Impact of inter-subunit interactions on the dimeric arginine kinase activity and structural stability

Archives of Biochemistry and Biophysics 512 (2011) 61–68

Contents lists available at ScienceDirect

Archives of Biochemistry and Biophysics journal homepage: www.elsevier.com/locate/yabbi

Impact of inter-subunit interactions on the dimeric arginine kinase activity and structural stability Qing-Yun Wu a,b, Feng Li d, Xiao-Yun Wang c,⇑, Zheng Jun Chen a a State Key Laboratory of Molecular Biology, Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai 200031, China b Graduate University of Chinese Academy of Sciences, Beijing 100049, China c College of Life Sciences, State Key Laboratory of Crop Biology, Shandong Agricultural University, Tai’an, Shandong 271018, China d State Key Laboratory of Biosafety, Nanjing Institute of Environmental Sciences, Ministry of Environmental Protection of China, Nanjing 210042, China

a r t i c l e

i n f o

Article history: Received 7 March 2011 and in revised form 20 April 2011 Available online 29 April 2011 Keywords: Arginine kinase Aggregation Conformational change Substrate synergism Structural stability

a b s t r a c t Arginine kinase (AK) is a key enzyme for cellular energy metabolism, catalyzing the reversible phosphoryl transfer from phosphoarginine to ADP in invertebrates. In this study, the inter-subunit hydrogen bonds between the Q53 and D200 and between D57 and D200 were disrupted to explore their roles in the activity and structural stability of Stichopus japonicus (S. japonicus) AK. Mutating Q53 and/or D57 to alanine (A) can cause pronounced loss of activity and substrate synergism, and cause distinct conformational changes. Spectroscopic experiments indicated that mutations destroying the inter-subunit hydrogen bonds impaired the structure of dimer AK, and resulted in a partially unfolded state. The inability to fold to the functional compact state made the mutants prone to be inactivated and aggregate under environmental stresses. Restoring hydrogen bonds in Q53E and D57E mutants could rescue the loss of activity and substrate synergism, and conformational changes. All those results suggested that the inter-subunit interactions played a key role in keeping the activity, substrate synergism and structural stability of dimer AK. The result herein may provide a clue in understanding the folding and self-assembly processes of oligomeric proteins. Ó 2011 Elsevier Inc. All rights reserved.

Introduction Arginine kinase (ATP: L-arginine phosphotransferase EC 2.7.3.3) (AK)1 catalyzes the reversible phosphorylation of arginine by ATP, yielding the phosphoarginine [1]. As a member of the phosphogen kinase (PK) family and analog of creatine kinase (CK) in vertebrates, AK is widely distributed in invertebrates, and plays a key role in cells, by buffering the ATP concentration according to cellular energy requirements [2–4]. Due to its presumed prominent role in energy metabolism and absence in vertebrates, AK could be chosen as a target to screen effective and harmless pesticide in agriculture [5,6]. In contrast to other PKs which are mostly dimeric or octameric, AKs are typically functional as monomers [6–8]. Unlike the mono-

⇑ Corresponding author. Fax: +86 538 8248696. E-mail address: [email protected] (X.-Y. Wang). Abbreviations used: AK, arginine kinase; CK, creatine kinase; PK, phosphogen kinase; IPTG, isopropyl-D-thiogalactopyranoside; ANS, 1-anilinonaphtalene-8-sulfonate; SEC, size exclusion chromatography; CD, circular dichroism; GdnHCl, guanidine hydrochloride; MG, molten globule intermediate; E. coli, Escherichia coli; SDS, sodium dodecyl sulfate; WT, wild type; [h]MRW, mean residue ellipticity; Emax, emission maximum wavelength of the intrinsic fluorescence. 1

0003-9861/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.abb.2011.04.015

meric 40 kDa AKs from molluscs and arthropods [5,6], Stichopus japonicus AK is dimeric [7–10], the same as cytoplasmic isoenzymes of the vertebrate CKs. The S. japonicus AK has raised interest recently because of its special position in evolution. Sequence analysis indicated that the dimer AK was evolutionarily closer to CK, while the conserved amino acids in its active sites are more like those of AKs though still significantly different from other AKs. Thus, it has been proposed that S. japonicus AK evolves at least twice during the evolution of PK: first at an early stage of PK evolution (its descendants are molluscan and arthropod AK), and second, from CK at a later time in metazoan evolution [10,11]. Therefore, studying the amino acid residues critical to catalytic activity of dimer AK may provide clues in understanding the difference of monomer AK and CK. Just like other PK, the catalytic mechanism of dimer AK belongs to the random-sequential bi mechanism as indicated in Scheme 1 [12,13]. If the binding of the first substrate facilitates the binding of the second substrate, synergism occurs. The parameter Kd/Km is often used to denote synergism, where Kd and Km are the dissociation constant in the absence and in the presence of the second substrate, respectively. A higher ratio of Kd/Km (Kd/Km > 1) indicates stronger synergism [12]. The synergism in substrate binding was suggested to be associated with substrate-induced conformational

