METHOD FOR PRODUCING VARIANTS HAVING AN FC WITH IMPROVED SIALYLATION

Abstract

The present invention relates to a method for increasing the sialylation of an Fc fragment, including a mutation step of at least one amino acid selected from among the amino acids in positions 240 to 243, 258 to 267 and 290 to 305 of the Fc fragment, the numbering being that of the EU index or equivalent in Kabat. The present invention also relates to a method for producing a variant of a parent polypeptide including an Fc fragment, the variant having improved sialylation of the Fc fragment relatively to the parent polypeptide, which includes a mutation step of at least one amino acid selected from among the amino acids in positions 240 to 243, 258 to 267 and 290 to 305 of the Fc fragment, the numbering being that of the EU index or equivalent in Kabat.

Claims

1-18. (canceled)

19. A method for increasing the sialylation of an Fc fragment, comprising a mutation step of at least one amino acid selected from among amino acids in positions 240 to 243, 258 to 267 and 290 to 305 of said Fc fragment, the numbering being that of the EU index or equivalent in Kabat.

20. A method for producing a variant of a parent polypeptide comprising an Fc fragment, said variant having improved sialylation of said Fc fragment relatively to the sialylation of the Fc fragment of the parent polypeptide, which comprises a mutation step of at least one amino acid selected from among the amino acids in positions 240 to 243, 258 to 267 and 290 to 305 of said Fc fragment, the numbering being that of the EU index or equivalent in Kabat.

21. The method according to claim 19, wherein the sialylation of said Fc fragment is increased by at least 10% relatively to the sialylation of said Fc fragment before the mutation step or of said Fc fragment of the parent polypeptide.

22. A method for producing a variant of a parent polypeptide comprising an Fc fragment, said variant having at least one effector activity mediated by said Fc fragment reduced relatively to the effector activity of the parent polypeptide, further comprising a mutation step of at least one amino acid selected from among the amino acids in positions 240 to 243, 258 to 267 and 290 to 305 of said Fc fragment, the numbering being that of the EU index or equivalent in Kabat.

23. The method according to claim 22, wherein the effector activity mediated by the Fc fragment is selected from among cell cytotoxicity dependent on antibodies (ADCC), cytotoxicity depending on the complement (CDC) and cell phagocytosis dependent on the antibodies (ADCP).

24. The method according to claim 22, wherein the variant is without any effector activity mediated by the Fc fragment.

25. A method for producing a variant of a parent polypeptide comprising an Fc fragment, said variant having an affinity mediated by said Fc fragment reduced relatively to the affinity of the parent polypeptide, for at least one of the receptors of the Fc region (FcR), and the method comprising a mutation step of at least one amino acid selected from among the amino acids in positions 240 to 243, 258 to 267 and 290 to 305 of said Fc fragment, the numbering being that of the EU index or equivalent in Kabat.

26. The method according to claim 25, wherein the variant has an affinity mediated by said Fc fragment reduced relatively to the affinity of the parent polypeptide, for the receptor FcgRIIIa (CD16a) and the receptor FcgRIIa (CD32a).

27. The method according to claim 19, wherein the mutation is selected from among an insertion, a substitution, and a deletion.

28. The method according to claim 19, wherein the mutation is carried out on at least one amino acid located in position 240, 241, 242, 243, 258, 259, 260, 261, 262, 263, 264, 265, 266, 267, 290, 291, 292, 293, 294, 295, 296, 297, 298, 299, 300, 301, 302, 303, 304 or 305, the numbering being that of the EU index or equivalent in Kabat.

29. The method according to claim 19, wherein the mutation is carried out on at least one amino acid located in position 240, 241, 242, 243, 258, 259, 260, 261, 263, 265, 266, 267, 290, 291, 292, 293, 294, 295, 296, 298, 299, 300, 301, 302, 303, 304 or 305.

30. The method according to claim 19, wherein the mutation is carried out on the amino acid located in position 293 or 294, the numbering being that of the EU index or equivalent in Kabat.

31. The method according to claim 19, wherein the Fc fragment of the parent polypeptide already comprises at least one additional mutation, selected from among a combination of additional mutations selected from among P230S/N315D/M428L/N434Y, T256N/A378V/S383N/N434Y, V2591/N315D/N434Y and N315D/A330V/N361D/A378V/N434Y.

