Antibody method for treatment of a disease in which the target cells are cells which express CD20

Abstract

The present invention relates to a monoclonal antibody method directed against CD20 antigen including administration of an anti CD20 antibody wherein each of the light chains thereof has the murine-human chimeric amino acid sequence SEQ ID NO: 28, and each of the heavy chains thereof has the murine-human chimeric amino acid sequence SEQ ID NO: 20.

Claims

1. A method for the treatment of a disease in which the target cells are cells which express CD20, which comprises administering to a patient an effective amount of a monoclonal antibody directed against the CD20 antigen, wherein each of the light chains thereof has the murine-human chimeric amino acid sequence SEQ ID No. 28, and each of the heavy chains thereof has the murine-human chimeric amino acid sequence SEQ ID No. 20.

2. The method of claim 1, wherein the disease in which the target cells are cells which express CD20 is an immune dysfunction disease involving B lymphoid cells.

3. The method of claim 2, wherein the immune dysfunction disease involving B lymphoid cells is selected from auto-immune diseases.

4. The method of claim 1, wherein the disease in which the target cells are cells which express CD20 is chronic graft-versus-host disease.

5. The method of claim 1, wherein the disease in which the target cells are cells which express CD20 is organ transplant rejection.

6. The method of claim 5, wherein the organ transplant rejection is kidney transplant rejection.

7. The method of claim 1, wherein each of the light chains thereof is encoded by murine-human chimeric nucleic acid sequence SEQ ID No. 27, and each of the heavy chains thereof is encoded by murine-human chimeric nucleic acid sequence SEQ ID No. 19.

8. The method of claim 1, wherein the antibody is produced by a rat hybridoma cell line.

9. The method of claim 8, wherein the antibody is produced in the rat hybridoma cell line YB2/3HL.P2.G11.16Ag.20, registered at the American Type Culture Collection under ATCC number CRL-1662.

10. The method of claim 9, wherein the antibody is the EMAB603 antibody produced by clone R603, registered under registration number CNCM I-3529 at the Collection Nationale de Cultures de Microorganismes (CNCM).

11. The method of claim 1, wherein the antibody is produced in a CHO line.

12. A method for the treatment of a disease in which the target cells are cells which express CD20, which comprises administering to a patient an effective amount of a monoclonal antibody directed against the CD20 antigen and one or more further antibodies, wherein each of the light chains of the monoclonal antibody directed against the CD20 antigen has the murine-human chimeric amino acid sequence SEQ ID No. 28, and each of the heavy chains of the monoclonal antibody directed against the CD20 antigen has the murine-human chimeric amino acid sequence SEQ ID No. 20.

13. The method of claim 12, wherein the disease in which the target cells are cells which express CD20 is an immune dysfunction disease involving B lymphoid cells.

14. The method of claim 13, wherein the immune dysfunction disease involving B lymphoid cells is selected from auto-immune diseases.

15. The method of claim 12, wherein the disease in which the target cells are cells which express CD20 is chronic graft-versus-host disease.

16. The method of claim 12, wherein the disease in which the target cells are cells which express CD20 is organ transplant rejection.

17. The method of claim 16, wherein the organ transplant rejection is kidney transplant rejection.

Description

DESCRIPTION OF THE FIGURES

(1) FIG. 1: Schematic representation of the CKHu vector used for the chimerisation of the light chain kappa of antibodies EMAB6 and EMAB603.

(2) FIG. 2: Schematic representation of the light chain pEF-EMAB6-K expression vector used for the production of antibody EMAB6.

(3) FIG. 3: Schematic representation of the G1Hu vector used for the chimerisation of the heavy chain of antibodies EMAB6 and EMAB603.

(4) FIG. 4: Schematic representation of the heavy chain pEF-EMAB6-H expression vector used for the production of antibody EMAB6.

(5) FIG. 5: Competition by the chimeric EMAB6 antibody for the binding of the murine antibody produced by CAT-13.6E12 (CAT13) to CD20 expressed on Raji cells.

(6) FIG. 6A: Complement-dependent cytotoxic activity of the anti-CD20 antibodies on Raji cells. Rituxan?: open triangle, EMAB6: closed lozenge. Cell lysis is estimated by measuring the intracellular LDH released into the supernatant. Results are expressed as percentage lysis, with 100% being the value obtained with Rituxan? (at 5,000 ng/mL anti-CD20 antibody). Mean of 5 tests. FIG. 6B: Complement-dependent cytotoxic activity of the anti-CD20 antibodies on Raji cells. Comparison of the complement-dependent cytotoxic activities of EMAB6 (closed lozenge) and EMAB603 (open lozenge).

