Langmuir 2006, 22, 1775-1781
1775
Surface Thermodynamics Reveals Selective Structural Information Storage Capacity of c-Fos-phospholipid Interactions Graciela A. Borioli* and Bruno Maggio Departamento de Quı´mica Biolo´ gica-CIQUIBIC, Facultad de Ciencias Quı´micas, UniVersidad Nacional de Co´ rdoba-CONICET, Haya de la Torre y Medina Allende, Ciudad UniVersitaria X5000HUA, Co´ rdoba, Argentina ReceiVed September 14, 2005. In Final Form: December 13, 2005 The surface properties of c-Fos, a regulator of normal and pathologic cell growth and a modulator of phospholipid metabolism, suggest that it has the potential to transduce information through molecular reorganization, placing the nature of its interaction with phospholipids at the basis of its possible effects at the membrane level. Previous studies established that c-Fos induces condensation and depolarization of PIP2 films and expansion and hyperpolarization of PC. We have now explored more in depth the thermodynamic aspects of these lipid-protein interactions, finding that the mixtures have associated hysteresis. The analysis of the excess thermodynamic functions provides evidence of entropic-enthalpic compensations that result in a favorable enthalpic contribution derived from the interaction of c-Fos with PIP2, which exceeds the unfavorable configurational entropy. On the contrary, favorable entropy terms dominate the interaction of c-Fos with PC over the unfavorable enthalpy. The free energy of hysteresis is stored as excess free energy. A shift in molecular packing-dependent surface reorganization, compared to that of ideally mixed films, indicates a gain in information content at the lipid-protein interface in mixed films of c-Fos with PIP2 but not with PC. It is postulated that the free energy stored in these mixtures could act as a bidirectional structural information transducer for dynamic compression-expansion processes occurring on the membrane surface.
Introduction The oncogene product c-Fos is an early constituent of the signaling cascade triggered by various stimuli in eukaryotic cells. Its best recognized function is to regulate the transcription of downstream genes involved in normal and pathologic cell growth,1-3 but a recently ascribed role as a modulator of phospholipid synthesis, linked to the endoplasmic reticulum and independent of transcription,4,5 is also relevant for cellular growth. Another oncogene product, c-Jun, is the most common partner of c-Fos in AP-1 transcription dimers6 which are strongly stabilized by an interface as an equimolar complex.7 Both proteins are able to interact with clean, phospholipid interfaces with comparable affinity, but whereas c-Jun’s interaction with phospholipids is nonspecific7 c-Fos is highly selective, being capable of discerning the phospholipid polar headgroups and charges.8 Although it is equally stabilized at the interface by phosphatidylserine, -glycerol, -choline and phosphatidylinositol-4,5-bisphosphate (PIP2), the most favorable interaction occurs with the latter. A more detailed study reveals that the thermodynamic driving force for these interactions is different depending on the nature of the phospholipid, determining the favorableness or unfavorableness of mixing.9 These features indicate that c-Fos has the potential to participate in the transduction of molecular information at the membrane level. * Corresponding author. E-mail:
[email protected]. Phone: 54-351-4334171. Fax: 54-351-4334074. (1) Angel, P.; Karin, M. Biochim. Biophys. Acta 1991, 1072, 129-157. (2) Shaulian, E.; Karin, M. Nat. Cell Biol. 2002, 4, E131-E136. (3) Eferl, R.; Wagner, E. F. Nat. ReV. Cancer 2003, 3, 859-868. (4) Bussolino, D. F.; Guido, M. E.; Gil, G. A.; Borioli, G. A.; Renner, M. L.; Grabois, V. R.; Conde, C. B.; Caputto, B. L. FASEB J. 2001, 15, 556-558. (5) Gil, G. A.; Bussolino, D. F.; Portal, M. M.; Pecchio, A. A.; Renner, M. L.; Borioli, G. A.; Guido, M. E.; Caputto, B. L. Mol. Biol. Cell 2004, 15, 18811894. (6) Shaulian, E.; Karin, M. Oncogene 2001, 20, 2390-2400. (7) Del Boca, M.; Caputto, B. L.; Maggio, B.; Borioli, G. A. J. Colloid Interface Sci. 2005, 287, 80-84. (8) Borioli, G. A.; Caputto, B. L.; Maggio, B. Biochem. Biophys. Res. Commun. 2001, 280, 9-13. (9) Borioli, G. A.; Caputto, B. L.; Maggio, B. Biochim. Biophys. Acta 2005, 1668, 41-52.
