4162
Langmuir 2004, 20, 4162-4167
Interaction of Bovine Serum Albumin and Human Blood Plasma with PEO-Tethered Surfaces: Influence of PEO Chain Length, Grafting Density, and Temperature Willem Norde*,†,‡ and Dick Gage† Laboratory of Physical Chemistry and Colloid Science, Wageningen University, P.O. Box 8038, 6700 EK Wageningen, The Netherlands, and Department of Biomedical Engineering, University of Groningen, P.O. Box 196, 9700 AD Groningen, The Netherlands Received November 17, 2003. In Final Form: February 17, 2004 Solid surfaces are modified by grafting poly(ethylene oxide), PEO, to influence their interaction with indwelling particles, in particular molecules of bovine serum albumin and human plasma proteins. As a rule, the grafted PEO layers suppress protein adsorption. The suppression is most effective when the PEO layer is in a molecular brush conformation having a reciprocal grafting density (area per grafted PEO chain) less than the dimensions of the protein molecules. Nevertheless, the protein molecules may penetrate the PEO brush to some extent. For a given grafting density, the penetration is facilitated by increasing thickness of the brush. Tenuous brushes of reciprocal grafting densities exceeding the protein molecular dimensions enhance protein adsorption. The results point to a weak attractive interaction between PEO and protein. The protein repellency of a densely PEO-brushed surface is ascribed to a high activation energy for the protein molecules to enter the brush. Varying the temperature between 22 and 38 °C does not significantly affect the range of grafting density over which the brush changes from protein-attractive to protein-repellent.
Introduction Water-soluble polymers grafted on a surface render that surface less accessible for deposition of particles and (macro)molecules from the adjoining aqueous phase. Such surface modification is relevant for various purposes, especially biotechnological and biomedical applications. For instance, tethered polymer molecules on surfaces may strongly retard and suppress biofouling, that is, the unintentional formation of a layer of organic materials on surfaces.1 Biofouling usually starts with the adsorption of proteins, and this, in turn, triggers the deposition of biological cells, bacteria, or other micro-organisms to form a so-called biofilm. A biofilm may cause serious microbial contamination problems, for example, in the food and pharmaceutical industries and also in water purification and the production of potable water. Moreover, biofilms on various processing units may retard transport of heat and matter. Biofilms on ship hulls promote the deposition of algae and shells. This may enhance corrosion and perturbation of the streamlining of the vessel. Similarly, biofouling may occur when body fluids come into contact with synthetic materials that are used for artificial body implants, extracorporal assist devices, and medical utensils. Another application of grafting polymers to surfaces is to protect liposomes against phagocytosis.2 This is crucial when liposomes are used as carriers for therapeutic agents in drug targeting and drug delivery systems. Poly(ethylene oxide) (PEO), alternatively called poly(ethylene glycol) (PEG), is most commonly used in the before-mentioned applications. This is, in the first place, because in an aqueous environment PEO molecules are * Corresponding author. † Wageningen University. ‡ University of Groningen. (1) Leckband, D.; Sheth, S.; Halperin, A. J. Biomater. Sci., Polym. Ed. 1999, 10, 1125. (2) Allen, T. M. Adv. Drug Delivery Rev. 1994, 13, 285.
