Protein Science Sheba protein
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 QUICK SEARCH:   [advanced]


     


Protein Science (2005), 14:341-350. Published by Cold Spring Harbor Laboratory Press. Copyright © 2005 The Protein Society
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Furutani, M.
Right arrow Articles by Maruyama, T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Furutani, M.
Right arrow Articles by Maruyama, T.
Social Bookmarking
 Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

An engineered chaperonin caging a guest protein: Structural insights and potential as a protein expression tool

Masahiro Furutani1, Jun-Ichi Hata1, Yasuhito Shomura2,7, Keisuke Itami3, Takao Yoshida4, Yoshitaka Izumoto1, Akiko Togi1, Akira Ideno1, Takuo Yasunaga3, Kunio Miki2,5 and Tadashi Maruyama4,6

1 Research and Development Institute, Sekisui Chemical Co. Ltd., Shimamoto-cho, Mishima-gun, Osaka 618-8589, Japan
2 Department of Chemistry, Graduate School of Science, Kyoto University, Sakyo-ku, Kyoto 606-8502, Japan
3 Department of Bioscince and Bioinfomatics, Kyushu Institute of Technology, Iizuka-shi, Fukuoka 820-8502, Japan
4 Japan Agency for Marine-Earth Science and Technology (JAMSTEC), Yokosuka 237-0061, Japan
5 RIKEN Harima Institute/SPring-8, Mikazuki-cho, Sayo-gun, Hyogo 679-5148, Japan
6 Marine Biotechnology Institute, Kamaishi, Iwate 026-0001, Japan

Reprint requests to: Masahiro Furutani, Sekisui Chemical Co. Ltd., Hyakuyama 2-1, Shimamoto-cho, Mishima-gun, Osaka 618-8589, Japan; e-mail: furutani002{at}sekisui.jp; fax: 81-75-962-8903.

(RECEIVED August 12, 2004; FINAL REVISION October 21, 2004; ACCEPTED October 22, 2004)


    Abstract
 TOP
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 References
 
The structure of a chaperonin caging a substrate protein is not quite clear. We made engineered group II chaperonins fused with a guest protein and analyzed their structural and functional features. Thermococcus sp. KS-1 chaperonin {alpha}-subunit (TCP) which forms an eightfold symmetric double-ring structure was used. Expression plasmids were constructed which carried two or four TCP genes ligated head to tail in phase and a target protein gene at the 3' end of the linked TCP genes. Electron microscopy showed that the expressed gene products with the molecular sizes of ~120 kDa (di-TCP) and ~230 kDa (tetra-TCP) formed double-ring complexes similar to those of wild-type TCP. The tetra-TCP retained ATPase activity and its thermostability was significantly higher than that of the wild type. A 260-kDa fusion protein of tetra-TCP and green fluorescent protein (GFP, 27 kDa) was able to form the double-ring complexes with green fluorescence. Image analyses indicated that the GFP moiety of tetra-TCP/GFP fusion protein was accommodated in the central cavity, and tetra-TCP/GFP formed the closed-form similar to that crystallographically resolved in group II chaperonins. Furthermore, it was suggested that caging GFP expanded the cavity around the bottom. Using this tetra-TCP fusion strategy, two virus structural proteins (21–25 kDa) toxic to host cells or two antibody fragments (25–36 kDa) prone to aggregate were well expressed in the soluble fraction of Escherichia coli. These fusion products also assembled to double-ring complexes, suggesting encapsulation of the guest proteins. The antibody fragments liberated by site-specific protease digestion exhibited ligand-binding activities.

Keywords: chaperonin; structure; archaea; fusion protein; protein folding; protein expression

Abbreviations: TCP, Thermococcus sp. KS-1 (JCM 11816) chaperonin {alpha}-subunit • hCR, human antibody heavy chain constant region (CH1–CH2–CH3) • HEL, hen egg lysozyme • HBs, hepatitis B surface antigen • HCc, hepatitis C core antigen

Article and publication are at http://www.proteinscience.org/cgi/doi/10.1110/ps.041043905.


    Introduction
 TOP
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 References
 
Chaperonin has a unique quaternary structure and can function as a folding chamber for both nascent and nonnative polypeptides (Hartl and Hayer-Hartl 2002). Chaperonins (~60 kDa) form seven- to ninefold symmetric double-ring complexes of ~850–1000 kDa, resulting in a central cavity in each ring. Chaperonins promote protein folding by several sequential steps: binding of partially folded polypeptides to the apical domain by hydrophobic contact, encapsulating them into the central cavity, progress of protein folding in the cavity, and releasing of the folded polypeptides coupled with ATP hydrolysis (Hartl and Hayer-Hartl 2002). The most studied chaperonin Escherichia coli GroEL-GroES complex (GroE) is called "Anfinsen cage," because it allows substrate polypeptides to fold in the same manner, as it would be in free solution in the classic Anfinsen’s refolding experiment (Ellis 2001). It was also reported that an archaeal group II chaperonin from Thermococcus sp. KS-1, promotes protein folding in the central cavity with the closed form in which the built-in lid regions cap the cavity (Yoshida et al. 2002b). However, structure of chaperonins during encapsulating a substrate polypeptide is not quite clear. Preparation of a chaperonin–substrate complex in which the substrate protein is caged in the cavity has been difficult. Crystal structures of bacterial and archaeal chaperonins have shown that the N- and C- termini are closely located in the central cavity (Braig et al. 1994; Ditzel et al. 1998; Shomura et al. 2004). It was also reported that the covalent connection of seven adjacent GroEL subunits does not disturb its double-ring structure or function (Farr et al. 2000). These reports present the possibility that a fusion protein comprising the covalently linked chaperonin subunits and a guest protein could form ring complexes, thus enabling the fused guest protein moiety to be folded in the chaperonin cavity.

To test this idea, we used the {alpha}-subunit of chaperonin (TCP) (Yoshida et al. 1997) from a hyperthermophilic archaeum Thermococcus sp. KS-1 (JCM 11816). Archaeal group II chaperonins including TCP do not require a cofactor like GroES in their protein folding activities (Yoshida et al. 1997; Furutani et al. 1998), and their apical domains probably function as a built-in lid (Horwich and Saibil 1998; Yoshida et al. 2002b). The TCP has an eightfold symmetric ring structure (Shomura et al. 2004), which allows us to utilize TCP subunit monomer, linked dimer, linked tetramer, or linked octamer as a fusion partner for a guest protein. The resultant fusion proteins are expected to assemble into the double-ring complex like that of wild-type TCP.