The ability of c-Fos to modulate membrane activities that involve phospholipids is well illustrated by its differential influence on phospholipases A2, C, and sphingomyelinase. Depending on surface packing, small amounts of c-Fos mixed with substrate (dilauroylphosphatidylcholine -PC- or sphingomyelin) have clear and distinct effects on the activation and catalytic phases of the three phospholipases.10 Some of these effects can be explained by changes in the substrate interfacial organization induced by the protein.11 The concept that c-Fos can influence membrane processes, based on selective interactions with phospholipids, calls for a closer look at their essence. Previous work showed that c-Fos undergoes reversible molecular rearrangements in organized interfaces that depend on the lateral surface pressure. At least two major changes in the protein organization occur over defined ranges of surface pressure that are independently revealed by marked alterations of the compression behavior, the in-plane elasticity, which is assessed by the surface compressional modulus, and the reorientation of the transverse dipole potential density.9 Further evidence of this molecular reorganization is provided by Brewster angle microscopy showing that the reflectance curve of a c-Fos monolayer during compression has an inflection that is coincident with the low surface pressure reorganization point.9 The protein intermolecular arrangement also exhibits considerable reversible hysteresis under successive compression-expansion cycles, which implies a high potential capacity for the storage of surface free energy.8 c-Fos exhibits nonideal composition-dependent interactions with PC and PIP2. These interactions involve thermodynamically unfavorable mean molecular area expansion and dipole potential hyperpolarization in binary mixtures with PC and thermodynamically favorable intermolecular condensation and dipole depolarization with PIP2.9 c-Fos contributes to most of the alterations of the mean molecular parameters of the mixed lipidprotein interface when its proportion is below 7 mol %. At higher (10) Borioli, G. A.; Fanani, M. L.; Caputto, B. L.; Maggio, B. Biochem. Biophys. Res. Commun. 2002, 295, 964-969. (11) Borioli, G. A.; Caputto, B. L.; Maggio, B. FEBS Lett 2004, 570, 82-86.
10.1021/la0525168 CCC: $33.50 © 2006 American Chemical Society Published on Web 01/24/2006
1776 Langmuir, Vol. 22, No. 4, 2006
proportions, the surface behavior becomes dominated by the protein whereas the lipid component mainly accounts for the changes in mean molecular area and dipole potential observed in protein-enriched films.9 In addition, 2D phase diagrams in combination with direct visualization of the interface by epifluorescence microscopy and the determination of reflectance by Brewster angle microscopy revealed long-range supramolecular structuring with the formation of segregated, differentially condensed, and optically thick domains in mixed films of c-Fos with PIP2. This lipid has a strong influence on the surface pressuredependent points for the reorganization of c-Fos, facilitating the acquisition of the low-pressure protein molecular arrangement and impairing the high-pressure one. By contrast, PC has no influence on the former but abolishes the latter.9 In this work, we have analyzed in detail the surface thermodynamics of binary mixed films of c-Fos with PIP2 and with PC in different proportions under precisely controlled conditions regarding surface free energy, intermolecular packing, lipid-protein composition, and surface electrostatics. The dissection of thermodynamic parameters provides insight into the energetics of lipid-protein interactions in our system. Materials and Methods Compression-expansion isotherms were obtained for synthetic dilauroylphosphatidylcholine (PC) or porcine brain PIP2 (Avanti Polar-Lipids Inc., Alabaster, AL) mixed with recombinant c-Fos (purified by affinity chromatography as described in ref 8, assumed apparent molecular mass 57 kDa) in different proportions ranging from 0.01 to 0.1 mol fraction of protein. The aqueous protein stocks were in buffer (pH 7.9) containing 60 mM imidazole/8 M urea. Mixed lipid-protein monolayers were spread from premixed solutions as previously described.9 All measurements were performed at room temperature in an 80 cm2 compartment of a specially designed circular Monofilmmeter Teflon trough (Mayer Feintechnik, Germany) filled with 74 mL of 145 mM NaCl. Details of the equipment used were described previously.9 The surface pressure (π) and surface potential (∆V) were automatically recorded as a function of the mean molecular area during compression-expansion at a speed of 0.45-0.60 nm2‚molecule-1‚min-1 for at least two cycles. The reproducibility of duplicate experiments was within the maximum S.E.M of (1 mN/m for surface pressure, (30 mV for surface potential, and (0.04 nm2 for molecular area. Monolayer stability was proved by negligible (less than 10%) variations in the limiting mean molecular area and dipole potential at the collapse pressure in successive compression-expansion cycles. Similar to the behavior of other lipids,12 the pure lipids showed no hysteresis (less than a 15% gap between compression-expansion), whereas the pure protein exhibited considerable hysteresis represented by a free-energy gap of about 5 kcal/mol between the free energy of compression and expansion.8 All thermodynamic quantities were derived from the measured surface pressures, mean molecular areas, and dipole potentials and from the theoretical (ideal) values calculated for the mixtures at each proportion using the corresponding values of the pure components.9 The functions analyzed were the excess free energy, the entropy, and the enthalpy of mixing; similar parameters were obtained regarding hysteresis on the basis of the film behavior under compression and expansion for binary monolayers of PC or PIP2 containing different proportions of c-Fos. Enthalpic-entropic compensations in the lipid-protein films were deduced using a similar approach to that described elsewhere.13 Thermodynamic Functions of Mixing. Free Energy of Mixing for a Binary Monolayer. ∆Gm is derived13-15 from the addition of the ideal free energy of mixing (eq 2) and the measured excess free energy of mixing (eq 3): (12) Lusted, D. Biochim. Biophys. Acta 1973, 307, 270-278. (13) Bellomio, A.; Oliveira, R. G.; Maggio, B.; Morero, R. D. J. Colloid Interface Sci. 2005, 285, 118-124.
