Anti-human ninjurin-1 (NINJ-1) antibodies and methods for detecting human NINJ-1

Abstract

The present invention provides: an antibody which specifically bonds to human NINJ-1; and a fragment thereof. The antibody or the fragment thereof according to the present invention has very high bonding affinity and bonding specificity with respect to a human NINJ-1 or a homogeneous binding site of the protein, and does not exhibit cross-reactivity with NINJ-1 proteins that are derived from other organisms and have high protein similarity. Accordingly, the present invention provides significant advantages with respect to accuracy and sensitivity and the like, not only in diagnosing disease related to NINJ-1 proteins but also in inhibiting pathological conditions involving NINJ-1 proteins. In particular, the antibody provided according to the present invention has a remarkable effect of inhibiting attachment between immunocytes and human cerebral endothelial cells, and thus has an effect of treating multiple sclerosis.

Claims

1. An antibody or antigen-binding fragment thereof that specifically binds to a residue region between 26th and 37th in a human-derived Ninjurin-1 (NINJ-1) protein sequence defined by SEQ ID NO: 25, wherein the antibody or antigen-binding fragment thereof comprises: an antibody light chain variable region (VL) comprising a complementarity determining region (CDR) L1 comprising the amino acid sequence defined by SEQ ID NO: 7, a complementarity determining region (CDR) L2 comprising the amino acid sequence defined by SEQ ID NO: 8, and a complementarity determining region (CDR) L3 comprising the amino acid sequence defined by SEP ID NO: 9; and an antibody heavy chain variable region (VH) comprising a complementarity determining region (CDR) H1 comprising the amino acid sequence defined by SEQ ID NO: 10, a complementary determining region (CDR) H2 comprising the amino acid sequence defined by SEQ ID NO: 11, and a complementarity determining region (CDR) H3 comprising the amino acid sequence defined by SEP ID NO: 12.

2. The antibody or antigen-binding fragment thereof of claim 1, wherein the antigen-binding fragment is selected from the group consisting of diabody, Fab, Fab′, F(ab)2, F (ab′) 2, Fv and scFv of the antibody.

3. A composition comprising the antibody or antigen-binding fragment thereof of claim 1 as an active ingredient.

4. An isolated polynucleotide molecule comprising a polynucleotide encoding the antibody or antigen-binding fragment thereof of claim 1.

5. A vector comprising the polynucleotide molecule of claim 4.

6. An isolated host cell comprising the vector of claim 5.

7. A method for preparing an antibody or an antigen-binding fragment thereof that specifically binds to a human NINJ-1 protein defined by the amino acid sequence of SEQ ID NO: 25, the method comprising the steps of: culturing the host cell of claim 6 in a medium to express and produce the antibody or antigen-binding fragment thereof; and recovering the antibody or antigen-binding fragment thereof from the host cell or the medium in which the host cell has been cultured, wherein the antibody or antigen-binding fragment thereof comprises the VL comprising the CDRL1-3 of SEQ ID NOs: 7-9 respectively and the VH comprising the CDR1-3 of SEQ ID NOs: 10-12 respectively.

8. A method of specifically detecting human NINJ-1, the method comprising the steps of: contacting the antibody or antigen-binding fragment thereof of claim 1 with a biological sample expressing or comprising human NINJ-1 defined by the amino acid sequence of SEQ ID NO: 25; and detecting the presence of the complex of the antibody or antigen-binding fragment thereof binding to the human NINJ-1; wherein the biological sample is a tissue or cell or is isolated from a tissue or cell.

Description

BRIEF DESCRIPTION OF DRAWINGS/FIGURES

(1) FIG. 1 shows each domain and a domain important for homologous binding ability in the human NINJ-1 (Ninjurin-1) protein sequence (SEQ ID NO: 25), and shows the number of colonies by each time during the bio-panning process using the antigen corresponding to the homologous binding domain.

(2) FIG. 2 shows the results of confirming and screening the affinity of HBAg1 (P26-N37 region) through ELISA, for the phage candidates selected through the 3rd round bio-panning process.

