Antibody Fc variants for increased blood half-life

Abstract

The present invention relates to a polypeptide including an Fc variant produced by substituting a portion of the amino acid sequence of the Fc domain of a human antibody with a different amino acid sequence. The present invention also relates to an antibody including the polypeptide. The Fc variant can find application in a wide range of antibodies and Fc-fusion constructs. In one aspect, the antibody or Fc fusion construct of the present invention is a therapeutic, diagnostic or laboratory reagent, preferably a therapeutic reagent. The Fc variant is suitable for use in the treatment of cancer because its in vivo half-life can be maximized by optimization of the portion of the amino acid sequence. The antibody or Fc fusion construct of the present invention is used to kill target cells that bear a target antigen, for example cancer cells. Alternatively, the antibody or Fc fusion construct of the present invention is used to block, antagonize or agonize a target antigen. For example, the antibody or Fc fusion construct of the present invention may be used to antagonize a cytokine or a cytokine receptor.

Claims

1. A polypeptide comprising a human antibody with Fc variant wherein the Fc variant consists of, as amino acid substitutions, M428L and Q311R or M428L and L309G according to the Kabat EU numbering system in the Fc domain of the wild-type human antibody, and wherein the Fc variant has an increased half-life of the human antibody compared to the wild type.

2. The polypeptide according to claim 1, wherein the antibody is an IgG antibody.

3. The antibody according to claim 1, wherein the antibody is a monoclonal antibody, a bispecific antibody, an antibody conjugate, or a human antibody.

4. A composition comprising the polypeptide according to claim 1, an antibody comprising the polypeptide, a nucleic acid molecule comprising the polypeptide or vector comprising the nucleic acid molecule.

5. The composition according to claim 4, wherein the composition increases the blood half-life of the antibody for therapy in vivo.

6. The composition according to claim 4, wherein the composition is a pharmaceutical composition for treating cancer.

7. The composition according to claim 6, wherein the antibody recognizes a cancer antigen.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 shows expression vectors for expression and purification of tetrameric FcRn and dimeric FcRn and SDS-PAGE gels after purification.

(2) FIG. 2 is a schematic diagram of a library constructed such that 18 amino acids are contained at positions M252 and M428.

(3) FIG. 3 shows a 2M library search process and a sorted M428L variant.

(4) FIG. 4 schematically shows an error library and a point library constructed based on M428L.

(5) FIG. 5 shows FACS fluorescence intensities of variants sorted from (5a) an error library and (5b) a point library.

(6) FIG. 6 shows plasmids for expressing trastuzumab heavy and light chains in animal cells.

(7) FIG. 7 shows expression and purification results for wild-type trastuzumab.

(8) FIG. 8 compares the physical properties of commercial trastuzumab with those of in-house trastuzumab (a: CE-cIEF, b: SEC).

(9) FIG. 9 compares the physical properties of commercial trastuzumab with those of in-house trastuzumab by N-glycan profiling.

(10) FIG. 10 shows expression and purification results for 10 trastuzumab Fc variants (a: affinity chromatography, b: SDS-PAGE analysis, c: list of final yield).

(11) FIG. 11 shows SEC characterization results for trastuzumab Fc variants.

(12) FIG. 12 shows binding forces of trastuzumab Fc variants to FcRn, which were measured by ELISA.

(13) FIG. 13 shows binding forces of trastuzumab Fc variants to hFcRn at pH values of 6.0 and 7.4, which were measured using a BiaCore instrument (a: pH 6.0 (capture method) b: pH 7.4 (avid format)).

(14) FIG. 14 compares the pharmacokinetics of commercial trastuzumab with those of in-house trastuzumab in regular mice (C57BL/6J (B6)) and human FcRn Tg mice.

(15) FIG. 15 shows the results of pharmacokinetic analysis for Fc variants in human FcRn Tg mice (after intravenous injection of the variants (5 mg/kg each), n=5).

(16) FIG. 16 shows binding forces of trastuzumab Fc variants to FcγRs, which were measured by ELISA.

(17) FIG. 17 compares the effector functions of trastuzumab Fc variants with those of normal IgG and trastuzumab as a control group (ADCC assay).

(18) FIG. 18 compares the effector functions (ADCC) of trastuzumab Fc variants.

(19) FIG. 19 shows binding forces of trastuzumab Fc variants to C1q, which were measured by ELISA.

BEST MODE FOR CARRYING OUT THE INVENTION

(20) The present invention will be explained in more detail with reference to the following examples. It will be evident to those skilled in the art that the scope of the present invention is not limited by these examples according to the gist of the present invention.

