USE OF CASD1 AS A BIOMARKER OF A CANCER EXPRESSING THE O-ACETYLATED-GD2 GANGLIOSIDE

Abstract

The use of CASD1 as a biomarker of a cancer expressing the O-acetylated-GD2 ganglioside. Also, a method of diagnosing a cancer expressing the O-acetylated-GD2 ganglioside, a method of selecting a subject suffering from a cancer expressing the O-acetylated-GD2 ganglioside for treatment targeting the cancer, or a method of monitoring the response of a subject suffering from a cancer expressing the O-acetylated-GD2 ganglioside to a treatment targeting the cancer.

Claims

1-14. (canceled)

15. An in vitro method of selecting a subject suffering from a cancer expressing the O-acetylated-GD2 ganglioside for treatment targeting said cancer, said method comprising the steps of: a1) measuring CASD1 expression level as a biomarker in a biological sample of the subject; a1) optionally, measuring the expression level of at least one biomarker selected from the group consisting of GD2S, GD3S, OAcGD2, GD2, CERK, PIK3C2A, PDK3, MERTK and NME3, in a biological sample of the subject; and b) based on the level measured at step a1), and optionally at step a1), selecting said subject to undergo treatment targeting said cancer expressing the O-acetylated-GD2 ganglioside.

16. An in vitro method of monitoring the response of a subject suffering from a cancer expressing the O-acetylated-GD2 ganglioside to a treatment targeting said cancer, said method comprising the steps of: a1) measuring CASD1 expression level as a biomarker in a biological sample of the subject; a1) optionally, measuring the expression level of at least one biomarker selected from the group consisting of GD2S, GD3S, OAcGD2, GD2, CERK, PIK3C2A, PDK3, MERTK and NME3, in a biological sample of the subject; and b) based on the level measured at step a1), and optionally at step a1), monitoring the response of said subject to said treatment.

17. An in vitro method of diagnosing a cancer expressing the O-acetylated-GD2 ganglioside in a subject, said method comprising the steps of: a1) measuring CASD1 expression level as a biomarker in a biological sample of the subject; a1) optionally, measuring the expression level of at least one biomarker selected from the group consisting of GD2S, GD3S, OAcGD2, GD2, CERK, PIK3C2A, PDK3, MERTK and NME3, in a biological sample of the subject; and b) based on the level measured at step a1), and optionally at step a1), diagnosing a cancer expressing the O-acetylated-GD2 ganglioside in said subject.

18. The in vitro method according to claim 15, further comprising a step a2) of comparing the expression level measured at step a1), and/or optionally at step a1), with a threshold value.

19. The in vitro method according to claim 18, wherein the subject is selected to undergo a treatment targeting said cancer expressing the O-acetylated-GD2 ganglioside, or is diagnosed as suffering from a cancer expressing the O-acetylated-GD2 ganglioside, if the expression level measured at step a1), and/or optionally at step a1), is higher than the threshold value.

20. The in vitro method according to claim 15, wherein the biological sample is selected from the group consisting of a blood sample, a serum sample, a plasma sample, a urine sample, a tissue sample from a biopsy and a cell sample from a biopsy.

21. The in vitro method according to claim 15, wherein the expression level measured at step a1), and/or optionally at step a1), is measured at the DNA or RNA level, preferably by RT-PCR, RT-qPCR, Northern Blot, hybridization techniques, microarrays or sequencing.

22. The in vitro method according to claim 15, wherein the expression level measured at step a1), and/or optionally at step a1), is measured at the protein level, preferably by FACS, immunohistochemistry, mass spectrometry, western blot associated with cell fractionation, enzyme-linked immunosorbent assay (ELISA), sandwich ELISA, fluorescent-linked immunosorbent assay (FLISA), enzyme immunoassay (EIA), radioimmunoassay (RIA) or image analysis.

23. The in vitro method according to claim 15, wherein said treatment comprises an antibody that binds to the O-acetylated-GD2 ganglioside.

24. The in vitro method according to claim 15, wherein said cancer expressing the O-acetylated-GD2 ganglioside is characterized by the presence of cells expressing the O-acetylated-GD2 ganglioside at their cell surface in the subject.

25. The in vitro method according to claim 15, wherein said cancer expressing the O-acetylated-GD2 ganglioside is selected from the group consisting of neuroblastoma, glioma (including glioblastoma), retinoblastoma, Ewing's family of tumors, sarcoma (including rhabdomyosarcoma, osteosarcoma, leiomyosarcoma, liposarcoma, and fibrosarcoma), lung cancer (including small cell lung cancer), breast cancer, melanoma (including uveal melanoma), metastatic renal carcinoma, head and neck cancer, hematological cancers (including leukemia, Hodgkin lymphoma, non-Hodgkin lymphoma and myeloma), colorectal cancer, pancreatic cancer, prostate cancer, liver cancer, bladder cancer, gastric/stomach cancer, cervical cancer, endometrial cancer, neuroendocrine cancer, esophageal cancer, ovarian cancer, skin cancer, kidney cancer, soft tissue sarcoma, adrenal cancer, testicular cancer, thymic cancer (including thymoma) and thyroid cancer.

26. A method for treating a cancer expressing the O-acetylated-GD2 ganglioside in a subject, comprising administering to a subject in need thereof a therapeutical effective amount of an inhibitor of CASD1.

27. A method for treating a cancer expressing the O-acetylated-GD2 ganglioside in a subject, comprising administering to a subject in need thereof a therapeutical effective amount of an inhibitor of a biomarker selected from the group consisting of CERK, PIK3C2A, PDK3, MERTK, NME3 and EZH2, in combination with a therapy targeting the O-acetylated-GD2 ganglioside.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0365] FIG. 1 corresponds to photographs of Western-blot showing that CASD1 induces the O-acetylation of GD2 and GD3 in CHO cells. Total gangliosides were extracted from CHO-WT or CHO?Casd1 cells transfected with empty vector (mock) or a plasmid encoding the indicated synthases. Gangliosides were separated by thin-layer chromatography and stained with the indicated antibodies. Pure gangliosides and their in vitro generated 9-O-acetylated forms were used as standards (left panel). Please note that the OAcGD2 standard contains residual amounts of GD2. FIG. 1A: CASD1-dependent formation of 9-OAcGD2. FIG. 1B: CASD1-dependent formation of 9-O-Ac-GD3.

[0366] FIG. 2 is a graph showing CASD1 expression in neuro-ectoderm derived cancer cells. CASD1 mRNA expression was determined by qPCR in breast cancer cell lines. SK-MEL-28 melanoma cell line and LAN-1 neuroblastoma cell line were used as controls. Results were normalized to the expression of HPRT (hypoxanthine phosphoribosyl transferase) mRNA. Each bar represents the mean?SD of n=3 experiments.

[0367] FIG. 3 is a set of graphs showing the reduced OAcGD2 expression in SUM159PT cells depleted for CASD1 expression using siRNA strategy. qPCR quantification of GD2S (FIG. 3A) and CASD1 (FIG. 3B) expression in transiently transfected and control SUM159PT cells (n=3). Results were normalized to the expression of HPRT mRNA. Quantification of the mean fluorescence intensity of GD2 (FIG. 3C) and OAcGD2 (FIG. 3D) expression using immunocytochemistry and confocal microscopy in SUM159PT cells (n=3). Statistical difference using unpaired t-test: *p<0.5; ****p<0.0001; ns: not significant.

[0368] FIG. 4 is a set of graphs showing increased OAcGD2 expression in CASD1 overexpressing SUM159PT cells using plasmid transfection (CADS1+). qPCR quantification of GD2S (FIG. 4A) and CASD1 (FIG. 4B) genes in transiently transfected SUM159PT cells (n=3). Results were normalized to the expression of HPRT mRNA. Quantification of mean fluorescence intensity of GD2 (FIG. 4C) and OAcGD2 (FIG. 4D) expression by immunochemistry and confocal microscopy in breast cancer cells (n=3). Statistical difference using unpaired t-test: **** p<0.0001; ns: non-significant.

[0369] FIG. 5 is a set of graphs showing expression of CASD1 mRNA and quantification of OAcGD2 and GD2 expression in SUM159PT CASD1+clones. FIG. 5A is a graph showing RT-qPCR quantification of CASD1 gene expression in stably transfected and control SUM159PT cells (n=3). Results were normalized to HPRT mRNA expression. FIG. 5B is a graph representing the quantification of mean fluorescence intensity of GD2. Statistical difference using unpaired t-test: **** p<0.0001. FIG. 5C is a graph representing the quantification of mean fluorescence intensity of OAcGD2. Statistical difference using unpaired t-test: **** p<0.0001.

[0370] FIG. 6 is a set of graphs demonstrating the biological properties of SUM159PT CASD1+clones. The growth of control and SUM159PT CASD1+#19 and #26 clones was assessed after 0 h, 24 h, 48 h, 72 h and 96 h using MTS reagent (Promega) in media containing 5% (FIG. 6A), 1% (FIG. 6B) or 0% (FIG. 6C) of fetal calf serum (FCS). The migration (FIG. 6D) and invasion (FIG. 6E) capabilities of control and SUM159PT CASD1+clones #19 and #26 were assessed after 48 h by Transwell assay in serum free media. Statistical difference using one-way anova: **** p<0.0001; ** p<0.002; * p<0.02.

[0371] FIG. 7 is a set of chemical structures representing GD2 and two illustrative examples of OAcGD2 gangliosides. FIG. 7A represents the chemical structure of GD2 (Neu5Ac?2-8Neu5Ac?2-3[GalNAc?1-4]LacCer). FIG. 7B represents the chemical structure of a first example of a 9-O-acetylated GD2 isomer (Neu5,9Ac.sub.2?2-8Neu5Ac?2-3[GalNAc?1-4]LacCer). FIG. 7C represents the chemical structure of a second example of a 9-O-acetylated GD2 isomer (Neu5Ac?2-8Neu5,9Ac.sub.2?2-3[GalNAc?1-4]LacCer). FIG. 7B and 7C are examples of possible GD2 O-acetylation isomers. However, GD2 may also be O-acetylated at other positions within the molecule.

[0372] FIG. 8A-O is a set of forest plots showing the hazard ratio and 95% confidence intervals in patients having high and low expression levels of CASD1, CERK, PIK3C2A, B4GALTN1, ST8SIA1 genes and combinations thereof, in TCGA cohorts. Hazard ratio was calculated in populations computationally identified as having high or low expression of the gene of interest, based on individual signature expression in TCGA datasets by SurvExpress and tcgasurvival optimized algorithm. The datasets analyzed were Sarcoma (SARC); Pheochromocytoma and Paraganglioma (PCPG); Uterine Corpus Endometrial Carcinoma (UCEC); Thyroid carcinoma (THCA); Thymoma (THYM); Testicular Germ Cell Tumors (TGCT); Stomach adenocarcinoma (STAD); Skin Cutaneous Melanoma (SKCM); Prostate adenocarcinoma (PRAD); Pancreatic adenocarcinoma (PAAD); Ovarian serous cystadenocarcinoma (OV); Lung squamous cell carcinoma (LUSC); Lung adenocarcinoma (LUAD); Liver hepatocellular carcinoma (LIHC); Kidney PAN cancer (KIPAN); Acute Myeloid Leukemia (LAML); Head and Neck squamous cell carcinoma (HNSC); Uveal Melanoma (UVM); Esophageal carcinoma (ESCA); Colon and Rectum adenocarcinoma (COADREAD); Cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC); Breast invasive carcinoma (BRCA); Gliomas (GBM and LGG); Bladder Urothelial Carcinoma (BLCA); Cholangiocarcinoma (CHOL); Adrenocortical carcinoma (ACC). P-value is indicated on the graph (*** p<0.001, ** p<0.01, * p<0.05). Statistically significant hazard ratios are indicated in black, while non-significant hazard ratios are in dark grey. FIG. 8A represents B4GALNT1, FIG. 8B represents ST8SIA, FIG. 8C represents CASD1, FIG. 8D represents CASD1/B4GALNT1, FIG. 8E represents CASD1/ST8SIA, FIG. 8F represents CASD1/B4GALNT1/ST8SIA, FIG. 8G represents CERK, FIG. 8H represents CERK/B4GALNT1, FIG. 8I represents CERK/ST8SIA, FIG. 8J represents CERK/B4GALNT1/ST8SIA, FIG. 8K represents PIK3C2A, FIG. 8L represents PIK3C2A/B4GALNT1, FIG. 8M represents PIK3C2A/ST8SIA, FIG. 8N represents PIK3C2A/B4GALNT1/ST8SIA, FIG. 8O represents CASD1, FIG. 8P represents CASD1/CERK, FIG. 8Q represents CASD1/PIK3C2A.

