Hindered Diffusion in Polymer-Tethered Membranes:A Monolayer

present fluorescence recovery after photobleaching experiments (FRAP) on mixed phospholipid- ... In contrast to recent FRAP experiments on polymer- te...
2 downloads 0 Views 107KB Size
Download PDF
Biomacromolecules 2001, 2, 1097-1103

1097

Hindered Diffusion in Polymer-Tethered Membranes:A Monolayer Study at the Air-Water Interface C. A. Naumann,*,†,‡,⊥ W. Knoll,§ and C. W. Frank*,†,| Department of Chemical Engineering, Stanford University, 381 North-South Mall, Stanford, California 94305-5025; Department of Chemistry, Indiana University Purdue University Indianapolis, Indianapolis, Indiana 46202-3274; and MPI fu¨r Polymerforschung Mainz, Ackermannweg 10, 55128 Mainz, Germany Received January 29, 2001; Revised Manuscript Received August 9, 2001

Polymer-tethered phospholipid bilayers, which are based on a phospholipid-lipopolymer mixture, represent a very promising approach to stabilize complex biomimicking composite membranes. Furthermore, they are interesting model systems to study problems of hindered diffusion in two-dimensional liquids. Here, we present fluorescence recovery after photobleaching experiments (FRAP) on mixed phospholipid-lipopolymer monolayers of DMPC and DSPE-EO45 at the air-water interface. In contrast to recent polymer-tethered bilayer experiments where the hydrophobic lipopolymer anchors behaved as immobile obstacles within the fluid phospholipid matrix,1 this paper investigates the influence of mobile lipopolymer obstacles on the lateral diffusion of phospholipids. We found that the lateral diffusion of phospholipids with D ) 7.1 ( 0.5 µm2/s is independent of the lipopolymer obstacle concentration if adjacent polymer chains do not interact with each other. However, the diffusion coefficient of nontethered phospholipids gradually decreases from D ) 7.1 ( 0.5 µm2/s to D ) 3.4 ( 0.1 µm2/s in the case of increasing polymer-polymer interactions based on frictional coupling. This can be understood by a slowing down of the obstacle mobility. While phospholipids still show a significant lateral diffusion as long as the polymer moieties interact with each other only via frictional coupling, they become rather immobile (D ) 0.9 ( 0.1 µm2/s) if lipopolymers form a two-dimensional physical network. Introduction Solid-supported biomembranes, which are able to mimic such important properties of biological membranes as the two-dimensional fluidity of the lipid matrix and the lateral mobility of membrane proteins under nondenaturing conditions, have attracted the interest of many research groups in recent years.2 They not only represent an excellent system for use in the study of membrane-related processes, but are also relevant for some technological applications, e.g., biosensors.2-4 While lipid properties related to structure, thermodynamics, or dynamics can be mimicked fairly well using a solid-supported phospholipid bilayer with a lubricating water layer of 10-20 Å,5-8 proteins, especially transmembrane proteins embedded in the lipid matrix, require more complex membrane architectures. This is because the solid substrate suppresses protein mobility and might even result in protein denaturation.9-11 To prevent this, a hydrophilic polymer cushion can be introduced between the bilayer and solid substrate.12,13 While these composite materials show promising decoupling of the bilayer from the solid substrate, they lack sufficient stability if not stabilized via attractive forces at both the polymer-substrate and polymer-bilayer * Corresponding authors. † Stanford University. ‡ Indiana University Purdue University Indianapolis. § MPI fu ¨ r Polymerforschung Mainz. | E-mail: [email protected]. ⊥ E-mail: [email protected].