62

Q.-Y. Wu et al. / Archives of Biochemistry and Biophysics 512 (2011) 61–68

Scheme 1.

changes within the tertiary complex [6,14]. However, little is known about the effect of amino acid residues outside of active sites on the synergism. The amino acid sequence of a protein determines the threedimensional conformation of the functional protein [15]. S. japonicus AK, as a special dimer protein, is a valuable model for studying the subunit dissociation and inter-subunit interactions of oligomeric proteins. Understanding the folding and self-assembly processes of oligomeric proteins remains a challenge. For oligomeric proteins, folding is generally thought to be hierarchical, the individual domains with different stabilities fold autonomously and independently [16]. However, the folding mechanism of oligomeric proteins may be much more complicated than monomer proteins because it involves both the folding of the individual subunit and the docking of these subunits [16]. To achieve the native tertiary structures, adjacent domains need to recognize each other through the domain-domain and the inter-subunit interface(s) [17–20]. Previous studies presumed that the inter-subunit interactions of S. japonicus AK may be crucial in keeping its activity and structural stability [21–23]. Furthermore, the hydrogen bonds between D200 and Q-53 and between D-200 and D-57 were deduced to play a key role in the dimerization of dimer AK [23]. However, little is known about the roles of inter-subunit hydrogen bonds in dimer AK activity, folding and structural stability. In this research, our study suggested that the mutations Q53A, D57A and Q53A/D57A can cause pronounced loss of AK activity and substrate synergism, and distinct conformational changes. Moreover, those mutations led to the conformational changes by destroying the inter-subunit hydrogen bonds and hindered the dimerization of AK. The partially unfolded state of mutant AKs made them susceptible to environmental stresses and prone to be inactivated and unfolded, form insoluble aggregates. All those results suggested that the inter-subunit interactions played a key role in the dimer AK activity, substrate synergism and structural stability.

GTGTCGCTAATCCCGATTTC-30 and 50 -GAA ATCGGGATTAGCGACACC GTTCTG-30 ; for D57E, 50 -CAGAACGGTGTCGAAAATCCCGATTTC-30 and 50 -GAAATCGGGATTTTCGACACCGTTCTG-30 mutated sequences are underlined. As for the double mutant, the Q53A mutant was used as template and the primers for D57A were used to introduce the second mutation. Then the cDNA of the mutants was cloned into expression vector pET-28a, sequenced and transformed into the Escherichia coli (E. coli) BL21 (DE3) codon plus. The WT and the mutant AKs fusion protein was expressed in E. coli BL21 and purified as described previously [24]. The purity was checked by SDS–PAGE. Protein concentration was determined using Bradford’s method [25]. The size exclusion chromatography (SEC) analysis was carried out using a FPLC system (General Electric Company) with a Superdex 200HR column at 25 °C. Each time, a total of 120 ll (0.4 mg/ml) sample was injected into the column pre-incubated with the standard buffer (10 mM glycine-NaOH, 1 mM DTT at pH 8.1).

Enzyme assay and determination of kinetic parameters AK activity (phosphoarginine synthesis) was assayed as previously described with some modification [26,27]. The assay mixture for AK determination consisted of 100 mM Tris, pH 8.0, 10 mM Larginine, 8 mM ATP-Na, 10 mM mercapto-ethanol and 10 ll of 0.01 mM enzyme solution. The absorbance at 660 nm was measured at 30 °C using an Ultrospec4300 pro UV–vis Spectrophotometer. The two-substrate graphical method was used to obtain the kinetic parameters [6]. The activity assays were carried out at the optimum pH (pH 8.1) and temperature (30 °C) with different concentrations of ATP and arginine. All the experiments were repeated at least four times.

Thermal stability of AK The thermal stability of WT and mutant AKs was determined by activity assay after being incubated at different temperatures. The enzyme solutions were incubated at given temperatures varying from 25 to 65 °C for 10 min, then cooled on ice and the activity was measured at 30 °C. The data were normalized to the activity measured at 25 °C. The aggregation of AKs at a given temperature was monitored by measuring the turbidity at 400 nm. The final protein concentrations of WT and mutant AKs were all adjusted to 2.3 lM.

