Article pubs.acs.org/jpr

Characterization of Binding Epitopes of CA125 Monoclonal Antibodies Lara Marcos-Silva,†,‡,§ Yoshiki Narimatsu,†,∥ Adnan Halim,† Diana Campos,†,‡,§ Zhang Yang,†,∥ Mads A. Tarp,†,# Pedro J. B. Pereira,⊥ Ulla Mandel,† Eric P. Bennett,† Sergey Y. Vakhrushev,† Steven B. Levery,†,▽ Leonor David,*,‡,§ and Henrik Clausen*,†,∥ †

Copenhagen Center for Glycomics, Department of Cellular and Molecular Medicine and School of Dentistry, Faculty of Health Sciences, University of Copenhagen, Blegdamsvej 3, DK-2200 Copenhagen N, Denmark ‡ IPATIMUP, Institute of Molecular Pathology and Immunology of the University of Porto, Rua Dr. Roberto Frias s/n, 4200-465 Porto, Portugal § Faculty of Medicine of the University of Porto, Al. Prof. Hernâni Monteiro, 4200-319 Porto, Portugal ∥ The Novo Nordisk Foundation Center for Biosustainability, Technical University of Denmark, Kogle Alle 6, 2970 Hørsholm, Denmark ⊥ IBMC, Instituto de Biologia Molecular e Celular, Universidade do Porto, Rua do Campo Alegre 823, 4150-180 Porto, Portugal S Supporting Information *

ABSTRACT: The most used cancer serum biomarker is the CA125 immunoassay for ovarian cancer that detects the mucin glycoprotein MUC16. Several monoclonal antibodies (mAbs) including OC125 and M11 are used in CA125 assays. However, despite considerable efforts, our knowledge of the molecular characteristics of the recognized epitopes and the role played by glycosylation has remained elusive. Here a comprehensive set of recombinant MUC16 tandem repeats (TRs) expressed in glycoengineered mammalian cells and E. coli, together with overlapping peptides, was used to probe antigen-binding epitopes. We present a complete analysis of N- and O-glycosylation sites of a MUC16 TR expressed in CHO cells and demonstrate that neither N- nor O-glycosylation appear to substantially influence binding of OC125 and M11 mAbs. A series of successive N- and C-terminal truncations of a MUC16 TR construct expressed in E. coli narrowed down the epitopes for OC125 and M11 to a segment containing parts of two consecutive SEA domains with a linker. Thus, a complete SEA domain is not required. These findings suggest that binding epitopes of mAbs OC125 and M11 are dependent on conformation but not on glycosylation. The availability of recombinant TR constructs with and without aberrant glycosylation now opens the way for vaccine studies. KEYWORDS: cancer vaccine, mass spectrometry, ETD, GALNT

■

INTRODUCTION

The CA125 ovarian cancer-associated serum antigen was initially reported in 1981 by Bast and colleagues,2 and not until early 2000 were its molecular characteristics defined by partial cloning of the large gene encoding a mucin-like glycoprotein designated MUC16.3,4 This protein belongs to the family of tethered mucins and is heavily glycosylated with both O- and N-glycans, although actual sites of glycan attachment remain unknown.5,6 MUC16 is a type-I transmembrane protein and consists of a very large extracellular domain with a N-terminal region of over 12 000 residues without TRs or apparent structure but with high density of PST residues that are predicted to be heavily O-glycosylated,7,8 a C-terminal region

Ovarian cancer is the seventh leading cause of cancer-related death among women, and because of the absence of clinical symptoms in early stages, patients are often diagnosed late and have poor survival.1 Currently, the best biomarker for ovarian cancer is CA125, which shows elevated levels at time of diagnosis in ∼80% of cases. CA125 is approved for monitoring progression and treatment response of ovarian cancer patients but not for diagnostic purposes due to a relatively high frequency of elevated levels in benign disorders. Nevertheless, CA125 is one of the most used cancer biomarkers today. Surprisingly, while a number of antibodies have been produced to MUC16 and are used in different CA125 clinical assays, the molecular epitopes recognized by these monoclonal antibodies (mAbs) are poorly defined. © XXXX American Chemical Society

Received: March 4, 2014

A

dx.doi.org/10.1021/pr500215g | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

We also tried to narrow down the binding epitope further using a large panel of synthetic peptides and sequential truncations of E. coli expressed MUC16 TR constructs, and while we were unable to identify a binding peptide, we did demonstrate that the binding of CA125 mAbs does not require a complete folded SEA domain per se, as recently reported.24

consisting of approximately 60 TRs of around 156 residues, and a small apparent degenerate repeat sequence linked to a short transmembrane stretch.9 The intracellular portion is a short cytoplasmic sequence with 35 residues with potential phosphorylation sites, which is known to interact with the actin cytoskeleton through binding to the ezrin/radixin/moesin (ERM) family of proteins.10 The MUC16 TRs consist of a SEA domain of ∼120 amino acids with low PST density, two cysteine residues, and three potential N-glycosylation sites9 and also appear to have a flanking region of ∼30 residues with high density of PST residues that may be O-glycosylated. SEA domains are found in sea urchin sperm protein, enterokinase, and agrin but have also been found in, for example, the MUC1 cell membrane mucin.11 The SEA domain in the murine MUC16 folds into a α/β sandwich, with a four-stranded antiparallel β-sheet and a short α-helix on one side and two αhelices and a two-stranded short β-sheet on the other.12 The SEA domains have been proposed to function in diverse recognition events, and N-glycosylation in the SEA domain is reported to be essential for the interaction of MUC16 with mesothelin.13,14 The CA125 serum assay relies on immunodetection of circulating MUC16, presumably shed from cells by regulated proteolytic cleavage in the stem region of the membrane-bound glycoprotein.3 Most CA125 assays use two monoclonal antibodies, OC125 and M11.2,15 The original OC125 mAb used to identify the CA125 antigen was obtained following immunization of mice with an ovarian cancer cell line (OVCA433),16 while mAb M11 was obtained by immunizing mice with partially purified CA125 antigen derived from ascitic fluid.17,18 Analysis of a larger group of mAbs directed to CA125 has demonstrated that these all react with the TR region and can be classified into three groups using competition assays: (i) OC125-like (group A); (ii) M11-like (group B); and (iii) OV197-like (group C).19,20 However, a more detailed analysis of the epitopes of these mAbs is missing. Originally, Lloyd and colleagues21 used a rabbit polyclonal antibody to isolate the first cDNA fragment coding for the stem and a few tandem repeats of MUC16, and in a later study they demonstrated that expressing this cDNA in mammalian cells led to reactivity of mAbs OC125 and M11.4 mAbs OC125, M11, and most other antibodies prepared against ovarian cancer cell extracts were long believed to be directed at complex, glycosylationdependent, and/or conformational-dependent epitopes in the TRs.3,19,22,23 During the preparation of this manuscript, an elegant study demonstrated that several of these mAbs indeed react with MUC16 TRs expressed in E. coli by SDS-PAGE Western blotting and that the minimum epitopes appeared to be confined to a 128 amino acid segment encompassing a complete SEA domain because further truncation abolished reactivity.24 While this study clearly shows that CA125 mAbs can bind nonglycosylated MUC16 TRs, it is still unclear to what extent glycosylation plays a role in binding and why binding requires an entire TR. In the present study, we sought to characterize sites of Nand O-glycosylation of a MUC16 TR (TR number 5) and analyze the role of glycosylation for binding of MUC16 mAbs. We identified four N-glycosylation and 19 O-glycosylation sites in a 1.2 TR construct expressed in CHO cells. Comparative analysis of the same construct expressed in CHO cells and in E. coli showed that there was no substantial difference in mAb binding, suggesting that glycosylation does not significantly influence binding affinities of at least M11 and OC125 mAbs.

