Method for inhibiting intracellular activated RAS using intact immunoglobulin-type antibody having cytosol-penetrating ability and use thereof

Abstract

The present disclosure relates to a method for inhibiting intracellular activated (GTP-bound) RAS using an intact immunoglobulin-type antibody having the ability to penetrate the cytosol, and to the use thereof. The disclosure further relates to a heavy-chain variable region (VH) which induces an intact immunoglobulin-type antibody to penetrate the cytosol and bind to activated RAS in the cytosol, and to an antibody comprising the same. The disclosure correspondingly provides a method for inhibiting the growth of cancer or tumor cells using the antibody, and a method for treating cancer or tumor.

Claims

1. An anti-RAS antibody, wherein the antibody comprises a heavy chain variable region (VH) that binds specifically to RAS-GTP comprising: a VH CDR1, VH CDR2 and VH CDR3, respectively comprising the amino acid sequences selected from the group consisting of: VHCDR1 SEQ ID NO: 11, VHCDR2 SEQ ID NO: 12, and VHCDR3 SEQ ID NO: 13; VHCDR1 SEQ ID NO: 14, VHCDR2 SEQ ID NO: 15, and VHCDR3 SEQ ID NO: 16; VHCDR1 SEQ ID NO: 17, VHCDR2 SEQ ID NO: 18, and VHCDR3 SEQ ID NO: 19; VHCDR1 SEQ ID NO: 20, VHCDR2 SEQ ID NO: 21, and VHCDR3 SEQ ID NO: 22; VHCDR1 SEQ ID NO: 23, VHCDR2 SEQ ID NO: 24, and VHCDR3 SEQ ID NO: 25; and VHCDR1 SEQ ID NO: 26, VHCDR2 SEQ ID NO: 27, and VHCDR3 SEQ ID NO: 28; and a light-chain variable region (VL) that penetrates the cell membrane comprising: a VL CDR1, VL CDR2, and VL CDR3, respectively comprising the amino acid sequences selected from the group consisting of: VLCDR1 SEQ ID NO: 32, VLCDR2 SEQ ID NO: 33, VHCDR3 SEQ ID NO: 34; and VLCDR1 SEQ ID NO: 32, VLCDR2 SEQ ID NO: 33, VHCDR3 SEQ ID NO: 40; wherein 2nd and 4th amino acids starting from the N-terminus of the light-chain variable region are respectively substituted with leucine (L) and methionine (M), (wherein the positions of the amino acids are numbered according to the Kabat numbering system).

2. The antibody of claim 1, wherein the antibody binds specifically to RAS-GTP in the cytosol of a cell.

3. The antibody of claim 1, wherein the 9th, 10th, 13th, 15th, 17th, 19th, 21st, 22nd, 42nd, 45th, 58th, 60th, 79th and 85th amino acids starting from the N-terminus of the light-chain variable region (VL) are serine (S), serine (S), alanine (A), valine (V), aspartic acid (D), valine (V), isoleucine (I), threonine (T), lysine (K), lysine (K), valine (V), serine (S), glutamine (Q) and threonine (T), respectively (wherein the positions of the amino acids are numbered according to the Kabat numbering system).

4. The antibody of claim 3, wherein the 89th and 91st amino acids starting from the N-terminus of the light-chain variable region (VL) are glutamine (Q) and tyrosine (Y), respectively (wherein the positions of the amino acids are numbered according to the Kabat numbering system).

5. The antibody of claim 1, wherein the heavy chain variable region (VH) comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 2 to 7.

6. The antibody of claim 5, wherein the antibody binds specifically to the RAS-GTP in the cytosol of a cell.

7. The antibody of claim 1, wherein the light-chain variable region (VL) comprises an amino acid sequence selected from the group consisting of SEQ ID NOs: 29, 30 and 31.

8. The antibody of claim 1, wherein the antibody is fused to a biologically active molecule selected from the group consisting of peptides, proteins, small-molecule drugs, nanoparticles and liposomes.

9. The antibody of claim 8, wherein the biologically active molecule is a peptide, wherein the peptide is RGD4C comprising an amino acid sequence as set forth in SEQ ID No: 41, or RGD4C comprising an amino acid sequence as set forth in SEQ ID No: 42.

10. A polynucleotide that encodes the antibody of claim 1.

11. The antibody of claim 1, wherein the antibody comprises a full-length heavy chain variable region (VH).

12. The antibody of claim 11, wherein the antibody is an intact immunoglobulin-type antibody.

13. The antibody of claim 11, wherein the antibody binds specifically to the RAS-GTP in the cytosol of a cell.

14. The antibody of claim 12, wherein the antibody binds specifically to the RAS-GTP in the cytosol of a cell.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee.

(2) FIG. 1 is a schematic view showing a strategy of inducing cytotoxicity specific for Ras mutant cells by use of a monoclonal antibody (anti-Ras. GTP iMab: internalizing & interfering monoclonal antibody) which is constructed by replacing the heavy-chain variable region (VH) of the immunoglobulin-type antibody cytotransmab (having only cytosol-penetrating ability) with a heavy-chain variable region (VH) binding specifically to GTP-bound KRas and which penetrates cells and binds specifically to GTP-bound Ras in the cytosol.

(3) FIG. 2 is a schematic view showing a method of constructing anti-RasGTP iMab by replacing the heavy-chain variable region (VH) of cytotransmab, which has only cytosol-penetrating ability, with a heavy-chain variable region (VH) which binds specifically to GTP-bound KRas.

(4) FIG. 3 is a schematic view showing a library screening strategy for obtaining a humanized antibody heavy-chain variable single domain having high affinity only for GTP-bound KRas G12D protein.

(5) FIG. 4 shows the results of FACS analysis of binding, performed under a condition of GTP-bound KRas G12D alone and a condition competitive with GTP-bound KRas G12D in each step of the above-described process for obtaining a high affinity for GTP-bound KRas G12D.

(6) FIG. 5A shows the results of analysis of a sequence including a clone used in a process of obtaining the improved, cytosol-penetrating humanized light-chain variable single domain hT3 VL, which binds stably to a humanized antibody heavy-chain variable region, from the mouse light-chain variable region m3D8 VL.

(7) FIG. 5B compares model structures using the WAM modeling of m3D8 VL, the humanized light-chain variable single domain hT0 VL and its mutants (hT2 VL and hT3 VL) by a superimposing method.

(8) FIG. 6A shows the results of confocal microscopy observation of the cytosol-penetrating ability of light-chain variable single domains.

(9) FIG. 6B shows the results of confocal microscopy observation performed to verify the cytosol-penetrating mechanisms of light-chain variable single domains.

(10) FIG. 7A shows the results of analyzing the amino acid sequence of hT3 VL together with the amino acid sequences of light-chain variable regions (VLs) of conventional human antibody Adalimumab (Humira) and humanized antibody Bevacizumab (Avastin) in order to confirm whether or not hT3 VL can be applied to a variety of human antibody heavy-chain variable regions.

(11) FIG. 7B shows the results of analyzing interface residues between variable regions in order to construct stable cytotransmab that optimally interacts with a human antibody heavy-chain variable region.

(12) FIG. 8 is a schematic view showing a method of replacing a light-chain variable region having no cell-penetrating ability with a humanized light-chain variable region having cytosol-penetrating ability in order to construct cytotransmab.

(13) FIG. 9A shows the results of observing 1-2 cells in various cell lines by confocal microscopy in order to verify the cytosol-penetrating ability of cytotransmabs having a light-chain variable region replaced with the cytosol-penetrating light-chain region hT4 VL.

(14) FIG. 9B shows the results of examining cytosol-penetrating ability for several cells, performed at a reduced magnification in order to examine cell-penetrating efficiency in the cytosol-penetrating ability examination experiment by confocal microscopy observation as shown in FIG. 7A.

(15) FIG. 10A is a graph showing the results obtained by treating HeLa and PANC-1 cell lines with cytotransmab and evaluating in vitro the inhibition of growth of the cells.

(16) FIG. 10B is an image showing the results obtained by treating HeLa and PANC-1 cell lines with cytotransmab and evaluating in vitro the inhibition of growth of the cells.

(17) FIG. 11 shows the results of analyzing anti-RasGTP iMab RT4 by 12% SDS-PAGE under reductive or non-reductive conditions after purification.

(18) FIG. 12 shows the results of ELISA performed to measure affinity for GTP-bound and GDP-bound wild-type KRas and GTP-bound and GDP-bound KRas mutants (KRas G12D, KRas G12V, and KRas G13D).

(19) FIG. 13 shows the results of analyzing the affinity of anti-RasGTP iMab RT4 for GTP-bound KRAS G12D by use of SPR (BIACORE 2000) (GE Healthcare).

(20) FIG. 14 shows the results of confocal microscopy observation performed to examine the cytosol-penetrating ability of anti-RasGTP iMab RT4.

(21) FIG. 15 shows the results obtained by treating NIH3T3, NIH3T3 KRas G12V and NIH3T3 HRas G12V cell lines with anti-RasGTP iMab RT4 and evaluating in vitro the inhibition of growth of the cells.

(22) FIG. 16 shows the results of evaluating the inhibition of growth of non-adherent cells in an NIH3T3 HRas G12V cell line.

(23) FIG. 17 shows the results of confocal microscopy observation of whether anti-Ras RT4 is superimposed with activated HRas G12V mutants in cells.

(24) FIG. 18 shows the results of confocal microscopy observation of whether anti-Ras RT4 is superimposed with GTP-bound KRas G12V mutants in cells.

(25) FIG. 19 shows the results obtained by treating HCT116 and PANC-1 cell lines with RGD-TMab4 and RGD-RT4 and evaluating in vitro the inhibition of growth of the cells.

(26) FIG. 20A shows the results of analyzing the tumor growth inhibitory effect of RGD-fused anti-RasGTP iMab RT4 in mice xenografted with HCT116 cells.

(27) FIG. 20B is a graph showing the results of measuring the body weight of mice in order to examine the non-specific side effects of RGD-fused anti-RasGTP iMab RT4.

(28) FIG. 21A shows a strategy of constructing a human heavy-chain variable region library to improve the affinity of RT4.

(29) FIG. 21B is a schematic view showing a method of constructing a designed library by a PCR technique and transforming the constructed library onto the yeast surface by homologous combination with a heavy-chain single yeast surface display vector (pYDS-H) treated with the restriction enzymes NheI and ApaI.

(30) FIG. 22 shows the results of FACS analysis performed to determine binding to GTP-bound KRas G12D and GDP-bound KRas G12D for library-expressing yeast in each step in order to confirm enrichment specific for GTP-bound KRas G12D in the above-described library screening process.

(31) FIG. 23 shows the results of sequencing of individual clones using three libraries.

(32) FIG. 24 shows the results of analyzing anti-RASGTP iMab having improved affinity by 12% SDS-PAGE under a reductive or non-reductive condition.

(33) FIG. 25 shows the results obtained by replacing the heavy-chain variable region of anti-RasGTP iMab with a RasGTP-specific heavy-chain variable region having improved affinity and then performing confocal microscopic observation to confirm whether or not the anti-RasGTP iMab has the ability to penetrate cells.

(34) FIG. 26A shows the results of ELISA performed to measure the affinity anti-RasGTP iMab having improved affinity for GTP-bound KRas G12D and GDP-bound KRas G12D.

(35) FIG. 26B shows the results of ELISA analysis performed to confirm the highly specific affinity of RT11, selected based on the ELISA-based binding analysis, for various GTP-bound Ras mutants.

