. Pommier, M. Amor-Guéret

13 Writing, review and/or revision of the manuscript: H. Mameri, I. Bièche, D. Meseure, E.

14 Marangoni, Y. Pommier, M. Amor-Guéret

15 Administrative, technical or material support (i.e., reporting or organizing data,

16 constructing databases): G. Buhagiar-Labarchède, A. Nicolas, S. Vacher, R. Onclercq-Delic,

17 V. Rajapakse, S. Varma, W. Reinhold

18 Study supervision: M Amor-Guéret

20 Acknowledgments

21 We thank D. Gentien, E. Henry, B. Albaud, F. Reyal, V. Maire, B. Marty-Provost, T. Dubois,

22 S. Roman-Roman (Translational Research Department of the Institut Curie) F. Radvanyi

23 (UMR144CNRS/Institut Curie) and O. Delattre (INSERM U380, Institut Curie), for

24 providing us with the complete study of the breast cancer cell lines and the TCGA network

25 for making available the high throughput data. We thank S. Vagner for the melanoma cell

1 lines and for critical reading of the manuscript. We also thank M. Dutreix for stimulating

2 discussions.

4 Grant support

5 This work was supported by the Curie Institute (Translational cancer research grant 2014 and

6 PICSysBio), CNRS, Cancéropôle Ile de France, Ligue Nationale contre le Cancer (Comité

7 de l’Essonne), the Association pour la Recherche sur le Cancer, the Agence Nationale de la

8 Recherche, and the Center for Cancer Research (BC 006161), the Intramural Program of the

9 US National Cancer Institute, National Institutes of Health. HM was supported by a

10 postdoctoral fellowship from Cancéropôle Ile de France and Association pour la Recherche

11 sur le Cancer.

1 References

2 1. Chen Z, Shi T, Zhang L, Zhu P, Deng M, Huang C, et al. Mammalian drug efflux 3 transporters of the ATP binding cassette (ABC) family in multidrug resistance: A review 4 of the past decade. Cancer Lett. 2016;370:153–64.

5 2. Conticello SG, Thomas CJF, Petersen-Mahrt SK, Neuberger MS. Evolution of the 6 AID/APOBEC family of polynucleotide (deoxy)cytidine deaminases. Mol Biol Evol. 7 2005;22:367–77.

8 3. Conticello SG. The AID/APOBEC family of nucleic acid mutators. Genome Biol. 9 2008;9:229.

10 4. Chabosseau P, Buhagiar-Labarchède G, Onclercq-Delic R, Lambert S, Debatisse M, 11 Brison O, et al. Pyrimidine pool imbalance induced by BLM helicase deficiency 12 contributes to genetic instability in Bloom syndrome. Nat Commun. 2011;2:368.

13 5. Gemble S, Ahuja A, Buhagiar-Labarchède G, Onclercq-Delic R, Dairou J, Biard DSF, et 14 al. Pyrimidine Pool Disequilibrium Induced by a Cytidine Deaminase Deficiency 15 Inhibits PARP-1 Activity, Leading to the Under Replication of DNA. PLoS Genet. 16 2015;11:e1005384.

17 6. Demontis S, Terao M, Brivio M, Zanotta S, Bruschi M, Garattini E. Isolation and 18 characterization of the gene coding for human cytidine deaminase. Biochim Biophys 19 Acta. 1998;1443:323–33.

20 7. Cacciamani T, Vita A, Cristalli G, Vincenzetti S, Natalini P, Ruggieri S, et al. 21 Purification of human cytidine deaminase: molecular and enzymatic characterization and 22 inhibition by synthetic pyrimidine analogs. Arch Biochem Biophys. 1991;290:285–92.

23 8. Neff T, Blau CA. Forced expression of cytidine deaminase confers resistance to cytosine 24 arabinoside and gemcitabine. Exp Hematol. 1996;24:1340–6.

25 9. Weizman N, Krelin Y, Shabtay-Orbach A, Amit M, Binenbaum Y, Wong RJ, et al. 26 Macrophages mediate gemcitabine resistance of pancreatic adenocarcinoma by 27 upregulating cytidine deaminase. Oncogene. 2014;33:3812–9.

