Method of analyzing binding aspect of membrane protein in a living cell

Abstract

The present invention relates to a method for analyzing the pattern of live intercellular membrane protein binding. The method for analyzing the pattern according to the present invention can analyze accurately, sensitively, quickly, and readily the binding pattern of a target membrane protein and a candidate substance to be specifically bound therewith without tagging to a ligand, and thus measure directly and accurately the position and quantitative information of the binding of the membrane protein and the target substance. Such effects make it possible to apply the method for various uses such as dissociation constant, mutant study, complex formation, and signal transduction. Moreover, it is expected to use the method for searching out undiscovered membrane proteins and target substances.

Claims

1. A method of analyzing a binding aspect between a candidate material and a target membrane protein in a living cell, the method comprising: obtaining diffusion coefficients of the target membrane protein before and after treatment with the candidate material in the living cell expressing the target membrane protein, and analyzing a change in diffusion coefficient of the target membrane protein obtained thereby.

2. The method of claim 1, wherein the analysis of the binding aspect between the target membrane protein and the candidate material includes the analysis of binding between the target membrane protein and the candidate material, a ratio of target membrane proteins that the candidate material binds to among the total target membrane proteins, a molecular weight of the candidate material bound to the target membrane protein, a dissociation constant between the target membrane protein and the candidate material, or a process of forming a complex between the target membrane protein and the candidate material.

3. The method of claim 1, wherein the target membrane protein is selected from the group consisting of an integral membrane protein, a peripheral membrane protein, a transmembrane protein, a membrane glycoprotein and a lipid anchored membrane protein.

4. The method of claim 1, wherein the candidate material includes at least one selected from the group consisting of a compound, a nucleic acid, a saccharide, a carbohydrate, a lipid, a peptide, and a protein.

5. The method of claim 1, wherein the diffusion coefficient of the target membrane protein is obtained by detecting the motion of the membrane protein in a cell membrane by single particle tracking (SPT).

6. The method of claim 5, wherein the diffusion coefficient is obtained by Equations 1 and 2: $\begin{array}{cc}MSD\left(\Delta \right)=\frac{1}{N-\Delta }\underset{n=1}{\overset{N-\Delta }{.Math.}}\left({\left(x_{n+\Delta }-x_n\right)}^2+{\left(y_{n+\Delta }-y_n\right)}^2\right)& \left[\mathrm{Equation}1\right]\end{array}$ where Δ is the step size between coordinates of the target membrane protein particle, wherein Δ is a positive integer, MSD(Δ) is the mean square displacement of a target membrane protein particle with respect to the step size between coordinates of the target membrane protein particle, N is the total number of a pair of coordinates in one trajectory for the target membrane protein particle, (x.sub.n, y.sub.n) are the coordinates of the target membrane protein particle at an n.sup.th-numbered position in one trajectory, (x.sub.1, y.sub.1) are the coordinates of the target membrane protein particle at the start point in one trajectory, (x.sub.N,y.sub.N) are the coordinates of the target membrane protein particle at the end point in one trajectory, and (x.sub.n+Δ, y.sub.n+Δ) are the coordinates of the target membrane protein particle at an n+Δ-numbered position in one trajectory, however, n+Δ is the same as or smaller than N; and
MSD(Δ)=4DΔ  [Equation 2] where D is the diffusion coefficient, and Δ is the step size between coordinates of the target membrane protein particle.

7. The method of claim 1, wherein the change in the diffusion coefficient is obtained by Equation 3:
Change in diffusion coefficient (%)=100*|1−(D.sub.c2/D.sub.c1)|  [Equation 3] where D.sub.c1 is the diffusion coefficient of the target membrane protein at a concentration c1 of the candidate material in a peripheral environment of the cell at a single cell level, and D.sub.c2 is the diffusion coefficient of the target membrane protein at a concentration c2 of the candidate material in a peripheral environment of the cell at a single cell level.

8. The method of claim 7, wherein, when the change in the diffusion coefficient (%) obtained by Equation 3 is 5% or more, the candidate material is determined as a ligand binding to the target membrane protein.

9. The method of claim 5, wherein the detection of the movement of the membrane protein in the cell membrane is performed by detecting fluorescence signal of a fluorescent protein in a cell expressing a fusion protein of the target membrane protein and the fluorescent protein.

10. The method of claim 9, wherein the fluorescent protein includes one or more selected from the group consisting of a green fluorescent protein (GFP) type, a blue fluorescent protein (BFP) type, a cyan type, a yellow fluorescent protein (YFP) type, a red fluorescent protein (RFP) type, an orange type, a far-red type, a near-IR, a photoactivatable protein, a photoconvertible protein, and a photoswitchable protein.

11. The method of claim 9, wherein the fluorescent protein includes one or more selected from enhanced green fluorescent protein (EGFP), Emerald, Superfolder GFP, azami green mWasabi, TagGFP, AcGFP, T-sapphire, mUKG, Clover, mNeonGreen, enhanced blue fluorescent protein (EBFP), EBFP2, Azurite, mTagBFP, mKalama1, Sirius, enhanced cyan fluorescent protein (ECFP), monomeric ECFP (mECFP), Cerulean, mTurquoise, mTurquoise2, CyPet, TagCFP, mTFP1 (Teal), SCFP3A, monomeric Midoriishi Cyan, enhanced yellow fluorescent protein (EYFP), Topaz, Benus, mCitrine, YPet, TagYFP, PhiYFP, mBanana, SYFP2, mRuby, mRuby2, mApple, mStrawberry, mRFP1, mCherry, mRaspberry, dKeima-Tandem (monomeric version), HcRed-Tandem (monomeric version), mPlum, mKate2, mNeptune, mKate2, mNeptune, TagRFP657, IFP1.4, PA-GFP, PAmCherry1, PaTagRFP, PS-CFP2, mEos2, mEos3.2, PSmOrange and Dronpa.

12. The method of claim 1, wherein the cell expressing the fusion protein of the target membrane protein and the fluorescent protein is selected from the group consisting of a human embryonic kidney (HEK) cell, HEK 293 cell, 3T3-L1 cell, C6 cell, Chinese hamster ovary (CHO) cell, CHOK1 cell, NIH/3T3 cell, baby hamster kidney (BHK) cell, COS1 cell, COS7 cell, HaCaT cell, HeLa cell, HeLa S3 cell, HepG2 cell, HL-60 cell, HUV-EC-C cell, Jurkat cell, K-562 cell, L6 cell, MCF7 cell, MDCK cell, NIH/3T3 cell, RAW 264.7 cell, RBL-1 cell, SH-SY5Y cell and U-937 cell.

13. The method of claim 5, wherein the detection of the movement of the membrane protein in the cell membrane is performed with a fluorescent material-conjugated probe specifically binding to the target membrane protein.

14. The method of claim 13, wherein the probe is an antibody, an aptamer, or a non-antibody protein scaffold.

15. The method of claim 13, wherein the fluorescent material is an organic fluorescent dye.

16. The method of claim 15, wherein the organic fluorescent dye is selected from Atto 488, Alexa Flour 488, Dy505, Rhodamine 123, Atto 520, Dy 530, ATTO 532, Alexa Fluor 532, Fluorescein, FITC, Cy2, Cy3B, Alexa Flour 568, TAMRA, Cy3, Cy3.5, SNAP-Cell TMR-Star, Atto 565, Atto 590, Alexa Fluor 647, Cy5, Atto 647, Atto 647N, Dyomics 654, Atto 655, TMP-ATTO 655, Atto 680, Cy5.5, Atto 680, Alexa Fluor 680, Atto 700, Alexa Fluor 700, DyLight 750, Cy7, Alexa Flour 750, Atto 740, Alexa Flour 790, and IRDye 800 CW.

17. The method of claim 1, further comprising: measuring the diffusion coefficient of the target membrane protein after the treatment with the candidate material at several time points, and analyzing a change in a diffusion coefficient of at least one target membrane protein obtained thereby over time.

