Adsorption of Candida rugosa lipase at water-polymer interface: The case of poly(dl )lactide

Surface Science 605 (2011) 2017–2024

Contents lists available at ScienceDirect

Surface Science j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s u s c

Adsorption of Candida rugosa lipase at water-polymer interface: The case of poly(DL)lactide Gihan Kamel a,⁎, Federico Bordi b, Laura Chronopoulou c, Stefano Lupi a, Cleofe Palocci c, Simona Sennato b, Pedro V. Verdes d a

Dipartimento di Fisica, Sapienza Università di Roma, Piazzale A. Moro 2, 00185 Roma, Italy Dipartimento di Fisica and CNR-IPCF, Sapienza Università di Roma, Piazzale A. Moro 2, 00185, Roma, Italy Dipartimento di Chimica, Sapienza Università di Roma, Piazzale Aldo Moro 5, 00185 Rome, Italy d Soft Matter and Molecular Biophysics Group, Department of Applied Physics, Campus Vida, Faculty of Physics, University of Santiago de Compostela, 15782 Santiago de Compostela, Spain b c

a r t i c l e

i n f o

Article history: Received 30 March 2011 Accepted 27 July 2011 Available online 7 August 2011 Keywords: Protein adsorption Nanostructured polymers Langmuir films

a b s t r a c t Insights into the interactions between biological macromolecules and polymeric surfaces are of great interest because of potential uses in developing biotechnologies. In this study we focused on the adsorption of a model lipolytic enzyme, Candida rugosa lipase (CRL), on poly-(D,L)-lactic acid (PDLLA) polymer with the aim to gain deeper insights into the interactions between the enzyme and the carrier. Such studies are of particular relevance in order to establish the optimal conditions for enzyme immobilization and its applications. We employed two different approaches; by analyzing the influence of adsorbed CRL molecules on the thermodynamic behavior of Langmuir films of PDLLA deposited at air–water interface, we gained interesting information on the molecular interactions between the protein and the polymer. Successively, by a systematic analysis of the adsorption of CRL on PDLLA nanoparticles, we showed that the adsorption of a model lipase, CRL, on PDLLA is described in terms of a Langmuir-type adsorption behavior. In this model, only monomolecular adsorption takes place (i.e. only a single layer of the protein adsorbs on the support) and the interactions between adsorbed molecules and surface are short ranged. Moreover, both the adsorption and desorption are activated processes, and the heat of adsorption (the difference between the activation energy for adsorption and desorption) is independent from the surface coverage of the adsorbing species. Finally, we obtained an estimate of the number of molecules of the protein adsorbed per surface unit on the particles, a parameter of a practical relevance for applications in biocatalysis, and a semi-quantitative estimate of the energies (heat of adsorption) involved in the adsorption process. © 2011 Elsevier B.V. All rights reserved.

1. Introduction Polylactic acid (PLA) is one of the most commonly used polymers for biotechnological applications [1,2]. In fact, it possesses features like non-toxicity, biodegradability and biocompatibility that make it an ideal material for a wide number of biotechnological applications [3]. Moreover, the biodegradation of PLA leads to pharmacologically inactive substances, which are absorbed by the body or removed by metabolism [4]. PLA is derived from renewable sources, such as corn starch and sugar cane, and its use in the preparation of bioplastics, usefulness for producing loose-fill packaging, compost bags, food packaging, and disposable tableware, qualify it as a sustainable alternative to petrochemical-derived products [5,6]. PLA monolayer isotherms have been recently studied [7]. They show a feature that looks similar to the well-known monolayer main transition of fatty acids or

⁎ Corresponding author at: Dipartimento di Fisica, Sapienza Università di Roma, Piazzale Aldo Moro, 2-00185 Rome, Italy. Tel.: +39 06 4991 3506; fax: +39 06 44 63 158. E-mail address: [email protected] (G. Kamel). 0039-6028/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.susc.2011.07.021

phospholipids, with analogous temperature dependence, although PLA has a completely different molecular structure. The dependence of the main transition on the length of the tails of the molecules is mirrored by the dependence on the length of the polymer chain (or molecular weight). Materials of different molecular weight are miscible in monolayers and the main transition like feature is shifted in proportion to the concentration. This is similar to the behavior of mixtures of biologically relevant lipids and suggests the possible use of polymers as membrane materials for artificial applications because polymers should be able to provide much more durable membranes than natural or small molecule films can afford. All cellular properties are strongly dependent from the specific functional features provided to the subcellular space by interfaces. Lipases are soluble enzymes comprising a category of the most frequently studied interfacial enzymes [8,9], mainly acting at the oil–water interface (surface of oil droplets) and not at the level of cellular membranes, on the contrary to some phospholipases. Having a catalytic action which is strictly dependent upon the presence of a water–lipid interface, lipases are an example of the importance of studying protein adsorption and protein–surface interactions [10–14]. However most of the published