(3) FIG. 3A shows the human NINJ-1 extracellular domain nucleic acid sequence and mouse NINJ-1 extracellular domain nucleic acid sequence introduced into the pGEX-4T-1 expression vector. From top to bottom: SEQ ID NOS: 33-37.

(4) FIG. 3B is a schematic diagram showing a cleavage site of a vector for the introduction of a human NINJ-1 extracellular domain nucleic acid sequence or a mouse NINJ-1 extracellular domain nucleic acid sequence, and a finally prepared expression vector.

(5) FIG. 4 shows the results of SDS-PAGE analysis of GST/huNINJ-1 protein expression and GST/musNINJ-1 protein expression in the transformed E. coli (FIG. 4A: expression, FIG. 4B: Sonication, FIG. 4C: Purification).

(6) FIG. 5 is a result of final selection of four affinity groups (D9, D12, E4 and G11) having high affinity to GST/HuNINJ-1 and GST/musNINJ-1 among candidates showing affinity to HBAg1, showing comparatively affinity to GST (Control), GST/huNINJ-1, GST/musNINJ-1 represented by each antibody (The absorbance value at 430 nm indicated by each candidate group was normalized to the value of B1 (Non-binding active control)).

(7) FIG. 6 is a schematic diagram showing a process for producing an IgG expression vector containing an antibody heavy chain variable region (VH) or an antibody light chain variable region (VL) of scFv.

(8) FIG. 7 shows comparative results regarding the affinity of each test group antibody to HBAg1 (P26-N37 region) verifying that, a D12 Ig antibody of the present invention (FIG. 7a) has a significantly higher affinity for HBAg than the anti-NINJ-1 antibodies (FIG. 7a, FIG. 7b) sold by other companies.

(9) FIG. 8a shows comparative results regarding that the D9, D12, E4 and G11 Ig antibody of the present invention among the various candidate groups have a very high specificity in binding to native human NINJ-1 protein expressed in normal cells.

(10) FIG. 8B shows results confirming that the D12 Ig antibody of the present invention does not share an antigen recognition site with an anti-NINJ-1 antibody (R&D systems, Cat No. MAB51051, labeled R**) available from other companies.

(11) FIG. 9a is a schematic diagram showing a procedure for obtaining a recombinant vector for the production of a doxycycline-inducible human NINJ-1 overexpressing glioblastoma cell line.

(12) FIG. 9b shows the result of Western blotting detecting the expression level of the human NINJ-1 in the U87MG pLVX NINJ-1 glioblastoma cell line according to the treatment concentration of doxycycline.

(13) FIG. 10 shows the results of comparing the affinity of the D9, D12 E4 and G11 Ig antibody of the present invention and the two kinds of anti-NINJ-1 antibodies (BD bioscience Cat No. BD610776 (labeled B**), and SantaCruz Cat No. sc-136295 (labeled S**)) against the human NINJ-1 protein expressed in the U87MG pLVX NINJ-1 glioblastoma cell line.

(14) FIG. 11 shows results confirming whether the D9, D12, E4 and G11 Ig antibody of the present invention inhibits the adhesion between the immune cells and the human cerebral endothelial cells in order to confirm whether or not the antibody has utility as a therapeutic agent for multiple sclerosis.

MODE FOR CARRYING OUT INVENTION

(15) Hereinafter, the present invention will be described in detail.

(16) However, the following examples are only illustrative of the present invention, and the present invention is not limited to the following examples.

EXAMPLE 1

Identification of Antibodies Using P26-N37 Region of Human NINJ-1 Protein

(17) 1-1. Bio-Panning

(18) In order to identify antibodies specific to the P26-N37 region of human NINJ-1 (Ninjurin-1, SEQ ID NO: 25), the following experiment was performed using phage display technology based on the human scFv library. The fragment peptide for the P26-N37 region is hereinafter referred to as HBAg1.

(19) Biotin-conjugated peptides at the ends of HBAg1 were synthesized, followed by binding to Dynabeads M-270 streptavidin (GE Healthcare Life Science). Subsequently, the scFv phage of the library was reacted at 37° C. for 1 hour and 30 minutes.

