Polycation-globular protein complex: Ionic strength and chain length effects on the structure and properties

1 Introduction

In recent years, several studies have shown that the associative interaction between proteins and polyelectrolytes makes it possible to control the texture, structure and properties of many food products (1, 2, 3, 4, 5, 6, 7, 8). As a result, various protein/polyelectrolyte systems have been studied for their practical applications in food technology such as protein separation, increasing thermal stability or acid of proteins, and stabilization of foam emulsions. In fact, the addition of polyelectrolytes to a solution composed of mineral particles can also stabilize these latter or encapsulate them for delivery. The encapsulation procedures are also useful in food processing, where these particles are proteins. An abundant literature proves that the character of the protein/polyelectrolyte interactions is affected by environmental factors, in particular pH and ionic strength (9, 10, 11, 12). Indeed, the formed complex is based on electrostatic interactions between opposite charges of polyelectrolyte and protein.

Dubin and co-author, in a study of complexes formed by polycations and proteins, have shown that cationic polyelectrolytes can provide a stronger binding affinity and in addition a large binding amount for acid proteins (13, 14, 15, 16, 17). The strength of electrostatic interactions depends hugely on the variation of pH and ionic strength. Indeed, three critical pH-induced structural transitions, pH_c (formation of primary soluble complex), pH_c’ (end of primary soluble complex) and pH_φ (first appearance of microcoacervate droplets), were identified (13, 14, 15, 16, 17).

In this paper, the objective is to characterize the interactions between proteins and cationic polyelectrolytes using the Bovine Serum Albumin (BSA)-synthetic Polyethylenemine (PEI) system.

Recently, Polyethylenimine is used as a polymer for neutralization of the excessive anionic charges in particular in colloidal suspensions and as cosmetic products, as for example, as thickener ingredient (viscosity adjuster) (18,19). Still, PEI was intensely used as a drug deliverysystem in recent anti-cancer therapies (20,21). The effect of pH and ionic strength on binding of BSA protein onto cationic PEIs was observed by turbidimetric titration, dynamic light scattering (QELS). QELS method provides information about the overall hydrodynamic structure of the complex (hydrodynamic dimension). In particular, the computational method of the electrostatic potential of BSA is another way to know and to identify the functional sites situated at the surface of a protein. This later may be visualized by color-coding the molecular surface versus the potential values. In fact this part of extended work have as objective to know with precisely how PEI binds to BSA where these macromolecules have the same sign of charges.

2 Materials and methods

3 Results

3.1 Effect of salt on rate charge

Generally, the dissociation of the polycation strongly depends on the rate charge and allowed to quantitatively account the charge evolution as a function of the pH. For this, before studying the effect of salt on the complex formed by PEI and BSA, we studied the rate charge of the polyelectrolyte alone versus pH. In solution, the polyelectolyte BOH may be dissolved into ions B⁺ and OH⁻ according to the following chemical reaction and Henderson-Hasselbalch equation:

(6)BOH<=>B++OH−

The rate charge f can be defined as:

(7)f=β+10−pHCmB(10−2pH−1)

where, β is the neutralization rate equal to, β=[NaOH][BOH]+[B+]=[NaOH]CmB.

We present, in Figures 1a and 1b the evolution of the rate of charge f versus pH in the presence and in the absence of the salt (NaCl) for two molar masses of polycation, 2.5 kDa and 25 kDa. The rate charge is inversely proportional to pH i.e. an increase in pH causes a decrease of f. We remarked that the polymer, with Mw = 25 kDa, not affected by the presence of salt at 0.15 M, Figure 1b unlike the case where the molar mass equal to 2.5 kDa. In fact, the effect of presence of salt depends on the length of polyethelenimine.

Figure 1

Evolution of rate of charge f versus pH in the presence (●, [NaCl] = 0.15 M) and in the absence (●, [NaCl] = 0 M) of the salt (NaCl) for molar mass of polycation equal to 2.5 kDa.

