HK1070080B - Ganglioside-associated recombinant antibodies and the use thereof in the preparation of medicament for the diagnosis and treatment of tumors - Google Patents

Description

Ganglioside related recombinant antibody and its application in preparing medicine for diagnosing and treating tumor

Technical Field

The present invention relates to the field of biotechnology, in particular to novel recombinant antibodies obtained by genetic engineering, in particular to chimeric and humanized antibodies obtained from the murine monoclonal antibody P3(MAb P3) and the anti-idiotype murine monoclonal antibody 1E10(MAbai 1E 10).

More particularly, the invention relates to antibodies that bind to gangliosides containing N-glycolylated (N-glycosylated) sialic acid, but which bind neither to the acetylated form of the ganglioside nor to the neutral glycolipid. Gangliosides containing N-glycolylated sialic acid are antigens that are widely expressed in breast cancer and melanoma. On the other hand, the antitumor effect of MAbai 1E10 has also been demonstrated in experimental models.

The invention also relates to pharmaceutical compositions comprising the aforementioned recombinant antibodies useful in the diagnosis and treatment of cancer, particularly breast cancer and melanoma.

Prior Art

Gangliosides are sialic acid-containing Glycosphingolipids and are present in the cytoplasmic membranes of vertebrate cells (Stults et al (1989): Glycosphingolipids: structural, biological source and properties (Glycosphingolipids: structures, biological sources and properties), Methods Enzymology, 179: 167-. Some reports of this molecule as tumor-associated or tumor-marker antigens have been found in the literature (Hakomori et al (1991): possible functions of tumor-associated carbohydrate antigens (Possibleinteractions of tumor associated carbohydrate antigens), Current. Optin. Immunol., 3: 646-653), for which the use of anti-ganglioside antibodies has been considered useful in the diagnosis and treatment of cancer (Hougton et al (1985): murine monoclonal antibody IgG3antibody detecting GD3 ganglioside; phase I test in patients with malignant melanoma (Mouse monoclonal antibody IgG3antibody detecting detection of GD3 ganglioside: to I tertiary antibodies with tumor antigens; PNAS USA, 82: 1242 1246; use of Zkomori et al as a tumor-targeting antigen for histochemical screening of tumors I. for histochemical antigens), int.j.cancer, 73: 42-49).

Sialic acids more frequently expressed in animals are N-acetylation (NeuAc) and N-glycolylation (NeuGc) (Corfield et al (1982): sialic acid production (Occurence of clinical acids), cell. biol. monogr., 10: 5-50). In general, NeuGc is not expressed in normal human and chicken tissues, but is widely distributed among other vertebrates (Leeden and Yu, (1976): Sialic acid Chemistry and analysis (Chemistry and nomenclature of clinical acids), Biological Role of clinical acids, Rosemamberg A and Shengtrun CL (Eds); Plenum Press, New York, 1-48; Kawai et al (1991): Quantitative determination of N-glycolylneuraminic acid as tumor-associated Sialic acid expressed in human Cancer tissues and avian lymphoma cell lines using a combination of gas chromatography-mass spectrometry (Quantitative determination of N-glycolylneuraminic acid expression in human Cancer tissues and avian lymphoma cell lines) and Cancer cells of the same type using a protein 12412451;. Cancer cells 1242. sample 1246). However, it has been reported that anti-NeuGc antibodies recognize certain human tumors and tumor cell lines (Higashi et al (1988): Detection of gangliosides as N-glycolylneuraminic acid-specific tumor-associated Hanganutziu-Deicher antigens in human retinoblastoma cells (Detection of gangliosides as N-glycolylneuraminic acid tumor-associated Hanganutziu-Deicher antigens), JJ.Cancer Res.79: 956; Fukuiti et al (1989): Detection of glycoproteins as tumor-associated Hanganutziu-Deicher antigens in NuGC4 (Detection of glycoproteins as tumor-associated Hanganutzi-Deicher antigens, Biochemical 115160. Biochemical). Increased levels of GM3(NeuGc) Gangliosides have been found in human breast Cancer (Marquina et al (1996): Gangliosides expressed in human breast Cancer (Cancer Research, 1996; 56: 5165-.

Monoclonal antibody (Mab) P3 (European patent EP 0657471B 1), a murine monoclonal antibody of the IgM isotype, obtained when spleen cells of BALB/c mice inoculated with liposomes containing GM3(NeuGc) and tetanus toxoid and a murine myeloma cell line P3-X63-Ag8.653 were fused, was produced using the cell line deposited under number ECACC 94113026. This Mab P3 reacted strongly with gangliosides containing N-glycolylated sialic acid but neither with the acetylated form of ganglioside nor with neutral glycolipids. Immunocytochemistry and immunohistochemistry studies using cell lines and tissues from benign and malignant tumors demonstrated that Mab P3 recognizes breast cancer (V a zzoz et al (1995): production of N-glycolylneuraminic acid-directed murine monoclonal antibodies, which also recognize glycolipids sulfate (Generation of a polysaccharide monoclonal antibody for N-glycosylated amino acid-binding polypeptides), Hybridoma, 14: 551-556) and melanoma.

Mab P3 induced an anti-idiotypic immune response in BALB/c mice (Syngeneic model) even without adjuvant and carrier proteins (V a zquez et al (1998): Syngeneic anti-idiotypic monoclonal antibodies against monoclonal antibodies containing gangliosides against NeuGc (Syngeneic anti-idiotypic monoclonal antibodies to an anti-NeuGc-containing ganglioside monoclonal antibodies); Hybridoma, 17: 527-534). Immunochemical analysis suggested negatively charged groups: sialic acid (for gangliosides) and SO3- (for sulfatides), a role in the recognition properties of this antibody (Moreno et al (1998): characterization of epitopes recognized with specific antibodies against gangliosides containing N-glycolylneuraminic acid (delication of epigenetic by an antibody specific for N-glycosylated amino acid-binding polypeptides), Glycobiology, 8: 695-valent 705).

