J. Biol. Chem.-2008-Kaneko-jbc.M806776200 by wuxiangyu


									    JBC Papers in Press. Published on December 15, 2008 as Manuscript M806776200
       The latest version is at http://www.jbc.org/cgi/doi/10.1074/jbc.M806776200
Mesothelin and CA125/MUC16, Page 1

Osamu Kaneko, Lucy Gong, Jingli Zhang, Johanna K. Hansen, Raffit Hassan, Byungkook
                                       Lee and Mitchell Ho
   From the Laboratory of Molecular Biology, Center for Cancer Research, National Cancer
                  Institute, National Institutes of Health, Bethesda, MD 20892
                         Running head: Mesothelin and CA125/MUC16
 *This work was supported in part by the Intramural Research Program of the NIH, National
  Cancer Institute, Center for Cancer Research, in part by the Ovarian Cancer Research Fund
  Individual Investigator Award (M. Ho) and in part by the Mesothelioma Applied Research
                     Foundation Grant in Honor of Craig Kozicki (M. Ho).
  Address correspondence to: Dr. Mitchell Ho, Laboratory of Molecular Biology, National
        Cancer Institute, 37 Convent Drive, Room 5002C, Bethesda, MD 20892-4264
          Phone: (301) 451-8727; Fax: (301) 402-1344; E-mail: homi@mail.nih.gov

     Ovarian     cancer      and      malignant    new therapeutic agent for preventing or
mesothelioma      frequently     express   both    treating peritoneal malignant tumors.

                                                                                                         Downloaded from www.jbc.org by guest, on August 14, 2011
mesothelin and CA125 (also known as MUC16)
at high levels on the cell surface. The                 Ovarian cancer largely is confined to the
interaction between mesothelin and CA125 may       peritoneal cavity for much of its natural history
facilitate the implantation and peritoneal         (1). Peritoneal mesothelioma is a highly invasive
spread of tumors by cell adhesion, whereas the     tumor originating from the mesothelial linings of
detailed nature of this interaction is still       the peritoneum (2). The development of effective
unknown. Here, we used truncated mutagenesis       drug regimens against ovarian cancer and
and alanine replacement techniques to identify     mesothelioma has proven extremely difficult.
a binding site on mesothelin for CA125. We              Mesothelin was first identified in 1992 by the
examined the molecular interaction by Western      monoclonal antibody (mAb)1 K1 that was
blot overlay assays and further quantitatively     generated by the immunization of mice with
analyzed by enzyme-linked immunosorbent            human ovarian carcinoma (OVCAR-3) cells (3).
assay. We also evaluated the binding on cancer     The mesothelin gene encodes a 71-kDa precursor
cells by flow cytometry. We identified the         protein that is processed to a 40-kDa protein
region (296-359) consisting of 64 amino acids at   termed mesothelin, which is a glycosyl-
the N terminal of cell surface mesothelin as the   phosphatidylinositol (GPI)-anchored glycoprotein
minimum fragment for complete binding              present on the cell surface (4). Mesothelin is a
activity to CA125. In addition, we found that      differentiation antigen that is present on a
substitution of tyrosine-318 with an alanine       restricted set of normal adult tissues such as the
abolished the CA125-binding. Replacement of        mesothelium. In contrast, it is overexpressed in a
tryptophan-321 and glutamic acid-324 with          variety of cancers including mesothelioma,
alanine could partially decrease binding to        ovarian cancer, and pancreatic cancer (5). In
CA125, while mutation of histidine-354 had no      addition, mesothelin is also expressed on the
effect. These results indicate that a
conformation-sensitive structure of the region
(296-359) is required and sufficient for the
binding of mesothelin to CA125. In addition, we     The abbreviations used are: mAb, monoclonal
have shown that a single chain monoclonal          antibody; GPI, glycosyl-phosphatidylinositol;
antibody (SS1) recognizes this CA125-binding       FDA, Food and Drug Administration; TR, tandem
domain and blocks the mesothelin-CA125             repeat; ELISA, enzyme-linked immunosorbent
interaction on cancer cells. The identified        assay; DMEM, Dulbecco’s modified eagle
CA125-binding domain significantly inhibits        medium; rFc, rabbit Fc; hnMSLN, human
cancer cell adhesion and merits evaluation as a    mesothelin; RT, room temperature; scFv, single
                                                   chain Fv antibody; mnMSLN, mouse mesothelin
Mesothelin and CA125/MUC16, Page 2

surface of non-small cell lung cancer cells (6, 7),           CA125 was originally used as a biomarker in
especially most lung adenocarcinomas (8).                ovarian cancer due to its high expression in
     We and others have shown that mesothelin is         ovarian carcinomas and that it is shed into the
shed from tumor cells (9, 10), and antibodies            serum (23). A majority (88%) of mesotheliomas
specific for mesothelin are elevated in the sera of      are also CA125 positive on the cell membrane
patients with mesothelioma and ovarian cancer            (24). It was shown that 25% of peritoneal
(11). Shed serum mesothelin has been approved by         mesotheliomas have high CA125 expression (25).
the U.S. Food and Drug Administration (FDA) as           The intensity of CA125 membranous expression is
a new diagnostic biomarker in mesothelioma. In a         indistinguishable between ovarian carcinomas and
Phase I clinical study of an intrapleural IFN-β          peritoneal mesotheliomas. Gene expression
gene transfer using an adenoviral vector in patients     analysis using SAGE tag database has shown that
with mesotheliomas, we found that antitumor              mesothelioma has the second highest co-
immune responses targeting mesothelin were               expression of CA125 and mesothelin after ovarian
elicited in several patients (12). A recent study        cancer (26). Rump and colleagues have shown that
indicated that anti-mesothelin antibodies and            mesothelin binds to CA125 and that this
circulating mesothelin relate to the clinical state in   interaction may mediate cell adhesion (26).
ovarian cancer patients (13). Pastan and colleagues      Scholler     et    al.   recently    showed     that
                                                         CA125/mesothelin-dependent cell attachment

