WO2009031834A2

WO2009031834A2 - Hybridization method between leucine rich repeat family proteins

Info

Publication number: WO2009031834A2
Application number: PCT/KR2008/005220
Authority: WO
Inventors: Jie-Oh Lee; Ho-Min Kim; Beom Seok Park; Mi Sun Jin; Ook Jun Yoo
Original assignee: Korea Advanced Institute Of Science And Technology
Priority date: 2007-09-04
Filing date: 2008-09-04
Publication date: 2009-03-12
Also published as: WO2009031834A3

Abstract

Disclosed is a method for fusing proteins. More specifically, disclosed is a method of fusion between LRR (Leucine Rich Repeat )-famiIy proteins, the method comprising fusing the LRR-family proteins with each other at the following consensus sequence position: LxxLxxLxLxxN, wherein L represents alanine, glycine, phenylalanine, tyrosine, isoleucine, valine or tryptophane, N represents asparagine, glutamine, serine, cysteine or threonine, and X represents any amino acid. According to the disclosed fusion LRR technique, a fusion protein having high crystallinity can be obtained by fusing an LRR-family protein having high crystallinity with another LRR-family protein having low crystallinity. Thus, the fusion LRR technique is useful to determine the structures of proteins having low crystallinity. Furthermore, because the active regions and mechanisms of proteins can be understood from the structural analysis of the proteins, the fusion LRR technique can be effectively to understand the in vivo action of the proteins and develop therapeutic drugs. In addition, according to the fusion LRR technique of the present invention, fusion proteins, which have high crystallinity while maintaining the function of the respective proteins, can be produced in large amounts. Thus, the fusion LRR technique is useful for the development of protein drugs.

Description

[DESCRIPTION]

[Invention Title]

HYBRIDIZATION METHOD BETWEEN LEUCINE RICH REPEAT FAMILY PROTEINS

[Technical Field]

<i> The present invention relates to a method of fusing proteins, and more particularly to a protein fusion method based on the characteristic sequences and structures of LRR-family proteins, in which the method can produce LRR- fami Iy proteins in large amounts, can easily crystallize the proteins, does not cause changes in the structures and functions of the proteins, and thus can be used to determine the structures of the proteins and develop protein drugs .

[Background Art]

<2> The leucine-rich repeat (LRR) family of proteins has the following sequence characteristics: (1) it has at least LRR module; (2) the LRR module consists of 20-30 amino acids; (3) the LRR has a conserved pattern of LxxLxxLxLxxN, wherein L represents a hydrophobic amino acid such as alanine, glycine, phenylalanine, tyrosine, isoleucine, valine or tryptophane, N represents asparagine, glutamine, serine, cysteine or threonine, and X represents any amino acid; and the LRR family of proteins has a horseshoe- like three-dimensional structure. It is known that the LRR family of proteins is involved in diverse physiological processes such as signal transduction, cell cycle regulation, apoptosis and immune response. <3> The immune system of the human body is broadly divided in the innate immune system and the adaptive immune system. The innate immune system provides a rapid and nonspecific response that acts as the first line of defense against pathogens. On the other hand, the adaptive immune . system provides a strong and effective response to pathogens that the body has previously encountered through vaccinations or infections, but it is impotent against new infections. If the body's immune system is weak, it is weak against viral infections, and if the immune response is excessive and uncontrolled, it can lead to septicemia with high mortality. <4> Toll-like receptors (TLRs), which are members of the LRR protein family, play an important role in innate immunity and serve to link innate immunity with adaptive immunity. Their extracellular domains recognize and bind to broad but highly conserved structural patterns on bacteria, fungi and viruses. So far, 13 TLR families have been found in mammals, and the functions and sequences thereof have been elucidated for 6 subfamilies.

<5> Variable Lymphocyte Receptors (VLRs) are immune receptors of jawless fish and similar to immune receptors of jawed vertebrates. The lymphocyte cells of sea lamprey, a kind of jawless fish, express a unique VLR protein (VLRB.61) protein in a monoallelic manner that can specifically recognize the corresponding antigen in the humoral response. The VLR protein is a member of the LRR family of proteins and has high crystallinity.

<6> VLRs and TLRs generally include an LRR domain consisting of a signal peptide, an N-terminal cap (LRRNT), an LRR module and a C-terminal cap (LRRCT) in the extracellular domain (FIG. 1). FIG IA shows the arrangement of the VLR domain. In FIG. IA, the red color indicates the signal peptide, the blue color indicates LRRNT, the green color indicates the LRR , module (LRRl, LRRV, LRRVe, LRRCP), the orange color indicates LRRCT, and the pink color indicates the stalk. FIG. IB shows the arrangement of the TLR domain.

<7> In FIG. IB, the red color indicates the signal peptide, the blue color indicates LRRNT, the green color indicates the LRR module (LRR1-LRR22), the orange color indicates LRRCT, and the grey color indicates the TIR region. The LRR-family proteins have a horseshoe-like structure, in which the concave surface is formed by parallel β strands, and the convex surface is formed by loops, 3io helixes and α -helixes.

