- Email: [email protected]
PDI family protein ERp29 recognizes P-domain of molecular chaperone calnexin
Biochemical and Biophysical Research Communications 487 (2017) 763e767
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
PDI family protein ERp29 recognizes P-domain of molecular chaperone calnexin Hitomi Nakao a, Akira Seko b, Yukishige Ito b, c, Masafumi Sakono a, * a
Department of Applied Chemistry, University of Toyama, 3190 Gofuku, Toyama, Toyama 930-855, Japan Japan Science and Technology Agency (JST), ERATO Ito Glycotrilogy Project, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan c Synthetic Cellular Chemistry Laboratory, RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 April 2017 Accepted 25 April 2017 Available online 26 April 2017
Endoplasmic reticulum (ER) resident lectin chaperone calnexin (CNX) and calreticulin (CRT) assist folding of nascent glycoproteins. Their association with ERp57, a member of PDI family proteins (PDIs) which promote disulfide bond formation of unfolded proteins, has been well documented. Recent studies have provided evidence that other PDIs may also interact with CNX and CRT. Accordingly, it seems possible that the ER provides a repertoire of CNX/CRT-PDI complexes, in order to facilitate refolding of various glycoproteins. In this study, we examined the ability of PDIs to interact with CNX. Among them ERp29 was shown to interact with CNX, similarly to ERp57. Judging from the dissociation constant, its ability to interact with CNX was similar to that of ERp57. Results of further analyses by using a CNX mutant imply that ERp29 and ERp57 recognize the same domain of CNX, whereas the mode of interaction with CNX might be somewhat different between them. © 2017 Elsevier Inc. All rights reserved.
Keywords: Protein disulfide isomerase ER resident molecular chaperone Protein-protein interaction
1. Introduction Nascent glycoproteins form higher-order structures by the assistance of endoplasmic reticulum (ER) resident folding machinery that comprises various enzymes and lectin chaperones [1]. As accumulation of incorrectly structured proteins causes unfolded protein response and trigger cellular stress [2], the role of glycoprotein quality control in the ER is important in minimizing the production of misfolded proteins [3]. ER resident molecular chaperones calnexin (CNX) and calreticulin (CRT) recognize immature glycoproteins having Glc1Man9GlcNAc2 glycoform and assist in their formation of correct structures [4,5]. Both of them are able to interact with ERp57, which is a member of protein disulfide isomerase (PDI) family proteins [6e8]. PDI family proteins (PDIs) preserve the PDI homologue sequence. Emblematic among them is protein disulfide isomerase (PDI), which conserves four types of thioredoxin sequence (a, a0 , b, b0 domain), behaves as an oxidoreductase to promote disulfide bond formation by virtue of its thioredoxin activity. All other PDIs are known to possess at least one type of thioredoxin sequences [9e11]. As a and a0 domains include redox active CXXC motif, PDIs
* Corresponding author. E-mail address: [email protected] (M. Sakono). http://dx.doi.org/10.1016/j.bbrc.2017.04.139 0006-291X/© 2017 Elsevier Inc. All rights reserved.
containing these domains are potentially able to catalyze redox reactions. On the other hand, b and b0 domains are regarded to be functional in peptide binding. The b0 domain of ERp57 in the a-b-b0 a0 quadruplet is engaged in complexation with CNX and CRT [12]. Although many of human PDIs have been reported to show oxidoreductase activity, it has been unclear why as many as twenty kinds of PDIs exist in the ER. Whilst ERp57 bound with chaperones is considered to function as co-chaperone to facilitate intramolecular disulfide bond formation of unfolded proteins [13], recent studies indicated that additional PDIs interact with CNX and CRT like [14e16]. As a matter of fact, we previously reported the interaction of ERp29 with CRT in vitro, which was found by using the BIACORE system [14]. As the interaction of ERp27, ERp44, ERp46 and PDIp with CRT was not detected, these PDIs were excluded from the members of co-chaperones for CRT. The dissociation constant of ERp29 with CRT calculated from BIACORE sensorgram was the almost identical to that of ERp57, suggesting that presence of the CRT-ERp29 complex in the ER might be of biological relevance. Analogously, it seemed probable that CNX also associates with ERp29. This issue warrants experimental verification, as thioredoxin-related transmembrane oxidoreductase which is one of PDIs is able to recognize CNX but not CRT [15], implying that CNX and CRT would interact with distinct set of PDIs. In this study, we explored PDIs ability to interact with CNX. Our
764
H. Nakao et al. / Biochemical and Biophysical Research Communications 487 (2017) 763e767
results indicate that ERp29 interacts with CNX similarly to CRT, and its affinity to CNX is almost identical to that of ERp57. A mutant of CNX which had diminished interaction with ERp57 exhibited reduced affinity to ERp29. This result implies that the ERp29 and ERp57 share the same region of CNX for binding. However, reduction of the affinity caused by the mutation was clearly less prominent for ERp29, suggesting that modes of interaction with CNX are subtly different between ERp29 and ERp57.
2. Materials and methods 2.1. Materials General regents were purchased from Wako Pure Chemical Industries, Ltd. Tris(hydroxymethyl)aminomethane was obtained from RIKAKEN. HEPES was purchased from Dojindo.