Materials and methods Cloning, site-directed mutagenesis and expression of the mutant S. japonicus AKs Total RNAs were isolated from the muscle of S. japonicus by using the TRIzol reagent (Invitrogen). First strand cDNAs were prepared by reverse transcription of total RNAs with random primers. For amplification of AK cDNA, the primers 50 -GCTGGATCCCCGGTGTTAATCATGGCAAA-30 and 50 -GCACTCGAGGTCCCCAAGTAAACGGCT-30 were used, in which the restriction sites introduced were underlined. Then the PCR product was purified and inserted into the pET-28a vector. Five mutations (Q53A, D57A, Q53A/D57D, Q53E and D57E) were introduced into the template of the pET-28a-WT AK by overlap PCR using mutation primers. The sequences of the mutation primers were as follows: for Q53A, 50 -CTGGACAGAGCCATAGCTAACGGTGTCGAT-30 and 50 -ATCGACACCGTTAGCTATGGCTCTGTCCAG-30 ; for D53E, 50 -CTG GACAGAGCCATAGAAAACGGTGTCGAT-30 and 50 -ATCGAC ACCGTTTTCTATGGCT CTGTCCAG-30 ; for D57A, 50 -CAGAACG

Unfolding and refolding experiments For the unfolding experiment, the WT and mutant AKs were added to the standard buffer (pH 8.1) with different concentrations of GdnHCl dissolved for 24 h at equilibrium state. The refolding experiment was initiated by diluting the denatured AKs into the standard buffer (pH 8.1) with final GdnHCl concentrations ranging from 0.1 to 3 M. The intrinsic Trp fluorescence spectra of unfolding and refolding AKs were collected on an F-4500 spectrofluorometer using a 1-cm path-length cuvette. For ANS-fluorescence measurements, 10-fold molar excess of ANS was added to the samples. The samples were equilibrated for 30 min in the dark, and then the extrinsic fluorescence was measured on an F-4500 spectrofluorometer using a 1-cm path-length cuvette. Far-UV circular dichroism (CD) spectra were recorded on a Jasco 715 spectrophotometer with a 1 mm path-length cell. The final concentration of the enzyme for spectroscopic experiments was 2.3 lM and all experiments were carried out at 25 °C.

63

Q.-Y. Wu et al. / Archives of Biochemistry and Biophysics 512 (2011) 61–68

Parameter A and phase diagram analysis of intrinsic fluorescence data To more clearly characterize the effects of mutations on AK folding, Parameter A and phase diagram analysis, which are sensitive tools to identify folding intermediates, were used to compare the folding pathways of the WT and mutant AKs. Parameter A, which reflects the spectral shape of the intrinsic Trp fluorescence [28], was obtained by dividing the fluorescence intensity at 320 nm (I320) by the intensity at 365 nm (I365) during unfolding and refolding. The ‘‘phase diagram’’ analysis which is a sensitive tool to detect folding intermediates was carried out as described previously [29]. The phase diagram was constructed using I320 versus I365 for WT and mutant AKs in buffers containing different concentrations of GdnHCl. In the phase diagram, each straight line in the phase diagram reflects an ‘‘all-or-none’’ process, and the joint position of two lines indicates that an intermediate appeared at the corresponding GdnHCl concentration. Modeling the structure of mutant AKs In order to analyze the effect of the inter-subunit interaction on the AKs, both the SPDBV (http://swissmodel.expasy.org/) and the VMD modeling procedures (version 1.8.7, http://www.ks.uiuc. edu/Research/vmd/) [30] were used to model the structure of mutant AKs based on PDB files (PDB ID numbers 3JU5 and 3JU6) for S. japonicus AK from the Protein Data Bank. The energy was all minimized after introducing the mutations by SPDBV and VMD modeling procedures.

Results and discussion

decline in catalysis efficiency, compared with that (29.93 s1 mM1) of WT AK. For Q53E and D57E mutant AKs, most of the kinetic parameters were similar to those of WT AK (Table 1). A small decrease in Kd/Km and kcat =K Arg m values indicated a slight loss of substrate synergism and catalytic efficiency. Our results are consistent with what was reported by Wu and his colleagues [23]. They showed that mutations in Q53 and D57 led to remarkably decreased activity and substrate binding affinity, though the absolute values are different from ours. The K ATP of dim mer AK, for instance, is 0.823 ± 0.043 mM in our hands, which is similar to the results of some other groups (0.81 and 1.04 mM) [31,32], while in Wu’s report it is 6.1 ± 1.9 mM [23]. In this study and some other groups two-substrate graphical method was used to determine the kinetic parameters [31,32], while the data were fitted to the Hill function in Wu’s report [23]., since different assays were used in different studies [31,32], absolute values may be different. The synergism in substrate binding was found to be a common feature in PK, however, its molecular basis has not been established systematically. Previous studies implied that substitution of the amino acid residues near the active sites may decrease the substrate synergism by affecting the correct active sites conformation [13,32]. Our previously studies also suggested that the decrease of the substrate synergism was related to the conformational changes in the active sites [6]. In this study the decreased substrate synergism might be caused by the mutations destroyed AK dimerization and affected the conformation of active sites. Since Q53 and D57 do not participate in the catalysis or substrate binding of AK [23], the reason for the low activity of the mutant AKs might due to its impaired tertiary structure.