18. The method of claim 17, further comprising: determining membrane protein-specific endocytosis by analyzing the change in the diffusion coefficient of the target membrane protein over time.

19. The method of claim 1, wherein the candidate material is a drug.

Description

DESCRIPTION OF DRAWINGS

(1) FIG. 1 is a schematic diagram of a system that can only specifically observe a target protein among various membrane proteins present in a cell membrane by labeling the target membrane protein with a fluorescent protein, under the environment of which binding of a ligand makes diffusion of the target protein slow.

(2) FIG. 2 is a diagram illustrating a process of tracking the motion of a target membrane protein of one cell attached to a round coverslip by TIRF and spatio-temporally averaging the tracking result to analyze the diffusion coefficient of the target membrane protein in a single cell.

(3) FIG. 3 illustrates randomly selected trajectories of PMT-mEos3.2 and EGFR-mEos3.2 before and after treatment with cetuximab to the host cells expressing each protein respectively, in which PMT has almost no change in length of the trajectory before and after the treatment with cetuximab, but EGFR shows absolutely shorter length of the trajectory after the treatment with cetuximab.

(4) FIG. 4 illustrates the means and standard deviations of diffusion coefficients of PMT and EGFR before and after the treatment with cetuximab for 10 cycles according to Example 1.5.

(5) FIG. 5 illustrates the quantitatively summarized results of FIG. 3.

(6) FIG. 6 illustrates that β2-AR statically is not bound with a Gi-protein and thus has no difference even by the treatment with Pertussis toxin (PTX), but as FZD1 is bound with a Gi-protein, when treated with PTX, ADP of the Gi-protein is ribosylated, thereby inhibiting the binding with FZD1.

(7) FIG. 7 illustrates changes in diffusion coefficients of PMT, FZD1 and β2-AR before and after PMT, FZD1 and β2-AR are treated with PTX according to Example 2.4.

(8) FIG. 8 illustrates absolute values of averaged diffusion coefficients of PMT, β2-AR, EGFR according to Example 2.4.

(9) FIG. 9 illustrates changes of diffusion coefficients of EGFR measured in different environments according to Example 3, in which batch#1, batch#2 and batch #3 represent changes in diffusion coefficient of EGFR in COS7 cells prepared in three different batches, and batch#3-1 and batch#3-2 are in the same batch, but represent changes in diffusion coefficient of EGFR measured at different glasses.

(10) FIG. 10 illustrates changes in diffusion coefficient of EGFR in HEK293, HELA and CHO-K1 cells according to Examples 4 to 6.

(11) FIG. 11 is a diagram illustrating antibodies capable of binding to EGFR at different sites, for example, cetuximab, mAb 199.12, mAb 528 and mAb R-1.

(12) FIG. 12 illustrates western blotting results according to Comparative Example 1, in which NT represents a non-treatment state, and EGF+/− represents treatment/non-treatment of EGF.

(13) FIG. 13 illustrates changes in diffusion coefficients of PMT and EGFR according to Example 7.

(14) FIG. 14 illustrates cetuximab specifically binding to EGFR, and trastuzumab specifically binding to ErbB2.

(15) FIG. 15 illustrates flow cytometry results according to Comparative Example 2, in which, particularly, the x axis represents the intensity of a fluorescent signal, and the y axis represents a relative ratio of the cell population having a specific level of the intensity of a fluorescent signal.

(16) FIG. 16 is a graph illustrating changes in diffusion coefficients before and after PMT and EGFR are treated with cetuximab, trastuzumab, and an anti-actin antibody according to Example 8.

(17) FIG. 17 is diagrams illustrating that EGFR is bound with four different ligands having different molecular weights at the same binding position.

(18) FIG. 18 illustrates changes in diffusion coefficient of EGFR before and after PMT and EGFR are treated with cetuximab, cetuximab Fab, cetuximab F(ab′)2, and EGFR-specific binding aptamer according to Example 9.

(19) FIG. 19 illustrates that the molecular weight of the ligand and the change in diffusion coefficient have a linear relationship according to Example 9.

(20) FIG. 20 is a diagram illustrating that EGFR-bound cetuximab increases in number as the concentration of cetuximab is increased according to Example 10.1.

(21) FIG. 21 illustrates reduction in diffusion coefficient of EGFR as the concentration of cetuximab is increased from 0.02 to 1.28 μg/ml in a single cell.

(22) FIG. 22 is a graph of a dissociation constant and cooperativity obtained by calculating the ratio of EGFR-bound cetuximab by increasing the concentration of cetuximab.

(23) FIG. 23 illustrates that EGFR wild-type (WT, wild-type) can be bound with both of cetuximab and an mAb R-1 antibody, but a mutant, EGFRvIII, is bound with only cetuximab, and not an mAb R-1 antibody.

(24) FIG. 24 illustrates that PMT is not bound with two antibodies such as cetuximab and mAb R-1, and thus has no reduction in diffusion coefficient of PMT, but EGFRvIII is not bound with mAb R-1 but bound with cetuximab and thus shows different change in diffusion coefficient.

(25) FIG. 25 is a graph of a dissociation constant and cooperativity obtained by calculating a ratio of EGFRvIII-bound cetuximab by the method described in Example 10.1.

(26) FIG. 26 illustrates that cetuximab, a secondary antibody, and a tertiary antibody that bind to EGFR in turn.

(27) FIG. 27 illustrates that the diffusion coefficient of EGFR is gradually reduced when EGFR is treated with the cetuximab, secondary antibody, and tertiary antibody in turn.

(28) FIG. 28 illustrates the changes in diffusion coefficient of EGFR when the EGFR is sequentially treated with the cetuximab, tertiary antibody, and secondary antibody.

(29) FIG. 29 illustrates a process of detecting endogenous EGFR by conjugating a photoswitchable fluorescent organic dye, Alexa Fluor 647, to a Fab fragment capable of specifically binding to the endogenous EGFR.

(30) FIG. 30 illustrates that the binding of mAb 199.12 to the endogenous EGFR does not interfere with the binding of cetuximab.

(31) FIG. 31 illustrates a specific binding aspect between the endogenous EGFR and cetuximab.

(32) FIG. 32 is the quantitative results of the means and standard deviations of diffusion coefficients changed by treatment of an immunoglobulin antibody and cetuximab in five different cell lines.

(33) FIG. 33 is images of mouse primary cells classified by marker expressions of cell types in lung cell tissues and autofluorescence levels.

(34) FIG. 34 illustrates the changes in diffusion coefficient depending on a binding aspect of cetuximab to endogenous EGFRs present in four types of cells selected in FIG. 33, which are BASCs, endothelial cells, AT2 cells, and Clara cells.

BEST MODE

(35) Hereinafter, the present invention will be described in further detail with reference to the following examples. However, the following examples are merely provided to illustrate the present invention, and the scope of the present invention is not limited to the following examples.

EXAMPLES

Example 1. Measurement of Target Membrane Protein: EGFR

(36) 1.1 Plasmid for Preparing EGFR-mEos3.2 Fusion Protein

(37) EGFR WT (Addgene plasmid #11011) used in the experiment was provided by Dr. Matthew Meyerson, and mEos3.2 fluorescent protein was provided by Dr. Tao xu. The given fluorescent protein was subcloned from pEGFP-N1/mEos3.2 to form pcDNA3.1/mEos3.2-his in order to facilitate binding with a membrane protein. To prepare a protein in which the membrane protein EGFR WT (SEQ ID NO: 1) is linked with the fluorescent protein mEos3.2 (SEQ ID NO: 2), pcDNA3.1/EGFR WT-mEos3.2-His was constructed using restriction enzymes such as Xbal (SEQ ID NO: 3, TCTAGA) and NotI (SEQ ID NO: 4, GCGGCCGC).

