Fluorescence properties and conformational stability of the β-hemocyanin of Helix pomatia

Biochimica et Biophysica Acta 1764 (2006) 807 – 814 http://www.elsevier.com/locate/bba

Fluorescence properties and conformational stability of the β-hemocyanin of Helix pomatia Krassimira Idakieva a,⁎, Nurul I. Siddiqui b , Katja Parvanova a , Peter Nikolov a , Constant Gielens b a

b

Institute of Organic Chemistry, Bulgarian Academy of Sciences, Akad. G. Bonchev-Str. Bl. 9, Sofia 1113, Bulgaria Laboratory of Biochemistry, Chemistry Department, Katholieke Universiteit Leuven, Celestijnenlaan 200 G, 3001 Leuven-Heverlee, Belgium Received 3 November 2005; received in revised form 1 December 2005; accepted 4 December 2005 Available online 29 December 2005

Abstract The β-hemocyanin (β-HpH) is one of the three dioxygen-binding proteins found freely dissolved in the hemolymph of the gastropodan mollusc Helix pomatia. The didecameric molecule (molecular mass 9 MDa) is built up of only one type of subunits. The fluorescence properties of the oxygenated and apo-form (copper-deprived) of the didecamer and its subunits were characterized. Upon excitation of the hemocyanins at 295 or 280 nm, tryptophyl residues buried in the hydrophobic interior of the protein determine the fluorescence emission. This is confirmed by quenching experiments with acrylamide, cesium chloride and potassium iodide. The copper–dioxygen system at the binuclear active site quenches the tryptophan emission of the oxy-β-HpH. The removal of this system increases the fluorescence quantum yield and causes structural rearrangement of the microenvironment of the emitting tryptophyl residues in the apo-form. Time-resolved fluorescence measurements show that the oxygenated and copper-deprived forms of the β-HpH and its subunits exist in different conformations. The thermal stability of the oxy- and apo-β-HpH is characterized by a transition temperature (Tm) of 84 °C and 63 °C, respectively, obtained by differential scanning calorimetry. Increase of the temperature influences the active site at lower temperatures than the environments of tryptophans and tyrosines causing a loss of oxygen bound to the copper atoms. This process is, at least partially, reversible as after cooling of the protein samples, around 60% reinstatement of the copper-peroxide band has been observed. The results confirm the role of the copper– dioxygen complex for the stabilization of the hemocyanin structure in solution. The other important stabilizing factor is oligomerization of the hemocyanin molecule. © 2005 Elsevier B.V. All rights reserved. Keywords: Hemocyanin; Mollusca; Helix pomatia; Fluorescence spectroscopy; Differential scanning calorimetry

1. Introduction Hemocyanins (Hcs) are copper-containing respiratory proteins, freely dissolved in the hemolymph of many arthropod and mollusc species. The hemocyanin molecule in molluscs has the shape of a hollow cylinder (35 nm diameter) [1] and is constituted of ten (cephalopods) or twenty (gastropods) subunits with molecular mass of 350–450 kDa. In gastropodan Hcs the quaternary structure (didecamer with molecular mass of ∼9 MDa and cylinder height of 38 nm [1] consists of two Abbreviations: DSC, differential scanning calorimetry; EDTA, ethylenediaminetetraacetic acid; Hc, hemocyanin; HpH, Helix pomatia hemocyanin; MOPS, 3-[N-Morpholino]-propanesulfonic acid; Ac-Trp-NH2, N-acetyltryptophanamide; Tris/HCl, tris (hydroxymethyl) amino-methane hydrochloride ⁎ Corresponding author. Tel.: +359 2 9606 190; fax: +359 2 8700 225. E-mail address: [email protected] (K. Idakieva). 1570-9639/$ - see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.bbapap.2005.12.001

axially assembled decamers. The subunits themselves are folded into eight (seven for Hc from octopi) covalently linked functional units (FUs), indicated by the letters a–h (a–g) from the N-terminus on. These have a molecular mass of ∼50 kDa and carry a binuclear copper active site, capable of reversibly binding one dioxygen molecule [2,3]. Under non-physiological conditions Hcs can loose their quaternary structure. The most classical way to obtain dissociation into subunits, without loss of the ability to reversibly bind dioxygen, is an increase of the pH (to ∼9) and removal of alkaline-earth cations (by treatment with EDTA) [2,3]. Gastropodan Hc, however, can also be dissociated under the influence of high pressure (2 kbar) [4]. Data on the conformational stability of Hcs and subunits have been obtained by temperature rise [5,6] and by treatment with chemical agents like guanidinium chloride and urea [7–9]. For a gastropod Hc

