Prevention of Aggregation and Renaturation of Carbonic Anhydrase

Dec 11, 2014 - ABSTRACT: The prevention of aggregation during renaturation of urea- denatured carbonic anhydrase B (CAB) via hydrophobic and Coulomb...
0 downloads 0 Views 1MB Size
Article pubs.acs.org/Langmuir

Prevention of Aggregation and Renaturation of Carbonic Anhydrase via Weak Association with Octadecyl- or Azobenzene-Modified Poly(acrylate) Derivatives Nicolas Martin,†,‡,§ Juliette Ruchmann,† and Christophe Tribet*,†,‡,§ †

Département de Chimie, Ecole Normale Supérieure-PSL Research University, 24, rue Lhomond, 75005, Paris, France Sorbonne Universités, UPMC Univ. Paris 06, Pasteur, F-75005, Paris, France § CNRS, UMR 8640 Pasteur, F-75005, Paris, France ‡

S Supporting Information *

ABSTRACT: The prevention of aggregation during renaturation of ureadenatured carbonic anhydrase B (CAB) via hydrophobic and Coulomb association with anionic polymers was studied in mixed solutions of CAB and amphiphilic poly(acrylate) copolymers. The polymers were derivatives of a parent poly(acrylic acid) randomly grafted with hydrophobic side groups (either 3 mol % octadecyl group, or 1−5 mol % alkylamidoazobenzene photoresponsive groups). CAB:polymer complexes were characterized by light scattering and fluorescence correlation spectroscopy in aqueous buffers (pH 7.75 or 5.9). Circular dichroism and enzyme activity assays enabled us to study the kinetics of renaturation. All copolymers, including the hydrophilic PAA parent chain, provided a remarkable protective effect against CAB aggregation during renaturation, and most of them (but not the octadecyl-modified one) markedly enhanced the regain of activity as compared to CAB alone. The significant role of Coulomb binding in renaturation and comparatively the lack of efficacy of hydrophobic association was highlighted by measurements of activity regain before and after in situ dissociation of hydrophobic complexes (achieved by phototriggering the polarity of azobenzene-modified polymers under exposure to UV light). In the presence of polymers (CAB:polymer of 1:1 w/w ratio) at concentration ∼0.6 g L−1, the radii of the largest complexes were similar to the radii of the copolymers alone, suggesting that the binding of CAB involves one or a few polymer chain(s). These complexes dissociated by dilution (0.01 g L−1). It is concluded that prevention of irreversible aggregation and activity recovery were achieved when marginally stable complexes are formed. Reaching a balanced stability of the complex plays the main role in CAB renaturation, irrespective of the nature of the binding (by Coulomb association, with or without contribution of hydrophobic association).



(e.g., RNase,8 chymotrypsin and amylases with Eudragit9), or polyanions.10 The use of neutral chains associated with proteins via hydrophobic and hydrogen bonds is also recognized (e.g., polyvinylpyrrolidone mixed with carbonic anhydrase,11 nonionic nanogels of hydrophobized polysaccharides with carbonic anhydrase12 or peroxydase,13 hydrophobic complexes with integral membrane proteins4,14). Assemblies of proteins and polyelectrolyte, including amphiphilic ones, may however destabilize3 or precipitate proteins15 trapping non native conformers16,17 and irreversible aggregates.18 Recovering or preserving the activity of proteins implies therefore that protective sequestration under the form of complexes is not irreversible and is compatible with dissociation to release free, native proteins. This is generally achieved by modification of the buffer conditions (pH, ionic strength, temperature, or by introducing competitive association with the polymer). This step has been called “stripping”.19 When

INTRODUCTION Since proteins and enzymes have become essential active compounds of many drugs and food products, their processing and preservation under a stable, active form in solutions is extensively studied.1,2 In water, colloidal or interfacial assemblies with polymers or particles provide generic means to control protein stability and activity. Complexes with macromolecules (in particular, amphiphilic polyelectrolyte derivatives) were recently tailored to achieve on/off switching of protein activity,3 enhanced solubility of membrane proteins,4 and prevention of stress-induced aggregation,5 or on demand control of adsorption of proteins.6 Several synthetic polymers or dispersions of microgels, often called artificial chaperones, have been designed to sequester noncovalently misfolded proteins and/or intermediate folding states. Sequestration is expected to help the unstable conformers to escape aggregation pathways. In addition, immobilization of proteins under the form of complexes with macromolecules is generally believed to provide higher thermal stability, and tolerance to denaturants.7 Reported examples of water-dispersed protective assemblies include cationic or anionic proteins mixed with polycations © 2014 American Chemical Society

Received: September 11, 2014 Revised: December 5, 2014 Published: December 11, 2014 338

DOI: 10.1021/la503643q Langmuir 2015, 31, 338−349

Article

Langmuir Scheme 1. Chemical Structures of the Poly(acrylate) Derivatives Used in This Study

oxide) chains (PEO).31 Extensive study by Cleland at al. showed the formation of complexes, presumably due to hydrophobic association, between intermediate molten globule and short PEO molecules (Mw < 8000 g mol−1), which enhances solubility and helps renaturation. Yet longer PEO that develop repulsions with proteins32 were ineffective to help renaturation, showing the importance of balanced CAB:polymer interaction. Here, we studied enzyme activity and secondary structure (by circular dichroism measurements) of urea-denatured CAB abruptly diluted in aqueous refolding buffers, with or without polymer additives. The formation of complexes with polymers and aggregation were characterized by static and dynamic light scattering, and by fluorescence correlation spectroscopy (FCS) using fluorescein-labeled CAB (FITC-CAB) and unlabeled polymers. Circular dichroism (CD) and enzymatic activity assays were used to monitor the kinetics of renaturation. The contributions of Coulomb and hydrophobic associations are discussed on the basis of comparison of polymers derived from the same parent chain (poly(acrylic acid)), having or not octadecyl or azobenzene side groups (the later groups being subjected to in situ phototriggered polarity switch). An important outcome of the work is that Coulomb association is sufficient to help correct refolding and that hydrophobic association is not the main origin of chaperon-like properties of CAB:polymer mixed systems.

protective sequestration was based on interaction with hydrophobic self-assemblies (e.g., micelles,20 or hydrophobic clusters polymers12,21−23), the stripping was achieved by addition of cyclodextrin that dissociates the complexes by competitive association with the hydrophobic moieties of polymers or surfactants. Stepwise variations of temperature were used as stripping conditions, when the sequestration was based on association with temperature-sensitive polymers (e.g., poly(propylene oxide),24 poly(N-isopropylacrylamide)25). Light-triggered stripping has also been described in one pioneering study by Akyioshi et al.26 To optimize such systems, it is important to clarify whether enhanced refolding, and regaining of activity after exposure to stressful conditions in the presence of polymer “sequestering” agents, depends on the nature of the interactions developed between the protein and polymer additives. One hypothesis is that polymer-based protection may predominantly be due to a physical isolation of the proteins and minimization of interprotein collisions, irrespective of the hydrophobic/-philic nature of the binding into protective assemblies. Here, we studied the refolding of one of the most studied model enzymes, bovine carbonic anhydrase B (CAB). CAB was renatured in the presence of amphiphilic macromolecules, including light-responsive ones, derived from the same parent poly(acrylic acid). This model enzyme enabled us to assess the role of Coulombic and (light-switchable) hydrophobic interactions with polymers. CAB is a small (MW ≈ 29 kDa) zinc metalloenzyme that catalyzes the hydration of carbon dioxide and hydrolysis of esters.27,28 It has been shown to refold slowly in water (half time of several minutes up to ca. 10 h, with a molten globule intermediate state) and to undergo intermolecular association and irreversible aggregation when it is unfolded.29,30 Assemblies with amphiphiles (surfactants) enhance dramatically the solubility of CAB during its renaturation from urea solutions, and can yield high (>80%) regains of activity,20 suggesting that control of hydrophobic associations contributes to correct refolding. Ionic surfactants are however significantly more efficient than nonionic ones, suggesting a contribution of Coulombic repulsion. CAB has also been successfully refolded in the presence of poly(ethylene



