The present invention relates to antibodies binding with high affinity to Carcinoembryonic antigen (CEA) and CEACAM1 (CEA Cell Adhesion Molecule 1 ).

Background

Conventional pharmacological approaches for the therapy of tumours suffer from poor selectivity, thus compromising dose escalation to therapeutically active levels. Therefore, the development of selective, and better tolerated, cancer therapeutics represents an important goal in the research of new and more effective treatment. One way to improve the selectivity of therapeutic molecules is to target the tumour site, thereby sparing normal tissues. Antibody-based cancer treatments offer a solution to this problem and have given promising results in several malignancies (Goydel RS, Rader C Antibody-based cancer therapy. Oncogene 2021 , 40:3655-3664).

Carcinoembryonic antigen is an attractive target for immunotherapeutic purposes because of its expression profile, its role in tumour progression, and its immunogenicity. CEA belongs to the CD66 immunoglobulin superfamily that also comprises CEACAM1 . CEACAM1 was found to be involved in progression and metastatic potential of melanoma and lung cancer. CEACAM1 re-expression often occurs in the advanced stages of many malignancies such as bladder cancer, thyroid cancer, gastric carcinoma, pancreatic cancer, metastatic colon cancer and multiple myeloma. (Fiori V, et al. Ann 1st Super Sanita 2012, 48:161 -71 ). The involvement of CEACAM1 in cancer progression can be partially explained by its central role in the angiogenic switch of tumours. CEACAM1 is involved in the switch from non-invasive and non- vascularised to invasive and vascularised bladder cancer (Oliveira-Ferrer L, et al. Dual role of carcinoembryonic antigen-related cell adhesion molecule 1 in angiogenesis and invasion of human urinary bladder cancer. Cancer Res. 2004, 15:8932-8). In addition, CEACAM1 is emerging as a novel immune checkpoint target involved in the inhibition of antitumor immune responses (Dankner M, et al. CEACAM1 as a multipurpose target for cancer immunotherapy. Oncoimmunology 2017, 6:7). WO201 1160859A1 describes a single chain variable fragment specific for each of carcinoembryonic antigen (CEA) and CEA related Cell Adhesion Molecule 1 (CEACAM1 ), scFvDIATHIS-1 .

EP3909601 describes anti CEACAMs antibodies, demonstrating same properties in term of dose-response curve, kinetics, and Kd determination when the CDRs were in mouse monoclonal IgG antibodies or in human lgG1 , lgG2a or lgG4 antibodies.

An important need persists for a mean to target in a selective and efficient manner CEA and CEACAM1 to gain the full benefit of these therapeutic agents.

Description

Description of the figures

Figure 1 : Creation of lgG1 and lgG4 anti-CEACAM1 mAbs stable producing clones in HEK293 cells. (A) lgG1 pool selection by ELISA assay, black column: selected pool; (B) lgG1 clone selection by ELISA assay, black column: selected clone; (C) lgG1 subclone selection by ELISA assay, black column: selected subclone; (D) lgG4pool selection by ELISA assay, black column: selected pool; (E) lgG4 clone selection by ELISA assay, black column: selected clone.

Figure 2: SEC-HPLC analysis of (A) lgG1 , clone 12 and (B) lgG4, clone 2 anti-CEACAM1 mAbs.

Figure 3: SDS-PAGE of lgG1 anti-CEACAM1 mAb clone 12.3.

Figure 4: ELISA titration assays showing CEACAM1 recognition by lgG1 (black) and lgG4 (grey) purified after a transient transfection in HEK293T cells.

Figure 5: ELISA titration assays showing CEACAM1 recognition by (A) lgG1 clone 12 (black line, according to the invention) and lgG4 clone 2 (grey line, comparative); (B) scFvDIATHIS-1 (comparative).