(20) Then, the scFv phage having the specific binding was obtained and infected (input) with ER2537 E. coli (NEB) cultured to be 0.5 to 0.7 at OD600, and then ancillary phages were added and cultured in solid medium at 37° C. for 12 to 16 hours. For some, shake culture was performed in liquid medium at 37° C. and 160 rpm for 12 to 16 hours. Thereafter, the number of colony on the solid medium was measured and the output was confirmed. Poly ethylene glycol (PEG, Sigma aldrich) was added to the culture solution to precipitate the phage, and the same experiment was further repeated two more times to perform final third round of panning.

(21) The number of colony of the phage decreased in the second round compared to the first round, whereas it increased in the third round. This result is shown in FIG. 1.

(22) 1-2. Selection of Human NINJ-1 scFv Candidates

(23) After the host cells (ER2537, NEB) were infected with the phage recovered in the 3rd bio-panning of Example 1-1, 95 colonies were selected on LB solid medium, and then cultured in liquid medium at 700 rpm at room temperature. Then, each well was treated with Isopropyl B-D-1-thiogalactopyranoside (IPTG) to a final concentration of 1 mM, and cultured at 30° C. for 12-16 hours at 200 rpm. The culture solution was centrifuged at 3,000 rpm for 20 minutes to remove the supernatant, and osmotic cell lysis was performed through TES buffer to obtain intracellular phage particles. The plate was coated with HBAg1 or Phosphate buffered saline (PBS, control) at a concentration of 10 ug/ml in a 96-well plate (Thermo scientific, Cat No. 436014) previously coated with streptavidin, and each 50 μl of recovered phage particles was added to the coated plate, followed by reaction at room temperature for 1 hour. Then, the anti-HA (hemagglutinin) secondary antibody (Santa cruz biotechnology) conjugated with horseradish peroxidase was added and reacted again at room temperature for 1 hour. Then, a color development reaction using the TMB solution was induced for 5 minutes, and the reaction was stopped with 1N H.sub.2SO.sub.4 solution, and the absorbance was measured at 430 nm.

(24) When the results of the control and HBAg1-coated plates were compared, as shown in FIG. 2, DNA sequence analysis of colonies of the wells in which the reaction was observed only in HBAg1 was performed. This resulted in a number of scFv candidates (indicated by green in FIG. 2) that could specifically bind to human NINJ-1, particularly HBAg1.

EXAMPLE 2

Expression and Purification of Recombinant Proteins of NINJ-1

(25) 2-1. Preparation of Expression Vector for the Production of the Extracellular Domain Protein of NINJ-1

(26) In order to confirm whether the scFv candidates obtained in Example 1 can structurally bind to the P26-N37 region of the human NINJ-1 protein and can bind to the extracellular domain portion of NINJ-1, a nucleotide sequence containing two transcription termination codons in the extracellular domain portion of the human NINJ-1 or mouse NINJ-1 gene was inserted between the multi-cloning sites BamHI and XhoI in the pGEX-4T-1 expression vector to construct a recombinant protein expression vector, respectively. The specific nucleotide sequence is shown in FIG. 3A, the specific cloning region is shown in FIG. 3B, and each expression vector was named pGST/huNINJ1 and pGST/musNINJ1, respectively. The expression vectors were transfected into E. coli DH5a (enzynomics) and cultured. After the MIDI prep, the insertion of GST/huNINJ-1 and GST/musNINJ-1 was finally confirmed through DNA sequencing.

(27) 2-2. Expression and Purification of NINJ-1 Extracellular Domain Protein

(28) E. coli DH5a transformed with the expression vector pGST/huNINJ-1 or pGST/musNINJ-1 prepared in Example 2-1 was subjected to MIDI prep. After sufficient plasmid DNA was recovered, the following experiment was performed for the expression and purification of recombinant proteins in E. coli. E. coli BL21DE3 was transformed with pGST/huNINJ-1 or pGST/musNINJ1 plasmid DNA, respectively, and cultured in LB medium. Then, a single colony was cultured in a 5 ml LB liquid medium containing ampicillin so that the OD 600 value was 0.6 to 0.8. Then, the medium was treated with 0 mM, 0.5 mM, and 1 mM of IPTG, respectively cultured at 37° C. for 3 hours, and centrifuged at 8,000 rpm for 10 minutes to recover only cells. SDS-PAGE was performed on the cell lysate, and expression was confirmed by treating with Coomassie stain (Sigma Aldrich) (FIG. 4A).

