Introduction

Ionic liquids (ILs) are molten salts at room temperature because of low melting point. ILs have many distinctive properties such as low vapor pressure, high ionic conductivity, wide liquid range, high thermal stability, and low toxicity. They are able to dissolve a variety of solutes. During the last two decades, ILs have emerged as biocompatible solvents for organic synthesis, biocatalysis, and other biological systems [1,2,3,4,5,6,7,8,9,10,11,12]. They have been widely used in protein separation, extraction, and purification due to their high thermal stability and excellent biocompatibility [13,14,15,16,17]. To evaluate the effect of ILs on the biological reaction processes, the structure, stability, and activity of proteins in ILs have been extensively investigated [18,19,20,21,22,23,24,25,26,27,28,29,30,31,32].

Protein stability can be tuned by the environment surrounding it. The ILs offer perfect environments that can be tuned to alter the structure and physicochemical property of biomacromolecules. Understanding the mechanism how ILs affect the protein stability is crucial for the biological reaction processes taking place in ILs, including enzymatic reactions, bioengineering, protein extraction, and purification. Although ILs can alter protein stability and function, the mechanistic understanding of protein stability in ILs requires to be clarified. The effects of ILs on protein stability have been extensively studied [33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49]. For instance, some ILs were reported to increase the stability of proteins [33, 34, 37, 38, 40, 47, 48]. Some ILs were shown to destabilize proteins [39, 43, 45, 49]. However, other ILs were investigated to stabilize or destabilize proteins depending on the concentrations, species, polarity, and hydrophobicity of ILs [35, 36, 41, 42, 44, 46]. The interactions between proteins and ILs including electrostatic interactions and hydrophobic interactions may play a role in the effects of ILs on protein stability.

NMR spectroscopy has been utilized to investigate direct interaction between ILs and proteins through NMR chemical shift perturbations in 2D 1H-15N HSQC spectra [39, 40, 50]. Cabrita and coworkers showed that the preferential binding of ionic liquids with Im7 is crucial to modulate its stability in ILs [39]. Kaar and coworkers studied the direct ion interactions between protein and ILs and its effect on protein stability in ILs [40]. Varga and coworkers used HR-MAS NMR to probe the effects of [C4-mim]Br ILs on GB1 structure and dynamics [50].

Although the effects of ILs on protein stability have been extensively studied, few studies concentrated on the equilibrium thermodynamics of protein stability in ILs. Quantifying the enthalpic and entropic changes of protein contributes to a better understanding of protein equilibrium thermodynamics in ILs. Here, we use two small globular proteins KH1 (MW 9.4 kDa and pI 6.8) and SH3 (MW 6.9 kDa and pI 4.6), a domain of human K-homology splicing regulator protein and a domain of Drosophila signal transduction protein drk, respectively. We have utilized 19F NMR to examine the effects of [C4-mim]Br ILs on the equilibrium thermodynamics of KH1 and SH3 stability. 19F is a nucleus with almost 100% natural abundance and its intrinsic sensitivity is approximately 83% of the proton. The large chemical shift range and no background interference of 19F NMR make it an ideal method to probe and monitor the stability and conformational transitions of protein in complex systems [51, 52].

The modified standard state Gibbs free energy of protein unfolding, \( \Delta {G}_{\mathrm{u},T}^{0^{\prime }} \), equals − RTln(Ku), where R is the gas constant, T is the absolute temperature, and Ku is the equilibrium constant (Ku = [unfolded]/[folded]). The \( \Delta {G}_{\mathrm{u},T}^{0^{\prime }} \) can be separated into enthalpic and entropic components.

$$ \Delta {G}_{\mathrm{u},T}^{o^{\prime }}(T)=\Delta {H}_{\mathrm{u}}^{o^{\prime }}(T)-T\ \Delta {S}_{\mathrm{u}}^{o^{\prime }}(T) $$
(1)

\( \Delta {H}_{\mathrm{u}}^{o^{\prime }}(T) \) and \( \Delta {S}_{\mathrm{u}}^{o^{\prime }}(T) \) are the temperature-dependent modified standard-state enthalpy and entropy of unfolding, respectively.