Figure 6: lgG1 and lgG4 reactivity on CEACAM1 antigen expressed on metastatic melanoma cells by Flow cytometry. MelC cells were reacted with 50, 25 or 12 pg/ml of lgG1 or lgG4. Negative control (CTR-) is represented by MelC cells reacted with secondary antibodies only. (A) Cytograms display the results obtained with the indicated antibodies. (B) Mean Fluorescence values representing the affinity of the binding

Figure 7: lgG1 and lgG4 reactivity on CEACAM1 antigen expressed on metastatic melanoma cells by Flow cytometry. MelP5 cells were reacted with 50, 25 or 12 pg/ml of lgG1 or lgG4 anti-CEACAM1 mAb. Negative control is represented by MelP5 cells reacted with secondary antibodies only (CTR-). (A) Cytograms display the results obtained with the indicated antibodies. (B) Mean Fluorescence values representing the affinity of the binding

Figure 8: Analysis of lgG1 clone 12, lgG4 clone 2 (comparative), scFv (comparative) reactivity by Flow cytometry, mean fluorescence values, on CEACAM1 and CEA antigens expressed on (A) bladder cancer cells; (B) colorectal cancer cells.

Figure 9: Analysis of lgG1 purified from the stable clone 12.3, reactivity on CEACAM1 and CEA antigens expressed on metastatic bladder cancer cells or colorectal cancer cells by Flow cytometry. Cells were reacted with 50, 20, 10 or 5 pg/ml of DIA-12.3 lgG1. Negative control represents cells reacted with secondary antibodies only. (A) Cytograms display the results obtained with the indicated antibody concentrations. (B) Mean Fluorescence values, representing the affinity of the binding.

Figure 10: ELISA titration assays showing CEACAM1 recognition by lgG1 (grey) and lgG1 N297A mutated (black).

Figure 11: LDH assay (A) on MelC target cells co-incubated with IL-2 activated NK-92 effector cells at the indicated ratio, after incubation with DIA-12.3 Ab at the indicated concentration. (B) on MelC, TCCSLIP, 5637, HT29, HN target cells, respectively, co-incubated with IL-2 activated NK- 92 effector cells at E:T ratio of 10:1 , after incubation with 20 pg/ml DIA- 12.3 Ab.

Figure 12: DIA-12.3 lgG1 reactivity on CEACAM1 antigen expressed on neutrophils by Flow cytometry. Neutrophils were reacted with 10 pg/ml of DIA-12.3 lgG1 mAb. Negative control is represented by cells reacted with secondary antibodies only (Anti-human Fc antibody). Cytograms display the results obtained with the indicated antibodies.

Figure 13: FITC Annexin V+PI staining on neutrophils, exposed or not to DIA-12.3 lgG1. (A) % live cells; (B) % early proapoptotic cells; (C) late apoptotic cells.

Figure 14: DIA-12.3 lgG1 and DIA-12.3 N297A lgG1 reactivity on CEACAM1 antigen expressed on MelC by Flow cytometry. Binding percentages are indicated in the graphs.

Detailed description

It forms an object of the present invention a composition comprising a fully human antibody specific for CEA and CEACAM1 , comprising the light chain variable region sequence defined by SEQ ID NO: 1 (LCVR1 ); the heavy chain variable region sequence defined by SEQ ID NO: 2 (HCVR2) and the human antibody Immunoglobulin G constant region sequences; the light chain variable region sequence being:

SELTQDPAVSVALGQTVRITCQGDSLRSSYASWYRQRPGQAPVPVIY GKNNRPSGIPDRFSGSSSGNTASLTITGAQAEDEADYYCX6SSYAWL PYVVFGGGTKLTVLG (SEQ ID NO: 1 ) wherein X6 is any naturally occurring amino acid. In an embodiment, X6 is N or Q or A or L. In a preferred embodiment, X6 is Q.

The heavy chain variable region sequence being:

EVQLAESGGGLVQPGGSLRLSCAASGFTFSSDAMSWVRQAPGKGLE VWSAISGSGGSTYYADSVKGRFTISRDNSKNTLYLQMNSLRAEDTAV YYCAKSNEFLFDYWGQGTLVTVSR (SEQ ID NO: 2).

In an embodiment, the light chain constant region sequences are defined by SEQ ID NO: 8, GQPKAX1 PX2VTLFPPSSEELQANKATLVCLISDFYPGAVTVAWKADX 3SPVKAGVETTX4PSKQSNNKYAASSYLSLTPEQWKSHRSYSCQVTH EGSTVEKTVAPTECS, wherein X1 , X2, X3, X4 are any naturally occurring amino acid.

In an embodiment, X1 is N or A, X2 is T or S, X3 is G or S, X4 is K or T. In an embodiment, the heavy chain constant region sequences are defined by SEQ ID NO: 7, ASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALT SGVHTFPAVLQSSGLYSLSSWTVPSSSLGTQTYICNVNHKPSNTKVD KKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVT CVWDVS H EDP EVKFN WYVDGVEVH NAKTKPRE EQYX5STYRWSV LTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPP SRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDS DGSFFLYSKLTVDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPG K, wherein X5 any naturally occurring amino acid In an embodiment, X5 is N or A.