EXAMPLES

Example 1: Expression and Purification of Neonatal Fc Receptor (FcRn) for Searching Library of Fc Variants

(21) Tetrameric FcRn and dimeric FcRn for searching Fc variants with improved pH-dependent binding force to FcRn were expressed and purified. To this end, expression vectors were prepared (FIG. 1). pMAZ-β2microglobulin-GSlinker-FcRnα-chain-streptavidin-His was constructed as a DNA plasmid to obtain tetrameric FcRn. The DNA was co-transfected into HEK 293F cells and temporarily expressed at a level of 300 ml. The resulting culture medium was centrifuged at 7000 rpm for 10 min. The collected supernatant was equilibrated with 25×PBS and filtered with a 0.2 m bottle top filter (Merck Millipore). After equilibration with PBS, FcRn was allowed to bind to Ni-NTA resin (Qiagen) at 4° C. for 16 h. The FcRn-bound resin was loaded onto a column and the column was eluted with 50 ml of wash-1 buffer (PBS), 25 ml of wash-2 buffer (PBS+10 mM imidazole), 25 ml of wash-3 buffer (PBS+20 mM imidazole), and 200 μl of wash-4 buffer (PBS+250 mM imidazole) to remove proteins other than tFcRn. Then, 2.5 ml of elution buffer (PBS+250 mM imidazole) was allowed to flow through the column to obtain tFcRn. The buffer was replaced with a new one using centrifugal filter units (Merck Millipore). Dimeric FcRn was obtained from pcDNA-FcRnα-chain-GST-32 microglobulin plasmid, which was received from the University of Oslo. The DNA was co-transfected into HEK 293F cells and temporarily expressed at a level of 300 ml. The resulting culture medium was centrifuged at 7000 rpm for 10 min. The collected supernatant was equilibrated with 25×PBS and filtered with a 0.2 m bottle top filter (Merck Millipore). After equilibration with PBS, FcRn was allowed to bind to Glutathione Agarose 4B (incospharm) at 4° C. for 16 h. The FcRn-bound resin was loaded onto a column and the column was eluted with 10 ml of wash buffer (PBS) to remove proteins other than dFcRn. Then, 2.5 ml of elution buffer (50 mM Tris-HCl+10 mM GSH pH 8.0) was allowed to flow through the column. The buffer was replaced with a new one using centrifugal filter units 3K (Merck Millipore). The sizes of the tetrameric FcRn and dimeric FcRn after purification were determined using SDS-PAGE gels (FIG. 1). The purified tetrameric FcRn and dimeric FcRn were fluorescently labeled with Alexa 488 for fluorescence detection.

Example 2: Construction of 2M Library of Fc Variants

(22) pMopac12-NlpA-Fc-FLAG was constructed from the gene (SEQ ID NO: 29) of the Fc domain of trastuzumab using SfiI restriction enzyme. Based on the vector, library inserts were constructed using pMopac12-seq-Fw, Fc-M252-1-Rv, Fc-M252-2-Rv, Fc-M252-3-Rv, Fc-M428-Fw, Fc-M428-1-Rv, Fc-M428-2-Rv, Fc-M428-3-Rv, Fc-M428-frg3-Fw, and pMopac12-seq-Rv primers such that two Met residues in the Fc were substituted with 18 different amino acids except Cys and Met (Table 1 and FIG. 2). The inserts were treated with SfiI restriction enzyme and ligated with the vector treated with the same SfiI. Thereafter, the ligated inserts were transformed into E. coli Judel ((F′[Tn10(Tet.sup.r)proAB.sup.+lacI.sup.qΔ(lacZ)M15] mcrAΔ(mrr-hsdRMS-mcrBC)Φ80dlacZΔM15 ΔlacX74 deoR recA1 araD139Δ(ara leu)7697 galU galKrpsLendA1nupG) to establish a large 2M library of Fc variants (library size: 1×10.sup.9).