interfaces.14 While Israelachvili and co-workers have recently introduced an approach that leads to a higher stability via attractive electrostatic interactions,15,16 Wagner and Tamm17 and our group have pursued the stabilization via controlled covalent tethering, where the covalent tethers are immobile because the polymer layer is also covalently attached to the solid substrate.1,12,18 Because the latter approach does not rely on electrostatic interactions, it has the advantage of being independent of salt strength and pH. This becomes extremely important if different proteins are incorporated without changing membrane properties, because each protein typically requires specific environmental conditions. On the other hand, it is important to understand how the covalent tethers change the diffusive transport behavior. Fluorescence recovery after photobleaching (FRAP) experiments on polymer-tethered phospholipid bilayers have revealed the relationship between tethering density at the polymer-lipid interface and lateral mobility of the bilayer.1 While the lateral diffusion coefficient of D ) 17.7 µm2/s measured at low tethering density (5 mol % lipopolymer) showed that Brownian motion occurs within the bilayer, corresponding values at moderate (10 mol %) and high (30 mol %) tethering densities of D ) 9.7 µm2/s and D ) 1.1 µm2/s, respectively, revealed significant hindrance of diffusion. Interestingly, Wagner and Tamm did not observe the described diffusion behavior on their almost identical membrane architecture.17 Additional measurements, in which the diffusional properties of the inner and outer leaflets of

10.1021/bm010022t CCC: $20.00 © 2001 American Chemical Society Published on Web 10/09/2001

1098

Biomacromolecules, Vol. 2, No. 4, 2001

the bilayer were measured separately, showed different diffusion properties for both leaflets. Beside their relevance as novel membrane systems, polymer-tethered bilayers allow study of problems in hindered two-dimensional diffusion. In this case, lipopolymers can be seen as obstacles within the phospholipid matrix, with the lateral mobility of nontethered phospholipids dependent on the obstacle size, the obstacle density, and the obstacle mobility. In contrast to recent FRAP experiments on polymertethered bilayers, where the lipopolymers represent immobile obstacles within the phospholipid bilayer, this study investigates the problem of mobile obstacles. We present FRAP experiments on mixed PEG-lipopolymer/phospholipid monolayers at the air-water interface in which we investigate the effect of the lipopolymer molar concentration on the phospholipid mobility. The following questions will be addressed: (1) How is the lateral mobility of nontethered phospholipids dependent on the concentration of polymerdecorated phospholipids (lipopolymers)? (2) What is the relationship between the lateral mobility of nontethered phospholipids and the strength of polymer-polymer interactions among adjacent polymer chains? (3) How does the physical gelation within the polymer moieties discovered recently 19-21 affect the phospholipid diffusion? (4) How is the lateral mobility of nontethered phospholipids influenced if lipopolymers and phospholipids start to microphase separate above their high-film-pressure transition? Materials and Methods The lipopolymer investigated was 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[poly(ethylene glycol) 2000] (DSPE-EO45). The phospholipid studied was 1,2-dimyristoyl3-glycero-phosphocholine (DMPC). Both amphiphiles were purchased from Avanti Polar Lipids (Alabaster, AL). The fluorescence label used, N-(7-nitrobenz-2-oxa-1,3-diazol-4yl)-1,2-dihexadecanoyl-sn-glycero-3-phospho-ethanolamine, triethylammonium salt (NBD-DPPE), was purchased from Molecular Probes (Eugene, OR). The corresponding chemical structures are summarized in Figure 1. Chloroform was used as a spreading solvent for preparing the monolayers at the air-water interface. Milli-Q Water (pH ) 5.5, 18 MΩcm resistivity) was used as a subphase material for all experiments. Fluorescence imaging and fluorescence recovery after photobleaching (FRAP) were performed using a Nikon E800 upright microscope equipped with a CCD camera. To minimize the surface flow of the monolayer, the film balance was completely sealed against the outside using plastic wrap. To prevent temperature gradients within the sample, experiments have been conducted at room temperature. The size of the bleaching spot, which was determined on a completely rigid sample with a 40× objective, was set by the field stop of the microscope to a diameter of 144 µm; the bleaching time was set to 10 s. The fluorescence recovery was observed by switching to a 10× objective. Observation increments were set to 10 or 15 s, and background-corrected intensities of the bleaching spot were determined for each image taken. The normalized fluorescence recovery curves were fitted by

Naumann et al.