32. The method according to claim 20, wherein the parent polypeptide consists in an Fc fragment.

33. The method according to claim 20, wherein the parent polypeptide consists in a sequence of amino acids fused in the N- or C-terminal to an Fc fragment.

34. The method according to claim 20, wherein said parent polypeptide is an immunoglobulin or an antibody.

35. The method according to claim 19, wherein the Fc fragment of the parent polypeptide is an Fc fragment of an IgG1, corresponding to the sequence SEQ ID NO: 1.

36. The method according to claim 20, wherein the mutation step is obtained as follows: i) a nucleic acid sequence coding for the parent polypeptide comprising the Fc fragment is provided; ii) the nucleic sequence provided in i) is modified in order to obtain a nucleic sequence coding for the variant; and iii) the nucleic sequence obtained in ii) is expressed in a host cell, and the variant is recovered.

37. The method of claim 25, wherein the at least one of the receptors is selected from among the complement C1q and the receptors FcgRIIIa (CD16a), FcgRIIa (CD32a), and FcgRIIb (CD32b).

38. The method of claim 27, wherein the substitution is point like.

Description

FIGURES

[0089] FIG. 1: alignments of native human IgG1 sequences referring to the positions 216 to 447 according to the EU index:

[0090] FIG. 1 shows alignments of native human IgG1 sequences referring to the positions 216 to 447 (according to the EU index) with the corresponding sequences of human IgG2 (SEQ ID NO: 2 and 7), human IgG3 (SEQ ID NO: 3 and 8) and human IgG4 (SEQ ID NO: 4 and 9). The IgG1 sequences refer to the allotype G1m1,17 (SEQ ID NO: 1 and 6) and to the allotype G1m3 (SEQ ID NO: 5 and 10). The lower hinge CH2-CH3 IgG1 domain begins at position 226 (see arrow). The CH2 domain is highlighted in grey and the CH3 domain is in italics.

[0091] FIG. 2: Half-life of anti-CD20 and anti-Rhesus D antibodies produced in YB2/0: The persistence of immunoglobulins in the serum of transgenic mice for human FcRn was evaluated; Two antigenic specificities were tested; the anti-CD20 IgGs and the deleted anti-RhD IgGs in position 294 were tested as a comparison with the corresponding IgG WT. [0092] A) Illustrates the time-dependent change in the concentration of plasma IgGs; [0093] B) Ilustrates the half-life observed for both IgGs deleted in position 294 and of the IgG WTs.

EXAMPLES

[0094] The following examples are given with view for illustrating diverse embodiments of the invention.

Example 1

Production of Variants Deleted in Position 294

[0095] The inventors analyzed the sialylation of several variants according to the invention, notably deleted in position 294 (EU index or equivalent in Kabat). From among the analyzed Del294 variants, several variants comprise a combination of additional mutations from among the combinations described for providing an optimized bond to the FcRn in patent application EP 0 233 500.

[0096] The identification and the obtaining of such FcRn optimized variants, may be accomplished according to the methods described in the prior art, in particular in the

[0097] European patent application EP 0 233 500, which describes the obtaining of such mutants according to the so-called MutaGen™ Technique.

[0098] Typically, this method comprises the following steps:

A/Building an Fc Bank

[0099] The human gene Fc coding for the residues 226 to 447 (according to the EU index of Kabat and illustrated in FIG. 1) derived from the heavy chain of a human IgG1 is cloned in a suitable vector, such as the phagemid vector pMG58 according to standard procedures well known to one skilled in the art.

B/Mutagenesis

[0100] Several banks are then generated, according to the procedure described in WO 02/038756, which uses human polymerase DNAs of low reliability with the purpose of introducing random mutations homogenously on the entire target sequence. More specifically, three distinct mutases (pol β, η and τ) were used under different conditions for generating profiles of complementary mutations.

C/Expression of the Fc Banks by Phage-Display and Selection of the Variants Having an Improved Bond to the Neonatal Receptor FcRn

[0101] The Fc banks are expressed by using the Phage-display technique according to standard procedures, for being used for selecting Fc fragments. The selection may be accomplished according to the detailed procedure in European patent application EP 2 233 500, notably by selection on FcRn in a solid or liquid phase, and then determination of the binding characteristics of the fragments to FcRn with ELISA.