                                                                                                                Downloaded from www.jbc.org by guest, on August 14, 2011
developed an immunotoxin (SS1P) with a Fv for
mesothelin (14). Two Phase I clinical trials were        could be blocked with anti-CA125 antibodies (27).
completed at the National Cancer Institute (NIH,         Since mesothelin is present on peritoneal
Bethesda, MD) and there was sufficient antitumor         mesothelium, there may be an important role for
activity of SS1P to justify a Phase II trial. A          the mesothelin-CA125 interaction in the
chimeric antibody containing the mouse SS1 Fv            tumorigenesis of ovarian cancer and mesothelioma
for mesothelin was also developed and is currently       in the peritoneal cavity. The mesothelin binding
examined in a Phase I clinical trial for ovarian         site on CA125 may lie within the 156 amino acid
cancer, mesothelioma, pancreatic cancer, and non-        TR units, indicating multimeric binding of
small cell lung cancer (15).                             mesothelin to CA125. It has been found that the
     Mucins are heavily glycosylated proteins            extraordinarily abundant N-glycans on CA125,
found in the mucus layer or at the cell surface of       presumably in the TR region, are required for
many epitheliums (16). There are two structurally        binding to both glycosylated and non-glycosylated
distinct families of mucins, secreted and                mesothelin (28).
membrane-bound forms. CA125 (also known as                    Here, we identified the binding site of CA125
MUC16) was first identified in 1981 by OC125, a          on mesothelin by use of truncated mutagenesis and
mAb that had been developed from mice                    alanine replacement approaches. We measured
immunized with human ovarian cancer cells (17).          binding qualitatively by Western blot overlay
The first cDNA clones were reported in 2001 (18,         assays and quantitatively by enzyme-linked
19). CA125 is a very large membrane-bound cell           immunosorbent assay (ELISA). We also evaluated
surface mucin, with an average molecular weight          the interaction of CA125 and mesothelin on cancer
between 2.5 and 5 million Dalton. It is also             cells by flow cytometry. Furthermore, we have
heavily glycosylated with both O-linked and N-           shown that a single chain mAb (SS1) recognized
linked oligosaccharides (20). The peptide                the CA125-binding domain and blocked the
backbone of CA125 is composed of the N-                  mesothelin-CA125 interaction on cancer cells. The
terminal region, extensive Ser/Thr/Pro-rich              identified CA125-binding domain-Fc fusion
tandem repeats (TR) with 156 amino acids each            protein also significantly inhibited cancer cell
with both N- and O-glycosylations, a SEA domain          adhesion. Our results suggest that conformation-
with high levels of O-glycosylation and a C-             sensitive structures of the region (296-359) are
terminal region with a short cytoplasmic tail (19).      required and sufficient for the specific binding of
The SEA domain was first identified as a module          mesothelin to CA125. The domain proteins or the
commonly found in sea urchin sperm protein,              antibodies that block the mesothelin-CA125
enterokinase and agrin (21, 22). The significance        interaction merit evaluation as new therapeutic
of the SEA domain in CA125 is not clear.                 agents in treating peritoneal malignant tumors.
Mesothelin and CA125/MUC16, Page 3

                                                     phosphate buffer pH 5.0, and eluted with 100 mM
EXPERIMENTAL PROCEDURES                              glycine-HCl pH 3.0, and neutralized in 1M Tris
    Cell Culture - OVCAR-3 (ovarian) cells were      pH8.0. Fractions were collected using the AKTA
grown in RPMI 1640 (Dulbecco) supplemented           FPLC system (GE Life Sciences) and pooled and
with 20% FBS, 1% penicillin-streptomycin, 1% L-      concentrated. Final protein concentration was
Glutamine, and 0.2% human insulin. YOU               measured using Coomassie Plus Protein Assay
(mesothelioma) cells were grown in RPMI 1640         Reagent (Pierce). Fractions of the dominant peak
(Dulbecco) supplemented with 10% FBS, 1%             were run on a SDS-PAGE gel under non-reducing
penicillin-streptomycin, and 1% L-Glutamine.         and reducing conditions. To verify that generated
HEK 293T cells were grown in 100 mm tissue           proteins were not aggregated, mesothelin and its
culture dishes (Falcon) with Dulbecco’s Modified     mutants were subsequently run over a TSK gel-
Eagle Medium (DMEM) and supplemented with            filtration size exclusion column (TOSOH
10% FBS, 1% penicillin-streptomycin, and 1% L-       Bioscience LLC, Montgomeryville) at 0.5 ml/min
Glutamine.                                           in PBS pH7.5.
    Truncated Mutant Constructs - Full-length and         Flow Cytometry - To determine binding of
fragments of mesothelin were amplified by PCR        mesothelin fragments to CA125 on the cell
from pMH107 (GenBank Accession No.                   surface, OVCAR-3 or YOU cells were grown until
AY743922) (29). Primers were designed to             confluent, detached, and then incubated with 1

                                                                                                            Downloaded from www.jbc.org by guest, on August 14, 2011
incorporate flanking EcoRI and NotI restriction      µg/mL of mesothelin or its fragments in FACS
enzyme sites to facilitate in-frame cloning into a   buffer (5% BSA, 0.01% NaN3) for 1 h on ice.
modified pSecTag2 vector (Invitrogen) (Table 1).     Bound fragments were detected by incubating with
Constructs contained an Ig-κ leader sequence         a 1:200 dilution of Goat anti-Rabbit IgG-PE
followed by the rabbit IgG Fc and the full-length    (Biosource) secondary antibody in FACS buffer
sequence of extracellular domain of mesothelin       for 0.5 h on ice. Cells were analyzed using
(pMH113) or its fragments, followed by a myc         FACSCalibur (BD Biosciences). Each binding
epitope and 6xHis tag. The rabbit IgG Fc (rFc) and   experiment was repeated three to five times.
mesothelin fragments were separated by a                  In inhibition assays, cells were incubated with
thrombin cleavage site and a flexible linker. The    Flag-tagged mesothelin and excess amount (10-
rFc-mouse mesothelin fusion (pMH117) was             fold) of mesothelin or mesothelin fragments
constructed using the same strategy.                 without a Flag tag for 1 h on ice. Bound Flag-
    Alanine mutants were generated by PCR            tagged mesothelin proteins were detected by
mutagenesis       of   pMH107-derived       human    incubating with 1:100 dilution of an anti-Flag tag
mesothelin (hnMSLN) cDNA followed by cloning         mAb (Sigma) followed by PE conjugated goat
into the modified pSecTag2 vector. Mutations         anti-mouse IgG (Biosource).
were introduced in either forward or reverse              Sandwich ELISA - Nunc MaxiSorp 96 well
primers or by using a two-step overlap-extension     flat-bottomed plates were incubated overnight
PCR reaction (Table 2).                              with 5 µg/mL goat anti-rabbit IgG (Jackson
    Transfections and Mesothelin Mutant-Fc           ImmunoResearch Laboratories) in PBS, followed
Fusion Protein Production - HEK 293T cells were      by an overnight block with 5% BSA, 0.01% NaN3
grown until 60% confluent on 100 mm tissue           in PBS. Purified Fc mesothelin fragments were
culture dishes (Falcon). Constructs encoding         diluted to 1 µg/mL in ELISA buffer (0.01%
mesothelin-Fc fusion proteins were transiently       Tween 20, 10% Pierce SuperBlock) and incubated
transfected using Lipofectamine (Invitrogen) in 6    on plate for 1 h at room temperature (RT). Plates
mL serum free media. Three to 5 h later, 6 mL of     were then incubated with OVCAR-3 supernatant
20% FBS DMEM was added to each dish and              containing CA125 for 1 h at RT. To detect bound
incubated for 48 h. Media was harvested              CA125 a 1:200 dilution of anti-CA125 OC125
subsequently on a daily basis and replaced with      mAb (Zymed) was incubated for an additional
fresh medium. Fc fusion proteins were purified       hour at RT; subsequently a 1:1500 dilution of goat
from the media using columns containing Protein      anti-mouse IgG HRP conjugate (Biosource) was
A Sepharose (Amersham Biosciences). One mL           added for 1 h at RT. The plates were washed four
columns were loaded, washed with citrate-            times with ELISA buffer between each coating.
Mesothelin and CA125/MUC16, Page 4