Figure 1. Molecular structures of amphiphiles investigated. It shows the phospholipid, DMPC, the fluorescence-labeled phospholipid NBDDMPE, and the lipopolymer, DSPE-EO45.

a single-exponential function providing the intensities at both t ) 0 s, I0, and at t ) ∞, I∞, as well as t1/2, the diffusion time at half-recovery. The lateral diffusion coefficient, D, was then determined from the FRAP curves by using the following equation:22 D ) 0.224r2/t1/2

(1)

with the radius of the bleaching spot, r, and the half-time of the fluorescence recovery, t1/2. This model assumes Brownian motion with an immobile fraction. The immobile fraction (IF) representing immobile molecules was determined via IF ) [1 - I∞] * 100%

(2)

where I∞ is the fluorescence intensity for t ) 8. Unfortunately, it is very difficult to observe a bleaching spot at the air-water interface over a period of time of more than 5 min. Therefore, our FRAP curves do not reach the saturation values providing I∞ and IF directly. Both parameters could only be determined from fitted curves, which makes their absolute values quite erroneous (up to 20% error). Because t1/2 depends on IF, our absolute diffusion values might also show a significant experimental error (5-10%). We emphasize, however, that our discussion is mostly based on relative changes among these parameters, which provide much more reliable results. Results Figure 2 compares FRAP curves of a DMPC monolayer at the air-water interface (T ) 21 °C) at different areas per phospholipid molecule. At Alipid ) 90 Å2, the DMPC monolayer shows values of D ) 16.1 µm2/s and IF ) 5%, which are in good agreement with analogous measurements on a DPPC monolayer (D ) 17.3 µm2/s and IF ) 3%) for the same area. In contrast, the mobility of the DMPC monolayer is significantly reduced at Alipid ) 65 Å2, as the values of D ) 6.5 µm2/s and IF ) 20% indicate. This reduction in D to 38% of the value found at Alipid ) 90 Å2 is in good agreement with FRAP experiments on DLPC

Polymer-Tethered Membranes

Figure 2. Fluorescence recovery curves comparing phospholipid monolayers at different areas per lipid of Alipid ) 65 Å2 and Alipid ) 90 Å2. While DMPC and DPPC monolayers show a very similar fluorescence recovery at Alipid ) 90 Å2, the DMPC mobility obtained at Alipid ) 65 Å2 is significantly different. Note that the monolayer behavior at Alipid ) 90 Å2 is identical to that of a polymer-tethered bilayer at 5 mol % lipopolymer concentration (Alipid∼65 Å2) which is also included. This indicates that experimental data obtained from monolayer studies cannot be extrapolated to bilayer systems in a straightforward way.

monolayers at the air-water interface, where D at Alipid ) 65 Å2 was about 40% of that found at Alipid ) 90 Å2.23 As Figure 2 also shows, a polymer-tethered DMPC bilayer shows at low tethering density of 5% with D ) 17.7 µm2/s and IF ) 0% almost the same fluorescence recovery behavior as a DMPC monolayer at Alipid ) 90 Å2.1 We note that the latter value is similar to results obtained by single-moleculeimaging experiments on free-standing membranes with D ) 20 µm2/s.24 Figure 3 shows Langmuir isotherms of DSPE-EO45/ DMPC mixed monolayers at different lipopolymer molar concentrations of 2, 5, 10, 20, and 30 mol %. The isotherms represent the relationship between film pressure, π, and area per lipopolymer, Alipo. All π-A isotherms show a characteristic low-film-pressure transition at about 10 mN/m, which is related to the desorption of PEG chains from the airwater interface into the subphase.25 To evaluate the effect of the tether mobility on the lateral diffusion of DMPC molecules, FRAP experiments have been conducted at different lipopolymer-phospholipid molar ratios, thereby keeping the area per phospholipid fixed at Alipid ) 65 Å2 (circles in Figure 3). Alipid and Alipo are related to each other via Alipo ) Alipid(n + 1) with n being the lipopolymerphospholipid molar ratio. At the lipopolymer molar concentration of 30 mol %, a high-film-pressure transition can be found around Afb ) 230 Å2 (see insert) which is related to a first-order-like alkyl chain condensation among adjacent lipopolymer molecules.20,26 Recent surface rheology experiments, furthermore, revealed a physical gelation among polymer chains at about Arheo ) 170 Å2.20 To investigate how the high-film-pressure transition and the physical gelation affect the lateral mobility within the monolayer, we performed FRAP experiments for 30 mol % lipopolymer at different areas per lipopolymer (squares in Figure 3). Figure 4 compares FRAP curves of DSPE-EO45/DMPC mixtures at different lipopolymer molar concentrations of 0, 2, 5, 10, 20, and 30% where the area per lipid is kept