[0373] FIG. 9 is a bar plot showing the quantification of the mean fluorescence intensity of O-acetylated GD2 ganglioside in breast cancer cell line MDA-MB-231 GD3S+treated with a CERK inhibitor for 2, 4, 6, 8, 20, 24 or 48 hours and stained by immunocytochemistry with an anti-OAcGD2 antibody. Mean fluorescence intensity was measured by confocal microscopy.

[0374] FIG. 10 is a bar plot showing the migration capacity of the breast cancer cell line MDA-MB-231 GD3S+treated with a CERK inhibitor. Cells were placed in a Transwell in presence or absence of CERK inhibitor and counted after 24 hours of treatment.

[0375] FIG. 11 is a bar plot showing the quantitative real-time PCR quantification of CERK mRNA expression in transiently transfected breast cancer cell line MDA-MB231 GD3S+. MDA-MB-231 GD3S+cells were transfected with either a control siRNA (siControl), or a siRNA targeting CERK (siCERK2 or siCERK4). Results were normalized to the expression of HPRT mRNA.

[0376] FIG. 12A-D are representative confocal microscopy photographs of the analysis of OAcGD2 expression in transiently transfected breast cancer cell line MDA-MB-231 GD3S+. Cells were stained by immunocytochemistry using an anti-OAcGD2 antibody. FIG. 12A shows cells transfected with a control siRNA (siControl). FIG. 12B shows cells transfected with a siRNA targeting CASD1 (siCASD1). FIG. 12C shows cells transfected with a siRNA targeting CERK (siCERK2). FIG. 12D shows cells transfected with a siRNA targeting CERK (siCERK4).

[0377] FIG. 13 is a bar plot showing the quantification of the mean fluorescence intensity of O-acetylated GD2 ganglioside in transiently transfected breast cancer cell line MDA-MB-231 GD3S+. MDA-MB-231 GD3S+cells were transiently transfected either with a control siRNA (siControl), a siRNA targeting CASD1 (siCASD1), or one of two different siRNA targeting CERK (siCERK2 or siCERK4). Cells were stained by immunocytochemistry using an anti-OAcGD2 antibody. Mean fluorescence intensity was quantified by confocal microscopy.

[0378] FIG. 14 is a bar plot showing the migration capacity of transiently transfected breast cancer cell line MDA-MB-231 GD3S+. MDA-MB-231 GD3S+cells were transiently transfected either with a control siRNA (siControl), a siRNA targeting CASD1 (siCASD1), or one of two different siRNA targeting CERK (siCERK2 or siCERK4). Cells were placed in a Transwell and counted after 24 hours of treatment.

[0379]