Visualization was achieved with TMB detection         with Dunnett's and Newman-Keuls multiple
reagent (KPL) and absorbance was read at 450 nm       comparison post tests. P values < 0.05 were
with SpectraMax Plus plate reader (Molecular          considered statistically significant.
     Western Blots - Purified Fc mesothelin           RESULTS
fragments (500 ng) were mixed with Laemmli                 Generation of Mesothelin Mutants - Truncated
Sample Buffer (Biorad) supplemented with 5% B-        mutants of mesothelin were generated to
Mercaptoethanol. Samples were boiled for 2 min        sequentially narrow down the binding domain to
and separated on 4-20% SDS-PAGE gels                  CA125. As shown in Fig. 1, portions of
(Invitrogen). After transfer for 4 h at 30V, the      mesothelin were PCR amplified to incorporate
PVDF membrane was blocked overnight at 4°C in         NotI and EcoRI restriction sites and cloned into a
1% Western Blocking Reagent (Roche) in Tris           modified pSecTag 2B vector containing an N-
buffered saline (50 mM Tris-HCl, 150 mM NaCl).        terminal rFc fragment. HEK 293T cells were
The membrane was then incubated with OVCAR-           transfected and Fc mesothelin proteins were
3 supernatant + 0.5% blocking solution for 1 h at     collected and purified from the supernatant over
RT. This was followed by incubation with 1:200        the course of 8-10 days. Constructs encoding
dilution of OC125 mAb for 1 h at RT. Detection        amino acid residues 296-390 (Region I), 391-486
was performed with goat anti-mouse IgG-HRP            (Region II), and 487-581 (Region III) of

                                                                                                           Downloaded from www.jbc.org by guest, on August 14, 2011
conjugate (Biosource) 1:1000 for 1 h.                 mesothelin were initially generated and tested.
     Determination of Affinity Constants (KD) – As    Constructs encoding smaller fragments within
previously described (29), equilibrium constants      Region I , Region IAB (296-359), Region IBC
and Scatchard plots were determined by using the      (328-405), Region IA (296-337), Region IB (328-
Marquardt-Levenberg algorithm for nonlinear           369), and Region IC (360-405) were also
regression with the Prism software (version 5.0,      generated. It was found that secretion of the IA
GraphPad Software, San Diego, CA).                    (296-337) fragment was prevented by protein
     Mammalian Cell Display – As previously           aggregation; however, this problem was
described (30), single chain Fv antibody (scFv)       circumvented by moving the rFc to the C-terminus
SS1 was cloned into an expression vector              of the mesothelin fragment.
(pMH112) for cell surface expression of scFv on            PCR mutagenesis was used to generate a panel
HEK-293 cells.                                        of alanine mutants within region 296-359 (Region
     Heterotypic Cancer Cell Adhesion Assay – We      IAB). As shown in Fig. 2A, both human and
followed a recently developed protocol (15).          mouse mesothelin (mnMSLN) proteins bind
Briefly, OVCAR-3 or YOU cells (1 x 105) were          human CA125. It was reasoned that those residues
seeded in triplicate in microplates, incubated        conserved between the two species would be more
overnight at 37˚C/5% CO2 . The following day, H9      likely to be involved in the interaction. Specific
(A431-Mesothelin+) (11) were harvested and            amino acids were targeted based on the homology
loaded with Calcein AM cell dye (Invitrogen) as       between human and murine mesothelin (Fig. 2B).
per the instruction manual. The OVCAR-3 or            In addition, we hypothesized that these residues
YOU monolayers were washed once with 200 µL           are likely to be involved in the carbohydrate
10% complete RPMI and pre-incubated for 1 h at        binding. In total, eight alanine mutant constructs
4oC with purified mesothelin Fc fusion proteins.      were made: Y318A, W321A, E324A, F344A,
The labeled H9 (2 x 105) cells were added to          E347A, K353A, H354A, and K355A. Of these
triplicate wells for 1 h at 4˚C. Wells were gently    mutants, four were secreted at high enough levels
washed five times with 200 µL PBS by inverting        to purify using a Protein A column – Y318A,
the plate on paper towels. Cell adherence and its     W321A, E324A, and H354A. The purity and
inhibition were quantitated using VICTOR3             molecular weight of each purified protein was
Multilabel Counter model 1420 (Perkin Elmer).         confirmed on SDS-PAGE. The other four mutants,
     Statistical Analysis - The data obtained was     F344A, E347A, K353A and K355A, were not
entered in Prism (version 5) for Windows              secreted into the culture supernatants due to
(GraphPad Software,) for statistical analysis. Flow   aggregation     inside     cells   according    to
cytometry raw data were analyzed by ANOVA             immunoblotting of whole cell lysates (data not
Mesothelin and CA125/MUC16, Page 5

shown), indicating that mutations of these residues     for the mesothelin-CA125 interaction is consistent
may cause misfolding of mesothelin. For all             with the value (~5 nM) previously obtained on
proteins purified, a distinct peak was found on a       OVCAR-3 cells by flow cytometry (28). Three
TSK size exclusion column (Fig. 3). The rFc –           smaller fragments (296-337, 328-369, 360-405)
extracellular portion of mesothelin 296-581 fusion      within Region I showed no binding to CA125 in
protein (full-length) was estimated to be               ELISA (data not shown), indicating the first 64
approximately 75-kDa, while truncated mutants           residues at the N-terminus of cell surface
were relatively smaller in molecular weight (~50-       mesothelin is the irreducible binding domain on
kDa for Regions I, II and III and ~40-kDa for IAB,      mesothelin for the CA125 protein. It is striking
IBC and alanine mutants of IAB).                        that substitution of the tyrosine at position 318
     Binding of CA125 to Mesothelin and its             with an alanine (Y318A) completely disrupted the
Mutants - To examine the interaction of CA125           interaction with CA125. Alanine mutations at Glu-
and mesothelin mutants, we used Western blot            324 (E324A; KD = 42.4 nM) and Trp-321
overlay analysis. We ran equal amounts (500 ng)         (W321A; KD = 19.5 nM) partially reduced the
of each protein an SDS-PAGE gel and transferred         binding of mesothelin to CA125. The alanine
to a PVDF membrane. Membrane was blotted with           mutation at His-354 (H354A) did not change the
OVCAR-3 supernatant             containing CA125        mesothelin-CA125 interaction (KD = 2.71 nM).
followed by OC125, an anti-CA125 mAb. As                     To verify that fragments were specifically

