Structure and stability of recombinant bovine odorant-binding protein: III. Peculiarities of the wild type bOBP unfolding in crowded milieu

Materials

GdnHCl (Nacalai Tesque, Japan), ANS (ammonium salt of 8-anilinonaphtalene-1-sulfonic acid; Fluka, Switzerland) and crowding agents (PEG600, PEG4000 and PEG12000; Sigma-Aldrich, USA) were used without further purification. The protein concentration was 0.1–0.2 mg/mL. The experiments were performed in 20 mm Na-phosphate-buffered solution at pH 7.8.

Gene expression and protein purification

The plasmid pT7-7-bOBP which encodes bOBP with a poly-histidine tag were used to transform Escherichia coli BL21(DE3) host (Invitrogen) (Stepanenko et al., 2014b). The protein expression was induced by incubating the cells with 0.3 mm of isopropyl-beta-D-1-thiogalactopyranoside (IPTG; Fluka, Switzerland) for 24 h at 37 °C. The recombinant protein was purified with Ni+-agarose packed in HisGraviTrap columns (GE Healthcare, Sweden). The protein purity was determined through SDS-PAGE in 15% polyacrylamide gel (Laemmli, 1970).

Fluorescence spectroscopy

Fluorescence experiments were performed using a Cary Eclipse spectrofluorimeter (Varian, Australia) with microcells FLR (10 × 10 mm; Varian, Australia). Fluorescence intensity was corrected on the primary inner filter effect (Fonin et al., 2014). Fluorescence lifetime were measured using a “home built” spectrofluorimeter with a nanosecond impulse (Stepanenko et al., 2012; Stepanenko et al., 2014b; Turoverov et al., 1998) as well as micro-cells (101.016-QS 5 × 5 mm; Hellma, Germany). Tryptophan fluorescence in the protein was excited at the long-wave absorption spectrum edge (λ_ex = 297 nm), wherein the tyrosine residue contribution to the bulk protein fluorescence is negligible. The fluorescence spectra position and form were characterized using the parameter A = I₃₂₀/I₃₆₅, wherein I₃₂₀ and I₃₆₅ are the fluorescence intensities at the emission wavelengths 320 and 365 nm, respectively (Turoverov & Kuznetsova, 2003). The values for parameter A and the fluorescence spectrum were corrected for instrument sensitivity. The tryptophan fluorescence anisotropy was calculated using the equation $r=(I_V^V-G{I}_H^V)/(I_V^V+2G{I}_H^V)$, wherein $I_V^V$ and $I_H^V$ are the vertical and horizontal fluorescence intensity components upon excitement by vertically polarized light. G is the relationship between the fluorescence intensity vertical and horizontal components upon excitement by horizontally polarized light $(G=I_V^H/I_H^H)$, λ_em = 365 nm (Turoverov et al., 1998). The fluorescence intensity for the fluorescent dye ANS was recorded at λ_em = 480 nm (λ_ex = 365 nm). Protein unfolding was initiated by manually mixing the protein solution (40 μl) with a buffer solution (510 μl) that included the necessary GdnHCl concentration and crowding agent concentration. The GdnHCl concentration was determined by the refraction coefficient using an Abbe refractometer (LOMO, Russia; Pace (1986)). The dependences of different fluorescent characteristics bOBP on GdnHCl were recorded following protein incubation in a solution with the appropriate denaturant concentration at 4 °C for different times (see in the text). The protein refolding was initiated by diluting the pre-denatured protein (in 3.0 M GdnHCl, 40 μl) with the buffer or denaturant solutions at various concentrations (510 μl), containing crowding agent. The spectrofluorimeter was equipped with a thermostat that holds the temperature constant at 23 °C.

Circular dichroism measurements

The CD spectra were generated using a Jasco-810 spectropolarimeter (Jasco, Japan). Far-UV CD spectra were recorded in a 1 mm path length cell from 260 nm to 190 nm with a 0.1 nm step size. Near-UV CD spectra were recorded in a 10 mm path length cell from 320 nm to 250 nm with a 0.1 nm step size. For the spectra, we generated 3 scans on average. The CD spectra for the appropriate buffer solution were recorded and subtracted from the protein spectra.