Here, we present a chaperonin-guest complex using the TCP subunit linked oligomer as a fusion partner, and discuss the structural and functional features of this complex caging a guest protein in the TCP cavity. The potential of this chaperonin fusion strategy as a protein expression tool is also described.


    Results
 TOP
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 References
 
Design and expression of mono-, di-, tetra-, and octa-TCP
Crystallography of TCP revealed that the N- and C-terminal resolvable residues are Val 9 and Ala 526, respectively (Shomura et al. 2004), and that these residues are closely located in the central cavity (Fig. 1AGo). This suggests that the covalent connection of TCP subunits between the C-terminus of a subunit and the N-terminus of its adjacent subunit does not disturb the double-ring structure. In order to cage guest proteins with different molecular sizes in the TCP cavity, various covalently linked TCP subunits fused with a guest protein were designed (Figs. 1BGo, 2Go). The ratio of the TCP subunit to a guest protein is possible to be from 1:1 to 8:1. Increasing the ratio of the TCP subunit against a guest protein would increase the molecular size of a guest accommodated. For example, when the ratio of the TCP subunit to a target protein is 4:1, two molecules of a 30-kDa guest protein are expected to be encapsulated in the single TCP cavity, of which inside room is probably enough for a 60-kDa globular protein.



View larger version (31K):
[in this window]
[in a new window]
 
Figure 1. Structural basis of strategy for tandem joining of TCP subunits and caging a guest protein in the central cavity. (A) C{alpha} trace of a TCP ring at the level of its equatorial domain, showing the crystallographically resolvable termini of the subunits. Val 9 and Ala 526 are numbered. Eight N-terminal residues preceding the Val 9 and 22 C-terminal residues beyond Ala 526 are not crystallographically resolvable (Shomura et al. 2004). (B) Schematic diagram of the possible arrangements of the covalently linked TCP subunits and guest proteins with different molecular sizes. Increasing the ratio of TCP subunit to a guest protein allows the increase of the molecular size of the encapsulated guest protein. Blue and pink circles indicate the covalently linked TCP subunits and guest proteins, respectively. (C) Close-up view of two adjacent TCP subunits showing their terminal residues in the TCP ring. Distances between Ala 526 of a TCP subunit and N-terminal residues Val 9, Ile 11, and Glu 14 of the next subunit are 20.0, 16.5, and 13.5 Å. (A) and (C) were created with DS Viewer5.0 (Accelrys) using the TCP structure (Protein Data Bank entry, 1Q2V [PDB] ).

 


View larger version (32K):
[in this window]
[in a new window]
 
Figure 2. Construction of vectors expressing the covalently linked TCP subunits and TCP/guest fusion proteins. PT7, TT7, PS, CS, and 6His indicates T7 promoter, T7 terminator, PreScission protease digestion site, the cloning site, and 6 histidines tag, respectively. TCP (n), (n) indicates a number of ligated TCP genes (n = 1, 2, 4, or 8).

 
To connect two adjacent TCP subunits via Ala 526 of the preceding subunit, Val 9 of the following subunit was selected, because the hydrophobic residues from Val 9 to Pro 13 form hydrophobic contacts with the adjacent subunit (Fig. 1CGo). A covalent connection of adjacent TCP subunits with a short rigid peptide may disturb the quaternary structure of TCP. Conversely, a connection via a too long flexible peptide may increase the likelihood of degradation by host proteases. Because these residues are 20.0 Å apart (Fig. 1CGo), TCP subunits were covalently connected with a flexible octapeptide linker of serine and glycine (GGSGGGGS).

While mono-TCP, di-TCP, and tetra-TCP were expressed as dominant proteins in the soluble fraction (Fig. 3AGo, lanes 2–4), the linked octamer (octa-TCP) was not well expressed (data not shown). The expressed TCPs were purified to homogeneity by chromatographies with a Ni-chelating sepharose column and a RESOURCE Q column (Fig. 3AGo, lanes 5–7).



View larger version (83K):
[in this window]
[in a new window]
 
Figure 3. Double-ring formation and dissociation of mono-, di-, and tetra-TCP. (A) Expression and purification of mono-, di-, and tetra-TCP. Soluble extracts of control E. coli cells with pET3-d (lane 1), soluble cell extracts induced with the expression of mono-, di-, and tetra-TCP (lanes 24), purified mono-, di-, and tetra-TCP (lanes 57). (B) Gel filtration profiles of the purified mono-, di-, and tetra-TCP. Elution was conducted in 25 mM HEPES-KOH (pH 7.0) containing 50 mM MgCl2 and 200 mM KCl. The mono-TCP was incubated with 1 mM ATP before the gel filtration. Open and solid arrowheads indicate the peaks corresponding to the TCP double-ring (R.T.; 11.5 min) and the monomeric form of mono-TCP (R.T.; 16.8 min), respectively. (C) Electron micrographs of mono-, di-, and tetra-TCP. White arrowheads, side-on views of double-rings. (D) Gel filtration profiles showing dissociation of di- and tetra-TCP upon removal of Mg2+. Elution was conducted in 25 mM HEPES-KOH (pH 7.0) containing 1 mM EDTA and 200 mM KCl. Open arrowhead, the peak of the TCP double-ring.

 
Double-ring formation and dissociation of mono-, di-, and tetra-TCP
The purified mono-, di-, and tetra-TCP were applied to a TSK gel G3000SWXL column and eluted with HEPES-KOH (pH 7.0) buffer containing 50 mM MgCl2 and 200 mM KCl. Most of the mono-TCP was eluted at 16.8 min and was thought to be a monomeric form (data not shown). After the addition of ATP (1.0 mM), a new peak appeared at 11.5 min immediately after the void volume (Fig. 3BGo). Electron microscopy of this peak fraction showed that mono-TCP formed the double-ring structure similar to that of wild-type TCP (Yoshida et al. 1997; Fig. 3CGo). Gel filtration of di- and tetra-TCP showed peaks at 11.5 min, corresponding to the double ring without the addition of ATP (Fig. 3BGo). Electron microscopy of these peak fractions of di- and tetra-TCP confirmed the formation of double rings (Fig. 3CGo). These results indicate that the covalent connection of the adjacent TCP subunits with the octapeptide linker (GGSGGGGS) did not disturb the quaternary structure of TCP.

