Antibodies against carcinoembryonic antigen for cancer therapy and diagnosis

Abstract

The invention relates to antibodies against carcinoembryonic antigen (CEA) which have a direct cell growth inhibition activity on tumor cells expressing CEA and to their use for the treatment and diagnosis of cancer.

Claims

1. An antibody against carcinoembryonic antigen (CEA) which has a direct tumor cell growth inhibition activity on tumor cells expressing CEA and which comprises light-chain (VL) and heavy-chain (VH) variable domains complementarity-determining region (CDR) sequences selected from the group consisting of: (a) the VL-CDR1 sequence of SEQ ID NO: 1; the VL-CDR2 sequence AAT; the VL-CDR3 sequence of SEQ ID NO: 2; the VH-CDR1 sequence of SEQ ID NO: 3; the VH-CDR2 sequence of SEQ ID NO: 4; the VH-CDR3 sequence of SEQ ID NO: 5; and (b) the VL-CDR1 sequence of SEQ ID NO: 6; the VL-CDR2 sequence YAS; the VL-CDR3 sequence of SEQ ID NO: 7; the VH-CDR1 sequence of SEQ ID NO: 8; the VH-CDR2 sequence of SEQ ID NO: 9; the VH-CDR3 sequence SEQ ID NO: 10.

2. The antibody according to claim 1, which has a variable region formed by the association of a VL domain of SEQ ID NO: 11 and a VH domain of SEQ ID NO: 12 ora VL domain of SEQ ID NO: 13 and a VH domain of SEQ ID NO: 14.

3. The antibody according to claim 1, which is a human/mouse chimeric antibody.

4. The antibody according to claim 1, which is an IgA.

5. The antibody according to claim 1, which is a polymeric antibody.

6. The antibody according to claim 1, which is a secretory antibody.

7. The antibody according to claim 1, which is a polymeric or secretory IgA antibody.

8. The antibody according to claim 1, which is coupled to a labeling agent.

9. A pharmaceutical composition comprising at least an antibody according to claim 1 and a pharmaceutically acceptable vehicle.

10. A method of treating cancer overexpressing CEA, comprising administering to an individual a therapeutically effective amount of the antibody according to claim 1.

11. An in vitro method of diagnostic of cancer overexpressing CEA, comprising detecting CEA expression in a tissue sample from an individual using the antibody according to claim 1.

12. The method according to claim 10, wherein said cancer is a mucosal epithelium cancer.

13. The method according to claim 10, wherein said cancer is selected from the group consisting of: colorectal, gastric, thyroid, lung, breast, pancreas, gallbladder, urinary bladder, ovary and endometrium cancers.

14. The method according to claim 11, wherein said cancer is a mucosal epithelium cancer.

15. The method according to claim 11, wherein said cancer is selected from the group consisting of: colorectal, gastric, thyroid, lung, breast, pancreas, gallbladder, urinary bladder, ovary and endometrium cancers.

Description

(1) In addition to the above arrangements, the invention also comprises other arrangements, which will emerge from the description which follows, which refers to exemplary embodiments of the subject of the present invention, with reference to the attached drawings in which:

(2) FIG. 1: Western-Blot analysis of IgA monoclonal antibody anti-CEA (clone #15B3) using anti-human alpha-heavy chain antibody as a probe.

(3) FIG. 2: In vitro direct cancer cell growth inhibition by anti-CEA monoclonal antibody in human colorectal cancer target cells (WiDr− CEA.sup.+).

(4) A. Inhibition of cancer cell proliferation by increasing concentrations (6.25, 12.5 and 25 μg/ml) of culture supernatant of IgA anti-CEA positive clones #15B3 and #14G8, compared with a negative clone (#17C7). B. Direct induction of apoptosis in target cells by increasing concentrations (5, 20 and 40 μg/ml) of culture supernatant of IgA anti-CEA clone #15B3, compared with IgG anti-CEA clone #15B3 and irrelevant IgA anti-peanut (IgA #15F11) culture supernatant. *p value≤0.05; **p value≤0.01

(5) FIG. 3: In vitro complement-dependent cytotoxicity (CDC) induction by IgA monoclonal antibody anti-CEA in human colorectal cancer target cells (WiDr− CEA.sup.+).

(6) Inhibition of cell proliferation in target cells was assayed with increasing concentrations (6.25, 12.5 and 25 μg/ml) of culture supernatant of IgA anti-CEA positive clones #15B3 and #14G8 and negative clone #17C7, in the presence (IgA mab(#15B3)-CDC; IgA mab(#14G8)-CDC; IgA mab(#17C7)-CDC) or absence (IgA mab(#15B3); IgA mab(#14G8); IgA mab(#14G8) of complement. *p value≤0.05; **p value≤0.01.