(23) TABLE-US-00001 TABLE 1 pMopac12-seq-Fw 5′-CCAGGCTTTACACTTTATGC-3′ Fc-M252-1-Rv 5′-CCTCAGGGGTCCGGGAGATGWAGAGGGTGTCCTTGGGTTTTGGG-3′ Fc-M252-2-Rv 5′-CCTCAGGGGTCCGGGAGATKNBGAGGGTGTCCTTGGGTTTTGGG-3′ Fc-M252-3-Rv 5′-CCTCAGGGGTCCGGGAGATCCAGAGGGTGTCCTTGGGTTTTGGG-3′ Fc-M428-Fw 5′-ATCTCCCGGACCCCTGAGG-3′ Fc-M428-1-Rv 5′-GTAGTGGTTGTGCAGAGCCTCATGGWACACGGAGCATGAGAAGACGTTCC-3′ Fc-M428-2-Rv 5′-GTAGTGGTTGTGCAGAGCCTCATGKNBCACGGAGCATGAGAAGACGTTCC-3′ Fc-M428-3-Rv 5′-GTAGTGGTTGTGCAGAGCCTCATGCCACACGGAGCATGAGAAGACGTTCC-3′ Fc-M428-frg3-Fw 5′-CATGAGGCTCTGCACAACCACTAC-3′ pMopac12-seq-Rv 5′-CTGCCCATGTTGACGATTG-3′ fC-Sub#0-Rv 5′-GTCCTTGGGTTTTGGGGGGAAG-3′ Fc-Sub#1-1-Fw 5′-CTTCCCCCCAAAACCCAAGGACNNKCTCATGATCTCCCGGACCCCTGAGGTCACATGCG-3′ Fc-Sub#1-2-Fw 5′-CTTCCCCCCAAAACCCAAGGACACCNNKATGATCTCCCGGACCCCTGAGGTCACATGCG-3′ Fc-Sub#1-3-Fw 5′-CTTCCCCCCAAAACCCAAGGACACCCTCATGNNKTCCCGGACCCCTGAGGTCACATGCG-3′ Fc-Sub#1-4-Fw 5′-CTTCCCCCCAAAACCCAAGGACACCCTCATGATCNNKCGGACCCCTGAGGTCACATGCG-3′ Fc-Sub#1-5-Fw 5′-CTTCCCCCCAAAACCCAAGGACACCCTCATGATCTCCNNKACCCCTGAGGTCACATGCG-3′ Fc-Sub#1-Rv 5′-GACGGTGAGGACGCTGACC-3′ Fc-Sub#2-1-Fw 5′-GGTCAGCGTCCTCACCGTCNNKCACCAGGACTGGCTGAATGGCAAGGAGTACAAGTGCAAGG-3′ Fc-Sub#2-2-Fw 5′-GGTCAGCGTCCTCACCGTCCTGCACNNKGACTGGCTGAATGGCAAGGAGTACAAGTGCAAGG-3′ Fc-Sub#2-3-Fw 5′-GGTCAGCGTCCTCACCGTCCTGCACCAGGACTGGNNKAATGGCAAGGAGTACAAGTGCAAGG-3′ Fc-Sub#2-Rv 5′-CACGGAGCATGAGAAGACGTTCC-3′ Fc-Sub#3-1-Fw 5′-GGAACGTCTTCTCATGCTCCGTGCTGCATNNKGCTCTGCACAACCACTACACGCAGAAGAGCCTCTCCCTG-3′ Fc-Sub#3-2-Fw 5′-GGACGTCTTCTCATGCTCCGTGCTGCATGAGGCTNNKCACAACCACTACACGCAGAAGAGCCTCTCCCTG-3′ Fc-Sub#3-3-Fw 5′-GGAACGTCTTCTCATGCTCCGTGCTGCATGAGGCTCTGCACNNKCACTACACGCAGAAGAGCCTCTCCCTG-3′ Fc-Sub#3-4-Fw 5′-GGAACGTCTTCTCATGCTCCGTGCTGCATGAGGCTCTGCACAACCACNNKACGCAGAAGAGCCTCTCCCTG-3′ ep-Fc-Fw 5′-CCAGCCGGCCATGGCG-3′ ep-Fc-Rv 5′-GAATTCGGCCCCCGAGGCCCC-3′ Primers used for cloning (SEQ ID NOS: 1-27)

Example 3: Search Against the 2M Library of Fc Variants Based on Bacterial Culture and Flow Cytometry

(24) In this example, a search was conducted against the established 2M library of Fc variants. Specifically, 1 ml of Fc variant library cells transformed into E. coli Jude 1 cells were cultured with shaking in Terrific broth (TB) medium supplemented with 2% (w/v) glucose and chloramphenicol (40 μg/mL) as an antibiotic at 37° C. and 250 rpm for 4 h. After shaking culture, the library cells were inoculated into TB medium in a ratio of 1:100 and cultured with shaking at 250 rpm and 37° C. until an OD600 of 0.5 was reached. Thereafter, culture was further performed at 25° C. for 20 min for cooling and 1 mM isopropyl-1-thio-β-D-galactopyranoside (IPTG) was added to induce expression. After completion of the culture, the collected cells were divided into equal amounts based on OD600 normalization, followed by centrifugation at 14000 rpm for 1 min. The harvested cells were resuspended in 1 ml of 10 mM Tris-HCl (pH 8.0) and washed twice by centrifugation for 1 min. Cells were resuspended in 1 ml of STE (0.5 M sucrose, 10 mM Tris-HCl, 10 mM EDTA (pH 8.0)) and centrifuged at 37° C. for 30 min to remove the outer membrane. The supernatant was discarded by centrifugation and 1 ml of Solution A (0.5 M sucrose, 20 mM MgCl.sub.2, 10 mM MOPS (pH 6.8)) was added, followed by resuspension and centrifugation. Cells were resuspended in 1 ml of a mixture of 1 ml of Solution A and 20 μl of 50 mg/ml lysozyme solution, followed by centrifugation at 37° C. for 15 min to remove the peptidoglycan layer. The supernatant was removed and cells were resuspended in 1 ml of PBS. 300 μl of the suspension was added with 700 μl of PBS and fluorescently labeled tetrameric FcγRIIIa-Alexa 488 fluor probe and centrifuged at room temperature to label the fluorescent probe with spheroplast. After the labeling, cells were washed once with 1 ml of PBS and sorting was performed by flow cytometry (S3 sortor (Bio-rad)) to collect the top 3% highly fluorescent cells. The sorted cells were resorted for higher purity. For the resorted sample, genes were amplified by PCR using Taq polymerase (Biosesang) with pMopac12-seq-Fw and pMopac12-seq-Rv primers, followed by a series of processes, including treatment with SfiI restriction enzyme, ligation, and transformation, to construct sub-libraries in which the genes of the sorted cells were amplified. A total of 2 rounds of this procedure was performed. Thereafter, the resulting 40 clones were individually analyzed and an M428L variant with higher affinity for FcRn at pH 5.8 than the wild-type Fc was sorted (FIG. 3).