(36) FIG. 27A shows the results of analyzing the affinity of anti-RasGTP iMab RT11 for GTP-bound KRas G12D by use of SPR (BIACORE 2000) (GE Healthcare).

(37) FIG. 27B is a sensorgram showing the results of analyzing the affinity of RT11 for GTP- or GDP-bound KRas G12D at the highest concentration (1000 nM).

(38) FIG. 28 shows the results of a competitive ELISA performed to confirm whether anti-RasGTP iMab RT11 can inhibit the binding between the effector molecule Raf and intracellular KRas.

(39) FIG. 29 shows the results of confocal microscopic observation performed to confirm whether anti-RasGTP iMab having improved affinity has the ability to penetrates various types of tumor cells.

(40) FIG. 30 shows the results of confocal microscopic observation performed using a non-cell-penetrating, self-quenching dye (calcein (Sigma)) to observe the cytosol-remaining ability of anti-RasGTP iMab having improved affinity.

(41) FIG. 31 shows the results obtained by treating various Ras wild-type and Ras mutant cell lines with anti-RasGTP iMab RT11 and evaluating in vitro the inhibition of growth of the cells.

(42) FIG. 32 are a set of images showing the results of polarizing microscopic observation performed to determine the cell density of each cell line.

(43) FIG. 33 shows the results of confocal microscopic observation performed to examine whether RT11 is superimposed with activated KRas G12V mutants in cells.

(44) FIG. 34 shows the results of an immunoprecipitation assay performed to examine whether RT11 binds to activated Ras in cells.

(45) FIGS. 35A and 35B show the results of an immunoprecipitation assay performed to examine whether or not RT11 inhibits the binding between RasGTP and effector proteins.

(46) FIG. 36 shows the ELISA results obtained by measuring the affinities of the constructed RGD10 peptide-fused RT11 (RGD10-RT11) for a variety of GTP-bound and GDP-bound Ras mutants.

(47) FIGS. 37 and 38 show the results obtained by treating Colo320DM, HCT116, PANC-1, SW480 and DLD-1 cell lines with RGD10-TMab4 and RGD10-RT11 and evaluating in vitro the inhibition of growth of the cells.

(48) FIG. 39 shows the results of analysis performed to examine whether or not RGD10-TMab4 and RGD10-RT11 bind specifically to integrin 3 on the cell surface.

(49) FIG. 40 shows the results of confocal microscopic observation performed to examine whether or not RGD10-RT11 is superimposed with an activated KRas G12V mutant in cells.

BEST MODE FOR CARRYING OUT THE INVENTION

(50) Hereinafter, the present invention will be described in further detail with reference to examples. It will be obvious to a person having ordinary skill in the art that these examples are illustrative purposes only and are not to be construed to limit the scope of the present invention.

Example 1: Selection of Heavy-Chain Variable Region (VH), which Binds Specifically to GTP-Bound KRas, by High-Diversity Human VH Library

(51) FIG. 1 is a schematic view showing a strategy of inducing cytotoxicity specific for Ras mutant cells by use of a monoclonal antibody (anti-Ras. GTP iMab: internalizing & interfering monoclonal antibody) which is constructed by replacing the heavy-chain variable region (VH) of an IgG-type cytotransmab (having only cytosol-penetrating ability) with a heavy-chain variable region (VH) binding specifically to GTP-bound KRas and which penetrates cells and binds specifically to GTP-bound Ras in the cytosol.

(52) FIG. 2 is a schematic view showing a method of constructing anti-RasGTP iMab by replacing the heavy-chain variable region (VH) of an intact IgG-type cytotransmab, which has only cytosol-penetrating ability, with a heavy-chain variable region (VH) which binds specifically to GTP-bound KRas.

(53) Specifically, the FR (framework) of the library used was the V gene IGHV3-23*04, J.sub.H4 which is most commonly used in conventional antibodies, and the CDR3 in the library had 9 residues. The construction of the library and a yeast surface display method are described in detailed in a previously reported paper (Baek and Kim, 2014).

(54) In order to select a stable humanized heavy-chain variable single domain (VH) antibody fragment which is to be introduced into the anti-RasGTP iMab and which binds specifically to GTP-bound KRas, a yeast display VH library constructed in a previous studies was used.

Example 2: Preparation of GTP-Bound KRas G12D Protein

(55) Expression in E. coli and purification, performed to prepare GTP-bound KRas G12D antigen for library screening and affinity analysis, are described in detail in a previously reported paper (Tanaka T et al., 2007).

(56) Specifically, a DNA encoding residues 1 to 188, which comprises the CAAX motif of each of wild-type KRas and mutant KRas G12D, KRas G12V and KRas G13D (listed in the order of higher to lower mutation frequency), was cloned into the E. coli expression vector pGEX-3X by use of the restriction enzymes BamHI/EcoRI. Herein, the expression vector was designed to have a T7 promoter-GST-KRas. All KRas mutations were induced using an overlap PCR technique, and the expression vector was constructed using the above-described method. The pGEX-3X-KRas vector was transformed into E. coli by electroporation, and selected in a selection medium. The selected E. coli was cultured in LB medium in the presence of 100 g/ml of an ampicillin antibiotic at 37 C. until the absorbance at 600 nm reached 0.6. Then, 0.1 mM IPTG was added thereto for protein expression, and then the E. coli cells were further cultured at 30 C. for 5 hours. Thereafter, the E. coli cells were collected by centrifugation, and then disrupted by sonication (SONICS). The disrupted E. coli cells were removed by centrifugation, and the remaining supernatant was collected and purified using glutathione resin (Clontech) that specifically purifies GST-tagged protein. The glutathione resin was washed with 50 ml of washing buffer (140 mM NaCl, 2.7 mM KCl, 10 mM NaH.sub.2PO.sub.4, 1.8 mM KH.sub.2PO.sub.4, 1 mM EDTA, 2 mM MgCl.sub.2 pH 7.4) (SIGMA), and then protein was eluted with elution buffer (50 mM Tris-HCl pH8.0, 10 mM reduced glutathione, 1 mM DTT, 2 mM MgCl.sub.2) (SIGMA). The eluted protein was dialyzed to replace the buffer with storage buffer (50 mM Tris-HCl pH8.0, 1 mM DTT, 2 mM MgCl.sub.2) (SIGMA). The purified protein was quantified by measuring the absorbance at a wavelength of 280 nm and the absorption coefficient. SDS-PAGE analysis indicated that the protein had a purity of about 98% or higher.

(57) Next, in order to bind a GTPS (Millipore) or GDP (Millipore) substrate to KRas protein, KRas and a substrate at a molecular ratio of 1:20 were reacted in a reaction buffer (50 mM Tris-HCl pH8.0, 1 mM DTT, 5 mM MgCl.sub.2, 15 mM EDTA) (SIGMA) at 30 C. for 30 minutes, and 60 mM MgCl.sub.2 was added thereto to stop the reaction, and then stored at 80 C.

Example 3: Selection of Heavy-Chain Variable Region (VH) Specific for GTP-Bound KRas G12D

(58) FIG. 15 is a schematic view showing a library screening strategy for obtaining a humanized antibody heavy-chain variable single domain having a high affinity only for GTP-bound KRas G12D protein.

(59) Specifically, GTP-bound KRas G12D purified in Example 14 was biotinylated (EZ-LINK Sulfo-NHS-LC-Biotinylation kit (Pierce Inc., USA)), and then reacted with a heavy-chain variable region library displayed on the yeast cell surface at room temperature for 1 hour. The heavy-chain variable region library on the yeast cell surface, which reacted with the biotinylated GTP-bound KRas G12D, was reacted with Streptavidin (Microbead (Miltenyi Biotec) at 4 C. for 20 minutes, and then yeast displaying a heavy-chain variable region having a high affinity for the GTP-KRAS G12D was enriched using MACS (magnetic activated cell sorting). The selected library-displaying yeast was cultured in a selection medium and cultured in SG-CAA+URA (20 g/L Galactose, 6.7 g/L Yeast nitrogen base without amino acids, 5.4 g/L Na2HPO.sub.4, 8.6 g/L NaH.sub.2PO.sub.4, 5 g/L casamino acids, 0.2 mg/L Uracil) (SIGMA) medium to induce protein expression. Next, the yeast was incubated with a yeast displaying the library competitively with GTP-bound KRas G12D alone or non-biotinylated GTP-bound KRas G12D antigen at a concentration 10-fold higher than GTP-bound KRas G12D, at room temperature for 1 hour, after which it was reacted with PE-conjugated Streptavidin (Streptavidin-R-phycoerythrin conjugate (SA-PE) (Invitrogen), and enriched by FACS (fluorescence activated cell sorting) (FACS Caliber) (BD biosciences). After selection of screening conditions by FACS analysis, antigen was bound to the yeast displaying the enriched library under the same conditions as described, and then the yeast was enriched using a FACS aria II sorter. The humanized heavy-chain region library enriched by the first MACS and first FACS screening was mated with a yeast secreting the cytosol-penetrating light-chain variable single domain (hT4 VL), and displayed on the yeast surface in the form of Fab, and then subjected to second FACS and third FACS screening.

(60) Specifically, in order to construct a yeast which is to be mated with the heavy-chain variable domain (VH) library and which secretes the cytosol-penetrating light-chain variable domain (VL), a DNA encoding the cytosol-penetrating hT4 VL was cloned into the light-chain variable domain yeast secretion vector pYDS-K by the restriction enzymes NheI and BsiWI, thereby obtaining pYDS-K-hT4 VL. The obtained pYDS-K-hT4 VL was transformed into the mating -type yeast mating strain YVH10 by electroporation, and mated with a yeast cultured in the selection medium SD-CAA+Trp (20 g/L Glucose, 6.7 g/L Yeast nitrogen base without amino acids, 5.4 g/L Na2HPO.sub.4, 8.6 g/L NaH.sub.2PO.sub.4, 5 g/L casamino acids, 0.4 mg/L tryptophan) (SIGMA).

(61) Specifically, in the case of yeast mating, there are 110.sup.7 yeast cells when the absorbance at 600 nm is 1. Among the cultured yeast cells, 1.510.sup.7 yeast cells expressing the selected heavy-chain variable domain library and 1.510.sup.7 yeast cells containing hT4 VL were added to GTP-bound KRas G12D, and washed three times with YPD YPD (20 g/L Dextrose, 20 g/L peptone, 10 g/L yeast extract, 14.7 g/L sodium citrate, 4.29 g/L citric acid, pH 4.5) (SIGMA). Then, the yeast cells were re-suspended in 100 l of YPD, and dropped onto an YPD plate so as not to spread, after which these yeast cells were dried and cultured at 30 C. for 6 hours. Next, the dried yeast-coated portion was washed three times with YPD medium, and then incubated in the selection medium SD-CAA at 30 C. for 24 hours to a final yeast concentration of 110.sup.6 cells or less, and only mated yeast cells were selected. The selected yeast cells were incubated in SG-CAA medium to induce expression of a humanized antibody Fab fragment, and enriched by second and third FACS such that the yeast cells would be 100-fold competitive with GDP-bound KRas G12D at a GTP-bound KRas G12D concentration of 100 nM.

(62) FIG. 16 shows the results of FACS analysis of binding under a condition of GTP-bound KRas G12D alone and a condition competitive with GTP-bound KRas G12D in each step of the above-described screening process for obtaining a high affinity for GTP-bound KRas G12D. Accordingly, it was found that it is possible to select a library that can bind specifically to GTP-bound KRas G12D in a manner dependent on the heavy-chain variable domain (VH).

(63) Through the high-throughput screening as described above, an RT4 clone was finally selected from the library having a high affinity and specificity for GTP-bound KRas G12D protein by individual clone analysis.