(7) FIG. 4: In vivo ADCC induction by anti-CEA IgA antibody in human colorectal cancer target cells (WiDr− CEA.sup.+). Target cells were injected into the peritoneal cavity of transgenic mice expressing human CD89 (Fc-alpha-Receptor) in the presence of purified IgG or purified polymeric IgA anti-CEA or irrelevant IgA anti-peanut (n=4 mice per group). A. Target cell count. B. Percentage of doublet target cells (CEA.sup.+)/Effector cells (CD89.sup.+) in total target cells. *p value≤0.05.

(8) FIG. 5: Biodistribution of .sup.99mTc-anti-CEA IgA monomeric-SH, .sup.99mTc-anti-CEA IgA polymeric-SH and .sup.99mTc-anti-CD20 IgG-SH in healthy Balb/c mice at 4, 8, 18, 24 and 48 h, expressed as the percentage of injected (intravenous; IV) dose per gram, % ID/g (values represent means±SD of the % ID/g). *p value≤0.05.

(9) FIG. 6: Biodistribution of .sup.99mTc-anti-CEA IgA polymeric-SH and irrelevant .sup.99mTc-anti-PEANUT (anti-ARA) in nude mice bearing intracaecal tumours, at 8 h, expressed as the percentage of the injected dose per gram, % ID/g (values represent means±SD of the % ID/g). ***p value≤0.001.

(10) FIG. 7: Colorectal tumor growth inhibition by anti-CEA IgA treatment in orthotopic mouse model of human colorectal cancer. 8 days after tumor cell implantation in nude mice, polymeric IgA anti-CEA (IgA anti-CEA) and irrelevant dimeric IgA anti-peanut (IgA anti-ARA) were administered by the intravenous route (0.2 or 0.3 mg/injection for 5 consecutive days) and IgA anti-CEA (IgA anti-CEA) was also administered by the intravenous (1 mg) and intraperitoneal (1 mg) routes, whereas secretory IgA anti-CEA and irrelevant secretory IgA anti-peanut were administered by the oral route (0.135 mg/day for 11 consecutive days) or oral (0.6 mg) and intraperitoneal (1 mg) routes. 10 weeks after tumor cell implantation, caecum weight of IgA anti-CEA treated mice was compared with that of irrelevant IgA treated controls.

(11) FIG. 8: Colorectal tumor growth inhibition by anti-CEA IgA or anti-EGFR IgG (cetuximab) treatment in orthotopic mouse model of human colorectal cancer. Seven weeks after tumor cell implantation, polymeric IgA anti-CEA (IgA) and anti-EGFR IgG (cetuximab) were administered by the intravenous route (4 mg for each antibody). Ten weeks after tumor cell implantation, caecum weight of IgA anti-CEA and IgG/cetuximab treated mice was compared with that of untreated controls.

EXAMPLE 1: MATERIAL AND METHODS

(12) —Immunization

(13) Immunization was performed in HAMIGA™ transgenic mice (EP Patent 1 680 449 B1), a transgenic mouse strain producing human/mouse chimeric IgAs consisting of human IgA heavy chain constant regions and mouse light chain constant regions and (heavy and light chain) mouse variable regions. HAMIGA™ transgenic mice (4 mice) were immunized by intraperitoneal route twice at two weeks interval with human recombinant CEA (Eurobio-Abcys) or irrelevant antigen (peanut), in Freund adjuvant (10 μg/mouse/injection, ratio 1:1 CFA (SIGMA) and 10 μg/mouse/injection, ratio 1:1 IFA (SIGMA).

(14) —Preparation of Monoclonal IgA

(15) Human chimeric monoclonal IgA (IgA1) against CEA or irrelevant antigen (peanut) were prepared by immortalization of B-cell lymphocytes from CEA-immunized HAMIGA™ mice, according to standard protocol (Kohler, G. & Milstein, C., European Journal of Immunology, 1976, 6, 511-9). Briefly, all splenocytes from CEA-immunized mice were harvested and pelleted before being fused with mice myeloma cells (P3X63 Sp2/0:AG14; ATCC CRL-1581; cellular ration 1 SP2/0 for 3 splenocytes). The hybridomas were subcloned on 96-wells plate. After three weeks of culture, supernatants of hybridoma clones were harvested and tested for their biological activity or antigen affinity. Each clone was cryopreserved in DMSO 10%/SVF 20%/DMEM media in liquid nitrogen.