$$ \Delta {H}_{\mathrm{u}}^{o^{\prime }}(T)=\Delta {H}_{\mathrm{u}}^{o^{\prime }}\left({T}_{\mathrm{ref}}\right)+{\int}_{T_{\mathrm{ref}}}^T\Delta {C}_{\mathrm{p}}^{o^{\prime }} dT $$
(2)
$$ \Delta {S}_{\mathrm{u}}^{0^{\prime }}(T)=\Delta {S}_{\mathrm{u}}^{o^{\prime }}\left({T}_{\mathrm{ref}}\right)+{\int}_{T_{\mathrm{ref}}}^T\frac{\Delta {C}_p^{o^{\prime }}}{T} dT $$
(3)

Tref is a reference temperature and \( \Delta {C}_{\mathrm{p}}^{o^{\prime }} \) is the modified standard-state heat capacity of unfolding. \( \Delta {C}_{\mathrm{p}}^{o^{\prime }} \) is assumed to be temperature-independent over the range studied. Combining Eqs. 1–3 gives Eq. 4.

$$ \Delta {G}_{\mathrm{u}}^{o^{\prime }}(T)=\Delta {H}_{\mathrm{u}}^{o^{\prime }}\left({T}_{\mathrm{ref}}\right)-T\ \Delta {S}_{\mathrm{u}}^{o^{\prime }}\left({T}_{\mathrm{ref}}\right)+\Delta {C}_{\mathrm{p}}^{o^{\prime }}\left[T-{T}_{\mathrm{ref}}-T\ln \frac{T}{T_{\mathrm{ref}}}\right] $$
(4)

The temperature dependences of \( \Delta {G}_{\mathrm{u},T}^{0^{\prime }} \) for globular proteins are bell-shaped with a maximum at Ts, where denaturation is purely enthalpic, and \( \Delta {S}_{\mathrm{u}}^{o^{\prime }} \) equals zero [53], whereas Tm is the melting temperature, which is the point where the \( \Delta {G}_{\mathrm{u},T}^{0^{\prime }} \) equals zero.

Material and methods

Protein expression and purification

The plasmids encoding the his-tagged KH1 domain protein and the N-terminal SH3 domain of drk (SH3) were transformed into Escherichia coli BL21 (DE3) competent cells. Expressions of S193A KH1 and SH3 were based on the described protocols [54, 55]. For 3-19F-tyrosine-labeled S193A KH1, 70 mg of D,L-m-fluorotyrosine, 60 mg of L-phenylalanine, 60 mg of L-tryptophan, and 0.5 g glyphosate were added into 1 L of M9 minimal medium [56]. The inducer isopropyl β-D-thiogalactopyranoside (IPTG) was added to a final concentration of 1 mM to induce expression. Cells were induced for 20 h at 20 °C and harvested by centrifugation. For 5-19F-tryptophan-labeled SH3, 60 mg 5-fluoroindole was added into 1 L of M9 minimal medium [57]. Cells were induced by 1 mM IPTG for 2 h at 37 °C and harvested by centrifugation.

For S193A KH1, the cell pellet was resuspended in Ni-column buffer A (50 mM Tris, 300 mM NaCl, 10 mM imidazole, pH 8.0) for sonication. The supernatant was collected after centrifugation at 16000g at 4 °C for 30 min. KH1 was purified as reported [54] using Ni-affinity chromatography then followed by size exclusion chromatography. For SH3, the cell pellet was resuspended in buffer (50 mM Tris, pH 7.5) and protease inhibitors were added for sonication. Purification of SH3 was accomplished as the described protocols [55], which involved two chromatography steps using an anion exchange chromatography followed by size exclusion chromatography. Purified S193A KH1 and SH3 were applied to desalting columns, lyophilized for 24 h, and stored at − 20 °C.

Ionic liquids

[C4-mim]Br was purchased from Aladdin Ltd. and used without further purification. Solutions of various concentrations of ILs were prepared by dissolving [C4-mim]Br in phosphate buffer (50 mM sodium phosphate buffer, pH 7.5). The ILs were then used to dissolve the lyophilized protein.

NMR spectroscopy

19F spectra were acquired between 283 and 331 K on Bruker 600 MHz and Bruker 500 MHz spectrometers equipped with 5-mm H/F (C, N) triple resonance cryoprobes. 19F spectra were acquired with spectral widths of 30 ppm with a duty cycle delay of 2.0 s.