highly mobile3 and strongly hydrated, attaining extremely large exclusion values.4 In addition, PEO is compatible with living cells5 and it can therefore be used in tissue engineering and artificial organs. Many papers are published in which the effects of grafted PEO layers on protein adsorption and cell adhesion are discussed. A survey of this literature is given in, for example, ref 6. It appears that experimental results on protein adsorption and cell adhesion on PEO-coated surfaces are not always unambiguous, if not controversial. This is, at least partly, due to the various ways the PEO layer has been applied, that is, by physical adsorption of diblock, triblock, or comblike polymers or by different ways of chemical grafting. In most reports, the grafting density and, consequently, the thickness of the tethered PEO layer are only poorly controlled. It is therefore not always clear whether the polymer layer is present in a mushroom or in a brush conformation. See Figure 1. In relatively few publications,7-10 the PEO density in, and the thickness of, the tethered layer are well-defined. As expected, deposition of indwelling particles (including protein molecules) at the sorbent surface decreases with increasing grafting density of the polymer. Also, as a rule, particle repellency increases with increasing length of the (3) Nagaoka, S.; Mori, Y.; Takyuchi, H.; Yokota, K.; Tanzawa, H.; Nishiumi, S. In Polymers as Biomaterials; Shalaby, S. W., Hoffman, A. S., Ratner, B. D., Horbett, T. A., Eds.; Plenum Press: New York, 1985; p 361. (4) Ryle, A. P. Nature 1965, 206, 1256. (5) Albertsson, P.-Å. Partition of Cell Particles and Macromolecules, 3rd ed.; Wiley: New York, 1986. (6) Currie, E. P. K.; Norde, W.; Cohen Stuart, M. A. Adv. Colloid Interface Sci. 2003, 100-102, 205-265. (7) Schroe¨n, C. G. P. H.; Cohen Stuart, M. A.; van der Voort Maarschalk, K.; van der Padt, A.; van’t Riet, K. Langmuir 1995, 11, 3068. (8) Sofia, S. J.; Premnath, V.; Merrill, E. W. Macromolecules 1998, 31, 5059. (9) Currie, E. P. K.; van der Gucht, J.; Borisov, O. V.; Cohen Stuart, M. A. Pure Appl. Chem. 1999, 71, 1227. (10) Efremova, N. E.; Bondurant, B.; O’Brien, D. F.; Leckband, D. E. Biochemistry 2000, 39, 3441.
10.1021/la030417t CCC: $27.50 © 2004 American Chemical Society Published on Web 04/09/2004
Interaction of Proteins with PEO-Tethered Surfaces
Figure 1. Conformational states of tethered polymer chains: mushroom (left) and brush (right).
Figure 2. Interactions between a particle and a brush-coated surface: (a) short-range particle-surface contact, (b) long-range Lifshits-van der Waals attraction, and (c) repulsive osmotic interaction. The solid line represents the overall interaction (a + b + c).
grafted polymer chains. Moreover, it appears that the protein and particle resistance increases sharply when the PEO grafting density reaches a value corresponding to the onset of brush formation, that is, when the PEO chains are forced to stretch out from the surface, as shown in Figure 1. In a brush conformation, the separation d between neighboring grafting points is less than the characteristic dimension, for instance, the end-to-end distance 〈r2〉0.5, of the coiled PEO molecule dissolved in water. Brushes offer a unique possibility of modifying the conditions for particle deposition with respect to both equilibrium and dynamics. The fundamental features may be predicted by generic models such as those proposed by Halperin1,11 and by Szleifer.12-14 In Halperin’s model, the interaction between a particle (which has no affinity for the polymer chains) and a brush-coated surface includes three contributions: (a) short-range particle-surface contact (e.g., hydrogen bonding, ion pairing, hydrophobic interaction), (b) (long-range) Lifshits-van der Waals attraction between the particle and the sorbent, and (c) repulsive osmotic interaction between the particle and the brush. The resulting profile for a particle at a brushed surface is qualitatively indicated in Figure 2. The model predicts two possible adsorption modes: primary adsorption at the grafting surface and secondary adsorption at the outer periphery of the brush. Primary adsorption represents the equilibrium state, whereas at the secondary (11) Halperin, A. Langmuir 1999, 15, 2525. (12) Szleifer, I. Biophys. J. 1997, 72, 595. (13) McPherson, T.; Kidane, A.; Szleifer, I.; Park, K. Langmuir 1998, 14, 176. (14) Carigano, M. A.; Szleifer, I. Colloids Surf., B 2000, 18, 169.