                                                                                                               Downloaded from www.jbc.org by guest, on August 14, 2011
shown in Fig. 4, full-length extracellular domain       binding to CA125, an ELISA assay was employed
of mesothelin (296-581), Region I (296-390) and         wherein captured mesothelin fragments were
Region IAB (296-359) bound CA125. A 64 amino            incubated with either OVCAR-3 supernatant
acid fragment (IAB) at the N-terminus of                containing CA125 or supernatant from the
mesothelin (296-359) retained 100% binding              OVCAR-3 with knockdown of CA125 expression.
capability to CA125. However, three smaller             No signal was detected from those fragments
fragments, Region IA (296-337), Region IB (328-         incubated with the supernatant from CA125-
369), and Region IC (360-405), consisting of            knockdown cells (data not shown), indicating that
approximately 42 amino acids covering all the           the binding between mesothelin and CA125 in
residues within Region I showed no binding to           ELISA was specific.
CA125 in Western blot. The alanine mutation at               Cell Binding Assays by Flow Cytometry - To
His-354 did not change the mesothelin-CA125             assess binding to CA125 on cancer cells, wild-type
interaction. Interestingly, alanine mutations at Tyr-   mesothelin and its mutants were incubated with
318 (Y318A) and Glu-324 (E324A) abolished the           cells. As shown in Fig. 6, Region I (a 95 amino
binding of mesothelin to CA125. Alanine mutation        acid fragment consisting of residues 296-390 at
at Trp-321 (W321A) partially reduced the binding        the N-terminus) of mesothelin was found to bind
of mesothelin to CA125.                                 to OVCAR-3 cells, while fragments Region II
     Kinetic Studies by ELISA - To precisely            (391-486) and Region III (487-581) showed no
characterize the interaction between CA125 and          binding. The smallest fragment that still contained
mesothelin and to determine a binding affinity          most (~90%) of binding activity to CA125 was
(KD) for each mesothelin fragment, we used a            Region IAB. Three smaller fragments within
quantitative ELISA-based assay. ELISA plates            Region I, Region IA (296-337), IB (328-369), and
were coated with a goat anti-rabbit IgG antibody        IC (360-405), were also tested. Only Region IB
overnight for capturing the Fc mesothelin mutant        had modest (~10%) CA125-binding activity.
fusion proteins. OVCAR-3 supernatant containing         These data indicate that IAB, the first 64 residues
soluble CA125 was then added, followed by the           at the N-terminus of cell surface mesothelin, is the
OC125 mAb. Overall results from the ELISA were          minimum region which retains the most binding
consistent with those seen in the Western blot          activity to CA125. The IAB binding domain was
overlay assays except for the E324A mutant of           found to bind with comparable affinity, when
mesothelin (Fig. 5). The average KD for the             compared to the full-length mesothelin (FULL).
binding of the Fc fusion protein of wild-type           This suggests that it is primarily the N-terminus of
mature mesothelin (296-598), Region I (296-390)         cell surface mesothelin that is involved in the
and IAB (296-359) to CA125 was ~3 nM. The KD            interaction with CA125 and the minimum
Mesothelin and CA125/MUC16, Page 6

sequence for CA125 binding activity is Region          CA125 interaction but not induce cell adhesion
IAB (328-405).                                         and signaling. To this end, we examined if any of
     Four alanine mutants (Y318A, W321A,               the mesothelin truncated or alanine mutants
E324A, and H354A) generated within the region          reported here can compete with the binding of
296-359 were similarly assessed for their ability to   wild-type mesothelin to CA125. We co-incubated
bind to CA125 on the surface of OVCAR-3 cells.         a Flag-tagged wild-type mesothelin with a panel of
It was found that the substitution of alanine for      our mesothelin mutants (Fig. 8). Region IAB can
tyrosine at residue 318 completely ablated binding     effectively      inhibit  the    mesothelin-CA125
(Fig. 6). The H354A mutant conversely showed           interaction. Other constructs such as Region I and
comparable binding to CA125 as the wild-type           H354A can also inhibit the mesothelin-CA125
region 296-359. Alanine mutants of the tryptophan      interaction while Regions II, III and IBC, mutants
at 321 and glutamic acid at 324 also demonstrated      Y318A, E324A and W321A cannot. We have
decrease in binding (10-20%) to CA125 on the           demonstrated that full-length mesothelin (FULL),
surface of OVCAR-3 cells.                              Region I, IAB and the H354A mutant significantly
     Using ANOVA with Dunnett's and Newman-            inhibited the binding of mesothelin with a Flag tag
Keuls multiple comparison post tests, we have          to CA125 on OVCAR-3 cells, as compared to
demonstrated that full-length mesothelin (FULL),       Regions II, III, IBC and the Y318A, W321A and
Region I, IAB and the H354A mutant significantly       E324A mutants (p < 0.05). As compared to Region

