Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 158 (2016) 49–55
Contents lists available at ScienceDirect
Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa
Determination of structural elements on the folding reaction of mnemiopsin by spectroscopic techniques Forough Hakiminia a, Khosrow Khalifeh b, Reza H. Sajedi c, Bijan Ranjbar a,⁎ a b c
Department of Biophysics, Faculty of Biological Sciences, Tarbiat Modares University, P.O. Box 14115-175, Tehran, Iran Department of Biology, Faculty of Sciences, University of Zanjan, P.O. Box 45195-313, Zanjan, Iran Department of Biochemistry, Faculty of Biological Sciences, Tarbiat Modares University, P.O. Box 14115-175, Tehran, Iran
a r t i c l e
i n f o
Article history: Received 8 May 2015 Received in revised form 19 November 2015 Accepted 14 January 2016 Available online 15 January 2016 Keywords: Photoprotein Thermodynamic stability Refolding Unfolding Kinetics Intermediate
a b s t r a c t Mnemiopsin 1 is a member of photoprotein family, made up of 206 amino acid residues. These Ca2+-regulated photoproteins are responsible for light emission in a variety of marine cnidarians and ctenophores. They composed of an apoprotein, a single polypeptide chain of 25 kDa, molecular oxygen and the non-covalently bound chromophore. In this study, we examined whether three mutations, namely R39K, S128G and V183T affect the thermodynamic stability as well as refolding and unfolding kinetics of mnemiopsin 1. Conformational stability measurements using fluorescence and far-UV CD spectroscopies revealed that all variants unfold in multi-step manner in which the secondary and tertiary structures are lost in different steps. However kinetic studies showed that point mutation S128G destabilizes both kinetic intermediate and native conformation; while, these structural elements are stabilized in V183T. We also found that the stability of folded and intermediate states increases in R39K. We concluded that the initial packing of helical segments within the protein structure is more facilitated when Lys with smaller side chain is present in the protein chain. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Ca2+-regulated photoproteins are responsible for light emission in a variety of marine cnidarians and ctenophores. These photoproteins consist of an apoprotein, a single polypeptide chain of 25 kDa, molecular oxygen and the non-covalently bound chromophore called peroxycoelenterazine which is located in the highly hydrophobic cavity. All Ca2+-binding photoproteins have the same compact globular structure containing three “EF-hand” calcium-binding sites [1–3]. Upon binding of Ca2+ ions with the Ca2+-binding EF-hand motif, the tertiary structure of protein changes which results in the oxidation of the non-covalently bound coelenterazine to coelenteramide, yielding carbon dioxide (CO2) and blue light. Aequorin and obelin are two well-known and highly-studied photoproteins isolated from cnidarians Aequorea and Obelia, respectively. However, photoproteins from ctenophores remain largely unexplored [4]. The applications of photoproteins span a wide spectrum including tracking the location and concentration of Ca2+, nucleic acid hybridization assays, investigating the signal transduction pathways, protein– protein interaction, imaging of living cells, and discovery of novel drugs [5–7].
⁎ Corresponding author. E-mail address: [email protected] (B. Ranjbar).
http://dx.doi.org/10.1016/j.saa.2016.01.020 1386-1425/© 2016 Elsevier B.V. All rights reserved.
Photoproteins have attracted much research interest in both basic and diagnostic areas because of their unique biophysical properties such as low background noise, superior detection sensitivity and harmless applications including lack of cellular toxicity, hazard-free handling, and non-invasive nature [8,9]. Mnemiopsin was first isolated and characterized from the luminous ctenophore Mnemiopsis sp. in 1970s [10–12]. Additionally, molecular cloning and expression of cDNA coding for two isoforms of mnemiopsins from Mnemiopsis leidyi and further characterization of related photoproteins have recently been performed [13]. In an attempt to understand the structural properties related to bioluminescence in ctenophore photoproteins, we previously investigated functional properties of a variety of critical residues in mnemiopsin 1 using site directed mutagenesis [14]. It has 206 amino acids and bears high homology (sequence identity 90%) to berovin, as the only structurally determined ctenophore photoprotein (PDB ID: 4MN0) [4,12]. The crystal structure of native aequorin (PDB ID: 1EJ3) also reveals a highly hydrophobic core containing 21 residues that stabilize the chromophore. Three sets of tyrosine, histidine and tryptophan residues form three triads that interact with the chromophore through hydrogen bonds and π–π interactions and are responsible for the stabilizing of coelenterazine but in mnemiopsin tryptophan, methionine and phenylalanine form this triad. Due to the fact that bioluminescence activity is highly dependent on the stability of chromophore within the pocket, bioluminescent properties of photoproteins may be substantially altered by changing these critical residues [15–18].
50
F. Hakiminia et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 158 (2016) 49–55
Nowadays equilibrium unfolding experiments on wild-type (WT) and mutants are widely used for determining the conformational stability of proteins. This approach evaluates how a mutation changes the free energy of unfolding of a protein (ΔΔGequilibrium) [19]. Protein engineering analysis with stopped-flow fluorescence measurements and chevron plot analysis may also be used to characterize transition states and intermediate structures of refolding reaction. These methods can compare the effect of mutations on the conformational stability of native and intermediate states as well as folding rate interpreted in terms of the changes in the free-energy barrier (ΔΔGTS) [20–22]. In this study, we performed thermodynamic stability and kinetics of refolding and unfolding measurements on the WT and the three mutants of mnemiopsin 1. Sequence alignment of mnemiopsin, obelin as well as aequorin indicates that the corresponding residues in the positions 128 and 183 of mnemiopsin in aforementioned proteins are Gly and Thr; respectively. So, in our previous studies, Ser128 and Val183 in mnemiopsin were replaced by Gly and Thr, respectively. In the other mutant, Arg39 was replaced by Lys as another charged residue with smaller side chain. Hence; the selection of mutants is based on the evolutionary conservation of selected residues according to the resulting data of multiple sequence alignment [18]. In comparison with the WT protein, these modifications led to an increase in the activity of R39K and a decrease in the activities of S128G and V183T. The aim of this study is to determine the importance of the above-mentioned residues on the folding pathway and conformational stability of mnemiopsin. According to thermodynamic studies using fluorescence and far-UV CD spectroscopy measurements, we found that all protein variants are unfolded in different steps in urea denaturation experiments so that the tertiary structure disarrays prior to the secondary contacts. Our kinetic data indicate that the stability of intermediate state increases in R39K and V183T mutants, while in S128G variant it is the same as WT protein. The conversion of intermediate state to native structure as the rate limiting step in the refolding reaction is also speeding up in V183T and R39K relative to WT protein. 2. Materials and methods 2.1. Expression, purification and activity determination of His-tagged apomnemiopsin Expression and purification of His-tagged apo-mnemiopsins and its luminescence activity were carried out as described previously [14]. pET28a expression vector containing apo-mnemiopsin 1 gene (GenBank accession No. GQ231544) from M. leidyi was obtained from the Caspian Sea, northern Iran [12] and its mutants were used for overexpression of the apo-photoproteins in Escherichia coli BL21 (DE3). 2.2. Stability measurements Fluorescence and far-UV CD spectroscopies were used for monitoring urea-induced unfolding of mnemiopsin. The buffer used contains 50 mM NaH2Po4, 300 mM NaCl, pH 8.0 and 25 °C. At first, urea stock solutions (0–9 M) were prepared in buffer. Then, 100 μg/ml concentration of protein was incubated at different concentrations of urea (0–8 M). Fluorescence measurement was performed by fluorescence spectroscopy with excitation at 280 nm and emission spectra were recorded between 300 and 450 nm (both slits of excitation and emission were set to 5 nm). For CD measurements, the concentration of protein was 200 μg/ml [23,24]. It is noticeable that agents such as urea at high concentrations absorb too strongly CD data below 210 nm even using cells of short pathlength. This, of course, is not a problem if changes in the CD signals at 222 or 225 nm are used to assess the unfolding of a protein. Hence, using data of 222 nm wavelengths in this study is not problematic for CD investigations [25].