2.2. Preparation of recombinant Halo protein fusion calnexin A Halo tag protein encoding DNA was prepared by PCR employing pH6HTN His6 HaloTag T7 Vector (Promega) as the template using forward primer 50 TTTGAGCTCGCAGAAATCGGTACTGG-30 and reverse primer 50 e TTTAAGCTTGAATTCACTCCCTCCGCCACCACTCCCTCCGCCACCACTAGTGGTTGGCTCGCCG -30 . The reverse primer contained EcoRI site and DNA sequence encoding serine-glycine repeat as a linker for protein fusion. The amplified fragment was digested by SacI and HindIII, and then the purified fragment was ligated into pCold I vector (Takara Bio inc.) digested with SacI and HindIII. CNX (residues 205e1562) encoding DNA was amplified by PCR using forward primer 50 - TTTGAATTCGAGGCTCATGATGGACAT -30 and reverse primer 50 - TTTCTGCAGCTACGGGCGCTCTTCAGCT -30 , and re-cloned into above Halo sequence-inserted pCold I vector using EcoRI and PstI sites. Halo-fusion CNX were expressed in Escherichia coli BL21 and purified over Ni-NTA agarose (Qiagen) according to the manufacturer's instructions.
2.4. Preparation of human Halo-CNX M346A mutant The alignment program ClustalX was used to align sequences of human CNX, in comparison with canine CNX. All sequences in FASTA format were obtained from Genbank database. Their accession numbers were the following: canine CNX (NP_001003232.1) and human CNX (AAA21013.1). Met346 of CNX was mutated to alanine by PrimeSTAR Max DNA Polymerase (Takara Bio inc.) using forward primer 50 -GAAGACGCTGATGGAGAATGGGAGGCT-30 and reverse primer 50 - TCCATCAGCGTCTTCATCCCAATCCTC-30 . After PCR reaction, the mixture was treated with DpnI (Takara Bio inc.). Preparation of Halo-CNX M346A protein was same as above mention. 3. Results and discussion To find interactions between PDIs and CNX, pull-down assay using CNX-immobilized resin was performed. Immobilization of CNX onto the resin was carried out through binding reaction of Halo protein to Halo ligand [19]. Halo protein-fused CNX (Halo-CNX) was expressed in E. coli, and the prepared protein was purified by NiNTA column. The production of Halo-CNX was confirmed by SDSPAGE (Fig. 1), revealing a single band corresponding to the desired protein. Although the band migrated slightly slower than expected from the molecular weight (ca. 90 kDa), the expression and purification of Halo-CNX was considered to be carried out successfully, as this deviation was rationalized by the presence of a large proportion of acidic amino acids in CNX as described in our previous report [5]. The prepared Halo-CNX was immobilized on resins modified with Halo ligands. As a control, unmodified Halo protein was immobilized on resin in the same manner. Recombinant PDIs prepared as substrates for pull-down assay include ERp18, ERp27, ERp29, ERp44, ERp57, PDI and PDIp. At first, aptness of the system was verified by using ERp57 and PDI as positive and negative controls [6,14,20]. As shown in Fig. 2(A), the CNX-immobilized resin was able to bind ERp57 although the interaction of PDI was not detected, as anticipated. Subsequently, interaction of various PDIs (ERp18, ERp27, ERp29,
2.3. Pull-down assay Fifty mL of HaloLink Resin (Promega) was washed three times with the binding buffer (100 mM Tris-HCl buffer (pH 7.6) including 150 mM NaCl and 0.1% v/v TritonX-100). Thirty mM of Halo-CNX was added into the binding buffer suspended the resin after washing, and then the mixture was rotated for 1 h at room temperature. CNX-immobilized resin was washed five times with 500 mL of reaction buffer (30 mM HEPES (pH 7.5) including 0.1% v/v TritonX-100 and 1 mM CaCl2). The resin was mixed with 40 mM PDI family proteins (ERp18, ERp27, ERp29, ERp44, PDI, PDIp) in the reaction buffer and rotated for 1 h at room temperature. After incubation, unbound protein was removed by washing five times with reaction buffer. The resin was suspended into SDS-PAGE sample buffer and subjected to SDS-PAGE. After electrophoresis, separated proteins were detected by CBB staining. Pull-down assay of ERp57 was performed with reaction buffer containing 0.5 mM CaCl2. The intensities of the protein bands were analyzed with NIH ImageJ. The dissociation constants were estimated via a non-linear fitting of the following binding equation using Origin 2015 (Origin lab corp.) [17,18].
Y ¼ Bmax ½PDI family protien=ðKd þ ½PDI family proteinÞ Y and Bmax indicated complexation yield of CNX with PDI family proteins and the maximum binding yield, respectively.
Fig. 1. SDS-PAGE profiles of recombinant proteins. Halo means Halo protein.
H. Nakao et al. / Biochemical and Biophysical Research Communications 487 (2017) 763e767
Fig. 2. Confirmation of interaction between CNX and various PDIs ((A) ERp57 and PDI, (B) ERp18, ERp27, ERp29, ERp44 and PDIp) by pull-down assay. The CNX-immobilized resin was mixed with 40 mM PDIs in the reaction buffer and rotated for 1 h at room temperature. The resin was subjected to SDS-PAGE, and then separated proteins were detected by CBB staining after electrophoresis. Abbreviation ctrl means control sample whose assay was performed using Halo protein-immobilized resin.