(29) The cell lysate, which had been induced to express with 0.5 mM and 1.0 mM IPTG, was collected and centrifuged to recover cells. Cells were homogenized with 1 ml PBS and were lysed using an ultrasonicator (BRANSON). Then, the dissolved cells were centrifuged to separate inclusion bodies and water-soluble proteins, and the separated samples were subjected to SDS-PAGE and Coomassie stain treatment to confirm the water-solubility of the recombinant proteins (FIG. 4B). Then, the water-soluble protein portion was reacted with Glutathione Sepharose™ 4B resin (GE healthcare Life Sciences) for 1 hour at 4° C., centrifuged to separate the protein bound to the resin. Then, only the GST-tagged recombinant protein was recovered using 50 mM Tris-HCl, 10 mM reduced glutathione and pH 8.0 buffer. SDS-PAGE was performed on the recovered protein, and purification was performed by the same method using a coomassie stain to confirm the size of the recovered protein (FIG. 4C).

EXAMPLE 3

Affinity Measurement of GST/huNINJ-1 and GST/musNINJ-1 by ELISA

(30) In order to screen scFv candidates capable of specifically binding to the recombinant proteins GST/huNINJ-1 and GST/musNINJ-1 obtained in Example 2, the following experiment was performed. Experiments were performed to confirm the affinity of scFv candidate groups that specifically showed binding activity to HBAg1 in Example 1 and GST/huNINJ-1 and GST/musNINJ-1, using B1 in which no specific binding to HBAg1 was observed as a control. Recombinant proteins GST/huNINJ-1, GST/musNINJ-1 or GST were coated at a concentration of 10 ug/ml in a 96 well plate (corning 3690 flat bottom, half-area plate), respectively. The scFv phage particles in the cells were separated from the scFv candidate group selected in Example 1 through the same experimental method as in Example 1-2. Then, ELISA was performed using the coated plates.

(31) The absorbance values at 430 nm indicated by each candidate group were normalized to the values of B1 (Non-binding active control), and finally, four candidates (D9, D12, E4 and G11) with high affinity to GST/HuNINJ-1 and GST/musNINJ-1 were selected among candidates showing affinity to HBAg1. As shown in FIG. 5, the affinities of human and mouse NINJ-1 were confirmed in four candidate groups, and CDR sequences of these candidate groups are shown in Table 1 below.

(32) TABLE-US-00001 TABLE 1 CDRH1 CDRH2 CDRH3 CDRL1 CDRL2 CDRL3 D9 DYSMS GIYPDDSNTY DPVHCERSV RGSSSNI ADSHR GAWDS YADSVKG CYYADAMDV GSNYVT PS SLNA D12 NYDMS WISPDGSNIY YRITPMSWL TGSSSNI SDSKR GSWDY YADSVKG SYYDDAMDV GSNSVT PS SLNA E4 DYAMS GIYHGGGNT DPMHSERIT SGSSSNI DNSKR GTWDY YYADSVKG FDY GSNNVS PS SLSA G11 DYAMS AIYYDSGSIY DPMTSLALT SGSSSNI DNSQR GAWDA ADSVKG Y FDY GSNYVS PS SLNG

EXAMPLE 4

Preparation of Expression Vector for Human Anti-NINJ-1 Antibody Production and Antibody Production

(33) In order to express the immunoglobulin G (IgG) antibody having a molecular weight of 160 kDa containing the VH and VL nucleotide sequences of the selected four scFvs (D9, D12, E4 and G11), the following cloning was performed.

