| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Experimental Therapeutics, Preclinical Pharmacology |
1 Department of Obstetrics and Gynecology and 2 Division of Diagnostic Pathology, School of Medicine, Keio University, Tokyo, Japan; 3 Department of Chemistry, Faculty of Science, Ochanomizu University, Tokyo, Japan; 4 Pharmaceutical Research Laboratory, Kirin Brewery Co., Ltd., Gunma, Japan; and 5 Glycobiology Program, The Burnham Institute, La Jolla, California
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental Design: Transchromosomal KM mice were immunized with the human uterine endometrial cancer cell line SNG-S. Hybridomas were constructed between spleen cells from KM mice and mouse myeloma cells. Reactivity of the antibody was evaluated by immunohistochemistry of pathological specimens of gynecological cancers. Cytotoxicity of HMMC-1 against SNG-S cells was tested by in vitro cytotoxicity assays. The epitope of HMMC-1 was determined by transfection with a panel of glycosyltransferase cDNAs and by inhibition assays with chemically synthesized oligosaccharides.
Results: HMMC-1 is a human IgM monoclonal antibody that reacts positively with müllerian duct-related carcinomas with positive rates of 54.6% against uterine endometrial adenocarcinoma, 76.9% against uterine cervical adenocarcinoma, and 75.0% against epithelial ovarian cancer. HMMC-1 does not react with normal endometrium at proliferative or secretory phases, normal uterine cervix, or normal and malignant tissue from other organs, whereas it reacts weakly with the epithelium of the gall bladder and the collecting duct of the kidney. HMMC-1 exhibits antigen-dependent and complement-mediated cytotoxicity. Upon cotransfection with cDNAs encoding two glycosyltransferases required for fucosylated extended core 1 O-glycan, mammalian cells express HMMC-1 antigen. Finally, binding of HMMC-1 to SNG-S cells is inhibited by synthetic Fuc
1
2Galß14GlcNAcß13Galß13GalNAc1-octyl.
Conclusions: These results indicate that HMMC-1 specifically recognizes a novel O-glycan structure. The unique specificity and cytotoxicity of HMMC-1 strongly suggest a therapeutic potential of this antibody.
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Attempts to generate humanized monoclonal antibodies have been made by generating cross-species hybridomas, as exemplified by our antibody HMST-1 raised against endometrial adenocarcinoma (6) . However, this method has a major obstacle: hybrid cells between human lymphocytes and mouse myelomas survive poorly in culture. Alternatively, recombinant mouse-human chimeric monoclonal antibodies have been produced, as exemplified by Trastuzumab (7) against breast cancer and Rituximab (8) against non-Hodgkin’s lymphoma. However, generation of chimeric monoclonal antibodies is still laborious, preventing rapid production of humanized monoclonal antibodies.
Recently, Tomizuka et al. (9
, 10)
succeeded in producing a transchromosomic KM mouse line in which human chromosomes 2 and 14 are stably introduced. Thus, these KM mice express the complete human immunoglobulin heavy chain and 
light chain (11)
. We immunized KM mice with SNG-S cells, a human uterine endometrial cancer cell line established in our laboratory (6
, 12)
, and produced a human monoclonal antibody against SNG-S cells. This antibody, termed HMMC-1 (human monoclonal antibody against müllerian duct-related carcinoma), reacts specifically with carcinomas of müllerian duct origin. Furthermore, we found that HMMC-1 reacts specifically to a novel mucin type O-glycan carbohydrate structure and exhibits HMMC-1 antigen-dependent and complement-mediated cytotoxicity. These findings strongly indicate the potential utility of HMMC-1 in clinical therapies of these gynecological malignancies.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Immunization of KM Mice and Screening of Hybridomas.