TABLE-US-00002 TABLEOFSEQUENCES SEQ ID NO Sequencefunction Sequence 1 CASD1humanprotein MAALAYNLGKREINHYFSVRSAKVLALVAVLLLAACHLASRRYRGNDS CEYLLSSGRFLGEKVWQPHSCMMHKYKISEAKNCLVDKHIAFIGDSRIRQ LFYSFVKIINPQFKEEGNKHENIPFEDKTASVKVDFLWHPEVNGSMKQCI KVWTEDSIAKPHVIVAGAATWSIKIHNGSSEALSQYKMNITSIAPLLEKLA KTSDVYWVLQDPVYEDLLSENRKMITNEKIDAYNEAAVSILNSSTRNSKS NVKMFSVSKLIAQETIMESLDGLHLPESSRETTAMILMNVYCNKILKPVD GSCCQPRPPVTLIQKLAACFFTLSIIGYLIFYIIHRNAHRKNKPCTDLESGEE KKNIINTPVSSLEILLQSFCKLGLIMAYFYMCDRANLFMKENKFYTHSSFF IPIIYILVLGVFYNENTKETKVLNREQTDEWKGWMQLVILIYHISGASTFL PVYMHIRVLVAAYLFQTGYGHFSYFWIKGDFGIYRVCQVLFRLNFLVVV LCIVMDRPYQFYYFVPLVTVWFMVIYVTLALWPQIIQKKANGNCFWHFG LLLKLGFLLLFICFLAYSQGAFEKIFSLWPLSKCFELKGNVYEWWFRWRL DRYVVFHGMLFAFIYLALQKRQILSEGKGEPLFSNKISNFLLFISVVSFLTY SIWASSCKNKAECNELHPSVSVVQILAFILIRNIPGYARSVYSSFFAWFGKI SLELFICQYHIWLAADTRGILVLIPGNPMLNIIVSTFIFVCVAHEISQITNDL AQIIIPKDNSSLLKRLACIAAFFCGLLILSSIQDKSKH 2 GD2synthasehuman MWLGRRALCALVLLLACASLGLLYASTRDAPGLRLPLAPWAPPQSPRRP protein ELPDLAPEPRYAHIPVRIKEQVVGLLAWNNCSCESSGGGLPLPFQKQVRA IDLTKAFDPAELRAASATREQEFQAFLSRSQSPADQLLIAPANSPLQYPLQ GVEVQPLRSILVPGLSLQAASGQEVYQVNLTASLGTWDVAGEVTGVTLT GEGQADLTLVSPGLDQLNRQLQLVTYSSRSYQTNTADTVRFSTEGHEAA FTIRIRHPPNPRLYPPGSLPQGAQYNISALVTIATKTFLRYDRLRALITSIRR FYPTVTVVIADDSDKPERVSGPYVEHYLMPFGKGWFAGRNLAVSQVTTK YVLWVDDDFVFTARTRLERLVDVLERTPLDLVGGAVREISGFATTYRQL LSVEPGAPGLGNCLRQRRGFHHELVGFPGCVVTDGVVNFFLARTDKVRE VGFDPRLSRVAHLEFFLDGLGSLRVGSCSDVVVDHASKLKLPWTSRDAG AETYARYRYPGSLDESQMAKHRLLFFKHRLQCMTSQ 3 GD3synthasehuman MSPCGRARRQTSRGAMAVLAWKFPRTRLPMGASALCVVVLCWLYIFPV protein YRLPNEKEIVQGVLQQGTAWRRNQTAARAFRKQMEDCCDPAHLFAMT KMNSPMGKSMWYDGEFLYSFTIDNSTYSLFPQATPFQLPLKKCAVVGNG GILKKSGCGRQIDEANFVMRCNLPPLSSEYTKDVGSKSQLVTANPSIIRQR FQNLLWSRKTFVDNMKIYNHSYIYMPAFSMKTGTEPSLRVYYTLSDVGA NQTVLFANPNFLRSIGKFWKSRGIHAKRLSTGLFLVSAALGLCEEVAIYG FWPFSVNMHEQPISHHYYDNVLPFSGFHAMPEEFLQLWYLHKIGALRMQ LDPCEDTSLQPTS 4 CERKhumanprotein MGATGAAEPLQSVLWVKQQRCAVSLEPARALLRWWRSPGPGAGAPGA DACSVPVSEIIAVEETDVHGKHQGSGKWQKMEKPYAFTVHCVKRARRH RWKWAQVTFWCPEEQLCHLWLQTLREMLEKLTSRPKHLLVFINPFGGK GQGKRIYERKVAPLFTLASITTDIIVTEHANQAKETLYEINIDKYDGIVCV GGDGMFSEVLHGLIGRTQRSAGVDQNHPRAVLVPSSLRIGIIPAGSTDCV CYSTVGTSDAETSALHIVVGDSLAMDVSSVHHNSTLLRYSVSLLGYGFY GDIIKDSEKKRWLGLARYDFSGLKTFLSHHCYEGTVSFLPAQHTVGSPRD RKPCRAGCFVCRQSKQQLEEEQKKALYGLEAAEDVEEWQVVCGKFLAI NATNMSCACRRSPRGLSPAAHLGDGSSDLILIRKCSRFNFLRFLIRHTNQQ DQFDFTFVEVYRVKKFQFTSKHMEDEDSDLKEGGKKRFGHICSSHPSCC CTVSNSSWNCDGEVLHSPAIEVRVHCQLVRLFARGIEENPKPDSHS 5 PIK3C2Ahuman MAQISSNSGFKECPSSHPEPTRAKDVDKEEALQMEAEALAKLQKDRQVT protein DNQRGFELSSSTRKKAQVYNKQDYDLMVFPESDSQKRALDIDVEKLTQA ELEKLLLDDSFETKKTPVLPVTPILSPSFSAQLYFRPTIQRGQWPPGLPGPS TYALPSIYPSTYSKQAAFQNGFNPRMPTFPSTEPIYLSLPGQSPYFSYPLTP ATPFHPQGSLPIYRPVVSTDMAKLFDKIASTSEFLKNGKARTDLEITDSKV SNLQVSPKSEDISKFDWLDLDPLSKPKVDNVEVLDHEEEKNVSSLLAKDP WDAVLLEERSTANCHLERKVNGKSLSVATVTRSQSLNIRTTQLAKAQGH ISQKDPNGTSSLPTGSSLLQEVEVQNEEMAAFCRSITKLKTKFPYTNHRTN PGYLLSPVTAQRNICGENASVKVSIDIEGFQLPVTFTCDVSSTVEIIIMQAL CWVHDDLNQVDVGSYVLKVCGQEEVLQNNHCLGSHEHIQNCRKWDTEI RLQLLTFSAMCQNLARTAEDDETPVDLNKHLYQIEKPCKEAMTRHPVEE LLDSYHNQVELALQIENQHRAVDQVIKAVRKICSALDGVETLAITESVKK LKRAVNLPRSKTADVTSLFGGEDTSRSSTRGSLNPENPVQVSINQLTAAIY DLLRLHANSGRSPTDCAQSSKSVKEAWTTTEQLQFTIFAAHGISSNWVSN YEKYYLICSLSHNGKDLFKPIQSKKVGTYKNFFYLIKWDELIIFPIQISQLP LESVLHLTLFGILNQSSGSSPDSNKQRKGPEALGKVSLPLFDFKRFLTCGT KLLYLWTSSHTNSVPGTVTKKGYVMERIVLQVDFPSPAFDIIYTTPQVDR SIIQQHNLETLENDIKGKLLDILHKDSSLGLSKEDKAFLWEKRYYCFKHP NCLPKILASAPNWKWVNLAKTYSLLHQWPALYPLIALELLDSKFADQEV RSLAVTWIEAISDDELTDLLPQFVQALKYEIYLNSSLVQFLLSRALGNIQIA HNLYWLLKDALHDVQFSTRYEHVLGALLSVGGKRLREELLKQTKLVQL LGGVAEKVRQASGSARQVVLQRSMERVQSFFQKNKCRLPLKPSLVAKEL NIKSCSFFSSNAVPLKVTMVNADPMGEEINVMFKVGEDLRQDMLALQMI KIMDKIWLKEGLDLRMVIFKCLSTGRDRGMVELVPASDTLRKIQVEYGV TGSFKDKPLAEWLRKYNPSEEEYEKASENFIYSCAGCCVATYVLGICDRH NDNIMLRSTGHMFHIDFGKFLGHAQMFGSFKRDRAPFVLTSDMAYVING GEKPTIRFQLFVDLCCQAYNLIRKQTNLFLNLLSLMIPSGLPELTSIQDLKY VRDALQPQTTDAEATIFFTRLIESSLGSIATKFNFFIHNLAQLRFSGLPSND EPILSFSPKTYSFRQDGRIKEVSVFTYHKKYNPDKHYIYVVRILREGQIEPS FVFRTFDEFQELHNKLSIIFPLWKLPGFPNRMVLGRTHIKDVAAKRKIELN SYLQSLMNASTDVAECDLVCTFFHPLLRDEKAEGIARSADAGSFSPTPGQ IGGAVKLSISYRNGTLFIMVMHIKDLVTEDGADPNPYVKTYLLPDNHKTS KRKTKISRKTRNPTFNEMLVYSGYSKETLRQRELQLSVLSAESLRENFFL GGVTLPLKDFNLSKETVKWYQLTAATYL 6 PDK3humanprotein MRLFRWLLKQPVPKQIERYSRFSPSPLSIKQFLDFGRDNACEKTSYMFLR KELPVRLANTMREVNLLPDNLLNRPSVGLVQSWYMQSFLELLEYENKSP EDPQVLDNFLQVLIKVRNRHNDVVPTMAQGVIEYKEKFGFDPFISTNIQY FLDRFYTNRISFRMLINQHTLLFGGDTNPVHPKHIGSIDPTCNVADVVKD AYETAKMLCEQYYLVAPELEVEEFNAKAPDKPIQVVYVPSHLFHMLFEL FKNSMRATVELYEDRKEGYPAVKTLVTLGKEDLSIKISDLGGGVPLRKID RLFNYMYSTAPRPSLEPTRAAPLAGFGYGLPISRLYARYFQGDLKLYSME GVGTDAVIYLKALSSESFERLPVFNKSAWRHYKTTPEADDWSNPSSEPRD ASKYKAKQ 7 MERTKhuman MGPAPLPLLLGLFLPALWRRAITEAREEAKPYPLFPGPFPGSLQTDHTPLL protein SLPHASGYQPALMFSPTQPGRPHTGNVAIPQVTSVESKPLPPLAFKHTVG HIILSEHKGVKFNCSISVPNIYQDTTISWWKDGKELLGAHHAITQFYPDDE VTAIIASFSITSVQRSDNGSYICKMKINNEEIVSDPIYIEVQGLPHFTKQPES MNVTRNTAFNLTCQAVGPPEPVNIFWVQNSSRVNEQPEKSPSVLTVPGLT EMAVFSCEAHNDKGLTVSKGVQINIKAIPSPPTEVSIRNSTAHSILISWVPG FDGYSPFRNCSIQVKEADPLSNGSVMIFNTSALPHLYQIKQLQALANYSIG VSCMNEIGWSAVSPWILASTTEGAPSVAPLNVTVFLNESSDNVDIRWMK PPTKQQDGELVGYRISHVWQSAGISKELLEEVGQNGSRARISVQVHNAT CTVRIAAVTRGGVGPFSDPVKIFIPAHGWVDYAPSSTPAPGNADPVLIIFG CFCGFILIGLILYISLAIRKRVQETKFGNAFTEEDSELVVNYIAKKSFCRRAI ELTLHSLGVSEELQNKLEDVVIDRNLLILGKILGEGEFGSVMEGNLKQED GTSLKVAVKTMKLDNSSQREIEEFLSEAACMKDFSHPNVIRLLGVCIEMS SQGIPKPMVILPFMKYGDLHTYLLYSRLETGPKHIPLQTLLKFMVDIALG MEYLSNRNFLHRDLAARNCMLRDDMTVCVADFGLSKKIYSGDYYRQGR IAKMPVKWIAIESLADRVYTSKSDVWAFGVTMWEIATRGMTPYPGVQN HEMYDYLLHGHRLKQPEDCLDELYEIMYSCWRTDPLDRPTFSVLRLQLE KLLESLPDVRNQADVIYVNTQLLESSEGLAQGSTLAPLDLNIDPDSIIASCT PRAAISVVTAEVHDSKPHEGRYILNGGSEEWEDLTSAPSAAVTAEKNSVL PGERLVRNGVSWSHSSMLPLGSSLPDELLFADDSSEGSEVLM 8 NME3humanprotein MICLVLTIFANLFPAACTGAHERTFLAVKPDGVQRRLVGEIVRRFERKGF KLVALKLVQASEELLREHYAELRERPFYGRLVKYMASGPVVAMVWQGL DVVRTSRALIGATNPADAPPGTIRGDFCIEVGKNLIHGSDSVESARREIAL WFRADELLCWEDSAGHWLYE 9 CASD1PCRprimer GCTCGGGATCCGCGGCTCTGGCCTACAACCTG 10 CASD1PCRprimer GCTCGCTCGAGATGTTTTGATTTATCTTGAATGGATG 11 V5epitopepre- AGCTTCGAATGGGTAAGCCTATCCCTAACCCTCTCCTCGGTCTCGATT hybridized CTACGG oligonucleotide 12 V5epitopepre- GATCCCGTAGAATCGAGACCGAGGAGAGGGTTAGGGATAGGCTTACC hybridized CATTCGA oligonucleotide 13 Mycepitopepre- TCGAGGAACAAAAACTCATCTCAGAAGAGGATCTGAATTAAT hybridized oligonucleotide 14 Mycepitopepre- CTAGATTAATTCAGATCCTCTTCTGAGATGAGTTTTTGTTCC hybridized oligonucleotide 15 S94Amutagenesis GCATTTATTGGAGATGCCAGAATTCGTCAATTG primer 16 S94Amutagenesis CAATTGACGAATTCTGGCATCTCCAATAAATGC primer 17 ST8SIA1PCRprimer GCTAAGCTTCGAATGGGTAAGCCTATCCCTAACCCTCTCCTCGGTCTC (G3DS) GATTCTACGGGTACCAGCCCCTGCGGGCGGGC 18 ST8SIA1PCRprimer GCTGCGGCCGCCTAGGAAGTGGGCTGGAGTG (G3DS) 19 self-cleaving2A QCTNYALLKLAGDVESNPGP peptideofequine rhinitisAvirus 20 GD3S/self-cleaving2A ATAGCGGCCGCATGAGCCCCTGCGG peptide/GD2SPCR primer 21 GD3S/self-cleaving2A GCTCTCTAGATCACTCGGCGGTCATGCAC peptide/GD2SPCR primer 22 exon2-specifictarget TTGCATTTATCGGAGATTCCAGG sequence 23 TargetregionPCR GCTGTGCCTAACAGTTTG primer 24 TargetregionPCR TGGCAAGTTTTTCCATGAG primer 25 TargetregionPCR TGAAGCAAAGAATTGCCTTGTAGA primer 26 TargetregionPCR CTTATTTCCTTCTTCTTTAAACTGGG primer 27 CASD1PCRprimer ATGTTCACAACGCCACGG 28 CASD1PCRprimer CAGGAACCATCCACAGGC 29 NeuDPCRprimer CGCCGCGGATCCGAAAAAATAACCTTAAAATGC 30 NeuDPCRprimer GTCCGCTCGAGTTAAAATAGATTAAAAATTTTTTTTGATTTTAG 31 CASD1PCRprimer GTGGATTTTCTGTGGCATCC 32 CASD1PCRprimer AAGCGCTTCACTGCTACCAT 33 B4GALNT1PCR CAGCGCTCTAGTCACGATTGC primer(G2DS) 34 B4GALNT1PCR CCACGGTAACCGTTGGGTAG primer(G2DS) 35 ST8SIA1PCRprimer GCGATGCAATCTCCCTCCT (G3DS) 36 ST8SIA1PCRprimer TTCCCGAATTATGCTGGGAT (G3DS) 37 8B6mAbLC-CDR1 QSLLKNNGNTFL 38 8B6mAbLC-CDR3 SQSTHIPYT 39 8B6mAbHC-CDR1 EFTFTDYY 40 8B6mAbHC-CDR2 IRNRANGYTT 41 8B6mAbHC-CDR3 ARVSNWAFDY 42 Humanizedanti- D/EV/IVMTQSPL/AS/TLP/SV/L/AS/TL/P/VGD/Q/EQ/P/RA/VS/TI/LS/TCRS/ OAcGD2VL ASQSL/VL/VKN/SN/QG/A/SN/Y/ST/N/SF/YLH/N/S/A/Y/GWY/FL/QQK/RPGQ/ consensussequence KS/A/VPK/Q/RL/R/VLIYK/G/LV/A/GSN/TRL/D/AS/TGV/IPD/A/ SRFSGSGSGTY/DFTLK/TIS/NR/SV/LE/QA/PEDL/V/FG/AV/TYF/YCS/M/QQS/ AT/YH/Q/NI/T/QP/SYTFGG/QGTKVEIK 43 Humanizedanti- E/QVQLV/LESGGGLVQ/KPGG/RSLRLSCA/TT/ASE/GFTFT/S/GDY/HYMT/ OAcGD2VH H/N/SWV/IRQAPGKGLEWL/VG/SF/YI/TRNR/K/SA/SN/SG/A/SY/GT/IT/ consensussequence IE/YYN/AP/A/DSVKGRFTISRDN/GS/AKS/NI/S/TL/T/AYLQMNSLR/K/QT/ AEDTAV/I/LYYCA/TRVSNWA/YFDYWGQGTT/LL/VTVSS 44 Anti-OAcGD2VH49A EVQLVESGGGLVQPGRSLRLSCTTSEFTFTDYYMTWVRQAPGKGLEWL GFIRNRANGYTTEYNPSVKGRFTISRDNSKSILYLQMNSLKTEDTAVYYC ARVSNWAFDYWGQGTLVTVSS 45 Anti-OAcGD2VH72A EVQLVESGGGLVQPGGSLRLSCATSEFTFTDYYMTWVRQAPGKGLEWL GFIRNRANGYTTEYNPSVKGRFTISRDNSKNSLYLQMNSLKTEDTAVYY CARVSNWAFDYWGQGTLVTVSS 46 Anti-OAcGD2 EVQLVESGGGLVQPGRSLRLSCTTSEFTFTDYYMTWVRQAPGKGLEWL VH49BHS GFIRNKANGYTTEYNPSVKGRFTISRDNSKSILYLQMNSLKTEDTAVYYC ARVSNWAFDYWGQGTLVTVSS 47 Anti-OAcGD2 EVQLVESGGGLVQPGGSLRLSCATSEFTFSDYYMTWVRQAPGKGLEWL VH72BHNPS GFIRNKANGYTTEYNPSVKGRFTISRDNSKNSLYLQMNSLKTEDTAVYY CARVSNWAFDYWGQGTLVTVSS 48 Anti-OAcGD2VL30A DVVMTQSPLSLPVTLGQPASISCRSSQSLLKNNGNTFLHWYQQRPGQSPR LLIYKVSNRLSGVPDRFSGSGSGTDFTLKISRVEAEDVGVYFCSQSTHIPY TFGGGTKVEIK 49 Anti-OAcGD2VL28A DVVMTQSPLSLPVTPGEPASISCRSSQSLLKNNGNTFLHWYLQKPGQSPQ LLIYKVSNRLSGVPDRFSGSGSGTDFTLKISRVEAEDVGVYFCSQSTHIPY TFGQGTKVEIK 50 Anti-OAcGD2 DVVMTQSPLSLPVTPGEPASISCRSSQSLLKSNANTFLHWYLQKPGQSPQ VL28Bs01/A2 LLIYKVSNRLSGVPDRFSGSGSGTDFTLKISRVEAEDVGVYYCMQSTHIP YTFGQGTKVEIK 51 Humanizedanti- DV/IVMTQSPLSLPVS/TL/PGD/Q/EQ/PASISCRSSQSLL/VKN/SNG/ANTF/ OAcGD2VL YLHWY/FL/QQK/RPGQSPK/Q/RLLIYKVSNRL/ASGVPDRFSGSGSGTY/ consensussequence DFTLKISRVEAEDL/VGVYF/YCS/MQSTHIPYTFGG/QGTKVEIK 52 Humanizedanti- EVQLVESGGGLVQPGG/RSLRLSCA/TT/ASE/GFTFT/S/GDY/HYMT/ OAcGD2VH SWVRQAPGKGLEWLGFIRNR/KAN/SG/S/AY/GTT/IE/YYN/AP/ consensussequence ASVKGRFTISRDNSKS/NI/SL/AYLQMNSLR/KTEDTAVYYCA/ TRVSNWAFDYWGQGTT/LL/VTVSS 53 EZH2humanprotein MGQTGKKSEKGPVCWRKRVKSEYMRLRQLKRFRRADEVKSMFSSNRQ KILERTEILNQEWKQRRIQPVHILTSVSSLRGTRECSVTSDLDFPTQVIPLK TLNAVASVPIMYSWSPLQQNFMVEDETVLHNIPYMGDEVLDQDGTFIEE LIKNYDGKVHGDRECGFINDEIFVELVNALGQYNDDDDDDDGDDPEERE EKQKDLEDHRDDKESRPPRKFPSDKIFEAISSMFPDKGTAEELKEKYKEL TEQQLPGALPPECTPNIDGPNAKSVQREQSLHSFHTLFCRRCFKYDCFLH RKCNYSFHATPNTYKRKNTETALDNKPCGPQCYQHLEGAKEFAAALTA ERIKTPPKRPGGRRRGRLPNNSSRPSTPTINVLESKDTDSDREAGTETGGE NNDKEEEEKKDETSSSSEANSRCQTPIKMKPNIEPPENVEWSGAEASMFR VLIGTYYDNFCAIARLIGTKTCRQVYEFRVKESSIIAPAPAEDVDTPPRKK KRKHRLWAAHCRKIQLKKDGSSNHVYNYQPCDHPRQPCDSSCPCVIAQ NFCEKFCQCSSECQNRFPGCRCKAQCNTKQCPCYLAVRECDPDLCLTCG AADHWDSKNVSCKNCSIQRGSKKHLLLAPSDVAGWGIFIKDPVQKNEFIS EYCGEIISQDEADRRGKVYDKYMCSFLFNLNNDFVVDATRKGNKIRFAN HSVNPNCYAKVMMVNGDHRIGIFAKRAIQTGEELFFDYRYSQADALKY VGIEREMEIP 54 HPRTPCRprimer GCCAGACTTTGTTGGATTTG (forward) 55 HPRTPCRprimer CTCTCATCTTAGGCTTTGTATTTTG (reverse)

[0380] In the consensus sequences of SEQ ID NO: 42, 43, 51 and 52, residuel/residue2 at a given position means that the residue at that position is either residuel or residue2. CDRs are represented in bold.