(38) 1.2 Plasmid for Preparing PMT-mEos3.2 Fusion Protein

(39) A plasmid in which PMT is linked with a fluorescent protein was prepared to be used as a control group for Example 1.1. DNA (SEQ ID NO: 5) corresponding to PMT was inserted into the subcloned pcDNA3.1/mEos3.2-his using restriction enzymes such as Xbal (SEQ ID NO: 3) and NotI (SEQ ID NO: 4), thereby constructing pcDNA3.1/PMT-mEos3.2-His.

(40) 1.3 Preparation of Host Cells for Expressing Fusion Protein

(41) COS7 cells provided by ATCC were grown in DMEM (Dulbecco's Modified Eagle Medium, Lonza) under conditions of 37° C., 5% CO.sub.2, and 95% humidity, with 10% FBS (Gibco), and COS7 cells were transfected with each of the plasmids constructed in Examples 1.1 and 1.2 using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instruction. To reduce the influence of protein overexpression, cells showing especially low expression level were sorted from the cells grown for 24 hours after transfection through flow cytometry. In detail, the transfected cells were washed with PBS to remove debris, detached from a bottom using a detaching buffer, and collected and strained with a cell strainer (40 μm, BD Bioscience). Afterward, the cells showing low expression level were specifically sorted using a MoFlo™ XDP cell sorter (Beckman Coulter).

(42) The sorted cells were seeded on a 25-mm-diameter coverslip. Before use, the coverslip was prepared by procedures of sonication in acetone at 42° C. for 30 minutes to minimize autofluorescence of the coverslip, washing with distilled water three times, sonication again in 1% hydrofluoric acid at 42° C. for 10 minutes, washing with distilled water about 15 to 20 times to completely remove hydrofluoric acid, and finally soaking in 100% ethanol under UV light for 30 minutes to be sterilized. As a result, the coverslips on which host cells expressing the EGFR-mEos fusion protein and host cells expressing the PMT-mEos fusion protein were seeded were obtained.

(43) 1.4. Identification of Membrane Protein to which Ligand Did not Bind

(44) The coverslip on which the cells were seeded as described in Example 1.3 was placed onto a microscope with a live cell chamber (37° C., 5% CO.sub.2) and an electron multiplying charge coupled device (EM-CCD) and observed the cells at a single cell level. The microscope, the Olympus IX71 model, is designed based on total internal reflection fluorescence (TIRF), and can be used to observe only fluorescent molecules at a plasma membrane of a cell closely attached (within about 200 nm from a coverslip) to a round coverslip. mEos3.2 fluorescent proteins were randomly converted from a green form into a red from by cleaving their backbone using 405 nm laser, and an image of the randomly converted red-form mEos3.2 fluorescent protein was obtained by an electron multiplying charge coupled device (EM-CCD), ixon3 897, manufactured by Andor Technology, which can detect a signal emitted from a single fluorescent molecule using 561 nm laser.

(45) The fluorescent images were sequentially obtained at regular intervals of time (about 150 ms), and a centroid was found from the signal of a single molecule present on each image through centroid fitting, and the moving distance per unit time and the trajectory of the fluorescent protein were determined by using multiple tracking between consecutive images, and a mean square displacement (MSD) was calculated by Equation 1.

(46) FIG. 1 is a schematic diagram of a system that can specifically observe a target protein among many membrane proteins present in a cell membrane by labeling the target membrane protein with a fluorescent protein, under an environment in which binding of a ligand makes diffusion of the target protein slow.

(47) FIG. 2 is a diagram illustrating a process of tracking the motion of a target membrane protein in a single cell attached to a round coverslip by TIRF and spatio-temporally averaging the tracking result to analyze the diffusion coefficient of the target membrane protein in a single cell. A ligand treatment gave a different environment in a single cell, and the trajectory of a single target membrane protein over time was acquired. To obtain statistical significance, the above procedures were repeated, and the values obtained therefrom were spatio-temporally averaged, thereby overcoming heterogeneity of the cell itself, and enabled to be compared with diffusion of the target membrane protein measured under different conditions.

(48) 1.5. Identification of Ligand-Bound Target Membrane Protein

(49) Each group of the host cells expressing PMT-mEos3.2 and the host cells expressing EGFR-mEos3.2 obtained in Example 1.3 was treated with 20 μg/ml of cetuximab, and the changes of diffusion were observed before and after the treatment by the method described in Example 1.4. The results are shown in FIG. 3.

(50) FIG. 3 illustrates randomly selected trajectories in the host cells expressing PMT-mEos3.2 and the host cells expressing EGFR-mEos3.2 before and after treatment with cetuximab, in which PMT has almost no change in length of the trajectory before and after the treatment with cetuximab, but EGFR shows an obviously shorter length of the trajectory after the treatment with cetuximab. Non-treatment (NT) shown in FIG. 3 represents a state in which nothing is applied to an external environment, and cetuximab represents the trajectories measured after the treatment with cetuximab. In each image, the enlarged image represents the trajectories of several fluorescent molecules randomly extracted from a corresponding region. It can be seen that in non-treated NT (left), both of the PMT and EGFR membrane proteins are relatively freely diffused.

(51) However, it can be seen that, in the cetuximab (right), the motion patterns are different. Since PMT do not bind with cetuximab, there is no significant change in the length of the trajectory of the fluorescent molecule. However, since EGFR directly binds with cetuximab, it can be observed that the length of the trajectory of the fluorescent molecule becomes obviously shorter.

(52) To prove that the result of the change shown in FIG. 3 is not caused by a certain factor generated during the experiment, but by EGFR and an antibody binding thereto, changes of other factors occurring before and after the treatment with cetuximab were observed. In each experiment, a total of 4,000 images were obtained, trajectories are obtained from these images, diffusion coefficients of PMT and EGFR were calculated by Equations 1 and 2, and then the mean and standard deviation of the diffusion coefficients are shown in FIG. 4.

(53) FIG. 4 illustrates the means and standard deviations of diffusion coefficients of PMT and EGFR before and after the treatment with cetuximab for 10 cycles, and it can be seen from FIG. 4 that the diffusion coefficient of EGFR is changed by cetuximab, and not by other imaging factors during measurement.

(54) FIG. 5 illustrates the quantitatively summarized results of FIG. 3, which show changes in diffusion coefficients (%) of PMT and EGFR after treatment with cetuximab, compared with the diffusion coefficients of non-treated PMT and EGFR. While PMT had almost no change in diffusion coefficient before and after the treatment with cetuximab, EGFR showed a change in diffusion coefficient of about 40% after the treatment with cetuximab. In detail, the diffusion coefficients may be obtained by Equations 1 and 2, and when a ratio of the change in diffusion coefficient is calculated by Equation 3, provided that diffusion coefficients obtained under different conditions were D.sub.c1 and D.sub.c2, PMT has almost the same D.sub.c1 and D.sub.c2 values, thereby having a very small change in diffusion coefficient, but the D.sub.c2 value of EGFR is about 60% of the D.sub.c1 value thereof, and thus it can be seen that the diffusion coefficient is reduced by about 40%.

(55) As shown in FIGS. 1 to 5, an increase in size caused by the binding between the target membrane protein and the ligand leads to reduction in diffusion coefficient of the target membrane protein, and a degree of the reduction can be quantified. It could be confirmed that PMT that cannot bind to cetuximab had no significant change in diffusion coefficient before and after the treatment with cetuximab, but EGFR was reduced in diffusion coefficient after the treatment with cetuximab, and the reduction degree was about 40%.

Example 2. Measurement of Target Membrane Protein: GPCR

(56) 2.1 Plasmid for Preparing β2-AR-mEos3.2 Fusion Protein

(57) β2-AR (Addgene plasmid #14697) used in the experiment was provided by Dr. Robert Lefkowitz, and β2-AR DNA (SEQ ID NO: 6) was inserted into the pcDNA3.1/mEos3.2-his subcloned in Example 1.1 using restriction enzymes such as BamHI and Xbal, thereby constructing pcDNA3.1/β2-AR-mEos3.2-His.