2018

G. Kamel et al. / Surface Science 605 (2011) 2017–2024

studies on lipase–monolayer interface interactions concern the interactions with lipid films that is with the substrate. Moreover, lipase immobilization for biocatalytic applications is particularly important and represents a very active research area [15–17]. When lipases are used as immobilized enzymes, their substrate may also be solubilized in organic solvent, thus increasing dramatically the number of their potential applications [18–20]. Candida rugosa lipase (CRL) is one of the most important industrial enzymes, thanks to its ability to produce chiral chemicals with high enantiomeric purity [21,22]. Having the GRAS status, its multiple applications of commercial interest cover a broad range of industrial fields, from food, to chemical synthesis, to skin care [23]. It is well established that a hydrophobic interface can improve lipase stability. Moreover, recent studies have demonstrated that the interaction of lipolytic enzymes with nanostructured materials can enhance the activity of the adsorbed protein [24]. The mechanisms likely to occur during the adsorption of lipase onto monomolecular films at the air–water interface have been studied to some extent [25–27]. Further investigations of protein interactions with synthetic interfaces may inform the design of protein-based biosensors and, since adsorbed proteins mediate the interaction of cells with a surface, complement our understanding of cell–surface interactions. Adsorption of proteins at an interface can be investigated by using a number of methods. The Langmuir balance is a useful technique for model studies of molecular interactions at interfaces because surface pressure-area isotherms reflect the intermolecular forces operating in two-dimensional arrays of macromolecules, and also provide information about their organization and conformational changes [28–31]. When lipases are used as immobilized enzymes, their substrate is usually solubilized in organic solvents. Therefore, a double interest in studying lipase adsorption arises: one can study adsorption in relation with the mode of action of the enzyme on its natural substrate, but also in relation with enzyme immobilization for other purposes. In this context, the adsorption of an enzymatic protein, candida rugosa lipase (CRL), on poly-D,L-lactic acid (PDLLA)-based polymeric surfaces, was investigated. Qualitative information on the interactions between the protein and the polymeric surface as well as on their mutual organization was obtained. CRL adsorption isotherms on PDLLA nanoparticles were also studied and analyzed to gain quantitative thermodynamic information and an estimate of the heat of adsorption. 2. Experimental section 2.1. Materials The polylactide employed in this study, a copolymer of poly-D and acid, PDLLA, (MW 75–120 kDa), and the lipase from Candida rugosa (CRL), type VII (MW 57 kDa) were purchased from SigmaAldrich (St. Louis MO). Bradford Reagent and bovine serum albumin (BSA), chloroform, dimethylformamide and other solvents, analytical grade, were also purchased from Sigma. Na2HPO4 and KH2PO4 were purchased from Carlo Erba Analyticals (Milano, Italy). All the chemicals were used without further treatment. Water used in all experiments and cleaning procedures was purified using a Milli-Q system (Billerica, MA), to a specific resistance of 18 MΩ cm − 1. Phosphate buffered saline (PBS, pH 7.6, I 0.1) was used. Protein samples were centrifuged at 14000 rpm for 3 min at 4 °C. Following the centrifugation, a UV/Visible spectrophotometer Pharmacia Biotech Ultrospec 4000 (Stockholm, SE) was used to determine the protein concentration by the Bradford method using bovine serum albumin (BSA) as a standard. L-lactic

2.2. Methods 2.2.1. Thermodynamic measurements on Langmuir films Surface measurements were carried out using a thermostated KSV Langmuir Mini-trough system (KSV LTD, Finland), placed on an anti-

vibration table and enclosed in a Plexiglas box to avoid impurities and dust deposition. In this system, compression is achieved with the symmetric movement of two opposing barriers. In all experiments the compression rate was 20 cm 2/min. A rectangular trough with a total area of 246.4 cm 2 was used. Trough and barriers were thoroughly cleaned before each measurement with appropriate solvents, and rinsed with ultrapure water. Prior to film deposition, the surface was cleaned repeatedly by slowly sweeping the barriers and vacuum aspirating the surface in between, until no change in surface pressure was detectable comparing the closed and open positions. Due to the rather small effect of the presence of PDLLA on the surface pressure, the cleaning process was a critical step of the measurement procedure. PDLLA was dissolved in chloroform. A known amount of the solution was carefully spread with a micro-syringe onto the clean air– water interface. All the measurements were performed at 25 ± 0.2 °C. The solvent was allowed to evaporate for about 10 min before starting the compression. The desired subphase temperature was controlled by a water circulating bath (C25, Haake, Karlsruhe, Germany). Surface pressure measurements were carried out by the Wilhelmy method, using a roughened platinum plate, to an accuracy of 1 mN/m. Due to unavoidable experimental uncertainties, the maximum difference between every two repeated curves is contained in a band of ~±2.5 mN/m on the average. Surface potential measurements were performed using the noncontact vibrating plate capacitor method, originally introduced by Kelvin and improved by Yamins and Zisman [32,33]. We used a computer-controlled device (SPOT1, KSV LTD, Finland), with a 17 mm diameter active electrode, placed at less than 3 mm above the air– water interface, and a stainless steel reference electrode immersed in the subphase [34]. Surface pressure and surface potential were measured simultaneously during film compression; at the beginning of each experiment, the surface potential of the aqueous phase was measured, and this value was assumed as a reference. Before starting the compression, the solvent was allowed to evaporate for about 10 min for a good stabilization of the initial potential value. The probe parameter setting was adjusted in order to reduce the stray capacitance effect, due to the small variation of the electrode distance from the interface in different measurements, to a negligible extent. The reproducibility of surface potential measurements was within 5 mV. The interaction of CRL with PDLLA films was studied by comparing the thermodynamic behavior of PDLLA films spread on pure PBS subphase and on CRL-containing subphase. In a different series of experiments, changes in the surface parameters of PDLLA induced by the CRL were investigated by monitoring the effect of injecting different concentrations of the protein beneath the compressed films at different target pressures of compression (5, 10, and 15 mN/m). More in detail, a fixed volume of CRL buffered solution (at the concentration required to obtain the final desired one in the whole subphase volume) was carefully injected in the subphase by a syringe ending in a thin Teflon tube placed in the subphase well below the film, thus avoiding any perturbation of the film [28]. The increase in surface pressure upon CRL injection was monitored until it reached an equilibrium value. All experiments were repeated at least three times. 2.2.2. Adsorption on PDLLA-nanoparticles PDLLA nanoparticles were prepared by using a recently patented methodology [35]. In brief, the commercially available polymer was dissolved in dimethylformamide. The obtained solution (5 mg/ml) was then transferred in a dialysis bag and dialyzed against water, which is a non-solvent for the polymer (solvent/non solvent ratio 1:20). After 5 days at 4 °C the precipitate was recovered, rinsed several times in water, centrifuged and freeze-dried. The size and morphology of the obtained nanoparticles (average diameter 220 ± 5 nm) were characterized by scanning electron microscopy and dynamic light scattering as described elsewhere [24]. Adsorption experiments of CRL onto PDLLA nanoparticles were performed in Pyrex tubes containing a known