ERp44 and PDIp) was analyzed as indicated in Fig. 2(B). Similarly to PDI, ERp18, ERp27, ERp44 and PDIp did not show specific interaction with CNX. By contrast, an obvious band was observed when ERp29 was employed as a target of pull-down assay, indicating that CNX, similarly to CRT, was able to interact with ERp29. This result is in line with a previous report which implied the possibility of CNXERp29 complexation by immunoprecipitation of extract of the thyroid cells [21]. To detail the interaction of ERp57 and ERp29 with CNX, calculation of dissociation constants between them were measured. Fig. 3(A) indicates complexation yields of CNX-ERp57 under various concentrations. A theoretical equation of 1:1 complex formation was well-fitted to the plots of complexation yield against ERp57 concentration. From the fitting curve, the dissociation constant of
(A)
765
CNX-ERp57 was calculated to be 4.4 mM. This value is in excellent agreement with previous report which employed ITC (Kd ¼ 5 mM) [12]. In the same manner, the dissociation constant of CNX-ERp29 was calculated. The theoretical curve showed good fitting to 1:1 binding equation, similar with ERp57 (Fig. 3(B)). The value was estimated to be 3.7 mM from the curve, suggesting that the strength of CNX-ERp29 interaction was highly similar to that of CNX-ERp57. Above results indicate that CNX bind more tightly to co-chaperons ERp57 and ERp29 than CRT [14]. Next, the interacting manner of CNX-ERp29 complex was investigated. Previously, ERp57 was reported to interact with CNX through the latter's tip regions of the P-domain [12,20,22]. Especially, Met347 located in this region seems to be important role, as mutant M347A of canine CNX was reported to exhibit reduced interaction with ERp57 [12]. With this in mind, a similar mutant of human CNX was prepared, to narrow down the location of CNX interacting with ERp29. From the alignment of amino acid sequences of both CNX, M347 of canine CNX was deemed to correspond to M346 of human CNX (Fig. 4(A)). The pull-down assay using prepared M346A CNX to ERp29 and ERp57 was performed. As shown in Fig. 4(B), the interaction of M346A mutant with ERp57 was decreased to about 10% compared to CNXwt in line with previous report [12]. The mutation also caused reduction of interaction with ERp29, suggesting that Met346 in the P-domain of CNX participated in interaction with both ERp57 and ERp29. For ERp29, however, effect of the M346A mutation was clearly more marginal, implying that the role of the Met346 on ERp29-CNX is less significant. ERp29 conserves only b domain among the thioredoxin like domains, and a and a0 domains which have CXXC motif for disulfide bond formation are lacking [9,23]. A b0 domain including recognition site of CNX is also not conserved in ERp29, suggesting that the interacting manner of ERp29 to CNX differs from that of ERp57. Intriguingly, we previously observed the similar difference between ERp29 and ERp57, when we analyzed their interactions with CRT [14]. From above results, we conclude that ERp29 is able to interact with both CNX and CRT. The dissociation constant between ERp29 and CNX was almost same as that of ERp57-CNX complex. The recognition site of CNX toward ERp29, similarly to ERp57, was revealed to locate at the P-domain. However, the mode of interaction to CNX might be different between ERp29 and ERp57. A few reports implied involvement of ERp29 in folding of proteins such as thyroid hormone and thyroglobulin [24]. Therefore, the influence of
(B) 100
CNX-ERp29 complexation yield [%]
CNX-ERp57 complexation yield [%]
100 80 60 40 20 0
0
20
40 60 80 ERp57 conc. [μM]
100
80 60 40 20 0
0
20
40 60 80 ERp29 conc. [μM]
100
Fig. 3. Complexation of CNX-immobilized resin with various concentrations of (A) ERp57 and (B) ERp29. The CNX-immobilized resin was mixed with various concentrations of PDIs in the reaction buffer and rotated for 1 h at room temperature. After staining, the intensities of the bands were analyzed with NIH ImageJ. The dissociation constants were estimated via a non-linear fitting of the 1:1 binding equation.
766
H. Nakao et al. / Biochemical and Biophysical Research Communications 487 (2017) 763e767
Fig. 4. Effect of CNX mutation on interaction with ERp57 and ERp29. (A) Alignment of amino acid sequence of canine CNX (334e364) and human CNX (333e363). Canine CNX €dinger). The sequences were aligned using ClustalX. Arrow indicates canine CNX Met347. (B) (C) Complexation yield structure (PDB code: 1JHN) was rendered in PyMOL (Schro between CNX mutant and PDIs. The human M347A CNX-immobilized resin was mixed with 20 mM PDIs ((B) ERp57, (C) ERp29) in the reaction buffer and rotated for 1 h at room temperature. After staining, the intensities of the bands were analyzed with NIH ImageJ. The error bars shown ±SE (n ¼ 3).
ERp29 on protein folding is promising new direction of research even though it is devoid of the CXXC motif. Further investigation is expected to reveal the role of ERp29, as well as that of other PDIs, in the CNX/CRT cycle. Conflicts of interest The authors have no conflicts of interest to declare. Acknowledgement This work was supported by Grant-in-Aid for Specially Promoted Research (16H06290) from Japan Society for the Promotion of Science (to Y.I.). References [1] J.D. Schrag, D.O. Procopio, M. Cygler, D.Y. Thomas, J.J.M. Bergeron, Lectin control of protein folding and sorting in the secretory pathway, Trends Biochem. Sci. 28 (2003) 49e57. [2] R.V. Rao, D.E. Bredesen, Misfolded proteins, endoplasmic reticulum stress and neurodegeneration, Curr. Opin. Cell Biol. 16 (2004) 653e662. [3] R. Sitia, I. Braakman, Quality control in the endoplasmic reticulum protein factory, Nature 426 (2003) 891e894. [4] F.E. Ware, A. Vassilakos, P.A. Peterson, M.R. Jackson, M.A. Lehrman, D.B. Williams, The molecular chaperone calnexin binds Glc1Man9GlcNAc2 oligosaccharide as an initial step in recognizing unfolded glycoproteins, J. Biol. Chem. 270 (1995) 4697e4704.