3.2 Study of complex formation BSA-PEI (25 kDa)

3.2.1 Turbidity variation of the complex BSA-PEI

The variation of the turbidity of the mixture of PEI (25 kDa) and BSA for a ratio r = Pr:Pl = 5:1 versus pH is presented in Figure 2a In fact, we remind, that turbidity is given by the following relation:

Figure 2

Acid base titration at various salt concentrations of mixture of BSA/PEI (25 kDa) at Pr:Pl = 5:1. (a) Turbidity of the complex vs pH. (b) Hydrodynamic radius of the complex vs pH. The arrows show the pH transitions. ●, [NaCl] = 0 M; ●, [NaCl] = 0.15 M.

(8)A=100−T%

where A is the absorbance and T is the transmission.

As a first step, for pH values lower than 4.6, the turbidity is stable and not changed significantly by the quantity of the added NaOH. By adding more NaOH to bring the pH to value above 4.6, the turbidity increases gradually until pH value about equal to 7.3. After this last value of pH i.e. above ~7.3, the turbidity significantly evolves and reaches rapidly a value equal to 8.7. Finally, the turbidity varies weakly, reaches almost a plateau for pH > 8.7, and regarded as the endpoint of the titration.

We can estimate that pH = 4.6, below the isoelectric point pH_i 4.9, corresponds to the pH_c, which corresponds to the formation of soluble complex.

It has shown that the formation of macromolecular complexes can be detected at pH_c lower than the pH_i of the protein, where both macromolecule species have the same sign of charges. Such a phenomenon had already been reported by Mattison when studying the interaction between poly (dimethyldially1ammonium chloride) (PDMDAAC) and Bovine serum (17). The formation of complexes, under these conditions, is attributed to the attraction between polycation charges and oppositely charged patches on the protein surface (9).

The value of pH = 7.3 represents the phase separation, where the soluble complex formation is completed, and called pH_c’ and pH_φ corresponds to formation of the first appearance of microcoacervate droplets.

In fact, four regions can be defined in Figure 2a These four transition regions (pH_c, pH_c’ and pH_φ) were determined by the intersections of the two tangents. The region 1 corresponds to pH < pH_c where the Coulomb repulsive forces between the positive charges carried by the protein and the polycation, prevent the formation of the complex, resulting macroscopically in a roughly constant turbidity. In region 2 (pH_c < pH < pH_c’) PEI and BSA are ionized and bear opposite charges, marked by the formation of soluble complexes and causing the turbidity to increase but very slightly. In the region 3 (pH_c’ < pH < pH_φ) the turbidity increases rapidly. The last region corresponds to pH > pH_φ, where constant values of turbidity are observed, which is indicative of the phase separation of the system and therefore the presence of coacervates.

This result is in agreement with the study by Kaibara et al. who analyzed the BSA/ poly-(dimethyldiallylammonium chloride) (PDADMAC) mixture and showed the existence of four distinct regions of pH as illustrated by the evolution of turbidity as a function of pH (Figure 2a) (25).

3.2.2 Variation of the hydrodynamic radius of the complex BSA / PEI

Figure 2b shows the variation of apparent hydrodynamic radius R_H of mixture BSA/PEI (25 kDa) versus pH. The apparent hydrodynamic radius variation consists of three branches:

1. At pH values between 3.2 and 4.6, the R_H is constant and roughly equal to 4.2 nm. This value corresponds to the size of the BSA protein alone.

2. At pH values between 4.6 and 7, R_H increases linearly from 4.2 to 5.4 nm. This result means that from the pH = 4.6 the soluble complex begins to establish. In addition, since the molar mass of the added polycation is low, compared to that of the BSA, the size of the primary complex obtained in this pH range is small. We can still estimate that pH ≅ 4.6 is the pH_c and pH_c’≅ 7.

3. At 7< pH < 8.7, the hydrodynamic radius increases rapidly and reaches 50 nm.

4. At pH > 8.7, the solution becomes very turbid and with dynamic light scattering technique, it is difficult to determine with precisely the particle size, due to the multiple scattering effect or strong light absorption.

3.2.3 Effect of salt on complex formation

Like pH, the addition of an amount of salt in a medium is an important factor, which can affect the degree of ionization and therefore the complexation and phase separation of the weak polyelectrolytes.

In order to analyze the salt effect on the formation of complex BSA-polycation, we added a 0.15 M of NaCl to the samples.