Biomacromolecules, Vol. 2, No. 4, 2001 1099

Figure 3. Pressure-area isotherms of phospholipid-lipopolymer mixtures at the air-water interface at different lipopolymer molar concentartions of 2, 5, 10, 20, and 30 mol %. The marker represents experimental conditions chosen for FRAP experiments. These conditions are characterized by a constant value of the area per lipid of Alipid ) 65 Å2 (circles). Note that all FRAP experiments have been performed well above the low-film-pressure transition at around 10 mN/m, which is related to a submerging of polymer chains from the air-water interface into the aqueous subphase. The insert shows the region around the high-film-pressure transition found at a lipopolymer molar concentration of 30 mol % and FRAP experiments performed at this lipopolymer concentration (squares). Both the characteristic areas per lipopolymer at the high-film-pressure transition, Afb, and at the physical gelation transition, Arheo, are indicated.

Figure 4. Fluorescence recovery curves of phospholipid-lipopolymer monolayers at different lipopolymer molar concentrations of 2, 5, 10, 20, and 30 mol %. For comparison, results from the DMPC monolayer at Alipid ) 65 Å2 are also included. While the fluorescence recovery remains unchanged for lipopolymer molar concentrations of 0, 2, 5, and 10 mol %, it begins to change between 10 and 20 mol % and is significantly slower at 30 mol %.

constant at Alipid ) 65 Å2. The data obtained are summarized in Table 1. We distinguish between two different situations. At tethering concentrations less than 10%, both the diffusion coefficient, D, and the immobile fraction, IF, show no significant dependence on the tethering density. The values determined are D ) 7.1 ( 0.5 µm2/s and IF ) 23 ( 3%. In contrast, a significant dependence of D and IF on the

1100

Biomacromolecules, Vol. 2, No. 4, 2001

Naumann et al.

Table 1. Diffusion Coefficients, D, and Immobile Fractions, IF, of Phospholipid-Lipopolymer Monolayers at Different Lipopolymer Molar Concentrationsa DSPE-EO45 molar concentration [mol %]

diffusion coeff D (µm2/s)

immobile fraction IF (%)

0 2 5 10 20 30

6.6 7.3 7.6 7.2 4.5 3.4

20 22 26 24 20 39

a Note, the area per lipid, A lipid, is kept constant for all experiments at Alipid ) 65 Å2.

Figure 6. Diffusion coefficients, D, and immobile fractions, IF, as a function of the area per lipopolymer, Alipo, obtained from FRAP data shown in Figure 4 and Figure 5. The lateral diffusion can be characterized by the following regions: (I) independence of Alipo; (II) moderate decrease of lateral mobility with decreasing Alipo; (III) sharp decrease of lateral mobility with decreasing Alipo. While the transition between regions I and II occurs at a specific area per lipopolymer Alipo*, that between regions II and III is characterized by a specific value Alipo**. Note that open and filled symbols represent different experimental conditions regarding Alipid. While Alipid is kept constant in one case (filled symbols), it is varied in the other one (open symbols).