■

MATERIAL AND METHODS

Production of Recombinant MUC16 1.2TR in CHO SimpleCells

We used CHO SimpleCells (SCs) with zinc finger nuclease (ZFN) targeted knockout of the cosmc gene to express MUC16 TRs with truncated simple O-glycans. The establishment of the CHO SC cell line with cosmc knockout and deficient core 1 enzyme activity was performed essentially as previously8 and will be described in detail elsewhere. Cells were stably transfected with a MUC16 1.2TR construct encoding the entire TR number 5 (protein accession NP_078966.2, residues 12 665−12 858) with the 124 amino acid SEA domain flanked by sequences with high PST residue density from TR number 4 on one side (IPVPTSSTPGTSTVDLGSGTPSSLP) and TR number 6 on the other (SPTSAGPLLVPFT). A synthetic codon-optimized MUC16 1.2TR gene construct (Genewiz) including nucleotides 38 195−38 778 of GenBank entry NM_024690.2 was inserted downstream of the previously described pcDNA3-γ-interferon vector.25 The gene construct included a C-terminal myc-tag and N- and C-terminal 6xHis tags. Cells were grown in EX CELL CD CHO Fusion medium (Sigma) supplemented with 2% glutamine. Transfection was performed with 3 μg of pcDNA3-γ-MUC16 1.2TR construct using the Nucleofector IITM and the human MSC NucleofectorKitV and U-024 program (Amaxa Biosystems). Cells were seeded into 96-well plates as 0.7 cells per well in 200 μL of plating medium containing 80% EX CELL CHO Cloning Medium (Sigma) and 20% EX CELL CD CHO Fusion medium (Sigma) with 2% glutamine and 0.32 mg/mL G418 (Invitrogen). Positive clones were selected by screening with anti-His mAb (His-probe, Santa Cruz Biotechnology) or antiMUC16 mAb clone M11 (IgG1, Dako). The secreted Histagged MUC16 1.2TR was purified by affinity chromatography on nickel-nitrilotriacetic acid (Ni-NTA) agarose (Qiagen), followed by reversed-phase chromatography on a Jupiter 5 μm C4 300 Å column (Phenomenex) with an 1100 HewlettPackard system (Avondale, PA) using 0.1% trifluoroacetic acid (TFA) and a linear gradient of 10−100% acetonitrile. Production of Recombinant MUC16 TR Fragments in E. coli

N- or C-terminally 6xHis-tagged recombinant fragments of MUC16 were produced in E. coli. A synthetic codon-optimized MUC16 1.2TR gene construct (Eurofins MWG Operon), including nucleotides 38 195−38 778 of NM_024690.2, was inserted into the BamHI/NotI site of pET28 (Novagen). The repeat fragment from MUC16 1.7TR was generated by PCR, including nucleotides 37 978−38 770 of GenBank accession XR_079401.1, and was inserted into the BamHI/NotI site of pET28. A series of truncated MUC16 TR constructs 1/4, 5/2, 5/7, 5/8, 5/9, 5/10, and 11/2 were generated by PCR of the MUC16 1.7TR-containing pET28. K292 and K293 were produced as synthetic DNA constructs (Eurofins MWG Operon) and inserted into pET28. The MUC16 F/R construct was generated by insertion of the two reannealed ssDNAoligo’s muc16F/R (Table S1 in the Supporting Information) into the BamHI/NotI site of pET28. The sequences of all constructs B

dx.doi.org/10.1021/pr500215g | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

9.6). Plates were then blocked in PLI-P buffer pH 7.4 for 1 h at RT, followed by incubation with 1 μg/mL of mAbs OC125 or M11 for 1 h at RT. Bound antibodies were detected with HRPconjugated goat antimouse IgG antibody (Southern Biotech). Plates were developed with TMB+ substrate (Dako), reactions were stopped with 0.5 M H2SO4, and absorbance was read at 450 nm. Recombinant MUC16 proteins (1.7TR; 1.2TR; 5/2; 5/7) and a 33 amino acid (aa) peptide based on human MUC2 were used as controls.

generated were verified using BigDye chemistry and ABI Avant3100 instrumentation (Life Technologies). Proteins were expressed in Rosetta 2(DE3) cells (Novagen) with IPTG induction and purified under denaturing conditions using NiNTA agarose (Qiagen) as previously reported.26 Release of N- and O-Linked Glycans from Recombinant MUC16 1.2TR

PNGase F digestion was performed in denaturing conditions according to the manufacturer’s instructions using 100 μg of purified MUC16 1.2TR produced in CHO SC and 3000 U endoglycosidase (New England Biolabs), followed by reversedphase purification, as previously described. As control, an equivalent amount of glycoprotein was subjected to the same conditions but without the addition of PNGase F enzyme. For the release of GalNAc O-glycans, recombinant purified Elizabethkingia meningosepticum α-GalNAcase27 was used in a 40 μL reaction mixture at neutral pH containing 10 μg of MUC16 1.2TR and 15 μg of enzyme. Reactions were incubated at room temperature (RT) for 17 h. Subsequently, samples were reduced with 25 mM dithiothreitol in 50 mM NH4HCO3 and 1 M urea for 30 min at 60 °C and then alkylated with 75 mM iodoacetamide in 50 mM NH4HCO3 for 30 min in the dark. Next, samples were subjected to electrophoresis on Novex Tris-glycine 18% gels (Invitrogen) and blotted onto nitrocellulose, followed by incubation in 5% BSA in Tris-buffered saline (TBS) plus 0.05% Tween 20. Membranes were then incubated with primary antibody (mouse anti-CA125 M11 clone and OC125 clone (IgG1, Cell Marque)), biotinylated Vicia villosa lectin (VVA-B; Vector Laboratories), or negative control mouse IgG1 (Dako) overnight at 4 °C. Blots were developed with 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium substrate after incubation with secondary alkaline phosphatase-conjugated antibodies for 1 h at RT.

Mass Spectrometry Analysis

Prior to proteolytic digestion, purified MUC16 1.2TR produced in CHO SC was enzymatically de-N-glycosylated. To distinguish between natural deamidation of the asparagine residue and deamidation caused by PNGase F, the enzymatic reaction was performed in 18O water. For this purpose, 20 μg of the protein was dried and resuspended in H218O (95% 18O, Sigma), reduced and alkylated as previously described, treated with 0.5 U of PNGase F (Roche), and finally digested using either trypsin or endoproteinase Lys-C (1 μg; Roche). The digestion reaction was quenched with TFA and purified using homemade Stage Tips (Empore disk-C18, 3M). All samples were analyzed by mass spectrometry using a system composed of an EASY-nLC 1000 (Thermo Fisher Scientific) interfaced via a nanospray Flex ion source to an LTQ-OrbitrapVelos Pro hybrid spectrometer (Thermo Fisher Scientific), equipped for both HCD- and ETD-MS2 modes. nLC was operated in a single analytical column setup using PicoFrit Emitters (New Objectives, 75 μm inner diameter) packed in-house with Reprosil-Pure-AQ C18 phase (Dr. Maisch; 3 μm particle size, 19 cm column length) with a flow rate of 200 nL min−1 using a linear gradient from 5 to 30% buffer B in either 10 or 30 min and from 30 to 80% buffer B in 5 min, followed by isocratic elution at 80% buffer B for 10 min, where buffer A is 0.1% formic acid (FA) in water and buffer B is 0.1% FA in acetonitrile. The nanospray Flex ion source was operated at a spray voltage of 2.1 kV and the heated capillary was operated at 275 °C. A data-dependent precursor selected acquisition was used. A precursor MS1 scan (m/z 350−1700) of intact peptides was acquired in the Orbitrap at a nominal resolution setting of 30 000, followed by Orbitrap HCD-MS2 (m/z 100−2000, nominal resolution 15 000) of the three most abundant multiply charged precursors above 5000 counts in the MS1 spectrum or ETD-MS2 (m/z 100−2000, nominal resolution 15 000) (“top three” method). Activation times were 30 and 100−200 ms for HCD and ETD fragmentation, respectively; isolation width was 4 mass units, and usually 1 microscan was collected for each spectrum. Automatic gain control targets were 100 000 ions for Orbitrap MS1 and 10 000 for MS2 scans, and the automatic gain control for fluoranthene ion used for ETD was 300 000. Supplemental activation (20%) of the charge-reduced species was used in the ETD analysis to improve fragmentation. Dynamic exclusion for 30 s was used to prevent repeated analysis of the same components. Polysiloxane ion at m/z 445.12003 was used as a lock mass in all runs. The raw data were processed, in a manner similar to previous publications,28,29 using Proteome Discoverer 1.4 software (Thermo Fisher Scientific) and searched against MUC16 TR FASTA sequence file. HCD and ETD data were searched using the SEQUEST node in PD 1.4 (algorithm version 28 build 60; node version 1.13). In all cases, the precursor mass tolerance was set to 10 ppm and fragment ion mass tolerance to 50