EXAMPLES

[0381] The present invention is further illustrated by the following examples.

Example 1: CASD1 is Essential for O-Acetylation of GD2

Materials and Methods

Antibodies

[0382] The anti-GD3 R24 mouse IgG3 was purchased from Abcam (Cambridge, MA, USA). The mouse IgM anti-9-OAcGD3 mAb M-T6004 was from Thermo Scientific (Waltham, USA). The anti-GD2 mAb 14.18 mouse IgG3/k and the anti-OAcGD2 mAb 8B6 mouse IgG3/k were produced in CHO cells by OGD2 Pharma (Nantes, France). The mouse IgG2a anti-GD2 mAb ME361 used for immune-TLC experiments was from Kerafast (Winston-Salem, USA). The secondary antibodies Alexa Fluor 488 donkey anti-mouse IgG and Alexa Fluor 546 donkey anti-rabbit IgG were purchased from Invitrogen (Cergy Pontoise, France).

Mammalian Expression Plasmids

[0383] To generate a construct encoding full-length CASD1 with an N-terminal V5 and a C-terminal Myc epitope (V5-CASD1-Myc), the coding region of human CASD1 (accession no. NM_022900) was amplified by PCR using the primers 5-GCTCGGGATCCGCGGCTCTGGCCTACAACCTG-3 (SEQ ID NO: 9) and 5-GCTCGCTCGAGATGTTTTGATTTATCTTGAATGGATG-3 (SEQ ID NO: 10) containing BamHI and XhoI restriction sites (underlined), respectively, and the resulting PCR product was ligated into the corresponding restriction sites of the vector pcDNA3 (Invitrogen). Sequences encoding the epitope tags were inserted by adapter ligation. For the V5 epitope, the pre-hybridized oligonucleotide pair 5-AGCTTCGAATGGGTAAGCCTATCCCTAACCCTCTCCTCGGTCTCGATTCTACGG-3 (SEQ ID NO: 11) and 5-GATCCCGTAGAATCGAGACCGAGGAGAGGGTTAGGGATAGGCTTACCCATTCG A-3 (SEQ ID NO: 12) was ligated into the HindIII and BamHI sites of pcDNA3. For the Myc epitope, the pre-hybridized oligonucleotide pair 5-TCGAGGAACAAAAACTCATCTCAGAAGAGGATCTGAATTAAT-3 (SEQ ID NO: 13) and 5-CTAGATTAATTCAGATCCTCTTCTGAGATGAGTTTTTGTTCC-3 (SEQ ID NO: 14) was ligated into the XhoI and XbaI sites of pcDNA3, resulting in the plasmid pcDNA3-V5-CASD1(wt)-Myc.

[0384] Site-directed mutagenesis was performed by PCR using the QuikChange site-directed mutagenesis kit (Stratagene) and pcDNA3-V5-CASD1(wt)-Myc as template. To introduce the amino-acid exchange S94A, the mutagenesis primers 5-GCATTTATTGGAGATGCCAGAATTCGTCAATTG-3 (SEQ ID NO: 15) and 5-CAATTGACGAATTCTGGCATCTCCAATAAATGC-3 (SEQ ID NO: 16) were used, resulting in the plasmid pcDNA-V5-CASD1(S94A)-Myc.

[0385] For expression of the human sialyltransferase ST8SIA I, a full-length construct encoding an N-terminal V5 epitope was generated by amplification of the coding region of human ST8SIA1 (I.M.A.G.E. clone IRCMp5012B0613D, ImaGenes) with the primers 5-GCTAAGCTTCGAATGGGTAAGCCTATCCCTAACCCTCTCCTCGGTCTCGATTCT ACGGGTACCAGCCCCTGCGGGCGGGC-3 (SEQ ID NO: 17) and 5-GCTGCGGCCGCCTAGGAAGTGGGCTGGAGTG-3 (SEQ ID NO: 18) containing the sequence encoding the V5 epitope (bold), HindIII and NotI restriction sites (underlined). The resulting PCR fragment was ligated into the HindIII and NotI sites of the expression vector pcDNA3.1-zeo (Invitrogen). For efficient co-expression of GD3S and GD2S, the inventors generated a plasmid that carries the coding sequence of GD3S (accession no. NM_011374.2) without stop-codon fused to a sequence stretch that encodes the self-cleaving 2A peptide of equine rhinitis A virus (QCTNYALLKLAGDVESNPGP (SEQ ID NO: 19)) and the coding sequence of GD2S (accession no. NM_008080.5). The entire tripartite sequence was generated by gene synthesis (Eurofins MWG Operon), amplified by PCR using the primers 5 -ATAGCGGCCGCATGAGCCCCTGCGG-3 (SEQ ID NO: 20) and 5 -GCTCTCTAGATCACTCGGCGGTCATGCAC-3 (SEQ ID NO: 21), and the obtained PCR product was ligated into the Notl and Xbal restriction sites of the vector pcDNA3 (Invitrogen). The identity of the final construct was verified by sequencing.

Mammalian Cell Culture

[0386] Cell culture reagents were purchased from Lonza (Verviers, Belgium). The human breast cancer cell SUM159PT was obtained by the American Tissue Culture Collection (ATCC, Rockville, MD, USA). Cells were routinely grown in monolayer culture and maintained at 37? C. in an atmosphere of 5% CO2. Chinese Hamster Ovary (CHO) cells were cultivated in Dulbecco's Modified Eagle's Medium (DMEM)/Ham's F12 1:1 (PAN-Biotech) supplemented with 5% fetal calf serum (FCS) (Sigma-Aldrich) and maintained at 37? C. and 5% CO2. SUM159PT cells were grown in DMEM/F12 (1:1) containing 5% heat-inactivated fetal calf serum (FCS), 2 mM L-glutamine, 1 ?g/mL hydrocortisone and 5 ?g/mL insulin.

Transfection of CHO Cells

[0387] For transient transfections, CHO cells were cultivated in 10 cm dishes until they reached 70-80% confluency. A mixture of 12 ?l PEI MAX (Polysciences) and 12 ?g of plasmid DNA was prepared in 1.2 ml Opti-MEM (Gibco), incubated for 20 mM at room temperature and added drop-wise to a cell culture containing 12 ml of culture medium. After 6 h, transfections were stopped by removal of the transfection mixture and the addition of fresh culture medium. Transfections in 24-well plates were performed accordingly using a mixture of 0.5 ?l PEI and 0.5 ?g DNA in 50 ?l of OptiMEM that was added to cells maintained in 500 ?l of culture medium.

siRNA Transfection of SUM159PT Cells

[0388] Depletion of CASD1 was performed using siRNA strategy by a double transfection. The second transfection was performed 48 h after the first one using the same conditions. Cells were grown in six-well plates and transfections were performed with 2 ?M of siRNA-targeting CASD1 (L-016926-01-0010, Horizon) or a scramble sequence and 8 ?L RNAimax (#137781, Thermo-Fisher Scientific) in 1 mL of UltraMem (Lonza). After 5 h, transfection was stopped by adding 1 mL of DMEM/F12 media supplemented with 5% FCS. Cells were collected at 72 h for quantitative polymerase chain reaction (qPCR) and immunocytochemistry experiments.

shRNA Transfection of SUM159PT Cells

[0389] Stable depletion of CASD1 was performed using shRNA strategy. ShRNA encoding plasmids were from EZyvec (Loos, France). Cells were grown in six-well plates and transfection was performed with 500 ng of shRNA plasmid targeting-CASD1 (A236.1b) or a scramble sequence in 4 ?L lipofectamine 2000 (Invitrogen). The selection of stable transfectants was performed by adding hygromycin at 500 ?g/ml 48 h after transfection.

Transfection of SUM159PT Cells with CASD1 or GD3 Synthase-Encoding Expression Vector

[0390] Transfection of SUM159PT cells was performed with RNAimax transfection reagent (#137781, Thermo-Fisher Scientific). Cells were grown in six-well plates, washed twice with UltraMem and transfected with 2 ?g of plasmid DNA and 4 ?L of RNAimax in 1 mL of UltraMem (Lonza). After 5 h, transfection was stopped by adding 1 mL of DMEM/F12 media supplemented with 5% of FCS. For the selection of stable transfectants, 500 ?g/mL of hygromycin was added per well 48 h post-transfection. Clones were isolated by limited dilution. Positive clones were selected by qPCR and immunocytochemistry-confocal microscopy experiments.

CRISPR/Cas-Mediated Genome Editing

[0391] CHO cells carrying a selective Casd1 gene knockout (CHO?Casd1) were generated by introducing a frameshift mutation in exon 2 of Casd1 by CRISPR/Cas9-mediated genome editing. Exon 2 of hamster Casd1 corresponds to exon 3 of human CASD1 and encodes the active site serine. A plasmid encoding a respective Casd1-specific guide RNA was generated on the basis of the bicistronic vector pX330-U6-Chimeric_BB-CBh-hSpCas9 (Addgene plasmid # 42230; http://n2t.net/addgene:42230; RRID:Addgene_42230). Following the protocol provided in Cong et al., 2013, Science 339, 819-23, the exon 2-specific target sequence 5-TTGCATTTATCGGAGATTCCAGG-3 (PAM sequence underlined) (SEQ ID NO: 22) was inserted into the Bbsl sites of the vector. The final plasmid allowed co-expression of the RNA-guided nuclease Cas9 from Streptococcus pyogenes and the Casd1-specific guide RNA. Transient transfections in CHO cells were performed in 24-well plates using 0.375 jag of the CRISPR/Cas9-plasmid and 0.125 ?g of a reporter plasmid (pEGFP-C1, Clontech) that allowed expression of the enhanced green fluorescent protein (EGFP). After 24 h, cells were cloned by limiting dilution and colonies grown from EGFP-expressing single-cell clones were expanded and screened for frameshift mutations. This included amplification of the target region by PCR using two primer sets (5-GCTGTGCCTAACAGTTTG-3 (SEQ ID NO: 23)/5-TGGCAAGTTTTTCCATGAG-3 (SEQ ID NO: 24) and 5-TGAAGCAAAGAATTGCCTTGTAGA-3 (SEQ ID NO: 25)/5-CTTATTTCCTTCTTCTTTAAACTGGG-3 (SEQ ID NO: 26)) and sequencing of the obtained PCR product. CHO clones carrying homozygous or heterozygous frameshift mutations in exon 2 of Casd1 were subcloned by limiting dilution and re-analyzed. In this step, frameshift mutations were confirmed on the genomic level as described above and additionally verified on the transcript level by amplification of Casd1 transcripts by RT-PCR and analysis of the PCR products by sequencing. As gene-specific primers, the following multiple intron-spanning primer pair was used: 5-ATGTTCACAACGCCACGG-3 (exon 1) (SEQ ID NO: 27) and 5-CAGGAACCATCCACAGGC-3 (exon 8) (SEQ ID NO: 28). The CHO?Casd1 clone used in this study contains a 2 bp insertion on one allele and a 4 bp deletion on the second allele. Both frameshift mutations occurred at the 5-end of the triplet encoding Asp-60. This eliminated the triplet that encodes the catalytic residue Ser-61 and resulted in the formation of a premature stop codon in exon 2.

