myoglobin aqueous solutions: A dynamic light scattering study

Journal of Molecular Liquids 209 (2015) 294–300

Contents lists available at ScienceDirect

Journal of Molecular Liquids journal homepage: www.elsevier.com/locate/molliq

Study of the dynamical behavior of sodium alginate/myoglobin aqueous solutions: A dynamic light scattering study Caterina Branca ⁎, Ulderico Wanderlingh, Giovanna D'Angelo, Cristina Crupi, Simona Rifici Dipartimento di Fisica e Scienze della Terra, Università di Messina, Viale Stagno D'Alcontres 31, 98166 Messina, Italy

a r t i c l e

i n f o

Article history: Received 9 April 2015 Received in revised form 28 May 2015 Accepted 1 June 2015 Available online 11 June 2015 Keywords: Polysaccharide Protein Dynamic light scattering Dilute solutions

a b s t r a c t In the last years, protein/polysaccharide systems have received increasing research interest since they play an essential role in stabilizing food formulations, pharmaceutical preparations and so on. In this study, the diffusive properties of both sodium alginate and sodium alginate/myoglobin water solutions were characterized by dynamic light scattering. By analyzing the effects of concentration, pH and ionic strength on the diffusion coefficient, it emerged that sodium alginate forms aggregates which diffuse as pH sensitive rigid spheres. In the presence of the protein, the polysaccharide is less prone to self-aggregation favoring the formation of more solvent exposed protein/polysaccharide aggregates. These aggregates resulted to be both pH and ionic strength sensitive since their formation arises from a balance between coexisting repulsive and attractive interactions that are strongly influenced by the environmental conditions. By Fourier-Transform Infrared spectroscopy we evidenced the structural changes that take place in myoglobin during the protein–alginate complex formation. The observed increase in the solvent exposed extended chain conformation supported the hypothesis of a more open conformation adopted by the polysaccharide–protein complex with a consequent reorganization of hydrogen bonding. © 2015 Elsevier B.V. All rights reserved.

1. Introduction In the last decades much interest has been focused on protein– polysaccharide complexes in response to an increasing demand for novel functional ingredients in pharmaceutical and food industry. Proteins and polysaccharides are for example used together in gels, foams and emulsions [1–3]; they are also used to formulate blood derived products, vaccines and drug delivery systems [4,5]. In the field of pharmaceuticals, one of the greatest challenges is to improve protein stabilization by embedding it in a polysaccharide matrix. Even if this technique is now widely applied, understanding the mechanisms underlying the protein stability is still a complex problem. Previous studies evidenced the presence of aggregative processes mainly driven by electrostatic forces even if hydrogen, hydrophobic or covalent bonds may also contribute significantly [2,6–8]. Consequently, the optimal stability ranges for proteins will depend on a variety of conditions and any study on protein–polysaccharide complexes must take this into account. One of the most frequently used polysaccharides is sodium alginate, NaAlg, (NaC6H7O6)n, the salt of the long chain carbohydrate alginic acid [9]. It can be found naturally at low concentrations in seawater, where it is secreted by several species of brown algae. Sodium alginate can be characterized as an anionic copolymer comprised of mannuronic acid ⁎ Corresponding author. E-mail address: [email protected] (C. Branca).

http://dx.doi.org/10.1016/j.molliq.2015.06.002 0167-7322/© 2015 Elsevier B.V. All rights reserved.

(M block) and guluronic acid (G block) units arranged along the chain in an irregular pattern of varying proportion of GG, MM and MG blocks [10]. Alginates have COO– and COOH groups along the chain conferring different charge densities depending on pH. The hydrophilic and hydrophobic units along the chain can be altered by the protonation and deprotonation of carboxyl groups in the backbone chain [11]. Alginate is both biopolymer and polyelectrolyte and is considered to be biocompatible, non-toxic, non-immunogenic and biodegradable, which make it an attractive candidate for biomedical applications [12]. Commercially available alginates are used in a variety of technical applications like thickening and gelling agents in the food industry or to immobilize cells in the biotechnological industries [13]. For these reasons, a majority of studies have focused on its gelling properties whereas few studies have investigated the interactions between sodium alginate and surfactants [11,14], or proteins, like BSA [7] and β-lactoglobulin [8]. Starting from these considerations, in this paper we focused our attention on the changes induced by the presence of myoglobin on the diffusive and conformational properties of sodium alginate solutions. Myoglobin has been usually used as a model protein to check binding mechanisms and to reveal structural changes in its native conformation due to its peculiar structure and function. The structure of myoglobin, consisting of a single polypeptide chain of 154 amino acids [15], was first delineated by John Kendrew and colleagues over 40 years ago [16–18]. From these and subsequent works, it emerged that the secondary structure is unusual in that it contains a very high proportion (75%) of α-helical secondary structure. Myoglobin is found mainly in