[5] M. Sakono, A. Seko, Y. Takeda, J.I. Aikawa, M. Hachisu, A. Koizumi, K. Fujikawa, Y. Ito, Glycan specificity of a testis-specific lectin chaperone calmegin and effects of hydrophobic interactions, Biochim. Biophys. Acta - Gen. Subj. 1840 (2014) 2904e2913. [6] J.D. Oliver, H. Llewelyn Roderick, D.H. Llewellyn, S. High, ERp57 functions as a subunit of specific complexes formed with the ER lectins calreticulin and calnexin, Mol. Biol. Cell. 10 (1999) 2573e2582. [7] L. Ellgaard, P. Bettendorff, D. Braun, T. Herrmann, F. Fiorito, I. Jelesarov, P. Güntert, A. Helenius, K. Wüthrich, NMR structures of 36 and 73-residue fragments of the calreticulin P-domain, J. Mol. Biol. 322 (2002) 773e784. [8] E.-M. Frickel, R. Riek, I. Jelesarov, A. Helenius, K. Wü, L. Ellgaard, TROSY-NMR reveals interaction between ERp57 and the tip of the calreticulin P-domain, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 1954e1959. €tta €nen, D.Y. Thomas, K. Gehring, A structural overview of the [9] G. Kozlov, P. M€ aa PDI family of proteins, FEBS J. 277 (2010) 3924e3936. [10] J.J. Galligan, D.R. Petersen, The human protein disulfide isomerase gene family, Hum. Genomics 6 (2012) 6. [11] L. Ellgaard, L.W. Ruddock, The human protein disulphide isomerase family: substrate interactions and functional properties, EMBO Rep. 6 (2005) 28e32. [12] G. Kozlov, P. Maattanen, J.D. Schrag, S. Pollock, M. Cygler, B. Nagar, D.Y. Thomas, K. Gehring, Crystal structure of the bb0 domains of the protein disulfide isomerase ERp57, Structure 14 (2006) 1331e1339. [13] C.E. Jessop, T.J. Tavender, R.H. Watkins, J.E. Chambers, N.J. Bulleid, Substrate specificity of the oxidoreductase ERp57 is determined primarily by its interaction with calnexin and calreticulin, J. Biol. Chem. 284 (2009) 2194e2202. [14] M. Sakono, A. Seko, Y. Takeda, Y. Ito, PDI family protein ERp29 forms 1:1 complex with lectin chaperone calreticulin, Biochem. Biophys. Res. Commun. 452 (2014) 27e31. [15] Y. Matsuo, H. Masutani, A. Son, S. Kizaka-Kondoh, J. Yodoi, Physical and functional interaction of transmembrane thioredoxin-related protein with major histocompatibility complex class I heavy chain: redox-based protein quality control and its potential relevance to immune responses, Mol. Biol. Cell. 20 (2009) 4552e4562.
H. Nakao et al. / Biochemical and Biophysical Research Communications 487 (2017) 763e767 €a €tta €nen, A. Rosenauer, F. Zheng, [16] G. Kozlov, S. Bastos-Aristizabal, P. Ma A. Killikelly, J.F. Trempe, D.Y. Thomas, K. Gehring, Structural basis of cyclophilin B binding by the calnexin/calreticulin P-domain, J. Biol. Chem. 285 (2010) 35551e35557. [17] Y. Lad, P. Jiang, S. Ruskamo, D.S. Harburger, J. Yl€ anne, I.D. Campbell, D.A. Calderwood, Structural basis of the migfilin-filamin interaction and competition with integrin b tails, J. Biol. Chem. 283 (2008) 35154e35163. €nne, Binding properties and stability of the Ras-association [18] H. Takala, J. Yla domain of Rap1-GTP interacting adapter molecule (RIAM), PLoS One 7 (2012) e31955. [19] C.G. England, H. Luo, W. Cai, HaloTag technology: a versatile platform for biomedical applications, Bioconjug. Chem. 26 (2015) 975e986. [20] S. Pollock, G. Kozlov, M.-F. Pelletier, J.-F. Trempe, G. Jansen, D. Sitnikov, J.J.M. Bergeron, K. Gehring, I. Ekiel, D.Y. Thomas, Specific interaction of ERp57 and calnexin determined by NMR spectroscopy and an ER two-hybrid system,
767
EMBO J. 23 (2004) 1020e1029. [21] S. Park, K.H. You, M. Shong, W.G. Tae, Y.Y. Eun, W.K. Seok, O.Y. Kwon, Overexpression of ERp29 in the thyrocytes of FRTL-5 cells, Mol. Biol. Rep. 32 (2005) 7e13. [22] M.R. Leach, M.F. Cohen-Doyle, D.Y. Thomas, D.B. Williams, Localization of the lectin, ERp57 binding, and polypeptide binding sites of calnexin and calreticulin, J. Biol. Chem. 277 (2002) 29686e29697. € ling, ERp28, a human [23] D.M. Ferrari, P. Nguyen Van, H.D. Kratzin, H.-D. So endoplasmic-reticulum-lumenal protein, is a member of the protein disulfide isomerase family but lacks a CXXC thioredoxin-box motif, Eur. J. Biochem. 255 (1998) 570e579. [24] E. Sargsyan, M. Baryshev, L. Szekely, A. Sharipo, S. Mkrtchian, Identification of ERp29, an endoplasmic reticulum lumenal protein, as a new member of the thyroglobulin folding complex, J. Biol. Chem. 277 (2002) 17009e17015.