D/Production of Variants as an Entire Ig and Deletion in Position 294

[0102] Several combinations of FcRn optimized mutations were selected for being used as a base for the production of mutants deleted in position 294. The following combinations were selected: [0103] N315D/A330V/N361D/A378V/N434Y (T5A-74) [0104] T256N/A378V/S383N/N434Y (C6A-78) [0105] V2591/N315D/N434Y (C6A-74)

1—Production of IgG Variants in HEK Cells

[0106] The Fc fragment sequence SEQ ID NO: 1 was cloned in a generic eukaryotic expression vector derived from pCEP4 (Invitrogen) and containing the heavy chain of an anti-CD20 chimeric antibody according to standard PCR procedures. The lightweight chain of this antibody was inserted into a similar pCEP4-derived vector. All the mutations of interest in the Fc fragment were inserted into the expression vector containing the anti-CD20 heavy chain by overlap PCR. For example, the variant 294Del was obtained by using two sets of primers adapted for integrating the deletion in position 294 on the heavy chain contained in the expression vector.

[0107] The thereby obtained fragments by PCR were associated and the resulting fragment was amplified by PCR by using standard procedures. The PCR product was purified on 1% agarose gels (w/v), digested with the suitable restriction enzymes and cloned in the expression vector of the anti-CD20 heavy chain.

[0108] The HEK 293 cells were co-transfected with the expression vectors of the lightweight chain and of the heavy chain of the anti-CD20 IgG in equimolar amounts according to standard procedures (Invitrogen). The cells were cultivated so as to produce antibodies in a transient way. The produced antibodies were able to be isolated and purified according to current techniques of the art, with view to their characterization.

2—Production of IgG Variants in YB2/0 Cells

[0109] The Fc variants were prepared in an entire IgG format in the cell line YB2/0 (ATCC, CRL-1662) with the anti-CD20 and anti-RhD specificity. For this, the heavy and lightweight chain of the IgGs were cloned in a bicistronic HKCD20 vector optimized for production in YB2/0. The production was made in stable pools of YB2/0 cells. The production steps by cell cultivation and cultivation for purifying the antibodies were carried out according to current techniques of the art, with view to their characterization.

[0110] The inventors verified that the deletion in position 294 did not have any significant impact on the binding to FcRn. The variants deleted in position 294 preserve their binding to FcRn relatively to the parent IgG (IgG WT or IgG comprising FcRn optimized mutations).

Example 2

Analysis of the Sialylation of Different Proteins

Operating Method: Preparation of the Sample

1 Desalting and N-Deglycosylation

[0111] In a first phase, the sample to be analyzed was salted out according to standard procedures so as to remove all the potentially present free reducing carbohydrates as well as the substances which may interfere during the subsequent steps (salts and excipients). After salting out, the sample was dried and then the glycans were released by enzymatic action of N-Glycanase under denaturation and reducing conditions, in order to maximize the yield of N-deglycosylation. For the N-deglycosylation of the Igs, the dry sample was taken up with 45 μL of the digestion solution PNGase F diluted to 1/5. 1.5 μL of a 10% (v/v) β-mercaptoethanol in ultra-pure water was added with stirring and incubation for 15 minutes at room temperature. Next, 1 μL of the PNGase F solution (2.5 mU/μL) was added before stirring and incubation in a water bath at 37° C. for 12 to 18 hours. Next the glycans were separated from the deglycosylated proteins by precipitation with cold EtOH.

[0112] The obtained glycan extract was then distributed into 4 fractions before being treated with exoglycosidases.

2 Quantification of the Fucosylation and Intercalating GlcNAc Levels, and of the Galactosylation Index of N-Glycans

[0113] Each dried alcoholic sub-fraction N, containing the equivalent of 100 μg of glycoprotein, were respectively digested (1) with α-sialidase, β-galactosidase and N-acetyl-β-hexosaminidase, in order to determine the fucosylation level; (2) with α-sialidase, β-galactosidase and α-fucosidase, for calculating the level of intercalating GlcNAc; and (3) with α-sialidase and α-fucosidase, for determining the galactosylation index.

[0114] These deglycosylations were carried out at 37° C. for 12 to 18 hours.