Fitting of denaturation curves

The equilibrium dependences of the parameter A on the GdnHCl concentration were fit using a two-state model (Staiano et al., 2007):(1)$S=\frac{S_N+S_U\alpha K_{N-U}}{1+\alpha K_{N-U}},$(2)$K_{N-U}=\exp \left(\frac{-\Delta G_{N-U}^0+m_{N-U}[D]}{RT}\right),$(3)$K_{N-U}=F_U/F_N=(1-F_N)/F_N,$taking into account(4)$S_N=a_N+b_N[D],$(5)$S_U=a_U+b_U[D],$where S is the parameter A at the measured GdnHCl concentration; [D] is the guanidine concentration; m is the linear dependence of ΔG_N−U on the denaturant concentration; $\Delta G_{N-U}^0$ is the free energy of unfolding at 0 M denaturant; F_N and F_U are the fractions of native and unfolded molecules, respectively; S_N and S_U are the signal of the native and unfolded states, respectively; a_N, b_N, a_U and b_U are constants needed to fit linear dependences of the S_N and S_U signals on the GdnHCl concentration; and $\alpha =\frac{I_{U,365}}{I_{N,365}}$ with I_N,365 and I_U,365 being fluorescence intensity at 365 nm for the native and unfolded protein. Fitting was performed using a nonlinear regression with Sigma Plot.

Previously, to evaluate conformational stability of the studied proteins we took into account that the formation of the native dimeric state of bOBP occurred at moderate GdnHCl concentration is followed by full protein unfolding while conformational perturbations of bOBP at low denaturant concentrations were not attributed to the unfolding of the protein globule (Stepanenko et al., 2016c). As bOBP unfolding is fully reversible the transition from native to unfolded state of the protein was used to calculate ΔG_N−U value. Conformational stability of the bOBP in the crowded environment was evaluated similarly as presence of crowding agents resulted in flattering of denaturing curve of bOBP.

It is important to emphasize here that the maximal achievable concentrations of denaturant in the presence of crowding agents, especially at their highest tested concentrations, were limited by the solubility of the protein–denaturant–crowder systems. This limitation determined the number of data-points within the post-transition region.

bOBP unfolding in the presence of PEG-600

Previously, we have shown that denaturing curves describing GdnHCl-induced unfolding of bOBP have a complex shape with two clearly distinguishable regions where the pattern of the different protein characteristics diverges significantly (Stepanenko et al., 2014b). In the region above 1.6 M GdnHCl, the bOBP unfolding took place as indicated by significant and simultaneous changes of all protein characteristics. The moderate structural perturbations of the bOBP with local minimum at 0.5 M GdnHCl (Figs. 2–4, red symbols and lines) in the region below 1.6 M GdnHCl were designated to the bOBP transition from a mixture of monomeric and dimeric molecules in the absence of denaturant to a native dimeric state through the local reorganization of the bOBP structure in the intermediate state at 0.5 M GdnHCl (Stepanenko et al., 2014b). The conformational stability of bOBP was described in terms of the half-transition values (2.1 ± 0.1 M GdnHCl, see Table 1) (Stepanenko et al., 2016c).

In other words, the formation of the native dimeric state of bOBP takes place at moderate GdnHCl concentration and is followed by the complete unfolding of this protein, whereas conformational perturbations of bOBP induced by low denaturant concentrations are not attributed to the unfolding of the protein globule. In the absence of GdnHCl, the recombinant bOBP is in a stable state with features similar to the native dimeric bOBP. Still, recombinant bOBP in the absence of GdnHCl is characterized by a less ordered secondary structure compared with the wild-type bOBP crystallographic data and a more rigid microenvironment of tryptophan residues. These structural perturbations are responsible for the decreased capability of the recombinant bOBP for dimerization in buffered solutions. We designated this stable recombinant bOBP state in buffered solution as a “trapped” state with incorrect α-helical and β-sheet packing in the protein globule, which may interfere with the formation of the dimeric bOBP native state. The reasons for accumulation of this “trapped” state may lie in a relatively complex domain-swapping mechanism which is required for the monomers to be correctly folded. As a result, in this trapped state bOBP exists as a mixture of monomers and dimers. On the other hand, the intermediate state accumulated at 0.5 M GdnHCl is characterized by the reorganized the bOBP structure, having fewer ordered secondary structure elements, both α-helices and β-strands, compared to the recombinant bOBP both in a buffered solution and in solution containing 1.5 M GdnHCl.

Our analysis revealed that in the presence of low concentrations of PEG-600 (80 mg/mL), shapes of the curves describing the GdnHCl-induced unfolding of the recombinant bOBP were similar to shapes of the corresponding curves recorded in the absence of crowder. However, the half-transition points for the unfolding curves measured in the presence of PEG-600 were shifted towards the higher GdnHCl concentrations (C_m = 2.4 ± 0.1 M, see Fig. 2A; Table 1). Table 2 shows that the values of the parameter A and fluorescence anisotropy r measured for the recombinant bOBP in the presence of 80 mg/mL PEG-600 were somewhat higher than the corresponding values measured in the absence of crowder. The increase in the PEG-600 concentration to 150 mg/mL resulted in the more pronounced increase in the parameter A and fluorescence anisotropy r values. Figure 2B shows that when the 150 mg/mL of PEG-600 are added to the solution of the recombinant bOBP, the pre-transition region of the unfolding curve flattens and the transition happens at higher GdnHCl concentrations than the unfolding in the presence of the 80 mg/mL PEG-600 (C_m = 2.8 ± 0.1 M, Table 1).