The gel filtration profiles of di- and tetra-TCPs also showed some retarded minor peaks in addition to the main peak at 11.5 min (Fig. 3BGo). Those minor peaks were likely to represent misassembled di- and tetra-molecules. Upon removal of Mg2+ with 1 mM EDTA, areas of the 11.5 min peaks of double-ring forms of di- and tetra-TCP significantly reduced (Fig. 3DGo). These observations indicate that the double-ring structure of the linked TCP subunits is unstable in the absence of Mg2+ and ATP.

ATPase activity of tetra-TCP
The optimum temperature of ATPase activity of tetra-TCP was 80°C, with the ATP hydrolysis rate of 153 mol of ATP hydrolysis/mol of TCP double ring/min (Fig. 4Go), while the optimum temperature of ATPase activity of the wild-type TCP is reported to be 60°C (Yoshida et al. 2002a; Fig. 4Go). While heat inactivation of tetra-TCP was observed at 90°C, 78% of the activity at 80°C was retained (Fig. 4Go). Extensive heat inactivation of the wild-type is observed at 90°C (Yoshida et al. 2002a; Fig. 4Go). These suggest that the thermostability of tetra-TCP was significantly enhanced compared to that of the wild-type TCP.



View larger version (21K):
[in this window]
[in a new window]
 
Figure 4. Effect of temperature on ATPase activity of tetra-TCP. The tetra-TCP was incubated in the ATPase assay mixture for 30 min at the indicated temperature, and liberated Pi was measured by the malachite green method as described in Materials and Methods. Open circles indicate tetra-TCP. Filled circles indicate the wild-type TCP of which data have been previously reported (Yoshida et al. 2002a).

 
Double-ring formation of tetra-TCP/GFP fusion protein
The cavity size of the TCP ring is 148,000 Å3 (Shomura et al. 2004) equivalent to a globular protein of ~49 to ~87 kDa, assuming a perfect fit of a guest protein in the cavity (1.7–3.0 Å3 /Da) (Matthews 1968). To investigate whether the TCP ring could be formed when a guest protein was fused with the covalently linked TCP subunits, the wild-type 27 kDa-green fluorescent protein (GFP) (from Aequorea victoria) was employed as a model guest protein. Using the crystal structures of GFP (Ormo et al. 1996) and TCP (Shomura et al. 2004), we simulated packing of two GFP molecules into the TCP cavity with BOBSCRIPT (Esnouf 1999), and found that two GFP molecules could fit the cavity without steric collision (data not shown). This modeling suggested that tetra-TCP/GFP fusion protein could form a double-ring structure without steric hindrance. The tetra-TCP/GFP was expressed and purified by a Ni-chelating sepharose column and a RESOURCE Q column (Fig. 5AGo). Gel filtration analysis showed that tetra-TCP/GFP was eluted at the same position as double-rings (at 11.5 min) of mono-, di-, and tetra-TCP. It was also demonstrated that the GFP moiety emitted fluorescence at 510 nm with excitation at 380 nm (Fig. 5AGo). Electron microscopy also revealed that tetra-TCP/GFP formed the double-ring complex (Fig. 5BGo). While free GFP (Fig. 5CGo, lanes 3,6) or GFP in a mixture of free GFP and tetra-TCP (230 kDa) (Fig. 5CGo, lanes 2,5) was found in the immunoprecipitated fraction with anti-GFP antibodies immobilized on Protein A/G beads, most of tetra-TCP/GFP (260 kDa) was detected in the unbound fractions (Fig. 5CGo, lanes 1,4). These results indicate that anti-GFP antibodies could not access the GFP moiety of the fusion protein in the cavity easily.



View larger version (71K):
[in this window]
[in a new window]
 
Figure 5. Double-ring formation of tetra-TCP/GFP fusion protein. (A) Gel filtration profile of tetra-TCP/GFP fusion protein. Soluble cell extract induced with the expression of tetra-TCP/GFP fusion protein (inlet, lane 1), purified tetra-TCP/GFP after a RESOURCE Q column (inlet, lane 2). Elution was conducted in 25 mM HEPES-KOH (pH 7.0) containing 50 mM MgCl2 and 200 mM KCl, and was monitored with UV absorbance (solid line) at 280 nm and fluorescence (dotted line) at 510 nm with excitation at 380 nm. Open arrowhead indicates the peak corresponding to the TCP double ring. (B) Electron micrograph of tetra-TCP/GFP double rings. White arrowheads indicate side-on views of double-rings. (C) Immunoprecipitation of tetra-TCP/GFP fusion protein and other control proteins. Immunoprecipitation was conducted in PBS containing 25 mM MgCl2. B and U, anti-GFP immobilized Protein A/G-bound and -unbound fractions, respectively. Anti-GFP immobilized Protein A/G-bound fractions of tetra-TCP/GFP (lane 1), tetra-TCP and GFP (lane 2), and GFP (lane 3). Anti-GFP immobilized Protein A/G-unbound fractions of tetra-TCP/GFP (lane 4), tetra-TCP and GFP (lane 5), and GFP (lane 6). Solid and open arrowheads, tetra-TCP and GFP, respectively. The band slightly above 50 kDa and the series of weak bands around 25 kDa correspond to heavy chains and light chains of polyclonal anti-GFP antibodies, respectively (lanes 13). The band slightly below 75 kDa probably corresponds to albumin from anti-GFP serum (lanes 46).