[0115] The isolation of the products of exoglycosidase degradations was achieved by cold alcohol extraction by adding 60 μL (3 volumes) of absolute ethanol equilibrated at −20° C., before stirring and then incubation at −20° C. for 15 minutes. Centrifugation at 10,000 rpm was achieved for 10 minutes at +4° C., and the supernatant was immediately transferred into a microtube of 0.5 mL before being dried in vacuo. The obtained oligosaccharides were then marked with a fluorochrome, APTS, and then separated and quantified in HPCE-LIF.

3 Utilization of the Results

[0116] The identification of the N-glycan peaks was achieved by means of a reference glycoprotein standard, the N-glycosylation of which is perfectly known, by comparison of the migration times of its N-glycans with those of the species observed on the electrophoretic profiles of the samples to be analyzed. Further, the migration times of the oligosaccharides standards are converted into glucose units (GUs) after analysis of a heterogeneous mixture of a glucose homopolymer (Glc ladder). These values of GUs will then be compared with those of a few standard oligosaccharides of known GUs, and will give the possibility of increasing the confidence level of the identifications.

Results

[0117] A-Variants produced in YB2/0

[0118] The following polypeptides were analyzed:

TABLE-US-00002 Name Mutations Anti-CD20 Del294 Del294 Anti-CD20-C6A_78-Del294 T256N/A378V/S383N/N434Y/Del294 Anti-CD20-C6A_74-Del294 V259I/N315D/N434Y/Del294 Anti-RhD — Anti-RhD Del294 Del294 Anti-RhD-C6A_78 T256N/A378V/S383N/N434Y Anti-RhD-C6A_78-Del294 T256N/A378V/S383N/N434Y/Del294

Anti-CD20 Del294:

[0119] The electropherograms obtained show biantenna glycan structures. These structures are in majority sialylated.

[0120] 87.98% of the structures seem to be sialylated. The calculated fucosylation level is 48.24%.

TABLE-US-00003 A2 11.9 A2F 19.8 A1 5.5 A1F 1.6 G0 0.84 G0B 0.96 G1(1.6) + G0F 0.53 G1(1.3) + G0BF 0.19 G1(1.6)B 2.44 G1(1.6)F 0.14 G2 + G1(1.3)F 2.09 G2B 0.51 G2F 0.16 G2FB 3.84 Sialylated structures unidentified 49.18 % tage of sialylated structures: 87.98 Fucosylation % tage: 48.24

Anti-CD20-C6A 78-Del294:

[0121] The obtained electrophoregrams show biantenna glycan structures. These structures are in majority sialylated.

[0122] 88.69% of the structures seem to be sialylated. The calculated fucosylation level is 51.87%

TABLE-US-00004 A2 12.94 A2F 20.61 A1 7.52 A1F 3.12 G0 1.57 G0B 0.61 G1(1.6) + G0F 1.38 G1(1.3) + G0BF 0.58 G1(1.6)B 1.99 G1(1.6)F 0.51 G2 + G1(1.3)F 3.56 G2B 0.26 G2F 0.6 G2FB 0 NI sialylated structures 44.5 % age of sialylated structures: 88.69 Fucoslylation % age: 51.87

Anti-CD20-C6A 74-Del294:

[0123] The obtained electrophoregrams show biantenna glycan structures. These structures are in majority sialylated.

[0124] 93.48% of the structures seem to be sialylated. The calculated fucosylation level is 51.24%.

TABLE-US-00005 A2 13.39 A2F 23.01 A1 5.57 A1F 1.93 G0 0.59 G0B 0.73 G1(1.6) + G0F 0.43 G1(1.3) + G0BF 0.18 G1(1.6)B 2 G1(1.6)F 0.11 G2 + G1(1.3)F 2.05 G2B 0.15 G2F 0.1 G2FB 0 NI sialylated structures 49.58 % age of sialylated structures 93.48 Fucoslylation % age: 51.24

Anti-RhD:

[0125] The obtained electrophoregrams show biantenna glycan structures in majority consisting of short non-fucosylated agalactosylated short structures (G0: 52.06%). The fucosylated structures are a minority. A few structures having a GlcNac in a bissecting position (GOB, GOFB) are observed.