(34) As shown in FIG. 6, VH and VL nucleotide sequences of each scFv were inserted into IgG expression vectors (Modified pOptiVEC (IgG VL-LC.sub.K notation, light chain constant region comprised; Thermo Fisher Scientific, pOptiVEC-TOPO TA cloning kit, Cat No. 12744-017), Modified pcDNA3.3 TOPO (IgG VH-HC notation, heavy chain constant region comprised; Thermo Fisher Scientific, pcDNA3.3-TOPO TA cloning kit, Cat No. K830001) ClaI and NheI regions using the primers shown in Table 2 below to prepare heavy chain expression vectors (IgG VH-HC) and light chain expression vectors (IgG VL-LC.sub.K) (See FIG. 6). Then, the vectors were transformed into E. coli; cultured on a liquid medium, and plasmid DNA was extracted through a midi prep. The extracted plasmid DNA was co-transfected into the FreeStyle 293F cell line according to the experimental method provided in Freestyle 293 Expression system (Thermo Fisher Scientific) to induce the expression of the immunoglobulin G type antibody. The cell culture was centrifuged to separate the cells, and then the antibodies were separated and purified by affinity chromatography using MabSelect SuRe™ Protein A (GE Healthcare Life Sciences).

(35) TABLE-US-00002 TABLE 2  Information of primers SEQ ID Primer Sequence NO. IgG-VH-HC ATTCGATCGATATGGAGACAGACACACTCCTGCTA forward  TGGGTACTGCTGCTCTGGGTTCCAGGTTCCACGTG 28 primer GGAGGTGCAGCTGTTGGAGTCT IgG-VH-HC CTTGGTGCTAGCTGAGCTCACGGTGACCAGTGT 29 reverse  ATTCGATCGATATGGAGACAGACACACTCCTGCTA primer IgG-VL-LC TGGGTACTGCTGCTCTGGGTTCCAGGTTCCACGTG 30 forward  GCAGTCTGTGCTGACTCAGCCA primer IgG-VL-LC AGCCACCGTACGTAGGACCGTCAGCTTGGTGCC 31 reverse  primer

EXAMPLE 5

Affinity Measurement of Human Anti-NINJ-1 IgG Against HBAg1

(36) D12 was used as a representative Ig antibody prepared in Example 4, and affinity to the P26-N37 region was measured using NINJ-1 antibodies (MAB51051, BD610776, SC-136295) that are commercially available from other companies as a control.

(37) Specifically, HBAg1 was coated on 96-well plates (Thermo scientific, Cat No. 436014) coated with streptavidin at a concentration of 10 ug/ml overnight at 4° C. After washing three times with TBS buffer containing 0.05% Tween 20, 3% bovine serum albumin (BSA) was dissolved again in the same buffer, and then reacted at room temperature for 1 hour. Various anti-NINJ-1 antibodies that are commercially available from other companies and the Ig antibodies of the present invention were diluted 2-fold at a concentration of 5 ug/ml or 25 ug/ml, reacted at room temperature for 1 hour in each well, and then a secondary antibody conjugated with horseradish peroxidase was added and reacted for 1 hour. After reacting with TMB for 20 minutes, the reaction was stopped with 1N H.sub.2SO.sub.4 and the absorbance was measured at 430 nm.

(38) As a result, as shown in FIGS. 7A and 7B, it was confirmed that the D12 Ig antibody of the present invention was found to have a significantly higher affinity for the human P26-N37 peptide than the anti-NINJ-1 antibody that are commercially available from other companies.

EXAMPLE 6

Normal Cell-Based Anti-NINJ-1 IgG Screening

(39) It was confirmed whether the four candidate groups selected above and the experimental groups belonging to the candidate group of Example 1-2 but excluded later were specifically bound to the NINJ-1 protein expressed in actual human cell lines. Also, it was confirmed by using a flow cytometer whether the antigen recognition sites of the anti-NINJ-1 antibodies that are commercially available from other companies are the same.