(7) FIG. 7A: ADCC activity induced by anti-CD20 antibodies in the presence of Raji cells. Rituxan?: open triangle, EMAB6: closed lozenge. Cell lysis is estimated by measuring the intracellular LDH released into the supernatant. Results are expressed as percentage lysis, with 100% being the value obtained with Rituxan? (at 250 ng/mL anti-CD20 antibody). Mean of 3 tests. FIG. 7B: ADCC activity induced by anti-CD20 antibodies in the presence of Raji Cells. Comparison of ADCC induced by EMAB6 (closed lozenge) and EMAB603 (open lozenge).

(8) FIG. 8: ADCC activity induced by anti-CD20 antibodies in the presence of B lymphocytes from patients with B-CLL. Rituxan?: open triangle, EMAB6: closed lozenge. E/T ratio=15. Cell lysis is estimated by measuring the calcein released into the supernatant. Results are expressed as percentages, with 100% being the value obtained with Rituxan? (at 500 ng/mL anti-CD20 antibody). Mean of 4 experiments corresponding to 4 different patients.

(9) FIG. 9A: Activation of CD16 (Fc?RIIIA) induced by anti-CD20 antibodies in the presence of Raji cells. Rituxan?: open triangle, EMAB6: closed lozenge. Results are expressed as percentage of IL-2, as measured in supernatants using ELISA; with 100% being the value obtained with Rituxan? (at 2,500 ng/mL anti-CD20 antibody). Mean of 4 tests. FIG. 9B: Activation of CD16 (Fcg RIIIA) induced by anti-CD20 antibodies in the presence of Raji cells. Comparison between the activation of CD16 (Fc?RIIIA) as induced by EMAB6 (closed lozenge) and EMAB603 (open lozenge).

(10) FIG. 10: Activation of CD16 (Fc?RIIIA) induced by anti-CD20 antibodies in the presence of B lymphocytes from patients with B-CLL. Rituxan?: open triangle, EMAB6: closed lozenge. Results are expressed as percentage of IL-2, as measured in the supernatants using ELISA; with 100% being the value obtained with Rituxan? (at 2,500 ng/mL anti-CD20 antibody). Mean of 12 patients.

(11) FIG. 11: Production of IL-2 induced by the CAT-13.6E12 murine antibody in the presence of Jurkat-CD16 cells (Fc?RIIIA).

(12) FIG. 12: Schematic representation of the heavy chain and light chain pRSV-HL-EMAB603 expression vector used for the production of antibody EMAB603.

EXAMPLES

Example 1: Construction of Expression Vectors for Anti-CD20 Chimeric Antibodies EMAB6 and EMAB603

(13) A. Determination of the Sequence of the Variable Regions of the CAT-13.6E12 Murine Antibody

(14) Total RNA from murine hybridoma CAT-13.6E12 cells (supplier: DSMZ, ref. ACC 474), which produces an IgG2a, ?-type immunoglobulin, was isolated (RNAeasy kit, Qiagen ref. 74104). After reverse transcription, the variable domains of the light (V?) and heavy (VH) chains of the CAT-13.6E12 antibody were amplified using the 5RACE technique (Rapid Amplification of cDNA Ends) (GeneRacer kit, Invitrogen ref. L1500-01). The primers used for the two steps were the following:

(15) 1. Reverse Transcription Primers

(16) a. Murine Kappa Specific Antisense Primer (SEQ ID No. 1)

(17) TABLE-US-00001 5-ACTGCCATCAATCTTCCACTTGAC-3
b. Murine G2a Specific Antisense Primer (SEQ ID No. 2)

(18) TABLE-US-00002 5-CTGAGGGTGTAGAGGTCAGACTG-3
2. 5RACE PCR Primers
a. Murine Kappa Specific Antisense Primer (SEQ ID No. 3)

(19) TABLE-US-00003 5-TTGTTCAAGAAGCACACGACTGAGGCAC-3
b. Murine G2a Specific Antisense Primer (SEQ ID No. 4)

(20) TABLE-US-00004 5-GAGTTCCAGGTCAAGGTCACTGGCTCAG-3

(21) The resulting VH and V? PCR products were cloned into vector pCR4Blunt-TOPO (Zero blunt TOPO PCR cloning kit, Invitrogen, ref. K2875-20) and sequenced. The nucleotide sequence of the V? region of the murine CAT-13.6E12 antibody is shown as sequence SEQ ID No. 5 and the deduced peptide sequence is sequence SEQ ID No. 6. The V? gene belongs to the V?4 class [Kabat et al. (1991) Sequences of Proteins of Immunological Interest. NIH Publication 91-3242].