All experiments were also measured by circular dichroism technique at far-UV region (far-UV CD). In these wavelengths, we can monitor the changes of the secondary structure of WT and mutants of mnemiopsin [26]. The analysis of CD equilibrium denaturation curves was the same as fluorescence-based curves. Standard deviations are calculated based on four or five replicates of experiments. Mnemiopsin is a monomer and only one protein concentration (100 μg/ml for fluorescence and 200 μg/ml for CD experiments) was used for the final analysis. However, several concentrations in the range of experimental conditions were tested to verify that the protein does not aggregate at higher protein concentrations. Results show that the equilibrium denaturation curves of different concentrations were superimposable, demonstrating that the stability of protein is independent of protein concentration (from 100 to 200 μg/ml concentration of protein) [27]. The observed sigmoid-like equilibrium curves were fitted by KaleidaGraph analysis software into Eq. (1); considering a two state model [28]: F350 ¼
fðαN þ βN ½UreaÞ þ ðαD þ βD ½UreaÞg expððmDN ð½Urea ½Urea50% ÞÞ=RTÞ f1 þ expððmDN ð½Urea ½Urea50% ÞÞ=RTÞg
ð1Þ where F350 is the fluorescence intensity at 350 nm as a function of [Urea], αN and αD are the intercepts, and βN and βD are the slopes of the baselines. [Urea]50% is the concentration of urea at which half of the protein is denatured, R is the gas constant and T is the temperature in Kelvin. mD − N is defined as a constant that is proportional to difference in the solvent accessible surface area between the native and denatured states. The relationship between the free energy of unfolding in the urea and buffer as well as the concentration of denaturants is given by Eq. (2) [29]: ½Urea
H2 O mDN ½Urea: ΔGDN ¼ ΔGDN
ð2Þ
Moreover, the free energy of unfolding at the absence of urea, ΔGDcan be calculated by Eq. (3):
H2O , N
H2 O ¼ mDN ½Urea50% : ΔGDN
ð3Þ
2.3. Kinetic experiments A Biologic μ-SFM-20 fluorescence detected stopped-flow, equipped with a 0.8 cm cuvette (FC-08) (excitation 280 nm, emission 350 nm) was used for kinetic measurements. For refolding studies, unfolded protein in 50 mM NaH2Po4, 300 mM NaCl, pH 8.0 containing high concentration of urea was diluted with refolding buffer containing different concentrations of urea ranging from 0 to 5 M. For unfolding experiments, protein was unfolded by mixing one volume of protein in 50 mM NaH 2Po 4, 300 mM NaCl, pH 8.0 with 6 volumes of buffer containing different concentrations of urea. Kinetic traces were analyzed by fitting to exponential function of Eq. (4) using Biokine software (Ver. V4.49-1): N
FðtÞ ¼ at þ b þ ∑ Ci expð ki tÞ
ð4Þ
i
where F(t) is the fluorescence signal at time t, Ci is the amplitude, ki is the rate constant, a is the slope of the drift and b is the offset of kinetic curve corresponding to the baseline. According to the biological meaning and the accuracy of fitting, kinetic traces may be fitted to a single or double exponential function by using simplex method and setting N = 1 or N = 2 into Eq. (4), respectively [19].
F. Hakiminia et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 158 (2016) 49–55
Kinetic data as the logarithms of rate constants for unfolding and refolding versus urea concentration (chevron plots) displayed curved arms in refolding region and at low [urea]. We analyzed these plots in two different models. In the first model of analysis, we assumed two-state folding in which only 2 states accumulate significantly (Scheme 1): kNU
N→ ← U:
ðScheme 1Þ
51
that mnemiopsin 1 has high homology with berovin (sequence identity 90%) [14]. So the structure of this protein (PDB ID: 4MN0) has been used as template for constructing 3D structural models of mnemiopsin 1. Additionally, aequorin structural information (PDB ID: 1EJ3) was also used in order to insert a coelenterazine molecule within the constructed models [33] and the quality of the models was evaluated by SaliLab Model Evaluation Server at http://modbase.compbio. ucsf.edu/evaluation/. [34–36].
kUN
3. Results and discussion In this model, we tried to fit chevron plots to the following equation: H O logðkUN2
logkobs ¼
mUN ½Urea mNU ½Urea H O exp þ kNU2 exp : RT RT ð5Þ
In the second model of analysis, we used 3-state folding mechanism to describe the behavior of proteins as the following form: kNI
kIU
kIN
kUI
→ N→ ← I ← U:
ðScheme 2Þ
In this model we assume that unstable on-pathway intermediates (I) are populated in the refolding arm of chevron plot. In our analysis we also assumed that the formation of I is essential for achieving the native state (N), and the formation of I (kUI) occurs much more quickly than the rate-limiting step to the native state formation (kIN). On the other hand, a rapid equilibrium between the unfolded state (U) and I is assumed, leading to the accumulation of intermediate state with an equilibrium constant KUI = [I] / [U], which was used in Eq. (6) in order to fit the experimental data of chevron plots to the three state on-pathway model [30]. 3 mUI ½Urea H2 O KUI exp 7 6 RT 7 ¼ log6 4 mUI ½Urea 5 H2 O 1 þ KUI exp " !RT !# ‡ m ½ Urea m‡NI ½Urea H2 O H2 O IN kIN exp þ kNI exp RT RT 2
logkobs
ð6Þ
where KUI is the equilibrium constant between the unfolded and intermediate states. kij is the rate constant of conversion i to j extrapolated to water and mij is the difference in exposed surface area between i and j states. Eq. (7) was used to calculate the free energy of transition state between I and N: kIN ¼
kB T
. h
exp ΔG‡I RT :
3.1. Description of mutations The structure of mnemiopsin 1 was modeled with MODELLER program using berovin (PDB ID: 4MN0) as template (sequence identity 90%). Fig. 1 shows a ribbon diagram of mnemiopsin 1 in which the position of mutations is shown. The coelenterazine binding position is specific for this family of proteins. Hence, the only charged residue in the cavity (Arg39) was replaced by Lys as another charged residue (R39K mutation). In V183T mutation, Val183 was replaced by Thr. According to bioinformatic studies Ser128 is surrounded by water molecules and participates in the formation of hydrogen bonds. In S128G mutant, this residue was replaced by Gly. 3.2. Urea-induced unfolding monitored by fluorescence and CD spectroscopies The stability of the secondary and tertiary structures of mnemiopsin and its mutants was investigated using a combination of spectroscopic methods. Intrinsic fluorescence spectroscopy provides an excellent probe for the tertiary structure of a protein and circular dichroism at far UV region can give information concerning the secondary structure of a protein. Furthermore, we used fluorescence and far-UV CD spectroscopies for monitoring the changes in the secondary and tertiary structures of all variants in unfolding equilibrium experiments. Fig. 2 shows the equilibrium denaturation curves of WT and mutant proteins monitored by fluorescence and far-UV CD spectroscopies. Thermodynamic parameters obtained from analysis of these curves are given in Table 1. Comparison of ΔG(H2O) as a measure of conformational stability as well as [Urea]50% shows that the resulting data of CD and fluorescence
ð7Þ
Considering all reactions of Scheme 2 and the effect of different rates on populating the I molecules at the pre-equilibrium state and by regarding the conversion of I to N as the rate limiting step, the free energy of intermediate state as a reflection of its population is determined by following thermodynamic equation: ΔGUI ¼ RTlnKUI :
ð8Þ
At the end, all data can be normalized by considering the unfolded state as reference [31,32]. 2.4. Bioinformatic studies The MODELLER program (Ver. 9.15) was used for constructing 3D structural models of mnemiopsin and mutants. According to Similarity search against Protein Data Bank using BLAST program, it was shown
Fig. 1. Ribbon diagram of the structural model of mnemiopsin 1 and representation of interactions of R39, S128 and V183. In three dimensional structure of WT mnemiopsin, triad of Phe, Met and Trp forms a part of the substrate binding site which is close to Arg39 and is shown in figure by dotted surface. The Trp residues as chromophores in fluorescence studies are also shown.