(58) 2.2 Plasmid for Preparing FZD1-mEos3 Fusion Protein

(59) FZD1 (Addgene plasmid #16819) used in the experiment was provided by Dr. Randall Moon, and FZD1 DNA (SEQ ID NO: 7) was inserted into the pcDNA3.1/mEos3.2-his subcloned in Example 1.1 using restriction enzymes such as BamHI and Xbal, thereby constructing pcDNA3.1/FZD1-mEos3.2-His.

(60) 2.3 Preparation of Host Cells for Expressing Fusion Protein

(61) For an experiment to inactivate GPCR, COS7 cells were transfected with β2-AR-mEos3.2, FZD1-mEos3.2, and PMT-mEos3.2 (refer to Example 1.2) by the same method as described in Example 1.3, and each group of the transfected COS7 cells were starved in 1% serum (minimum condition)-containing medium for 16 hours. Also, for a pertussis toxin (PTX) experiment, prior to measurement of a membrane protein, each group of the transfected COS7 cells were treated with PTX at a concentration of 100 ng/ml for 6 hours.

(62) 2.4 Identification of Ligand-Bound Target Membrane Protein

(63) Diffusion coefficients of PMT, FZD1 and β2-AR were measured by the same method described in Example 1.5, and the results are shown in FIGS. 7 and 8.

(64) FIG. 6 illustrates that β2-AR basically do not bind with a Gi-protein, but as FZD1 bind with a Gi-protein, when treated with PTX, ADP of the Gi-protein is ribosylated, thereby inhibiting the binding with FZD1. According to the above-described mechanism, regardless of the presence of PTX, β2-AR has no change in diffusion coefficient, and FZD increases in diffusion coefficient after the treatment with PTX.

(65) FIG. 7 illustrates changes in diffusion coefficients of PMT, FZD1, β2-AR before and after treatment with PTX.

(66) FIG. 8 shows absolute values of mean diffusion coefficients of PMT, β2-AR, and EGFR, in which since β2-AR is faster than EGFR, the binding to β2-AR is more sensitively measured than EGFR.

(67) As shown in FIGS. 7 and 8, the method can be even applied to GPCR without limitation to EGFR. This means that it can be applied to other membrane proteins, and also can detect intracellular binding as well as extracellular binding. Also, FIGS. 7 and 8 show that the method can detect the binding between EGFR and cetuximab, and the dissociation of FZD1 from the Gi-protein.

Example 3. Measurement of Target Membrane Protein in Different Environments

(68) To prove technical reproducibility of the method described in Example 1, three groups of transfected COS7 cells were prepared on different days (batch#1, batch#2, and batch#3), two groups of transfected COS7 cells (batch#3-1 and batch#3-2) were prepared independently but on the same day, and then coverslips on which these batches were seeded were prepared. The host cells were treated with cetuximab by the same method as described in Example 1.4, and then a change in diffusion coefficient of EGFR was measured, and the results are shown in FIG. 9.

(69) FIG. 9 illustrates the changes in diffusion coefficient of EGFR measured by the above-described method, in which batch#1, batch#2, and batch#3 represent the changes in diffusion coefficient of EGFR in the COS7 cells prepared in three different batches, and batch#3-1 and batch#3-2 represent changes in diffusion coefficient of EGFR measured at different glasses, but in the same batch.

(70) As shown in FIG. 9, all of the batches prepared on different days and the batches independently prepared on the same day showed no difference in change in diffusion coefficient of EGFR, and it was confirmed by ANOVA analysis that there is no statistical difference between the batches (p=0.768).

Examples 4-6. Identification of Target Membrane Protein (Cell Lines: HEK293, HELA, and CHO-K1)

(71) To demonstrate that target membrane proteins can be measured in other cell lines than COS7 cells, host cells expressing PMT-mEos3.2 and host cells expressing EGFR-mEos3.2 were prepared using HEK293, HELA, and CHO-K1 cells provided by ATCC instead of COS7 cells used in Example 1.3 by the method described in Example 1.4, and changes in diffusion coefficients of EGFR and PMT were measured by treating each group of the host cells with cetuximab by the method described in Example 1.5, and the results are shown in FIG. 10. The results obtained using HEK293, HELA, and CHO-K1 cells as host cells were described in Examples 4 to 6, respectively.

(72) FIG. 10 illustrates changes in diffusion coefficient of EGFR in HEK293, HELA, and CHO-K1 by the above-described method, and like the results obtained with COS7 cells, in all cell lines, a reduced change in diffusion coefficient of EGFR caused by cetuximab was able to be measured.

Comparative Example 1. Measurement of Binding with Ligand: Western Blotting

(73) As a comparative example to analyze the binding between a target membrane protein and a ligand, western blotting was carried out.

(74) Ligand mAb 199.12 was purchased from Invitrogen, ligand mAb 528/mAb R-1 was purchased from Santa Cruz, anti-pEGFR (Y1068) antibody was purchased from Invitrogen, and anti-total EGFR and anti-pErk1/2 antibodies were purchased from Signaling, and an anti-actin antibody was purchased from MP biomedical.

(75) COS7 cells were lysed with radioimmunoprecipitation (RIPA) buffer. The cell lysate was loaded in a 6 to 16% gradient SDS-PAGE gel to allow proteins to migrate in an electrical field. The proteins separated in the SDS-PAGE gel were transferred to a nitrocellulose membrane, treated with a primary antibody at 4° C. overnight, and then treated with a secondary antibody tagged with horseradish peroxidase (HRP) or IRDye800CW (Li-COR) at room temperature for 1 hour. The presence and amount of the proteins were detected using chemiluminescence (ECL system, Pierce) or infra-red fluorescence (Odyssey system, Li-COR), and the results are shown in FIG. 12.

(76) FIG. 11 illustrates cetuximab, mAb 199.12, mAb 528 and mAb R-1 which can bind to EGFR at different sites.

(77) FIG. 12 illustrates western blotting results, in which NT represents no treatment with an antibody, and EGF+/− represents treatment/non-treatment of EGF. As shown in FIG. 12, cetuximab and mAb 528 prevent phosphorylation of EGFR, and also prevent phosphorylation of Erk1/2 in a different manner, and mAb 199.12 and mAb R-1 do not effectively prevent the phosphorylation of EGFR. In detail, EGFR is activated by EGF, and pEGFR may be detected using an antibody that can specifically bind to the activated EGFR.

(78) Since EGFR is not activated without EGF, it was not detected using pEGFR antibody, but in an EGF-existing environment, EGFR is activated by EGF, and the activated EGFR was detected using the pEGFR antibody, and therefore the same results as described are shown. pErk1/2 is a main marker to confirm downstream signaling during signal transduction, which maintains a minimal signal transduction in the EGF-free environment, and increases in the amount in the EGF-existing environment, and therefore it can be seen that the downstream signal transduction is effectively performed.

(79) Actin is one of the main proteins composing a cell, which serves as a marker to confirm that a similar amount of proteins are used in each experiment when several western blotting analyzes are performed.

(80) When EGF, and cetuximab or mAb 528 are present at the same time, EGF binding site on EGFR is interfered with cetuximab and mAb 528, and as a result, it inhibits the EGFR activation, whose tendency is seen through western blotting. However, mAb 199.12 and mAb R-1 do not show the tendency of inhibiting EGFR activation in the EGF-existing environment through the western blotting result, and thus it can be noted that mAb 199.12 and mAb R-1 do not have an influence on the EGFR activation caused by EGF.

Example 7. Measurement of Bonding with Ligand

(81) The host cells expressing PMT-mEos3.2 and the host cells expressing EGFR-mEos3.2, which were obtained in Example 1.3, were treated with cetuximab, mAb 199.12, mAb 528, and mAb R-1, and diffusion coefficients were measured by the method described in Example 1.5. In detail, images were obtained in a non-treatment state (about 10 cycles), and after each type of the antibodies was mixed with the cell culture medium up to the final concentration of 20 μg/ml to induce binding of the antibody to a membrane protein present on the cell surface, images were obtained (about 10 cycles). Changes of diffusion coefficient before and after the host cells expressing PMT-mEos3.2 and the host cells expressing EGFR-mEos3.2 were treated with the corresponding antibodies were calculated by Equation 3, and the results are shown in FIG. 13.