G. Kamel et al. / Surface Science 605 (2011) 2017–2024

amount of the particles dispersed in 2 ml of a phosphate buffer solution (PBS 0.1 M, pH 7.6) of lipase (50 mg/ml), under magnetic stirring (600 rpm) and at controlled temperature. After following the time course of the adsorption for 6 h, the samples were filtered through cellulose nitrate filters (Whatman, pore diameter 0.025 μm) to recover the particles. The amount of immobilized enzyme was obtained by standard Bradford assay of the initial lipase solutions, the supernatants, and the washing solutions after immobilization. Protein concentration was determined spectrophotometrically by measuring the adsorption peak at 595 nm. 3. Results and discussion 3.1. PDLLA Langmuir films Typical surface pressure, π, vs. mean molecular area, A, isotherms of PDLLA monolayers at air–water interface are shown in Fig. 1. PDLLA films were deposited on pure PBS subphase (solid line) and on CRLcontaining subphase (at a concentration of 0.12 mg/ml, dashed line). For comparison, the typical “isotherm” obtained by simply sweeping the barriers on a subphase containing CRL (at a concentration of 0.12 mg/ml), but without spreading any film on its surface, is also shown (dotted line). The presence of a variable amount of protein at the interface that adsorbs from the bulk is clearly apparent. The isotherm of PDLLA on pure PBS subphase is qualitatively similar to those reported in literature for PLLA and PDLA homopolymers [7,36,37]. However, the smoothed step that is observed for PLLA and PDLA in the mean molecular area (Mma) range between 20 and 30 Å2/molecule, resembling the feature that in fatty acids monolayers would be typical of the main phase transition, is absent here. It should be noted that the area here refers to the lactic acid monomers. On the contrary, the breakdown (where the monolayer breaks down and its fragments slide one over the other beginning to form a multilayer) appears to occur at similar values of area per molecule and surface pressure (~18 Å 2/molecule and ~10 mN/m; 17–18 Å 2/molecule and same pressure for PLLA [7]). Increasing compression beyond this point, pressure of the PDLLA film increases rather gradually and no plateau is observed, again as in the homopolymers isotherms. By using atomic force microscopy techniques, the main-transitionlike step that appears in the isotherms of lactide homopolymers has been recently shown to be due to a genuine disorder/order or liquid expanded-to-condensed (LE/LC) phase transition [38], possibly favored by the helix conformation that the homopolymers tend to assume at the interface under the effect of compression [36]. Both the

Fig. 1. π-A isotherms of PDLLA deposited on a pure PBS subphase (solid line) and on a CRL-containing subphase (dashed line). The isotherm obtained by compressing the interface of the pure CRL subphase, without any film deposition is also shown (dotted line).