Borioli and Maggio ∆Gm ) ∆Gim + ∆Gex m
(1)
The ideal free energy of mixing ∆Gim is ∆Gim ) RT(X1 ln X1 + X2 ln X2)
(2)
where X1 and X2 are the mol fractions of components 1 and 2. The excess free energy of mixing ∆Gex m is computed as the difference between the area under the experimental compression or expansion π-area isotherms of a given mixture and the area under the ideal isotherm, integrated between π1 ) 1 and π2 ) 40 mN/m. This avoids the rather variable gaseous region of the isotherm and inconsistencies that can arise from irreproducibilities related to variations of compressibility in films approaching collapse.15,16 Thus, the excess free energy of mixing is expressed as [∆Gex m )
∫
π2
π1
(Am - X1A1 - X2A2) dπ]X
(3)
where π1 ) 1 mN/m, π2 ) 40 mN/m, Am is the mean molecular area of the experimental mixed monolayer, and Ai ) (X1A1 + X2A2) is the mean molecular area of an ideally mixed film at each surface pressure and composition. Entropy of Mixing for a Binary Monolayer. ∆Sm is derived from the addition of the ideal entropy of mixing (eq 5) and the excess entropy of mixing (eq 6): ∆Sm ) ∆Sim + ∆Sex m
(4)
The ideal entropy of mixing ∆Sim is ∆Sim ) -R(X1 ln X1 + X2 ln X2)
(5)
where X1 and X2 are the mol fractions of components 1 and 2; by assuming that the excess entropy of mixing ∆Sex m is due solely to ex configurational entropy (∆Sex m ) ∆Scf ) resulting from the isothermal area condensation/expansion,13 it can be derived17 from the measured mean molecular areas of experimental and ideal isotherms as
[
∆Sex m ) R ln
]
Am Ai
(6)
π,X
It should be pointed out that several entropic contributions from different molecular factors are intrinsically included within the experimentally measured mean molecular area and its variations with film composition and surface pressure. ∆Sex m for a mixture with a fixed proportion of components is represented by the sum of individual configurational entropy terms taken at discrete lateral pressure steps; from these, the entropic contribution to the excess free energy of mixing (T∆Sex m ) can be obtained. Enthalpy of Mixing for a Binary Monolayer. ∆Hm is derived from the addition of the ideal enthalpy of mixing (equal to zero in an ideal mixture) and the experimental excess enthalpy of mixing ∆Hex m: ∆Hm ) ∆Him + ∆Hex m
(7)
The second term in eq 7 is obtained by adding the experimental values of excess free energy and the corresponding entropic contribution:13 ex ex [∆Hex m ) ∆Gm + T∆Sm ]π,X
(8)
(14) Sykora, J. C.; Neely, W. C.; Vodyanoy, V. J. Colloid Interface Sci. 2004, 276, 60-67. (15) Maget-Dana, R. Biochim. Biophys. Acta 1999, 1462, 109-140. (16) Maggio, B. Chem. Phys. Lipids 2004, 132, 209-224. (17) Atkins, P. W. Physical Chemistry; W. H. Freeman: New York, 1990; Chapter 4.