808

K. Idakieva et al. / Biochimica et Biophysica Acta 1764 (2006) 807–814

evidence was provided for conformational differences between fragments of subunits [10]. The Hc of the gastropodan mollusc Helix pomatia (vineyard snail) consists of three components. Next to two α-components (αD- and αN-Hc), also a β-component (here referred to as βHpH) is present [2]. This Hc differs from the α-Hcs by its ability to precipitate or crystallize during dialysis against sodium acetate buffer, pH 5.3, at low ionic strength (10 mM), and by the subunit composition, consisting of only one type of polypeptide chain (β subunits) compared with two types (α and α' subunits) in each of the two α-Hcs [11]. Because of this subunit homogeneity structural investigations have mainly been performed on β-HpH. The primary structure of the FUs d [12] and g [13] has been obtained through amino acid sequencing. For each FU the carbohydrate composition has been determined [14] and for most FUs the carbohydrate attachment sites have been localized [15–17]. Also in β-HpH (for the first time in a molluscan Hc), the presence was discovered of an unusual thioether bridge between a cysteine residue and one of the three histidine residues involved in the coordination of the copper A atom of the binuclear copper group, and in FUs d and g, the location of the disulphide bridges was determined [18]. For the αD-Hc, a definite influence of the binding of oxygen on the stability of the didecamer was observed. The oxy-form is less stable than the deoxy-form and dissociates faster when calcium is removed at high pH [19]. We have performed this study in order to obtain more information about the structure in solution and the conformational stability of the β-Hc of Helix pomatia and the structural subunits constituting the Hc molecule. The influence of dissociation and of removal of copper was investigated. 2. Materials and methods 2.1. Hemocyanin β-HpH was purified from the hemolymph of terrestrial snails Helix pomatia as described [13,20]. The apo-form (copper-deprived Hc) was prepared by dialysis against 50 mM Tris/HCl, containing 25 mM KCN, pH 7.2, for 48 h, at 4 °C, according to [21]. The samples were then dialyzed against 50 mM Tris/HCl, containing 10 mM EDTA at pH 7.2, and finally against 50 mM Tris/HCl, pH 7.2. The buffer conditions used to study the β-HpH and its apo-form in either the associated or dissociated state were chosen on the basis of the pH-stability regions, described in [22].

the void volume) containing the protein was analyzed both for protein content (absorption at 278 nm) and copper content (atomic absorption at 324.7 nm). 2.2.2. Steady-state fluorescence measurements Fluorescence spectra were recorded with a Perkin Elmer model LS 5 spectrofluorimeter equipped with a thermostatic cell compartment and a Data Station model 3600. The optical density of the solutions was lower than 0.05 at the excitation wavelength to avoid inner filter effects. The relative fluorescence quantum yields were calculated using the fluorescence standard N-acetyltryptophanamide (Ac-Trp-NH2) [23]. The results of the quenching reactions between the excited tryptophyl side chains and acrylamide, CsCl or KI were analyzed according to the Stern–Volmer equation [24]: F0 =F ¼ 1 þ KSV ½X 

where F0 and F are the fluorescence intensities in the absence and in the presence of quencher, respectively; KSV is the collisional quenching constant and [X] is the quencher concentration. A small amount of Na2S2O3 was added to the iodide solutions to prevent I−3 formation. The inner filter effect due to the acrylamide was corrected by the factor: Y ¼ antilogðdA þ dB Þ=2

ð2Þ

where dA and dB are the optical densities at the excitation and emission wavelength, respectively. Static quenching with acrylamide was separated from the total effect by the equation [25]: F0 =F:eV ½X  ¼ 1 þ KSV ½X 

ð3Þ

where V is the static quenching constant which is related to the probability of finding a quencher molecule close enough to the chromophore at the moment of excitation, to quench it statically. The temperature dependence of the tryptophyl fluorescence was determined in 50 mM Tris/HCl buffer, pH 8.2 (20 °C). The samples were equilibrated at each temperature for 10 min before the measurements. The data were analyzed according to the equation [26]: lnðQ1  1Þ ¼ lnðfi =kf Þ  Ea =ðRT Þ

ð4Þ

where Q is the fluorescence quantum yield, fi is the frequency factor for the nonradiative deactivation processes, kf is the temperature independent rate constant for the fluorescence emission, Ea is the apparent activation energy, R is the gas constant and T is the absolute temperature. 2.2.3. Time-resolved fluorescence measurements Time-resolved fluorescence studies were performed at 20 °C using a nanosecond single-proton-counting spectrofluorimeter (system PRA 2000) and a nitrogen-filled flash lamp with a full width at half-maximum of ∼2.5 ns. The protein samples were dissolved in 50 mM Tris/HCl buffer, containing 5 mM CaCl2 and 5 mM MgCl2, pH 7.2. For measurements of the subunits 50 mM Tris/ HCl buffer, containing 10 mM EDTA, pH 9.0, was used. The data were analyzed by convoluting the instrument response function L(t′) with an assumed decay function P(t), as

2.2. Spectroscopic methods 2.2.1. UV-VIS absorption and atomic absorption spectrophotometry Absorption spectra were recorded with a Shimadzu UV-1601 spectrophotometer, equipped with a refrigerated circulator F-12MP Julabo. Protein concentration was determined spectrophotometrically using the absorption −1 coefficient A0.1% ml cm−1. 278 = 1.416 mg Protein samples were dialyzed against 50 mM MOPS, pH 7.2, containing 5 mM CaCl2 and 5 mM MgCl2 (didecameric Hc) or against 50 mM borax buffer, pH 9.2 (subunits) [14], for measurements of their thermal stability. The protein concentrations were 1.2 mg/ml and 1.5 mg/ml, respectively. The possible liberation of copper from the protein upon temperature increase of the Hc solution was examined using Perkin-Elmer 372 atomic absorption spectrophotometer. After heating, EDTA was added to the sample in a final concentration of 10 mM in order to complex freed copper. The sample was passed through a PD-10 column (Amersham) and the flow-through fraction (at

ð1Þ

Z

t

Yc ðtÞ ¼

Lðt VÞPðt  t VÞdt V

ð5Þ

0

and comparing Yc (t) with the experimental time-dependence Rm(t) using a nonlinear least-squares iterative convolution method based on the Marquardt algorithm. The decay curves contained 104 counts at the maxima and the timeresolution for these curves was 100 ps per channel. The goodness of the fit was assessed by the weighted residuals and the reduced χ2-test.