 
Image-analyses of the double-ring structure of tetra-TCP/GFP fusion protein and tetra-TCP
The averaged images of tetra-TCP/GFP and tetra-TCP are shown in Figure 6A and 6BGo, respectively. These images were obtained after classification and directional refinement of about 300 particles extracted from electron micrographs. They were classified into top views (Fig. 6A-a,A-b,A-d,A-e,B-a–dGo) and side views (Fig. 6A-f,B-eGo). We examined the change of the density and the statistical significance of the differences in the top or side views (Fig. 6CGo). These analyses reveal that the higher density region in the cavity exists in tetra-TCP/GFP and the region is likely to be the fused GFP moiety. The size of the higher density region surrounded with blue lines in the cavity, 2–3 nm, is consistent with that of two expected GFPs (Fig. 6C-aGo). The higher density regions surrounded by blue or red lines were observed around the equatorial domain (Fig. 6C-bGo). These observations suggest that the fused GFP was at the bottom of the cavity. Furthermore, we observed the lower density regions (surrounded with green lines) between each TCP subunit (Fig. 6C-aGo), and between the equatorial domains of tetra-TCP/GFP double-ring (i.e., in the gap of double-ring) (Fig. 6C-bGo). On the other hand, the higher density regions (surrounded with red lines) appeared in a peripheral region of the TCP ring (Fig. 6C-a,C-bGo). This suggests that structural change of each TCP subunit was induced and the cavity was expanded in tetra-TCP/GFP; we cannot detect the significant conformational change around the lid, i.e., open–close conformational change, between tetra-TCP and tetra-TCP/GFP. The both lids of the double-rings of tetra-TCP and tetra-TCP/GFP seem to be closed. The distances (indicated with an orange or a red couple head arrow) between the particle center and the higher, or lower density region in the top view analysis (Fig. 6C-aGo) are consistent with those represented near the bottom of the cavity in the side view analysis (Fig. 6C-bGo). These observations suggest that caging of GFP caused expansion of the bottom area of the cavity.



View larger version (134K):
[in this window]
[in a new window]
 
Figure 6. Image analyses of tetra-TCP/GFP fusion protein and tetra-TCP. (A) Averaged images of tetra-TCP/GFP. The images were obtained after cluster analysis and refinement. The number of images to be averaged was 41, 34, 29, 17, 9, or 11 for af. The images a,b,d,e appear to be top views and f, a side view. The scale bar is 5 nm. (B) Averaged images of tetra-TCP. The images were obtained after cluster analysis and refinement. The number of images to be averaged was 77, 52, 30, 24, or 49 for ae. The images ad appear to be top views and e, a side view. The scale bar is 5 nm. (C) Superimposed images of difference between tetra-TCP/GFP and tetra-TCP; typical top (C-a) and side (C-b) views, respectively. The background gray-scale images in C-a and C-b show the top and side views of averaged images of tetra-TCP, respectively: the same images as B-c and B-e. The blue contour lines represent the higher density region in the images of tetra-TCP/GFP as shown in A-d and A-f. Red or green contour lines indicate the statistical difference by t-test. The regions surrounded with red lines show that the density of tetra-TCP/GFP is significantly (0.1%, a, 10%, b, by t-test) higher than that of tetra-TCP, while those surrounded with green ones show that tetra-TCP/GFP is significantly (10% in both a and b by t-test) lower than that of tetra-TCP. The white and black arrowheads (C-b) represent the higher (red lines) and lower (green lines) density regions in tetra-TCP/GFP compared to tetra-TCP, respectively. The orange and red couple head arrows indicate the distances between the particle center and the higher, or lower density, region in the top and side views, respectively. Their distances are consistent between the top (C-a) and side (C-b) views.

 
Expression of virus structural proteins and antibody fragments
Using the chaperonin ring of tetra-TCP as a cage, we tried to express several guest proteins, which were reportedly difficult to express in the soluble fraction of E. coli cytoplasm. For the expression of virus structural proteins that had been reported to be toxic to E. coli, the genes of hepatitis B surface antigen (HBs, 25 kDa) and hepatitis C core antigen (HCc, 21 kDa) were cloned to the tetra-TCP expression vector. For the expression of aggregation-prone guest proteins, the genes of mouse anti-HEL (hen egg lysozyme) single-chain Fv antibody (anti-HEL scFv, 25 kDa) (Iba et al. 1997) and the human antibody heavy chain constant region comprising CH1, CH2, and CH3 domains (hCR, 36 kDa) were used.

All of the four tetra-TCP fusion proteins were highly expressed in the soluble fraction of E. coli cytoplasm (Fig. 7A-aGo). Only small part of each of the four tetra-TCP fusion proteins was produced in the insoluble fraction (data not shown). Densitometric analysis with Figure 7AGo revealed that the expression levels of these tetra-TCP/guest fusion proteins were estimated to be 67–94 mg/L-culture. Production of HBs, HCc, and hCR moieties were confirmed by immunoblotting (Fig. 7A-bGo). Electron microscopy revealed that all of four tetra-TCP fusion proteins assembled into double-ring complexes (Fig. 7BGo).



View larger version (105K):
[in this window]
[in a new window]
 
Figure 7. Expression and double-ring formation of tetra-TCP fused with virus structural proteins and antibody fragments. (A) Expression of fusion proteins. (A-a) SDS-PAGE analysis of soluble cell extracts of E. coli. Control cells with pET-3d (lane 1), E. coli cells expressing tetra-TCP, tetra-TCP/HBs, tetra-TCP/HCc, tetra-TCP/anti-HEL scFv, and tetra-TCP/hCR (lanes 26). (A-b) Western-blotting of the guest proteins. Tetra-TCP and tetra-TCP/HBs (lanes 1 and 2) reacted with anti-HBs polyclonal antibody. Tetra-TCP and tetra-TCP/HCc (lanes 3,4) reacted with anti-HCc monoclonal antibody (mAb). Tetra-TCP and tetra-TCP/hCR (lanes 5,6) reacted with anti-human antibody Fc region mAb. (B) Electron micrographs of tetra-TCP fusion proteins. (B-a) tetra-TCP/HBs. (B-b) tetra-TCP/ HCc. (B-c) tetra-TCP/anti-HEL scFv. (B-d) tetra-TCP/hCR. White arrowheads indicate side-on views of double rings. (C) Ligand-binding of the liberated antibody fragments. Digestion by PreScission protease and ligand-binding tests were conducted in PBS in the absence of Mg2+ and ATP. (C-a) SDS-PAGE analyses. Undigested tetra-TCP/anti-HEL scFv, PreScission protease-digested tetra-TCP/anti-HEL scFv, HEL-bound fraction and HEL-unbound fraction (lanes 14). Undigested tetra-TCP/hCR, PreScission protease-digested tetra-TCP/hCR, Protein G-bound fraction and Protein G-unbound fraction (lanes 58). Proteins were stained with Coomassie Brilliant Blue. Asterisks, PreScission protease. Solid and open arrowheads indicate anti-HEL scFv and hCR, respectively. (C-b) Western-blotting of hCR. Undigested tetra-TCP/hCR, PreScission protease-digested tetra-TCP/hCR and Protein G-bound fraction (lanes 13) detected with anti-human antibody Fc region mAb. Open arrowhead indicates hCR.