                                                                                                              Downloaded from www.jbc.org by guest, on August 14, 2011
bound to CA125 on OVCAR-3 cells, as compared           IAB, the inhibition of the methelin-CA125
to Regions II, III, IBC, IA, IB, IC and the Y318A,     interaction by full-length mesothelin (FULL) and
W321A and E324A mutants (p < 0.05). As                 Region I was not significant (p > 0.05), clearly
compared to Region IAB, the binding by full-           indicating that IAB is the minimal sequence (296-
length mesothelin (FULL), the H354A mutant and         359) of CA125-binding domain.
Region I was not significant (p > 0.05), indicating         Epitope Mapping of scFv SS1 by Mammalian
that IAB is the minimal sequence of CA125-             Cell Display – Since Fv SS1-derived antibody
binding domain.                                        drugs are currently in several multi-center clinical
     Since surface protein expression of mesothelin    trials to treat mesothelioma and ovarian cancer, we
and CA125 are also found frequently in malignant       decided to examine its epitope using the
mesothelioma, we then examined the binding of          mesothelin fragments we have made. To this end,
mesothelin and its mutants to the YOU                  we used a new method (called “mammalian cell
mesothelioma cell line (Fig. 7). The binding           display”) recently developed in the laboratory (30)
patterns were similar to those seen in OVCAR-3         to express the SS1 scFv on HEK-293 cells. We
cells except for Region I. As shown in Fig. 8,         incubated the HEK-293 cells expressing scFv SS1
wild-type mesothelin (FULL), Regions I and IAB         with the mesothelin and its fragments or mutants.
bound CA125 on YOU cells significantly stronger        As shown in Fig. 9, the epitope of scFv SS1
than Regions II, III and IBC (p < 0.05). Unlike        overlaps the CA125-binding site on mesothelin.
what we saw in OVCAR-3 cells, Region I retained             Inhibition      of   the    Mesothelin-CA125
about 40% of the CA125 binding activity on YOU         Interaction by the SS1 scFv – We investigated
cells. Nevertheless, the binding of Region IAB to      whether or not the SS1 scFv could block the
CA125 on YOU cells is comparable to the full-          mesothelin-CA125 interaction. We made the SS1
length mesothelin (FULL) (p > 0.05). We also           in the format of a scFv SS1-PE38 fusion protein
found that Region IB had modest (~10%) binding         (8). We co-incubated the SS1 or HA22, a scFv-
activity. The alanine replacements at positions        PE38 specific for CD22 (29) with mesothelin and
318, 321 and 324 showed significant decrease of        OVCAR-3 or YOU cells. As shown in Fig. 10, the
CA125 binding activities (10-30%) on YOU cells.        SS1 scFv remarkably inhibited the mesothelin-
     Inhibition     of    the     Mesothelin-CA125     CA125 interaction while the HA22 scFv did not.
Interaction by the CA125-binidng Domain - The               Inhibition of Cancer Cell Adhesion by the
molecular mechanisms underlying the cell               CA125-binding Domain – Finally we explored the
adhesion and signaling induced by the mesothelin-      possibility that the CA125-binding domain can
CA125 interaction are not clear. An ideal              functionally block cancer cell adhesion. We used
antagonist drug would disrupt the mesothelin-          the assay system recently established by Hassan et
Mesothelin and CA125/MUC16, Page 7

al. (15). We measured adhesion of fluorescently             Here we have shown that it is primarily the N-
labeled mesothelin-expressing H9 cells (11) onto       terminus of the extracellular domain of mesothelin
CA125 positive OVCAR-3 or YOU cancer cells.            (residues 296-359) that is sufficient and necessary
Fig. 11 shows the results of this study. Significant   for binding to CA125. A fragment (IAB)
heterotypic cell binding was seen on a monolayer       consisting of the first 64 amino acids displayed
of OVCAR-3 or YOU cells with the control               binding to CA125 on ovarian cancer and
CD22-Fc fusion protein or a control fragment           mesothelioma cells as detected by flow cytometry
(Region IBC). The CA125-binding domain (IAB)-          as well as in ELISA and Western blot assays.
Fc fusion protein completely abolished H9 cell         Other regions of mesothelin (391-486; 487-581)
adhesion onto OVCAR-3 or YOU cells. A                  showed no affinity for CA125. Efforts to further
statistically significant inhibition with the CA125-   narrow down the binding domain with the
domain was detected with concentrations as low as      truncated mutant approach, however, showed that
1 µg/mL (Fig. 11).                                     this fragment was irreducible; smaller fragments
     Structure of CA125-binding Domain - Since a       296-337, 328-369, and 360-405 showed no or only
three-dimensional structure of mesothelin is           modest binding to CA125. This may be due to a
currently not available, the secondary structure       specific conformation that is required for binding
was evaluated by the algorithms PROF                   that is lost when mesothelin is truncated smaller
(http://www.predictprotein.org) and APSSP2             than the 64 amino acid length.

                                                                                                               Downloaded from www.jbc.org by guest, on August 14, 2011
(http://www.imtech.res.in/raghava/apssp2).       The        Our efforts to further find specific amino acids
CA125-binding domain is primarily composed of          within CA125-binding domain (296-359) of
helix-turn-helix repeats (Fig. 12). Interestingly,     mesothelin that are involved in the interaction with
Tyr-318 whose alanine replacement significantly        CA125 led us to mutate several suspected amino
reduces the CA125 binding is located at the coil       acid residues to alanine to see if it would affect
between two helical secondary structures. The          binding affinity. We chose residues that were
tyrosine seems a critical residue which either         homologous between mouse and human and those
directly binds the N-glycan on CA125 or indirectly     likely to be involved in the carbohydrate binding.
plays an important role by maintaining a               The replacement of tyrosine with alanine in the
conformation required for CA125 binding. The           IAB fragment resulted in no binding to CA125 on
partial loss of CA125 binding activities of the        ELISA and Western blot and significantly reduced
W321A and E324A mutants may indicate that              most CA125-binding activity on cancer cells by
other residues near 318 are also involved. The fact    flow cytometry indicating the tyrosine at position
that Region IB alone had only a modest CA125-          318 is critical for either direct binding to CA125
binding activity (~10%) indicates that the CA125-      or maintaining the conformation of mesothelin to
binding domain requires Region IA for its full         allow binding to occur. Alanine mutants of W321
activity.                                              and E324 in the vicinity of the tyrosine at 318 also
                                                       showed decreased binding to CA125 in all assays
DISCUSSION                                             but were not nearly as dramatic. Mutating the
     Cancer cells commonly spread within the           histidine at 354 had no effect on the mesothelin-
peritoneal cavity via seeding to mesothelium-lined     CA125 interaction.
structures. The interaction between CA125, a                During the preparation of SDS-PAGE,
mucin present on a majority of ovarian cancer and      proteins are typically reduced and denatured by
mesothelioma cells, and mesothelin, a GPI-             treatment with the Laemmli sample buffer. Many
anchored glycoprotein present on the mesothelial       protein functions depend on protein secondary and
cells lining along the peritoneal cavity, has been     tertiary structures, which could be disrupted under
suggested to facilitate implantation and metastasis    reducing and denaturing conditions. The fact that
of tumors (26, 28). Blocking the mesothelin-           CA125 bound to mesothelin (296-581), Region I
CA125 interaction may prevent or reverse               (296-390) and IAB (296-359) in Western blot
metastasis and lead to overall improved survival in    overlay assays may indicate that the mesothelin
cancer patients. A better understanding of how         proteins were renatured rapidly during the PVDF
these two proteins interact may eventually aid in      membrane transfer. The renaturing may allow
developing such a therapy.
Mesothelin and CA125/MUC16, Page 8