Kinetic parameters analysis

Mutations impaired the structure of AK

All the recombinant enzymes were successfully expressed as soluble fusion proteins and purified as described in the methods. The recombinant WT AK showed similar enzymatic characteristics to native AK, indicating the His6-tag portion had no effects on its activity (Table 1). As is shown in Table 1, all the mutated AKs retained 14.9–77.6% of the WT AK activity (kcat) and displayed decreased substrate affinity. The K Arg values of Q53A and D57A mutant AKs (1.036– m 1.571 mM) was 2 to 4 fold higher than that of recombinant WT AK (0.421 mM), while their K Arg values (2.279–2.828 mM) were d 1.8–2.2 fold higher than that of WT AK (1.275 mM), resulting in a Kd/Km value of 1.81–2.21. The changes in Kd/Km indicated that the synergism in substrate binding decreased significantly. As for the Q53A/D57A double mutant AK, the changes in kinetic parameters as well as the loss of substrate synergism were more severe than that of Q53A and D57A mutant AKs (Table 1). In addition, the double mutant had the lowest affinity for ATP (K ATP m = 4.021 mM), and its kcat value (4.38 s1 mM1) showed a 7 fold

In the SEC experiment, the SEC profile of WT AK showed a sharp peak at the elution volume of about 11.64 ml, corresponding to the native dimer AK with a molecular weight about 84 kDa (Fig. 1A). The SEC profile of Q53A and D57A mutant AKs centered at the elution volume of 12.50 and 12.85 ml, respectively, while the Q53E and D57E mutant AKs showed a sharp peak centered at the elution volume of 11.70 and 11.80 ml, respectively. As to the double mutant Q53A/D57A, the SEC profile revealed a widest peak centered at an elution volume of 13.55 ml. The wide peak and more lag time might be the apparent peak of the fast equilibrium of dimeric and monomeric species. This result suggested that the Q53A, D57A and Q53A/D57A mutations impaired the compact structure or the dimerization of AK. This was similar with the dissociation of CK, when inter-subunit interactions were destroyed [19,20]. The fast equilibrium of dimer and monomer observed in mutant AKs implied that residues Q53 and D57 in dimer AK might contribute to the stabilization of the dimer interface as residue R152 or D210 did in CK, since the destruction of interactions did not dissociate

Table 1 Comparison of kinetic parameters for the forward reaction of recombinant WT and mutants of AKs. Mutations a

Native AK Recombinant WT AK Q53A D57A Q53A/D57A Q53E D57E

Vmax (lmol Pi min1 mg1)

kcat(s1)

K Arg m (mM)

(mM) K Arg d

K ATP (mM) m

K ATP mM d

K d =K m

38.5 ± 1.894 44.61 ± 2.35 19.37 ± 1.03 15.18 ± 0.72 6.49 ± 0.31 34.52 ± 1.84 23.61 ± 1.18

25.7 ± 1.26 29.93 ± 1.56 13.09 ± 0.63 10.26 ± 0.48 4.38 ± 0.21 23.33 ± 1.16 15.95 ± 0.82

0.413 ± 0.0593 0.421 ± 0.024 1.036 ± 0.069 1.571 ± 0.086 2.726 ± 0.153 0.578 ± 0.032 0.744 ± 0.043

1.249 ± 0.231 1.275 ± 0.125 2.279 ± 0.114 2.828 ± 0.142 3.544 ± 0.182 1.561 ± 0.076 1.793 ± 0.087

0.814 ± 0.032 0.823 ± 0.043 1.321 ± 0.072 1.631 ± 0.099 3.092 ± 0.187 0.988 ± 0.056 1.132 ± 0.061

2.46 ± 0.0968 2.497 ± 0.096 2.908 ± 0.197 2.936 ± 0.153 4.021 ± 0.223 2.678 ± 0.135 2.939 ± 0.45

3.02 ± 0.15 3.03 ± 0.24 2.21 ± 0.13 1.81 ± 0.09 1.30 ± 0.07 2.72 ± 0.16 2.40 ± 0.13

Note: Kinetic parameters were obtained from at least three runs of the reaction. a Kinetic parameters cited from the article [32].

64

Q.-Y. Wu et al. / Archives of Biochemistry and Biophysics 512 (2011) 61–68

Fig. 1. Effects of mutations on AK structures detected by SEC (A), CD (B), intrinsic fluorescence spectra (C) and ANS fluorescence spectra (D). In Fig. 1 curve (1–7) represents spectra of WT AK, Q53E, D57E, Q53A, D57A, Q53A/D57A and ANS solution, respectively. The molecular weight standards for SEC were yeast alcohol dehydrogenase (150 kDa), Bovine serum albumin (65 kD) and half-mer hemoglobin (32 kD). The final concentration of AKs for spectroscopic experiments was 2.5 lM and all experiments were carried out at 25 °C.