2019

absence of an evident first-order transition in the low pressure zone, and the continuous rise of the pressure beyond the breakdown (observed here for the PDLLA copolymer) are consistent with those findings. Actually, the intrinsic disorder that characterizes the copolymer, where stretches of all-D or all-lthat can organize in helices, are separated by coiled regions where the D and l monomers are at random, reflects in a much lower ordering of the monolayer and a greater plasticity of the whole structure. The presence of CRL in the subphase (Fig. 1; dashed line) notably increases disorder and plasticity of the PDLLA film. Consequently, even the breakdown is completely absent. In addition, this isotherm, compared to that of the PDLLA film spread on pure PBS, is characterized by higher values of pressure in the whole compression range. This finding suggests that, upon compression, the protein that initially is at the air–water interface is not completely squeezed out into the subphase, but at least partially remains at the interface, even at the highest compression, forming a CRL-PDLLA mixed film. That the protein itself is surface active and tends to stay at the air– water interface, is apparent from the isotherm obtained by just sweeping the barriers over the pure CRL subphase without depositing any PDLLA film on the surface (Fig. 1; dotted line), where a notable rise of the pressure is observed upon compression. In this case, CRL molecules were not deposited on the surface but spontaneously adsorbed at the interface from the bulk subphase (Gibbs isotherm). Since the number of adsorbed molecules is not known and varies in principle upon compression, the variable shown on the horizontal axis should simply be the area between the barriers. However, in order to plot the curve for the pure CRL-isotherm, and to make a comparison with the PDLLA isotherms possible, we chose to give this area in terms of a fictitious “area per monomer”, as if the same amount of PDLLA used for the other isotherms in the figure had been deposited at the interface. Another and more subtle effect suggests a rather strong affinity of CRL for the PDLLA film. As usual, the measured pressure is set to zero after cleaning the surface and immediately before film deposition. With this procedure, on a pure PBS subphase, when the proper amount of PDLLA is spread, and with the barrier in the initial position, the pressure remains close to zero (Fig. 1; solid line). On the contrary, when the same amount of PDLLA, after having zeroed the pressure, is spread on a CRLcontaining subphase, a vertical (i.e. with motionless barriers) rapid increase of the pressure is observed. Such a sudden increase is clearly due to the interaction of PDLLA with CRL at the interface, and possibly to the recruitment of further CRL molecules from the bulk to the interface due to this interaction. In order to gain qualitative information on the mechanical properties of the PDLLA-CRL mixed film we calculated the static elasticity, ε = −A (∂π/∂A)T, from the isotherms shown in Fig. 1. When PDLLA is spread on a pure PBS subphase (Fig. 2; solid line), ε increases steadily as the film is compressed. Initially, the polymer segments do not overlap and the molecular arrangements of the polymer, due to its strongly hydrophilic character, lie mostly in the subphase [38]. However, upon compression, the hydrophobic methyl groups direct themselves to the air/water interface producing a progressively more closely packed monolayer of a cohesive structure that finally breaks down at 18 Å2/molecule. Again, there is no trace here for the net LE/LC phase transition which is clearly visible in the homopolymer isotherms, confirming the intrinsic disorder of the film formed by the DL copolymer. Intrinsic disorder and plasticity is also pointed out by the small, but significantly different from zero and gradually increasing value of the elasticity after the breakdown, as expected when the structure does not collapse uniformly. As the compression continues, ε increases rapidly, which corresponds to the formation of multilayers. The rather sharp peak shown by the isotherm in the low area region corresponding to pure CRL (Fig. 2; dotted line) could be indicative of an order–disorder transition or of an experimental artifact, however the detailed analysis of the behavior of Gibbs isotherms of pure CRL was beyond the scope of the present work and we did not investigate further

2020

G. Kamel et al. / Surface Science 605 (2011) 2017–2024

Fig. 2. ε-A isotherms of PDLLA deposited on a pure PBS subphase (solid line) and on a CRL-containing subphase (dashed line). Also shown the isotherm of pure CRL subphase (dotted line).

this effect. What concerns here is that after a first steady rise of the elasticity (suggesting a cohesive structure) followed by an abrupt decrease (suggesting the collapse of that structure), the static elasticity remains stable, suggesting that in this whole compression range the protein is only partially squeezed out of the interface into the bulk subphase. Examining the elasticity pattern of the PDLLA-CRL mixed film (Fig. 2; dashed line), it is clear that it behaves in a much different way than pure PDLLA or CRL films. The mixed film seems to exhibit new characteristics owing to the mutual interactions between materials. After a sharp peak similar but higher than that observed in the CRL isotherm, and probably sharing the same origin (i.e., the coordination of protein molecules), now, favored by the polymer presence, an extended region is observed where static elasticity remains stable. This region is followed by a rapid increase of the elasticity and finally by a collapse. This behavior is consistent with the hypothesis of a mixed film that is compressible down to ~ 10 Å 2/molecule (of PDLLA), where the structure becomes much more cohesive. Notably, when PDLLA is deposited on a pure PBS subphase, at this same value of area per molecule, the formation of multilayer begins (Fig. 1). The transition zone around 15 Å 2/molecule and 10 Å 2/molecule, where the slope changes more gradually, probably corresponds to the protein being partially expelled from the interface. The above described scenario, characterized by a rather strong affinity of CRL for the PDLLA film, was supported by surface potential measurements. Surface potential (ΔV) and dipole moment (μn) curves of PDLLA monolayers are shown in Fig. 3. The measured surface potential, ΔV, is defined as the difference between the measured potential of monolayer-covered and monolayer-free subphase, which was taken as reference. This potential can be related to the normal component of the dipole moment μn of the molecules forming the film through the Helmholtz equation: ΔV = μn = ðAε0 εr Þ + φ0

ð1Þ

where εr and ε0 are the “effective” dielectric constant within the layer and the permittivity of free space, respectively; μn is the normal component of the molecular dipole moment; A is the area occupied by each molecule, or monomer; and φ0 is the double-layer contribution, that can be neglected since CRL, the only charged species, is only weakly charged at the pH of the measurement (pH 7.6). The dipole moment μn results from the contribution of different components: the polar groups of adsorbed molecules bathed by the surface; the reorientation of the few layers of water molecules induced by the presence of the film; the hydrophobic part of adsorbed