EXPERIMENTAL SECTION

Material. Carbonic anhydrase II from bovine erythrocytes (CAB), urea, 2-amino-2-(hydroxymethyl)-1,3-propanediol (Tris), 2-(Nmorpholino)ethanesulfonic acid (MES), sodium hydrogenocarbonate (NaHCO3), potassium carbonate (K2CO3), Sephacryl 300 HR resin, anhydrous dimethyl sulfoxide (DMSO), p-nitrophenyl acetate (pNPA), tetradecylsulfate (STS), and methyl-β-cyclodextrin (Me-βCD) were purchased from Sigma and used without further purification. Fluorescein 5-isothiocyanate (FITC) was supplied by Fluka. Poly(acrylic acid) (PAA150) of weight-average molecular weight 150 000 g mol−1 was purchased from Polyscience Inc., Warrington (number-average molecular weight Mn 60 000 g mol−1, PI ≈ 6 measured by GPC in 0.5 M LiNO3 with a triple detection GPC Viscotek Malvern instrument, France, equipped with A6000 M 190 PEHMA columns). The octadecyl-modified polymer, PAA150-3C18, containing ca. 3 mol % octadecyl side groups was prepared by coupling 339

DOI: 10.1021/la503643q Langmuir 2015, 31, 338−349

Article

Langmuir the PAA150 parent chain with octadecylamine in N-methylpyrrolidone by a reported procedure.33 The azobenzene-modified PAAs were synthesized by a similar procedure coupling the PAA150 parent chain with 6-amino-N-(4-phenylazophenyl)hexanamide as described in ref 34 yielding random copolymers of varying degree, y%, of modification with the azobenzene (PAA150-yC6azo, Scheme 1). For simplicity, the copolymer containing ca. 3.5 mol % azobenzene will be named PAA150-3C6azo. Preparation of FITC-Labeled CAB. A solution of CAB (2 g L−1) in 0.5 M NaHCO3−Na2CO3 buffer pH 9.5 was dialyzed against the same buffer for 1 h. An aliquot of a solution of FITC (5 g L−1) in anhydrous DMSO was added to the protein solution (50 μg of FITC/ 1 mg of CAB), and the reaction mixture was kept at room temperature for 3 h. It was eluted through a gel permeation chromatograph (Sephacryl 300 HR) with Tris-HCl 50 mM pH 7.75 elution buffer to remove unreacted FITC. The concentration of FITC-CAB in the collected fractions was assessed by spectrophotometry according to the relationship: [CAB] (g L−1) = (A280 − 0.33 × A495)/1.83 (where A280 and A495 are the absorbances at 495 and 280 nm, 0.33 is the correction factor to account for the absorption of the dye at 280 nm, and 1.83 mg−1 mL cm−1 is the extinction coefficient of CAB).35 A final molar ratio FITC:CAB of ca. 2:1 was determined from the relationship: FITC/CAB = (A495/69 000) × 29 000/[CAB] (g L−1) (with 69 000 mol−1 L cm−1 being the molar extinction coefficient of FITC and 29 000 g mol−1 the molar mass of CAB) (application note from Pierce FITC Labeling kit, Thermoscientific). Denaturation and Refolding Protocol. The denaturation and refolding protocol was adapted from Hanson and Gellman.36 Ureadenatured CAB stock solution (25 g L−1) was prepared by dissolving freeze-dried CAB in 10 M urea aqueous solution. Mixed CAB/ polymers stock solutions (at 1:1 w/w ratio) were prepared by dissolving an aliquot of the freeze-dried polymer in the urea-denatured CAB stock solution to reach the final polymer concentration of 25 g L−1. CAB and mixed CAB/polymer solutions were incubated for 24 h at room temperature, and heated (70 °C for 6 min) just before being abruptly diluted (×40 to ×830-fold) in a refolding buffer down to a final CAB concentration of 0.1 g L−1 (CD experiments) or 0.6 g L−1 (for light scattering measurements), and 0.03 or 0.6 g L−1 (for activity and solubility assays). Except when specified, refolding buffer was 10 mM Tris-HCl pH 7.75, or 10 mM MES pH 5.9. As a reference blank experiment, we implemented the detergent-assisted refolding previously reported by Hanson and Gellman:36 the urea-denatured CAB solution, heated at 70 °C for 6 min, was diluted to reach the CAB concentration of 0.043 g L−1 in 50 mM Tris buffer pH 7.75 containing STS at 0.57 mmol L−1. Mixed CAB/STS solution was incubated for 2 h, and then supplemented with Me-β-CD added in excess (as the STS stripping agent, final concentrations 0.03 g L−1 CAB, 0.4 mmol L−1 STS, 4 mmol L−1 Me-β-CD). The mixed solution was incubated for 2 additional hours prior to solubility and enzymatic activity assays. CAB concentration was determined by UV−vis measurement on a Thermo Scientific Evolution Array spectrophotometer using an extinction coefficient of 1.83 (g L−1 protein)−1 cm−1 at 280 nm35 after subtraction of the absorbance value at 314 nm (to limit the contribution of light scattering). Circular Dichroism. CD measurements were carried out at 20 °C with a Jasco J/815 spectrophotometer using quartz cells of 1 mm path length. The specific ellipticity [θ] was calculated according to the equation: [θ ] =

0.1 × θ × MR l×c

wavelength of 637 nm. The time at which we diluted urea-denatured mixed CAB:polymer or CAB solutions (filtered through a Millex 0.2 μm syringe filter) in the filtered renaturation buffer was defined as time 0. The diluted CAB solutions were kept at 25 °C, and variations in the scattering intensity were measured as a function of incubation time. Measurements of hydrodynamic radius were performed on the same apparatus in the dynamic mode. Fluorescence Correlation Spectroscopy. FCS measurements were performed on a home-built two-photon excitation system equipped with a mode-locked Ti:sapphire laser (Mira900, Coherent, Auburn, CA) pumped by a solid-state laser at 532 nm (Verdi, Coherent). The laser beam (780 nm, ∼100 fs pulse width) was focused into the sample using a 60× water immersion microscope objective (1.2 NA, UPlanApo, Olympus). The power was kept below 10 mW by means of neutral filters. The fluorescence signal was collected through the same objective lens, filtered, and reflected by dichroic filters to select fluorescein fluorescence (580 ± 30 nm). The collected light was then separated by a beam splitter and focused on the 200 μm2 working surfaces of two APDs (SPCM-AQR-14, PerkinElmer, Vaudreuil, Canada). At the concentrations used in this Article, the typical signal on each APD was about 3−10 kHz. The signal outputs of the APD modules (TTL pulses) were acquired by a digital autocorrelator module (ALV-6000, ALV-GmbH, Langen, Germany), which computed the cross-correlation function of the fluorescence fluctuations, g(t). The data were fitted to a simple two-mode decay equation, representing a mixture of two populations with different diffusion coefficients: g (t ) =

g01 1+

t τ1

+

g02 1+

t τ2

(2)

where the diffusion times, τ1 and τ2, are related to the diffusion coefficients, D1 and D2, of fluorescent objects (fluorescently labeled proteins and/or aggregates) through the equation, τι = ω2xy/8Di, where ωxy is the radial size of the excited volume. Calibration was performed with solutions of fluorescein in water at pH 10, using the diffusion coefficient of 4.25 × 10−10 m2 s−1.37 The hydrodynamic radius was calculated via the Stokes−Einstein equation. Measurements were performed at 25 °C on FITC-CAB solutions (0.01 g L−1 in 10 mM Tris-HCl pH 7.75, or 10 mM MES pH 5.9, or in 10 M urea) equilibrated overnight with or without polymer (1:1 wt ratio to CAB). Refolded FITC-CAB was characterized as follows: a 2 μL aliquot of a preheated mixed CAB:polymer solution (1.35 g L−1 in 10 M urea) was mixed with 198 μL of the renaturation buffer (either Tris-HCl 10 mM pH 7.75 or MES 10 mM pH 5.9) and then incubated at room temperature. At t = 5 h, a 2 μL sampling of this mixed solution was deposited in the observation chamber of the FCS apparatus to perform measurements of radii (25 °C, 10 acquisitions each of 60 s). Solubility and Enzymatic Activity Assays. Solubility measurements were performed at time 5 h after dilution in refolding buffer (CAB concentration 0.6 g L−1). The solubility was assessed by ultracentrifugation of the CAB solution diluted 20-fold in 50 mM TrisHCl buffer pH 7.75 (1 mL, 10 min at 200 000g). The concentration of CAB prior and after centrifugation was determined from the absorbance at 280 nm (after subtracting the absorbance at 314 nm to remove the contribution of turbidity). To determine the activity of CAB after incubation in refolding buffers (0.6 g L−1 CAB in either 10 mM Tris-HCl pH 7.75 or 10 mM MES pH 5.9), measurements of the initial rate of hydrolysis of pnitrophenyl acetate (pNPA) were performed in Tris-HCl 50 mM pH 7.75 using an excess of pNPA substrate.35 Because the enzymatic activity of CAB depends on pH and the protein is completely inactive at pH 5.9,35 activity assays were performed at pH 7.75 (by 20× dilution into a 50 mM Tris-HCl pH 7.75 buffer of aliquots of the CAB solutions incubated in refolding buffer). Typically, a 25 μL sampling of the CAB preparation at 0.6 g L−1 (ca. 21 μM) was diluted in 500 μL of 50 mM Tris-HCl buffer pH 7.75, and supplemented at time zero with 5 μL of 100 mM pNPA in acetonitrile (1 mmol). The absorbance at 400 nm was monitored for 200 s in an Evolution Array

(1)

where θ is the measured ellipticity in millidegrees, MR is the mean residue molar mass (MR = 111.9 g mol−1 because the protein is comprised of 259 residues and has a molar mass of ∼29 kDa), l is the path length (in cm) of the cell, and c is the CAB concentration (in g L−1). Static and Dynamic Light Scattering. An index of aggregation of CAB was assessed by measuring the intensity of the light scattered at a fixed angle of 90° using a BI-200SM System (Brookhaven Instruments Corp.) equipped with a 30 mW laser LED operating at a 340

DOI: 10.1021/la503643q Langmuir 2015, 31, 338−349

Article

Langmuir

Figure 1. Far-UV CD spectra of 0.1 g L−1 CAB (A) in 10 M urea, and of (B) native CAB at pH 7.75, or (C) at pH 5.9 in the absence or presence of polymers at 1:1 w/w ratio.

Figure 2. Time-dependence of far-UV CD spectra of CAB during refolding in the presence of PAA150 (A,C) or PAA150-3C18 (B,D) at 1:1 wt ratio, either at pH 7.75 (A,B) or at pH 5.9 (C,D). The protein, initially at 25 g L−1 in 10 M urea, with or with no polymer, was diluted at time zero in aqueous buffer (10 mM Tris-HCl pH 7.75 or 10 mM MES pH 5.9) to reach the final concentration of 0.1 g L−1. spectrophotometer (ThermoScientific), and we determined the slope, S, of initial variation of the absorbance with time. Similar measurements without CAB or with freshly solubilized native CAB gave the background self-hydrolysis rate, S0 (buffer only), and the maximal rate, Sref (solution of 0.03 g L−1 CAB freshly prepared in Tris-HCl buffer pH 7.75 from the stock protein powder). The % activity recovery is given by (S − S0)/(Sref − S0).

PAA150 or PAA150-3C18 added at 1:1 wt ratio. The CD spectra of urea-denatured CAB in the absence of polymer consisted of a single and pronounced minimum at 200 nm ([θ] ≈ −14 500 deg cm2 dmol−1 residue−1) characteristic of a predominant fraction of random-coil conformation.38 The CD spectrum in the presence of PAA150 did not differ significantly from that in the polymer-free case. In the presence of PAA1503C18, the minimum at 200 nm was shifted ([θ] ≈ −11 500 deg cm2 dmol−1 residue−1), and a shoulder appeared between 215 and 230 nm, suggesting that the hydrophobically modified PAA150-3C18 interacts with urea-unfolded CAB and favors partial folding (Figure 1A). Spectra of the native CAB in either Tris-HCl buffer pH 7.75 or MES buffer pH 5.9 (no urea) were broad and reached minimum values in the wavelength range



RESULTS AND DISCUSSION Secondary Structure of CAB Studied by Circular Dichroism. CD spectra of CAB incubated for 24 h in reference conditions (native CAB in renaturation buffer, or urea-denatured CAB in aqueous 10 M urea solution) are reported in Figure 1, in the absence or presence of polymer 341

DOI: 10.1021/la503643q Langmuir 2015, 31, 338−349

Article

Langmuir

Figure 3. Time-dependence of enzyme activity recovered in mixed urea-denatured CAB:polymer (1:1 w/w) solutions. Aliquot of stock solution (25 g L−1 CAB in 10 M urea, with or with no polymer) was diluted at time zero (A) down to 0.03 g L−1 CAB or (B,C) down to 0.6 g L−1 CAB. Dilution buffers were (A,B) 10 mM Tris buffer 10 mM pH 7.75 and (C) 10 mM MES pH 5.9. Reference measurement in the absence of polymers is quoted “CAB” in the figures. Lines are guides to the eye.

(Figure S1 in the Supporting Information), which was likely due to the gradual aggregation of a fraction of the protein. In the presence of PAA150 or PAA150-3C18, the renaturation procedure yielded solutions that were essentially transparent by eye (Figure S2 in the Supporting Information), and free of micrometer-large aggregates, making it possible to collect relevant CD signals. CD spectra in Figure 2 indicate that the secondary structure gradually evolved for hours. The refolding rate depended on the nature of polymer and pH. At pH 7.75 (Figure 2A,B), the ellipticity increased faster in the presence of PAA150 than in the presence of PAA150-3C18 (between t = 0 and t = 1 h). A single minimum at 205 nm was observed in spectra with PAA150, whereas a marked shoulder was visible at 220 nm in the presence of the hydrophobically modified PAA150-3C18. The minimum value of the mean residue ellipticity at time t = 0 (in practice at ca. 3 min after dilution) was lower in the presence of PAA150 than with PAA150-3C18. These results suggest that different folding intermediates are formed according to the hydrophobic or hydrophilic nature of the polymer. Similar comments apply to the data at pH 5.9 (Figure 2C,D), even if the difference between the two polymers was less pronounced in this case. Loss of cooperativity of folding as determined by measurements of CD and intrinsic fluorescence in solutions of increasing urea concentrations (Figures S5 and S6 in the Supporting Information) was an additional indication of the marked effect of PAA150-3C18 on the conformational stability of CAB. At pH 7.75, days-long incubation was not sufficient to recover the CD spectrum of native CAB, suggesting that a fraction of CAB was trapped in misfolded states. This may occur either because of oligomerization (object of radii < 40 nm were detected by DLS, see below) or due to association with polymers. The reference “native” spectrum corresponds here to a solution of native CAB without polymer. Interestingly, the CD spectrum obtained after a week of incubation with PAA150 at pH 5.9 closely resembled that of the native CAB, indicating

210−220 nm. Native CAB is known to contain a central 10stranded β-sheet as the dominant secondary structure element, surrounded with short α-helices, turns, and coil structures39,40 accounting for a total amount of 15% α-helices and 30% βsheets (cf., Saito et al.41). These fractions are consistent with estimates of % secondary structures as determined by Dichroweb analysis of experimental CD spectra42 that returned values of 11% (respectively 13%) α-helices and 36% (respectively 34%) β-sheets at pH 7.75 (respectively 5.9) (CONTIN program, Set 4 database). A shift of CD toward lower wavelengths and lower ellipticities occurred at both pH’s in the presence of the amphiphilic polymer PAA150-3C18, pointing here to the destabilization of the native structure by this polymer (Figure 1B,C). Both the polymers and the protein (isolectric point 5.9) are negatively charged at pH 7.75, which suggests that the destabilization of CAB occurs upon hydrophobic association. The presence of PAA150-3C18 had a more pronounced effect at pH 5.9 as compared to pH 7.75, which is compatible with a strengthening of associations due to decreasing Coulomb repulsion at pH near the pI of CAB. CD Study of the Refolding Kinetics. To estimate the refolding rate of CAB in the presence of poly(acrylate) derivatives, we recorded CD spectra as a function of incubation time after a 250-fold dilution of urea-unfolded CAB (25 g L−1, with or without polymer) into aqueous renaturation buffers (cf., Experimental Section). Two buffer conditions were considered, 10 mM Tris-HCl pH 7.75 or 10 mM MES pH 5.9 (Figure 2). In the absence of polymers, measurements could not be carried out at pH 5.9 because of a rapid aggregation of the protein resulting in high turbidity. At pH 7.75, the renaturation procedure without polymers yielded less turbid solutions, and CD spectra were acquired (shown in Figure S1 in the Supporting Information) but not further analyzed as residual scattering may distort spectra.43 Without polymers, the CD signal gradually diminished during incubation, and ellipticity was always lower than the one measured for native CAB 342

DOI: 10.1021/la503643q Langmuir 2015, 31, 338−349

Article

Langmuir that the recovery of the secondary structure is possible, but slow. At the same pH of 5.9, and in the presence of PAA1503C18, the CD spectrum at time 7 days approached the spectrum of mixed native CAB:PAA150-3C18 that is different from the spectrum of native CAB (Figure 1). Hydrophobic association with PAA150-3C18 presumably hinders a complete recovery of the native structure. In conclusion, results from CD highlight the importance of both Coulombic and hydrophobic associations and a significant role of PAA backbone in the interactions with CAB, including at pH 7.75 when the polymer chains and CAB are both negatively charged. Kinetics of Activity Recovery. To assess the recovery of the native state, we measured the rate of hydrolysis of pNPA at fixed concentration of CAB and pH 7.75 (see Experimental Section for details). The initial rate was normalized by the rate measured with a fresh aqueous solution of native CAB (i.e., that had never been unfolded) to obtain the degree of activity recovery in the presence or absence of polymers (1:1 weight ratio with CAB). In practice, unassisted (polymer-free) renaturation of CAB at pH 7.75 yielded batch to batch fluctuations of the activity recovered. The high sensitivity of the degree of renaturation to small variations of the unfolding and incubation conditions in urea has been reported,29,36 and it was ascribed to temperature-dependent and slow evolution of persistent secondary structures of CAB in urea. To limit batch to batch fluctuations, we considered only batches that enabled us to reach similar activity recovery in the absence of polymer at pH 7.75 (ca. 30% here), and measurements with and without polymers were carried out in parallel on the same batch of CAB. The recovery of activity was monitored as a function of incubation time in mixed CAB:polymer solutions (0.03 or 0.6 g L−1) in renaturation buffers at pH 5.9 or 7.75 (Figure 3). The activity gradually increased and reached a pseudoplateau at incubation times > 5 h. In the absence of polymers, the recovery at time 5 h was generally ∼30% (pH 7.75) or ∼10% (pH 5.9). Similarly low degrees of recovery were reached when incubations were performed in the presence of polymers and high dilution of the CAB:polymer mixture (0.03 g L−1). In contrast, incubation with PAA150 at a final dilution of 0.6 g L−1 clearly improved the degree of renaturation, which could reach 58% in mixed CAB:PAA150 solutions (Table 1). In the presence of PAA150-3C18, activity was generally lower than in the absence of polymer, suggesting that associations with this amphiphilic polymer hampered a correct folding or competed with the substrate for binding on the active site. Measurements of the activity of native CAB (freshly dissolved in aqueous buffer) in the presence of PAA150-3C18 indicated that this amphiphilic copolymer affects the enzyme activity (17% decrease of activity, see Table S1 in the Supporting Information), although to a much lower extent than the almost inactivation observed in the renaturation from urea solutions. The lack of activity during the renaturation procedure is thus attributed to a lack of refolding due to interactions developed by PAA150-3C18 with unfolded, or partially folded, forms of CAB. At fixed protein concentration (0.6 g L−1), the recovery of activity was dependent on the PAA150:CAB weight ratio (Figure 4). The activity increased gradually with increasing amount of PAA150 up to a threshold CAB:polymer ratio of ca. 1:1, indicating that the 1:1 w/w used all along the present report is fully effective at enhancing the renaturation. Effect of the Chemical Structure of Polymers, LightResponsive Hydrophobic Groups. To compare the

Table 1. Enzyme Activity and Solubility after 5 h of Incubation of Urea-Denatured CAB Diluted 40× at Time Zero (0.6 g L−1) in Aqueous Renaturation Buffersa buffer

additive

activity (%)

solubility (%)

pH 7.75

no additive (1) STS, (2) MeβCDb PAA150 PAA150-3C18 PAA150-1C6azo PAA150-3C6azo PAA150-5C6azo no additive PAA150 PAA150-3C18

30 63 58 4−7 31 52 30 7−16 52−56 1−7

19−25 67 87 92 nd nd nd 19 92 83

pH 5.9

a

When present, polymers were added at 1:1 wt ratio. Buffers were 10 mM Tris-HCl pH 7.75 or 10 mM MES pH 5.9. Experimental error on the activity measurements were within ±2%. nd: Solubility of CAB could not be determined because of the high absorbance of azobenzene-containing polymers. bThe line quoted STS, MeβCD corresponds to the procedure of renaturation with sodium tetradecylsulfate surfactant as implemented by Hanson et al.36

Figure 4. Dependence of enzyme activity on CAB:polymer ratio after a 5-h long renaturation of urea-denatured CAB (0.6 g L−1). The polymer was added in the 10 M urea-denatured CAB solution prior to dilution at time zero of the protein in renaturation buffer (either TrisHCl 10 mM pH 7.75 or MES 10 mM pH 5.9). The lines are guides to the eye.

efficiency of copolymers carrying different hydrophobicities (octadecyl or azobenzene hydrophobic side groups, Scheme 1), we fixed somewhat arbitrarily the composition of mixed CAB:polymer solutions (1:1 wt ratio, 25 g L−1 in 10 M urea) and incubation conditions (dilution to 0.6 g L−1 and incubation for 5 h). The measured degree of activity recovery in the presence of copolymers varied from 4% to 60%, pointing in Table 1 to the importance of the polymer structure. Using PAA150 yielded a recovery similar to that of the reference surfactant-based procedure (incubation in STS aqueous buffer, and cyclodextrin stripping step as reported by Hanson et al.36). Yet at variance with renaturation in the presence of surfactants, no stripping step (i.e., a step of sequestration of the refolding additive) was needed with PAA150. This indicates that either CAB is active under the form of complexes with the polymer chains or CAB:PAA150 complexes spontaneously dissociate. In the presence of PAA150, the % activity recovery was significantly increased as compared to CAB alone (irrespective of pH), and was accompanied by the regain of almost full solubility of CAB (the soluble fraction in Table 1 was measured as the % CAB in the supernatant of ultracentrifuged solutions, see Experimental Section). In the presence of hydrophobically 343

DOI: 10.1021/la503643q Langmuir 2015, 31, 338−349

Article

Langmuir

Figure 5. Activity recovery of urea-denatured CAB in the presence of PAA150-3C6azo (CAB:polymer 1:1 w/w). CAB (25 g L−1) in 10 M urea was abruptly diluted to 0.6 g L−1 in aqueous buffers (A) 10 mM Tris-HCl pH 7.75 or (B) 10 mM MES pH 5.9. “UV 2h” (respectively “UV 24h”) refers to samples irradiated with UV light after a 2-h (respectively 24-h) incubation in the dark. The lines are guides to the eye.

Figure 6. Time-dependence of the intensity scattered after 40× dilution at time zero of a urea-denatured mixed CAB:polymer solution (1:1 w/w) into the renaturation buffer at pH 7.75 (A,C) or pH 5.9 (B,D). “Dark” refers to samples incubated in the dark during the experiment, and “UV 2h” (respectively “UV 24h”) refers to samples irradiated with UV light after 2 h (respectively 24 h) of incubation in the dark (arrows mark the beginning of UV irradiation). Final protein concentration in the refolding buffer: 0.6 g L−1. The lines are guides to the eye.

To assess the importance of varying hydrophobic interaction during renaturation, we studied the effect of shining UV on CAB:azobenzene-modified PAA mixtures. Azobenzene-containing amphiphiles are known to undergo UV-triggered trans to cis isomerization that can markedly affect their hydrophobic assembly. Lee et al. reported that hydrophobic association of azo-surfactants with proteins can be modulated by exposure to light.44,45 Cationic azobenzene-containing surfactants under their apolar trans form were shown to favor an activity loss of CAB,44 or the unfolding of bovine serum albumin,45 whereas the cis form did not affect the conformation of albumin. Using azobenzene-containing PAA, we showed that the cis−trans isomerization and hydophobic assembly are effectively controlled in situ by exposure to UV light46 and can modulate association with native bovine serum albumin33 or cytochrome c.34 All of these effects were ascribed to higher hydrophobic interactions between the blue-adapted (predominantly trans) azobenzene isomer as compared to the UV-adapted (predominantly cis) one. Here, kinetics of activity recovery in the

modified PAA150-3C18, the solubility was similarly enhanced, but the activity was lower than the activity measured in the absence of polymer. Although the presence of PAA150-3C18 inhibits slightly the activity of CAB (decrease by 17% of native CAB activity, Table S1 in the Supporting Information), inhibition cannot explain the almost lack of activity regain. The absence of renaturation is therefore attributable to hydrophobic interactions with non-native conformers formed during the renaturation. The low degree of renaturation in the presence of amphiphilic polymers is not a general rule. The presence of azobenzene-containing polymers was not detrimental to activity recovery (e.g., in mixed CAB:PAA1503C6azo solutions). We note however that PAA150-3C6azo did not improve the regain of enzyme activity as compared to that reached in CAB:PAA150 solutions devoid of hydrophobic groups. In addition, no simple correlation could be established between renaturation yield and increasing degree of PAA modification with azobenzene (Table 1), suggesting a complicated mechanism. 344

DOI: 10.1021/la503643q Langmuir 2015, 31, 338−349

Article

Langmuir presence of the trans or cis isomers were assessed in mixed solutions with PAA150-3C6azo (Figure 5). Light was shone at time 2 h (i.e., during the initial fast regain of activity at a time corresponding to about half-recovery in the dark, with transazobenzene), or at time 24 h after dilution in the renaturation buffer (either pH 7.75 or pH 5.9). Immediately after exposure to light (exposure at 365 ± 10 nm, 1 mW cm−2), the degree of renaturation was above the one reached in the dark at the same incubation time. Longer incubation with the polymer kept under its cis isomer form by a constant exposure to UV did not however increase the activity regain. Long incubation times are thus more favorable in the dark, due to a gradual increase of renaturation yield that does not occur with cis isomers. Earlier exposure to UV and formation of cis-azobenzene is accordingly expected to be detrimental to the final degree of renaturation, a point that was experimentally confirmed by higher activity recovery reached at time 2 h in the dark (∼18%) as compared to sample exposed to UV immediately after dilution (∼8−9% activity similar to the blank CAB without polymer, Figure S4 in the Supporting Information). It is concluded that hydrophobic association and photoswitch of hydrophobicity did not enable better enhancement of the degree of renaturation as compared to CAB:PAA150 (i.e., predominant Coulomb association). The azobenzene groups may however be of practical interest to contribute to the control of the stability of complexes with CAB (the impact of azobenzene cis−trans isomerization on formation of complexes with CAB is shown in the next section). Study by DLS of the Size of CAB and CAB:Polymer Complexes. Variations of intensity scattered (fixed incident laser intensity, fixed setup, and angle of 90°) by urea-denatured CAB (with or without polymer) diluted in refolding buffers were measured as an index of the presence or absence of aggregates during renaturation. Massive aggregation of CAB alone occurred after dilution to 0.6 g L−1 of urea-unfolded CAB with no polymer in the renaturation buffers. At pH 5.9, large aggregates were visible by the naked eye (they sedimented rapidly, Figure S2 in the Supporting Information). Because of aggregation, dilution of polymer-free urea-denatured CAB into renaturation buffers led to an abrupt jump in scattered intensity (more than 100-fold the intensity scattered by native CAB at the same concentration). The presence of polymer was very effective to prevent this aggregation. Dilution of a mixed ureadenatured CAB:polymer solution into the refolding buffers increased the scattering by typically 4−5× in comparison with the intensity scattered by a mixed native CAB:polymer solution at the same concentration (Figure 6, Table 2, and Table S2 in the Supporting Information). This initial intensity indicated the absence of very large aggregates, but with hydrophobically modified polymers it was significantly higher than the intensity scattered by native CAB:polymer mixed solutions, pointing to the presence of complexes that did not form with the native protein. In mixed CAB:PAA150 solutions, the scattering intensity was constant during 1 week and was comparable to the intensity scattered by mixtures of native CAB with PAA150. At pH 5.9, this scattered intensity betrays the presence of complexes with native CAB, and at pH 7.75 the scattering by mixed native CAB:PAA150 was equal to the sum of intensities scattered by CAB alone and PAA150 alone, suggesting that no or weak complexes were formed (see Table S2 in the Supporting Information). In CAB:PAA150-3C18 mixture, the intensity was initially up to 5-fold higher than in the CAB:PAA150 solutions.

Table 2. Characteristic Features Determined by LightScattering Measurements of Mixed Urea-Denatured CAB:Polymer Solutions (1:1 w/w), Diluted to 0.6 g L−1 in 10 mM Tris-HCl pH 7.75 or 10 mM MES pH 5.9 and Incubated for 1 week (t = 168 h) condition

additive

intensity at t = 168 h (kcps)

pH 7.75

PAA150 PAA150-3C18 PAA150-3C6azo Dark PAA 150-3C6azo UV PAA150 PAA150-3C18 PAA150-3C6azo Dark PAA150-3C6azo UV

39b 107b 47b 34 74b 276b 62b 44b

pH 5.9

normalized intensitya

Rh at t = 168 h (nm)

1.34 1.67 1.24 1.06 0.98 1.05 1.00 1.05

28.6 41.6 155 nd 22.6 38.1 48.5 nd

a

The normalized intensity is the raw scattered intensity divided by the intensity scattered by mixed native CAB:polymer solutions at the same concentrations. bIndicates that the presence of complexes was evidenced by a markedly higher scattering than the sum of scattered intensities of native CAB alone and polymer alone.

It decreased slowly to reach the intensity of the mixed native CAB:PAA150-3C18 solution after 1 week, suggesting that aggregates that formed at time zero have slowly disappeared to yield complexes having properties similar to those of native CAB:polymer complexes. The scattered intensity of a mixture of native CAB and PAA150-3C18 was up to 5 times higher than the sum of intensities scattered by native CAB alone and the polymer alone at pH 5.9 (1.3× at pH 7.75), indicating that complexes with native CAB are formed (see Table S2 in the Supporting Information). Scattering by mixed CAB:PAA1503C6azo solutions showed similarly the initial formation of complexes and/or aggregates that gradually reached the size of complexes formed between native CAB and the polymer. The apparent hydrodynamic radius of the scattering species was determined by dynamic light scattering measurements. In the presence of polymers, objects of Rh ≈ 20−40 nm were the predominant scatterers in solution and persisted after incubation for several days (Table 2). The radii of PAA150 alone (12.2 nm at pH 7.75 and 13.5 nm at pH 5.9) and PAA150-3C18 (38.8 nm at pH 7.75 and 40.7 nm at pH 5.9) measured by DLS are in the same range as the largest objects present in the CAB:polymer mixed solutions, which is likely due to the stabilization of CAB under the form of associates with one or few polymer chains. When UV light (365 nm, 1 mW cm−2) was shone on CAB:PAA150-3C6azo mixtures in the refolding buffer, the intensity was further decreased by ∼30% as compared to samples kept in the dark (Figure 6C,D). A photoinduced intensity drop occurred in both renaturation buffers at pH 5.9 or 7.75, irrespective of a 2-h or 24-h preincubation in the dark. The magnitude of a UV-triggered decrease of scattered intensity was markedly higher in CAB:polymer mixtures (20−30 kcps) than in solutions of the polymer alone (3−4 kcps, Supporting Information Table S2). The slight decrease observed in the polymer solution without CAB was likely due to photovariation of polarity (trans−cis isomerization), but it is negligibly small as compared to the drop of scattering in the presence of CAB:polymer complexes. Complexes between the polymer and CAB are thus dissociated upon trans to cis isomerization of the azobenzene side groups (either the one formed in mixed native CAB:polymer, Table S2 in the 345

DOI: 10.1021/la503643q Langmuir 2015, 31, 338−349

Article

Langmuir

Table 3. Hydrodynamic Radius Rh of Unstressed, 10 M Urea-Unfolded, and Renatured Urea-Unfolded FITC-CAB As Measured by FCS in the Absence or Presence of Polymers at 1:1 w/w Ratioa pH 7.75

pH 5.9

sample

10 M urea

unstressedb

renaturedc

unstressedb

renaturedc

CAB CAB/PAA150 CAB/PAA150-3C18

3.2 ± 0.6 2.9 ± 0.7 2.5 ± 0.6

2.3 ± 0.2 2.1 ± 0.2 2.7 ± 0.6

2.1 ± 0.3 2.4 ± 0.2 2.1 ± 0.5 and 39.7 ± 1

2.9 ± 0.3 2.7 ± 0.4 3.1 ± 0.2 and 106 ± 2

2.2 ± 0.4 2.1 ± 0.3 2.5 ± 0.6 and 57.2 ± 2

a Protein concentration: 0.01 g L−1. bRefers to aqueous solutions of FITC-CAB that were not unfolded (no urea, no thermal stress). cAt t = 5 h after dilution in the refolding buffer.

result indicates the persistence of at least a fraction of the association formed with PAA150-3C18 during renaturation, and thus an incomplete release of FITC-CAB. Discussion on the Role of Coulomb and Hydrophobic Associations. The present results provide a clear indication that poly(acrylate) derivatives prevent massive aggregation of CAB by forming CAB:polymer complexes during refolding. Complexes may form with native CAB (e.g., PAA150-3C18, or most polymers at pH 5.9), but the protective effect is also present when polymer chains have a low affinity for the native CAB (as evidenced at pH 7.75 by the scattered intensities of native CAB:PAA150, or CAB:PAA-3C6Azo, mixed solutions almost equal within experimental errors to the sum of scattered intensities by CAB alone and polymer alone, cf., Table S2 in the Supporting Information). Interestingly, a predominant Coulomb interaction (with PAA150) is sufficient to achieve high yield of renaturation. In this case, prevention of aggregation during folding (chaperone-like effect) is likely due to the formation of complexes between the polyanionic PAA and quasi neutral (pH 5.9) or anionic (pH 7.75) CAB. We recently reported that poly(acrylate) derivatives confer colloidal stability to heat-unfolded immunoglobulins G near their isoelectric point, predominantly via electrostatic interactions.49 A few other experimental reports point out that renaturation is facilitated in systems predominantly involving Coulomb association between proteins and polyelectrolytes. For instance, urea-denatured lipase could be refolded by addition of alginate with high guluronic acid content,9b,50 and a copolymer of methylacrylic acid and acrylic acid enhanced the renaturation yield of lysozyme, a basic protein, by providing a negatively charged surface for an electrostatic interaction with the denatured protein.51 At pH close to pI, such stabilization can be rationalized by invoking interaction with patches of charges opposite to the whole charge of the protein (Scheme 2). Binding of polyelectrolytes confers a high total charge to the complex, limiting interprotein contacts at pH ≈ pI, and therefore preventing aggregation. Persistence of positive patches at pH > pI (Scheme 2) can provide attractive interaction to maintain a similar protective effect of PAA, despite that the association strength is certainly weakened with increasing pH. These qualitative features are based on the native structure of CAB, which is not the unstable form. Aggregation comes from irreversible bridging pathways involving intermediate folding states. Because the fully unfolded CAB is unlikely to display any charge patches, we propose that the observed protection by PAA150 essentially comes from the formation of soluble complexes with molten globules or compact intermediate states whose surfaces can be highly charged and possibly more tightly bound than the native form thanks to a higher flexibility of the interface. As compared to molecular additives that are known to affect solubility (e.g., molar amount of kosmotropic/chaotropic

Supporting Information, or that formed in mixed urea-unfolded CAB:polymer solutions, Table 2). Measurements of Sizes by FCS (FITC-CAB). To obtain size measurements of CAB-containing objects, and discriminate the contributions coming from CAB and from the polymer chains that are additive in light scattering measurements, fluorescence correlation spectroscopy (FCS) was carried out using fluorescently labeled FITC-CAB (0.01 g L−1, see Experimental Section) and unlabeled polymers. Because the polymers are not fluorescent, CAB-free chains are invisible and do not account in the average radii determined by FCS. The detected, FITC-CAB-containing species include CAB monomers, oligomers, aggregates, or complexes with polymer chains. The hydrodynamic radius of native FITC-CAB alone (no polymer) was below 3 nm (Table 3), which is close to values reported in the literature.29,47 Within experimental uncertainties, the same Rh was recorded in mixed FITC-CAB:polymer solutions at pH 7.75. Neither PAA150 nor PAA150-3C18 associated with FITC-CAB at pH 7.75 (and FICT-CAB at 0.01 g L−1 concentration). At pH 5.9, larger species were detected, pointing to the formation of complexes with FITC-CAB and PAA150-3C18 (Rh = 106 nm), but not with PAA150. In 10 M urea and room temperature, the radius of FITC-CAB in the absence of polymer was slightly increased to 3.2 nm, as expected upon unfolding of the protein. In the presence of polymers, Rh remained close to that of CAB alone regardless of the polymer, pointing out that polymers at 0.01 g L−1 do not associate with the unfolded protein in 10 M urea. FCS measurements were then performed on renatured samples, that is, urea-denatured FITC-CAB (1.35 g L−1 in 10 M urea, brought to 70 °C for 6 min) that was diluted and incubated for 5 h in renaturation buffers (same procedure as for CD and activity measurements except that a 135-fold dilution into the renaturation buffer was used). The final concentration of 0.01 g L−1 in FCS measurements was required to avoid saturation of the detector at higher FITC-CAB concentrations. In the absence of polymer, and regardless of pH, the measured radius was ∼2−3 nm, which means that aggregation did not occur (Table 3). Protein concentration is known to play an important role in CAB aggregation, and high dilutions can prevent aggregation.48 The higher dilution as compared to experimental conditions used for CD measurements, and the covalent attachment of FITC, presumably favored a higher solubility and dispersion of the protein alone. In the presence of PAA150, Rh was comparable to Rh of the unbound FITC-CAB, suggesting the absence of complexes between FITC-CAB and PAA150 after renaturation. In contrast, both small (2−3 nm) and larger objects (∼50 nm) were present in the renatured FITC-CAB:PAA150-3C18 mixed system. These large species persisted for 1 week (not shown). Because we showed that at pH 7.75 the polymers at the same concentration do not bind to urea-denatured CAB (in urea), neither to native CAB, this 346

DOI: 10.1021/la503643q Langmuir 2015, 31, 338−349

Article

Langmuir Scheme 2. Electrostatic Surface Isocontour (±1 kT/e) of CAB Calculated from Its Protein Data Bank Structure 1V9E, Using Adaptative Poisson−Boltzman Solver Software,53 and 10 mM Ionic Strength, pH 7.75 (Left) or pH 5.9 (Right)a

pathways (e.g., lower renaturation yields reached with PAA1501C6Azo, PAA150-3C18 as compared to PAA150). The present results point that Coulomb complexes reached here a high chaperon-like efficacy that is not improved by additional contribution of hydrophobic associations.



CONCLUSION Studies of mixed CAB:polymer solutions by FCS, light scattering, and CD measurements have contributed to decipher the origin of polymer-based protection of CAB against aggregation and of the activity recovery during refolding of urea-denatured CAB. Results point to the role of Coulomb association between the poly(acrylate) chain and CAB, even at pH above the pI of the protein. Weak binding of non-native CAB onto the anionic chain of PAA150 presumably hinders aggregation (e.g., between intermediate folding states), while the weak stability of the complexes enables (hours-long) release of native CAB and/or dissociation of the complexes upon dilution. Polymers bearing hydrophobic moieties form more stable associations with CAB, including with urea-denatured CAB, and also prevent aggregation in the aqueous renaturation buffer. Tight binding with C18-modified PAA however impaired the activity recovery and perturbed the secondary structure in aqueous solutions. More balanced conditions of hydrophobic binding were reached in mixed CAB:PAA1503C6azo solutions, which can undergo a light-controlled CAB:polymer dissociation (upon trans−cis photoisomerization of the azobenzene side groups). This photoswitch triggers on demand fast dissociation of CAB:polymer complexes that otherwise persists for more than 1 week during renaturation. Phototriggered release does not however contribute to the activity recovery. This photocontrol may find applications when Coulomb interactions are too weak to protect the protein of interest, which was not the case here. The dominant scheme favoring urea-unfolded CAB renaturation is noncovalent attachment of polymer chains due to formation of labile Coulomb complexes in the aqueous renaturation buffer, and hours-long spontaneous refolding occurred together with slow (stripping-free) dissociation.

a

PyMol was used as a visualization tool. Positive regions are drawn in blue, and red is used for negatively charged ones.

osmolytes), a practical advantage of polymers is the low amount needed (ca. 1:1 wt/wt and micromolar concentration). Under these diluted concentration conditions, polymers do not affect the activity of water as do kosmotropic agents. The origin of protection presumably relies on the formation of a limited numbers of attachment points on the protein, which differs from the more global variation of solubility and hydration achieved with conventional molecular protectants. At variance with surfactant-based renaturation, no stripping step (i.e., no removal of protein-bound additive) was needed, which indicates that the binding was dynamic and compatible with release of the native CAB. A poor stability of complexes is probably playing a role in spontaneous (stripping-free) release of native CAB, and weak binding of the polymer (instead of tight complexes) appears desirable to facilitate renaturation. This is not general in complexes of polyelectrolyte and proteins. In the case of highly charged partners of opposite sign, complexes can be trapped in frozen colloidal aggregates (e.g., lysozyme and polystyrenesulfonate studied by Cousin et al.52). In the present experimental work, concentration conditions enabled formation of complexes (as seen by DLS at 0.6 g L−1), although exposure to light (PAA150-36Azo) or a 60-fold dilution (PAA150) would dissociate them. Using the protein near neutrality strengthens the association as compared to CAB in basic pH condition, but FCS of native CAB:PAA150 showed that complexes did not survive to dilution down to 0.01 g L−1. As regards species formed during refolding (ureaunfolded CAB abruptly diluted in aqueous buffer), DLS showed that large species formed at short times gradually dissociate during incubation. Stabilizing the complexes by introduction of long hydrophobic side groups (complexes present at 0.01 g L−1) impaired activity recovery, because tight hydrophobic association (with PAA150-3C18) presumably traps the protein in an inactive state. In situ photocontrol of complex formation with azobenzene-modified chains did not however increase the maximum degree of renaturation as compared to the (weak) complexes formed with PAA150 devoid of hydrophobes. It stopped the regain of activity when photodissociation occurred early in the renaturation buffer. We conclude that a balanced (here photocontrolled) hydrophobic association is either a marginal contributor to CAB renaturation (PAA150-3C6Azo) or may orient refolding into unproductive



ASSOCIATED CONTENT

S Supporting Information *

Additional data and figures. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the French National Research Agency (program Blanc International, grant ANR 2010-INT 1501, program Investissement d’Avenir ANR-11-LABX-001101, and France-BioImaging infrastructure ANR-10-INSB-04 “Investments for the future”). We are grateful to L. Jullien and Th. Le Saux for their advice in FCS measurements.



REFERENCES

(1) Fu, K.; Klibanov, A. M.; Langer, R. Protein stability in controlledrelease systems. Nat. Biotechnol. 2000, 18, 24−25.

347

DOI: 10.1021/la503643q Langmuir 2015, 31, 338−349

Article

Langmuir (2) Clark, E. d. B. Protein refolding for industrial processes. Curr. Opin. Biotechnol. 2001, 12, 202−207. (3) Kurinomaru, T.; Tomita, S.; Kudo, S.; Ganguli, S.; Nagasaki, Y.; Shiraki, K. Improved complementary polymer pair system: Switching for enzyme activity by PEGylated polymers. Langmuir 2012, 28, 4334−4338. (4) Tribet, C.; Diab, C.; Dahmane, T.; Zoonens, M.; Popot, J. L.; Winnik, F. M. Thermodynamic characterization of the exchange of detergents and amphipols at the surfaces of integral membrane proteins. Langmuir 2009, 25, 12623−12634. (5) Minsky, B. B.; Zheng, B.; Dubin, P. L. Inhibition of antithrombin and bovine serum albumin native state aggregation by heparin. Langmuir 2014, 30, 278−287. (6) Hollmann, O.; Czeslik, C. Characterization of a planar poly(acrylic acid) brush as a materials coating for controlled protein immobilization. Langmuir 2006, 22, 3300−3305. (7) Secundo, F. Conformational changes of enzymes upon immobilisation. Chem. Soc. Rev. 2013, 42, 6250−6261. (8) Ganguli, S.; Yoshimoto, K.; Tomita, S.; Sakuma, H.; Matsuoka, T.; Shiraki, K.; Nagasaki, Y. Improving the heat resistance of ribonuclease A by the addition of poly(N,N-diethylaminoethyl methacrylate)-graft-poly(ethylene glycol) (PEAMA-g-PEG). Macromol. Biosci. 2010, 10, 853−859. (9) (a) Roy, I.; Gupta, M. N. pH-responsive polymer-assisted refolding of urea- and organic solvent-denatured alpha-chymotrypsin. Protein Eng. 2003, 16, 1153−1157. (b) Mondal, K.; Raghava, S.; Barua, B.; Varadarajan, R.; Gupta, M. N. Role of stimuli-sensitive polymers in protein refolding: alpha-amylase and CcdB (controller of cell division or death B) as model proteins. Langmuir 2007, 23, 70−75. (10) Shalova, I. N.; Asryants, R. A.; Sholukh, M. V.; Saso, L.; Kurganov, B. I.; Muronetz, V. I.; Izumrudov, V. A. Interaction of polyanions with basic proteins, 2(a) - Influence of complexing polyanions on the thermoaggregation of oligomeric enzymes. Macromol. Biosci. 2005, 5, 1184−1192. (11) Jiang, Y.; Yan, Y.-B.; Zhou, H.-M. Polyvinylpyrrolidone 40 assists the refolding of bovine carbonic anhydrase B by accelerating the refolding of the first molten globule intermediate. J. Biol. Chem. 2006, 281, 9058−9065. (12) Nomura, Y.; Ikeda, M.; Yamaguchi, N.; Aoyama, Y.; Akiyoshi, K. Protein refolding assisted by self-assembled nanogels as novel artificial molecular chaperone. FEBS Lett. 2003, 553, 271−276. (13) Sawada, S.; Sasaki, Y.; Nomura, Y.; Akiyoshi, K. Cyclodextrinresponsive nanogel as an artificial chaperone for horseradish peroxidase. Colloid Polym. Sci. 2011, 289, 685−691. (14) Dahmane, T.; Rappaport, F.; Popot, J. L. Amphipol-assisted folding of bacteriorhodopsin in the presence or absence of lipids: functional consequences. Eur. Biophys. J. Biophys. Lett. 2013, 42, 85− 101. (15) Li, X.; Fang, Y.; Al-Assaf, S.; Phillips, G. O.; Yao, X.; Zhang, Y.; Zhao, M.; Zhang, K.; Jiang, F. Complexation of bovine serum albumin and sugar beet pectin: Structural transitions and phase diagram. Langmuir 2012, 28, 10164−10176. (16) Ivinova, O. N.; Izumrudov, V. A.; Muronetz, V. I.; Galaev, I. Y.; Mattiasson, B. Influence of complexing polyanions on the thermostability of basic proteins. Macromol. Biosci. 2003, 3, 210−215. (17) Stogov, S.; Izumrudov, V.; Muronetz, V. Structural changes of a protein bound to a polyelectrolyte depend on the hydrophobicity and polymerization degree of the polyelectrolyte. Biochemistry (Moscow) 2010, 75, 437−442. (18) (a) Zhang, H.-M.; Lou, K.; Cao, J.; Wang, Y.-Q. Interaction of a hydrophobic-functionalized PAMAM dendrimer with bovine serum albumin: Thermodynamic and structural changes. Langmuir 2014, 30, 5536−5544. (b) Cousin, F.; Gummel, J.; Clemens, D.; Grillo, I.; Boue, F. Multiple scale reorganization of electrostatic complexes of poly(styrenesulfonate) and lysozyme. Langmuir 2010, 26, 7078−7085. (19) Rozema, D.; Gellman, S. H. Artificial chaperone-assisted refolding of carbonic anhydrase B. J. Biol. Chem. 1996, 271, 3478− 3487.

(20) Rozema, D.; Gellman, S. H. Artificial chaperones: Protein refolding via sequential use of detergent and cyclodextrin. J. Am. Chem. Soc. 2005, 117, 2373−2374. (21) Nomura, Y.; Sasaki, Y.; Takagi, M.; Narita, T.; Aoyama, Y.; Akiyoshi, K. Thermoresponsive controlled association of protein with a dynamic nanogel of hydrophobized polysaccharide and cyclodextrin: Heat shock protein-like activity of artificial molecular chaperone. Biomacromolecules 2004, 6, 447−452. (22) Morimoto, N.; Endo, T.; Iwasaki, Y.; Akiyoshi, K. Design of hybrid hydrogels with self-assembled nanogels as cross-linkers: interaction with proteins and chaperone-like activity. Biomacromolecules 2005, 6, 1829−1834. (23) Cavalieri, F.; Chiessi, E.; Paradossi, G. Chaperone-like activity of nanoparticles of hydrophobized poly(vinyl alcohol). Soft Matter 2007, 3, 718−724. (24) Yoshimoto, N.; Hashimoto, T.; Felix, M. M.; Umakoshi, H.; Kuboi, R. Artificial chaperone-assisted refolding of bovine carbonic anhydrase using molecular assemblies of stimuli-responsive polymers. Biomacromolecules 2003, 4, 1530−1538. (25) (a) Tao, Q.; Li, A.; Liu, X.; Ma, R.; An, Y.; Shi, L. Protecting enzymes against heat inactivation by temperature-sensitive polymer in confined space. Phys. Chem. Chem. Phys. 2011, 13, 16265−16271. (b) Lu, D.; Liu, Z. X.; Zhang, M. L.; Liu, Z.; Zhou, H. M. The mechanism of PNIPAAm-assisted refolding of lysozyme denatured by urea. Biochem. Eng. J. 2005, 24, 55−64. (c) Lu, D. N.; Liu, Z. X.; Zhang, M. L.; Wang, X. G.; Liu, Z. Dextran-grafted-PNIPAAm as an artificial chaperone for protein refolding. Biochem. Eng. J. 2006, 27, 336−343. (26) Hirakura, T.; Nomura, Y.; Aoyama, Y.; Akiyoshi, K. Photoresponsive nanogels formed by the self-assembly of spiropyranebearing pullulan that act as artificial molecular chaperones. Biomacromolecules 2004, 5, 1804−1809. (27) Thorslund, A.; Lindskog, S. Studies of the esterase activity and the anion inhibition of bovine zinc and cobalt carbonic anhydrases. Eur. J. Biochem. 1967, 3, 117−123. (28) Verpoorte, J. A.; Mehta, S.; Edsall, J. T. Esterase activities of human carbonic anhydrases B and C. J. Biol. Chem. 1967, 242, 4221− 4229. (29) Cleland, J. L.; Wang, D. I. C. Refolding and aggregation of bovine carbonic anhydrase B: quasi-elastic light scattering analysis. Biochemistry 1990, 29, 11072−11078. (30) Prokhorov, D. A.; Timchenko, A. A.; Uversky, V. N.; Khristoforov, V. S.; Kihara, H.; Kimura, K.; Kutyshenko, V. P. Dynamics of oligomer formation by denatured carbonic anhydrase II. Biochim. Biophys. Acta, Proteins Proteomics 2008, 1784, 834−842. (31) Cleland, J. L.; Hedgepeth, C.; Wang, D. I. C. Polyethyleneglycol enhanced refolding of bovine carbonic anhydrase-B - Reaction stoichiometry and refolding model. J. Biol. Chem. 1992, 267, 13327− 13334. (32) Hamada, H.; Arakawa, T.; Shiraki, K. Effect of additives on protein aggregation. Curr. Pharm. Biotechnol. 2009, 10, 400−407. (33) Pouliquen, G.; Tribet, C. Light-triggered association of bovine serum albumin and azobenzene-modified poly(acrylic acid) in dilute and semidilute solutions. Macromolecules 2005, 39, 373−383. (34) Sun, J.; Ruchmann, J.; Pallier, A.; Jullien, L.; Desmadril, M.; Tribet, C. Unfolding of cytochrome c upon interaction with azobenzene-modified copolymers. Biomacromolecules 2012, 13, 3736−3746. (35) Pocker, Y.; Stone, J. T. The catalytic versatility of erythrocyte carbonic anhydrase. III. Kinetic studies of the enzyme-catalyzed hydrolysis of p-nitrophenyl acetate. Biochemistry 1967, 6, 668−678. (36) Hanson, P. E.; Gellman, S. H. Mechanistic comparison of artificial-chaperone-assisted and unassisted refolding of urea-denatured carbonic anhydrase B. Folding Des. 1998, 3, 457−468. (37) Culbertson, C. T.; Jacobson, S. C.; Ramsey, J. M. Diffusion coefficient measurements in microfluidic devices. Talanta 2002, 56, 365−373. 348

DOI: 10.1021/la503643q Langmuir 2015, 31, 338−349

Article

Langmuir (38) Kelly, S. M.; Jess, T. J.; Price, N. C. How to study proteins by circular dichroism. Biochim. Biophys. Acta, Proteins Proteomics 2005, 1751, 119−139. (39) Håkansson, K.; Carlsson, M.; Svensson, L. A.; Liljas, A. Structure of native and apo carbonic anhydrase II and structure of some of its anion-ligand complexes. J. Mol. Biol. 1992, 227, 1192− 1204. (40) Stams, T.; Christianson, D. W. X-ray crystallographic studies of mammalian carbonic anhydrase isozymes. EXS 2000, 90, 159−174. (41) Saito, R.; Sato, T.; Ikai, A.; Tanaka, N. Structure of bovine carbonic anhydrase II at 1.95 A resolution. Acta Crystallogr., Sect. D 2004, 60, 792−795. (42) Whitmore, L.; Wallace, B. A. DICHROWEB, an online server for protein secondary structure analyses from circular dichroism spectroscopic data. Nucleic Acids Res. 2004, 32, W668−W673. (43) Urry, D. W. Protein conformation in biomembranes - Optical rotation and absorption of membrane suspensions. Biochim. Biophys. Acta 1972, 265, 115−168. (44) Mirarefi, P.; Lee, C. T., Jr. Photo-induced unfolding and inactivation of bovine carbonic anhydrase in the presence of a photoresponsive surfactant. Biochim. Biophys. Acta, Proteins Proteomics 2010, 1804, 106−114. (45) Lee, C. T.; Smith, K. A.; Hatton, T. A. Photocontrol of protein folding: The interaction of photosensitive surfactants with bovine serum albumin. Biochemistry 2005, 44, 524−536. (46) Ruchmann, J.; Fouilloux, S.; Tribet, C. Light-responsive hydrophobic association of surfactants with azobenzene-modified polymers. Soft Matter 2008, 4, 2098−2108. (47) Piaggio, M. V.; Peirotti, M. B.; Deiber, J. A. Effect of background electrolyte on the estimation of protein hydrodynamic radius and net charge through capillary zone electrophoresis. Electrophoresis 2005, 26, 3232−3246. (48) Cleland, J. L.; Wang, D. I. C. Transient association of the first intermediate during the refolding of bovine carbonic anhydrase B. Biotechnol. Prog. 1992, 8, 97−103. (49) Martin, N.; MA, D.; Herbet, A.; Boquet, D.; Winnik, F. M.; Tribet, C. Prevention of thermally induced aggregation of IgG antibodies by noncovalent interaction with poly(acrylate) derivatives. Biomacromolecules 2014, 15, 2952−2962. (50) Mondal, K.; Bohidar, H. B.; Roy, R. P.; Gupta, M. N. Alginatechaperoned facile refolding of Chromobacterium viscosum lipase. Biochim. Biophys. Acta, Proteins Proteomics 2006, 1764, 877−886. (51) Ye, C.; Ilghari, D.; Niu, J.; Xie, Y.; Wang, Y.; Wang, C.; Li, X.; Liu, B.; Huang, Z. A comprehensive structure−function analysis shed a new light on molecular mechanism by which a novel smart copolymer, NY-3-1, assists protein refolding. J. Biotechnol. 2012, 160, 169−175. (52) Cousin, F.; Gummel, J.; Ung, D.; Boue, F. Polyelectrolyteprotein complexes: Structure and conformation of each specie revealed by SANS. Langmuir 2005, 21, 9675−9688. (53) Baker, N. A.; Sept, D.; Joseph, S.; Holst, M. J.; McCammon, J. A. Electrostatics of nanosystems: Application to microtubules and the ribosome. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 10037−10041.

349

DOI: 10.1021/la503643q Langmuir 2015, 31, 338−349