Curiously, the curves describing the recombinant bOBP refolding from the completely unfolded state and recorded in the presence of 80 or 150 mg/mL of PEG-600 did not coincide with the quasi-equilibrium unfolding curves recorded under the similar conditions. However, these refolding curves were close to the curves describing unfolding and refolding of the recombinant bOBP alone (i.e., in the absence of crowding agent; Figs. 2A and 2B).

Figure 2C shows that the transition curves describing equilibrium unfolding and refolding of the recombinant bOBP in the presence of 300 mg/mL PEG-600 coincide and have sigmoidal shape. Furthermore, these transitions happened at significantly higher GdnHCl concentrations than transitions recorded in the presence of 80 or 150 mg/mL of this crowder (C_m = 2.9 ± 0.1 M, Table 1; Fig. 6). Table 2 shows that the values of parameter A and fluorescence anisotropy r determined in solutions containing 300 mg/mL PEG-600 were further increased compared to values of these parameters measured at lower PEG concentrations or in the absence of crowding agent. We also observed a slight decrease in the fluorescence lifetime of recombinant bOBP with increasing concentration of PEG-600 from 80–300 mg/mL (Table 1). These data, together with the observed changes in parameter A and fluorescence anisotropy r values, suggested that some compaction of the protein globule took place in the presence of the crowding agent, which resulted in the decrease in a distance between the quenching groups of the protein and its tryptophan residues.

Interestingly, the ANS fluorescence intensity, added to the protein solution in the presence of denaturant and PEG-600 at all concentrations tested, remained substantially unchanged (Fig. 2). These data are likely to reflect the fact that the presence of this crowding agent prevents the possibility of the direct interaction of the molecules of low molecular weight dye ANS and the protein.

bOBP unfolding in the presence of PEG-4000 and PEG-12000

Addition of the increasing concentrations of PEG-4000 and PEG-12000 was accompanied by the increase in the values of the parameter A and fluorescence anisotropy r, as well as the value of fluorescence lifetime (see Table 1). It is worth noting that the value of fluorescence lifetime for recombinant bOBP in the presence of 80 mg/mL of PEG-4000 significantly below the corresponding value for this parameter for bOBP in the presence of 80 mg/mL of PEG-12000, and especially in the presence of 80 mg/mL of PEG-600. However, at elevating the concentration of PEG-4000 and PEG-12000 up to 300 mg/mL the value of the fluorescence lifetime of recombinant bOBP increased to the values typical of the protein in the presence of 300 mg/mL of PEG-600. These data may reflect the different effect of the crowding agents with diverse molecular weights on the structure of the protein.

Curiously, when the unfolding-refolding process of the recombinant bOBP was analyzed in the presence of 80 mg/mL of PEG-4000 or PEG-12000, the corresponding transitions curves coincided with each other and with curve describing the equilibrium unfolding-refolding processes in the recombinant bOBP alone (C_m = 2.3 ± 0.1 M, Table 1; Figs. 3A and 4A). Subsequent increase in concentration of PEG-4000 and PEG-12000 to 150 mg/mL did not change the shape of corresponding curves, but lead to an insignificant and equal for both crowding agents shift of the unfolding transition to higher GdnHCl concentrations and slight flattering of the pre-transition regions (see Figs. 3B and 4B). Furthermore, the half-transition point for the bOBP unfolding in the presence of 150 mg/mL of PEG-4000 or PEG-12000 is observed at a significantly lower GdnHCl concentrations than in the presence of 150 mg/mL of PEG-600 (C_m = 2.6 ± 0.1 M, Table 1; Fig. 6).

The half-transition value evaluated in the presence of maximal studied concentration of PEG-4000 and PEG-12000 can not be determined because of the GdnHCl concentrations needed for the complete unfolding of this protein cannot be not reached due to the high viscosity of solutions. Still, Fig. 3C shows that when PEG-4000 concentration was increased to 300 mg/mL the curve describing the equilibrium unfolding-refolding transitions of bOBP became sigmoidal and coincided with the corresponding curve describing equilibrium unfolding-refolding of this protein in the presence of 300 mg/mL PEG-600 (see also Fig. 6). Although the GdnHCl-induced unfolding curve of the recombinant bOBP in the presence of 300 mg/mL PEG-12000 was also sigmodal (see Fig. 4C), the corresponding transition occurred at significantly lower GdnHCl concentrations (Fig. 6).