(40) Human cerebral microvascular endothelial cell line (hCMEC/D3) were treated with 10% fetal bovine serum (Hyclone) for 20 minutes at 4° C., and then mouse-derived anti-NINJ-1 antibody (R&D systems, MAB51051), Isotype control (R&D systems, MAB002), and the selected human anti-NINJ-1 antibody (D9, D12, E4, G11 and 8 additional antibodies, all tested in Ig form) were reacted at 4° C. for 90 minutes at a concentration of 1 ug per 10,000 cells. After completion of the reaction, the cells were washed three times with PBS, and then the mouse-derived antibody was reacted with anti-mouse IgG with a secondary antibody conjugated with FITC, Human-derived antibodies (antibodies of the invention) were reacted with anti-human IgG secondary antibody conjugated with FITC for 1 hour at 4° C. This assay was performed using flow cytometry (BD FACS Calibur™)

(41) As shown in FIG. 8A, it was shown that the four antibodies of the present invention specifically bind to human NINJ-1 expressing in the cells, while D9, D12, and G11 among the four selected antibodies showed particularly high binding ability.

(42) In addition, using the D12 of the candidate groups, it was confirmed whether or not it shares the antigen recognition site with the anti-NINJ-1 antibody (R&D systems) commercially available from other companies. Each of the antibodies was treated at the same concentration of 1 ug per 10,000 cells in the immortalized-Human cerebral microvascular endothelial cell line (HBMEC), and then, the anti-human IgG secondary antibody conjugated with FITC was reacted at 4° C. for 1 hour. This assay confirmed binding using a flow cytometer (BD FACS Calibur™). As shown in FIG. 8B, it was confirmed that the antibody D12 of the present invention does not share the antigen recognition sites with the anti-NINJ-1 antibody (R&D systems) commercially available from other companies.

EXAMPLE 7

Confirmation of Diagnostic Specificity for NINJ-1

(43) In the treatment of doxycycline, the specific binding ability of the four candidate groups (D9, D12, E4, G11) tested through Example 3 and Example 4 to human NINJ-1 protein was verified using a glioblastoma cell line overexpressing NINJ-1.

(44) 7-1. Construction of Expression Vector and Stable Expression Cells for Doxycycline-Inducible Human NINJ-1 Over-Expressing Cell Lines

(45) In order to construct a human NINJ-1 over-expressing cell line, the cloning positions in the cDNA sequence of NINJ-1 (Korea Human Gene Bank, hMU007113) were selected from CDS 21 to 476 (ATGGACTC to AGCAGTAG), and cloning primers were prepared using BamHI and EcoRI in the MCS portion of the pLVX-Tet-On puro vector.

(46) PCR (95° C./5 min, (95° C./30 sec, 60° C./45 sec, 72° C./30 sec) 30 times, 72° C./10 min) was performed using a forward primer 5′-TATGGATCCCTACTGCTGGGGTGCCATG-3′ (SEQ ID NO:32) and a reverse primer 5′-TATGAATTCATGGACTCGGGAACCGAGG-3′ (SEQ ID NO:33), and then electrophoresis on 0.7% agarose gel to recover only specific fragments, and the vector and PCR products were reacted with BamHI and EcoRI for 1 hour at 37° C. After completion of the reaction, the sample was recovered and purified, and the inducible expression vector was completed through T4 ligase (This process is shown in FIG. 9A). The inducible expression vector was transfected into a glioblastoma cell line (U87MG), and puromycin was repeatedly treated at a concentration of 1 ug/ml for 3 days to establish a stable cell line. 1, 5, 10, and 50 ng/ml of doxycycline was added to the growth medium of the stable cells, and the increase of the expression rate of NINJ-1 was confirmed by Western blotting (see FIG. 9B). It was shown that the expression level of NINJ1 protein increased according to the concentration of doxycyclin (see FIG. 9B). These cell lines are hereinafter referred to as U87MG pLVX NINJ1.

(47) 7-2. Measurement of Binding Capacity Based on Doxycycline-Inducible Human NINJ-1 Over-Expressing Cell Lines

(48) In order to compare the binding ability with the human anti-NINJ-1 IgG candidate group finally selected in Example 4 (D9, D12, E4, G11) and the anti-NINJ-1 antibody developed and marketed by another company (BD bioscience, SantaCruz), the following experiment was performed using a glioblastoma cell line (U87MG pLVX NINJ1) in which human NINJ-1 overexpression was induced by the doxycycline prepared in Example 7-1.