In an embodiment, X1 is N, X2 is T, X3 is G, X4 is K.

In an alternative embodiment, X1 is A, X2 is S, X3 is S, X4 is T.

In an embodiment, the light chain variable region sequence is LCVR1 , X6 being N, the heavy chain variable region sequence is HCVR2, the light chain constant region sequence is defined by SEQ ID NO: 3; the heavy chain constant region sequence is defined by SEQ ID NO: 4, wherein SEQ ID NO: 3 is GQPKANPTVTLFPPSSEELQANKATLVCLISDFYPGAVTVAWKADGS PVKAGVETTKPSKQSNNKYAASSYLSLTPEQWKSHRSYSCQVTHEG STVEKTVAPTECS and SEQ ID NO: 4 is ASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALT SGVHTFPAVLQSSGLYSLSSWTVPSSSLGTQTYICNVNHKPSNTKVD KKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVT CVWDVSHEDPEVKFNWYVDGVEVHNAKTKPREEQYNSTYRWSVL TVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPS RDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSD GSFFLYSKLTVDKSRWQQGNVFSCSV MHEALHNHYTQKSLSLSPGK. In an embodiment, the light chain variable region sequence is LCVR1 , X6 being Q, the heavy chain variable region sequence is HCVR2, the light chain constant region sequence is defined by SEQ ID NO: 3; the heavy chain constant region sequence is defined by SEQ ID NO: 4.

In an embodiment, the light chain variable region sequence is LCVR1 , X6 being N, the heavy chain variable region sequence is HCVR2, the light chain constant region sequence is defined by SEQ ID NO: 3; the heavy chain constant region sequence is defined by SEQ ID NO: 5, wherein SEQ ID NO: 5 is ASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALT SGVHTFPAVLQSSGLYSLSSWTVPSSSLGTQTYICNVNHKPSNTKVD KKVEPKSCDKTHTCPPCPAPELLGGPSVFLFPPKPKDTLMISRTPEVT CVWDVS H EDP EVKFN WYVDGVEVH NAKTKPRE EQYASTYRWSVL TVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTLPPS RDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSD GSFFLYSKLTVDKSRWQQGNVFSCSV MHEALHNHYTQKSLSLSPGK. In an embodiment, the light chain variable region sequence is LCVR1 , X6 being Q, the heavy chain variable region sequence is HCVR2, the light chain constant region sequence is defined by SEQ ID NO: 3; the heavy chain constant region sequence is defined by SEQ ID NO: 5.

In an embodiment, the light chain variable region sequence is LCVR1 , X6 being N, the heavy chain variable region sequence is HCVR2, the light chain constant region sequence is defined by SEQ ID NO: 6; the heavy chain constant region sequence is defined by SEQ ID NO: 4, wherein SEQ ID NO: 6 is

GQPKAAPSVTLFPPSSEELQANKATLVCLISDFYPGAVTVAWKADSSP VKAGVETTTPSKQSNNKYAASSYLSLTPEQWKSHRSYSCQVTHEGS TVEKTVAPTECS.

In an embodiment, the light chain variable region sequence is LCVR1 , X6 being Q, the heavy chain variable region sequence is HCVR2, the light chain constant region sequence is defined by SEQ ID NO: 6; the heavy chain constant region sequence is defined by SEQ ID NO: 4.

In an embodiment, the heavy chain constant region sequence is defined by SEQ ID NO: 9, wherein SEQ ID NO: 9 is ASTKGPSVFPLAPCSRSTSESTAALGCLVKDYFPEPVTVSWNSGALT SGVHTFPAVLQSSGLYSLSSWTVPSSSLGTKTYTCNVDHKPSNTKV DKRVESKYGPPCPPCPAPEFLGGPSVFLFPPKPKDTLMISRTPEVTCV WDVSQEDPEVQFNWYVDGVEVHNAKTKPREEQFNSTYRVVSVLTV LHQDWLNGKEYKCKVSNKGLPSSIEKTISKAKGQPREPQVYTLPPSQ EEMTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDG SFFLYSRLTVDKSRWQEGNVFSCSVMHEALHNHYTQKSLSLSLGK.