Storage Capacity of c-Fos and Phospholipids
Langmuir, Vol. 22, No. 4, 2006 1777
For a mixture with a fixed composition, the experimental excess enthalpy of mixing is represented by the sum of the enthalpy terms taken at each lateral pressure interval. The excess enthalpy of mixing may be independently calculated by assuming a simple model that considers pair London-van der Waals and dipole-dipole interaction energies: ex ex [∆Hex m ) ∆HvdW + ∆Hµ ]π,X
(9)
The first term in eq 9 can be calculated on the basis of a model described for the interaction between two parallel cylinders with radii R1 and R2 and height L, separated by a distance D18
[
∆Hex vdW ) -
)]
(
R1R2 AL 3/2 R + R 1 2 12x2D
(10)
π,X
where the Hamaker constant A is taken as 14.448 kcal/mol. The value for L, the interfacial thickness,18 is taken as twice the diameter of a protein molecule with a limiting mean molecular area (cylindrical cross section perpendicular to the interface),8,19,20 measured for c-Fos at each lateral surface pressure. The distance D of maximum molecular proximity is taken as twice the diameter of a CH2 group (3.5 × 10-10 m). The second term in eq 9 accounts for the repulsive interaction energy between two equal parallel resultant dipoles perpendicular to the interface21
[
∆Hex µ )
14.3µ2 4πor3
]
kcal/mol
(11)
π,X
where r ) 2 xA/π; A is the mean molecular area at each surface pressure, and µ is the resultant dipole moment along the direction perpendicular to the interface, both of which are experimentally measured parameters. The value of unity was taken as the dielectric constant.22 The calculated total excess enthalpy of mixing for a mixture with a fixed proportion is the sum of the enthalpy terms taken at each lateral pressure step. Thermodynamic Functions of Hysteresis. Ideal films show no his his hysteresis, so ∆Ghis i ) 0, ∆Si ) 0, and ∆Hi ) 0. his The free energy of hysteresis ∆Gm , the configurational entropy of hysteresis ∆Shis m (from which the entropic contribution to the free energy of hysteresis T∆Shis m can be obtained), and the enthalpy of hysteresis ∆Hhis m are defined by eqs 12-14, respectively, in a manner similar to that described for the thermodynamic functions of mixing and are obtained from the experimental compression and expansion isotherms. It is important to notice that hysteresis was consistently observed through at least two compression-expansion cycles and is thus reversible. ∆Ghis m ) ∆Gexpans - ∆Gc
[
∆Shis m ) R ln
]
Aexpans Ac
(12) (13)
π,X
his his ∆Hhis m ) ∆Gm + T∆Sm
(14)
All excess thermodynamic functions of mixing are derived from the difference between the values corresponding to the real mixtures (18) Israelachvili, J. N. Intermolecular and Surface Forces; Academic Press: London, 1991. (19) Carrizo, M. E.; Miozzo, M. C.; Maggio, B.; Curtino, J. A. FEBS Lett. 2001, 509, 323-326. (20) Fidelio, G. D.; Maggio, B.; Cumar, F. A. Biochim. Biophys. Acta 1986, 854, 231-239. (21) Gabler, R. Electrical Interactions in Molecular Biophysics; Academic Press: New York, 1978. (22) Gaines, G. L. Insoluble Monolayers at Liquid-Gas Interfaces; WileyInterscience: New York, 1966; pp 136-207.
Figure 1. Compression-expansion behavior of monolayers of c-Fos with PC or PIP2. Plots show the surface pressure-mean molecular area compression and expansion isotherms of mixtures of c-Fos with PC (panel A) or PIP2 (panel C) and the surface compressional modulus-mean molecular area for the same films under compression and expansion (panels B and D, respectively). The mol fraction of protein was 0.05 in both cases. The insets in panels A and C show the variation of the mean molecular area as a function of composition (mol fraction of c-Fos in the mixture) for films under compression (upper inset) or expansion (lower inset); the insets in panels B and D show the variation in mean dipole potential/molecule as a function of the film composition for films under compression (upper insets) or expansion (lower insets). Symbols in the insets correspond to surface pressures of 10 (b) and 30 mN/m (9). All dashed lines represent the corresponding ideal behavior. The hysteresis is reversible, with less than an 8% difference between successive cycles. The curves corresponding to the second cycle have been omitted for clarity. and the ideally mixed films, as a reference. For the thermodynamic functions of hysteresis, the difference is taken between the values of expansion and those of compression, as a reference.
