Generation of a monoclonal antibody recognizing the CEACAM glycan structure and inhibiting adhesion using cancer tissue-originated spheroid as an antigen.

Spheroids cultured directly from tumours can better reflect in vivo tumour characteristics than two-dimensional monolayer culture or three-dimensional culture of established cell lines. In this study, we generated antibodies by directly immunizing mice with primary-cultured living spheroids from human colorectal cancer. We performed phenotypic screening via recognition of the surface of the spheroids and inhibition of their adhesion to extracellular matrices to identify a monoclonal antibody, clone 5G2. The antibody inhibited cell migration in two-dimensional culture and promoted cell detachment. Western blotting and immunohistochemistry detected the 5G2 signal in many colorectal cancer spheroids, as well as patient tumours, but failed to detect in various cell lines examined. We found that 5G2 recognized the Le(a) and Le(c) on N-glycan, and their major carrier proteins were CEACAM5 and CEACAM6. Pre-incubation of the spheroids with 5G2 impaired translocation of integrin β4 from the lateral membrane to the contact interface between the extracellular matrix when embedded in it. As we successfully obtained a functional antibody, which antigen was glycan structures and lost in cell lines, cancer tissue-originated spheroids can be a useful antigen for generating novel anti-cancer antibodies.

4˚C. The supernatants (sup.1) were centrifuged at 10,000!g for 10min at 4˚C, and the resultant supernatants were further centrifuged at 105,000!g for 60min at 4˚C. The pellets were re-suspended to suspension buffer (50mM Tris-HCl, pH 7.4 with cOmplete, EDTA free) to make membrane rich fraction and subjected to glycosidase digestion analysis.

IP-MAS analysis
Sup.1 described above were centrifuged at 105,000!g for 1h at 4˚C, and the pellets were re-suspended to suspension buffer with Triton (20mM Tris-HCl, pH 8.0 with cOmplete, EDTA free proteinase inhibitor, 1% Triton) and incubated at 4˚C. Solubilized proteins were collected as supernatant of centrifugation (14,000 rpm for 15min at 4˚C), and diluted to five times with 20mM Tris-HCl, pH8.4. Unbound fraction to anion exchange resin (Bio-Rad Lab. Inc., Hercules, CA) was collected, and pre-cleared by incubation with mouse IgG3 kappa and Protein G sepharose (GE Healthcare), then incubated with 5µg/ml of 5G2 mAb at 4˚C for overnight. Immunoprecipitate was collected by incubation with protein G sepharose, and subjected to SDS-PAGE after boiling in SDS-PAGE sample buffer. The gel was stained with Simple Blue Safe Stain (Invitrogen), cut out by molecular weight upper than 75kDa, and applied to in-gel digestion 1 and LC-MS/MS analysis. Digested peptides were analyzed using an LTQ-Orbitrap XL mass spectrometer (Thermo fisher scientific) 2 . The nano LC gradient is delivered at 500 nL/min and consists of a linear gradient of Buffer B developed from 5 to 35% B in 45 min. A spray voltage of 2000 V is applied. MS data was searched against human uniprot database using Proteome Discoverer 1.3 and Mascot v2.4. The precursor mass tolerance is set to 7 ppm and a fragment ion mass tolerance is set to 0.6 Da for CID. The search parameters allow for one missed cleavage for trypsin, fixed modifications (carbamidomethylation at cysteine), and variable modifications (oxidation at methionine, acetylation at protein N-terminal). Peptides identified at a threshold with 1% FDR are accepted. Candidates of antigen proteins were selected, which detected 8 times more than HCT116 derived samples, and not detected in IgG control (Table S2).

Microarray analysis
Total RNA was extracted from CTOSs or tumours with TRIzol reagent (Life Technologies) according to the manufacturer's instructions. One microgram of total RNA was reverse-transcribed to obtain cDNA using Superscript III (Life Technologies).
Microarray hybridizations were performed at Hokkaido System Science (Sapporo, Japan) using SurePrint G3 Human GE 8x60K Ver.1 (Agilent Technologies, Santa Clara, CA). The microarray slides were scanned and the gene expression profiles were analyzed at Hokkaido System Science according to the manufacturer's protocol. Each dataset was standardized using the z-score transformation method. Microarray data can be viewed using the NCBI Gene Expression Omnibus (GEO). For hierarchical clustering analysis, microarray data for glycosylation-related enzyme genes were picked up and hierarchical clustering performed with Pearson correlation using average linkage clustering after log 2 -transformation by MeV.  List of proteins detected by mass spectrometry from 5G2 immunoprecipitated samples along with cDNA microarray data.

Supplementary Table S2
5G2 C45 !score: total score of each proteins detected by mass spectrometry in immunoprecipitated samples with 5G2 mAb from floating cultured C45 CTOSs. MS fold change C45/HCT116: the ratio of total score of immunprecipitated samples with 5G2 mAb from C45 to that from HCT116. S indicates proteins detected specifically in C45. MA C45/HT29: the ratio of expression levels of mRNA coding each protein in floating cultured C45 to HT29 cells analyzed by cDNA microarray analysis.