In an embodiment, the human Immunoglobulin G is Class I or Class IV IgG where the class IV IgG is stabilized with the S228P mutation, reference is made to SEQ ID NO: 9. In a preferred embodiment, said human Immunoglobulin G is Class I.

In an embodiment, said antibody is DIA-12.3 lgG1 , wherein the light chain variable region sequence is SEQ ID NO: 1 , X6 being N, the heavy chain variable region sequence is SEQ ID NO: 2, the light chain constant region sequence is SEQ ID NO: 3; the heavy chain constant region sequence is SEQ ID NO: 4.

In an embodiment, said antibody is DIA-12.3 lgG1 N87A, wherein the light chain variable region sequence is SEQ ID NO: 1 , X6 being A, the heavy chain variable region sequence is SEQ ID NO: 2, the light chain constant region sequence is SEQ ID NO: 3; the heavy chain constant region sequence is SEQ ID NO: 4.

In an embodiment, said antibody is DIA-12.3 lgG1 N297A, wherein the light chain variable region sequence is SEQ ID NO: 1 , the heavy chain variable region sequence is SEQ ID NO: 2, the light chain constant region sequence is SEQ ID NO: 3; the heavy chain constant region sequence is SEQ ID NO: 5. n an embodiment, said antibody is DIA-12.3 lgG1 N87A, N297A wherein the light chain variable region sequence is SEQ ID NO: 1 , X6 being A, the heavy chain variable region sequence is SEQ ID NO: 2, the light chain constant region sequence is SEQ ID NO: 3; the heavy chain constant region sequence is SEQ ID NO: 5.

The DIA-12.3 lgG1 N297A is a particularly preferred embodiment, wherein the introduced mutation at the glycosylation site N297, known to reduce effector function, does not impact the affinity of the antibody for the target.

In an embodiment, said antibody is DIA-2.2(3) lgG-4, wherein the light chain variable region sequence is LCVR1 , the heavy chain variable region sequence is HCVR2, the light chain constant region sequence is defined by SEQ ID NO: 3; the heavy chain constant region sequence is defined by SEQ ID NO: 9.

In an embodiment, said antibody is DIA-2.2 lgG4, wherein the light chain variable region sequence is LCVR1 , the heavy chain variable region sequence is HCVR2, the light chain constant region sequence is defined by SEQ ID NO: 6; the heavy chain constant region sequence is defined by SEQ ID NO: 9.

The antibody according to the present invention demonstrated capable to enhances NK cell mediated cytotoxicity against tumor cells. Data reported in Fig. 11 demonstrates the impact of DIA-12.3 lgG1 on Melanoma cells (panel A) and on bladder, colorectal and head and neck cancer cells (panel B). Moreover, DIA-12.3 lgG1 binds CEACAM1 antigen, when expressed on neutrophils isolated from healthy donors, as shown in Figure 12, without exerting any toxicity on the same cells, Figure 13.

In another embodiment, the antibody specific for CEA and CEACAM1 according to the present invention is conjugated to at least one diagnostic and/or therapeutic agent.

Non limiting examples of said at least one therapeutic agent are an antibody, an antigen-binding antibody fragment, a cytotoxic agent, a drug, a toxin, a radionuclide, boron atoms, an immunomodulator, a photoactive therapeutic agent, an immunoconjugate, an oligonucleotide and a hormone.

Non limiting examples of said at least one diagnostic agent are a radionuclide, a radiological contrast agent, a paramagnetic ion, a metal, a fluorescent label, a chemiluminescent label, an ultrasound contrast agent and a photoactive agent.

In another embodiment, the antibody specific for CEA and CEACAM1 according to the present invention is combined with at least one diagnostic and/or therapeutic agent.

In an embodiment, it is here claimed a pharmaceutical composition comprising at least one antibody according to the present invention.

In an embodiment, said pharmaceutical composition further comprises an additional therapeutic and/or diagnostic agent. In yet another embodiment, a method for diagnosing or treating a patient comprises the step of administering in an appropriate regimen the conjugate or the combination of the previous preferred embodiments.

In yet another embodiment, said method comprises the step of administering the pharmaceutical composition according to the present invention in combination with conventional tumor therapy, wherein the conventional tumor therapy is selected from the group consisting of: radiotherapy, chemotherapy, antibody therapy, and surgical tumor removal.

