Protein epitopes in carcinoembryonic antigen. Report of the ISOBM TD8 workshop.

To characterize antigenic sites in carcinoembryonic antigen (CEA) further and to investigate whether there are differences between colon tumor CEA and meconium CEA (NCA-2) that can be detected by anti-CEA monoclonal antibodies (MAb), 19 new anti-CEA MAb were analyzed with respect to specificity, epitope reactivity and affinity. Their reactivities were compared with 10 anti-CEA MAb with known CEA-domain binding specificity that have previously been classified into five nonoverlapping epitope groups, GOLD 1-5. Cross-inhibition assays with antigen-coated microtiter plates and immunoradiometric assays were performed in almost all combinations of MAbs, using conventionally purified CEA (domain structure: N-A1B1-A2B2-A3B3-C) from liver metastasis of colorectal carcinomas, recombinant CEA, meconium CEA (NCA-2), truncated forms of CEA and NCA (CEACAM6) as the antigens. The affinity of the MAbs for CEA was also determined. The new MAbs were generally of high affinity and suitable for immunoassays. Three new MAbs were assigned to GOLD epitope group 5 (N-domain binding), 3 MAbs to group 4 (A1B1 domain), 1 to group 3 (A3B3 domain), 3 to group 2 (A2B2 domain) and 3 to group 1 (also the A3B3 domain). Three MAbs formed a separate group related to group 4, they were classified as GOLD 4' (A1B1 domain binding). The remaining 3 MAbs appear to represent new subspecificities with some relationship to GOLD groups 1, 2 or 4, respectively. Five MAbs, all belonging to epitope group 1 and 3, reacted strongly with tumor CEA but only weakly or not at all with meconium CEA, demonstrating that the two products of the CEA gene differ from each other, probably due to different posttranslational modifications.     

Monoclonal antibodies identify a CEA crossreacting antigen of 95 kD (NCA-95) distinct in antigenicity and tissue distribution from the previously described NCA of 55 kD

During the selection of monoclonal antibodies (MAb) raised against purified carcinoembryonic antigen (CEA), two MAbs were identified which immunoprecipitated a glycoprotein of 95 kD present both in perchloric acid extracts of normal lung and on the surface of normal granulocytes. This antigen was distinct from the previously reported normal glycoprotein crossreacting with CEA (NCA) which had a molecular weight of 55 kD. The difference between the smaller and the larger crossreacting antigens termed NCA-55 and NCA-95, respectively, was demonstrated by SDS-polyacrylamide gel electrophoresis, by elution from Sephadex-G200 and by selective binding to a series of anti-CEA MAb. Out of six MAb which all bound CEA purified from colon carcinoma, three did not react with these two crossreacting antigens, one bound only NCA-95, one reacted only with NCA-55 and one reacted with both NCA-55 and NCA-95. Immunoadsorbent purified preparations of 125I labelled NCA-95 and NCA-55 were found useful for the screening of new anti-CEA MAb. In addition, when tested on frozen sections of colon carcinoma, normal spleen, normal lung and pancreas, each type of MAb gave a clearly different pattern of reactivity. The three anti-CEA MAb which did not bind any of the crossreacting antigens stained only the colon carcinoma cells; the MAb binding to either one of the two types of NCA gave a similar pattern of reactivity both on colon carcinoma cells and on granulocytes. However, on normal lung and pancreas, the MAb binding NCA-55 stained granulocytes as well as bronchiolar and alveolar epithelial cells in lung and inter- and intra-lobular duct epithelial cells in pancreas, whereas the MAb binding only NCA-95 stained only the granulocytes. Thus, the newly identified NCA-95 appears to differ from NCA-55 not only in terms of molecular size and antigenicity but also by the fact that in normal lung and pancreas it is found in granulocytes but not in epithelial cells.     