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
PDI family protein ERp29 recognizes P-domain of molecular chaperone calnexin Hitomi Nakao a, Akira Seko b, Yukishige Ito b, c, Masafumi Sakono a, * a
Department of Applied Chemistry, University of Toyama, 3190 Gofuku, Toyama, Toyama 930-855, Japan Japan Science and Technology Agency (JST), ERATO Ito Glycotrilogy Project, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan c Synthetic Cellular Chemistry Laboratory, RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 April 2017 Accepted 25 April 2017 Available online 26 April 2017
Endoplasmic reticulum (ER) resident lectin chaperone calnexin (CNX) and calreticulin (CRT) assist folding of nascent glycoproteins. Their association with ERp57, a member of PDI family proteins (PDIs) which promote disulfide bond formation of unfolded proteins, has been well documented. Recent studies have provided evidence that other PDIs may also interact with CNX and CRT. Accordingly, it seems possible that the ER provides a repertoire of CNX/CRT-PDI complexes, in order to facilitate refolding of various glycoproteins. In this study, we examined the ability of PDIs to interact with CNX. Among them ERp29 was shown to interact with CNX, similarly to ERp57. Judging from the dissociation constant, its ability to interact with CNX was similar to that of ERp57. Results of further analyses by using a CNX mutant imply that ERp29 and ERp57 recognize the same domain of CNX, whereas the mode of interaction with CNX might be somewhat different between them. © 2017 Elsevier Inc. All rights reserved.
Keywords: Protein disulfide isomerase ER resident molecular chaperone Protein-protein interaction
1. Introduction Nascent glycoproteins form higher-order structures by the assistance of endoplasmic reticulum (ER) resident folding machinery that comprises various enzymes and lectin chaperones [1]. As accumulation of incorrectly structured proteins causes unfolded protein response and trigger cellular stress [2], the role of glycoprotein quality control in the ER is important in minimizing the production of misfolded proteins [3]. ER resident molecular chaperones calnexin (CNX) and calreticulin (CRT) recognize immature glycoproteins having Glc1Man9GlcNAc2 glycoform and assist in their formation of correct structures [4,5]. Both of them are able to interact with ERp57, which is a member of protein disulfide isomerase (PDI) family proteins [6e8]. PDI family proteins (PDIs) preserve the PDI homologue sequence. Emblematic among them is protein disulfide isomerase (PDI), which conserves four types of thioredoxin sequence (a, a0 , b, b0 domain), behaves as an oxidoreductase to promote disulfide bond formation by virtue of its thioredoxin activity. All other PDIs are known to possess at least one type of thioredoxin sequences [9e11]. As a and a0 domains include redox active CXXC motif, PDIs
* Corresponding author. E-mail address: [email protected] (M. Sakono). http://dx.doi.org/10.1016/j.bbrc.2017.04.139 0006-291X/© 2017 Elsevier Inc. All rights reserved.
containing these domains are potentially able to catalyze redox reactions. On the other hand, b and b0 domains are regarded to be functional in peptide binding. The b0 domain of ERp57 in the a-b-b0 a0 quadruplet is engaged in complexation with CNX and CRT [12]. Although many of human PDIs have been reported to show oxidoreductase activity, it has been unclear why as many as twenty kinds of PDIs exist in the ER. Whilst ERp57 bound with chaperones is considered to function as co-chaperone to facilitate intramolecular disulfide bond formation of unfolded proteins [13], recent studies indicated that additional PDIs interact with CNX and CRT like [14e16]. As a matter of fact, we previously reported the interaction of ERp29 with CRT in vitro, which was found by using the BIACORE system [14]. As the interaction of ERp27, ERp44, ERp46 and PDIp with CRT was not detected, these PDIs were excluded from the members of co-chaperones for CRT. The dissociation constant of ERp29 with CRT calculated from BIACORE sensorgram was the almost identical to that of ERp57, suggesting that presence of the CRT-ERp29 complex in the ER might be of biological relevance. Analogously, it seemed probable that CNX also associates with ERp29. This issue warrants experimental verification, as thioredoxin-related transmembrane oxidoreductase which is one of PDIs is able to recognize CNX but not CRT [15], implying that CNX and CRT would interact with distinct set of PDIs. In this study, we explored PDIs ability to interact with CNX. Our
764
H. Nakao et al. / Biochemical and Biophysical Research Communications 487 (2017) 763e767
results indicate that ERp29 interacts with CNX similarly to CRT, and its affinity to CNX is almost identical to that of ERp57. A mutant of CNX which had diminished interaction with ERp57 exhibited reduced affinity to ERp29. This result implies that the ERp29 and ERp57 share the same region of CNX for binding. However, reduction of the affinity caused by the mutation was clearly less prominent for ERp29, suggesting that modes of interaction with CNX are subtly different between ERp29 and ERp57.
2. Materials and methods 2.1. Materials General regents were purchased from Wako Pure Chemical Industries, Ltd. Tris(hydroxymethyl)aminomethane was obtained from RIKAKEN. HEPES was purchased from Dojindo.