Langmuir, Vol. 20, No. 10, 2004 4163
minimum the particles are kinetically trapped but will eventually, by an activated process, reach the primary minimum. Hence, to repress particle deposition altogether it is required to suppress both the primary and the secondary minimum. The depth of the primary minimum depends strongly on the grafting density of the brush. If the grafting density is high, a dense compressed layer of polymer chains will prevent the particle making intimate contact with the surface. Thus, for a given particle size primary adsorption is much less probable for higher brush densities. Furthermore, primary adsorption can be kinetically repressed by increasing the osmotic repulsion, which, in turn, for a given particle is also mainly determined by the grafting density. Repression of secondary adsorption can only be achieved in a thermodynamic way, that is, by sufficiently reducing the depth of the secondary minimum. This is, for a given particle size, realized by increasing the thickness of the brush which is determined by the length of the polymer chains and the grafting density. According to Szleifer’s theory, the polymer density profile in the brush is described by self-consistent field models. The interaction between an incoming particle and a brushed sorbent results from the influence of the polymer chains on the attraction between the particle and the sorbent. Because of a tradeoff between particle resistance and particle attraction, that both increase with increasing particle dimensions, it could well be that the optimal particle-resisting brush density is not very sensitive to the particle size. It must be realized that each model is an incomplete description of a more complex reality. For instance, the surface of the particles is often heterogeneous, as is the case for protein molecules and most biological cells. Also, the conformation of the polymer chains in the brush and, consequently, their interaction with the solvent and the particles may depend on environmental conditions such as temperature and pressure.15,16 Furthermore, in an aqueous environment the asymmetry of the water molecules leads to orientation-dependent interactions with the surface, the polymer chains, and the particles. All these complications and subtleties are not accounted for in the theories mentioned above. This paper reports on the interaction between proteins and surfaces grafted with PEO chains. The proteins are supplied from a solution containing one type of protein, that is, bovine serum albumin (BSA), or from whole (diluted) human blood plasma, containing a variety of proteins. The PEO chains are anchored in different ways, using either diblock or triblock copolymers. The length of the PEO chains, the grafting density, and the temperature are taken as the main experimental variables. Because the grafted layers are well characterized, the experimental data allow unambiguous interpretation and, hence, contribute to our understanding of how proteins interact with PEO-tethered interfaces. Materials and Methods Silicon (Si) wafers were bought from Wacker Chemitronic GmbH, Munich, Germany. Triblock copolymers of the type poly(ethylene oxide)-polypropylene oxide-poly(ethylene oxide), (PEO)n-(PPO)m-(PEO)n, were donated by ICI, Rotterdam, The Netherlands. The indices n and m indicate the degrees of polymerization. The (PEO)n parts are well soluble in water, whereas the (PPO)m part is water-insoluble. Vinyl-terminated polystyrene, vinyl-(PS)200, and the diblock copolymers of the type (PS)38-(PEO)n were obtained from Polymer Source, Inc., (15) Goldstein, R. E. J. Chem. Phys 1984, 80, 5340. (16) Halperin, A. Eur. Phys. J. 1998, 3, 359.
4164
Langmuir, Vol. 20, No. 10, 2004
Dorval, Quebec, Canada. Dichloro-dimethylsilane (DDS) was from BDH Chemicals Ltd., Poole, U.K. BSA was a product from Sigma Chemical Co., catalog number A-7030, and human blood plasma was obtained from the Central Laboratory of the Red Cross, Amsterdam, The Netherlands. The water used in this study was purified by percolation through a mixed bed ion exchange column followed by an activated carbon column and a microfilter. All other chemicals used were of analytical grade. Si wafers were covered with a layer of SiO2 of approximately 100 nm, as determined by ellipsometry, formed by thermal oxidation of the Si surface at 1000 °C for 90 min. The SiO2 layer is essential for obtaining a high sensitivity in the reflectometer experiments. (PEO)n-grafted Si/SiO2 wafers were prepared in two ways. (a) Strips of Si/SiO2 surfaces (70 mm × 10 mm) were cleaned by UV-ozone treatment for 30 min and subsequently hydrophobized by immersion in 1% (v/v) DDS in toluene for 15 min. Thereafter, the surfaces were extensively rinsed with ethanol and water. Silanization changed the water contact angle from about 5° (for a clean silica surface) to about 90°. The surfaces thus prepared were kept under distilled water or in a vacuum and used within a week. (PEO)n-(PPO)m-(PEO)n triblock copolymers were adsorbed