Example 3: Construction of Error Library and Point Library of Fc Variants

(25) Two additional libraries were constructed using the sorted M428L as a template. First, mutations were introduced into Fc by error prone PCR to construct an error library. The library (size: 2×10.sup.8) was constructed using ep-Fc-Fw and ep-Fc-Rv primers at such an error rate that 0.3% error (2.04 bp) was contained in Fc (680 bp). Second, M428L was used as a template to construct a point library. The library was constructed using pMopac12-seq-Fw, pMopac12-seq-Rv, Fc-Sub #0-Rv, Fc-Sub #1-1-Fw, Fc-Sub #1-2-Fw, Fc-Sub #1-3-Fw, Fc-Sub #1-4-Fw, Fc-Sub #1-5-Fw, Fc-Sub #1-Rv, Fc-Sub #2-1-Fw, Fc-Sub #2-2-Fw, Fc-Sub #2-3-Fw, Fc-Sub #2-Rv, Fc-Sub #3-1-Fw, Fc-Sub #3-2-Fw, Fc-Sub #3-3-Fw, and Fc-Sub #3-4-Fw primers such that mutations were randomly introduced into selected regions where Fc were bound to FcRn (FIG. 4). Thereafter, a library of Fc variants was established by transformation into Judel in the same manner as described above.

Example 4: Search Against the Error and Point Libraries of Fc Variants Based on Bacterial Culture and Flow Cytometry and Sorting of Variants, Including PFc3, PFc29, PFc41, EFc29, EFc41, EFc82, and EFc88

(26) The above sorting and resorting procedure was performed for the additional error and point libraries constructed based on the sorted M428L. 5 rounds of sorting and resorting were repeated for the error library and only one round of sorting was performed for the point library. A group of about 100 clones from each of the two libraries were individually analyzed and Fc variants having high affinity for FcRn at pH 5.8 and low affinity for FcRn at pH 7.4 were sorted. FACS analysis revealed that EFc6, EFc29, EFc41, EFc46, EFc70, EFc90 EFc82, and EFc88 sorted from the error library showed higher fluorescence intensities at pH 5.8 than the wild-type Fc and conventional variants, including YTE from Medimmune (Gabriel J. Robbie et al., Antimicrob Agents Chemother. 2013 December; 57(12): 6147-6153) and LS from Xencor (U.S. Pat. No. 8,324,351). EFc6, EFc29, EFc41, EFc82, and EFc88 were found to show lower fluorescence intensities at pH 7.4 than LS. In addition, PFc3, PFc29, and PFc41 variants sorted from the point library showed higher fluorescence intensities at pH 5.8 than YTE and LS. PFc30 showed a lower fluorescence intensity at pH 5.8 than YTE and LS. PFc29 and PFc41 showed lower fluorescence intensities at pH 7.4 than LS. Finally, EFc6, EFc29, EFc41, EFc82, EFc88, PFc3, PFc29, and PFc41 were selected because they are expected to increase blood half-lives (Table 2 and FIG. 5).

(27) TABLE-US-00002 TABLE 2 Name of Fc variant Positions of Fc variants and substituted amino acids PFc 3 P228L/L309R/M428L/N434S (SEQ ID NO: 30) PFc 29 Q311R/M428L (SEQ ID NO: 31) PFc 41 L309G/M428L (SEQ ID NO: 32) EFc 6 P228L/V264M/L368Q/E388D/V422D/M428L/P445S (SEQ ID NO: 33) EFc 29 P228L/R292L/T359A/S364G/M428L (SEQ ID NO: 34) EFc 41 P228L/L234F/E269D/Q342L/E338D/T394A/M428L (SEQ ID NO: 35) EFc 82 P230Q/F243Y/K246E/N361S/N384I/M428L (SEQ ID NO: 36) EFc 88 P230S/Q295L/K320M/D356E/F405I/M428L (SEQ ID NO: 37) Point mutations of the sorted variants

(28) The positions of the mutations are numbered according to the Kabat EU numbering system, as described in Kabat et al., “Sequences of Proteins of Immunological Interest”, 5th Ed., U.S. Department of Health and Human Services, NIH Publication No. 91-3242, 1991).

Example 5: Production and Purification of Control Trastuzumab for Introduction of the Fc Variants

(29) Trastuzumab (Herceptin®), a representative IgG1 therapeutic antibody, was selected as a control group. In the subsequent examples, the sorted Fc variants were introduced into trastuzumab.

(30) The heavy and the light chain variable regions of wild-type trastuzumab were synthesized (Genscript) from the corresponding amino acid sequences from DrugBank Online (https://go.drugbank.com/) through mammalian codon optimization simultaneously with back-translation. The synthesized trastuzumab heavy and light chain genes were sub-cloned into pOptiVEC-Fc and pDNA3.3 vectors, respectively (FIG. 6). Animal cell expression plasmids encoding the trastuzumab heavy and light chains were prepared, expressed in HEK293 cells, and purified.

