- Email: [email protected]
Phosvitin–calcium aggregation and organization at the air–water interface
Available online at www.sciencedirect.com
Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
Phosvitin–calcium aggregation and organization at the air–water interface Corinne Belhomme a , Elisabeth David-Briand a , Catherine Gu´erin-Dubiard b , V´eronique Vi´e c , Marc Anton a,∗ a
UR1268 Biopolym`eres Interactions Assemblages, Equipe Interfaces et Syst`emes Dispers´es, INRA, F-44316 Nantes cedex 3, France b 2 UMR Agrocampus Rennes-INRA, 65 rue de Saint-Brieuc, CS 84215, 35042 Rennes cedex, France c UMR Groupe Mati` ere Condens´ee et Mat´eriaux, Universit´e de Rennes I, Campus Beaulieu, 263 Avenue du G´en´eral Leclerc, 35042 Rennes cedex, France Received 17 July 2007; accepted 30 October 2007 Available online 4 November 2007
Abstract Phosvitin, an egg yolk protein constituted by 50% of phosphorylated serines, presents good emulsifying properties whereas its interfacial properties are not yet clearly elucidated and remain object of discussion. Phosvitin has a high charge density and naturally forms aggregates through phosphocalcic bridges in egg yolk. This high charge density, doubled by this capacity to aggregate, limits the adsorption of the protein at the air–water interface. In this work, we investigated the aggregation impact by calcium ions on the organization of the phosvitin interfacial film using the atomic force microscopy. Phosvitin interfacial films without calcium ions are compared to phosvitin interfacial films formed in the presence of calcium ions in the subphase. We demonstrated that phosvitin is able to anchor at air–water interfaces in spite of its numerous negative charges. In the compression isotherm a transition was observed just before 28 mN/m signifying a possible modification of the interfacial film structure or organization. Calcium ions induce a reorganization towards a greater compaction of the phosvitin interfacial film even at low surface pressure. In conclusion we suggest that, in diluted regime, phosvitin molecules could adsorb by their two hydrophobic extremities exhibiting loops in the aqueous phase, whereas in concentred regime (high interfacial concentration) it would be adsorbed at the interface by only one extremity (brush model). © 2007 Elsevier B.V. All rights reserved. Keywords: Phosvitin; Calcium; Aggregation; Air–water interface; Spread film; Atomic force microscopy
1. Introduction Hen egg yolk phosvitin is a 35 kDa phosphoglycoprotein that represents respectively, 4% and 11% of yolk dry matter and proteins [1,2]. Of the known proteins, it is the most phosphorylated one with 50% of phosphoserines [3,4] inducing an important negative net charge of about −179 [5]. The phosphoserines are arranged in a singular way, forming numerous blocks in the centre of the sequence that can carry up to 15 consecutive residues, whereas the C-terminal and Nterminal extremities are relatively rich in hydrophobic amino acids [6,7]. These characteristics limit the folding of the protein which thus presents an elongated configuration in aqueous solution because of its numerous internal electrostatic repulsions. Consequently thanks to its two hydrophobic extremities
∗
Corresponding author. Tel.: +33 2 40 67 50 80; fax: +33 2 40 67 50 84. E-mail address: [email protected] (M. Anton).
0927-7765/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2007.10.012
and its high charge density in the central part, phosvitin can be designed as a tribloc model: a polyelectrolyte with two well defined apolar extremities. Furthermore, in egg yolk, phosvitin naturally forms calcium aggregates, called granules, through phosphocalcic bridges with an other egg yolk protein [8–12]. Phosvitin is the centre of a controversy concerning its emulsifying properties. It has been demonstrated that this protein has excellent emulsifying properties [13–16], which can be modulated by environmental conditions such as sodium chloride (NaCl) [17]. At the same time few articles showed that, because of its high charge density, phosvitin presents a negligible adsorption capacity at oil–water [17] and air–water interfaces [18]. Even at high concentration (i.e. 0.1 mg/mL), phosvitin exhibited a slow kinetic of adsorption (i.e. 20 h) to reduce significantly the oil–water interfacial tension at equilibrium. Recently, we have investigated this discrepancy taking into account the effect of charge screening by NaCl addition, and aggregation by calcium chloride (CaCl2 ) addition [19]. Our
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
main conclusion were that phosvitin is able to adsorb at the air–water interface in spite of its high negative charge and its small hydrophobic parts. The presence of increasing NaCl concentrations improved interfacial properties of phosvitin and the interfacial film formed was more dense. The better adsorption of the phosvitin is attributed to the charge screening due to NaCl which permits a higher spreading of the protein at the interface. Concerning the aggregation effect, calcium–phosvitin aggregates favour the contact between the protein and the interface and act on the interfacial film morphology inducing the formation of a thicker and a more rigid interfacial phosvitin film. Calcium seems to create a huge network through intermolecular lateral interactions between phosvitin molecules. However, we have characterised this film using Brewster angle microscopy (BAM) that does not allow a fine description of film organization as this technique is not enough resolutive (few m). Furthermore, we have showed, establishing Langmuir balance compression isotherms of phosvitin films with calcium ions, an inflexion point around 28 mN/m revealing a possible transition in the interfacial film structure [19]. Consequently, in order to clarify this possible structure modification of air–water interfacial films in relation with the adsorption mechanism of phosvitin, we have decided to characterise their topography by employing a near field microscopy: Atomic force microscopy (AFM). This technique allows a nanometer lateral resolution and gives an information in 3D where the XYZ coordinates are measurable. So in this study we will try to establish the relationships between topography of interfacial phosvitin film and its compression isotherms at the air–water interface. We have spread phosvitin solution at air–water interfaces with or without calcium in the subphase and then made, at selected film pressures (before and after the inflexion point), Langmuir–Blodgett and Langmuir–Schaeffer transfers of the film onto a mica plate for the characterisation of its topography by AFM. 2. Material and methods 2.1. Materials Eggs were obtained from local wholesale distributor. All chemicals (analytical grade) were purchased from Sigma (Saint Quentin-Fallavier, France), excepted hydrochloric acid (HCl) 37% (pro-analysis) which was purchased from Carlo Erba Reagenti (Val de Reuil, France) and CaCl2 from Merck (Nantes, France). Ultrapure water was obtained from a Millipore system (Millipore, Saint Quentin, France). 2.2. Phosvitin isolation Hen eggs were manually broken, and yolks were carefully freed from adhering white and chalazae by rolling on a filter paper (Whatman). The vitellin membrane was punctured with a lancet and the content was collected in a beaker cooled in iced water. After that, temperature was maintained at 4 ◦ C all through
13
Fig. 1. Spreading of phosvitin monolayers at the air–water interface at pH 7.3. Temperature: 20 ◦ C. The measurements were made with a 5 g/L of phosvitin solution, 600 L were deposed. The measurements were made without calcium (A) or with calcium (B) in the subphase. The transfers of the interfacial film are indicated by arrows.