Stability measurements

Topspin 3.2 (Bruker), MestRe-C, and Origin8.0 were utilized to process and analyze data. MestRe-c was used to process the free induction decays and transform the decays into ASCII text files, which can be read by Origin8.0. Origin8.0 was applied to fit the peaks in 19F spectra. Peak fitting and integration were accomplished by using Lorentzian functions in Origin8.0. Peak integrations were used to calculate the populations of folded state (F) and unfold ensembles (U). \( \Delta {G}_{\mathrm{u}}^{0^{\prime }} \) was calculated from the integrals of the two peaks. Three measurements at one temperature were conducted to estimate the uncertainty expressed as the standard deviation (SD) of the mean. The uncertainties in \( \Delta {G}_{\mathrm{u}}^{0^{\prime }} \) were utilized to calculate the uncertainties in TS, \( \Delta {H}_{\mathrm{u}}^{0^{\prime }} \)(Ts), Tm, \( \Delta {H}_{\mathrm{u}}^{0^{\prime }} \)(Tm), and \( \Delta {C}_{\mathrm{p}}^{0^{\prime }} \) via a weighted fit to Eq. 4.

Results and discussion

We labeled S193A KH1’s sole tyrosine and SH3’s sole tryptophan with fluorine. Fluorine labeling makes it easy to quantify the folded state (F) and unfolded ensemble (U) [55, 58]. 19F NMR is a quantitative method with high sensitivity and no background interference. It is an ideal technique to probe protein stability and structure in complex environment [55]. The peak area under each 19F resonance is proportional to its concentration. We acquired 19F spectra (Fig. 1) of S193A KH1 and SH3 in buffer and various concentrations of ionic liquids at 298 K and 288 K, respectively. The 19F spectrum of S193A KH1 in buffer at 298 K shows mainly one 19F resonance, which corresponds to the folded state (F). The 19F spectrum of SH3 in buffer at 288 K shows two 19F resonances, one from the folded state (F), the other from unfolded ensemble (U). The peak from U gradually increases as the concentrations of [C4-mim]Br ILs are elevated for both S193A KH1 and SH3. These data indicate that S193A KH1 and SH3 are destabilized in ILs.

Fig. 1
figure 1

Spectra of fluorine-labeled S193A KH1 (a) and SH3 (b) in buffer and various concentrations of [C4-mim]Br ionic liquids at 298 K and 288 K, respectively

Full size image

In order to assess the effect of ionic strengths on protein stability, we also acquired the 19F spectra (Fig. 2) of S193A KH1 and SH3 in buffer and various concentrations of NaBr solutions at 298 K and 283 K, respectively. For S193A KH1, the peak intensity from U marginally increases as the concentrations of NaBr are elevated. The peak intensity from U in 0.5 M NaBr is significantly weaker than that in 0.5 M [C4-mim]Br ILs. For SH3, the peak intensity from F marginally increases as the concentrations of NaBr are elevated, which means that SH3 are slightly stabilized in NaBr solutions. These results suggest that the effect of ionic strengths on protein stability does not dominate in the destabilizations of proteins by [C4-mim]Br ILs.

Fig. 2
figure 2

Spectra of fluorine-labeled S193A KH1 (a) and SH3 (b) in buffer and various concentrations of NaBr solutions at 298 K and 283 K, respectively

Full size image

To investigate the equilibrium thermodynamics of protein stability in ILs, we measured the temperature dependence of S193A KH1 and SH3 stability (Fig. 3) in buffer alone, 0.13 M, 0.33 M, and 0.53 M [C4-mim]Br IL. The \( \Delta {G}_{\mathrm{u}}^{{\mathrm{o}}^{\prime }} \) values of S193A KH1 and SH3 in [C4-mim]Br ILs are smaller than those in buffer, which means that the [C4-mim]Br ILs destabilized S193A KH1 and SH3.