2.3. Differential scanning calorimetry Calorimetric measurements were performed on a high-sensitivity differential scanning microcalorimeter DASM-4 (Biopribor, Pushchino, Russia), with sensitivity greater than 0.017 mJ/K and a noise level less than ± 0.05 μW [27]. A constant pressure of 2 atm was maintained during all DSC experiments to

K. Idakieva et al. / Biochimica et Biophysica Acta 1764 (2006) 807–814 prevent possible degassing of the solution on heating. Each sample run was preceded by a baseline run with buffer-filled cells. The protein solution in the calorimetric cell was reheated after the cooling from the first run to estimate the reversibility of the thermally induced transitions. In all cases, the thermal denaturation was found to be irreversible, therefore the thermogram corresponding to the reheating run was used as instrumental base line. The transitions were corrected for the difference in heat capacity between the initial and final state by using a linear chemical base line. The calorimetric data were evaluated using the ORIGIN (MicroCal Software) program package. The temperature at the maximum of the excess heat capacity curve was taken as the transition temperature (Tm). The protein samples were dialyzed extensively against the buffer used in the scanning experiment. In the calculation of molar quantities, the molecular mass used for the protein was 9,000,000 Da. Calorimetric experiments were carried out in 20 mM MOPS, containing 0.1 M NaCl, 5 mM CaCl2, 5 mM MgCl2, pH 7.2 (20 °C).

3. Results and discussion 3.1. Steady-state fluorescence measurements Fluorescence spectroscopy is one of the most sensitive methods for studying the changes in the protein conformation in solution. Fluorescence parameters of the oxy- and apo-forms of the β-HpH and its structural subunits are summarized in Table 1. After excitation at 280 nm, where is the main absorption maximum of the tyrosyl and tryptophyl residues, or at 295 nm, where the tryptophyl side chains are selectively excited, the fluorescence spectra of the oxy-forms of β-HpH and its subunits have maxima at 325 ± 1 nm and 327 ± 1 nm, respectively. These maxima are typical for “buried” tryptophyl side chains in nonpolar environment [28]. The apo-forms of β-HpH and its subunits have fluorescence maxima at 328 ± 1 nm and 331 ± 1 nm, respectively. These shifts in the position of the emission maxima indicate that removal of the copper atoms at the active sites and dissociation into subunits of the copper-deprived Hc change the microenvironment of the emitting tryptophyl residues. In addition to the small wavelength shifts, pronounced differences in the quantum yields, already reported in other Hcs [7,8], were observed. While the tryptophyl emission quantum yield of the oxy-form of β-HpH is low (0.01) (Table 1), a quantum yield of 0.05 has been determined for the apo-β-HpH. The observed effect can be explained by radiationless energy transfer from the excited indole rings to the copper–dioxygen system in the oxy-form. Recently, Erker et al. [29] presented a detailed study, at atomic level resolution, that Förster transfer is Table 1 Steady-state fluorescence parameters of β-HpH and its subunits Sample

Oxy β-HpH Oxy β-HpH-subunits Apo β-HpH Apo β-HpH-subunits Ac-Trp-NH2 a

Data from [37].

Emission λmax [nm] (excitation at 295 nm)

Relative quantum yield

Acrylamide quenching KSV (M−1)

V (M−1)