the process. Granules were extracted from yolk according to the method of McBee and Cotterill [20]. Yolk was diluted with an equal mass of a 0.17 M NaCl solution and mixed with a magnetic stirrer. After 1 h, the solution was centrifuged at 10,000 × g for 45 min in a Jouan centrifuge (model GR 2022, St Herblain, France) and the pellet (granules) was collected and dissolved in a 1.74 M NaCl solution (10% wt/vol%). The mixture was stirred to complete dissolution keeping the pH adjusted to 7.25. The solution was then dialysed against several changes of distilled water for 24 h and centrifuged at 10,000 × g for 30 min. The highdensity lipoproteins precipitated. The supernatant was diluted with a 0.9 M magnesium sulphate solution to obtain a 0.2 M final concentration of this salt. After centrifugation (10,000 × g for 30 min), a precipitate of phosvitin was collected at the bottom of the tubes and was dialysed against distilled water and freeze-dried. 2.3. Phosvitin free-metal preparation In order to purify the phosvitin, the lyophilised protein isolate was dissolved in a 0.05 M Tris–HCl buffer solution (pH 7.3) and then applied on a cation exchange resin. Before the application of the sample on the resin, this one was equilibrated with a 0.05 M Tris–HCl buffer solution (pH 7.3) and then with a 2 M NaCl solution in order to charge the resin with the counter-ion Na+. After its elution, phosvitin was dialysed against ultrapure water for at least 72 h at 4 ◦ C. The final prod-
14
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
uct, phosvitin without metal ions, was obtained after its freezedrying.
effects due to the barrier displacement. At least three isotherms were performed for each sample.
2.4. Surface measurements
2.5. Film transfers
Measurements of the surface pressure (π)–surface area (A) isotherms were performed by the Wilhelmy plate method on an isolated and fully automated Langmuir-type film trough (Nima Technology 601B). The maximum and minimum areas of the trough were 350 × 10−3 and 40 × 10−3 cm2 , respectively. It was filled with the buffer employed for solution preparation and temperature was kept at 20 ◦ C by water circulation from a thermostat Bioblock Ministat (Illkirch, France). An aliquot of 600 L of phosvitin solution (5 mg/mL) was spread on the surface by means of a micrometric syringe. The spreading was made on different buffers containing or not CaCl2 . To allow the spreading of the protein, samples were let to stand 1 h before compression. The compression speed was maintained constant at 40 cm2 /min, which is sufficiently low to prevent secondary
At selected surface pressures, an interfacial film on a Langmuir trough was transferred (Fig. 1), at a speed of 0.1 mm/s, either onto freshly cleaved mica using Langmuir–Blodgett technique, or onto a hydrophobed mica (deposit of dipalmitoylphosphatidylcholine monolayer) using Langmuir–Schaeffer technique. The transferred film was imaged using an AFM system to visualise the structure of the interfacial film. 2.6. Atomic force microscopy We used an AFM Pico-plus (Agilent, Phoenix, US). Images were obtained in contact mode. The film transfers were done from a 350 cm2 Nima 601 Langmuir trough (Nima Technology
Fig. 2. AFM images on hydrophilic mica of phosvitin interfacial film spread on a subphase without calcium (Tris–HCl buffer, 50 mM, pH 7.5). The images show the topography in a contact mode in the air. The transfers were made at 15 mN/m (A) and 28 mN/m (B, C). The size of the images is of 10 m × 10 m (A and B) and are representative of all the surface of the sample, and of 5 m × 5 m (C). The white lines on the images indicate the zone used for the height difference measurements.
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
Ltd., Coventry, UK). The phosvitin films were transferred at pH 7.5, at surface pressures of 15 and 28 mN/m, in the presence of calcium or not. The films were imaged in air or in buffer according to the film transfer method. On AFM images, the height of objects is reflected by the grey level: the darker the image the lower the objects areand inversely.
15
3. Results 3.1. Hydrophilic mica
2.7. Statistical analyses
When we transferred the interfacial film on hydrophilic mica, we compared the phosvitin deposit on a subphase without calcium (Tris–HCl) or with calcium ions. In both cases, the interfacial film was transferred at 15 and 28 mN/m.
The data presented were the mean values of triplicate experiments. Analysis of variance was performed to determine variations among the different conditions.
3.1.1. Phosvitin deposit on a subphase without calcium ions At 15 mN/m (Fig. 2) the phosvitin interfacial film looks globally homogeneous. The film seems constituted of small
Fig. 3. AFM images on hydrophilic mica of phosvitin interfacial film spread on a subphase containing 3 mM of calcium (Tris–HCl buffer, 50 mM, pH 7.5). The images show the topography in a contact mode in the air. The transfers were made at 15 mN/m (A–C). The size of the images is of 10 m × 10 m (A) and are representative of all the surface of the sample, and of 5 m × 5 m (B, C). The white lines on the images indicate the zone used for the height difference measurements.
16
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
Fig. 4. AFM images on hydrophilic mica of phosvitin interfacial film spread on a subphase containing 3 mM of calcium (Tris–HCl buffer, 50 mM, pH 7.5). The images show the topography in a contact mode in the air. The transfers were made at 28 mN/m (A, B). The size of the images is of 20 m × 20 m (A) and are representative of all the surface of the sample, and of 10 m × 10 m (B). The white lines on the images indicate the zone used for the height difference measurements.
individually entities (100 nm) more or less lengthened. These entities could be an arrangement of several phosvitin molecules. These entities are maintained at a constant distance by electrostatic repulsions. Furthermore, we can remark the presence of some less dense zones. The height difference measured between the two zones with different densities (Fig. 2A, 1) is of 4–6 nm. The less dense zones contained several objects with the same height that the most homogeneous zone (4–6 nm) and the forms of these objects look not lengthened but circular. We can also see on the homogeneous zone some large and high isolated objects (Fig. 2A, 2). They have a high difference of 8–10 nm in comparison with the first homogeneous zone. At 28 mN/m (Fig. 2B) the interfacial film looks like that at 15 mN/m. The film is constituted of a homogeneous compact background with some large and less dense zones. The homogeneous zone seems much more compact than that observed at 15 mN/m because we cannot distinguish the individuality of the small entities. The objects constituting this zone are closer from each other but no modification of height was noted as compared with the less dense zones in comparison with that measured at 15 mN/m (4–6 nm). These less dense zones at 28 mN/m (Fig. 2C) present the same organization than that observed at 15 mN/m
with circular objects. Nevertheless the grey levels in this zone at 28 mN/m indicate two height levels of objects whereas only one level was measured at 15 mN/m. The greyest objects are larger and have a height difference of 4–6 nm with the black background. These objects seem to be the same that those constituting the homogeneous and compact zone because they presented the same height variation. The other objects whiter, circular and less large have a height difference of 11–16 nm with the black background. These objects are higher than those constituting the homogeneous zone and appear during the compression of the interfacial film. In spite of the compression we can notice that the average surface of the less dense zones is still similar to that measured at 15 mN/m. If the less dense zones are not covered during the compression, therefore the frontier surrounding this zone should be formed of compact and rigid molecules. 3.1.2. Phosvitin deposit on a subphase with calcium ions At 15 mN/m the phosvitin interfacial film is homogeneous (Fig. 3A) with some black holes by places. Fig. 3B and C show the granular aspect of the interfacial film with few less dense zones. The film looks structured as a real network. The homo-
Fig. 5. AFM images on hydrophobic mica of phosvitin interfacial film spread on a subphase without calcium ions (Tris–HCl buffer, 50 mM, pH 7.5). The images show the topography in a contact mode in liquid phase (Tris–HCl buffer). The transfers were made at 28 mN/m (A, B). The size of the images is of 20 m × 20 m (A) and of 50 m × 50 m (B). The white lines on the images indicate the zone used for the height difference measurements.