(82) FIG. 13 illustrates changes in diffusion coefficient, in which PMT do not specifically bind with cetuximab, mAb 199.12, mAb 528 and mAb R-1, and thus there is no change in diffusion coefficient before and after the treatment with the antibodies, but as EGFR specifically binds with all of cetuximab, mAb 199.12, mAb 528, and mAb R-1, the diffusion coefficient is changed by about 40% before and after the treatment.

(83) As shown in FIG. 13, PMT has no change in diffusion coefficient before and after the treatment with the antibodies, but EGFR is changed in diffusion coefficient by about 40% before and after the treatment with the antibodies. This shows that the change in diffusion coefficient measured by the above-described method is influenced by direct binding, and is capable of being measured regardless of binding sites. These advantageous characteristics cannot be obtained using fluorescence resonance energy transfer (FRET) and protein fragment complementation assay (PCA).

Comparative Example 2. Molecular Specificity Distinction Test: Flow Cytometry

(84) As a comparative example to analyze a molecular specificity distinction test between a target membrane protein and a ligand, flow cytometry was carried out.

(85) Alexa Fluor 647 fluorescent molecules were conjugated to immunoglobulins (anti-Human (A21445)/anti-Mouse (A21235). Invitrogen), cetuximab, trastuzumab (Roche), and an anti-actin antibody ((691001), MP Biomedicals) to be employed for flow cytometry, and the flow cytometry was performed using Gallios manufactured by BD bioscience. In detail, COS7 cells were treated with the Alexa Fluor 647-conjugated antibody to induce binding with a membrane protein, and a fluorescent signal attached to the surface of the cell was observed by flow cytometry, and the results are shown in FIG. 15.

(86) FIG. 14 illustrates cetuximab specifically binding to EGFR, and trastuzumab specifically binding to ErbB2.

(87) FIG. 15 illustrates flow cytometry results, in which the x axis represents the intensity of a fluorescent signal, and the y axis represents a relative ratio of the cell populations having a specific level of intensity of a fluorescent signal. That is, as the graph goes to the right, the fluorescent signal becomes stronger, which is interpreted that more fluorescent dyes are attached to the cell surface.

(88) As shown in FIG. 15, it can be seen that cetuximab or trastuzumab is relatively well attached to the cell surface (the immunoglobulin or actin serves as a control that does not specifically bind to a membrane protein present on the cell surface), but this approach cannot show to which membrane protein molecules cetuximab or trastuzumab binds. That is, the flow cytometry can reveal the presence or absence of a corresponding membrane protein using a fluorescent molecule-labeled antibody for each of EGFR and ErbB2 endogenously expressed in COS7 cells, but cannot reveal that, especially, cetuximab and trastuzumab specifically bind to EGFR and ErbB2, respectively.

Example 8. Molecular Specificity Distinction Test

(89) Diffusion coefficients of PMT and EGFR were measured for the host cells expressing PMT-mEos3.2 and the host cells expressing EGFR-mEos3.2 obtained in Example 1.3 by the method described in Example 1.5, and in detail, each group of the host cells were treated with cetuximab, trastuzumab or an anti-actin antibody, and a change in diffusion coefficient was measured by Equation 3 using the values of the diffusion coefficients before and after the treatment with the antibody. The results are shown in FIG. 16.

(90) FIG. 16 is a graph illustrating changes in diffusion coefficients before and after the host cells expressing PMT-mEos3.2 and the host cells expressing EGFR-mEos3.2 were treated with the cetuximab, trastuzumab, and anti-actin antibody, which were measured by the above-described method.

(91) As shown in FIG. 16, when the diffusion coefficients of PMT and EGFR were measured by the above-described method, PMT had no change in diffusion coefficient by the treatment with the cetuximab, trastuzumab, or anti-actin antibody, but it was able to be seen that EGFR was changed in diffusion coefficient due to cetuximab. In detail, PMT has almost uniform diffusion coefficients regardless of the treatment of the antibodies as shown in the previous experimental results, and thus has a very small change in diffusion coefficient. However, since cetuximab specifically binds to EGFR, EGFR has a change in diffusion coefficient, but it does not show any changes in diffusion coefficient by trastuzumab and the anti-actin antibody which do not specifically bind to EGFR.

(92) Compared to the flow cytometry results obtained in Comparative Example 2, the flow cytometry shows that cetuximab and trastuzumab bind to membrane proteins present on the cell surface, but cannot determine whether this binding is specific. However, when the diffusion coefficient is measured by the method of the present invention, it can be seen that EGFR is specifically bound with cetuximab, not trastuzumab.

Example 9. Sensitivity to Molecular Weight of Measuring Method of Target Membrane Protein

(93) When a target membrane protein binds with a ligand whose molecular weight is not exactly known, as well as the ligand whose molecular weight is known such as cetuximab, to prove that information about a molecular weight of the ligand which has not been known can be estimated from a degree of change in diffusion coefficient, materials having total four different molecular weights (17 kDa to 150 kDa) such as cetuximab (150 kDa), cetuximab Fab (50 kDa), cetuximab F(ab′).sub.2 (90 kDa)) and an EGFR-specific binding aptamer (17 kDa) were bound with EGFR, and then an experiment of observing changes in diffusion coefficient of EGFR was performed.

(94) Cetuximab was purchased from Merck Serono, cetuximab Fab and F(ab′)2 were prepared using preparation kits (Pierce, 44685 and 44688), and the EGFR-specific binding aptamer was made using SELEX. As shown in Example 1.5, COS7 cells transfected with EGFR-mEos3.2 were treated with each of the four prepared ligands (cetuximab, cetuximab Fab, cetuximab F(ab′).sub.2, and the EGFR-specific binding aptamer), and changes in diffusion coefficient of EGFR were measured before and after the treatment by the same method as described in Example 1.5, and the results are shown in FIGS. 18 and 19.

(95) FIG. 17 is a diagram illustrating the bindings between the four different ligands with different molecular weights and EGFR, which are made at the same binding site.

(96) FIG. 18 illustrates changes in diffusion coefficients of EGFR and PMT before and after the host cells expressing PMT-mEos3.2 and the host cells expressing EGFR-mEos3.2 were treated with cetuximab, cetuximab Fab, cetuximab F(ab′).sub.2, and EGFR-specific binding aptamer.

(97) FIG. 19 illustrates that the molecular weight of a ligand and the change in diffusion coefficient have a linear relationship.

(98) As shown in FIGS. 17 and 19, it was confirmed that as the molecular weight of a ligand was increased as 17 kDa, 50 kDa, 90 kDa, and 150 kDa, the change in diffusion coefficient of EGFR was also increased by about 7%, 16%, 27%, and 38%, and it can be noted that such relationship is linear. In detail, when the diffusion coefficients were calculated by Equations 1 and 2, the molecular weight of a corresponding ligand can be estimated using the fact that the diffusion coefficient becomes smaller as the molecular weight of the ligand is increased.

Example 10. Measurement of Dissociation Constant

(99) 10.1 Measurement of Dissociation Constant of EGFR-WT

(100) When there is insufficient treatment with cetuximab, EGFRs are classified into two states: a cetuximab-bound EGFR state and an unbound EGFR state. If these two different EGFR states exist at the same time, a ratio of the two groups can be estimated by measuring a diffusion coefficient of EGFR by the method described in Example 1. Since the ligand-bound and unbound groups have different sizes, a difference between the diffusion coefficients of EGFR in these groups can be distinguished by the method described in Example 1, and therefore the dissociation constant between EGFR and cetuximab in a single living cell can be quantitatively measured using this ratio.