mesothelin to recover their secondary and tertiary             Furthermore, we have shown that a single
structures specific for CA125 binding.                    chain mAb (SS1) recognizes the CA125-binding
    Mesothelin is predicted to be primarily               domain and blocks the mesothelin-CA125
composed        of      helix-turn-helix       repeats.   interaction on cancer cells. Most interestingly, the
Interestingly, Tyr-318 whose alanine replacement          CA125 binding domain-Fc fusion significantly
abolishes the CA125 binding of IAB is located at          inhibits heterotypic cancer cell adhesion. The SS1
the coil between two helical secondary structures         Fv-derived immunotoxin and a chimeric (the SS1
(Fig. 12), indicating that binding could be               mouse Fv/human Fc) antibody are currently in
dependent on the conformation of this region.             clinical trials for mesothelioma, ovarian cancer,
    We further showed that the CA125-binding              pancreatic cancer and non-small cell lung cancer.
domain (IAB) or its alanine mutant could                       Here, we have shown that mesothelin binds to
effectively block the binding of wild-type                CA125 primarily with the N-terminal portion of
mesothelin to CA125, indicating the CA125                 cell surface mesothelin. The CA125-binding
binding domain Fc-fusion protein is a good                domain of 64 residues, likely a helical structure, is
candidate to be used as a potential antagonist to         irreducible. We have further shown that a tyrosine
inhibit    the    mesothelin-CA125         interaction.   at position 318 is critical for binding to CA125.
Moreover, the Fc portion of the molecule with             The CA125-binding domain significantly inhibits
antibody dependent cell mediated cytotoxicity or          cancer cell adhesion and merits evaluation as a

                                                                                                                  Downloaded from www.jbc.org by guest, on August 14, 2011
complement dependent cytotoxicity could recruit           new therapeutic agent in preventing or treating
NK cells or complements to kill the CA125-                peritoneal malignant tumors.
positive tumor cells. Since Enbel (TNFR-Fc), an
FDA-approved Fc fusion drug, many peptide-Fc              ACKNOWLEDGMENTS
fusion proteins are currently in clinical trials (32).    We thank Ira Pastan for helpful discussions, and
                                                          Anna Mazzuca for editorial assistance.


1.     Runowicz, C. D. (2008) Cancer J. 14, 7-9
2.     Palumbo, C., Bei, R., Procopio, A., and Modesti, A. (2008) Curr. Med. Chem. 15, 855-867
3.     Chang, K., Pastan, I., and Willingham, M. C. (1992) Int. J. Cancer 50, 373-381
4.     Chang, K., and Pastan, I. (1996) Proc. Natl. Acad. Sci. USA 93, 136-140
5.     Hassan, R., and Ho, M. (2008) Eur. J. Cancer 44, 46-53
6.     Miettinen, M., and Sarlomo-Rikala, M. (2003) Am. J. Surg. Pathol. 27, 150–158
7.     Ordonez, N. G. (2003) Am. J. Surg. Pathol. 27, 1418–1428
8.     Ho, M., Bera, T. K., Willingham, M. C., Onda, M., Hassan, R., FitzGerald, D., and Pastan, I.
       (2007) Clin. Cancer Res. 13, 1571-1575
9.     Hellstrom, I., Raycraft, J., Kanan, S., Sardesai, N. Y., Verch, T., Yang, Y., and Hellstrom, K. E.
       (2006) Cancer Epidemiol. Biomarkers Prev. 15, 1014-1020
10.    Ho, M., Onda, M., Wang, Q., Hassan, R., and Pastan, I. (2006) Cancer Epid. Biomarkers Prev. 15,
11.    Ho, M., Hassan, R., Zhang, J., Wang, Q. C., Onda, M., Bera, T., and Pastan, I. (2005) Clin. Cancer
       Res. 11, 3814-3820
12.    Sterman, D. H., Recio, A., Carroll, R. G., Gillespie, C. T., Haas, A., Vachani, A., Kapoor, V., Sun,
       J., Hodinka, R., Brown, J. L., Corbley, M. J., Parr, M., Ho, M., Pastan, I., Machuzak, M.,
       Benedict, W., Zhang, X. Q., Lord, E. M., Litzky, L. A., Heitjan, D. F., June, C. H., Kaiser, L. R.,
       Vonderheide, R. H., Albelda, S. M., and Kanther, M. (2007) Clin. Cancer Res. 13, 4456-4466
13.    Hellstrom, I., Friedman, E., Verch, T., Yang, Y., Korach, J., Jaffar, J., Swisher, E., Zhang, B.,
       Ben-Baruch, G., Tan, M. C., Goedegebuure, P., and Hellstrom, K. E. (2008) Cancer Epidemiol.
       Biomarkers Prev. 17, 1520-1526
14.    Pastan, I., Hassan, R., Fitzgerald, D. J., and Kreitman, R. J. (2006) Nat. Rev. Cancer 6, 559-565
Mesothelin and CA125/MUC16, Page 9

15.   Hassan, R., Ebel, W., Routhier, E. L., Patel, R., Kline, J. B., Zhang, J., Chao, Q., Jacob, S.,
      Turchin, H., Gibbs, L., Phillips, M. D., Mudali, S., Iacobuzio-Donahue, C., Jaffee, E. M., Moreno,
      M., Pastan, I., Sass, P. M., Nicolaides, N. C., and Grasso, L. (2007) Cancer Immun. 7, 20
16.   Desseyn, J. L., Tetaert, D., and Gouyer, V. (2008) Gene 410, 215-222
17.   Bast, R. C., Jr., Feeney, M., Lazarus, H., Nadler, L. M., Colvin, R. B., and Knapp, R. C. (1981) J.
      Clin. Invest. 68, 1331-1337
18.   Yin, B. W. T., and Lloyd, K. O. (2001) J. Biol. Chem. 276, 27371–27375
19.   O'Brien, T. J., Beard, J. B., Underwood, L. J., Dennis, R. A., Santin, A. D., and York, L. (2001)
      Tumour Biol. 22, 348-366
20.   O’Brien, T. J., Beard, J. B., Underwood, L. J., and Shigemasa, K. (2002) Tumour Biol. 23, 154-
21.   Bork, P., and Patthy, L. (1995) Protein Sci. 4, 1421-1425
22.   Maeda, T., Inoue, M., Kshiba, S., Yabuk, T., Aoki, M., Nunokawa, E., Seki, E., Matsuda, T.,
      Motoda, Y., Kobayashi, A., Hiroysau, F., Shirouzu, M., Tereda, T., Hayami, N., Ishizuka, Y.,
      Shinya, N., Tatsuguchi, A., Yoshida, M., Hirota, H., Matsuo, Y., Tani, K., Arakawa, T., Carninci,
      P., Kawai, J., Hayashizaki, Y., Kigawa, T., and Yokoyama, S. (2004) J. Biol. Chem. 279, 13174-
23.   Bast, R. C., Klug, T. L., St. John, E., Jenison, E., Niloff, J. M., Lazarus, H., Berkowitz, R. S.,