C. Branca et al. / Journal of Molecular Liquids 209 (2015) 294–300

the muscle tissue where it serves as an intracellular storage site for oxygen [15]. Previous studies showed that globular proteins can interact with anionic polysaccharides to form soluble or insoluble complexes that can be stabilized by electrostatic, ion–dipole or hydrophobic interactions [8]. Several mechanisms of interactions were hypothesized for protein–alginate system, and among them noncovalent electrostatic interactions were mostly studied [19,20], but notwithstanding much effort, they are far from being satisfactorily elucidated. The mechanisms of interactions have obviously a strong influence on the dynamical properties of these systems that up to date were scarcely investigated. Since the mobility, and hence the flexibility of these systems, influences directly the functional activity of the protein, studying the diffusion dynamics of protein–polysaccharide systems under different environmental conditions is of crucial importance. This should lead to the development of novel applications and support the understanding of protein–polysaccharide interactions in general. In the present study, the dynamical properties of both sodium alginate and sodium alginate/myoglobin water solutions were investigated by dynamic light-scattering (DLS) under different environmental conditions. Moreover, in order to relate the differences in the dynamical properties to structural changes occurring in the systems, Fourier Transform Infrared (FTIR) spectroscopy was also employed. The results were compared with previous investigations on aqueous solutions of sodium alginate tested under similar conditions [21–23]. Such information is of crucial importance for the formulation and optimization of new biocompatible products. 2. Materials and methods 2.1. Materials and solution preparation Alginic acid sodium salt and horse skeletal muscle myoglobin were purchased from Aldrich (Milwaukee, WI) and Sigma (St Louis, MO), respectively. According to the specifications from the manufacturer, the sodium alginate sample has a weight-average molecular weight of 120,000–190,000 g/mol and a guluronic acid to mannuronic acid (M/G) ratio of 1.56. As reported in the product information sheet by Sigma, and as also detected by Sheng and Pawliszyn [24], two components can be identified for the myoglobin used in this work; a major component with a pI value of 7.3 and a minor component with a pI of 6.8. Nevertheless, according to recent studies on the same Sigma product [25,26], a value of 7.2 can be considered as a good estimation of the myoglobin pI. For this reason we refer to this value for our following considerations. Like all the globular proteins, at a pH value above the isoelectric point the protein shows a negative net charge. All commercially solvents and reagents used were of analytical grade and no further purifications were made. The aqueous solutions of sodium alginate were investigated at different concentrations, c (from 0.25 mg/ml up to 5 mg/ml). Each solution was prepared by dissolving a weighed amount of sodium alginate in doubly distilled water; to ensure complete dispersion, the solutions were heated at 30 °C and stirred for 24 h. To free these solutions from dust, they were filtered, through 0.8 μm Millipore filters. All concentrations were prepared at all the desired pH values (3.5, 4.5, 5.5, 6.5, 8, 10.5) using HCl or NaOH. NaCl was added to these solutions in order to have the same starting ionic strength. Then, for each pH value, the ionic strength of the media was varied by adding more NaCl. The contribution to the ionic strength of the sodium counterions from the alginate, about 20 mmol/dm3 according to the Manning theory [27], was not considered. A stock solution was prepared by dissolving myoglobin (0.1 mg/ml) in water. The protein/polysaccharide mixtures were prepared by mixing the sodium alginate solution (0.3 mg/ml) at different pHs with the protein stock solution. The starting solutions were then diluted

295

obtaining concentrations of 0.14 mg/ml, 0.2 mg/ml, 0.3 mg/ml, and 0.4 mg/ml at pH values of 5.5, 6.5, 7.5 and 9. The ionic strength of the mixed solution was varied by the addition of NaCl (10 mM–100 mM). High concentrations of myoglobin (40 mg/ml) and sodium alginate solutions were used to obtain accurate infrared spectra. In more details, we prepared a protein/polysaccharide mixture with a final concentration of 10 mg/ml myoglobin and 3% w/v sodium alginate. This concentration was chosen to maintain roughly the same polysaccharide/ protein ratio used for DLS measurements. The pH of each solution was maintained at approximately 7.5. Measurements were performed at 25 °C. 2.2. Dynamic light scattering DLS experiments were performed by using an Argon laser source (λ = 488 nm). A standard scattering apparatus with a Brookhaven BI-2030 correlator was employed to analyze the scattered light. Temperature was set at 25 °C and controlled within 0.01 °C by a water circulating apparatus. Every sample was measured several times over the angular range 40°–140°. In a DLS experiment the measured quantity is the normalized intensity autocorrelation function [28–30]: g 2 ðQ; t Þ ¼

bIðQ ; 0ÞI ðQ; t ÞN bI ðQ; 0ÞN2

ð1Þ

where Q = 4πn/λsin(θ/2) is the exchanged wave vector, n being the refraction index of the medium and θ the scattering angle. If the scattered field, E(Q ,t), obeys a Gaussian statistic, g2(Q , t) is related to the normalized field autocorrelation function g 1 ðQ; t Þ ¼