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
geneous zone (Fig. 3B, 1) has a difference of height of 20–25 nm with the black background. This homogeneous zone also contains some isolated and thin objects, higher and whiter with a height difference of 50–60 nm (Fig. 3B, 2). The less dense zones (Fig. 3C) show two thickness levels: the first with a dark aspect (Fig. 3C, 1) has a height difference of 4–6 nm with the black background, the second clearer level shows a thickness difference of 8–9 nm with the black background (Fig. 3C, 2). At 28 mN/m the interfacial film differs from that at 15 mN/m. Large domains of several micrometers, linked together, are observable on a more homogeneous background (Fig. 4A). This homogeneous background presents a height difference of about 10 nm with the black background (Fig. 4B, 1). And the large domains present some differences of thickness between 20 and 50 nm as compared to the homogeneous background (Fig. 4B, 2).
17
Between the two surface pressures (15 and 28 mN/m) we observed a modification in the aspect of the phosvitin interfacial film spread on a subphase containing calcium ions. At 15 mN/m we measured four thickness levels whereas only two levels were measured at 28 mN/m. If we add the thicknesses measured at the two surface pressures the addition at 15 mN/m is superior to that at 28 mN/m. The less dense zones at 15 mN/m (4–6 nm and 8–9 nm) seem to not resist to the compression. The first thickness at 4–6 nm could be covered by the thickness of 8–9 nm when compressing. The homogeneous background which results from it presents a difference of thickness of an average of 10 nm. Then the compression of the phosvitin interfacial film spread on a subphase containing calcium ions induces some rearrangements in the interfacial film and/or the sinking of one or several thicknesses in the subphase.
Fig. 6. AFM images on hydrophobic mica of phosvitin interfacial film spread on a subphase containing 3 mM of calcium (Tris–HCl buffer, 50 mM, pH 7.5). The images show the topography in a contact mode in liquid phase (Tris–HCl buffer, 3 mM of calcium). The transfers were made at 15 (A) and 28 mN/m (B, C). The size of the images is of 20 m × 20 m (A, B) and of 5 m × 5 m (C). The white lines on the images indicate the zone used for the height difference measurements.
18
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
3.2. Hydrophobed mica 3.2.1. Phosvitin deposit on a subphase without calcium ions We transferred the phosvitin interfacial film at 28 mN/m. The Fig. 5A shows a homogeneous background with more or less lenghtened repeated motifs and with some hollowed breaking lines. These ones have a difference of height of 10–20 nm with the homogeneous background (Fig. 5A, 1) and the homogeneous background shows a difference of thickness of 25–30 nm (Fig. 5A, 2) with the less dense zones. The latter make holes (Fig. 5A, 3) randomly among the homogeneous background. Other small holes are visible in the less dense zones with a difference of high of 4–6 nm comparing to the larger holes (Fig. 5A, 3). Onto the homogeneous background we observe other isolated objects with an important thickness distributed uniformly (Fig. 5B). These objects show a difference of height of 80–150 nm as compared to the homogeneous background (Fig. 5A, 1). 3.2.2. Phosvitin deposit on a subphase with calcium ions At 15 mN/m the phosvitin interfacial film in the presence of calcium looks homogeneous with some hollowed break-
ing lines uniformly present onto the mica sheet (Fig. 6A). The homogeneous background seems constituted of interlinked units which the difference of thickness is of 4 nm with the darker background (Fig. 6A, 1). Some isolated and thicker objects are randomly present. They have a difference of height of 30 nm with the homogeneous background (Fig. 6A, 2). This organization of the phosvitin interfacial film in the presence of calcium at 15 mN/m looks like that of the phosvitin interfacial film at 28 mN/m with a subphase without calcium (Fig. 5). When the compression goes up to 28 mN/m we observed an intensification of the densification of the phosvitin interfacial film (Fig. 6B). The hollowed breaking lines at 15 mN/m seems to fill at 28 mN/m. The material is organized in network between the breaking lines. The network shows a difference of thickness of 40–80 nm with the darker background (Fig. 6B and C). The full breaking lines are not larger than the organized structures observed between these lines. The network organization also presents some areas unoccupied by the protein. Nevertheless the organization of these breaking lines induces a small difference of thickness of 20–40 nm in comparison with the structured network between the lines (Fig. 6C).
Fig. 7. Hypothesis of phosvitin positioning on mica sheets according to the two theories of hanging on at the air–water interface, 1 and 2: sequential steps of transfers, ( ) hydrophilic part of phosvitin, ( ) hydrophobic part of phosvitin.
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
4. Discussion Whatever the presence or not of calcium ions, the topography of phosvitin interfacial films is different according to whether we observe the hydrophobic (transfer on hydrophilic mica) or the hydrophilic part (transfer on hydrophobed mica) of the phosvitin molecule. Fig. 7 illustrates the possible phosvitin spreading onto the mica sheets according to the two hypotheses of hanging on at the air–water interface by only one or both hydrophobic ends. It is not well established if the phosvitin is anchored at the interface by only one or both hydrophobic ends. If the phosvitin is hung on by both ends it would make a loop in the subphase [21]. This hypothesis results from a modelling of phosvitin adsorption using the Sheutjens–Fleer theory (self-consistency theory) [22]. This model permitted to calculate pH, ionic strength and concentration effects on the interfacial distribution of the different segments of phosvitin. These authors calculated that phosvitin made a loop (residues 50–200) in the aqueous phase whereas the extremities (residues 0–50 and 205–216) were in contact with the air–water interface. The second hypothesis favoured by Damodaran and Xu [5] is that phosvitin is hung on at the interface by only one hydrophobic end and the remainder of the molecule would be unfold in the subphase. For these authors, this could explain the adsorption of phosvitin in spite of the high electrostatic energy barrier due to its polyanionic character. According to Fig. 7 it seems that the interfacial film transfer onto the hydrophobed mica could better reflect its adsorption at the air–water interface. Inversely, on the hydrophilic mica, whatever if the phosvitin hangs on the interface by one or two hydrophobic ends it could probably spread of its entire hydrophilic part. The negative charges of the phosphoserin residues would probably interact by the same manner with the hydrophilic part of the mica. 4.1. Without calcium ions Concerning the transfer on hydrophilic mica of the phosvitin interfacial film, we measured only one height of 4–6 nm. If the hydrophilic part of the phosvitin spreads on the hydrophilic mica as we supposed on, then the objects measured at 4–6 nm should correspond to the phosvitin hydrophobic segments (Fig. 7). The phosvitin N-terminal segment constitutes the 50 first aminoacids of phosvitin primary structure [23]. When considering the 217 aminoacids of phosvitin and the molecule length equal to 28 nm [24], the N-terminal hydrophobic segment should measure a length of 6.3 nm. The high of 4–6 nm measured on the phosvitin interfacial film could correspond to the phosvitin hydrophobic segments. Between the two surface pressures, 15 and 28 mN/m, few modifications in the film topography were observable in spite of the presence of a small inflexion point at 28 mN/m visible on previous Langmuir isotherms (Fig. 1, [19]) suggesting a possible transition of phosvitin interfacial film structure or reorganization. In the two cases a homogeneous zone and a less dense zone are present. The height differences measured are the same excepted for 28 mN/m where a supplementary height is measured. A different roughness and a higher compactness