28 10. Ciccolini J, Dahan L, André N, Evrard A, Duluc M, Blesius A, et al. Cytidine deaminase 29 residual activity in serum is a predictive marker of early severe toxicities in adults after 30 gemcitabine-based chemotherapies. J Clin Oncol Off J Am Soc Clin Oncol. 31 2010;28:160–5.

32 11. Sugiyama E, Kaniwa N, Kim S-R, Kikura-Hanajiri R, Hasegawa R, Maekawa K, et al. 33 Pharmacokinetics of gemcitabine in Japanese cancer patients: the impact of a cytidine 34 deaminase polymorphism. J Clin Oncol Off J Am Soc Clin Oncol. 2007;25:32–42.

35 12. Yue L, Saikawa Y, Ota K, Tanaka M, Nishimura R, Uehara T, et al. A functional single- 36 nucleotide polymorphism in the human cytidine deaminase gene contributing to ara-C 37 sensitivity. Pharmacogenetics. 2003;13:29–38.

1 13. Farrell JJ, Bae K, Wong J, Guha C, Dicker AP, Elsaleh H. Cytidine deaminase single- 2 nucleotide polymorphism is predictive of toxicity from gemcitabine in patients with 3 pancreatic cancer: RTOG 9704. Pharmacogenomics J. 2012;12:395–403.

4 14. Monzo M, Brunet S, Urbano-Ispizua A, Navarro A, Perea G, Esteve J, et al. Genomic 5 polymorphisms provide prognostic information in intermediate-risk acute myeloblastic 6 leukemia. Blood. 2006;107:4871–9.

7 15. Fitzgerald SM, Goyal RK, Osborne WRA, Roy JD, Wilson JW, Ferrell RE. 8 Identification of functional single nucleotide polymorphism haplotypes in the cytidine 9 deaminase promoter. Hum Genet. 2006;119:276–83.

10 16. Okazaki T, Javle M, Tanaka M, Abbruzzese JL, Li D. Single nucleotide polymorphisms 11 of gemcitabine metabolic genes and pancreatic cancer survival and drug toxicity. Clin 12 Cancer Res Off J Am Assoc Cancer Res. 2010;16:320–9.

13 17. Mahlknecht U, Dransfeld C-L, Bulut N, Kramer M, Thiede C, Ehninger G, et al. SNP 14 analyses in cytarabine metabolizing enzymes in AML patients and their impact on 15 treatment response and patient survival: identification of CDA SNP C-451T as an 16 independent prognostic parameter for survival. Leukemia. 2009;23:1929–32.

17 18. Ye F-G, Song C-G, Cao Z-G, Xia C, Chen D-N, Chen L, et al. Cytidine Deaminase Axis 18 Modulated by miR-484 Differentially Regulates Cell Proliferation and Chemoresistance 19 in Breast Cancer. Cancer Res. 2015;75:1504–15.

20 19. Zauri M, Berridge G, Thézénas M-L, Pugh KM, Goldin R, Kessler BM, et al. CDA 21 directs metabolism of epigenetic nucleosides revealing a therapeutic window in cancer. 22 Nature. 2015;524:114–8.

23 20. Helleday T. Poisoning Cancer Cells with Oxidized Nucleosides. N Engl J Med. 24 2015;373:1570–1.

25 21. Lorenzi PL, Reinhold WC, Varma S, Hutchinson AA, Pommier Y, Chanock SJ, et al. 26 DNA fingerprinting of the NCI-60 cell line panel. Mol Cancer Ther. 2009;8:713–24.

27 22. CellMiner - Cell Line Information [Internet]. [cited 2016 Jul 21]. Available from: 28 https://discover.nci.nih.gov/cellminer/celllineMetadata.do

29 23. Baldeyron C, Brisson A, Tesson B, Némati F, Koundrioukoff S, Saliba E, et al. TIPIN 30 depletion leads to apoptosis in breast cancer cells. Mol Oncol. 2015;9:1580–98.