Results and Discussion Compression-Expansion Isotherms. The compressionexpansion surface pressure-mean molecular area isotherms and the packing-dependent in-plane elasticity of binary equimolar mixtures of c-Fos with PC (Figure 1A, B) and with PIP2 (Figure 1C, D) are examined. The corresponding deviations of the mean molecular area and dipole potential/molecule from that of ideally mixed films are shown at two characteristic surface pressures (insets) taken below and above the first protein reorganization point that is indicated by the variation of the in-plane elasticity. The previously reported expansion-hyperpolarization (for mixed films with PC) and the condensation-depolarization (for mixed films with PIP2) occurring during compression are clearly evident.9,11 The behavior of protein-containing films under expansion denotes the large hysteresis in both the mean molecular area (Figure 1) and dipole potential/molecule (insets in Figure 1B,
1778 Langmuir, Vol. 22, No. 4, 2006
Figure 2. Molecular packing of phospholipid/c-Fos films with different proportions. The surface pressure is shown as a function of the mean molecular area for pure phospholipids (curve a), pure c-Fos (curve d), and mixed films with mol fractions of c-Fos of 0.03 (curve b) and 0.06 (curve c). The panels show compression (upper panels) or expansion (lower panels) isotherms. Left panels correspond to PC/c-Fos mixtures, and right panels correspond to PIP2/c-Fos mixtures.
D). The alterations of the mean molecular parameters of the individual components with respect to the ideal mixed films of c-Fos with PIP2 are of the same quality as for compression: area condensation and dipole potential depolarization are evidenced by the negative deviations from ideal behavior (cf. insets in Figure 1C, D). By contrast, the quality of deviations from ideality observed in the films with PC is different. Positive deviations in area under film compression change to area condensation under expansion, and although hyperpolarization is reduced under expansion, PC fails to induce interfacial depolarization (insets in Figure 1 A, B). The local molecular interactions of the protein with each phospholipid are not the same, as indicated by the different deviations of the surface parameters from ideal behavior. The excess free energy of compression11 is negative in mixtures with PIP2 but positive with PC, indicating overall thermodynamically favorable and unfavorable interactions, respectively. Therefore, the information content (see below) stored under compression that is evidenced as hysteresis in the two systems must be derived from different enthalpic-entropic balances and/ or energy sources. There is a clear dependence of compression and expansion film behavior on the proportion of protein. Figure 2 illustrates the compression (upper panels) and expansion (lower panels) of mixtures with PC (left panels) or PIP2 (right panels). It can be seen that, with both phospholipids and independently of the protein content, all of the mixed films retain a similar condensed molecular organization (and dipole potential, not shown for clarity) and high in-plane elasticity (obtained from the slopes of the curves9,22)
Borioli and Maggio
Figure 3. Excess thermodynamic functions for mixed monolayers of c-Fos with PC or PIP2. Data correspond to the experimental excess free energy of compression (curve 1); the experimental configurational entropy contribution T∆Sex m to the excess free energy of compression (curve 2); the experimental enthalpy contribution to the excess free energy of compression (curve 3); the calculated enthalpy contribution to the excess free energy of compression (curve 4); and the calculated excess free energy of compression based on the calculated enthalpy contribution (curve 5). The values are shown as a function of the film composition for mixed films of c-Fos with PC (upper panel) and with PIP2 (lower panel). For the calculated functions, see Materials and Methods.