2.2. Preparation of recombinant Halo protein fusion calnexin A Halo tag protein encoding DNA was prepared by PCR employing pH6HTN His6 HaloTag T7 Vector (Promega) as the template using forward primer 50 TTTGAGCTCGCAGAAATCGGTACTGG-30 and reverse primer 50 e TTTAAGCTTGAATTCACTCCCTCCGCCACCACTCCCTCCGCCACCACTAGTGGTTGGCTCGCCG -30 . The reverse primer contained EcoRI site and DNA sequence encoding serine-glycine repeat as a linker for protein fusion. The amplified fragment was digested by SacI and HindIII, and then the purified fragment was ligated into pCold I vector (Takara Bio inc.) digested with SacI and HindIII. CNX (residues 205e1562) encoding DNA was amplified by PCR using forward primer 50 - TTTGAATTCGAGGCTCATGATGGACAT -30 and reverse primer 50 - TTTCTGCAGCTACGGGCGCTCTTCAGCT -30 , and re-cloned into above Halo sequence-inserted pCold I vector using EcoRI and PstI sites. Halo-fusion CNX were expressed in Escherichia coli BL21 and purified over Ni-NTA agarose (Qiagen) according to the manufacturer's instructions.
2.4. Preparation of human Halo-CNX M346A mutant The alignment program ClustalX was used to align sequences of human CNX, in comparison with canine CNX. All sequences in FASTA format were obtained from Genbank database. Their accession numbers were the following: canine CNX (NP_001003232.1) and human CNX (AAA21013.1). Met346 of CNX was mutated to alanine by PrimeSTAR Max DNA Polymerase (Takara Bio inc.) using forward primer 50 -GAAGACGCTGATGGAGAATGGGAGGCT-30 and reverse primer 50 - TCCATCAGCGTCTTCATCCCAATCCTC-30 . After PCR reaction, the mixture was treated with DpnI (Takara Bio inc.). Preparation of Halo-CNX M346A protein was same as above mention. 3. Results and discussion To find interactions between PDIs and CNX, pull-down assay using CNX-immobilized resin was performed. Immobilization of CNX onto the resin was carried out through binding reaction of Halo protein to Halo ligand [19]. Halo protein-fused CNX (Halo-CNX) was expressed in E. coli, and the prepared protein was purified by NiNTA column. The production of Halo-CNX was confirmed by SDSPAGE (Fig. 1), revealing a single band corresponding to the desired protein. Although the band migrated slightly slower than expected from the molecular weight (ca. 90 kDa), the expression and purification of Halo-CNX was considered to be carried out successfully, as this deviation was rationalized by the presence of a large proportion of acidic amino acids in CNX as described in our previous report [5]. The prepared Halo-CNX was immobilized on resins modified with Halo ligands. As a control, unmodified Halo protein was immobilized on resin in the same manner. Recombinant PDIs prepared as substrates for pull-down assay include ERp18, ERp27, ERp29, ERp44, ERp57, PDI and PDIp. At first, aptness of the system was verified by using ERp57 and PDI as positive and negative controls [6,14,20]. As shown in Fig. 2(A), the CNX-immobilized resin was able to bind ERp57 although the interaction of PDI was not detected, as anticipated. Subsequently, interaction of various PDIs (ERp18, ERp27, ERp29,
2.3. Pull-down assay Fifty mL of HaloLink Resin (Promega) was washed three times with the binding buffer (100 mM Tris-HCl buffer (pH 7.6) including 150 mM NaCl and 0.1% v/v TritonX-100). Thirty mM of Halo-CNX was added into the binding buffer suspended the resin after washing, and then the mixture was rotated for 1 h at room temperature. CNX-immobilized resin was washed five times with 500 mL of reaction buffer (30 mM HEPES (pH 7.5) including 0.1% v/v TritonX-100 and 1 mM CaCl2). The resin was mixed with 40 mM PDI family proteins (ERp18, ERp27, ERp29, ERp44, PDI, PDIp) in the reaction buffer and rotated for 1 h at room temperature. After incubation, unbound protein was removed by washing five times with reaction buffer. The resin was suspended into SDS-PAGE sample buffer and subjected to SDS-PAGE. After electrophoresis, separated proteins were detected by CBB staining. Pull-down assay of ERp57 was performed with reaction buffer containing 0.5 mM CaCl2. The intensities of the protein bands were analyzed with NIH ImageJ. The dissociation constants were estimated via a non-linear fitting of the following binding equation using Origin 2015 (Origin lab corp.) [17,18].
Y ¼ Bmax ½PDI family protien=ðKd þ ½PDI family proteinÞ Y and Bmax indicated complexation yield of CNX with PDI family proteins and the maximum binding yield, respectively.
Fig. 1. SDS-PAGE profiles of recombinant proteins. Halo means Halo protein.
H. Nakao et al. / Biochemical and Biophysical Research Communications 487 (2017) 763e767
Fig. 2. Confirmation of interaction between CNX and various PDIs ((A) ERp57 and PDI, (B) ERp18, ERp27, ERp29, ERp44 and PDIp) by pull-down assay. The CNX-immobilized resin was mixed with 40 mM PDIs in the reaction buffer and rotated for 1 h at room temperature. The resin was subjected to SDS-PAGE, and then separated proteins were detected by CBB staining after electrophoresis. Abbreviation ctrl means control sample whose assay was performed using Halo protein-immobilized resin.