(22) The nucleotide sequence of the VH region of CAT-13.6E12 is sequence SEQ ID No. 7 and the deduced peptide sequence is sequence SEQ ID No. 8. The VH gene belongs to the VH1 class [Kabat et al. (1991) Sequences of Proteins of Immunological Interest. NIH Publication 91-3242].

(23) B. Construction of Heavy and Light Chain Expression Vectors for Chimeric Antibodies EMAB6 and EMAB603

(24) 1. Light Chain Kappa Vector

(25) 1.1. Light Chain Vector for Antibody EMAB6

(26) The V? sequence cloned into the pCR4Blunt-TOPO sequencing vector was amplified using the following cloning primers:

(27) a) V? Sense Primer (SEQ ID No. 9)

(28) TABLE-US-00005 5-CTCAGTACTAGTGCCGCCACCATGGATTTTCAAGTGCAGATTTTC AG-3

(29) The underlined sequence corresponds to the SpeI restriction site, the sequence in bold lettering corresponds to a Kozak consensus sequence, the ATG initiator is in italics.

(30) b) V? Antisense Primer (SEQ ID No. 10)

(31) TABLE-US-00006 5-TGAAGACACTTGGTGCAGCCACAGTCCGGTTTATTTCCAGCCTGG T-3

(32) This primer joins the murine V? sequences (in italics) to the human constant region (C?) (in bold). The underlined sequence corresponds to the DraIII restriction site.

(33) The resulting V? PCR product contains the sequence which codes for the natural signal peptide of the CAT-13.6E12 murine antibody. This V? PCR product was then cloned between the SpeI and DraIII sites of the light chain chimerisation vector (FIG. 1), which corresponds to sequence SEQ ID No. 11, at 5 in the human constant region C?, the nucleic acid sequence of which is sequence SEQ ID No. 21 and the deduced peptide sequence of which is sequence SEQ ID No. 22. The human C? sequence of this chimerisation vector had been modified beforehand by silent mutagenesis in order to create a DraIII restriction site to allow cloning of murine V? sequences to take place. This chimerisation vector contains an RSV promoter and a bGH (bovine Growth Hormone) polyadenylation sequence together with the dhfr (dihydrofolate reductase) selection gene.

(34) The light chain sequence of the chimeric EMAB6 antibody encoded by this vector is shown as SEQ ID No. 13 for the nucleotide sequence and corresponds to the deduced peptide sequence SEQ ID No. 14.

(35) 1.2. Light Chain Vector for Antibody EMAB603

(36) The protocol is the same as for the light chain vector for the EMAB6 antibody (see Example 1, B-1.1), apart from the V? antisense primer which is:

(37) b) V? Antisense Primer (SEQ ID No. 29)

(38) TABLE-US-00007 5-TGAAGACACTTGGTGCAGCCACAGTCCG TTTATTTCCA GCCTGGT-3

(39) This primer joins the murine V? sequences (in italics) to the human constant region (C?) (in bold). The underlined sequence corresponds to the DraIII restriction site.

(40) This primer also introduces the mutation AAC.fwdarw.AAA (framed nucleotide in the antisense primer sequence SEQ ID No. 29), which corresponds to mutation N106K (see nucleotide sequence and deduced peptide sequence SEQ ID No. 25 and SEQ ID No. 26) relative to the natural V? sequence of CAT-13.6E12 (see. SEQ ID No. 5 and SEQ ID No. 6).

(41) The light chain sequence of the chimeric EMAB603 antibody encoded by this vector is shown as SEQ ID No. 27 for the nucleotide sequence and corresponds to the deduced peptide sequence SEQ ID No. 28.

(42) 2. Heavy Chain Vector

(43) A similar approach was applied to the chimerisation of the heavy chains of the EMAB6 and EMAB603 antibodies.

(44) The VH sequence cloned into the pCR4Blunt-TOPO vector was first of all amplified using the following cloning primers:

(45) a) VH Sense Primer (SEQ ID No. 15)

(46) TABLE-US-00008 5-CTCAGTACTAGTGCCGCCACCATGGGATTCAGCAGGATCTTTCT C-3

(47) The underlined sequence corresponds to the restriction site SpeI, the sequence in bold lettering corresponds to a Kozak consensus sequence, the ATG initiator is in italics.