that was acquired after the surface pressure increased beyond the high-pressure reorganization point. Only at relatively high protein proportions and below surface pressures of about 10-15 mN/m do the films show some increased intermolecular packing and become less elastic under expansion (lower panels in Figure 2). Excess Thermodynamic Functions. Enthalpic-entropic compensations could account for the differences in the excess free energy of compression previously reported in mixed films of c-Fos with PIP2 and with PC.11 To investigate this possibility, we analyzed the contribution to the excess free energy derived from the variation of the configurational entropy due to the changes in mean molecular packing with respect to ideal behavior. Because in an ideally mixed monolayer intermolecular interactions are nonexistent, the excess enthalpy is null, and the free energy of mixing is derived entirely from the entropy of mixing; in the range of compositions studied, the ideal free energy of mixing is below 0.2 kcal/mol. Clearly, there are enthalpic-entropic compensations in these mixtures so as to account for the sign and magnitude of the excess free energy of mixing, depending on the protein content. Figure 3A shows that the unfavorable (positive) excess compression free energy of mixing in mixed films of c-Fos interacting with PC is partially compensated for by a favorable (positive) configurational entropy term; this is brought about by the large mean molecular area expansion reflecting increased configuration possibilities with respect to the ideal mixture. This is similar to the behavior of a small peptide with egg PC.13 The opposite
Storage Capacity of c-Fos and Phospholipids
occurs for mixtures with PIP2 (Figure 3B), for which the favorable (negative) excess compression free energy of mixing is partially counterbalanced by an entropically unfavorable (negative) energy term contributed by the configurational entropy because of the ordering effect reflected by the mean molecular area condensation compared to that of the ideally mixed film. Thus, the experimental excess compression enthalpy that can be derived from the measured free energy and entropy excess functions indicates the existence of energy terms involving molecular interactions that, on balance and compared to ideal behavior, must be thermodynamically unfavorable (positive) or favorable (negative) in the mixed films of the protein with PC and PIP2, respectively, so as to exceed the entropy terms in each case. No further information about these energy terms can be obtained with the experimental setup employed unless a model is used to gain insight into the effects occurring at the molecular level. As a first approximation to the possible origin of the enthalpic contributions that are implicitly related to molecular interactions, we used a simple, black-box-type model that combines distance-dependent attractive and repulsive pair interactions in an interaction potential function to obtain a calculated excess compression enthalpy. At the simplest level, the model assumes that the interactions between two neighboring and equal (average) molecules (with mean cross sections and overall dipole moment perpendicular to the interface as determined by the measured mean molecular area and mean dipole potential/ molecule at each surface pressure and at each composition) are nonretarded, additive, and result solely from van der Waals effects. The model also assumes that the overall resultant dipole-dipole interactions result from attractive and repulsive terms of the pair potential function18 (Materials and Methods). In this black-box approach, all other possible contributions such as those derived from charge-charge interactions, reorientation of water dipoles, hydration changes associated with the hydrophilic portions of the molecule, molecular tilting, and so forth are assumed to be already incorporated and balanced, hidden within the phenomenological values of the experimentally determined parameters of the surface pressure-dependent mean molecular area and dipole potential/molecule. Despite the model simplicity and probably because of a complex compensation of several molecular effects, the thermodynamic functions calculated on this basis appear to describe the variations of the excess compression enthalpies satisfactorily; this leads to roughly coincident calculated and measured excess compression free energies (Figure 3). Another interesting observation emerges from a close inspection of the variation of the experimentally obtained or calculated excess compression enthalpy (∆Hex m ) and of the configurational entropy contribution to the excess compression free energy (T∆Sex m ) with the mean molecular area in the mixed films in different proportions (Figure 4). The variation with packing of these functions for the mixed film of c-Fos with PIP2 shows defined features reflecting the relative changes in favorable excess enthalpy (negative ∆Hex m ) and unfavorable entropy contributions (negative T∆Sex m ) due to the composition-dependent protein reorganization affected by PIP211 at low and high surface pressures (Figure 4B, D). In the ranges of mean molecular area corresponding to the low-pressure protein-reorganization point, the film energetics indicates that the interactions become less enthalpically favored but entropically facilitated (lines curving toward less-negative values over the range of relatively large mean molecular area) with respect to neighboring packing ranges at all compositions; however, a protein proportion above about 4 mol % is required to induce similar thermodynamic changes in the more closely packed range of mean molecular areas
Langmuir, Vol. 22, No. 4, 2006 1779
Figure 4. Molecular packing- and composition-dependent variation of the excess enthalpy and entropy contribution in mixed monolayers of c-Fos with PIP2 and PC. The calculated enthalpy contribution to the excess free energy of compression is shown for mixed films of c-Fos with PC (panel A) or with PIP2 (panel B) as a function of the mean molecular area. The calculated configurational entropy contribution T∆Sex m to the excess free energy of compression is shown for mixed films of c-Fos with PC (panel C) or with PIP2 (panel D). The numbers on the curves indicate the mol % of c-Fos in the films. The insets in panels C (for mixed films with PC) and D (for mixed films with PIP2) show the variation with the mean molecular area of the contributions to the experimental enthalpy from the electrostatic (dipole-dipole) interaction energy term (Materials and Methods) for films with the mol % of c-Fos indicated on the curves.