Production of the Sialyl-9-O-Acetyltransferase NeuD of Campylobacter jejuni

[0392] The coding sequence of NeuD (orf11) was amplified from genomic DNA of the Campylobacter jejuni (C. jejuni) strain MK104 (ATCC 43446) in a PCR reaction with the primers 5-CGCCGCGGATCCGAAAAAATAACCTTAAAATGC-3 (SEQ ID NO: 29) and 5-GTCCGCTCGAGTTAAAATAGATTAAAAATTTTTTTTGATTTTAG-3 (SEQ ID NO: 30). The obtained PCR product was ligated into the BamHI and XhoI sites of a pET32a (Novagen) vector that carries a sequence encoding the maltose binding protein (MBP), an (S)3(N)10-linker and a thrombin cleavage site (LVPRGS) that was inserted into the NdeI and Xhol sites, with the last two triplets encoding the most C-terminal amino acids of the cleavage site (GS) creating a unique BamHI restriction site. The identity of the resulting construct was confirmed by sequencing and the encoded MBP-NeuD fusion protein was expressed in E. coli BL21(DE3). Transformed cells were cultivated at 37? C. in Power Broth (AthenaES) until an optical density at 600 nm of 1.5 was reached. The expression was induced with 1mM isopropyl-?-D-thiogalactopyranoside (IPTG) and cultivation at 15? C. for 20 h. Cells were harvested and resuspended in binding buffer (20 mM Tris-HCl pH 7.4, 200 mM NaCl, 1 mM EDTA) containing 40 ?g/ml bestatin, 1 ?g/ml pepstatin and 1 mM PMSF, and were disrupted by sonication. Recombinant protein was purified on 1 ml MBPTrap HP columns (GE Healthcare) using 10 mM D-(+)-maltose in binding buffer for elution. Affinity purified protein was dialyzed against 50 mM MES pH 7.0 containing 100 mM NaCl (Slide-A-Lyzer, ThermoFisher, 3.5 kDa cutoff) and concentrated using an Amicon Ultra-4 centrifugal filter device (Merck Millipore, 50 kDa cutoff).

Generation of 9-O-Acetylated Gangliosides as Standards for TLC

[0393] The 9-O-acetylated forms of GD2 and GD3 were generated by enzymatic in vitro synthesis using NeuD from C. jejuni, which allows the site-selective introduction of an O-acetyl group at position C9 of a terminal ?2,8-linked sialic acid. GD3 (Sigma-Aldrich, 345752) and GD2 (Sigma-Aldrich, 345743) gangliosides (1 mM) were dissolved in MES buffer (50 mM) pH 6.5 or pH 7, containing acetyl-Coenzyme A (1 mM), MgCl.sub.2 (10 mM) and dithiothreitol (1 mM) (Fluka, Buchs, Germany) Various concentrations of sodium cholate (Sigma-Aldrich, Steinheim, Germany), ranging from 0 to 0.2% (w/v) were added to the reaction. NeuD (100 mU) was added and the reaction mixture was incubated at 37? C. for 3 hours with stirring (300 rpm). The reaction was stopped by adding an equal volume of methanol and gangliosides were purified on Chromabond C18 columns (Macherey-Nagel), dried under a nitrogen stream and dissolved in chloroform/methanol (1:2, v/v).

Extraction of Gangliosides

[0394] Total gangliosides were extracted from transfected CHO cells by mixing 107 cells with 3 ml chloroform/methanol (1:2, v/v) and sonic dispersion. After twenty pulses given by a Sonifier S-450 equipped with a cup horn (Branson), samples were incubated for 15 min in a bath sonicator. Debris were removed by centrifugation (1,600?g for 10 min) and the supernatant was transferred into a new tube. After adjusting a final ratio of chloroform/methanol/water of 4:8:5 (v/v/v), samples were centrifuged (1,600?g for 10 min) and the upper phase containing the ganglioside fraction was desalted on a Chromabond C18 column (Macherey-Nagel). Gangliosides were dried under a nitrogen stream, dissolved in 20 ?l of chloroform/methanol (1:2, v/v) and stored at ?20? C.

High-Performance Thin-Layer Chromatography (HPTLC) and Immunostaining

[0395] Total gangliosides of an equivalent of 2?10.sup.6 cells or 0.2 ?g of the indicated ganglioside standards were spotted on Nano-DURASIL-20 (0.2 mm silica gel 60) HPTLC plates (Macherey-Nagel) and chromatographed in chloroform/methanol/H.sub.2O (50:40:10, v/v/v) containing 0.05% calcium chloride. HPTLC plates were dried and chromatographed twice in 0.5% poly(isobutyl methacrylate) (Sigma-Aldrich) in hexane, which was prepared from a 25% stock solution in chloroform (w/v). Plates were dried and incubated overnight at 37? C. in PBS. After blocking with 2% BSA (w/v) in PBS for 1 h at room temperature, plates were incubated with the following primary antibodies diluted in PBS: Mouse IgG3 anti-9-OAcGD2 mAb 8B6 (10 ?g/ml; OGD2 Pharma), mouse IgM anti-9-OAcGD3 mAb M-T6004 (1:40; Thermo Scientific, MA1-34707), mouse IgG2a anti-GD2 mAb ME361 (15 ?g/ml; Kerafast EWI023), or mouse IgG3 anti-GD3 mAb R24 (10 ?g/ml; purified by protein A affinity chromatography from cell culture supernatant of R24 hybridoma cells ATCC HB-8445). HPTLC plates were washed three times with PBS and incubated for 1 h at room temperature with goat anti-mouse IgM IRDye 800CW-conjugate (1:20,000; LI-COR Biosciences, 926-32280) or goat anti-mouse IgG IRDye 800CW-conjugate (1:10,000; LI-COR Biosciences, 926-32210). HPTLC plates were washed with PBS and bound antibodies were detected by infrared imaging using an Odyssey Imaging System (LI-COR Biosciences).

RNA Extraction, cDNA Synthesis and qPCR

[0396] Gene expression was evaluated using real-time qPCR analysis after RNA extraction and cDNA synthesis. Total RNA was extracted using the Nucleospin RNA II kit (Macherey-Nagel, Duren, Germany) The amount of extracted RNA was quantified using a DeNovix DS-11 spectrophotometer (DeNovix Inc., Wilmington, DE, USA) and the purity of the RNA was checked by the ratio of the absorbance at 260 and at 280 nm. Total RNA was subjected to reverse transcription using the Maxima First Strand cDNA Synthesis Kit (ThermoFisher Scientific, Villeneuve d'Ascq, France) according to the protocol provided by the manufacturer. The oligonucleotide sequences (Eurogentec, Seraing, Belgium) used as primers for the PCR reactions are given in SEQ ID NO: 31 and SEQ ID NO: 32. qPCR and subsequent data analysis were performed using the Mx3005p Quantitative System (Stratagene, La Jolla, CA, USA). PCR reaction (25 ?L) contained 12.5 ?L of the 2X Brilliant SYBR Green qPCR Mastermix (Thermo Fischer Scientific, Rockford, USA), 300 nM of primers and 4 ?L of cDNA (1:40). DNA amplification was performed with the following thermal cycling profile: initial denaturation at 94? C. for 10 min, 40 cycles of amplification (denaturation at 94? C. for 30 s, annealing at Tm for 30 s, and extension at 72? C. for 30 s) and a final extension at 72? C. for 5 min. Hypoxanthineguanine PhosphoRibosylTransferase (HPRT) gene was used to normalize the expression of genes of interest. The fluorescence monitoring occurred at the end of each cycle. The analysis of amplification was performed using the Mx3005p software. The specificity of the amplification was checked by recording the dissociation curves. The efficiency of amplification was checked by serial dilutions of cDNA from SK-MEL-28 cells and was between 97 and 102%. All experiments were performed in triplicate. The quantification was performed by the method described by Pfaffl (Pfaffl, M.W., 2001, Nucleic Acids Res. 29(9), e45).

Immunocytochemistry and Confocal Microscopy

[0397] Transfected cells were grown on glass coverslips fixed for 15 min in 4% paraformaldehyde in 0.1 M sodium phosphate buffer. Cells were washed thrice with PBS and membrane permeabilization was performed in 5 ?g/mL digitonin in PBS for 20 min. After saturation in blocking buffer, cells were incubated with either with the anti-GD2 or anti-OAcGD2, or anti-V5-tag mAbs at 20 ?g/mL for 1 h followed by the secondary antibody for 1 h. Cells were washed and mounted in fluorescent mounting medium (Dako, Carpetaria, CA, USA). Stained slides were analyzed under a Zeiss LSM 700 confocal microscope. The same settings were used for all acquisitions to ensure the comparability of the data obtained.

MTS Assay

[0398] Cell growth was analyzed using the MTS reagent (Promega, Charbonni?res-les-bains, France) according to the manufacturer's instructions. Briefly, cells were seeded in 96 well plates in 0%, 1% or 5% FCS containing media in which MTS reagents were added. The proliferation rate was measured by the absorbance of MTS reagent at 490 nm at 24 h, 48 h, 72 h, and 96 h after seeding.

Transwell Assays

[0399] Migration and invasion properties of cells were measured by transwell assays using migration chambers or invasion chambers (Dutscher, Brumath, France). Cells were seeded in 24-well plates containing either migration or invasion chambers in serum-free media. After 24 h incubation at 37? C., cells were fixed 4% paraformaldehyde in 0.1 M sodium phosphate buffer and non-migratory/invasive cells were swapped with cotton swabs. Nuclei were counterstained with DAPI and membrane were mounted on the slide with fluorescent mounting medium (Dako, Carpetaria, CA, USA). Nuclei were counted under Leica microscope.

Statistical Analysis

[0400] Statistical difference was assessed using unpaired t-test or ordinary one-way Anova.

Results

CASD1 is Essential for the 9-O-Acetylation of GD2

[0401] Prior to deciphering the role of CASD1 in breast cancer cells, the inventors dissected the biosynthesis of 9-OAcGD2 in CHO cells, a well-defined cellular system. CHO cells display mainly the mono-sialyl ganglioside GM3, are easy to transfect and known to produce 9-OAcGD3 upon expression of GD3S. Using CRISPR/Cas9-mediated genome editing, the inventors generated CHO?Casd1 cells by introducing a frameshift mutation in exon 2. To produce GD2, CHO wild type (WT) and CHO?Casd1 cells were transiently transfected with a bicistronic plasmid that allows co-expression of GD3S and GD2S. Total gangliosides were extracted from transfected cells and analyzed by thin-layer chromatography (TLC). Upon transfection of the expression plasmid, but not of empty vector (mock), GD2 was detected in both CHO-WT and CHO?Casd1 cells (FIG. 1A, lower panel). The formation of 9-OAcGD2 was observed in CHO-WT, but not in Casd1-deficient cells (FIG. 1A, upper panel), demonstrating that the biosynthesis of 9-OAcGD2 critically relies on CASD1. In addition, the formation of GD3 and 9-OAcGD3 in GD3S expressing CHO cells was monitored (FIG. 1B). The deletion of Casd1 in CHO cells also prevented the formation of 9-OAcGD3.

CASD1 Expression is Ubiquitous Among Breast Cancer Cells

[0402] Results in CHO cells suggest that the expression of 9-OAcGD2, the major O-acetylated ganglioside species in breast cancer cells, is CASD1 dependent. CASD1 expression in breast cancer cells was next studied. The human protein atlas reveals that CASD1 is expressed in almost all healthy and cancer tissues (http://www.proteinatlas.org/ENSG00000127995 -CASD1/tissue). qPCR experiments were performed in order to quantify the expression of CASD1 in different breast cancer cells. The inventors used SUM159PT, Hs578T, and 2 clones derived from MDA-MB-231 (MDA-MB-231 GD3S+) and MCF-7 (MCF-7 GD3S+) breast cancer cell lines overexpressing GD3 synthase and high levels of complex gangliosides. SK-MEL-28 melanoma cells and LAN-1 neuroblastoma cells expressing high levels of O-acetylated gangliosides were used as controls. The results presented in FIG. 2 indicate that CASD1 expression is ubiquitous among breast cancer cells confirming the human protein atlas data. Among breast cancer cell lines tested, CASD1 is more expressed in MCF-7 and MCF-7 GD3S+, compared to MDA-MB-231, MDA-MB-231 GD3S+SUM159PT and Hs578T cells. The level of CASD1 expression is distinctly higher in SK-MEL-28 and LAN-1 compared to breast cancer cells. SUM129PT, a triple negative breast cancer cell line derived from anaplastic carcinoma, was chosen for this study. The inventors' previous data show a moderate expression of GD2 and OAcGD2, and CASD1 (FIG. 2) suggesting that SUM149PT is suitable for both depletion and overexpression of CASD1.