ELISA

An initial titration was performed to determine the optimal dilution (A450nm ≈ 1) of OVCAR-3 wt, and SC8 harvested and cleared spent growth medium after 40 h growth. Plates were coated overnight at 4 °C with 1 μg/mL mAb M11 in a carbonate−bicarbonate buffer (pH 9.6) and blocked in PLI-P (PO4, Na/K, 1% Triton, 1% BSA) buffer pH 7.4 for 1 h at RT. Following washing, wells were incubated with the respective serially diluted supernatants for 1 h at RT. Binding of native shed MUC16 derived from human cancer cell lines was detected by incubating wells with 1 μg/mL biotinylated X52 anti-MUC16 (Thermo Scientific) for 1 h at RT. Subsequently, wells were incubated with streptavidin-HRP for 1 h at RT and visualized by incubating with TMB+ (Dako). Development was stopped by the addition of 0.5 M H2SO4 and absorbance read at 450 nm. Inhibition ELISA was performed as previously described. In brief, after coating with M11 and blocking, wells were incubated with each recombinant MUC16 1.2TR serially diluted from an initial concentration of 2 μg/mL and incubated for 1 h at RT. OVCAR3 wt and SC supernatants were then diluted (1:8) in PLI-P buffer and incubated for 1 h at RT. The rest of the protocol was performed as described in the preceding paragraph. As controls, each isolated component of the inhibition ELISA was substituted by PLI-P buffer, while all others were used solely at the maximum concentration. Peptides and purified recombinant proteins were serially diluted from 1 μg/mL and used to coat MaxiSorp plates (Nunc) overnight at 4 °C in carbonate−bicarbonate buffer (pH C

dx.doi.org/10.1021/pr500215g | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Figure 1. Human MUC16. (A) Schematic depiction of the domain structure of MUC16. TR, Tandem repeat; TM, transmembrane domain; CT, cytoplasmic tail. (B) Detail of TR5 with its SEA domain and C-linker region as well as upstream linker region of TR4 and downstream SEA domain fragment of following TR6. Identified N- and O-glycosites found in the MUC16 1.2TR produced by CHO SC are illustrated with blue trees and lollipops, respectively. Identified O-glycosites with unambiguous position assignment and those with ambiguous assignment are marked with lollipops in yellow and white, respectively. Dotted arrow lines underneath mark the peptide sequence within which ambiguously assigned Oglycosites were identified. Green lollipops mark predicted O-glycosites according to sequence alignments (Figure S1 in the Supporting Information). Vertical dotted lines mark protein coverage (spanning residues Thr12,702-Thr12,858). (C) Homology models for two consecutive SEA domains of human MUC16 (NP_078966.2) were generated independently using the SWISS-MODEL server45 and the 3-D coordinates of a murine SEA domain as template (PDB entry 1IVZ12). Linker segments connecting the two SEA domains or preceding the first of these were modeled as random coils using the BAX Group PDB Utilities Server (http://spin.niddk.nih.gov/bax/nmrserver/pdbutil/ext.html). The four individual elements were positioned manually with COOT46 to minimize spatial clashes, and the geometry of the resulting model was regularized with PHENIX.47 The complete model includes the sequence spanning residues 12 660 to 12 969. The N-terminal linker region in TR4 (Residues 12 660 to 12 689) and the C-terminal linker region in TR5 (12 814−12 845) are marked in red and orange, respectively. The SEA5 (12 690−12 813) and SEA6 domains (12 846−12 969) are shown in salmon and wheat, respectively. The four N-glycosylation sites are highlighted in cyan, and the eight unambiguously identified O-glycosites are marked in yellow. Putative O-glycosites are shown in green. The model was prepared with PyMOL (http://pymol.org).

millimass units. Carbamidomethylation on cysteine residues was used as a fixed modification; methionine oxidation and HexNAc attachment to serines and threonines were used as variable modifications; a maximum of 12 variable modifications were allowed per peptide.

■

The secreted His-tagged 1.2TR was purified by nickel chromatography, followed by reversed-phase chromatography. For simultaneous identification of N- and O-glycosylation sites, purified MUC16 1.2TR was treated with PNGase F in the presence of 18O water, digested with trypsin or endoproteinase Lys-C, and analyzed by nLC−MS/MS, which enabled essentially almost complete sequence coverage with the exception of the N-terminal His-tag and the adjacent sequence covering the N-terminal linker region of TR4 (Figure 1B). Applying HCD and ETD fragmentation techniques, a total of 19 O-glycosites were identified, where 8 were unambiguously assigned to short sequence stretches by ETD MS/MS (Figure 1B and Table S2 in the Supporting Information). A total of 15 O-glycosites were located in TR5, where 5 were located within the SEA domain and 10 were found in the C-terminal putative linker region. In the flanking region of TR6, four O-glycosites were identified as well. Four canonical N-glycosites were identified and located in TR5, three of which were in the SEA domain. In some cases where the quality of ETD spectra was not sufficient for unambiguous site assignment, HCD MS/MS was used to define the sequence of glycopeptides and numbers of GalNAc residues attached. Distribution of the cleavage sites across the MUC16 1.2TR construct accessible for proteolysis by trypsin and endoproteinase Lys-C revealed that part of the construct could only be covered by large glycopeptides (up to 52-mer), where the degree of glycosylation is in the range from 5 to 14. Partial construct degradation resulting in shorter peptides enabled us to detect some glycosites of this stretch unambiguously.

RESULTS

Expression and Characterization of a MUC16 1.2TR Produced in CHO SimpleCells

The TRs of MUC16 are predicted to contain both N- and Oglycosylation, but actual sites of glycosylation have not been previously characterized. We chose to express a complete TR (TR5) with putative N- and C-terminal flanking sequences from TR4/6 (MUC16 1.2TR) in CHO SC cells to identify sites of O-glycosylation (Figure 1B and Figure S1 in the Supporting Information). CHO SC cells produce truncated O-glycans that consist only of the first GalNAc residue, which simplifies isolation and identification of O-glycosites.7,8,28 A CHO SC clone (2D7) secreting the MUC16 1.2TR construct was isolated by selection and limiting dilution. The clone reacted by immunocytology with anti-CA125 mAbs, M11, and OC125 but not with another anti-MUC16 mAb X52 (Figure S2 in the Supporting Information). X52 is a M11-like antibody (group B) that reacts with CA125 derived from cervical mucus and ascitic fluid.30 However, it did not react with our 1.2TRcontaining TR5 (Figure S2 in the Supporting Information), which is in agreement with a recent study demonstrating that this mAb shows differential reactivity with different MUC16 TRs.24 D

dx.doi.org/10.1021/pr500215g | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Figure 2. Binding of mAbs OC125 and M11 was not affected by N- and O-linked glycosylation of MUC16. Purified recombinant MUC16 1.2TR was treated with PNGase F and α-GalNAcase, as indicated, and analyzed by Western blot analysis. (A) NuPAGE Bis-Tris analysis of digestion visualized by Coomassie staining as well as Western blotting with mAb OC125. (B) SDS-PAGE Tris-glycine analysis of digestion analyzed by Western blotting with mAb M11 and VVA-B lectin. Asterisk indicates the α-N-acetylgalactosaminidase. Control included E. coli produced 1.2TR.