bE  ðQ; 0ÞEðQ ; t ÞN bI ðQ; 0ÞN

ð2Þ

through the Siegert's relation g2(Q, t) = 1 + |α ⋅ g1(Q, t)|2, α being a constant depending on the experimental setup. For diffusing monodisperse spherical scatterers, the normalized scattered electric field autocorrelation function takes a simple exponential form with decay rate Γ: g1(Q , t) = exp(− Γt). Because the investigated solutions were slightly polydisperse, the autocorrelation functions were analyzed in terms of a standard cumulant analysis [28–30]: ln jg 1 ðQ ; t Þj ¼ −bΓNt þ 1=2!μ 2 t 2 –1=3!μ 3 t 3 þ …

ð3Þ

where μn represents the moments of the distribution of the decay rates. If the first moment, bΓN, is found to be proportional to Q2, an effective diffusion coefficient can be determined as Deff = bΓN/Q2. By fitting ln|g1(Q ,t)| to a quadratic in t, the mean, bΓN, and the variance, μ2, were obtained. The ratio of variance to the square of the mean is a measure of the polydispersity index, PI, of the diffusion coefficient, PI = μ2/bΓN2. Before experimentation, the DLS setup was tested using standard polystyrene microbead solutions; precise RH values have been obtained for samples with particle size between 1 nm and 1 μm with an error of ±2%. For these solutions, it was found that the detector accepts about one coherence area, α ≈ 0.96. To judge the data quality, only those for which the percent difference between the calculated and measured baseline, reported by the software correlator, was no more than 0.05% were considered. A reduced chi-square parameter less than 0.0001 was used as a measure of goodness-of-fit of the cumulant analysis. Each DLS experiment was repeated in triplicate. Finally, the Stokes–Einstein relation was used to estimate an average radius of hydration, bRHN = kBT/6πηD0, kB being the Boltzmann's constant, T the absolute temperature and η the solvent viscosity; D0 was obtained from an extrapolation of Deff to zero concentration.

296

C. Branca et al. / Journal of Molecular Liquids 209 (2015) 294–300

400

2.3. Fourier Transform Infrared spectroscopy

3. Results and discussion 3.1. DLS Firstly, we investigated the concentration and pH dependence of the diffusive properties of sodium alginate in water solution. In Fig. 1 typical plots of the normalized correlation functions for a solution of sodium alginate at pH 5.5 and at a concentration c = 1.4 mg/ml are shown at two different scattering angles, θ, 40° and 140°. All the functions were fitted using the cumulant expansion truncated at the second order, Eq. (3), and from the fitting procedure the mean decay rate b Γ N was obtained. As an example, in Fig. 2 the mean relaxation rates were plotted against the square of the scattering vector, Q2, for three concentrations (0.7 mg/ml, 1.4 mg/ml and 4.2 mg/ml) at the same pH value of 5.5. As can be seen from the figure, relaxation rates exhibit nearly linear dependence on Q2; the linear regression lines pass through the origin suggesting the existence of a Brownian diffusion of spherical and quite rigid macromolecules. From the slope of the straight line we determined the effective diffusion coefficient, Deff. Notwithstanding Deff was found to be nearly independent of the scattering angle for all the investigated

c=0.7 mg/ml c=1.4 mg/ml c=4.2 mg/ml

300 -1

<Γ> (sec )

All FT-IR spectra were recorded by a Nicolet Nexus 5770 and performing 512 scans with a resolution of 2 cm−1. All samples were placed between two CaF2 windows, with a 0.015 mm Teflon spacer. Water vapor was subtracted from the spectra, and the sloping background originating mainly from water's combination band was subtracted by using the multipoint spline function routine. The background subtraction was carefully inspected for flatness in the 2000–1800 cm−1 region. As it is well known, the identification of the various structures is complicated by the presence of water: in fact, it has strong IR absorbance with a bending vibrational band at 1645 cm−1 that overlaps directly with the amide I vibration. However, since we are interested in studying the protein polysaccharide system in the most native environment possible, no water subtraction was attempted. Rather basing on a band fitting of the amide regime I, the fractional areas of the fitted components, directly proportional to the relative proportions of structure that they represent, were evaluated. The amide I band was deconvoluted using standard algorithms with all fit parameters having a 95% confidence interval. To avoid any ambiguity as to the number of components, we monitored both χ2 and the sum of the fit residual and selected the number of Gaussian components that minimized both the parameters. The assignments used in this work for myoglobin are in agreement with those used by other investigators [31–34].