(16) —Ig cloning and sequencing

(17) Ig cloning and sequencing was performed using standard cloning and sequencing techniques.

(18) —Preparation of Recombinant IgG1

(19) Recombinant human-mouse chimeric IgG1 anti-CEA was synthesised after cloning the mouse variable regions of the heavy and light chains of anti-CEA human-mouse chimeric IgA1. It was then produced in human embryonic kidney cells (HEK 293).

(20) —Ig radiolabelling with [.sup.99mTc(CO).sub.3(H.sub.2O).sub.3].sup.+

(21) IgA radiolabelling method was adapted from the radiolabelling method previously described for IgG (Carpenet et al., PLoS One, 2015 Oct. 6; 10(10):e0139835). Briefly, the first step was thiol-derivatisation of Ig with 2-iminothiolane. Next, 0.5 to 2.2 nmol IgA and IgG (300 μL in PBS) were incubated with 2-IT (3.8 μM, 25° C., 120 min). The solutions were purified by size exclusion chromatography. The number of thiol groups was determined by a micromethod using Ellman's reagent (5.5′-dithiobis-2-nitrobenzoic acid, DTNB). The second step was synthesis of the tricarbonyl precursor [.sup.99mTc(CO).sub.3(H.sub.2O).sub.3].sup.+. Next, 0.8-1 mL of freshly eluted [Na.sup.99mTcO.sub.4] (CisBio, Codolet, France) in fixed activities (2,220-3,700 MBq) was added to the IsoLink® kit (Covidien, Petten, The Netherlands) and incubated for 25 min at 100° C. Radiochemical purity (RCP) analysis was performed by thin-layer chromatography (TLC) using two systems to separate the [.sup.99mTc(CO).sub.3(H.sub.2O).sub.3].sup.+ from free [.sup.99mTc]-pertechnetate, reduced .sup.99mTc and hydrolysed [.sup.99mTc(OH).sub.n(H.sub.2O).sub.y] (Baker-flex aluminium, MeOH/HCl (95/5 v/v); instant thin layer chromatography-silica gel (ITLC-SG), MeOH; JT Baker Inc., Phillipsburg, N.J., USA). The .sup.99mTc-Isolink® labelling yields were superior to 98%. The third and last step was the radiolabelling of native or derivatised Ig with [.sup.99mTc(CO).sub.3(H.sub.2O).sub.3].sup.+. A total of 0.5-2.2 nmol of non-derivatised IgA or derivatised IgA-SH, or 1.5 nmol IgG-SH in 300 μL of PBS, was incubated for 120 min (25° C.) with 150 μL (148-185 MBq) of a .sup.99mTc-tricarbonyl solution, previously neutralised to pH 7.0 (0.5 M HCl). RCP was determined by TLC with ITLC-SG/NaCl 0.9%.

(22) —Ig Purification

(23) The antibodies were purified by affinity chromatography using a Tricorn Column 5/100 with protein A-Sepharose at a flow rate of 1.0 mL/min (GE Healthcare, Waukesha, Wis., USA) and were eluted with glycine (0.1 M pH 2.7) equilibrated in Tris/base (1.0 M). Subsequently, IgA and IgG were dialysed against phosphate-buffered saline (PBS) by centrifugation (1,000×g, 15 min) using Amicon 30 kDa (Millipore, Saint-Quentin, France). The protein concentrations were determined before and after radiolabelling using Micro Bicinchoninic Acid (BCA™) Protein Assay kit (ThermoFisher Scientific, Elancourt, France), using bovine serum albumin (BSA) as a standard with quantification limits of 2.5 and 100 μg/mL.

(24) —Purification of Monomeric and Polymeric IgA

(25) Total IgA were purified using CAPTURESELECT™ IgA Affinity Matrix (ThermoFisher Scientific), according to manufacturer's instructions. Monomeric and polymeric forms of IgA were then separated by size exclusion column chromatography using HILOAD™ 26/600 SUPERDEX™ 200 pg (GE Healthcare), according to manufacturer's instructions. An enriched fraction of the monomeric (mIgA) or polymeric (pIgA) form was obtained (purities of 95% and 85%, respectively).