We show in Figures 2a and 2b the variation of turbidity and hydrodynamic radius of the system, formed by the mixture of BSA and polycation, with and without salt in the solution.

From these figures, it can be seen that the turbidity and the hydrodynamic radius begin to evolve at pH < pH_i values of BSA for the two systems in the absence and presence of salt. In addition, the turbidity increases in a similar way to that of R_H when the pH increases.

By comparing the results obtained for the variation of the hydrodynamic radius in the presence and absence of the salt as well as those of the variation of the turbidity, it noted that the curves present a small shift. We can conclude that the electrostatic screening has weak effect. Indeed, pH_c, pH_c’ and pH_φ are increased to a value about 0.3 as the NaCl concentration increases. This result is quite contrary to that obtained by Clitor J.F. Souza et al. and Bram L.H.M. Sperber et al. (26,27), even if the shift is the same, where they have studied the effect of ionic strength on a mixture of protein with polyanion. In addition, Andrea Salis et al. showed, in a study on the effect of NaCl on charge of BSA, that the pH_i increases by increasing the ionic strength (see Figure 4b) (28). In another study, Tolstoguzov showed that the main forces responsible for the interactions between proteins and polyelectrolytes are electrostatic in nature, so the ionic strength has a great effect on interactions (29). The concentration of salts also influences the electrostatic interactions and the compatibility of the molecules since the ions can block the reagents sites of the polymers, reducing their rate charge and favoring the association between proteins, occasioning the incompatibility between the two polyelectrolytes (30). Finally, after these interpretations, we can still conclude that polycation plays an inverse role as a polyanion.

3.3 Study of the formation of the complex BSA/PEI (2.5 kDa)

In the previous part, we examined the effect of salt on a system containing BSA-PEI mixture for a molar mass of PEI equal to 2.5 kDa, the ratio r = 5 and T = 25°C. We represent in the following Figure 3, the evolution of turbidity (100-T %) and hydrodynamic radius R_H in the presence and absence of the salt:

Figure 3

Acid base titration at various salt concentrations of mixture of BSA/PEI (2.5 kDa) at Pr:Pl = 5:1. (a) Turbidity of the complex vs pH. (b) Hydrodynamic radius of the complex vs pH. The arrows show the pH transitions. ●, [NaCl] = 0 M; ●, [NaCl] = 0.15 M.

As Figure 2aFigure 3a shows four regions, separated by pH_c, pH_c’ and pH_φ, and which have been still determined by the intersections of the two tangents. In fact, these regions represent the stability and instability of a mixture of PEI/BSA. In this figure, not all pH transition are changed, except in presence of NaCl where the pH gap is enormous (pH_{0 M} – pH_{0.15 M} ≅ 2). If we compare the result with that obtained in Figure 2a this phenomenon may be due to presence of weak molecular mass of PEI (2.5 kDa) and high salt concentration, where the BSA protein loss their attractive forces and is released from PEI. Indeed, the turbidity and the hydrodynamic radius curves of the mixture at 0.15 M of NaCl show that the electrostatic screening is clearly noticeable. As mentioned before, in presence of salt, the turbidity and R_H evolutionsstart at a pH much higher than that without salt. These evolutions are noticeable for pH values higher than pH_i of BSA (4.6), regardless the presence or not of salt. Indeed, at pH_c, a small increase in turbidity was observed, which indicates the start of the formation of soluble complex. At this region, the binding of BSA onto PEI may due to the attraction between positive PEI and negative ’’charge patch’’ of BSA i.e. the change in the pKa of BSA induced by the electric field of the PEI (9). For pH_c < pH < pH_c’, the turbidity and the hydrodynamic radius increase slightly when the pH increases. In this region, the size of the soluble complex is the same compared to that obtained for the molar mass of PEI = 25 kDa. For pH_c’ < pH < pH_φ an abrupt increase in turbidity values and on the other hand, the hydrodynamic radius not be detectable with precision because the solution is very turbid (not transparent). For pH > 8 and pH > 9, the solution becomes turbid and the formation of coacervates obtained for the cases of 0 M and 0.15 M of salt respectively. Moreover, with the standard scattering technique used in our work, the hydrodynamic radius cannot be determined with precision due the problem cited before. In fact, at very high pH and salt concentration, the polycation can behave like an aggregate and this result has been confirmed with that obtained by Kimberly A Curtis et al. (31). Indeed, these authors, in a study of the effect of salt and pH on the change in PEI conformation 2.5 kDa, showed that this polycation passes from a high intra-chain repulsion conformation to low intra-chain repulsion conformation or aggregate when pH and salt increase and the polycation charges increase with decreasing pH, whereas the charge repulsion is screened by increasing the salt concentration.