325 ± 1 327 ± 1 328 ± 1 331 ± 1 350 ± 1

0.01 0.01 0.05 0.05 0.130 a

1.57 2.49 3.11 2.65 16.33 a

– – 0.6 –

809

responsible for the oxygen-dependent quenching of the tryptophan fluorescence in tarantula (arthropod) Hc. The tryptophans transfer their excitation energy to the oxygenated active sites. The absence of tyrosyl emission in the fluorescence spectra of the investigated Hcs after excitation at 280 nm can be explained by a singlet–singlet radiationless energy transfer from phenol groups (donors) to indole rings (acceptors) according to Förster's theory of electronic energy transfer in donor–acceptor systems [30]. Similar explanation has been given at the fluorescence measurements of the Hc of the related gastropod Rapana thomasiana (RtH) [31]. Quenching experiments with acrylamide confirm the conclusion that the indole groups in the oxy- and apo-forms of the β-HpH and its subunits are located in the interior of the Hc molecules. Acrylamide is an efficient neutral quencher of tryptophyl fluorescence and provides topographical information about the emitting chromophores [25]. It can discriminate between “exposed” and “buried” tryptophyl side chains and the results are not influenced by the charge of the chromophore microenvironment. The ability to quench collisionally the excited indole rings depends on its ability to penetrate the protein matrix. Quenching of “buried” tryptophans by acrylamide has been explained in terms of structural fluctuations of the protein molecules that facilitate the inward diffusion of the quencher [32]. The fluorescence quenching with acrylamide of the oxy-forms of β-HpH and the subunits followed the classical Stern–Volmer Eq. (1) (Fig. 1A). The observed linearity of the Stern–Volmer plot can be explained by the similarity of the individual KSV constants of the tryptophyl fluorophores [32]. This means that tryptophyl residues of the protein differ only slightly in accessibility, although discrimination among the indole fluorophores could be expected as β-HpH contains 5–6 tryptophyl residues per functional unit of 50 kDa [11,12]. Similar linear plots have also been observed for the oxy-forms of the Rapana thomasiana Hc and its subunits, RtH1 and RtH2 [31]. The slope of the plots at a low concentration of acrylamide reflects to a large extent the quenching of the more accessible residues and selective quenching can be observed only if the KSV constants differ sufficiently [32]. KSV values of 1.57 M−1 and 2.49 M−1 were calculated (Table 1). These constants are significantly lower than the value for Ac-Trp-NH2, i.e. for tryptophan in aqueous solution, and are indicative of a very low efficiency of quenching. With apo-β-HpH the value of KSV is 3.11 M −1 or approximately twofold higher than that for the oxy-form. The Stern–Volmer plot for the apo-form of β-HpH subunits tryptophan fluorescence quenching with acrylamide showed upward curvature (Fig. 1B). In this case, the fluorescence of all tryptophyl residues is equally quenched (nearly equally accessible chromophores), or only part of them are fluorescent [25]. The best fit of the curve to the linear plot was obtained using the modified Stern–Volmer relationship of Eq. (3) and a value of 0.6 M−1 for the static quenching constant. The parameter V is related to the probability of finding the quencher molecule close enough to the excited chromophore to quench it with 100% efficiency. A value of V = 2.0 M−1 has been obtained

810

K. Idakieva et al. / Biochimica et Biophysica Acta 1764 (2006) 807–814

ronment of the dioxygen-binding site, at a distance of 5–6 Å from the copper atom CuA [16]. Hence, its emission should be very sensitive to conformational changes in the region of the active site. The other indole groups are located at a distance of ≥10 Å from this site and are buried in the hydrophobic interior of the globule. The emission of the buried tryptophyl residues, like in the case of the investigated β-HpH, should respond to conformational changes in the protein, which will change the polarity of their environment. 3.2. Effect of pH on the conformational stability

Fig. 1. (A) Acrylamide quenching of the tryptophyl fluorescence of oxy-forms of β-HpH (●–●) and its subunits (○–○), according to Eq. (1). (B) Acrylamide quenching of the tryptophyl fluorescence of apo-subunits of β-HpH (■–■). Linear plot (□–□) was obtained using Eq. (3) and a value of 0.6 for the static quenching constant.

for the quenching of the Ac-Trp-NH2 fluorescence by acrylamide [25]. The value calculated for the apo-form of βHpH thus indicates that the local concentration of acrylamide molecules in the proximity of the emitting tryptophans is moderate, suggesting that the emitting residues are only partially buried. It thus can be concluded that the tryptophyl residues in the apo-forms of the β-HpH and its subunits become more exposed than in the oxy-Hcs. The removal of the copper at the binuclear active sites causes a structural rearrangement of the microenvironment of the emitting tryptophyl residues and they are more accessible to the quencher molecules than in the respective oxygenated form of the proteins. Quenching experiments were also performed with Cs+ and − I . Ionic quenchers like Cs+ and I− are charged and hydrated. In contrast with acrylamide, which can penetrate the protein matrix, they are able to quench only surface fluorophores and are effective in discriminating between “exposed” and “buried” chromophores, as well as in revealing charge effects. Exposure of neither oxy- nor apo-forms of β-HpH and its subunits to increasing Cs+ or I− concentrations [0.02–0.8 M] had effect on their fluorescence after excitation at 295 nm. This confirms the conclusion that the tryptophyl side chains are deeply buried in the interior of the Hc molecule. All these observations are in agreement with the structural data obtained by X-ray crystallography on FU RtH2-e isolated from the Hc of Rapana thomasiana [33]. Inspection of the structure of this Hc shows that Trp-69, which is completely conserved in molluscan Hcs, is in the immediate microenvi-

The effect of pH on the tryptophyl fluorescence quantum yield of the oxy- and apo-forms of the β-HpH is shown in Fig. 2. Practically no change in the emission of oxy-β-HpH was observed in the pH region 5.0–8.0. The increase of the fluorescence quantum yield at pH values above 8 can be attributed to the titration of an ionizable group with apparent pKa of 8.1, which is within the ionization region of the αamino group. Alternatively, the quenching process may involve electron transfer from an excited tryptophyl residue to the α-amino group or to an ε-amino group with abnormal pK, which can occur over an appreciable distance. In the case of the apo-Hc the quantum yield was constant at pH 7.0–9.0. The transition with a midpoint at pH 6.0 most probably represents the apparent pKa of imidazole groups of histidyl residues. The emission of tryptophyl residues in the apo-βHpH is thus most likely quenched by nearby protonated imidazole groups, which are able to form complexes with indole. The transition at pH ≈10.5, also observed for the oxyform, can be ascribed to ionization of tyrosyl residues and efficient radiationless energy transfer from the excited indole rings to ionized phenol groups. The small increase of the fluorescence quantum yield of the oxy-form below pH 5.0 is probably a result of a destruction of the copper–dioxygen system at the active site, which quenches the tryptophyl