(26) —Secretory IgA Preparation

(27) Anti-CEA secretory IgA was produced by in vitro covalent bond with recombinant human secretory component (hSC), as previously described (Rindisbacher et al., JBC, 1995, 270, 14220-14228; Koteswara et al., PNAS, 1997, 94, 6364-6368). Briefly, human secretory component cDNA was amplified by PCR from ileum RNA preparation and inserted into mammalian expressing vector (pCDNA.3, Invitrogen). Recombinant hSC was expressed in HEK-293 cells and purified by affinity chromatography. In vitro covalent assembly of hSC and dimeric IgA was performed by incubating hSC and dimeric IgA during 1 h at 37° C. (at a protein mass ration 1:1).

(28) —Antigen Specificity Analysis

(29) CEA specific IgA were assessed by ELISA using Maxisorp® 96-wells plates (NUNC) coated with 1 to 5 μg/mL of antigens overnight at 4° C. Crude supernatants of unpurified IgA (diluted in PBS/Gelatin 0.2%) were incubated 2 hours at 37° C. Specific IgA binding was revealed with an AP-conjugated goat anti-human IgA antibody (1/2000 diluted, Beckman Coulter).

(30) —Purified Monomeric and Polymeric IgA and Secretory IgA Concentration Titration by ELISA

(31) Purified IgA (purified by affinity chromatography) and secretory IgA were titered by ELISA. Briefly, 96-well plates (NUNC, Maxisorp®) were coated with 1 μg/mL of goat anti-human IgA (Beckman Coulter) in PBS buffer at 4° C. overnight. Wells were saturated in PBS buffer containing 2% BSA during 30 minutes at 37° C. Incubation of the samples (secretory IgA, diluted 10 times in PBS containing 0.2% BSA; purified IgA, diluted 100 times in PBS containing 0.2% BSA) was performed at 37° C. during 2 h. Human IgA calibrator range (from [control hIgA]=0.2 mg/ml to 1.56 ng/mL) was incubated following the same protocol and revealed by an Alkaline-Phosphatase (AP) labelled goat anti-hIgA polyclonal antibody (Beckman Coulter; diluted 2000 times in PBS containing 0.2% BSA).

(32) —Cell Culture

(33) WiDr, a human colorectal cell line expressing CEA, derived from a primary adenocarcinoma of the rectosigmoid, was purchased from ATCC (VA, USA) (WiDr ATCC® CCL-218™). HT-29, human cell line from colorectal adenocarcinoma was purchased from ATCC (HTB-38™). The cell lines were grown in DMEM or in RPMI, supplemented with 10% fetal calf serum (FCS), 1% sodium pyruvate, 1% (100 U/ml) penicillin-streptomycin (100 μg/ml). Medium was also supplemented with 1% of nonessential amino acids and 1% of glutamine.

(34) —In Vitro Cellular Growth Inhibition

(35) Cells were harvested when reached the log phase growth stage. Cells were then plated at 50,000 cells/well in 1004 and incubated overnight. Culture media was removed and replaced by culture media (DMEM/FCS10%) containing various concentrations of the antibodies (anti-CEA IgA, irrelevant anti-Peanut IgA, recombinant IgG anti-CEA). After 48 h of incubation, ALAMARBLUE™ was added aseptically in an amount equal to 10% of the culture volume. Cultures were replaced to incubator. At various time, fluorescence/absorbance was measured. Absorbance was measured at a wavelength of 600 nm. To evaluate the impact of fresh human complement elements on cellular growth inhibition, culture media was removed and replaced by media (DMEM) containing 10% of fresh human sera with various concentrations of antibodies (anti-CEA IgA, anti-Peanut IgA, recombinant IgG anti-CEA).

(36) —Apoptosis Assay

(37) Apoptosis assay was performed using PE Annexin V Apoptosis Detection kit I, according to manufacturer's instructions (BD PHARMINGEN™). Briefly, cells harvested when reached the log phase growth stage were plated at 50,000 cells/well in 1004 and incubated overnight. Culture media was removed and replaced by culture media (DMEM/SVF10%) containing various concentrations of the antibodies (anti-CEA IgA, irrelevant anti-Peanut IgA, recombinant IgG anti-CEA). After 48 h of incubation, cells were harvested, washed several times in PBS and diluted at 10.sup.6 cells/mL in Binding Buffer (BD PHARMINGEN™). Cells are incubated with PE-conjugated Annexin V (BD PHARMINGEN™) during 15 minutes at room temperature. Cells were washed in Binding Buffer and resuspended in 1× Binding Buffer. 7AAD Viability Staining Solution (BD PHARMINGEN™) was added just prior analyzing by flow cytometry.