Anti-idiotypic Mab1E10 (Mabai 1E10) of the IgG1 subtype was obtained from BALB/c mice immunized with Mab P3 conjugated to KLH (U.S. Pat. No. 6,063,379, cell line deposited under accession number ECACC 97112901). Mabai 1E10 specifically recognized Mab P3 and did not bind to other anti-ganglioside IgM antibodies. Also, Mabai 1E10 prevented specific binding of Mab P3 to GM3(NeuGc) and to the cell line MDA-MB-435 derived from ductal breast cancer (positive for binding of MabP 3). Upon immunization of mice in either a Syngeneic or allogeneic (allogenic) model, Mabai 1E10 induced a strong immune response to Ab 3antibody, which Ab 3antibody did not show the same specificity as Mab P3 even when the unique epitope it carries was similar to that carried by Ab1 antibody (V a zquez et al (1998): Syngeneic anti-idiotypic monoclonal antibodies against ganglioside monoclonal antibodies containing anti-NeuGc (Syngenic anti-idiotypic monoclonal antibodies to anti-NeuGc-neu monoclonal antibodies), Hybridoma, 17: 527-534). Mabai 1E10 induced a strong antitumor effect in syngeneic as well as allogeneic mice. In the case of BALB/c mice vaccinated, the growth of the breast cancer cell line F3II was greatly reduced with repeated doses of Mabai 1E10 conjugated to KLH in Freund's adjuvant. The number of spontaneous lung metastases was also reduced after vaccination. Intravenous administration of Mabai 1E10 to vaccinated C57BL/6 mice resulted in a dramatic reduction in the number of lung metastases 10-14 days after intravenous vaccination with melanoma cells B16 compared to mice treated with irrelevant IgG. This result suggests the initiation of more than one anti-tumor mechanism of action. (V.zquez et al (2000) anti-tumor properties of anti-idiotypic monoclonal antibodies associated with N-glycolyl-containing gangliosides (anti properties of anti-inflammatory monoclonal antibodies in relation to N-glycyl-stabilizing lipids), Oncol. Rep.7: 751-756, 2000).

Hybridoma technology has been developed over the course of 15 years (Koehler and Milstein (1975): continuous culture of fused cells secreting predetermined specific antibodies (Nature, 256: 495-. This is mainly due to its short half-life in blood and the failure of murine effector functions to the human immune system and the human anti-murine antibody immune response (HAMA response).

In addition, genetic engineering techniques have allowed the development of potential uses for MAbs since the manipulation of immunoglobulin genes has made it possible to obtain modified antibodies with reduced antigenicity and to improve the effector functions thereof for the treatment or diagnosis of specific pathologies. A fundamental task of methods for reducing the immunogenicity of immunoglobulins is to reduce the differences between murine and human immunoglobulins without altering the specificity of antigen recognition (Morrison and Oi (1989): Genetically engineered antibody molecules, Adv Immunol., 44: 65-92).

Several approaches have recently been developed to humanize mouse or rat antibodies that reduce the allogeneic immune response to foreign proteins when injected into humans. One of the earliest methods to reduce antigenicity was a chimeric antibody in which the variable region of a murine protein was inserted into the constant region of a human molecule which exhibits the same specificity but reduced immunogenicity as compared to its murine counterpart, (Morrison et al (1984): chimeric human antibody molecule: murine antigen binding domains with human constant regions (PNAS USA, 81: 6851-. Immune responses to rodent variable regions are often observed even when the chimeric antibody has the same specificity as the murine counterpart.

In an attempt to further reduce the immunogenicity of chimeric antibodies, only the CDRs from rodent monoclonal antibodies were grafted into human framework regions, which hybrid variable regions were expressed together with human constant regions (Jones et al (1986): replacement of the complementarity determining regions in human antibodies with murine complementarity determining regions (lacing the complementary-determining regions in a human antibody with human breast mouse), Nature 321: 522-524; Verhoeyen et al (1988): reconstituted human antibodies: grafted anti-lysozyme activity (Science 239, 1534-1536). However, this method has several disadvantages: the resulting antibodies often have reduced affinity and a large number of framework residues must be back mutated to the corresponding murine residues to restore binding capacity (Rietchmann et al (1988): reconstituted human antibodies for therapy (Reshaping human antibodies for therapy), Nature, 332: 323-. Furthermore, durable immunogenicity is often observed in CDR grafted antibodies.

Mateo and co-workers (U.S. Pat. No. US 5712120) describe a method for reducing the immunogenicity of murine antibodies. According to this method, the modification is limited to the variable regions, in particular to the murine FRs of the chimeric antibody. Furthermore, substitutions are made only in those regions where FRs have amphipathic sequences and are therefore potential epitopes for recognition by T cells. This approach involves judicious substitution of several amino acid residues by the most homologous human counterpart, with the substituted amino acid residues being located in the potential immunogenic epitope, the amino acids that contribute primarily to the canonical structure and the residues in the immediate neighborhood of the CDRs or in the Vernier zone having to be retained.

The resulting antibody retains its antigen binding specificity and is less immunogenic than its murine or chimeric precursor (Mateo et al (2000): Removal of T cell epitopes from a genetically engineered antibody Production of modified immunogenically reduced immunoglobulins (produced by the Production of modified immunoglobulins with reduced immunogenicity), hybrid 19: 463-71), a property that enhances its therapeutic utility. With this new approach, only a few mutations need to be completed, but of course, fewer genetic manipulations are required.