Fig. 2. pH-dependence of the tryptophyl fluorescence quantum yield of oxy-βHpH (●–●) and apo-β-HpH (○–○). The following buffers were used: 50 mM sodium citrate (pH 3.0–7.0); 50 mM Tris/HCl (pH 7.0–9.0); 50 mM carbonate/ bicarbonate (pH 9.0–11.0).

K. Idakieva et al. / Biochimica et Biophysica Acta 1764 (2006) 807–814 Table 2 Fluorescence decay parameters of β-HpH and its subunits after excitation at 297 nm Sample

τ1 [ns]

A1 [%]

τ2 [ns]

A2 [%]

χ2

Oxy β-HpH Oxy β-HpH subunits Apo β-HpH Apo β-HpH subunits

0.25 0.27 0.10 0.14

68 73 61 55

2.44 2.26 3.11 2.98

32 27 39 45

1.31 1.28 1.30 1.23

fluorescence. There is not such a system in the apo-β-HpH and for this reason, no increase of the fluorescence quantum yield is observed in this case. 3.3. Time-resolved fluorescence measurements For the further characterization of the conformational states of oxy- and apo-forms of β-HpH and its structural subunits, we used time-resolved fluorescence spectroscopy. This method provides information about the molecular environment of the fluorophores. The fluorescence decay of the Hcs was investigated upon excitation at 297 nm, where the light is selectively absorbed by the tryptophyl chromophores, and was well fitted by bi-exponential decay function. The theoretical fluorescence intensity at time t is given by the function IF(t) = A1.exp (−t/τ1) + A2.exp (−t/τ2), where A1 and A2 are amplitudes. The value of the parameter χ2 was 1.2–1.3 (Table 2); the residuals between the theoretical and experimental decay curves as well as autocorrelation plots were flat and random, which demonstrates a good quality of the fit. Fig. 3 shows the fluorescence decay of oxy-β-HpH. The calculated lifetimes, the relative amplitudes and the χ2-values are shown in Table 2. The comparison of the data for the oxy-form of β-

811

HpH and its subunits shows that the shorter lifetime, τ1, is very similar −0.25 ns and 0.27 ns, respectively. Also the apo-form of β-HpH and its subunits have practically similar shorter lifetime −0.10 ns and 0.14 ns (Table 2). The same behavior was observed for the longer lifetime components τ2: 2.44 ns and 2.26 ns for the oxy-β-HpH and its subunits and 3.11 ns and 2.98 ns for the respective apo-forms (Table 2). However, the τ1values for the oxy-Hcs are significantly higher than those for the respective apo-forms. The oxy- and apo-states of the βHpH and its subunits also differ in the values of the τ2, which are higher for the apo-Hcs. These results clearly show that no significant changes in the microenvironment of the tryptophyl chromophores have to be expected between the oxy-form of the β-HpH and its subunits as well as between the apo-Hc and aposubunits. On the other hand, the parameters obtained from the analysis of the fluorescence decay curves (Table 2) demonstrate conformational changes in the native β-HpH and its subunits upon the transition from oxy- to the apo-forms. These results are in agreement with the conclusions made on the basis of the steady-state fluorescence measurements. Dynamic light scattering and time-resolved fluorescence measurements of the Rapana thomasiana Hc and its substructures also showed that the proteins adopt different conformations in their native, oxidized or copper-deprived form [34]. 3.4. Thermostability measurements Investigation of the temperature-dependence of protein fluorescence provides information on thermally induced conformational changes. These studies were carried out on apo-Hc only because the increase of the temperature causes a loss of oxygen bound to the copper active site influencing the

Fig. 3. Fluorescence decay of oxy-β-HpH in 50 mM Tris/HCl buffer, containing 5 mM CaCl2 and 5 mM MgCl2, pH 7.2, λex = 297 nm, λem = 350 nm. Curve A— experimental data for the fluorescence decay; curve B—convoluted results with bi-exponential decay function; curve C—instrumental response function.