 
Ligand binding of the liberated antibody fragments
In order to liberate anti-HEL scFv or hCR from tetra-TCP ring, tetra-TCP fusion proteins obtained from a Ni-chelating chromatography were dialyzed against a protease digestion buffer containing 1 mM EDTA without Mg2+ and ATP to destabilize the quaternary structure. Incubation of these dialysates with PreScission protease in the same buffer led to the liberation of anti-HEL scFv or hCR moiety, respectively (Fig. 7C-aGo, lanes 2,6). Liberated anti-HEL scFv specifically bound to lysozyme-immobilized beads while PreScission protease existed only in the unbound fraction (Fig. 7C-aGo, lanes 3,4). Liberated hCR specifically bound to Protein G beads (Fig. 7C-aGo, lanes 7,8). Liberated hCR was also specifically detected by immunoblotting with antihuman antibody Fc region monoclonal antibody (Fig. 7C-bGo, lanes 2,3). These results indicate that most of expressed anti-HEL scFv and hCR were produced in active forms. However, some of undigested tetra-TCP fusion proteins seemed to bind to these affinity beads (Fig. 7C-aGo, lanes 3,7).


    Discussion
 TOP
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 References
 
Structure of a chaperonin during encapsulation of a substrate protein is not well understood. In order to approach to this subject, we made engineered chaperonin-guest fused complexes and analyzed their structural and functional features. The {alpha}-subunit chaperonin (TCP) from a hyperthermophilic archaeum, Thermococcus sp. KS-1, which forms an eightfold symmetric double-ring structure was used.

At first, we designed mono-, di-, tetra-, and octa- TCP (Figs. 1Go, 2Go) and expressed them in E. coli (Fig. 3AGo). Purified di- and tetra-TCP formed double-ring complexes, which were morphologically indistinguishable from those of wild-type TCP in the buffer containing MgCl2 without ATP (Fig. 3B,CGo). However, double-ring formation of mono-TCP required ATP in addition to Mg2+. This suggests that ATP stabilizes the assembly of TCP subunits. The double-ring complexes of di- and tetra-TCP were dissociated in the absence of Mg2+ and ATP (Fig. 3DGo). This suggests that Mg2+ also stabilizes the assembly of TCP subunits. Therefore, the guest protein moiety caged in the TCP cavity is probably exposed in the medium in the absence of Mg2+ and ATP, and can be liberated by the protease digestion of specific amino acid sequence in the linker (Fig. 2Go).

As shown in Figure 4Go, the optimum temperature of ATPase activity of tetra-TCP was 80°C while that of the wild-type is reported to be 60°C (Yoshida et al. 2002a; Fig. 4Go). The ATPase activity of tetra-TCP under the optimum condition was also extensively higher than that of the wild type (Fig. 4Go). This suggests that thermostability of tetra-TCP was significantly enhanced compared to those of the wild-type TCP. The wild-type TCP dissociates to ATPase-inactive subunit monomers above 80°C; the fold of TCP subunit is stabilized by formation of the double-ring form (Yoshida et al. 2002a; The covalent connection of the TCP subunit seemed to stabilize the double-ring structure and to confer higher thermostability and ATPase activity at high temperatures (>60°C).

The purified tetra-TCP/GFP fusion protein was shown to form the double-ring complex, in which GFP moiety fluoresced (Fig. 5A,BGo). This result suggested that the GFP moiety existed in the TCP cavity, and correctly folded. Image-analyses indicated that the GFP moiety was caged in the TCP cavity (Fig. 6A,C-aGo) and the structure of tetra-TCP/ GFP was similar to the closed-form (Ditzel et al. 1998; Shomura et al. 2004) crystallographically resolved in archaeal group II chaperonins (Fig. 6A-f,C-bGo). This is also supported by immunoprecipitation showing inaccessibility of anti-GFP antibodies to the GFP moiety of tetra-TCP/GFP (Fig. 5CGo). Furthermore, the subtraction analyses of the images between tetra-TCP and tetra-TCP/GFP suggested that caging GFP expanded the cavity from the top-view analysis (Fig. 6C-aGo), and that the equatorial portion seemed to be expanded around the bottom from the side-view analysis (Fig. 6C-bGo). This is the first report suggesting the structural change of the equatorial domain of a group II chaperonin induced with caging a guest protein in the cavity. The TCP-mediated GFP folding progresses in the closed-form in which the built-in lid regions of the apical domains cap the cavity (Yoshida et al. 2002b). The cavity inside of the closed-form of group II chaperonins is hydrophilic and suitable for protein folding (Horwich and Saibil 1998; Shomura et al. 2004). Folding of the GFP moiety of tetra-TCP/GFP seemed to be completed because the GFP moiety fluoresced (Fig. 5BGo). These suggest that the presented structure of tetra-TCP/GFP may be similar to the closed form of a wild-type group II chaperonin in which folding of a substrate is completed.

The lids of both rings of tetra-TCP also seem to be closed (Fig. 6B-e,C-bGo). Unexpectedly, tetra-TCP was not able to bind to the folding intermediates of GFP either at 60 or 37°C (data not shown), although the wild-type TCP is able to bind to the folding intermediates of GFP and release them in an ATP-dependent manner at 60°C (Yoshida et al. 2002b). These results suggest that even at 60°C the double-ring complex of tetra-TCP cannot take the open-form suitable for binding to a substrate polypeptide. Crystallography, 3D electron microscopy, and their combination would be required to clarify the more detailed structure of tetra-TCP and its fusion protein.