52
F. Hakiminia et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 158 (2016) 49–55
Fig. 2. The equilibrium denaturation curves of wild-type and mutants of mnemiopsin monitored by different spectroscopic probes. (a) Normalized fluorescence intensity at 350 nm; (b) normalized far-UV CD signal at 222 nm. The continuous line is a theoretical curve upon fitting experimental data to Eq. (1) with KaleidaGraph analysis software. The equilibrium parameters obtained are given in Table 1.
measurements are different for all variants (Table 1). On the other hand, the secondary structure of protein is more stable than the tertiary one; for example, the value of ΔG(H2O) of CD measurement in WT protein is 3.72 kJ/mol higher than that of the fluorescence one. The differences in far UV CD and fluorescence data suggest that the secondary and tertiary structures of all variants denature in different steps in which the tertiary structure of protein denatured first, followed by the loss of the secondary structure. According to equilibrium data in Table 1, both the secondary and tertiary structures of S128G variant are destabilized in comparison to WT protein. However, the conformational stability of V183T increases when compared with WT protein. Substitution of Arg by Lys in R39k mutant brought about changes on the stability of the secondary and tertiary structures of protein toward stabilization. Comparison of [Urea]50% and thermodynamic m-values in Table 1 indicates that the effect of replacing Val by Thr on the conformational stability of protein is related to both m-value and [Urea]50%. The m-value has been shown to be proportional to the difference in the solvent accessible surface area (SASA) between the denatured and folded states of protein [9]. Hence, the compactness of more stabilized V183T in the folded state may be greater than that of the WT protein. In the case of S128G mutant, decreasing the m-value may be attributed to the increasing solvent accessible surface area of folded state; this is due to the fact that this variant is destabilized compared to the WT protein. This observation may also be related to the solvation effects originated from free energy transfer of Gly from hydrophobic core of protein to urea during unfolding of protein from the folded state.
3.3. Kinetics of folding and unfolding The kinetics of unfolding and refolding can be used for detecting transition states as well as folding intermediates which are not stable enough to be observed under equilibrium conditions. For this purpose, the rate constants for folding and unfolding at different concentrations of urea were determined by recording the intrinsic Trp fluorescence as a function of time. Fig. 3 shows a kinetic trace illustrating the unfolding time course of WT protein measured by stopped-flow fluorescence spectroscopy as obtained by rapid mixing of one volume of protein in buffer with 6 volumes of buffer containing 3 M urea at pH 8.0 and 25 °C. The final concentration of urea was 2.57 M. An excitation wavelength of 280 nm was used, and change in fluorescence emission was monitored at 350 nm. Solid line represents a curve of best fit to the data and the fit quality is represented by the residual data in the upper panel; see “Materials and methods” for a complete experimental description. Upon fitting experimental kinetic data to Eq. (4), describing exponential function, the rate constants were obtained from stopped-flow fluorescence data. Plots of the natural logarithm of rate constants versus urea concentration known as chevron plot can be fitted globally to suitable equation for obtaining kinetic data of folding and unfolding reactions. As shown in Fig. 4 for WT protein, the chevron plots were first fitted to Eq. (5) describing a two state model as depicted in Scheme 1. It is evident from Fig. 4 that refolding arm of the chevron plot shows a deviation from linearity which becomes evident at low urea concentrations,
Table 1 Thermodynamic parameters for wild-type and mutants of mnemiopsin obtained by analysis of equilibrium denaturation curves of Fig. 2; a and b, using Eq. (1). Variants
WT R39K S128G V183T a b
Equilibrium parameters of fluorescence measurementsa,b
Equilibrium parameters of CD measurementsa,b
m-Valuea
[Urea]50%
ΔG(H2O)a
m-Valuea
[Urea]50%
ΔG(H2O)a
4.14 ± 0.005 4.19 ± 0.008 3.36 ± 0.003 4.45 ± 0.003
3.43 ± 0.004 3.79 ± 0.007 3.71 ± 0.003 3.56 ± 0.002
14.20 ± 0.004 15.90 ± 0.006 12.46 ± 0.004 15.86 ± 0.004
4.93 ± 0.003 5.08 ± 0.005 4.70 ± 0.009 5.11 ± 0.008
3.63 ± 0.002 3.87 ± 0.004 3.48 ± 0.007 3.80 ± 0.006
17.92 ± 0.003 19.65 ± 0.007 16.34 ± 0.008 19.40 ± 0.006
Standard deviations are calculated based on four or five replicates of experiments. m-Value in kJ/mol−2; [Urea]50% in M and ΔG(H2O) in kJ/mol.
F. Hakiminia et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 158 (2016) 49–55
53
Fig. 3. Representative kinetic trace illustrating the unfolding time course of wild-type protein measured by stopped-flow fluorescence spectroscopy.
showing a clear systematic deviation from the expected values for a two-state behavior. Using this model for fitting, we found that the rate H2 O 2 O constant of unfolding and refolding at 0 M Urea (kH NU and kUN ) was imprecise because of the paucity of data in the refolding limb and the very long extrapolation to 0 M urea. So, the rollover in the refolding arm of chevron plot may not be accounted for by Hammond behavior within a two state model [37]. The low quality of fitting to a two state model indicates that chevron plots should be fitted to other models rather than the two state one. In a three-state model, we assumed that the chevron plot curvature at low urea concentrations indicates the transient accumulation of a partially folded species with a partially non-native topology, which is
Fig. 4. Experimental data of kinetic studies of WT protein and fit of chevron plot to Eq. (5) describing a two-state model as depicted in Scheme 1. Continuous line represents best fit of experimental data to Eq. (5). A deviation from linearity becomes evident, particularly at low urea concentrations. The results indicate that other models rather than the two-state one should be considered.
less sensitive to urea. In other words, there is a molecular species in this region which is relatively less compact and less sensitive to urea relative to more compact folded state. Hence, the observed chevron plots were fitted to a three-state model (Scheme 2 and Fig. 5) assuming the presence of on-pathway intermediate (see Materials and methods). Fig. 6 shows the results of fitting experimental data of chevron plots to Eq. (6) describing a three-state model. Kinetic parameters of all variants are shown in Table 2. According to the resulting data of Table 2 and comparison of equilibrium constants for the formation of intermediate state from unfolded structure (KUI), it reveals that accumulation of intermediate state during refolding reaction is greater in R39K and V183T relative to the WT and S128G variants. Furthermore, m-values of intermediate formation from the unfolded state for the WT protein is larger than those for mutants, indicating that the change in surface area should be larger when starting from the unfolded structure to the partially folded intermediate state in the WT protein. Larger m-value in the WT protein may be attributed to more extended polypeptide chain in the unfolded state or more compactness of the intermediate state. According to our model described in the Materials and methods, the formation of the native structure from the intermediate state is considered as the rate limiting step in protein folding. In other words, the formation of native structure from the intermediate state (kIN) is
Fig. 5. Free energy profile for the folding of protein according to Scheme 2 and via one major intermediate (I) and transition state (TS). The right hand side of the figure as the energy level of N is obtained from fluorescence-based equilibrium data of Table 1. The energy levels of I and TS are obtained from kinetic studies.
54
F. Hakiminia et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 158 (2016) 49–55
Fig. 6. Chevron plots showing the urea dependence of rate constants for unfolding and refolding of wild-type and mutants of mnemiopsin. The experiments at the left-hand side of the curves consist of protein solutions in high concentrations of urea diluted into refolding buffer in the stopped-flow fluorescence apparatus. In this region, kobs is the rate constant for refolding. At the right-hand side of the curves, folded protein in aqueous buffer was mixed with urea solutions, and so the kinetics of unfolding is recorded. In protein unfolding/ refolding studies, the kinetic parameters are reported at 0 M concentration of denaturant. This condition is a reference state for comparing the parameters. However, doing unfolding and/or refolding experiments at 0 M urea is impossible, hence the parameters at 0 M were obtained by extrapolation of experimental data to 0 M urea which is carried out by mathematical modeling and fitting experimental data to appropriate equations. The data of chevron plots were fitted to a three-state on-pathway intermediate model (Scheme 2 and Fig. 5). Continuous line represents best fit of experimental data to Eq. (6).
accompanied by a further molecular rearrangement of molecule and the values of rate constants of reverse reaction (kNI) are negligible when compared with those of forward reaction (kIN). Our data in Table 2 suggest that the rate limiting step in all mutants is accelerated relative to WT protein and the largest differences in this process were observed in V183T and R39K mutants. For better understanding of kinetic and thermodynamic consequences of mutations on the stability and kinetics of refolding and unfolding, stability measurements (fluorescence based data in equilibrium unfolding experiments) and kinetic data were normalized by considering the unfolded state as the reference and the resulting data were presented as difference free energy diagrams (Fig. 7). To do this, rate constants of reaction (kIN) should be converted to the free energy change of transition state using Eq. (7). Similarly, by applying Eq. (8) free energy change of intermediate states as a measure of the population of I molecules or intermediate stability at pre-equilibrium can be achieved.