Serologic mapping and biochemical characterization of the carcinoembryonic antigen epitopes using fourteen distinct monoclonal antibodies

A series of 14 monoclonal anti-carcinoembryonic antigen (CEA, Mr 180,000) antibodies (MAbs) that show a strong degree of selective reactivity for human colon carcinomas versus normal adult tissues were used to construct a serological map of the CEA molecule. The MAbs were generated using extracts of colon carcinomas as immunogen and are thus given a COL designation. None of the 14 COL-MAbs tested were reactive with purified non-specific cross-reacting antigen (NCA, Mr 55,000) from normal lung, although some showed reactivity to human granulocytes. All the COL-MAbs tested were reactive with normal fecal antigen-2 (NFA-2, Mr 170,000); however, many of the COL-MAbs demonstrated a higher affinity constant to CEA than to NFA-2. Cross-competition radioimmunoassays classified the 14 COL-MAbs into 5 groups. The chemical nature of the COL-binding domains was tested using chemically or enzymatically treated CEA; all reacted with periodate-treated CEA and deglycosylated CEA, indicating that the COL-reactive epitopes appear to be of a proteinaceous nature. Heat treatment, reduction, alkylation, pepsin digestion or pronase treatment of CEA, however, gave differential results with respect to COL binding. Antibody titration experiments were carried out to define differential reactivities to colorectal carcinoma versus NCA-containing granulocyte extracts; these results were compared with results obtained using several anti-CEA MAbs that have been used in clinical trials. Granulocyte binding and biochemical studies showed that the COL MAbs may distinguish as many as 7 to 10 CEA epitopes.     

Immunohistochemical localization and molecular characteristics of three monoclonal antibody-defined epitopes detectable on carcinoembryonic antigen (CEA)

Three murine monoclonal antibodies (MAbs) reactive to different epitopes on CEA were selected according to their ability to bind to various human tissue sections. The most selective MAb, BW 431/31, bound to the majority of colon carcinomas but only faintly to normal colon mucosa. In addition to the tissues stained by MAb BW 431/31, MAb BW 250/183 reacted with granulocytes, colon mucosa and faintly with pancreatic ducts. The third MAb, BW 374/14, reacted with granulocytes, colon mucosa, strongly with pancreatic ducts and with alveolar and bronchial epithelium. The antigenic determinants recognized by the 3 MAbs in human tissue sections were resistant to formaldehyde fixation and paraffin embedding as well as to periodic acid oxidation and neuraminidase treatment. The last two treatments suggest that the epitopes are protein in nature. Using MAb affinity chromatography, 3 antigen preparations were isolated from a human colon carcinoma xenograft with an approximate molecular weight of 180 kd. These preparations were shown to bear the epitopes from each of the MAbs and from a polyclonal antiserum specific for purified CEA. Furthermore, the ability of MAb BW 431/31 to localize its antigenic determinant in vivo on a human colon carcinoma xenograft is evaluated and its possible application in the patient is suggested.     

An idiotypic replica of carcinoembryonic antigen inducing cellular and humoral responses directed against human colorectal tumours.

A monoclonal anti-idiotypic antibody (anti-Id MAb) 708 (IgG2b), which inhibited the binding of NCRC23 (IgG1) MAb to CEA and prevented radiolabelled CEA from binding to MAb NCRC23, was produced. No recognition of 3 other anti-CEA antibodies, 3 other IgG1 or 2 IgM MAbs was observed with this anti-idiotypic antibody. When an immunoblotting technique was used, 708 anti-Id MAb failed to bind to isolated heavy or light chains of MAb NCRC23, whereas binding was observed with intact antibody. Mouse, rat and human lymphocytes (in vitro) were immunized with 708 anti-Id MAb and the resultant Ab3 antibodies all inhibited binding of labelled 708 anti-Id MAb to MAb NCRC23 and also reacted with CEA, showing that 708 anti-Id MAb induced anti-CEA antibody responses. Similarly, mice immunized with 708 anti-Id MAb could be restimulated in vitro with either CEA or tumour cells expressing CEA which induced specific T-cell proliferative responses. Human tumour-infiltrating lymphocytes isolated from colorectal tumours or peripheral blood T cells from cancer patients were stimulated in vitro with 708 anti-Id MAb or an irrelevant IgG2b antibody. Six days later both sets of lymphocytes were restimulated with CEA, and lymphocytes primed to 708 anti-Id MAb proliferated in response to CEA. These results suggest that 708 anti-Id MAb can act as an idiotypic replica of CEA and stimulate cellular and humoral anti-CEA immune responses. It is therefore of great interest as an idiotypic vaccine against colorectal cancer.     