19
appear when the film is compressed at 28 mN/m. This result could explain the inflexion point observed at 28 mN/m. Then the compression of the interfacial film when the phosvitin is spread on a subphase without calcium induces minor changes in its topography. Concerning the transfer of phosvitin film on hydrophobed mica, we noticed, at 28 mN/m, the presence of a homogeneous background with a difference of height of 25–30 nm. If the phosvitin molecule anchored at the interface by only one hydrophobic extremity, this homogeneous background could correspond to the phosvitin molecules (Fig. 7). Calculating the length of the hydrophilic part it will be equal to an average of 22 nm which is really close to the 25–30 nm that we measured. For technical reasons, we were not able to make the interface transfer onto the hydrophobed mica at 15 mN/m, but we can speculate that at low surface pressure, when the interfacial protein concentration is low, as observed before to compress the surface in Langmuir experiments (diluted regime), phosvitin adsorbs with the two hydrophobic extremities (Fig. 7). In semidiluted regime (compression up to 15 mN/m), phosvitin loops could start to compact but intermolecular electrostatic repulsions could prevent narrow interactions. In concentrated regime like that obtained at 28 mN/m (interfacial protein concentration saturation), compaction forces can probably desorb C-ter hydrophobic extremity (the smallest one) passing from a loop model to a brush model (Fig. 7). The differences of height at 25–30 nm correspond to the phosvitin molecules hang on the air–water interface by only one hydrophobic end. 4.2. With calcium ions Phosvitin interfacial film is considerably more cohesive whatever the hydrophobic and/or the hydrophilic side. We observed large domains interlinked on the hydrophobic side and a real structured network on the hydrophilic side, putting in relief that calcium ions interact with phosvitin at the interface through phosphocalcic bridges which induce a large and structured network. Furthermore the differences of height that we measured underline the aggregation complexity between calcium ions and phosvitin. The high thicknesses measured prove that the aggregation takes place in three dimensions. In the hypothesis of the entire phosvitin spreading onto the hydrophilic mica, the first difference of height (4–6 nm at 15 mN/m) could represent a homogeneous monolayer of phosvitin. The isolated objects with a z2 = 7–8 nm (15 mN/m) (Fig. 2A) could be some aggregated phosvitin molecules which stay at the interface after the deposit. When we spread the phosvitin on a subphase containing calcium ions we measured a number and values of differences of thicknesses higher than when we spread phosvitin on a subphase without calcium. In the presence of calcium the links intra- and inter-molecules of phosvitin can be more abundant according to the hanging on of phosvitin by one and/or two ends (depending of the interfacial concentration). But it is impossible to define the organization and the arrangement of phosvitin molecules among the different thicknesses measured. Nevertheless such thicknesses could reveal a multilayer formation in the presence of calcium ions.
20
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
5. Conclusion This study showed that phosvitin is able to anchor at air–water interfaces in spite of its high negative charges. At low interfacial concentration (diluted regime), phosvitin molecules could probably anchor by their two hydrophobic extremities exhibiting loops in the aqueous phase, whereas at high interfacial concentrations (concentred regime) phosvitin would be anchored at the interface by only one extremity (brush model). Calcium ions induce a reorganization towards a greater compaction of the phosvitin interfacial film even at low surface pressure. Calcium ions interact with phosvitin through intermolecular lateral interactions by making phosphocalcic bridges and then provoke a real aggregation and a structured network in three dimensions. This film reorganization and reinforcement could have a key role on the stability of dispersed systems stabilised with phosvitin. To better understand interfacial phosvitin–calcium interactions, further studies using Fourier transformation infra-red and X-ray diffraction are envisaged to complete this work.
[3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18]
Acknowledgments
[19]
This research was supported by INRA and the R´egion Pays de la Loire.
[20] [21] [22]
References [1] M. Anton, Recent Res. Dev. Agric. Food Chem. 2 (1998) 839. [2] W.D. Powrie, S. Naka¨ı, The chemistry of eggs and egg products, in: W.J. Stadelman, O.J. Cotterill (Eds.), Egg Science and Technology, The Avi Publishing Company Inc, Westport, 1986.
[23] [24]
G. Taborsky, Adv. Inorg. Biochem. 5 (1983) 235. R.C. Clark, J. Biochem. 17 (1985) 983. S. Damodaran, S. Xu, J. Colloid Interface Sci. 178 (1996) 426. B.M. Byrne, A.D. van Het Schip, J.A.M. van de Klundert, A.C. Arnberg, M. Gruber, A.B. Geert, Biochemistry 23 (1984) 4275. N. Mabuchi, J.I. Yamamura, T. Adachi, N. Aoki, R. Nakamura, T. Nakamura, EMBL/GenBank/DDBJ databases (P87498), 1996. R.W. Burley, W.H. Cook, Can. J. Biochem. Physiol. 39 (1961) 1295. M.W. Radomski, W.H. Cook, Can. J. Biochem. 42 (1964) 395. O. Castellani, C. Gu´erin-Dubiard, E. David-Briand, M. Anton, Food Chem. 85 (2004) 569. K. Grizzuti, G.E. Perlmann, Biochemistry 12 (1973) 4399. K. Grizzuti, G.E. Perlmann, Biochemistry 14 (1975) 2171. S.L. Chung, L.K. Ferrier, J. Food Sci. 56 (1991) 1259. S.L. Chung, L.K. Ferrier, J. Food Sci. 57 (1992) 40. S. Damodaran, in: S. Damodaran, A. Paraf (Eds.), Food Proteins and Their Applications, Marcel Dekker, New York, 1997. O. Castellani, E. David-Briand, C. Gu´erin-Dubiard, M. Anton, Food Hydrocolloids 19 (2005) 769. O. Castellani, C. Belhomme, E. David-Briand, C. Gu´erin-Dubiard, M. Anton, Food Hydrocolloids 20 (2006) 35. E. Dickinson, J.A. Hunt, D.G. Dalgleish, Food Hydrocolloids 4 (1991) 403. C. Belhomme, E. David-Briand, M.H. Ropers, C. Gu´erin-Dubiard, M. Anton, Food Hydrocolloids 21 (2007) 896. L.E. McBee, O.J. Cotterill, J. Food Sci. 44 (1979) 657. E. Dickinson, V.J. Pinfield, D.S. Horne, J. Colloid Interface Sci. 187 (1997) 539. G.J. Fleer, M.A. Cohen-Stuart, J.M.H.M. Scheujtens, T. Cosgrove, B. Vincent, in: Chapman and Hall (Eds.), Polymers at Interfaces, London, 1993. F. McRitchie, A.E. Alexander, J. Colloid Sci. 18 (1963) 453. F.J. Joubert, W.H. Cook, Can. J. Biochem. Physiol. 36 (1956) 399.
Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
Phosvitin–calcium aggregation and organization at the air–water interface Corinne Belhomme a , Elisabeth David-Briand a , Catherine Gu´erin-Dubiard b , V´eronique Vi´e c , Marc Anton a,∗ a
UR1268 Biopolym`eres Interactions Assemblages, Equipe Interfaces et Syst`emes Dispers´es, INRA, F-44316 Nantes cedex 3, France b 2 UMR Agrocampus Rennes-INRA, 65 rue de Saint-Brieuc, CS 84215, 35042 Rennes cedex, France c UMR Groupe Mati` ere Condens´ee et Mat´eriaux, Universit´e de Rennes I, Campus Beaulieu, 263 Avenue du G´en´eral Leclerc, 35042 Rennes cedex, France Received 17 July 2007; accepted 30 October 2007 Available online 4 November 2007
Abstract Phosvitin, an egg yolk protein constituted by 50% of phosphorylated serines, presents good emulsifying properties whereas its interfacial properties are not yet clearly elucidated and remain object of discussion. Phosvitin has a high charge density and naturally forms aggregates through phosphocalcic bridges in egg yolk. This high charge density, doubled by this capacity to aggregate, limits the adsorption of the protein at the air–water interface. In this work, we investigated the aggregation impact by calcium ions on the organization of the phosvitin interfacial film using the atomic force microscopy. Phosvitin interfacial films without calcium ions are compared to phosvitin interfacial films formed in the presence of calcium ions in the subphase. We demonstrated that phosvitin is able to anchor at air–water interfaces in spite of its numerous negative charges. In the compression isotherm a transition was observed just before 28 mN/m signifying a possible modification of the interfacial film structure or organization. Calcium ions induce a reorganization towards a greater compaction of the phosvitin interfacial film even at low surface pressure. In conclusion we suggest that, in diluted regime, phosvitin molecules could adsorb by their two hydrophobic extremities exhibiting loops in the aqueous phase, whereas in concentred regime (high interfacial concentration) it would be adsorbed at the interface by only one extremity (brush model). © 2007 Elsevier B.V. All rights reserved. Keywords: Phosvitin; Calcium; Aggregation; Air–water interface; Spread film; Atomic force microscopy
1. Introduction Hen egg yolk phosvitin is a 35 kDa phosphoglycoprotein that represents respectively, 4% and 11% of yolk dry matter and proteins [1,2]. Of the known proteins, it is the most phosphorylated one with 50% of phosphoserines [3,4] inducing an important negative net charge of about −179 [5]. The phosphoserines are arranged in a singular way, forming numerous blocks in the centre of the sequence that can carry up to 15 consecutive residues, whereas the C-terminal and Nterminal extremities are relatively rich in hydrophobic amino acids [6,7]. These characteristics limit the folding of the protein which thus presents an elongated configuration in aqueous solution because of its numerous internal electrostatic repulsions. Consequently thanks to its two hydrophobic extremities
∗
Corresponding author. Tel.: +33 2 40 67 50 80; fax: +33 2 40 67 50 84. E-mail address: [email protected] (M. Anton).
0927-7765/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2007.10.012
and its high charge density in the central part, phosvitin can be designed as a tribloc model: a polyelectrolyte with two well defined apolar extremities. Furthermore, in egg yolk, phosvitin naturally forms calcium aggregates, called granules, through phosphocalcic bridges with an other egg yolk protein [8–12]. Phosvitin is the centre of a controversy concerning its emulsifying properties. It has been demonstrated that this protein has excellent emulsifying properties [13–16], which can be modulated by environmental conditions such as sodium chloride (NaCl) [17]. At the same time few articles showed that, because of its high charge density, phosvitin presents a negligible adsorption capacity at oil–water [17] and air–water interfaces [18]. Even at high concentration (i.e. 0.1 mg/mL), phosvitin exhibited a slow kinetic of adsorption (i.e. 20 h) to reduce significantly the oil–water interfacial tension at equilibrium. Recently, we have investigated this discrepancy taking into account the effect of charge screening by NaCl addition, and aggregation by calcium chloride (CaCl2 ) addition [19]. Our
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
main conclusion were that phosvitin is able to adsorb at the air–water interface in spite of its high negative charge and its small hydrophobic parts. The presence of increasing NaCl concentrations improved interfacial properties of phosvitin and the interfacial film formed was more dense. The better adsorption of the phosvitin is attributed to the charge screening due to NaCl which permits a higher spreading of the protein at the interface. Concerning the aggregation effect, calcium–phosvitin aggregates favour the contact between the protein and the interface and act on the interfacial film morphology inducing the formation of a thicker and a more rigid interfacial phosvitin film. Calcium seems to create a huge network through intermolecular lateral interactions between phosvitin molecules. However, we have characterised this film using Brewster angle microscopy (BAM) that does not allow a fine description of film organization as this technique is not enough resolutive (few m). Furthermore, we have showed, establishing Langmuir balance compression isotherms of phosvitin films with calcium ions, an inflexion point around 28 mN/m revealing a possible transition in the interfacial film structure [19]. Consequently, in order to clarify this possible structure modification of air–water interfacial films in relation with the adsorption mechanism of phosvitin, we have decided to characterise their topography by employing a near field microscopy: Atomic force microscopy (AFM). This technique allows a nanometer lateral resolution and gives an information in 3D where the XYZ coordinates are measurable. So in this study we will try to establish the relationships between topography of interfacial phosvitin film and its compression isotherms at the air–water interface. We have spread phosvitin solution at air–water interfaces with or without calcium in the subphase and then made, at selected film pressures (before and after the inflexion point), Langmuir–Blodgett and Langmuir–Schaeffer transfers of the film onto a mica plate for the characterisation of its topography by AFM. 2. Material and methods 2.1. Materials Eggs were obtained from local wholesale distributor. All chemicals (analytical grade) were purchased from Sigma (Saint Quentin-Fallavier, France), excepted hydrochloric acid (HCl) 37% (pro-analysis) which was purchased from Carlo Erba Reagenti (Val de Reuil, France) and CaCl2 from Merck (Nantes, France). Ultrapure water was obtained from a Millipore system (Millipore, Saint Quentin, France). 2.2. Phosvitin isolation Hen eggs were manually broken, and yolks were carefully freed from adhering white and chalazae by rolling on a filter paper (Whatman). The vitellin membrane was punctured with a lancet and the content was collected in a beaker cooled in iced water. After that, temperature was maintained at 4 ◦ C all through
13
Fig. 1. Spreading of phosvitin monolayers at the air–water interface at pH 7.3. Temperature: 20 ◦ C. The measurements were made with a 5 g/L of phosvitin solution, 600 L were deposed. The measurements were made without calcium (A) or with calcium (B) in the subphase. The transfers of the interfacial film are indicated by arrows.
