Vol. 2, 301-306, March 2003
Molecular Cancer Therapeutics
© 2003 American Association for Cancer Research
Selection and Identification of Mimic Epitopes for Gastric Cancer-associated Antigen MG7 Ag1
Li Xu,
Bo-quan Jin and
Dai-ming Fan2
Institute of Gastroenterology, Xijing Hospital [L. X., D. F.], and Department of Immunology, Fourth Military Medical University [B. J.], Xi’an, Shaanxi, 710032, People’s Republic of China
2 To whom requests for reprints should be addressed, at Institute of Gastroenterology, Xijing Hospital, Fourth Military Medical University, No.15 West Chang Le Road, Xi’an, Shaanxi, 710032, People’s Republic of China. Phone: 86-029-3375221; Fax: 86-029-2539041; E-mail: fandaim{at}fmmu.edu.cn
 |
Abstract
|
|---|
Phage display technology is now well established as an important experimental approach in designing new reagents for diagnosis of diseases and development of novel vaccines. Aiming at identifying possible immunogenic mimotopes of gastric cancer-associated antigen, we used a monoclonal antibody against gastric cancer, named monoclonal antibody MG7, to screen two phage-displayed nanopeptide libraries. All selected phages were used to immunize BALB/c mice separately. By immunohistochemical staining the tissue sections with the immunized mice sera, we identified one phage that induced gastric cancer-specific antibody response. Consistent with our reported results before, we demonstrate that specific immunogenic epitopes can be isolated by screening phage-displayed library even when natural antigens are not readily available.
|
Introduction
|
|---|
Gastric cancer is the most frequent malignancy of the digestive tract in China. Much effort has been dedicated to the development of simpler and more practical diagnostic and immunotherapeutic methods for clinical application. Because the invention of hybridoma technique by Kohler and Milstein, our laboratory has generated a panel of murine mAbs3 against gastric cancer, designated as MG series (1, 2). The corresponding antigens recognized by these antibodies are named MGAgs (3, 4). Previous studies have evaluated the diagnostic and immunological significance of these monoclonal antibodies (5–7). However, many questions remain unanswered before their corresponding antigens and antigenic epitopes are identified.
The phage display system is well suited for antigenic epitope mapping (8–10). The high copy number of recombinant pVIII molecules per phage particle displayed provides a highly sensitive system, and linear and constrained peptide libraries in pVIII have been successfully used for the selection of specific ligands for several different molecules. Therefore, we applied phage display technology in our research to identify the potential antigen epitopes of gastric cancer.
|
Materials and Methods
|
|---|
mAbs and Peptide Libraries.
The monoclonal antibody against human gastric cancer cell line KATO-III, mAb MG7 (IgG2a, k), was established in our laboratory. Immunohistochemistry results suggested that its corresponding antigen localized in cytoplasm and cell membrane. mAb MG7 was purified from the murine ascitic fluid by ammonium sulfate precipitation and chromatographed on a DE52 ion exchange column before being used to screen libraries.
The two phage peptide libraries used in this study, pVIII-9aa and pVIII-9aa-cys, were kindly provided by Felici and Luzzago (11, 12). The phage vector that these two libraries were constructed in was named PC88. Both libraries were derived from 
1010 independent clones carrying nine-amino acid-long random peptide inserts in the NH2-terminal region of the phage major coat protein (pVIII). In pVIII-9aa-cys library, the random inserts were flanked by two cysteine residues and could be cyclically constrained.
Affinity Selection.
Specific phage clones were isolated from the libraries by one round of affinity selection after biopanning procedures (8, 13). Microwells of a 96-well polystyrene plate were preincubated with mAb MG7 at a concentration of 0.1 µg/ml in coating buffer [50 mM NaHCO3 (pH 9.6)] for 12 h at 4°C. Washed with PBS, the wells were blocked with 10 mg/ml BSA in TTBS for 4 h at room temperature, then incubated with an excess of M13K07-UV-killed phage (2 x 1013 phage particles) in TTBS containing 1 mg/ml BSA for 4 h at 4°C before being washed with PBS three times. The phage mixture was added to the wells and incubated for 12 h at 4°C before the wells were washed extensively with TTBS at 4°C again. The bound phages were eluted by incubating the wells with elution buffer [0.1 M HCl (pH 2.2) with glycine] for 10 min with gentle agitation at room temperature. The eluate was transferred to a polypropylene tube and neutralized with 2 M Tris-HCl (pH 9.0). The percentage of clones containing a productive insert was determined by plating infected bacteria on X-Gal/isopropyl-1-thio-ß-D-galactopyranoside indicator plates.