Fig. 3. Surface potential ΔV (a) and dipole moment μn (b) as a function of the mean molecular area of PDLLA deposited on a pure PBS subphase (solid line) and on a CRL-containing subphase (dashed line) and also for the pure CRL subphase (dotted lines). The arrows point to the scale quantifying each curve.

molecules which extend from the surface to the air. However, a simple surface potential measurement cannot discriminate between these different contributions. The quantity shown in Fig. 3, ΔVA (Eq. (1)), determined experimentally, should hence be regarded as an overall effective dipole that takes into account all these effects. Nevertheless, from the comparison of the curves in Fig. 3, (panel b), interesting information can be obtained. As far as PDLLA on pure PBS (solid line) is concerned, during the compression that precedes the collapse at approximately 18 Å2/molecule, the potential ΔV increases steadily. However, in the same area range, the dipole moment per monomer μn is almost constant; suggesting that the potential increase may be just due to the increasing density of the polymer on the surface when the area is reduced. Nonetheless, the density increase does not affect the orientation or the ordering of molecules. After the collapse, the effective dipole per molecule decreases with its value almost halved (from ~0.34 to ~ 0.13V Å 2/molecule), consistently with the formation of a double layer. The final slight decrease of the dipole moment is probably due to a change in the arrangement of the superimposed layers induced by compression.

G. Kamel et al. / Surface Science 605 (2011) 2017–2024

Dipole moment behavior of PDLLA on a CRL-containing subphase is shown in Fig. 3 (panel b, dashed line). In fact, after an extended range of area variation where the dipole remains almost constant (from ~ 20 down to ~ 11 Å 2/molecule), it decreases upon further compression. The decrease of the dipole moment per molecule which is observed upon compression, above ~ 15 Å 2/molecule and ~ 12 Å 2/molecule for “pure CRL” and “mixed film”, respectively, after an extended zone where its value remains constant, may be attributed to a decrease of the density of molecules at the surface. However, it can be also due to a conformational/positional rearrangement that modifies the dipole moment or its vertical component. This decrease of the effective dipole values occurs, in the presence of PDLLA, at even lower values of the surface area (although, due to the presence of the polymer, the surface area available to the protein is smaller) and is comparatively reduced, suggesting, again, the existence of attractive interactions between the PDLLA and the protein, that stabilize the mixed film and reduce the expulsion of the protein from the interface. 3.2. Injection of the protein into the subphase Fig. 4 shows the CRL injection into the PBS subphase beneath the stabilized PDLLA films compressed to a target pressure. The increase of the film pressure, Δπ, is measured upon CRL injection at the chosen target pressure of the PDLLA film after its relaxation. In Fig. 5, the increase of film pressure, Δπ, measured after CRL injection, is shown for three different target pressures of the PDLLA film (5, 10, and 15mN/m), as a function of the nominal concentration of protein in the subphase, i.e. the concentration that would be reached in the subphase if the protein was homogeneously distributed in the whole volume. The reproducibility in the measured Δπ (Figs. 4 and 5) is better

a T=25OC

2021

than ±0.5 mN/m away from the “coexistence zone” (above or below ~10 mN/m) and rather erratic in this zone (~10 mN/m, filled circles in Fig. 5). Represented points of this erratic behavior at ~10 mN/m are only an example of a series of measurements. Other series measured at the same zone are completely different yet still erratic. That is, the measured Δπ is non-reproducible and the uncertainties are very large that any further analysis is meaningless. Notably, while for π = 5 and 15 mN/m, Δπ increases smoothly towards a plateau with increasing protein concentration, the observed behavior for π = 10 mN/m is erratic and poorly reproducible. This behavior is consistent with the fact that at this pressure (corresponding to the plateau in the isotherm: see Fig. 1, solid line) the PDLLA monolayer has already collapsed but a complete bilayer has not formed yet, so that the interface is now an inhomogeneous mixture of monolayer and multilayer fragments that slowly reorganize under the combined effect of pressure and interactions with the protein. In other words, the protein adsorbs “regularly” causing a larger variation of the pressure of the film when this is a monolayer, than when it is a bilayer, however this is probably due only to the greater rigidity (see Fig. 2) of the bilayer rather than to a different interaction of the protein with the film surface. When the film is at the collapse, the film instability cannot be decoupled from the effect of the adsorption, and the pressure variation is apparently erratic. A final comment is in order. In all the experiments at 15 mN/m, after reaching the target pressure, a large variation of the measured pressure is observed. This relaxation is probably due to the rearrangement of the film fragments after the collapse, as suggested also by the large hysteresis that appears when the films are compressed above the collapse and decompressed again (data not shown, but see [36]). The presence of such a large relaxation could raise doubts about the pressure that has to be considered, if the target one or the pressure measured at the injection. However, while the target pressure characterizes the physical state that has been reached by the film upon compression (at 15 mN/m the formation of an at least partial bilayer), by measuring the pressure during the relaxation one only monitors the mechanical rearrangement of the different zones (mono and bilayers), to be sure that the protein is injected when the pressure is stabilized, but this pressure is not indicative of the physical state of the film fragments, that is assumed to remain substantially unaltered after the compression. When the layers at the surface are homogeneous, a regular increase of Δπ with CRL concentration is observed, and both data set at π = 5 and 15 mN/m fit reasonably with the expression:   C = C0 Δπ = A 1  e

ð2Þ

where C is CRL concentration and A and C0 are constants.

b T=25OC

Fig. 4. CRL injection into the subphase beneath the PDLLA film compressed at 25 °C to a target pressure (a) 5 mN/m, (b) 15 mN/m. Surface pressure increment, Δπ, induced by CRL injection is shown in dotted line.