the process. Granules were extracted from yolk according to the method of McBee and Cotterill [20]. Yolk was diluted with an equal mass of a 0.17 M NaCl solution and mixed with a magnetic stirrer. After 1 h, the solution was centrifuged at 10,000 × g for 45 min in a Jouan centrifuge (model GR 2022, St Herblain, France) and the pellet (granules) was collected and dissolved in a 1.74 M NaCl solution (10% wt/vol%). The mixture was stirred to complete dissolution keeping the pH adjusted to 7.25. The solution was then dialysed against several changes of distilled water for 24 h and centrifuged at 10,000 × g for 30 min. The highdensity lipoproteins precipitated. The supernatant was diluted with a 0.9 M magnesium sulphate solution to obtain a 0.2 M final concentration of this salt. After centrifugation (10,000 × g for 30 min), a precipitate of phosvitin was collected at the bottom of the tubes and was dialysed against distilled water and freeze-dried. 2.3. Phosvitin free-metal preparation In order to purify the phosvitin, the lyophilised protein isolate was dissolved in a 0.05 M Tris–HCl buffer solution (pH 7.3) and then applied on a cation exchange resin. Before the application of the sample on the resin, this one was equilibrated with a 0.05 M Tris–HCl buffer solution (pH 7.3) and then with a 2 M NaCl solution in order to charge the resin with the counter-ion Na+. After its elution, phosvitin was dialysed against ultrapure water for at least 72 h at 4 ◦ C. The final prod-
14
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
uct, phosvitin without metal ions, was obtained after its freezedrying.
effects due to the barrier displacement. At least three isotherms were performed for each sample.
2.4. Surface measurements
2.5. Film transfers
Measurements of the surface pressure (π)–surface area (A) isotherms were performed by the Wilhelmy plate method on an isolated and fully automated Langmuir-type film trough (Nima Technology 601B). The maximum and minimum areas of the trough were 350 × 10−3 and 40 × 10−3 cm2 , respectively. It was filled with the buffer employed for solution preparation and temperature was kept at 20 ◦ C by water circulation from a thermostat Bioblock Ministat (Illkirch, France). An aliquot of 600 L of phosvitin solution (5 mg/mL) was spread on the surface by means of a micrometric syringe. The spreading was made on different buffers containing or not CaCl2 . To allow the spreading of the protein, samples were let to stand 1 h before compression. The compression speed was maintained constant at 40 cm2 /min, which is sufficiently low to prevent secondary
At selected surface pressures, an interfacial film on a Langmuir trough was transferred (Fig. 1), at a speed of 0.1 mm/s, either onto freshly cleaved mica using Langmuir–Blodgett technique, or onto a hydrophobed mica (deposit of dipalmitoylphosphatidylcholine monolayer) using Langmuir–Schaeffer technique. The transferred film was imaged using an AFM system to visualise the structure of the interfacial film. 2.6. Atomic force microscopy We used an AFM Pico-plus (Agilent, Phoenix, US). Images were obtained in contact mode. The film transfers were done from a 350 cm2 Nima 601 Langmuir trough (Nima Technology
Fig. 2. AFM images on hydrophilic mica of phosvitin interfacial film spread on a subphase without calcium (Tris–HCl buffer, 50 mM, pH 7.5). The images show the topography in a contact mode in the air. The transfers were made at 15 mN/m (A) and 28 mN/m (B, C). The size of the images is of 10 m × 10 m (A and B) and are representative of all the surface of the sample, and of 5 m × 5 m (C). The white lines on the images indicate the zone used for the height difference measurements.
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
Ltd., Coventry, UK). The phosvitin films were transferred at pH 7.5, at surface pressures of 15 and 28 mN/m, in the presence of calcium or not. The films were imaged in air or in buffer according to the film transfer method. On AFM images, the height of objects is reflected by the grey level: the darker the image the lower the objects areand inversely.
15
3. Results 3.1. Hydrophilic mica
2.7. Statistical analyses
When we transferred the interfacial film on hydrophilic mica, we compared the phosvitin deposit on a subphase without calcium (Tris–HCl) or with calcium ions. In both cases, the interfacial film was transferred at 15 and 28 mN/m.
The data presented were the mean values of triplicate experiments. Analysis of variance was performed to determine variations among the different conditions.
3.1.1. Phosvitin deposit on a subphase without calcium ions At 15 mN/m (Fig. 2) the phosvitin interfacial film looks globally homogeneous. The film seems constituted of small
Fig. 3. AFM images on hydrophilic mica of phosvitin interfacial film spread on a subphase containing 3 mM of calcium (Tris–HCl buffer, 50 mM, pH 7.5). The images show the topography in a contact mode in the air. The transfers were made at 15 mN/m (A–C). The size of the images is of 10 m × 10 m (A) and are representative of all the surface of the sample, and of 5 m × 5 m (B, C). The white lines on the images indicate the zone used for the height difference measurements.
16
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
Fig. 4. AFM images on hydrophilic mica of phosvitin interfacial film spread on a subphase containing 3 mM of calcium (Tris–HCl buffer, 50 mM, pH 7.5). The images show the topography in a contact mode in the air. The transfers were made at 28 mN/m (A, B). The size of the images is of 20 m × 20 m (A) and are representative of all the surface of the sample, and of 10 m × 10 m (B). The white lines on the images indicate the zone used for the height difference measurements.
individually entities (100 nm) more or less lengthened. These entities could be an arrangement of several phosvitin molecules. These entities are maintained at a constant distance by electrostatic repulsions. Furthermore, we can remark the presence of some less dense zones. The height difference measured between the two zones with different densities (Fig. 2A, 1) is of 4–6 nm. The less dense zones contained several objects with the same height that the most homogeneous zone (4–6 nm) and the forms of these objects look not lengthened but circular. We can also see on the homogeneous zone some large and high isolated objects (Fig. 2A, 2). They have a high difference of 8–10 nm in comparison with the first homogeneous zone. At 28 mN/m (Fig. 2B) the interfacial film looks like that at 15 mN/m. The film is constituted of a homogeneous compact background with some large and less dense zones. The homogeneous zone seems much more compact than that observed at 15 mN/m because we cannot distinguish the individuality of the small entities. The objects constituting this zone are closer from each other but no modification of height was noted as compared with the less dense zones in comparison with that measured at 15 mN/m (4–6 nm). These less dense zones at 28 mN/m (Fig. 2C) present the same organization than that observed at 15 mN/m
with circular objects. Nevertheless the grey levels in this zone at 28 mN/m indicate two height levels of objects whereas only one level was measured at 15 mN/m. The greyest objects are larger and have a height difference of 4–6 nm with the black background. These objects seem to be the same that those constituting the homogeneous and compact zone because they presented the same height variation. The other objects whiter, circular and less large have a height difference of 11–16 nm with the black background. These objects are higher than those constituting the homogeneous zone and appear during the compression of the interfacial film. In spite of the compression we can notice that the average surface of the less dense zones is still similar to that measured at 15 mN/m. If the less dense zones are not covered during the compression, therefore the frontier surrounding this zone should be formed of compact and rigid molecules. 3.1.2. Phosvitin deposit on a subphase with calcium ions At 15 mN/m the phosvitin interfacial film is homogeneous (Fig. 3A) with some black holes by places. Fig. 3B and C show the granular aspect of the interfacial film with few less dense zones. The film looks structured as a real network. The homo-
Fig. 5. AFM images on hydrophobic mica of phosvitin interfacial film spread on a subphase without calcium ions (Tris–HCl buffer, 50 mM, pH 7.5). The images show the topography in a contact mode in liquid phase (Tris–HCl buffer). The transfers were made at 28 mN/m (A, B). The size of the images is of 20 m × 20 m (A) and of 50 m × 50 m (B). The white lines on the images indicate the zone used for the height difference measurements.