812

K. Idakieva et al. / Biochimica et Biophysica Acta 1764 (2006) 807–814

Fig. 4. Thermal dependence of the tryptophyl fluorescence of apo-β-HpH (●–●) and apo-subunits (○–○). The fluorescence quantum yields were determined with Ac-Trp-NH2 as a standard.

fluorescence. Fig. 4 demonstrates the thermal dependence of the tryptophyl fluorescence quantum yield of the apo-forms of βHpH and the subunits. The critical temperature (Tc), that is the point at which the curve deviates from linearity, was 71 °C for the apo-form of β-HpH and 66 °C for the subunits. This result highlights the different stabilities of the didecameric β-HpH and its subunits. Ea of 9.73 kJ/mol was determined for the thermally deactivated protein fluorophores in apo-β-HpH. Such a value of Ea is typical for buried tryptophans [15] and is in agreement with the results obtained from the quenching experiments (Table 1). The Ea value determined for the thermally deactivated fluorophores in the apo-form of the subunits was 14.01 kJ/mol. In both cases the thermal denaturation resulted in a more or less pronounced shift of the tryptophyl fluorescence spectrum to longer wavelength (red shift) indicating that tryptophans become more “exposed”. The thermal stability of the oxy- and apo-forms of β-HpH was studied by high-sensitivity differential scanning calorimetry. This method has been widely used to study folding– unfolding processes in proteins. The thermograms are shown in Fig. 5. A single transition, with apparent transition temperature (Tm) at 84 °C, was obtained for the oxy-β-HpH at neutral pH (Fig. 5, curve A). Similar value of Tm was found for the oxy-Hc of Rapana thomasiana [6]. The thermal denaturation of the protein was irreversible, because no thermal effect was observed in a second heating of the protein solution. The sample extracted from the calorimetric cell showed aggregation. Irreversibility of the thermal denaturation was observed also in the DSC measurements of the Hcs from Rapana thomasiana [6], lobster Palinurus vulgaris [5] and tarantula Eurypelma californicum [35] and is a common property of the large hemocyanin molecules. The specific calorimetrical enthalpy ΔHcal (per mass of protein), calculated by the integration of the heat capacity curves, is 190 MJ mol−1. This value is similar to the ΔHcal obtained by DSC measurements of Rapana thomasiana Hc [6]. The thermogram for the apo-form of β-HpH is more complex (Fig. 5, curve B)

Fig. 5. Heating thermograms of oxy-β-HpH (A) and apo-β-HpH (B) in 50 mM MOPS, pH 7.2, recorded at 1 K/min heating rate; protein concentration of 4.65 mg/ml oxy-β-HpH and 5.4 mg/ml apo-β-HpH.

with an apparent Tm of the main transition at 63.5 °C, which is ∼20 °C lower than the Tm obtained for the oxy-Hc. The total ΔHcal obtained at the thermal denaturation of the apo-Hc is 125 MJ mol−1. This result is a demonstration for the contribution of the copper–dioxygen system at the binuclear active sites to the stability of the Hc structure in solution. Circular dichroism studies of nine Hcs from arthropod and gastropod organisms also have shown that the hemocyanin thermostability is highly dependent on the presence or absence of the copper-containing dioxygen-binding sites [36]. The oxyHcs have been more stable by 3–16 degrees in their melting points in comparison with the apo-Hcs [36]. It can be concluded that the binuclear active site has a considerable stabilizing effect on the protein structure in solution and its removal causes structural rearrangement and decreases the thermal stability of the molecule.

Fig. 6. Absorption spectra of oxy-β-HpH in 50 mM MOPS, pH 7.2, containing 5 mM CaCl2 and 5 mM MgCl2, recorded with temperature rise. The protein concentration is 1.2 mg/ml. Inset: Temperature dependence of the absorbance at 345 nm of the oxy-β-HpH (●–●) and of the oxy-β-HpH subunits (○–○).

K. Idakieva et al. / Biochimica et Biophysica Acta 1764 (2006) 807–814

The absorption spectrum of β-HpH shows a band with maximum at 278 nm, typical for proteins containing aromatic amino acids, and a band at 345 nm due to the copper(II)– peroxide complex. We studied the changes of the characteristic band at 345 nm with a temperature rise at pH 7.2 (didecameric, non-dissociated Hc), and at pH 9.2 (Hc dissociated into subunits). The increase of the temperature caused a decrease of the intensity of this band (Fig. 6). Sigmoid curves were obtained, when the absorbance at 345 nm was plotted as a function of the temperature (Fig. 6, inset). A Tm value of 65.7 °C was determined for the oxy-form of βHpH in 50 mM MOPS, pH 7.2, containing 5 mM CaCl2 and 5 mM MgCl2. Lower Tm of 60.4 °C was determined after dissociation of the β-HpH into subunits at pH 9.2. After the temperature rise up to 75 °C, no free copper was detected by atomic absorption. Indeed, after passing through a PD-10 column in order to remove possibly liberated copper, the protein fraction still contained 0.23% (w/w) copper, characteristic for native gastropodan Hc. The increase of the temperature causes a loss of oxygen bound to the binuclear copper active sites in the Hc molecule. This process is, at least partially, reversible as after cooling of the protein samples, around 60% reinstatement of the copper–peroxide band has been observed. The thermal denaturation of the proteins leads to disturbance of the environment of tryptophyl and tyrosyl residues. We observed also the effect on the absorbance at 278 nm on increasing the temperature; here any change in this parameter would indicate the beginning of the thermal unfolding of the Hc. Comparing the decrease in the absorption at both wavelengths, 278 and 345 nm (Fig. 6), the increase of the temperature influences the active site at lower temperatures than the environments of tryptophans and tyrosines. Compared to the protein dissociated into subunits, the loss of oxygen and the changes in the microenvironment of the tryptophyl and tyrosyl residues occur at higher temperature in the didecameric β-HpH. In conclusion, the present investigations allow physicochemical characterization of the β-HpH and the constituent structural subunits. Tryptophyl residues buried in the hydrophobic interior of the protein molecule determine the fluorescence emission of the oxy and apo-forms of the investigated Hc. The oxygenated and copper-deprived forms of the β-HpH and its subunits exist in different conformations. Both factors, oligomerization and the copper–dioxygen system at the active site, are important for stabilizing the structure of the hemocyanin molecule.