(101) To demonstrate it, a diffusion coefficient of EGFR was measured in each environment in which the COS7 cells expressing EGFR-mEos3.2 prepared in Example 1.3 was treated with cetuximab at concentrations increased gradually from 0.02, 0.04, 0.08, 0.16, 0.32, 0.64 up to 1.28 μg/ml, which is the saturation level. The obtained diffusion coefficients of EGFR were converted into bound/unbound ratios between EGFR and cetuximab using Equations 4 to 8.

(102) In the following Equations 4 to 8, U and B are random variables of a diffusion coefficient of ligand-unbound or ligand-bound target membrane proteins, respectively. For a simple model, it is assumed that there is no conversion between the two states, and the concentration of a ligand is represented as c.
M.sub.(C=c)=α.sub.(C=c)*U+β.sub.(C=c)*B  [Equation 4]

(103) M.sub.(C=c) is the diffusion coefficient of a target membrane protein when the concentration of a ligand is c,

(104) α.sub.(C=c) is the ratio of ligand-unbound membrane proteins to the total target membrane proteins, when the concentration of a ligand in one cell is c, and

(105) β.sub.(C=c) is the ratio of ligand-bound target membrane proteins to the total target membrane proteins, when the concentration of a ligand in one cell is c, and α.sub.(C=c)+β.sub.(C=c) is 1.

(106) U is the diffusion coefficient of a ligand-unbound target membrane protein,

(107) B is the diffusion coefficient of a ligand-bound target membrane protein.
E(M.sub.(C=c))=α.sub.(C=c)*E(U)+β.sub.(C=c)*E(B)  [Equation 5]

(108) E(M.sub.(C=c)) is the mean diffusion coefficient of a target membrane protein when the concentration of a ligand is c,

(109) E(U) is the mean diffusion coefficient of a ligand-unbound target membrane protein,

(110) E(B) is the mean diffusion coefficient of a ligand-bound target membrane protein.

(111) Therefore, the ratio of a ligand-bound membrane protein leads to the following Equation 6.
β.sub.(C=c)=(E(M.sub.(C=c))−E(M.sub.(C=0)))/(E(M.sub.(C=s))−E(M.sub.(C=0)))=(D.sub.(C=c)−D.sub.(C=0))/(D.sub.(C=s)−D.sub.(C=0))
E(U)=E(M.sub.(C=0)), E(B)=E(M.sub.(C=s))  [Equation 6]

(112) D.sub.(C=c) is the diffusion coefficient of a target membrane protein when the ligand concentration in one cell is c. Therefore, the dissociation constant (K.sub.d) is defined as Equation 7.
K.sub.d=([L][R])/([RL])=(c*α.sub.(C=c))/β.sub.(C=c)  [Equation 7]

(113) [L] is the concentration of a ligand,

(114) [R] is the concentration of a target membrane protein (receptor),

(115) [LR] is the concentration of the binding complex of a ligand and a target membrane protein.

(116) Equation 8 may be obtained by Equations 6 and 7.
(D.sub.(C=c)−D.sub.(C=0)))/(D.sub.(C=s)−D.sub.(C=0)))=c/(K.sub.d+c)  [Equation 8]

(117) Cooperativity as well as a dissociation constant between EGFR and cetuximab was confirmed by Equations 4 to 8 from the bound/unbound ratio between EGFR and cetuximab at a single cell level. In addition, it was confirmed that the dissociation constant measured as described above and the cooperativity values obtained from a scatchard plot were similar to those obtained by an actual experiment carried out in vitro, and the detail results are shown in FIGS. 20 to 22.

(118) FIG. 20 is a diagram illustrating that cetuximab binding to EGFR increases in number as the concentration of cetuximab is increased.

(119) FIG. 21 illustrates the reduction of a diffusion coefficient of EGFR as the concentration of cetuximab in a single one cell is increased from 0.02 to 1.28 μg/ml. A reduced value of the diffusion coefficient of EGFR was normalized by the value in non-treatment condition.

(120) FIG. 22 is a graph illustrating a ratio of cetuximab binding to EGFR as the concentration of cetuximab is increased based on the result of FIG. 21, and thereby the dissociation constant between EGFR and cetuximab can be calculated. Also, the inner graph of FIG. 22 shows that there is no cooperativity in the binding of cetuximab according to the scatchard plot.

(121) As shown in FIGS. 20 to 22, when the diffusion coefficient of EGFR was measured by the above-described method, the size of the ligand-bound or ligand-unbound group is not the same as each other, and therefore a difference between the diffusion coefficients of EGFR can be distinguished by the degree of binding between EGFR and cetuximab, and it can be confirmed that the dissociation constant between EGFR and cetuximab in a single living cell can be quantitatively measured using this ratio.

(122) 10.2 Measurement of Dissociation Constant of Mutant Form, EGFRvIII

(123) The binding between a ligand and a fluorescence-labeled target membrane protein can only be observed by the method described in Example 1. In other words, even in an environment in which a variety of membrane proteins are mixed, only the fluorescence-labeled target membrane protein can be observed.

(124) To demonstrate this, an experiment was performed by a method of treating EGFRvIII (SEQ ID NO: 8), which is a mutant form of EGFR from which a part of an extracellular domain is deleted, compared to EGFR, with cetuximab and mAb R-1 antibody. In detail, transfected cells were prepared by the method described in Example 1, except that, instead of EGFR-mEos3.2, EGFRvIII-mEos3.2 was expressed in COS7 cells, and PMT-expressed COS7 cells were prepared by the same method as used above. Also, the experiment was performed with mAb R-1 that binds to EGFR but does not bind to EGFRvIII by the same method as described in Example 10.1, and the results are shown in FIG. 24.

(125) FIG. 23 illustrates that EGFR WT can bind to both cetuximab and the mAb R-1 antibody, but the mutant form, EGFRvIII, only binds with cetuximab, not mAb R-1. FIG. 24 illustrates that since PMT does not have the binding of two antibodies such as cetuximab and mAb R-1, a reduced diffusion coefficient of PMT is not shown, but EGFRvIII binds with cetuximab, but not with mAb R-1.

(126) As shown in FIG. 24, it can be confirmed that a reduced diffusion coefficient of EGFRvIII due to cetuximab is obvious, but reduced diffusion coefficient of EGFRvIII due to mAb R-1 is insignificant. This means that EGFR vIII, which is a mutant made by deleting a mAb R-1 binding site from EGFR WT, binds to cetuximab, but not to mAb R-1.

(127) The dissociation constant between EGFR vIII and cetuximab was measured by a dose-dependent test, and in detail, the diffusion coefficient of EGFRvIII was measured by the method described in Example 1.4 while the concentration of cetuximab was increased as 0.02, 0.04, 0.08, 0.16, 0.32, 0.64, 1.28 μg/ml by the same method as described in Example 1.4, the dissociation constant between EGFRvIII and cetuximab was measured as described in Example 10.1 using the ratio of binding with cetuximab and analyzed by plotting a scatchard plot, and then the results are shown in FIG. 25.

(128) FIG. 25 illustrates the dissociation constant and cooperativity between EGFRvIII and cetuximab using the method described in Example 10.1. As shown in FIG. 25, it was able to be confirmed from the scatchard plot that there is no cooperativity in the binding between EGFRvIII and cetuximab.

(129) As shown in FIGS. 20 to 25, the method can obtain the dissociation constant and cooperativity of a ligand specifically binding to a target membrane protein to be observed in a live cell.

Example 11. Analysis of Process of Forming Complex of Target Membrane Protein

(130) When the diffusion coefficients of target membrane proteins including EGFR are measured by the method described in Example 1, the size of a molecule becomes larger during the formation of a complex of the target membrane proteins, and is directly connected to the diffusion coefficient. Therefore, the process of forming the complex may be known based on the diffusion coefficient.