31 24. Marangoni E, Vincent-Salomon A, Auger N, Degeorges A, Assayag F, Cremoux P de, et 32 al. A New Model of Patient Tumor-Derived Breast Cancer Xenografts for Preclinical 33 Assays. Clin Cancer Res. 2007;13:3989–98.

34 25. Nogales V, Reinhold WC, Varma S, Martinez-Cardus A, Moutinho C, Moran S, et al. 35 Epigenetic inactivation of the putative DNA/RNA helicase SLFN11 in human cancer 36 confers resistance to platinum drugs. Oncotarget. 2016;7:3084–97.

37 26. The Cancer Genome Atlas Home Page [Internet]. Cancer Genome Atlas - Natl. Cancer 38 Inst. [cited 2016 Jul 25]. Available from: http://cancergenome.nih.gov.gate1.inist.fr/

1 27. Broad Institute TCGA Genome Data Analysis Center. Analysis-ready standardized 2 TCGA data from Broad GDAC Firehose stddata__2015_11_01 run [Internet]. Broad 3 Institute of MIT and Harvard; 2016 [cited 2016 Jun 23]. Available from: 4 http://dx.doi.org/10.7908/C1571BB1

5 28. FireBrowse [Internet]. [cited 2016 Jul 25]. Available from: http://firebrowse.org/

6 29. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, et al. The cBio cancer 7 genomics portal: an open platform for exploring multidimensional cancer genomics data. 8 Cancer Discov. 2012;2:401–4.

9 30. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, et al. Integrative 10 analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci 11 Signal. 2013;6:pl1.

12 31. cBioPortal for Cancer Genomics [Internet]. [cited 2016 Jul 25]. Available from: 13 http://www.cbioportal.org/

14 32. Gentien D, Henry E, Maire V, Marty-Provost B, Albaud B, Radvanyi F, et al. Breast 15 cancer cell lines: complete study [Internet]. [cited 2016 Jul 25]. Available from: 16 http://microarrays.curie.fr/publications/recherche_translationnelle/plateforme_genomiqu 17 e/

18 33. CellMiner - Analysis Tools [Internet]. [cited 2016 Jul 25]. Available from: 19 https://discover.nci.nih.gov/cellminer/analysis.do

20 34. Broad-Novartis Cancer Cell Line Encyclopedia [Internet]. [cited 2016 Jul 25]. Available 21 from: http://portals.broadinstitute.org/ccle/home

22 35. Shin G, Kang T-W, Yang S, Baek S-J, Jeong Y-S, Kim S-Y. GENT: gene expression 23 database of normal and tumor tissues. Cancer Inform. 2011;10:149–57.

24 36. GENT : Gene Expression across Normal and Tumor tissue [Internet]. [cited 2016 Jul 25 25]. Available from: http://medical-genome.kribb.re.kr/GENT/

26 37. Home - GEO - NCBI [Internet]. [cited 2016 Jul 25]. Available from: 27 http://www.ncbi.nlm.nih.gov.gate1.inist.fr/geo/

28 38. den Boon JA, Pyeon D, Wang SS, Horswill M, Schiffman M, Sherman M, et al. 29 Molecular transitions from papillomavirus infection to cervical precancer and cancer: 30 Role of stromal estrogen receptor signaling. Proc Natl Acad Sci U S A. 31 2015;112:E3255-3264.

32 39. Gilbert JA, Salavaggione OE, Ji Y, Pelleymounter LL, Eckloff BW, Wieben ED, et al. 33 Gemcitabine pharmacogenomics: cytidine deaminase and deoxycytidylate deaminase 34 gene resequencing and functional genomics. Clin Cancer Res Off J Am Assoc Cancer 35 Res. 2006;12:1794–803.

36 40. Abraham A, Varatharajan S, Abbas S, Zhang W, Shaji RV, Ahmed R, et al. Cytidine 37 deaminase genetic variants influence RNA expression and cytarabine cytotoxicity in 38 acute myeloid leukemia. Pharmacogenomics. 2012;13:269–82.