Example 4: Rationale for Development of Cytosol-Penetrating Humanized Light-Chain Variable (VL) Single Domain

(64) FIGS. 5A and 5B are schematic views showing the concept of an intact immunoglobulin antibody, named cytotransmab, which penetrates a cell and localizes in the cytosol. To realize this antibody and understand the cytosol-penetrating ability of humanized antibody light-chain variable regions, reference was made to conventional studies on the correlations between the cytosol-penetrating ability of the mouse light-chain variable single domain m3D8 VL and CDRs corresponding to light-chain variable region fragments (Lee et al., 2013).

(65) FIG. 5A shows the results of analysis of a sequence including a clone used in a process of obtaining the improved, cytosol-penetrating humanized light-chain variable single domain hT3 VL, which binds stably to a humanized antibody heavy-chain variable region, from the mouse light-chain variable region m3D8 VL.

(66) Specifically, based on a comparison of cytosol-penetrating ability between the mouse light-chain variable single domain m3D8 VL and hT0 VL obtained by humanizing the single domain m3D8 VL by use of CDR-grafting technology, it was confirmed that the cytosol-penetrating ability was lost even though the CDR1 sequence of the light-variable variable region (VL) was conserved.

(67) Thus, in order to improve the structure of CDR1 to have a structure similar to that of m3D8 VL to thereby restore the cytosol-penetrating ability of the humanized antibody light-chain variable single domain, CDR regions (Vernier zones) in the FR (framework) were comparatively analyzed. As a result, it was found that residues 2 and 4 differ from those of mouse m3D8 VL having cytosol-penetrating ability. Particularly, because residues 2 and 4 act as an upper core that greatly influence the CDR1 structure (Vernizer zone), hT2 VL having a CDR1 structure similar to that of m3D8 VL was developed by reverse mutations of hT0 VL (see FIG. 5A).

(68) Next, in order to construct stable cytotransmab and to create a pair between VH3 and Via subgroups (that are highly prevalent in stable antibodies) to thereby develop a light-chain variable region that complementarily stably binds to a variety of human antibody heavy-chain variable regions and retains its ability to penetrate into the cytosol, the FR (framework) of hT2 VL and the light-variable region FR (framework) of the humanized therapeutic monoclonal antibody Trastuzumab (Herceptin), which has VH3 and V1 subgroups and is very stable, were comparatively analyzed. As a result, it was shown that 14 residues in the FR (framework) of hT2 VL differ from those in the light chain-variable region FR (framework) of Trastuzumab. These 14 residues were mutated with the sequence of the light chain-variable region FR (framework) of Trastuzumab, thereby developing hT3 VL (see FIG. 5A).

(69) FIG. 5B compares model structures using the WAM modeling of m3D8 VL, the humanized light-chain variable single domain hT0 VL and its mutants (hT2 VL and hT3 VL) by a superimposing method. It was found that, through reverse mutations at residues 2 and 4 as described above, the structural difference of the CDR1 region from that of m3D8 VL was reduced.

Example 5: Expression and Purification of Humanized Light-Chain Variable (VL) Single Domain Having Cytosol-Penetrating Ability

(70) To compare the actual cytosol-penetrating abilities of hT2 VL and hT3 VL designed in the above Example 4, humanized light-chain variable (VL) single domains were purified.

(71) Specifically, the cytosol-penetrating light-chain variable single domain containing a Pho A signal peptide at the N-terminus and a protein A tag at the C-terminus was cloned into a pIg20 vector by NheI/BamHI restriction enzymes, and then the vector was transformed into E. coli BL21(DE3)plysE for protein expression by electroporation. The E. coli was cultured in LBA medium containing 100 ug/ml of ampicillin at 180 rpm and 37 C. until the absorbance at 600 nm reached 0.6-0.8. Then, the culture was treated with 0.5 mM of IPTG (isopropyl -D-1-thiogalactopyronoside, and then incubated at 23 for 20 hours to express the protein. After expression, the culture was centrifuged by a high-speed centrifuge at 8,000 rpm for 30 minutes, and the supernatant was collected, and then reacted with IgG-Sepharose resin (GE Healthcare). The resin was washed with 50 ml of TBS (Tris-HCl, 137 mM NaCl, 2.7 mM KCl, pH 7.4), and then washed with 5 ml of 5 mM NH.sub.4Ac (pH 5.0) buffer. Next, the protein was eluted from the resin by use of 0.1 M HAc (pH 3.0) buffer, and the buffer was replaced with TBS (pH 7.4) by dialysis. Then, the concentration of the protein was measured by a BCA (bicinchoninic acid (Pierce)) assay, and the purity of the protein was analyzed by SDS-PAGE.

Example 6: Verification of Cytosol-Penetrating Ability and Cell Penetration Mechanism of Cytosol-Penetrating Humanized Light-Chain Variable (VL) Single Domain

(72) FIG. 6A shows the results of confocal microscopy observation of the cytosol-penetrating ability of light-chain variable single domains.

(73) Specifically, in order to verify the cytosol-penetrating abilities of m3D8 VL, hT0 VL, hT2 VL and hT3 VL, a cover slip was added to 24-well plates, and 510.sup.4 HeLa cells per well were added to 0.5 ml of 10% FBS (Fetal bovine Serum)-containing medium and cultured for 12 hours under the conditions of 5% CO.sub.2 and 37 C. When the cells were stabilized, each well was treated with 10 M of m3D8 VL, hT0 VL, hT2 VL or hT3 VL in 0.5 ml of fresh medium, and incubated for 6 hours under the conditions of 37 C. and 5% CO.sub.2. Next, the medium was removed, and each well was washed with PBS, and then treated with a weakly acidic solution (200 mM glycine, 150 mM NaCl, pH 2.5) to remove proteins from the cell surface. Next, each well was washed with PBS, and the cells were fixed in 4% paraformaldehyde at 25 C. for 10 minutes. After washing with PBS, each well was incubated with PBS buffer containing 0.1% saponin, 0.1% sodium azide and 1% BSA at 25 C. for 10 minutes to form pores in the cell membranes. After washing with PBS, each well was incubated with PBS buffer c containing 2% BSA at 25 C. for 1 hour to eliminate nonspecific binding. Then, each well was treated with rabbit-IgG (Sigma) that recognizes the protein A tag of the light-chain variable single domain, and each well was incubated at 25 C. for 2 hours, washed three times with PBS, and then treated with red fluorescence (TRITC)-labeled anti-rabbit antibody (Sigma), followed by incubation at 25 C. for 1 hour. Finally, the nucleus was blue-stained with Hoechst33342 and observed with a confocal microscope. As a result, it was shown that m3D8 VL, hT2 VL and hT3 VL, except for hT0 VL, had cell-penetrating ability.

(74) FIG. 6B shows the results of confocal microscopy observation performed to verify the cytosol-penetrating mechanisms of light-chain variable single domains.

(75) Specifically, when HeLa cells were prepared as shown in FIG. 6A and stabilized, a dilution of 10 M of m3D8 VL, hT2 VL or hT3 VL and 10 ug/ml of Alexa Fluor 488-transferrin (TF, green fluorescence), FITC-cholera toxin B (Ctx-B, green fluorescence) or Oregon green-dextran (Dextran, green fluorescence) in 0.5 ml of fresh medium was added to each well and incubated for 2 hours under the conditions of 37 C. and 5% CO.sub.2. Next, the light-chain variable single domains were stained as shown in FIG. 3A. As shown in FIG. 3B, all the light-chain variable single domains were superimposed with cholera toxin-B, indicating that these domains penetrate the cytosol by caveolae.

Example 7: Development of Cytosol-Penetrating Humanized Light-Chain Variable (VL) Single Domain that Easily Interacts with Human Antibody Heavy-Chain Variable Domain

(76) FIG. 7A shows the results of analyzing the amino acid sequence of hT3 VL together with the amino acid sequences of light-chain variable domains (VLs) of conventional human antibody Adalimumab (Humira) and humanized antibody Bevacizumab (Avastin) in order to confirm whether or not hT3 VL can be applied to a variety of human antibody heavy-chain variable domains.

(77) Specifically, VH-VL interface residues that are involved in the interaction between heavy-chain and light-chain variable domains were analyzed. As a result, it was found that lysine (K) at position 89 and serine (S) at position 91 of the CDR3 of the VL domain are consistent with glutamine (Q) at position 89 and tyrosine (Y) in human antibodies.

(78) To construct a strategy for improving the residues, the effects of VH-VL interface residues on the CDRs of the heavy-chain variable domain and the light-chain variable region were analyzed in more detail.

(79) FIG. 7B shows the results of analyzing interface residues between variable regions in order to construct stable cytotransmab that optimally interacts with a human antibody heavy-chain variable region.

(80) Specifically, based on Information about the positions of interface residues between human antibody variable regions, the frequency of binding to specific interface residues located in opposite variable regions, and the abundance of interface residues in human antibodies, which were reported in the literature, hT3 VL and the interface residues between the heavy chain and light chain variable regions of Bevacizumab (Avastin) and Adalimumab (Humira), which are antibodies approved by the FDA, were analyzed (Vargas-Madrazo and Paz-Garcia, 2003). The results of the analysis indicated that, in the mouse CDRs of hT3 VL, residues 89 and 91 in CDR3 that is involved in association between variable regions are highly abundant in human antibodies and can influence the CDR3 structure of the heavy-chain variable region (VH). The two residues were mutated with amino acids that are highly abundant in human antibodies, thereby hT4 VL that can optimally bind to human antibody heavy-chain variable regions.

(81) Tables 1 and 2 below show the sequences of the designed human antibody light-chain variable regions having cytosol-penetrating ability. Table 1 shows the full-length sequences of the human antibody light-chain variable regions, numbered according to the Kabat numbering system, and Table 2 shows the CDR sequences of the antibody sequences shown in Table 1.

(82) TABLE-US-00005 TABLE1 Full-lengthsequencesofcytosol-penetrating humanantibodylight-chainvariableregions Namesof light-chain variable SEQ regions Sequence IDNO: hT2VL 11020abc 29 DLVMTQSPATLSLSPGERATLSCKSSQSLF def304050 NSRTRKNYLAWYQQKPGQAPRLLIYWASTR 607080 ESGIPDRFSGSGSGTDFTLTISSLEPEDFA 90100 VYYCKQSYYHMYTFGQGTKVEIKR hT3VL 11020abc 30 DLVMTQSPSSLSASVGDRVTITCKSSQSLF def304050 NSRTRKNYLAWYQQKPGKAPKLLIYWASTR 607080 ESGVPSRFSGSGSGTDFTLTISSLQPEDFA 90100 TYYCKQSYYHMYTFGQGTKVEIKR hT4VL 11020abc 31 DLVMTQSPSSLSASVGDRVTITCKSSQSLF def304050 NSRTRKNYLAWYQQKPGKAPKLLIYWASTR 607080 ESGVPSRFSGSGSGTDFTLTISSLQPEDFA 90100 TYYCQQYYYHMYTFGQGTKVEIKR

(83) TABLE-US-00006 TABLE2 CDRsequencesofcytosol-penetratinghuman antibodylight-chainvariableregions. Namesof light-chain variable CDR1 SEQ CDR2 SEQ regions Sequence ID Sequence ID KabatNo. 24 25 26 27 27a 27b 27c 27d 27e 27f 28 29 30 31 32 33 34 NO: 50 51 52 53 54 55 56 NO: hT2VL K S S Q S L F N S R T R K N Y L A 32 W A S T R E S 33 hT3VL K S S Q S L F N S R T R K N Y L A 35 W A S T R E S 36 hT4VL K S S Q S L F N S R T R K N Y L A 38 W A S T R E S 39 Namesof light-chain variable CDR3 SEQ regions Sequence ID KabatNo. 89 90 91 92 93 94 95 96 97 NO: hT2VL K Q S Y Y H M Y T 34 hT3VL K Q S Y Y H M Y T 37 hT4VL Q Q Y Y Y H M Y T 40

Example 8: Development of Cytotransmab by Substitution with Cytosol-Penetrating Humanized Light-Chain Region (VL), and Expression and Purification of Cytotransmab

(84) FIG. 8 is a schematic view showing a method of substituting a light-chain variable region having no cell-penetrating ability with a humanized light-chain variable region having cytosol-penetrating ability in order to construct cytotransmab.