<8> In the microscopic world of protein molecules, the therapeutic effects of a drug can be explained by a process in which the drug adheres to a specific protein of causing or treating disease, like a key and a lock, to turn on or off the functional switch of the protein. Accordingly, the three- dimensional structure of a protein is very important in understanding the in vivo action of the protein and developing therapeutic drugs. This is because if the arrangement and three-dimensional structure of atoms constituting a polymer protein is understood, the location of the functional switch of a protein involved in disease and immunity can be determined.

<9> Because each protein is very small such that visible light cannot collide with the protein, the three-dimensional structure of proteins cannot be observed microscopically. For this reason, X-rays having a wavelength which is much shorter than that of ultraviolet rays are used to observe the structure of proteins. In addition, it is difficult to recognize a single protein due low X-ray diffraction intensity, a crystal consisting of repeats of the same protein must be prepared for protein analysis. X-ray diffraction analysis is the most useful method for analyzing the three-dimensional structure of protein, which comprises preparing a protein into a protein crystal, radiating the protein crystal with X-rays, and then analyzing the information of light diffracted. However, for X-ray diffraction analysis, there is a limitation in that a protein must be prepared into a protein crystal .

<10> It is a common issue in the biological and medical fields to produce a protein in a water-soluble form in large amounts and obtain a crystal of the protein. In many proteins, several domains are linked with each other by flexible loops, and the characteristic function of each protein is attributable to some domains having a length which is much shorter than the full length. Accordingly, if the non-essential domain of a protein is removed, the crystallization of the protein can become easier. However, the above-described method cannot be applied to the LRR-family proteins, because hydrophobic cores are distributed through the molecules of the proteins. The LRR modules have two modules, LRRNT and LRRCT, which cover the hydrophobic cores. If the LRR module is removed, the LRRNT or LRRCT will also be removed, and thus the hydrophobic cores of the proteins will be exposed to the outside, making the proteins very unstable.

<11> Accordingly, there has been a need to develop a method which can produce easy-to-crystal lize LRR family proteins in large amounts in a manner different from the prior art. [Disclosure] [Technical Problem]

<12> The present invention has been made in order to solve the above- described problems occurring in the prior art, and it is an object of the present invention to provide a method for preparing a fusion protein, which can produce LRR-family proteins in large amounts, can easily crystallize the proteins, does not cause changes in the structures and functions of a target protein, and thus can be used to determine the structure of proteins and develop protein drugs.

<13> Another object of the present invention is to provide a method of analyzing the structure of a target protein by analyzing the structure of said fusion protein. [Technical Solution]

<14> To achieve the above objects, the present invention provides a method for preparing a fusion protein by fusion between two proteins belonging to the LRR family, the method comprising fusing a first LRR-family protein with a second LRR-family protein at the following characteristic consensus sequence position: LxxLxxLxLxxN, wherein L represents alanine, glycine, phenylalanine, tyrosine, isoleucine, valine or tryptophane, N represents asparagine, glutamine, serine, cysteine or threonine, and X represents any amino acid. As used herein, the term "amino acid" refers to all naturally occurring L-α-amino acids, including norleucine, ornithine and homocysteine. The method of preparing the fusion protein according to the present invention will hereinafter be referred to as "fusion LRR technique".

<15> In the present invention, one of the LRR-family proteins is preferably a toll-like receptor (TLR) protein or a variable lymphocyte receptor (VLR) protein. Furthermore, one of the LRR-family proteins is preferably an LRR- family protein having high crystallility.

<16> In another aspect, the present invention relates to a fusion protein obtained according to said preparation method, a DNA fragment encoding said fusion protein, and a vector including said DNA fragment.

<17> In still another aspect, the present invention provides a method for analyzing the structures of LRR proteins, the method comprising the steps of: (A) crystallizing a fusion protein and any one of LRR-family proteins constituting the fusion protein, and analyzing the structures of the crystallized proteins; and (B) excluding the crystallized LRR-family protein from the fusion protein to determine the structure of another protein constituting the fusion protein.

<18> The present invention relates to a method of fusing a first LLR-family protein fragment with a second LRR-family protein fragment (fusion LRR technique). According to the present invention, fusion LRR proteins can be substantially infinitely prepared without structural changes by changing the kinds of LRR proteins to be fused and the position of fusion of LRR proteins.

<19> In the fusion of LRR proteins, if the hydrophobic cores of the LRR proteins do not accurately coincide with fusion positions, there is a problem in that the resulting fusion protein has unstable hydrophobic nuclei. Meanwhile, even though the structures of proteins have been found in detail, it is very difficult to fuse two hydrophobic nuclei with each other. This is partially attributable to the flexibility of protein residues.

<20> The present inventors have performed various tests and, as a result, have found that the "LxxLxxLxLxxN" region of the LRR module is well conserved, the structure thereof does not substantially change, and thus the structure of fusion proteins can be reliably predicted. Based on this finding, the "LxxLxxLxLxxN" region was selected as the best fusion position. The relative positions of conserved residues in said region were strictly conserved in the fusion position.

<21> In the present invention, in order to evaluate the effectiveness of "fusion LRR technique", human TLR4, human TLR2, mouse TLR2 and human TLRl were selected as the first LRR-family protein, and hagfish VTR was selected as the second LRR-family protein.

<22> Fusion LRR proteins were obtained through the following processes. <23> (1) Extraction and fusion of first LRR-family protein and second LRR- fami Iy protein

<24> When the first LRR-family protein gene and the second LRR-family protein gene are conceptually indicated, the first LRR-family protein gene can be indicated as [first protein 5' region-LxxLxxLxLxxN-first protein 3' region], and the second LRR-family protein gene can be indicated as [second protein 5' region-LxxLxxLxLxxN-second protein 3' region] (FIG. 2A).

<25> In the present invention, fusion proteins, in which the front ends and back ends of the first and second proteins are crossl inked with each other, like [first protein 5' region-LxxLxxLxLxxN-second protein 3' region] and [second protein 5' region-LxxLxxLxLxxN-first protein 3' region] (FIG. 2B), are obtained using given templates and primer sets and conventional genetic engineering techniques including PCR methods. Herein, various fusion genes can be obtained by suitably selecting the position of LxxLxxLxLxxN to be fused in each protein.

<26> Although primer sets and restriction enzymes are specifically presented in Examples below, various primer sets and restriction enzymes other than those shown in Examples may also be used. However, the (internal) cleavage of the first LRR-family protein and second LRR-family protein genes is preferably performed using the same restriction enzyme in view of the crossl inking of both the protein genes.

<27> (2) Introduction of tag for primary purification of fusion protein

<28> As a tag for the purification of the fusion protein to be expressed, the Fc region gene of human IgGl is linked to the 3'-end of the fusion gene. At this time, a thrombin cleavage site is introduced between the fusion gene and the Fc tag.

<29> It is to be understood that the Fc tag may be linked to the following vector, and then the fusion gene may be linked thereto, because the relative position is important in the linkage of the Fc tag with the fusion gene.

<30> FIG. 3 illustrates examples of the gene fragment structures of the fusion gene having the Fc tag linked thereto. <31> (3) Construction and expression of vector and purification of fusion protein

<32> The fusion gene is introduced into a specific vector to construct an expression vector, which is then introduced into host cells to express the fusion protein. In Examples below, the fusion gene was introduced into a vector to construct a transfer vector, which was then mixed with baculovirus replicable/expressible in insect cells. The fusion protein was inserted into baculovirus by a fusion reaction, thus constructing "recombinant baculovirus". The recombinant baculovirus was infected into insect cells to express the fusion protein. However, it is evident that the fusion protein can be expressed using various host systems and expression systems.

<33> The expressed fusion protein is treated with thrombin to remove the Fc tag, and then various kinds of fusion proteins, including [first protein 5' region-LxxLxxLxLxxN-second protein 3' region] and [second protein 5' region- LxxLxxLxLxxN-first protein 3' region], are purified.

<34> The host cells that are used in the present invention may be bacterial cells such as E. coli, yeast cells such as Pichia pastoris, insect cells such as Spodoptera frugiperda, or mammalian cells such as African green monkey fibroblast-derived COS cells or CHO (Chinese Hamster Ovary) cells.

<35> The fusion proteins can be purified by ion-exchange chromatography, affinity chromatography, different sugar chromatography, hydrophobic interaction chromatography, reverse phase chromatography or gel filtration, or other various methods.

<36> It was observed that 44 of 77 fusion proteins obtained by the above- described method were water-soluble (Table 1-10).

<37> The structures of the obtained TLR/VLR fusion proteins were analyzed and, as a result, it could be seen that the structures of the TLR and VLR modules did not change due to fusion and, in addition, the structures thereof were strictly conserved not only in the overall backbone, but also in the side chains and module fusion sites having structural flexibility.

<38> Accordingly, it can be seen that the fusion LRR technique of the present invention can be used to analyze the structures of proteins through crystallization by preparing fusion proteins of difficult-to-crystallize LRR- family proteins with other LRR-family proteins such as VLR having high crystal Unity.

<39> Furthermore, the functions of the fusion proteins of the present invention were analyzed and, as a result, the fusion proteins maintained not only the structures of proteins, but also the functions of proteins before fusion. This suggests that the fusion proteins can be applied in the preparation and analysis of protein drugs. [Advantageous Effects]

<40> As described above, according to the fusion LRR technique of the present invention, it is possible to effectively fuse two LRR-family proteins with each other while maintaining the structure and function of each of the proteins.

<4i> According to the fusion LRR technique of the present invention, a fusion protein having high crystal Unity can be obtained by fusing an LRR- family protein having high crystal Unity with another LRR-family protein having low crystal Unity. Accordingly, the fusion LRR technique of the present invention is useful to determine the structures of proteins having low crystal Unity. Furthermore, because the active regions and mechanisms of proteins can be understood from the structural analysis of the proteins, the fusion LRR technique of the present invention can be effectively used to understand the in vivo action of the proteins and develop therapeutic drugs.

<42> In addition, according to the fusion LRR technique of the present invention, fusion proteins, which have high crystal Unity while maintaining the function of the respective proteins while having high crystal Unity, can be produced in large amounts. Thus, the fusion LRR technique of the present invention is useful for the development of protein drugs. [Description of Drawings]

<43> FIG. 1 is a schematic diagram showing the molecular structures of VLR of jawless fish and TLR4. <44> FIG. 2 is a conceptual diagram showing the structures of LRR-family protein and second LRR-family protein genes and fusion proteins according to the present invention. <45> FIG. 3 is a schematic diagram showing the cloning of TLR/VLR fusion proteins. <46> FIG. 4 photographically shows a process of purifying mTLR2/VLR fusion proteins. <47> FIG. 5 photographically shows a process of purifying mTLRl/VLR fusion proteins. <48> FIG. 6 shows X-ray diffraction analysis results illustrating the structure of hTLR4/VLR fusion proteins. <49> FIG. 7 shows X-ray diffraction analysis results illustrating the structures of hTLRl/VLR, hTLR2/VLR and mTLR2/VLR. <50> FIG. 8 shows the super imposition of X-ray diffraction analysis results for the VLR fragment region of a TLR4-VLR fusion protein and full-length

VLRB.61. <5i> FIG. 9 shows the super imposition of X-ray diffraction analysis results for the VTR fragment regions of hTLR2-VLR, mTLR2-VLR and hTLRl-VLR fusion proteins and full-length. <52> FIG. 10 shows super imposition of the overlapping regions of TLR4-VLR fusion proteins. <53> FIG. 11 shows super imposition of the overlapping regions of the hT₂V_2 and hT₂V_9 Ca traces of a hT₂V_2/hT₂V_9 fusion protein.

<54> FIG. 12 is a photograph showing the results of SDS-PAGE electrophoresis conducted after expression of a TLR4-VLR fusion protein or co-expression of TLR4 and MD-2. <55> FIG. 13 depicts a graph and electrophoresis photograph showing the results of chromatography of a complex of TLR4 or TLR4/VLR fusion protein with MD-2. <56> FIG. 14 is an electrophoresis photograph showing the change in migration in Native-PAGE, when a TLRl-TLR2-Pam₃CSK4 complex was formed.

<57> FIG. 15 is a chromatogram showing that peaks in Superdeχ-200 gel filtration chromatography (GE Healthcare) shift forward, when a TLR1-TLR2- Pam₃CSK₄ complex was formed.

[Best Mode] <58> Hereinafter, the present invention will be described in detail with reference to examples. However, these examples are for illustrative purposes, and the scope of the present invention is not limited thereto. <59> In order to prove the effectiveness of the fusion LRR technique according to the present invention, each of human TLR4(hTLR4), human

TLR2(hTLR2), mouse TLR2(mTLR2) and human TLRl(hTLR2) was fused with hagfish

VLR.

<60> Example 1: Preparation of fusion proteins <6i> TLR and a hagfish VLRB.61 fragment were fused with each other at the position 'LxxLxLxxN' , in which the relative positions of leucine and asparagine were strictly conserved. A transfer vector containing each of the fusion proteins was constructed in the following manner. <62> The fusion protein consisting of [TLRx 5' region-LxxLxxLxLxxN~VLR 3' region] is briefly expressed as TxV, and [VLR 5' region-LxxLxxLxLxxN-TLRx 3' region] as VTx. Herein, "x" is an integer ranging from 1 to 4. <63> (1) Preparation of hTLR4/VLR fusion proteins <64> The full-length amino acid sequences of hTLR4 and VLR are shown in SEQ

ID NO: 1 and SEQ ID NO: 2, respectively. <65> To prepare an hTLR4/VLR fusion protein, the positions of fusion of hTLR4 and VLR, characterized by the sequence 'LxxLxLxxN' as shown in Table 1, were selected.

<66> I)-I : Construction of transfer vector <67> (i) Construction of hT₄V:

<68> First, a gene fragment having a structure of [Xbal site-TLR4 5' region-Nhel site] was obtained by performing PCR using a 293T HEK (human embryonic kidney) cDNA library as a template with primer sets shown in Table 2. Also, a gene fragment having a structure of [Nhel site-VLR 3' region-Notl site] was obtained by performing PCR using a VLRB.61 gene (obtained from Masanori Kasahara, Department of Pathology, Hokkaido University Graduate School of Medicine) as a template with primer sets shown in Table 2. The two obtained gene fragments were linked with each other by a Nhel restriction enzyme site, and then inserted between Xbal and Not I of a pAcGP67A expression vector (BD Biosciences).

<69> Meanwhile, a human IgGl Fc gene fragment (tag for protein purification) was isolated by performing PCR using a 293T HEK cell cDNA library as a template with primer sets shown in Table 2. Then, the isolated gene fragment was inserted between Not I and BgIII of the pAcGP67A (BD Bioscience, CAT #:554756), thus constructing a transfer vector. Herein, between the fusion gene and the Fc tag, a thrombin cleavage site was present.

<70> (ii) Construction of hVT₄:

<71> Gene fragments having structures of [Xbal site-VLR 5' region-Nhel site] and [Nhel site~TLR4 3' region-Notl site], respectively, were obtained in the same manner as in the above section (i), and a transfer vector was constructed in the same manner as the above section (i).

<72> l)-2: Expression and purification of fusion proteins

<73> Each of the above-constructed transfer vectors was mixed with baculogold DNA (linear baculovirus genome, BD Biosciences) at a ratio of 1:1 and infected into Sf9 insect cells, which were then cultured in Grace's insect medium at 27 °C for 5 days, thus constructing "recombinant baculovirus" having the fusion gene inserted therein. The cultured recombinant baculovirus was infected into fresh Sf9 insect cells, which were then cultured for 2 days. The cultured recombinant baculovirus was stored at 4 °C before use in experiments.

<74> The TLR4/VLR fusion protein was expressed using Hi5 insect cells (Invitrogen). For this purpose, Hi5 cells grown in Sf900 II SFM (GIBCO) medium were placed in a flask at a concentration of 60 x 10⁴ /mL, and after 16 hours, the cells were inoculated with the prepared recombinant baculovirus in an amount of 40 mL/L of Hi5 and expressed at 27 °C at 100 rpm for 3 days to express the fusion protein.

<75> After completion of the culture, the cell culture was purified by protein A Sepharose (GE Healthcare) affinity chromatography to purify hTLR4/VLR-Fc fusion proteins. Then, the cell culture was treated with thrombin (Haematologic Technologies) at 4 °C for 12 hours to remove the Fc tag, and then purified again by ion exchange (Q Sepharose) and Superdeχ-200 gel filtration chromatography (GE Healthcare). During the purification process, the pH of the cell culture was maintained with a buffer containing 20 mM Tris (pH8.0) and 200 mM NaCl. Among 15 expressed fusion proteins, 7 water-soluble fusion proteins were collected (Table 1). The sequences of the 7 obtained fusion proteins are shown in SEQ ID NOS: 35-41.

<76> [table 1]

<77>

<78> [table 2]

<79>

<80> (2) Preparation of mTLR2/VLR fusion proteins <81 > The full-length amino acid sequence of mTLR2 is shown in SEQ ID NO:

42.

<82> To prepare an mTLR2/VLR fusion protein, the positions of fusion of mTLR2 and VLR, characterized by the sequence 'LxxLxLxxN' as shown in Table 3, were selected.

<83> The construction of transfer vectors and the purification of a fusion protein were performed in the same manner as in Example 1-(1), except that mTLR2 was used instead of hTLR4, primer sets in Table 4 were used instead of the primer set shown Table 2, and pVL1393 (BD Biosciences, CAT #: 554745) was used instead of the pAcGP67A (FIG. 4). However, for mTLR2/VLR fusion proteins, HiTrapQ was used instead of Q sepharose in the purification process. The sequences of the obtained fusion proteins are shown in SEQ ID NOS: 53-59.

<84> [table 3]

The amino acid residue added according to the limitation enzyme insertion at the cloning

<85> process was indicated in parenthesis. <86> [table 4]

<87>

<88> (3) Preparation of hTLR2/VLR fusion proteins

<89> The full-length amino acid sequence of hTLR2 is shown in Table 60. <90> To prepare hTLR2/VLR fusion proteins, the positions of fusion of hTLR2 and VLR, characterized by the sequence 'LxxLxLxxN' , were selected.

<91> The construction of transfer vectors and the purification of fusion proteins were performed in the same manner as in Example 1-(1) above, except that hTLR2 was used instead of hTLR4, and primer sets in Table 6 were used instead of the primer sets shown in Table 2 (FIG. 4). The sequence of the obtained fusion proteins are shown in SEQ ID NOS: 75-85. <92> [table 5]

*The amino acid residue added according to the limitation enzyme insertion at the cloning

<93> process was indicated in parenthesis.

<94> [table 6]

<95>

<96> (4) Preparation of mTLRl/VLR fusion proteins

<97> The full-length amino acid sequence of mTLRl is shown in SEQ ID NO:

86.

<98> To prepare mTLRl/VLR fusion proteins, the positions of fusion of mTLRl and VLR, characterized by the sequence 'LxxLxLxxN' as shown in Table 7, were selected.

<99> The construction of transfer vectors and the purification of fusion proteins were performed in the same manner as in Example l-(2) above, except that mTLRl was used instead of mTLR2, and primer sets in Table 8 were used instead of the primer sets shown in Table 4 (FIG. 5). However, for mTLRl/VLR fusion proteins, HiTrapSP (GE Healthcare) was used instead of HiTrapQ in the purification process. The sequences of the obtained fusion proteins are shown in SEQ ID NOS: 96-101.

<100> [table 7]

* The am no ac d residue added according to the limitation enzyme nser on a e c on ng process was indicated in parenthesis.

<101> <102> [table 8]

<103>

<104> (5) Preparation of hTLRl/VLR fusion proteins

<105> The full-length amino acid sequence of hTLRl is shown in SEQ ID NO:

102.

<106> To prepare hTLRl/VLR fusion proteins, the positions of fusion of hTLRl and VLR, characterized by the sequence 'LxxLxLxxN' , were selected. <107> The construction of transfer vectors and the purification of fusion proteins were performed in the same manner as in Example 1-(1), except that hTLRl was used instead of hTLR4, and primer sets in Table 10 were used instead of the primer sets shown in Table 2 (FIG. 5). The sequences of the obtained fusion proteins are shown in SEQ ID NOS: 112-118.

<108> [table 9]

The amino acid residue added according to the limitation enzyme insertion at the cloning process was indicated in parenthesis.

<109> <110> [table 10]

<111>

<112> Example 2: Analysis of structure of fusion protein <113> In order to confirm whether the structure of the TLR module or the VLR module changes due to the fusion of proteins, the fusion proteins prepared in Example 1 were crystallized, and the structures thereof were determined using X-ray diffraction analysis.

<114> (1) Crystallization <115> Among the fusion proteins prepared in Example 1, the fusion proteins shown in Table 11 were crystallized using the hanging-drop vapor diffusion method. Optimized crystallization conditions for each of the fusion proteins are shown in Table 11. In Table 11, the complex of the TLR/VLR fusion protein with another protein means conditions for the crystallization of protein complexes obtained in Example 3 below. More specifically, 1μl of protein solution was mixed with 1μl of crystallization buffer and crystallized using the hanging-drop vapor diffusion method for 1 week. To the obtained crystals, the freezing buffers shown in Table 11 were added, and the solutions were stored -170 °C . <i i6> [table 11]

<117>

<118> (2) X-ray diffraction analysis

<119> Diffraction data were obtained using the 4A beamline of the Pohang Accelerator Laboratory, the ID14-2 beamline of SPring-8, the beamline of ESRF, or BL5.0.1 of ALS. HKL2000 package (Otwinowski and Minor, Macromolecular Crystallography, part A, 28: 307-326, 1997) and MOSFLM/SCALA program (WinnJ.SynchrotronRadiat . 10: 23-25, 2003) were used in diffraction data analysis.

<120> The initial phase was calculated by molecular replacement using the program PHASER (McCoy AJ, et al . , Acta Crystal logr. D Biol. Crystal logr. 61 : 458-464, 2005). The atomic model was built by iterative modeling and refinement using the program 0 and CNS (Brunger AT, et al . , Acta Crystallogr. D Biol. Crystal logr. 54 : 905-921, 1998; Jones TA, et al., Acta Crystal logr. A 47 (Pt 2): 110-119, 1991). The final models were further refined using the program REFMAC (Murshudov GN et al , ActaCrystal logr . D Biol. Crystallogr. 54: 905-921, 1998; Jones TA et al , Acta Crystallogr. 53, 240-255, 1997) with the TLS program parameters generated by the TLSMD server (Painter J and Merritt EA, Acta Crystallogr. 62, 439-450, 2006).

<121> Among the fusion proteins crystallized in the method of Example 2-(l), the crystal structures of hT₄V_3, hT₄V_8, hVT4_3, hT₂V_2, hT₁V_8-hT₂V_9-

Pam3CSK4, mT₂V_6-Pam3CSK4 and mT₂V_6-Pam2CSK4 were analyzed, and the analyzed results are shown in FIGS. 6 and 7.

<122> FIG. 6 shows the structure of the TLR4/VLR fusion. In FIG. 6, the blue color indicates the TLR4 fragment, and the grey color indicates the VLRB.61 fragment.

<123> FIG. 7 shows the overall structures of the hTLRl/VLR fusion protein and the hTLR2/VLR fusion protein. In FIG. 7, the hTLRl fragment, the hTLR2 fragment and the VLR fragment are indicated by green, blue and grey, respectively. The central region is indicated by light green and sky blue, the Pam3CSK4 lipopeptide is indicarted by red, and the disulfide bridge is indicated by yellow. The region corresponding to the TLR1/VLR fusion protein is indicated by the symbol "*".

<124> (3) Analysis of whether VLR module structure changed

<125> The VLR structure obtained in the method of Example 2-(2) was superimposed on the VLR module portion among the structure of the fusion protein obtained in Example 2-(2), and the superimposed results are shown in FIG. 8 and FIG. 9. FIG. 8 shows the super imposition of the VLRB.61 structure on the VRL module of the TLR4/VLR fusion protein, and FIG. 9 shows the super imposition of the VLRB.61 structure on the VLRB.61 structures of the TLR1/VLR and TLR2/VLR fusion proteins.

<126> As can be seen in FIGS. 8 and 9, the backbone atoms of VLRB.61 were substantially the same as those of the VLR modules of the fusion proteins, and the C rms difference after super imposition was 0.3-0.9 A, suggesting that the structure of the VLR module did not substantially change after the fusion of the proteins. In addition, the structures of not only the backbone atoms, but also side chains, did not greatly change even at the fusion boundaries, suggesting that the structures were practically identical to the original structure of VLRB.61.

<127> (4) Analysis of whether TLR module structure changed <128> Because the structures of full-length human TLR4 and TLR2 are unknown, it was impossible to directly compare the structures of TLR4 and TLR2 fragments of the fusion protein with hTLR4 and hTLR2. For this reason, to deduce the entire structure of TLR4, the structures of overlapping regions of three TLR4/VLR fusion proteins shown in FIG. 6 were compared with each other, and the results are shown in FIG. 10.

<129> As can be seen in Table 1 above, amino acid residues 27-227 of TLR4 included in the hT4V_3 fusion protein were included in the hT₄V_8 fusion protein, and 145 amino acid residues 383-527 of TLR4 were common to the hVT₄_3 and hT₄V_8 fusion proteins. As can be seen in FIG. 10, the overlapping regions had average C rms distances of 0.39 A and 0.28 A, suggesting that they had the same backbone structure. As in the VLR module of Example 2-(2), the structural homology was not limited only to the backbone structures and could also be confirmed at the positions of fusion of flexible side chains which were a distance from the backbone structures.

<130> In the case of the TLR2 fragment, as can be seen in Table 5, amino acid residues 1-234 of TLR2 included in the hT₂V_3 fusion protein were included in the hT₂V_9 fusion proteins. As shown in FIG. 11, the overlapping regions of the hT₂V_2 and hT₂V_9 fusion proteins were structurally substantially the same even at the fusion boundary of hT₂V_2,.

<131> Accordingly, it can be seen that, even when the TLR4 and TLR2 fragments are fused with the VLR fragment, the change in the structures thereof is negligibly small.

<132> From the results of Examples 2-3) and 2-4), it could be seen that, in the fusion proteins produced by fusion of LRR regions, neither VLR nor TLR underwent a significant structural change caused by fusion.

<133> Example 3: Analysis of functions of fusion proteins

<134> (1) Analysis of whether hTLR4/VLR functions maintained their functions

<135> TLR4 and MD-2 forms a heterodimer which recognizes the LPS of gram- negative bacteria.

<136> Whether the function of TLR4 in the TLR4 fusion protein is maintained was analyzed by testing whether the TLR4 fusion protein forms a complex with MD-2 and whether the complex also binds to LPS or LPS antagonist Eritoran.

<137> 1) Binding to MD-2

<138> MD-2 (obtained from 293T HEK cell cDNA) having a protein A tag fused thereto was expressed in Hi5 cells together with hTLR4 or the fusion protein hT₄V_3, hT₄V_4, hT₄V_5, hT₄V_8 or hT₄V_9, and the cells were purified by IgG

Sepharose (GE Healthcare) affinity chromatography to purify only MD-2 and MD- 2-bound proteins. The protein A tag of the purified proteins was removed using thrombin, and then the proteins were purified again by ion exchange (S Sepharose) and gel filtration chromatography. The proteins were separated by 12% SDS PAGE. <139> As can be seen in FIG. 12, not only full-length TLR4, but also hT₄V_3, hT₄V_4, hT₄V_5, hT₄V_8 and hT₄V_9, having MD-2-binding site, were bound to MD-

2, suggesting that the function of full-length TLR4 was maintained even in the fusion proteins of TLR4.

<140> 2) Binding of TLR4-MD-2 complex with LPS or Eritoran <141> E.coli LPS Ra (Sigma, L9641) or Eritoran (Eisai Inc.) was sonicated for 10 minutes and cultured with 1 mg/ml of the TLR4-MD-2 complex or hT₄V_8-

MD2 complex, obtained in Example 3-(2), at 37 °C for 3 hours, the weight ratio of the protein complex to the LPS or Eritoran protein was maintained at 10:1. The binding of LPA RA or Eritoran to the protein complex was observed by 8% native and SDS PAGE and Superdex 200 gel filtration chromatography (GE Healthcare), and the observation results are shown in FIG. 13.

<142> The left figures of FIG. 13 are graphic diagrams showing molecular weights corrected after gel filtration chromatography. In the left figures of FIG. 13, the block color shows a gel filtration chromatogram of the culture with the TLR4-MD-2 complex or hT₄V_8-MD2 complex alone, the red color shows a gel filtration chromatogram of the culture with LPS, and the sky blue shows a gel filtration chromatogram of the culture with Eritoran. In the left figures, the surface molecular weights of the complexes, calculated from elution volumes, are shown together with expected molecular weights in brackets. The right figures of FIG. 13 shows the results of native PAGE and SDS PAGE analysis. The dimeric complexes migrated slower than the monomeric complexes and are indicated by the symbol "*".

<143> As can be seen in FIG. 13, the binding of Eritoran to hTLR4-MD~2 and hT4V_8~MD2 did not change the elution volume of gel filtration chromatography and caused only a slight change in migration in native gel electrophoresis. The Eritoran/hTLR4 protein complex migrated slightly faster than the monomer ic complex due to the negative charge of the ligand.

<144> However, the binding of LPS caused a great change in the aggregation of the TLR4-MD-2 and hT₄V_8-MD2 complexes. The TLR4-MD-2 and hT₄V_8-MD2 complexes bound with LPS were all rapidly eluted in gel filtration chromatography and migrated upward in native gel electrophoresis. In gel filtration chromatography, the elution volumes of the LPS-bound proteins coincided with the expected elution volumes of the heterotetramers. <145> From the above results, it can be seen that hT₄V_8-MD2 that is a complex of the fusion protein with MD-2 was the same as TLR4-MD-2 with respect to the responses to LPS and Eritoran, suggesting that the structure thereof in the fusion protein with LRR was maintained and the function thereof was also retained. <146> (2) Analysis of whether ligand-binding function of fusion protein is retained <147> To test if the functions of the fusion proteins are retained, ligand- binding tests of the TLRl fusion proteins and the TLR2 fusion proteins were performed. <148> To form complexes, a synthetic R-isomer of PaIn₃CSK₄ (EMC

Microcol lections GmbH) as a ligand was incubated with 5 mg/ml of purified hT₂V_9 at room temperature for 3 hr in a buffer containing 20 mM Tris HCl (pH

8.0) and 200 mM NaCl. Whether the ligand has been bound was observed in native-PAGE. It could be seen that, when hT₂V_9 was bound to Pam₃CSK₄, bands on native-PAGE gels shifted upward slightly (FIG. 14). <149> A hT₁V_8-hT₂V_9-Pam₃CSK4 complex was prepared by the stepwise binding of hTiV_8 with hT₂V_9. First, hT₂V_9 and Pam₃CSK₄ were incubated at room temperature for 2 hours at a buffer containing 20 mM Tr is HCl (pH 8.0) and 200 mM NaCl. Then, hT₁V_8 was added thereto, and the resulting material was incubated at 37 °C for 2 hours. The concentration of each of hT₁V_8 and hT₂V_9 was 2.5 mg/ml, and the molar ratio of Pam₃CSK₄ to the proteins was maintained at 5:1. To remove unbound TLR protein and an excess of PaIn₃CSK₄, the incubated material was purified by Superdeχ-200 gel filtration chromatography (FIG. 15). Because gel filtration chromatography performs purification using the difference in size between proteins, the Pam₃CSK₄-bound hT₁V_8-hT₂V_9 complex was rapidly eluted in gel filtration chromatography and migrated upward in native gel electrophoresis (FIG. 14 and FIG. 15). In gel filtration chromatography, the elution volume of the hT₁V_8-hT2V_9-Pam₃CSK₄ complex coincided with the expected elution volume of the heterodimer. <150> A complex of an R-isomer of Pam₂CSK₄ (EMC Microcol lections GmbH) with mT₂V_6 was prepared in the same manner as described above. mT₂V_6 and PaIn₂CSK₄ were incubated at room temperature for 2 hours in a buffer containing 20 mM Tr i s HCl (pH 8.0) and 200 mM NaCl . The molar rat io of Pam₃ CSK₄ to the protein was maintained at 3:1.

<151> As described above, the fusion proteins according to the present invention showed the same ligand-binding properties as those of the native proteins. Accordingly, it can be deduced that these fusion proteins also retain functions like those of the native proteins. [Industrial Applicability]

<152> As described above, according to the fusion LRR technique of the present invention, LRR proteins having low crystal linity can be produced in large amounts in the form of fusion proteins having high crystal linity while maintaining the structures and functions of the respective proteins. Thus, the fusion LRR technique is useful for the development of protein drugs.

<153> In addition, because the fusion LRR technique can be used advantageously used to determine the structures of LRR proteins having low crystal linity, it can be effectively used to understand the in vivo action of the proteins and develop therapeutic drugs.

Claims

[CLAIMS] [Claim 1]

<155> A method of fusion between LRR (Leucine Rich Repeat )-fami Iy proteins, the method comprising fusing the LRR-family proteins with each other at the following consensus sequence position: <156> LxxLxxLxLxxN

<157> wherein L represents alanine, glycine, phenylalanine, tyrosine, isoleucine, valine or tryptophane, N represents asparagine, glutamine, serine, cysteine or threonine, and X represents any amino acid.

[Claim 2]

<158> The method of Claim 1, wherein one of the LRR-family proteins is selected from among toll-like receptor (TLR) proteins.

[Claim 3]

<159> The method of Claim 1, wherein one of the LRR-family proteins is LRR- family protein having high crystallinity.

[Claim 4]

<160> The method of Claim 3, wherein the LRR-family protein is variable lymphocyte receptor (VLR) protein.

[Claim 5] <161> A fusion protein prepared according to any one of Claims 1 to 5.

[Claim 6] <162> A nucleic acid molecule encoding the fusion protein of Claim 6.

[Claim 7] <163> A method for analyzing the structures of LLR proteins, the method comprising the steps of:

<164> (A) crystallizing the fusion protein of Claim 1 and any one of LRR- family proteins constituting the fusion protein, and analyzing the structures of the crystallized proteins; and

<165> (B) excluding the crystallized LRR-family protein from the fusion protein to determine the structure of another protein constituting the fusion protein.

[Claim 8]

<166> The method of Claim 7, wherein the analysis of the structures is performed by X-ray diffraction analysis.

[Claim 9]

<167> A vector containing a gene fragment encoding the fusion protein of Claim 5.