1 41. Sugiyama E, Kaniwa N, Kim S-R, Kikura-Hanajiri R, Hasegawa R, Maekawa K, et al. 2 Pharmacokinetics of gemcitabine in Japanese cancer patients: the impact of a cytidine 3 deaminase polymorphism. J Clin Oncol Off J Am Soc Clin Oncol. 2007;25:32–42.

4 42. Nogales V, Reinhold WC, Varma S, Martinez-Cardus A, Moutinho C, Moran S, et al. 5 Epigenetic inactivation of the putative DNA/RNA helicase SLFN11 in human cancer 6 confers resistance to platinum drugs. Oncotarget. 2015;

7 43. Christman JK. 5-Azacytidine and 5-aza-2’-deoxycytidine as inhibitors of DNA 8 methylation: mechanistic studies and their implications for cancer therapy. Oncogene. 9 2002;21:5483–95.

10 44. Samulitis BK, Pond KW, Pond E, Cress AE, Patel H, Wisner L, et al. Gemcitabine 11 resistant pancreatic cancer cell lines acquire an invasive phenotype with collateral 12 hypersensitivity to histone deacetylase inhibitors. Cancer Biol Ther. 2015;16:43–51.

13 45. Yoshida T, Endo Y, Obata T, Kosugi Y, Sakamoto K, Sasaki T. Influence of cytidine 14 deaminase on antitumor activity of 2’-deoxycytidine analogs in vitro and in vivo. Drug 15 Metab Dispos Biol Fate Chem. 2010;38:1814–9.

16 46. Reinhold WC, Sunshine M, Varma S, Doroshow JH, Pommier Y. Using CellMiner 1.6 17 for Systems Pharmacology and Genomic Analysis of the NCI-60. Clin Cancer Res Off J 18 Am Assoc Cancer Res. 2015;21:3841–52.

19 47. Search of: Aminoflavone - List Results - ClinicalTrials.gov [Internet]. [cited 2016 Jul 20 21]. Available from: 21 https://clinicaltrials.gov/ct2/results?term=Aminoflavone&Search=Search

22 48. Loaiza-Pérez AI, Kenney S, Boswell J, Hollingshead M, Alley MC, Hose C, et al. Aryl 23 hydrocarbon receptor activation of an antitumor aminoflavone: basis of selective toxicity 24 for MCF-7 breast tumor cells. Mol Cancer Ther. 2004;3:715–25.

25 49. Terzuoli E, Puppo M, Rapisarda A, Uranchimeg B, Cao L, Burger AM, et al. 26 Aminoflavone, a ligand of the aryl hydrocarbon receptor, inhibits HIF-1alpha expression 27 in an AhR-independent fashion. Cancer Res. 2010;70:6837–48.

28 50. Stark K, Burger A, Wu J, Shelton P, Polin L, Li J. Reactivation of estrogen receptor α by 29 vorinostat sensitizes mesenchymal-like triple-negative breast cancer to aminoflavone, a 30 ligand of the aryl hydrocarbon receptor. PloS One. 2013;8:e74525.

31 51. Bhatla D, Gerbing RB, Alonzo TA, Conner H, Ross JA, Meshinchi S, et al. Cytidine 32 deaminase genotype and toxicity of cytosine arabinoside therapy in children with acute 33 myeloid leukemia. Br J Haematol. 2009;144:388–94.

34 52. Mahfouz RZ, Jankowska A, Ebrahem Q, Gu X, Visconte V, Tabarroki A, et al. 35 Increased CDA expression/activity in males contributes to decreased cytidine analog 36 half-life and likely contributes to worse outcomes with 5-azacytidine or decitabine 37 therapy. Clin Cancer Res Off J Am Assoc Cancer Res. 2013;19:938–48.

38 53. Benhura MA. Metabolism of cytosine arabinoside in Tetrahymena pyriformis. Comp 39 Biochem Physiol B. 1985;80:821–5.