(85) Specifically, in order to construct a heavy-chain expression vector for producing an intact immunoglobulin-type monoclonal antibody, a DNA encoding a heavy chain comprising an antibody heavy-chain variable region (Bevacizumab VH, Adalimumab VH, or humanized hT0 VH) and a heavy-chain constant region (CH1-hinge-CH2-CH3), which has a secretion signal peptide-encoding DNA fused to the 5 end, was cloned into a pcDNA3.4 vector (Invitrogen) by NotI/HindIII. Furthermore, in order to construct a vector that expresses a light chain, a DNA encoding either a cytosol-penetrating light-chain variable region (hT4 VL) or the light-chain variable region (Bevacizumab VL, or Adalimumab VL) and light-chain constant region (CL) of a model antibody, which a secretion signal peptide-encoding DNA fused to the 5 end, was cloned into a pcDNA3.4 vector (Invitrogen) by use of NotI/HindIII.

(86) The light-chain and heavy-chain expression vectors were transiently transfected, and the proteins were expressed and purified, followed by comparison of the yield of the proteins. In a shaking flask, HEK293-F cells (Invitrogen) suspension-growing in serum-free FreeStyle 293 expression medium (Invitrogen) were transfected with a mixture of plasmid and polyethylenimine (PEI) (Polyscience). After 200 mL transfection in a shaking flask (Corning), HEK293-F cells were seeded into 100 ml of medium at a density of 2.010.sup.6 cells/ml, and cultured at 150 rpm and in 8% CO.sub.2. To produce each monoclonal antibody, a suitable heavy-chain and light-chain plasmid were diluted in 10 ml of FreeStyle 293 expression medium (Invitrogen) (125 g heavy chain, 125 g light chain, a total of 250 g (2.5 g/ml)), and the dilution was mixed with 10 ml of medium containing 750 g (7.5 g/ml) of PEI, and the mixture was incubated at room temperature for 10 minutes. The incubate medium mixture was added to 100 ml of the seeded cell culture which was then cultured at 150 rpm in 8% CO.sub.2 for 4 hours, after which 100 ml of FreeStyle 293 expression was added to the cell culture, followed by culture for 6 days. In accordance with the standard protocol, the protein was purified from the collected cell culture supernatant. The antibody was applied to a Protein A Sepharose column (GE Healthcare), and washed with PBS (pH 7.4). The antibody was eluted using 0.1 M glycine buffer (pH 3.0), and then immediately neutralized with 1M Tris buffer. The eluted antibody fraction was concentrated while the buffer was replaced with PBS (pH 7.4) by dialysis. The purified protein was quantified by measuring the absorbance at 280 nm and the absorption coefficient.

(87) Table 3 below shows the yields of purified cytotransmabs and proteins produced per liter of culture volume. Three measurements were statistically processed, and indicates standard deviation values. With respect to the yields of the obtained proteins, cytotransmabs, including hT4 VL improved to facilitate its interaction with a human heavy-chain variable region (VH), did not greatly differ from the wild-type monoclonal antibodies.

(88) TABLE-US-00007 TABLE 3 Comparison of the purification yields of Cytotransmabs with those of wild-type IgG-type monoclonal antibodies (Adalimumab, and Bevacizumab) IgG purification yield (mg/1-liter of transfected IgG clone VH VL cells) TMab2 h3D8 VH hT2 VL 8.0 0.7 TMab3 h3D8 VH hT3 VL 8.2 0.5 TMab4 h3D8 VH hT4 VL 10.8 1.0 Adalimumab Adalimumab VH Adalimumab VL 11.6 0.3 HuT2 Adalimumab VH hT2 VL 2.1 0.6 HuT3 Adalimumab VH hT3 VL 3.5 0.8 HuT4 Adalimumab VH hT4 VL 10.9 0.8 Bevacizumab Bevacizumab VH Bevacizumab VL 8.8 0.4 AvaT4 Bevacizumab VH hT4 VL 8.0 1.1

(89) These results indicate that the humanized light-chain variable region (hT4 VL) obtained by additionally modifying interface residues can optimally interact with a humanized antibody heavy-chain variable region, and thus can be stably expressed and purified.

Example 9: Verification of Cytosol-Penetrating Abilities of Cytotransmab

(90) FIG. 9A shows the results of observing 1-2 cells in various cell lines by confocal microscopy in order to verify the cytosol-penetrating abilities of cytotransmabs having a light-chain variable region replaced with the cytosol-penetrating light-chain region hT4 VL.

(91) Specifically, in a 24-well plate, 510.sup.4 HeLa, PANC-1, HT29 or MCF-7 cells per well were added to 0.5 ml of 10% FBS-containing medium, and cultured for 12 hours under the conditions of 5% CO.sub.2 and 37 C. When the cells were stabilized, each well was incubated with a dilution of each of 1 M of TMab4, Adalimumab (Humira), Bevacizumab (Avastin), HuT4 or AvaT4 in 0.5 ml of fresh medium for 6 hours under the conditions of 37 C. and 5% CO.sub.2. Next, the medium was removed, and each well was washed with PBS, and then treated with a weakly acidic solution (200 mM glycine, 150 mM NaCl (pH 2.5)) to remove proteins from the cell surface. After washing with PBS, the cells were fixed in 4% paraformaldehyde at 25 C. for 10 minutes. Next, each well was washed with PBS, and incubated with PBS buffer containing 0.1% saponin, 0.1% sodium azide and 1% BSA at 25 C. for 10 minutes to pores in the cell membranes. Next, each well was washed with PBS, and then incubated with PBS buffer containing 2% BSA at 25 C. for 1 hour in order to eliminate nonspecific binding. Thereafter, each well was incubated with FITC (green fluorescence)-labeled antibody (Sigma), which specifically recognizes human Fc, at 25 C. for 1.5 hours, and the nucleus was blue-stained with Hoechst33342, and observed with a confocal microscope. Unlike IgG-type monoclonal antibodies (Adalimumab and Bevacizumab) which target extracellularly secreted proteins, TMab4, HuT4 and AvaT4 showed green fluorescence in the cells.

(92) FIG. 9B shows the results of examining cytosol-penetrating ability for several cells, performed at a reduced magnification in order to examine cell-penetrating efficiency in the cytosol-penetrating ability examination experiment by confocal microscopy observation as shown in FIG. 9A.

(93) It was shown that the cytotransmab introduced with the cytosol-penetrating humanized light-chain variable region penetrated the cytosol of all the cells and localized in the cytosol.

Example 10: Evaluation of Cytotoxicity of Cytotransmabs

(94) In order to examine whether or not the cytotransmabs confirmed to have cytosol-penetrating ability in Example 7 would have cytotoxicity in vitro, HeLa or PANC-1 cells were treated with each of TMab4, HuT4, Adalimumab, AvaT4 and Bevacizumab, and the inhibition of growth of the cells was examined by an MTT assay (Sigma).

(95) Specifically, in a 96-well plate, 110.sup.4HeLa or PANC-1 cells per well were cultured in 0.1 ml of 10% FBS-containing medium for 12 hours under the conditions of 37 C. and 5% CO.sub.2. Then, each well was treated with 1 M of each of TMab4, HuT4, Adalimumab, AvaT4 and Bevacizumab for 20 hours or 44 hours, and then 20 l of MTT solution (1 mg/ml PBS) was added to each well, followed by incubation for 4 hours. The formed formazan was dissolved in 200 l of DMSO (dimethyl sulfoxide), and the absorbance at 595 nm was measured to determine cell viability.

(96) FIG. 10A is a graph showing the results obtained by treating HeLa and PANC-1 cell lines with cytotransmab and evaluating the inhibition of growth of the cells in vitro). FIG. 10B is an image showing the results obtained by treating HeLa and PANC-1 cell lines with cytotransmab and evaluating the degree of inhibition of the cells in vitro. As shown in FIGS. 10A and 10B, all the antibodies showed no cytotoxicity.

Example 11: Expression and Purification of Anti-RasGTP iMab, and Analysis of Affinity of Anti-RasGTP iMab for KRas Mutants

(97) The heavy-chain variable region (VH) of cytotransmab, which has the property of penetrating cells and localizing in the cytosol, was replaced with RT4 VH selected in Example 3, thereby constructing anti-RasGTP iMab which can penetrate cells and specifically target GTP-bound Ras in the cytosol. The constructed anti-RasGTP iMab was expressed in animal cells.

(98) Specifically, in order to construct a heavy-chain expression vector for producing an intact immunoglobulin-type monoclonal antibody, a DNA, which has a secretion peptide-encoding DNA fused to the 5 end and encodes a heavy chain comprising an RT11 heavy-chain variable region (RT11 VH) and a heavy-chain constant region (CH1-hinge-CH2-CH3), was cloned into a pcDNA3.4 (Invitrogen) vector by NotI/HindIII. In addition, in order to construct a light-chain expression vector, a DNA, which has a secretion peptide-encoding DNA fused to the 5 end and encodes a light chain encoding a cytosol-penetrating light-chain variable region (hT4 VL) and a light-chain constant region (CL), was cloned into a pcDNA3.4 (Invitrogen) vector by a NotI/HindIII.

(99) The light-chain and heavy-chain expression vectors were transiently transfected, and the proteins were expressed and purified, followed by comparison of the yield of the proteins. In a shaking flask, HEK293-F cells (Invitrogen) suspension-growing in serum-free FreeStyle 293 expression medium (Invitrogen) were transfected with a mixture of plasmid and polyethyleneimine (PEI) (Polyscience). In the case of transfection of 200 mL in a shaking flask (Corning), HEK293-F cells were seeded into 100 ml of medium at a density of 2.010.sup.6 cells/ml, and cultured at 150 rpm and in 8% CO.sub.2. To produce each monoclonal antibody, suitable heavy-chain and light-chain plasmids were diluted in 10 ml of FreeStyle 293 expression medium (Invitrogen) (125 g heavy chain, 125 g light chain, a total of 250 g (2.5 g/ml)), and the dilution was mixed with 10 ml of medium containing 750 g (7.5 g/ml) of PEI, and the mixture was incubated at room temperature for 10 minutes. The incubated medium mixture was added to 100 ml of the seeded cell culture which was then cultured at 150 rpm in 8% CO.sub.2 for 4 hours, after which 100 ml of FreeStyle 293 expression medium was added to the cell culture, followed by culture for 6 days. In accordance with the standard protocol, the protein was purified from the collected cell culture supernatant. The antibody was applied to a Protein A Sepharose column (GE Healthcare), and the column was washed with PBS (pH 7.4). The antibody was eluted using 0.1 M glycine buffer (pH 3.0), and then immediately neutralized with 1M Tris buffer. The eluted antibody fraction was concentrated while the buffer was replaced with PBS (pH 7.4) by dialysis. The purified protein was quantified by measuring the absorbance at 280 nm and the absorption coefficient.