Fig. 5. Pressure increase of the PDLLA film vs protein concentration in the subphase measured after CRL injection, at different initial pressures of the PDLLA film: (■) 5 mN/m; (●) 10 mN/m; (▲) 15 mN/m.

2022

G. Kamel et al. / Surface Science 605 (2011) 2017–2024

that allows an easier check of the hypothesis that this model fits the experimental data: C C  ΔHads = KB T C = e + LE LM LM

Fig. 6. Linear fitting of the pressure increase the of PDLLA films vs protein concentration upon CRL injection in the subphase, at different initial pressures of the PDLLA film: π = 5 mN/m (■) and π = 15 mN/m (▲).

Assuming, as a first approximation, that the surface pressure increase is proportional to the number of lipase molecules that adsorb on the PDLLA film, the qualitative exponential behavior of the curves shown in Fig. 5, well satisfied for target surface pressures 5 and 15 mN/m, suggests that the adsorption could be described as Langmuir-type, following the equation:

LE =

LM CC e−ΔHads = KB T 1+

C C

e−ΔHads = KB T

ð3Þ

Here, LE is the mass of substance adsorbed per unit area of the adsorbing surface when C is its concentration in the bulk solution, and LM the mass of substance required to saturate the adsorption sites of the unit surface area. ΔHads = Ea − Ed(Ea b Ed) is the heat of adsorption, i.e., the difference between the energies needed for adsorption, Ea, and desorption Ed, C* is a constant with dimensions of a concentration, and finally, KB is Boltzman constant (1.380 × 10 − 23 J/K). Within this model, the adsorption rate is simply written as the product between the rate at which the adsorbing molecules collide with the surface (assumed proportional to their concentration in the bathing solution) and the probability (1-LE/LM) of striking a vacant site, with an activation term e −Ea/KBT. Similarly, the desorption rate is given by the product between the already covered surface fraction (LE/LM), and the corresponding activation term e −Ed/KBT. At equilibrium the two rates must be equal. By simple algebra, Eq. (3) can be put in a linear form

ð4Þ

Assuming again the proportionality between Δπ and LE, a plot of C/Δπ vs. C should give a straight line of slope 1/Δπ* and intercept C  ΔHads = KB T , where 1/Δπ* is the pressure increment reached for the Δπ  e maximum lipase adsorption. Actually, both data set, plotted in this way (Fig. 6), show a well-defined linear behavior so the hypothesis of a Langmuir-type adsorption of CRL on PDLLA seems to be confirmed, where for π = 5 and 15 mN/m, the slopes are 0.073 ± 0.002 and 0.089 ± 0.003 m/mN, and intercepts are (7.3 ± 0.7) × 10 − 4 and (9.9 ± 1.5)× 10− 4 mg m/ml mN, respectively (the regression coefficients R, are 0.997 and 0.998). From the slopes, the maximum pressure increments, Δπ*, can be calculated, which would be expected when the PDLLA films were completely “saturated” by the adsorbed lipase. At the two target pressures, 5 and 15 mN/m, these increments are 14.3 and 11.1 mN/m, respectively. The smaller value of the increment corresponding to the higher target pressure could indicate both that the protein interact differently with a monolayer (π = 5 mN/m) or a bilayer (π = 15 mN/m), or that the multilayered structure, being more “resistant” and compact, is less prone to the “expansion” induced by the protein adsorption. Interestingly, the products of the inverse slope and the intercept, that is in terms of Eq. (4), the quantities C*eΔHads/KBT, (0.0104 and 0.0109 mg/ml for π = 5 and 15 mN/m, respectively) are coincident within the errors (~4.5%) for the two target pressures. This should be expected for a Langmuir-type adsorption, since within this model both the heat of adsorption, ΔHads, and the parameter C*, depend only on the local interaction of the adsorbing substance with the surface, and hence, in our case, are expected to be independent from the mono- or multilayered structure of the film. This simple qualitative finding seems to exclude that lipase interaction with a monolayer or a multilayer should be different. Such a qualitative result, the identification of the model, is a prerequisite to use the model for quantitative predictions (for example, the adsorption enthalpy). However, due to the unknown proportionality factor between the measured pressure increment Δπ and the mass of the enzyme adsorbed per unit area, LE a more, quantitative analysis does not appear feasible. Conversely, such quantitative results can be obtained by using a slightly different system. In fact, as reported in the following section, the thermodynamic study of the adsorption of CRL on PDLLA nano-particles, analyzed in terms of Langmuir adsorption isotherms,

▪

◂

▸

Fig. 7. Adsorption isotherms of CRL on PDLLA nanoparticles at different temperatures: (●) 10 °C; ( ) 20 °C; (▲) 25 °C; (▼) 30 °C; ( ) 40 °C and ( ) 50 °C. (Uncertainties within ~5%).