Reduction of CASD1 Expression

[0403] The reduction of CASD1 expression in SUM159PT was performed by transient transfection using siRNA strategy. The expression levels of GD2S (B4GALNT1) and CASD1 genes were determined by qPCR experiments and normalized to HPRT gene expression. Transfected cells exhibit a decrease of CASD1 gene expression (FIG. 3A) (up to 50%) while GD2 synthase gene expression is unchanged compared to control cells (FIG. 3B). The effect of CASD1 depletion on OAcGD2 expression was evaluated by immunofluorescence and confocal microscopy experiments. OAcGD2 expression was reduced in CASD1-depleted cells compared to control cells. The mean fluorescence intensity calculated based on multiple images showed that transfected cells exhibit an increased GD2 expression (FIG. 3C), but a 75% decrease in OAcGD2 expression compared to control cells (FIG. 3D). The inventors concluded that a 50% reduction of CASD1 gene expression lead to a 75% decrease of OAcGD2 expression in transiently transfected cells compared to SUM159PT control cells. The stable depletion of CASD1 expression using shRNA strategy was performed twice. Nevertheless, transfected cells did not grow after several passages in antibiotic-containing medium (data not shown) and stable CASD1 depletion could not be achieved in SUM159PT breast cancer cells.

Transient Overexpression of CASD1 in SUM159PT Breast Cancer Cells

[0404] Overexpression of CASD1 (CASD1+) in SUM159PT cells was performed using a plasmid that allows the expression of human CASD1 with an N-terminal V5-epitope. In these experiments, CASD1 and GD2 synthase (GD2S) gene expression was assessed by qPCR experiments and the effect of CASD1 overexpression on OAcGD2 expression was studied by immunocytochemistry and confocal microscopy. CASD1 mRNA expression level showed approximately a 3000-fold increase in transfected cells compared to control cells (FIG. 4B). GD2 synthase expression remained unchanged between controls and transfected cells (FIG. 4A). The efficiency of transfection was checked using an anti-V5-tag antibody and ganglioside expression with either anti-GD2 or anti-OAcGD2 antibodies. CASD1 transfected cells exhibited an increase in OAcGD2 and GD2 expression compared to control cells. Mean fluorescence intensity quantified for each condition showed that overexpression of CASD1 increased both GD2 (FIG. 4C) and OAcGD2 (FIG. 4D) expression by 60% and 55%, respectively. The inventors concluded that, as observed for the transient inhibition of CASD1 gene expression, the transient overexpression of CASD1 in SUM159PT showed an effect on OAcGD2 expression. Since the stable depletion of CASD1 by shRNA in SUM159PT cells remained unsuccessful (data not shown), stable overexpression was considered.

Stable Overexpression of CASD1 in SUM159PT Breast Cancer Cells

[0405] Stable transfectants overexpressing CASD1 (SUM159PT CASD1+) was produced using the plasmid pcDNA3.1 V5-tag-CASD1-cMyc and clones were isolated after antibiotic selection and limiting dilution cloning. From the 28 clones pre-selected, 12 clones were maintained during proliferation monitoring. CASD1 expression levels in these clones were assessed by qPCR experiments, confirming the overexpression of CASD1 in CASD1+clones compare to controls (data not shown). Selection of CASD1+clones among the 12 clones isolated has been performed by the analysis of GD2 and OAcGD2 expression using immunocytochemistry and confocal microscopy. Two CASD1+clones exhibiting high CASD1 gene expression and OAcGD2 ganglioside expression were used to study the biological properties. The level of expression of CASD1, OAcGD2 and GD2 of the two selected clones (clone #19 and clone #26) is depicted in FIG. 5. CASD1 mRNA expression was 2-fold and 3-fold-increased in clone #19 and in clone #26 compared to control cells, respectively (FIG. 5A). Mean fluorescence intensity quantified shows an increased level of OAcGD2 expression in clones #19 and #26 compared to the control (FIG. 5C), whereas the expression of GD2 remained unchanged (FIG. 5B).

Biological Properties of SUM159PT CASD1+

[0406] Biological properties of the SUM159PT CASD1+clones were studied by MTS and Transwell assays, to assess their proliferation and migration/invasion capabilities, respectively. SUM159PT CASD1+clones did not exhibit differential growth properties compared to their control counterpart, regardless of the percentage of fetal calf serum in the culture medium (FIGS. 6A, B, C). However, both clones showed increased migration (FIG. 6D) and invasion (FIG. 6E) capabilities in serum free media. The migration capabilities of SUM159PT CASD1+clones increased twice compared to their control counterpart (FIG. 6D). The invasion activity of clone #26 was doubled compared to control while this activity increased up to 10 folds in clone #19 compared to control (FIG. 6E).

Discussion

[0407] Ganglioside O-acetylation results from the enzymatic action of a SOAT on a sialic acid residue. Recent studies have highlighted the importance of OAcGD2 as a marker and therapeutic target of interest in neuro-ectoderm derived cancers, including breast cancer. Deciphering GD2 O-acetylation mechanisms and the involvement of CASD1 in OAcGD2 biosynthesis in breast cancer is therefore of utmost importance.

CASD1 is Involved in GD2 9-O-Acetylation in CHO Cells and in SUM159PT Cells.

[0408] In this study, the inventors first used CHO cell lines that do not naturally express b-series gangliosides, as a model to study CASD1 activity on gangliosides. Ganglioside expression can be modulated in these CHO cell lines, either by overexpressing GD3S required for GD3 expression, or both GD3S and GD2S for more complex ganglioside biosynthesis. Consequently, the CHO WT and CHO?Casd1 cell lines are suitable models to study CASD1 SOAT activity on different gangliosides. The use of these cell lines allowed us to conclude that no O-acetylated ganglioside was detected in CHO?Casd1 cells, highlighting the critical role of CASD1 in both GD3 and GD2 9-O-acetylation. These data also demonstrate that CASD1 is the unique SOAT involved in GD3 and GD2 9-O-acetylation in CHO cells.

[0409] Since there are no breast cancer cellular models available with a knockout for CASD1, the modulation of CASD1 expression was adopted as the strategy to assess the potential SOAT activity of CASD1 on GD2 O-acetylation in SUM159PT breast cancer cell line. Transient overexpression or depletion of CASD1 in SUM159PT cells modulated OAcGD2 expression: RNAi silencing of CASD1 induced a 70% decrease of OAcGD2 expression, whereas CASD1 overexpression increased OAcGD2 expression (50% increase). GD2 levels were either decreased (when CASD1 is overexpressed) or unchanged (when CASD1 is depleted). The inventors' previous structural analysis had allowed to identify 9-OAcGD2 as the major O-acetylated ganglioside species. Altogether, these data show that CASD1 is essential for GD2 9-O-acetylation in breast cancer cells, as demonstrated in CHO cells.

Influence of CASD1 and OAcGD2 on Breast Cancer Cell Properties

[0410] 30 clones overexpressing CASD1 have been isolated and assessed for OAcGD2 expression. Two clones were selected according to their level of OAcGD2/CASD1 overexpression. These clones exhibited higher migrative and invasive capacities with no modification of their proliferation rates, suggesting a role of OAcGD2 in breast cancer migration and invasion. Although O-acetylated gangliosides such as OAcGD3 and OAcGD2 are now considered as TACAs, there is very little data in the literature regarding their roles in cancer cell biology. OAcGD3 protects leukemic blasts, Jurkat cells and glioblastoma cells from apoptosis. Moreover, increased levels of 9-O-acetylated Neu5Ac corresponding notably to elevated 9-OAcGD3 were detected in acute lymphocytic leukemia (ALL) cells that developed resistance against vincristine or nilotinib, two drugs with different cytotoxic mechanisms. Treatment of ALL cells by a sialate acetyl esterase that cleaved the 9-O-acetyl residues from sialic acids made these cells more sensitive to both drugs. SIAE overexpression in hamster melanoma cells induced a loss of OAcGD3, altered cell morphology, a slower growth rate, and lower melanogenesis activity compared to controls. Previous studies suggest a role of OAcGD2 in cancer cell properties, for example an anti-OAcGD2 mAb c.8B6 monoclonal antibody inhibited glioblastoma and neuroblastoma cell proliferation in vitro and in vivo. The inventors described here higher migrative and invasive capacities of SUM159PT clones overexpressing CASD1 and 9-OAcGD2, with no modification in their proliferation rates. Importantly, CASD1 overexpression could modulate the expression of other O-acetylated gangliosides or sialylated glycosphingolipids (globo, lacto/neolacto series), which could also modify the biological properties of cancer cells.

[0411] CASD1 is ubiquitously expressed in all tissues and cells according to the Human Protein Atlas. In agreement, all breast cancer cell lines tested in this study express CASD1 at variable levels.

[0412] For now, CASD1 is mentioned only in very few publications in Pubmed (NCBI), showing the limited knowledge available regarding the physiological role of CASD1. The difficulties encountered for cloning and isolation of SOAT render the deciphering of O-acetylated ganglioside biosynthesis mechanisms complicated. The inventors' data indicate a role of CASD1 in GD2 O-acetylation in breast cancer cells and a CASD1-dependent pathway for both 9-OAcGD2 and 9-0AcGD3 in SUM159PT breast cancer cells and in CHO cells. In addition, increased tumorigenic properties of breast cancer cells over-expressing CASD1 and OAcGD2 were observed.

[0413] Altogether, the inventors' data allow to identify new markers and therapeutic targets for cancer treatment.

Example 2: Biomarkers of OAcGD2 Expression

Materials and Methods

Cell Culture

[0414] Cell culture reagents were purchased from Lonza (Verviers, Belgium). Cells were routinely grown in monolayer culture and maintained at 37? C. in an atmosphere of 5% CO2. The human breast cancer cell line MDA-MB-231 GD3S+cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-inactivated fetal calf serum, 2 mmol/L L-glutamine and 1 mM sodium pyruvate.

SiRNA Reverse Transfection

[0415] 2.5 ?L/well of 500 nM siGenome siRNA (Dharmacon/Horizon) was printed into blackwalled 384 well cell carrier ultra plates (Perkin Elmer) with Velocity 11 (Agilent) as described before (Chia et al., 2012, Mol. Syst. Biol., 8: 629.). Reverse siRNA transfection was performed by pre-mixing 0.1 ?L of Dharmafect 1 transfection reagent (#T-2001-03, Horizon) with 7.4 ?L of Optimem for 5 minutes. Addition of mixture to the siRNA plate was then performed with small multidrop combi cassette (Thermo-Fisher) and was left for complexation for 20 minutes with shaking. Addition of 2750 cells in 40 ?L of DMEM supplemented with 10% FCS per well was then performed with multidrop combi standard cassette at medium speed (Thermo-Fisher). siRNA targeting-GD2S (L-011279-00-0020, Horizon), Polo like kinase 1 (PLK1; L-003290-00-0020, Horizon) and on-targeting pool (D-001810-10-20, Horizon) were used as screen controls.