(100) FIG. 11 shows the results of analyzing anti-RasGTP iMab RT4 by 12% SDS-PAGE under reductive or non-reductive conditions after purification.

(101) Specifically, in a non-reductive condition, a molecular weight of about 150 kDa appeared, and in a reductive condition, a heavy-chain molecular weight of about 50 kDa and a light-chain molecular weight of about 25 kDa appeared. This indicates that the expressed and purified anti-RasGTP iMab is present as a monomer in a solution state free of a non-covalent bond, and does not form a dimer or an oligomer by a non-natural disulfide bond.

(102) FIG. 12 shows the results of ELISA performed to measure affinity for GTP-bound and GDP-bound wild-type KRas and GTP-bound and GDP-bound KRas mutants (KRas G12D, KRas G12V, and KRas G13D).

(103) Specifically, each of GTP-bound KRas mutants and GDP-bound KRas mutants, which are target molecules, was incubated in a 96-well EIA/RIA plate (COSTAR Corning) at 37 C. for 1 hour, and then the plate was washed three times with 0.1% TBST (0.1% Tween20, pH 7.4, 137 mM NaCl, 12 mM Tris, 2.7 mM KCl, 5 mM MgCl.sub.2) (SIGMA) for 10 minutes. Next, each well of the plate was incubated with 4% TBSB (4% BSA, pH7.4, 137 mM NaCl, 12 mM Tris, 2.7 mM KCl, 10 mM MgCl.sub.2) (SIGMA) for 1 hour, and then washed three times with 0.1% TBST for 10 minutes. Thereafter, each well was incubated with anti-RasGTP iMab RT4 (and cytotransmab TMab4 having cytosol-penetrating ability only without Ras-binding ability) diluted in 4% TBSB at various concentrations, after which each well was washed three times with 0.1% PBST for 10 minutes. As a marker antibody, goat alkaline phosphatase-conjugated anti-human mAb (SIGMA) was used. Each well was treated with pNPP (p-nitrophenyl palmitate) (SIGMA), and the absorbance at 405 nm was measured.

(104) In order to further quantitatively analyze the affinity of anti-RasGTP iMab RT4 for GTP-bound KRas G12D, SPR (Surface plasmon resonance) was performed using a Biacore 2000 instrument (GE healthcare).

(105) Specifically, anti-RasGTP iMab RT4 was diluted in 10 mM Na-acetate buffer (pH 4.0), and immobilized on a CM5 sensor chip (GE Healthcare) at a concentration of about 1100 response units (RU). For analysis, Tris buffer (20 mM Tris-HCl, pH 8.0, 100 mM NaCl, 5 mM MgCl2, 0.005% Tween 20) was flushed at a flow rate of 30 l/min, and GTP-bound KRas G12D was used at a concentration ranging from 1000 nM to 62.5 nM. After analysis of association and dissociation, regeneration of the CM5 chip was performed by flushing a buffer (10 mM NaOH, 1M NaCl, pH10.0) at a flow rate of 30 l/min for 1.5 minutes. Each of sensorgrams obtained at 3 min of association and 3 min of dissociation was normalized and subtracted from a blank cell, thereby determining affinity.

(106) FIG. 13 shows the results of analyzing the affinity of anti-RasGTP iMab RT4 for GTP-bound KRAS G12D by use of SPR (BIACORE 2000) (GE Healthcare).

Example 12: Examination of Cytosol-Penetrating Ability of Anti-RasGTP iMab RT4

(107) FIG. 14 shows the results of confocal microscopy observation performed to examine the cytosol-penetrating ability of anti-RasGTP iMab RT4. In cells lines (PANC-1, and HCT116) having mutant KRas and cell lines (HT29, HeLa) having wild-type KRas, the cell-penetrating ability of anti-RasGTP iMab RT4 was analyzed.

(108) Specifically, each cell line was added to a 24-well plate at a density of 510.sup.4 cells per well and cultured in 0.5 ml of 10% FBS-containing medium for 12 hours under the conditions of 5% CO.sub.2 and 37 C. When the cells were stabilized, each of TMab4 and RT4, diluted in 0.5 ml of fresh medium at a concentration of 1 M, was added to each well, followed by incubation for 6 hours under the conditions of 37 C. and 5% CO.sub.2. Next, the medium was removed, and each well was washed with PBS, and then treated with a weakly acidic solution (200 mM glycine, 150 mM NaCl (pH 2.5)) to remove proteins from the cell surface. After washing with PBS, the cells were fixed in 4% paraformaldehyde at 25 C. for 10 minutes. Next, each well was washed with PBS, and incubated with PBS buffer containing 0.1% saponin, 0.1% sodium azide and 1% BSA at 25 C. for 10 minutes to pores in the cell membranes. Next, each well was washed with PBS, and then incubated with PBS buffer containing 2% BSA at 25 C. for 1 hour in order to eliminate nonspecific binding. Thereafter, each well was incubated with FITC (green fluorescence)-labeled antibody (Sigma), which specifically recognizes human Fc, at 25 C. for 1.5 hours, and the nucleus was blue-stained with Hoechst33342, and observed with a confocal microscope. It was observed that anti-RasGTP iMab showed fluorescence in the cells, indicating that cytotransmab did not lose its cytosol-penetrating ability, even after it was substituted with the heavy-chain variable region that binds specifically to GTP-bound KRas.

Example 13: Evaluation of Cytotoxicity of Anti-RasGTP iMab RT4

(109) (1) Evaluation of the Effect of Anti-RasGTP iMab on Inhibition of Growth of Adherent Cells

(110) FIG. 15 shows the results obtained by treating NIH3T3, NIH3T3 KRas G12V and NIH3T3 HRas G12V cell lines with anti-RasGTP iMab RT4 and evaluating the inhibition of growth of the cells in vitro.

(111) Specifically, in order to examine whether anti-RasGTP iMab has cytotoxicity specific for KRas mutant-dependent cells in vitro, wild-type KRas NIH3T3 mouse fibroblast cells, NIH3T3 KRas G12V cells having artificially overexpressed Ras mutant, NIH3T3 HRas G12V mutant cells, and KRas G13D mutant human pancreatic cells (PANC-1), were treated with 1 M of each of TMab4 and RT4, and the inhibition of growth of adherent cells was evaluated.

(112) Specifically, each type of NIH3T3 and PANC-1 cells was added to a 24-well plate at a density of 210.sup.3 cells per well and cultured in 0.5 ml of 10% FBS-containing medium for 12 hours under the conditions of 37 C. and 5% CO.sub.2. Next, the cells were treated twice with 1 M of TMab4 or RT4 for 72 hours each time and observed for a total of 144 hours, and then the number of viable cells was counted, thereby determining the degree of growth of the cells.

(113) As shown in FIG. 21, the cells treated with TMab4 showed no cytotoxicity, whereas RT4 inhibited the growth of the KRas mutant cell lines (NIH3T3 KRas G12V, and NIH3T3 HRas G12V), and the NIH3T3 cells showed no cytotoxicity. In addition, the growth of the KRas G13D mutant PANC-1 cells was inhibited. Thus, TMab4 had no cytotoxicity, whereas RT4 inhibited cell growth.

(114) (2) Evaluation of the Effect of Anti-RasGTP iMab RT4 on Inhibition of Growth of Non-Adherent Cells

(115) FIG. 16 shows the results of evaluating the inhibition of growth of non-adherent cells in an NIH3T3 HRas G12V cell line.

(116) Specifically, in order to examine whether anti-RasGTP iMab inhibits the growth of non-adherent cells in KRas mutant cells, NIH3T3 HRas G12V mutant cells were analyzed by a colony typeion assay. Specifically, a mixture of 0.5 ml of 2DMEM medium and 0.5 ml of 1% agrose solution was plated on a 12-well plate and hardened to form 0.5% gel. Then, 0.4 ml of 2DMEM medium, 0.5 ml of 0.7% agarose, and 0.05 ml of 110.sup.3 NIH3T3 HRas G12V cells were mixed with 0.05 ml (20 M) of PBS, TMab4, RT4 or Lonafarnib (20 M), and the mixture was plated on the 0.5% agarose gel and hardened. Thereafter, the 0.35% agarose gel was treated with a dispersion of 1 M of PBS, TMab4, RT4 or Lonafarnib in 0.5 ml of 1DMEM at 3-day intervals for a total of 21 days. On day 21, the cells were stained with NBT (nitro-blue tetrazolium) solution, and then the number of colonies was counted.

(117) Similarly to the results of the above-described experiment on the inhibition of growth of adherent cells, RT4 inhibited colony typeion, whereas TMab4 did not inhibit colony typeion.

(118) The above results indicate that anti-Rasab4lue te RT4 bind specifically to Ras mutants in the cytosol and inhibits the growth of adherent and non-adherent cells.

Example 14: Examination of Whether Anti-RasGTP iMab RT4 Binds Specifically to GTP-Bound KRas in Cells

(119) FIG. 17 shows the results of whether anti-RasGTP iMab RT4 is superimposed with activated HRas G12V mutants in cells. FIG. 18 shows the results of confocal microscopy observation of whether anti-RasG12V te is superimposed with GTP-bound KRas G12V mutants in cells.

(120) Specifically, 24-well plates were coated with fibronectin (Sigma), and then a dilution of 0.5 ml of NIH3T3 cells expressing mCherry (red fluorescence) HRas G12V or mCherry (red fluorescence) KRas G12V was added to the plate at a density of 210.sup.4 cells per well, and cultured for 12 hours under the conditions of 37 C. and 5% CO.sub.2. Next, the cells were treated with 2 M of each of TMab4 and RT4 and cultured at 37 C. for 12 hours. Next, the medium was removed, and each well was washed with PBS, and then treated with a weakly acidic solution (200 mM glycine, 150 mM NaCl (pH 2.5)) to remove proteins from the cell surface. After washing with PBS, the cells were fixed in 4% paraformaldehyde at 25 C. for 10 minutes. Next, each well was washed with PBS, and incubated with PBS buffer containing 0.1% saponin, 0.1% sodium azide and 1% BSA at 25 C. for 10 minutes to pores in the cell membranes. Next, each well was washed with PBS, and then incubated with PBS buffer containing 2% BSA at 25 C. for 1 hour in order to eliminate nonspecific binding. Thereafter, each well was incubated with FITC (green fluorescence)-labeled antibody (Sigma), which specifically recognizes human Fc, at 25 C. for 1.5 hours, and the nucleus was blue-stained with Hoechst33342, and observed with a confocal microscope.

(121) As shown in FIGS. 17 and 18, green fluorescent RT4 was superimposed with the cellular inner membrane in which red-fluorescent activated Ras was located, whereas TMab was not superimposed.

(122) The above experimental results indicate that anti-Ras Rasabhst3334 RT4 bind specifically to GTP-bound Ras in the cells.

Example 15: Evaluation of Cytotoxicity of RGD-Fused Anti-Ras1181t3334 RT4

(123) For in vivo experiments, it is required to impart tumor tissue specificity. Conventional cytotransmabs bind to HSPG on the cell surface, and have no specificity for any other tumor tissue, and for this reason, cannot specifically inhibit the growth of tumors in in vivo experiments. To overcome this problem, an RGD4C peptide (CDCRGDCFC; SEQ ID NO: 41) having specificity for integrin 3 which is overexpressed in angiogenetic cells and various tumors was fused to the N-terminus of the light chain via one GGGGS linker by a genetic engineering method. The RGD4C peptide is characterized in that it has affinity higher than conventional RGD peptides and can be fused using a genetic engineering method, and the specific structure thereof can be maintained even when it is fused to the N-terminus (Koivunen E et al., 1995).