dimer AK to monomer AK completely [19,20]. The effects of mutations on the dimerization of AK could be responsible for the decreased activity and the substrate synergism which is in agreement with the previous report that the dimeric structure is necessary for the high activity of dimer AK, although the isolated subunit also has limited activity [21,22]. The crystal structure of dimer AK suggested that the hydrogen bond between D200 and Q53 and between D200 and D57 constituted a hydrogen bond cluster between the inter-subunit [23]. Considering the fact that neither Q53 nor D57 is close to the active center, one can deduce that the hydrogen bonds between D200 and Q53 and between D200 and D57 may be important for keeping AK activity due to its roles in the stabilization of dimer AK. The secondary structure assayed by CD spectra is consistent with the SEC results. The mean residue ellipticity of CD spectra at 222 nm of the Q53A, D57A and Q53A/D57A mutant AKs was smaller than that of Q53E and D57E mutant and WT AKs at the same concentration (Fig. 1B). The changes of the mean residue ellipticity at 222 nm of the mutant and WT AKs showed the similar

trend to that of SEC profile changes. This result suggested that mutating D to A at Q53 and/or D57 impaired the inter-subunit interactions, which led to the distinct loss of regular secondary structures, while the Q53E and D57E mutations which rescued the inter-subunit interactions, only showed minor decrease of regular secondary structures. The decreased secondary structures can be explained by the hypothesis that Q53A, D57A and Q53A/D57A mutations impaired the inter-subunit interactions which not only destroyed the compact secondary structures interaction in the inter-subunit in the compact dimer structure but also influenced the primary regular secondary structure in the subunit itself. This result was also in agreement with Feng’s report showing that the mutation D54G destroyed the inter subunit hydrogen bond and the secondary structure [19]. There are four Trp residues (at position 208 and 218 in each subunit) in dimer AK located in the C-terminal domain. Intrinsic fluorescence spectra of Trp showed the tertiary structures were impaired by the mutations. The emission maximum of the intrinsic fluorescence (Emax) was found to red shift to 327 nm for Q53E,

Q.-Y. Wu et al. / Archives of Biochemistry and Biophysics 512 (2011) 61–68

328 nm for D57E, 331 nm for Q53A, 332 nm for D57A and 338 nm for Q53A/D57A compared to the WT AK (330 nm), which suggested that the Trp residues in the Q53A, D57A and Q53A/D57A mutant AKs were more exposed to water (Fig. 1C). Interestingly, intrinsic fluorescence spectra also suggested that the mutations also affected the structural compactness of the C-terminal domain, as all the four Trp residues in dimer AK are located on the C-terminal domain which is particularly important to AK structural stability [23,31, 33]. The ANS fluorescence spectra further validated the above results. The intensities of Q53A, D57A and Q53A/D57A mutant AKs were higher than that of WT AK at the same concentration (Fig. 1D), which reflected that the mutant AKs had more hydrophobic exposure than that of WT AK to allow the binding of the ANS molecules. These spectroscopic experiments clearly indicated that the mutations might impair both the structures (secondary and tertiary structures) and the dimerization of AK which were important for the formation of the compact functional enzyme. Previous studies also suggested that the dissociation of the subunits not only diminished the possible cooperation of the two subunits and reduced the activity [34], but also resulted in a less compact structure with more hydrophobic exposure [23,31,33]. All these results indicated that the inter-subunit interactions played a key role in keeping the activity, compact structure and the dimerization of AK. Mutations reduced AK thermal stability and led to AK thermal aggregation The activity of WT AK changed a little after heat treatment at the temperatures below 45 °C. A steep decrease of activity was observed between 45 and 60 °C, and a complete loss of activity occurred above 65 °C (Fig. 2A). Mutants (Q53A, D57A and Q53A/ D57A) with hydrogen bonds destroyed were much less stable than WT or mutants with hydrogen bonds (Q53E and D57E). The midpoints of thermal inactivation of WT and Q53E, D57E, Q53A, D57A and Q53A/D57A mutant AKs, were about 52.5 ± 0.5 °C, 50.9 ± 0.5 °C, 49.2 ± 0.5 °C, 43.5 ± 0.5 °C, 41.3 ± 0.5 °C and 35.9 ± 0.5 °C, respectively. Thermal inactivation of AK was often accompanied with aggregation and even minor conformational changes could initiate the formation of aggregates [18,30]. So aggregation experiments could sensitively detect changes of the structural stability of AK. As shown in Fig. 2B, only little aggregation of AK occurred after being heated at 48 °C for 60 min in case of WT and mutants with hydrogen bond (Q53E and D57E), although the aggregate amount of the two mutants was slightly increased when compared to WT AK. Differently, considerable aggregation was observed for mutants with destroyed hydrogen bond (Q53A, D57A and Q53A/D57A) at 48 °C