Structural examination of protein indicates that the helix containing Arg39 is short and it seems that it loses few contacts upon mutation to Lys, which is physico-chemically similar to Arg. Besides, thermodynamic data suggest that replacement of Arg with Lys has similar effect on the conformational stability of the secondary and tertiary structures of protein. Yet the higher stability of intermediate state in R39K in comparison to WT and S128G suggests that initial packing of helical segments within the protein structure is more facilitated when Lys with smaller side chain is present in the protein chain. The region of protein containing Val183 is a loop and usually the small number of contacts is made by loops on the structure of proteins. Surprisingly, the free energy change of unfolding in V183T variant (Table 1) toward stabilization suggests that new interactions may be formed upon mutation of Val to Thr. As indicated in Table 1, protein is destabilized upon mutation of Ser to Gly. Examination of structural models indicates that upon replacement of Ser by Gly in S128G variant, the hydrogen bonds between main chain of Ser128 and side chain of
Table. 2 Kinetic and thermodynamic parameters for the wild-type and mutant proteins obtained from chevron plot analysis.
WT R39K S128G V183T
2O KH UI
2O kH IN
2O kH NI
m‡IN
mUI
m‡NI
123 ± 34 600 ± 66 116 ± 19 438 ± 102
0.56 ± 0.20 3.20 ± 0.60 1.67 ± 0.44 5.89 ± 1.46
0.03 ± 0.001 0.03 ± 0.001 0.06 ± 0.003 0.01 ± 0.0005
−1.95 ± 0.40 −0.04 ± 0.18 −1.10 ± 0.29 −0.08 ± 0.25
4.24 ± 0.36 3.52 ± 0.13 3.59 ± 0.22 3.10 ± 0.16
0.32 ± 0.005 0.34 ± 0.006 0.23 ± 0.005 0.44 ± 0.006
−1 H2O 2O is the rate The chevron plots were fitted to a three state on-pathway intermediate model. KH UI is the equilibrium constant between the unfolded and intermediate states. kij in S constant regarding conversion of i-state to j-state. mij refers to differences in surface area between i and j states.
F. Hakiminia et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 158 (2016) 49–55
55
References
Fig. 7. Difference free energy diagrams. ΔΔG(H2O) is the absolute value of free energy difference of structural elements between WT and all protein variants, so the difference energy for WT protein becomes zero for all structural states.
Leu182 as well as main chain of Val183 with side chain of Ser128 are lost. So, decreasing the conformational stability of S128G mutant may be a consequence of removing interactions made by Ser as well as increasing the flexibility of protein upon replacement of Ser by Gly which has more conformational space. According to Table 2, the effect of S128G mutant on the stability of intermediate state is negligible. In summary, thermodynamic and kinetic data together indicate that all mutants follow essentially the same reaction scheme as WT protein for folding from the urea-denatured state. Also, overall conformational stability of intermediate and folded states of protein increases in both V183T and R39K.
Acknowledgments Financial support for this work was provided by the Research Council of Tarbiat Modares University (2244604). There is no conflict of interest.
[1] Y. Takenaka, A. Yamaguchi, N. Tsuruoka, M. Torimura, T. Cojobori, Y. Shingeri, Mol. Biol. Evol. 29 (2006) 1669–1681. [2] Z.J. Liu, G.A. Stepanyuk, E.S. Vysotski, J. Lee, S.V. Markova, N.P. Malikova, B.C. Wang, Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 2570–2575. [3] L. Deng, E.S. Vysotski, S.V. Markova, Z.J. Liu, J. Lee, J. Rose, B.C. Wang, Protein Sci. 14 (2005) 663–675. [4] G.A. Stepanyuk, Z.J. Liu, L.P. Burakova, J. Lee, J. Rose, E.S. Vysotski, B.C. Wang, Biochim. Biophys. Acta 1834 (2013) 2139–2146. [5] J.R. Blinks, Environ. Health Perspect. 84 (1990) 75–81. [6] C.J. Lewis, S. Daunert, Fresenius J. Anal. Chem. 366 (2000) 760–768. [7] Y. Ohmiya, T. Hirano, Chem. Biol. 3 (1996) 337–347. [8] C.H. Contag, S.D. Spilman, P.R. Contag, M. Oshiro, B. Eames, P. Dennery, D.K. Stevenson, D.A. Benaron, Photochem. Photobiol. 66 (1997) 523–531. [9] E. Gonzalez, R.Y. Tslen, Chem. Biol. 4 (1997) 269–277. [10] W.W. Ward, H.H. Seliger, Biochemistry 13 (1974) 1500–1510. [11] W.W. Ward, H.H. Seliger, Biochemistry 13 (1974) 1491–1499. [12] M.R. Aghamaali, V. Jafarian, R. Sariri, M. Molakarimi, B. Rasti, M. Taghdir, R.H. Sajedi, S. Hosseinkhani, Protein J. 30 (2011) 566–574. [13] V. Jafarian, R. Sariri, S. Hosseinkhani, M.R. Aghamaali, R.H. Sajedi, M. Taghdir, S. Hassannia, Biochem. Biophys. Res. Commun. 413 (2011) 164–170. [14] A. Mahdavi, R.H. Sajedi, S. Hosseinkhani, M. Taghdir, R. Sariri, Photochem. Photobiol. Sci. 12 (2013) 467–478. [15] J.F. Head, S. Inouye, K. Teranishi, O. Shimomura, Nature 405 (2000) 372–376. [16] E. Dikici, X. Qu, L. Rowe, L. Millner, C. Logue, S.K. Deo, M. Ensor, S. Daunert, Protein Eng. Des. Sel. 22 (4) (2009) 243–248. [17] S. Toma, K.T. Chong, A. Nakagawa, K. Teranishi, S. Inouye, O. Shimomura, Protein Sci. 14 (2) (2005) 409–416. [18] A. Mahdavi, R.H. Sajedi, S. Hosseinkhani, M. Taghdir, Photochem. Photobiol. Sci. (2014) 792–800. [19] A.R. Fersht, A. Matouschek, L. Serrano, J. Mol. Biol. 224 (1992) 805–818. [20] H.S. Chanand, K. Dill, Proteins 30 (1998) 2–33. [21] S. Blond, M.E. Goldberg, Protein 1 (1986) 247–255. [22] A.R. Fersht, A. Matouschek, L. Serrano, J. Mol. Biol. 224 (1992) 771–782. [23] F. Hillger, D. Nettels, S. Dorsch, B. Schuler, J. Fluoresc. 17 (2007) 759–765. [24] M. Lindgren, K. Sörgjerdy, P. Hammarström, Biophys. J. 88 (2005) 4200–4212. [25] S.M. Kelly, T.J. Jess, N.C. Price, Biochim. Biophys. Acta 1751 (2005) 119–139. [26] P. Gill, B. Ranjbar, Chem. Biol. Drug Des. 74 (2009) 101–120. [27] J. Walters, S.L. Milam, A.C. Clark, Methods Enzymol. 455 (2009) 1–39. [28] C.N. Pace, Methods Enzymol. 131 (1986) 266–279. [29] J.K. Myers, C.N. Pace, J.M. Scholtz, Protein Sci. 4 (1995) 2138–2148. [30] A. Zarrine-Afsara, A.R. Davidson, Methods 34 (2004) 41–50. [31] H. Eyring, Chem. Rev. 17 (1935) 65–77. [32] A.R. Fersht, A. Matouschek, M. Bycroft, J.T. Kellis, L. Serrano, Pure Appl. Chem. 63 (2) (1991) 187–194. [33] N. Eswar, M.A. Marti-Renom, B. Webb, M.S. Madhusudhan, D. Eramian, M. Shen, U. Pieper, A. Sali, John Wiley& Sons, Inc., Supplement 15 (2006) 5.6.1–5.6.30. [34] D. Eramian, N. Eswar, M.Y. Shen, A. Sali, Protein Sci. 17 (2008) 1881–1893. [35] M.Y. Shen, A. Sali, Protein Sci. 15 (2006) 2507–2524. [36] F. Melo, A. Sanchez, A. Sali, Protein Sci. 11 (2002) 430–448. [37] A. Matouschek, D.E. Otzen, L.S. Itzhaki, S.E. Jackson, A.R. Fersht, Biochemistry 17 (41) (1995) 13356–13362.
