Production and Characterization of a Murine/Human Chimeric Anti-Idiotype Antibody That Mimics Ganglioside1



Similar documents
pcas-guide System Validation in Genome Editing

Hands on Simulation of Mutation

UNIVERSITETET I OSLO Det matematisk-naturvitenskapelige fakultet

GENEWIZ, Inc. DNA Sequencing Service Details for USC Norris Comprehensive Cancer Center DNA Core

( TUTORIAL. (July 2006)

DNA Sample preparation and Submission Guidelines

Inverse PCR & Cycle Sequencing of P Element Insertions for STS Generation

10 µg lyophilized plasmid DNA (store lyophilized plasmid at 20 C)

KMS-Specialist & Customized Biosimilar Service

Chapter 2 Antibodies. Contents. Introduction

pcmv6-neo Vector Application Guide Contents

The p53 MUTATION HANDBOOK

(A) Microarray analysis was performed on ATM and MDM isolated from 4 obese donors.

Applications of Ab Molecules. Chapter 4 Monoclonal Ab (p.99) Chapter 5 Ab genes and Ab Engineering (p.128)

Optimal Conditions for F(ab ) 2 Antibody Fragment Production from Mouse IgG2a

Part ONE. a. Assuming each of the four bases occurs with equal probability, how many bits of information does a nucleotide contain?

Mutations and Genetic Variability. 1. What is occurring in the diagram below?

HuCAL Custom Monoclonal Antibodies

Interim Progress Report R&D Project 348. Development of a Field Test Kit for Detection of Blue-Green Algal Toxins

CUSTOM ANTIBODIES. Fully customised services: rat and murine monoclonals, rat and rabbit polyclonals, antibody characterisation, antigen preparation

Chapter 5: Organization and Expression of Immunoglobulin Genes

Chapter 18: Applications of Immunology

Your partner in immunology

Antibody Function & Structure

The Use of Antibodies in Immunoassays

Guidance for Industry

Table S1. Related to Figure 4

Anti-ATF6 α antibody, mouse monoclonal (1-7)

Classic Immunoprecipitation

1.Gene Synthesis. 2.Peptide & Phospho-P. Assembly PCR. Design & Synthesis. Advantages. Specifications. Advantages

CHAPTER 6 ANTIBODY GENETICS: ISOTYPES, ALLOTYPES, IDIOTYPES

MAB Solut. MABSolys Génopole Campus 1 5 rue Henri Desbruères Evry Cedex. is involved at each stage of your project

Types, production of antibodies and Antibody/antigen interaction

Aviva Systems Biology

Geniron. Custom Antibody Services. Serum services Antibody Monoclonal. Purification Antibody Mono Y Genetic Immunization Genbody Polyclonal Antibody

Some terms: An antigen is a molecule or pathogen capable of eliciting an immune response

CHAPTER 8 IMMUNOLOGICAL IMPLICATIONS OF PEPTIDE CARBOHYDRATE MIMICRY

ELISA BIO 110 Lab 1. Immunity and Disease

Antigens & Antibodies II. Polyclonal antibodies vs Monoclonal antibodies

Molecular Facts and Figures

Supplementary Online Material for Morris et al. sirna-induced transcriptional gene

Pharmaceutical Biotechnology. Recombinant DNA technology Western blotting and SDS-PAGE

The immune response Antibodies Antigens Epitopes (antigenic determinants) the part of a protein antigen recognized by an antibody Haptens small

a. Ribosomal RNA rrna a type ofrna that combines with proteins to form Ribosomes on which polypeptide chains of proteins are assembled

Basics of Immunology

Molecular analyses of EGFR: mutation and amplification detection

Chapter 3 Contd. Western blotting & SDS PAGE

Next generation sequencing and proteomics. to study the antibody repertoire. and generate monoclonal antibodies

Antibody Production Price List

Expression and Purification of Recombinant Protein in bacteria and Yeast. Presented By: Puspa pandey, Mohit sachdeva & Ming yu

Next Generation Sequencing

Chapter 3.2» Custom Monoclonal

HuCAL Custom Monoclonal Antibodies

Aviva Systems Biology

Analyzing antibody sequence for recombinant antibody expression. Hangxing Yu, Ph.D Senior Scientist, GenScript May 20, 2015

Chapter 6: Antigen-Antibody Interactions

Gene Synthesis 191. Mutagenesis 194. Gene Cloning 196. AccuGeneBlock Service 198. Gene Synthesis FAQs 201. User Protocol 204

LESSON 3: ANTIBODIES/BCR/B-CELL RESPONSES

Antibody Structure, and the Generation of B-cell Diversity CHAPTER 4 04/05/15. Different Immunoglobulins

Antibody Purification and Labeling

SERVICES CATALOGUE WITH SUBMISSION GUIDELINES

Chapter 6. Antigen-Antibody Properties 10/3/2012. Antigen-Antibody Interactions: Principles and Applications. Precipitin reactions

PCR was carried out in a reaction volume of 20 µl using the ABI AmpliTaq GOLD kit (ABI,

MONOCLONAL ANTIBODY PRODUCTION

ISTEP+: Biology I End-of-Course Assessment Released Items and Scoring Notes

Custom Antibody Services

Custom Antibody Services

A. A peptide with 12 amino acids has the following amino acid composition: 2 Met, 1 Tyr, 1 Trp, 2 Glu, 1 Lys, 1 Arg, 1 Thr, 1 Asn, 1 Ile, 1 Cys

Final Review. Aptamers. Making Aptamers: SELEX 6/3/2011. sirna and mirna. Central Dogma. RNAi: A translation regulation mechanism.