(31) After culture in HEK 293F, the wild-type trastuzumab was purified by Protein A affinity chromatography (AKTA prime plus, cat #11001313) and gel permeation chromatography (HiTrap MabselectSure, GE, cat #11-0034-95). 7.7 mg of the wild-type trastuzumab was obtained in high purity from 300 ml of the culture medium (FIG. 7).

Example 6: Analysis and Comparison of Physical Properties of in-House Trastuzumab and Commercial Trastuzumab

(32) Unlike commercial trastuzumab produced by suspension culture in CHO cells, in-house trastuzumab was produced in HEK 293. Two basic characteristics of commercial trastuzumab and in-house trastuzumab antibodies were analyzed before introduction and function analysis of the sorted Fc variants. The pI values and charge variants of the samples were analyzed by capillary electrophoresis (CE: PA800 Plus, Beckman coulter) using Pharmalyte 3-10 carrier ampholytes (GE Healthcare, 17-0456-01) establishing a pH gradient of 3-10. The analytical results showed that no impurities were detected by size exclusion chromatography (SEC, Tskgel G3000swxl, Tosoh). For the commercial trastuzumab, the pI values by charge variants were 8.27-8.74 and the pI of the main peak was 8.62. For the in-house trastuzumab, the pI values by charge variants were 8.29-8.78 and the pI of the main peak was 8.65, which were almost the same as those for the commercial trastuzumab (FIG. 8). The pI values were measured by capillary electrophoresis (CE, PA800 Plus, Beckman coulter). However, cIEF analysis revealed that there were slight differences in the content of the in-house trastuzumab at the main peak and the other peaks. These differences are because of the different glycan patterns of the in-house trastuzumab produced in HEK293 cell line and the commercial trastuzumab produced in CHO cell line, leaving a possibility that the in-house trastuzumab might be oxidized by sialic acid. Thus, glycan analysis was also performed (FIG. 9).

(33) After cleavage of N-glycan from the protein with PNGase F (NEB, 186007990-1) and labeling with RapiFluor-MS reagent (Waters, 186007989-1), glycan analysis was performed using a UPLC system (Acquity UPLC I class, Waters, FLR detector). As a result of the glycan analysis, the glycan patterns were similar but different glycan contents of the compositions were observed, which seems to be not caused by sialic acid-induced oxidation but by the different production cell lines. Further, glycans were found top have no significant influence on binding force analysis and pharmacokinetic analysis (data not shown). Thus, the sorted Fc variants were introduced into the commercial trastuzumab and the in-house trastuzumab.

Example 7: Production and Purification of the Fc Variants and Analysis of Physical Properties of the Fc Variants

(34) Five control variants, including the commercial wild-type variant, the in-house wild-type variant, LS (XenCor), YTE (MedImmune), and 428L, and the sorted variants PFc29, PFc41, EFc29, EFc41, and EFc82 were transfected into HEK 293F animal cells. On the day before transfection, 300 ml of HEK293F cells were passaged at a density of 1×10.sup.6 cells/ml. On the next day, cells were transfected with polyethylenimine (PEI, Polyscience, 23966). First, a heavy chain gene and a light chain gene of each of the variants were mixed in a 2:1 ratio in 30 ml of Freestyle 293 expression culture medium (Gibco, 12338-018). Then, PEI and the variant genes were mixed in a 1:2 ratio, left standing at room temperature for 20 min, mixed with the cells that had been passaged on the previous day, cultured in a C02 shaking incubator at 125 rpm, 37° C., and 8% C02 for 6 days, and centrifuged. The supernatant only was collected.

(35) The proteins were purified from the supernatant by affinity chromatography using AKTA prime plus with a HiTrap MabselectSure column. 300 ml of the supernatant was allowed to flow through the column at a rate of 3 ml/min and washed with 100 ml of 1×PBS. Then, IgG Elution buffer (Thermo scientific, 21009) was allowed to flow through the column at a rate of 5 ml/min. Six fractions (5 ml each) were collected. Each fraction was neutralized with 500 μl of 1M Tris (pH 9.0). The fraction was determined for proteins using Bradford (BioRad, 5000001) and put in a new tube. The purified variants were concentrated using a 30K Amicon ultra centrifugal filter (UFC903096) and their physical properties were analyzed (FIGS. 10 and 11).

(36) Each of the Fc variants other than the in-house wild-type trastuzumab was purified with protein A and its purity (>90%) and molecular weight were determined by SDS-PAGE. SEC-HPLC (FIG. 11a) was used to obtain high-purity protein samples for efficacy evaluation and purity analysis (purity ≥97%). Analysis under isocratic conditions (mobile phase 1×PBS, pH 7.0, 1 ml/min flow rate) revealed that all Fc variants had the same retention time for the main peaks and had estimated molecular weights (FIG. 11b).