65

(Fig. 2B). When the heating temperature was increased to 55 °C, the aggregation rate of the mutants (Q53A, D57A and Q53A/ D57A) was much faster than that of WT AK. Combined with the spectroscopic results in Fig. 1, one can deduce that the looser structure and more hydrophobic exposure induced by the mutations might facilitate the protein aggregation at elevated temperatures. All those results indicated that the inter subunit interactions contributed to the structural stability of dimer AK. This is consistent with the previous reports showing that the stable proteins need the correct domain-domain reorganization and interactions [17–19]. Effects of the mutations on AK unfolding and refolding Spectroscopic spectra were used to investigate the effects of mutations on AK unfolding and refolding. In the unfolding process, the Emax of WT and mutant AKs showed a sharp red shift in the ranges of 0.1–0.5 M and 1.0–1.4 M GdnHCl concentrations (Fig. 3A). Meanwhile, the ANS intensity reached a peak at 0.4 M GdnHCl and a second peak at 1.2 M GdnHCl for the WT and mutant AKs (Fig 3C). Furthermore, in the CD spectra, the [h222]MRW also showed the similar transition, the [h222]MRW of WT and the mutant AKs decreased dramatically as GdnHCl concentration increased from 0 to 0.5 M. The ellipticity began to increase and reached a maximum at 1.0 M GdnHCl, with 59% secondary structure of the native enzyme, indicating that an intermediate with more secondary structure was induced (Fig 3E). These results suggested that two equilibrium intermediate states existed in the unfolding process of WT and mutant AKs. The first intermediate, which exists at about 0.4 M GdnHCl, is a compact inactive dimer lacking partial global structure, while the second intermediate, forming at 1.2 M GdnHCl, possesses characteristics similar to the globular folding intermediates described in the literature, which is quite consistent with previous observations [21,22]. The refolding transitions of WT and mutant AKs from the GdnHCl-denatured state were similar to those unfolding at GdnHCl concentrations above 0.5 M when monitored by all the three techniques (Fig. 3B, D and F). The deviation between the unfolding and refolding at GdnHCl concentrations below 0.5 M indicated that the folding of WT and mutant AKs was not fully reversible under our conditions. However, no large aggregates were observed as monitored by the turbidity at 400 nm (Fig. 2B). Thus irreversibility might be caused by the formation of soluble nonnative oligomers from the aggregation-prone intermediates [21,22,35]. Nevertheless, the transition curves of the mutant AKs were almost identical to that of WT AK at GdnHCl concentrations above 0.5 M, which suggested that the mutation did not affect the MGMU transitions. Previous studies implied that the dimeric-

Fig. 2. Thermal stability and aggregation of AKs. (A) Thermal stability of AKs. (B) Thermal aggregation of AKs at 48 °C.

66

Q.-Y. Wu et al. / Archives of Biochemistry and Biophysics 512 (2011) 61–68

Fig. 3. Effect of mutations on AKs folding investigated by Emax of the intrinsic fluorescence (A and B), the intensity of ANS fluorescence (C and D) and the mean residue ellipticity at 222 nm ([h222] MRW, 103 deg cm2 dmol1) of CD (E and F).

monomeric transition was mainly occurred at the GdnHCl concentrations from 0.4 to 1.2 M [21,31,33]. As shown in Fig. 3, no obvious difference was observed between WT and mutant AKs, which suggested that the mutations did not affect the dimeric-monomeric transition although they may affect the stability of the dimer AK. This conclusion is consistent with the previous reports showing that the dimerization is not the rate-limiting step in AK refolding [21,22,35]. The major difference observed in the spectroscopic spectra at low GdnHCl concentrations in folding process may be caused by the differences in their native structures (Fig. 1). Parameter A and phase diagram analysis were also used to clarify the effects of mutations on AK folding. As presented in Fig. 4A and B, the unfolding and refolding transition curves of mutant AKs were almost identical to that of WT AK for all conditions at GdnHCl concentrations varying from 0.5–6 M. Furthermore, as indicated by the phase diagram analysis in Fig. 4C and D, during the unfolding and refolding process, all the transitions could be fitted by three linear parts, which suggested that all the transitions composed three independent transitions processed with the appearance of two intermediates. Consistent with the spectro-

scopic experiments (Fig. 3), the joint position of the lines (Fig. 4C and D) was also at the same GdnHCl concentration (at 0.4 M and 1.2 M GdnHCl, respectively). These results coincided with those from direct spectra analysis in Fig. 3 which also suggested that the mutation did not affect the folding pathway of AK. The discrepancy between the WT and the mutant folding at low concentrations of GdnHCl was more likely to be caused by their different native structures, rather than the interference of the folding pathway by the mutations (Figs. 1, 3 and 4). This conclusion agrees with the previous observation that the dimerization is not the rate-limiting step in AK refolding [18,19,30]. We propose that the following mutations, Q53A, D57A and Q53A/D57A impaired the native ensembles of dimer AK which resulted in partially unfolded AK with higher free energy (Figs. 1 and 5). By calculating the energy of modeling structures of mutant AKs, we found that Q53A, D57A and Q53A/D57A mutations led to significant increase in energy, which means mutations result in unstable structures. The minimized free energy of the Q53A/D57A, D57A, Q53A, D57E, Q53E mutant AKs and the WT AK was 736.43, 449.55, 415.11, 66.53, 44.34 and 18.33 kJ mol1, respectively. The higher free energy might