Acknowledgements We wish to thank the Fund for Scientific Research, Flanders (Belgium) and the Bulgarian Academy of Sciences for financial support of the joint research project “Structure and conformational stability of gastropodan hemocyanin”. This work was also supported by research grant X-1209 from the NCSI of the Ministry of Education and Science,

813

Bulgaria. We are grateful to the Katholieke Universiteit Leuven (Belgium) for an IRO fellowship (N.I.S.). References [1] E.F.J. van Bruggen, W.G. Schutter, J.F.L. van Breemen, M.M.C. Bijlholt, T. Wichertjes, Arthropodan and molluscan hemocyanins, in: J.R. Harris (Ed.), Electron Microscopy of Proteins, vol. I, Academic Press, London, New York, 1981, pp. 1–38. [2] G. Préaux, C. Gielens, Hemocyanins, in: R. Lontie (Ed.), Copper Proteins and Copper Enzymes, vol. II, CRC Press, Boca Raton, FL, 1984, pp. 159–205. [3] K.E. van Holde, K.I. Miller, Hemocyanins, Adv. Protein Chem. 47 (1995) 1–81. [4] C.F.S. Bonafe, J.V. Araujo, J.L. Silva, Intermediate states of assembly in the dissociation of gastropod hemocyanin by hydrostatic pressure, Biochemistry 33 (1994) 2651–2660. [5] M. Guzman-Casado, A. Parody-Morreale, P.L. Mateo, J.M. Sanchez-Ruiz, Differential scanning calorimetry of lobster hemocyanin, Eur. J. Biochem. 188 (1990) 181–185. [6] K. Idakieva, K. Parvanova, S. Todinova, Differential scanning calorimetry of the irreversible denaturation of Rapana thomasiana (marine snail, Gastropod) hemocyanin, BBA 1748 (2005) 50–56. [7] R. Hübler, B. Fertl, N. Hellmann, H. Decker, On the stability of the 24meric hemocyanin from Eurypelma californicum, BBA 1383 (1998) 327–339. [8] R. Favilla, M. Goldoni, A. Mazzini, P. Di Muro, B. Salvato, M. Beltramini, Guanidinium chloride induced unfolding of a hemocyanin subunit from Carcinus aestuarii: I. Apo form, Biochim. Biophys. Acta 1597 (2002) 42–50. [9] R. Favilla, M. Goldoni, A. Mazzini, F. Del Signore, P. Di Muro, B. Salvato, M. Beltramini, Guanidinium chloride induced unfolding of a hemocyanin subunit from Carcinus aestuarii. I. Holo form, Biochim. Biophys. Acta 1597 (2002) 51–59. [10] A. Sabatucci, P. Vachette, M. Beltramini, B. Salvato, E. Dainese, Comparative structural analysis of low-molecular mass fragments of Rapana venosa hemocyanin obtained using two different procedures, J. Struct. Biol. 149 (2005) 127–137. [11] R. Lontie, Components, functional units, and active sites of Helix pomatia hemocyanin, Life Chem. Rep., Suppl. 1 (1983) 109–120. [12] R. Drexel, S. Siegmund, H.-J. Schneider, B. Linzen, C. Gielens, G. Préaux, R. Lontie, J. Kellermann, F. Lottspeich, Complete amino acid sequence of a functional unit from a molluscan hemocyanin (Helix pomatia), Biol. Chem. Hoppe-Seyler 368 (1987) 617–635. [13] X.-Q. Xin, C. Gielens, R. Witters, G. Préaux, Amino-acid sequence of functional unit g from the βc-hemocyanin of Helix pomatia, in: G. Préaux, R. Lontie (Eds.), Invertebrate Dioxygen Carriers, Leuven Univ. Press, Leuven, 1990, pp. 113–117. [14] E.J. Wood, M.F. Chaplin, C. Gielens, J. De Sadeleer, G. Préaux, R. Lontie, Relative molecular mass of the polypeptide chain of βc-hemocyanin of Helix pomatia and carbohydrate composition of the functional units, Comp. Biochem. Physiol. 82B (1985) 179–186. [15] R. Vanhoegaerden, C. Gielens, G. Préaux, Location of the carbohydrate in the functional unit d of the βc-hemocyanin of Helix pomatia, Arch. Int. Physiol. Biochem. 96 (1988) B64. [16] R. Vanhoegaerden, C. Gielens, G. Préaux, Location of the carbohydrate chains in functional unit g of the βc-hemocyanin of Helix pomatia, in: G. Préaux, R. Lontie (Eds.), Invertebrate Dioxygen Carriers, Leuven Univ. Press, Leuven, 1990, pp. 153–156. [17] C. Gielens, N. De Geest, F. Compernolle, G. Préaux, Glycosylation sites of hemocyanins of Helix pomatia and Sepia officinalis, Micron 35 (2004) 99–100. [18] C. Gielens, N. De Geest, X.-Q. Xin, B. Devreese, J. Van Beeumen, G. Préaux, Evidence for a cysteine-histidine thioether bridge in functional units of molluscan hemocyanins and location of the disilfide bridges in functional units d and g of the βc-hemocyanin of Helix pomatia, Eur. J. Biochem. 248 (1997) 879–888.