Targeted therapy of athymic mice bearing GW-39 human colonic cancer micrometastases with 131I-labeled monoclonal antibodies.

The therapeutic potential of radiolabeled antibodies is usually evaluated in experimental animal models bearing s.c. xenografts. We have established a micrometastatic model of the GW-39 human colonic carcinoma in the nude mouse lung (J. Natl. Cancer Inst., 83: 627-632, 1991) and presented preliminary findings on the efficacy of a 131I-anticarcinoembryonic antigen (CEA) antibody in this model. We now extend our observations on the use of radioiodinated labeled monoclonal antibodies (MAbs) to treat multiple small tumor nodules. Biodistribution and dosimetry analysis was performed for intact and F(ab')2 of NP-4 anti-CEA IgG, Mu-9 anti-colon-specific antigen IgG, isotype-matched irrelevant anti-AFP IgG, and intact MAb 34A anti-lung endothelial IgG antibody. Comparisons were made for rad dose delivered to small s.c. tumors, normal lung, lung with tumor nodules, and isolated tumor nodules. Survival curves were generated for tumor-bearing animals treated 1, 7, or 14 days after tumor cell implantation with these antibodies using the maximal tolerated dose for intact antibodies (275 microCi) and for F(ab')2 fragments (1.2 mCi). The studies established the following observations: (a) in contrast to previous results in a bulky tumor model in hamsters, intact antibodies are more therapeutic than MAb fragments for both NP-4 and Mu-9; (b) tumor nodule size, even on the microscopic level, affects therapeutic outcome; antibodies were more effective when administered 7 days postimplantation (mean nodule diameter, 150 microns) compared with treatment 14 days postimplantation (mean nodule diameter, 750 microns); (c) administration of radioiodinated Mu-9 was exquisitely effective on single avascular tumor cells that had seeded in lung; irrelevant antibody was minimally radiotoxic; (d) as in the bulky disease model, the anti-colon-specific antigen p antibody delivers a higher rad dose than the anti-CEA antibody and is significantly more therapeutic in the micrometastasis model; (e) a higher affinity anti-CEA antibody (MN-14) recognizing the same epitope on CEA as NP-4 was equally therapeutic; (f) the use of MAb directed against the lung endothelium was not as therapeutic as a tumor-associated antibody; and (g) all tumor-associated antibodies were more efficacious than administration of the maximal tolerated dose of 5-fluorouracil and leucovorin in this human tumor-xenograft model. These results provide further support for the use of radioimmunotherapy in the handling of minimal disease, probably as part of an adjuvant treatment regimen.     

Immunohistochemical analysis of pulmonary and pleural neoplasms using a monoclonal antibody (47D10) which reacts with nonspecific cross-reacting antigen

A series of 251 human pulmonary carcinomas were analyzed immunohistochemically for the antigens recognized by a new monoclonal antibody (MAb) 47D10. These antigens are part of a complex family of substances similar to, yet distinct from carcinoembryonic antigen (CEA), and are termed 'nonspecific cross-reacting antigens' (NCAs). The NCA epitope recognized by the MAb 47D10 is expressed on the cell surface and has previously been shown to be distinct from epitopes detected by several anti-CEA MAbs, as well as by MAbs 19-9 and Du-PAN-2. The NCA epitope recognized by MAb 47D10 is well preserved in formalin-fixed and paraffin-embedded tissues. Using immunohistochemistry, this epitope has been shown to have a limited biodistribution in normal tissues, and to be expressed by adenocarcinomas arising in the pancreas, colon, breast, ovary, prostate and lung. The frequency and pattern of NCA expression in human pulmonary neoplasms was found to correlate with the known distribution of CEA: and was often present in the non-small-cell carcinomas. In addition, the expression of CEA relative to NCA was evaluated in a select group of non-small-cell carcinoma cases using several anti-CEA MAbs, to directly compare the expression of CEA to NCA. In general, the NCA reaction pattern is more intense and expressed on more cells within the tumors than that of CEA expression.     