[0126] The predominant oligosaccharide structure is: G0 (52.06%). The calculated fucosylation level is 17.05%, the fucosylation level obtained with the run DSial+DGal+DhexNAc (*) is 13.07%. The level of forms having a bissecting GlcNac is 2.87%. The calculated galactosylation level is 40.5%.

TABLE-US-00006 Structure (%) HPCE-LIF Sialylated 0.00 Mono-sialylated 0.00 Bi-sialvlated 0.00 Bissecting 2.87 Fucosylated* 13.07 Fucosylated 17.05 A2 0.00 A2F 0.00 M3N2 0.00 M3N2F 0.00 A1 0.00 A1F 0.00 G2FB 0.00 G2F 0.47 G2B 0.00 G2 4.66 G1FB 0.00 G1F 5.65 G1(1.3)FB 0.00 G1(1.6)FB 0.00 G1(1.3)F 0.00 G1(1.6)F 5.65 G1B 0.00 G1 24.59 G1(1.3)B 0.00 G1(1.6)B 0.00 G1(1.3) 4.00 G1(1.6) 20.59 G0FB 1.24 G0F 9.69 G0B 1.63 G0 52.06 MAN-5 0.00 Identified (%) 99.99

Anti-RhD Del294:

[0127] The electrophoregrams obtained show biantenna glycan structures and a few triantenna structures. These structures are in majority sialylated.

[0128] 92.25% of the structures seem to be sialylated. The calculated fucosylation level is 37.08%.

TABLE-US-00007 A2 14.05  A2F 20.61  A1 6.79 A1F 1.59 G0 0.63 G0B 0.82 G1(1.6) + G0F 0   G1(1.3) + G0BF 0.72 G1(1.6)B 2.86 G1(1.6)F 0   G2 + G1(1.3)F 2.72 G2B 0   G2F 0   G2FB 0   NI sialylated structures 49.21  % age of sialylated structures 92 25    Fucoslylation % age: 37.08 

Anti-RhD -C6A 78:

[0129] The predominant oligosaccharide structure is: G0 (55.20%). The calculated fucosylation level is 12.37%, the fucosylation level obtained with the run DSial+DGal+DhexNAc (*) is 10.63%. The level of forms having a bissecting GlcNac is 2.27%. The calculated galactosylation level is 39.13%.

TABLE-US-00008 Structure (%) HPCE-LIF Sialylated 0.00 Mono-sialvlated 0.00 Bi-sialylated 0.00 Bissecting 2.27 Fucosylated* 10.63 Fucosylated 12.37 A2 0.00 A2F 0.00 M3N2 0.00 M3N2F 0.00 A1 0.00 A1F 0.00 G2FB 0.00 G2F 0.00 G2B 0.00 G2 4.00 G1FB 0.00 G1F 4.20 G1(1.3)FB 0.00 G1(1.6)FB 0.00 G1(1.3)F 0.00 G1(1.6)F 4.20 G1B 0.83 G1 26.10 G1(1.3)B 0.00 G1(1.6)B 0.83 G1(1.3) 5.67 G1(1.6) 20.43 G0FB 0.66 G0F 7.51 G0B 1.50 G0 55.20 MAN-5 0.00 Identified (%) 100.00

Anti-RhD -C6A 78-Del294:

[0130] The electrophoregrams obtained show biantenna glycan structures and a few triantenna structures. These structures are in majority sialylated.

[0131] 91.83% of the structures seem to be sialylated. The calculated fucosylation level is 57.81%.

TABLE-US-00009 A2 14.6 A2F 20.57 A1 7.72 A1F 3.38 G0 1.15 G0B 1.75 G1(1.6) + G0F 0 G1(1.3) + G0BF 0.62 G1(1.6)B 3.14 G1(1.6)F 0 G2 + G1(1.3)F 1.53 G2B 0 G2F 0 G2FB 0 NI sialylated structures 45.56 % tage of sialylated structures: 91.83 Fucosylation % tage: 57.81

B-Variants Produced in the HEK Lines

[0132] The following polypeptides were analyzed (Anti-CD20 IgG variants):

TABLE-US-00010 Name Mutations T5A-74 N315D/A330V/N361D/A378V/N434Y T5A-74H V264E/N315D/A330V/N361D/A378V/N434Y T5A-74Del294 E294del/N315D/A330V/N361D/A378V/N434Y WT /

[0133] The variant T5A-74H differs from the parent variant T5A-74 by the mutation V264E.