(48) b) VH Antisense Primer (SEQ ID No. 16)

(49) TABLE-US-00009 5-GACCGATGGGCCCTTGGTGGAGGCTGAGGAGACGGTGACTGAGGTT CC-3

(50) This primer joins the murine VH sequences (in italics) to the human G1 constant region (in bold). The underlined sequence corresponds to the ApaI restriction site.

(51) The amplified VH fragment contains the sequence which codes for the natural signal peptide of the CAT-13.6E12 murine antibody. This VH PCR product was then cloned between the SpeI and ApaI sites in the heavy chain chimerisation vector (FIG. 3) which corresponds to sequence SEQ ID No. 17, at 5 of the ?1 human constant region, the nucleic acid sequence of which is sequence SEQ ID No. 23 and the deduced peptide sequence of which is sequence SEQ ID No. 24. This chimerisation vector contains an RSV promoter and a bGH (bovine Growth Hormone) polyadenylation sequence as well as the neo selection gene.

(52) The heavy chain sequences of the chimeric EMAB6 and EMAB603 antibodies encoded by this vector are shown as SEQ ID No. 19 for the nucleotide sequence and as SEQ ID No. 20 for the deduced peptide sequence.

(53) 3. Final Expression Vectors

(54) 3.1. EMAB6 Antibody Expression Vectors

(55) For the expression of the EMAB6 antibody, the RSV promoter of the kappa light chain chimerisation vector (see Example 1, B-1.1) was replaced with the human EF-1 alpha promoter. The final light chain pEF-EMAB6-K expression vector is shown in FIG. 2 and corresponds to sequence SEQ ID No. 12.

(56) The light chain sequence of the chimeric EMAB6 antibody encoded by this vector is shown as SEQ ID No. 13 for the nucleotide sequence and corresponds to the deduced peptide sequence SEQ ID No. 14.

(57) For the expression of the EMAB6 antibody, the RSV promoter of the heavy chain chimerisation vector (see Example 1, B-2) was replaced with the human EF-1 alpha promoter. The thus-obtained final heavy chain pEF-EMAB6-H expression vector is shown in FIG. 4 and corresponds to sequence SEQ ID No. 18.

(58) 3.2. EMAB603 Antibody Expression Vector

(59) A unique expression vector containing both heavy chain and light chain transcription units of the anti-CD20 EMAB603 antibody was constructed from two light and heavy chain chimerisation vectors (see Example 1, B-1.2 and B2 respectively) by sub-cloning into the XhoI site of the heavy chain vector, a BglII-PvuII fragment of the light chain vector containing the light chain transcription unit and the dhfr gene. This pRSV-HL-EMAB603 expression vector includes two selection genes, i.e. neo (neo-phosphotransferase II) and dhfr (dihydrofolate reductase), together with two heavy chain and light chain transcription units under the control of an RSV promoter (FIG. 12).

Example 2: Production of Cell Lines Derived from the YB2/0 Line Producing Anti-CD20 Chimeric EMAB6 and EMAB603 Antibodies

(60) The rat YB2/0 cell line (ATCC # CRL-1662) was cultivated in EMS medium (Invitrogen, ref. 041-95181M) containing 5% foetal calf serum (JRH Biosciences, ref. 12107). For transfection, 5 million cells were electroporated (Biorad electroporator, model 1652077) in Optimix medium (Equibio, ref. EKITE 1) with 25 ?g of light chain vector pEF-EMAB6-K (FIG. 2), linearised with AatII, and 27 ?g of heavy chain vector pEF-EMAB6-H (FIG. 4), linearised with ScaI, for the expression of the EMAB6 antibody, or with vector pRSV-HL-EMAB603, for the expression of the EMAB603 antibody. The electroporation conditions applied were 230 Volts and 960 microFarads in a 0.5-mL cuvette. Each electroporation cuvette was then distributed over 5 P96 plates at a density of 5,000 cells/well.

(61) Placement in a selective RPMI medium (Invitrogen, ref 21875-034) containing 5% dialysed serum (Invitrogen, ref. 10603-017), 500 ?g/mL G418 (Invitrogen, ref. 10131-027) and 25 nM methotrexate (Sigma, ref. M8407), was carried out 3 days after transfection.

(62) The supernatants from the resistant transfection wells were screened for the presence of chimeric immunoglobulin (Ig) by applying an ELISA assay specific to the human Ig sequences.

(63) The 10 transfectants producing the largest amount of antibody were amplified on P24 plates and their supernatants re-assayed using ELISA to estimate their productivity and select, by limited dilution (40 cells/plate), the best three producers for cloning.