Rat monoclonal antibodies produced against rat colorectal adenocarcinomas define tumor- and colon-associated, auto-immunogenic antigens.

Monoclonal antibodies (MAbs) were derived from rats immunized against allo- and syngeneic rat colon carcinomas. Screening was performed by immunohistochemistry modified for MAbs of the same species as the tissue used for frozen sections. From 2 fusions, several MAbs were found that bound to syngeneic tumor tissue but showed little or no staining of normal adult tissues. Another group of MAbs demonstrated an intracellular staining of tumor cells and staining of mucus and goblet cells in the distal 2/3 of the normal colon. A third group of MAbs showed staining of both tumor and normal colon tissue. Staining patterns were reproduced with 6 MAbs selected from the first 2 groups after purification and biotinylation. One MAb, 10B12, recognized an antigen expressed in 10/11 colon carcinomas, the lowest parts of colonic crypts, intracellularly in a subpopulation of pancreatic acinar cells and mucus in antrum. It was used for affinity purification of a high-molecular-weight antigen, to which 3 of the other rat MAbs also bound. This antigen was also bound by antibodies to blood group A and isogloboseries carbohydrate determinants. Competition with the labelled 10B12 MAb for binding to the purified antigen was demonstrated in sera of tumor-bearing and immune rats. Thus, this high-molecular-weight glycoprotein expressed both auto-antigenic, tumor-associated and tissue-type-restricted epitopes. The 10B12 MAb homogeneously stained the majority of colorectal carcinomas and the antigenic determinant was expressed on the cell surface of 12/13 tissue-cultured colon-carcinoma clones. Modulation of the cell-surface antigen phenotype by recombinant rat interferon-gamma markedly increased expression of this determinant.     

Penetration and binding of radiolabeled anti-carcinoembryonic antigen monoclonal antibodies and their antigen binding fragments in human colon multicellular tumor spheroids

The binding and penetration of two 125I-labeled anti-carcinoembryonic antigen (CEA) monoclonal antibodies (MAb) and their F(ab')2 and Fab fragments were measured in multicellular spheroids of poorly (HT29) and moderately well differentiated (Co112) human colon adenocarcinomas which express different amounts of CEA. Spheroids cultured in vitro model tumor microenvironments where poor vascular supply may modulate antigen expression and accessibility. The two MAb studied, 202 and 35, were shown previously to react with different CEA epitopes and to have high affinities of 1.2 and 5.8 X 10(9) M-1, respectively. MAb 202 has also been shown to cross-react with antigens present on human granulocytes and normal epithelial cells from human lung and pancreas. Specific binding of intact MAb and fragments of both antibodies was demonstrated for both types of human colon carcinoma spheroids compared to mouse colon carcinoma (CL26) and mammary tumor (EMT6/Ro) spheroids. Total binding of MAb and fragments was greater (1.5- to 2.5-fold) after 4 h compared to 1 h of exposure; the amount of binding compared to control IgG1 was 5- to 30-fold greater after 1-h incubation and 15 to 200 times greater after 4 h. This binding was stable as demonstrated by short and long wash experiments at 37 degrees and 4 degrees C. The binding of F(ab')2 and Fab fragments of the anti-CEA MAb 35 to spheroids of human colon Co112 was almost 2-fold greater than that of the intact MAb. However, for MAb 202, the binding of intact MAb and F(ab')2 was greater than that of Fab fragments. In addition the binding of both intact and F(ab')2 fragments of MAb 202 was greater than that obtained with MAb 35. Specific binding of both antibodies to HT29 spheroids, which express less CEA, was decreased for MAb and fragments of both 202 and 35. Autoradiography and immunoperoxidase experiments were performed to determine the penetration of MAb and fragments after incubation with intact spheroids. Comparisons were made with labeled MAb directly applied to frozen sections of spheroids. F(ab')2 and Fab fragments of both antibodies were bound at the surface of intact spheroids and penetrated to eight to ten cells, but the intact MAb were localized mainly at the spheroid surface and the outer one to three cell layers. There was much less binding at the surfaces of HT29 compared to Co112 spheroids. An enzyme immunoassay using MAb 35 and 202 demonstrated that Co112 spheroids produced about 8-fold more CEA/mg of cell protein than did monolayer cultures.(ABSTRACT TRUNCATED AT 400 WORDS)     