corresponding to the high-surface-pressure protein reorganization. This behavior suggests less-favorable interactions and surface organization (relative to neighboring packing ranges) occurring in concert with the protein-reorganization points that become more important as the protein content in the mixture increases. The inset in Figure 4D shows that the variation of the electrostatic interactions (which are calculated on the basis of an independently measured experimental parameter, namely, the dipole potential) reflects an increasingly favorable dipolar matching in the lowpressure protein-reorganization point (over the range of large mean molecular areas), relative to the neighboring packing ranges, whereas an increasingly closer packing gradually opposes this trend, especially in films with a high protein proportion. However, the variation with packing of the same functions for the mixed films of c-Fos with PC (Figure 4A, C) reveals only a rather monotonic change reflecting unfavorable interactions (positive ∆Hex m ) that cannot be fully compensated for by a favorable entropic contribution (positive T∆Sex m ). Comparatively, the energetics in these films indicates, not surprisingly, that the surface organization gradually becomes more enthalpically favored (decreasing ∆Hex m ) but entropically unfavored (decreasing T∆Sex m ) under compression because of increasingly unfavorable electrostatic interactions (inset in Figure 4C).
1780 Langmuir, Vol. 22, No. 4, 2006
Figure 5. Hysteresis thermodynamic function for the interactions of c-Fos with PC and PIP2. The free energy of compression (circles), free energy of expansion (squares), and free energy gap (free energy of hysteresis) between compression and expansion (open bars) is shown as a function of the film composition for mixed films of c-Fos with PC (panel A, white symbols) and PIP2 (panel B, black symbols). Panel C shows the comparison of the experimental excess free energy of mixing and the free energy of hysteresis for the mixed films with c-Fos at the different proportions indicated in the abscissa as follows: experimental compression excess free energy of mixing (squares), experimental expansion excess free energy of mixing (triangles), and free energy of hysteresis (circles) for films with PC (white symbols); experimental compression excess free energy of mixing (squares), experimental expansion excess free energy of mixing (triangles), and free energy of hysteresis (circles) for films with PIP2 (black symbols). Please note that the sign of the free energy of hysteresis for films with PC is negative, but is plotted with positive values to facilitate comparison.
Thermodynamics of Hysteresis. The thermodynamic functions for hysteresis, derived from the difference between the behavior under expansion with respect to compression, are obviously negative in all cases and indicate the retention (negative ∆Ghis m ) of a considerable amount of free energy during compression. This energy storage remains the same (within 8% variation) in successive cycles (not shown), indicating the reversibility of hysteresis. Figure 5A, B and Figure 6 show that this amount increases slightly with the protein proportion in the films. Interestingly, the values of the thermodynamic functions for hysteresis are quite similar (Figures 5 and 6) in the mixed films of c-Fos with either PC or PIP2 despite the entirely different behavior of the lipid-protein films with both lipids in terms of interactions and excess thermodynamic functions of compression. This is possible only if the hysteresis is due to the storage of similar amounts of energy acquired during the compression of the film but derived from different sources in the mixed films with PIP2 and with PC.
Borioli and Maggio
Figure 6. Hysteresis thermodynamic functions in mixed monolayers of c-Fos with PC or PIP2. The experimental free energy of hysteresis (O), the experimental configurational entropy contribution T∆Sex m to the free energy of hysteresis (0), the experimental enthalpy contribution to the free energy of hysteresis (4), the calculated enthalpy contribution (see Materials and Methods) to the free energy of hysteresis (3) and the calculated free energy of hysteresis on the basis of the calculated enthalpy contribution ()) are plotted. The values are shown as a function of the film composition for mixed films of c-Fos with PC (upper panel) and with PIP2 (lower panel).
Performing similar calculations of intermolecular interaction energies to those described above and comparing the values for real and ideal mixtures but using the values of expansion relative to compression, we find that the following pattern emerges for hysteresis (Figures 5 and 6): in mixed films with PIP2, the energy stored as the free energy of hysteresis represents the balance of favored interactions that enthalpically exceed the contributions provided by the unfavorable ordering (involving condensed packing and dipolar matching) reflected in the configurational entropy term (Figure 6). As a consequence, a spontaneous thermodynamically favored intermolecular organization is acquired under compression in which excess compression free energy provided by the favorable interactions is stored and subsequently released with hysteresis. Note that this amount is similar in magnitude for the excess expansion free energy and the free energy of hysteresis (Figure 5C). However, in mixed films of c-Fos with PC the thermodynamic functions indicate a different origin of the free energy of hysteresis, which can be accounted for by the favorable contributions reflected by the positive configurational entropy term (increased molecular area expansion aided by hyperpolarized dipole repulsion) and by external energy provided by the mechanically imposed area reduction during compression (positive excess compression free energy) that exceeds the excess compression enthalpy and that was employed to pack the molecules overcoming the enthalpically unfavorable interactions (Figure 6A). It is this excess energy that
Storage Capacity of c-Fos and Phospholipids
Langmuir, Vol. 22, No. 4, 2006 1781
films with different protein proportions. It increases monotonically when driven by compression, approaching values near unity only at the closest packing. By contrast, the mixed films of c-Fos with PIP2 reflect entropy information content that is always above unity (Figure 7, lower panel) and depends on the protein content. Their variation further indicates that along the packing ranges corresponding to the low-pressure protein reorganization point some information content is lost while a relatively small transient increase in entropy information occurs in the ranges of mean molecular area corresponding to the high-pressure protein reorganization point, especially in films with higher proportions of c-Fos.