from a 10 mg/L aqueous solution for 30 min. (b) The surfaces of Si/SiO2 strips (70 mm × 10 mm) that were cleaned by UV-ozone treatment (30 min) were covered with a solution (1 g/L) of vinyl-(PS)200 in chloroform. The chloroform was evaporated in a stream of nitrogen after which the strips were placed under a vacuum in an oven and heated at 150 °C overnight. By this treatment, the (PS)200 is chemically attached to the Si/SiO2 surface.17 Excess polymer is removed by thoroughly washing with chloroform whereafter the surfaces are dried in a stream of nitrogen. This modification resulted in a stable PS coating on the surface having a thickness of 2-3 nm as measured by ellipsometry.18 Monolayers of (PS)38-(PEO)n block copolymers at an air/water interface were prepared in a Langmuir trough as described in refs 9, 19, and 20, and pressure-area isotherms were determined. At a predetermined constant area per polymer molecule, the PS-coated strips were dipped (air f water) and retracted (water f air) through the (PS)38-(PEO)n monolayer at a speed of 1 mm2 per second, while keeping the interfacial pressure and, hence, the area per polymer molecule in the monolayer constant. Thus, a single layer of (PS)38-(PEO)n was transferred from the monolayer at the air/water interface onto the PS-coated strips. Usually, a transfer ratio of unity was reached, resulting in a grafting density at the strip that is the same as the one selected at the air/water interface. Continuous washing with water did not remove the (PS)38-(PEO)n layer from the surface, as probed by reflectometry. The samples were heated for 5 min at 95 °C (which is just beyond the glass temperature of PS) so that the PS block of the copolymer fuses with the PS coating. This firmly attaches the (PEO)n buoy on the surface. Most of the BSA and plasma component adsorption experiments were performed in an aqueous environment at 22 °C. To investigate the temperature dependence of adsorption on the PEO-tethered surfaces, additional experiments were conducted at 38 °C. The adsorption process was monitored in real time using a home-built reflectometer. The design of the reflectometer and the measuring procedure have been described in detail elsewhere.21 The proteins were transported toward the surface by a stagnation point flow.22 The flow of the solution into the reflectometer was set at 1 mL/min. (17) Maas, J. H.; Cohen Stuart, M. A.; Sieval, A. B.; Zuilhof, H.; Sudho¨lter, E. J. R. Thin Solid Films 2003, 426, 135. (18) Azzam, R. M. A.; Bashara, N. M. Ellipsometry and Polarized Light; North-Holland: Amsterdam, 1989. (19) Currie, E. P. K.; Sieval, A. B.; Avena, M.; Zuilhof, H.; Sudho¨lter, E. J. R.; Cohen Stuart, M. A. Langmuir 1999, 15, 7116. (20) Currie, E. P. K.; Sieval, A. B.; Fleer, G. J.; Cohen Stuart, M. A. Langmuir 2000, 16, 8324. (21) Dijt, J. C.; Cohen Stuart, M. A.; Fleer, G. J. Adv. Colloid Interface Sci. 1994, 50, 79. (22) Dabros, T.; van de Ven, T. G. M. Colloid Polym. Sci. 1983, 26, 694.
Norde and Gage Ellipsometric measurements were made with an ellipsometer from Sentec Instruments GmbH, Berlin, Germany.
Theory The end-to-end distance 〈r2〉0.5 of a coiled polymer molecule in a good solvent is related to the length of the polymer chain through
〈r2〉0.5 ) Np0.6lp
(1)
where Np is the number of freely jointed chain segments of persistence length lp and molar mass mp. Hence,
Np )
M mp
(2)
with M being the molar mass of the polymer. The equilibrium thickness L of a well-soluble polymer layer, that is in a brush conformation tethered at a surface, may be estimated by the mean field analysis described by Alexander23 and de Gennes24 assuming a homogeneous, constant segment density distribution throughout the brush, the so-called box-model:
(6z)
(3)
Np3/2σv lp
(4)
L2 ) Np lp2
2/3
where
z)
with σ being the grafting density of the polymer at the surface ()number of polymer chains per unit surface area) and v the excluded volume parameter, given by
v ) w1/2(1 - 2χ)
(5)
where χ is the Flory-Huggins parameter for the polymer segment-solvent interaction and
( )
v˜ mo w ) c∞ NAv
2
(6)
in which c∞ is the number of chain bonds per segment,
c∞ )
lp lo
(7)
and mo and lo are the molar mass and length per bond in the polymer chain, v˜ is the specific volume of the polymer, and NAv is Avogadro’s number. Thus, if for a given system M, mo, lo, and lp are known, 〈r2〉0.5 can be derived. If for soluble polymer molecules that are terminally attached to the surface d ) σ-0.5 < 2〈r2〉0.5, the polymer chains are assumed to be in a brush conformation. Knowing the values of σ, v˜ , and χ allows evaluation of the equilibrium thickness L of the polymer brush. Results and Discussion For PEO in water, the following values for the various parameters are taken (Table 6.1 in ref 25): lo ) 0.148 nm, (23) Alexander, S. J. Phys. France 1977, 38, 983. (24) De Gennes, P. G. Macromolecules 1980, 13, 1069. (25) Russel, W. B.; Saville, D. A.; Schowalter, W. R. Colloidal Dispersions; Cambridge University Press: Cambridge, U.K., 1989; p 176.