ERp44 and PDIp) was analyzed as indicated in Fig. 2(B). Similarly to PDI, ERp18, ERp27, ERp44 and PDIp did not show specific interaction with CNX. By contrast, an obvious band was observed when ERp29 was employed as a target of pull-down assay, indicating that CNX, similarly to CRT, was able to interact with ERp29. This result is in line with a previous report which implied the possibility of CNXERp29 complexation by immunoprecipitation of extract of the thyroid cells [21]. To detail the interaction of ERp57 and ERp29 with CNX, calculation of dissociation constants between them were measured. Fig. 3(A) indicates complexation yields of CNX-ERp57 under various concentrations. A theoretical equation of 1:1 complex formation was well-fitted to the plots of complexation yield against ERp57 concentration. From the fitting curve, the dissociation constant of
(A)
765
CNX-ERp57 was calculated to be 4.4 mM. This value is in excellent agreement with previous report which employed ITC (Kd ¼ 5 mM) [12]. In the same manner, the dissociation constant of CNX-ERp29 was calculated. The theoretical curve showed good fitting to 1:1 binding equation, similar with ERp57 (Fig. 3(B)). The value was estimated to be 3.7 mM from the curve, suggesting that the strength of CNX-ERp29 interaction was highly similar to that of CNX-ERp57. Above results indicate that CNX bind more tightly to co-chaperons ERp57 and ERp29 than CRT [14]. Next, the interacting manner of CNX-ERp29 complex was investigated. Previously, ERp57 was reported to interact with CNX through the latter's tip regions of the P-domain [12,20,22]. Especially, Met347 located in this region seems to be important role, as mutant M347A of canine CNX was reported to exhibit reduced interaction with ERp57 [12]. With this in mind, a similar mutant of human CNX was prepared, to narrow down the location of CNX interacting with ERp29. From the alignment of amino acid sequences of both CNX, M347 of canine CNX was deemed to correspond to M346 of human CNX (Fig. 4(A)). The pull-down assay using prepared M346A CNX to ERp29 and ERp57 was performed. As shown in Fig. 4(B), the interaction of M346A mutant with ERp57 was decreased to about 10% compared to CNXwt in line with previous report [12]. The mutation also caused reduction of interaction with ERp29, suggesting that Met346 in the P-domain of CNX participated in interaction with both ERp57 and ERp29. For ERp29, however, effect of the M346A mutation was clearly more marginal, implying that the role of the Met346 on ERp29-CNX is less significant. ERp29 conserves only b domain among the thioredoxin like domains, and a and a0 domains which have CXXC motif for disulfide bond formation are lacking [9,23]. A b0 domain including recognition site of CNX is also not conserved in ERp29, suggesting that the interacting manner of ERp29 to CNX differs from that of ERp57. Intriguingly, we previously observed the similar difference between ERp29 and ERp57, when we analyzed their interactions with CRT [14]. From above results, we conclude that ERp29 is able to interact with both CNX and CRT. The dissociation constant between ERp29 and CNX was almost same as that of ERp57-CNX complex. The recognition site of CNX toward ERp29, similarly to ERp57, was revealed to locate at the P-domain. However, the mode of interaction to CNX might be different between ERp29 and ERp57. A few reports implied involvement of ERp29 in folding of proteins such as thyroid hormone and thyroglobulin [24]. Therefore, the influence of
(B) 100
CNX-ERp29 complexation yield [%]
CNX-ERp57 complexation yield [%]
100 80 60 40 20 0
0
20
40 60 80 ERp57 conc. [μM]
100
80 60 40 20 0
0
20
40 60 80 ERp29 conc. [μM]
100
Fig. 3. Complexation of CNX-immobilized resin with various concentrations of (A) ERp57 and (B) ERp29. The CNX-immobilized resin was mixed with various concentrations of PDIs in the reaction buffer and rotated for 1 h at room temperature. After staining, the intensities of the bands were analyzed with NIH ImageJ. The dissociation constants were estimated via a non-linear fitting of the 1:1 binding equation.
766
H. Nakao et al. / Biochemical and Biophysical Research Communications 487 (2017) 763e767
Fig. 4. Effect of CNX mutation on interaction with ERp57 and ERp29. (A) Alignment of amino acid sequence of canine CNX (334e364) and human CNX (333e363). Canine CNX €dinger). The sequences were aligned using ClustalX. Arrow indicates canine CNX Met347. (B) (C) Complexation yield structure (PDB code: 1JHN) was rendered in PyMOL (Schro between CNX mutant and PDIs. The human M347A CNX-immobilized resin was mixed with 20 mM PDIs ((B) ERp57, (C) ERp29) in the reaction buffer and rotated for 1 h at room temperature. After staining, the intensities of the bands were analyzed with NIH ImageJ. The error bars shown ±SE (n ¼ 3).
ERp29 on protein folding is promising new direction of research even though it is devoid of the CXXC motif. Further investigation is expected to reveal the role of ERp29, as well as that of other PDIs, in the CNX/CRT cycle. Conflicts of interest The authors have no conflicts of interest to declare. Acknowledgement This work was supported by Grant-in-Aid for Specially Promoted Research (16H06290) from Japan Society for the Promotion of Science (to Y.I.). References [1] J.D. Schrag, D.O. Procopio, M. Cygler, D.Y. Thomas, J.J.M. Bergeron, Lectin control of protein folding and sorting in the secretory pathway, Trends Biochem. Sci. 28 (2003) 49e57. [2] R.V. Rao, D.E. Bredesen, Misfolded proteins, endoplasmic reticulum stress and neurodegeneration, Curr. Opin. Cell Biol. 16 (2004) 653e662. [3] R. Sitia, I. Braakman, Quality control in the endoplasmic reticulum protein factory, Nature 426 (2003) 891e894. [4] F.E. Ware, A. Vassilakos, P.A. Peterson, M.R. Jackson, M.A. Lehrman, D.B. Williams, The molecular chaperone calnexin binds Glc1Man9GlcNAc2 oligosaccharide as an initial step in recognizing unfolded glycoproteins, J. Biol. Chem. 270 (1995) 4697e4704.