(124) FIG. 19 shows the results obtained by treating HCT116 and PANC-1 cell lines with RGD-TMab4 and RGD-RT4 and evaluating the inhibition of growth of the cells in vitro.

(125) In order to examine whether RGD-TMab4 and RGD-RT4 themselves have cytotoxicity in vitro, human colorectal cancer HCT116 cells having a KRas G13D mutant, and human pancreatic cancer PANC-1 cells having a KRas G12D mutant, were treated with each of RGD-TMab4 and RGD-RT4, and the inhibition of growth of the cells was evaluated.

(126) Specifically, each type of HCT116 and PANC-1 cells was added to a 24-well plate at a density of 510.sup.3 cells per well, and cultured in 0.5 ml of 10% FBS-containing medium for 12 hours under the conditions of 37 C. and 5% CO.sub.2. Next, the cells were treated twice with 1 M of each of RGD-TMab4 and RGD-RT4 for 72 hours each time, and observed for a total of 144 hours, and then the number of the cells was counted, thereby determining the degree of growth of the cells.

(127) As shown in FIG. 19, RGD-TMab4 inhibited the growth of HCT116 cells by about 20% and inhibited the growth of PANC-1 cells by about 15%, and RGD-RT4 inhibited the growth of HCT116 and PANC-1 cells by about 40% and about 50%, respectively. According to previous studies, the RGD4C peptide has an affinity for integrin 5, which is about 3 times lower than that for integrin 3. However, integrin 3 is overexpressed mainly in angiogenetic cells, and integrin 5 is expressed in various tumor cells. Thus, the RGD4C peptide has the ability to bind 5 of HCT116 and PANC-1 cells to thereby inhibit cell adhesion (Cao L et al., 2008).

(128) Thus, RGD4C peptide-fused TMab4 does not appear to have cytotoxicity. In addition, a comparison between RGD-TMab4 and RGD-RT4 indirectly confirmed that TMab4 can inhibit Ras-specific cell growth even when the RGD is fused thereto.

Example 16: Examination of the Effect of RGD-Fused Anti-Ras1181b4alC on Inhibition of Tumor Growth

(129) FIG. 20A shows the results of analyzing the tumor growth inhibitory effect of RGD-fused anti-RasGTP iMab RT4 in mice xenografted with HCT116 cells. FIG. 20B is a graph showing the results of measuring the body weight of mice in order to examine the non-specific side effects of RGD-fused anti-RasGTP iMab RT4.

(130) Specifically, in order to examine the tumor growth inhibitory effect of RGD-RT4 in vivo based on the in vitro experiment results of Example 15, KRas G13D mutant human colorectal HCT116 cells were injected subcutaneously into Balb/c nude mice at a density of 510.sup.6 cells per mice. After about 6 days when the tumor volume reached about 50 mm.sup.3, the mice were injected intravenously with 20 mg/kg of each of PBS, RGD-TMab4 and RGD-RT4. The injection was performed a total of 9 times at 2-day intervals, and the tumor volume was measured using a caliper for 18 days.

(131) As shown in FIG. 20A, unlike the control PBS, RGD-TMab4 and RGD-RT4 inhibited the growth of cancer cells, and RGD-RT4 more effectively inhibited tumor growth compared to RGD-TMab4. In addition, as shown in FIG. 20B, there was no change in the body weight of the test group treated with RGD-RT4, indicating that RGD-RT4 has no other toxicities.

Example 17: Construction and Screening of Library for Improving Affinity of Anti-RasGTP iMab RT4

(132) Anti-RasGTP iMab RT4 shows Ras-specific biological activity, but the affinity thereof determined by SPR analysis is about 110 nM. Thus, it has a very low affinity for antigen, even though it is an IgG-type antibody. In order to improve this shortcoming and to allow anti-RasGTP iMab RT4 to exhibit increased biological activity even at low concentration, it is required to improve the affinity of anti-RasGTP iMab RT4.

(133) FIG. 21A shows a strategy of constructing a human heavy-chain variable region library to improve the affinity of RT4. To improve affinity, CDR3 (residues 95 to 100a) playing an important role in binding to antigen was designed to have lengths of 6 residues (library 6), 7 residues (library 7) and 9 residues (library 9), and a degenerated codon (NNK) capable of encoding all amino acid residues was used. In addition, to improve the affinity and retain the antigen-binding site of RT4, a spiked oligomer capable of maintaining wild-type RT4 residues at a ratio of 50% was used for CDR1 (residues 31 to 33) and CDR2 (residues 50 and 52 to 56), which show high solvent accessibility. In this technology, a primer is designed such that the percentage of wild-type nucleotides in three nucleotides encoding an amino acid for each residue is maintained at 79% and the percentage of the remaining nucleotides is 7% so that wild-type amino acids in a PCR process will be maintained at 50%.

(134) FIG. 21B is a schematic view showing a method of constructing a designed library by a PCR technique and transforming the constructed library onto the yeast surface by homologous combination with a heavy-chain single yeast surface display vector (pYDS-H) treated with the restriction enzymes NheI and ApaI.

(135) Specifically, a DNA encoding each of the designed libraries was amplified by a PCR technique, and then enriched by ethanol precipitation. A pYDS-H heavy-chain yeast surface display vector for homologous recombination was treated with NheI and ApaI restriction enzymes, after which it was purified by agarose gel extraction and enriched by ethanol precipitation. For 12 g of each library-encoding DNA, 5 g of a vector was transformed into mating type A yeast JAR200 for yeast surface display by electroporation (Baek D. S and Kim Y. S, 2014; Lorenzo B et al., 2010), followed by serial dilution. The number of colonies in the selection medium SD-CAA+URA (20 g/L glucose, 6.7 g/L yeast nitrogen base without amino acids, 5.4 g/L Na2HPO.sub.4, 8.6 g/L NaH.sub.2PO.sub.4, 5 g/L casamino acids, 0.2 mg/L uracil) was measured to determine the size of the library.

(136) In each library screening process, according to the method as shown in Example 3 and 4, 1.sup.st MACS was performed for GTP-bound KRas G12D at an antigen concentration of 100 nM using the yeast library displaying the heavy-chain variable region alone. Then, for Fab libraries by yeast mating, clones specific for GTP-bound KRas G12D were selected through competitive binding to GDP-bound KRas G12D that was not biotinylated in 1.sup.st, 2.sup.nd and 3.sup.rd FACS.

(137) FIG. 22 shows the results of FACS analysis performed to determine the affinity of library 6 (which is a library having a CDR3 length of 6 residues) for GTP-bound KRas G12D and GDP-bound KRas G12D for library-expressing yeast in each step in order to confirm enrichment specific for GTP-bound KRas G12D in the above-described library screening process. As shown therein, the screened library did bind specifically to GTP-bound KRas G12D, and showed a higher affinity than RT4 used as a template.

(138) FIG. 23 shows the results of sequencing of individual clones using the three libraries. As shown therein, only residues in the CDR region having mutations induced by the library were mutated.

(139) Table 4 shows the human antibody heavy-chain variable region (VH) sequences of individual clones selected from the libraries having improved affinity by use of RT4 as a template, and, Table 5 below show the sequences of CDR1, CDR2 and CDR3 of the selected heavy-chain variable region (VH) sequences specific for RasGTP.

(140) TABLE-US-00008 TABLE4 Human antibody heavy-chain variable region (VH) sequences showing specific affinity for Ras .Math. GTP, used in anti-Ras .Math. GTP iMab Namesof heavy- chain variable SEQID regions Sequence NO: RT4 102030 SEQID EVQLVESGGGLVQPGGSLRLSCAASGFTFS NO:1 4050a SYAMSWVRQAPGKGLEWVSTISRSGHSTYY 607080abc ADSVKGRFTISRDNSKNTLYLQMNSLRAED 90a101110 TAVYYCAKRFGSIVFDYWGQGTLVTVSS RT11 102030 SEQID EVQLVESGGGLVQPGGSLRLSCAASGFTFS NO:2 4050a SYSMSWVRQAPGKGLEWVSYISRTSHTTYY 607080abc ADSVKGRFTISRDNSKNTLYLQMNSLRAED 90a101110 TAVYYCARGFF---MDYWGQGTLVTVSS RT13 102030 SEQID EVQLVESGGGLVQPGGSLRLSCAASGFTFS NO:3 4050a TFSMSWVRQAPGKGLEWVSYISRTSHTTYY 607080abc ADSVKGRFTISRDNSKNTLYLQMNSLRAED 90a101110 TAVYYCARGTFG--FDYWGQGTLVTVSS RT14 102030 SEQID EVQLVESGGGLVQPGGSLRLSCAASGFTFS NO:4 4050a TFSMSWVRQAPGKGLEWVSYISRTSHTTYY 607080abc ADSVKGRFTISRDNSKNTLYLQMNSLRAED 90a101110 TAVYYCARPRGW--FDYWGQGTLVTVSS RT15 102030 SEQID EVQLVESGGGLVQPGGSLRLSCAASGFTFS NO:5 4050a TFSMSWVRQAPGKGLEWVSYISRTSHTTYY 607080abc ADSVKGRFTISRDNSKNTLYLQMNSLRAED 90a101110 TAVYYCAKRFGS--FDYWGQGTLVTVSS RT16 102030 SEQID EVQLVESGGGLVQPGGSLRLSCAASGFTFS NO:6 4050a TFSMSWVRQAPGKGLEWVSYISRTSHTTYY 607080abc ADSVKGRFTISRDNSKNTLYLQMNSLRAED 90a101110 TAVYYCARSSGRFVFDYWGQGTLVTVSS RT17 102030 SEQID EVQLVESGGGLVQPGGSLRLSCAASGFTFS NO:7 4050a TFSMSWVRQAPGKGLEWVSYISRTSHTTYY 607080abc ADSVKGRFTISRDNSKNTLYLQMNSLRAED 90a101110 TAVYYCAKGRFGSVFDYWGQGTLVTVSS

(141) TABLE-US-00009 TABLE5 CDR sequences of human antibody heavy-chain variable region (VH) showing specififc affinity for Ras .Math. GTP, used in anti-Ras .Math. GTP iMab Namesof heavy- chain variable regions CDR1 SEQ CDR2 SEQ Kabat Sequence ID Sequence ID No. 31 32 33 34 35 NO: 50 51 52 52a 53 54 55 56 57 58 59 60 61 62 63 64 65 NO: RT4 S Y A M S 8 T I S R S G H S T Y Y A D S V K G 9 RT11 S Y S M S 11 Y I S R T S H T T Y Y A D S V K G 12 RT13 T F S M S 14 Y I S R T S H T T Y Y A D S V K G 15 RT14 T F S M S 17 Y I S R T S H T T Y Y A D S V K G 18 RT15 T F S M S 20 Y I S R T S H T T Y Y A D S V K G 21 RT16 T F S M S 23 Y I S R T S H T T Y Y A D S V K G 24 RT17 T F S M S 26 Y I S R T S H T T Y Y A D S V K G 27 Namesof heavy- chain variable regions CDR3 SEQ Kabat Sequence ID No. 95 96 97 98 99 100 100a 101 102 NO: RT4 R F G S I V F D Y 10 RT11 G F F M D Y 13 RT13 G T F G F D Y 16 RT14 P R G W F D Y 19 RT15 R F G S F D Y 22 RT16 S S G R F V F D Y 25 RT17 G R F G S V F D Y 28

Example 18: Expression and Purification of Anti-RasGTP iMab Having Improved Affinity

(142) As described in Example 11, a heavy chain comprising the heavy-chain variable region, obtained by library screening and having an improved affinity for RasGTP, a heavy-chain constant region (CH1-hinge-CH2-CH3), was cloned into an animal expression vector. The expression vector and a vector expressing a cytosol-penetrating humanized light-chain were transiently co-transfected into HEK293F protein-expression cells. Anti-RasGTP iMab was expressed in the cells and purified in the same manner as described in Example 11.