Example 8: Measurement of Binding Forces of the Fc Variants to FcRn by ELISA

(37) ELISA was conducted to measure the pH-dependent binding forces of the prepared variants to FcRn and the binding forces of the variants to FcγRs and C1q, which allow the variants to exhibit effector functions. First, the pH-dependent binding forces of the variants to FcRn were investigated. To this end, 50 μl of each of the IgG Fc variants diluted to 4 μg/ml with 0.05 M Na.sub.2CO.sub.3 (pH 9.6) was immobilized onto a flat-bottom polystyrene high-bind 96-well microplate (costar) at 4° C. for 16 h, blocked with 100 μl of 4% skim milk (GenomicBase) (in 0.05% PBST pH 5.8/pH 7.4) at room temperature for 2 h, and washed four times with 180 μl of 0.05% PBST (pH 5.8/pH 7.4). Thereafter, 50 μl of FcRn serially diluted with 1% skim milk (in 0.05% PBST pH 5.8/pH 7.4) was plated in each well and the reaction was carried out at room temperature for 1 h. After washing, an antibody reaction with 50 μl of anti-GST-HRP conjugate (GE Healthcare) was allowed to proceed at room temperature for 1 h. The plate was washed and developed with 50 μl of 1-Step Ultra TMB-ELISA Substrate Solution (Thermo Fisher Scientific). The reaction was quenched with 2 M H.sub.2SO.sub.4 (50 μl each). Then, the reaction product was analyzed using an epoch microplate spectrophotometer (BioTek). The sorted variants had binding forces to FcRn at pH 5.8 similar to the variant LS and were more easily dissociated at pH 7.4 than LS (FIG. 12).

Example 9: Measurement and Comparison of Binding Forces of the Trastuzumab Fc Variants to Monomeric hFcRn at pH 6.0 and pH 7.4

(38) In this example, the pH-dependent binding forces of the commercial trastuzumab, the in-house trastuzumab, and the sorted Fc variants, which were analyzed and investigated for physical properties, to human FcRn were compared. Specifically, K.sub.D values were measured using a Biacore T200 instrument (GE Healthcare). At pH 6.0, human FcRn was used as an analyte in an antigen-mediated antibody capture format, as disclosed in the literature (Yeung Y A. et al., J. Immunol, 2009). Each Fc variant as a ligand was diluted in running buffer (50 mM phosphate, pH 6.0, 150 mM NaCl, 0.005% surfactant P20, pH 6.0), injected at a level of ˜300 response units (RUs) into the surface of CM5 chip on which HER2 ECD domain was immobilized to a level of ˜3,000 RUs, and captured. For binding force measurement, monomeric FcRn (Sinobiological inc., CT009-H08H) as an analyte was serially diluted from 125 nM in FcRn running buffer, injected at a flow rate of 30 μl/min for 2 min, followed by dissociation for 2 min. In each cycle, regeneration was performed with 10 mM glycine (pH 1.5) at a flow rate of 30 ml/min for 30 sec. Sensograms were fit to a 1:1 binding model using the BIAevaluation software (Biacore). As a result, the Fc variants had higher binding forces (PFc 3: 5.6 nM, PFc29: 6.8 nM, PFc 41: 5.9 nM, etc.) than the commercial trastuzumab (15 nM) and the in-house trastuzumab (16.9 nM) as control groups and 428L (9 nM) as the backbone. However, the Fc variants had rather lower binding forces than YTE (5.7 nM) and LS (4.1 nM) whose binding forces are known to be the highest values in the world, but their differences were almost the same within the error range. Since the ligands were less bound to the analyte at pH 7.0, dissociation was evaluated using an avid format in which monomeric hFcRn was directly immobilized and different concentrations of the Fc variants were injected (Zalevsky J et al. Nat. Biotechnol, 2010). Human FcRn ECD domain (Sino Biological) was immobilized to a level of ˜1,500 RUs onto the surface of a CM5 chip. The Fc variants were serially diluted from 3000 nM in HBS-EP (pH 7.4) and injected at a flow rate of 5 ml/min into the FcRn-immobilized chip surface for 2 min. The bound Fc variants were dissociated for 2 min. After each cycle was finished, the chip surface was regenerated with 100 mM Tris (pH 9.0) (conatat time 30 sec; flow rate 30 l/min). Particularly, the Fc variants PFc29 and PFc41 maintained their high binding forces at pH 6.0 and were more rapidly dissociated at pH 7.4 than YTE and LS. These results were in agreement with the results obtained by ELISA and suggest long expected half-lives of the Fa variants (FIG. 13). In practice, in vivo pharmacokinetic experiments were conducted in human FcRn Tg mice.

Example 10: Analysis and Comparison of In Vivo PK Experiments of the Commercial Trastuzumab and the in-House Trastuzumab in Regular B6 Mice and hFcRn Tg Mice

(39) PK analysis was conducted on regular B6 mice (Jungang Experimental Animal Resource Center, C57BL/6J(B6)) whose genetic background is identical to that of human FcRn Tg mice. As a result, the affinity of the Fc of a human antibody for regular mouse FcRn was found to be higher than that for human FcRn, as reported in the literature. There was a variation in the PK values between the in-house antibody and the commercial antibody in the regular mice, and the in-house antibody appeared to be unstable. Further, the AUC in the Tg mice (B6.Cg-Fcgrttm1Dcr Prkdcscid Tg (Jackson lab, CAGFCGRT)276Dcr/DcrJ) was slightly lower than that in the regular mice but the in-house antibody and the commercial antibody showed similar pharmacokinetic tendencies. demonstrating that no problems were encountered in experiments using the in-house Fc variants produced in HEK293 to analyze the actual in vivo pharmacokinetics of the Fc variants in the Tg mice (FIG. 14).