MUC16 mAbs M11 and OC125 Are Not Affected by N- and O-Glycosylation

apparent difference in capture of MUC16 with and without truncated O-glycans.

To evaluate the role of glycosylation for binding of CA125 mAbs,we expressed similar MUC16 1.2TR constructs in CHO SC and in E. coli. The purified CHO expressed glycoprotein migrated as a diffuse band in the range of 50 kDa on SDSPAGE, while the nonglycosylated protein produced in E. coli migrated as a single band of ∼28 kDa (Figure S3 in the Supporting Information). Enzymatic deglycosylation of the CHO expressed 1.2TR with PNGase F or α-GalNAcase to remove N-glycans and GalNAc O-glycans, respectively, resulted in substantial shifts in migration with PNGase F and αGalNAcase, equally enhancing mobility corresponding to an apparent loss of ∼9 kDa, while the combination further enhanced mobility, corresponding to the loss of an additional 9 kDa (Figure 2). Treatment with α-GalNAcase abolished reactivity with the α-GalNAc binding lectin VVA (Figure 2B). The band with lowest molecular mass after digestion with only PNGase F or with both enzymes migrates closely to the recombinant protein produced in E. coli and was not detected by VVA staining. Western blot analysis with mAb OC125 showed comparable reactivity with glycosylated and partially deglycosylated CHO expressed 1.2TR as well as with the construct expressed in E. coli (Figure 2A). Similar results were also observed with M11 (Figure 2B), suggesting that binding of the antibodies was not substantially influenced by glycosylation. We further sought to confirm this by developing a competitive inhibition ELISA assay enabling better quantification (Figure 3). We first semiquantified all MUC16 constructs by SDS-PAGE with Coomassie blue staining (Figure S3 in the Supporting Information). An initial titration was also performed to determine optimal dilution of the spent media of OVCAR3 wt and SC, which was 1:8 (A450nm ≈ 1; Figure 3A). Other cell lines were tested negative (data not shown). The assay was designed to evaluate the capacity of mAb M11 to capture shed endogenous MUC16 with and without truncated O-glycosylation produced in the OVCAR3 cell line in competition with increasing concentrations of purified recombinant 1.2TRs. In this way, we could evaluate the inhibitory capacity of a glycosylated recombinant TR in direct comparison with the unglycosylated version. Because mAb X52 did not react with the recombinant 1.2TRs, we used this mAb to probe the captured endogenous MUC16 (Figure 3B,C). We also tested mAb X306 (OC125-like), and this exhibited poorer binding capacity toward endogenous MUC16 (data not shown). This assay showed that the 1.2TR glycoprotein produced in CHO cells was approximately two times more effective in inhibiting the binding of M11 to native MUC16 than the nonglycosylated 1.2TR. Furthermore, there was no

MUC16 mAbs M11 and OC125 React with a Complex Conformational Epitope That Does Not Require One Complete SEA Domain

We sought to further define the peptide epitopes of mAb OC125 and M11 by C- and N-terminal truncations of the MUC16 construct produced in E. coli. We used the 1.7TR that includes the entire TR5 along with half of the SEA domain of TR6. The 1.2TR construct is fully contained in this construct (Figure 4A). All constructs were purified by Ni-chromatography and shown to migrate as homogeneous bands in SDSPAGE, except for one construct designated 11/2 that was difficult to express and keep in solution during purification probably due to its basic pI (Figure 4B).31 Western blot analysis showed that mAbs OC125 and M11 reacted with both the 1.7TR and 1.2TR constructs, as expected, as well as with an Nterminal truncated construct 5/2, although the interaction was weaker for M11 (Figures 4B and 5). The 5/2 construct does not include a complete SEA domain but comprises half of the SEA domain of TR5 and half of the SEA domain of TR6, including one of the two cysteine residues in each domain. mAb OC125 also showed lower binding to a truncated construct (5/ 7) that excludes the most C-terminal part of the 5/2 construct, including one cysteine residue. Although we did not observe reactivity with construct 11/2, which excludes the most Nterminal part of the 5/2 construct, because we experienced problems with expression and purification of this particular construct, we cannot completely rule out reactivity (Figures 4B and 5). While this manuscript was in preparation, Bressan and colleagues24 reported that mAbs OC125 and M11-bound E. coli produced MUC16 TR constructs, and they concluded that the entire SEA domain contained in a fragment of 128 residues was the minimal epitope for mAbs OC125, M11, and OV197. Our results confirm these findings but also extend them by showing that although a large construct is required for binding the complete SEA domain with its two cysteine residues enabling folding into the native SEA domain structure12 does not appear to be necessary. Given these findings, we further tested a library of 38 overlapping 17 aa peptides covering the 5/2 construct by direct ELISA (Figure 5). Two synthetic biotinylated 25 aa peptides based on the sequence of peptides 6 and 29 were also tested because we found these to represent the immunodominant regions of the MUC16 TR (results to be reported elsewhere). mAbs M11 and OC125 did not bind to any of the peptides, while both antibodies reacted with the 1.7TR, 1.2TR, and 5/2 constructs and mAb OC125 reacted very weakly with 5/7. E

dx.doi.org/10.1021/pr500215g | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Maeda and colleagues12 found two cysteine residues in close proximity in two adjacent β-strands, although a disulfide bond was not formed between them, which could be due to the use of reducing agents (1 mM DTT) in protein buffer. More recently, Bressan and colleagues24 showed by deletion analysis of a TR expressed in E. coli, using both chimeric and nonchimeric constructs, that the minimal region for antibody binding includes an entire SEA domain. The authors also demonstrated that the cysteine loop in the SEA domain proposed by O’Brien and colleagues3 was not sufficient for binding. While these recent studies suggest that glycosylation is not involved in the binding epitopes for mAbs to the TR regions of MUC16, it has in the past been shown that treatment of sera with PNGase F or O-glycanase resulted in a lower reactivity with mAb OC125 to CA125.32 Another report analyzing glycosylation of CA125 derived from OVCAR3 also demonstrated a lower binding of mAb OC125 with CA125 after treatment of PNGase F.6 To clarify the controversy on glycosylation dependence of antibodies M11 and OC125 binding, we undertook more detailed studies and expressed a TR (number 5) of MUC16 in both E. coli and CHO cells for comparative analysis and confirmed that binding of the antibodies was not substantially influenced by glycosylation. We could demonstrate a minor increase in binding for the glycosylated TR by competitive inhibition ELISA, but by the binding of mAbs, M11 and OC125 appeared to be largely unaffected by glycosylation. Moreover, we evaluated MUC16 expression in OVCAR3 wt and SC using three additional commercial mAbs: X52 (M11-like), X306 (OC125-like), and clone 197 (OV197-like) (Figure S2B in the Supporting Information). For all three mAbs, there were no apparent differences in cell staining regardless of the induced changes in O-glycosylation. Using a series of truncated constructs of the MUC16 TR expressed in E. coli and a large panel of peptides, we could confirm that binding of MUC16 mAbs requires an unusually large portion of the TR. However, we demonstrated that the minimal epitope is not limited to an entire SEA domain per se and hence to the global fold of this domain. Instead, we found that binding of mAbs could be accommodated within a construct (5/2) bridging parts of two consecutive SEA domains. Overall, these results suggest that the binding epitopes are composed of an unusually stable local conformation in an area of the TR that is not influenced by reduction and denaturation in SDS and urea or by glycosylation. We also tested several alternative disruptive strategies including high concentration of urea (2 M), TCEP (5 mM), and 2-mercaptoethanol (10 mM), and these did not disrupt the binding of mAb M11 (data not shown). The large MUC16 mucin is predicted to be heavily N- and O-glycosylated, but so far direct experimental evidence has been missing. Using our SimpleCell shotgun mass spectrometry sequencing approach, we previously identified a number of Oglycosylation sites (270 sites) located mainly in the N-terminal region of MUC16. These glycosites were identified from total extracts or secretomes of Capan1 SC7 and more recently OVCAR3 SC and HeLa SC.8 Interestingly, this approach has not led to the identification of many O-glycosites in the TR region of MUC16. In fact, we have only identified two Oglycosites in the putative TR linker regions, one in TR12 (LGASKTPASIF13,781) and one site in TR13 (THRSSVPTTSTGVVSEEPF13,926), and one site in an apparently degenerated SEA domain (QPTSSSSTQHF14,323).8 However, sequence analysis of the TR region may suggest additional O-glycosites