200 100 0 0

200

400 600 800 1000 1200 2 -2 Q (μm )

Fig. 2. Mean decay rate, bΓ N, of sodium alginate aqueous solutions (pH = 5.5) as a function of Q2 at different concentrations. In the plot the linear fits of data are reported.

samples, all the Deff values reported in the following were obtained by linear least squares extrapolation to a zero angle at all values of Q2. For all the samples, the polydispersity was also evaluated by extrapolation to zero angle. As a result of this extrapolation, we obtained a polydispersity index ranging between a minimum of 0.05 and a maximum of about 0.11. As an example, in Fig. 3 the concentration dependence of both the polydispersity index and the effective diffusion coefficient was reported for the sodium alginate solutions at pH 5.5. As expected, the polydispersity increases with concentration, reaching a value of about 0.11 for concentrations higher than 3 mg/ml. We can therefore consider negligible the influence of polydispersity on the observed changes of diffusivity. For all the pH values investigated a similar concentration dependence for the diffusion coefficient was observed (Fig. 4). This behavior is in agreement with previous studies according to which, for highly charged macromolecules in solution, different concentration regimes were distinguished [35]. At very low concentrations (dilute regime), the DLS measurements evidenced a fast relaxation mode independent of the concentration of the polyelectrolyte [36]. This fast mode was interpreted as a result of a coupled diffusion of polyions and counterions [36,37]. At higher concentrations but still below the overlap concentration, c*, a transition regime of semiflexible polyions was identified. In this regime, beside a fast diffusion, that was found to follow a power law, [38–41] a slow mode was observed in many systems even if its origin is still under debate. According to the most accredited proposals, this slow mode is related with the presence of multichain domains [36,37,42,43] whose formation is highly influenced by many factors (such as polyelectrolyte molecular weight [36,37,44] and concentration

1.0 -1

Deff (10 cm sec )

0.6

-9

0.4 0.2 0.0

-4

10

0.10

2

θ

4

-3

10

-2

10

0.08

PI

g1 (Q,t)

pH=5.5

θ

0.8

2

0.06

-1

10

0 Fig. 1. Correlation functions of sodium alginate aqueous solution for a concentration c = 1.4 mg/ml at pH = 5.5 at two scattering angles. The continuous lines are the results obtained by using the second order cumulant expansion (see text for details).

1

2 3 c (mg/ml)

4

5

Fig. 3. Concentration dependence of the diffusion coefficient and of the polydispersity index of sodium alginate solutions at pH = 5.5.

C. Branca et al. / Journal of Molecular Liquids 209 (2015) 294–300

Fig. 4. Concentration dependence of the diffusion coefficient for sodium alginate solutions at different pH values. The solid lines are fits to a power law behavior (see text for details).

[37,44]); notwithstanding some controversies, the slow diffusion coefficient was found to decrease with concentration [36,37,44]. Above c*, the overlapping flexible polymer coils form a transient network and polyelectrolyte systems were found to behave similarly to that of uncharged polymer systems. The slow relaxation mode becomes more pronounced in this more concentrated regime [45,46] which, depending on its Q dependence may be interpreted in various ways [47]. For example, a Q2 dependent behavior has been observed both for semidilute polyelectrolyte systems [44–46,48], as well as for semidilute systems of uncharged polymers, associating polymers and for gelling polymer systems [47]. As above stated, the experimental conditions we have chosen in the present study allowed us to identify, at any pH value, only one slow diffusion characterized by a diffusion coefficient, Deff, that is almost constant, within the experimental error, in the very dilute region (c b 0.5 mg/ml), and then drops sharply with increasing concentration, thus suggesting the existence of two concentration regimes. In view of a previous study on the same system [22], it is reasonable to assume that the concentration of 0.5 mg/ml marks the cross-over from the dilute to the semi-dilute regime in which the diffusion of the aggregates becomes slower and slower with increasing concentration. The observed pronounced Q2 angular dependence of the diffusion coefficient, together with the condition QR ≫ 1, support this interpretation [37]. On the other hand, the presence of structures on scales considerably larger than the chain size was also hypothesized to account for the enormous scattering at very low wavevectors that dominates the small-angle neutron scattering profile of low ionic strength polyelectrolyte solutions [43]. More specifically, the presence of large aggregates was observed in previous works on sodium alginate solutions [6,49,50]. The observed slowing of the diffusion in the semidilute regime with increasing concentration conforms to the most common observations for the slow diffusion coefficient in polyelectrolyte solutions, even if its analytical dependence on concentration is still under debate and a number of proposals can be found in literature. In most of the works on polyelectrolyte solutions, it was described in terms of a power law [37,44,45] with an exponent, that resulted to be strongly dependent on the system, the preparation method, the environmental conditions and, therefore, on the type of interactions that are dominant in the macromolecular solutions. In the present case, it was found that, as revealed by the straight lines reported in Fig. 4, Deff satisfies a power law on c with exponent close to 0.7 for all the pH values investigated. Finally, we tried to obtain a rough estimation of an average dimension of the aggregates formed within the sodium alginate solution. Under the conditions of the present work, the usage of the StokesEinstein formula, to obtain an apparent hydrodynamic radius, RH, is obviously questionable, especially because the viscosity experienced by the large aggregates could not be the viscosity of the solvent [51]. Anyway, since we were interested in following the relative pH-