814

K. Idakieva et al. / Biochimica et Biophysica Acta 1764 (2006) 807–814

[19] M. Brouwer, C. Bonaventura, J. Bonaventura, Effect of oxygen and allosteric effectors on structural stability of oligomeric hemocyanins of the arthropod, Limulus polyphemus, and the mollusc, Helix pomatia, Biochemistry 20 (1981) 1842–1848. [20] C. Gielens, L.J. Verschueren, G. Préaux, R. Lontie, Localization of the domains in the polypeptide chain of βc-hemocyanin of Helix pomatia, in: J. Lamy, J. Lamy (Eds.), Invertebrate Oxygen Binding Proteins: Structure, Active Site, and Function, Marcel Dekker, NY, 1981, pp. 295–303. [21] M. Beltramini, F. Ricchelli, B. Salvato, The kinetics of the reaction of Octopus vulgaris hemocyanin with cyanide. Its significance for the structure of the 11S subunit of molluscan hemocyanin, Inorg. Chim. Acta 92 (1984) 209–217. [22] R. Witters, R. Lontie, Stability regions and amino acid composition of gastropod haemocyanins, in: F. Ghiretti (Ed.), Physiology and Biochemistry of Haemocyanins, Academic Press, London, 1968, pp. 61–73. [23] E.P. Kirby, R.F. Steiner, The tryptophan microenvironment in apomyoglobin, J. Biol. Chem. 245 (1970) 6300–6306. [24] S.S. Lehrer, Solute perturbation of protein fluorescence. The quenching of the tryptophyl fluorescence of model compounds and of lysozyme by iodide ion, Biochemistry 10 (1971) 3254–3263. [25] M.R. Eftink, C.A. Ghiron, Exposure of tryptophyl residues in proteins. Quantitative determination by fluorescence quenching studies, Biochemistry 15 (1976) 672–680. [26] E.P. Kirby, R.F. Steiner, The influence of solvent and temperature upon the fluorescence of indole derivatives, J. Phys. Chem. 74 (1970) 4480–4490. [27] P.L. Privalov, V.V. Plotnikov, V.V. Filimonov, Scanning microcalorimeters for studying macromolecules, J. Chem. Thermodyn. 7 (1975) 41–47. [28] E.A. Burstein, N.S. Vedenkina, M.N. Ivkova, Fluorescence and the

[29]

[30] [31]

[32] [33]

[34]

[35]

[36]

[37]

location of tryptophan residues in protein molecules, Photochem. Photobiol. 18 (1973) 263–279. W. Erker, R. Hübler, H. Decker, Structure-based calculation of multi-donor multi-acceptor fluorescence resonance energy transfer in the 4 × 6-mer tarantula hemocyanin, Eur. Biophys. J. 33 (2004) 386–395. L. Stryer, Fluorescence energy transfer as a spectroscopic ruler, Annu. Rev. Biochem. 47 (1978) 819–846. P. Dolashka, N. Genov, K. Parvanova, W. Voelter, M. Geiger, S. Stoeva, Rapana thomasiana grosse (gastropoda) hemocyanin: spectroscopic studies of the structure and conformational stability of the native protein and its structural subunits, Biochem. J. 315 (1996) 139–144. M.R. Eftink, C.A. Ghiron, Fluorescence quenching studies with proteins, Anal. Biochem. 114 (1981) 199–227. M. Perbandt, E. Guthöhrlein, W. Rypniewski, K. Idakieva, S. Stoeva, W. Voelter, N. Genov, Ch. Betzel, The structure of a functional unit from the wall of a gastropod hemocyanin offers a possible mechanism for cooperativity, Biochemistry 42 (2003) 6341–6346. D. Georgieva, D. Schwark, P. Nikolov, K. Idakieva, K. Parvanova, K. Dierks, N. Genov, Ch. Betzel, Conformational states of the Rapana thomasiana hemocyanin and its substructures by dynamic light scattering and time-resolved fluorescence spectroscopy, Biophys. J. 88 (2005) 1276–1282. R. Sterner, T. Vogl, H.-J. Hinz, F. Penz, R. Hoff, R. Föll, H. Decker, Extreme thermostability of tarantula hemocyanin, FEBS Lett. 364 (1995) 9–12. D.N. Georgieva, S. Stoeva, A.A. Ali, A. Abbasi, N. Genov, W. Voelter, Circular dichroism study of the hemocyanin thermostability, Spectrochim. Acta, A 54 (1998) 765–771. N. Genov, M. Shopova, R. Boteva, G. Jori, F. Ricchelli, Chemical, photochemical and spectroscopic characterization of an alkaline proteinase from Bacillus subtilis variant DY, Biochem. J. 207 (1982) 193–200.