(38) —In Vivo Antibody-Dependent Cellular Cytotoxicity (ADCC)

(39) A transgenic SCID-CD89 mouse model (provided by Pr Jeannette Leusen, Utrecht University, Netherland), in which neutrophils and monocytes express the human CD89 has been used. 10.sup.7 cells were incubated at 10.sup.6 cells/mL with polymeric IgA anti-CEA, polymeric IgA anti-Peanut or recombinant IgG anti-CEA (at 20 μg/mL) during 1 h at 4° C. Target cells (CEA.sup.positive-WiDr adenocarcinoma) were pelleted, resuspended in PBS and injected into the peritoneal cavity. 16 to 18 h post-inoculation, cells were harvested by peritoneal cold washing and cell populations were analyzed by flow cytometry (FACS) using FITC-anti-CD89 antibody (BioLegend) staining. Effector cells were identified by selective expression of CD89.sup.positive on cell surface and cancer cells were gated by their characteristics of cellular size and structure.

(40) —Orthotopic Mouse Model of Human Colorectal Cancer (CRC)

(41) All in vivo experiments were performed in accordance with animal ethical regulations and all efforts were made to minimize suffering. Direct orthotopic cell microinjection (OCMI) was performed according to Cespedes method (Cespedes et al., Am. J. Pathol., 2007, 170, 1077-1085). Briefly, seven-week-old female nude mice (athymic nude, HARLAN Laboratories) or transgenic SCID-CD89 mice (provided by Pr Jeannette Leusen, Utrecht University, Netherland) were anesthetized with ketamine (80 mg/kg; Imalgene (100 mg/ml), MERIAL) and xylazine (9.6 μg/kg; 2% Rompun, BAYER) to exteriorize their caecum by a laparotomy. WiDr cells (2.10.sup.5 cells suspended in 10 μl of PBS in a sterile micropipette) were slowly injected between the mucosa and the muscularis externa layers of the caecum wall, under a binocular lens, with an approximate 30° angle. After injection, the caecum was returned to the abdominal cavity. Animals were treated by veterinary antibiotics to prevent intraperitoneal infection. Peritoneal cavity was closed by surgical laparotomy. If animals demonstrated clinical alteration or weight loss, animals were euthanized by anesthesia and cervical dislocation.

(42) —Biodistribution of .sup.99mTc-Anti-CEA IgA-SH and .sup.99mTc-Anti-CEA IgG-SH in Normal Mice and Human CRC Mouse Model

(43) Biodistribution experiments were carried out in 7-week-old male BALB/c mice (Charles River Laboratories, chalaronne, the L'ARBRESLE Cedex) or xenografted nude mice. .sup.99mTc-IgA-SH monomeric or .sup.99mTc-IgA-SH polymeric or 99mTc-IgG-SH (40 MBq, 170 μg of antibody) was injected intravenously (tail vein) to BALB/c mice. 6 to 8 weeks after OCMI procedure, xenografted nude mice were divided into 2 groups. The first group (n=6) received intravenously 170 μg .sup.99mTc-anti-CEA pIgA-SH and the second (n=6) 170 μg of .sup.99mTc-anti-PEANUT pIgA-SH (All mice received an activity of 35-37 MBq). Nude mouse controls received the same .sup.99mTc-anti-CEA pIgA-SH. Animals were euthanized by anesthesia and cervical dislocation, at different times after administration (4 h, 8 h, 18 h, 24 h, 48 h post-injection for BALB/c mice; 4 h and 8 h for xenografted nude mice). Selected tissues were excised, rinsed, and weighed, and their radioactivity levels were measured with a gamma-counter. The uptake of radioactivity in these organs was expressed as a percentage of the injected dose per gram of tissue (% ID/g) after correcting for radioactive decay for each time point. Blood cells, plasma, and feces were also collected and measured. Faeces refers to faecal matter collected in the small intestine and colon during dissection. Furthermore, the caecum was longitudinally opened washed with PBS and countered separately from caecal feces to evaluate luminescent IgA. The caecum and lungs of xenografted nude mice were fixed with buffered formalin during radioactive decay (48 h).

(44) —Histological Analysis of Human Colorectal Orthotopic Grafts

(45) Mice organs were transferred to 4% formalin and include in paraffin after an automated cycling of dehydration system. 4 μm sections were prepared using microtome. For histological analysis, slides were stained with hematoxylin eosin and saffron (HES analysis), with alcian blue (secreting mucus analysis). For vascularization analysis, CD31 staining was made using VENTANA robot.