Immunofluorescence Staining and Automated Images Acquisition

[0416] After 72 h of incubation, transfected cells were fixed with 50 ?L per well of 4% paraformaldehyde in 2% sucrose and 0.1M sodium phosphate buffer, 15 mM at 37? C. Cells were washed once with Hepes buffer 0.2M pH 7.4 and membrane permeabilization was performed in 5 ?g/ml of digitonine in Hepes buffer for 20 min. Blocking was performed for 1 h with blocking buffer containing 0.2% gelatin, 2% BSA and 2% FCS. Aspiration of any liquid was performed with a 384 channels aspiration manifold at constant distance height from bottom of the well (V&P Scientific Inc). Sodium hydroxide treatment at 1 mM was added to selective control non-transfected well for the deacetylation of sialic acid. Staining of OAcGD2 was performed by incubation of 8B6 mAbs followed by suitable anti-mouse conjugated Alexa Fluor 488 secondary antibody at 1/500 dilution (Thermo Fisher Scientific). Each antibody was incubated successively for 1 h each on a 1 cm-span orbital shaker at 150 rpm. Nuclei were counterstained with 1 ?g/ml Hoescht Thermo Fisher Scientific. Washings after antibody incubation were performed three times with 0.2M Hepes at pH 7.4 and 5 minutes shaking. All dispensing steps were performed with multidrop combi standard cassette. Stained plates were subjected to sequential channel acquisition for Hoechst/Alexa 488 with high content spinning disk confocal imager: phenix Opera (Perkin Elmer). Eight fields per well were acquired with 20X NA 1.0 water immersion objective with default laser power and exposure settings.

Segmentation Pipeline for the Selection of the Hits

[0417] An OAcGD2 expression metric was derived with total cell thresholded fluorescence intensity obtained by immunodetection with 8B6 mAb in MDA-MB-231 GD3S+cells and was normalized with Hoeschst nuclei counts. Pools of four siRNAs per gene were arrayed in a series of 384-well plates. The segmentation pipeline applied for analysis of data derived with Columbus (Perkin Elmer) image analysis software and consisted of few module blocks: a basic flatfield correction for each image. Nuclei count detection with method B excluding nuclei object<50 ?m.sup.2 and with a 0.4 common intensity threshold. OAcGD2 signal detection with Image region-based algorithm with a threshold of 0.6 and with multiple objects detection, file hole algorithm on objects and exclusion of object size<2000 square pixel. The calculation of OAcGD2 fluorescent signal metric was then derived with the sum of pixel intensity for all objects over 2000 square pixels size divided by the nuclei number for all 8 fields image per well. The exclusion of objects with area less than 2000 square pixel size was applied to subfilter antibody artefacts.

Results

Key Controls

[0418] Key controls were designed to validate consistency of our workflow and to normalize plate to plate variations. In this screening, siRNA Non-Targeting pool (siNT) was added to empty wells of each 384-well plate as a negative control, siRNA targeting-Polo like kinase 1 (siPLK1) was used as siRNA transfection control. Finally, siRNA targeting GD2 synthase (siGD2S) was used as a modulator of OAcGD2 staining fluorescent signal. siPLK1 induced over 95% decrease in nuclei count as compared to NT transfected wells and confirmed efficient siRNA transfection in all plates tested. Nuclei count between siNT-transfected wells or non-transfected control wells were very similar highlighting the specific siPLK1 killing mediated effect and the very low level of transfection toxicity induced by our transfection reagents.

[0419] SiGD2S mediates a reduced intensity of OAcGD2 fluorescent signal in all plates tested when compared to signal in siNT control wells but showed some changes in silencing performance between the first screen replicate versus the second screen replicate. Due to this variability, the chemical treatment was used to control the modulation of OAcGD2 fluorescent signal. Sodium hydroxide has been shown previously to deacetylate all acetyl groups present on the cell surface and blocked efficiently antigen recognition by 8B6 mAb. Fixed cells in selected control wells were thus treated with NaOH 0.1M before primary antibody staining. OAcGD2 staining obtained on NaOH-treated wells was consistently abolished when compared to siNT transfected wells. The Z factor for siGD2S versus siNT was equal to 0.30 whereas the Z factor for NaOH treated wells versus siNT was around 0.70. Since Z factor readout with siNT and NaOH treated wells showed better consistency, these 2 key controls were used to calibrate screen data for normalization.

Formatting Results

[0420] SiRNA that showed high toxicity in the wells (total nuclei counts<1000) were excluded from the analysis. The number of wells affected by toxicity constituted fewer than 5% of total siRNA tested. To minimize variations between plate data, each datapoint was normalized with the alternative score dependent on plate mean values of control siNT and plate mean values of NaOH treated controls wells (Moreau et al., 2011, Cell, 146:303-317) by applying the following formula:

[00001]$\mathrm{Alternative}\mathrm{score}=\frac{\mathrm{Xi}-{\stackrel{\_}{X}}_{\mathrm{siNT}}}{{\stackrel{\_}{X}}_{\mathrm{siNT}}-{\stackrel{\_}{X}}_{\mathrm{NaOH}}}$

[0421] The cutoff for the selection of OAcGD2 up or downregulating hits was defined with the first derivative approach (Moreau et al., 2011, Cell, 146:303-317). Genes were ranked according to their alternative score value from the minimum to the maximum. The cutoffs were designed before the largest spike at lowest ranks and highest ranks of the first derivative.

Selection of Hits

[0422] Pearson correlation (r or R.sup.2) factor was calculated on the basis of alternative scores on both replicates datapoints. In this screening experiment, we obtained r=0.79 and R.sup.2=0.63 showing that the linear correlation between the two replicates was acceptable and that the screen outcome was reasonably reproducible. Five genes were identified as upregulating OAcGD2 expression (Table 1).

TABLE-US-00003 TABLE 1 Genes modulating OAcGD2 expression in MDA-MB-231 GD3S+ cells Replicate 1 alt Replicate 2 alt OAcGD2 Gene NMID score score modulation CERK NM_022766 5.24 6.19 Upregulation PIK3C2A NM_002645 13.10 8.13 Upregulation PDK3 NM_005391 5.84 2.23 Upregulation MERTK NM_006343 4.99 2.44 Upregulation NME3 NM_002513 5.52 5.46 Upregulation

Discussion

[0423] The OAcGD2 siRNA screen was analyzed based on the fluorescence intensity obtained by immunodetection using 8B6 mAb in MDA-MB-231 GD3S+cells. Results were replicated and analyzed by combining first derivatives cutoff method and visual confirmation of hits on both replicates leading to the identification of 5 hits upregulating OAcGD2 expression. Results obtained could be interpreted based on the identification of hits but also on the images acquired. Images obtained from the transfection of the MDA-MB-231 GD3S+using siRNA targeting the different genes selected from our screen revealed significant variations of cellular morphology for the hits upregulating OAcGD2 with an intracellular and membrane staining pattern. Cells transfected with siRNA like siCERK and siPI3KC2A showed extended shape. Modifications of cell morphology after siRNA transfection can occur frequently depending on the depleted gene.

Example 3: Transcriptomic Analyses

Materials and Methods

Transcriptomics Analyses

[0424] Data are TCGA datasets obtained from SurvExpress. Analyses were performed using SurvExpress optimized algorithm. Hazard ratio was calculated in patient populations computationally identified as high or low expression level of the gene of interest (i.e., CASD1, CERK, PIK3C2A, B4GALTN1, ST8SIA1) based on individual signature expression in TCGA datasets by SurvExpress optimized algorithm. The datasets analyzed were Sarcoma (SARC); Pheochromocytoma and Paraganglioma (PCPG); Uterine Corpus Endometrial Carcinoma (UCEC); Thyroid carcinoma (THCA); Thymoma (THYM); Testicular Germ Cell Tumors (TGCT); Stomach adenocarcinoma (STAD); Skin Cutaneous Melanoma (SKCM); Prostate adenocarcinoma (PRAD); Pancreatic adenocarcinoma (PAAD); Ovarian serous cystadenocarcinoma (OV); Lung squamous cell carcinoma (LUSC); Lung adenocarcinoma (LUAD); Liver hepatocellular carcinoma (LIHC); Kidney PAN cancer (KIPAN); Acute Myeloid Leukemia (LAML); Head and Neck squamous cell carcinoma (HNSC); Uveal Melanoma (UVM); Esophageal carcinoma (ESCA); Colon and Rectum adenocarcinoma (COADREAD); Cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC); Breast invasive carcinoma (BRCA); Gliomas (GBM and LGG); Bladder Urothelial Carcinoma (BLCA); Cholangiocarcinoma (CHOL); Adrenocortical carcinoma (ACC).

Results

[0425] The relationship between CASD1, B4GALNT1, ST8SIA1, CERK, PIK3C2A gene expression and patient overall survival was investigated in the TCGA datasets using the SurvExpress online tool. Hazard ratios represent the probability of patient death, where a hazard ratio of two means that a patient from the high expression group has twice the probability of dying compared to a patient from the low expression group. We analyzed the impact of CASD1, B4GALNT1, ST8SIA1, CERK, PIK3C2A gene expression, alone or in combination, on patient survival.

[0426] As shown on FIG. 8A, high B4GALNT1 gene expression was correlated with poor prognosis in 11 cancer types out of 22 including uterine corpus endometrial carcinoma (UCEC), Lung squamous cell carcinoma (LUSC), lung adenocarcinoma (LUAD), kidney cancer (KIPAN), head and neck squamous cell carcinoma (HNSC), Uveal Melanoma (UVM), Cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC), gliomas (LGG), bladder urothelial carcinoma (BLCA), Cholangiocarcinoma (CHOL), and Adrenocortical carcinoma (ACC).

[0427] As shown on FIG. 8B, high ST8SIA1 gene expression was correlated with poor prognosis in 12 cancer types out of 21 which are Sarcoma (SARC), Prostate adenocarcinoma (PRAD), Pancreatic adenocarcinoma (PAAD), Ovarian serous cystadenocarcinoma (OV), Lung adenocarcinoma (LUAD), Kidney PAN cancer (KIPAN), Head and Neck squamous cell carcinoma (HNSC), Uveal melanoma (UVM), Colon and Rectum adenocarcinoma (COADREAD), Breast invasive carcinoma (BRCA), gliomas (LGG) and Bladder Urothelial Carcinoma (BLCA).

[0428] As shown on FIG. 8C, high CASD1 gene expression was associated with poorer survival in 17 cancer types out of 27 including sarcoma (SARC), lung adenocarcinoma (LUAD), colon and rectum adenocarcinoma (COADREAD), breast invasive carcinoma (BRCA), and glioblastoma (GBM and LGG). We analyzed CASD1 gene expression in combination with one or two glycosyltransferases encoding genes: B4GALNT1 and ST8SIA1.

[0429] Most cancer types kept a poor prognostic (14 of 17) when CASD1 and B4GALNT1 genes were both highly expressed (see FIG. 8D). Moreover, this combination induced an increase of the hazard ratio in 5 other cancer types, in which high gene expression of both CASD1 and B4GALNT1 was associated with poorer survival: uterine corpus endometrial carcinoma (UCEC), liver hepatocellular carcinoma (LIHC), lung squamous cell carcinoma (LUSC), head and neck squamous cell carcinoma (HNSC), and esophageal carcinoma (ESCA).

[0430] As shown on FIG. 8E, combination of high expression of both CASD1 and ST8SIA1 genes was associated with poorer outcome than high CASD1 gene expression alone in the same cancer types. High CASD1 and ST8SIA1 gene expression increased the hazard ratio in ovarian serous cystadenocarcinoma (OV), head and neck squamous cell carcinoma (HNSC) and esophageal carcinoma (ESCA).

[0431] High expression of the three CASD1, B4GALNT1 and ST8SIA genes was associated with poorer prognosis in 19 out of 27 TCGA datasets (see FIG. 8F).

[0432] As shown on FIG. 8G, high CERK gene expression was associated with poorer survival in 13 cancer types of 26 including sarcoma (SARC), uterine corpus endometrial carcinoma (UCEC), Skin Cutaneous Melanoma (SKCM), Pancreatic adenocarcinoma (PAAD), Ovarian serous cystadenocarcinoma (OV), Lung squamous cell carcinoma (LUSC), Liver hepatocellular carcinoma (LIHC), Kidney PAN cancer (KIPAN), Acute Myeloid Leukemia (LAML), Head and Neck squamous cell carcinoma (HNSC), Breast invasive carcinoma (BRCA), gliomas (LGG), and Bladder Urothelial Carcinoma (BLCA).

[0433] As shown on FIG. 8H, high expression of both CERK and B4GALNT1 genes worsened the prognostic of 4 cancer types: Lung adenocarcinoma (LUAD), uveal melanoma (UVM), Colon and Rectum adenocarcinoma (COADREAD), and adrenocortical carcinoma (ACC).