In yet another embodiment, the antibody specific for CEA and CEACAM1 , conjugates or compositions thereof is claimed for use in a method for the treatment of a pathologies selected in the group comprising: medullary thyroid cancer (MTC), colorectal cancer, hepatocellular carcinoma, liver cancer, gastric cancer, oesophageal cancer, lung cancer, non-small cell lung cancer, breast cancer, pancreatic cancer, ovarian cancer, uterine cancer, cervical cancer, endometrial cancer, head-and-neck cancer, bladder cancer, urothelial cancer, prostate cancer, hematopoietic cancer, leukaemia or melanoma. In yet another embodiment, the antibody specific for CEA and CEACAM1 , conjugates or compositions thereof is claimed for use in a method for the diagnosis of cancer.

Another preferred embodiment comprises the DNA sequence of the heavy and light chain sequences described above.

Other objects, features and advantages of the present invention will become apparent from the appended claims.

In a particularly preferred embodiment of the present invention, the light chain variable region sequence defined by SEQ ID NO: 1 (LCVR1 ); the heavy chain variable region sequence defined by SEQ ID NO: 2 (HCVR2) are grafted into a fully human antibody.

In this context, human antibody refers to any antibody that occurs in a human or an engineered antibody that has been designed, in some respect, to be compatible with the human immune system.

Particularly preferred for this purpose are antibodies that, broadly, do not elicit an adverse immune response in a patient.

Advantages

The authors of the present invention have surprisingly found IgG antibodies which bind with high efficiency and in a highly selective manner CEA and CEACAM1. The here described IgG antibodies are surprisingly better than the already available scFv.

In a preferred embodiment, the IgG is a Class I IgG. Surprisingly, this embodiment has been demonstrated to have a higher affinity than the embodiment where the IgG is a Class IV IgG.

In a further preferred embodiment, the IgG is a Class I IgG and a mutation has been introduced at the conserved amino acid N297. In this preferred embodiment, the effector function is reduced, and the affinity is maintained.

Surprisingly, this is further improved in the embodiment wherein a mutation is introduced in the light chain variable region sequence, a position 87. Preferably, the conserved amino acid N87 is substituted with a Glutamine, or with an Alanine, or with a Leucine. To note, there are no indication in the state of the art to introduce mutation in the variable region sequence.

The examples that follows have not limiting scope, wherein the scope of protection of the present invention is defined by the claims.

Experimental section

Example 1 : Generation of monoclonal lgG1 and lqG4 antibodies Transient production of lgG1 and lqG4 antibodies in HEK 293T cells HEK293T and HEK293 (Human Embryonic Kidney) cells were purchased from DSMZ (Germany). HEK293T cells were kept in culture in DMEM medium (Dulbecco's Modified Eagle Medium) (Carlo Erba, Italy), supplemented with 10% decomplemented foetal bovine serum (FBS, Sigma Aldrich, USA), 2mM L-glutamine (Sigma Aldrich, USA). All cell lines were grown in T-Flask (VWR, USA) and incubated at 37 °C in an incubator with 5% CO2 in a humidified atmosphere. HEK293T cells were transiently co-transfected with plasmids encoding the Heavy chain (HC) and the Light chain (LC) of the antibodies containing the same variable regions as the scFv DIATHIS-1 fused to the lgG1 or lgG4 with the S228P mutation subclasses constant regions. Briefly, the day before the transfection 2x10 ⁶ HEK293T cells in 10 ml of media were seeded in each of 20 T75-flask. The day after, the cells were co-transfected with a mixture of HC and LC plasmids, total of 10 pg of plasmid DNA and the transfection reagent GeneCellin (Eurobio, France). Cells from 10 flasks were co-transfected with HC and LC plasmids belonging to the lgG1 subclass and cells from 10 flasks with HC and LC plasmids belonging to the lgG4 subclass.

Transfected cells were grown for 7 days at 37°C in CO2 incubator and the cell culture medium was then collected for antibodies purification with protein A affinity chromatography. The yield of purified antibody was similar but higher for the lgG1 compared to the lgG4 (1.52 mg vs 1.25 mg).

Selection of stable clones for the production of IgG 1 and lgG4

For the selection of a stable cell line expressing the lgG1 or the lgG4(S228P) antibodies, a bicistronic vector containing both the HC and LC genes was created for each antibody. The expression vector used contains the gene for the Neomycin resistance allowing the use of this antibiotic for the selection of stable clones.