SNG-S cells were injected every 2 weeks (1 x 107 cells/injection) into cross-bred KM mice. Spleen cells from immunized mice were fused to Sp2/0-Ag14 mouse myeloma cells (American Type Culture Collection No. CRL-1581) using S-clone cloning medium (Sankoh Jyunyaku Co., Tokyo, Japan), Hybridoma Cloning Factor (IGEN International, Inc., Gaithersburg, MD), HAT (ICN Biochemicals, Costa Mesa, CA), and HT media (ICN Biochemicals). Clones positive for SNG-S cells and negative for the control SKG-IIIa cells, derived from squamous cervical cell carcinoma, were selected by screening culture supernatants by ELISA (14)
. The positive hybridoma clones were additionally verified by immunohistochemistry with human uterine endometrial adenocarcinoma specimens. All of the animals were maintained and handled in accordance to the rules and regulations established by the Animal Care Committee of School of Medicine, Keio University.
Tissue Preparation and Immunohistochemistry.
Formalin-fixed, paraffin-embedded specimens of normal and malignant human tissues were selected from the pathology files of the Department of Obstetrics and Gynecology and the Division of Diagnostic Pathology, School of Medicine, Keio University. Three-micrometer sections were used for immunohistochemistry. Specimens were blocked with goat F(ab')2 fragment to human chain (bound; Cappel, Aurora, OH) followed by incubation with the hybridoma culture supernatant (5 µg HMMC-1/mL). Biotin-conjugated goat antihuman chain and peroxidase-conjugated avidin (ABC kit; Vector Laboratories, Burlingame, CA) were used. Counterstaining was performed using hematoxylin. Staining patterns in tissue specimens were evaluated by staining intensity and frequency of positive cells. Staining intensity was graded on an arbitrary scale of +, ++, or +++, whereas frequency was classified as <10%, 10% to 50%, and >50%, based on the percentage of positive glandular or carcinoma cells in each section. Combining both intensity and frequency, the reactivity of each specimen was categorized as weak, moderate, or strong positive (15)
.
Complement-Mediated Cytotoxicity Assay.
SNG-S cells and SNG-II cells were grown in 96-well plates. Cells were reacted with HMMC-1 hybridoma supernatant at 4°C for 1 hour. Antibody sensitized cells were washed with cold gelatin veronal buffer followed by incubation with diluted (1:5 to 1:40) active human serum complement or C6-depleted inactive complement (Sigma, St. Louis, MO) at 37°C for 30 minutes. A supernatant of the cells was subjected to a lactate dehydrogenase assay using Cytotox96 kit (Promega, Madison, WI).
Western Blot Analysis.
SNG-S cells were cultured in 60-mm plates for 24-hours in medium containing 0, 5, and 10 mmol/L benzyl N-acetyl -galactosaminide (Sigma), which inhibits glycosylation of O-linked glycoproteins (16)
. SNG-S cells were lysed with 50 mmol/L Tris-HCl buffer (pH 8.0), containing 150 mmol/L NaCl, 0.1% SDS, 0.5% sodium deoxycholate, 1% Triton X-100, and 1 mmol/L phenylmethylsulfonylfluoride, and the extract was applied to Western blot analysis with HMMC-1 antibody. For SDS-PAGE, proteins were resolved by electrophoresis on a Multi SDS-PAGE Gel 10/20 (Daiichi Pure Chemicals, Tokyo, Japan) and were transferred to phenylmethylsulfonylfluoride membrane (Millipore, Bedford, MA). The membrane was soaked in PBS (pH 7.4) containing Block Ace Powder (Dainippon Pharmaceutical, Osaka, Japan) and 0.05% Tween 20. The membrane was reacted with diluted monoclonal antibody in blocking solution (PBS containing 2% skim milk) for 2 hours at room temperature. The membrane was then reacted with horseradish peroxidase-conjugated goat antihuman IgM (Kierkegaard and Perry Laboratories, Gaithersburg, MD) for 1 hour at room temperature. Immunoreactive bands were detected using the enhanced chemiluminescence reagent (American Biosciences, Buckinghamshire, United Kingdom).
Transfection and Immunocytochemistry.