Contents lists available at ScienceDirect
Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa
Determination of structural elements on the folding reaction of mnemiopsin by spectroscopic techniques Forough Hakiminia a, Khosrow Khalifeh b, Reza H. Sajedi c, Bijan Ranjbar a,⁎ a b c
Department of Biophysics, Faculty of Biological Sciences, Tarbiat Modares University, P.O. Box 14115-175, Tehran, Iran Department of Biology, Faculty of Sciences, University of Zanjan, P.O. Box 45195-313, Zanjan, Iran Department of Biochemistry, Faculty of Biological Sciences, Tarbiat Modares University, P.O. Box 14115-175, Tehran, Iran
a r t i c l e
i n f o
Article history: Received 8 May 2015 Received in revised form 19 November 2015 Accepted 14 January 2016 Available online 15 January 2016 Keywords: Photoprotein Thermodynamic stability Refolding Unfolding Kinetics Intermediate
a b s t r a c t Mnemiopsin 1 is a member of photoprotein family, made up of 206 amino acid residues. These Ca2+-regulated photoproteins are responsible for light emission in a variety of marine cnidarians and ctenophores. They composed of an apoprotein, a single polypeptide chain of 25 kDa, molecular oxygen and the non-covalently bound chromophore. In this study, we examined whether three mutations, namely R39K, S128G and V183T affect the thermodynamic stability as well as refolding and unfolding kinetics of mnemiopsin 1. Conformational stability measurements using fluorescence and far-UV CD spectroscopies revealed that all variants unfold in multi-step manner in which the secondary and tertiary structures are lost in different steps. However kinetic studies showed that point mutation S128G destabilizes both kinetic intermediate and native conformation; while, these structural elements are stabilized in V183T. We also found that the stability of folded and intermediate states increases in R39K. We concluded that the initial packing of helical segments within the protein structure is more facilitated when Lys with smaller side chain is present in the protein chain. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Ca2+-regulated photoproteins are responsible for light emission in a variety of marine cnidarians and ctenophores. These photoproteins consist of an apoprotein, a single polypeptide chain of 25 kDa, molecular oxygen and the non-covalently bound chromophore called peroxycoelenterazine which is located in the highly hydrophobic cavity. All Ca2+-binding photoproteins have the same compact globular structure containing three “EF-hand” calcium-binding sites [1–3]. Upon binding of Ca2+ ions with the Ca2+-binding EF-hand motif, the tertiary structure of protein changes which results in the oxidation of the non-covalently bound coelenterazine to coelenteramide, yielding carbon dioxide (CO2) and blue light. Aequorin and obelin are two well-known and highly-studied photoproteins isolated from cnidarians Aequorea and Obelia, respectively. However, photoproteins from ctenophores remain largely unexplored [4]. The applications of photoproteins span a wide spectrum including tracking the location and concentration of Ca2+, nucleic acid hybridization assays, investigating the signal transduction pathways, protein– protein interaction, imaging of living cells, and discovery of novel drugs [5–7].
⁎ Corresponding author. E-mail address: [email protected] (B. Ranjbar).
http://dx.doi.org/10.1016/j.saa.2016.01.020 1386-1425/© 2016 Elsevier B.V. All rights reserved.
Photoproteins have attracted much research interest in both basic and diagnostic areas because of their unique biophysical properties such as low background noise, superior detection sensitivity and harmless applications including lack of cellular toxicity, hazard-free handling, and non-invasive nature [8,9]. Mnemiopsin was first isolated and characterized from the luminous ctenophore Mnemiopsis sp. in 1970s [10–12]. Additionally, molecular cloning and expression of cDNA coding for two isoforms of mnemiopsins from Mnemiopsis leidyi and further characterization of related photoproteins have recently been performed [13]. In an attempt to understand the structural properties related to bioluminescence in ctenophore photoproteins, we previously investigated functional properties of a variety of critical residues in mnemiopsin 1 using site directed mutagenesis [14]. It has 206 amino acids and bears high homology (sequence identity 90%) to berovin, as the only structurally determined ctenophore photoprotein (PDB ID: 4MN0) [4,12]. The crystal structure of native aequorin (PDB ID: 1EJ3) also reveals a highly hydrophobic core containing 21 residues that stabilize the chromophore. Three sets of tyrosine, histidine and tryptophan residues form three triads that interact with the chromophore through hydrogen bonds and π–π interactions and are responsible for the stabilizing of coelenterazine but in mnemiopsin tryptophan, methionine and phenylalanine form this triad. Due to the fact that bioluminescence activity is highly dependent on the stability of chromophore within the pocket, bioluminescent properties of photoproteins may be substantially altered by changing these critical residues [15–18].
50
F. Hakiminia et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 158 (2016) 49–55
Nowadays equilibrium unfolding experiments on wild-type (WT) and mutants are widely used for determining the conformational stability of proteins. This approach evaluates how a mutation changes the free energy of unfolding of a protein (ΔΔGequilibrium) [19]. Protein engineering analysis with stopped-flow fluorescence measurements and chevron plot analysis may also be used to characterize transition states and intermediate structures of refolding reaction. These methods can compare the effect of mutations on the conformational stability of native and intermediate states as well as folding rate interpreted in terms of the changes in the free-energy barrier (ΔΔGTS) [20–22]. In this study, we performed thermodynamic stability and kinetics of refolding and unfolding measurements on the WT and the three mutants of mnemiopsin 1. Sequence alignment of mnemiopsin, obelin as well as aequorin indicates that the corresponding residues in the positions 128 and 183 of mnemiopsin in aforementioned proteins are Gly and Thr; respectively. So, in our previous studies, Ser128 and Val183 in mnemiopsin were replaced by Gly and Thr, respectively. In the other mutant, Arg39 was replaced by Lys as another charged residue with smaller side chain. Hence; the selection of mutants is based on the evolutionary conservation of selected residues according to the resulting data of multiple sequence alignment [18]. In comparison with the WT protein, these modifications led to an increase in the activity of R39K and a decrease in the activities of S128G and V183T. The aim of this study is to determine the importance of the above-mentioned residues on the folding pathway and conformational stability of mnemiopsin. According to thermodynamic studies using fluorescence and far-UV CD spectroscopy measurements, we found that all protein variants are unfolded in different steps in urea denaturation experiments so that the tertiary structure disarrays prior to the secondary contacts. Our kinetic data indicate that the stability of intermediate state increases in R39K and V183T mutants, while in S128G variant it is the same as WT protein. The conversion of intermediate state to native structure as the rate limiting step in the refolding reaction is also speeding up in V183T and R39K relative to WT protein. 2. Materials and methods 2.1. Expression, purification and activity determination of His-tagged apomnemiopsin Expression and purification of His-tagged apo-mnemiopsins and its luminescence activity were carried out as described previously [14]. pET28a expression vector containing apo-mnemiopsin 1 gene (GenBank accession No. GQ231544) from M. leidyi was obtained from the Caspian Sea, northern Iran [12] and its mutants were used for overexpression of the apo-photoproteins in Escherichia coli BL21 (DE3). 2.2. Stability measurements Fluorescence and far-UV CD spectroscopies were used for monitoring urea-induced unfolding of mnemiopsin. The buffer used contains 50 mM NaH2Po4, 300 mM NaCl, pH 8.0 and 25 °C. At first, urea stock solutions (0–9 M) were prepared in buffer. Then, 100 μg/ml concentration of protein was incubated at different concentrations of urea (0–8 M). Fluorescence measurement was performed by fluorescence spectroscopy with excitation at 280 nm and emission spectra were recorded between 300 and 450 nm (both slits of excitation and emission were set to 5 nm). For CD measurements, the concentration of protein was 200 μg/ml [23,24]. It is noticeable that agents such as urea at high concentrations absorb too strongly CD data below 210 nm even using cells of short pathlength. This, of course, is not a problem if changes in the CD signals at 222 or 225 nm are used to assess the unfolding of a protein. Hence, using data of 222 nm wavelengths in this study is not problematic for CD investigations [25].