Using tetra-TCP as a fusion partner, two toxic virus structural proteins HBs and HCc (Sheu and Lo 1995; Seong et al. 1996), and two aggregation-prone antibody fragments anti-HEL scFv and hCR (Maynard and Georgiou 2000; Simmons et al. 2002) with the molecular sizes of 21–36 kDa, were well expressed in the soluble fraction of E. coli cytoplasm (Fig. 7AGo). All of the four tetra-TCP fusion proteins assembled into double-ring complexes (Fig. 7BGo). This suggests that each guest protein moiety could be expressed being caged in the TCP cavity and isolated from cytoplasmic environment. Expression of mono-TCP/HBs or di-TCP/ HBs was undetectable (data not shown). The trans-membrane domains of HBs are toxic to E. coli (Sheu and Lo 1995). In these fusion proteins, eight or four HBs moieties (total molecular weight is 200 kDa or 100 kDa, respectively) are probably too large for the cavity and might hinder the double-ring formation. These HBs moieties might be exposed in the host cytoplasm and exert their toxicity to host cells. The expressed antibody fragments with ligand-binding activities were liberated from tetra-TCP by digestion with PreScission protease in the absence of Mg2+ and ATP (Fig. 7CGo). These support the above view that caging those toxic or aggregation prone guest proteins in the chaperonin cavity improve the expression efficiency in E. coli.

For expression of recombinant proteins in host cells, it is important to prevent formation of inclusion body, proteolysis by host proteases, and possible toxicity to host cells. Because the target protein is always caged in the chaperonin cavity in host cells, this strategy (we call cCAP system; chaperonin-caged protein expression system) would be applicable to wide variety of proteins, including proteins toxic to host cells or trans-membrane proteins. The expressed target protein could be liberated from TCP by a site-specific proteolysis in the absence of Mg2+ and ATP under which assembly of the TCP subunits is unstable. The fusion product generated from cCAP system has a potential to be applied to several biotechniques, for example, to protein chips, immobilized enzyme technology for bioreactors, immunogens for antibody preparation, or three-dimensional structural analyses of target proteins using the crystallization protocols for a chaperonin.


    Materials and methods
 TOP
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 References
 
Guest genes
The anti-HEL (hen egg lysozyme) single-chain Fv (anti-HEL scFv) and the wild-type GFP genes were obtained from expression plasmids, pAALSC (Iba et al. 1997) and pIVEX2.3-GFP (Roche Diagnostics Co.), respectively. The custom-made synthetic genes of hepatitis B surface antigen S-region (HBs; GenBank accession number, AY247032 [GenBank] ) and hepatitis C core antigen (HCc; GenBank accession number, U10234 [GenBank] ) were purchased from Nippon Flour Mills Co., Ltd. (Kanagawa, Japan). A plasmid encoding the human antibody heavy-chain constant region (hCR) comprising CH1, CH2, and CH3 domains was kindly donated by Dr. Y. Akahori (Fujita Health University). The guest gene having SpeI and HpaI sites at its 5' and 3' ends, respectively, was made by PCR and was inserted into the SpeI–HpaI site of the expression vectors (Fig. 2Go).

Vector construction
The TCP gene having BglII and BamHI sites at its 5'- and 3'-termini, respectively, was designed and amplified by PCR from the genomic DNA of Thermococcus sp. KS-1 (JCM 11816) (Hoaki et al. 1994) using the forward primer 5'-TAAGATCTGTTGTTATT CTGCCTGAGGG-3' and the reverse primer 5'-ATGGATCCAC CGCCACCAGAACCGCCGGCCTTCGCAGCTATGACA-3'. The amplified TCP gene encodes the region of Val 9 to Ala 526 of TCP sequence followed by the additional C-terminal octapeptide (GGSGGGGS). The amplified TCP gene digested with BamHI and BglII, was self-ligated to make tandemly linked subunit genes. The resultant TCP subunit monomer (mono-TCP), linked dimer (di-TCP), linked tetramer (tetra-TCP), and linked octamer (octa-TCP) genes were respectively introduced into an expression vector pET3-d (Novagen) between T7 promoter and terminator-system (Fig. 2Go). All expression products had a 6His-tag at their C-termini and a PreScission protease (Amersham Biosciences) digestion site immediately after the covalently linked TCP subunits sequence.

Protein purification
The E. coli BL21 star (DE3) (Invitrogen) cell pellet from a 1-liter culture, in which the target protein expression was induced with 1 mM IPTG was sonicated in 100 mL of 50 mM Tris-HCl (pH 7.5) containing 0.2 mM EDTA and the protease inhibitor cocktail (No. 25954–21, nacalai tesque, Japan). After centrifugation, the supernatant was dialyzed against 1 mM imidazole/50 mM Tris-HCl (pH 7.5). The dialysate was applied to a Hi-Trap Ni-chelating sepha-rose column (5 mL) (Amersham Biosciences). The proteins were eluted using a linear gradient of 1 to 300 mM imidazole in 500 mM NaCl/50 mM Tris-HCl (pH 7.5). The TCP fraction was dialyzed against 5 mM MgCl2/0.5 mM ATP/50 mM Tris-HCl (pH 7.5), and then the dialysate was centrifuged at 100,000 x g for 1 h. The supernatant was applied to a RESOURCE Q column (6 mL) (Amersham Biosciences) and proteins were fractionated using a linear gradient of 0–500 mM NaCl in 5 mM MgCl2/50 mM Tris-HCl (pH 7.5). Protein samples were analyzed by 4–20% SDS-PAGE gel and visualized with Coomassie Brilliant Blue.

Gel filtration analyses
The mono-, di-, tetra-TCP, and tetra-TCP/GFP fusion proteins (2.5–3.0 mg/mL) purified with a RESOURCE Q column were applied to the gel filtration with a TSK gel G3000SWXL column (TOSOH, Japan) in 50 mM MgCl2/200 mM KCl/25 mM HEPES-KOH (pH 7.0). To confirm the double-ring formation of tetra-TCP/GFP fusion protein, elution was monitored with both OD280 nm and fluorescence at 510 nm with the excitation at 380 nm. To stabilize the double-ring structure of mono-TCP, it was mixed with ATP (final concentration: 1 mM) before the gel filtration. To investigate the structural stability of the di- and tetra-TCP doublering structure in the absence of Mg2+, they were dialyzed against 1 mM EDTA/25 mM HEPES-KOH (pH 7.0). After dialysis, the samples (3 mg/ml) were applied to a TSK gel G3000SWXL column in 1 mM EDTA/200 mM KCl/25 mM HEPES-KOH (pH 7.0).