(64) After cloning, the R509.6A4 clone (R509-33903/046-6H1(1)6A4, productivity: 17 ?g/10.sup.6 cells), hereafter referred to as R509, as well as the R603 clone were selected for the production of the chimeric EMAB6 and EMAB603 antibodies respectively and progressively acclimated to the CD Hybridoma production medium (Invitrogen, ref. 11279-023).

(65) The production of the chimeric EMAB6 and EMAB603 antibodies was achieved by expanding, in CD Hybridoma medium, the acclimated culture obtained by dilution to 3?10.sup.5 cells/mL in 75-cm.sup.2 and 175-cm.sup.2 vials and then dilution to 4.5?10.sup.5 cells/mL in roller flasks. Once the maximum volume (1 L) was achieved, culture was continued until the cell viability was only 20%. After production, the chimeric EMAB6 and EMAB603 antibodies were purified using protein-A affinity chromatography (HPLC estimated purity <95%) and checked by polyacrylamide gel electrophoresis.

Example 3: Characterisation of the Functional Activity of Chimeric Antibodies EMAB6 and EMAB603

(66) A. Specificity

(67) Specificity of the antigen recognition of the chimeric EMAB6 antibody was evaluated by studying the competition with the murine antibody CAT-13.6E12 (CAT13) for binding the CD20 antigen expressed by Raji cells.

(68) For that purpose, the EMAB6 antibody (10 ?L at 0.5 to 50 ?g/mL) was incubated at 4? C. with a fixed quantity of CAT-13.6E12 murine antibody (10 ?L at 5 ?g/mL) for 20 minutes in the presence of Raji cells (50 ?L at 4?10.sup.6 cells/mL). After washing, a mouse anti-IgG antibody coupled to phycoerythrin (PE) was added to the Raji cells so as to specifically detect the binding of the CAT-13.6E12 murine antibody. The Median Fluorescence Intensities (MFIs) obtained in the presence of various concentrations of EMAB6 are converted to percentages, with 100% corresponding to binding to CAT-13.6E12 cells in the absence of the EMAB6 antibody.

(69) An inhibition curve is thus obtained for binding of the CAT-13.6E12 (CAT13) antibody to Raji cells in the presence of increasing concentrations of EMAB6 (FIG. 5).

(70) This study demonstrates that the chimerisation process has not adversely affected the specificity of the EMAB6 antibody, which does compete with the parental CAT-13.6E12 murine antibody for binding to CD20 expressed on the surface of Raji cells.

(71) The antigen recognition specificity of the EMAB603 antibody is comparable with that of the EMAB6 antibody.

(72) B. Complement-Dependent Cytotoxic Activity

(73) Complement-dependent cytotoxic activity of the EMAB6 and EMAB603 antibodies was examined with Raji cells in the presence of young rabbit serum as a source of complement; the anti-CD20 chimeric antibody Rituxan? was included in one test, for comparison.

(74) For this test, the Raji cells were adjusted to 6?10.sup.5 cells/mL in IMDM (Iscove's Modified Dulbecco's Medium) 5% FCS (Foetal Calf Serum). The antibodies were diluted with IMDM+0.5% FCS. The reaction mixture was made up of 50 ?L antibody, 50 ?L young rabbit serum (1/10 IMDM+0.5% FCS dilution of Cedarlane CL 3441 reagent), 50 ?L target cells and 50 ?L IMDM+0.5% FCS medium. The final antibody concentrations were 5,000, 1,250, 250 and 50 ng/mL. A control without antibodies was included in the test. After 1 hr incubation at 37? C. in a 5% CO.sub.2 atmosphere, the plates were centrifuged and the levels of intracellular LDH released into the supernatant estimated using a specific reagent (Cytotoxicity Detection Kit 1 644 793).

(75) The percentage lysis was estimated using a calibration range obtained using various dilutions of target cells lysed using triton X100 (2%) corresponding to 100, 50, 25, and 0% lysis respectively.

(76) The results shown in FIG. 6(A) demonstrate that EMAB6 and Rituxan? both induce complement-dependent lysis of the Raji cells. Nevertheless, EMAB6 complement activity appeared to be slightly less than that of Rituxan?. This difference is greater at the low concentrations of antibody used in this test. Thus for concentrations of 50 and 250 ng/mL, the activity of EMAB6 is of the order of 45% of that of Rituxan?. This difference becomes smaller as the antibody concentration is increased, with the % complement-dependent cytotoxic activity of the EMAB6 antibody representing 92% of that of Rituxan? at the highest concentration tested, i.e. 5,000 ng/mL.