[0134] The mutant T5A-74De1294 differs from the parent variant T5A-74 by deletion of the amino acid in position 294.

[0135] The profile of glycosylation of these variants was subsequently analyzed. The results are summarized in the table hereafter (in percentages):

TABLE-US-00011 T5A-74 T5A-74H T5A-74Del294 WT A1 0 1.8 5.64 0 A1F 0 6.72 9.18 0 Sialylated unidentified 0 29.28 19.56 0 G0 0 2.37 0 3.34 G0B 2.13 0.65 0 1.77 G1(1.6) 0 1 0 0 G0F 81.44 10.27 28.76 79.47 G1(1.3) 0 1.21 0 0 G0FB 1 0 4.92 0.88 G1(1.6)B 0.51 0 0 0.71 G1(1.6)F 10.08 12.15 6.64 9.23 G2 0 1.85 0 0 G1(1.3)F 4 3.26 10.63 3.68 G1(1.6)FB 0 0 0 0 G2B 0 6.41 2 0 G2F 0.84 5.57 5.94 0.91 G2FB 0 1.65 1.57 0 Galactosylation 16.27 >58.36 >51.11 15.44 Sialylation 0 >37.8 >34.38 0 Fucosylation 97.36 >40.88 >67.64 94.17

Example 3

Analysis of the Half-Life of IgGs Deleted in Position 294

[0136] The persistence of immunoglobulins in the serum of transgenic mice for the human FcRn was evaluated. Two antigenic specificities were tested; the anti-CD20 IgGs and the anti-RhD IgGs deleted in position 294 were tested in comparison with the corresponding IgG WTs.

[0137] Pharmacokinetic experiments were thereby conducted in hFcRn mice which are homozygotes for an allele KO of the murine and heterozygote FcRn for a transgene of human FcRn (mFcRn.sup.−/− hFcRnTg).

[0138] For these pharmacokinetic studies, each animal received a single intravenous injection of IgG at 5 mg/kg at the retro-orbital sinus, in a procedure similar to the one described earlier (Petkova SB, et al. Enhanced half-life of genetically engineered human IgG1 antibodies in a humanized FcRn mouse model: potential application in humorally mediated autoimmune disease. Int Immunol 2006). Blood samples were taken from the retro-orbital sinus at multiple points of time and the IgGs titrated with ELISA.

Results:

[0139] In this test, both IgGs deleted in position 294 showed an increase in the half-life with a ratio (variant half-life/WT) of 1.7 (FIG. 2).

[0140] The analyzed parameters are grouped in the table below:

TABLE-US-00012 removal C0 AUC0-t AUCinf half-life Vd Cl Molecules (μg/mL) (h .Math. μg/mL) (h .Math. μg/mL) (h) (mL) (mL/h) Anti-CD20 WT 121 1423 1425 29.2 4.07 0.097 Anti-CD20 Del294 120 3482 3673 48.8 2.64 0.037 Anti-RhDWT 150 7491 8030 65.7 1.62 0.017 Anti-RhD Del294 141 11142 13354 111 1.66 0.010
The analyzed parameters are defined below: [0141] C0: Maximum concentration extrapolated to T0 [0142] AUCO−t: Area under the time/plasma concentration curve (from time T0 to the last time [0143] t where the antibody is still quantifiable) [0144] AUCinf: Area under the time/plasma concentration curve from T0 to infinity (=AUCO−t+extrapolation up to infinity) [0145] T½: Half-life [0146] Vd: Distribution volume [0147] Cl: Clearance

Example 4

Production of Additional Fc Variants by Directed Mutagenesis and Binding Tests on the Fc Receptors:

1. Building Fc Variants:

[0148] Each mutation of interest in the fragment Fc was independently inserted into an expression vector containing the anti-CD20 heavy chain by overlap PCR by using two sets of primers adapted for integrating a deletion or a degenerate codon (NNN or NNK) in the targeted position (240 to 243, 258 to 267, 290 to 305). The fragments thereby obtained by PCR were associated and the resulting fragment was amplified by PCR by using standard procedures. The PCR product was purified on 1% (w/v) agarose gels, digested with suitable restriction enzymes and cloned in the eukaryotic expression vector pMGM05-CD20 (pCEP4 InvitroGen), which contains cloning sites for the Fc fragment (BamHl and Notl) and the variable chain VH of the anti-CD20 antibody. This construct causes mutation of two amino acids in the Fc (aa224 and 225, HT changed to GS) and the addition of the EFAAA sequence at the C-terminal of the Fc, but gives the possibility of very rapidly testing a very large number of clones. In a first phase, it was verified that these mutations do not modify the binding of IgG-WT to the different receptors. Subsequently, positive controls were cloned in this system in order to validate it: [0149] IgG1-5239D, 1332E, from the anti-CD19 antibody XmAb5574 from Xencor (C1): positive control for CD16a; [0150] IgG1-G236A, from Xencor (C4): positive control for CD32aH/R; [0151] IgG1-K326W, E333S, from Abgenix/Genentech (C3): positive control for C1q; and [0152] IgG1-5267E, L328F, from the anti-CD19 antibody XmAb5574 from Xencor (C5): positive control for CD32b.

[0153] The DNA of the isolated clones was sequenced after PCR on colonies. After bio-computer analyses, the clones including new mutations were frozen to −80° C. into a bacterium XL1-Blue and the sequences included in our database. Thus, 268 variants were built.

2. Production of IgG Variants in HEK293 Cells:

[0154] The light chain of the anti-CD20 was inserted into a pCEP4 vector identical with the vector used for the heavy chain, noted as pMGM01-CDC20 (pCEP4 InvitroGen). HEK293-F Freestyle™ (Invitrogen) cells, cultivated in 24-well plates, were co-transfected with the vectors pMGM01-CD20 and pMGM05-CD20 (Fc-WT and variants) in equimolar amounts (250 ng/ml) with a transfection reagent (1 μl/ml) by using standard procedures (Invitrogen). The cells were cultivated suspended in a medium without any serum for 7-9 days post-transfection and the supernatants (1 ml) containing IgGs were harvested after centrifugation of the cells at 100 G for 10 min. The IgGs secreted in the supernatants were quantified by using an ELISA test (FastELISA, R&D biotech).

3. Recombinant Fc Receptors Used:

[0155] CD16a is an activator receptor which has a polymorphism V/F in position 158, on the binding site to Fc. The affinity is better for CD16aV. CD16aV is commercially available (R&D system).

[0156] CD32a is an activator receptor which has a polymorphism H/R in position 131, on the binding site to Fc. The affinity is better for CD32aH. CD32aH was produced by PX′Therapeutics. CD32aR and CD32b are commercially available (R&D system).

4. ELISA Tests of IgG Variants Produced in the Supernatants of HEK293-F Cells:

[0157] The IgG variants were tested for their binding to several human FcRs and to the FcRn with ELISA. Maxisorp immunoplates were coated with 0.1 μg of CD32aH/well, or 0.2 μg CD16aV/well in PBS or 0.25 μg of FcRn in P6 (sodium phosphate 100 mM, sodium chloride 50 mM pH6.0). NiNTA plates (HisGrab Pierce) were coated with 0.05 μg of CD32aR/well, or 0.2 μg CD32b/well in PBS. After coating overnight at 4° C., the plates were washed twice with PBS (or P6)/0.05% Tween-20 and saturated with PBS/4% of BSA (or P6 4% in skimmed-milk) for 2 hours at 37° C. In parallel, the supernatants were diluted in PBS (or P6 for the test on FcRn) at a final concentration of 0.5 μg of IgG/ml and mixed with F(ab′)2 of anti-human goat HRP IgG at the same concentration for 2 hours at room temperature. The IgGs aggregated with the F(ab′)2 were then incubated with gentle stirring for 1 hour at 30° C. on saturated ELISA plates without any dilution for CD16aV, CD32aR and CD32b (i.e. IgG at 0.5 μg/ml), diluted in PBS at 0.25 pg/ml for CD32aH and diluted in P6 at 0.035 μg/ml for FcRn. The plates are then developed with TMB (Pierce) and the absorbence is read at 450 nm.

[0158] By means of this ELISA test, the built variants were tested in comparison with the wild type Fc (Fc-WT) and their variant/Fc-WT ratio was calculated, as indicated in tables 1 to 3.