Example 11: Pharmacokinetics of Four Species, Including LS, YTE, and Variants PFc29 and PFc41 in hFcRn Tg Mice

(40) The binding forces measured at pH 6.0 and pH 7.4 using an ELISA system and a BiaCore instrument were found to be constant. Based on these results, PFc29 and PFc41 were sorted due to their high binding forces comparable to that of LS under acidic conditions (pH 6.0) and higher dissociation forces at pH 7.4 than that of LS. Simultaneously with this, LS mutant from Xencor and YTE from MedImmune as control groups, which are currently known to be most effective in the world, and the two trastuzumab Fc variants were injected into 20 human FcRn Tg mice (5 animals per group, 5 mg/kg I.V (tail vein)). After injection, blood samples were collected a total of 12 times (0, 30 min, 1 hr, 6 hr, 24 hr, 3 day, 7 day, 14 day, 21 day, 28 day, 35 day, 42 day, and 50 day) from the facial vein. The concentrations of the Fc variants in the blood samples were analyzed by ELISA and then non-compartmental analysis (NCA) was conducted using WinNonlin. As expected from the results of ELISA and BiaCore analysis, the Fc variants PFc29 and PFc41 showed increased in vivo half-lives. Particularly, the half-life of PFc29 was longer than that of conventional LS (FIG. 15 and Table 3).

(41) TABLE-US-00003 TABLE 3 Trastuzumab Trastuzumab Parameter PFc29 PFc41 YTE LS (in-house) (commercial) t.sub.1/2 (day) 15.99 ± 9.57 10.87 ± 3.77 11.16 ± 7.21 11.72 ± 7.20 6.92 ± 1.06  6.57 ± 1.06 T.sub.max (h)  0.70 ± 0.27  0.70 ± 0.27  0.70 ± 0.27  0.70 ± 0.27 0.75 ± 0.27  0.79 ± 0.27 C.sub.0 (μg/mL)  85.11 ± 11.31  69.72 ± 11.42  89.26 ± 12.54 82.72 ± 8.18 75.60 ± 4.11  64.98 ± 40.08 C.sub.max (μg/mL) 78.19 ± 9.55  65.22 ± 10.60  82.55 ± 11.36 75.92 ± 7.07 70.40 ± 4.35 61.54 ± 9.29 AUC.sub.last (μg/mLxday)  418.30 ± 120.22 466.04 ± 51.57  520.26 ± 214.61  522.25 ± 182.53 249.60 ± 26.43 257.39 ± 43.61 AUC.sub.inf (μg/mLxday)  450.03 ± 158.08 489.92 ± 68.62  574.14 ± 264.60  574.95 ± 228.28 250.84 ± 27.05 258.74 ± 44.32 AUC.sub.%Extrap (%)  5.36 ± 6.20  4.55 ± 3.33  6.86 ± 6.57  7.24 ± 6.73  0.48 ± 0.24  0.49 ± 0.34 V.sub.z (mL/kg)  292.54 ± 271.40 157.05 ± 41.29 126.70 ± 39.27 135.94 ± 52.09 158.91 ± 14.44 184.34 ± 22.30 CL (mL/day/kg) 12.29 ± 4.31 10.36 ± 1.41 10.91 ± 5.96 10.21 ± 4.92  0.67 ± 0.07  0.83 ± 0.14 Non-compartmental analysis of pharmacokinetic parameters of trastuzumab Fc variants after intravenous administration (5 mg/kg) to mice (data are expressed as mean ±SD (n = 5)). t.sub.1/2, terminal half-life; T.sub.max, time at maximal concentration; C.sub.0, extrapolated zero time concentration; C.sub.max, maximal concenration, AUC.sub.last, area under the curve from administration to the last measured concentration; AUC.sub.inf, area under the curve from administration to infinity; AUC.sub.%Extrap, percentage of the extrapolated area under the curve at the total area under the curve; V.sub.z, volume of distribution; CL, clearance.

Example 12: Measurement of Binding Forces of the Fc Variants to FcγRs by ELISA

(42) In this example, the binding forces of the Fc variants to FcγRs were measured. Specifically, 50 μl of each of the IgG Fc variants diluted to 4 μg/ml with 0.05 M Na.sub.2CO.sub.3 (pH 9.6) was immobilized onto a flat-bottom polystyrene high-bind 96-well microplate (costar) at 4° C. for 16 h, blocked with 100 μl of 4% skim milk (GenomicBase) (in 0.05% PBST pH 7.4) at room temperature for 2 h, and washed four times with 180 μl of 0.05% PBST (pH 7.4). Thereafter, 50 μl of FcγRs serially diluted with 1% skim milk (in 0.05% PBST pH 7.4) was plated in each well and the reaction was carried out at room temperature for 1 h. After washing, an antibody reaction with 50 μl of anti-GST-HRP conjugate (GE Healthcare) was allowed to proceed at room temperature for 1 h. The plate was washed and developed with 50 μl of 1-Step Ultra TMB-ELISA Substrate Solution (Thermo Fisher Scientific). The reaction was quenched with 2 M H.sub.2SO.sub.4 (50 μl each). Then, the reaction product was analyzed using an epoch microplate spectrophotometer (BioTek). Each experiment was conducted in duplicate. FIG. 16 shows the binding forces of the Fc variants to FcγRs (FcγRI, FcγRIIa(H), FcγRIIa(R), FcγRIIb, FcγRIIIa(V), and FcγRIIIa(F)), which were measured by ELISA.