(143) FIG. 24 shows the results of analyzing anti-RASGTP iMab having improved affinity by 12% SDS-PAGE under a reductive or non-reductive condition.

(144) Specifically, as described in Example 11, in a non-reductive condition, a molecular weight of about 150 kDa appeared, and in a reductive condition, a heavy-chain molecular weight of about 50 kDa and a light-chain molecular weight of about 25 kDa appeared. This indicates that the expressed and purified anti-RasGTP iMab is present as a monomer in a solution state and does not form a dimer or an oligomer by a non-natural disulfide bond.

Example 19: Examination of Cell-Penetrating Ability of RasGTP iMab Having Improved Affinity

(145) FIG. 25 shows the results obtained by replacing the heavy-chain variable region of anti-RasGTP iMab with a RasGTP-specific heavy-chain variable region having improved affinity and then performing confocal microscopic observation to confirm whether or not the anti-RasGTP iMab has the ability to penetrate cells.

(146) Specifically, HeLa cells were added to each well of a 24-well plate at a cell density of 510.sup.4 cells per well with 0.5 ml of 10% FBS-containing medium, and cultured for 12 hours under the conditions of 5% CO.sub.2 and 37 C. When the cells were stabilized, a 1:17 dilution of each of TMab4, RT11, RT13, RT14, RT15, RT16 and RT17 in 0.5 ml of fresh medium was added to each well which was then incubated for 6 hours under the conditions of 5% CO.sub.2 and 37 C. A subsequent process was performed in the same manner as the RT4 staining process described in Example 14. The intracellular fluorescence of RT11, RT13, RT14, RT15, RT16 and RT17, which are anti-Ras. GTP iMab having improved affinity, were observed, indicating that they have the ability to penetrate cells.

Example 20: Analysis of GTP-Bound Ras-Specific Affinity of Anti-RasGTP iMab Clones Having Improved Affinity

(147) FIG. 26A shows the results of ELISA performed to measure the affinity of the anti-RasGTP iMab clones having improved affinity for GTP-bound KRas G12D and GDP-bound KRas G12D.

(148) Specifically, according to the same method as described in Example 11, each of GTP-bound KRas mutants and GDP-bound KRas mutants, which are target molecules, was incubated in a 96-well EIA/RIA plate (COSTAR Corning) at 37 C. for 1 hour, and then the plate was washed three times with 0.1% TBST (0.1% Tween20, pH 7.4, 137 mM NaCl, 12 mM Tris, 2.7 mM KCl, 5 mM MgCl.sub.2) (SIGMA) for 10 minutes. Next, each well of the plate was incubated with 4% TBSB (4% BSA, pH7.4, 137 mM NaCl, 12 mM Tris, 2.7 mM KCl, 10 mM MgCl.sub.2) (SIGMA) for 1 hour, and then washed three times with 0.1% TBST for 10 minutes. Thereafter, each well was incubated with each of the anti-RasGTP iMab clones diluted in 4% TBSB at various concentrations, after which each well was washed three times with 0.1% TBST for 10 minutes. As a marker antibody, goat alkaline phosphatase-conjugated anti-human mAb (SIGMA) was used. Each well was incubated with Ultra TMB-ELISA substrate solution (Thermo Scientific), and then the absorbance at 450 nm was measured.

(149) As shown in FIG. 26A, among the anti-Ras. GTP iMab clones having improved affinity, RT11 was selected as a clone having a high specific affinity for GTP-bound KRas G12D.

(150) FIG. 26B shows the results of ELISA analysis performed to confirm the affinity of RT11, selected based on the ELISA-based binding analysis, for various GTP-bound Ras mutants.

(151) Specifically, using the same ELISA method used in the above-described analysis of the affinity of anti-RasGTP iMab having improved affinity, the affinities of anti-RasGTP iMab RT11 for GTP- or GDP-bound wild-type KRas, KRas G12D, KRas G12V, KRas G13D, wild-type HRas and HRas G12V, were analyzed.

(152) As shown in FIG. 26B, anti-RasGTP iMab RT11 did bind to various GTP-bound Ras mutants.

Example 21: Quantitative Analysis of the Affinity of Anti-RasGTP iMab RT11 for KRas G12D

(153) In order to quantitatively analyze the affinity of anti-RasGTP iMab RT11 for GTP-bound KRas G12D, SPR (surface plasmon resonance) was performed using a Biacore 2000 instrument (GE Healthcare).

(154) FIG. 27A shows the results of analyzing the affinity of anti-RasGTP iMab RT11 for GTP-bound KRas G12D by use of SPR (BIACORE 2000) (GE Healthcare).

(155) FIG. 27B is a sensorgram showing the results of analyzing the affinity of RT11 for GTP- or GDP-bound KRas G12D at the highest concentration (1000 nM).

(156) Specifically, according to the same method as described in Example 11, anti-RasGTP iMab RT11 was immobilized on a CM5 sensor chip (GE Healthcare) at a concentration of about 1100 response units (RU). For analysis, Tris buffer (20 mM Tris-HCl, pH 8.0, 100 mM NaCl, 5 mM MgCl.sub.2, 0.005% Tween 20) was flushed at a flow rate of 30 l/min, and GTP- or GDP-bound KRas G12D was used at a concentration ranging from 1000 nM to 62.5 nM.

(157) As a result, it was shown that RT11 did bind to KRas G12D with high affinity (12.9 nM).

Example 22: Analysis of the Ability of Anti-RasGTP iMab RT11 to Inhibit Binding Between GTP-Bound KRas and Raf

(158) FIG. 28 shows the results of a competitive ELISA performed to confirm whether anti-RasGTP iMab RT11 can inhibit the binding between the effector molecule Raf and intracellular KRas.

(159) Specifically, the Ras binding site (RBD: 1-149) fragment of the effector protein cRaf (NM_002880.2) was cloned into the E. coli expression vector pGEX-3X by the restriction enzymes BamHI/EcoRI, and then expressed and purified according to the same method as described in Example 2. The purified cRaf-RBD was incubated in a 96-well EIA/RIA plate (COSTAR Corning) at 37 C. for 1 hour, and then the plate was washed three times with 0.1% TBST (0.1% Tween20, pH 7.4, 137 mM NaCl, 12 mM Tris, 2.7 mM KCl, 5 mM MgCl.sub.2) (SIGMA) for 10 minutes. Each well of the plate was incubated with 4% TBSB (4% BSA, pH7.4, 137 mM NaCl, 12 mM Tris, 2.7 mM KCl, 10 mM MgCl.sub.2) (SIGMA) for 1 hour, and then washed three times with 0.1% TBST for 10 minutes. Next, each concentration of anti-RasGTP iMab RT11 diluted in 4% TBSB at various concentrations (ranging from 1 M to 5.64 pM) was incubated with 1 M of biotinylated GTP-bound KRas G12D, and then each well was washed three times with 0.1% TBST for 10 minutes. As a marker antibody, goat alkaline phosphatase-conjugated anti-human mAb (SIGMA) was used. Each well was incubated with pNPP (p-nitrophenyl palmitate) (SIGMA), and the absorbance at 405 nm was measured.

(160) As shown in FIG. 28, anti-RasGTP iMab RT11 showed the ability to inhibit binding to the effector protein cRaf (IC.sub.50=35 nM).

Example 23: Examination of the Ability of Anti-RasGTP iMab RT11 to Penetrate Various Tumor Cells

(161) FIG. 29 shows the results of confocal microscopic observation performed to confirm whether anti-RasGTP iMab having improved affinity has the ability to penetrates various types of tumor cells. Various tumor cell lines, including human colorectal cancer cell lines (SW480 (KRasG12V mutant), PANC-1 (KRas G12D mutant), DLD-1 (KRas G13D mutant), HCT116 (KRas G13D mutant)), and a human fibrosarcoma cell line (HT1080 (NRas Q61L mutant), were used as Ras mutant cell lines, and a human breast cancer cell line (MCF7) and human colorectal cancer cell lines (HT29, CaCo2, Colo320DM) were used as Ras wild-type cell lines.

(162) Specifically, each of the above-described Ras mutant and Ras wild-type cell lines was added to each of a 24-well plate at a density of 510.sup.4 cells with 0.5 ml of 10% FBS-containing medium, and cultured for 12 hours under the conditions of 5% CO.sub.2 and 37 C. When the cells were stabilized, each of TMab4 and RT11 diluted in fresh well at a concentration of 2 M was added to each well which was then incubated for 12 hours under the conditions of 37 C. and 5% CO.sub.2. A subsequent process was performed in the same manner as the RT4 staining process described in Example 14. As a result, anti-RasGTP iMab RT11 having improved affinity showed fluorescence in various types of tumor cells, indicating that it has the ability to penetrate various tumor cell lines, in the same manner as TMab4.

Example 24: Examination of the Ability of Anti-RasGTP iMab RT11 to Remain in Cytosol

(163) FIG. 30 shows the results of confocal microscopic observation performed using a non-cell-penetrating, self-quenching dye (calcein (Sigma)) to observe the cytosol-remaining ability of anti-RasGTP iMab having improved affinity.

(164) Specifically, HCT116 cells were added to each well of a 24-well plate at a density of 510.sup.4 cells per well with 0.5 ml of 10% FBS-containing medium, and cultured for 12 hours under the conditions of 37 C. and 5% CO.sub.2. Next, each well was treated with 1 M of TMab4 and RT4 for 4 hours, and then treated with 100 M of Calcein for 2 hours. Thereafter, the medium was removed, and each well was washed with PBS, and then treated with a weakly acidic solution (200 mM glycine, 150 mM NaCl pH 2.5) to remove calcein from the cell surface. After washing with PBS, the cells were fixed with 4% paraformaldehyde at 25 C. for 10 minutes. Next, each well was washed with PBS, and the nucleus was blue-stained with Hoechst33342 and observed with a confocal microscope. As shown in FIG. 30, both the anti-RasGTP iMab RT11 and the cytotransmab TMab4 showed calcein fluorescence throughout the cytosol. However, PBS showed only vesicle-shaped fluorescence. Such results indicate that anti-RasGTP iMab RT11 remained in the cytosol.

Example 25: Evaluation of Cytotoxicity of Anti-RasGTP iMab RT11

(165) FIG. 31 shows the results obtained by treating various Ras wild-type and Ras mutant cell lines with anti-RasGTP iMab RT11 and evaluating in vitro the inhibition of growth of the cells, and FIG. 32 are a set of images showing the results of polarizing microscopic observation performed to the cell density of each cell line.

(166) Specifically, in order to examine in vitro whether anti-RasGTP iMab RT11 has cytotoxicity specific for Ras mutant cell lines, the inhibition of growth of cells was evaluated using mouse NIH3T3 fibroblast cells and human colorectal cancer Colo320DM cells as Ras wild-type cell lines and using mouse NIH3T3 KRas G12V mutant cells, human colorectal cancer cell lines (HCT116 cells (KRas G13D), HCT116 (KRas G13D), SW480 (KRas G12V), DLD-1 (KRas G13D)) and a human pancreatic cell line (PANC-1 (KRas G12D)).