1 54. Serdjebi C, Milano G, Ciccolini J. Role of cytidine deaminase in toxicity and efficacy of 2 nucleosidic analogs. Expert Opin Drug Metab Toxicol. 2015;11:665–72.

3 55. Meng L, Kohlhagen G, Liao Z, Antony S, Sausville E, Pommier Y. DNA-protein cross- 4 links and replication-dependent histone H2AX phosphorylation induced by 5 aminoflavone (NSC 686288), a novel anticancer agent active against human breast 6 cancer cells. Cancer Res. 2005;65:5337–43.

7 56. Meng L, Shankavaram U, Chen C, Agama K, Fu H, Gonzalez FJ, et al. Activation of 8 aminoflavone (NSC 686288) by a sulfotransferase is required for the antiproliferative 9 effect of the drug and for induction of histone gamma-H2AX. Cancer Res. 10 2006;66:9656–64.

11 57. Meng L-H, Kohn KW, Pommier Y. Dose-response transition from cell cycle arrest to 12 apoptosis with selective degradation of Mdm2 and p21WAF1/CIP1 in response to the 13 novel anticancer agent, aminoflavone (NSC 686,288). Oncogene. 2007;26:4806–16.

14 58. McLean L, Soto U, Agama K, Francis J, Jimenez R, Pommier Y, et al. Aminoflavone 15 induces oxidative DNA damage and reactive oxidative species-mediated apoptosis in 16 breast cancer cells. Int J Cancer J Int Cancer. 2008;122:1665–74.

17 59. Nextbio [Internet]. [cited 2016 Jul 25]. Available from: 18 https://www.nextbio.com/b/authentication/login.nb

19 60. Oncomine [Internet]. [cited 2016 Jul 25]. Available from: https://www.oncomine.org

1 Figure legends

2 Figure 1: CDA expression levels in cancer cell lines and tissues. A, Transcriptomic data

3 sets in log 2 values for Curie Institute breast cancer cell lines (n=34) publicly available (see

4 Materials and Methods section) (left panel) and from NCI 60 cancer cell lines data miner (46)

5 (right panel). Mean and median values are shown as dashed and solid lines, respectively. B,

6 Real-time RT-qPCR and western blot analysis of CDA expression in a set of 26 cell lines for

7 three different cancers (breast, non-small lung and ovary) representative of the Curie Institute

8 (left panel) and NCI60 (right panel) panels. Hsp90 and β-actin were used as loading controls

9 for western blotting; TBP and GAPDH were used for RT-qPCR data normalization. Western

10 blotting and RT-qPCR data were reproduced at least twice C, Real-time RT-qPCR

11 quantification of human (Hm in dark) and mouse (Mm in gray) CDA transcripts relative to

12 human and mouse TBP housekeeping gene transcripts in 66 breast-derived xenografts. D,

13 IHC analysis of CDA expression in six different tumor tissues (n=50 per tissue).

14 Representative images for each tissue (upper panels) and quantitative representations of CDA

15 protein levels in each tissue (lower panels) are shown. The results are presented as

16 percentages of tumors expressing low (black) and high (gray) levels of CDA on the basis of

17 the scores obtained (Low: scores 0-1 and High: scores 2-3). For all images, the scale bar is 50

18 µm. The percentage of low-CDA cancer tissues is indicated in the black bars E, Scatter dot

19 plot with mean ± SD for transcriptomic data for CDA transcripts, comparing unmatched

20 normal and tumor tissues for the liver (GSE14520), esophagus (GSE13898), cervix

21 (GSE9750) and colon (GSE9348). The data were retrieved from the Nextbio (59) and

22 Oncomine (60) data sources and downloaded from GEO and presented as log 2 intensities. F,

23 CDA transcript levels relative to TBP, as quantified by RT-qPCR in a mini-cohort of

24 cancerous and non-cancerous colon tissues. Error bars indicate the SD. The P values

25 calculated in unpaired two-tailed t-tests are considered statistically significant if <0.05.