All experiments were also measured by circular dichroism technique at far-UV region (far-UV CD). In these wavelengths, we can monitor the changes of the secondary structure of WT and mutants of mnemiopsin [26]. The analysis of CD equilibrium denaturation curves was the same as fluorescence-based curves. Standard deviations are calculated based on four or five replicates of experiments. Mnemiopsin is a monomer and only one protein concentration (100 μg/ml for fluorescence and 200 μg/ml for CD experiments) was used for the final analysis. However, several concentrations in the range of experimental conditions were tested to verify that the protein does not aggregate at higher protein concentrations. Results show that the equilibrium denaturation curves of different concentrations were superimposable, demonstrating that the stability of protein is independent of protein concentration (from 100 to 200 μg/ml concentration of protein) [27]. The observed sigmoid-like equilibrium curves were fitted by KaleidaGraph analysis software into Eq. (1); considering a two state model [28]: F350 ¼
fðαN þ βN ½UreaÞ þ ðαD þ βD ½UreaÞg expððmDN ð½Urea ½Urea50% ÞÞ=RTÞ f1 þ expððmDN ð½Urea ½Urea50% ÞÞ=RTÞg
ð1Þ where F350 is the fluorescence intensity at 350 nm as a function of [Urea], αN and αD are the intercepts, and βN and βD are the slopes of the baselines. [Urea]50% is the concentration of urea at which half of the protein is denatured, R is the gas constant and T is the temperature in Kelvin. mD − N is defined as a constant that is proportional to difference in the solvent accessible surface area between the native and denatured states. The relationship between the free energy of unfolding in the urea and buffer as well as the concentration of denaturants is given by Eq. (2) [29]: ½Urea
H2 O mDN ½Urea: ΔGDN ¼ ΔGDN
ð2Þ
Moreover, the free energy of unfolding at the absence of urea, ΔGDcan be calculated by Eq. (3):
H2O , N
H2 O ¼ mDN ½Urea50% : ΔGDN
ð3Þ
2.3. Kinetic experiments A Biologic μ-SFM-20 fluorescence detected stopped-flow, equipped with a 0.8 cm cuvette (FC-08) (excitation 280 nm, emission 350 nm) was used for kinetic measurements. For refolding studies, unfolded protein in 50 mM NaH2Po4, 300 mM NaCl, pH 8.0 containing high concentration of urea was diluted with refolding buffer containing different concentrations of urea ranging from 0 to 5 M. For unfolding experiments, protein was unfolded by mixing one volume of protein in 50 mM NaH 2Po 4, 300 mM NaCl, pH 8.0 with 6 volumes of buffer containing different concentrations of urea. Kinetic traces were analyzed by fitting to exponential function of Eq. (4) using Biokine software (Ver. V4.49-1): N
FðtÞ ¼ at þ b þ ∑ Ci expð ki tÞ
ð4Þ
i
where F(t) is the fluorescence signal at time t, Ci is the amplitude, ki is the rate constant, a is the slope of the drift and b is the offset of kinetic curve corresponding to the baseline. According to the biological meaning and the accuracy of fitting, kinetic traces may be fitted to a single or double exponential function by using simplex method and setting N = 1 or N = 2 into Eq. (4), respectively [19].
F. Hakiminia et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 158 (2016) 49–55
Kinetic data as the logarithms of rate constants for unfolding and refolding versus urea concentration (chevron plots) displayed curved arms in refolding region and at low [urea]. We analyzed these plots in two different models. In the first model of analysis, we assumed two-state folding in which only 2 states accumulate significantly (Scheme 1): kNU
N→ ← U:
ðScheme 1Þ
51
that mnemiopsin 1 has high homology with berovin (sequence identity 90%) [14]. So the structure of this protein (PDB ID: 4MN0) has been used as template for constructing 3D structural models of mnemiopsin 1. Additionally, aequorin structural information (PDB ID: 1EJ3) was also used in order to insert a coelenterazine molecule within the constructed models [33] and the quality of the models was evaluated by SaliLab Model Evaluation Server at http://modbase.compbio. ucsf.edu/evaluation/. [34–36].
kUN
3. Results and discussion In this model, we tried to fit chevron plots to the following equation: H O logðkUN2
logkobs ¼
mUN ½Urea mNU ½Urea H O exp þ kNU2 exp : RT RT ð5Þ
In the second model of analysis, we used 3-state folding mechanism to describe the behavior of proteins as the following form: kNI
kIU
kIN
kUI
→ N→ ← I ← U:
ðScheme 2Þ
In this model we assume that unstable on-pathway intermediates (I) are populated in the refolding arm of chevron plot. In our analysis we also assumed that the formation of I is essential for achieving the native state (N), and the formation of I (kUI) occurs much more quickly than the rate-limiting step to the native state formation (kIN). On the other hand, a rapid equilibrium between the unfolded state (U) and I is assumed, leading to the accumulation of intermediate state with an equilibrium constant KUI = [I] / [U], which was used in Eq. (6) in order to fit the experimental data of chevron plots to the three state on-pathway model [30]. 3 mUI ½Urea H2 O KUI exp 7 6 RT 7 ¼ log6 4 mUI ½Urea 5 H2 O 1 þ KUI exp " !RT !# ‡ m ½ Urea m‡NI ½Urea H2 O H2 O IN kIN exp þ kNI exp RT RT 2
logkobs
ð6Þ
where KUI is the equilibrium constant between the unfolded and intermediate states. kij is the rate constant of conversion i to j extrapolated to water and mij is the difference in exposed surface area between i and j states. Eq. (7) was used to calculate the free energy of transition state between I and N: kIN ¼
kB T
. h
exp ΔG‡I RT :
3.1. Description of mutations The structure of mnemiopsin 1 was modeled with MODELLER program using berovin (PDB ID: 4MN0) as template (sequence identity 90%). Fig. 1 shows a ribbon diagram of mnemiopsin 1 in which the position of mutations is shown. The coelenterazine binding position is specific for this family of proteins. Hence, the only charged residue in the cavity (Arg39) was replaced by Lys as another charged residue (R39K mutation). In V183T mutation, Val183 was replaced by Thr. According to bioinformatic studies Ser128 is surrounded by water molecules and participates in the formation of hydrogen bonds. In S128G mutant, this residue was replaced by Gly. 3.2. Urea-induced unfolding monitored by fluorescence and CD spectroscopies The stability of the secondary and tertiary structures of mnemiopsin and its mutants was investigated using a combination of spectroscopic methods. Intrinsic fluorescence spectroscopy provides an excellent probe for the tertiary structure of a protein and circular dichroism at far UV region can give information concerning the secondary structure of a protein. Furthermore, we used fluorescence and far-UV CD spectroscopies for monitoring the changes in the secondary and tertiary structures of all variants in unfolding equilibrium experiments. Fig. 2 shows the equilibrium denaturation curves of WT and mutant proteins monitored by fluorescence and far-UV CD spectroscopies. Thermodynamic parameters obtained from analysis of these curves are given in Table 1. Comparison of ΔG(H2O) as a measure of conformational stability as well as [Urea]50% shows that the resulting data of CD and fluorescence
ð7Þ
Considering all reactions of Scheme 2 and the effect of different rates on populating the I molecules at the pre-equilibrium state and by regarding the conversion of I to N as the rate limiting step, the free energy of intermediate state as a reflection of its population is determined by following thermodynamic equation: ΔGUI ¼ RTlnKUI :
ð8Þ
At the end, all data can be normalized by considering the unfolded state as reference [31,32]. 2.4. Bioinformatic studies The MODELLER program (Ver. 9.15) was used for constructing 3D structural models of mnemiopsin and mutants. According to Similarity search against Protein Data Bank using BLAST program, it was shown
Fig. 1. Ribbon diagram of the structural model of mnemiopsin 1 and representation of interactions of R39, S128 and V183. In three dimensional structure of WT mnemiopsin, triad of Phe, Met and Trp forms a part of the substrate binding site which is close to Arg39 and is shown in figure by dotted surface. The Trp residues as chromophores in fluorescence studies are also shown.