                                                                                                             Downloaded from www.jbc.org by guest, on August 14, 2011
      Leavitt, T., Griffiths, C. T., Parker, L., Zurawski, V. R., and Knapp, R. C. (1983) N. Engl. J.
      Med. 309, 883–887
24.   Bateman, A. C., al-Talib, R. K., Newman, T., Williams, J. H., and Herbert, A. (1997)
      Histopathology 30, 49-56
25.   Attanoos, R. L., Webb, R., Dojcinov, S. D., and Gibbs, A. R. (2002) Histopathology 40, 237-244
26.   Rump, A., Morikawa, Y., Tanaka, M., Minami, S., Umesaki, N., Takeuchi, M., and Miyajima, A.
      (2004) J. Biol. Chem. 279, 9190-9198
27.   Scholler, N., Garvik, B., Hayden-Ledbetter, M., Kline, T., and Urban, N. (2007) Cancer Lett. 247,
28.   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., and Patankar, M. S. (2006) Mol. Cancer 5, 50
29.   Ho, M., Kreitman, R. J., Onda, M., and Pastan, I. (2005) J. Biol. Chem. 280, 607-617
30.   Ho, M., Nagata, S., and Pastan, I. (2006) Proc. Natl. Acad. Sci. USA 103, 9637-9642
31.   Kojima, T., Oh-eda, M., Hattori, K., Taniguchi, Y., Tamura, M., Ochi, N., and Yamaguchi, N.
      (1995) J. Biol. Chem. 270, 21984-21990
32.   McGregor, D. P. (2008) Curr. Opin. Pharmacol. 8, 616-619
Mesothelin and CA125/MUC16, Page 10

Primers used to construct truncated mutants of mesothelin. The restriction enzyme sites are

Fragments             Primers (5’ -> 3’)
Region I              Forward:
Region II             Forward:
Region III            Forward:

                                                                                              Downloaded from www.jbc.org by guest, on August 14, 2011
Region IAB            Forward:
Region IBC            Forward:
Region IA             Forward:
Region IB             Forward:
Region IC             Forward:
Mesothelin and CA125/MUC16, Page 11

Primers used to generate alanine replacement mutants of mesothelin. The restriction enzyme sites
are underlined.

Mutants   Primers (5’ -> 3’)
Y318A     Forward:
W321A     Forward:
E324A     Forward:

                                                                                                   Downloaded from www.jbc.org by guest, on August 14, 2011
F344A     Forward:
E347A     Forward:
K353A     Forward:
H354A     Forward:
K355A     Forward:
Mesothelin and CA125/MUC16, Page 12


FIGURE 1. Generation of truncated and alanine replacement mutants of mesothelin. A. The
mesothelin (MSLN) gene encodes a precursor protein of 622 amino acids. On translocation into the
endoplasmic reticulum the N-terminal signal peptide (red; residues 1–33) and the C-terminal GPI anchor
addition signal (blue; a predicted cleavage site: Ser598) are removed and the latter replaced with a GPI
anchor. The MSLN precursor (71-kDa) is cleaved into two products, the 30-kDa megakaryocyte
potentiating factor (MPF; residues Ser34–Arg286) (31) and the 41-kDa GPI-anchored membrane-bound
mature MSLN (orange) starting from Glu296. The proteolytic cleavage region (green) contains a furin
cleavage site at Arg295, and other protease cleavage sites including a trypsin cleavage site at Arg286. The
four predicted N-linked glycans (black lollipops; Asn57, Asn388, Asn488 and Asn515) on mesothelin are
indicated. Truncated mutants (Regions I, II, III, IAB, IBC, IA, IB and IC) were generated as rabbit Fc
fusion proteins to sequentially narrow down the CA125 binding domain of mesothelin. B. Mesothelin and
truncated mutants were generated using a modified pSecTag2B vector that when transfected into HEK
293T cells created secreted rabbit Fc-fusion proteins. A CMV promoter (PCMV) drove the expression of an
Ig-κ signal, followed by a rabbit Fc fragment, a thrombin cleavage site and the desired portion of
mesothelin. A 6xHis tag was added at the C-terminal of the construct.

                                                                                                              Downloaded from www.jbc.org by guest, on August 14, 2011
FIGURE 2. Key residues proposed for alanine mutations. A. OVCAR-3 cells were incubated with 1
µg/mL of human mesothelin (hnMSLN) or mouse mesothelin (mnMSLN) rFc fusion proteins. The
mesothelin binding on OVCAR-3 cells was detected by a goat anti-rabbit IgG Fc PE conjugate. B. Eight
residues identical between hnMSLN and mnMSLN in Region IAB were selected for alanine replacement.
Region IAB starts at E296 and ends at L359 (*) in hnMSLN. Alanine mutants within Region IAB (296-
359) were expressed. Alanine mutants Y318A, W321A, E324A and H354A (arrows) were secreted and
purified for analysis. The other four mutants, F344A, E347A, K353A and K355A, were not secreted and
found aggregated inside transfected HEK 293T cells.

FIGURE 3. Chromatograms from TSK size exclusion columns. Purified mesothelin (A) and truncated
mutants of mesothelin (B-I) and alanine mutants within Region IAB (J-K) were run over a TSK size
exclusion column to verify that the generated proteins were not aggregated. Truncated mutants and
alanine mutants had a dominant peak that matched with the wild-type fragment, indicating that there was
minimal aggregation. A. Full-length mature MSLN 296-581; B. Region I 296-390; C. Region II 391-486;
D. Region III 487-581; E. Region IAB 296-359; F. Region IBC 328-405; G. Region IA 296-337; H.
Region IB 328-369; I. Region IC 360-405; J. Y318A; K. W321A; L. E324A; and M. H354A.

FIGURE 4. Western blots of mesothelin fragments and alanine mutants. Each protein (500 ng) was
run on an SDS-PAGE gel and transferred to a PVDF membrane. Membrane was blotted with OVCAR-3
supernatant containing CA125 followed by OC125 (anti-CA125 mAb). A. Full-length extracellular
domain of mesothelin (FULL, 296-581), Region I (296-390) and Region IAB (296-359) bind CA125.
Regions II (391-486), III (487-581), IBC (327-390), IA (296-337), IB (328-369) and IC (360-405) do not
bind CA125. B. Alanine mutants within Region IAB (296-359) show differential binding. Alanine
mutations at Tyr-318 (Y318A) and Glu-324 (E324A) abolish the binding of mesothelin to CA125.
Alanine mutation at Trp-321 (W321A) partially reduce the binding of mesothelin to CA125. The alanine
mutation at His-354 does not change the mesothelin-CA125 interaction.
Mesothelin and CA125/MUC16, Page 13

FIGURE 5. Biniding kinetics of mesothelin mutants and CA125. Scatchard plots (top left corner) were
made. ELISA plates captured the Fc mesothelin mutant fusion proteins at various concentrations (x axis;
see Experimental Procedures). OVCAR-3 supernatant containing soluble CA125 was then added,
followed by the OC125 mAb and a goat anti-mouse IgG HRP. Visualization was achieved with TMB
detection reagent and absorbance was read at 450 nm (y axis). The full-length mature form of mesothelin
(FULL) bound to CA125 with an approximate affinity of 1.54 nM. Region IAB wild-type (wt) (296-359)
had a KD of 3.35 nM. Substitution of the tyrosine at position 318 with an alanine (Y318A) completely
disrupted the interaction with CA125. Alanine mutation at Glu-324 (E324A; KD = 42.4 nM) and Trp-321
(W321A; KD = 19.5 nM) reduce the binding of mesothelin to CA125. The alanine mutation at His-354
(H354A) does not change the mesothelin-CA125 interaction (KD = 2.71 nM).

