ARTICLE pubs.acs.org/Langmuir
pH-Dependent Peroxidase Activity of Yeast Cytochrome c and Its Triple Mutant Adsorbed on Kaolinite Antonio Ranieri,† Fabrizio Bernini,† Carlo Augusto Bortolotti,† Alois Bonifacio,‡ Valter Sergo,‡ and Elena Castellini*,† † ‡
Department of Chemistry, University of Modena and Reggio Emilia, Via Campi 183, I-41125 Modena, Italy CENMAT, DI3, University of Trieste, Via Valerio 6/a, I-34127 Trieste, Italy
bS Supporting Information ABSTRACT: The peroxidase activity of wild-type yeast cytochrome c and its triple mutant K72AK73AK79A adsorbed onto kaolinite was investigated as a function of pH and temperature. Both adsorbed proteins displayed an appreciable catalytic activity, which remained constant from pH 7 to pH 10, decreased below pH 7, and showed a remarkable increase at pH values lower than 4. In the whole pH range investigated the catalytic activity of the adsorbed wild-type cytochrome c was higher than that of the mutant. Both diffuse-reflectance UVvis and resonance Raman spectroscopies applied on solid samples were used to probe the structural features responsible for the catalytic activity of the immobilized proteins. At neutral and alkaline pH values a six-coordinate low-spin form of cytochrome c was observed, while at pH < 7 the formation of a high-spin species occurred whose population increased at decreasing pH. The orientation and exposure of the heme to the substrate—strictly dependent on adsorption—was found to affect the peroxidase activity.
1. INTRODUCTION Peroxidases are widely diffused heme enzymes found in bacteria, fungi, plants, and animals which catalyze a variety of oxidative reactions of organic and inorganic substrates, exploiting hydrogen peroxide as an oxidizing agent.1 As a consequence, they also act as antioxidant protective enzymes.2 These enzymes are applied in industrial and environmental biocatalysis. As an example, horseradish peroxidase is employed to remove phenols in the presence of H2O2 from contaminated aquifers by polymerization and precipitation in porous media.3 Moreover, a bienzymatic system made of a glucose oxidase and horseradish peroxidase immobilized on phospholipid-templated silica nanocapsules has been recently exploited as a catalyst in the reaction of 4-aminoantipyridine and phenol with in situ produced H2O2.4 Peroxidases are also known to be effective in the degradation of dyes using hydrogen peroxide:5 dyes recalcitrant to common chemical bleaches were successfully treated using horseradish and soybean peroxidases.6 The concomitant use of redox mediators facilitated the peroxidase-catalyzed color removal from textile effluents and dyes.7 However, peroxidases as catalysts present a number of disadvantages, mainly related to their low stability, the difficulty to undergo site-directed mutagenesis to optimize their catalytic activity, and, finally, the high cost. Therefore, the development of stable proteins imparted with peroxidase activity, which can be tuned through a mutational approach, would open new interesting perspectives. In this view, cytochrome c is a very good r 2011 American Chemical Society
candidate. In fact, the ability of cytochrome c to induce lipid peroxidation and its involvement in hydroperoxide cleavage has been known since 1960.8 In 50 years the peroxidase activity of cytochrome c toward several substrates in the presence of hydrogen peroxide or organic hydroperoxides has been demonstrated.9 Moreover, cytochrome c is remarkably stable in vitro, it can be easily engineered and purified,10 and the heme prosthetic group is firmly bound to the polypeptide chain,9 thus preventing the loss of the heme in organic solvents. Furthermore, cytochrome c retains its catalytic activity over a wide pH range (211) and at high concentrations of organic solvents.9 Diederix and co-workers11 demonstrated that peroxidase activity of cytochrome c requires a free coordination position on the heme iron. The activity increase upon unfolding can be ascribed to the loss of the native Met ligand and increased access to the heme iron. Also the improved peroxidase activity which is observed in the presence of denaturing agents is attributed to partial protein unfolding. The peroxidase activity of cytochromes c specifically relates to non-native five-coordinate species11 or native five-coordinate species.12 Therefore, induction of peroxidase activity in cytochrome c requires a conformational transition to turn the cytochrome c architecture into a peroxidase-like structure. This involves disruption of the sixth axial ligation and Received: May 19, 2011 Revised: July 14, 2011 Published: July 21, 2011 10683
dx.doi.org/10.1021/la201876k | Langmuir 2011, 27, 10683–10690
Langmuir
ARTICLE
opening of a channel in the heme crevice to allow H2O2 and substrate access to the heme while retaining the integrity of the heme pocket.12 The peroxidase activity of cytochrome c can be induced by protein adsorption on a proper supporting solid material.10 The immobilizing material must possess a wide accessible surface area to increase the extent of enzymesubstrate interaction. This strategy also allows recovery of the solid catalyst from the medium after reaction and surface renewal, also preventing the contamination of the medium by the protein.4 In a previous paper, we investigated the thermodynamic aspects of the adsorption of cytochromes c and Lys-to-Ala mutants on kaolinite clay.10 The kaolinite was selected a priori as an adsorbing material due to the availability of a pure mineral, the high specific surface area, the absence of swelling layers, the high chemical stability, and the low cost. In the present work, we have studied the peroxidase activity of yeast cytochrome c and of its triple mutant K72AK73AK79A adsorbed on kaolinite as a function of pH. The main purpose is to determine the low- and high-pH limits of the peroxidase activity of the two adsorbed species. Moreover, both diffuse-reflectance UVvis and resonance Raman spectroscopies were employed to gain insight into how adsorption onto kaolinite affects the catalytic activity of cytochrome c, acting on the orientation and exposure of the heme to the solution. Wang and Waldeck13 reported that the folding of cytochrome c is influenced by immobilization onto SAM films. Moreover, it has been recently reported by Hung et al.14 that the tertiary structure of cytochrome c may be affected by adsorption onto different SAMcoated metal nanoparticle surfaces.
one. At pH 9.5 YCC in solution is characterized by the substitution of the axial Met 80 ligand with a lysine residue (alkaline form).21 The starting YCC and TMut solutions were prepared by dilution respectively of the salt (0.15 mg/mL) and of the mother solution in the wanted buffer solution (pH 3.5 and 9.5) and then were passed through a preconditioned exclusion column (Sephadex G-15), and their concentrations were checked spectrophotometrically (UVvis, JASCO model V-570) and adjusted to reach an absorbance value of 0.62 at the λmax of the Soret band. In fact, preliminary measurements showed that this absorbance value corresponds to a starting cytochrome concentration which provides the maximum coverage of the Nak surface. The batches of suspensions were prepared by mixing 2 mg of Nak with 1 mL of cytc solution followed by shaking at 250 rpm in an orbital incubator (Stuard Scientific Orbital Incubator SI50) for 2 h at 298 K, as described in ref 10. The suspensions were then allowed to stand in the thermostated incubator. After solidliquid separation, the supernatant was centrifuged at 12500 rpm (14400g) (Labnet Spectrafuge 24D) for 2 min. UVvis spectra of the clarified supernatants were recorded using a Jasco V-570 spectrophotometer. The Soret band was selected for quantitative analysis to calculate the moles of adsorbed cytc. The shape of the UVvis spectrum invariably showed that all the cytochromes were in their oxidized form. The solid phases were washed three times with the same buffer used in the adsorption procedure. Washing caused no loss of cytochrome c. These washed solids constitute the solid catalysts NakYCC3.5, Nak-YCC9.5, and Nak-TMut3.5. 2.2.2. Determination of the Peroxidase Activity of Nak-YCC3.5, NakYCC9.5, and Nak-TMut3.5 as a Function of pH. The peroxidase activity (PA) of adsorbed YCC and TMut was determined through guaiacol (Gc) oxidation to tetraguaiacol (tetraGc) in the presence of H2O2, following the reaction22
2. EXPERIMENTAL SECTION
UVvis spectra were used to follow the formation of tetraGc (λ = 480 nm, ε480 = 26600 M1 cm1). A 1.46 104 M Gc solution in the buffer at selected pH values was used to measure the peroxidase activity. This solution was also employed as a blank for the UVvis measurement. A 2 mL volume of the Gc solution was added to the solid catalyst, and as soon as an homogeneous suspension was obtained, 50 μL of H2O2 (0.5 M) was rapidly added to the suspension. After 15 s (Hanhart chronometer) from H2O2 injection, the spectrum of the products was instantaneously recorded on the filtered liquids (Φ = 0.45 μm). The same procedure was carried out using kaolinite lacking the adsorbed cytc (Nak, control sample). The H2O2 concentration of 12.2 mM was selected on the basis of measurements showing that this concentration is far from the values at which the reaction rate is independent of the substrate concentration (catalyst saturation). From the absorbance of the product tetraGc at 480 nm (A480nm) and the measured time of contact between the adsorbed cytc and H2O2 (Δt), the initial reaction rate (V0) was calculated. The initial reaction rate is defined as the concentration of reagent Gc, expressed in nanomolarity, which was consumed per second and per micromole of adsorbed cytc (μmolcyt), namely
2.1. Materials. Well-ordered kaolinite KGa-1b from The Clay Minerals Society (Source Clays Repository, University of Missouri, Columbia, MO) was completely sodium exchanged as previously described.15 The exchanged kaolinite is termed Nak hereafter. Sodium phosphate buffer solutions at pH 7 and I = 5 mM were obtained from Na2HPO4 and NaH2PO4, buffer solutions at pH 3.5 were prepared from HAc glacial, and the stocks of buffer solutions at pH 2, 3, 3.5, 4, 4.5, 5, 5.5, 6, and 6.5 and at pH 7.5, 8, 8.5, 9, 9.5, and 10 were obtained by adding HClO4 or NaOH, respectively, to the buffer solutions at pH 7. All chemicals were reagent grade: HAc glacial from Riedel-de Ha€en, pellets of NaOH (99%), NaCl, and HClO4 (65%) from Carlo Erba, Na2HPO4 and NaH2PO4 from J. T. Baker, and guaiacol (2-methoxyphenol) and H2O2 from Sigma. Wild-type recombinant untrimethylated C102T iso1-cytochrome c from Saccharomyces cerevisiae (YCC) and its triple mutant K72AK73AK79A (TMut) were expressed in Escherichia coli and purified following the procedure described elsewhere.1618 A threonine residue replaces the cysteine in position 102 in all species. This substitution prevents dimerization and minimizes autoreduction, without affecting the spectral and functional properties of the protein.19,20 Doubly distilled water was used throughout. 2.2. Methods. 2.2.1. Preparation of the Solid Catalysts: NakYCC3.5, Nak-YCC9.5, and Nak-TMut3.5. The adsorption of YCC was performed on Nak at pH 3.5 and 9.5, while that of TMut was performed at pH 3.5. Cytochrome c (cytc) in solution is known to undergo a conformational change (known as acid transition) with a pK value of approximately 2.5, which involves detachment of the axial His and Met ligands from the heme iron and their replacement with H2O molecules.21 Although the Met ligand begins to leave the heme already at pH values higher than 2.5, this transition is reversible,21 and we can confidently assume that the prevailing form of cytc at pH 3.5 is still the native
4Gc þ 4H2 O2 f tetraGc þ 8H2 O
ð1Þ
V0 ¼ ½Gc=ðΔt 3 μmolcyt Þ
ð2Þ
½V0 ¼ nMGc =ðs 3 μmolcyt Þ
ð3Þ
V0 is obtained from the experimental data as follows: V0 ¼ ð4 109 ÞΔAλ ¼ 480nm =ðΔt 3 εtetraGc, 480 3 d 3 μmolcyt Þ
ð4Þ
where ΔAλ=480nm is the difference between the Aλ=480nm values recorded at t = Δt and at t = 0, εtetraGc,480 is the molar extinction coefficient of the tetraGc at λ = 480 nm (εtetraGc,480 = 26600 M1cm1), d is the optic pathway of the UVvis cell (1 cm), the coefficient 4 109 is needed to 10684
dx.doi.org/10.1021/la201876k |Langmuir 2011, 27, 10683–10690
Langmuir
ARTICLE
transform tetraGc molarity into Gc nanomolarity, μmolcyt is the number of micromoles of adsorbed cytc obtained as described in section 2.2.1. The average time of contact (15 s) is within the linear dependence of Aλ=480nm vs t. The PA of the adsorbed cytochromes is defined as the concentration of Gc, expressed in nanomolarity, which is consumed per second, per micromole of adsorbed cytc, and per unit of starting concentration of H2O2, expressed in millimolarity, in reaction 1: PA ¼ ½Gc=ðΔt 3 μmolcyt 3 ½H2 O2 0 Þ
ð5Þ
½PA ¼ nMGc =ðs 3 μmolcyt 3 mMH2 O2 Þ
ð6Þ
Therefore PA ¼ V0 =½H2 O2 0
ð7Þ
2.2.3. Determination of the Initial Rate (V0) of Reaction 1 Catalyzed by Nak-YCC3.5 and Nak-TMut3.5 as a Function of the Substrate Concentration ([H2O2]). V0 has been measured at different pH values as a function of the H2O2 concentration. A fresh solid catalyst was used for each substrate concentration. The solid catalysts were obtained as described in section 2.2.1. Each set of measurements was carried out at a fixed pH value, given by the pH of the Gc solution, as a function of the H2O2 concentration. The standard experimental conditions were as follows: catalysts, Nak-YCC3.5 and Nak-TMut3.5; [Gc]0 = 1.46 104 M; [H2O2]0 = 3, 6, 9, 12, 15, 18, 21, and 30 mM; time of contact between the solid catalyst and the substrate H2O2, 15 s; T = 25 °C. In particular, Nak-YCC3.5 catalyst was investigated at pH 2, 3, 4, 5, 7, and 10, while Nak-TMut3.5 was investigated only at pH 3.5 and 7. 2.2.4. Raman Spectroscopy. Raman spectra were collected in backscattering geometry, with an InVia Raman microscope (Renishaw plc, Wotton-under-Edge, U.K.) equipped with a 405 nm diode laser (Bluephoton model, Omicron GmbH, Rodgau-Dudenhofen, Germany) and a 514.5 nm Ar ion laser (Laser-Physics, West Jordan, UT), delivering 2 and 5 mW of laser power at the sample, respectively. A Leica 60 water immersion microscope objective (NA = 1.00) focused the laser on the samples, which consisted of small amounts of the solid catalysts NakYCC3.5 and Nak-TMut3.5 and of kaolinite without adsorbed cytochrome c put on a CaF2 microscope slide and immersed in a few drops of buffers at different pH values. A 2400 L/mm (for 405 nm excitation) or 1800 L/mm (for 514.5 nm excitation) grating yielded a spectral resolution of approximately 4 cm1. A thermoelectrically cooled chargecoupled device (CCD) camera was used for detection. The spectrograph was calibrated using the lines of a Ne lamp. Single spectra were collected with an exposure time of 90 s (for 405 nm excitation) or 150 s (for 514.5 nm excitation). During each measurement, the sample was kept moving using a motorized microscope stage to renew the part of the sample illuminated by the laser, thus avoiding overheating and/or photodegradation of the proteins under investigation. 2.2.5. Diffuse-Reflectance UVVis Spectroscopy. Diffuse-reflectance (DR) UVvis spectra were recorded on a V-570 Jasco instrument equipped with an integrating sphere attachment (Jasco model ISN-470) in the 190800 nm range using a bandwidth of 0.5 nm. Kaolinite soaked in the same buffer of the measurement was used as a standard instead of BaSO4. The samples consisted of the solid catalyst Nak-YCC3.5 wetted with drops of buffers at different pH values.
3. RESULTS 3.1. Peroxidase Activity of Nak-YCC3.5 as a Function of pH (at 20 and 35 °C). Adsorption of cytc was carried out at pH 3.5.
At this pH, the charge of the kaolinite edges is positive, as their
Figure 1. PA in nMGc/(s 3 μmolcyt 3 mMH2O2) vs pH of Nak-YCC3.5 at T = 20 °C and T = 35 °C.
Figure 2. PA in nMGc/(s 3 μmolcyt 3 mMH2O2) vs pH of Nak-YCC3.5 and Nak-TMut3.5 at T = 20 °C.
isoelectric point is much higher (e.iep = 5.6).23 Since the adsorption driving force is mainly electrostatic in nature,10 working at pH 3.5 drives the immobilization process of the positively charged cytc selectively into the kaolinite faces, whose adsorption sites are always negatively charged and pH-independent, because they originate from isomorphic substitution. cytc is strongly adsorbed on kaolinite, and no appreciable desorption occurs, as confirmed by the absence of the Soret band in the UVvis spectra collected in the entire pH range of 210 to measure the PA of Nak-YCC3.5. It is also conceivable that the cytc molecules reside on the kaolinite faces over the entire pH range. Figure 1 shows that Nak-YCC3.5 is catalytically active in the entire pH range investigated at 20 and 35 °C. At 20 °C, maximum activity is observed from pH 7 to pH 10 and at pH 2. At intermediate pH values PA decreases and reaches a minimum at pH 45 (at which PA is approximately halved). At 35 °C, the PA in the pH range 79 is constant and approximately 2 times that measured at 20 °C. At higher pH values, PA decreases dramatically. 3.2. Peroxidase Activity of Nak-TMut3.5 as a Function of pH (at 20 °C). Nak-TMut3.5 shows PA at 20 °C in the entire pH range investigated (Figure 2), showing a minimum at pH 5 (40 nMGc/(s 3 mMH2O2 3 μmolcyt)). As for Nak-YCC3.5, the UVvis spectra taken on the reaction mixture show the absence of free TMut in solution, indicating that TMut does not desorb from Nak under these conditions. 10685
dx.doi.org/10.1021/la201876k |Langmuir 2011, 27, 10683–10690
Langmuir
ARTICLE
Table 1. KM and Vmax Values Obtained by Best Fitting (Sigma Plot v. 10.0) Applied to the Couples of Data (V0; [H2O2]0) Relative to Reaction 1 Conducted at Different pH Values, Catalyzed by Nak-YCC3.5 or Nak-TMut3.5 pH 2
pH 3
pH 4
pH 5
pH 7
pH 10
Nak-YCC3.5
KM ( 7% (mM)
18.9
18.8
21.1
23.6
27.3
26.8
4.07 104
3.10 104 16.1
2.21 104
1.47 104
4.13 104 28.3
4.22 104
Nak-TMut3.5
Vmax ( 5% (nMGc/(s 3 μmolcyt)) KM ( 7% (mM) Vmax ( 5% (nMGc/(s 3 μmolcyt))
Figure 3. KM of reaction 1 vs pH for the solid catalyst Nak-YCC3.5 at 25 °C.
The PA of Nak-TMut3.5 at T = 20 °C shows a pH dependence similar to that of Nak-YCC3.5 at the same temperature, showing a minimum at pH 5 and a maximum of activity at pH extremes. Nak-YCC3.5 is invariably more catalytically active than NakTMut3.5 except at pH 3.5 (Figure 2). 3.3. Analysis of the Peroxidase Activity by the Michaelis Menten Model. V0 of reaction 1 for Nak-YCC3.5 and NakTMut3.5 as a function of [H2O2]0 follows a typical Michaelis Menten behavior, showing substrate saturation over all the pH range investigated, from 2 to 10 (Figures S1S8, Supporting Information). The KM and Vmax values for Nak-YCC3.5 and Nak-TMut3.5 at different pH values are listed in Table 1. The plots of KM and Vmax vs pH are shown in Figures 3 and 4, respectively. KM increases with a sigmoidal behavior, having an apparent pKa value of approximately 5.5. Vmax shows the same pH dependence as PA, reaching a minimum at pH 5. 3.4. Raman Spectroscopy Measurements. The Raman spectra of kaolinite with and without adsorbed cytc are shown in Figure 5. The bands present in the 200900 cm1 region of both spectra can be assigned to kaolinite.24 Those in the 1200 1700 cm1 region, observed only in the presence of adsorbed cytc, can be undoubtedly assigned to cytc, as they are typical of a heme resonance Raman (RR) spectrum obtained with an excitation in resonance with a heme absorption Q-band.25 In particular, the band at 1371 cm1 can be assigned to a ν4 marker band for a ferric cytc, whereas the ν10 spin marker band at 1634 cm1 suggests that the heme is in a low-spin state.25 To gain information on the oxidation, spin, and coordination states of the adsorbed wild-type protein and its triple mutant at different pH values, we collected RR spectra using a laser capable of Soret band excitation (i.e., at 405 nm, Figure 6). Spin and coordination marker bands in fact are more clearly visible in heme Soret-excited RR spectra than in Q-band-excited ones. The ν4
2.50 104
2.13 104
Figure 4. Vmax in nMGc/(s 3 μmolcyt) of reaction 1 vs pH for the solid catalyst Nak-YCC3.5 at 25 °C.
marker band at 1371 cm1 in the Soret-excited RR spectra shows that both YCC and TMut contain ferric heme at all pH values investigated. At pH 6.88, however, the ν4 band for TMut is slightly asymmetric, with a weak shoulder at 1361 cm1, suggesting the presence of some ferrous heme as well. The overall RR intensity decreases with pH for both YCC and TMut, suggesting a change in the absorption spectrum of the heme chromophore (as confirmed by DR UVvis measurements; see below), with a consequent change in the resonance condition. This intensity decrease is particularly pronounced for the YCC protein at pH 3.54. The spin and coordination marker bands at 1501 cm1 (ν3), 1583 cm1 (ν2), and 1634 cm1 (ν10) in the Soret-excited RR spectra (Figure 6) show that, at all pH values investigated, both YCC and TMut contain a six-coordinate low-spin heme center (6cLS), in agreement with the Q-excited spectra. Previous spectroscopic studies26 associate these values with a heme iron featuring a methionine and histidine as axial ligands. Despite the fact that the low signal-to-noise ratio does not allow a more precise analysis of other species present, the occurrence of a shoulder of the ν2 band at 1568 cm1 in both YCC and TMut spectra at pH 5.14 suggests the presence of a high-spin species (6cHS). According to the literature, such HS species contains a heme iron with a water molecule as the axial ligand in place of the methionine ligand.26 At pH 3.54, the ratio between the ν2 band of the HS species at 1568 cm1 and the other ν2 band for the LS species at 1583 cm1 increases for YCC, suggesting an increased population of the HS species with respect to the TMut. 3.5. Diffuse-Reflectance UVVis Spectroscopy Measurements. The DR UVvis spectra recorded at pH 7.5 for ferriYCC immobilized on kaolinite at pH 3.5 (Figure 7) are very similar to those for the protein in solution.27 They contain the 10686
dx.doi.org/10.1021/la201876k |Langmuir 2011, 27, 10683–10690
Langmuir
ARTICLE
Figure 7. DR UVvis spectra at different pH values of YCC adsorbed on kaolinite at pH 3.5. Figure 5. Regions (2001800 cm1) of Raman spectra of kaolinite with (thick line) and without (thin line) adsorbed YCC at pH 6.88. The two insets show details of two regions of the baseline-corrected spectra where bands typical of kaolinite (top left) and YCC (bottom right) are found. The excitation wavelength is 514.5 nm.
A pH decrease causes progressive changes in the DR UVvis absorption spectra of immobilized ferricytochrome c. At pH 3 the Soret band and Q-band show both a slight blue shift (with new maxima at 404 and 529 nm, respectively) as well as a decrease in absorbance, while a new well-defined CT band appears at 627 nm. Moreover, the Soret band and the Q-band show weak shoulders at about 386 and 496 nm, respectively. As observed above for the Raman spectra, the behavior at pH 5 is intermediate between those of pH 7 and pH 3: the Soret band and Q-band undergo a smaller blue shift compared to that at pH 3 and the new CT band at 627 appears, but less pronounced than at pH 3. Previous spectroscopic measurements performed on cytochrome c in solution at pH 2 indicate the presence of a high-spin form of the protein with a Soret band at 394 nm, a Q-band at 495 nm, and a CT band at 620 nm.27 Therefore, the changes observed upon lowering the pH are consistent with an increase of the highspin heme-containing species to the detriment of the low-spin species.27
4. DISCUSSION
Figure 6. Regions (12501700 cm1) of the Soret-excited resonance Raman spectra at different pH values of YCC (thick line) and TMut (thin line) adsorbed on kaolinite at pH 3.5. The excitation wavelength is 405 nm.
Soret absorption band at 409 nm and the Q-band at 531 nm (shoulder at about 555 nm), while the charge transfer (CT) band at 695 nm is not observed, probably due to its very low intensity. No remarkable changes occur with increasing pH up to 10.
4.1. Peroxidase Activity of Nak-YCC3.5 and Nak-TMut3.5. It is known that the native six-coordinate (Met, His) form of cytc does not show peroxidase activity due to the lack of suitable sites around the heme center for the interaction with H2O2 and subsequent formation of the ferryl group.1 The PA observed for adsorbed cytc must be ascribed to the effects of protein unfolding following adsorption, which induce either the formation of a five-coordinate heme center or a pronounced weakening of the “sixth” axial bond to the heme iron. These catalytically active forms of cytc are formed at low pH11,13 or due to the action of denaturing agents11 or upon immobilization on a surface.10,13 Here, the unfolding effect responsible for the observed PA probably arises from the electrostatic adsorption of the protein on the kaolinite surface. Moreover, it is conceivable that the increase (decrease) in the proteinkaolinite electrostatic interaction induces an increase (decrease) of the unfolding effect and therefore a destabilization (stabilization) of the heme pocket structure. In this view, the charge carried by the kaolinite surface as a function of pH plays a key role in modulating the unfolding effect. It is wellknown that the charge on the kaolinite particle is pH-dependent. It is negative at the e.iep (pH 5.6)23 and decreases further with increasing pH due to the deprotonation of silanol and aluminol groups of the edges. The main ionization effect, however, occurs 10687
dx.doi.org/10.1021/la201876k |Langmuir 2011, 27, 10683–10690
Langmuir between the e.iep and neutral pH, while at higher pH values the increase in the negative charge is rather limited.28 The PA of both Nak-YCC3.5 and Nak-TMut3.5 remains unchanged with pH from 7 to 10 (Figure 2). This indicates that the folding of both adsorbed proteins is unaffected by the residual deprotonation of the edges. The data reported in Figure 2 also indicate that Nak-YCC3.5 has a larger catalytic activity than NakTMut3.5. YCC and TMut bear different charges (+6 and +3, respectively), resulting in a stronger electrostatic interaction for YCC with the kaolinite faces.10 According to the above interpretation, the observed larger PA of YCC can be attributed, therefore, to the more extensive unfolding induced by the electrostatic effect which is responsible for an increased heme pocket exposition to solvent. A relevant finding of this work is the catalytic behavior of the adsorbed proteins—in particular that of YCC—at pH values around 10. It is known that YCC in solution at about pH 8.2 undergoes the structural change called “alkaline transition”,21 which involves substitution of the axial ligand Met 80 with a Lys residue, causing a dramatic decrease of the peroxidase activity.11 The alkaline transition is not expected to occur for TMut, as this variant lacks all the lysine residues which can substitute the axial ligand Met 80.21 The evidence (Figure 2) is that both the adsorbed proteins retain PA at pH 10 and 20 °C. This finding indicates that the alkaline transition does not occur for YCC adsorbed on kaolinite at 20 °C. This point was confirmed by a set of measurements in which the effect of the rise in temperature—a condition favoring alkaline transition—on the PA measured at pH 10 and 20 °C was specifically studied, keeping both NakYCC3.5 and Nak-TMut3.5 on a phosphate buffer at pH 10 and 35 °C for 30 min. While Nak-TMut3.5 indeed exhibited PA at pH 10 and t = 20 °C and the obtained value was almost the same as that of the standardized method, the detected value for NakYCC3.5 was very low (