52
F. Hakiminia et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 158 (2016) 49–55
Fig. 2. The equilibrium denaturation curves of wild-type and mutants of mnemiopsin monitored by different spectroscopic probes. (a) Normalized fluorescence intensity at 350 nm; (b) normalized far-UV CD signal at 222 nm. The continuous line is a theoretical curve upon fitting experimental data to Eq. (1) with KaleidaGraph analysis software. The equilibrium parameters obtained are given in Table 1.
measurements are different for all variants (Table 1). On the other hand, the secondary structure of protein is more stable than the tertiary one; for example, the value of ΔG(H2O) of CD measurement in WT protein is 3.72 kJ/mol higher than that of the fluorescence one. The differences in far UV CD and fluorescence data suggest that the secondary and tertiary structures of all variants denature in different steps in which the tertiary structure of protein denatured first, followed by the loss of the secondary structure. According to equilibrium data in Table 1, both the secondary and tertiary structures of S128G variant are destabilized in comparison to WT protein. However, the conformational stability of V183T increases when compared with WT protein. Substitution of Arg by Lys in R39k mutant brought about changes on the stability of the secondary and tertiary structures of protein toward stabilization. Comparison of [Urea]50% and thermodynamic m-values in Table 1 indicates that the effect of replacing Val by Thr on the conformational stability of protein is related to both m-value and [Urea]50%. The m-value has been shown to be proportional to the difference in the solvent accessible surface area (SASA) between the denatured and folded states of protein [9]. Hence, the compactness of more stabilized V183T in the folded state may be greater than that of the WT protein. In the case of S128G mutant, decreasing the m-value may be attributed to the increasing solvent accessible surface area of folded state; this is due to the fact that this variant is destabilized compared to the WT protein. This observation may also be related to the solvation effects originated from free energy transfer of Gly from hydrophobic core of protein to urea during unfolding of protein from the folded state.
3.3. Kinetics of folding and unfolding The kinetics of unfolding and refolding can be used for detecting transition states as well as folding intermediates which are not stable enough to be observed under equilibrium conditions. For this purpose, the rate constants for folding and unfolding at different concentrations of urea were determined by recording the intrinsic Trp fluorescence as a function of time. Fig. 3 shows a kinetic trace illustrating the unfolding time course of WT protein measured by stopped-flow fluorescence spectroscopy as obtained by rapid mixing of one volume of protein in buffer with 6 volumes of buffer containing 3 M urea at pH 8.0 and 25 °C. The final concentration of urea was 2.57 M. An excitation wavelength of 280 nm was used, and change in fluorescence emission was monitored at 350 nm. Solid line represents a curve of best fit to the data and the fit quality is represented by the residual data in the upper panel; see “Materials and methods” for a complete experimental description. Upon fitting experimental kinetic data to Eq. (4), describing exponential function, the rate constants were obtained from stopped-flow fluorescence data. Plots of the natural logarithm of rate constants versus urea concentration known as chevron plot can be fitted globally to suitable equation for obtaining kinetic data of folding and unfolding reactions. As shown in Fig. 4 for WT protein, the chevron plots were first fitted to Eq. (5) describing a two state model as depicted in Scheme 1. It is evident from Fig. 4 that refolding arm of the chevron plot shows a deviation from linearity which becomes evident at low urea concentrations,
Table 1 Thermodynamic parameters for wild-type and mutants of mnemiopsin obtained by analysis of equilibrium denaturation curves of Fig. 2; a and b, using Eq. (1). Variants
WT R39K S128G V183T a b
Equilibrium parameters of fluorescence measurementsa,b
Equilibrium parameters of CD measurementsa,b
m-Valuea
[Urea]50%
ΔG(H2O)a
m-Valuea
[Urea]50%
ΔG(H2O)a
4.14 ± 0.005 4.19 ± 0.008 3.36 ± 0.003 4.45 ± 0.003
3.43 ± 0.004 3.79 ± 0.007 3.71 ± 0.003 3.56 ± 0.002
14.20 ± 0.004 15.90 ± 0.006 12.46 ± 0.004 15.86 ± 0.004
4.93 ± 0.003 5.08 ± 0.005 4.70 ± 0.009 5.11 ± 0.008
3.63 ± 0.002 3.87 ± 0.004 3.48 ± 0.007 3.80 ± 0.006
17.92 ± 0.003 19.65 ± 0.007 16.34 ± 0.008 19.40 ± 0.006
Standard deviations are calculated based on four or five replicates of experiments. m-Value in kJ/mol−2; [Urea]50% in M and ΔG(H2O) in kJ/mol.
F. Hakiminia et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 158 (2016) 49–55
53
Fig. 3. Representative kinetic trace illustrating the unfolding time course of wild-type protein measured by stopped-flow fluorescence spectroscopy.
showing a clear systematic deviation from the expected values for a two-state behavior. Using this model for fitting, we found that the rate H2 O 2 O constant of unfolding and refolding at 0 M Urea (kH NU and kUN ) was imprecise because of the paucity of data in the refolding limb and the very long extrapolation to 0 M urea. So, the rollover in the refolding arm of chevron plot may not be accounted for by Hammond behavior within a two state model [37]. The low quality of fitting to a two state model indicates that chevron plots should be fitted to other models rather than the two state one. In a three-state model, we assumed that the chevron plot curvature at low urea concentrations indicates the transient accumulation of a partially folded species with a partially non-native topology, which is
Fig. 4. Experimental data of kinetic studies of WT protein and fit of chevron plot to Eq. (5) describing a two-state model as depicted in Scheme 1. Continuous line represents best fit of experimental data to Eq. (5). A deviation from linearity becomes evident, particularly at low urea concentrations. The results indicate that other models rather than the two-state one should be considered.
less sensitive to urea. In other words, there is a molecular species in this region which is relatively less compact and less sensitive to urea relative to more compact folded state. Hence, the observed chevron plots were fitted to a three-state model (Scheme 2 and Fig. 5) assuming the presence of on-pathway intermediate (see Materials and methods). Fig. 6 shows the results of fitting experimental data of chevron plots to Eq. (6) describing a three-state model. Kinetic parameters of all variants are shown in Table 2. According to the resulting data of Table 2 and comparison of equilibrium constants for the formation of intermediate state from unfolded structure (KUI), it reveals that accumulation of intermediate state during refolding reaction is greater in R39K and V183T relative to the WT and S128G variants. Furthermore, m-values of intermediate formation from the unfolded state for the WT protein is larger than those for mutants, indicating that the change in surface area should be larger when starting from the unfolded structure to the partially folded intermediate state in the WT protein. Larger m-value in the WT protein may be attributed to more extended polypeptide chain in the unfolded state or more compactness of the intermediate state. According to our model described in the Materials and methods, the formation of the native structure from the intermediate state is considered as the rate limiting step in protein folding. In other words, the formation of native structure from the intermediate state (kIN) is
Fig. 5. Free energy profile for the folding of protein according to Scheme 2 and via one major intermediate (I) and transition state (TS). The right hand side of the figure as the energy level of N is obtained from fluorescence-based equilibrium data of Table 1. The energy levels of I and TS are obtained from kinetic studies.
54
F. Hakiminia et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 158 (2016) 49–55
Fig. 6. Chevron plots showing the urea dependence of rate constants for unfolding and refolding of wild-type and mutants of mnemiopsin. The experiments at the left-hand side of the curves consist of protein solutions in high concentrations of urea diluted into refolding buffer in the stopped-flow fluorescence apparatus. In this region, kobs is the rate constant for refolding. At the right-hand side of the curves, folded protein in aqueous buffer was mixed with urea solutions, and so the kinetics of unfolding is recorded. In protein unfolding/ refolding studies, the kinetic parameters are reported at 0 M concentration of denaturant. This condition is a reference state for comparing the parameters. However, doing unfolding and/or refolding experiments at 0 M urea is impossible, hence the parameters at 0 M were obtained by extrapolation of experimental data to 0 M urea which is carried out by mathematical modeling and fitting experimental data to appropriate equations. The data of chevron plots were fitted to a three-state on-pathway intermediate model (Scheme 2 and Fig. 5). Continuous line represents best fit of experimental data to Eq. (6).
accompanied by a further molecular rearrangement of molecule and the values of rate constants of reverse reaction (kNI) are negligible when compared with those of forward reaction (kIN). Our data in Table 2 suggest that the rate limiting step in all mutants is accelerated relative to WT protein and the largest differences in this process were observed in V183T and R39K mutants. For better understanding of kinetic and thermodynamic consequences of mutations on the stability and kinetics of refolding and unfolding, stability measurements (fluorescence based data in equilibrium unfolding experiments) and kinetic data were normalized by considering the unfolded state as the reference and the resulting data were presented as difference free energy diagrams (Fig. 7). To do this, rate constants of reaction (kIN) should be converted to the free energy change of transition state using Eq. (7). Similarly, by applying Eq. (8) free energy change of intermediate states as a measure of the population of I molecules or intermediate stability at pre-equilibrium can be achieved.