Mammalian expression vectors for 1, 3/4 fucosyltransferase (17)
, core 3ß N-acetylglucosaminyltransferase (18)
, core 2ß N-acetylglucosaminyltransferase-L (19)
, core 2ß N-acetylglucosaminyltransferase-M (20)
, and core 1 ß1, 3-N-acetylglucosaminyltransferase (21)
were kindly provided by Drs. Junya Mitoma and Hiroto Kawashima at the Burnham Institute. An expression vector for 1,2-fucosyltransferase-1 (22)
was kindly provided by Dr. Assou El Battari, Glycobiologie et thérapies antitumorales, INSERM (Marseille, France). The cDNA encoding galactosyltransferase-5 (23)
for synthesis of blood group type 1 structure was kindly provided by Dr. Hisashi Narimatsu, Institute Molecular and Cellular Biology, Tsukuba (Ibaraki, Japan).
SNG-II, COS, and CHO cells were grown on glass coverslips and transfected with cDNAs using LipofectAMINEplus (Invitrogen, Carlsbad, CA). Two days after transfection, expression of HMMC-1 antigen was examined by immunocytochemistry using Histostain-SP kit (Zymed, South San Francisco, CA).
HMMC-1 Inhibition Assay with Synthetic Oligosaccharides.
Synthetic oligosaccharides Galß14GlcNAcß13Galß13GalNAc1octyl (extended core 1 O-glycan), and Galß14(Fuc13)GlcNAcß1octyl (Lewis X antigen) have been described (21)
. Galß14GlcNAcß16GalNAc1para-nitrophenol (core 2 O-glycan) was purchased from Toronto Research Chemicals (North York, Ontario, Canada). Fucose was added to extended core 1 O-glycan and to core 2 O-glycan using soluble recombinant 1,2-fucosyltransferase-1 (22)
. ELISA inhibition assay was performed on SNG-S cells grown on 3 mm2 glass coverslips (SurgePath, Richmond, IL; ref. 24
). Fixed cells were incubated with HMMC-1 with or without oligosaccharide inhibitor, followed by fixation with 1% paraformaldehyde in PBS and inactivation of internal peroxidase with hydrogen oxide in methanol. HMMC-1 bound to SNG-S cells were then reacted with biotinylated antihuman IgM antibody and peroxidase-conjugated avidin. Each coverslip was then incubated in a 96-well plate containing 100 µL peroxidase substrate ABTS (Pierce), and the absorbance at 405 nm was recorded by an ELISA reader.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Immunohistochemistry performed with the human immunoglobulin chain as the secondary antibody showed positive immunostaining, whereas immunostainings using mouse chain as the secondary antibody were negative (data not shown), indicating that HMMC-1 is a human antibody. The immunoglobulin subclass was determined by immunoblotting of the hybridoma supernatant with human IgM and IgG antibodies, which identified HMMC-1 as a member of the human IgM subclass (data not shown).
Immunohistochemistry with HMMC-1.
HMMC-1 did not react with normal endometrium and endocervical glands (Fig. 1, A and B)
, whereas it reacts with both the luminal surface and the cytoplasm of adenocarcinoma cells in uterine endometrial (Fig. 1C)
and endocervical specimens. The luminal surface of ovarian serous adenocarcinoma and peritoneal serous carcinoma was strongly stained with HMMC-1 (Fig. 1, D and H)
. Both the luminal surface and cytoplasm were positive in ovarian mucinous, endometrioid, and clear cell adenocarcinoma (Fig. 1, E–G)
. Serous adenocarcinoma of the fallopian tube was weakly stained with HMMC-1.
|
, no positive reactivity was detected with normal endometrium and only slightly with normal endocervical glands (see Fig. 2
for weak, moderate, and strong immunostainings). On the other hand, HMMC-1 was highly reactive with uterine endometrial adenocarcinoma (54.6%), uterine cervical adenocarcinoma (76.9%), epithelial ovarian carcinoma (75.0%), and peritoneal serous carcinoma (87.5%). The positive rates for cervical squamous cell carcinoma and for serous adenocarcinoma of the fallopian tube were relatively low (26.3% and 25%, respectively). The positive rate for hyperplasia of endometrium was 20.6% and that for atypical hyperplasia was 31.2%. The histologic grades of endometrial adenocarcinomas were as follows: 60.9% for well-differentiated cases, 61.7% for moderately differentiated cases, and 31.8% for poorly differentiated cases. The positive rates for different histopathologic types of uterine cervical adenocarcinoma were 85.1% for mucinous adenocarcinoma and 68.0% for endometrioid adenocarcinoma. The positive rate of each histopathologic type of ovarian carcinoma was 75.0% for serous adenocarcinoma, 87.5% for endometrioid adenocarcinoma, 75.0% for mucinous adenocarcinoma, and 62.5% for clear cell adenocarcinoma.