HEK293 cells were used for the stable pools generation and for clones selection. Briefly, HEK293 cells were transfected with the lgG1 or lgG4(S228P) bicistronic plasmid. The plasmid was linearized prior the transfection. Two days after the transfection, cells were expanded in medium containing neomycin (G418) at a concentration of 500pg/ml. The cells were fed with the selective media every 3-4 days. After 14 days of selection, neomycin resistant pools were tested for antibody expression analyzing the reactivity by ELISA against the CEACAM1 antigen.100 pl of culture medium from each pool were analyzed in duplicate with the method described below. Fig 1A is exemplificative of the expression levels in lgG1 transfected pools, and Fig 1 D of lgG4 transfected pools. The best expressing pools were used for cloning through limiting dilution in 96 well plates. Selected clones were screened for lgG1 or lgG4 expression in the culture media by ELISA.

As shown in Figure 1 , lgG1 pools and clones have a much higher reactivity compared to the lgG4 counterparts. IgG 1 clone 12 (Fig. 1 B) and lgG4 clone 2 (Fig. 1 E) that showed the highest reactivity against CEACAM1 , were expanded for antibodies production. Briefly, the cells were seeded in T-182 flasks at the density of 250.000 cells/ml. Cells were grown for 2 days at 30°C followed by further 7 days at 37°C in CO2 incubator and the culture media harvested and used for antibodies purification with protein A affinity chromatography. The yield values were 9.2 mg/L for lgG1 clone 12 and 4.6 mg/L for lgG4 clone 2.

In SEC-HPLC analysis performed under isocratic conditions, both lgG1 purified clone 12 and lgG4 purified clone 2 eluted as a single peak with a molecular weight of about 150kDa corresponding to the monomeric form of IgG (Fig.2), confirming the integrity of the antibodies. Exclusion chromatography was performed using the TSKgel G3000SWXL column (Tosoh Bioscience, Japan), with eluent consisting of 0.1 mol I L Na2SO4 + 0.05% NaNs in 0.1 mol I L phosphate buffer (pH = 6.7). Molecular weights were calculated based on a standard curve obtained by calibrating the system with molecules of known molecular weight: Thyroglobulin 670 kDa, Y-globulin 150 kDa, Ovalbumin 44.3 kDa, Ribonuclease A 43.7 kDa, Para-aminobenzoic acid 137.1 Da.

IgG 1 clone 12 was used for a further subcloning step with limiting dilution in 96 well plates. Results are showed in Figure 1 C. Among the subclones arised from the limiting dilution, the 12.3 showing the highest reactivity with CEACAM1 was expanded and used for antibody production as described above. The yield of purified mAb was 10.4 mg/L. SDS-PAGE, 10% acrylamide, showed the purity of the antibody (Figure 3).

Example 2: ELISA titration Purified antibodies were analysed for their reactivity against CEACAM1 antigen in ELISA titration assay as follow:

96 wells microtiter plates (MAXISORB NUNC) were coated with 100 pl/well of CEACAM1 antigen (Diatheva) solution 1 pg/ml diluted in PBS pH 7.3 and incubated overnight (ON) at 37°C. Plates were washed five times with PBS added with 0.05 % (v/v) Tween 20 pH 7.3 (PBST) (the same washing procedure was repeated after each incubation step during the assays) and then blocked with 1 % (w/v) BSA dissolved in PBS (150 pl/well) and maintained at 37°C for 60 min. After washing, serial dilutions of lgG1 or lgG4 in PBS containing 2% (w/v) not-fat dry milk (PBSM) were added to the plate (100 pl/well) and incubated for 90 min at 37°C. After washes, plates were incubated at 37°C for 60 min with an anti-human Fc antibody HRP conjugated (Meridian) diluted 1 :500 v/v in PBSM. Finally, the detection was performed after addition of 2,2'-azino-bis(3- ethylbenzthiazoline-6-sulphonic acid) (ABTS; Roche, 10102946001 ) as substrate. The absorbance values were read by a microplate reader (BioRad Laboratories) at 405 nm after 45 min. In Figure 4 are reported the results obtained with lgG1 (black line) and lgG4 (grey line) antibodies purified after transient expression in HEK293T cells. The absorbance of the mock sample is shown by the dotted line. The ELISA assay highlighted a higher affinity for the CEACAM1 antigen of the antibody belonging to the lgG1 subclass compared to the same antibody with the lgG4 (S228P) constant region. The EC50 were 12.79 and 26.71 pg/ml for the lgG1 and the lgG4, respectively.