Detailed Description

The present invention relates to recombinant antibodies obtained by genetic engineering techniques. In particular, the present invention relates to a chimeric antibody obtained from the murine monoclonal antibody P3, which was produced with the hybridoma cell line deposited under accession number ECACC 94113026. MAB P3 recognizes an antigen expressed in breast cancer cells and melanoma. MAB P3 is characterized by the following sequences of the heavy and light chain highly variable regions (CDRs):

heavy chain

CDR1：RYSVH

CDR2：MIWGGGSTDYNSALKS

CDR3：SGVREGRAQAWFAY

CADENA LIGERA

CDR1：KASQDVSTAVA

CDR2：SASYRYT

CDR3：QQHYSTPWT

Preferred FRs sequences for the heavy and light chains are as follows:

heavy chain

FR1：QVQLKESGPGLVAPSQSLSITCTVSGFSLS

FR2：WVRQPPGKGLEWLG

FR3：RLSISKDNSKSQVFLKMNSLQTDDTAMYYCAR

FR4：WGQGTLV

Light chain

FR1：DIVMTQSHKFMSTSVGDRVSITC

FR2：WYQQKPGQSPKLLIY

FR3：GVPDRFTGSGSGTDFTFTISSVQAEDLAVYYC

FR4：FGGGTKL

In a preferred embodiment, the chimeric antibody of the invention comprises the constant region of human IgG1 heavy chain and the constant region of human Ck light chain. In another aspect, the present invention relates to a humanized antibody derived from Mab P3 produced by the hybridoma cell line deposited under accession number ECACC94113026, characterized by containing the constant region of the human IgG1 heavy chain and the constant region of the human light chain Ck and the FRs region of its light chain containing any of the following point mutations:

light chain

Position 8: replacement of His with Pro

Position 9: replacement of Lys with Ser

Position 10: replacement of Phe with Ser

Position 11: replacement of Met by Leu

Position 13: replacement of Thr with Ala

In another aspect, the invention relates to a chimeric antibody derived from the murine monoclonal antibody 1E10 produced with the hybridoma cell line deposited under accession number ECACC 97112901 and which is an anti-idiotypic antibody recognizing Mab P3. MAbai 1E10 was characterized by the following sequences of the heavy and light chain highly variable regions (CDRs):

heavy chain

CDR1：SYDIN

CDR2：WIFPGDGSTKYNEKFKG

CDR3：EDYYDNSYYFDY

Light chain

CDR1：RASQDISNYLN

CDR2：YTSRLHSG

CDR3：QQGNTLPWT

Preferred FRs sequences for the heavy and light chains are as follows:

heavy chain

FR1：QVQLQQSGAELVKPGASVKLSCKASGYTFT

FR2：WVRQRPEQGLEWIG

FR3：KATLTTDKSSSTAYMQLSRLTSEDSAVYFCAR

FR4：WGQGTTLTV

Light chain

FR1：DIQMTQTTSSLSASLGDRVTISC

FR2：WYQQKPDGTVKLLIY

FR3：VPSRFSGSGSGTDYSLTISNLEQEDIATYFC

FR4：FGGGTKLESK

In a preferred embodiment, the chimeric antibody of the invention comprises the constant region of the heavy chain of human IgG1 and the constant region of the light chain of human Ck. In another aspect, the invention relates to a humanized antibody derived from Mab1E10 produced by the hybridoma cell line deposited under accession number ECACC 97112901, characterized in that it contains the constant region of the human IgG1 heavy chain and the constant region of the human Ck light chain and the heavy and light chain FRs regions contain any of the following point mutations:

light chain

Position 7: replacement of Thr with Ser

Position 8: replacement of Thr with Pro

Position 15: replacement of Leu by Val

Heavy chain

Position 5: replacement of Gln with Val

Position 40: replacement of Arg by Ala

Position 42: replacement of Glu with Gly

Position 87 (numbering 83 according to Kabat): replacement of Thr with Arg

In another aspect, the invention relates to cell lines expressing the chimeric and humanized antibodies; in addition the invention relates to pharmaceutical compositions containing said antibodies.

Preferably, it relates to pharmaceutical compositions containing said antibodies and suitable excipients for the treatment of tumors of the mammary gland, lung, digestive system, urogenital system, melanoma, sarcoma and neuroectodermal layer, as well as their metastases and relapses.

In another expression of the invention, the pharmaceutical compositions containing said antibodies can be used for in vivo localization and diagnosis of breast, lung, digestive, urogenital, melanoma, sarcoma and neuroectodermal tumors, as well as their metastases and relapses.

The synthesis and gene amplification of cDNA of Mab P3 and Mabai 1E10 variable regions was performed by PCR (polymerase chain reaction).

From about 10⁶The P3 (murine IgM MAb, recognizing GM 3N-glycolylated ganglioside) or 1E10 (anti-idiotypic antibody against P3) hybridoma cells. RNA was extracted using TRIZOL reagent (GIBCO BRL, Grand Island, NY) according to the manufacturer's instructions.

Mu.g of RNA, 25 picomoles of Vh (complementary to the constant region of murine IgM of VHP3 and having the constant region of murine IgG1 of VH1E 10) or Vk (complementary to the constant region of murine kappa of both antibodies), 2.5mM of various dNTPs, Tris-HCl 50mM at pH 7.5, 75mM KCl, 10mM DTT, 8mM MgCl2 were mixed and 15 units of RNase inhibitor were added per 50. mu.l of the reaction mixture to carry out the cDNA synthesis reaction. Heated at 70 ℃ for 10 minutes and slowly cooled to 37 ℃. Then, 100 units of MLV reverse transcriptase was added and incubation was continued at 42 ℃ for one hour.

cDNAs for the variable regions VK and VH were amplified by PCR. Briefly, 5. mu.l of either VH or VK cDNA was mixed with 25 pmoles of specific primers, 2.5mM of each dNTP, 5. mu.l of 10-fold Taq DNA polymerase buffer, and 1 unit of the enzyme composition. The samples were subjected to 25 thermal cycles of 94 ℃ for 30 seconds, 50 ℃ for 30 seconds, 72 ℃ for 1 minute, and finally incubated at 72 ℃ for 5 minutes.