Q.-Y. Wu et al. / Archives of Biochemistry and Biophysics 512 (2011) 61–68

67

Fig. 4. Parameter A (A and B) and phase diagram (C and D) analysis of intrinsic fluorescence data in WT and mutant AKs folding. Parameter A was obtained by dividing the fluorescence intensity at 320 nm (I320) by the intensity at 365 nm (I365). The phase diagram was constructed by I320 versus I365 for the folding of AKs.

indicate the more instability and the more amounts of monomer AK in the monomeric and the dimeric species. The above predicted results matched the SEC analysis results shown in the Fig. 1A. The partially folded Q53A, D57A and Q53A/D57A mutant AKs tend to be inactivated and unfolded under environmental stresses, and are prone to form insoluble aggregates. Structural analysis of mutant AKs

Fig. 5. Overall crystal structures of dimer AK (A) and the detail Q53 and D57 interactions (B). The hydrogen bonding was indicated by the green dot line. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

The structure of the mutant AKs was modeled to investigate the effects of inter-subunit interactions on the structural stability of dimer AK. The result suggested that in the Q53A, D57A and Q53A/ D57A mutant AKs, the mutations destroyed the inter-subunit interactions which led dimer AK to a loose, partially folded structure (Figs. 1 and 5). While in the Q53E and D57E mutant AKs, the mutations could rescue the inter-subunit interactions resulting in a more compact structure (Figs. 1 and 5). Combined with the kinetic parameters, spectroscopic and protein folding results, one can deduce that the destruction of inter-subunit interactions and incompact structure were responsible for the decreased substrate synergism, activity and structural stability, which were consistent with previous studies [6,31,33]. As the folding of oligomeric proteins is complex, understanding the folding and self-assembly processes of oligomeric proteins remains a challenge [16,18]. Our study suggested that the intersubunit interactions might play a key role in the domain-domain reorganization, interactions and self-assembly processes of dimer AK. To achieve the native tertiary structures, structurally adjacent domains need to recognize each other through the inter-subunit interactions and inter-subunit interface(s) [17–20]. All in all, the results herein may provide a clue in understanding the folding and self-assembly processes of oligomeric proteins. AK has been proposed as a target for innovative pesticide compounds [36–38] due to its presumed prominent role in energy

68

Q.-Y. Wu et al. / Archives of Biochemistry and Biophysics 512 (2011) 61–68

metabolism and absence in vertebrates. Although CK and AK displayed homologous amino acid sequences, similar crystal structure and the identical catalytic mechanism [6–8], they have different substrate recognition mechanism and functional type [32,39]. The dimeric structure is necessary for the high activity of CK and dimer AK [21,22], while AKs are typically functional as monomers [6–8]. Previous studies indicated that there is a proportional relationship between the size of the deletion in the guanidino specificity (GS) region and the mass of the guanidine substrate used [23,32,39]. AK has a five amino acid deletion in this region and use relatively large guanidine substrate, while CK has a one-amino-acid deletion [32,39]. Besides, specificity might be mediated not only by steric interactions, but also through the ability of substrates of appropriate length to bridge the active sites, and through interactions at each end, initiating the conformational changes required for activity [32,39]. Moreover, the AKs in the agricultural herbivorous insect and the major indoor allergens insects and pets’ parasites are monomeric and absence in vertebrates [6–8,36–38], while dimer formation is important for CKs to function [21,22]. Those differences were the basis for innovative pesticide compounds specific to AK. This study investigating the amino acid residues involved in the dimer AK activity and substrate synergism provides a clue for designing insecticides specific to monomeric AKs. In conclusion, destruction of inter-subunit hydrogen bonds in Q53 and D57 can cause pronounced loss of activity and structural stability, and distinct substrate synergism alteration. Besides, the partially unfolded state was prone to form insoluble aggregates. Our results suggested that the inter-subunit interactions played a key role in maintaining the dimer AK substrate synergism, activity and structural stability. The difference of inter-subunit interactions and amino acid residues sustaining the stability of the active sites conformation between monomer AKs and the CKs may provide a clue for the development of insecticides specific to agricultural herbivorous and the indoor allergens insects.

Acknowledgments The present investigation was supported by Grants from the National Key Scientific Program of China (2007CB914504) and Program for Changjiang Scholar and Innovative Research Team in University (IRT0635).