ELISA assay was then performed with antibodies isolate from stable clones: lgG1 clone 12 and lgG4 clone 2 and, for comparative purpose only, to detect scFvDIATHIS-1 reactivity (as described in WO201 1160859A1 ). In this case, the plates were incubated at 37°C for 60 min with an anti-His6 monoclonal antibody (100 pl/well; AbD Serotec, MCA1396) diluted 1 :1 ,000 (v/v) in PBSM. Then, a goat anti-mouse polyclonal antibody HRP conjugated (Bio-Rad, 172-1011 EDU) diluted 1 :1 ,000 v/v in PBSM was dispensed (100 pl/well) and incubated for 60 min at 37°C. Results shown on Figure 5, panel A, confirmed a higher affinity for the lgG1 (EC50 1.27 pg/ml) compared to the lgG4 (EC502.554 pg/ml). In addition, both lgG1 and lgG4 have a higher reactivity in ELISA compared to scFvDIATHIS-1 (Figure 5 panel B) that showed an EC50 of 4.155 pg/ml, i.e., doubled compared to the lgG4 and 4 fold that of the igGi .

A further experiment has been performed wherein the lgG1 subclone 12.3 (i.e., DIA-12.3 lgG1 ) antibody has been tested by ELISA together with the N297A mutant (i.e. DIA-12.3 lgG1 N297A). Results are shown in Figure 10. The N297A mutated antibody (black line) shows an overlapping performance than is naive counterpart (grey line).

Example 3: flow cytometry assay on target cells

The antibodies purified after transient expression in HEK293T cells, were also evaluated for the binding of CEACAM1 expressed on the surface of metastatic melanoma cells MelC and MelP5. The analysis is carried out by dispensing 5x10 ⁵ MelC cells (primary metastatic melanoma cells) or MEL- P5 melanoma cell lines in 1 mL of medium per tube. The cells are then washed twice in PBS by centrifuging at 1 ,600 rpm for 6 min and finally resuspended in 200 pl of a blocking solution consisting of 1X PBS and 1 % w/v BSA and incubated for 30 minutes at room temperature (RT). After this phase, the antibody diluted in 200 pl of block solution is added and incubated for 60 minutes at room temperature. At the end of the incubation, two washes are carried out in PBS and each cell pellet is resuspended in 200 pl of a solution consisting of an anti-human-FITC antibody diluted 1 : 1000 v/v in blocking solution, the incubation is maintained for 60 minutes at RT. Cells are then washed once with PBS and centrifuged as above, the supernatant is removed, and pellet resuspended in 500 pl of PBS. The samples were subjected to flow cytometric analysis with FacScan (Becton Dickinson). In a first assay, MelC cells were reacted with 50, 25 or 12pg/ml of anti CEACAM1 lgG1 or lgG4. Results are reported in Figure 6.

The assay was repeated on a different cell line, the human melanoma cell line MEL-P5. Results are reported in Figure 7.

On the two cell lines, the percentage of positive cells as well as the values of the mean fluorescence intensity (MFI), representing a measure of the antibody affinity, were higher for the lgG1 compared to the lgG4 at every concentration tested confirming the results obtained in ELISA.

In a third assay, HT29 colorectal cancer cells and TCCSLIP bladder cancer cells were reacted with 20 or 10 pg/ml of lgG1 clone 12 or lgG4 clone 2 and, for comparative purposes, scFv DIATHIS-1. Results of the MFI are reported in Figure 8, with reference to bladder cancer cells, panel A, and with reference to colorectal cancer cells, panel B. In the two experimental setting, i.e. , antibody obtained from transiently expressing cells or from stable clones, it is evident that the affinity obtained with the lgG1 antibody and lgG4 antibody according to the present invention is higher with respect to the scFv. lgG1 is confirmed by this assay, too, as the most preferred embodiment.

The selected DIA-12.3 lgG1 of the present invention was further characterized for the binding to different tumor cell lines by flow cytometry. The results reported on Figure 9 showed a high percentage of positive cells and high MFI values at every concentration tested in two metastatic bladder cancer cells (TTCSLIP and 5637 cell lines) and for the colorectal cancer cell line HT29.