ATPase assay
The ATPase assay mixture (100 µL) contained 2 mM ATP, 5 mM MgCl2, 300 mM KCl, and 25 µg of tetra-TCP in 50 mM HEPES-KOH (pH 6.8). The assay mixture was incubated at 37–90°C for 30 min. The reaction was terminated by adding 10% (w/v) perchloric acid. After centrifugation, released Pi in supernatant was measured using a 96-well microplate by malachite green method (Lanzetta et al. 1979) with a slight modification. The Pi assay mixture contained 5 µL of the terminated reaction mixture, 25 µL of HEPES-KOH (pH 6.8), 200 µL of color reagent (0.034% [w/v] malachite green hydrochloride, 1.05% [w/v] ammonium molybdate, and 0.1% Triton X-100), and 25 µL of 34% (w/v) sodium citrate dihydrate. The Pi assay mixture was incubated at room temperature for 30 min, and then absorbance at 650 nm was measured. Because nonenzymatic Pi release during the reactions was not negligible, ATPase activity was calculated by subtracting spontaneous Pi release.

Electron microscopy and image processing
After a RESOURCE Q column chromatography, the purified protein was separated with a TSK gel G3000 SWXL column in 50 mM MgCl2/200 mM KCl/25 mM HEPES-KOH (pH 7.0). The fraction corresponding to the chaperonin double ring was collected and then applied to a carbon-coated copper grid. Chaperonins immobilized on the grids were washed with 25 mM HEPES-KOH (pH 7.0), and negatively stained with 2.0% uranyl acetate. Images were recorded at a magnification of 40,000x with a transmission electron microscope JEOL JEM-1010 operated at 75 kV. Electron micrographs were digitized by a scanner (Leafscan45, Scitex Corporation) at a pixel size of 5 µm (0.125 nm/pixel). Images (301 tetra-TCP/GFP fusion protein and 292 tetra-TCP particles) were randomly selected as possible. After optimization of the crosscorrelation coefficient of the images by rotation with decreasing angular steps from 2.5 to 0.25 degree, the images were classified into five or six subsets by cluster analysis based on crosscorrelation coefficients. The rotational and translational parameters of the images in each of clusters were iteratively refined using the crosscorrelation method to be selected and averaged, respectively. Direction of each double-ring particles, top or side view, was assigned by comparing with the crystal structure of TCP (Shomura et al. 2004), and the representative images of top or side views of tetra-TCP/GFP fusion protein or tetra-TCP were selected, respectively. Their statistical difference in structure was examined by Student’s t-test. The image analysis was performed using Eos packages (Yasunaga and Wakabayashi 1996; Yokoyama et al. 2000).

Immunoprecipitation
Anti-GFP antibody serum (25 µL) (BD Biosciences Clontech) was incubated with 200 µL of Protein A/G beads (Amersham Biosciences) in PBS, and then anti-GFP-Protein A/G beads were collected by centrifugation with a spin filter (CytoSignal). The collected anti-GFP-Protein A/G beads were suspended in 200 µL of PBS, and then 50 µL of the suspension was incubated with a mixture containing, tetra-TCP/GFP fusion protein (30 µg) containing 3 µg of the GFP moiety, both tetra-TCP (27 µg) and GFP (3 µg), or only GFP (3 µg) in 200 µL of 25 mM MgCl2– PBS for 30 min at room temperature. After incubation, anti-GFP-Protein A/G-bound and -unbound fractions were collected by centrifugation with a spin filter and a collection tube (Cytosignal). The anti-GFP-Protein A/G-bound fraction was dissolved in 100 µL of SDS-PAGE sample buffer (125 mM Tris-HCl, 4.3% SDS, 30% glycerol, 10% {beta}ME, and 0.01% BPB, pH 6.8). Proteins in the unbound fractions were precipitated with ninefold volume of acetone, and then precipitate was dried up and dissolved with 100 µL of SDS-PAGE sample buffer. Each sample (7.5 µL) was separated by 4–20% SDS-PAGE gel and visualized with Coomassie Brilliant Blue.

Ligand-binding of the liberated antibody fragments
The fusion protein fraction obtained from a Ni-chelating sepharose column was dialyzed against 1 mM EDTA/1 mM DTT/150 mM NaCl/Tris-HCl (pH 7.0). The dialysate was digested with PreScission protease (Amersham Biosciences) in the same buffer at 4°C for 12 h, and then dialyzed against PBS (pH 7.4) containing 1 mM EDTA. The ligand-binding activities of the liberated anti-HEL scFv and hCR were assessed using HEL-immobilized sepharose beads and Protein G beads (Amersham Biosciences), respectively. The digested fusion protein was incubated with the affinity beads at room temperature for 30 min in PBS containing 1 mM EDTA, and then bound and unbound fractions were separated by centrifugation. The both fractions were analyzed by SDS-PAGE.


    Footnotes
 
7 Present address: Department of Cellular Biochemistry, Max-Planck-Institute für Biochemie, D-82152 Martinsried, Germany. Back


    Acknowledgments
 
We thank Drs. Y. Akahori and Y. Iba (Fujita Health University) for providing plasmids coding the human antibody heavy chain constant region and the mouse anti-HEL single chain Fv, respectively; Drs. M. Tanokura (University of Tokyo), K. Yamamoto (International Medical Center of Japan), and J. Chiba (Science University of Tokyo) for helpful advice and discussion; and Ms. R. Tatsumi for excellent technical assistance.