2 Figure 2: Silencing of CDA gene expression by DNA methylation. A, Mean-centered CDA

3 transcript intensity data for the NCI60 panel of cell lines obtained with the NCI-Cell Miner

4 analysis tool (upper left panel), and mean CDA gene methylation levels in the data for the

5 NCI60 panel of cell lines extracted from GEO under accession number GSE66872 (lower left

6 panel), and representation of the correlation between CDA transcript intensity and CDA

7 promoter methylation for the cg04087271 and cg00784581 probes (Pearson correlation) (right

8 panel). B, RT-qPCR analysis of the induction of CDA expression relative to GAPDH in cells

9 initially with and without CDA expression, after 96 hours of treatment with 2.5 µM 5-Aza-

10 dC. Error bars represent means ± SD for at least 3 independent experiments. The P values

11 were calculated in paired t-test. All P values <0.05 were considered statistically significant.

12 C, Left panels: RT-qPCR analysis (upper panel) and western blot analysis (lower panels) of

13 the induction of CDA expression in HCC-1954 and IGROV-1 cell lines left untreated (white

14 bars) or treated with 1 µM 5-Aza-dC for 96h (black bars). Survival curves of the HCC-1954

15 (n=3; middle panel) and IGROV-1 (n=5; right panel) cell lines left untreated (control, blue

16 curve) or subjected to pretreatment for 96 h with 1 μM 5-Aza-dC (5-Aza-dC, red curve), and

17 then treated with various doses of gemcitabine for a further 72 hours. Cell viability was

18 assessed in the MTT assay. Error bars represent means ± SD for 3 or 5 experiments. The P

19 values were calculated in paired t-tests. P values <0.05 were considered statistically

20 significant. D, Scatter plots showing the Pearson correlation between mRNA seq data for

21 CDA expression and CDA CpG cg04087271 methylation for 10 different cancer samples

22 from TCGA. The data are publicly available and were retrieved from the Broad Institute

23 FireBrowse portal (27) and the CBioPortal for Cancer Genomics database (29,30). Mean CDA

24 expression is indicated by dashed vertical lines and mean methylation level is indicated by

25 dashed horizontal lines. All P values <0.05 were considered statistically significant.

2 Figure 3: Drug sensitivity of CDA-deficient cells. A, SCE frequency in CDA-deficient and

3 CDA-proficient cells (left panel) and representation of SCE frequency in cells classified on

4 the basis of their CDA expression status, low or high (right panel). P values were calculated

5 in Mann-Whitney tests for at least 3 independent experiments. P<0.05 was considered

6 statistically significant B, Scatterplot showing a significant negative correlation between

7 aminoflavone cytotoxicity and CDA expression (Pearson correlation) in the NCI60 panel of

8 cell lines. The colors indicate the origin of the cancer tissue. C, Isogenic HeLa cell lines

9 (HeLa control cells in gray and CDA-depleted HeLa cells in black) were treated for 72 hours

10 with the indicated concentrations of aminoflavone, and the percentage of cells surviving is

11 shown. D, Breast (MCF-7, MDA-MB-468 and MDA-MB-231) and ovarian (SKOV-3,

12 OVCAR-8 and IGROV-1) cancer cell lines were treated for 72 hours with the indicated

13 concentrations of aminoflavone. Survival curves of cell lines with low levels of CDA

14 expression are represented in black and the survival curves of cell lines with high levels of

15 CDA expression are shown in gray. For C and D, cell viability was assessed in MTT assays.

16 The error bars represent means ± SD for three independent experiments. P values <0.05 were

17 considered statistically significant.

Cytidine deaminase deficiency reveals new therapeutic opportunities against cancer

Hamza Mameri, Ivan Bieche, Dider Meseure, et al.

Clin Cancer Res Published OnlineFirst September 6, 2016.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-16-0626

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2016/09/03/1078-0432.CCR-16-0626.DC1

Author Author manuscripts have been peer reviewed and accepted for publication but have not yet been Manuscript edited.

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 Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/early/2016/09/03/1078-0432.CCR-16-0626. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.