4 Computational methods

In this part of work, we studied the electrostatic potential of protein (BSA). The electrostatic potential can identify the functional sites at the surface of a protein and can be showed by color-coding of the molecular surface as function of potential values. In reality, this work helps us to know with precision how polyelectrolyte binds with the protein mainly at pHc, where the charges of the protein and of the polyelectrolyte are the same.

Figure 4

Change of folding free energy (a) and total charge Zp (b) of BSA versus pH and at different NaCl salt concentrations (0 and 0.15 M).

In order to analyze the effects of ionic strength, the electrostatic potential was determined at 0 M and 0.15 M NaCl salt concentration and assuming +1 and −1 charges for counter-ions. The protein crystal structures with Protein Data Bank identifications 3V03 (BSA) were taken from the RCSB Protein Data Bank (http://www.rcsb.org)

In fact, according to the PARSE force field, electrostatic potential and van der Waals radius were done with PDB2PQR web portal (http://nbcr-222.ucsd.edu/pdb2pqr_2.1.1/) The probe radius r for ion accessibility and surface dielectric surface were set to be 2.0 Å and 1.4 Å, respectively. The system temperature T was fixed to 298.17 K. In addition, PDB2PQR uses PROPKA (32) to attribute protonation states at desired pH. The linear Poisson–Boltzmann Solver (APBS) through Opal web service, which is inserted in UCSF Chimera software (http://nbcr-222.ucsd.edu/opal2/services/apbs_1.3) was used to calculate the electrostatic potential. All isopotential contours were plotted at ± 1K_BT/e by taking 78.54 and 2 for dielectric constants of the solvent and the protein respectively.

The net charge and folding free energy of protein at different NaCl salt concentrations were estimated from Bluues web server (http://protein.bio.unipd.it/bluues/)

The Bluues version 2.0 generated plots showing (a) the change of folding free energy and (b) total charge Zp with the increase in pH (Figures 4a and 4b) All figures were obtained at different NaCl salt concentrations (0 and 0.15 M). The titration results showed that the folding free energy of the protein, was almost low in the pH range 5.0 –11.0 (Figure 4a) which indicates that the protein stable in this particular range. In a study on the stability of a human calcium/calmodulin-dependent kinase IV protein, Huma Naz et al. found the same result by MD simulations (33). In addition, in this range of pH, no remarkable effects when the salt concentration increased.

An alternative way of addressing the same problem is to calculate the total charges Zp of BSA (Figure 4b) In fact, at the extremes of pH range, this protein presents high density of charges (BSA highly charged), a huge positive charge at acidic pH and huge negative charge at basic pH. With an increasing salt concentration, the calculated values of total charge of the BSA increase at the edges of the pH range, which reflect a higher number of de/protonation of acidic and basic residues at higher and

lower pH values, respectively. Indeed, the pH values close to the point of zero charge (protein isoelectric point (IEP)) are presented in the inset of Figure 4b In presence of salt, the isoelectric point is practically shifted more than half a pH unit towards basic pH. Our results for bovine serum albumin protein total charge in the presence of salt concentration are in good agreement with the papers published by Jachimska and co-workers (34), where these authors found, by zeta potential (ζ) measurement that the IEP of BSA increases with increasing salt NaCl concentration.