(77) This lower complement-dependent cytotoxic activity of the AMAB6 antibody compared to that of Rituxan? may be regarded as an advantage, since it limits the potential in vivo toxicity of EMAB6 compared to Rituxan?, associated with the activation of the conventional complement pathway, which leads to the production of various molecules with undesirable inflammatory, allergic and vascular activities.

(78) The complement activity of the EMAB603 antibody is shown in FIG. 6(B).

(79) C. ADCC Activity

(80) The cytotoxicity of the chimeric EMAB6 antibody was evaluated in the presence of Raji cells or B lymphocytes from patients with CLL. The anti-CD20 chimeric antibody Rituxan? was included in the tests for comparison.

(81) The calcein-labelling ADCC measurement technique used was as follows:

(82) NK cells were isolated from PBMCs using the separation on magnetic beads (MACS) technique from Myltenyi. The cells were washed and re-suspended in IMDM+5% FCS (45?10.sup.5 cells/mL). The effector cells and target cells were used in a ratio of 15/1. The Raji cells or the PBMCs (Peripheral Blood Mononuclear Cells) from patients with B-CLL obtained after Ficoll treatment (>95% B cells) were labelled beforehand with calcein (1 mL cells at 3?10.sup.6 cells/mL in IMDM+5% FCS+20 ?L calcein (20 ?M), 20 min incubation at 37? C. and then washing with HBSS (Hank's Buffered Saline Solution)) and adjusted to 3?10.sup.5 cells/mL in IMDM+5% FCS. The antibodies were diluted with IMDM+0.5% FCS (final concentrations: 500; 50; 5; 0.5; 0.05 and 0.005 ng/mL).

(83) The reaction mixture was made up of 50 ?L antibody, 50 ?L effector cells, 50 ?L target cells and 50 ?L IMDM medium in a P96 microtitration plate. Two negative controls were used: Lysis without NK: NK effector cells were replaced with IMDM+5% FCS. Lysis without antibodies (Ab): antibodies were replaced with IMDM+5% FCS.

(84) After 4 hrs incubation at 37? C. in a 5% CO.sub.2 atmosphere, the plates were centrifuged and the fluorescence associated with the supernatant was measured using a fluorimeter (excitation: 485 nm, emission: 535 nm).

(85) The percentage lysis was estimated using a calibration range obtained using various dilutions of target cells lysed using Triton X100 (2%), corresponding to 100, 50, 25, and 0% lysis respectively.

(86) The results were first calculated using the following formula:
% lysis=(% lysis with Antibody and NK)?(% lysis without Antibody)?(% lysis without NK)
and then expressed as relative percentages, with 100% being the value obtained at the highest concentration of Rituxan?.

(87) The results obtained for the EMAB6 antibody on the Raji line cells shown in FIG. 7(A) demonstrate that, irrespective of the concentration being tested, the cytotoxicity induced by the EMAB6 antibody is greater than that induced by Rituxan?. This difference is particularly large at low antibody concentrations. Thus, at 0.5 ng/mL, the lysis percentages were 96% and 4% for EMAB6 and Rituxan? respectively. By increasing the dose 500-fold (250 ng/mL), the difference is still appreciable since the relative percentages of ADCC are 164% and 100% for EMAB6 and Rituxan? respectively. When the EC50s were calculated (antibody concentration corresponding to 50% of the E Max, the maximum effectiveness obtained at the highest antibody concentration and at the plateau) by graphical estimation (in ng/mL) and assuming that Rituxan? and EMAB6 attain the same E Max, the Rituxan? EC50/EMAB6 EC50 ratio in this test was then equal to 300.

(88) The cytotoxicity of the chimeric EMAB603 antibody was evaluated in the presence of Raji cells using the same procedure as for the EMAB6 antibodies. Its activity was comparable with that of the EMAB6 antibody (see FIG. 7(B)).

(89) With the lymphocytes from patients with B-CLL, the results obtained, shown in FIG. 8, demonstrate that, irrespective of the concentration being tested, the cytotoxicity induced by the EMAB6 antibody is greater than that induced by Rituxan?. As already observed with the Raji cells, this difference is particularly large at low antibody concentrations. A concentration of 0.5 ng/mL EMAB6 induces the same percentage lysis as 500 ng/mL Rituxan?, i.e. a concentration ratio of 1,000. At 5 ng/mL, the lysis percentages are 269% and 9% for EMAB6 and Rituxan? respectively. At the maximum dose tested (500 ng/mL), the difference is still very large since the relative percentages of ADCC are 350% and 100% for EMAB6 and Rituxan? respectively. An interesting result corresponds to the concentrations which give rise to 50% lysis. In this test, the Rituxan? EC50/EMAB6 EC50 ratio was estimated as 10,000 (graphical estimate in ng/mL for EC50 assuming that Rituxan? and EMAB6 attain the same E Max).