|
|
. However, no staining was seen in normal tissues from stomach, esophagus, colon, liver, pancreas, lung, thyroid gland, testis, or brain specimens, and no staining was detected in cancer specimens from these organs (data not shown).
|
. SNG-II cells, which are negative for HMMC-1 antigen, did not show cytolysis by this assay (data not shown). These results indicate that HMMC-1 can kill specifically the target cells through complement-mediated cytolysis.
|
, lane 1). When SNG-S cells were cultured in medium containing benzyl N-acetyl -galactosaminide, an inhibitor of mucin type O-glycan synthesis (16)
, the levels of HMMC-1 antigen were significantly reduced (Fig. 4A
, lanes 2 and 3). These results suggest that HMMC-1 antigen is carried by O-linked carbohydrate.
|
-GalNAc (Fig. 4, B and C)
. This is in contrast to Lewis B antigen, which is partially carried by O-glycan but also carried by glycolipids (13)
. These results suggest that HMMC-1 determinant is an O-glycan-specific structure.
Identification of Glycosylation Enzymes Required for Synthesis of HMMC-1 Antigen.
SNG-S cells strongly express HMMC-1 antigen, whereas the parental line, SNG-II (6
, 12)
, barely expresses HMMC-1 antigen. This observation suggested that SNG-II cells might become HMMC-1 positive if supplemented with one or two glycosyltransferases. As shown in Fig. 5A
, panel a, SNG-II cells cotransfected with a mixture of 7 glycosyltransferases (see legend for details) became positive for HMMC-1 antigen, demonstrating that enzymes within this group play a key role in synthesizing the HMMC-1 epitope. When -1,2 fucosyltransferase-1 was deleted from this mixture, HMMC-1 antigen did not appear in transfected SNG-II cells (Fig. 5A
, panel b). Similarly, a mixture of 6 glycosyltransferases without core 1 ß1, 3-N-acetylglucosaminyltransferase could not synthesize HMMC-1 antigen (Fig. 5A
, panel c). These results indicate that 1,2-fucosyltransferase-1 and core 1 ß1, 3-N-acetylglucosaminyltransferase are necessary for expression of HMMC-1 antigen. To determine whether these two enzymes are sufficient for HMMC-1 antigen expression, SNG-II cells were cotransfected with -1,2 fucosyltransferase-1 and core 1 ß1, 3-N-acetylglucosaminyltransferase. As shown in Fig. 5A
, panel d, SNG-II cells cotransfected with these two enzymes were strongly positive for HMMC-1, demonstrating that these enzymes are sufficient for SNG-II cells to synthesize the HMMC-1 epitope.
|
-1,2 fucosyltransferase-1 and core 1 ß1, 3-N-acetylglucosaminyltransferase. These cells became positive for HMMC-1 antigen (Fig. 5A
, panels e and f). These results strongly suggest that the HMMC-1 epitope is determined by a fucosylated and extended core 1 O-glycan, Fuc12Galß14GlcNAcß13Galß13GalNAc1Ser/Thr.
Effect of Synthetic Oligosaccharides on HMMC-1.
To additionally evaluate the specificity of HMMC-1, we chemically synthesized the predicted HMMC-1 antigen oligosaccharide and tested the effect of the oligosaccharide on the reactivity of HMMC-1 to SNG-S cells. An inhibition ELISA assay demonstrated that fucosylated extended core 1 oligosaccharide (Fig. 5B)
specifically inhibited binding activity of HMMC-1. This result establishes the epitope of HMMC-1 to be Fuc12Galß14GlcNAcß13Galß13GalNAc1.
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and 2
; Tables 1
and 2
). The in vivo expression pattern of HMMC-1 antigen in cancer cells is closely related to their cell types during embryonic development. Müllerian carcinoma is often used as clinical terminology to denote the broad category of ovarian epithelial carcinomas, peritoneal carcinomas, and carcinomas of the fallopian tube, as these malignancies are similar in clinical management. In embryonic origin, however, the uterus and fallopian tube originate from the müllerian duct, which itself is derived from the celomic epithelium. Similarly, the surface epithelium of the ovary also originates from the celomic epithelium. Thus, the HMMC-1 antigen is closely related to a celomic epithelium-derived carcinoma. Therefore, we have assigned the term müllerian-duct related carcinomas to designate both müllerian carcinomas as well as uterine carcinomas, all gynecologic organ carcinomas that express the HMMC-1 antigen.
The present study shows that the HMMC-1 epitope is determined by an O-glycan carbohydrate with a novel structure. The hypothesis that the HMMC-1 epitope is determined by an O-glycan is supported by findings that HMMC-1 antigen is abolished upon culturing SNG-S cells with benzyl--GalNAc (Fig. 4)
, an inhibitor of O-glycan synthesis (16)
. Many monoclonal anticarbohydrate antibodies are reactive to the carbohydrate moiety of glycolipid (2)
. However, it is difficult to determine the carbohydrate epitope if an antibody does not react with the glycolipid. For example, it was not possible to determine the MECA79 antibody epitope until core 2 GnT gene knockout mouse was produced (21)
. The present study shows the usefulness of transfection with cDNAs encoding glycosyltransferases to elucidate the epitope structure presented by O-glycans.
Although core 1 O-glycan is widely expressed in mammalian epithelial cells, an extended core 1 structure may be rare, as this structure has been reported only in pig oocyte zona pellucida glycoproteins (25
, 26)
, in high endothelial venules (21)
, and in colon cancer cells (27)
. Thus, coexpression of 1,2-fucosyltransferase-1 and core 1 ß1, 3-N-acetylglucosaminyltransferase in vivo may be limited, leading to an expression pattern of HMMC-1 antigen specific to müllerian duct-related carcinomas.
Human monoclonal antibodies specific to cancer cell surfaces have great therapeutic potential. Thus, the recombinant mouse-human chimeric IgG Trastuzumab increases the clinical benefit of first-line chemotherapy in metastatic breast cancer overexpressing HER2, in addition to its direct actions toward metastatic disease (7
, 28)
. Although our KM mouse produces completely human IgG (11)
, IgMs produced by our KM mice contain mouse J-chain. In this regard, HMMC-1, which is IgM, is a humanized antibody. Nonetheless, HMMC-1 may be an excellent candidate for cancer therapy, because it is capable of initiating complement-mediated cytotoxicity (Fig. 3)
. Given that anticarbohydrate IgM antibody suppresses cancers in vivo (29)
, HMMC-1 could potentially exhibit anticancer activity in vivo. Due to the i.p. spread demonstrated by most HMMC-1-positive müllerian ductrelated carcinomas, HMMC-1 could be particularly effective upon i.p. administration to cancer patients. We are, however, aware of the cautions in predicting clinical application of IgM antibody. The utility of such antibodies depends in part on the expression pattern of the HMMC-1 antigen in humans, which requires additional analysis. Also, the in vivo activity of HMMC-1 should be determined using tumor-bearing immunodeficient mice. Such analysis will allow us to determine whether this new human monoclonal antibody represents a novel strategy against gynecological malignancies.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: S. Nozawa and D. Aoki contributed equally to this work; F. Belot is currently at the Institut Pasteur, Unité de Chimie Organique, URA Centre National de la Recherche Scientifique, Paris, France.
Requests for reprints: Shiro Nozawa, Department of Obstetrics and Gynecology, School of Medicine, Keio University, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan. Phone: 81-3-5363-3817; Fax: 81-3-3358-6957; E-mail: shiron{at}sc.itc.keio.ac.jp
Received 4/23/04; revised 7/ 7/04; accepted 7/15/04.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