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
geneous zone (Fig. 3B, 1) has a difference of height of 20–25 nm with the black background. This homogeneous zone also contains some isolated and thin objects, higher and whiter with a height difference of 50–60 nm (Fig. 3B, 2). The less dense zones (Fig. 3C) show two thickness levels: the first with a dark aspect (Fig. 3C, 1) has a height difference of 4–6 nm with the black background, the second clearer level shows a thickness difference of 8–9 nm with the black background (Fig. 3C, 2). At 28 mN/m the interfacial film differs from that at 15 mN/m. Large domains of several micrometers, linked together, are observable on a more homogeneous background (Fig. 4A). This homogeneous background presents a height difference of about 10 nm with the black background (Fig. 4B, 1). And the large domains present some differences of thickness between 20 and 50 nm as compared to the homogeneous background (Fig. 4B, 2).
17
Between the two surface pressures (15 and 28 mN/m) we observed a modification in the aspect of the phosvitin interfacial film spread on a subphase containing calcium ions. At 15 mN/m we measured four thickness levels whereas only two levels were measured at 28 mN/m. If we add the thicknesses measured at the two surface pressures the addition at 15 mN/m is superior to that at 28 mN/m. The less dense zones at 15 mN/m (4–6 nm and 8–9 nm) seem to not resist to the compression. The first thickness at 4–6 nm could be covered by the thickness of 8–9 nm when compressing. The homogeneous background which results from it presents a difference of thickness of an average of 10 nm. Then the compression of the phosvitin interfacial film spread on a subphase containing calcium ions induces some rearrangements in the interfacial film and/or the sinking of one or several thicknesses in the subphase.
Fig. 6. AFM images on hydrophobic mica of phosvitin interfacial film spread on a subphase containing 3 mM of calcium (Tris–HCl buffer, 50 mM, pH 7.5). The images show the topography in a contact mode in liquid phase (Tris–HCl buffer, 3 mM of calcium). The transfers were made at 15 (A) and 28 mN/m (B, C). The size of the images is of 20 m × 20 m (A, B) and of 5 m × 5 m (C). The white lines on the images indicate the zone used for the height difference measurements.
18
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
3.2. Hydrophobed mica 3.2.1. Phosvitin deposit on a subphase without calcium ions We transferred the phosvitin interfacial film at 28 mN/m. The Fig. 5A shows a homogeneous background with more or less lenghtened repeated motifs and with some hollowed breaking lines. These ones have a difference of height of 10–20 nm with the homogeneous background (Fig. 5A, 1) and the homogeneous background shows a difference of thickness of 25–30 nm (Fig. 5A, 2) with the less dense zones. The latter make holes (Fig. 5A, 3) randomly among the homogeneous background. Other small holes are visible in the less dense zones with a difference of high of 4–6 nm comparing to the larger holes (Fig. 5A, 3). Onto the homogeneous background we observe other isolated objects with an important thickness distributed uniformly (Fig. 5B). These objects show a difference of height of 80–150 nm as compared to the homogeneous background (Fig. 5A, 1). 3.2.2. Phosvitin deposit on a subphase with calcium ions At 15 mN/m the phosvitin interfacial film in the presence of calcium looks homogeneous with some hollowed break-
ing lines uniformly present onto the mica sheet (Fig. 6A). The homogeneous background seems constituted of interlinked units which the difference of thickness is of 4 nm with the darker background (Fig. 6A, 1). Some isolated and thicker objects are randomly present. They have a difference of height of 30 nm with the homogeneous background (Fig. 6A, 2). This organization of the phosvitin interfacial film in the presence of calcium at 15 mN/m looks like that of the phosvitin interfacial film at 28 mN/m with a subphase without calcium (Fig. 5). When the compression goes up to 28 mN/m we observed an intensification of the densification of the phosvitin interfacial film (Fig. 6B). The hollowed breaking lines at 15 mN/m seems to fill at 28 mN/m. The material is organized in network between the breaking lines. The network shows a difference of thickness of 40–80 nm with the darker background (Fig. 6B and C). The full breaking lines are not larger than the organized structures observed between these lines. The network organization also presents some areas unoccupied by the protein. Nevertheless the organization of these breaking lines induces a small difference of thickness of 20–40 nm in comparison with the structured network between the lines (Fig. 6C).
Fig. 7. Hypothesis of phosvitin positioning on mica sheets according to the two theories of hanging on at the air–water interface, 1 and 2: sequential steps of transfers, ( ) hydrophilic part of phosvitin, ( ) hydrophobic part of phosvitin.
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
4. Discussion Whatever the presence or not of calcium ions, the topography of phosvitin interfacial films is different according to whether we observe the hydrophobic (transfer on hydrophilic mica) or the hydrophilic part (transfer on hydrophobed mica) of the phosvitin molecule. Fig. 7 illustrates the possible phosvitin spreading onto the mica sheets according to the two hypotheses of hanging on at the air–water interface by only one or both hydrophobic ends. It is not well established if the phosvitin is anchored at the interface by only one or both hydrophobic ends. If the phosvitin is hung on by both ends it would make a loop in the subphase [21]. This hypothesis results from a modelling of phosvitin adsorption using the Sheutjens–Fleer theory (self-consistency theory) [22]. This model permitted to calculate pH, ionic strength and concentration effects on the interfacial distribution of the different segments of phosvitin. These authors calculated that phosvitin made a loop (residues 50–200) in the aqueous phase whereas the extremities (residues 0–50 and 205–216) were in contact with the air–water interface. The second hypothesis favoured by Damodaran and Xu [5] is that phosvitin is hung on at the interface by only one hydrophobic end and the remainder of the molecule would be unfold in the subphase. For these authors, this could explain the adsorption of phosvitin in spite of the high electrostatic energy barrier due to its polyanionic character. According to Fig. 7 it seems that the interfacial film transfer onto the hydrophobed mica could better reflect its adsorption at the air–water interface. Inversely, on the hydrophilic mica, whatever if the phosvitin hangs on the interface by one or two hydrophobic ends it could probably spread of its entire hydrophilic part. The negative charges of the phosphoserin residues would probably interact by the same manner with the hydrophilic part of the mica. 4.1. Without calcium ions Concerning the transfer on hydrophilic mica of the phosvitin interfacial film, we measured only one height of 4–6 nm. If the hydrophilic part of the phosvitin spreads on the hydrophilic mica as we supposed on, then the objects measured at 4–6 nm should correspond to the phosvitin hydrophobic segments (Fig. 7). The phosvitin N-terminal segment constitutes the 50 first aminoacids of phosvitin primary structure [23]. When considering the 217 aminoacids of phosvitin and the molecule length equal to 28 nm [24], the N-terminal hydrophobic segment should measure a length of 6.3 nm. The high of 4–6 nm measured on the phosvitin interfacial film could correspond to the phosvitin hydrophobic segments. Between the two surface pressures, 15 and 28 mN/m, few modifications in the film topography were observable in spite of the presence of a small inflexion point at 28 mN/m visible on previous Langmuir isotherms (Fig. 1, [19]) suggesting a possible transition of phosvitin interfacial film structure or reorganization. In the two cases a homogeneous zone and a less dense zone are present. The height differences measured are the same excepted for 28 mN/m where a supplementary height is measured. A different roughness and a higher compactness