(46) —In Vivo Inhibition of Tumor Growth

(47) Experiments were performed in the orthotopic mouse model of human colorectal cancer (CRC) described above. The IgA anti-CEA antibody and the irrelevant IgA anti-peanut antibody were administered 8 days after implantation of human tumor cells in the caecum of Balb/c Nude mice (n=12 per group). Two routes of administration have been evaluated: the intravenous (by administration of 0.2 or 0.3 mg/injection for 5 consecutive days) for the polymeric IgA and the oral (by administration of 0.135 mg/day for 11 consecutive days) for the secretory IgA. 8 weeks after the end of the treatment, animals were euthanized by anesthesia and cervical dislocation and the caecum were excised, emptied, rinsed and weighed.

EXAMPLE 2: IDENTIFICATION OF ANTI-CEA MONOCLONAL ANTIBODIES HAVING DIRECT GROWTH-INHIBITORY ACTIVITY ON CEA EXPRESSING TUMOR CELLS

(48) IgA monoclonal antibodies (Mabs) anti-CEA were generated by immunization of HAMIGA™ transgenic mice, a transgenic mouse line producing human/mouse chimeric antibodies having human IgA heavy chain constant regions and mouse light chain constant regions and (heavy and light chain) mouse variable regions. Hybridomas were produced after fusion of B cells of the immunized HAMIGA™ mice with the Sp2/0 mouse myeloma cell line. As the Sp2/0 line is derived from murine myeloma expressing murine J chain necessary for the dimerization of the IgA, the hybridomas which were obtained produce distinct forms of IgA MAbs anti-CEA: a monomeric form (without J chain) and a polymeric form which contain J chain(s) and includes dimeric form and higher polymeric forms of IgA (FIG. 1). The level of expression of the various forms of IgA depends on the selected hybridoma. The different forms can be purified and separated by chromatography.

(49) Hybridoma clones were selected for direct cancer cell growth-inhibitory activity on various CEA expressing human tumor cell lines using ALAMARBLUE™ assay.

(50) Anti-CEA IgA of clone 15B3 (also named #15B3) and clone 14G8 also named #14G8) block cell growth of two cell lines derived from human colorectal adenocarcinoma (WiDr, 18.2%±1.9% for 15B3 and 18.1±1.2 for 14G8 at 25 μg/mL (culture supernatant); FIG. 2). These positive clones were selected for further analysis. Variable regions from IgA heavy and light chains (VH and VL) were cloned and sequenced. The VL and VH amino acid sequences correspond to SEQ ID NO: 11 and 12 and SEQ ID NO: 13 and 14, respectively for 15B3 and 14G8. The nucleotide sequences encoding said amino acid sequences correspond to SEQ ID NO: 15 and 16 and SEQ ID NO: 17 and 18, respectively for 15B3 and 14G8.

(51) A recombinant anti-CEA IgG1 derived from clone 15B3 was constructed using the cloned VH and VL domains and expressed in Sp2/0 or HEK cell line. RecIgG1 anti-CEA #15B3 shows a counterpart level of growth inhibition in comparison to IgA anti-CEA (clone #15B3). The affinity of the antibody (mainly carried by the variable regions forming the “paratope”) is preserved during the process of transformation of an IgA1 towards an IgG1.

(52) Increasing doses of anti-CEA IgA (5, 20 and 40 μg/mL, clone #15B3) induces early mechanisms of apoptosis (fixation of Annexin V, permeabilization of the membrane visualized by labelling by 7AAD+(FIG. 2B, significantly higher than for other conditions (recIgG recombinant #15B3, irrelevant IgA1 (anti-peanut #15F11)). The higher valence of a polymeric IgA would allow a potential cross-linking of CEA on the surface of target cells WiDr, inducing an unexpected biological effect (direct apoptosis) stronger and faster than a monomeric form of IgA, as in the case of recIgG1 #15B3.

(53) IgA recruits elements of the complement leading to lysis of the cell target via the alternate pathway. Same tests of cell growth inhibition of WiDr cancer cells were performed in the presence of human sera (human sera alone did not affect biologically WiDr cell culture). A significant rise of growth inhibition was recorded in the presence of human complement for the two selected anti-CEA IgA clones (clone #15B3 and clone #14G8; FIG. 3).