Interaction of Proteins with PEO-Tethered Surfaces
Langmuir, Vol. 20, No. 10, 2004 4165 Table 2. Characteristics of (PEO)n-Tethered Surfaces Obtained by Adsorption of (PEO)n-(PPO)m-(PEO)n Triblock Copolymers on Hydrophobic Surfacesa sample
M(PEO)n (Da)
σ-1 (nm2)
L (nm)
2〈r2〉0.5/d
1 2 3 4
5600 4360 345 2625
11.70 12.53 6.81 20.34
4.2 3.2 (0.3) 1.6
5.3 4.4 1.3 2.6
a
Properties of the copolymer samples are given in Table 1.
Figure 3. Adsorption of the triblock copolymer (PEO)127(PPO)48-(PEO)127 from a flowing aqueous solution at the stagnation point on a hydrophobized (DDS-coated) silica surface. Triblock copolymer solution, 10 mg/L; solution flow rate, 1 mL/ min; temperature, 22 °C.
Figure 4. Plateau values for the adsorption of (PEO)n(PPO)m-(PEO)n on hydrophobized silica surfaces. Sample information is given in Table 1. Conditions are as in Figure 3. Table 1. The (PEO)n-(PPO)m-(PEO)n Triblock Copolymers (PEO)n sample (number in Figure 1) 1 2 3 4
(PPO)m
n
molar mass (Da)
〈r2〉0.5 (nm)
m
molar mass (Da)
127 100 8 60
5600 4360 345 2625
9.1 7.9 1.7 5.8
48 56 47 90
2800 3280 2760 5250
mo ) 14.7 × 10-3 kg mol-1, lp ) 0.60 nm, mp ) 60.0 × 10-3 kg mol-1, v˜ ) 0.79 × 10-3 m3 kg-1, and χ ) 0.40.26 Surfaces Grafted with (PEO)n-(PPO)m-(PEO)n. Table 1 summarizes the triblock copolymers that are adsorbed on the silicon wafers. Values for 〈r2〉0.5 of the (PEO)n parts are included. By way of example, Figure 3 shows the rate of adsorption of (PEO)127-(PPO)48-(PEO)127 from a 10 mg/mL aqueous solution. The adsorbed mass per unit surface area Γ reaches a plateau value Γpl within a few minutes. As commonly observed for polymers, the adsorption is irreversible with respect to dilution, that is, rinsing with water does not result in significant desorption. The Γ(t) curves for the other samples were similar but reached different plateau values. These values are given in Figure 4. It is to be expected and indeed experimentally confirmed7 that the (PEO)n-(PPO)m-(PEO)n triblock copolymers adsorb with their hydrophobic (PPO)m part onto the hydrophobic DDS layer, whereas the (PEO)n parts dangle into the aqueous solution. Comparison of samples 1, 2, and 3, that contain (PPO)m parts of similar size, shows that Γpl decreases with decreasing PEO chain length. The small adsorbed amount of sample 4 results from the relatively low grafting density (26) Amu, T. C. Polymer 1982, 23, 1775.