Structural examination of protein indicates that the helix containing Arg39 is short and it seems that it loses few contacts upon mutation to Lys, which is physico-chemically similar to Arg. Besides, thermodynamic data suggest that replacement of Arg with Lys has similar effect on the conformational stability of the secondary and tertiary structures of protein. Yet the higher stability of intermediate state in R39K in comparison to WT and S128G suggests that initial packing of helical segments within the protein structure is more facilitated when Lys with smaller side chain is present in the protein chain. The region of protein containing Val183 is a loop and usually the small number of contacts is made by loops on the structure of proteins. Surprisingly, the free energy change of unfolding in V183T variant (Table 1) toward stabilization suggests that new interactions may be formed upon mutation of Val to Thr. As indicated in Table 1, protein is destabilized upon mutation of Ser to Gly. Examination of structural models indicates that upon replacement of Ser by Gly in S128G variant, the hydrogen bonds between main chain of Ser128 and side chain of
Table. 2 Kinetic and thermodynamic parameters for the wild-type and mutant proteins obtained from chevron plot analysis.
WT R39K S128G V183T
2O KH UI
2O kH IN
2O kH NI
m‡IN
mUI
m‡NI
123 ± 34 600 ± 66 116 ± 19 438 ± 102
0.56 ± 0.20 3.20 ± 0.60 1.67 ± 0.44 5.89 ± 1.46
0.03 ± 0.001 0.03 ± 0.001 0.06 ± 0.003 0.01 ± 0.0005
−1.95 ± 0.40 −0.04 ± 0.18 −1.10 ± 0.29 −0.08 ± 0.25
4.24 ± 0.36 3.52 ± 0.13 3.59 ± 0.22 3.10 ± 0.16
0.32 ± 0.005 0.34 ± 0.006 0.23 ± 0.005 0.44 ± 0.006
−1 H2O 2O is the rate The chevron plots were fitted to a three state on-pathway intermediate model. KH UI is the equilibrium constant between the unfolded and intermediate states. kij in S constant regarding conversion of i-state to j-state. mij refers to differences in surface area between i and j states.
F. Hakiminia et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 158 (2016) 49–55
55
References
Fig. 7. Difference free energy diagrams. ΔΔG(H2O) is the absolute value of free energy difference of structural elements between WT and all protein variants, so the difference energy for WT protein becomes zero for all structural states.
Leu182 as well as main chain of Val183 with side chain of Ser128 are lost. So, decreasing the conformational stability of S128G mutant may be a consequence of removing interactions made by Ser as well as increasing the flexibility of protein upon replacement of Ser by Gly which has more conformational space. According to Table 2, the effect of S128G mutant on the stability of intermediate state is negligible. In summary, thermodynamic and kinetic data together indicate that all mutants follow essentially the same reaction scheme as WT protein for folding from the urea-denatured state. Also, overall conformational stability of intermediate and folded states of protein increases in both V183T and R39K.
Acknowledgments Financial support for this work was provided by the Research Council of Tarbiat Modares University (2244604). There is no conflict of interest.
[1] Y. Takenaka, A. Yamaguchi, N. Tsuruoka, M. Torimura, T. Cojobori, Y. Shingeri, Mol. Biol. Evol. 29 (2006) 1669–1681. [2] Z.J. Liu, G.A. Stepanyuk, E.S. Vysotski, J. Lee, S.V. Markova, N.P. Malikova, B.C. Wang, Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 2570–2575. [3] L. Deng, E.S. Vysotski, S.V. Markova, Z.J. Liu, J. Lee, J. Rose, B.C. Wang, Protein Sci. 14 (2005) 663–675. [4] G.A. Stepanyuk, Z.J. Liu, L.P. Burakova, J. Lee, J. Rose, E.S. Vysotski, B.C. Wang, Biochim. Biophys. Acta 1834 (2013) 2139–2146. [5] J.R. Blinks, Environ. Health Perspect. 84 (1990) 75–81. [6] C.J. Lewis, S. Daunert, Fresenius J. Anal. Chem. 366 (2000) 760–768. [7] Y. Ohmiya, T. Hirano, Chem. Biol. 3 (1996) 337–347. [8] C.H. Contag, S.D. Spilman, P.R. Contag, M. Oshiro, B. Eames, P. Dennery, D.K. Stevenson, D.A. Benaron, Photochem. Photobiol. 66 (1997) 523–531. [9] E. Gonzalez, R.Y. Tslen, Chem. Biol. 4 (1997) 269–277. [10] W.W. Ward, H.H. Seliger, Biochemistry 13 (1974) 1500–1510. [11] W.W. Ward, H.H. Seliger, Biochemistry 13 (1974) 1491–1499. [12] M.R. Aghamaali, V. Jafarian, R. Sariri, M. Molakarimi, B. Rasti, M. Taghdir, R.H. Sajedi, S. Hosseinkhani, Protein J. 30 (2011) 566–574. [13] V. Jafarian, R. Sariri, S. Hosseinkhani, M.R. Aghamaali, R.H. Sajedi, M. Taghdir, S. Hassannia, Biochem. Biophys. Res. Commun. 413 (2011) 164–170. [14] A. Mahdavi, R.H. Sajedi, S. Hosseinkhani, M. Taghdir, R. Sariri, Photochem. Photobiol. Sci. 12 (2013) 467–478. [15] J.F. Head, S. Inouye, K. Teranishi, O. Shimomura, Nature 405 (2000) 372–376. [16] E. Dikici, X. Qu, L. Rowe, L. Millner, C. Logue, S.K. Deo, M. Ensor, S. Daunert, Protein Eng. Des. Sel. 22 (4) (2009) 243–248. [17] S. Toma, K.T. Chong, A. Nakagawa, K. Teranishi, S. Inouye, O. Shimomura, Protein Sci. 14 (2) (2005) 409–416. [18] A. Mahdavi, R.H. Sajedi, S. Hosseinkhani, M. Taghdir, Photochem. Photobiol. Sci. (2014) 792–800. [19] A.R. Fersht, A. Matouschek, L. Serrano, J. Mol. Biol. 224 (1992) 805–818. [20] H.S. Chanand, K. Dill, Proteins 30 (1998) 2–33. [21] S. Blond, M.E. Goldberg, Protein 1 (1986) 247–255. [22] A.R. Fersht, A. Matouschek, L. Serrano, J. Mol. Biol. 224 (1992) 771–782. [23] F. Hillger, D. Nettels, S. Dorsch, B. Schuler, J. Fluoresc. 17 (2007) 759–765. [24] M. Lindgren, K. Sörgjerdy, P. Hammarström, Biophys. J. 88 (2005) 4200–4212. [25] S.M. Kelly, T.J. Jess, N.C. Price, Biochim. Biophys. Acta 1751 (2005) 119–139. [26] P. Gill, B. Ranjbar, Chem. Biol. Drug Des. 74 (2009) 101–120. [27] J. Walters, S.L. Milam, A.C. Clark, Methods Enzymol. 455 (2009) 1–39. [28] C.N. Pace, Methods Enzymol. 131 (1986) 266–279. [29] J.K. Myers, C.N. Pace, J.M. Scholtz, Protein Sci. 4 (1995) 2138–2148. [30] A. Zarrine-Afsara, A.R. Davidson, Methods 34 (2004) 41–50. [31] H. Eyring, Chem. Rev. 17 (1935) 65–77. [32] A.R. Fersht, A. Matouschek, M. Bycroft, J.T. Kellis, L. Serrano, Pure Appl. Chem. 63 (2) (1991) 187–194. [33] N. Eswar, M.A. Marti-Renom, B. Webb, M.S. Madhusudhan, D. Eramian, M. Shen, U. Pieper, A. Sali, John Wiley& Sons, Inc., Supplement 15 (2006) 5.6.1–5.6.30. [34] D. Eramian, N. Eswar, M.Y. Shen, A. Sali, Protein Sci. 17 (2008) 1881–1893. [35] M.Y. Shen, A. Sali, Protein Sci. 15 (2006) 2507–2524. [36] F. Melo, A. Sanchez, A. Sali, Protein Sci. 11 (2002) 430–448. [37] A. Matouschek, D.E. Otzen, L.S. Itzhaki, S.E. Jackson, A.R. Fersht, Biochemistry 17 (41) (1995) 13356–13362.