7 -1

1 c (mg/ml)

2

pH4.5 pH6.5 pH10.5

pH 5.5 pH 7.5

pH 6.5 pH 9

6

-9

1

pH3.5 pH5.5 pH8

induced size changes, we used the solvent viscosity in the Stokes– Einstein formula just to obtain an estimate of the radius of a hypothetical sphere that diffuses at the same speed as the aggregates being measured. From this analysis, it resulted that, starting from an initial value of 290 nm for pH 3.5, bRHN increased, reaching a value of 550 nm at pH 5.5, and then decreased again to 300 nm at pH = 10.5. Taking into account the experimental uncertainties, it emerged that the most significant changes in size can be observed below pH 5.5 and above pH 7. The decrease observed under alkaline conditions can be attributed to the depolymerization of the polysaccharide chains [21], whereas as the pH approaches the pKa of the molecule (pKa = 3.2–3.65), the reduction in size of the aggregates is a consequence of the protonation of the alginic acid that determines a decrease of the intramolecular electrostatic repulsion. This was also observed by Avaltroni and colleagues [21], which ascribed the reduction in size by lowering pH to the progressive protonation of the alginic acid with the appearance of hydrophobic segments in alginate chains. The presence of hydrophilic–hydrophobic aggregates in aqueous solutions depending on pH was also observed by Cao and colleagues [52]. Taking in mind all the above information on sodium alginate, we then investigated the diffusive properties of the polysaccharide/protein mixture under different pH and ionic strength conditions in the dilute regime. For this reason, a mixture having an initial concentration of 0.3 mg/ml of sodium alginate and 0.1 mg/ml of myoglobin was prepared. All the measurements were performed above pH 5.5 to avoid segregative phase separation [53–57]. The starting mixed solutions, prepared at different pH values, were gradually diluted and the corresponding correlation functions were recorded. Similarly to the sodium alginate solutions, the intensity autocorrelation functions of the mixed solutions, all slightly deviating from single exponential behavior, were analyzed by fitting to a cumulant expansion truncated at the second order. In Fig. 5 the concentration dependence of the effective diffusion coefficient for the sodium alginate/myoglobin water solution at pH values of 5.5, 6.5, 7.5 and 9 is reported. As can be seen from this figure, the diffusion coefficient increases by increasing concentration suggesting that, in the presence of the protein, the polysaccharide is less prone to aggregate, preferring the formation of more solvent exposed protein/polysaccharide complexes. We can therefore hypothesize that the surface exposure of the buried charged sites of myoglobin is enhanced thus favoring the formation on the protein surface of patches oppositely charged. Accordingly to previous studies on polysaccharide–protein interactions in solution [58–61], the formation of the protein/polysaccharide complex can be considered as the result of an electrostatic balance between these patches and the negatively charged sodium alginate. As observed for the polysaccharide solution, also in this case the diffusive properties

Deff (10 cm sec )

-9

2

-1

Deff (10 cm sec )

10

297

5 0.1

0.2 0.3 c (mg/ml)

0.4

Fig. 5. Concentration dependence of the effective diffusion coefficient, Deff, for a sodium alginate/myoglobin water solution at different pH values.

298

C. Branca et al. / Journal of Molecular Liquids 209 (2015) 294–300

Absorbance (a.u.)

are pH sensitive; attractive electrostatic interactions predominate below the isoelectric point of the protein (7.2) where the protein and the polysaccharide, oppositely charged, form aggregates that move slower. To better evidence the role played by the electrostatic interactions, the effect of the addition of a salt on the diffusion properties of the protein/polysaccharide complex was considered. As one can see from Fig. 6, for the protein/polysaccharide complex, Deff resulted to be significantly affected by the variation of the ionic strength, especially at the lowest NaCl concentrations. Contrarily, according to previous studies [21,22], for the sodium alginate solutions Deff resulted to be independent, within the experimental uncertainties, of the salt concentration in the range 10–50 mM (see the inset of Fig. 6). Once more, the behavior observed for the protein/polysaccharide complex can be explained by taking into account the electrostatic nature of the existing interactions that are screened at increasing salt concentration. It must be also noted that this screening effect is more marked far from the isoelectric point of the protein; in fact, at this pH point, the electrostatic interactions are minimized and, consequently, the diffusion is less sensitive to ionic strength changes.