Figure 5. Adsorption of BSA from a flowing aqueous solution at the stagnation point on (PEO)n-tethered surfaces (obtained by adsorption of (PEO)n-(PPO)m-(PEO)n on hydrophobized silica surfaces). Influence of the conformation of the preadsorbed copolymer layer. Information on the copolymer samples (numbers in brackets) is given in Table 1. BSA solution, 50 mg/L; solution flow rate, 1 mL/min; temperature, 22 °C.
of PEO, as large PPO blocks are attached to the surface. From the adsorbed mass and the molar mass of the (PEO)n-(PPO)m-(PEO)n sample, the grafting density σ of the PEO chains and their average separation d ()σ-0.5) are calculated. These separations are 3.42, 3.54, 2.61, and 4.51 nm, and hence the values for 2〈r2〉0.5/d are 5.3, 4.4, 1.3, and 2.6 for samples 1-4, respectively. It is concluded that the PEO chains of samples 1 and 2 are forced to extend into the solution implying a brush conformation. Sample 4 has only a weak brushy character, and this applies even more so to sample 3, if a brush is formed at all with that sample. Table 2 summarizes some characteristics of the tethered PEO layers. The value of L for sample 3 is given between brackets because of the uncertainty of brush formation and, hence, of the applicability of eq 3. After preadsorption of the triblock copolymers, a 50 mg/L aqueous BSA solution was supplied to the surfaces and the adsorption was monitored by reflectometry. BSA adsorption reached plateau values within a few minutes. Figure 5 shows these plateau values as a function of 2〈r2〉0.5/d. The value for 2〈r2〉0.5/d ) 0 represents BSA adsorption on the DDS-coated Si/SiO2 surface. BSA adsorption is reduced by the presence of (PEO)n-(PPO)m(PEO)n at the surface, the more so the more the conformation of the tethered PEO chains changes from a mushroom to a brush. In the insert in Figure 5, Γpl BSA is plotted against the reciprocal grafting density σ-1. As expected for a PEO brush (samples 1, 2, and 4), Γpl BSA decreases with increasing grafting density. The deviating result for sample 3 supports the idea that the PEO layer is in a mushroom rather than in a brush conformation. The shape and dimensions of a (hydrated) BSA molecule
4166
Langmuir, Vol. 20, No. 10, 2004
Norde and Gage
Table 3. (PS)38-(PEO)n Diblock Copolymers (PEO)n sample
(PS)38 molar mass (Da)
n
molar mass (Da)
〈r2〉0.5 (nm)
L (nm) at σ-1 ) 10 nm2
1 2 3 4 5
4000 4000 4000 4000 4000
90 148 250 445 700
3950 6500 11000 19600 30800
7.4 10.0 13.7 19.4 25.4
3.1 5.1 8.6 15.3 24.1
may be approximated by an ellipsoid of rotation having axes of 14 nm × 4 nm × 4 nm.27 Hence, according to Halperin’s model,1,11 BSA molecules may penetrate the brush of sample 4 (σ-1 ) 20 nm2) in an end-on orientation and deposit in the primary minimum. It is unrealistic to assume that adsorption in the secondary minimum at the outer periphery of the PEO layer contributes significantly to the observed adsorbed amount of 0.25 mg m-2. The attractive interaction between the protein and the sorbent at the outer edge of the PEO layer results from Londonvan der Waals forces. For a planar geometry, the resulting Gibbs energy GL-vdW may be calculated using28
GL-vdW ) -
A 12πL2
Figure 6. Pressure-area isotherm for a (PS)38-(PEO)700 monolayer at an air/water interface. Temperature, 22 °C.
(8)
where A is the Hamaker constant for BSA interacting with the silicon wafer across the PEO layer (which mainly consists of water). Taking a value for A of the order of 10-21 J,29 GL-vdW is of the order of -0.1 kT per BSA molecule, which is too low to capture the molecules. The more dense layers of samples 1 and 2 are more difficult to penetrate by BSA, resulting in smaller values of Γpl BSA. For those samples, secondary adsorption is also excluded since their equilibrium thicknesses are even larger than that of sample 4. The residual adsorbed amounts of BSA in the primary minimum of the polymer-coated surface, shown in Figure 5, are plateau values of Γ(t) curves, reached within a few minutes and followed for 30 min. Yet, because adsorption in the primary minimum is an activated process, it is possible and even probable that BSA adsorption continues in time, but, apparently, at a rate that is too slow to be detected by our method. The initial fast adsorption may be due to homogeneity imperfections of the brush. Surfaces Grafted with (PS)38-(PEO)n. Some characteristics of the (PS)38-(PEO)n samples are presented in Table 3. A known amount of the diblock copolymer was spread from a 1 g/L solution in chloroform at an air/water interface in a Langmuir trough. For each of the samples, the pressure (π)-area (σ-1) isotherm at room temperature was determined. By way of example, the π-σ-1 isotherm of (PS)38-(PEO)700 is shown in Figure 6. Compression forces the monolayer into a condensed state with the (PS)38 blocks at the interface and the (PEO)n parts protruding in the aqueous subphase. For the five samples, LangmuirBlodgett transfer onto PS-coated silicon wafers was carried out at σ-1 ()area per molecule) of 10 nm2/molecule. At this grafting density, the average separation d between the tethered PEO chains is 3.16 nm. In view of the values for 2〈r2〉0.5/d, it is concluded that for each of the samples the PEO layer is in a brush conformation. The thickness L of the brush derived from the box model as described in the foregoing theoretical section is for each sample given (27) Squire, P. G.; Moser, P.; O’Konski, C. T. Biochem. J. 1968, 7, 426. (28) Norde, W. Colloids and Interfaces in Life Sciences; Marcel Dekker: New York, 2003; Chapter 16. (29) Nir, S. Prog. Surf. Sci. 1977, 8, 1.