(54) In humans, IgA recruits also immune effector cells that express a high affinity IgA receptor (Fc-alphaR or CD89), mainly neutrophils and monocytes, two particularly numerous cells in blood cell population. To demonstrate the ability of effector cells recruitment mediated by the anti-CEA IgA #15B3, a transgenic mouse model in which neutrophils and monocytes express the human CD89 has been used. Flow cytometry (FACS) analysis shows the presence of a new population of doubly selected cells (Effector Cell (CD89.sup.+): WiDr target cells (CEA.sup.+) in mice treated with the anti-CEA IgA (31%±15%) compared with mice treated with irrelevant IgA (anti-peanut IgA; 10.2%±3.5%, FIG. 4). These results demonstrate that the anti-CEA IgA recognizes in vivo CEA on the surface of WiDr cancer cells and induces the recruitment of cells (monocytes/macrophages and polynuclear cells) via the CD89 receptor. In the presence of anti-CEA IgA (#15B3), a very powerful cytotoxic effect eliminates more than 80% of the cancer cell population in 16 h to 18 h (FIG. 4). Compared to the WiDr population harvested in mice treated with the irrelevant IgA, only 16.4%±3.6% of WiDr target cells are harvested after treatment with the specific anti-CEA IgA (#15B3).

(55) Cell types recruited preferentially by IgG and IgA are different: it is known that the IgG-mediated ADCC is induced by the NK cells, strongly expressing the FcgammaR. The recombinant IgG generated from the variable regions of the IgA anti-CEA clone (#15B3) induces a cytotoxic cell lysis of WiDr-CEA.sup.+ as strong as the “original” IgA as only 11.3%±8.6% of WiDr were harvested by peritoneal washing after treatment with recIgG anti-CEA (FIG. 4).

(56) The similar cytotoxic effect of the two antibodies supports the demonstration that the IgA is able to induce the ADCC mechanisms as quickly and effectively as the IgG. It also validates that the transformation of the IgA towards IgG does dot modify the affinity of the antibody for its target.

(57) In addition, immunofluorescence analysis on tissues (liver and stomach, data not shown) and flow cytometry analysis on leukocytes (data not shown) demonstrated that the anti-CEA antibody #15B3 does not cross-react with CEACAM-6 (NCA or CD66c).

(58) In conclusion, the anti-CEA antibodies according to the invention have a direct cell growth inhibition effect on cancer cells expressing CEA. This unique antitumor effect is further enhanced by their ability to recruit the complement pathway and the immune cell-effectors of antibody dependent cell cytotoxicity (ADCC) leading to cancer cell lysis. All these antitumor effects are specific for the targeted tumor cells because the antibodies are specific for CEA.

EXAMPLE 3: BIODISTRIBUTION OF ANTI-CEA ANTIBODY IN NORMAL MICE AND ORTHOTOPIC MOUSE MODEL OF HUMAN COLORECTAL CARCINOMA

(59) Bio-distribution and pharmaco-kinetic studies were performed using different forms of Ig, IgA (monomeric (mIgA) and polymeric (pIgA)) and IgG, all labelled with technetium 99m (.sup.99mTc). Very quickly, from 8 h post-injection (by i.v.) to normal Balb/c mice, monomeric IgA concentration dropped in sera and it's even more marked for the polymeric IgA to the detectable limit (FIG. 5). Conversely, the concentration of the polymeric IgA (and monomer, to a lesser extent) is quantifiable as early as 4 h post-iv injection, in the caecum, the lung and the liver (FIG. 5). Hepatic-biliary cycle of the IgA is particularly patent here, the circulating IgA is quickly captured by the liver, directed to the gallbladder to be excreted with bile into the lumen of the digestive tract. This is what explains the strong labelling of the liver on the one hand and of the intestinal fluid and stools on the other hand. As early as 4 h post-injection, the secretory IgA is detectable in the lumen of the mucosal organs. In this healthy murine model, the polymeric IgA tropism for the caecum is the strongest and the fastest, superior to the tropism of the monomeric IgA, but also that of the IgG, already described to persist longer in blood.

(60) In conclusion, bio-distribution studies of .sup.99mTc-anti-CEA IgA monomeric-SH and .sup.99mTc-anti-CEA IgA polymeric-SH in normal Balb/c mice confirmed rapid and strong mucosal tropism of pIgA and, to a lesser extent, mIgA.

(61) To evaluate IgA targeting potency, a CRC tumour model was created in which human cancer cells were grafted in the mucosal environment. Pathological microscopic analysis clearly revealed a structural glandular architecture of the grafted tumour and the presence of large vacuoles in the WiDr cell line, consistent with muco-secretions in the lamina propria layer. Cancer cells invaded the normal caecum, under the muscle layer through the lamina propria, to produce protruding polyps in the lumen. Depending on the delay after direct orthopic microinjection, different stages of CRC have been observed from localised tumours to metastasis in the lungs. Immunohistochemical analysis in tumours revealed that tumour cells were present within tumour vessels, suggesting cellular dissemination by the vascular system. All of these factors led to consider colorectal orthotopic grafts as being useful models of human CRC, because they share the same characteristics as human tumours.