3.2. FT-IR In order to validate the hypothesis suggested by the DLS investigation of formation of solvent exposed protein/polysaccharide complexes, the FTIR amide I band was investigated; in fact, it is rightly considered uniquely useful for the analysis of protein secondary structural composition and conformational changes. Fig. 7 shows the absorption spectra in the region 1400–1800 cm−1 of myoglobin, sodium alginate and myoglobin/sodium alginate mixture at a pH value of 7.5, where the prominent conformational changes were supposed to occur according to the DLS data. In all of them the contribution assigned to the bending of water is present (1645 cm−1); for the samples containing protein, it overlaps with the random coil contribution. For the myoglobin mixture, the best fit to the data was obtained through a deconvolution into five Gaussian components each assigned, according to previous works [31–34], to a secondary structure component. The fractional area of each of these components (see Table 1) was calculated by dividing the analytical area of the corresponding Gaussian component by the total area of the amide I band. As it can be observed, two major components are present in the amide I region: the just discussed band at 1645 cm−1, and the band at 1652 cm−1,

1500 1600 1700 -1 Wavenumber (cm ) Fig. 7. FT-IR absorption spectra at T = 25 °C, along with their deconvolution into Gaussian bands, of (from top to bottom) sodium alginate, myoglobin and sodium alginate/ myoglobin solutions, at the same water concentration, at pH = 7.5. Solid, dashed, and dotted lines represent the contributions from myoglobin, sodium alginate, and water, respectively.

typically assigned to α-helix. The obtained secondary structure is in agreement with that reported in literature [31–34]. In the same spectral region, the band component analysis for the sodium alginate water solution identifies three bands: a band centered at 1614 cm−1 attributed to the antisymmetric stretch of the carboxyl groups, a band at 1646 cm−1 assigned to the water bending mode and a high-frequency band, 1688 cm−1, that can be attributed to water coordinated to a cation such as Na+. The deconvolutions obtained for sodium alginate and myoglobin were used as a starting point to perform the fitting procedure on the sodium alginate/myoglobin sample. As one can see in Fig. 7, the spectrum of the resulting solution showed some differences in respect to the single components. In particular, whereas no variation in peak frequency can be detected for the components attributed to myoglobin, a slight decrease of the antisymmetric stretch of the carboxyl groups of sodium alginate can be observed (from 1614 to 1594 cm−1). This red-shift can be attributed to the formation of more hydrogen bonds between carboxyl groups and water molecules. This hypothesis is

Table 1 Values of the normalized integrated area for the secondary structures present in the native myoglobin and in the sodium alginate/myoglobin mixture. Percentage of structure (%)

Fig. 6. Dependence of the diffusion coefficient, Deff, on NaCl content for a sodium alginate/ myoglobin water solution at different pH values. The protein and polysaccharide concentrations are 0.1 mg/ml and 0.3 mg/ml, respectively. In the inset the dependence of Deff on the NaCl concentration for a sodium alginate solution at a concentration c = 0.3 mg/ml and at different pH values is shown.

Myoglobin Myoglobin/sodium alginate a b c

α-Helix+ randoma

Extended chainb

β-Turnc

68 51

19 38

12 11

Band at 1650 cm−1 for α-helix and at 1645 cm−1 for random structure. Band at 1624 and 1632 cm−1. Band at 1668 cm−1.

C. Branca et al. / Journal of Molecular Liquids 209 (2015) 294–300

supported by the DLS observations above reported, according to which the conformation adopted by the protein/polysaccharide complex in water enhances the solute–solvent interactions of the different functional groups. As for myoglobin, also for the protein/polysaccharide mixture the fractional area of each secondary structure component was calculated; the obtained values are also reported in Table 1. According to previous studies, [33,62,63] the observed increase of the percentage of extended chain, bands at 1624 and 1632 cm−1, can be ascribed to a higher solvent exposure of the extended chain conformation and, then, to the just cited formation and reorganization of hydrogen bonding. It is reasonable hypothesizing that the conformation adopted by the polysaccharide–protein complex, favoring the solvation of the hydrophobic patches of the protein, prevents protein– protein interactions, that can cause demixing and protein crosslinking. This is confirmed by the absence of the components at 1618 and 1683 cm−1 which are characteristic, according to previous studies, for intermolecular antiparallel β-sheet aggregation [64–67]. Concomitant to the intensity increase of the extended chain contributions, a partial loss of the original α-helix+ random component is observed.

4. Conclusions DLS measurements were performed on sodium alginate and myoglobin/sodium alginate water solution at different environmental conditions. The autocorrelation functions of the scattered light were fitted with the help of a cumulant expansion and from the value of the first cumulant an effective diffusion coefficient was evaluated. For both sodium alginate and sodium alginate myoglobin water solutions, the presence of a single slow relaxation mode emerged. According to previous studies, this slow mode was interpreted as caused by aggregates or domains that, under the experimental conditions of the present work, form over all the concentration range investigated and whose dimensions resulted to be pH dependent. For the sodium alginate solution, two distinct concentration regimes were identified at any pH value investigated, with a concentration discontinuity around 0.9 mg/ml. Moreover, the concentration dependence of Deff was analyzed and an interesting transition between two different models was evidenced. The DLS results obtained for sodium alginate were used as reference in order to evidence any change in the dynamics of the polysaccharide induced by the presence of the protein. From this comparison it emerged that, whereas for sodium alginate the diffusion coefficient was found to decrease with increasing concentration and no significant dependence on ionic strength was observed, in the presence of the protein, Deff was found to slightly increase with concentration. Moreover, a marked ionic strength dependence was observed especially at the highest pH values, thus evidencing the strong electrostatic nature of the involved interactions. These results suggested that, in the presence of the protein, the polysaccharide is less prone to self-aggregation favoring the formation of more solvent exposed protein/polysaccharide aggregates. This was also confirmed by the FTIR analysis of the secondary structures of myoglobin in the presence of the polysaccharide, according to which, thanks to a reorganization of hydrogen bonding, the presence of sodium alginate favors the protein–solvent interaction. The higher solvent exposure and the consequent structural rearrangement are, in our opinion, the key factors needed to prevent protein aggregation and preserve, to the extent of possible, the α-helix shape of the secondary protein structure. Considering that to date there is no effective treatment that can prevent such protein assemblies from forming, the results in this study are potentially useful for further development of new biocompatible materials designed to preserve protein functionality, which can be