Mouse monoclonal antibodies which recognize a human (beta 1-4)galactosyl-transferase associated with tumor in body fluids.

Mouse monoclonal antibodies against human (beta 1-4)galactosyl-transferase (GalT) purified from human ovarian tumor effusion fluids were prepared and characterized. GalT purified from normal human plasma showed a single diffused band in nondenaturing polyacrylamide gel electrophoresis, but GalT purified from human ovarian tumor effusion fluids showed several oligomeric bands and a monomeric band in nondenaturing polyacrylamide gel electrophoresis. These oligomeric bands were dissociated into monomer by urea treatment and polymerized by a 2-mercaptoethanol treatment. Nine monoclonal antibodies (MAb) were prepared by immunization of purified GalT from human ovarian tumor effusion fluids and classified into three groups. Type I MAbs (MAb8611, MAb8913, and MAb8919) reacted only to the GalT monomer. Type II MAbs (MAb4880, MAb8507, and MAb8628) reacted to both the GalT monomer and the GalT polymer. Type III MAbs (MAb7907, MAb8513, and MAb8677) reacted only to the GalT polymer. These MAbs except MAb7907 could recover GalT enzyme activity from effusion fluids by immunoprecipitation. A fraction passed through MAb8513 affinity chromatography still showed reactivity to MAb8919, demonstrating that an epitope of MAb8513 resides on a minor part of GalT. A sandwich immunoassay (MAb8513-MAb8628HRP) was developed, and serum samples from ovarian cancer patients and benign ovarian patients were tested. The levels of sandwich immunoassay of serum samples from cancer were elevated significantly compared to those from benign and did not necessarily correlate to total GalT enzyme activity in serum samples. These results suggested that MAb8513 (Type III) might recognize a unique GalT associated with tumor (GAT).     

Tumor targeting with newly designed biparatopic antibodies directed against two different epitopes of the carcinoembryonic antigen (CEA)

In an attempt to improve tumor targeting and tumor retention time of monoclonal antibodies (MAbs), we prepared biparatopic antibodies (BpAbs) having the capability of binding 2 different non-overlapping epitopes on the same target antigen molecule, namely, the carcinoembryonic antigen (CEA). Six BpAbs were constructed by coupling 2 different Fab' fragments from 4 different specific anti-CEA MAbs recognizing 4 CEA epitopes (Gold 1-4). Demonstration of the double paratopic binding of these antibodies for CEA was confirmed in vitro by inhibition radioimmunoassay and cross-inhibition analysis by surface plasmon resonance (SPR; BIACORE) technology. Using the latter technique, the affinity constants for CEA immobilized onto the sensor chip were found to range from 0.37 to 1.54 x 10(9) M(-1) for the 4 parental F(ab')2 fragments and from 1.88 to 10.14 x 10(9) M(-1) for the BpAbs, demonstrating the advantage of biparatopic binding over conventional F(ab')2 binding. The Ka improvement was particularly high for BpAb F6/35A7 and BpAb F6/B17 with a 9.5- and 8.1-fold increase, respectively, as compared with the parental F(ab')2. In vivo, the 6 BpAbs were compared with their 2 respective parental F(ab')2 by injection of 131I-BpAb/125I-F(ab')2 parental fragments into nude mice xenografted with the human colon carcinoma T380. Dissection 72 hr post-injection demonstrated that BpAb B17/CE25 and BpAb F6/B17 gave higher tumor uptake than that of their parental F(ab')2. This finding is particularly interesting for BpAb F6/B17, which compared favorably with the F6 F(ab')2, one of the best parental F(ab')2 fragments used in our study.     