Figure 5 shows the anisotropies of the electrostatic potential contour around the Bovine Serum Albumin (3V03) as a function of pH (pH_c, pH_c’ and pH_φ) and salt NaCl concentration. These images portray BSA as “electrostatically seen” by a polycation and clearly reveal the dipolar asymmetry of the presented protein. The values of electrostatic potential are color-coded from red for −1 K_BT/e to blue for +1 K_BT/e on the molecular surface. In fact, these results help us to know, with precisely, the pH and ionic strength dependence of the complexation observed in Figures 2 and 3 and the comparison of the electrostatic potential on the surface of BSA. Indeed, to facilitate the comparison, the protein is presented in the same orientation for all the least cases. In Figure 5, we find that when the pH increases the number of negativelycharged surface residues was increased and at pH = pH_c lower than pI the negatively charged character of a BSA chain in acid pH was shown by little “patches” in red. This demonstrated that BSA interacts with PEI even with positive net charges. Still, we noticed a significant effect of added salt (change of colors at surface of BSA), for example, at pH ≥ 7.3, the appearance of remarkable weak symmetry between positive and negative regions with regard to both shape and magnitude. Indeed, in this region of pH, the protein prefers to bind to the polycation since it has a distinct sign of charges. Moreover, in the case of pH > pI and at high salt concentration (0.15 M), the Debye length (κ ⁻₁ ≈ 0.78 nm) is much less than the protein radius R_protein, indicating that all repulsions are screened. In fact, addition of salt acts to weaken attractions, corresponding to the decrease in affinity (35). Indeed, the change of color, i.e. charge sign, in presence of salt, at the surface of the protein (see Figure 5) explains well the shift founded in Figures 2 and 3. Still, in this region, the combination of short-range attractive interactions coupled with longue-range repulsive interactions is performed only if PEI binds to the negative domain of BSA.

Figure 5

Electrostatic potential contour +1 kT/e (blue) and −1 kT/e (red) around the BSA unimer at different salt concentrations ([NaCl]= 0 M and 0.15 M) and at pH transitions (pH_c, pH_c’ and pH_φ). Calculation was based on PDB ID of Bovine Serum Albumin (3V03 ’’unimer’’).

Based on the analysis performed by Bowman et al., we treated the BSA protein, as a dipole of negative charge Q₋ and positive charge Q₊ with relative distances with polycation of charge q₊, are R₋ and R₊, respectively. At pH > pI, Q₋ > Q₊ and the electrostatic binding of the polycation requires R₋ < R₊ (36). The electrostatic interaction energy can be written as a sum of repulsive and attractive terms:

where ε is the dielectric constant of solvent and κ⁻¹ is the Kuhn length.

The extremum κ_ext corresponding to the maximum in binding energy at a screening length κ−1(κ−1=1κext),can be determined from the first derivative of Eq. 10

Taking the same procedures followed by Dubin (35), where the rate of charges Q₊/Q₋ ranging from 1.2 to 1.5, the estimated screen length κ⁻¹ ranging from Rprotein2toRprotein3.These two values are valid for charges that are not in direct contact with the binding site and located along the skeleton of polycation or far from the surface of BSA.

5 Conclusion

In this paper, we have investigated the interactions of BSA with various molar masses of a linear polycation (PEI) versus pH and salt concentration using a dynamic light scattering, turbidimetry and computer simulations. These simulations were used to estimate the influence of pH and salt on the surface charge distributions and on the electrostatic potential of protein.

No influence of the salt before the formation of the complex (pH < pH_c on the mixture formed of BSA and PEI. Moreover, the molar mass of the polyethelenimine does not influence the hydrodynamic radius and the turbidity. By way of example where the pH ∼ 6 and for the two mixtures (BSA/PEI (2.5 kDa) and BSA/PEI (25 kDa)), the R_H ∼ 6 nm, and the turbidity ∼ 10. At pH = pHc, the Coulombic attraction forces between the concentrations of negative charges (i.e., the patches) of protein and other positively charged macromolecules could locally overwhelm the overall repulsions due to the overall positive charges of the two macromolecules at pH values lower than the pHi of the protein. Hence, the formation of macromolecular complexes is enabled. In addition, from the results mentioned before i.e. for pH > pH_c an increase in the size of the complex formed when the molecular mass of the polycations increases. However, two limiting cases to account for the influence of this parameter. Firstly, if the molecular mass of the polycation is too low, the interaction will be done by simple ion bridging rather than by the formation of intermolecular electrostatic complexes. Conversely, if the molecular weight is very high, the interaction will lead to the formation of precipitates rather than coacervates. Finally, the control of the charge density of the macromolecules provides a mean to regulate the stoichiometry of the macromolecules within the complexes.