Figure 7. Adsorption of BSA from a flowing aqueous solution at the stagnation point on (PEO)n-tethered surfaces (obtained by grafting (PS)38-(PEO)n on a PS-coated silica surface). Influence of the thickness of the (PEO)n brush at a grafting density of 10 nm2/(PEO)n-molecule. Conditions are as in Figure 5.
in the last column of Table 3. It is noted that the box model may underestimate the brush thickness. Using a self-consistent field model, Milner et al.30 and Zhulina et al.31 demonstrated that the segment density decreases parabolically from a finite value at the grafting plane to zero at the edge of the brush. The equilibrium thickness of the parabolic distribution scales in the same way with the grafting density and the length of the polymer chains, but it is higher by a factor of 1.34.10 In Figure 7, the plateau values, Γpl BSA, of Γ(t) for BSA adsorbed from a 50 mg/L solution in water onto the brushed surfaces are presented. These plateau values are reached within 3-5 min (Γpl BSA for n ) 0 is the adsorption at the bare PS-covered surface). The PEO brushes strongly suppress BSA adsorption. It is remarkable that, at least for the grafting density chosen (10 nm2/molecule), Γpl BSA plotted as a function of the thickness L of the brush passes through a minimum. The increasing trend in Γpl BSA(L) may be attributed to entrapment of BSA molecules in the PEO chains. This would imply attractive interaction between BSA and PEO. A similar conclusion has been arrived at (30) Milner, S. T.; Witten, T. A.; Cates, M. Macromolecules 1988, 21, 2610. (31) Zhulina, E. B.; Borisov, O. V.; Pryamitsin, V. A. J. Colloid Interface Sci. 1990, 137, 495.
Interaction of Proteins with PEO-Tethered Surfaces
Langmuir, Vol. 20, No. 10, 2004 4167
Table 4. Characteristics of (PEO)n-Tethered Surfaces Obtained by Grafting (PS)38-(PEO)700 on PS-Coated Surfaces σ-1 (nm2)
2〈r2〉0.5/d
L (nm)
σ-1 (nm2)
2〈r2〉0.5/d
L (nm)
3.24 3.92 4.41 5.70 6.17 7.07 8.20 9.23 10.00 12.00
28.2 25.6 24.2 21.3 20.4 19.1 17.7 16.7 16.0 14.7
35.1 32.9 31.6 29.1 28.1 26.9 25.6 24.7 24.0 22.8
13.20 14.00 14.90 15.80 17.20 20.00 23.00 24.50 33.00
14.0 13.6 13.1 12.8 12.2 11.3 10.6 10.3 8.8
21.8 21.5 21.2 20.6 20.2 19.3 18.3 18.0 16.1
Figure 8. Adsorption of plasma components on (PEO)ntethered surfaces (obtained as mentioned in Figure 7). Influence of the (PEO)n-grafting density and the temperature. Conditions are as in Figure 5.