299

widely applied in many fields such as controlled delivery systems and food formulations. References [1] E. Dickinson, An Introduction to Food Colloids, Oxford University Press, Oxford, New York, 1992. [2] E. Dickinson, Food Hydrocoll. 17 (2003) 25–39. [3] V.B. Tolstoguzov, Functional Properties of Food Proteins and Role of Protein– Polysaccharide Interaction, Protein-polysaccharide interactions, Marcel Dekker, New York, 1997. 171–198. [4] S. Sundar, J. Kundu, S.C. Kundu, Sci. Technol. Adv. Mater. 11 (2010) 014104–014113. [5] R. Langer, N.A. Peppas, AIChE J. 49 (2003) 2990–3006. [6] V. Ducel, P. Saulnier, J. Richard, F. Boury, Colloids Surf. B 41 (2005) 95–102. [7] Y. Zhao, F. Li, M.T. Carvajal, M.T. Harris, J. Colloid Interface Sci. 332 (2009) 345–353. [8] T. Harnsilawat, R. Pongsawatmanit, D.J. McClements, Food Hydrocoll. 20 (2006) 577–585. [9] T.H. Silva, A. Alves, B.M. Ferreira, J.M. Oliveira, L.L. Reys, R.J.F. Ferreira, R.A. Sousa, S.S. Silva, J.F. Mano, R.L. Reis, Int. Mater. Rev. 57 (2012) 276–306. [10] K.I. Draget, Handbook of Hydrocolloids, Woodhead Publishing, Cambridge, U.K., 2000. 379–395. [11] J. Yang, J. Zhao, Y. Fang, Carbohydr. Res. 343 (2008) 719–725. [12] K. Nakamura, Y. Nishimura, T. Hatakeyama, H. Hatakeyama, Thermochim. Acta 267 (1995) 343–353. [13] B.H.A. Rehm, Appl. Microbiol. Biotechnol. 48 (1997) 281–288. [14] J. Yang, S. Chen, Y. Fang, Carbohydr. Polym. 75 (2009) 333–337. [15] G.A. Ordway, D.J. Garry, J. Exp. Biol. 207 (2004) 3441–3446. [16] J.C. Kendrew, G. Bodo, H.M. Dintzis, R.G. Parrish, H. Wyckoff, Nature 181 (1958) 662–666. [17] J.C. Kendrew, R.E. Dickerson, B.E. Strandberg, R.G. Hart, D.R. Davies, D.C. Phillips, V.C. Shore, Nature 185 (1960) 422–427. [18] J.C. Kendrew, Science 139 (1963) 1259–1266. [19] V.B. Tolstoguzov, Food Hydrocoll. 4 (1991) 429–468. [20] G.R. Ziegler, E.A. Foegeding, Adv. Food Nutr. Res. 34 (1990) 203–298. [21] F. Avaltroni, M. Seijo, S. Ulrich, S. Stoll, K.J. Wilkinson, Biomacromolecules 8 (2007) 106–112. [22] K.A. Strand, A. Bøe, P.S. Dalberg, T. Sikkeland, O. Smidsrød, Macromolecules 15 (1982) 570–579. [23] E. Dingsøyr, O. Smidsrød, Br. Polym. J. 9 (1977) 56–61. [24] L. Sheng, J. Pawliszyn, Analyst 127 (2002) 1159–1163. [25] Y. Kuroda, H. Yukinaga, M. Kitano, T. Noguchi, M. Nemati, A. Shibukawa, T. Nakagawa, K. Matsuzaki, J. Pharm. Biomed. Anal. 37 (2005) 423–428. [26] C.A. Nesbitt, K.K.C. Yeung, Analyst 134 (2009) 65–71. [27] G.S. Manning, Acc. Chem. Res. 12 (1979) 443–449. [28] B.J. Berne, R. Pecora, Dynamic Light Scattering With Applications to Chemistry, Biology, and Physics, Wiley-Interscience, New York, 1976. [29] H.Z. Cummins, E.R. Pike, Photon Correlation and Light Beating Spectroscopy, Plenum Press, New York, 1974. [30] B. Chu, Laser Light Scattering-Basic Principle and Practice, Academic, San Diego, 1991. [31] D.M. Byler, H. Susi, Biopolymers 25 (1986) 469–487. [32] A. Dong, P. Huang, W.S. Caughey, Biochemistry 