Figure 3. ELISA assay comparing inhibitory activity of recombinant MUC16 1.2TRs produced in E. coli and CHO SC. (A) OVCAR3 wt express elongated O-glycans, while OVCAR3 SC express truncated Oglycans due to knockout of the COSMC gene.8 Initial titration ELISA to determine optimal dilution of OVCAR3 wt/SC supernatants (1:8 corresponding to A450 nm ≈ 1) for inhibition assay development. Plates were coated with M11, incubated with varying dilutions of spent culture medium from OVCAR3 cells, and binding of shed MUC16 was detected with mAb X52. (B) Competitive inhibition ELISA with varying amounts of MUC16 1.2TRs from E. coli and CHO SC, followed by incubation with 1:8 dilution of culture medium from OVCAR3 wt and final detection of bound MUC16 from OVCAR3 using mAb X52 that does not bind the 1.2TRs. (C) Same as panel B but with culture medium from OVCAR3 SC. MUC16 1.2TR produced in CHO SC was approximately two times more effective in inhibiting M11 binding than MUC16 1.2TR from E. coli.

■

DISCUSSION The binding epitopes of mAbs to MUC16 used in the CA125 ovarian cancer biomarker assays have long remained a puzzle. It has so far been impossible to narrow the binding epitopes down to linear peptides, and it has been suggested that the epitopes involve a cysteine loop region between two conserved residues (Cys-59 and Cys-79, O’Brien nomenclature) in TRs.3 In an experimental NMR model of a murine MUC16 SEA domain, F

dx.doi.org/10.1021/pr500215g | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Figure 4. SDS-PAGE Western blot analysis of MUC16 constructs expressed in E. coli. (A) Graphic depiction of MUC16 truncation constructs tested, including all domains represented within the 1.7TR sequence. The minimal epitopes identified by Western blot analysis are also indicated. (B) Coomassie staining of purified constructs (1 ug loaded/lane). (C) Western blot analysis with mAbs OC125 and M11.

residues (Figure 1C). Structural characterization by Fourier transform infrared spectroscopy of the 21 aa peptide delimited by the conserved cysteines (CRLTLLRPEKDGAATGVDAIC) showed that it is largely unstructured in solution at physiological pH, independently of the redox status of the thiol groups (i.e., whether the peptide was linear or cyclized through the formation of an intramolecular disulfide). Furthermore, it was observed that for peptides with an intramolecular disulfide bond, but not for the linear variants, replacement of the P residue at position 8 by S, as found in ∼25% of TRs, resulted in the formation of β-rich aggregates, likely reminiscent of the gel-like extracellular layers formed by naturally aggregating mucins.23 In TR5, position 8 is occupied by a serine residue (Ser12,758), which we found to be glycosylated in CHO SC. The structural effect of glycosylation was not addressed in previous reports, and potential structural changes due to this important post-translational modification cannot be disregarded.33 Indeed, it has been shown that GalNAc O-glycosylation on a synthetic MUC1 peptide promotes a shift from a type-I β-turn conformation toward a more extended state by restricting the conformational space of the peptide, inducing a more rigid structure, and by the establishment of novel peptide−sugar interactions.34 In our study, the linker region spanning residues 12 814 to 12 845 between the SEA domains of TR5 and TR6 was found to be highly glycosylated, probably resulting in a more conformational constrained (i.e., extended) configuration (Figure 1C). The sequence of this linker region is highly

in the SEA domains and in particular clustered O-glycosites in the putative interdomain linker regions. The TR regions also contain three highly conserved N-glycosylation consensus motifs positioned in the SEA domain and one more infrequent N-glycosylation motif in the linker region. Here we extensively characterized N- and O-glycosylation sites of the MUC16 TR5 with flanking linker sequences expressed in CHO SCs. TR5 has four canonical consensus motifs for N-glycosylation, three in the SEA domain and one in the linker region, and we could show that they were all utilized (Figure 1B). We further identified 15 O-glycosites within the boundaries of one full TR (TR5), where five sites were within the 124 amino acid SEA domain and the rest clustered in the linker region (Figure 1B). Thus, the glycosylation of the 10 000 amino acid long TR region of MUC16 can be proposed to contain at least three Nglycans and about 5 O-glycans in the 60 TRs and about 10 clustered O-glycans placed in adjoining mucin-like linker regions of 25−30 residues, which would extrapolate to over 180 N-glycans and 900 O-glycans in this region of MUC16. The N-terminal 12 000 amino acid long domain could harbor additionally over 4500 O-glycosites and up to 67 N-glycosites. We identified one O-glycosite (Ser12,758) between two conserved cysteine residues (Cys12,751 and Cys12,771) in TR5. In a homology model of the SEA domains of TR5 and 6, using the NMR structure of a murine CA125 SEA domain as template, the side chains of Cys12,751−Cys12,771 and of Cys12,907−Cys12,927 are in close proximity, suggesting the existence of covalent links between each consecutive pair of G

dx.doi.org/10.1021/pr500215g | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Figure 5. ELISA analysis of a panel of synthetic overlapping peptides covering MUC16 TR5/TR6. (A) Graphic depiction of peptide designs and sequence region analyzed. A library of 38 overlapping peptides (17 amino acids) with an offset of 4 residues covering region 12 757 to 12 923 of MUC16 TR5/TR6 were used. In addition, two 25 amino acid biotinylated peptides covering peptides numbers 6 and 29, respectively, were also included. (B) Binding of mAb OC125. (C) Binding of mAb M11. MUC16 recombinant fragments (1.7TR; 1.2TR; 5/2; 5/7) and MUC2 peptide were used as controls.