THE His Tag Antibody, mab, Mouse

Advanced BioDesign Outlines Solutions. Antibody Overview. by Advanced BioDesign. Project Start. Immunogenicity. Selecting Your Antigen

IV. -Amino Acids: carboxyl and amino groups bonded to -Carbon. V. Polypeptides and Proteins

QED Bioscience Inc. Revision 05/04/11 ADVANCED ANTIBODY TECHNOLOGIES

Mutation. Mutation provides raw material to evolution. Different kinds of mutations have different effects

Guide to Purification of Polyclonal Antibodies

Understanding the immune response to bacterial infections

Pipe Cleaner Proteins. Essential question: How does the structure of proteins relate to their function in the cell?

Elabscience Biotechnology Co.,Ltd

Rubisco; easy Purification and Immunochemical Determination

Molecular Biology Techniques: A Classroom Laboratory Manual THIRD EDITION

Custom Antibodies Services. GeneCust Europe. GeneCust Europe

Mouse Typer Sub-Isotyping Kit Instruction Manual

perfectprotein Antibodies ntelechon

Dot Blot Analysis. Teacher s Guidebook. (Cat. # BE 502) think proteins! think G-Biosciences

Methods for Protein Analysis

Protocol for Western Blotting

Amino Acids, Peptides, Proteins

Covalent Conjugation to Cytodiagnostics Carboxylated Gold Nanoparticles Tech Note #105

Advanced Medicinal & Pharmaceutical Chemistry CHEM 5412 Dept. of Chemistry, TAMUK

TABLE OF CONTENT. Page ACKNOWLEDGEMENTS. iii ENGLISH ABSTRACT THAI ABSTRACT. vii LIST OF TABLES LIST OF FIGURES. xvi ABBREVIATIONS.

DNA Bracelets

Chapter 9. Applications of probability. 9.1 The genetic code

STUDIES ON SEED STORAGE PROTEINS OF SOME ECONOMICALLY MINOR PLANTS

VLLM0421c Medical Microbiology I, practical sessions. Protocol to topic J10

ECL Western Blotting Substrate INSTRUCTIONS FOR USE OF PRODUCTS W1001 AND W1015.

GenScript Antibody Services

CHAPTER 9 IMMUNOGLOBULIN BIOSYNTHESIS

Provincial Exam Questions. 9. Give one role of each of the following nucleic acids in the production of an enzyme.

In Vitro And In Vivo Production Of Antibodies

Drosophila NK-homeobox genes

Transcription:

(CANCER RESEARCH 52. 1681-1686. April I. 1992] Production and Characterization of a Murine/Human Chimeric Anti-Idiotype Antibody That Mimics Ganglioside1 Alice Hastings, Sherie L. Morrison, Shinji Kanda, Romaine E. Saxton, and Reiko F. Irk- Department of Surgical Oncology, School of Medicine  A,H., S. K., R. E. S., R. F. I.] and Department of Microbiology and Molecular Genetics, The Molecular Biology Institute [A. H., S. L. M.J, University of California, Los Angeles-Jonsson Comprehensive Cancer Center, The University of California, Los Angeles, California 90024 ABSTRACT The \ i and VMfrom a murine anti-idiotypic antibody that mimics ganglioside have been cloned, sequenced, and expressed as a chimeric mouse/human IgGl antibody. The chimeric antibody retained a binding specificity indistinguishable from the original murine antibody. The V,, was a member of Vgam 3.8 family. The sequences are discussed in terms of ways in which proteins may mimic ganglioside epitopes. INTRODUCTION Since our initial report that gangliosides are immunogenic tumor cell surface antigens in cancer patients (1, 2), we have attempted to use monoclonal antibodies against several tumorassociated gangliosides in immunotherapy. Human IgM mono clonal antibodies that specifically recognize different ganglio sides of human malignant melanoma were developed in this laboratory (3,4). These antibodies induced complete regression of recurrent cutaneous melanoma after intratumor injection (5, 6). Our goal now is to extend our studies into active immuno therapy using ganglioside vaccines. However, ganglioside vac cines have several significant drawbacks: (a) tumor-associated gangliosides are present only in limited quantities and are relatively difficult to purify; and (A) in general, purified ganglio sides are poorly immunogenic. One way to produce better tumor specific immunogens is to exploit the idiotypic network. In this approach, the epitope responsible for eliciting tumor specific immunity is represented by an idiotypic determinant present on an antibody molecule. In the case of ganglioside antigens, the anti-idiotypic antibody is produced to an anti-ganglioside antibody. The ganglioside epitope, a carbohydrate determinant, is now represented by a protein epitope. In this context, the epitope should be more immunogenic because it should elicit a T-cell dependent im mune response. Additionally, the antigen is now available in potentially unlimited quantities and is easy to purify. Recently we have produced several murine monoclonal antiidiotypic antibodies specific for L612, a human monoclonal antibody that recognizes ganglioside GM3 on human mela noma (7). One, designated 4C10, bears the internal image of the original GM3 antigen, and will elicit an anti-gm3 antibody response when injected into syngeneic mice (7). Thus, the 4C10 peptide epitope mimics a carbohydrate determinant and can act as a surrogate antigen to induce GM3 specific antibodies. We now describe the cloning and characterization of the VM and V, from this murine antibody and their expression as a murine/lui man chimeric antibody. The VH of the anti-idiotypic antibody is a member of the under-represented Vgam 3.8 gene Received 10/7/91; accepted 1/20/92. 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. ' This study was supported by NIH/National Cancer Institute Grants CAI6858, â "lo whom CAI2582, requests CA42396. for reprints and CA30647. should be addressed, at: Department of Surgical Oncology, School of Medicine, The University of California, Los An geles, CA 90024. 1681 family (8). The structure of the variable region of the antibody is discussed in terms of ways in which the protein may mimic the carbohydrate determinant. This chimeric antibody has po tential as a human therapeutic agent if the principle response it elicits in humans is anti-idiotypic and hence antiganglioside. In addition, with the cloned genes available, they can now be manipulated to improve the potential efficacy of this protein as a tumor specific vaccine. MATERIALS AND METHODS Polymerase Chain Reaction Cloning of the Variable Regions of 4C10 RNA was prepared from the 4C10 mouse myeloma cell line using guanidinium thiocyanate (9) and the poly A-containing fraction isolated with oligo dt cellulose. One ug of polyadenylated* mrna was mixed with 100 ng of 3' primer. dntps were added to a final concentration of 200  im,mgcl2 to 1.5 mm, KC1 to 50 mm, Tris-Cl ph 8.3 to 10 HIM, and gelatin to 0.01%. The reaction mixture was heated to 70 Cand cooled, and 20 units of reverse transcriptase were added and incubated for l h at 37'C; 100 ng of the 5' primer were then added and amplification continued for 25 cycles. The primers used were: Heavy Chain Leader: CATAGGATATCCACCATGGGATGGAG- CTGGATC. This contains an EcoRV site to facilitate cloning into the promoter. Heavy Chain J Region: CTTGGTGCTAGCTGCAGAGACAGTGAC- CAG. This contains an Nhe\ site for cloning into CH1 of IgG. Light Chain Leader: CATAGGATATCCACCATGGAGACAGACA- CACTC. This contains an EcoRV site to facilitate cloning into the promoter. Light Chain J Region: GGAAGTCGACTTACGTTTGATTTCCAG- CTTGGAG. This contains a San site for cloning into the intron. After PCR' amplification, the products were digested with the ap propriate restriction endonucleases: EcoRV and Nhe\ for heavy chain and EcoRV and San for light chain. The heavy variable region was cloned into Bluescript containing an Nhe\ site that had been produced by ligating.\7icl linkers into the Smal site. The light chain was cloned into EcoRV and Sa/I cut Bluescript. The variable regions were initially sequenced in Bluescript to verify that they encode a functional domain. Then the PCR amplified mouse VMDNA was inserted into a psv2 human CH IgGl vector with an upstream EcoRV and Nhel site at the 5'-end of the CH1 constant region.'1 The PCR amplified mouse VL DNA was also inserted into a psv2 human C, in psv2 vector containig J intron separating V from C region.4 Three independent clones of each V region were isolated and sequenced to determine whether any PCR errors had occurred. A PCR generated C to G substitution was observed at amino acid 27d of the light chain and at amino acid 95 of the heavy chain variable regions; this resulted in a conservative Ser to Thr substitution in the light chain while the heavy chain mutation was silent. Transfection Vectors were linearized at the unique Pvu\ site; 1.1 x IO7 P3X63.Ag8.653 nonproducing myeloma cells were suspended in 1 ml 3The abbreviations used are: PCR, polymerase chain reaction; PBS, phos phate-buffered saline; ELISA, enzyme-linked immunosorbent assay; SDS, sodium dodecyl sulfate; TPBS, phosphate-buffered saline containing 0.05% Tween 20. 4 A. Hastings. L. A. Wims, M. J. Coloma, and S. L. Morrison, manuscript in preparation.

of PBS containing 10 pg of each vector. Cells were electroporated at 200 V, 960 fã Fusing a Gene Pulser (Bio-Rad Laboratories, Richmond, CA), diluted to 2.2 x 106/ml with Iscove's modification of Dulbecco's Medium (GIBCO Laboratories, Grand Island, NY) supplemented with 10% iron supplemented calf serum (Hyclone Laboratories, Inc., Logan, UT), and plated onto 96-well microtiter dishes, 125 ^I/well. After 48 h, histidinol (10) was added to 10 m\i and mycophenolic acid to 3 ng/ ml. To screen for producing clones, ELISA plates were coated with an anti-human *-chain antiserum (Sigma Chemical Co., St. Louis, MO). After adding culture supernatant s and washing off unbound antibodies, the plates were developed with alkaline phosphatase labeled anti-human 7-chain (Sigma). The frequency of surviving clones was 1.7 x 10~5;the frequency of clones secreting both heavy and light chains was 6.2 x io-6. Characterization of Chimeric Antibody To characterize the assembly, secretion, and molecular weight of the immunoglobulin, cells were labeled with "S-methionine and cytoplasmic lysates and secretions prepared. Antibody molecules were ini munoprecipitated with polyclonal rabbit Ab against human Fc and Staphylococcus aureus protein A (IgGsorb; The Enzyme Center, Mai den, MA) and analyzed by SDS-polyacrylamide gel electrophoresis (11) with and without reduction of the disulfide bonds. Purification of Chimeric Antibody Antibody secreting transfectomas were cultured in RPMI 1640 (GIBCO) supplemented with 5% fetal calf serum (Gemini Bioproducts, Calabasas, à A)and antibiotics (penicillin, streptomycin, and Fungizone; GIBCO) in humidified 5% CO2/95% air at 37 C.After 4 days in culture, one-tenth of the cells were transferred to fresh medium and maintained. The remaining cells were washed with serum free RPMI 1640 3 times and subcultured in serum-free medium, AIM-V medium (GIBCO), for an additional 4 days. The serum-free spent supernatant was then obtained by centrifugation at 2000 x g for 10 min and pelleted cells discarded. The chimeric antibody in pooled serum-free spent medium was precipitated by slow addition of solid ammonium sulfate to 50% saturation at 22'C. The protein precipitate was obtained by centrifugation at 4000 x g for 20 min. After resuspension and dialysis against PBS at 4'C, the chimeric antibody was purified on an affinity column (5 ml bed volume of protein G Sepharose 4B Fast Flow; Pharmacia 1KB Biotechnology, Inc., Piscataway, NJ) equilibrated with TPBS. The dialysate was applied repeatedly to the column at half bed volumes with protein G binding for l h at 22*C for each sample. After washing with 10 bed volumes of TPBS, the chimeric antibody was eluted with 0.1 M glycine HCl, ph 2.8, and neutralized immediately by adding a small volume of 1.5 M Tris-HCl, ph 8.8. The fractions containing the chimeric antibody were pooled, concentrated, and di alyzed against PBS. The quantity of purified IgG was calculated based on.-i.-.,,of 1.35 units for a concentration of 1mg/ml. Immunochemical Analysis SDS-polyacrylamide gradient (4-20%) gel electrophoresis (11) fol lowed by Western blotting (12) was carried out as described previously. For the detection of human or mouse IgGs, anti-human or anti-mouse IgG antibodies conjugated with peroxidase were used. For the detection of reactivity with the human IgM monoclonal antibody L612, the blotted strip was incubated with 20 jig/ml of L612 in TPBS at 22"C for 1.5 h, then bound L612 was detected with peroxidase conjugated anti-human IgM. 4-Chloro-l-naphthol was used as the substrate for the peroxidase reaction. ELISA and cell-elisa inhibition assays were used to test the specificity of the chimeric antibody as described previ ously (7). Anti-human IgG was obtained from Dako Corp., Carpenteria, CA. Other antisera were obtained from Boehringer Mannheim Biochemicals, Indianapolis, IN; human IgG and IgM were from Sigma, and mouse IgG was from Calbiochem Corp., La Jolla, CA. Human monoclonal antibodies L612 and L72 were purified as described pre viously (4). CHIMER1C ANTI-IDIOTYPE MIMICKING GANGLIOSIDE RESULTS Cloning of V. and Vn cdna Sequences. Direct sequencing of mrna prepared from 4C10 with a murine C, primer indicated that the light chain used J.I; the sequence of FR3 indicated that the light chain was in the V.III group of Kabat et al. (13). Many members of that group share similar or identical leader sequences. Therefore, a consensus leader primer was synthe sized (ATGGAGACAGACACACTC) and in conjunction with a J.I primer was used to amplify mrna, which had been reverse transcribed using a C, primer. Sequencing of 2 clones revealed one that was a productively rearranged light chain and used J. 1, except that Thr 102 (ACC) was replaced by Ser (TCC). In the data base of Kabat et al. (13), there is no other example of this conservative substitution. The light chain belongs to the V.21 family of Potter et al. (14); all members of this family contain 39 variable region residues up to invariant Trp 35 and thus all have a CDR1 of IS amino acids. Search of the nucleic acid data base yielded one germ-line V. gene with only 6 nucleotides different from the V. of the anti-idiotype ( 15). Three of these changes are silent (Fig. 1); at position 21 Met replaces He, at position 29 Val replaces Gly, and at position 66 Glu replaces Gly. One of the substitutions is within a complemen tary-determining region, while the other two are within frame works. Before cloning the heavy chain, we first determined the sequence of the entire V,, mrna using a mouse primer and a set of intermediate primers constructed based on partial se quence determination. Using the appropriate PCR primers the VH was amplified, cloned into both Bluescript and the expres sion vector, and sequenced (Fig. 2). The results show that VH -20-15 -10 Met Glu Thr Asp Thr Leu Leu Leu Trp Val Leu Leu Leu Trp Val ATG GAG ACA GAC ACA CTC CTG CTA TGG GTG CTG CTG CTC TGG GTT -5 Pro Qly Ser CCA QGT TCC -l l Thr Gly Asp ACA GGT GAC Ile Val ATC GTG T 5 10 Leu Thr Gin Ser Pro Ala Ser CTG ACC CAà TCT CCA GCT TCT 15 20 25 Leu Ala Val Ser Leu GLY Gin Arg Ala Thr SusteSer Cys Ara Ala TTG GCT GTG TCT CTA GGG CAO AGG GCC ACC ATG TCC TOC AGA GCC A 27 a b Ser Glu Ser val e d 28 30 35 ASP Ser Tvr y&s Asn Ser Phe Met Hi 8 Trp Tyr AGT GAA AGT GTT GAT AST TAT GTC AAT AGT TTT ATG CAC TGG TAC G Gin Gin Lys CAG CAO AAA 40 Pro Gly Gin CCA GGA CAG Pro Pro CCA CCC 45 50 Lys Leu Leu Ile Tyr Ara Ala AAA CTC CTC ATC TAT CGT OCA 55 60 65 Ser Asn Leu Glu Ser Gly Ile Pro Ala Arg Phe Ser Gly Ser  ta TCT AAC CTA GAA TCT GGG ATC CCT GCC AGG TTC AGT GGC AGT GAG C G 70 75 80 Ser Arg Thr Asp Phe Thr Leu Thr Ile Asn Pro Val Glu Ala Asp TCT AGG ACA GAC TTC ACC CTC ACC ATT AAT CCT GTG GAG GCT GAT 85 90 95 a Asp Val Ala Thr Tyr Tyr Cys Gin Gin Ser Asn Glu Asp Pro Thr GAT GTT GCA ACC TAT TAT TOT CAG CAA AGT AAT GAG GAT CCC ACG C 96 100 * 105 Tro Thr Phe Gly Gly Gly Ser Lys Leu Glu Ile Lys TGG ACG TTC GGT GGA GGC TCC AAG CTG GAA ATC AAA I- J«1-1 Fig. 1. Nucleotide and deduced amino acid sequence of the light chain of 4C 10. Underlined amino acids, the complementarity determining regions; solid line under the sequence, J. region. The site of leader cleavage is at amino acid / and the hydrophobic leader sequence is in italics. Nucleotides present in the germline VLand substituted in the VL of 4C10 are shown under the sequence. Amino acid substitutions from the germline gene are outlined; *, Unusual Ser in J.I. The nucleotide at which a PCR substitution occurred is outlined. 1682

-19-15 -10-5 Met Asp Trp Lea Trp Asn Leu Leu Phe Pro Met Ala Ala Ala Gin ATG GAT TOG CTG TOG AAC TTG CTA TTC CCG ATG GCA GCT GCC CAA -11 5 Ser Ile Gin Ala Gin Ile Gin Leu Val Gin AGT ATC CAA GCA GAG ATC CAG TTG GTG CAG 15 20 Lys Lya Pro Gly Glu Thr Val Lys Ile Ser AAG AAG CCT GGA GAG ACA GTC AAG ATC TCC 30 35 Tyr Thr Phe Thr Asn Tvr Glv Met Asn Trp TAT ACC TTC ACA AAC TAT GGA ATG AAC TGG Gly Lys Gly GGA AAG GGT 45 Leu ITA 60 Glu Pro Thr Tvr 'ate GAG CCA ACA TAT ACT Lys Trp Met Gly AAG TGG ATG GGC  3LB@iiB Phe G T C 75 Leu Glu Thr Ser Ala à sa Thr Ala TTG GAA ACC TCT GCC AAC ACT GCC G 83 85 90 Lys Asn Glu Asp Thr Ala Thr Tyr AAA AAT GAG GAG ACG GCT ACA TAT CHIMERIC ANTl-lDIOTYPE MIMICKING GANGLIOSIDE 10 Ser Gly Pro Glu Leu TCT GGG CCT GAG CTG A 25 Cys Lys Ala Ser Gly TOC AAG GCC TCT GGG T 40 Val Lys Gin Ala Pro GTG AAG CAG GCT CCA 50 52 a 53 55 Trp Ile Asn Thr  feâ 3&Thr Glv TGG ATA AAC ACC AAC ACT GGA T 65 70 Lvs Glv Arg Phe Ala Phe Ser GAA GAG TTC AAG GGA CGG TTT GCC TTC TCT 80 82 a b c Tyr Leu --Ile Asn Asn Leu TAT TTG CTG ATC AAC AAC CTC A 95 Phe Cys Ala Arg Glv Glu Glv TTC TOT GCA AGA GG GAA GGT 100 a b 101 105 110 His Ala Trn Glv Phe Ala Tvr Trp Gly Gin Gly Thr Leu Val Thr CAC GCG TGG GGG TTT GCT TAC TGG GGC CAA GGG ACT CTG GTC ACT Val Ser Ala GTC TCT OCA 1 I JH3 - Fig. 2. Nucleotide and deduced amino acid sequence of the heavy chain of 4C10. Underlined amino acids, the complementarity determining region. Solid line under the sequence, the JH region. The hydrophobic leader sequence is shown in italics. Nucleotides present in the prototype Vgam 8.3 and substituted in the VH of 4C10 are shown under the nucleotide sequence. Amino acid substitutions from the prototype sequence are outlined. The nucleotide at which a PCR substitution occurred is outlined. uses JH3; however, for the first residue the T normally present is replaced by a G, leading to a Trp to Gly replacement. With this replacement, the bulky side chain of Trp is removed, and extra flexibility of this region should result as a consequence of the glycine residue. The G addition is consistent with the action of terminal transferase. The sequence between the end of VH and the beginning of JH is GGCGAAGGTCACGCGTGG. This sequence does not directly match any of the published D se quences (13), but does contain a core element related to DSP2.4, DSP2.6, and DSP2.8 (Fig. 3). The nucleotides flank ing this core sequence are predominantly Gs, again consistent with the action of terminal transferase. Comparison of the VH sequence of 4C10 with the sequence of representatives of the 7 VH families defined by Brodeur and Riblet (16) shows that it is not as a member of any of these families. Instead it is a member of the infrequently expressed Vgam 3.8 family (8). Comparison of the 10 sequences from that gene family with the most similar sequence with the heavy chain of 4C10 reveals 6 positions at which amino acid substitutions have taken place compared to the prototype sequence (Fig. 2). Two substitutions occur in FR3, in which Ser 76 is replaced by Asn, and Gin 81 is replaced by Leu. The other 4 substitutions occur within CD2, in which Tyr 53 is replaced by Asn and Ala 60 by Thr (hydrophobic to polar), and at positions 61 and 62, where a pair of aspartic acids is replaced by a pair of glutamic acids. Transfection and Expression of Chimeric Anti-Idiotype. The PCR generated VL and VH cloned into drug marked expression vectors" were simultaneously transfected into nonproducing myeloma cell lines by electroporation and stably transfected cells selected. Transfectomas producing both chimeric heavy and light chains were identified, and one clone, TVE1, was amplified for further analysis. To characterize the chimeric protein, the transfectoma was labeled by growth in 15S-methionine, cytoplasmic extracts and secretion prepared, and the Ig precipitated using rabbit anti-human Fab and S. aureus protein A. The precipitates were analyzed by SDS-polyacrylamide gel electrophoresis both before and after reduction of the disulfide bonds (data not shown). The chimeric heavy and light chains were of the expected molecular weights but with the light chain showing slightly reduced mobility compared to other chimeric light chains. The chimeric protein was secreted as a fully assembled H2L2 molecule. Purification and Characterization of Chimeric Anti-Idiotype. Analysis of the purified TVE1 protein showed no significant difference in the size of the intact IgG molecules compared to the original mouse anti-idiotype 4C10, polyclonal murine IgG, or polyclonal human IgG (Fig. 4a). However, after reduction, the light chains of TVE1 migrated slightly more slowly than the other 3 light chains (Fig. 4b). Western blot analysis showed that the TVE1 chimeric antibody, unlike 4C10, reacted with anti-human IgG, but not with anti-murine IgG (Fig. 4, c and d). Like the original murine 4C10 antibody, the chimeric anti body showed specific anti-idiotype reactivity with the human L612 monoclonal (Fig. 4e). The specificity of the TVE1 chimeric antibody was also confirmed by ELISA. Fig. 5 shows the binding of TVE 1 and 4C10 to L612 (anti-gm3), to L72 (a human monoclonal anti body to ganglioside GD2) (3), and to human polyclonal IgM. TVE1 and 4C10 reacted only with L612 and did not react with L72 or human polyclonal IgM. The results were confirmed using a reversed ELISA in which the plates were coated with L612 and the binding of TVE1 and 4C10 tested alone or by competition (Fig. 6). Both TVE1 and 4C10 exhibited the ex pected concentration-dependent binding. In competitive bind 1683 ing assays, TVE1 and 4C10 displayed reciprocal inhibition with L612 at almost identical concentrations, suggesting equivalent anti-id affinity. The specificity of chimeric TVE1 was further examined by a cell-elisa inhibition assay using melanoma cell line UCLASO-M12, which expresses ganglioside GM3 on the cell membrane at high density (Fig. 7). ELISA plates were coated with melanoma cells and inhibition of binding of L612 by TVE1 and 4C10 antibodies tested. Both chimeric TVE1 and 4C10 antibodies inhibited the binding of L612 to the tumor cell membranes to a similar extent (Fig. 7). Taken together, these results show that the anti-id specificity of the chimeric mouse/ human antibody TVE1 is indistinguishable from that of the original mouse 4C10 monoclonal antibody and that the chi meric Ab should therefore bear the internal image of ganglioside GM3. DSP2.4 DSP2.6 CTATCTATGGTTACGAC CCTACTATGGTTACGAC 4C10 G G C G AA GGTCACG C G T G G DSP2.8 CCTAGTATGGTAACTAC Fig. 3. Comparison of the D region of 4C10 with other D regions. Residues in the D segments also found in 4C10 are shown by bold italics. Residues in 4C10 hypothesized to result from terminal transferase addition are underlined.

CHIMERIC ANTMDIOTYPE MIMICKING GANGL1OSIDE a. Non-reducing b. Reducing c. a-hu IgG d- a-mo IgG e. L612 + a-hu IgM 1234 1234 1234 1234 1234 Fig. 4. SDS-polyacrylamide gradient gel (4-20%) electrophoresis and Western blot analysis of purified protein. Proteins were stained with Coomassie brilliant blue R-2SO following separation under nonreducing (a) or reducing (b) conditions. Western blot analysis of unreduced proteins was performed using peroxidase conjugated anti-human IgG (c) and anti-murine IgG (d). To detect reactivity with L612 (e), the blotted membrane was incubated with 20 >ig/ml L612 in TPBS and binding of L612 detected with peroxidase conjugated anti-human IgM. The amount of antibody loaded in each lane was 2 jig for protein staining and 0.2 «igfor Western blot analysis. Lane I, polyclonal human IgG; Lane 2, polyclonal murine IgG; Lane 3, 4C10; Lane 4, TVE1. DISCUSSION Using PCR based expression vectors recently developed in the laboratory, we have now been successful in cloning and expressing a chimeric anti-idiotypic antibody that bears the internal image of GM3, a melanoma-associated ganglioside antigen. As has previously been observed for the numerous examples of chimeric antibodies produced to date, the combin ing specificity of the chimeric antibody is identical to that of the original murine hybridoma antibody (reviewed in Refs. 17 and 18). Both antibodies recognize the same epitope and on a weight basis compete in a reciprocal fashion (Figs. 5 and 6). The conservative Ser to Thr substitution at amino acid 27d therefore had no effect on the binding specificity. The chimeric antibody is assembled and secreted as a complete HI... mole cule. Its only unusual property is the apparent molecular weight of its «-lightchain on SDS-polyacrylamide gel electrophoresis. Light chains of the same apparent molecular weight frequently show different migration rates (19). The difference in migration rate cannot be due to N-linked carbohydrate because there are no carbohydrate addition sequences (Asn-X-Ser/Thr) within the light chain. Unlike the original murine hybridoma, the chimeric protein contains mostly human sequences as confirmed by Western blot analysis (Fig. 4). When injected into humans, the principle immune response should be directed against the variable region. It therefore has clinical potential as an idiotypic vaccine in cancer patients, and should induce an anti-gm3 specific im mune response that may be effective against tumors bearing this antigen. A question of fundamental interest is, how is this antiidiotypic antibody able to mimic ganglioside? Previous exam ples of idiotypic mimicry are more easily understood because protein structures mimic protein structures. A monoclonal antiidiotypic antibody directed against an antibody specific for the virus neutralizing epitope on the mammalian reovirus type 3 hemagglutinin has been shown to express an internal image of the receptor binding epitope of the reovirus type 3 (20). In this case, a stretch of homologous amino acids in CDR2 of VH 1684 could be combined with a second set of homologous amino acids in CDR2 of VL to create a region similar in primary sequence to the region on the hemagglutinin known to be near regions important for determining virulence. It seemed plausi ble that the combined VH-VL domain can function together to mimic the binding site of the reovirus hemagglutinin with molecular mimicry having a structural basis in shared primary sequences. In a second case in which 2 monoclonal antibodies were raised that bore an internal image of the VH al allotype of rabbits (21), mimicry appeared to be achieved by amino acids in the reverse order, and it was suggested that it is the confor mation of the amino acid side chains that is important. Again, it is an example of a protein mimicking a protein, so it is easy to visualize in structural terms. In our example, we must, however, explain how a protein structure can mimic a carbo hydrate structure. One of the difficulties in determining the important residues is that no cognate structures have been solved. a. TVE 1 b. 4C10 10" 10 10 \0 10" 10' 10' 10-' Concentration (ng/ml) Concentration (ng/ml) Fig. 5. Comparison of the reactivity of the murine and chimeric anti-idiotypic antibodies with L612. ELISA plates were coated with 80 nr/ml TVE1 (a) or 4C10 (b) in 0.1 M carbonate buffer, ph 9.6. After blocking the plates with 5% bovine serum albumin in TPBS, binding by the anti-gm3 antibody L612, the anti-gd2 monoclonal IgM L72, and polyclonal human IgM were determined over a range of concentrations.

CHIMERIC ANTl-lDIOTYPE MIMICKING GANGLIOS1DE 0.0 10 10" 10' 10' 10J Concentration (ng/ml) 0.8 o 0.6 at >0.4! 0.2 0.0 10" 101 102 103 10" Concentration of competitive antibody (ng/ml) Fig. 6. Direct binding and competitive inhibition assays of murine and chimeric anti-id binding to L612. ELISA plates were coated with L612 at a concentration of 80 / g/mlin 0.1 M carbonate buffer, ph 9.6. Direct binding by 4C10 and TVE1 added at varying concentrations was determined (left panel). For competitive binding (right panel), TVE1 was added at a concentration of 250 ng/ml and the concentration of 4C10 varied from 4 ng to 2 *<g/mlwith binding of TVE1 detected using peroxidase conjugated anti-human IgG (Q). Similarly, 4C10 was added at a concentration of 250 ng/ml and binding inhibited using TVE 1 at concentrations ranging from 4 ng to 2 /ig ml: 4C10 binding was detected by peroxidase conju gated ami murine IgG (â ). 1.0 E 0.8 o o> have been hypothesized to result from D region fusion (22, 23). In the present antibody, the D region is not an obvious conse quence of D region fusion (Fig. 3). The D sequence has 2 charged amino acids, Glu and His, a polar Trp, and 3 hydrophilic amino acids, 2 Gly and an Ala (Fig. 2). The D segments of some anti-idiotypes to anti-gat seem to have the "internal image" of the antigen (24). Another unusual aspect of CDR3 is that at the D-J junction the Trp encoded by J is replaced by Gly. The variable region of the antibody 4C10 remains a very intriguing example of a protein mimicking a nonprotein epitope. It is possible to theorize how this mimicry may be achieved, however, until we have solved structures to compare, such discussions of mimicry must by necessity be limited to speculation. Determining the structural basis for this mimicry is an important undertaking because, unlike proteins, ganglioside antigens cannot be made using genetic engineering tech niques. Therefore, there is no obvious way at this time to produce these important substances in large quantities. Peptide or protein mimics of gangliosides would be an important advance. ACKNOWLEDGMENTS The assistance of M. Josefina Coloma in the cloning and sequencing of the variable region is gratefully acknowledged. REFERENCES 1. Cahan, L. D., Irie, R. F., Singh, R., Cassidenti, A., and Paulson, J. C. Identification of a human neuroectodermal tumor antigen (OF-I-2) as gan I  '4 glioside GD2. Proc. Nati. Acad. Sci. USA, 79: 7629-7633, 1982. 2. Tai, T., Paulson, J. C.. Cahan, L. D., and Irie, R. F. Ganglioside GM2 as a human tumor antigen (OFA-I-1). Proc. Nati. Acad. Sci. USA, 80: 5392- E 5396, 1983. < 0.2 3. Irie, R. F., Sze, L. L., and Saxton, R. E. Human antibody to OFA-I, a tumor antigen, produced in vitro by Epstein-Barr virus-transformed human B- lymphoid cell lines. Proc. Nati. Acad. Sci. USA, 79: 5666-5670, 1982. 0.0 10 10" 10 10' 4. Irie, R. F., and Saxton, R. E. Human monoclonal antibodies: prospects for the therapy of human melanoma. In: K. Kano, S. Mori, T. Sugisaki, and M. Concentration (yg/ml) Torisu (eds.). Cellular, Molecular, and Genetic Aspects to Immunodiagnosis Fig. 7. Cell-ELISA inhibition assay of chimeric and murine anti-idiotypic and Immunotherapy, pp. 73-86. Tokyo: University of Tokyo Press, 1987. antibodies. Plates were coated with membranes prepared from melanoma cells 5. Irie, R. F., and Morton, D. L. Regression of cutaneous metastatic melanoma expressing GM3. L612 was added to each well at a concentration of 8 Mg/ml; by intralesional injection with human monoclonal antibody to ganglioside GD2. Proc. Nati. Acad. Sci. USA, 83: 8694-8698, 1986. 4C10 and TVE1 diluted from 16 to 0.125 >ig/ml were added as inhibitors and the binding of L612 determined using peroxidase conjugated anti-human IgM. 6. Irie, R. F., Matsuki, T., and Morton. D. L. Human monoclonal antibody to ganglioside GM2 for melanoma treatment. Lancet, /: 786-787, 1989. 7. Yamamoto, S., Yamamoto, T., Saxton, R. E., Hoon, D. S. B., and Irie, R. F. Anti-idiotype monoclonal antibody carrying the internal image of ganglio At the primary structure level, the light chain has 3 amino side GM3. J. Nati. Cancer Inst.. 82: 1757-60, 1990. acid substitutions from a known germ line gene (15). Of these, 8. Winter, E., Radbruck, A., and Krawinkel, U. Members of novel VH gene families are found in VDJ regions of polyclonally activated B-lymphocytes. the one in CDR2 is most likely to be exposed on the surface EMBO J., 4: 2861-2867, 1985. and thus may be a candidate for a residue important in deter 9. Chirgwin, J. M., Przybyla, A. E., McDonald. R. J., and Rutter, W. J. Isolation mining the idiotypic mimicry. However, VH is also probably of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry, 18: 5294-5299, 1979. important in forming the epitope mimicking ganglioside. 10. Hartman, S. C., and Mulligan, R. C. Two dominant-acting selectable markers Within VH, a likely contributor is CDR2, in which 4 amino for gene transfer studies in mammalian cells. Proc. Nati. Acad. Sci. USA, «5:8047-8051. 1988. acids, 3 in a row, are altered compared to the germline sequence. 11. Laemmli, U. k. Cleavage of structural proteins during the assembly of the CDR2 is frequently exposed on the surface of the variable head of bacteriophage T4. Nature (Lond.), 227: 680-685, 1970. region. It is therefore possible that these amino acids, like the 12. Towbin, H., Staehelin, T., and Gordon, J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some appli reovirus example, in conjunction with CDR2 of the light chain, cations. Proc. Nati. Acad. Sci. USA, 76:4350-4354, 1979. form the epitope corresponding to ganglioside. Exactly how 13. Rabat, E. A., Wu, T. T., Reid-Miller, M., Perry, H. M., and Gottesman, K. this mimicry would be effected is unclear. The twin glutamic Sequences of proteins of immunological interest. Washington, DC: United States Department of Health and Human Services, 1987. acids may serve as a mimic for sialic acid of the ganglioside, 14. Potter, M., Newell, J. B., Rudikoff, S., and Haber, E. Classification of mouse but how the other amino acids might mimic the other sugars V>.groups based on the partial amino acid sequence to the first invariant tryptophan: impact of 14 new sequences from IgG myeloma proteins. Mol. remains to be determined. Immunol., 19: 1619-1630, 1982. A second aspect of VH that must be considered for a role in 15. Heinrich, G., Traunecker, A., and Tonegawa, S. Somatic mutation creates mimicking ganglioside is the D region, which has an unusual diversity in the major group of mouse immunoglobulin Klight chains. J. Exp. Med., 759:417-435. 1984. sequence. The presence of unique D regions in Ab2s has been 16. Brodeur, P. H., and Riblet, R. The immunoglobulin heavy chain variable observed previously (22) and in some cases these D regions region (Igh-V) locus in the mouse. I. One hundred Igh-V genes comprise 1685

CHIMERIC ANTI-IDIOTYPE MIMICKING GANGLIOSIDE seven families of homologous genes. Eur. J. Immun»!..14: 922-930, 1984. anti-idiotype sequences in reverse orientation. Adv. Exp. Med. Hi»!..251: 17. Morrison, S. L., and Oi, V. T. Genetically engineered antibody molecules. 187-190, 1989. Adv. Immunol., 44: 65-92, 1989. 22. Meek, K., Hasemann, C., Pollok, B., Alkan, S. S., Brait, M., Slaoui, M., 18. Morrison, S. L. In vitro antibodies: strategies for production and application. Urbain, J., and Capra, J. D. Structural characterization of antiidiotypic Ann. Rev., in press, 1992. antibodies: evidence that Ab2s are derived from the germline differently than 19. Sharon, J., Kabat, E. A., and Morrison, S. L. Studies on mouse hybridomas Abls. J. Exp. Med., 169: 519-533, 1989. secreting IgM or IgA antibodies to a(lâ >6)linked dextran. Mol. Immunol., 23. Perfetti, V., Borden, P., Tao, M. Morrison, S. L., and Kabat, E. A. Specificity /S: 831-846, 1981. and variable region cdn A sequence of an isogeneic monoclonal anti-idiotype 20. Brack, C, Co, M. S., Slaoui, M., Gaulton, G. N., Smith, T., Fields, B. N., to an anti-a(l-6) dextran. Mol. Immunol., 28: 505-515, 1991. Mullins, J. I., and Greene, M. I. Nucleic acid sequence of an internal image- 24. Oilier, P., Rocca-Serra, J., Somme, G., Theze, J., and Fougerau, M. The bearing monoclonal anti-idiotype and its comparison to the sequence of the idiotypic network and the internal image: possible regulation of a germ-line external antigen. Proc. Nati. Acad. Sci. USA, 83: 6578-6582, 1986. network by paucigene encoded Ab2 (anti-idiotypic) antibodies in the GAT 21. Metzger, D. W., Naeve, C. W., and van Cleave, V. H. Epitope mimicry by system. EMBO J., 4: 3681-3688, 1985. 1686