29 (1990) 3303–3308. [33] F. Meersman, L. Smeller, K. Heremans, Biophys. J. 82 (2002) 2635–2644. [34] J. Kong, S. Yu, Acta Biochim. Biophys. Sin. 39 (2007) 549–559. [35] M. Mandel, Dynamic Light Scattering: The Method and Some applications, Clarendon Press, Oxford, 1993. [36] M. Sedálk, E.J. Amis, J. Chem. Phys. 96 (1992) 817–825. [37] M. Sedlák, E.J. Amis, J. Chem. Phys. 96 (1992) 826–834. [38] A.V. Dobrynin, R.H. Colby, M. Rubinstein, Macromolecules 28 (1995) 1859–1871. [39] R.S. Koene, M. Mandel, Macromolecules 16 (1983) 220–227. [40] R.S. Koene, T. Nicolai, M. Mandel, Regime, Macromolecules 16 (1983) 231–236. [41] W. Essafi, A. Abdelli, G. Bouajila, F. Boué, J. Phys. Chem. B 116 (2012) 13525–13537. [42] T. Costa, J.S. de Melo, M.D. Miguel, B. Lindman, K. Schillen, J. Phys. Chem. B 113 (2009) 6205–6214. [43] B.D. Ermi, E.J. Amis, Macromolecules 31 (1998) 7378–7384. [44] S. Förster, M. Schmidt, M. Antonietti, Polymer 31 (1990) 781–792. [45] M. Sedlák, Č. Koňák, P. Štěpánek, J. Jakeš, Polymer 28 (1987) 873–880. [46] E. Buhler, M. Rinaudo, Macromolecules 33 (2000) 2098–2106. [47] B. Nyström, B. Lindman, Macromolecules 28 (1995) 967–974. [48] J.J. Tanahatoe, M.E. Kuil, J. Phys. Chem. B 101 (1997) 9233–9239. [49] T. Windhues, W. Borchard, Eur. Polym. J. 38 (2002) 1219–1227. [50] T. Windhues, W. Borchard, Carbohydr. Polym. 52 (2003) 47–52. [51] S.M. Tosh, A.G. Marangoni, F.R. Hallett, I.J. Britt, Food Hydrocoll. 17 (2003) 503–513. [52] Y. Cao, X. Shen, Y. Chen, J. Guo, Q. Chen, X. Jiang, Biomacromolecules 6 (2005) 2189–2196. [53] A.N. Schechter, C.J. Epstein, J. Mol. Biol. 35 (1968) 567–589. [54] D. Puett, J. Biol. Chem. 248 (1973) 4623–4634. [55] R. Hoskins, I.D. Robb, P.A. Williams, P. Warren, J. Chem. Soc. Faraday Trans. 92 (1996) 4515–4520. [56] R. Hoskins, I.D. Robb, P.A. Williams, Biopolymers 45 (1998) 97–104. [57] J.-L. Doublier, C. Garnier, D. Renard, C. Sanchez, Curr. Opin. Colloid Interface Sci. 5 (2000) 202–214. [58] E. Dickinson, Soft Matter 4 (2008) 932–942. [59] J. Krägel, S.R. Derkatch, R. Miller, Adv. Colloid Interf. Sci. 144 (2008) 38–53. [60] T. Koupantis, V. Kiosseoglou, Food Hydrocoll. 23 (2009) 1156–1163.

300

C. Branca et al. / Journal of Molecular Liquids 209 (2015) 294–300

[61] A.R. Mackie, Curr. Opin. Colloid Interface Sci. 9 (2009) 357–361. [62] P.I. Haris, D. Chapman, Biopolymers 37 (1995) 251–263. [63] R. Gilmanshin, S. Williams, R.H. Callender, W.H. Woodruff, R.B. Dyer, Proc. Natl. Acad. Sci. U. S. A. 94 (1997) 3709–3713. [64] A.A. Ismail, H.H. Mantsch, P.T.T. Wong, Biochim. Biophys. Acta 1121 (1992) 183–188.

[65] A. Martinez, J. Haavik, T. Flatmark, J.L.R. Arrondo, A. Muga, J. Biol. Chem. 271 (1996) 19737–19742. [66] A.C. Dong, T.W. Randolph, J.F. Carpenter, J. Biol. Chem. 275 (2000) 27689–27693. [67] G. Damaschun, H. Damaschun, H. Fabian, K. Gast, R. Krober, M. Wieske, D. Zirwer, Proteins 39 (2000) 204–211.