Example 13: Measurement of Effector Functions of the Fc Variants by Antibody-Dependent Cell-Mediated Cytotoxicity (ADCC)

(43) The antibody-dependent cellular cytotoxicity (ADCC) activities of the trastuzumab Fc variants were evaluated using an ADCC reporter bioassay kit (Promega, G7010). Specifically, SKBR-3 cells as target cells were plated at a density of 5×10.sup.3 cells/100 μl in each well of a 96-well tissue culture plate and cultured in a C02 incubator at 37° C. for 20 h. Thereafter, 95 μl of the culture medium was removed from each well of the plate using a multi-pipette and 25 μl of ADCC assay buffer provided from the ADCC reporter bioassay kit was plated in each well. Normal IgG, trastuzumab, and the trastuzumab Fc variants were diluted to various concentrations with ADCC assay buffer. 25 μl of each dilution was plated in each well of the 96-well tissue culture plate containing the cells and left standing at room temperature until effector cells were added. Effector cells provided from the kit were dissolved in a thermostatic water bath at 37° C. for 2-3 min and 630 μl of the solution was mixed with 3.6 mL of ADCC assay buffer. 25 μl of the effector cells were plated in each well of the plate containing the target cells and the antibody dilution. The reaction was carried out in a C02 incubator at 37° C. for 6 h. After the lapse of a predetermined time, the plate was taken out of the incubator and placed at room temperature for 15 min. 75 μl of Bio-Glo™ Luciferase assay reagent was added to each well and the reaction was carried out at room temperature for 5 min. After completion of the reaction, the luminescence of each well was measured using a luminometer (Enspire multimode plate reader). The ADCC activity of each test antibody was determined by expressing the average of the experimental results as a fold induction, which was calculated by the following equation:
Fold induction=RLU(induced.sup.1−background.sup.2)/RLU(no antibody control.sup.3−background)

(44) induced.sup.1: RLU value acquired from the sample containing the target cells, the test antibody and the effector cells

(45) background.sup.2: RLU value acquired from the ADCC assay buffer

(46) no antibody control.sup.3: RLU value acquired from the sample containing the target cells and the effector cells only

(47) The ADCC activities of the trastuzumab Fc variants (LS, YTE, PFC29, and PFC41) for SKBR-3 were compared with that of trastuzumab (FIG. 17). As a result, the maximum ADCC activities of LS, PFc29, and PFc41 at their highest concentrations were ˜1.5, ˜1.18, and ˜1.27-fold higher than that of trastuzumab as the positive control group, respectively. In contrast, the maximum ADCC activity of YTE at its highest concentration was ˜4.2-fold lower than that of trastuzumab. In conclusion, the trastuzumab Fc variants PFc29 and PFc41 and the control variant LS achieved ADCC activities 1.18-1.5 times higher than that of the control trastuzumab. The EC50 values of the variants were measured. As shown in FIG. 18, the lower EC50 values of the Fc variant PFc29 (0.04543 μg/mL) and PFc41 (0.05405 μg/mL) than the control LS (0.05575 μg/mL) indicate that the efficacies of the Fc variants PFc29 and PFc41 were more stable.

Example 14: Measurement of Binding Forces of the Fc Variants to C1q by ELISA

(48) In this example, the binding forces of the Fc variants to C1q were measured. Specifically, 50 μl of each of the IgG Fc variants diluted to 4 μg/ml with 0.05 M Na.sub.2CO.sub.3 (pH 9.6) was immobilized onto a flat-bottom polystyrene high-bind 96-well microplate (costar) at 4° C. for 16 h, blocked with 100 μl of 4% skim milk (GenomicBase) (in 0.05% PBST pH 7.4) at room temperature for 2 h, and washed four times with 180 μl of 0.05% PBST (pH 7.4). Thereafter, 50 μl of Complement C1q Human (Millipore) serially diluted with 1% skim milk (in 0.05% PBST pH 7.4) was plated in each well and the reaction was carried out at room temperature for 1 h. After washing, an antibody reaction with 50 μl of anti-C1q-HRP conjugate (Invitrogen) was allowed to proceed at room temperature for 1 h. The plate was washed and developed with 50 μl of 1-Step Ultra TMB-ELISA Substrate Solution (Thermo Fisher Scientific). The reaction was quenched with 2 M H.sub.2SO.sub.4 (50 μl each). Then, the reaction product was analyzed using an epoch microplate spectrophotometer (BioTek). As a result of the analysis, the binding force of the sorted PFc29 to C1q was higher than those of conventional LS and YTE (FIG. 19).

(49) Although the particulars of the present invention have been described in detail, it will be obvious to those skilled in the art that such particulars are merely preferred embodiments and are not intended to limit the scope of the present invention. Therefore, the substantial scope of the present invention is defined by the appended claims and their equivalents.