(90) In these tests, the cytotoxic activities of EMAB6 and EMAB603 are therefore much greater than that of Rituxan?.

(91) D. Activation of CD16 (IL-2 Secretion)

(92) The activation of CD16 (Fc?RIIIA) induced by the chimeric EMAB6 antibody was determined in the presence of Raji cells or B lymphocytes from patients with CLL. This test evaluated the ability of the antibody to bind to CD16 (Fc?RIIIA) receptor expressed on the Jurkat-CD16 cells and to induce the secretion of IL-2. The anti-CD20 chimeric antibody Rituxan? is included in the tests for comparison.

(93) Measurement of CD16 activation was carried out in the following manner on the Jurkat-CD16 cell line in the presence of Raji cells or B lymphocytes from patients with CLL.

(94) Mixture in 96-well plate: 50 ?L antibody solution (dilution to 10,000, 1,000, 100 and 10 ng/mL with IMDM+5% FCS for B lymphocytes from patients with B-CLL and 10,000, 2,000, 1,000, 200, 100, 50 and 25 ng/mL for Raji cells), 50 ?L PMA (Phorbol Myristate Acetate, diluted to 40 ng/mL with IMDM+5% FCS), 50 ?L Raji or PBMCs from patients with B-CLL obtained after Ficoll treatment (>95% B cells) diluted to 6?10.sup.5/mL with IMDM+5% FCS, and 50 ?L Jurkat-CD16 cells (20?10.sup.6/mL in IMDM+5% FCS). Controls without antibodies were included in all tests. After incubation overnight at 37? C., the plates were centrifuged and the IL-2 contained in the supernatants estimated using a commercial kit (Quantikine from R/D). The OD readings were made at 450 nm.

(95) The results were initially expressed as IL-2 levels as a function of the antibody concentration (from 0 to 250 ng/mL final concentration), then as relative percentages, where 100% is the value obtained with Rituxan? at the highest test concentration.

(96) The results obtained with the Raji line cells shown in FIG. 9(A) demonstrate that, in the presence of EMAB6 and Rituxan?, the Jurkat-CD16 cells secrete IL-2, which indicates cell activation via binding of the Fc portion of the antibodies to CD16. The EMAB6 antibody, however, has an inductive activity which is much stronger than the Rituxan? antibody. Thus, at 6.25 ng/mL, the IL-2 percentages were 112% and 21% for EMAB6 and Rituxan? respectively. At 50 ng/mL, the difference is still large, with the percentages of IL-2 being 112% and 65% respectively. This difference decreases as concentration increases, with the respective percentages of IL-2 for EMAB6 and Rituxan? being 124% and 100% at 2,500 ng/mL. In this test, the Rituxan? EC50/EMAB6 EC50 ratio is estimated at 15 (graphical estimate in ng/mL for EC50 assuming that Rituxan? and EMAB6 attain the same E Max).

(97) These results confirm the ADCC results, both being CD16-dependant. They demonstrate that the binding to CD16 (Fc?RIIIA) by the Fc portion of the EMAB6 antibody is followed by a strong cellular activation which leads to the induction of effector functions.

(98) The activation of CD16 (Fc?RIIIA) induced by the chimeric EMAB603 antibody in the presence of Raji cells is comparable with that induced by the EMAB6 antibody.

(99) With lymphocytes from patients with B-CLL, the results obtained shown in FIG. 10 demonstrate that in the presence of anti-CD20 Rituxan? and EMAB6, the Jurkat-CD16 cells secrete IL-2, which indicates cell activation via binding of the Fc portion of the antibodies to CD16. The EMAB6 antibody, however, has an inductive ability which is much greater than the Rituxan? antibody. In fact, the IL-2 secretion induction activity of Rituxan? is close to the base line at concentrations of 2.5 and 25 ng/mL, whereas that of the EMAB6 antibody is significant. Thus at 25 ng/mL, the IL-2 percentages were 132% and 34% for EMAB6 and Rituxan? respectively. At the highest concentration (2,500 ng/mL), the IL-2 percentages were 148% and 100% respectively. The Rituxan? EC50/EMAB6 EC50 ratio in this test is greater than 100: it is estimated at 300 (graphical estimate in ng/mL for EC50 assuming that Rituxan? and EMAB6 attain the same E Max).