Example 4: LDH (Lactate Dehydrogenase) cytotoxicity assay

To evaluate the capability of the DIA-12.3 lgG1 antibody to affect the NK- 92 cell cytotoxicity against tumor cells, the LDH-Glo ^TM Cytotoxicity Assays (Promega) was performed. This assay is a bioluminescence method that quantifies the enzymatic activity of LDH, a widely used marker of cytotoxicity, which is released in the cytoplasm upon cell membrane destruction.

Target tumor cells were harvested, washed, and resuspended at the final density of 2x10 ⁵ cells/mL in basic DMEM with 5% v/v of FBS. Target cells were incubated alone or with the DIA 12.3 antibody for 30 minutes at RT. Meanwhile, NK-92 cells were firstly diluted at 2x10 ⁶ cells/mL in basic RPMI medium with 5% v/v of FBS and then serially diluted ranging routinely from 1x10 ⁵ to 1x10 ³ cells. 50 pl/well of each NK-92 cell suspension dilution were seeded in 96 white walled assay plates, in triplicate. Following the antibody treatment, 50 pl/well of the target cell suspension were then added to the effector NK-92 cells at increasing effector/target cell ratio (E:T) ranging from 1 :1 to 10:1 in triplicate. Positive controls were used by adding 2 pl/well of TritonX-100 to determine the maximum LDH release, while triplicate wells without cells were set up to serve as negative control to determine the culture medium background. The plate was incubated for 4 hours at 37°C. 5 pl of supernatants from each well were collected and diluted in 95 pl of LDH Storage buffer. The diluted samples were then stored at -20°C and the day of the assay they were further diluted in the LDH Storage Buffer to fit the linear range of the assay. 50 pl of the diluted samples were transferred into a 96-well opaque, non-transparent assay plate. 50pl of the LDH Detection Reagent were then added in each well and the plate was incubated at room temperature for 60 minutes. Luminescence was recorded from 30 to 60 minutes after the addition of the LDH reaction agent. The % of cytotoxicity was calculated as follows:

100 x (Experimental LDH Release - Medium Background)/(Maximum LDH Release Control - Medium Background).

Example 5: flow cytometry assay on effector cells

DIA-12.3 lgG1 was evaluated for the binding of CEACAM1 expressed on the surface of neutrophil cells isolated from healthy donors. The analysis is carried out by dispensing 5x10 ⁵ neutrophils cells in 1 mL of medium per tube. The cells are then washed twice in PBS by centrifuging at 1 ,600 rpm for 6 min and finally resuspended in 200 pl of a blocking solution consisting of 1X PBS and 1 % w/v BSA and incubated for 30 minutes at room temperature (RT). After this phase, the antibody diluted in 200 pl of block solution is added and incubated for 60 minutes at room temperature. At the end of the incubation, two washes are carried out in PBS and each cell pellet is resuspended in 200 pl of a solution consisting of an anti-human-FITC antibody diluted 1 : 1000 v/v in blocking solution, the incubation is maintained for 60 minutes at RT. Cells are then washed once with PBS and centrifuged as above, the supernatant is removed, and pellet resuspended in 500 pl of PBS. The samples were subjected to flow cytometric analysis with FacScan (Becton Dickinson). Neutrophils were reacted with 10 pg/ml of DIA-12.3 lgG1. Results are reported in Figure 12.

The binding to CEACAM1 does not impact viability of effector cells. At this end, neutrophils isolated from 3 patients and enriched were incubated in RPMI medium in the presence or in the absence of DIA-12.3 lgG1 for 1 , 2, 6, 18 and 20 h, at 37°C. At the end, the Annexin V/PI staining (Promega) has been performed, according to manufacturer’s instructions. The results, shown in Figure 13, do not show any difference between treated and untreated sample at any time points.

Example 6:

DIA-12.3 lgG1 and the mutated DIA-12.3 N297A lgG1 were tested for their reactivity on CEACAM1 expressed on melanoma cells by Flow cytometry. MelC cells were reacted with 20 or 10 pg/ml of DIA-12.3 lgG1 or with the lgG1 N297A mutated antibody. Negative control (CTR-) is represented by MelC cells reacted with secondary antibodies only. The N297A showed a reduced Fc effector functions, as highlighted in Figure 14.