Immunological Screening.
The phage colonies were immunobloted as described (14). Aliquots of 104 eluted phages derived from the first round of biopanning were added to 0.4 ml of an overnight culture of XL1-blue bacteria. After 1 h of incubation at 37°C, 10 ml of Luria-Bertani containing 50 µg Ap/ml were added, and the cultures were grown for 6 h in a 100-ml flask with vigorous shaking. Then, the bacteria were superinfected with 1011 particles of M13KO7 helper phage and incubated for 1 h at 37°C. After being washed twice with 15 ml of phage buffer, the bacteria were spread onto Luria-Bertani agar plates (diameter 10 cm) containing 50 µg Ap/ml and 10 µg Km/ml. After overnight culture, phage-infected bacterial lawns were lifted onto nitrocellulose blots, which were directly transferred into TBST milk (1 x TBS, 0.05% Tween 20, and 5% nonfat dry milk as blocking agent) for blocking. Finally, a standard immunological detection procedure was followed without lysis of the bacterial colonies. Immunostaining was performed by incubating the filters with horseradish peroxidase-conjugated antimouse IgG secondary Ab at a dilution of 1:2500 and subsequently developing with 3,3'-diaminobenzidine. Positive colonies were subcloned and identified again.
Dot Immunoblotting.
Dot immunoblotting was performed as described (15). 5 x 1010 PEG-precipitated phage particles were spotted on nitrocellulose blots. The blots were first blocked with 5% nonfat milk (in PBS/0.05% Tween 20) and subsequently incubated overnight at 4°C with mAb MG7. The filters were washed with PBS/0.05% Tween 20 and then incubated with horseradish peroxidase-conjugated antibody for 1 h at room temperature. After being washed, the filters were developed with enhanced chemiluminescence system. Phage vector PC88 was included as negative control. Only the clones giving the strongest dot-blot signal were chosen and used for further characterization.
ELISA Inhibition Assay.
Assay one: Different PEG-precipitated phage particles were preincubated with mAb MG7 at a concentration of 10 µg/ml. Besides, KATO-III cell suspension at a concentration of 105/ml was added to multiwell plates and fixed with acetone for 30 s before 5% BSA in PBS were added for blocking at 37°C for 2 h. The multiwell plates were then incubated with 100 µl of preincubated phage-mAb MG7 mixture at 37°C for 30 min. Then, after the procedure of avidin-biotin complex method kit (Vector), we developed the plates using TMB and calculated the inhibition rate: inhibition rate = 1 - Test OD450/MG7 OD450.
Assay two: Polystyrene beads (6.4 mm) were coated with mAb MG7 at a concentration of 10 µg/ml in coating buffer for 12 h at 4°C on a rotating wheel. After being washed with PBS, beads were blocked with 10 mg/ml BSA in TBS for 4 h at room temperature. Then, an increasing number of KATO-III cells and 5 x 1010 phages were added and incubated for 6 h at 37°C. Unbinding phages were removed by washing beads extensively with TTBS at 4°C, and binding phages were eluted by incubating the beads with elution buffer for 5 min with gentle agitation at room temperature. The eluate was transferred to a polypropylene tube and neutralized with 2 M Tris-HCl (pH 9.0). Eluted phages were titrated by infection of XL1-blue bacteria. Phage vector particles and mAb MGb2 were used as negative controls.
DNA Sequencing.
Single-stranded DNA was extracted from phage supernatant preparations using M13 purification kit (Promega). Sequencing reaction was conducted according to the protocol provided in ABI PRISM BigDye Terminator sequencing kit and sequenced on ABI PRISM 377 automatic sequencing equipment (PE Biosystems) with M13 "-40" primer. The amino acid sequences were derived from the nucleotide sequences. We used CTL epitope prediction software to predict if these selected phagotopes could also bind to HLA molecules.