G. Kamel et al. / Surface Science 605 (2011) 2017–2024

2023

results in a quantitative determination of the surface coverage of the particle by the protein and of the adsorption energy.

3.3. CRL adsorption on PDLLA-nanoparticles In order to gain quantitative data on CRL adsorption on PDLLA surfaces and to further explore the hypothesis of a Langmuir-type adsorption, we studied the adsorption of CRL on PDLLA nanoparticles with a diameter of 220 ± 5 nm suspended in an aqueous solution (PBS, pH = 7.6) at different temperatures (Fig. 7). In Fig. 8, the isotherms are shown in their linearized form. All curves have a similar profile which corresponds well to the Langmuir-type isotherm. In the low protein concentration range, protein loading increases by increasing concentration for all the investigated temperatures. At higher temperature, protein loading grows faster with concentration. Under the premises of Eq. (4) and based on the linear fit parameters shown in Fig. 8, we estimated that the mass of protein required for the complete coverage of PDLLA surface (LM) is ~66.1 mg per 100 mg of PDLLA, taking the average of LM values at the investigated temperatures (in agreement with the Langmuir model assumptions LM is largely independent of the temperature, see inset of Fig. 8). The average diameter of PDLLA nanoparticles as obtained from DLS measurements is 220 ± 5 nm intensity average, (150 ± 5 nm, if number average is used). Assuming for the density of PDLLA nanoparticles the value of 1.26 g/cm3 [39] and the value of 57.780 Da for the molecular weight of the protein, an average area per protein molecule in the adsorbed layer is calculated to be of ≈3.14 nm2 (4.6 nm2 using the number average value of the radius) with an error of 15%. This qualitative estimate appears reasonable since it compares very well with a “molecular diameter” of 5.2 nm of the CRL that can be deduced from its X-ray structure, suggesting a rather tight packing of the protein in the adsorbed layer [40]. As shown in Fig. 9, and according to Eq. (4), the heat of adsorption ΔHads, can be calculated from the equation: ln(a/b) = ΔHads/KB.1/T + lnc* by plotting the natural logarithm of the (intercept/slope) ratio, ln (a/b), as a function of the inverse absolute temperature. Linear fit shown in Fig. 9, gives ΔHads = (2 ± 1)10 3KB. This energy corresponds approximately to 5–6KBT at room temperature. This value appears reasonable, since the adsorption of a large protein such as CRL probably involves many contact points, each with interaction energy of a fraction of KBT. Conversely, the total energy is built up by all the contacts and results in several KBT. Based on this, CRL adsorption on PDLLA nanoparticles is practically irreversible.

Fig. 9. Logarithm of the (intercept/slope) ratio as a function of the inverse absolute temperature.

To better understand the interactions between CRL and PDLLA NPs, the kinetics of the adsorption phenomenon was investigated. By increasing the time of contact, the enzyme loading increases accordingly, until a plateau is reached after approximately 4.5 h, which resulted to be the optimal contact time. Desorption studies on CRL-PDLLA nanobioconjugates proved a good stability of the immobilization in PBS, in which, after a 24 h suspension, the amount of desorbed protein was approximately 10% (data not shown, see [24]). 4. Conclusions Advances in biotechnology have made proteins more and more available, including variants genetically modified for specific tasks, thus hugely expanding their practical use in the conversion of chemicals and materials. Several proteins have been shown to be able to self-assemble upon adsorption on polymeric surfaces, thus opening the possibility of bioactive surface patterning at the molecular level, particularly attractive for applications that include novel diagnostic or therapeutic agents on the single cell scale. The influence of CRL presence on the thermodynamic behavior of PDLLA films deposited at air–water interface was investigated providing qualitative information on the molecular interactions between the protein and the polymer films. By using the Langmuir model, we showed that the adsorbed lipase forms a rather compact layer on PDLLA surface, and obtained an estimate of the average area occupied in this layer by a single protein, that compares

▪

◂

Fig. 8. Linear fitting of the adsorption isotherms of CRL on PDLLA nanoparticles at different temperatures: (●) 10 °C; ( ) 20 °C; (▲) 25 °C; (▼) 30 °C; ( ) 40 °C and ( (Uncertainties within ~ 5%). LM dependence on temperature, T, is shown in the inset.

▸

) 50 °C.