[0434] As shown on FIG. 8I, high expression of both CERK and ST8SIA1 genes was associated with poorer outcome in 18 cancer types. High CERK/ST8SIA1 gene expression increased the hazard ratio of prostate adenocarcinoma (PRAD), Lung adenocarcinoma (LUAD), uveal melanoma (UVM), Colon and Rectum adenocarcinoma (COADREAD) and endocervical adenocarcinoma (CESC). High expression of the 3 CERK/B4GALNT1/ST8SIA genes was associated with poorer prognosis in 20 out of 25 TCGA datasets (see FIG. 8J).

[0435] As shown on FIG. 8K, high PIK3C2A gene expression was associated with poorer survival in 11 cancer types out of 27.

[0436] High expression of PIK3C2A gene in combination with B4GALNT1 (FIG. 8L) worsened the prognostic of 6 cancer types: uterine corpus endometrial carcinoma (UCEC), thymoma (THYM), stomach adenocarcinoma (STAD), Lung adenocarcinoma (LUAD), head and neck squamous cell carcinoma (HNSC), esophageal carcinoma (ESCA) and adrenocortical carcinoma (ACC).

[0437] High expression of both PIK3C2A and ST8SIA1 genes was associated with poorer outcome in 13 cancer types (FIG. 8M). High PIK3C2A/ST8SIA1 gene expression increased the hazard ratio of lung adenocarcinoma (LUAD), head and neck squamous cell carcinoma (HNSC) and cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC).

[0438] High expression of the 3 genes PIK3C2A/B4GALNT1/S T8SIA was associated with poorer prognosis in 20 out of 27 TCGA datasets (FIG. 8N).

[0439] High CASD1 gene expression was correlated with poor survival in 17 TCGA datasets (FIG. 8C and FIG. 8O), high CERK gene expression in 13 TCGA datasets (FIG. 8G) and high PIK3C2A gene expression in 11 TCGA datasets (FIG. 8K).

[0440] As shown on FIG. 8P, high expression of both CASD1 and CERK genes was associated with poorer prognosis in 19 out of 26 TCGA datasets. High expression of both CASD1 and CERK worsened the prognostic of 6 cancer types, as compared to high expression of CASD1 alone: Pheochromocytoma and Paraganglioma (PCPG), Uterine Corpus Endometrial Carcinoma (UCEC), Ovarian serous cystadenocarcinoma (OV), Lung squamous cell carcinoma (LUSC), Liver hepatocellular carcinoma (LIHC), and Head and Neck squamous cell carcinoma (HNSC).

[0441] As shown on FIG. 8Q, high expression of both CASD1 and PIK3C2A genes was associated with poor prognosis in 21 out of 27 TCGA datasets. High expression of both CASD1 and PIK3C2A worsened the prognosis of 5 cancer types, as compared to high expression of CASD1 alone: Pheochromocytoma and Paraganglioma (PCPG), Uterine Corpus Endometrial Carcinoma (UCEC), Thyroid carcinoma (THCA), Ovarian serous cystadenocarcinoma (OV), and Lung squamous cell carcinoma (LUSC).

[0442] In conclusion, our data show that high expression level of CASD1 correlated with poor survival in many cancer types, and may thus be used as a prognostic biomarker in these cancers.

[0443] Moreover, a combined high expression level of 2 or 3 genes correlated with poorer survival in more cancer types than the individual genes, showing that the use of combined biomarkers is relevant as a prognostic tool.

Example 4: Effect of CERK Inhibition on GD2 O-Acetylation and Migratory Properties in Breast Cancer Cells Line MDA-MB231 GD3S+

Materials and Methods

Cell Culture

[0444] MDA-MB231 GD3S+cells were obtained as described in Cazet et al. (Biol Chem. 2009; 390(7):601-609). Cells were routinely grown in monolayer culture and maintained at 37? C. in an atmosphere of 5% CO.sub.2. Cells were grown in Dulbecco's modified Eagle's medium (DMEM, Lonza) supplemented with 10% heat-inactivated fetal calf serum, 2 mmol/L L-Glutamine.

siRNA Transfection

[0445] Transfections were performed with 10 ?M of siRNA control non-targeting (D001810-10-20, Horizon), siRNA targeting CASD1 (L-016926-01-0010, Horizon), or two different siRNA targeting CERK: CERK2 (D-004061-02-0010, Horizon), or CERK4 (D-004061-02-0010, Horizon) and 4 ?L of RNAimax (#137781, Thermo fischer Scientific) in 500 ?L of UltraMEM (Lonza). In 6-well plates, 150,000 cells were grown in 1,5 mL of DMEM and transfection mix. Cells were collected 72 hours after transfection for qPCR or immunocytochemistry experiments.

RNA Extraction, cDNA Synthesis and qPCR

[0446] Total RNA was extracted from cells using the Nucleospin RNA II kit (Macherey-Nagel, Germany). The amount of extracted RNA was quantified using a DeNovix DS-11 spectrophotometer (DeNovix Inc., USA) and the purity of the RNA was checked by the ratio of the absorbance at 260 and 280 nm. Total RNA was subjected to reverse transcription using the Maxima First Strand cDNA Synthesis Kit (ThermoFisher Scientific, Villeneuve d'Ascq, France) according to the protocol provided by the manufacturer. The oligonucleotide sequences (Eurogentec, Seraing, Belgium) used as primers for the PCR reactions are given in SEQ ID NO: 31 and SEQ ID NO: 32. qPCR and subsequent data analysis were performed using the AriaMx Quantitative System (Stratagene, La Jolla, CA, USA). PCR reaction (25 ?L) contained 12.5 ?L of the 2X Luna Master Mix (NEB), 300 nM of primers and 4 ?L of cDNA (1:40). DNA amplification was performed with the following thermal cycling profile: initial denaturation at 94? C. for 10 min, 40 cycles of amplification (denaturation at 94? C. for 20 s, annealing at Tm for 20 s, and extension at 60? C. for 30 s). Hypoxanthine-guanine PhosphoRibosylTransferase (HPRT) gene was used to normalize the expression of genes of interest (oligonucleotides sequences used as PCR primers reactions are given in SEQ ID NO: 54 and SEQ ID NO: 55). The fluorescence monitoring occurred at the end of each cycle. The analysis of amplification was performed using the AriaMx 6.0 software. For each primer pair, the specificity of the amplification was checked by recording the dissociation curves. All experiments were performed in triplicate. The quantification was performed by the method described by Pfaffl (Nucleic Acids Research, 2001; 29(9):e45).

CERK Inhibitor Treatment

[0447] Cells were seeded in 6-well plates (150,000 cells/well) with coverslips in 2 mL of DMEM medium. After 24 hours, the medium was replaced with DMEM containing 1 ?M of the CERK inhibitor NVP231 (SigmaN9289). After 2 h, 4 h, 6 h, 8 h, 20 h, 24 h or 48 h of CERK inhibitor treatment, cells were collected for immunocytochemistry experiments.

Immunocytochemistry and Confocal Microscopy

[0448] Transfected cells were grown on glass coverslips and were fixed for 20 min in 4% paraformaldehyde. Cells were washed three times with PBS 1X and membrane permeabilization was performed in 5 ?g/mL digitonine in PBS 1X for 20 min. After 3 washes, cells were saturated in PBS 1X-BSA 0.5% blocking buffer. Coverslips were transferred in humid chamber and cells were incubated 2 hours with anti-OAcGD2 monoclonal antibody 8B6 mouse IgG3 (OGD2 Pharma, Nantes, France) at 20 ?g/mL. Cells were washed three times with PBS 1X-BSA 0.5% and incubated 1 hour with secondary antibody Alexa Fluor 488 donkey anti-mouse IgG at 3 ?g/mL (Invitrogen). After 3 washes, cells were incubated with 1 ?g/mL of DAPI (Sigma, #D9542) for 7 min. Coverslips were mounted in fluorescent mounting medium (Dako). Coverslips were observed under the A1 Nikon confocal microscope with a 60X oil immersion objective. The green fluorescence was acquired with ?ex=488 nm and ?em=500-530 nm, DAPI with ?ex=350 nm and ?em=460 nm. Images were processed with ImageJ and backgrounds generated by secondary antibody alone were deducted. Mean fluorescence intensity was calculated with macro using ImageJ.

Transwell Assays

[0449] Cells were seeded in 12-well plates containing migration chamber in serum-free medium. Below the chamber, medium with serum was added into the wells in the presence or absence of 1 ?M of NVP231. After 24 hours incubation at 37? C., wells were fixed with 4% paraformaldehyde and non-migratory cells were swapped with cotton swabs. Cells were washed three times and nuclei were stained with DAPI. Membranes were cut out and mounted between glass slide and coverslip with fluorescent mounting medium. Nuclei were counted using A1 Nikon confocal microscopy.

Results

[0450] In order to study the effect of CERK inhibition on O-acetylation of the GD2 ganglioside expression and on the migration capacity of the breast cancer cell line MDA-MB231 GD3S+, two strategies were used: CERK inhibition using a CERK inhibitory and using siRNA directed against CERK mRNA.

Effect of a CERK Inhibitor

[0451] Breast cancer cells MDA-MB-231 GD3S+, overexpressing the GD3 synthase, were treated with a CERK inhibitor for 2, 4, 6, 8, 20, 24 or 48 hours. At the end of the culture, cells were stained by immunocytochemistry using an antibody specifically recognizing the O-acetylated GD2 ganglioside, and the mean fluorescence intensity was measured by confocal microscopy. As shown on FIG. 9, CERK inhibition induced a transitory overexpression of OAcGD2, with a gradual increase of OAcGD2 expression from 2 to 8 hours, followed by a gradual decrease after 8 hours of treatment.

[0452] The effect of CERK inhibition on the migration capacity of the cells was also evaluated. Breast cancer cells MDA-MB-231 GD3S+were cultured in plates containing migration chamber in the presence or absence of the CERK inhibitor. After 24 hours of culture, cells were counted to evaluate migration. As shown on FIG. 10, treatment with the CERK inhibitor had no effect on the migration capacity of the cells. The number of cells counted in the bottom chamber of the Transwell plate was similar between cells treated with the CERK inhibitor or untreated.

[0453] These data demonstrate that CERK inhibition with the CERK inhibitor increased OAcGD2 expression in MDA-MB-231 GD3S+cells, but did not increase the migration capacity of the cells.

Effect of a siRNA Targeting CERK

[0454] Breast cancer cells MDA-MB-231 GD3S+were transiently transfected using either a control siRNA (siControl), a siRNA directed against CASD1 mRNA (siCASD1) or a siRNA directed against CERK mRNA (siCERK2 or siCERK4).

[0455] First, the efficiency of the siRNA directed against CERK mRNA (siCERK2 and siCERK4) was evaluated by quantification of CERK mRNA expression by qPCR. As shown on FIG. 11, both CERK siRNA, siCERK2 and siCERK4, were able to decrease the expression level of CERK mRNA as compared to the control siRNA (siControl).

[0456] Then, the expression of the O-acetylated GD2 ganglioside was measured on the cells by immunocytochemistry using an antibody specifically recognizing the O-acetylated GD2 ganglioside. Cells were analyzed by confocal microscopy. FIGS. 12A-D show representative confocal microscopy photographs of cells transfected with the following siRNA: control siRNA (FIG. 12A), siRNA CASD1 (FIG. 12B), siRNA CERK2 (FIG. 12C) and siRNA CERK4 (FIG. 12D). Mean fluorescence intensity was also measured for each condition and is plotted on FIG. 13. As shown on the images and by quantification, inhibition of CASD1 mRNA using a siRNA directed against CASD1 mRNA (FIGS. 12B and 13) decreased OAcGD2 expression, as compared to the control siRNA (siControl, FIGS. 12A and 13). On the other hand, inhibition of CERK mRNA using a siRNA directed against CERK mRNA (siCERK2, FIGS. 12C and 13, and siCERK4, FIGS. 12D and 13) increased the expression of the O-acetylated GD2 ganglioside, as compared to the control siRNA (siControl, FIGS. 12A and 13).

[0457] Additionally, the migration capacity of the cells transiently transfected with the siRNA was evaluated. As shown on FIG. 14, both CERK siRNA (siCERK2 and siCERK4) induced a significant decrease of the cell migration capacity, as compared to the control siRNA (siControl), while CASD1 siRNA (siCASD1) did not significantly impact cell migration.

[0458] These data demonstrate that CERK inhibition using CERK siRNA induced an increase of the O-acetylated GD2 ganglioside expression in breast cancer cell line MDA-MB231 GD3S+, but decreased the migratory capacity of these cells.