antitumor activity compared with 11D10. Epitope analysis of mAb 3A5 demonstrated that it binds to the TR region, although a conserved histidine residue (EENMQHPGS) seems to be critical for binding. However, Western blot analysis of different TRs expressed in E. coli by Bressan and colleagues24 indicates that binding of M11, OC125, and OV197 is not affected by this residue. More recently, Chekmasova and collaborators39 developed a chimeric antigen receptor (CAR) 4H11-28z that is able to lyse efficiently human ovarian cancer cell lines in vitro and also exhibits antitumor activity in mice bearing orthotopic human ovarian tumors. This mAb reacts with a nonglycosylated peptide in the juxtamembrane region out of the TR domain, which is not recognized by M11 and OC125. Moreover, immunohistochemistry with mAb 4H11 in different tumor tissues showed comparable results to OC125.40 The findings that current MUC16 antibodies recognize poorly defined conformational epitopes in the TR region have potential implications for the design of vaccines. Thus, the results presented here and those of Bressan24 do not support the design of simple peptide-based vaccines but suggest that larger peptides or recombinant TR proteins as produced here in E. coli or perhaps preferably CHO are required to elicit potent anti-MUC16 antibodies. The majority of existing MUC16 mAbs were produced to isolated native MUC16 glycoproteins, and although these are shown to react with unglycosylated TR protein fragments, more studies are needed to evaluate appropriate immunogens for eliciting MUC16 immune response. Ovarian cancer expresses truncated O-glycans

conserved between all TRs of MUC16 and on average contains 15 S/T residues in a stretch of ∼32 residues (Figure S1B in the Supporting Information). The TR of MUC16 has often been referred to as a mucin domain, but as demonstrated here, the TRs mainly consist of the SEA domains with three N-linked and 3−5 O-linked glycans, and only the linker regions are mucin-like with high density of O-glycans. In contrast, the large N-terminal domain carries high density of O-glycans, although the domain does not appear to have repeat structures. MUC16 is an interesting target for immunotherapeutic intervention.35 MUC16 is overexpressed on the membrane of most ovarian cancers,16 although similar to other cell membrane mucins such as MUC1, MUC16 is shed and found at some levels in the circulation. Several studies have explored targeting of MUC16 with mAbs. An example is the OC125-like mAb B43.1319 that binds specifically the TR of MUC16 with high affinity and forms immune complexes with circulating CA125, inducing cellular and humoral responses.36 A previous study showed that treatment with technetium 99mlabeled mAb B43.13 is able to induce antigen-specific immune responses in patients with recurrent ovarian cancer, resulting in an improved overall survival.37 Later, Chen and colleagues38 developed two drug-conjugated humanized mAbs: one (11D10) binds to a single epitope on a degenerate repeat sequence in the juxtamembrane domain and another (3A5) binds to multiple sites on the repeat region of MUC16. Their results showed that mAb 3A5 had an increased binding to OVCAR3 and ovarian tumor tissue and also exhibited higher H

dx.doi.org/10.1021/pr500215g | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Danish Research Councils, a program of excellence from the University of Copenhagen, and The Danish National Research Foundation (DNRF107), Fundaçaõ para a Ciência e a Tecnologia (FCT) − Programa Operacional Ciência e Inovaçaõ 2010 do Quadro Comunitário de Apoio III and FEDER, Programa COMPETE (PTDC/SAU-ONC/117216/ 2010). IPATIMUP is an Associate Laboratory of the Portuguese Ministry of Science, Technology and Higher Education and is partially supported by FCT. L.M.-S. acknowledges FCT for financial support through a Ph.D. fellowship (SFRH/BD/60536/2009). “Project NORTE - 070124-FEDER-000024 co-financed by Programa Operacional Regional do Norte (ON.2 − O Novo Norte), under Quadro de Referência Estratégico Nacional (QREN), by Fundo Europeu de Desenvolvimento Regional (FEDER).”

including the STn and Tn structures, although direct demonstration of Tn and STn on MUC16 in tissue is missing.41 We recently demonstrated that circulating MUC16 in cancer patients has exposed truncated O-glycans and that specific detection of the STn glycoform can increase specificity of the CA125 biomarker assay.42 The CHO SC-produced MUC16 1.2TR has homogeneous truncated Tn O-glycans and may potentially be used to generate antibodies to truncated Oglycopeptide epitopes. We have previously produced antibodies to MUC1 Tn-glycopeptides and shown that these exhibit exquisite cancer-specific reactivity and are highly immunogenic in man.43,44

■

CONCLUSIONS In summary, CA125 mAbs directed to the TR region appear to define local conformational epitopes that are largely unaffected by glycosylation. While the unglycosylated MUC16 TRs serve as antigens for CA125 mAbs, it remains to be explored if glycosylation is necessary to induce potent immunity. In this study, we generated a recombinant expression platform for a MUC16 TR vaccine with aberrant O-glycosylation, which should enable further testing and generation of cancer-specific immune response to aberrant glycopeptide epitopes. We hope this will pave the way for designing and testing vaccine strategies for immunotherapeutic intervention.

■

■

ABBREVIATIONS TR, tandem repeat; SEA, sperm protein enterokinase agrin; wt, wild-type; SC, simple cells; mAb, monoclonal antibody; VVAB, biotinylated Vicia villosa lectin

■

ASSOCIATED CONTENT

S Supporting Information *

Figure S1. Sequence alignment of the MUC16 SEA domains (A) and of linker regions (B) of human MUC16 (protein accession number NP_078966.2). Figure S2. Immunocytology of MUC16 expressing cell lines. Figure S3. Semiquantification of relative concentrations of MUC16 1.2TR recombinant proteins by coomassie blue staining. Table S1. Primers used for the amplification and cloning of N- and C-terminal truncations of MUC16 1.7TR. Table S2. N- and O-glyco sites identified in MUC16 1.2TR expressed in CHO SC. This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

(1) Jemal, A.; Bray, F.; Center, M. M.; Ferlay, J.; Ward, E.; Forman, D. Global cancer statistics. Ca-Cancer J. Clin. 2011, 61 (2), 69−90. (2) Bast, R. C.; Klug, T. L.; Stjohn, E.; Jenison, E.; Niloff, J. M.; Lazarus, H.; Berkowitz, R. S.; Leavitt, T.; Griffiths, C. T.; Parker, L.; Zurawski, V. R.; Knapp, R. C. A radioimmunoassay using a monoclonal-antibody to monitor the course of epithelial ovariancancer. N. Engl J. Med. 1983, 309 (15), 883−887. (3) O’Brien, T. J.; Beard, J. B.; Underwood, L. J.; Dennis, R. A.; Santin, A. D.; York, L. The CA 125 gene: An extracellular superstructure dominated by repeat sequences. Tumor Biol. 2001, 22 (6), 348−366. (4) Yin, B. W.; Dnistrian, A.; Lloyd, K. O. Ovarian cancer antigen CA125 is encoded by the MUC16 mucin gene. Int. J. Cancer 2002, 98 (5), 737−740. (5) Yin, B. W.; Lloyd, K. O. Molecular cloning of the CA125 ovarian cancer antigen: identification as a new mucin, MUC16. J. Biol. Chem. 2001, 276 (29), 27371−27375. (6) Kui Wong, N.; Easton, R. L.; Panico, M.; Sutton-Smith, M.; Morrison, J. C.; Lattanzio, F. A.; Morris, H. R.; Clark, G. F.; Dell, A.; Patankar, M. S. Characterization of the oligosaccharides associated with the human ovarian tumor marker CA125. J. Biol. Chem. 2003, 278 (31), 28619−28634. (7) Steentoft, C.; Vakhrushev, S. Y.; Vester-Christensen, M. B.; Schjoldager, K. T.; Kong, Y.; Bennett, E. P.; Mandel, U.; Wandall, H.; Levery, S. B.; Clausen, H. Mining the O-glycoproteome using zincfinger nuclease-glycoengineered SimpleCell lines. Nat. Methods 2011, 8 (11), 977−982. (8) Steentoft, C.; Vakhrushev, S. Y.; Joshi, H. J.; Kong, Y.; VesterChristensen, M. B.; Schjoldager, K. T.; Lavrsen, K.; Dabelsteen, S.; Pedersen, N. B.; Marcos-Silva, L.; Gupta, R.; Bennett, E. P.; Mandel, U.; Brunak, S.; Wandall, H. H.; Levery, S. B.; Clausen, H. Precision mapping of the human O-GalNAc glycoproteome through SimpleCell technology. EMBO J. 2013, 32 (10), 1478−1488. (9) Hattrup, C. L.; Gendler, S. J. Structure and function of the cell surface (tethered) mucins. Annu. Rev. Physiol. 2008, 70, 431−457. (10) Blalock, T. D.; Spurr-Michaud, S. J.; Tisdale, A. S.; Heimer, S. R.; Gilmore, M. S.; Ramesh, V.; Gipson, I. K. Functions of MUC16 in corneal epithelial cells. Invest. Ophthalmol. Visual Sci. 2007, 48 (10), 4509−4518. (11) Bork, P.; Patthy, L. The SEA module: a new extracellular domain associated with O-glycosylation. Protein Sci. 1995, 4 (7), 1421−1425. (12) Maeda, T.; Inoue, M.; Koshiba, S.; Yabuki, T.; Aoki, M.; Nunokawa, E.; Seki, E.; Matsuda, T.; Motoda, Y.; Kobayashi, A.; Hiroyasu, F.; Shirouzu, M.; Terada, T.; Hayami, N.; Ishizuka, Y.;

AUTHOR INFORMATION

Corresponding Authors

*L.D.: Tel: 351-225570700. Fax: 351-225570799. E-mail: [email protected]. *H.C.: Tel: 45-35326668. Fax: 45-35367980. E-mail: hclau@ sund.ku.dk. Present Address #

M.A.T.: Radiometer Medical ApS, Åkandevej 21, DK-2700 Brønshøj, Denmark. E-mail: [email protected]. Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest. ▽ Deceased February 2014.