2024

G. Kamel et al. / Surface Science 605 (2011) 2017–2024

reasonably with X-ray structural data reported for CRL in literature. An estimate of the heat of adsorption was also obtained. Acknowledgments P. V. Verdes acknowledges financial support from Ministerio de Ciencia e Innovación, Programa Nacional de Movilidad de Recursos Humanos del Plan Nacional de I-D+i 2008–2011. References [1] D. Lensen, K. van Breukelen, D.M. Vriezema, J.C.M. van Hest, Macromolecular Bioscience 10 (2010) 475. [2] M. Iannone, D. Cosco, F. Cilurzo, C. Celia, D. Paolino, V. Mollace, D. Rotiroti, M. Fresta, Neuroscience Letters 469 (2010) 93. [3] J. Vijayakumar, R. Aravindan, T. Viruthagiri, Chemical and Biochemical Engineering Quarterly 22 (2008) 245. [4] C. Contado, A. Dalpiaz, E. Leo, M. Zborowski, P.S. Williams, Journal of Chromatography. A 1157 (2007) 321. [5] C.I. Onwulata, A.E. Thomas, P.H. Cooke, Journal of Biobased Materials and Bioenergy 3 (2009) 172. [6] J. Sarasa, J.M. Gracia, C. Javierre, Bioresource Technology 100 (2009) 3764. [7] O. Albrecht, Colloids and Surfaces A: Physicochemical and Engineering Aspects 284–285 (2006) 175. [8] F. Nannelli, M. Puggelli, G. Gabrielli, Colloids and Surfaces. B, Biointerfaces 24 (2002) 1. [9] F. Hasan, et al., Biotechnology Advances 27 (2009) 782. [10] R. Narayanan, B.L. Stottrup, P. Wang, Langmuir 25 (2009) 10660. [11] P.M. Kosaka, Y. Kawano, O.A. El Seoud, D.F.S. Petri, Langmuir 23 (2007) 12617. [12] C. Nicolini, D. Bruzzese, V. Sivozhelezov, E. Pechkova, Biosystems 94 (2008) 228. [13] C. Nicolini, E. Pechkova, Nanomedicine 5 (2010) 677. [14] G. Niaura, et al., The Journal of Physical Chemistry. B 112 (2008) 4094. [15] I.V. Pavlidis, T. Tsoufis, A. Enotiadis, D. Gournis, H. Stamatis, Advanced Biomaterials (2010) B179. [16] M. Zoumpanioti, H. Stamatis, A. Xenakis, Biotechnology Advances 28 (2010) 395. [17] N. Miletic, C. Bos, K. Loos, Polymeric Materials (2009) 131.

[18] O. Barbosa, C. Ortiz, R. Torres, R. Fernandez-Lafuente, Journal of Molecular Catalysis B: Enzymatic 71 (2011) 124. [19] G. Celiz, M. Daz, Process Biochemistry 46 (2011) 94. [20] A. Cuetos, M.L. Valenzuela, I. Lavandera, V. Gotor, G.A. Carriedo, Biomacromolecules 11 (2010) 1291. [21] P. Gupta, S. Bhatia, A. Dhawan, S. Balwani, S. Sharma, R. Brahma, R. Singh, B. Ghosh, V.S. Parmar, A.K. Prasad, Bioorganic & Medicinal Chemistry 19 (2011) 2263. [22] D. Chavez-Flores, J.M. Salvador, Biotechnology Journal 4 (2009) 1222. [23] P. Dominguez de Maria, J.M. Sanchez-Montero, J.V. Sinisterra, A.R. Alcantara, Biotechnology Advances 24 (2006) 180; C.C. Akoh, G.C. Lee, J.F. Shaw, Lipids 39 (2004) 513. [24] L. Chronopoulou, G. Kamel, C. Sparago, F. Bordi, S. Lupi, M. Diociaiuti, C. Palocci, Soft Matter 7 (2011) 2653. [25] J.A. Laszlo, K.O. Evans, Journal of Molecular Catalysis B: Enzymatic 58 (2009) 169. [26] J.A. Laszlo, K.O. Evans, Journal of Molecular Catalysis B: Enzymatic 48 (2007) 84. [27] B.S. Chu, A.P. Gunning, G.T. Rich, M.J. Ridout, R.M. Faulks, M.S.J. Wickham, V.J. Morris, P.J. Wilde, Langmuir 26 (2010) 9782. [28] F. Bordi, C. Cametti, A. Motta, M.A. Molinari, Bioelectrochemistry and Bioenergetics 49 (1999) 51. [29] P. Dynarowicz-Latka, A. Dhanabalan, O.N.J. Oliveira, Advances in Colloid and Interface Science 91 (2001) 221. [30] M. Diociaiuti, I. Ruspantini, C. Giordani, F. Bordi, P. Chistolini, Biophysical Journal 86 (2004) 321. [31] F. Bordi, C. Cametti, F. De Luca, D. Gaudino, T. Gili, S. Sennato, Colloids and Surfaces. B, Biointerfaces 29 (2003) 149. [32] H.G. Yamins, W.A. Zisman, Journal of Chemical Physics 1 (1933) 656. [33] H. Brockman, Chemistry and Physics of Lipids 73 (1994) 57. [34] I.R. Peterson, The Review of Scientific Instruments 70 (1999) 3418. [35] L. Chronopoulou, I. Fratoddi, C. Palocci, I. Venditti, M.V. Russo, Langmuir 25 (2009) 11940. [36] J.M. Klass, R.B. Lennox, G.R. Brown, H. Bourque, M. Pezolet, Langmuir 19 (2003) 333. [37] H. Bourque, I. Laurin, M. Pezolet, J.M. Klass, R. Lennox, G.R. Brown, Langmuir 17 (2001) 5842. [38] S. Ni, W. Lee, B. Lia, A.R. Esker, Langmuir 22 (2006) 3672. [39] D.M. Yunos, Z. Ahmad, A.R. Boccaccini, Journal of Chemical Technology and Biotechnology 85 (2010) 768. [40] J.M. Mancheno, M.A. Pernas, M.J. Martinez, B. Ochoa, M.L. Rua, J.A. Hermoso, Journal of Molecular Biology 332 (2003) 1059.