(49) In advance, a cell line was prepared by treatment of the U87MG pLVX NINJ-1 glioblastoma cell line with a final concentration of 50 ng/ml for 24 hours at 37° C. Based on F4 (control), which was excluded from Example 3, the four candidates of the present invention and two anti-NINJ-1 antibodies of other companies (BD bioscience, SantaCruz) were analyzed, using a flow cytometry analyzer according to the same experimental method as in Example 6.

(50) As shown in FIG. 10, it was confirmed that the 4-candidate groups possess the high binding ability to NINJ-1 as compared with the antibody marketed by other companies (FIGS. 10A and 10B).

EXAMPLE 8

Inhibition of Adhesion of Cerebral Vascular Endothelial Cells and Immune Cells after Treatment with Anti-NINJ-1 IgG of the Present Invention

(51) Binding of immune cells to central nervous system endothelial cells and migration of immune cells into the central nervous system are important pathology in multiple sclerosis, and existing therapeutic strategies (for example natalizumab, etc.) for preventing such binding and migration have been used in the treatment of multiple sclerosis. It is known in the art that NINJ-1 acts as an adhesion molecule in the binding of immune cells to central nervous system endothelial cells and migration of the immune cells into the central nervous system. Therefore, if the activity of NINJ-1 is inhibited and binding of immune cells to central nervous system endothelial cells and migration of the immune cells is thus inhibited, the prevention and treatment of multiple sclerosis can be achieved.

(52) In order to confirm whether the human NINJ-1 Ig antibody prepared by the present invention can inhibit adhesion between the immune cell line and human cerebral endothelial cells, the following experiment was conducted.

(53) Collagen I Rat tail Protein was coated at a concentration of 50 ug/ml in a 96-well plate, and 1.0×10.sup.4 hCMEC/D3 cell lines were cultured in each well at 37° C. and 5% CO.sub.2 for 72 hours. Then, 200 IU/ml TNFα and IFNγ or 1 μg/ml PMA were treated for 16 hours. Then, the cells were washed twice with magnesium and calcium-free DPBS to remove inflammation inducers remaining in the cells. Thereafter, 10 ug/ml of each antibody was added to the wells and reacted. At the end of this procedure, the remaining antibody was removed by washing twice with DPBS again. Then, 1.0×10.sup.4 U937 cell lines (lymphoblast) expressing Green fluorescent protein (GFP) as immune cells were added, and the cells were washed three times with DPBS after inducing adhesion at 37° C. and 5% CO.sub.2 for 90 minutes. All procedures were performed using a real-time cell image analyzer (IncuCyte™ ZOOM) to measure the number of green-labeled cells. The number of cells finally remaining in each well was divided by the initial cell number, and then standardized based on the treatment with PMA (Sigma Aldrich).

(54) As shown in FIG. 11, adhesion between the immune cell line and human cerebral endothelial cells was decreased in the group treated with human NINJ-1 antibody (1 μg/ml PMA+D12, 1 μg/ml PMA+G11). Thus, it was confirmed that the antibodies prepared by the present invention inhibit the adhesion between the immune cells and the human cerebral endothelial cells through their specific binding ability to the P26-N37 region. Thus, it is apparent that the antibodies prepared in the present invention can exhibit a therapeutic effect on multiple sclerosis.

INDUSTRIAL APPLICABILITY

(55) As described above, the present invention provides an antibody or antigen-binding fragment thereof that specifically binds to a human NINJ-1 and the homologous binding site of the protein. The antibody or antigen-binding fragment thereof according to the present invention has a very high binding affinity and specificity to human NINJ-1 and does not show cross-reactivity with other origin-derived NINJ-1 proteins (especially mouse NINJ-1 protein) having high protein similarity. Thus, it provides significant advantages not only in the diagnosis of diseases related to NINJ-1 protein, but also in the accuracy, high sensitivity and the like in inhibiting the pathological conditions involved in NINJ-1 protein. In particular, the antibody provided by the present invention is remarkably effective in inhibiting adhesion between immune cells and human cerebral endothelial cells, and thus has therapeutic effect on multiple sclerosis. Therefore, the present invention is highly industrially applicable in the diagnostic and therapeutic industries where targeting characteristics are important.