Cloning and sequencing of amplified cDNA

PCR products of VH and VK (P3 and 1E10, respectively) were cloned into TA vectors (TA cloning kit. Promega, USA). The resulting clone was dideoxy sequenced with T7DNA polymerase (T7 sequencing kit Pharmacia, Sweden).

Construction of chimeric genes

The VH VK gene was excised from the TA vector by enzymatic digestion and cloned into respective expression vectors (Coloma et al (1992): Novel vectors for expressing antibody molecules using variable regions produced by the polymerase chain reaction (Novel vectors for the expression of antibody molecules using variable regions produced by the polymerase chain reaction), J.Immunol.meth.152: 89-104).

The VH gene was excised from the TA vector by enzymatic digestion with EcoRV and Nhel and cloned into an expression vector containing the human IgG1 variable region and the histidinol resistance gene (PAE 4604). The resulting constructs were P3VH-PAH4604 and 1E10VH-PAH 4604. The VK gene was excised from the TA vector by enzymatic digestion with EcoRV and SaII and cloned into an expression vector (PAG 4622). The vector includes a mycophenolic acid resistance gene and a human kappa constant region. The resulting constructs were P3VK-PAG4622 and 1E10VK-PAG 4622.

Expression of chimeric antibodies obtained from Mab P3 and Mabid 1E10

NS-O cells were electroporated with 10. mu. g P3VK-PAG4622 or 1E10VK-PAG4622, and clones expressing human kappa light chain were transfected with 10. mu. g P3VH-PAH4604 or 1E10VH-PAH 4604.

DNA digestion with PvulaseLinearized, precipitated with ethanol and dissolved in 50 μ l pbs. Centrifuging to obtain a product of about 10⁷The cells of (4) were resuspended in 0.5ml PBS together with the digested DNA in the electroporation cuvette. After 10 minutes on ice, the cells were pulsed with a current of 200 Volts 960. mu.F and placed on ice for another 10 minutes. Cells were plated in 96-well plates with D' MEM F12 supplemented with 10% fetal bovine serum. Two or four days later, selection medium (D' MEM F12 containing mycophenolic acid 0, 45. mu.g/ml or histidinol 10mM, respectively) was added. After 14 days, the transfected clones were visible to the naked eye.

The presence of human antibodies in the medium in the wells containing the transfected clones was determined by ELISA. Microtiter plate wells were covered with goat anti-human kappa light chain (for human kappa chain producing clones) or anti-human IgG (gamma chain specific) (for full antibody producing clones) antibodies. After washing with PBST (phosphate buffered saline containing 0.05% Tween 20), diluted medium containing the transfection product was added to each microtiter plate well for one hour at 37 ℃. The wells were washed with PBS-T and either goat anti-human kappa light chain conjugated to horseradish peroxidase (peroxidase of paramagnetic prepared) or goat anti-human IgG conjugated to alkaline phosphatase (gamma chain specific) was added followed by incubation at 37 ℃ for one hour. After the wells were washed with PBS-T, substrate buffers containing o-phenylenediamine or p-nitrophenylphosphate (p-nitrophenylphosphate) were added, respectively. Half an hour later, the absorption at 492 or 405nm, respectively, was measured.

Humanized antibodies P3hu and 1E10hu were constructed by humanizing the T cell epitope. Prediction of T cell epitopes.

The sequences of the P3 and 1E10 variable regions were analyzed using the AMPHI algorithm (Margalit et al (1987): Predictionof immunogenic helper T cell antigenic sites from the primary sequence), J.Immunol., 138: 2213-2229). Amphipathic (amphipathic) helical fragments with 7 or 11 amino acid residues associated with T immunogenicity were detected. The program SOHHA also predicted hydrophobic helical fragments. (Elliot et al (1987) An hypothesis on the binding of An amphetatic, alpha-helical sequence in li to the desotope of class IIanti, J.Immunol., 138: 2949-. Both algorithms predict which fragments of sequences in the variable regions of antibodies P3 and 1E10 are capable of being presented to helper T cells in the context of MHC class II molecules.

Homology analysis with human immunoglobulin

The amino acid sequence of the murine variable region was compared to immunoglobulin sequences contained in the GeneBank and EMBL databases (available from the Internet). For each antibody, the most homologous human variable region sequences were determined. The software used for sequence homology detection was PC-TWOHIBIO PROSIS 06-00 (Hitachi).

Assay for reduced immunogenicity

The aim of the method is to reduce immunogenicity with minimal changes, to destroy or humanize potentially immunogenic T epitopes. The method involves deliberate substitution of several amino acid residues located on the amphipathic helical segment. The amino acids that play a major role in the canonical structure must be retained, as well as residues in the immediate vicinity of the CDRs or Vernier zone.

According to the present method, the murine variable region sequence is compared to the most homologous human sequence and the amino acid residues that differ between the murine MAb and the most homologous human sequence at each position are determined, taking into account only the residues in the FRs (Kabat (1991), Sequences of proteins of immunological interest, fifth edition, national institute of Health), replacing the previously determined residues with those present in the most homologous human sequence. The substitution process is accomplished by site-directed mutagenesis techniques.

Residues involved in the three-dimensional structure of the binding site are not mutated; it may affect antigen recognition. Additional information about the effect of substitutions in tertiary structure can be obtained from molecular modeling of the antigen binding site.

The presence of proline residues in the amphipathic helical segment and the fact that specific murine residues do not occur at the same position in the human most homologous sequence but frequently occur in other human immunoglobulins must be kept in mind. For this reason, the murine collection of amino acids that are substituted into the framework is not unique. It is possible to use different numbers of substitutions to obtain different forms of modified antibody. The mutation was performed by overlapping PCR.

Humanized antibodies P3hu and 1E10hu were cloned and expressed.