Fig. 3
figure 3

Temperature dependence of S193A KH1 (a) and SH3 (b) stability in buffer and various concentrations of [C4-mim]Br ionic liquids

Full size image

Equation 4 was applied to extract values of Tm, Δ\( {H}_{\mathrm{u}}^{0^{\prime }} \)(Tm), Ts, Δ\( {H}_{\mathrm{u}}^{0^{\prime }} \)(Ts), and \( \Delta {C}_{\mathrm{p}}^{0^{\prime }} \) (Tables 1 and 2). Tm values were used to assess thermal stability. Tm of S193A KH1 and SH3 in 0.13 M [C4-mim]Br ILs were 8.1 K and 8.3 K lower than those in buffer alone, respectively, which means proteins are significantly destabilized in ILs.

Table 1 Thermodynamics parameters for S193A KH1 in buffer and ionic liquids
Full size table
Table 2 Thermodynamics parameters for SH3 in buffer and ionic liquids
Full size table

To quantify the enthalpic and entropic components, we assessed Δ\( {H}_{\mathrm{u}}^{0^{\prime }} \) and − TΔ\( {S}_{\mathrm{u}}^{0^{\prime }} \) at 308 K (Tables 3 and 4). For both S193A KH1 and SH3, the Δ\( {H}_{\mathrm{u}}^{0^{\prime }} \) values in ionic liquids are larger than those in buffer, showing that protein is enthalpically stabilized in ILs. The − TΔ\( {S}_{\mathrm{u}}^{0^{\prime }} \) values in ILs are smaller than those in buffer, indicating that protein is entropically destabilized in ILs. However, the enthalpic stabilization does not completely compensate the entropic destabilization in ILs, which contributes to an overall destabilization. The enthalpic stabilization and entropic destabilization of protein in ILs is consistent with the effect of trimethylamine N-oxide (TMAO) and other protective osmolytes on protein enthalpic and entropic changes [59,60,61]. However, the enthalpic stabilization compensates the entropic destabilization in TMAO and other protective osmolytes, which accounts for overall stabilization of protein in TMAO and other protective osmolytes.

Table 3 Thermodynamics parameters for S193A KH1 in buffer and ionic liquids at 308 K
Full size table
Table 4 Thermodynamics parameters for SH3 in buffer and ionic liquids at 308 K
Full size table

The stability difference of SH3 in [C4-mim]Br IL solutions and NaBr solutions indicates that the [C4-mim]+ dominates in the destabilization of proteins by ILs. At pH 7.5, both S193A KH1 (pI 6.8) and SH3 (pI 4.6) are negatively charged. The imidazolium cation will accumulate near the negatively charged residues of proteins. Cabrita and coworkers confirmed that the imidazolium cation would strongly bind to the non-polar amino acids [39]. The hydrophobic interactions between the [C4-mim]+ and protein hydrophobic residues will lead to the exposure of protein hydrophobic surface, which finally facilitates protein unfolding. Previous studies have demonstrated that the stability of protein in aqueous IL solutions does not necessarily obey the Hofmeister series [62]. Our data suggests that protein stability in ILs is the result of balance between hydrophobic interactions, electrostatic interactions, and hydration.

19F NMR is a quantitative method to study the protein stability in ILs. The equilibrium thermodynamics of protein folding and unfolding in ILs can be readily investigated by 19F NMR. Previous studies have shown that 2D 1H-15N HSQC spectra can be utilized to study the hydrophobic and electrostatic interaction between proteins and ILs through chemical shift perturbations. The combination of 19F NMR with 2D 1H-15N HSQC spectra will give a quantitative analysis on how the interactions between proteins and ILs affect the protein stability

Conclusions

For the first time, we used 19F NMR to study the effects of [C4-mim]Br ILs on equilibrium thermodynamics of protein. Two small globular proteins, S193A KH1 and SH3, are destabilized in ILs. Proteins are enthalpically stabilized and entropically destabilized in ILs compared to buffer. The enthalpic stabilization by ILs is counterbalanced by entropically destabilization, which leads to an overall destabilization. The preferential interactions between proteins and ILs are not strong enough to stabilize protein. Our results suggest that the preferential interactions between proteins and ILs are crucial to modulate protein stabilization in ILs. 19F NMR provides a simple and convenient approach to quantitatively analysis of protein stability in ILs. Compared to other analysis methods such as CD spectroscopy, fluorescence, UV-vis and FT-IR spectra, 19F NMR is able to study the structure and function of protein at atomic resolution with low background and facilitates the protein–protein interaction study in ILs.