■

ACKNOWLEDGMENTS This work was supported by Kirsten og Freddy Johansen Fonden, A.P. Møller og Hustru Chastine Mc-Kinney Møllers Fond til Almene Formaal, The Novo Nordisk Foundation, The I

dx.doi.org/10.1021/pr500215g | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

Shinya, N.; Tatsuguchi, A.; Yoshida, M.; Hirota, H.; Matsuo, Y.; Tani, K.; Arakawa, T.; Carninci, P.; Kawai, J.; Hayashizaki, Y.; Kigawa, T.; Yokoyama, S. Solution structure of the SEA domain from the murine homologue of ovarian cancer antigen CA125 (MUC16). J. Biol. Chem. 2004, 279 (13), 13174−82. (13) Palmai-Pallag, T.; Khodabukus, N.; Kinarsky, L.; Leir, S. H.; Sherman, S.; Hollingsworth, M. A.; Harris, A. The role of the SEA (sea urchin sperm protein, enterokinase and agrin) module in cleavage of membrane-tethered mucins. FEBS J. 2005, 272 (11), 2901−2911. (14) Gubbels, J. A.; Belisle, J.; Onda, M.; Rancourt, C.; Migneault, M.; Ho, M.; Bera, T. K.; Connor, J.; Sathyanarayana, B. K.; Lee, B.; Pastan, I.; Patankar, M. S. Mesothelin-MUC16 binding is a high affinity, N-glycan dependent interaction that facilitates peritoneal metastasis of ovarian tumors. Mol. Cancer 2006, 5 (1), 50. (15) Kenemans, P.; Vankamp, G. J.; Oehr, P.; Verstraeten, R. A. Heterologous double-determinant immunoradiometric assay CA-125II: reliable 2nd-generation immunoassay for determining CA-125 in serum. Clin Chem. 1993, 39 (12), 2509−2513. (16) Bast, R. C.; Feeney, M.; Lazarus, H.; Nadler, L. M.; Colvin, R. B.; Knapp, R. C. Reactivity of a monoclonal-antibody with human ovarian-carcinoma. J. Clin. Invest. 1981, 68 (5), 1331−1337. (17) Hardardottir, H.; Parmley, T. H.; Quirk, J. G.; Sanders, M. M.; Miller, F. C.; Obrien, T. J. Distribution of CA125 in embryonic-tissues and adult derivatives of the fetal periderm. Am. J. Obstet. Gynecol. 1990, 163 (6), 1925−1931. (18) O’Brien, T. J.; Raymond, L. M.; Bannon, G. A.; Ford, D. H.; Hardardottir, H.; Miller, F. C.; Quirk, J. G. New monoclonalantibodies identify the glycoprotein carrying the ca-125 epitope. Am. J. Obstet. Gynecol. 1991, 165 (6), 1857−1864. (19) Nustad, K.; Bast, R. C.; Obrien, T. J.; Nilsson, O.; Seguin, P.; Suresh, M. R.; Saga, T.; Nozawa, S.; Bormer, O. P.; deBruijn, H. W. A.; Nap, M.; Vitali, A.; Gadnell, M.; Clark, J.; Shigemasa, K.; Karlsson, B.; Kreutz, F. T.; Jette, D.; Sakahara, H.; Endo, K.; Paus, E.; Warren, D.; Hammarstrom, S.; Kenemans, P.; Hilgers, J. Specificity and affinity of 26 monoclonal antibodies against the CA 125 antigen: First report from the ISOBM TD-1 workshop. Tumor Biol. 1996, 17 (4), 196−219. (20) Nap, M.; Vitali, A.; Nustad, K.; Bast, R. C.; Obrien, T. J.; Nilsson, O.; Seguin, P.; Suresh, M. R.; Bormer, O. P.; Saga, T.; deBruijn, H. W. A.; Nozawa, S.; Kreutz, F. T.; Jette, D.; Sakahara, H.; Gadnell, M.; Endo, K.; Barlow, E. H.; Warren, D.; Paus, E.; Hammarstrom, S.; Kenemans, P.; Hilgers, J. Immunohistochemical characterization of 22 monoclonal antibodies against the CA125 antigen: 2nd report from the ISOBM TD-1 workshop. Tumor Biol. 1996, 17 (6), 325−331. (21) Lloyd, K. O.; Yin, B. W. T.; Kudryashov, V. Isolation and characterization of ovarian cancer antigen CA 125 using a new monoclonal antibody (VK-8): Identification as a mucin-type molecule. Int. J. Cancer 1997, 71 (5), 842−850. (22) Warren, D. J.; Nustad, K.; Beard, J. B.; O’Brien, T. J. Expression and epitope characterization of a recombinant CA 125 repeat: fourth report from the ISOBM TD-1 workshop. Tumor Biol. 2009, 30 (2), 51−60. (23) Berman, Z. T.; Moore, L. J.; Knudson, K. E.; Whelan, R. J. Synthesis and structural characterization of the peptide epitope of the ovarian cancer biomarker CA125 (MUC16). Tumor Biol. 2010, 31 (5), 495−502. (24) Bressan, A.; Bozzo, F.; Maggi, C. A.; Binaschi, M. OC125, M11 and OV197 epitopes are not uniformly distributed in the tandemrepeat region of CA125 and require the entire SEA domain. Dis. Markers 2013, 34 (4), 257−267. (25) Bennett, E. P.; Hassan, H.; Mandel, U.; Mirgorodskaya, E.; Roepstorff, P.; Burchell, J.; Taylor-Papadimitriou, J.; Hollingsworth, M. A.; Merkx, G.; van Kessel, A. G.; Eiberg, H.; Steffensen, R.; Clausen, H. Cloning of a human UDP-N-acetyl-alpha-D-galactosamine: polypeptide N-acetylgalactosaminyltransferase that complements other GalNAc-transferases in complete O-glycosylation of the MUC1 tandem repeat. J. Biol. Chem. 1998, 273 (46), 30472−30481. (26) Pedersen, J. W.; Blixt, O.; Bennett, E. P.; Tarp, M. A.; Dar, I.; Mandel, U.; Poulsen, S. S.; Pedersen, A. E.; Rasmussen, S.; Jess, P.;