References [1] W.R. Ellington, Annu. Rev. Physiol. 63 (2001) 289–325. [2] D.C. Watts, Evolution of phosphagen kinases, in: E. Schoffeniels (ed.), Biochemical Evolution and the Origin of Life North-Holland, Amsterdam, 1971, pp. 150-173. [3] D.C. Watts, Symp. Zool. Soc. Lond. 36 (1975) 105–127. [4] T. Suzuki, T. Furukohri, J. Mol. Biol. 237 (1994) 353–357. [5] K. Tanaka, S. Ichinari, K. Iwanami, S. Yoshimatsu, T. Suzuki, Insect Biochem. Mol. Biol. 37 (2007) 338–345. [6] Q.Y. Wu, F. Li, X.Y. Wang, Insect Biochem. Mol. Biol. 38 (2008) 59–65. [7] J.D. Seals, S.H. Grossman, Comp. Biochem. Physiol. 89B (1988) 701–707. [8] T. Suzuki, Y. Kawasaki, T. Furukohri, Biochem. J. 328 (1997) 301–306. [9] T. Suzuki, M. Kamidochi, N. Inoue, H. Kawamichi, Y. Yazawa, T. Furukohri, W.R. Ellington, Biochem. J. 340 (1999) 671–675. [10] T. Suzuki, Y. Kawasaki, Y. Unemi, Y. Nishimura, T. Soga, K. Kamidochi, Y. Yazawa, T. Furukohri, Biochim. Biophy. Acta 1388 (1998) 253–259. [11] S.R. Reddy, D.C. Watts, Comp. Biochem. Physiol Biochem. Mol. Biol. 108 (1994) 73–78. [12] W.W. Cleland, Methods Enzymol. 63 (1979) 103–138. [13] E.T. Maggio, G.L. Kenyon, J. Biol. Chem. 252 (1977) 1202–1207. [14] N. Fujimoto, K. Tanaka, T. Suzuki, FEBS Lett. 579 (2005) 1688–1692. [15] C.B. Anfinisen, Science 181 (1973) 223–230. [16] R. Jaenicke, Prog. Biophys. Mol. Biol. 71 (1999) 155–241. [17] R. Jaenicke, Prog. Biophys. Mol. Biol. 49 (1987) 117–237. [18] J.R. Garel, Protein Folding, W.H. Freeman and Co., New York, 1992. pp. 405–454. [19] S. Feng, T.J. Zhao, H.M. Zhou, Y.B. Yan, Int. J. Biochem. Cell Biol. 39 (2007) 392–401. [20] J.M. Cox, C.A. Davis, C. Chan, M.J. Jourden, A.D. Jorjorian, M.J. Brym, M.J. Snider, C.L. Borders, P.L. Edmiston, Biochemistry 42 (2003) 1863–1871. [21] Q. Guo, F. Zhao, Z. Guo, X. Wang, J. Biochem. 136 (2004) 49–56. [22] Q. Guo, J.L. Zhang, T.T. Liu, X.C. Wang, Int. J. Biol. Macromol. 41 (2007) 521–528. [23] X. Wu, S. Ye, S. Guo, W. Yan, M. Bartlam, Z. Rao, FASEB J. 24 (2010) 242–252. [24] S.Y. Guo, Z. Guo, Q. Guo, B.Y. Chen, X.C. Wang, Protein Expr. Purif. 29 (2003) 230–234. [25] M.M. Bradford, Anal. Biochem. 72 (1976) 248–254. [26] C.A. Pereira, G.D. Alonso, M.C. Paveto, J. Biol. Chem. 275 (2000) 1495–1501. [27] M. Li, X.Y. Wang, J.G. Bai, Protein Pept. Lett. 13 (2006) 405–410. [28] K.K. Turoverov, S.Y. Haitlina, G.P. Pinaev, FEBS. Lett. 62 (1976) 4–6. [29] N.A. Bushmarina, I.M. Kuznetsova, A.G. Biktashev, K.K. Turoverov, V.N. Uversky, Chembiochemistry 2 (2001) 819–821. [30] W. Humphrey, A. Dalke, K. Schulten, J. Mol. Graphics 14 (1996) 33–38. [31] J.W. Zhang, T.J. Zhao, S.L. Wang, Q. Guo, T.T. Liu, F. Zhao, X.C. Wang, Int. J. Biol. Macromol. 38 (2006) 203–210. [32] K. Uda, T. Suzuki, Protein J. 23 (2004) 53–64. [33] Q. Guo, F. Zhao, S.Y. Guo, X.C. Wang, Biochimie 86 (2004) 379–386. [34] Y. Degani, C. Degani, Biochemistry 18 (1979) 5917–5923. [35] Q. Guo; X.C. Wang, Protein Pept. Lett 12 (2005) 369–373. [36] H. Wang, L. Zhang, Q. Lin, N. Liu, Gene 430 (2009) 38–43. [37] X.Q. Wu, W.J. Zhu, Z.R. Lu, Y. Xia, J.M. Yang, F. Zou, X.Y. Wang, Int. J. Biol. Macromol. 44 (2009) 149–155. [38] A.E. Brown, S.H. Grossman, Arch. Insect Biochem. Physiol. 57 (2004) 166– 177. [39] T. Suzuki, Y. Kawasaki, T. Furukohri, W.R. Ellington, Biochim. Biophys. Acta 1343 (1997) 152–159.