19
appear when the film is compressed at 28 mN/m. This result could explain the inflexion point observed at 28 mN/m. Then the compression of the interfacial film when the phosvitin is spread on a subphase without calcium induces minor changes in its topography. Concerning the transfer of phosvitin film on hydrophobed mica, we noticed, at 28 mN/m, the presence of a homogeneous background with a difference of height of 25–30 nm. If the phosvitin molecule anchored at the interface by only one hydrophobic extremity, this homogeneous background could correspond to the phosvitin molecules (Fig. 7). Calculating the length of the hydrophilic part it will be equal to an average of 22 nm which is really close to the 25–30 nm that we measured. For technical reasons, we were not able to make the interface transfer onto the hydrophobed mica at 15 mN/m, but we can speculate that at low surface pressure, when the interfacial protein concentration is low, as observed before to compress the surface in Langmuir experiments (diluted regime), phosvitin adsorbs with the two hydrophobic extremities (Fig. 7). In semidiluted regime (compression up to 15 mN/m), phosvitin loops could start to compact but intermolecular electrostatic repulsions could prevent narrow interactions. In concentrated regime like that obtained at 28 mN/m (interfacial protein concentration saturation), compaction forces can probably desorb C-ter hydrophobic extremity (the smallest one) passing from a loop model to a brush model (Fig. 7). The differences of height at 25–30 nm correspond to the phosvitin molecules hang on the air–water interface by only one hydrophobic end. 4.2. With calcium ions Phosvitin interfacial film is considerably more cohesive whatever the hydrophobic and/or the hydrophilic side. We observed large domains interlinked on the hydrophobic side and a real structured network on the hydrophilic side, putting in relief that calcium ions interact with phosvitin at the interface through phosphocalcic bridges which induce a large and structured network. Furthermore the differences of height that we measured underline the aggregation complexity between calcium ions and phosvitin. The high thicknesses measured prove that the aggregation takes place in three dimensions. In the hypothesis of the entire phosvitin spreading onto the hydrophilic mica, the first difference of height (4–6 nm at 15 mN/m) could represent a homogeneous monolayer of phosvitin. The isolated objects with a z2 = 7–8 nm (15 mN/m) (Fig. 2A) could be some aggregated phosvitin molecules which stay at the interface after the deposit. When we spread the phosvitin on a subphase containing calcium ions we measured a number and values of differences of thicknesses higher than when we spread phosvitin on a subphase without calcium. In the presence of calcium the links intra- and inter-molecules of phosvitin can be more abundant according to the hanging on of phosvitin by one and/or two ends (depending of the interfacial concentration). But it is impossible to define the organization and the arrangement of phosvitin molecules among the different thicknesses measured. Nevertheless such thicknesses could reveal a multilayer formation in the presence of calcium ions.
20
C. Belhomme et al. / Colloids and Surfaces B: Biointerfaces 63 (2008) 12–20
5. Conclusion This study showed that phosvitin is able to anchor at air–water interfaces in spite of its high negative charges. At low interfacial concentration (diluted regime), phosvitin molecules could probably anchor by their two hydrophobic extremities exhibiting loops in the aqueous phase, whereas at high interfacial concentrations (concentred regime) phosvitin would be anchored at the interface by only one extremity (brush model). Calcium ions induce a reorganization towards a greater compaction of the phosvitin interfacial film even at low surface pressure. Calcium ions interact with phosvitin through intermolecular lateral interactions by making phosphocalcic bridges and then provoke a real aggregation and a structured network in three dimensions. This film reorganization and reinforcement could have a key role on the stability of dispersed systems stabilised with phosvitin. To better understand interfacial phosvitin–calcium interactions, further studies using Fourier transformation infra-red and X-ray diffraction are envisaged to complete this work.
[3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18]
Acknowledgments
[19]
This research was supported by INRA and the R´egion Pays de la Loire.
[20] [21] [22]
References [1] M. Anton, Recent Res. Dev. Agric. Food Chem. 2 (1998) 839. [2] W.D. Powrie, S. Naka¨ı, The chemistry of eggs and egg products, in: W.J. Stadelman, O.J. Cotterill (Eds.), Egg Science and Technology, The Avi Publishing Company Inc, Westport, 1986.
[23] [24]
G. Taborsky, Adv. Inorg. Biochem. 5 (1983) 235. R.C. Clark, J. Biochem. 17 (1985) 983. S. Damodaran, S. Xu, J. Colloid Interface Sci. 178 (1996) 426. B.M. Byrne, A.D. van Het Schip, J.A.M. van de Klundert, A.C. Arnberg, M. Gruber, A.B. Geert, Biochemistry 23 (1984) 4275. N. Mabuchi, J.I. Yamamura, T. Adachi, N. Aoki, R. Nakamura, T. Nakamura, EMBL/GenBank/DDBJ databases (P87498), 1996. R.W. Burley, W.H. Cook, Can. J. Biochem. Physiol. 39 (1961) 1295. M.W. Radomski, W.H. Cook, Can. J. Biochem. 42 (1964) 395. O. Castellani, C. Gu´erin-Dubiard, E. David-Briand, M. Anton, Food Chem. 85 (2004) 569. K. Grizzuti, G.E. Perlmann, Biochemistry 12 (1973) 4399. K. Grizzuti, G.E. Perlmann, Biochemistry 14 (1975) 2171. S.L. Chung, L.K. Ferrier, J. Food Sci. 56 (1991) 1259. S.L. Chung, L.K. Ferrier, J. Food Sci. 57 (1992) 40. S. Damodaran, in: S. Damodaran, A. Paraf (Eds.), Food Proteins and Their Applications, Marcel Dekker, New York, 1997. O. Castellani, E. David-Briand, C. Gu´erin-Dubiard, M. Anton, Food Hydrocolloids 19 (2005) 769. O. Castellani, C. Belhomme, E. David-Briand, C. Gu´erin-Dubiard, M. Anton, Food Hydrocolloids 20 (2006) 35. E. Dickinson, J.A. Hunt, D.G. Dalgleish, Food Hydrocolloids 4 (1991) 403. C. Belhomme, E. David-Briand, M.H. Ropers, C. Gu´erin-Dubiard, M. Anton, Food Hydrocolloids 21 (2007) 896. L.E. McBee, O.J. Cotterill, J. Food Sci. 44 (1979) 657. E. Dickinson, V.J. Pinfield, D.S. Horne, J. Colloid Interface Sci. 187 (1997) 539. G.J. Fleer, M.A. Cohen-Stuart, J.M.H.M. Scheujtens, T. Cosgrove, B. Vincent, in: Chapman and Hall (Eds.), Polymers at Interfaces, London, 1993. F. McRitchie, A.E. Alexander, J. Colloid Sci. 18 (1963) 453. F.J. Joubert, W.H. Cook, Can. J. Biochem. Physiol. 36 (1956) 399.