(131) Host cells expressing EGFR-mEos3.2 prepared in Example 1.3 were treated with cetuximab, a secondary antibody (goat anti-human immunoglobulin (IgG) antibody (81-7100), invitrogen) and a tertiary antibody (rabbit anti-goat immunoglobulin (A10537), invitrogen) in turn, and in order to figure out the process of forming the complex, the cetuximab, secondary antibody and tertiary antibody that sequentially bind to EGFR as shown in FIG. 26 are treated in turn to measure the diffusion coefficient of EGFR by the method described in Example 1.5. The results are shown in FIG. 27.

(132) FIG. 26 illustrates sequentially binding of the cetuximab, secondary antibody and tertiary antibody to the host cells expressing EGFR-mEos3.2, and FIG. 27 illustrates the diffusion coefficients of EGFR measured in the order of treating the host cells expressing EGFR-mEos3.2 with the cetuximab, secondary antibody and tertiary antibody.

(133) As shown in FIG. 27, when the host cells expressing EGFR-mEos3.2 were sequentially treated with the cetuximab, secondary antibody and tertiary antibody, the diffusion coefficients of EGFR also tended to be reduced in the same order. The size of the final complex formed by binding the cetuximab, secondary antibody and tertiary antibody to the target membrane protein was about 600 kDa or more, and the reduction of the diffusion coefficient of EGFR caused by binding the tertiary antibody to the cetuximab and secondary antibody complex was about 12%. As a result, it can be seen that such a complex forming process can also be applied to a larger complex having a size of 600 kDa or more.

(134) Contrarily, diffusion coefficients of EGFR were measured by sequentially treating the host cells expressing EGFR-mEos3.2 with the cetuximab, tertiary antibody and secondary antibody by the same method as described in Example 1.5, and the results are shown in FIG. 28.

(135) FIG. 28 illustrates the diffusion coefficients of EGFR measured by sequentially treating EGFR with the cetuximab, secondary antibody and tertiary antibody.

(136) As shown in FIG. 28, when the EGFR was sequentially treated with the cetuximab, tertiary antibody and secondary antibody, a complex was formed as soon as the secondary antibody binds to it, and therefore at this moment, a dramatic change in diffusion coefficient of EGFR, for example, about 30%, was shown.

(137) Comparing FIGS. 27 and 28, when EGFR was treated with the cetuximab and then the tertiary antibody, the diffusion coefficient of EGFR was not changed, but the secondary antibody first bound with the tertiary antibody, and then bound with the cetuximab, and therefore a much larger reduction in diffusion coefficient of EGFR was shown. Such specificity shows that the above-described method may become an effective means to study the process of forming the complex of target membrane proteins.

Example 12. Observation of Target Endogenous Membrane Protein: Endogenous EGFR Expressed in COS7 Cells

(138) EGFR-specific binding was observed using a mouse (monoclonal) Anti-Human Epidermal Growth Factor Receptor (mAb 199.12 (Invitrogen)) among various types of antibodies targeting EGFR. A target protein may be regulated by antibody binding, but mAb 199.12 is known as an antibody insignificantly affecting the activity of EGFR.

(139) 12.1 Preparation of Fab Fragment

(140) Fab fragments were generated using a Fab preparation kit (Pierce, 44685). First, a resin conjugated with a proteinase, papain, was mixed with an immunoglobulin antibody in a buffer effective for an enzyme reaction to occur and the reaction was conducted at 37° C. for 5 to 6 hours. After the reaction was completed, the antibody that is cleaved into two Fab fragments and an Fc fragment and the papain resin were centrifuged to separate only a supernatant. A protein A-conjugated resin and the supernatant were mixed, and then centrifuged again, thereby obtaining a purified Fab fragment.

(141) 12.2 Binding Between Antibody and Organic Fluorescent Dye

(142) Antibodies were conjugated with an organic fluorescent dye, Alexa Fluor 647, using an Alexa Fluor® 647 antibody labeling kit (Invitrogen). The antibodies were added in a 0.1M sodium bicarbonate solution, and mixed with a fluorescent dye to which functional groups such as a succinimidyl ester or tetrafluorophenyl ester was tagged to allow a reaction for 1 hour, thereby forming a stable complex between the fluorescent dye and a primary amine group in the antibody. It were loaded on a purifying resin to bind remaining non-binding fluorescent dyes to the resin, and then centrifuged to obtain a supernatant, from which only the antibodies conjugated with the fluorescent dyes were isolated.

(143) 12.3 Preparation of Coverslip and Cell Seeding

(144) Intact cells without any treatment were seeded on a 25-mm or 18-mm-diameter coverslip. These coverslips were prepared by procedures of sonication in acetone at 42° C. for 30 minutes to minimize autofluorescence of the coverslip, washing with distilled water three times, sonication again in 1% hydrofluoric acid at 42° C. for 10 minutes, washing with distilled water about 15 to 20 times to completely remove hydrofluoric acid, and finally soaking in 100% ethanol and exposure to UV light for 30 minutes to sterilize.

(145) When the cells were seeded onto the prepared coverslip, the coverslip was treated with a surface coating material to facilitate cells to be attached to the coverslip. As the surface coating material that can be used herein, collagen, fibronectin, gelatin or poly-L-lysine may be used.

(146) 12.4 Fixation of the Cell-Seeded Coverslip onto Microscope

(147) The coverslip on which cells were seeded as described in Example 12.3 was placed onto a microscope with a live cell chamber (37 C, 5% CO.sub.2) and an electron multiplying charge coupled device (EM-CCD) to observe the cells at a single cell level. The microscope, Olympus IX71 model, is designed based on total internal reflection fluorescence (TIRF), and therefore can observe only fluorescent molecules at a plasma membrane of a cell closely attached (within about 200 nm from a coverslip) to the round coverslip.

(148) 12.5 Treatment with Organic Fluorescent Dye-Binding Antibody

(149) Endogenous membrane proteins expressed in the cells seeded as described in Example 12.4 were specifically labeled with fluorescent dye-conjugated Fabs prepared in Examples 12.1 and 12.2. A degree of labeling the target proteins may be regulated by changing the concentration and treatment time of the fluorescent dye-conjugated Fab.

(150) After the endogenous protein was sufficiently labeled, unbound fluorescent dye-conjugated Fabs present in the solution were removed by washing with a growth medium 2 to 3 times.

(151) To induce photoswitching and inhibit photobleaching of an organic fluorescent dye under the condition in which cells were alive, 1 mM β-mercaptoethylamine (MEA), 0.2 u/ml protocatechuic acid (PCA) and 2.5 mM protocatechuate-3,4-dioxygenase (PCD) were added to a cell culture medium.

(152) 12.6 Observation of Endogenous Membrane Protein to which Ligand and Interaction Partner Did not Bind

(153) The fluorescence-labeled endogenous EGFR prepared in Example 12.5 was illuminated with 642 nm laser to temporarily turn off the fluorescence of Alexa Fluor 647 to make it possible to observe at a single molecule level, the electron state of the fluorescent molecule was changed with 405 nm laser to allow to randomly turn on the fluorescence, and then fluorescent images were sequentially obtained with 642 nm laser at regular intervals of time (about 50 ms) by an electron multiplying charge coupled device (EM-CCD) ixon3 897, manufactured by Andor Technology, which can detect a signal transmitted from a single fluorescent molecule. A signal of the single molecule present on each image was found, and then the images were compared to each other, thereby tracking and measuring the moving distance and trajectory of the fluorescent protein per unit time, and a mean square displacement (MSD) was calculated by Equation 1.

(154) FIG. 29 illustrates a process of detecting endogenous EGFR by conjugating the photoswitchable fluorescent organic dye, Alexa Fluor 647, to the Fab fragment that can specifically bind to the endogenous EGFR (left). This can be observed because, actually, Alexa Fluor 647 is photoactivated when binding to EGFR of a cell and emits a fluorescent signal as shown in the middle image. As the position information and trajectory of the fluorescent signal emitted from each single molecule are tracked down until the fluorescent signal is photobleached, a diffusion coefficient may be obtained. The right image illustrates the trajectories of tracked 10,000 single molecules.