Conclusions
Figure 7. Molecular packing-dependent variation of the entropy information content in mixed monolayers of c-Fos with phospholipids. The entropy information entropy content as a function of the mean molecular area during compression is shown for mixed films of c-Fos with PIP2 (upper panel) or with PC (lower panel) containing the mol % of protein indicated by the number on each curve. As a reference, the entropy information content of ideally mixed films is 1.
is stored in this case as the free energy of hysteresis (Figure 5C). As a consequence, the latter is again similar in magnitude to the excess compression and the excess expansion free energy (but, different from PIP2, it is opposite in sign). It is noteworthy that the source of the free energy of hysteresis is reflected in the way the molecules in the mixtures are arranged at the interface:9 a favorable enthalpic driving force induces condensed domains in PIP2/c-Fos mixtures because of the close interaction between lipid and protein molecules; however, molecular expansion and dipole repulsion lead to homogeneous films in mixtures with PC, consistent with the favorable entropic driving force. On the basis of entropy information theory,23 the variations in thermodynamic contributions from entropy can be interpreted as reflecting changes in information content related to the organization of the molecular system. Figure 7 illustrates composition and molecular packing-dependent variations in entropy information content; the intermolecular organization of the ideally mixed films is taken as a reference to which an entropy information content of 1 is assigned (zero excess information). Under this interpretation, the entropy information content of mixed films with PC (Figure 7, upper panel) is always less than that of the ideally mixed molecules in films of the same composition, and its values and variation are rather similar for (23) Moore, W. J. Physical Chemistry, 4th ed.; Prentice Hall: Englewood Cliffs, NJ, 1972; Chapter 5.
The behavior of lipid/c-Fos mixtures deviates from ideality depending on the lipid, on the history of the system (Figure 1), and on the proportion of protein in the film (Figure 2). The mixtures show hysteresis, with values (0.5 to 2.5 kcal/mol) that are significantly smaller than that of the pure protein (5 kcal/ mol) but are interestingly similar for each lipid despite the opposite nature of the interactions of the protein (Figure 5A, B). The excess thermodynamic functions show that entropicenthalpic compensations account for the previously reported favorable or unfavorable excess free energies of mixing for PIP2 or PC mixtures with c-Fos, respectively10 (Figure 3). These compensations depend on composition and vary with packing in a complex way in mixtures with PIP2 and monotonically with PC (Figure 4). The contribution of the enthalpic and entropic terms to the excess free energy of mixing differs according to the lipid: enthalpically favorable electrostatic interaction between PIP2 and c-Fos results from dipole matching overcoming entropically unfavorable ordering. The favorable configurational entropy that arises from PC/c-Fos interaction, on the contrary, exceeds the unfavorable enthalpy associated with hyperpolarization (Figure 1 insets). The excess free energy is stored in each case as the free energy of hysteresis (Figure 5C). The entropy information content of the mixtures depends on the lipid, and for mixed films of PC/c-Fos, it is less than that for an ideal mixture at any protein proportion. By contrast, the information content is always higher than the ideal for mixtures with PIP2, and the variations reflect the changes in molecular organization. Our findings indicate that c-Fos-phospholipid interactions involve different enthalpic-entropic compensations that can account for the surface behavior observed. The surface thermodynamics additionally points to the existence of defined modes for the acquisition and release of molecular information, mediated by selective protein-lipid interactions and configurational entropy changes that validate and explain the existence of hysteresis under dynamic compression-expansion. Within this context, the 2D behavior of c-Fos in organized interfaces emerges as a powerful, sensitive, and selective self-organizing molecular device capable of reversible storage and transduction of interfacial structural changes. Acknowledgment. This work belongs to a line of research that was initiated in the laboratory of Dr. Beatriz Caputto. We are profoundly grateful for her unconditional support and valuable contributions to this article. Financial support for this research was obtained from SECyT-UNC and FONCyT. LA0525168