Genetic constructs corresponding to P3hu and 1E10hu were cloned into expression vectors as described for the chimeric antibodies. The resulting constructs were P3Vkhu-PAG4622 or 1E10Vkhu-PAG4622 and P3VHhu-PAH4604 and 1E10VHhu-PAH 4604. They were transfected into NS-0 cells following the protocol described previously for chimeric antibodies.

And (5) purifying the recombinant antibody.

The recombinant antibody was purified by affinity chromatography using protein A (Pharmacia, Ussala, Sweden).

And (4) biological activity.

The biological activity of the recombinant antibodies was examined by specific binding to the antigen as determined by ELISA.

For recombinant MAb P3, microtiter plates were coated with GM3(NeuGc) ganglioside in methanol. After drying for one hour, non-specific binding was blocked with 1% Bovine Serum Albumin (BSA) in Tris-HCl buffer and incubated for one hour at 37 ℃. The wells were washed with PBS and incubated with purified recombinant Mab P3 for one hour at 37 ℃. Wells were washed with tris-HCl and goat anti-human antibody conjugated to alkaline phosphatase was added followed by one hour incubation at 37 ℃. Finally, the wells are washed and substrate buffer containing p-nitrophenyl phosphate is added. The absorption at 405 or 492nm, respectively, is measured after half an hour.

For recombinant Mabai 1E10, the other ELISA experimental conditions were similar except that wells were covered with Mab P3 and washed with PBS-0.05% Tween 20.

Examples

All enzymes, as well as reagents and materials used in the following examples were obtained from commercial sources, unless otherwise indicated.

Example 1 acquisition of chimeric MAb P3.

As mentioned previously, the synthesis of cDNA was obtained by a reaction using reverse transcriptase, starting with RNA obtained from the hybridoma producing Mab P3. The sequences of the specific primers used in this reaction are shown below:

for VH:

5′AGGTCTAGAA(CT)CTCCACACACAGG(AG)(AG)CCAGTGGATAGAC 3′

for VK:

5′GCGTCTAGAACTGGATGGTGGGAAGATGG 3′

the cDNA VHP3 and cDNAVKP3 were amplified by PCR using Taq polymerase and specific primers. The restriction sites included in the primers were ECORV/NHEI for VH and ECORV/SALI for VK. The sequences of the primers used were as follows:

for VH:

primer 1 (signal peptide):

5′GGGGATATCCACCATGG(AG)ATG(CG)AGCTG(TG)GT(CA)AT(CG)CTCTT 3′

primer 2(CH 1):

5′GGGGCTAGCTGCAGAGACAGTGACCAGAGT 3′

for VK:

primer 1 (signal peptide):

5′GGGGATATCCACCATGGAG(TA)CACA(GT)(TA)CTCAGGTCTTT(GA)T 3′

primer 2 (Ck):

5′AGCGTCGACTTACGTTT(TG)ATTTCCA(GA)CTT(GT)GTCCC 3′

the PCR product was cloned into a TA vector (TA cloning kit, Invitrogen). Twelve independent clones were sequenced by dideoxy using T7DNA Pol (Pharmacia). The most homologous sequence sets for VHP3 and VKP3 were determined using homology search analysis. According to the Kabat classification, VHP3 and VKP3 sequences (fig. 1 and fig. 2) have high homology to groups IB and V, respectively.

After digestion with the restriction enzymes ECORV and NHEI for VHP3 and ECORV and SALI for VKP3, they were cloned into expression vectors previously digested with the same enzymes, PAH4604 and PAG4622 for VH and VK, respectively. The expression vector is provided by SherieMorrison (UCLA, California, USA) and is suitable for expression of immunoglobulins in mammalian cells. The vector PAH4604 already contains the constant regions of human IgG1 and human PAG4622 (Coloma et al (1992): a Novel vector for expressing antibody molecules using variable regions produced by the polymerase chain reaction (Novel vectors for the expression of antibody molecules using variable regions produced by the polymerase chain reaction), J.Immunol.meth.152: 89-104). The resulting constructs were P3VH-PAH4604 and P3VK-PAG 4622.

NS-0 cells were transfected with 10. mu. g P3VK-PAG4622 and clones expressing the light chain were transfected with 10. mu. g P3VH-PAH4604, in both cases, before transfection, the DNA was linearized with PvuI, precipitated with ethanol and dissolved in 50. mu.l PBS.

Centrifuging to obtain a product of about 10⁷The cells of (4) were resuspended in 0.5ml PBS together with the digested DNA in the electroporation cuvette. After 10 minutes on ice, the cells were pulsed with a current of 200 Volts 960. mu.F and placed on ice for another 10 minutes. Cells were plated in 96-well plates with D' MEM F12 supplemented with 10% fetal bovine serum. Two or four days later, selection medium (D' MEM F12 containing mycophenolic acid 0, 45. mu.g/ml or histidinol 10mM, respectively) was added. After 14 days, the transfected clones were visible to the naked eye.

The presence of human antibodies in the medium in the wells containing the transfected clones was determined by ELISA. Microtiter plate wells were covered with goat anti-human kappa light chain (for clones producing human kappa chains) or anti-human IgG (gamma chain specific) (for clones producing complete antibodies) antibodies. After washing with PBST (phosphate buffered saline containing 0.05% Tween 20), diluted medium containing the transfection product was added to each microtiter plate well for one hour at 37 ℃. The wells were washed with PBS-T and either horseradish peroxidase conjugated goat anti-human kappa light chain or alkaline phosphatase conjugated goat anti-human IgG (gamma chain specific) was added followed by one hour incubation at room temperature. After the wells were washed with PBS-T, substrate buffers containing o-phenylenediamine or p-nitrophenyl phosphate were added, respectively. Half an hour later, the absorption at 492 or 405nm, respectively, was measured.