Immunization of BALB/c Mice.
Phages for immunization were prepared from XL1-blue-infected bacteria and purified twice by PEG according to the routine method (16). Five to 6-week-old female BALB/c mice were immunized by i.p. injection of 100-µl phage solution without added adjuvant every 3 weeks and bled at day 129. Each kind of phage particle was inoculated into five mice, and antiserum from each mouse was collected separately.
Immunohistochemical Analysis.
To analyze the immunized serum, tissue sections of gastric cancer with mAb MG7-positive staining were used. These formalin-fixed, paraffin-embedded sections were dewaxed in xylene, dehydrated in alcohol, and incubated in hydrogen peroxide to block endogenous peroxidase. After being extensively washed with PBS and blocked with horse serum, the sections were incubated with antiserum as primary antibody in a dilution of 1:100. Then, after the procedure of avidin-biotin complex method kit (Vector), we developed the sections using diaminobenzidine and counterstained with hematoxylin. Antiphage vector serum was used instead of primary antibody for negative control, whereas mAb MG7 was used instead of primary antibody for positive control; both were included in each staining batch.
|
Results
|
|---|
Selection and Screening of Phage Peptide Libraries.
mAb MG7 was used in two independent selection experiments to screen the pVIII-9aa and pVIII-9aa-cys libraries separately. The initial phage number from each library used for biopanning is 109. After one round of panning, the number of eluted phages was 2435 from pVIII-9aa library and 2759 from pVIII-9aa-cys library, indicating that the enriching rates were 4 x 105 and 3.6 x 105, respectively. After phage-colony immunoblotting, positive phages were further tested by dot immunoblotting to exclude false positive caused by bacterial protein. Only clones giving a strong dot-blot signal were chosen and used for further characterization. At this stage, we got 42 positive phages from two libraries, 12 from pVIII-9aa library, and 30 from pVIII-9aa-cys library. The selected phages did not show a reaction to secondary antibody and other monoclonal antibodies established in our lab, such as MGb2 (data not shown). Then, we designed two inhibition assays to study the relationship between phagotopes and natural antigen. First, we tested whether phagotopes could block the binding between natural antigens expressed on KATO-III cells and mAb MG7. As seen from the results, the inhibition rate of 3 phage particles was >80%, 22 between 75 and 80%, 11 between 70 and 75%, and 6 < 70%. (Table 1). Next, we tested whether the natural antigen would block the binding between mAb MG7 and phagotopes. For GC28, e.g. (other phages also got similar results), the curves demonstrated that as the number of cells increased, the number of phages binding to antibody-coated beads decreased (Fig. 1). Thus, the results implied that phage-displayed peptide was probably only a part of the binding site of natural antigen for antibody.
View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
|
Fig. 1. Inhibition of the mAb MG7 binding to the selected phage clones by KATO-III cell. The phage recovery was calculated as Ap® TU (ampicillin resistance transducing unit) by XL1-blue bacteria infection, and the values represent the mean of two independent experiments.
|
|
Sequence Analysis of the Selected Phage Clones.
The amino acid sequences of specific phage clones were obtained and aligned according to their sequence similarity. Sequenced clones could be classified into three groups (Table 2): (a) one group sharing the consensus PLX0–2S; (b) one group having the motif SAVR; and (c) another group containing the sequence XRMX. However, these homologies do not allow a precise assignment of residues involved in mAb MG7 binding, because no data about the three-dimensional structure of the MG7 antigen are available. We also searched the public expressed sequence tag databases for homology of these peptides, but no homology was found in any of the peptides. Because cellular immunization plays an important role in antitumor therapy, we wonder if these peptides could be recognized by HLA molecules. With the aid of HLA molecule-binding prediction software, we can see that most of the selected epitopes could bind with several HLA molecules (Table 3).
Immunogenicity of the Peptide Mimics.