Clausen, H.; Wandall, H. H. Seromic profiling of colorectal cancer patients with novel glycopeptide microarray. Int. J. Cancer 2011, 128 (8), 1860−1871. (27) Liu, Q. Y. P.; Sulzenbacher, G.; Yuan, H. P.; Bennett, E. P.; Pietz, G.; Saunders, K.; Spence, J.; Nudelman, E.; Levery, S. B.; White, T.; Neveu, J. M.; Lane, W. S.; Bourne, Y.; Olsson, M. L.; Henrissat, B.; Clausen, H. Bacterial glycosidases for the production of universal red blood cells. Nat. Biotechnol. 2007, 25 (4), 454−464. (28) Vakhrushev, S. Y.; Steentoft, C.; Vester-Christensen, M. B.; Bennett, E. P.; Clausen, H.; Levery, S. B. Enhanced Mass Spectrometric Mapping of the Human GalNAc-type O-Glycoproteome with SimpleCells. Mol. Cell. Proteomics 2013, 12 (4), 932−944. (29) Vester-Christensen, M. B.; Halim, A.; Joshi, H. J.; Steentoft, C.; Bennett, E. P.; Levery, S. B.; Vakhrushev, S. Y.; Clausen, H. Mining the O-mannose glycoproteome reveals cadherins as major O-mannosylated glycoproteins. Proc. Natl. Acad. Sci. U. S. A. 2013, 110 (52), 21018−21023. (30) Nustad, K.; Lebedin, Y.; Lloyd, K. O.; Shigemasa, K.; de Bruijn, H. W. A.; Jansson, B.; Nilsson, O.; Olsen, K. H.; rsquo; Brien, T. J. Epitopes on CA 125 from Cervical Mucus and Ascites Fluid and Characterization of Six New Antibodies. Tumor Biol. 2002, 23 (5), 303−314. (31) Mehlin, C.; Boni, E.; Buckner, F. S.; Engel, L.; Feist, T.; Gelb, M. H.; Haji, L.; Kim, D.; Liu, C.; Mueller, N.; Myler, P. J.; Reddy, J. T.; Sampson, J. N.; Subramanian, E.; Van Voorhis, W. C.; Worthey, E.; Zucker, F.; Hol, W. G. Heterologous expression of proteins from Plasmodium falciparum: results from 1000 genes. Mol. Biochem. Parasitol. 2006, 148 (2), 144−60. (32) Nagata, A.; Hirota, N.; Sakai, T.; Fujimoto, M.; Komoda, T. Molecular nature and possible presence of a membranous glycanphosphatidylinositol anchor of ca125-antigen. Tumor Biol. 1991, 12 (5), 279−286. (33) Jentoft, N. Why are proteins O-glycosylated? Trends Biochem. Sci. 1990, 15 (8), 291−294. (34) Schuman, J.; Campbell, A. P.; Koganty, R. R.; Longenecker, B. M. Probing the conformational and dynamical effects of Oglycosylation within the immunodominant region of a MUC1 peptide tumor antigen. J. Peptide Res. 2003, 61 (3), 91−108. (35) Singh, A. P.; Senapati, S.; Ponnusamy, M. P.; Jain, M.; Lele, S. M.; Davis, J. S.; Remmenga, S.; Batra, S. K. Clinical potential of mucins in diagnosis, prognosis, and therapy of ovarian cancer. Lancet Oncol. 2008, 9 (11), 1076−1085. (36) Noujaim, A. A.; Schultes, B. C.; Baum, R. P.; Madiyalakan, R. Induction of CA125-specific B and T cell responses in patients injected with MAb-B43.13–evidence for antibody-mediated antigenprocessing and presentation of CA125 in vivo. Cancer Biother. Radiopharm. 2001, 16 (3), 187−203. (37) Mobus, V. J.; Baum, R. P.; Bolle, M.; Kreienberg, R.; Noujaim, A. A.; Schultes, B. C.; Nicodemus, C. F. Immune responses to murine monoclonal antibody-B43.13 correlate with prolonged survival of women with recurrent ovarian cancer. Am. J. Obstet. Gynecol. 2003, 189 (1), 28−36. (38) Chen, Y.; Clark, S.; Wong, T.; Chen, Y.; Chen, Y.; Dennis, M. S.; Luis, E.; Zhong, F.; Bheddah, S.; Koeppen, H.; Gogineni, A.; Ross, S.; Polakis, P.; Mallet, W. Armed antibodies targeting the mucin repeats of the ovarian cancer antigen, MUC16, are highly efficacious in animal tumor models. Cancer Res. 2007, 67 (10), 4924−4932. (39) Chekmasova, A. A.; Rao, T. D.; Nikhamin, Y.; Park, K. J.; Levine, D. A.; Spriggs, D. R.; Brentjens, R. J. Successful eradication of established peritoneal ovarian tumors in SCID-Beige mice following adoptive transfer of T cells genetically targeted to the MUC16 antigen. Clin. Cancer Res. 2010, 16 (14), 3594−3606. (40) Dharma Rao, T.; Park, K. J.; Smith-Jones, P.; Iasonos, A.; Linkov, I.; Soslow, R. A.; Spriggs, D. R. Novel monoclonal antibodies against the proximal (carboxy-terminal) portions of MUC16. Appl. Immunohistochem. Mol. Morphol. 2010, 18 (5), 462−472. (41) Akita, K.; Yoshida, S.; Ikehara, Y.; Shirakawa, S.; Toda, M.; Inoue, M.; Kitawaki, J.; Nakanishi, H.; Narimatsu, H.; Nakada, H. Different levels of sialyl-Tn antigen expressed on MUC16 in patients J

dx.doi.org/10.1021/pr500215g | J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

Article

with endometriosis and ovarian cancer. Int. J. Gynecol Cancer 2012, 22 (4), 531−538. (42) Chen, K.; Gentry-Maharaj, A.; Burnell, M.; Steentoft, C.; Marcos-Silva, L.; Mandel, U.; Jacobs, I.; Dawnay, A.; Menon, U.; Blixt, O. Microarray Glycoprofiling of CA125 improves differential diagnosis of ovarian cancer. J. Proteome Res. 2013, 12 (3), 1408−1418. (43) Sorensen, A. L.; Reis, C. A.; Tarp, M. A.; Mandel, U.; Ramachandran, K.; Sankaranarayanan, V.; Schwientek, T.; Graham, R.; Taylor-Papadimitriou, J.; Hollingsworth, M. A.; Burchell, J.; Clausen, H. Chemoenzymatically synthesized multimeric Tn/STn MUC1 glycopeptides elicit cancer-specific anti-MUC1 antibody responses and override tolerance. Glycobiology 2006, 16 (2), 96−107. (44) Sabbatini, P. J.; Ragupathi, G.; Hood, C.; Aghajanian, C. A.; Juertzka, M.; Lasonos, A.; Hensley, M. L.; Spassova, M. K.; Ouerfelli, O.; Spriggs, D. R.; Tew, W. P.; Konner, J.; Clausen, H.; Abu Rustum, N.; Dansihefsky, S. J.; Livingston, P. O. Pilot study of a heptavalent vaccine-keyhole limpet hemocyanin conjugate plus QS21 in patients with epithelial ovarian, fallopian tube, or peritoneal cancer. Clin. Cancer Res. 2007, 13 (14), 4170−4177. (45) Arnold, K.; Bordoli, L.; Kopp, J.; Schwede, T. The SWISSMODEL workspace: a web-based environment for protein structure homology modelling. Bioinformatics 2006, 22 (2), 195−201. (46) Emsley, P.; Lohkamp, B.; Scott, W. G.; Cowtan, K. Features and development of Coot. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2010, 66 (Pt 4), 486−501. (47) Adams, P. D.; Afonine, P. V.; Bunkoczi, G.; Chen, V. B.; Davis, I. W.; Echols, N.; Headd, J. J.; Hung, L. W.; Kapral, G. J.; GrosseKunstleve, R. W.; McCoy, A. J.; Moriarty, N. W.; Oeffner, R.; Read, R. J.; Richardson, D. C.; Richardson, J. S.; Terwilliger, T. C.; Zwart, P. H. PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2010, 66, 213−221.

K

dx.doi.org/10.1021/pr500215g | J. Proteome Res. XXXX, XXX, XXX−XXX