(62) After 8 weeks, tumor cells were spread and identified by immunohistochemistry, as pulmonary metastases (no trace of hepatic colonization is detected in this model). Invasion of the lung capillaries and the presence of cancerous nodules in the lung tissue were particularly identifiable by HES staining. .sup.99mTc-anti-CEA IgA polymeric-SH or irrelevant .sup.99mTc-anti-PEANUT pIgA-SH were then administered intravenously. Very quickly (already at 4 h postinjection), the radioactive tracer is detected in the lung (FIG. 6). The anti-CEA IgA is able to recognize tumor cells very quickly and with a very strong affinity while metastases are small (observable only on immunohistochemistry slides, no macroscopic tumor is visible). The administration of an irrelevant IgA labelled with .sup.99mTc failed to detect any metastasis foci when administered to animals with lung metastases at the same stage of development as animals treated with the specific anti-CEA IgA. These studies confirm the fast mucosal tropism of the IgA antibody (8 h post-injection) and strong targeting power of early metastatic foci in vivo (in the lung).

EXAMPLE 4: ANTITUMOR EFFECT OF ANTI-CEA ANTIBODY IN ORTHOTOPIC MOUSE MODEL OF HUMAN COLORECTAL CARCINOMA

(63) The antitumor effect of the anti-CEA antibody (#15B3) was evaluated in the orthotopic mouse model of human colorectal carcinoma disclosed in example 1. Preliminary results showed a significant benefit of treatment with the polymeric form of anti-CEA IgA on the survival of the animals and the decrease in tumor growth.

(64) In a first experiment, the treatment was administered 8 days after the implantation of human tumor cells in the caecum of mouse (Balb/c Nude, Harlan) for all conditions. Two routes of administration have been evaluated: the intravenous (by administration of a dose/day for 5 days, 8 days after implantation of the tumor) for the polymeric IgA and the oral (by administration of a dose/day for 11 days, 8 days after tumor implantation) for the secretory IgA. When animals (n=12) are treated with a cumulative dose of 1 mg (IV, 0.2 mg/injection for 5 consecutive days), the reduction of the tumor mass is significant compared with the group treated with a cumulative dose of 1 mg (IV, 0.2 mg/injection for 5 consecutive days) of irrelevant polymeric anti-peanut IgA (**, p=0.001). The combination of intravenous administration of 1 mg and intraperitoneal administration of 1 mg of polymeric anti-CEA IgA prevents tumor growth as effectively as an only intravenous administration of 1 mg of anti-CEA IgA (FIG. 7). The reduced tumor mass is even more marked by intravenous administration a higher cumulative dose (1.5 mg) of polymeric anti-CEA IgA (0.3 mg/injection 5 days; ***, p=0.0002) compared to the group treated with an irrelevant polymeric anti-peanut IgA (FIG. 7). Secretory IgA is the unique isotype able to be administered orally without undergoing enzymatic degradation tumor growth was significantly reduced by oral administration of 1.5 mg of anti-CEA secretory IgA (0.135 mg/administration 11 days, 8 days after implantation) compared to oral administration of an irrelevant secretory IgA (sIgA anti-peanut) (FIG. 6; **, p=0.004).

(65) In a second experiment, seven weeks after the implantation of human tumor cells in the caecum of mouse (transgenic SCID-CD89), polymeric IgA anti-CEA (4 mg) or IgG anti-EGFR (cetuximab; 4 mg) was administered intravenously to engrafted mice (n=10 per group).

(66) The cecal tumor mass was significantly reduced (by about 30%) in anti-CEA IgA treated mice compared to cetuximab treated-mice (FIG. 8; **, p=0.0032).

(67) These results show the effectiveness of the anti-CEA IgA antibodies according to the invention in preventing growth of the primary tumor located in the intestinal mucosal environment. The Anti-CEA IgA demonstrates for the first time a greater therapeutic benefit than the Gold standard treatment for advanced colorectal cancer immunotherapy, the cetuximab-ERBITUX®/anti-EGFr IgG. In addition, high lung uptake of .sup.99mTc-anti-CEA pIgA-SH in the mouse tumour model (example 3) suggested efficient targeting potency of pIgA. Due to intrinsic mucosal tropism and biomarker affinity, polymeric IgA could reach its targets very effectively. This work clearly demonstrated the potential of the anti-CEA IgA antibodies according to the invention for the diagnosis and treatment of mucosal tumors.