Example 2. acquisition of different forms of humanized antibody P3.

The murine VHP3 and VKP3 sequences (fig. 1 and fig. 2) were compared to human sequences. The most homologous human sequences are shown in FIGS. 3 and 4. Amphipathic helical fragments or potential T cell epitopes were detected in the murine P3 variable region sequence and judicious strategies for amino acid substitutions were made to disrupt or humanize potential T cell epitopes in the murine sequence according to the method used.

Analysis of VHP3 resulted (fig. 3) in 2 amphipathic fragments, the first fragment comprising certain residues of CDR1, FR2 and CDR2, and the second fragment comprising the ends of FR3 and CDR 3. Major differences from the most homologous human and murine sequences were found in the CDRs or residues involved in the three-dimensional structure of the binding site. For this reason, it was decided not to replace any of the amino acids of murine VHP 3.

Analysis of VKP3 also yielded 2 amphipathic fragments (FIG. 4), the first fragment comprising FR1 and the second fragment comprising certain residues of CDR2 and FR 3. The residues at positions 8, 9, 10, 11 and 13 were determined to be replaced with residues at the same positions in the most homologous human sequence. The amino acids His, Lys, Phe, Met and Thr were replaced with Pro, Ser, Leu and Ala, respectively. The substitution was performed by overlap PCR using primers 1 and 2 and 3 and 4 (Kammann et al (1989) Rapid insertion mutagenesis (PCR) using DNA Polymerase Chain Reaction (PCR)), Nucleic Acids Res., 17: 5404) using the following sequences:

primer 1:

5′ATGACCCAGTCTCCTTCTTCTCTTTCCGCGTCAGTAGGAGAC 3′

primer 2:

5′AGCGTCGACTTACGTTT(TG)ATTTCCA(GA)CTT(GT)GTCCC 3′

primer 3:

5′GTCTCCTACTGACGCGGAAAGAGAAGAAGGAGACTGGGTCAT 3′

primer 4:

5′GGGGATATCCACCATGGAG(TA)CACA(GT)(TA)CTCAGGTCTTT(GA)T 3′

point mutations were examined by sequencing. The resulting construct was P3Vkhu and cloned into a PAG4622 expression vector. The resulting construct was P3Vkhu-PAG 4622. To express humanized antibody P3, NS-0 cells were transfected with P3VH-PAH4604 and P3Vkhu-PAG 4622.

The P3hu antibody was transfected according to the electroporation and detection methods described previously for chimeric antibodies.

Example 3: biological activity of chimeric MAb P3.

The biological activity of chimeric MAb P3 was tested by ELISA assay for specific binding to antigen.

For recombinant MAb P3, microtiter plates were coated with GM3(NeuGc) ganglioside in methanol. After drying at 37 ℃ for one hour, non-specific binding was blocked with 1% Bovine Serum Albumin (BSA) in Tris-HCl buffer and incubated at 37 ℃ for one hour. The wells were washed with PBS and incubated with purified recombinant Mab P3 for one hour at 37 ℃. Wells were washed with tris-HCl and goat anti-human antibody conjugated to alkaline phosphatase was added followed by one hour incubation at 37 ℃. Finally, the wells were washed with tris-HCl and substrate buffer containing p-nitrophenylphosphate was added. The absorption at 405nm was measured half an hour later.

Chimeric Mab T1 was used as a negative control.

FIG. 5 shows the specific binding of chimeric MAb P3 to antigen.

Example 4 preparation of chimeric MAb1E 10.

As mentioned previously, the synthesis of cDNA was obtained by a reaction using reverse transcriptase, starting with RNA obtained from the hybridoma producing Mab1E 10. The sequences of the specific primers used in this reaction are shown below:

for VH:

5′GGGGCTAGCTGAGGAGACTGTGAGAGTGGT 3′

for VK:

5′GCGTCTAGAACTGGATGGTGGGAAGATGGA 3′

the cDNAs VH1E10 and cDNAVK1E10 were amplified by PCR using Taq polymerase and specific primers.

For VH:

primer 1 (signal peptide):

5′GGGGATATCCACCATGG(AG)ATG(CG)AGCTG(TG)GT(CA)AT(CG)CTCTT3′

primer 2(CH 1):

5′GGGGCTAGCTGAGGAGACTGTGAGAGTGGT 3′

for VK:

primer 1 (signal peptide):

5′GGGGTTAACCACCATGAGG(GT)CCCC(AT)GCTCAG(CT)T(CT)CT(TG)GG(GA)3′

primer 2 (Ck):

5′AGCGTCGACTTACGTTT(TG)ATTTCCA(GA)CTT(GT)GTCCC3′

the PCR product was cloned into a TA vector (TA cloning kit, Invitrogen). Twelve independent clones were sequenced by dideoxy method using T7DNA Pol (Pharmacia) (FIGS. 7 and 8). The most homologous sequence groups for VH1E10 and VK1E10 were determined by homology search analysis. According to the Kabat classification, VH1E10 and VK1E10 sequences have high homology to promiscuous and V groups, respectively.

After digestion with the restriction enzymes ECORV and NHEI for VH1E10 and HincII and SALI for VK1E10, they were cloned into expression vectors previously digested with the same enzymes, PAH4604 and PAG4622 for VH and VK, respectively. The expression vector is supplied by Sherie Morrison (UCLA, California, USA) and is suitable for the expression of immunoglobulins in mammalian cells. The vector PAH4604 already includes the constant regions of human IgG1 and human PAG4622 (Coloma et al (1992): a novel vector for expressing antibody molecules using variable regions produced by the polymerase chain reaction (novelvers for the expression of antibody molecules using variable regions generated by polymerase chain reaction), J.Immunol.meth.152: 89-104). The resulting constructs were 1E10VH-PAH4604 and 1E10VK-PAG 4622.