(155) 12.7 Observation of Endogenous Membrane Protein to which Ligand and Interaction Partner Bind

(156) The cells in which the endogenous EGFR was labeled with the fluorescent dye, prepared in Example 12.5, were treated with 20 μg/ml of cetuximab. As shown in FIG. 30, it was confirmed that fluorescent dye-conjugated mAb 199.12 and cetuximab have different EGFR-specific binding sites, they do not affect each other. The cells before and after treatment with cetuximab were observed by the method described in Example 12.6, diffusion coefficients before and after the treatment were calculated by Equations 1 and 2, and means and standard deviations thereof are shown in FIG. 31.

(157) FIG. 30 illustrates the result of confirming that the bindings of mAb 199.12 and cetuximab to the endogenous EGFRs are not influenced by each other. 1 μg/ml each of cetuximab and mAb 199.12 antibody, which were labeled with a fluorescent dye, and 100 μg/ml of cetuximab, which was not labeled with a fluorescent dye, were mixed to allow binding to the endogenous EGFR expressed in the COS7 cells for 30 minutes, and a fluorescent signal depending on a binding degree was detected by flow cytometry. The flow cytometry results show that the binding degree of the mAb 199.12 antibody is not interfered even with the 100-fold higher concentration cetuximab, but the binding of the cetuximab antibody is inhibited by the high concentration of cetuximab to which a fluorescent dye is not conjugated, and therefore a reduced fluorescent signal was detected.

(158) FIG. 31 illustrates data obtained by observing a specific binding aspect between the endogenous EGFR and cetuximab. Fab of the mAb 199.12 antibody specifically binding to EGFR was purified, labeled with Alexa Fluor 647 fluorescent dye, and then added to a single living COS7 cell to bind to the endogenous EGFR expressed in the cell. A diffusion coefficient was obtained before the treatment with an immunoglobulin antibody and cetuximab, and a diffusion coefficient changed by the binding between the EGFR and the immunoglobulin antibody or cetuximab was calculated after the treatment with the immunoglobulin antibody and cetuximab. It can be observed that, since the immunoglobulin antibody does not specifically bind to EGFR, there is almost no change in diffusion coefficient, but since cetuximab specifically binds to EGFR, the diffusion coefficient is changed by about 40%.

(159) 12.8 Identification of Endogenous EGFRs Expressed in Various Cell Lines

(160) To observe the change in diffusion coefficient of endogenous EGFRs expressed in various cell lines by cetuximab, COS7, A431, MDA-MB-231, BT20, HCC827 cell lines were prepared as described in Example 12.4, endogenous EGFRs expressed in the cells were labeled with fluorescent dye using Fab fragments as described in Example 12.5, and a diffusion coefficient under a non-treatment condition was measured as described in Example 12.6. Afterward, the same cells as previously used were treated with an immunoglobulin antibody or cetuximab, and diffusion coefficients were measured by the same method as described above. Changes in diffusion coefficient were calculated by Equation 3, compared to that under the non-treatment condition, and the results are shown in FIG. 32.

(161) FIG. 32 illustrates the quantitative results of means and standard deviations of diffusion coefficients changed by the immunoglobulin antibody and cetuximab binding to five different cell lines. Compared with the diffusion coefficient of the endogenous EGFR in each cell under the non-treatment condition, there were almost no changes in diffusion coefficient before and after the immunoglobulin antibody was treated, but after the treatment with cetuximab, all cells were changed in diffusion coefficient by about 40%. Specifically, the diffusion coefficients may be obtained by Equations 1 and 2, and provided that the diffusion coefficients under different conditions are D.sub.c1 and D.sub.c2, a ratio of the change in diffusion coefficient is calculated by Equation 3. With the treatment of the immunoglobulin antibody, D.sub.c1 and D.sub.c2 values are almost the same, which means that the change in diffusion coefficient is very small. However, with the treatment of the cetuximab, D.sub.c2 value is about 60% of D.sub.c1 value, which means that the diffusion coefficient is reduced by about 40%.

Example 13. Observation of Endogenous Membrane Protein in Primary Cell Originating from Animal Model: Observation of Endogenous EGFRs in Various Types of Primary Cells Isolated from Mouse's Lung

(162) 13.1 Preparation of Cell Marker-Specific Antibody to which Fluorescent Dye is Conjugated

(163) Since, unlike a cell line, cells purified from tissues include a mixture of various types of cells, cells were labeled with primary antibodies specific to various cell membrane markers (Sca-1, PECAM) and fluorescent dye-conjugated secondary antibodies binding to the primary antibodies to determine the type of a cell to be observed. Here, to enable observation using lasers with different wavelengths, a type of the fluorescent marker was different depending on the marker (Pacific blue, Alexa Fluor 488, Dylight 549).

(164) 13.2 Purification of Primary Cells from Mouse's Lung Tissues and Cell Seeding

(165) A mixed liquid of 0.5 mg/ml Collagenase type II (Worthington) and 1 mg/ml Collagenase type IV (Worthington) was injected into a lung tissue of a 8-week-old C57BL/6 mouse model through a respiratory tract to dissolve the tissue, and the separated lung tissue was put into a collagenase mixed liquid at 37° C. for 30 minutes. After neutralization of collagenase using PBS, and the tissue sample was centrifuged to obtain only cells, which was strained with a cell strainer (BD falcon) to separate single cells. The separated cells were grown in a growth medium treated with a fibroblast growth inhibitor for 24 hours, and seeded on an 18-mm-diameter coverslip and fixed on a microscope by the methods described in Examples 12.3 and 12.4.

(166) 13.3 Classification of Primary Cells by Cell Labeling

(167) The fixed primary cells of murine lung were treated with a pacific blue fluorescent dye-conjugated antibody against Sca-1 protein (Invitrogen), a Dylight 549-conjugated secondary antibody which binds to primary antibody against Pecam protein (Santa cruz), and an Alexa Fluor 647 fluorescent dye-conjugated anti-EGFR protein antibody. Using lasers with different wavelengths suitable for respective fluorescent dyes (405, 488, 561, and 633 nm), the autofluorescence level, protein expression level and motion of each protein were observed by the method described in Example 12.6. Based on the results, compared with markers for the previously known types of cells in the lung cell tissue (Carla F. Bender Kim, et al., Cell, Vol. 121, 823-835, Jun. 17, 2005), four different types of cells that are mixed in lung cell groups separated from one mouse model were classified, and then a diffusion coefficient of endogenous EGFR in each cell type was measured, and a change in diffusion coefficient by cetuximab was measured.

(168) FIG. 33 illustrates images of the mouse primary cells classified by markers for the cell types in the lung cell tissue and autofluorescence levels. DIC is an image showing an original cell shape, and all of Sca-1, Pecam and autofluorescence are fluorescent images. Cell#1 in which Sca-1 was very highly expressed and Pecam was rarely expressed was classified as a Bronchio-alveolar stem cell (BASC), and cell#2 in which all of Sca-1 and Pecam were expressed was able to be classified as an endothelial cell. Cell#3 showed very low expression of Sca-1 and Pecam but very high autofluorescence at a 488 nm channel, and therefore it is considered as an AT2 cell, and cell#4 in which Sca-1 and Pecam were rarely expressed and the autofluorescence level was very low is considered as a Clara cell.

(169) FIG. 34 illustrates the results of observing changes in diffusion coefficient depending on patterns of binding cetuximab to endogenous EGFRs expressed in the four types of the cells classified in FIG. 33, such as the BASC, endothelial cell, AT2, and Clara cells. It was seen that the EGFRs expressed in the BASC cell and the Clara cell show a great change in diffusion coefficient due to cetuximab, for example, by about 40%, but the endothelial cell and the AT2 cell show the change in diffusion coefficient by less than 10%.

(170) It would be understood by those of ordinary skill in the art that the above descriptions of the present invention are exemplary, and the exemplary embodiments disclosed herein can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. Therefore, it should be interpreted that the exemplary embodiments described above are exemplary in all aspects, and are not limitative.