(167) Specifically, each type of the above-described cell lines was added to each well of a 24-well plate at a density of 2-510.sup.3 cells per well with 0.5 ml of 10% FBS-containing medium, and cultured for 12 hours under the conditions of 37 C. and 5% CO.sub.2. Next, each well was treated twice with each of TMab4 and RT11 for 72 hours each time and observed for a total of 144 hours, and then the number of viable cells was counted, thereby determining the degree of growth of the cells.

(168) As shown in FIGS. 31 and 32, TMab4 showed no cytotoxicity, whereas RT11 inhibited the growth of only Ras mutant cells (NIH3T3 KRas G12V, HCT116, PANC-1, SW480, and DLD-1), and showed no cytotoxicity in Ras wild-type cell lines (NIH3T3, and Colo320DM).

Example 26: Examination of the Abilities of Anti-RasGTP iMab RT11 to Bind Specifically to Intracellular Activated Ras and to Inhibit Binding Between Activated Ras and Effector Protein

(169) (1) Examination of the Ability of Anti-RasGTP iMab RT11 to Bind Specifically to Intracellular RasGTP

(170) FIG. 33 shows the results of confocal microscopic observation performed to examine whether RT11 is superimposed with activated KRas G12V mutants in cells.

(171) Specifically, 24-well plates were coated with fibronectin (Sigma), and then 0.5 ml of a dilution of NIH3T3 cells expressing mCherry (red fluorescence) HRas G12V was added to the plate at a density of 210.sup.2 cells per well, and cultured for 12 hours under the conditions of 37 C. and 5% CO.sub.2. Next, the cells were treated with 2 M of each of TMab4 and RT11 and cultured at 37 C. for 12 hours. Thereafter, the cells were stained under the same conditions as described in Example 14, and were observed with a confocal microscope.

(172) As shown in FIG. 33, green fluorescent RT11 was superimposed with the cellular inner membrane in which red-fluorescent activated Ras was located, whereas TMab was not superimposed.

(173) FIG. 34 shows the results of an immunoprecipitation assay performed to confirm whether RT11 binds to activated Ras in cells.

(174) Specifically, 10 ml of a dilution of each of a KRas G12V mutant-expressing NIH3T3 cell line and HCT116 cell line was added to a 100 mm.sup.3 plate at a density of 210.sup.6 cells per well, and cultured for 12 hours under the conditions of 37 C. and 5% CO.sub.2. Next, the cells were treated with 2 M of each of TMab4 and RT11 and cultured at 37 C. for 12 hours. Thereafter, the cells were lysed using a cell lysis buffer (25 mM Tris-Cl pH 7.4, 150 mM NaCl, 1% NP-40, 10 mM MgCl.sub.2, 10% glycerol, protease inhibitors), and the cell debris was removed by precipitation. Protein A/G agarose was added to the cell lysate and incubated for 2 hours, and then the antibody was precipitated. Next, Western blot analysis was performed using anti-KRas antibody (Santa Cruz) and human Fc antibody (Sigma).

(175) As shown in FIG. 34, KRas was observed only in RT11, but was not observed in TMab4 and PBS.

(176) Such experimental results indicate that RT11 binds specifically to intracellular activated Ras.

(177) (2) Examination of the Ability of Anti-RasGTP iMab RT11 to Inhibit Binding Between RasGTP and Effector Molecule

(178) FIGS. 35A and 35B show the results of an immunoprecipitation assay performed to examine whether or not RT11 inhibits the binding between RasGTP and effector proteins.

(179) Specifically, 10 ml of a dilution of each of a KRas G12V mutant-expressing NIH3T3 cell line and HCT116 cell line was added to a 100 mm.sup.3 plate at a density of 210.sup.6 cells per well, and cultured for 12 hours under the conditions of 37 C. and 5% CO.sub.2. Next, the cells were treated with 2 M of each of TMab4 and RT11 and cultured at 37 C. for 12 hours. Thereafter, the cells were lysed using a cell lysis buffer (25 mM Tris-Cl pH 7.4, 150 mM NaCl, 1% NP-40, 10 mM MgCl.sub.2, 10% glycerol, protease inhibitors), and the cell debris was removed by precipitation. The KRas G12V mutant cell lysate was incubated with anti-HA antibody (Covance) for 2 hours, and then treated with Protein A/G agarose to precipitate the anti-HA antibody. Raf-1 RBD agarose (Millipore) was added to the HCT116 cell lysate and incubated for 2 hours, and then precipitated. Next, Western blot analysis was performed using anti B-Rat C-Raf, PI3K and KRas antibodies (Santa Cruz) and human Fc antibody (Sigma).

(180) As shown in FIG. 35A, anti-RasGTP iMab RT11 inhibited the binding between Ras. GTP and effector proteins (B-Raf and C-Raf), whereas TMab4 did not inhibit the binding. Similarly, FIG. 35B shows that only the anti-RasGTP iMab RT11 inhibited the binding between the effector protein C-Raf and RasGTP, whereas TMab4 did not inhibit the binding.

(181) Such experimental results indicate that RT11 binds specifically to intracellular RasGTP to thereby inhibit the binding between RasGTP and the effector proteins (B-Raf, and C-Raf).

Example 27: Construction of RGD10 Peptide-Fused Anti-RasGTP iMab RT11 and Analysis of the Ability to Bind to RasGTP

(182) As described in Example 15, anti-RasGTP iMab RT11 penetrates by binding to HSPG on the cell surface. Thus, it is required to impart tissue specificity to anti-RasGTP iMab RT11 for in vivo experiments. For this, an RGD10 peptide (DGARYCRGDCFDG; SEQ ID NO: 42) having specificity for integrin 3 which is overexpressed in angiogenetic cells and various tumors was fused to the N-terminus of the light chain via a linker consisting of a total of 10 residues (GGGGSGGGGS) by a genetic engineering method. The RGD10 peptide will have an affinity for integrin, which is similar to that of a previous RGD4C peptide fused to RT4, and it has one disulfide bond in the peptide, and thus is expected to be more easily fused to the N-terminus of the antibody. Thus, the RGD10 peptide was fused to anti-RasGTP iMab RT11 by a genetic engineering method.

(183) FIG. 36 shows the ELISA results obtained by measuring the affinities of the constructed RGD10 peptide-fused RT11 for a variety of GTP-bound and GDP-bound Ras mutants.

(184) Specifically, according to the same method as described in Example 11, each of GTP-bound KRas G12D and GDP-bound Ras, which are target molecules, was incubated in a 96-well EIA/RIA plate (COSTAR Corning) at 37 C. for 1 hour, and then the plate was washed three times with 0.1% TBST (0.1% Tween20, pH 7.4, 137 mM NaCl, 12 mM Tris, 2.7 mM KCl, 5 mM MgCl.sub.2) (SIGMA) for 10 minutes. Next, each well of the plate was incubated with 4% TBSB (4% BSA, pH7.4, 137 mM NaCl, 12 mM Tris, 2.7 mM KCl, 10 mM MgCl.sub.2) (SIGMA) for 1 hour, and then washed three times with 0.1% TBST for 10 minutes. Thereafter, each well was incubated with each of the anti-RasGTP iMab clones diluted in 4% TBSB at a concentration of 10 nM, after which each well was washed three times with 0.1% TBST for 10 minutes. As a marker antibody, goat alkaline phosphatase-conjugated anti-human mAb (SIGMA) was used. Each well was incubated with Ultra TMB-ELISA substrate solution (Thermo Scientific), and then the absorbance at 450 nm was measured.

(185) As shown in FIG. 36, RGD10 peptide-fused RT11 (RGD10-RT11) showed the same affinity for GTP-bound Ras mutants.

Example 28: Evaluation of Cytotoxicity of RGD10-Fused Anti-RasGTP iMab RT11

(186) FIGS. 37 and 38 show the results obtained by treating Colo320DM, HCT116, PANC-1, SW480 and DLD-1 cell lines with RGD10-TMab4 and RGD10-RT11 and evaluating in vitro the inhibition of growth of the cells.

(187) In order to evaluate in vitro whether RGD10-TMab4 and RGD-RT11 themselves have cytotoxicity, the inhibition of growth of cells was evaluated using human colorectal cancer Colo320DM cells as a Ras wild-type cell line and using human colorectal cancer cell lines (HCT116 (KRas G13D), SW480 (KRas G12V), DLD-1 (KRas G13D)) and a human pancreatic cancer cell line (PANC-1 (KRas G12D)).

(188) Specifically, cells were added to each well of a 24-well plate at a density of 510.sup.3 cells per well with 0.5 ml of 10% FBS-containing medium, and cultured for 12 hours under the conditions of 37 C. and 5% CO.sub.2. Next, the cells were treated twice with 1 M of each of RGD10-TMab4 and RGD10-RT11 for 72 hours each time and observed for a total of 144 hours, and then the number of viable cells was counted to determine the degree of growth of the cells.

(189) As shown in FIG. 37, the comparison between RGD10-TMab4 and RGD10-RT11 indicated that the KRas mutant cell lines (HCT116, SW480, DLD-1, and PANC-1) showed a difference in cell growth of about 8-12%, whereas the Ras wild-type cell line showed no difference in cell growth. Thus, the comparison between RGD10-TMab4 and RGD10-RT11 indicated that RT11 can inhibit the growth of Ras-specific cells even when the RGD10 peptide is fused thereto.

Example 29: Examination of Whether RGD10-Fused Anti-RasGTP iMab RT11 Binds Specifically to Integrin 3

(190) FIG. 39 shows the results of analysis performed to examine whether or not RGD10-TMab4 and RGD10-RT11 bind specifically to integrin 3 on the cell surface.

(191) Specifically, each of a K562 cell line and a K562 integrin 3-overexpressing cell line was added to 1.5 ml at a density of 210.sup.5 cells, then washed twice with washing buffer (pH 7.4 PBS, 2% FBS). 100 nM of each of TMab4, RGD10-TMab4 and RGD10-RT11 100 nM was mixed with 300 IU/ml of heparin (Sigma), and the cells were incubated with the mixture at 4 C. for 1 hour. The cells were washed twice with washing buffer, and then stained with an Alexa488 (green fluorescence)-labeled antibody (Invitrogen) that specifically recognizes human IgG, at 4 for 1 hour. Next, the cells were washed twice with washing buffer, and analyzed by FACS.

(192) As shown in FIG. 39, unlike TMab4, RGD10-TMAb4 and RGD10-RT11 did bind specifically to the K562 integrin 3 cells. This suggests that the RGD10 peptide binds specifically to integrin 3.

Example 30: Examination of Whether Anti-RasGTP iMab RT11 Binds Specifically to Intracellular RasGTP

(193) FIG. 40 shows the results of confocal microscopic observation performed to examine whether or not RGD10-RT11 is superimposed with an activated KRas G12V mutant in cells.

(194) Specifically, a 24-well plate was coated with fibronectin (Sigma), and then 0.5 ml of a dilution of mCherry (red fluorescence) KRas G12V-expressing NIH3T3 cells were added to each well at a density of 210.sup.2 cells per well and cultured for 12 hours under the conditions of 37 C. and 5% CO.sub.2. Then, the cells were treated and incubated with 1 M of each of RGD10-TMab4 and RGD10-RT11 for 12 hours under the conditions of 37 C. and 12 hours. Next, the cells were stained under the same conditions as described in Example 14, and were observed with a confocal microscope.

(195) As shown in FIG. 40, green fluorescent RGD10-RT11 was superimposed with the cellular inner membrane in which red-fluorescent activated Ras was located, whereas RGD10-TMab was not superimposed.

(196) Such experimental results indicate that RGD10-RT11 binds specifically to intracellular activated Ras.