Enhancement of colorectal tumor targeting using a novel biparatopic monoclonal antibody against carcinoembryonic antigen in experimental radioimmunoguided surgery.

Biparatopic CEA, carcinoembryonic antigen (MAb) was newly designed and tested as to whether it enhanced the accuracy of tumor detection by reducing non-specific binding in experimental radioimmunoguided surgery. Biparatopic MAb was prepared by using cross-linking of reduced Fab' fragments from PR1A3 and T84.66. Fifty-nine tumors from 2 human colorectal carcinoma cell lines with high (KM-12c) and low (Clone A) carcinoembryonic antigen (CEA) expression were successfully implanted subcutaneously on the backs of 42 nude mice. Tumors were localized using 125I-labeled MAbs: IgG, F(ab')(2) and Fab' of PR1A3, and biparatopic MAb of PR1A3 and T84.66. Radioactivity counted on a portable radioisotope detector correlated well with that counted on a gamma counter (p < 0.001). Accumulations of radioactivity in control mice without tumorigenesis were the greatest in PR1A3 IgG-pretreated mice and the least in biparatopic MAb-pretreated mice. Tumors of 2 cell lines did not differ in the distribution of radiolabeled MAbs. Localization indices of the tumor in various organs revealed 1.3 to 4.1 in PR1A3 IgG-pretreated mice, 2.4 to 6.6 in fragment MAbs of PR1A3-pretreated mice and 2 to 4.6 in biparatopic MAb-pretreated mice. Silver grains and immune staining were predominantly distributed in tumor cells of all types of MAb-pretreated mice. Sensitivity and specificity of tumor localization by radioimmunoguided surgery (RIGS) were the highest in the biparatopic MAb-pretreated mice (90.9% and 94.5%, respectively) and the least in the PR1A3 IgG-pretreated mice (50% and 72%). The biparatopic MAb using 2 anti-CEA MAbs against different epitopes achieved a great affinity and avidity with accurate localization of colorectal carcinoma in experimental radioimmunoguided surgery.     

Fully human IgG and IgM antibodies directed against the carcinoembryonic antigen (CEA) Gold 4 epitope and designed for radioimmunotherapy (RIT) of colorectal cancers

Background  

Human monoclonal antibodies (MAbs) are needed for colon cancer radioimmunotherapy (RIT) to allow for repeated injections. Carcinoembryonic antigen (CEA) being the reference antigen for immunotargeting of these tumors, we developed human anti-CEA MAbs.     

Generation of a recombinant mouse-human chimaeric monoclonal antibody directed against human carcinoembryonic antigen

A procedure was devised for the identification and specific cloning of functionally rearranged variable region immunoglobulin (Ig) gene segments from genomic DNA of a murine hybridoma cell line which produces a high-affinity monoclonal antibody (MAb) directed against human carcinoembryonic antigen (CEA). The cloned, functionally-rearranged murine Ig H-chain and L-chain variable region gene segments were incorporated into plasmid vectors capable of directing the expression of a chimaeric mouse-human antibody molecule with human (gamma 4, kappa) constant region sequences. Expression plasmids were transfected into a mouse myeloma cell line by electroporation and transfectomas secreting functional chimaeric antibody selected. Chimaeric antibody generated by transfectomas was analysed and shown to compete effectively with its murine counterpart for binding to the CEA epitope, and to have an equivalent antigen-binding affinity. This anti-CEA recombinant antibody should find application in in vivo diagnosis by immunoscintigraphy of human colonic carcinoma, and possibly also in therapy of the disease, overcoming some of the difficulties associated with the repeated use of non-human immunoglobulins in human patients.     