FIGURE 6. Binding of truncated mutants and alanine mutants to CA125 on the cell surface of
OVCAR-3 cells. A. OVCAR-3 ovarian cancer cells were incubated with full-length extracellular domain
of mesothelin (FULL), Region I, IAB or IBC. The binding was visualized with a goat anti-rabbit IgG PE-
conjugated secondary antibody by flow cytometry (gray line). Light gray shaded plot: secondary antibody
only. B. A fluorescence intensity (geometrical mean) was used to quantitatively measure the CA125
binding. In each experiment, the binding of the full-length mature form of mesothelin (FULL) to CA125

                                                                                                          Downloaded from www.jbc.org by guest, on August 14, 2011
was determined as 100% of binding. The secondary antibody only was used as a negative (0%) control.
Full-length extracellular domain of mesothelin (296-581), Region I (296-390), Region IAB (296-359) and
the H354A mutant of IAB bound to CA125 on OVCAR-3 cells significantly stronger than any other
fragments or mutants on the figure (* p < 0.05).

FIGURE 7. Binding of truncated mutants and alanine mutants to CA125 on the cell surface of YOU
cells. A. YOU mesothelioma cells were incubated with full-length extracellular domain of mesothelin
(FULL) or a mutant of mesothelin. The binding was visualized with a goat anti-rabbit IgG PE-conjugated
secondary antibody by flow cytometry (gray line). Light gray shaded plot: secondary antibody only. B. A
mean fluorescence intensity (geometrical mean) was used to quantitatively measure the CA125 binding
on YOU cells. In each experiment, the binding of the full-length mesothelin (FULL) to CA125 was
determined as 100% of binding. The secondary antibody only was used as a negative (0%) control. Full-
length extracellular domain of mesothelin (296-581), Region I (296-390), Region IAB (296-359) and the
H354A mutant of IAB bound to CA125 on YOU cells significantly stronger than any other fragments or
mutants on the figure (* p < 0.05).

FIGURE 8. Inhibition of the mesothelin-CA125 interaction by mesothelin mutants. OVCAR-3 cells
were incubated with mesothelin or each mutant and the Flag-tagged mesothelin (see Experimental
Procedures). The binding of Flag-mesothelin (Flag-MSLN) to CA125 was detected by an anti-Flag mAb.
Full-length extracellular domain of mesothelin (296-581), Region I (296-390), CA125 binding domain or
Region IAB (296-359) and the H354A mutant of IAB (dark gray shading) can significantly block the
binding of Flag-tagged MSLN to CA125 on OVCAR-3 cells (* p < 0.05).

FIGURE 9. Epitope mapping of single chain SS1 monoclonal antibody by mammalian cell display.
HEK-293 cells expressing the anti-mesothelin SS1 single chain Fv were incubated with 5 µg/mL of with
full-length extracullular domain of human mesothelin (100% binding), mouse mesothelin (mnMSLN) or a
mutant of human mesothelin. The binding was visualized with a goat anti-rabbit IgG PE-conjugated
secondary antibody by flow cytometry. SS1 recognized the CA125 binding domain in mesothelin.
Mesothelin and CA125/MUC16, Page 14

FIGURE 10. Single chain Fv (scFv) SS1 blocks the mesothelin-CA125 interaction. OVCAR-3 (A, B)
or YOU (C, D) were incubated with an equimolar concentration of mesothelin and a scFv-PE38 fusion
protein (HA22 or SS1P). HA22 contains Fv specific for CD22. SS1P contains the SS1 Fv specific for
mesothelin. SS1P significantly blocked the binding of mesothelin to CA125 on OVCAR-3 cells (B) or
YOU cells (D) while HA22 did not (A and C).

FIGURE 11. The CA125-binding domain blocks the mesothelin/CA125-mediated cancer cell
adhesion. OVCAR-3 (A, C) or YOU cancer cells (B, D) formed monolayers. The OVCAR3 or YOU
cancer cell monolayer was pre-incubated with full-length mesothelin, the CA125-binding domain (IAB),
IBC or CD22 Fc fusion proteins before fluorescently labeled mesothelin-expressing H9 cells were added.
Mesothelin or IAB significantly blocked cancer cell adhesion with concentrations as low as 10 µg/mL on
OVCAR3 cells (A, C) or 1 µg/mL on YOU cells (B, D) as compared to the control protein (CD22 or IBC)
(* p < 0.05).

FIGURE 12. Interaction of CA125 and mesothelin. A. CA125 is heavily glycosylated with both O-
linked and N-linked oligosaccharides. The peptide backbone of CA125 is composed of the N-terminal
region, extensive Ser/Thr/Pro-rich tandem repeat (TR) containing 165 amino acids each with both N- and

                                                                                                             Downloaded from www.jbc.org by guest, on August 14, 2011
O-glycosylations, a SEA domain with high levels of O-glycosylation and a C-terminal region with a short
cytoplasmic tail. Previous studies have shown that the N-glycan (28) in the TR region (27) is required for
the binding of CA125 to mesothelin. B. The secondary structure of mesothelin was evaluated by the
algorithms PROF and APSSP2. The N-glycan moiety of CA125 repeat units binds to CA125-binding
domain (Region IAB), likely the helical structures around Tyr-318 at the N terminal of cell surface
mesothelin. Line: coil. Tube: helix.
Downloaded from www.jbc.org by guest, on August 14, 2011
A 1.50                                                                      B
 Cell adhesion (x 100%)




                          0.25       Mesothelin
                                     CD22                             *
                             0.001   0.01     0.1     1         10    100
                                            Fc fusion (µg/ml)

C1.25                                                                       D

Cell adhesion (x 100%)



                 0.25                   IAB

                    0.001            0.01     0.1      1        10    100
                                            Fc fusion ( µ g/ml)

                                                                Figure 11. Kaneko et al.

To top