Previously we showed that the recombinant bOBP exists as a mixture of monomeric and dimeric forms because of this protein is in a stable native-like state with reduced dimerization capability (Stepanenko et al., 2014b). The compact dimeric state of the recombinant bOBP is formed under the mild denaturing conditions, namely, in the presence of 1.5 M guanidine hydrochloride (GdnHCl). This process requires bOBP secondary and tertiary structure restructuring and is accompanied by the formation of a stable, more compact, intermediate state that is maximally populated at 0.5 M GdnHCl. In our unfolding-refolding experiment, this state is manifested as a local minimum at 0.5 M GdnHCl on the GdnHCl dependences of the bOBP fluorescent characteristics. It worth noting that the fluorescence anisotropy value measured for bOBP in this native dimeric state (i.e. in the presence of 1.5 M GdnHCl) exceeds that of bOBP in buffered solution in the absence of denaturant. The presence of crowding agents induced the increase of the fluorescence anisotropy values of bOBP even in the absence of GdnHCl. This may reflect that crowding agents are able to shift monomer-dimer equilibrium toward the formation of native dimeric bOBP. Flattering of the unfolding curves of bOBP in the presence of elevated concentrations of crowding agents indicates the unfolding of bOBP follows the two-state mechanism without accumulation of any intermediate states. These observations provide further support for the crowding agent-induced reorganization of bOBP to native dimeric state. This effect depends on the crowding agent used and on its concentration.

In the case of high concentrations of PEG-4000 and PEG-12000, the high values of parameter A and fluorescence anisotropy r, as well as the sigmoid shape of the corresponding unfolding curves testify for the fact that crowding agents stimulates preferential transition of the protein to its native dimeric form. As a result, under these conditions, bOBP unfolds according to the all-or-none model (Figs. 4 and 6). However, at lower concentrations of these crowding agents, only slight increase of part of native dimeric state of bOBP occurs.

The ANS fluorescence intensity in the presence of PEG-4000 or PEG-12000 shows almost no dependence on denaturant concentration (Figs. 3 and 4), which is further support for the interruption of any interaction of the ANS molecules and the protein in the presence of studied crowding agent.

Curiously, similar to the results reported in our previous study (Stepanenko et al., 2016c), analysis of the recombinant bOBP unfolding in the presence of various concentrations of different crowders revealed that the GdnHCl dependence of various structural characteristics depends on the incubation time of this protein in the presence of the denaturant (see Figs. 2–4). In fact, during the unfolding in crowded milieu, equilibrium and quasi-equilibrium values of the analyzed structural characteristics of the recombinant bOBP were reached after the incubation of this protein in the presence of the desired GdnHCl concentration for 72 hrs. This analysis also revealed the presence of noticeable hysteresis between the curves describing the unfolding and refolding of bOBP when the corresponding measurements were conducted after incubation of the corresponding solution for 1 hour before the measurements (data are not shown).

Figure 5A shows that the tertiary structure of the recombinant bOBP was not affected by low concentrations (80 mg/mL) of PEG-600, PEG-4000, and PEG-12000. However, although the near-UV CD spectra of this protein measured in the presence of high concentrations of crowding agents soon after mixing (∼1 h) were different from the corresponding spectrum measured for bOBP alone (see Figs. 5B and 5C), this structural difference disappeared after the prolonged incubation of this protein under the corresponding conditions. The secondary structure of the recombinant bOBP are not changed in the presence of PEG-600, PEG-4000 and PEG-12000, as evidenced by the coincidence of the values of the ellipticity in the far-UV spectrum region recorded for the protein in a buffer solution and in the presence of all crowding agents at all concentrations tested (Figs. 2–4).

The existence of some dependence of the bOBP structure on the time of incubation in the presence of crowders was further supported by the analysis of the intrinsic tryptophan florescence (see bottom panels in Fig. 5). Increase in the incubation time of the recombinant bOBP in the presence of 80 or 150 mg/mL of crowding agents generates fluorescence spectra that practically coincide with the spectrum of intrinsic fluorescence of the protein alone. However, when concentration of the crowding agents was increased to 300 mg/mL, the intensity of the tryptophan fluorescence was noticeably enhanced. In fact, the intensities of the fluorescence spectra measured in the presence of high concentrations of PEG-4000 and PEG-12000 were slightly higher, and spectra measured in the presence of 300 mg/mL PEG-600 markedly exceeded the bOBP fluorescence intensity in the solution without crowding agents.