Polyclonal and monoclonal anti-CEA antibodies in immunolocalization of colorectal cancer

Antibodies to carcinoembryonic antigen (CEA) from sheep and monkey were immunoadsorbent purified. Mouse monoclonal antibody (MAb) anti-CEA I-38S1 and Fab fragments of this antibody were prepared from mouse ascitic fluid. The IgG preparations were labelled with 123I or 131I, the Fab fragments with 131I. The antibody reactivity was unchanged after labelling. Patients with advanced colorectal carcinomas received an intravenous injection of 50-200 MBq 123I or 30-160 MBq 131I coupled to 250-500 micrograms antibody or antibody fragment. Patient examinations were performed using emission tomography (SPECT) and/or conventional gamma camera scintigraphy. The specific localization of labelled anti-CEA to tumor was compared to known tumor localized by CAT-scan, other x-ray methods or laparotomy, 50% of known tumors were accurately localized with sheep anti-CEA. In contrast, 70-80% of known tumor sites were correctly localized with polyclonal monkey anti-CEA antibodies, with monoclonal anti-CEA antibodies or with Fab fragments of the latter. A few previously unknown tumors were detected.     

Epitopes of carcinoembryonic antigen defined by monoclonal antibodies prepared from mice immunized with purified carcinoembryonic antigen or HCT-8R cells

A library of 18 monoclonal antibodies (MAbs) reactive with purified carcinoembryonic antigen (CEA) has been prepared. The specificity of these MAbs was tested and they have been separated into nine subgroups, each recognizing a different region of the CEA molecule. Seven MAbs from four of the groups also react with the nonspecific cross-reacting antigen. Some of the MAbs are directed against conformational determinants: three of the MAb groups bind poorly to sodium dodecyl sulfate-treated CEA, while five of the groups are not reactive with reduced and alkylated CEA. Three of the groups react with purified CEA but not with the cell surface CEA of HCT-8R cells, while the other groups react with both forms. The MAbs were tested for binding to fragments of CEA obtained by chemical cleavage and the groups of MAbs were found to react with different subsets of such fragments.     

Carcinoembryonic antigen continuous epitopes determined by the spot method.

Carcinoembryonic antigen (CEA) is a heavily glycosylated tumor-associated protein with an N-A1-B1-A2-B2-A3-B3 domain structure. Circulating CEA immunoassays are used for monitoring digestive cancer patients, and radiolabeled anti-CEA monoclonal antibodies (MAb) are used for the diagnosis and therapy of CEA-positive tumors. The five major nonoverlapping epitopes (Gold 1-5) have been broadly correlated with the domain organization, but there is no precise localization of the epitopes at the sequence level. In an attempt to identify the peptide sequences corresponding to the five Gold epitopes on the CEA molecule, we prepared a set of 227 overlapping fifteen-mer peptides corresponding to the complete CEA sequence with the SPOT method. Using five high affinity MAbs directed against the five CEA Gold epitopes, we demonstrated that none of these epitopes could be mimicked by a fifteen-mer peptide sequence. However, using rabbit and goat anti-CEA sera, we identified six major continuous antigenic regions. All are included in the Ig-like domains of the CEA: two in the A1 domain (residues 120-134 and 153-164), one each in the A2 (329-337) and A3 domains (508-513), one at the junction between the A3 and B3 domains (553-561) and one in the B3 domain (565-573). A very homologous sequence (common residues VSPRL) was mapped in each of the three A domains. Thus, in terms of occurrence of continuous epitopes, the Ig-like domains A1, A2, A3 and B3 seem to be the most antigenic parts of CEA. These peptide sequences should be good candidates for the future development of site-specific anti-CEA MAbs.     