(100) In conclusion, all the tests carried out on Raji cells demonstrate that the EMAB6 and EMAB603 antibodies, unlike Rituxan?, are highly cytotoxic and induce the activation of cells which express CD16 (Fc?RIIIA), especially at low antibody concentrations. On the contrary, under the same conditions, the complement-dependent cytotoxic activity of EMAB6 decreases by about 50% compared to that of Rituxan?.

(101) These results are confirmed by the studies carried out using cells isolated from patients with B-CLL, suggesting that the EMAB6 antibody is much more cytotoxic than Rituxan? towards B lymphocytes from patients with B-CLL. The differences between the two antibodies are more marked with lymphocytes from patients with B-CLL than with the Raji cells, which demonstrates the significant therapeutic interest of EMAB6 compared to Rituxan? for this condition.

(102) The reason of this increased difference may be, amongst other, the lower antigen expression of CD20 on B lymphocytes from patients with B-CLL compared to Raji cells.

(103) By analogy with Raji cells, it may be suggested that the complement-dependent cytotoxic activity of the EMAB6 antibody towards lymphocytes from patients with B-CLL must be less than that induced by Rituxan?, thus exhibiting the advantage of being less toxic in vivo as a result of the undesirable effects associated with a strong activation of the conventional complement pathway.

Example 4: Analysis of EMAB6 and EMAB603 Glycans by HPCE-LIF

(104) The N-glycan structure of the heavy chains of the EMAB6 and EMAB603 antibodies was analysed using HPCE-LIF. The N-glycan structure of the heavy chain of Rituxan? was also analysed for comparison.

(105) For that purpose, anti-CD20 monoclonal antibodies were desalted on a Sephadex G-25 column (HiTrap Desalting, Amersham Biosciences), evaporated and re-suspended in the hydrolysis buffer of PNGase F (Glyko) in the presence of 50 mM ?-mercaptoethanol. After 16 hrs incubation at 37? C., the protein fraction was precipitated by adding absolute ethanol and the supernatant, which contained the N-glycans, was evaporated. The resulting oligosaccharides were either directly labelled using a fluorochrome: APTS (1-amino-pyrene-3,6,8-trisulphonate), or subjected to the action of specific exoglycosidases before labelling with APTS. The resulting labelled oligosaccharides were injected onto an N-CHO capillary, separated and quantified by capillary electrophoresis with laser-induced fluorescence detection (HPCE-LIF).

(106) The estimation of the fucose level was carried out either by the addition of the isolated fucosylated forms, or more specifically after the simultaneous action of neuraminidase, ?-galactosidase and N-acetylhexosaminidase, which resulted in 2 peaks corresponding to the fucosylated or non-fucosylated pentasaccharide [GlcNac2-Man3] being obtained on the electrophoretogram:

(107) TABLE-US-00010 TABLE 1 Analysis of anti-CD20 EMAB603 and Rituxan? N-glycans Anti-CD20 % Fucose % Galactose Fuc/Gal EMAB603 15 37 0.4 Rituxan? 93 57 1.63

(108) The fucose level, expressed as %, was calculated using the following formula:

(109) $\mathrm{Fucose}\mathrm{level}=\frac{\mathrm{fucosylated}\left[\mathrm{GlcNac}2\text{-}\mathrm{Man}3\right]?100}{\left[\mathrm{GlcNac}2\text{-}\mathrm{Man}3+\mathrm{fucosylated}\mathrm{GlcNac}2\text{-}\mathrm{Man}3\right]}$

(110) The galactose level, expressed as %, was calculated by adding the percentages of the oligosaccharide forms containing galactose obtained after the action of neuraminidase and fucosidase. The formula used is as follows:
Galactose level=(G1+G1B)+2x(G2+G2B)

(111) The fucose/galactose ratio is obtained by dividing the fucose level by the galactose level, calculated as described above.

(112) From this analysis (see Table 1), it appears that the EMAB6 and EMAB603 antibodies are little fucosylated (% fucose<25%) compared to Rituxan? (% fucose=93%). In addition, the Fuc/Gal ratio (fucose/galactose ratio) for EMAB6 and EMAB603 is low (Fuc/Gal ratio<0.6), unlike the antibodies expressed in CHO cells such as Rituxan? (Fuc/Gal ratio=1.63).