[5] M. Sakono, A. Seko, Y. Takeda, J.I. Aikawa, M. Hachisu, A. Koizumi, K. Fujikawa, Y. Ito, Glycan specificity of a testis-specific lectin chaperone calmegin and effects of hydrophobic interactions, Biochim. Biophys. Acta - Gen. Subj. 1840 (2014) 2904e2913. [6] J.D. Oliver, H. Llewelyn Roderick, D.H. Llewellyn, S. High, ERp57 functions as a subunit of specific complexes formed with the ER lectins calreticulin and calnexin, Mol. Biol. Cell. 10 (1999) 2573e2582. [7] L. Ellgaard, P. Bettendorff, D. Braun, T. Herrmann, F. Fiorito, I. Jelesarov, P. Güntert, A. Helenius, K. Wüthrich, NMR structures of 36 and 73-residue fragments of the calreticulin P-domain, J. Mol. Biol. 322 (2002) 773e784. [8] E.-M. Frickel, R. Riek, I. Jelesarov, A. Helenius, K. Wü, L. Ellgaard, TROSY-NMR reveals interaction between ERp57 and the tip of the calreticulin P-domain, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 1954e1959. €tta €nen, D.Y. Thomas, K. Gehring, A structural overview of the [9] G. Kozlov, P. M€ aa PDI family of proteins, FEBS J. 277 (2010) 3924e3936. [10] J.J. Galligan, D.R. Petersen, The human protein disulfide isomerase gene family, Hum. Genomics 6 (2012) 6. [11] L. Ellgaard, L.W. Ruddock, The human protein disulphide isomerase family: substrate interactions and functional properties, EMBO Rep. 6 (2005) 28e32. [12] G. Kozlov, P. Maattanen, J.D. Schrag, S. Pollock, M. Cygler, B. Nagar, D.Y. Thomas, K. Gehring, Crystal structure of the bb0 domains of the protein disulfide isomerase ERp57, Structure 14 (2006) 1331e1339. [13] C.E. Jessop, T.J. Tavender, R.H. Watkins, J.E. Chambers, N.J. Bulleid, Substrate specificity of the oxidoreductase ERp57 is determined primarily by its interaction with calnexin and calreticulin, J. Biol. Chem. 284 (2009) 2194e2202. [14] M. Sakono, A. Seko, Y. Takeda, Y. Ito, PDI family protein ERp29 forms 1:1 complex with lectin chaperone calreticulin, Biochem. Biophys. Res. Commun. 452 (2014) 27e31. [15] Y. Matsuo, H. Masutani, A. Son, S. Kizaka-Kondoh, J. Yodoi, Physical and functional interaction of transmembrane thioredoxin-related protein with major histocompatibility complex class I heavy chain: redox-based protein quality control and its potential relevance to immune responses, Mol. Biol. Cell. 20 (2009) 4552e4562.
H. Nakao et al. / Biochemical and Biophysical Research Communications 487 (2017) 763e767 €a €tta €nen, A. Rosenauer, F. Zheng, [16] G. Kozlov, S. Bastos-Aristizabal, P. Ma A. Killikelly, J.F. Trempe, D.Y. Thomas, K. Gehring, Structural basis of cyclophilin B binding by the calnexin/calreticulin P-domain, J. Biol. Chem. 285 (2010) 35551e35557. [17] Y. Lad, P. Jiang, S. Ruskamo, D.S. Harburger, J. Yl€ anne, I.D. Campbell, D.A. Calderwood, Structural basis of the migfilin-filamin interaction and competition with integrin b tails, J. Biol. Chem. 283 (2008) 35154e35163. €nne, Binding properties and stability of the Ras-association [18] H. Takala, J. Yla domain of Rap1-GTP interacting adapter molecule (RIAM), PLoS One 7 (2012) e31955. [19] C.G. England, H. Luo, W. Cai, HaloTag technology: a versatile platform for biomedical applications, Bioconjug. Chem. 26 (2015) 975e986. [20] S. Pollock, G. Kozlov, M.-F. Pelletier, J.-F. Trempe, G. Jansen, D. Sitnikov, J.J.M. Bergeron, K. Gehring, I. Ekiel, D.Y. Thomas, Specific interaction of ERp57 and calnexin determined by NMR spectroscopy and an ER two-hybrid system,
767
EMBO J. 23 (2004) 1020e1029. [21] S. Park, K.H. You, M. Shong, W.G. Tae, Y.Y. Eun, W.K. Seok, O.Y. Kwon, Overexpression of ERp29 in the thyrocytes of FRTL-5 cells, Mol. Biol. Rep. 32 (2005) 7e13. [22] M.R. Leach, M.F. Cohen-Doyle, D.Y. Thomas, D.B. Williams, Localization of the lectin, ERp57 binding, and polypeptide binding sites of calnexin and calreticulin, J. Biol. Chem. 277 (2002) 29686e29697. € ling, ERp28, a human [23] D.M. Ferrari, P. Nguyen Van, H.D. Kratzin, H.-D. So endoplasmic-reticulum-lumenal protein, is a member of the protein disulfide isomerase family but lacks a CXXC thioredoxin-box motif, Eur. J. Biochem. 255 (1998) 570e579. [24] E. Sargsyan, M. Baryshev, L. Szekely, A. Sharipo, S. Mkrtchian, Identification of ERp29, an endoplasmic reticulum lumenal protein, as a new member of the thyroglobulin folding complex, J. Biol. Chem. 277 (2002) 17009e17015.