Selection of a DTPA chelate conjugate for monoclonal antibody targeting to a human colonic tumor in nude mice

Our previous studies with a 90Y-labelled antibody against carcinoembryonic antigen (CEA) conjugated to the cyclic anhydride-DTPA (CA-DTPA) indicated that the accretion of 90Y in the bone may limit the application of 90Y-labelled antibodies for therapy. In this report, we have compared the tumor targeting of CA-DTPA-conjugated antibody to antibody conjugated with 4 isothiocyanatobenzyl (ITC-Bz) derivatives of DTPA in nude mice bearing a human colonic tumor xenograft. In biodistribution studies using an 111In-labelled anti-CEA murine monoclonal antibody (MAb), the CA-DTPA-conjugated MAb showed lower tumor uptake, faster blood clearance, and higher accretion in the liver than any of the 4 ITC-Bz-DTPA-conjugated MAbs. There were smaller differences among the 4 ITC-Bz-DTPA conjugates. Whole-body autoradiography of animals given 90Y-MAb prepared with the CA-DTPA or the ITC-Bz-DTPA showed less radioactivity in the bone with the ITC-Bz-DTPA-MAb than the CA-DTPA-MAb. 90Y uptake in the bone corresponded with regions of low proliferative activity as defined by 3H-labelled thymidine, suggesting that the 90Y was in the cortex rather than the marrow. These studies clearly show an advantage of the ITC-Bz-DTPA derivatives for 90Y and 111In labelling of MAbs.     

Characterisation of a humanised bispecific monoclonal antibody for cancer therapy.

A humanised bispecific monoclonal antibody (bsMAb) with binding specificity for carcinoembryonic antigen (CEA) on one arm and a radiolabelled chelate (DTPA-90Y) on the other arm was generated by consecutively transfecting the humanised genes of an anti-CEA MAb and the chimerised genes of an anti-chelate MAb into eucaryotic BHK cells using the calcium-phosphate coprecipitation technique. The antibodies secreted were of IgG3 isotype with a shortened hinge region (delta gamma 3) and light chains. Double transfectomas were screened for the secretion of bsMAbs using a double determinant enzyme-linked immunosorbent assay (ELISA) based on solid phase attached HSA-benzyl-DTPA and an anti-idiotypic MAb selective for the CEA-specific arm. After purification on two immunoaffinity chromatography columns, the humanised bsMAbs were characterised by SDS-PAGE and a quantitative binding assay in antigen excess. The purification procedure resulted in 95% reactive bispecific MAb. This humanised bsMAb may be employed in two phase radioimmunotherapy, binding to the tumour via the anti-CEA arm and localising a radiolabelled chelate with the other arm, without inducing a strong immune response observed sometimes with murine MAbs.     

Xenogeneic monoclonal antibodies in the management of cancer: control of their in vivo immunogenicity and induction of specific unresponsiveness using an antibody-drug immunoconjugate

A bispecific mouse monoclonal antibody (mAb) that recognises carcinoembryonic antigen (CEA) with one binding site and vinblastine (VLB) with the other was used, and its in vivo immunosuppressive effect specific for anti-mouse immunoglobulin (Ig) was studied. The antibody was incubated with VLB at a molar ratio (MR) of 1:1, and administered i.v. to rabbits. Control animals received either the MAb alone, or the MAb with VLB covalently linked (MR 1:1), or the parental anti-CEA with equimolar amount of VLB. Seven days later, the rabbit anti-mouse Ig primary response was measured, and found to be almost 55% reduced in the animals that received the VLB 'loaded' MAb. In vivo kinetics and stability experiments revealed that the T1/2 of the MAb was 68 +/- 5 h, whereas free VLB disappeared within minutes. It was concluded that as soon as the drug dissociates from the antibody's binding site, it is rapidly removed. This problem was overcome by subcutaneously implanting osmotic mini-pumps containing VLB. The pumps released the drug at a constant rate for a period greater than 1 week, saturating the antibody's binding site. Under these conditions rabbits developed 80% less anti-mouse Ig antibodies when the bispecific antibody was administered (compared with the parental anti-CEA). The immunosuppression observed was specific for the mouse Ig, under conditions compatible with the full clinical therapeutic potential of the MAb. In conclusion, these experiments show, that it is possible to develop hybrid antibodies that can act as a 'lethal bait' to any specific lymphocyte in vivo, thus preventing undesirable responses against the xenogeneic MAb.