In earlier research, phage particles were shown to perform a "carrier function" that obviates the need for conjugation of peptides to carrier molecules or coinjection of adjuvant (17, 18). Therefore, we immunized mice with PEG-purified phages without adjuvant and used tissue sections of human gastric cancer expressing MG7 antigen to screen antisera by immunohistochemistry. Each group of phage particles was inoculated into 5 mice, and antiserum from each mouse was collected separately. Antisera to phage vector PC88 were collected and used as a negative control. Of 215 mice immunized, 12 mice died during the immunization. Immunohistochemistry assay showed that most of the immunized antisera failed to react with human gastric cancer, except that only antisera from three phage particles showed some encouraging results. For GC8, one antiserum showed positive staining among five antisera collected; for GC9, one antiserum showed positive staining among four antisera collected; and for GC28, four antisera showed positive staining (Fig. 2) among five antisera collected, whereas sections from normal tissue of the same patient failed to be stained (data not shown). We then used the positive antisera to examine the other nine tissue sections of human gastric cancer. From the results (Table 4), we found that among the mAb MG7-positive staining tissue sections, the positive rate of GC28 antisera was 75% (six of eight), and the positive rate of the other two were both <30%. On the basis of these assessments, we suggest that the peptide GC28 phage displayed could induce the response of antibody specific to human gastric cancer in mice.
View larger version (134K):
[in this window]
[in a new window]
[Download PPT slide]
|
Fig. 2. Immunohistochemical analysis of GC28 antisera. A tissue section from human gastric cancer was stained using GC28 antiserum as the primary antibody at a dilution of 1:100 and developed with 3,3'-diaminobenzidine.
|
|
|
Discussion
|
|---|
Over the last 10 years, we have dedicated a quest for new tumor-specific antigens for gastric cancer. Through immunization of mice with human gastric cancer cells, we prepared 14 different strains of hybridomas producing mAbs against gastric cancer (referred to as MG series). These monoclonal antibodies have been subjected to a variety of studies, and their overall positive rate in serum of gastric cancer patients is between 60 and 70%. It is noteworthy that one of these antibodies, mAb MG7, has high positive rate (1, 2). However, we failed to clone the gene encoding its corresponding antigen or antigenic epitopes thus far. Under such circumstances, mapping the mAb MG7 epitope could well contribute to the understanding of its clinical applications, especially the development of vaccine. For this reason, we applied phage-display technology to identify mAb MG7 epitope mimics.
After usual affinity enrichment and immunological screening, we purified selected phages and further tested the phages through dot blot. Then through ELISA inhibition assay, we examined the relationship between the natural antigen expressed on KATO-III cell and epitopes displayed by phage. From the inhibition rate, the result suggested that phagotopes could partially block the binding between natural antigens and mAb MG7, whereas from the titered curve, the result suggested that the natural antigen could completely block the binding between mAb MG7 and phagotopes. These primary results indicate that the binding site of positive phages is probably only part of the natural antigen.
Other previous studies have shown that phagotopes selected for their ability to bind monoclonal antibodies in a specific manner always prove to be good antigen mimics, although this property is not in every case consistent with the ability to induce antibodies recognizing the natural antigen. In general, when the monoclonal antibodies used for the selection are directed at continuous epitopes, a specific immune response against the natural antigen is induced. The situation appears to be more complex when directed at discontinuous epitopes. Because the antigen information of mAb MG7 is not available now, it might be interesting to observe if these selected phagotopes could display an immunogenic mimicry potential. We addressed this question through active immunization of mice. As no purified MG7 antigen was available, we applied immunohistochemical analysis in our study instead. First, we collected tissue sections of human gastric cancer expressing MG7 antigen, and then, using these tissue sections, we screened the immunized serum. Immunostaining shows that some antisera can react with gastric cancer tissues specifically. These results indicate that the peptide-elicited immune response could tell the difference between tumor and normal tissues. This is consistent with other reported results (19, 20) in showing that mimotopes can be suitable substitutes in cases where little is known about the antigen. The reason why most of the selected phages failed to induce effective antibody response is probably because of the nature of the mimotope itself, i.e., the kind of epitope it mimics and the way it is presented to the immune system. Both of these factors are expected to play important roles in immunization.
It is well known that the key to killing cancer cells is to elicit a CTL response, and this recognition process requires the binding of peptide fragments with MHC molecules. Earlier work concerning influenza hemagglutinin had indicated extensive commonality of the B- and T-cell repertoires and that single amino acid substitutions in mutant viruses affect both antibody recognition of the native molecule and T-cell recognition of processed antigen (19). Furthermore, evidence is accumulating that some of the tumor-associated antigens were found to be reactive with T cells, although they were first identified by their ability to react with antibodies, e.g., two HLA-A2-restricted epitopes have been identified on NY-ESO-1 antigen, which is recognized by its corresponding antibody (20). The other good example is MUC1 (21), an epithelial mucin that is aberrantly expressed on human tumor cells. Two antibody epitopes from MUC1 tandem repeat have also been reported to be able to bind with HLA-A2.1, A11, A1, and A3. These results show that, under certain circumstance, antibody epitopes can also serve as T-cell epitopes. Therefore, we are wondering if the sequenced peptides have the potential to bind with HLA molecules. Our prediction results based on HLA motif prediction demonstrate that some of them could bind with several HLA molecules. If these peptides can be further proved to be T-cell epitopes, the immune response these peptides elicited can be applied for novel vaccine design. However, much more extensive analysis in vivo or in vitro will have to be done to support this speculation. Recently, we have constructed a fusion DNA vaccine of GC28 peptide-HSP70. Primary results show that this fusion gene can induce specific humoral immune response and has some antitumor effect in vivo.4
Taken together, our study demonstrates that it is possible to derive mimotopes of antigens, which are not available for practical utility, and the results reported above can be useful for designing new diagnostic reagents and the development of novel vaccines, e.g., immunogenic peptide mimics can be combined to develop a multivalent subunit vaccine. Moreover, if a T lymphocyte epitope can be identified through this strategy, B- and T-cell subunit vaccines will be more effective. As phage particles might not be suitable for vaccination, additional work should be done using synthetic peptides. In the meantime, we have been working on the purification of the antigen recognized by mAb MG7 and will do more to find out the structure and function information of this gastric cancer-associated antigen.
|
Acknowledgments
|
|---|
We thank Dr. Franco Felici for kindly providing us his two peptide libraries and giving us valuable technical guidance. We also thank Prof. Zhu Chen and Sai-Juan Chen for sequencing.
|
Footnotes
|
|---|
1 Supported by the Chinese National Foundation of Natural Sciences Grants 30024002 and 39900068 and National Basic Research Project G198051203. 
3 The abbreviations used are: mAb, monoclonal antibody; TTBS, 1 x TBS, 0.05% Tween 20; PEG, polyethylene glycol; TBS, Tris-buffered saline.
4 X. ning, K. Wu, Y. Shi, L. Xu, and D. Fan. Immunologic properties of a fusion DNA vaccine of gastric cancer MG7-Ag mimotope and heat shock protein 70, manuscript in preparation.
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.
Received 4/25/02;
revised 10/ 9/02;
accepted 1/ 7/02.
|
References
|
|---|
Zhang, X. Y., Li, S., and Fan, D. M. Use of MG series monoclonal antibodies in the diagnosis and experimental targeting therapy of gastric cancer.Chin. Med. Sci. J.
,
6:1
–3,1990
.
Zhang, X. Y. Some recent works on diagnosis and treatment of gastric cancer.World J. Gastroenterology
,
5:1
–2,1999
.[Medline]
Fan, D. M., Zhang, X. Y., Chen, X. T., Mu, Z. X., Hu, J. L., Qiao, T. D., and Chen, B. J. Mouse and human monoclonal antibodies against gastric cancer.Chin. Med. J.
,
101:488
–489,1988
.[Medline]
Chao, C. Y., Lie, J., and Fan, D. M. Immunohistochemical study of monoclonal antibody MGD-1 in gastric carcinoma.Histopathology
,
15:523
–524,1989
.[Medline]
Li, G., Zhang, X. Y., Fan, D. M., and Hu, J. L. Detection of antigen-specific immune complexes in sera of gastric and esophageal cancer patients.Chin. Med. J.
,
107:189
–192,1994
.[Medline]
Ren, J., Fan, D. M., Zh, S. J., Zhang, X. Y., Yang, A. G., Li, M. F., and Chen, Z. Establishment of immuno-PCR technique for the detection of tumor associated antigen MG7-Ag on the gastric cancer cell line.Chin. J. Cancer. Res.
,
7:247
–249,1995
.
Liu, J., Hu, J. L., Zhang, X. Y., Qiao, T. D., Chen, X. T., Wu, K. C., Ding, J., Fan, D. M. The value of MG7 antigen in predicting cancerous change in dysplastic gastric mucosa.IJCP
,
56:169
–172,2002
.
Cortese, R., Felici, F., and Galfre, G. Epitope discovery using peptide librariesdisplayed on phage.Trends Biotechnol.
,
12:262
–265,1994
.[CrossRef][Medline]
Cortese, R., Monaci, P., Luzzago, A., Santini, C., Bartoli, F., Cortese, I., Fortugno, P., Galfre, G., Nicosia, A., and Felici, F. Selection of biologically active peptides by phage display of random peptide libraries.Curr. Opin. Biotechnol.
,
7:616
–619,1996
.[CrossRef][Medline]
Rodi, D. J., and Makowski, L. Phage-display technology–finding a needle in a vast molecular haystack.Curr. Opin. Biotechnol.
,
10:87
–92,1999
.[CrossRef][Medline]
Felici, F., Castagnoli, L., Musacchio, A., Jappelli, R., and Cesareni, G. Selection of antibody ligands from a large library of oligopeptides expressed on a multivalent exposition vector.J. Mol. Bio1.
,
222:301
–305,1991
.[CrossRef][Medline]
Luzzago, A., Felici, F., Tramontano, A., Pessi, A., and Cortese, R. Mimicking of discontinuous epitopes by phage-displayed peptides, I. Epitope mapping of human H ferritin using a phage library of constrained peptides.Gene
,
128:51
–56,1993
.[CrossRef][Medline]
Parmley, S. F., and Smith, G. P. Antibody-selectable filamentous fd phage vectors: affinity purification of target genes.Gene
,
73:305
–306,1988
.[CrossRef][Medline]
Felici, F., Luzzago, A., Folgori, A., and Cortese, R. Mimicking of discontinuous epitopes by phage displayed peptides, II. Selection of clones recognized by a protective monoclonal antibody against the Borderella pertussis toxin froin phage peptide libraries.Gene
,
128:21
–23,1993
.[CrossRef][Medline]
Sambrook, J., Fritsch, E. F., and Maniatis, T. (eds.). Molecular Cloning.A Laboratory Manual
, Ed. 2. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press,1989
.
Folgori, A., Luzzago, A., Monaci, P., Nicosia, A., Cortese, R., and Felici, F. Identification of disease-specific epitopes. Combinatorial peptide library protocols. In: S. Cabilly (ed.),Methods in Molecular Biology
,Vol. 87
, pp.195
–208. Totowa, NJ: Humana Press, Inc.,1989
.
Willis, A. E., Perham, R. N., and Wraith, D. Immunological properties of foreign peptides in multiple display on a filamentous bacteriophage.Gene
,
128:79
–83,1993
.[CrossRef][Medline]
Galfre, G., Monaci, P., Nicosia, A., Luzzago, A., Felici, F., and Cortese, R. Immunization with phage-displayed mimotopes.Methods Enzymol.
,
267:109
–115,1996
.[Medline]
Barbara, C. B., Christine, M. G., David, S. B., John, J. S., and Thomas, D. B. The immune response of balb/c mice to influenza hemagglutinin: commonality of the B cell and T cell repertories and their relevance to antigenic drift.Eur. J. Immunol.
19:515
–521,1989
.[Medline]
Rosenberg, S. A. A new era for cancer immunotherapy based on the genes that encode cancer antigens.Immunity
,
10:281
–283,1999
.[CrossRef][Medline]
Finn, O. J., and Gendler, S. J. Mimic 1 of MUC1.Nat. Biotechnol.
,
16:236
,1998
.[Medline]
This article has been cited by other articles:
|
|
|
|
 
Y. Zhao, Y. Jiang, R. Wang, X. Zheng, X. Wang, B. Jin, Y. Lu, T. Qiao, L. Hong, and D. Fan
Expression and prognostic value of MG7-Ag in patients with surgically resectable esophageal squamous cell carcinoma.
Ann. Surg. Oncol.,
September 1, 2007;
14(9):
2621 - 2627.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