    References
 TOP
 Abstract
 Introduction
 Results
 Discussion
 Materials and methods
 References
 
Braig, K., Otwinowski, Z., Hegde, R., Boisvert, D.C., Joachimiak, A., Horwich, A.L., and Sigler, P.B. 1994. The crystal structure of the bacterial chaperonin GroEL at 2.8 Å. Nature 371: 578–586.[CrossRef][Medline]

Ditzel, L., Löwe, J., Stock, D., Otto-Karl, S., Huber, H., Huber, R., and Steinbacher, S. 1998. Crystal structure of the thermosome, the archaeal chaperonin and homolog of CCT. Cell 93: 125–138.[CrossRef][Medline]

Ellis, R.J. 2001. Molecular chaperones: Inside and outside the Anfinsen cage. Curr. Biol. 11: R1038–R1040.[CrossRef][Medline]

Esnouf, R.M. 1999. Further additions to MolScript version 1.4, including reading and contouring of electron-density maps. Acta. Crystallogr. D55: 938–940.

Farr, G.W., Furtak, K., Rowland, M.B., Ranson, N.A., Saibil, H.R., Kirchhausen, T., and Horwich, A.L. 2000. Multivalent binding of nonnative substrate proteins by the chaperonin GroEL. Cell 100: 561–573.[CrossRef][Medline]

Furutani, M., Iida, T., Yoshida, T., and Maruyama, T. 1998. Group II chaperonin in a thermophilic methanogen, Methanococcus thermolithotrophicus: Chaperone activity and filament-forming ability. J. Biol. Chem. 273: 28399–28407.[Abstract/Free Full Text]

Hartl, F.U. and Hayer-Hartl, M. 2002. Molecular chaperones in the cytosol: From nascent chain to folded protein. Science 295: 1852–1858.[Abstract/Free Full Text]

Hoaki, T., Nishijima, M., Kato, M., Adachi, K., Mizobuchi, S., Hanzawa, N., and Maruyama, T. 1994. Growth requirements of hyperthermophilic sulfur-dependent heterotrophic archaea isolated from a shallow submarine geothermal system with reference to their essential amino acids. Appl. Environ. Microbiol. 60: 2898–2904.[Abstract/Free Full Text]

Horwich, A.L. and Saibil, H.R. 1998. The thermosome: Chaperonin with a built-in lid. Nat. Struct. Biol. 5: 333–336.[CrossRef][Medline]

Iba, Y., Ito, W., and Kurosawa,Y. 1997. Expression vectors for the introduction of highly diverged sequences into the six complementarity-determining regions of an antibody. Gene 194: 35–46.[CrossRef][Medline]

Lanzetta, P.A., Alvarez, L.J., Reinach, P.S., and Candia, O.A. 1979. An improved assay for nanomole amounts of inorganic phosphate. Anal. Biochem. 100: 95–97.[CrossRef][Medline]

Matthews, B.W. 1968. Solvent content of protein crystals. J. Mol. Biol. 33: 491–497.[Medline]

Maynard, J. and Georgiou, G. 2000. Antibody engineering. Annu. Rev. Biomed. Eng. 2: 339–376.[CrossRef][Medline]

Ormo M., Cubitt, A.B., Kallio, K., Gross, L.A., Tsien, R.Y., and Remington, S.J. 1996. Crystal structure of the Aequorea victoria green fluorescent protein. Science 273: 1392–1395.[Abstract]

Seong, Y.R., Lee, E.K., Choi, S., Chon, S.K., and Im, D.S. 1996. Overexpression and simple purification of a truncated, immunologically reactive GST-HCV core (1–123) fusion protein. J. Viol. Methods 59: 13–21.

Sheu, S.Y. and Lo, S.J. 1995. Deletion or alteration of hydrophobic amino acids at the first and the third transmembrane domains of hepatitis B surface antigen enhances its production in Escherichia coli. Gene 160: 179–184.[CrossRef][Medline]

Shomura, Y., Yoshida, T., Iizuka, R., Maruyama, T., Yohda, M., and Miki, K. 2004. Crystal structures of the group II chaperonin from Thermococcus strain KS-1: Steric hindrance by the substituted amino acid, and inter-subunit rearrangement between two crystal forms. J. Mol. Biol. 335: 1265–1278.[CrossRef][Medline]

Simmons, L.C., Reilly, D., Klimowski, L., Raju, T.S., Meng, G., Sims, P., Hong, K., Shields, R.L., Damico, L.A., Rancatore, P., et al. 2002. Expression of full-length immunoglobulins in Escherichia coli: Rapid and efficient production of aglycosylated antibodies. J. Immunol. Methods 263: 133–147.[CrossRef][Medline]

Yasunaga, T. and Wakabayashi, T. 1996. Extensible and object-oriented system Eos supplies a new environment for image analysis of electron micrographs of macromolecules. J. Struct. Biol. 116: 155–160.[CrossRef][Medline]

Yokoyama, K., Ohkuma, S., Taguchi, H., Yasunaga, T., Wakabayashi, T., and Yoshida, M. 2000. V-Type H+-ATPase/synthase from a thermophilic eu-bacterium, Thermus thermophilus. Subunit structure and operon. J. Biol. Chem. 275: 13955–13961.[Abstract/Free Full Text]

Yoshida, T., Yohda, M., Iida, T., Maruyama, T., Taguchi, H., Yazaki, K., Ohta, T., Odaka, M., Endo, I., and Kagawa, Y. 1997. Structural and functional characterization of homo-oligomeric complexes of alpha and beta chaperonin subunits from the hyperthermophilic archaeum Thermococcus strain KS-1. J. Mol. Biol. 273: 635–645.[CrossRef][Medline]

Yoshida, T., Ideno, A., Suzuki, R., Yohda, M., and Maruyama, T. 2002a. Two kinds of archaeal group II chaperonin subunits with different thermostability in Thermococcus strain KS-1. Mol. Microbiol. 44: 761–769.[CrossRef][Medline]

Yoshida, T., Kawaguchi, R., Taguchi, H., Yoshida, M., Yasunaga, T., Wakabayashi, T., Yohda, M., and Maruyama, T. 2002b. Archaeal group II chaperonin mediates protein folding in the cis-cavity without a detachable GroES-like co-chaperonin. J. Mol. Biol. 315: 73–85.[CrossRef][Medline]


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Furutani, M.
Right arrow Articles by Maruyama, T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Furutani, M.
Right arrow Articles by Maruyama, T.
Social Bookmarking
 Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS