- Email: [email protected]
D exchange mass spectrometry and isothermal titration calorimetry
Biochimica et Biophysica Acta 1854 (2015) 39–45
Contents lists available at ScienceDirect
Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbapap
Interaction of arginine with protein during refolding process probed by amide H/D exchange mass spectrometry and isothermal titration calorimetry Dawei Zhao a, Yongdong Liu a,⁎, Guifeng Zhang a, Chun Zhang b, Xiunan Li a, Qingqing Wang a, Hong Shi a, Zhiguo Su a,⁎⁎ a b
National Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China Key Laboratory of Drug Targeting and Drug Delivery System, Ministry of Education, West China School of Pharmacy, Sichuan University, Chengdu 610041, China
a r t i c l e
i n f o
Article history: Received 13 May 2014 Received in revised form 8 October 2014 Accepted 9 October 2014 Available online 16 October 2014 Keywords: HDX-MS ITC Refolding Arginine rhG-CSF
a b s t r a c t Arginine has been widely used as low molecular weight additive to promote protein refolding by suppressing aggregate formation. However, methods to investigate the role of arginine in protein refolding are often limited on protein's global conformational properties. Here, hydrogen/deuterium exchange mass spectrometry (HDX-MS) was used to study the effects of arginine on recombinant human granulocyte colony-stimulating factor (rhG-CSF) refolding at the scale of peptide mapping. It was found that deuteration levels of rhG-CSF refolded with arginine was higher than that without arginine during the whole refolding process, but they became almost the same when the refolding reached equilibrium. This phenomenon indicated that arginine could protect some amide deuterium atoms from being exchanged with hydrogen, but the protection diminished gradually along with refolding proceeding. Enzymatic digestion revealed six particular peptides of 16–47, 72–84, 84–93, 114–124, 145–153 and 154–162 were mainly responsible for the deuteration, and all of them dominantly located in protein's α-helix domain. Furthermore, thermodynamics analysis by isothermal titration calorimetry provided direct evidence that arginine could only react with denatured and partially refolded rhG-CSF. Taking all of the results together, we suggest that arginine suppresses protein aggregation by a reversible combination. At the initial refolding stage, arginine could combine with the denatured protein mainly through hydrogen bonding. Subsequently, arginine is gradually excluded from protein with protein's native conformation recovering. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Protein aggregation brings great troubles to the production and storage of protein based therapeutics [1]. Overexpressed heterologous recombinant proteins in prokaryotic and eukaryotic cells often form inactive and insoluble protein particles, which must be re-solubilized to recover their native conformation in vitro by refolding. Always low refolding efficiency is the bottlenecks for large production of such protein
Abbreviations: HDX, hydrogen deuterium exchange; ESI-MS, electrospray ionization mass spectrometry; ITC, isothermal titration calorimetry; rhG-CSF, recombinant human granulocyte colony-stimulating factor; IBs, inclusion bodies; β-ME, 2-mercaptoethanol; TFA, trifluoroacetic acid; Tris, Tris (hydroxymethyl) aminomethane; HPLC, high performance liquid chromatography ⁎ Correspondence to: Y. Liu National Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, No. 1 Bei-Er-Tiao Street, Haidian District, Beijing 100190, China. Tel.: +86 10 82545028; fax: +86 10 62561813. ⁎⁎ Correspondence to: Z. Su, National Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, No. 1 Bei-Er-Tiao Street, Haidian District, Beijing 100190, China. Tel.: +86 10 62561817; fax: +86 10 62561813. E-mail addresses: [email protected] (Y. Liu), [email protected] (Z. Su).
http://dx.doi.org/10.1016/j.bbapap.2014.10.007 1570-9639/© 2014 Elsevier B.V. All rights reserved.
therapeutics. Aggregation is the main instable factor during protein storage. Protein aggregate will not only decrease the bioactivity but also cause some degree of immunogenicity. To solve these problems, a variety of low molecular weight additives, such as sugars, polyols, detergents, salts, etc., have been used to improve protein refolding efficiency or increase protein solubility and long-term stability [2–5]. Among these additives, arginine is most widely used to prevent protein aggregation. Arginine was first used in the refolding of human tissue type plasminogen activator [6]. Subsequently, it was successfully applied to the refolding of casein II [7], Fab antibody fragments [8,9], interleukin-6 receptor [10], lysozyme [11], human p53 tumor suppressor protein [12], among others. To elucidate the mechanism by which arginine suppresses protein aggregation, various theories have been proposed. Arakawa et al. [13] compared the solubility of amino acids in aqueous arginine solution and suggested that interactions of arginine with aromatic amino acid side chains on protein surface contributed to its aggregation inhibition. Ito et al. [14] investigated the interaction between arginine and lysozyme with high resolution X-ray analysis and also suggested that shielding aromatic residues inhibited hydrophobic interactions on the protein surface. Recently, Vagenende et al. [15] carried out molecular
40
D. Zhao et al. / Biochimica et Biophysica Acta 1854 (2015) 39–45
dynamics simulation to probe the interactions of arginine with protein and reported that cluster formation of arginine cations at the specific protein surface loci might affect the apparent protein radius and even the apparent charge and hydrophobic properties of protein surface. Combined with dialysis/densimetry and vapor pressure osmometry techniques, their research group suggested that preferential interaction and attractive interactions between arginine molecules might also played a role in arginine behavior [16,17]. However, protein refolding is a complex procedure, limited experimental data for arginine action lead to the detailed pathway of arginine-assisted protein refolding remains ambiguous. Traditional methods for probing the effects of arginine on protein refolding, such as chromatographic analysis [18], fluorescence spectra analysis [19], and dynamic laser light scattering technique [20], only focused on the global conformational properties and couldn't provide detailed and direct information. Although X-ray technology could give an atomic resolution, it just provides static information. An alternative strategy, the use of hydrogen/deuterium exchange (HDX) has become attractive because it could characterize protein dynamics and conformation [21]. Hydrogen atoms located at the peptide backbone amide linkage (NHs) are most easily monitored for their relatively slow exchange rate with deuterium atoms in solvent. HDX correlates with two intertwined factors: accessibility and hydrogen bonding strength of amide hydrogen atoms in the protein structure. In general, exchange rate of NHs in folded proteins is many orders of magnitude slower than those in an unfolded version of the same sequence. Therefore, differences in the rate of HDX of NHs reflect alterations in protein conformation, structure, and intramolecular and intermolecular associations. Characterization of HDX behavior can be made easily by combining with mass spectrometry (MS) because the mass of protein will change one Da for each hydrogen/ deuterium exchange. In addition, the amounts of protein required by HDX-MS are orders of magnitude smaller than those of other biophysical methods for studying protein conformation and dynamics. To increase the spatial resolution of HDX, digestion with an acidic enzyme has been incorporated following exchange reaction quenching [22]. Thus, combination of HDX with liquid chromatographic separation of digested peptides and MS detection will provide detailed information regarding protein conformational mobility at the scale of peptide mapping. In addition to HDX-MS, isothermal titration calorimetry (ITC) is another powerful tool to probe biomolecular interactions. It has been used to study the thermodynamics of protein refolding assisted by protein disulfide isomerase [23], interaction between bovine heart cytochrome c and small molecule [24], and even protein–DNA/RNA interactions [25,26]. Although ITC also focus on proteins' global properties, it could provide direct thermodynamic information which was complementary to those obtained by other measurements [27]. For confirming the findings from HDXMS analysis, ITC was applied to investigate the effects of arginine on protein refolding as well. Escherichia coli-expressed recombinant human granulocyte colonystimulating factor (rhG-CSF) was selected as a model protein to investigate the effect of arginine on its refolding. The benefit of arginine during rhG-CSF refolding has been reported [18,19], but there is insufficient evidence revealing how arginine exerts its influence. Here, we obtained some detailed evidences to elucidate the mechanism of arginine-assisted refolding of rhG-CSF by HDX-MS. Peptide mapping was performed as well to compare the deuterium uptake at specific regions of rhG-CSF during the refolding with and without arginine assistance. Furthermore, the results of ITC measurements also provided supporting evidence for the hypothesis deduced from HDX-MS experiments. 2. Materials and methods 2.1. Materials Deuterium oxide (D2O) was purchased from Cambridge Isotope Laboratories (Andover, MA, USA). Ultrapure urea-d4, DCl and pepsin
were obtained from Sigma Chemical Co. (St. Louis, MO, USA). 2Mercaptoethanol (β-ME) was purchased from Fluka Chemical Corps (Buchs, Switzerland). All other chemicals were analytical grades and obtained from Shanghai Reagent Company (Shanghai, China). Water used for the experiments was ultrapure water from Milli-Q (Millipore, MA, USA). Deuterated L-arginine (D-Arg) was prepared by dissolving L-arginine in D2O, freezing and lyophilizing the solution, and repeating the process twice more. 2.2. Protein preparation An E. coli strain (DH5α) transformed with an expression plasmid pBV220 was used as the source of rhG-CSF. Inclusion bodies of rhGCSF were expressed, isolated and purified as previously described [19]. After reduction and denaturation, solubilized rhG-CSF was diluted into the refolding buffer (2 M urea, 20 mM Tris–HCl, 0.5 M Arg, pH 8.5) to recover its correct conformation. Then, the refolding solution was purified by one-step cation exchange chromatography with SP Sepharose™ Fast Flow (GE healthcare, Uppsala, Sweeden) and rhG-CSF with a purity of more than 99% upon SDS-PAGE analysis was obtained. Purified rhGCSF was lyophilized and stored at −80 °C for subsequent experiments. 2.3. Hydrogen/deuterium exchange Lyophilized rhG-CSF was dissolved with denaturing buffer containing 8 M urea-d4, 1% (vol/vol) β-ME, 20 mM Tris–DCl, pH 8.9 (pD 8.5) and incubated for 6 h at room temperature to completely denaturate rhG-CSF and fully deuterate the amide and other labile hydrogen sites simultaneously. Refolding was initiated by a rapid 100-fold dilution of 10 mg/mL denatured and deuterated rhG-CSF solution into the refolding buffer containing 2 M urea-d4, 20 mM Tris–DCl, pH 8.9 (pD 8.5) with or without 0.5 M D-Arg. Exchange-out reaction was conducted by diluting refolded samples 10-fold with back exchange refolding buffer of 2 M urea, 20 mM Tris–HCl, pH 8.5 and incubating for 10 s at 4 °C after refolding was initiated for 10, 30, 60, 180, 300, 420 and 540 min. Labeled samples were quenched by adding DCl to bring the pD to 2.3 and lowering the temperature to 0 °C by incubating in an ice bath, then subjected to liquid chromatography–mass spectrometry (LC–MS) analysis. The amide back exchange level during LC–MS analysis was less than 10%, as determined in control experiments on denatured and fully labeled rhG-CSF. The deuteration level was calculated using the following equation: deuteration level ¼
exchanged amide hydrogen atoms 100% exchangeable amide hydrogen atoms
where the exchanged amide hydrogen atoms equal to the increase of the molecular weight of deuterated protein and the exchangeable amide hydrogen atoms equal to the total number of protein's amino acids, minus proline residues minus 1 for N-terminal amide. For rhGCSF, the number of total amide hydrogen is 175 and the exchangeable amide hydrogen atoms are 163. 2.4. Peptic digestion, liquid chromatography and mass spectrometry Following exchange reaction quenching, refolded and deuterated samples were digested with pepsin and subjected to high performance liquid chromatography (HPLC)/electrospray ionization tandem mass spectrometry (ESI-MS/MS) using the Agilent 1100 HPLC system (Agilent Technologies, Palo Alto, CA, USA) with a reversed phase HPLC column (Shiseido Proteonavi-C4, 250 × 4.6 mm I.D., 5 μm, Japan) and the LCQ DECA XP ion-trap mass spectrometry system (Thermo Finnigan, San Jose, CA, USA). Pepsin digestion was performed in 0.1% DCl for 5 min on ice with 1:4 molar ratio of rhG-CSF to pepsin. The conditions for peptide separation were as follows: 15-min linear gradient, 20%–37% acetonitrile containing 0.1% trifluoroacetic acid (TFA), flow rate of
D. Zhao et al. / Biochimica et Biophysica Acta 1854 (2015) 39–45
41
1 mL/min. Buffers and the column used were cooled to 0 °C in an ice bath. ESI-MS/MS conditions were set as: 4.5 kV spray voltage, 300 °C for the heat capillary and m/z from 400 to 900 for the MS scan range. The assignment of the peptides produced by pepsin to specific partial sequences of rhG-CSF was implemented as follow: firstly, rhG-CSF was divided into the shortest segments at all possible sites by pepsin. Then, some successive segments were reconstructed into larger peptides to match the data of MS analysis. Finally, the peptides were calculated with Bioworks Browser 3.1 (Thermo Finnigan, San Jose, CA) and compared the results with those obtained from MS/MS analysis. The confidence score of N 95 for peptides identification was guaranteed by this procedure. As the comparisons of the deuteration levels of the peptides during refolding were relative experiments, no control experiment was performed for deuterium back exchange. The results presented for peptides were just relative deuteration levels as others previously reported [28]. In this study, each experiment for MS analysis of peptic peptides was conducted in three replicates. 2.5. Isothermal titration calorimetry (ITC) experiments Calorimetric measurements were performed at 25 °C with an iTC200 system (MicroCal, GE Healthcare, Piscataway, NJ, USA) interfaced with a personal computer. All solutions were thoroughly degassed before using by stirring under vacuum. Lyophilized rhG-CSF and L-arginine were dissolved with refolding buffer (2 M urea containing 20 mM Tris–HCl, pH 8.5) or denaturing buffer (8 M urea containing 20 mM Tris–HCl and 1% (vol/vol) β-ME, pH 8.5) to investigate the interaction between rhG-CSF and arginine at folded and denatured status, respectively. In a typical binding experiment on the iTC200 system, 10 mM arginine was injected in 19 aliquots of 2 μL (except the first injection, which was 0.4 μL) into a 0.2004 mL sample cell containing 20 μM rhG-CSF. Calorimetric data were analyzed using Origin for ITC version 7.0383. 3. Results 3.1. Differences in HDX behavior of rhG-CSF between refolding with and without arginine assistance Deuteration levels of exchangeable amide hydrogen atoms in rhGCSF were measured at different time points after initiation of refolding to verify the effects caused by arginine on the refolding process (Fig. 1). Deuteration levels reached the same margin of 65% at the end of rhGCSF refolding process whether arginine was present or not, but the changing profile of deuteration levels differed significantly between the two refolding conditions. In the presence of arginine, deuteration levels increased sharply after rhG-CSF refolding was initiated and reached a plateau when refolding had proceeded for 180 minutes. In contrast, deuteration levels increased more slowly and took 420 minutes to achieve deuteration equilibrium in the absence of arginine. At the same refolding time, the higher deuteration levels of refolded rhG-CSF in the presence of arginine indicated slower HDX rates than those in the absence of arginine. However, differences in deuteration levels between the two refolding conditions diminished as refolding proceeded. The results revealed that arginine put some effects on rhG-CSF refolding, but the influence weakened with refolding proceeding and disappeared after rhG-CSF retrieved the native conformation.
Fig. 1. Kinetic analysis of deuteration level variation over rhG-CSF refolding time in the presence and absence of arginine. Deuteration levels are shown for refolding times between 10 min and 660 min. Square = refolding in the absence of arginine; circle = refolding in the presence of 0.5 M arginine.
and 20 minutes. Fig. 3 shows the HDX map and secondary structure information of the completely refolded rhG-CSF. Different color blocks represent the peptides with different deuteration levels (the corresponding amino acid sequence is directly above the block). The deuteration level of each peptide was calculated by the same equation for the entire molecule which was presented in Materials and Methods. Only 11 of 175 residues were not detected, providing nearly 94% sequence coverage. It could be seen that deuteration was distributed at regions of the whole rhG-CSF molecule but noticeable differences in deuteration level existed among these peptic peptides, indicating various protection of amide deuterium atoms associated with structure-related factors, such as hydrogen boding within α-helices. Six proteolytic fragments spanning 16–47, 72–84, 84–93, 114–124, 145–153 and 154–162, which mainly locate in α-helix, were identified to possess more than 40% deuteration. Previous studies have suggested that hydrogen bonding might be related with the interaction between arginine and protein. Then the changing trend of deuterium level of the peptic peptides, especially of the six peptides mentioned above, was thus analyzed during rhG-CSF refolding with and without arginine assistance.
3.2. Deuteration levels of peptic peptides generated from completely refolded rhG-CSF To identify the deuteration regions, rhG-CSF was digested using pepsin after refolding and exchange-out reaction followed by HPLC–MS analysis. Fig. 2 shows a representative chromatogram obtained by pepsin digestion and RP-HPLC separation (peaks are labeled with their corresponding peptides). All peptides analyzed were eluted between 5
Fig. 2. Representative RP-HPLC chromatogram of peptides from completely refolded rhGCSF produced by pepsin digestion. Refolded rhG-CSF was digested with 1:4 molar ratio of rhG-CSF to pepsin. Peaks that correspond to the peptic peptides are labeled.
42
D. Zhao et al. / Biochimica et Biophysica Acta 1854 (2015) 39–45
Fig. 3. HDX map for completely refolded rhG-CSF. Denatured rhG-CSF refolded completely in 2 M urea-d4 containing 20 mM Tris–DCl, pD 8.5, then an exchange-out reaction was performed in 2 M urea containing 20 mM Tris–HCl, pH 8.5. Subsequently, refolded and deuterated samples were digested with pepsin and subjected to LC–MS analysis. Each block represents a pepsin-generated fragment of rhG-CSF. The deuteration level for each peptide is shown as a percentage (color-coded) of the theoretical maximum. Secondary structural information is also illustrated in the map.
3.3. Time dependence of deuterium exchange out of specific peptides generated from rhG-CSF
and arginine, but a combination occurred when denatured rhG-CSF and arginine coexisted in 8 M urea.
Deuterium exchange out analysis for the six peptides identified above was carried out during rhG-CSF refolding process in the presence or absence of arginine. At different refolding stage, samples after exchange-out and pepsin digestion were subjected to LC–MS analysis. As shown in Fig. 4, the deuteration profiles of the six selected peptides during the refolding process were similar to that of the entire rhG-CSF molecule. That is, higher deuteration levels with arginine addition could be observed at the same refolding point and finally reached the same levels with or without arginine. However, for those peptides with higher final deuteration levels such as peptide of 72–84 and 114–124, difference in deuteration level between refolding with and without arginine addition was more obvious before reaching the plateau. All the data indicated that these particular peptide regions were responsible for the influence of arginine on rhG-CSF refolding. The gradually vanished discrepancy in deuteration levels between the two refolding conditions indicated that the interaction between arginine and these peptides gradually weakened along with refolding proceeding. The same final deuteration levels suggested that there should be no interaction between arginine and the peptides after the protein recovered its correct structure.
4. Discussion
3.4. Interaction of arginine with rhG-CSF investigated by ITC To further investigate the effect of arginine on rhG-CSF refolding, ITC experiments were performed with native and denatured rhG-CSF, the two extreme states of the refolding process. In order to eliminate possible disturbance from arginine interacting with the components in the refolding or denaturing buffers, control experiments were carried out by titrating arginine into these two buffers with no protein (Fig. 5A and B, upper trace). Fig. 5A showed that heat uptake by the titration of arginine into the blank refolding buffer was obvious. Titration of arginine into the refolding buffer containing rhG-CSF gave a similar heat uptake, and almost no enthalpy for each injection was observed after offset calculation (Fig. 5C). This result suggested that no interaction between arginine and native rhG-CSF. Titration of arginine into the denaturing buffer also showed heat uptake, but the baseline was gradually increased. When arginine was titrated into the denaturing buffer containing protein, the power profile only showed as a gradually increased line and no obvious heat peak could be observed (Fig. 5B, lower trace). However, after data processing of deduction the blank control, a sigmoid shape curve of the enthalpy change during the titration process could be obtained (Fig. 5D). This result indicated that there is an endothermic reaction between arginine and denatured rhG-CSF. Isothermal titration calorimetry measurements provided direct evidences that confirmed no interaction existed between native rhG-CSF
As a low molecular additive, arginine is widely used to assist protein refolding. Our previous studies demonstrated that arginine not only inhibits aggregation, but also promotes the formation of correct structures [18,19]. However, the mechanism of how arginine exerts its effects on protein refolding requires more direct experimental data. In this study, HDX-MS and ITC were used to provide more informative and direct details for investigating arginine-assisted protein refolding. Denatured rhG-CSF was fully deuterated, thus more deuterium atoms remained after 10-second exchange-out reaction indicated lower HDX rate at the monitored points in refolding process (Fig. 1). HDX rate is related to two main factors: solvent exposure and hydrogen bonding. A slow exchange rate means solvent protection and/or hydrogen bonding [29]. Deuteration levels increased with protein refolding proceeding was mainly due to the formation of secondary and tertiary structures that protect amide deuterium atoms from being exchanged out. Previous reports have shown that arginine could not accelerate or even slow down protein refolding rate [30,31]. The protection of amide deuterium atoms induced by arginine thus could be explained by interacting with protein rather than accelerating native structure formation. In addition, deuteration levels of rhG-CSF reached the same extreme at the end of refolding whether arginine was present or not (Fig. 1). Therefore, we suggested that arginine put its effect on protein refolding by reversibly combining with denatured and partially refolded proteins, which provided protection for amide deuterium atoms from being exchanged with solvent hydrogen atoms; and then it was gradually excluded from protein as refolding proceeded and no connections existed when the protein refolded completely. Pepsin digested proteins into peptides in low pH values so that HDX quenching followed by enzymatic digestion could be introduced to determine the exact region of deuteration [32]. Pepsin digestion following HDX provided an approach to analyze the deuteration sites at peptide resolution. Peptides produced from completely refolded rhG-CSF by pepsin digestion possessed various deuteration levels which mainly resulted from structure-related protection for amide deuterium atoms (Fig. 3). Regions exposed to the solvent were generally deuterated rapidly to a high extent; whereas, parts of the protein buried in the core were labeled slower and to a lower degree. However, deuteration levels of each peptide reached the same margin at the end of the refolding process independent on the presence of arginine, confirming arginine can disassociate with rhG-CSF after reviving its native conformation which was consistent with that revealed by HDX of the whole molecule. Six peptides with final deuteration levels over 40% contributed more to the deuteration of rhG-CSF and thus were monitored during the
D. Zhao et al. / Biochimica et Biophysica Acta 1854 (2015) 39–45
43
Fig. 4. Time dependence of deuterium exchange out of selected peptides generated from rhG-CSF by pepsin proteolysis. Exchange-out reactions for refolded samples were performed at specific refolding time intervals. Results are shown for refolding times between 10 and 540 min. Data shown are the mean value of three parallel experiments ± s.d. Square = refolding in the absence of arginine; circle = refolding in the presence of arginine.
refolding process to analyze the kinetics of HDX for each (Fig. 4). Although each peptide possessed the same deuteration level after rhG-CSF molecule reached refolding equilibrium with and without arginine assistance, a substantial distinction was observed during the two refolding processes. The slow HDX rate accompanied arginine participation resulted from the sequence-related interaction between arginine and protein which protected amide deuterium atoms located in these binding sites from being exchanged. Combining the HDX results of the peptic peptides with three dimensional structures obtained from crystallographic and NMR approaches provide complementary information [33]. For example, HDX revealed rhG-CSF refolding dynamics in solution phase even at peptide resolution, whereas crystallographic and NMR could provide detailed spatial structures of these peptides. As shown in Fig. 6, crystallographic studies of rhG-CSF revealed that almost all the peptides identified above were located in the α-helix domain. Hydrogen bonds are distributed
abundantly within the α-helix. Furthermore, conjugation between the double bond and the nitrogen lone pairs in the guanidinium group that caps the distal end of arginine causes the positive charge to become delocalized, enabling the formation of multiple hydrogen bonds. Thus, arginine might participate in the formation of hydrogen bonds with denatured and partially refolded rhG-CSF. Hydrogen bonding is one of the factors of slow HDX rate, thus hydrogen bond formation could also explain our experimental results. That is, arginine, a hydrogen bond donor, contacts proteins at specific regions that act as hydrogen bond receivers to protect amide deuterium atoms located at these sites from being exchanged. Although the electrostatic and hydrophobic interactions between arginine and protein were presumed responsible for its aggregation inhibition property [34,35], results in this study provided direct experimental evidences that hydrogen bonding is an important factor contributing to the combination of arginine and protein during refolding process.
44
D. Zhao et al. / Biochimica et Biophysica Acta 1854 (2015) 39–45
Fig. 5. Calorimetric titration of rhG-CSF with arginine. (A, B) Representative profiles of the titration of rhG-CSF with arginine in refolding buffer (A) and denaturing buffer (B). Four-microliter aliquots of a 10 mM arginine solution were injected at time intervals of 150 seconds into a 20 μM solution of rhG-CSF (lower trace) or buffer without rhG-CSF (upper trace). (C, D) Integrated data of net heat exchange of panel A (C) and panel B (D) after subtraction of control data for titration into the buffer without rhG-CSF.
ITC experiments confirmed the interaction of rhG-CSF with arginine in terms of thermodynamics. Results of ITC provided direct evidences that arginine only interact with denatured protein, but not with the
native one. However, the high arginine/protein molar ratio of the combination saturation indicated a weak interaction between arginine and protein. 5. Conclusions Hydrogen/deuterium exchange mass spectrometry and isothermal titration calorimetry were used to probe the mechanism of how low molecular weight additive of arginine improves protein refolding. It was clearly shown that there existed reversible interaction between arginine and protein which were crucial to promote protein refolding. At the initial stage of refolding, arginine combined with denatured and partially refolded protein through hydrogen bonding. Then arginine was gradually excluded from protein as its structure became more compact. Finally, almost no interaction existed between arginine and protein when protein regained its native conformation. The weak interaction between protein and arginine minimized the possibility of aggregation caused by protein–protein collision but put no adverse effects on conformation transformation. The strategy of combining HDX-MS with ITC is powerful to analyse the effects of refolding additives, which might provide a promising perspective for elucidating the mechanism of efficient refolding and screening the potential additives for enhancing refolding. Acknowledgments
Fig. 6. Location of the segments spanning 16–47, 72–84, 84–93, 114–124, 145–153 and 154–162 in the X-ray crystal structure of rhG-CSF (PDB ID: 1GNC [36]). Regions with different deuteration levels are shown using colors in accordance with Fig. 3.
This research was financially supported by the National Natural Science Foundation of China (Grant Nos. 20976178, 21376248 and 21336010) and the National High Technology Research and Development Program of China (863 Program) (Nos. 2012AA021202 and 2014AA022109).
D. Zhao et al. / Biochimica et Biophysica Acta 1854 (2015) 39–45
References [1] M.C. Manning, D.K. Chou, B.M. Murphy, R.W. Payne, D.S. Katayama, Stability of protein pharmaceuticals: an update, Pharm. Res. 27 (2010) 544–575. [2] J.J. Li, Y.D. Liu, F.W. Wang, G.H. Ma, Z.G. Su, Hydrophobic interaction chromatography correctly refolding proteins assisted by glycerol and urea gradients, J. Chromatogr. A 1061 (2004) 193–199. [3] F.G. Meng, Y.D. Park, H.M. Zhou, Role of proline, glycerol, and heparin as protein folding aids during refolding of rabbit muscle creatine kinase, Int. J. Biochem. Cell Biol. 33 (2001) 701–709. [4] Z. Gu, Z. Su, J.C. Janson, Urea gradient size-exclusion chromatography enhanced the yield of lysozyme refolding, J. Chromatogr. A 918 (2001) 311–318. [5] M. Li, G. Zhang, Z. Su, Dual gradient ion-exchange chromatography improved refolding yield of lysozyme, J. Chromatogr. A 959 (2002) 113–120. [6] F.S. Rudolph R, Process for obtaining renatured proteins, 1990, U.S. patent, 4, 933, 434. [7] W.J. Lin, J.A. Traugh, Renaturation of casein kinase-II from recombinant subunits produced in Escherichia coli—purification and characterization of the reconstituted holoenzyme, Protein Expr. Purif. 4 (1993) 256–264. [8] J. Buchner, R. Rudolph, Renaturation, purification and characterization of recombinant fab-fragments produced in Escherichia coli, Nat. Biotechnol. 9 (1991) 157–162. [9] J. Buchner, U. Brinkmann, I. Pastan, Renaturation of a single-chain immunotoxin facilitated by chaperones and protein disulfide isomerase, Nat. Biotechnol. 10 (1992) 682–685. [10] T. Stoyan, U. Michaelis, H. Schooltink, M. Vandam, R. Rudolph, P.C. Heinrich, S. Rosejohn, Recombinant soluble human interleukin-6 receptor—expression in Escherichia coli, renaturation and purification, Eur. J. Biochem. 216 (1993) 239–245. [11] J.G.S. Ho, A.P.J. Middelberg, P. Ramage, H.P. Kocher, The likelihood of aggregation during protein renaturation can be assessed using the second virial coefficient, Protein Sci. 12 (2003) 708–716. [12] S. Bell, S. Hansen, J. Buchner, Refolding and structural characterization of the human p53 tumor suppressor protein, Biophys. Chem. 96 (2002) 243–257. [13] T. Arakawa, D. Ejima, K. Tsumoto, N. Obeyama, Y. Tanaka, Y. Kita, S.N. Timasheff, Suppression of protein interactions by arginine: a proposed mechanism of the arginine effects, Biophys. Chem. 127 (2007) 1–8. [14] L. Ito, K. Shiraki, T. Matsuura, M. Okumura, K. Hasegawa, S. Baba, H. Yamaguchi, T. Kumasaka, High-resolution X-ray analysis reveals binding of arginine to aromatic residues of lysozyme surface: implication of suppression of protein aggregation by arginine, Protein Eng. Des. Sel. 24 (2011) 269–274. [15] V. Vagenende, A.X. Han, M. Mueller, B.L. Trout, Protein-associated cation clusters in aqueous arginine solutions and their effects on protein stability and size, ACS Chem. Biol. 8 (2012) 416–422. [16] D. Shukla, B.L. Trout, Preferential interaction coefficients of proteins in aqueous arginine solutions and their molecular origins, J. Phys. Chem. B 115 (2010) 1243–1253. [17] C.P. Schneider, D. Shukla, B.L. Trout, Arginine and the hofmeister series: the role of ion–ion interactions in protein aggregation suppression, J. Phys. Chem. B 115 (2011) 7447–7458. [18] Y.D. Liu, J.J. Li, F.W. Wang, J. Chen, P. Li, Z.G. Su, A newly proposed mechanism for arginine-assisted protein refolding—not inhibiting soluble oligomers although promoting a correct structure, Protein Expr. Purif. 51 (2007) 235–242.
45
[19] J. Chen, Y.D. Liu, X.N. Li, Y.J. Wang, H. Ding, G.H. Ma, Z.G. Su, Cooperative effects of urea and L-arginine on protein refolding, Protein Expr. Purif. 66 (2009) 82–90. [20] E.M. Lyutova, A.S. Kasakov, B.Y. Gurvits, Effects of arginine on kinetics of protein aggregation studied by dynamic laser light scattering and tubidimetry techniques, Biotechnol. Prog. 23 (2007) 1411–1416. [21] J.R. Engen, Analysis of protein conformation and dynamics by hydrogen/deuterium exchange ms, Anal. Chem. 81 (2009) 7870–7875. [22] H.M. Zhang, S. Kazazic, T.M. Schaub, J.D. Tipton, M.R. Emmett, A.G. Marshall, Enhanced digestion efficiency, peptide ionization efficiency, and sequence resolution for protein hydrogen/deuterium exchange monitored by Fourier transform ion cyclotron resonance mass spectrometry, Anal. Chem. 80 (2008) 9034–9041. [23] Y. Liang, J. Li, J. Chen, C.C. Wang, Thermodynamics of the folding of D‐glyceraldehyde‐3‐ phosphate dehydrogenase assisted by protein disulfide isomerase studied by microcalorimetry, Eur. J. Biochem. 268 (2001) 4183–4189. [24] F. Yang, B.R. Zhou, P. Zhang, Y.F. Zhao, J. Chen, Y. Liang, Binding of ferulic acid to cytochrome c enhances stability of the protein at physiological pH and inhibits cytochrome c-induced apoptosis, Chem. Biol. Interact. 170 (2007) 231–243. [25] A. Ziegler, J. Seelig, High affinity of the cell-penetrating peptide HIV-1 Tat-PTD for DNA, Biochemistry 46 (2007) 8138–8145. [26] A. Loregian, E. Sinigalia, B. Mercorelli, G. Palu, D.M. Coen, Binding parameters and thermodynamics of the interaction of the human cytomegalovirus DNA polymerase accessory protein, UL44, with DNA: implications for the processivity mechanism, Nucleic Acids Res. 35 (2007) 4779–4791. [27] I. Jelesarov, H.R. Bosshard, Isothermal titration calorimetry and differential scanning calorimetry as complementary tools to investigate the energetics of biomolecular recognition, J. Mol. Recognit. 12 (1999) 3–18. [28] T.E. Wales, J.R. Engen, Hydrogen exchange mass spectrometry for the analysis of protein dynamics, Mass Spectrom. Rev. 25 (2006) 158–170. [29] T.M. Raschke, S. Marqusee, Hydrogen exchange studies of protein structure, Curr. Opin. Biotechnol. 9 (1998) 80–86. [30] J. Chen, Y.D. Liu, Y.J. Wang, H. Ding, Z.G. Su, Different effects of L-arginine on protein refolding: suppressing aggregates of hydrophobic interaction, not covalent binding, Biotechnol. Prog. 24 (2008) 1365–1372. [31] D.L. Hevehan, E.D. Clark, Oxidative renaturation of lysozyme at high concentrations, Biotechnol. Bioeng. 54 (1997) 221–230. [32] L. Cravello, D. Lascoux, E. Forest, Use of different proteases working in acidic conditions to improve sequence coverage and resolution in hydrogen/deuterium exchange of large proteins, Rapid Commun. Mass Spectrom. 17 (2003) 2387–2393. [33] S. Nakazawa, J. Ahn, N. Hashii, K. Hirose, N. Kawasaki, Analysis of the local dynamics of human insulin and a rapid-acting insulin analog by hydrogen/deuterium exchange mass spectrometry, Biochim. Biophys. Acta Protein Proteomics 1834 (2013) 1210–1214. [34] H. Hamada, K. Shiraki, L-Argininamide improves the refolding more effectively than L-arginine, J. Biotechnol. 130 (2007) 153–160. [35] Y. Nozaki, C. Tanford, The solubility of amino acids, diglycine, and triglycine in aqueous guanidine hydrochloride solutions, J. Biol. Chem. 245 (1970) 1648–1652. [36] T. Zink, A. Ross, K. Luers, C. Cieslar, R. Rudolph, T.A. Holak, Structure and dynamics of the human granulocyte-colony-stimulating factor determined by NMR-spectroscopy— loop mobility in a 4-helix-bundle protein, Biochemistry 33 (1994) 8453–8463.
Contents lists available at ScienceDirect
Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbapap
Interaction of arginine with protein during refolding process probed by amide H/D exchange mass spectrometry and isothermal titration calorimetry Dawei Zhao a, Yongdong Liu a,⁎, Guifeng Zhang a, Chun Zhang b, Xiunan Li a, Qingqing Wang a, Hong Shi a, Zhiguo Su a,⁎⁎ a b
National Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China Key Laboratory of Drug Targeting and Drug Delivery System, Ministry of Education, West China School of Pharmacy, Sichuan University, Chengdu 610041, China
a r t i c l e
i n f o
Article history: Received 13 May 2014 Received in revised form 8 October 2014 Accepted 9 October 2014 Available online 16 October 2014 Keywords: HDX-MS ITC Refolding Arginine rhG-CSF
a b s t r a c t Arginine has been widely used as low molecular weight additive to promote protein refolding by suppressing aggregate formation. However, methods to investigate the role of arginine in protein refolding are often limited on protein's global conformational properties. Here, hydrogen/deuterium exchange mass spectrometry (HDX-MS) was used to study the effects of arginine on recombinant human granulocyte colony-stimulating factor (rhG-CSF) refolding at the scale of peptide mapping. It was found that deuteration levels of rhG-CSF refolded with arginine was higher than that without arginine during the whole refolding process, but they became almost the same when the refolding reached equilibrium. This phenomenon indicated that arginine could protect some amide deuterium atoms from being exchanged with hydrogen, but the protection diminished gradually along with refolding proceeding. Enzymatic digestion revealed six particular peptides of 16–47, 72–84, 84–93, 114–124, 145–153 and 154–162 were mainly responsible for the deuteration, and all of them dominantly located in protein's α-helix domain. Furthermore, thermodynamics analysis by isothermal titration calorimetry provided direct evidence that arginine could only react with denatured and partially refolded rhG-CSF. Taking all of the results together, we suggest that arginine suppresses protein aggregation by a reversible combination. At the initial refolding stage, arginine could combine with the denatured protein mainly through hydrogen bonding. Subsequently, arginine is gradually excluded from protein with protein's native conformation recovering. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Protein aggregation brings great troubles to the production and storage of protein based therapeutics [1]. Overexpressed heterologous recombinant proteins in prokaryotic and eukaryotic cells often form inactive and insoluble protein particles, which must be re-solubilized to recover their native conformation in vitro by refolding. Always low refolding efficiency is the bottlenecks for large production of such protein
Abbreviations: HDX, hydrogen deuterium exchange; ESI-MS, electrospray ionization mass spectrometry; ITC, isothermal titration calorimetry; rhG-CSF, recombinant human granulocyte colony-stimulating factor; IBs, inclusion bodies; β-ME, 2-mercaptoethanol; TFA, trifluoroacetic acid; Tris, Tris (hydroxymethyl) aminomethane; HPLC, high performance liquid chromatography ⁎ Correspondence to: Y. Liu National Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, No. 1 Bei-Er-Tiao Street, Haidian District, Beijing 100190, China. Tel.: +86 10 82545028; fax: +86 10 62561813. ⁎⁎ Correspondence to: Z. Su, National Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, No. 1 Bei-Er-Tiao Street, Haidian District, Beijing 100190, China. Tel.: +86 10 62561817; fax: +86 10 62561813. E-mail addresses: [email protected] (Y. Liu), [email protected] (Z. Su).
http://dx.doi.org/10.1016/j.bbapap.2014.10.007 1570-9639/© 2014 Elsevier B.V. All rights reserved.
therapeutics. Aggregation is the main instable factor during protein storage. Protein aggregate will not only decrease the bioactivity but also cause some degree of immunogenicity. To solve these problems, a variety of low molecular weight additives, such as sugars, polyols, detergents, salts, etc., have been used to improve protein refolding efficiency or increase protein solubility and long-term stability [2–5]. Among these additives, arginine is most widely used to prevent protein aggregation. Arginine was first used in the refolding of human tissue type plasminogen activator [6]. Subsequently, it was successfully applied to the refolding of casein II [7], Fab antibody fragments [8,9], interleukin-6 receptor [10], lysozyme [11], human p53 tumor suppressor protein [12], among others. To elucidate the mechanism by which arginine suppresses protein aggregation, various theories have been proposed. Arakawa et al. [13] compared the solubility of amino acids in aqueous arginine solution and suggested that interactions of arginine with aromatic amino acid side chains on protein surface contributed to its aggregation inhibition. Ito et al. [14] investigated the interaction between arginine and lysozyme with high resolution X-ray analysis and also suggested that shielding aromatic residues inhibited hydrophobic interactions on the protein surface. Recently, Vagenende et al. [15] carried out molecular
40
D. Zhao et al. / Biochimica et Biophysica Acta 1854 (2015) 39–45
dynamics simulation to probe the interactions of arginine with protein and reported that cluster formation of arginine cations at the specific protein surface loci might affect the apparent protein radius and even the apparent charge and hydrophobic properties of protein surface. Combined with dialysis/densimetry and vapor pressure osmometry techniques, their research group suggested that preferential interaction and attractive interactions between arginine molecules might also played a role in arginine behavior [16,17]. However, protein refolding is a complex procedure, limited experimental data for arginine action lead to the detailed pathway of arginine-assisted protein refolding remains ambiguous. Traditional methods for probing the effects of arginine on protein refolding, such as chromatographic analysis [18], fluorescence spectra analysis [19], and dynamic laser light scattering technique [20], only focused on the global conformational properties and couldn't provide detailed and direct information. Although X-ray technology could give an atomic resolution, it just provides static information. An alternative strategy, the use of hydrogen/deuterium exchange (HDX) has become attractive because it could characterize protein dynamics and conformation [21]. Hydrogen atoms located at the peptide backbone amide linkage (NHs) are most easily monitored for their relatively slow exchange rate with deuterium atoms in solvent. HDX correlates with two intertwined factors: accessibility and hydrogen bonding strength of amide hydrogen atoms in the protein structure. In general, exchange rate of NHs in folded proteins is many orders of magnitude slower than those in an unfolded version of the same sequence. Therefore, differences in the rate of HDX of NHs reflect alterations in protein conformation, structure, and intramolecular and intermolecular associations. Characterization of HDX behavior can be made easily by combining with mass spectrometry (MS) because the mass of protein will change one Da for each hydrogen/ deuterium exchange. In addition, the amounts of protein required by HDX-MS are orders of magnitude smaller than those of other biophysical methods for studying protein conformation and dynamics. To increase the spatial resolution of HDX, digestion with an acidic enzyme has been incorporated following exchange reaction quenching [22]. Thus, combination of HDX with liquid chromatographic separation of digested peptides and MS detection will provide detailed information regarding protein conformational mobility at the scale of peptide mapping. In addition to HDX-MS, isothermal titration calorimetry (ITC) is another powerful tool to probe biomolecular interactions. It has been used to study the thermodynamics of protein refolding assisted by protein disulfide isomerase [23], interaction between bovine heart cytochrome c and small molecule [24], and even protein–DNA/RNA interactions [25,26]. Although ITC also focus on proteins' global properties, it could provide direct thermodynamic information which was complementary to those obtained by other measurements [27]. For confirming the findings from HDXMS analysis, ITC was applied to investigate the effects of arginine on protein refolding as well. Escherichia coli-expressed recombinant human granulocyte colonystimulating factor (rhG-CSF) was selected as a model protein to investigate the effect of arginine on its refolding. The benefit of arginine during rhG-CSF refolding has been reported [18,19], but there is insufficient evidence revealing how arginine exerts its influence. Here, we obtained some detailed evidences to elucidate the mechanism of arginine-assisted refolding of rhG-CSF by HDX-MS. Peptide mapping was performed as well to compare the deuterium uptake at specific regions of rhG-CSF during the refolding with and without arginine assistance. Furthermore, the results of ITC measurements also provided supporting evidence for the hypothesis deduced from HDX-MS experiments. 2. Materials and methods 2.1. Materials Deuterium oxide (D2O) was purchased from Cambridge Isotope Laboratories (Andover, MA, USA). Ultrapure urea-d4, DCl and pepsin
were obtained from Sigma Chemical Co. (St. Louis, MO, USA). 2Mercaptoethanol (β-ME) was purchased from Fluka Chemical Corps (Buchs, Switzerland). All other chemicals were analytical grades and obtained from Shanghai Reagent Company (Shanghai, China). Water used for the experiments was ultrapure water from Milli-Q (Millipore, MA, USA). Deuterated L-arginine (D-Arg) was prepared by dissolving L-arginine in D2O, freezing and lyophilizing the solution, and repeating the process twice more. 2.2. Protein preparation An E. coli strain (DH5α) transformed with an expression plasmid pBV220 was used as the source of rhG-CSF. Inclusion bodies of rhGCSF were expressed, isolated and purified as previously described [19]. After reduction and denaturation, solubilized rhG-CSF was diluted into the refolding buffer (2 M urea, 20 mM Tris–HCl, 0.5 M Arg, pH 8.5) to recover its correct conformation. Then, the refolding solution was purified by one-step cation exchange chromatography with SP Sepharose™ Fast Flow (GE healthcare, Uppsala, Sweeden) and rhG-CSF with a purity of more than 99% upon SDS-PAGE analysis was obtained. Purified rhGCSF was lyophilized and stored at −80 °C for subsequent experiments. 2.3. Hydrogen/deuterium exchange Lyophilized rhG-CSF was dissolved with denaturing buffer containing 8 M urea-d4, 1% (vol/vol) β-ME, 20 mM Tris–DCl, pH 8.9 (pD 8.5) and incubated for 6 h at room temperature to completely denaturate rhG-CSF and fully deuterate the amide and other labile hydrogen sites simultaneously. Refolding was initiated by a rapid 100-fold dilution of 10 mg/mL denatured and deuterated rhG-CSF solution into the refolding buffer containing 2 M urea-d4, 20 mM Tris–DCl, pH 8.9 (pD 8.5) with or without 0.5 M D-Arg. Exchange-out reaction was conducted by diluting refolded samples 10-fold with back exchange refolding buffer of 2 M urea, 20 mM Tris–HCl, pH 8.5 and incubating for 10 s at 4 °C after refolding was initiated for 10, 30, 60, 180, 300, 420 and 540 min. Labeled samples were quenched by adding DCl to bring the pD to 2.3 and lowering the temperature to 0 °C by incubating in an ice bath, then subjected to liquid chromatography–mass spectrometry (LC–MS) analysis. The amide back exchange level during LC–MS analysis was less than 10%, as determined in control experiments on denatured and fully labeled rhG-CSF. The deuteration level was calculated using the following equation: deuteration level ¼
exchanged amide hydrogen atoms 100% exchangeable amide hydrogen atoms
where the exchanged amide hydrogen atoms equal to the increase of the molecular weight of deuterated protein and the exchangeable amide hydrogen atoms equal to the total number of protein's amino acids, minus proline residues minus 1 for N-terminal amide. For rhGCSF, the number of total amide hydrogen is 175 and the exchangeable amide hydrogen atoms are 163. 2.4. Peptic digestion, liquid chromatography and mass spectrometry Following exchange reaction quenching, refolded and deuterated samples were digested with pepsin and subjected to high performance liquid chromatography (HPLC)/electrospray ionization tandem mass spectrometry (ESI-MS/MS) using the Agilent 1100 HPLC system (Agilent Technologies, Palo Alto, CA, USA) with a reversed phase HPLC column (Shiseido Proteonavi-C4, 250 × 4.6 mm I.D., 5 μm, Japan) and the LCQ DECA XP ion-trap mass spectrometry system (Thermo Finnigan, San Jose, CA, USA). Pepsin digestion was performed in 0.1% DCl for 5 min on ice with 1:4 molar ratio of rhG-CSF to pepsin. The conditions for peptide separation were as follows: 15-min linear gradient, 20%–37% acetonitrile containing 0.1% trifluoroacetic acid (TFA), flow rate of
D. Zhao et al. / Biochimica et Biophysica Acta 1854 (2015) 39–45
41
1 mL/min. Buffers and the column used were cooled to 0 °C in an ice bath. ESI-MS/MS conditions were set as: 4.5 kV spray voltage, 300 °C for the heat capillary and m/z from 400 to 900 for the MS scan range. The assignment of the peptides produced by pepsin to specific partial sequences of rhG-CSF was implemented as follow: firstly, rhG-CSF was divided into the shortest segments at all possible sites by pepsin. Then, some successive segments were reconstructed into larger peptides to match the data of MS analysis. Finally, the peptides were calculated with Bioworks Browser 3.1 (Thermo Finnigan, San Jose, CA) and compared the results with those obtained from MS/MS analysis. The confidence score of N 95 for peptides identification was guaranteed by this procedure. As the comparisons of the deuteration levels of the peptides during refolding were relative experiments, no control experiment was performed for deuterium back exchange. The results presented for peptides were just relative deuteration levels as others previously reported [28]. In this study, each experiment for MS analysis of peptic peptides was conducted in three replicates. 2.5. Isothermal titration calorimetry (ITC) experiments Calorimetric measurements were performed at 25 °C with an iTC200 system (MicroCal, GE Healthcare, Piscataway, NJ, USA) interfaced with a personal computer. All solutions were thoroughly degassed before using by stirring under vacuum. Lyophilized rhG-CSF and L-arginine were dissolved with refolding buffer (2 M urea containing 20 mM Tris–HCl, pH 8.5) or denaturing buffer (8 M urea containing 20 mM Tris–HCl and 1% (vol/vol) β-ME, pH 8.5) to investigate the interaction between rhG-CSF and arginine at folded and denatured status, respectively. In a typical binding experiment on the iTC200 system, 10 mM arginine was injected in 19 aliquots of 2 μL (except the first injection, which was 0.4 μL) into a 0.2004 mL sample cell containing 20 μM rhG-CSF. Calorimetric data were analyzed using Origin for ITC version 7.0383. 3. Results 3.1. Differences in HDX behavior of rhG-CSF between refolding with and without arginine assistance Deuteration levels of exchangeable amide hydrogen atoms in rhGCSF were measured at different time points after initiation of refolding to verify the effects caused by arginine on the refolding process (Fig. 1). Deuteration levels reached the same margin of 65% at the end of rhGCSF refolding process whether arginine was present or not, but the changing profile of deuteration levels differed significantly between the two refolding conditions. In the presence of arginine, deuteration levels increased sharply after rhG-CSF refolding was initiated and reached a plateau when refolding had proceeded for 180 minutes. In contrast, deuteration levels increased more slowly and took 420 minutes to achieve deuteration equilibrium in the absence of arginine. At the same refolding time, the higher deuteration levels of refolded rhG-CSF in the presence of arginine indicated slower HDX rates than those in the absence of arginine. However, differences in deuteration levels between the two refolding conditions diminished as refolding proceeded. The results revealed that arginine put some effects on rhG-CSF refolding, but the influence weakened with refolding proceeding and disappeared after rhG-CSF retrieved the native conformation.
Fig. 1. Kinetic analysis of deuteration level variation over rhG-CSF refolding time in the presence and absence of arginine. Deuteration levels are shown for refolding times between 10 min and 660 min. Square = refolding in the absence of arginine; circle = refolding in the presence of 0.5 M arginine.
and 20 minutes. Fig. 3 shows the HDX map and secondary structure information of the completely refolded rhG-CSF. Different color blocks represent the peptides with different deuteration levels (the corresponding amino acid sequence is directly above the block). The deuteration level of each peptide was calculated by the same equation for the entire molecule which was presented in Materials and Methods. Only 11 of 175 residues were not detected, providing nearly 94% sequence coverage. It could be seen that deuteration was distributed at regions of the whole rhG-CSF molecule but noticeable differences in deuteration level existed among these peptic peptides, indicating various protection of amide deuterium atoms associated with structure-related factors, such as hydrogen boding within α-helices. Six proteolytic fragments spanning 16–47, 72–84, 84–93, 114–124, 145–153 and 154–162, which mainly locate in α-helix, were identified to possess more than 40% deuteration. Previous studies have suggested that hydrogen bonding might be related with the interaction between arginine and protein. Then the changing trend of deuterium level of the peptic peptides, especially of the six peptides mentioned above, was thus analyzed during rhG-CSF refolding with and without arginine assistance.
3.2. Deuteration levels of peptic peptides generated from completely refolded rhG-CSF To identify the deuteration regions, rhG-CSF was digested using pepsin after refolding and exchange-out reaction followed by HPLC–MS analysis. Fig. 2 shows a representative chromatogram obtained by pepsin digestion and RP-HPLC separation (peaks are labeled with their corresponding peptides). All peptides analyzed were eluted between 5
Fig. 2. Representative RP-HPLC chromatogram of peptides from completely refolded rhGCSF produced by pepsin digestion. Refolded rhG-CSF was digested with 1:4 molar ratio of rhG-CSF to pepsin. Peaks that correspond to the peptic peptides are labeled.
42
D. Zhao et al. / Biochimica et Biophysica Acta 1854 (2015) 39–45
Fig. 3. HDX map for completely refolded rhG-CSF. Denatured rhG-CSF refolded completely in 2 M urea-d4 containing 20 mM Tris–DCl, pD 8.5, then an exchange-out reaction was performed in 2 M urea containing 20 mM Tris–HCl, pH 8.5. Subsequently, refolded and deuterated samples were digested with pepsin and subjected to LC–MS analysis. Each block represents a pepsin-generated fragment of rhG-CSF. The deuteration level for each peptide is shown as a percentage (color-coded) of the theoretical maximum. Secondary structural information is also illustrated in the map.
3.3. Time dependence of deuterium exchange out of specific peptides generated from rhG-CSF
and arginine, but a combination occurred when denatured rhG-CSF and arginine coexisted in 8 M urea.
Deuterium exchange out analysis for the six peptides identified above was carried out during rhG-CSF refolding process in the presence or absence of arginine. At different refolding stage, samples after exchange-out and pepsin digestion were subjected to LC–MS analysis. As shown in Fig. 4, the deuteration profiles of the six selected peptides during the refolding process were similar to that of the entire rhG-CSF molecule. That is, higher deuteration levels with arginine addition could be observed at the same refolding point and finally reached the same levels with or without arginine. However, for those peptides with higher final deuteration levels such as peptide of 72–84 and 114–124, difference in deuteration level between refolding with and without arginine addition was more obvious before reaching the plateau. All the data indicated that these particular peptide regions were responsible for the influence of arginine on rhG-CSF refolding. The gradually vanished discrepancy in deuteration levels between the two refolding conditions indicated that the interaction between arginine and these peptides gradually weakened along with refolding proceeding. The same final deuteration levels suggested that there should be no interaction between arginine and the peptides after the protein recovered its correct structure.
4. Discussion
3.4. Interaction of arginine with rhG-CSF investigated by ITC To further investigate the effect of arginine on rhG-CSF refolding, ITC experiments were performed with native and denatured rhG-CSF, the two extreme states of the refolding process. In order to eliminate possible disturbance from arginine interacting with the components in the refolding or denaturing buffers, control experiments were carried out by titrating arginine into these two buffers with no protein (Fig. 5A and B, upper trace). Fig. 5A showed that heat uptake by the titration of arginine into the blank refolding buffer was obvious. Titration of arginine into the refolding buffer containing rhG-CSF gave a similar heat uptake, and almost no enthalpy for each injection was observed after offset calculation (Fig. 5C). This result suggested that no interaction between arginine and native rhG-CSF. Titration of arginine into the denaturing buffer also showed heat uptake, but the baseline was gradually increased. When arginine was titrated into the denaturing buffer containing protein, the power profile only showed as a gradually increased line and no obvious heat peak could be observed (Fig. 5B, lower trace). However, after data processing of deduction the blank control, a sigmoid shape curve of the enthalpy change during the titration process could be obtained (Fig. 5D). This result indicated that there is an endothermic reaction between arginine and denatured rhG-CSF. Isothermal titration calorimetry measurements provided direct evidences that confirmed no interaction existed between native rhG-CSF
As a low molecular additive, arginine is widely used to assist protein refolding. Our previous studies demonstrated that arginine not only inhibits aggregation, but also promotes the formation of correct structures [18,19]. However, the mechanism of how arginine exerts its effects on protein refolding requires more direct experimental data. In this study, HDX-MS and ITC were used to provide more informative and direct details for investigating arginine-assisted protein refolding. Denatured rhG-CSF was fully deuterated, thus more deuterium atoms remained after 10-second exchange-out reaction indicated lower HDX rate at the monitored points in refolding process (Fig. 1). HDX rate is related to two main factors: solvent exposure and hydrogen bonding. A slow exchange rate means solvent protection and/or hydrogen bonding [29]. Deuteration levels increased with protein refolding proceeding was mainly due to the formation of secondary and tertiary structures that protect amide deuterium atoms from being exchanged out. Previous reports have shown that arginine could not accelerate or even slow down protein refolding rate [30,31]. The protection of amide deuterium atoms induced by arginine thus could be explained by interacting with protein rather than accelerating native structure formation. In addition, deuteration levels of rhG-CSF reached the same extreme at the end of refolding whether arginine was present or not (Fig. 1). Therefore, we suggested that arginine put its effect on protein refolding by reversibly combining with denatured and partially refolded proteins, which provided protection for amide deuterium atoms from being exchanged with solvent hydrogen atoms; and then it was gradually excluded from protein as refolding proceeded and no connections existed when the protein refolded completely. Pepsin digested proteins into peptides in low pH values so that HDX quenching followed by enzymatic digestion could be introduced to determine the exact region of deuteration [32]. Pepsin digestion following HDX provided an approach to analyze the deuteration sites at peptide resolution. Peptides produced from completely refolded rhG-CSF by pepsin digestion possessed various deuteration levels which mainly resulted from structure-related protection for amide deuterium atoms (Fig. 3). Regions exposed to the solvent were generally deuterated rapidly to a high extent; whereas, parts of the protein buried in the core were labeled slower and to a lower degree. However, deuteration levels of each peptide reached the same margin at the end of the refolding process independent on the presence of arginine, confirming arginine can disassociate with rhG-CSF after reviving its native conformation which was consistent with that revealed by HDX of the whole molecule. Six peptides with final deuteration levels over 40% contributed more to the deuteration of rhG-CSF and thus were monitored during the
D. Zhao et al. / Biochimica et Biophysica Acta 1854 (2015) 39–45
43
Fig. 4. Time dependence of deuterium exchange out of selected peptides generated from rhG-CSF by pepsin proteolysis. Exchange-out reactions for refolded samples were performed at specific refolding time intervals. Results are shown for refolding times between 10 and 540 min. Data shown are the mean value of three parallel experiments ± s.d. Square = refolding in the absence of arginine; circle = refolding in the presence of arginine.
refolding process to analyze the kinetics of HDX for each (Fig. 4). Although each peptide possessed the same deuteration level after rhG-CSF molecule reached refolding equilibrium with and without arginine assistance, a substantial distinction was observed during the two refolding processes. The slow HDX rate accompanied arginine participation resulted from the sequence-related interaction between arginine and protein which protected amide deuterium atoms located in these binding sites from being exchanged. Combining the HDX results of the peptic peptides with three dimensional structures obtained from crystallographic and NMR approaches provide complementary information [33]. For example, HDX revealed rhG-CSF refolding dynamics in solution phase even at peptide resolution, whereas crystallographic and NMR could provide detailed spatial structures of these peptides. As shown in Fig. 6, crystallographic studies of rhG-CSF revealed that almost all the peptides identified above were located in the α-helix domain. Hydrogen bonds are distributed
abundantly within the α-helix. Furthermore, conjugation between the double bond and the nitrogen lone pairs in the guanidinium group that caps the distal end of arginine causes the positive charge to become delocalized, enabling the formation of multiple hydrogen bonds. Thus, arginine might participate in the formation of hydrogen bonds with denatured and partially refolded rhG-CSF. Hydrogen bonding is one of the factors of slow HDX rate, thus hydrogen bond formation could also explain our experimental results. That is, arginine, a hydrogen bond donor, contacts proteins at specific regions that act as hydrogen bond receivers to protect amide deuterium atoms located at these sites from being exchanged. Although the electrostatic and hydrophobic interactions between arginine and protein were presumed responsible for its aggregation inhibition property [34,35], results in this study provided direct experimental evidences that hydrogen bonding is an important factor contributing to the combination of arginine and protein during refolding process.
44
D. Zhao et al. / Biochimica et Biophysica Acta 1854 (2015) 39–45
Fig. 5. Calorimetric titration of rhG-CSF with arginine. (A, B) Representative profiles of the titration of rhG-CSF with arginine in refolding buffer (A) and denaturing buffer (B). Four-microliter aliquots of a 10 mM arginine solution were injected at time intervals of 150 seconds into a 20 μM solution of rhG-CSF (lower trace) or buffer without rhG-CSF (upper trace). (C, D) Integrated data of net heat exchange of panel A (C) and panel B (D) after subtraction of control data for titration into the buffer without rhG-CSF.
ITC experiments confirmed the interaction of rhG-CSF with arginine in terms of thermodynamics. Results of ITC provided direct evidences that arginine only interact with denatured protein, but not with the
native one. However, the high arginine/protein molar ratio of the combination saturation indicated a weak interaction between arginine and protein. 5. Conclusions Hydrogen/deuterium exchange mass spectrometry and isothermal titration calorimetry were used to probe the mechanism of how low molecular weight additive of arginine improves protein refolding. It was clearly shown that there existed reversible interaction between arginine and protein which were crucial to promote protein refolding. At the initial stage of refolding, arginine combined with denatured and partially refolded protein through hydrogen bonding. Then arginine was gradually excluded from protein as its structure became more compact. Finally, almost no interaction existed between arginine and protein when protein regained its native conformation. The weak interaction between protein and arginine minimized the possibility of aggregation caused by protein–protein collision but put no adverse effects on conformation transformation. The strategy of combining HDX-MS with ITC is powerful to analyse the effects of refolding additives, which might provide a promising perspective for elucidating the mechanism of efficient refolding and screening the potential additives for enhancing refolding. Acknowledgments
Fig. 6. Location of the segments spanning 16–47, 72–84, 84–93, 114–124, 145–153 and 154–162 in the X-ray crystal structure of rhG-CSF (PDB ID: 1GNC [36]). Regions with different deuteration levels are shown using colors in accordance with Fig. 3.
This research was financially supported by the National Natural Science Foundation of China (Grant Nos. 20976178, 21376248 and 21336010) and the National High Technology Research and Development Program of China (863 Program) (Nos. 2012AA021202 and 2014AA022109).
D. Zhao et al. / Biochimica et Biophysica Acta 1854 (2015) 39–45
References [1] M.C. Manning, D.K. Chou, B.M. Murphy, R.W. Payne, D.S. Katayama, Stability of protein pharmaceuticals: an update, Pharm. Res. 27 (2010) 544–575. [2] J.J. Li, Y.D. Liu, F.W. Wang, G.H. Ma, Z.G. Su, Hydrophobic interaction chromatography correctly refolding proteins assisted by glycerol and urea gradients, J. Chromatogr. A 1061 (2004) 193–199. [3] F.G. Meng, Y.D. Park, H.M. Zhou, Role of proline, glycerol, and heparin as protein folding aids during refolding of rabbit muscle creatine kinase, Int. J. Biochem. Cell Biol. 33 (2001) 701–709. [4] Z. Gu, Z. Su, J.C. Janson, Urea gradient size-exclusion chromatography enhanced the yield of lysozyme refolding, J. Chromatogr. A 918 (2001) 311–318. [5] M. Li, G. Zhang, Z. Su, Dual gradient ion-exchange chromatography improved refolding yield of lysozyme, J. Chromatogr. A 959 (2002) 113–120. [6] F.S. Rudolph R, Process for obtaining renatured proteins, 1990, U.S. patent, 4, 933, 434. [7] W.J. Lin, J.A. Traugh, Renaturation of casein kinase-II from recombinant subunits produced in Escherichia coli—purification and characterization of the reconstituted holoenzyme, Protein Expr. Purif. 4 (1993) 256–264. [8] J. Buchner, R. Rudolph, Renaturation, purification and characterization of recombinant fab-fragments produced in Escherichia coli, Nat. Biotechnol. 9 (1991) 157–162. [9] J. Buchner, U. Brinkmann, I. Pastan, Renaturation of a single-chain immunotoxin facilitated by chaperones and protein disulfide isomerase, Nat. Biotechnol. 10 (1992) 682–685. [10] T. Stoyan, U. Michaelis, H. Schooltink, M. Vandam, R. Rudolph, P.C. Heinrich, S. Rosejohn, Recombinant soluble human interleukin-6 receptor—expression in Escherichia coli, renaturation and purification, Eur. J. Biochem. 216 (1993) 239–245. [11] J.G.S. Ho, A.P.J. Middelberg, P. Ramage, H.P. Kocher, The likelihood of aggregation during protein renaturation can be assessed using the second virial coefficient, Protein Sci. 12 (2003) 708–716. [12] S. Bell, S. Hansen, J. Buchner, Refolding and structural characterization of the human p53 tumor suppressor protein, Biophys. Chem. 96 (2002) 243–257. [13] T. Arakawa, D. Ejima, K. Tsumoto, N. Obeyama, Y. Tanaka, Y. Kita, S.N. Timasheff, Suppression of protein interactions by arginine: a proposed mechanism of the arginine effects, Biophys. Chem. 127 (2007) 1–8. [14] L. Ito, K. Shiraki, T. Matsuura, M. Okumura, K. Hasegawa, S. Baba, H. Yamaguchi, T. Kumasaka, High-resolution X-ray analysis reveals binding of arginine to aromatic residues of lysozyme surface: implication of suppression of protein aggregation by arginine, Protein Eng. Des. Sel. 24 (2011) 269–274. [15] V. Vagenende, A.X. Han, M. Mueller, B.L. Trout, Protein-associated cation clusters in aqueous arginine solutions and their effects on protein stability and size, ACS Chem. Biol. 8 (2012) 416–422. [16] D. Shukla, B.L. Trout, Preferential interaction coefficients of proteins in aqueous arginine solutions and their molecular origins, J. Phys. Chem. B 115 (2010) 1243–1253. [17] C.P. Schneider, D. Shukla, B.L. Trout, Arginine and the hofmeister series: the role of ion–ion interactions in protein aggregation suppression, J. Phys. Chem. B 115 (2011) 7447–7458. [18] Y.D. Liu, J.J. Li, F.W. Wang, J. Chen, P. Li, Z.G. Su, A newly proposed mechanism for arginine-assisted protein refolding—not inhibiting soluble oligomers although promoting a correct structure, Protein Expr. Purif. 51 (2007) 235–242.
45
[19] J. Chen, Y.D. Liu, X.N. Li, Y.J. Wang, H. Ding, G.H. Ma, Z.G. Su, Cooperative effects of urea and L-arginine on protein refolding, Protein Expr. Purif. 66 (2009) 82–90. [20] E.M. Lyutova, A.S. Kasakov, B.Y. Gurvits, Effects of arginine on kinetics of protein aggregation studied by dynamic laser light scattering and tubidimetry techniques, Biotechnol. Prog. 23 (2007) 1411–1416. [21] J.R. Engen, Analysis of protein conformation and dynamics by hydrogen/deuterium exchange ms, Anal. Chem. 81 (2009) 7870–7875. [22] H.M. Zhang, S. Kazazic, T.M. Schaub, J.D. Tipton, M.R. Emmett, A.G. Marshall, Enhanced digestion efficiency, peptide ionization efficiency, and sequence resolution for protein hydrogen/deuterium exchange monitored by Fourier transform ion cyclotron resonance mass spectrometry, Anal. Chem. 80 (2008) 9034–9041. [23] Y. Liang, J. Li, J. Chen, C.C. Wang, Thermodynamics of the folding of D‐glyceraldehyde‐3‐ phosphate dehydrogenase assisted by protein disulfide isomerase studied by microcalorimetry, Eur. J. Biochem. 268 (2001) 4183–4189. [24] F. Yang, B.R. Zhou, P. Zhang, Y.F. Zhao, J. Chen, Y. Liang, Binding of ferulic acid to cytochrome c enhances stability of the protein at physiological pH and inhibits cytochrome c-induced apoptosis, Chem. Biol. Interact. 170 (2007) 231–243. [25] A. Ziegler, J. Seelig, High affinity of the cell-penetrating peptide HIV-1 Tat-PTD for DNA, Biochemistry 46 (2007) 8138–8145. [26] A. Loregian, E. Sinigalia, B. Mercorelli, G. Palu, D.M. Coen, Binding parameters and thermodynamics of the interaction of the human cytomegalovirus DNA polymerase accessory protein, UL44, with DNA: implications for the processivity mechanism, Nucleic Acids Res. 35 (2007) 4779–4791. [27] I. Jelesarov, H.R. Bosshard, Isothermal titration calorimetry and differential scanning calorimetry as complementary tools to investigate the energetics of biomolecular recognition, J. Mol. Recognit. 12 (1999) 3–18. [28] T.E. Wales, J.R. Engen, Hydrogen exchange mass spectrometry for the analysis of protein dynamics, Mass Spectrom. Rev. 25 (2006) 158–170. [29] T.M. Raschke, S. Marqusee, Hydrogen exchange studies of protein structure, Curr. Opin. Biotechnol. 9 (1998) 80–86. [30] J. Chen, Y.D. Liu, Y.J. Wang, H. Ding, Z.G. Su, Different effects of L-arginine on protein refolding: suppressing aggregates of hydrophobic interaction, not covalent binding, Biotechnol. Prog. 24 (2008) 1365–1372. [31] D.L. Hevehan, E.D. Clark, Oxidative renaturation of lysozyme at high concentrations, Biotechnol. Bioeng. 54 (1997) 221–230. [32] L. Cravello, D. Lascoux, E. Forest, Use of different proteases working in acidic conditions to improve sequence coverage and resolution in hydrogen/deuterium exchange of large proteins, Rapid Commun. Mass Spectrom. 17 (2003) 2387–2393. [33] S. Nakazawa, J. Ahn, N. Hashii, K. Hirose, N. Kawasaki, Analysis of the local dynamics of human insulin and a rapid-acting insulin analog by hydrogen/deuterium exchange mass spectrometry, Biochim. Biophys. Acta Protein Proteomics 1834 (2013) 1210–1214. [34] H. Hamada, K. Shiraki, L-Argininamide improves the refolding more effectively than L-arginine, J. Biotechnol. 130 (2007) 153–160. [35] Y. Nozaki, C. Tanford, The solubility of amino acids, diglycine, and triglycine in aqueous guanidine hydrochloride solutions, J. Biol. Chem. 245 (1970) 1648–1652. [36] T. Zink, A. Ross, K. Luers, C. Cieslar, R. Rudolph, T.A. Holak, Structure and dynamics of the human granulocyte-colony-stimulating factor determined by NMR-spectroscopy— loop mobility in a 4-helix-bundle protein, Biochemistry 33 (1994) 8453–8463.