NS-0 cells were transfected with 10. mu.g of 1E10VK-PAG4622, and clones expressing light chains were transfected with 10. mu.g of 1E10VH-PAH4604, in both cases, before transfection, the DNA was linearized with PvuI, precipitated with ethanol and dissolved in 50. mu.l PBS.

Centrifuging to obtain a product of about 10⁷The cells of (4) were resuspended in 0.5ml PBS together with the digested DNA in the electroporation cuvette. After 10 minutes on ice, the cells were given a pulsed current of 200 Vo1ts 960 μ F and placed on ice for another 10 minutes. Cells were plated in 96-well plates with D' MEM F12 supplemented with 10% fetal bovine serum. Two or four days later, selection medium (D' MEM F12 containing mycophenolic acid 0, 45. mu.g/ml or histidinol 10mM, respectively) was added. After 14 days, the transfected clones were visible to the naked eye.

The presence of human antibodies in the medium in the wells containing the transfected clones was determined by ELISA. Microtiter plate wells were covered with goat anti-human kappa light chain (for clones producing human kappa chains) or anti-human IgG (gamma chain specific) (for clones producing complete antibodies) antibodies. After washing with PBST (phosphate buffered saline containing 0.05% Tween 20), diluted medium containing the transfection product was added to each microtiter plate well for one hour at 37 ℃. The wells were washed with PBS-T and either goat anti-human kappa light chain conjugated to horseradish peroxidase or goat anti-human IgG conjugated to alkaline phosphatase (gamma chain specific) was added followed by one hour incubation at room temperature. After the wells were washed with PBS-T, substrate buffers containing o-phenylenediamine or p-nitrophenyl phosphate were added, respectively. Half an hour later, the absorption at 492 or 405nm, respectively, was measured.

Example 5. acquisition of different forms of humanized antibody 1E 10.

The murine VH1E10 and VK1E10 sequences (fig. 6 and 7) were compared to human sequences. The most homologous human sequences are shown in FIGS. 8 and 9. Amphipathic helical fragments or potential T cell epitopes were detected in the murine 1E10 variable region sequence and judicious strategies for amino acid substitutions were made to disrupt or humanize potential T cell epitopes in the murine sequence according to the method used.

Analysis of VH1E10 yielded (fig. 8)3 amphipathic fragments, the first fragment comprising FR1, the second fragment comprising FR2, and the third fragment comprising FR 3. It was decided to replace the residues at positions 5, 40, 42 and 87 (numbering 83 according to Kabat) with the residues at the same position in the most homologous human sequence. The amino acids Gln, Arg, Glu are replaced by Val, Ala, Gly and Arg, respectively.

The substitution was performed by overlapping PCR using a series of different primers (Kammann et al (1989) Rapid insertion mutagenesis by DNA Polymerase Chain Reaction (PCR)) Nucleic Acids Res., 17: 5404.

The primers used for the mutation at position 5 of the heavy chain are 1 and 2 and 3 and 4, the sequences of which are as follows

Primer 1:

5′CAGGTTCAGCTGGTGCAGTCTGGAGCT 3′

primer 2:

5′GGGGCTAGCTGAGGAGACTGTGAGAGTGGT 3′

primer 3:

5′AGCTCCAGACTGCACCAGCTGAACCTG 3′

primer 4:

5′GGGGATATCCACCATGG(AG)ATG(CG)AGCTG(TG)GT(CA)AT(CG)CTCTT 3′

after checking the point mutation at position 5 by sequencing, mutations at positions 40 and 42 were introduced.

Primers for heavy chain position 40 and 42 mutations:

primer 1:

5′TGGGTGAGGCAGGCGCCTGGGCAGGGACTTGAG 3′

primer 2:

5′GGGGCTAGCTGAGGAGACTGTGAGAGTGGT 3′

primer 3:

5′CTCAAGTCCCTGCCCAGGCGCCTGCCTCACCCA 3′

primer 4:

5′GGGGATATCCACCATGG(AG)ATG(CG)AGCTG(TG)GT(CA)AT(CG)CTCTT 3′

after checking the point mutations at positions 40 and 42 by sequencing, a mutation at position 87 (numbering 83 according to Kabat) was introduced.

Primers for heavy chain position 87 (numbering 83 according to Kabat):

primer 1:

5′CTCAGCAGGCTGCGGTCTGAGGACTCT 3′

primer 2:

5′GGGGCTAGCTGAGGAGACTGTGAGAGTGGT 3′

primer 3:

5′AGAGTCCTCAGACCGCAGCCTGCTGAG 3′

primer 4:

5′GGGGATATCCACCATGG(AG)ATG(CG)AGCTG(TG)GT(CA)AT(CG)CTCTT 3′

since the residues are involved in the three-dimensional structure of the binding site, no other substitutions are made.

The point mutations were verified by sequencing. The resulting construct was 1E10VHhu and cloned into the PAH4604 expression vector. The resulting construct is 1E10VH-PAH 4604.

Analysis of VK1E10 also yielded 3 amphipathic fragments (FIG. 9), the first comprising FR1, the second comprising CDR1, and the third comprising FR 3. The residues at positions 7, 8 and 15 were determined to be replaced with residues at the same position in the most homologous human sequence. The amino acids Thr, Thr and Leu are replaced by Ser, Pro and Val, respectively. The replacement was carried out by overlap PCR using primers 1 and 2 and 3 and 4 (Kammann et al (1989) Rapid insertion of mutagenized DNA by Polymerase Chain Reaction (PCR)), Nucleic acids, sRs., 17: 5404), the sequences of the primers used were as follows:

primers for light chain position 7, 8 and 15 mutations:

primer 1:

5′CAGATGACACAGTCTCCTTCCTCCCTGTCTGCCTCTGTGGGAGACAGAGTC 3′

primer 2:

5′AGCGTCGACTTACGTTT(TG)ATTTCCA(GA)CTT(GT)GTCCC 3′

primer 3:

5′GACTCTGTCTCCCACAGAGGCAGACAGGGAGGAAGGAGACTGTGTCATCTG 3′

primer 4:

5′GGGGTTAACCACCATGAGG(GT)CCCC(AT)GCTCAG(CT)T(CT)CT(TG)GG(GA)3′

the point mutations were verified by sequencing. The resulting construct was 1E10Vkhu and cloned into a PAG4622 expression vector. The resulting construct was 1E10Vkhu-PAG 4622.

To express humanized antibody 1E10, NS-0 cells were transfected with 1E10VHhu-PAH4604 and 1E10Vkhu-PAG 4622.

1E10hu antibody was transfected according to the electroporation and detection methods described previously for chimeric antibodies.

Example 6: biological activity of chimeric MAb1E 10.

The biological activity of the chimeric MAb1E10 was tested by ELISA assay for specific binding to antigen.

For recombinant MAb1E10, microtiter plates were covered with MAb P3. After washing with PBST (phosphate buffered saline containing 0.05% Tween 20), non-specific binding was blocked with 1% Bovine Serum Albumin (BSA) in PBST and incubated at 37 ℃ for one hour. The wells were washed and incubated with purified recombinant Mab1E10 for one hour at 37 ℃. The wells were washed with PBST and goat anti-human antibody conjugated to alkaline phosphatase was added followed by one hour incubation at 37 ℃. Finally, the wells were washed with PBST and substrate buffer containing p-nitrophenylphosphate was added. The absorption at 405nm was measured half an hour later.

Chimeric Mab C5 was used as a negative control.

FIG. 10 shows the specific binding of chimeric MAb1E10 to MAb P3.

Brief description of the drawings:

FIG. 1: VHP3DNA and deduced amino acid sequences. The Sequences are arranged according to Kabat numbering (Kabat et al (1991), Sequences of proteins of immunological interest, fifth edition, National Institute of health), and the occurring CDRs are indicated by dashed lines.

FIG. 2: VKP3DNA and deduced amino acid sequences. The Sequences are arranged according to Kabat numbering (Kabat and co-workers (1991), Sequences of proteins of immunological interest, fifth edition, National Institute of health), and the occurring CDRs are indicated by dashed lines.

FIG. 3: VHP3 was aligned with the most homologous human sequence. The amphipathic segment is underlined and the CDRs are in bold.

FIG. 4: VKP3 was aligned with the most homologous human sequence. The amphipathic segment is underlined and the CDRs are in bold.

FIG. 5: specific binding of chimeric Mab P3 to GM3 (NeuGc). Different concentrations of Mab P3 and Mab T1 (negative control) were determined by ELISA. Microtiter plates were covered with GM3(NeuGc) and GM3(NeuAc) (negative control) gangliosides in methanol and assayed for specific binding.

FIG. 6: VH1E10 DNA and deduced amino acid sequence. The Sequences are arranged according to Kabat numbering (Kabat and co-workers (1991), Sequences of proteins of immunological interest, fifth edition, National Institute of health), and the occurring CDRs are indicated by dashed lines.

FIG. 7: VK1E10 DNA and deduced amino acid sequence. The Sequences are arranged according to Kabat numbering (Kabat et al (1991), Sequences of proteins of immunological interest, fifth edition, National Institute of health), and the occurring CDRs are indicated by dashed lines.

FIG. 8: VH1E10 was aligned with the most homologous human sequence. The amphipathic segment is underlined and the CDRs are in bold.

FIG. 9: VK1E10 was aligned with the most homologous human sequence. The amphipathic segment is underlined and the CDRs are in bold.

FIG. 10: specific binding of chimeric Mab1E10 to murine Mab P3. Different concentrations of Mab1E10 and Mab C5 (negative control) were determined by ELISA. Microtiter plates were covered with Mab P3 and Mab A3 (negative control) and their specific binding was determined.

Claims (5)

1. A chimeric monoclonal antibody obtained from the murine monoclonal antibody Mab P3, which recognizes gangliosides containing N-acetylalcoholized sialic acid, said Mab P3 being produced by the hybridoma cell line deposited under accession number ECACC94113026, wherein the sequences of the highly variable regions of its heavy and light chains are as follows:

heavy chain

CDR1：RYSVH

CDR2：MIWGGGSTDYNSALKS

CDR3：SGVREGRAQAWFAY

Light chain

CDR1：KASQDVSTAVA

CDR2：SASYRYT

CDR3：QQHYSTPWT。

2. The chimeric monoclonal antibody of claim 1, wherein the Framework Regions (FRs) of its heavy and light chains have the following sequences:

heavy chain

FR1：QVQLKESGPGLVAPSQSLSITCTVSGFSLS

FR2：WVRQPPGKGLEWLG

FR3：RLSISKDNSKSQVFLKMNSLQTDDTAMYYCAR

FR4：WGQGTLV

Light chain

FR1：DIVMTQSHKFMSTSVGDRVSITC

FR2：WYQQKPGQSPKLLIY

FR3：GVPDRFTGSGSGTDFTFTISSVQAEDLAVYYC

FR4：FGGGTKL。

3. The monoclonal antibody according to claim 1 or 2, wherein it comprises at least one of the following substitutions for its humanization and retention of binding properties to an antigen:

light chain

Position 8: replacement of His with Pro

Position 9: replacement of Lys with Ser

Position 10: replacement of Phe with Ser

Position 11: replacement of Met by Leu

Position 13: replacement of Thr with Ala.

4. The monoclonal antibody of claim 1 or 2, wherein the heavy chain constant region comprises the amino acid sequence of γ -1 chain, the light chain constant region comprises the amino acid sequence of kappa chain, and both γ -1 chain and kappa chain are derived from human immunoglobulin.

5. A cell line producing the monoclonal antibody of any one of claims 1 to 4.