by Currie et al.9 who observed enhanced BSA adsorption at surfaces precoated with brushes of long PEO chains and at not too high grafting densities, that is, with (PEO)445 and (PEO)700 at σ-1 J10 nm2. This suggests that the protein resilience of PEO-brushed surfaces results from a considerable activation energy for the protein to enter the brush rather than from repulsive forces between PEO and BSA. This is in accordance with the conclusion of Efremova et al.,32,33 based on their force-distance experiments with streptavidin at PEO-brushed surfaces, and it is corroborated by the work of Abbott et al.34 who showed weak attraction between BSA and PEO coils in an aqueous two-phase system. We selected (PS)38-(PEO)700 to investigate the influence of the grafting density on the adsorption of components from a 1% (v/v) dilution of human blood plasma. The grafting densities studied are listed in Table 4, together with the corresponding values of 2〈r2〉0.5/d (indicating the brush character of the PEO layers) and the equilibrium thickness L of the brushes. In Figure 8, adsorption plateau values, that are reached within a few minutes, of plasma components ()plasma proteins) are given as a function of σ-1. With densely packed brushes, that is, σ-1 j 8 nm2, plasma protein adsorption is suppressed below a detectable (32) Efremova, N. V.; Sheth, S. R.; Leckband, D. E. Langmuir 2001, 17, 7628. (33) Sheth, S. R.; Efremova, N. V.; Leckband, D. E. J. Phys. Chem. B 2000, 104, 7652. (34) Abbott, N. L.; Blankschtein, D.; Hatton, T. A. Macromolecules 1992, 25, 3932.
level. The protein repelling efficacy of the brush decreases with decreasing grafting density. For σ-1 J 10 nm2, plasma protein adsorption sharply increases with decreasing grafting density. This is ascribed to penetration of the brush and adsorption in the primary minimum. As discussed before, the brushes are far too thick to allow for considerable adsorption at the secondary minimum. For σ-1 J 20 nm2, plasma protein adsorption reaches a value of about 3 mg m-2. This value considerably exceeds plasma adsorption at the bare PS-covered surface which amounts to 2.0 mg m-2 at 22 °C and 1.7 mg m-2 at 38 °C. As is observed for BSA,9 (PEO)700 brushes of relatively low densities enhance adsorption of plasma proteins. pl (σ-1) data for BSA adsorption Comparison with Γplasma on surfaces tethered with (PEO)n-(PPO)m-(PEO)n, shown in Figure 5, reveals that for plasma the onset of adsorption occurs at a slightly higher grafting density, even though the brushes prepared by (PS)38-(PEO)700 are much thicker. This suggests the involvement of plasma proteins that are smaller than albumin (as the dimensions of human serum albumin and BSA are similar35). However, because of the uncertainties involved in comparing brushes that are prepared in different ways using different polymers, it might well be that the data presented in Figure 8 predominantly reflect adsorption of albumin from the human blood plasma. Figure 8 shows that the grafting density trajectory over which the brush changes from protein-adhesive to proteinresistant slightly shifts to lower grafting densities when going from 25 to 38 °C, if the shift is significant at all. In other words, even though at increased temperature the activation energy may be less of a barrier it does not result in reduced protein resistance of the brush, at least for a period of time that is not too long. Conclusions PEO grafted at surfaces affects adsorption of proteins from solution onto such surfaces. Under most conditions, protein adsorption is reduced. To attain effective repression of protein adsorption, the brush must be sufficiently dense, that is, with separations between the neighboring PEO chains well below the dimensions of the indwelling protein molecules. The conventional theories describing protein resistance of grafted PEO layers, that are based on (osmotic) repulsion between the protein and the polymer chains, explain only part of our observations. For instance, tenuous brushes tend to enhance protein adsorption in particular in the case of long PEO chains. This indicates the existence of a (weak) attractive force between PEO moieties and protein molecules. Hence, protein repellency of a densely brushed surface is the result of a high activation energy for the protein molecules to enter the brush. As a consequence, brushed surfaces may not withstand protein adsorption for a prolonged period of time and it is questionable whether such surfaces will remain free from the adhesion of biological cells, that is, biofilm formation, in long-term applications. Raising the temperature from 25 to 38 °C does not result in a decrease of the activation energy barrier that is sufficient to reduce the protein resilience of the PEO brush. LA030417T (35) Foster, J. F. Plasma Albumin. In The Plasma Proteins, Vol. I; Putnam, F. W., Ed.; Academic Press: New York, 1960.