Figure 5. Fluorescence recovery curves of phospholipid-lipopolymer monolayers at a lipopolymer molar concentration of 30 mol % at different areas per lipopolymer, Alipo, of 146, 180, 196, and 230 Å2. They provide information about the phospholipid mobility around the high film pressure transition, Afb ) 230 Å2, and the physical gelation at Arheo ) 170 Å2. Table 2. Diffusion Coefficients, D, and Immobile Fractions, IF, of Phospholipid-Lipopolymer Monolayers at a Lipopolymer Molar Concentration of 30 Mol % and Changing Areas Per Lipopolymer, Alipo area per lipopolymer Alipo (Å2)

diffusion coeff D (µm2/s)

immobile fraction IF (%)

230 196 180 146

3.9 3.4 2.3 0.9

30 38 60 81

tethering density can be found for lipopolymer molar concentrations ranging from 10 to 30 mol %. Figure 5 shows FRAP curves of DSPE-EO45/DMPC mixtures at a lipopolymer molar concentration of 30%. In this case, the monolayer is investigated at different areas per lipopolymer, Alipo, of 146, 180, 196, and 230 Å2 (cf. inset of Figure 3). The data obtained are summarized in Table 2. Figure 6 represents diffusion coefficients, D, and immobile fractions, IF, as a function of the area per lipopolymer, Alipo, obtained from FRAP data shown in Figure 4 and Figure 5. The lateral diffusion can be characterized by the following regions: (I) independence of Alipo; (II) moderate decrease of lateral mobility with decreasing Alipo; (III) sharp decrease of lateral mobility with decreasing Alipo. The immobile fraction is not affected by the area per lipopolymer until a value of Alipo ) 230 Å2 is reached, which represents the region around the high-film-pressure transition. While the

filled markers represent experimental data obtained from FRAP curves in Figure 4 (vary lipopolymer molar concentration, thereby keeping Alipid ) 65 Å2), open markers are determined from FRAP curves in Figure 5 (keep lipopolymer molar concentration at 30 mol %, thereby varying the area per lipopolymer). Discussion Recent FRAP experiments on polymer-tethered phospholipid bilayers, where some of the phospholipids within the inner leaflet of the bilayer are covalently attached to the polymer cushion (lipopolymers), revealed a significant slowing down of the lateral bilayer mobility if the tethering density was increased.1 These experimental findings point toward a general problem that should be considered for the design of more complex bioartificial membranessthe tradeoff between stability of the whole system and lateral mobility of the phospholipids in the bilayer. At the same time, such a tethered membrane represents a model system to study problems of hindered diffusion. Besides the frictional coupling among different layers,8,27 diffusion properties are also dependent on density, mobility, and size of obstacles within the bilayer.28,29 In polymer-tethered bilayers described recently,1 the hydrophobic moieties of the tethered phospholipids may be considered as immobile obstacles within the two-dimensional phospholipid matrix. (Note that the polymer moiety of lipopolymers is in this case covalently attached to the solid substrate!) In contrast, polymerdecorated phospholipids in mixed lipopolymer-phospholipid monolayers at the air-water interface represent the case of mobile obstacles within the phospholipid matrix. Our hypothesis is that the obstacle mobility is correlated to the strength of polymer-polymer interaction among adjacent polymer chains in the subphase. At this point, we note that data obtained from a bilayer in an aqueous environment and those from a monolayer at the

Polymer-Tethered Membranes

air-water interface should be compared with great care because interfacial conditions are quite different. Therefore, it is not straightforward to mimic bilayer conditions using monolayer experiments. This becomes obvious from Figure 2, where FRAP curves of a DMPC monolayer at Alipid ) 65 Å2 and Alipid ) 90 Å2 are plotted together with a FRAP curve representative for the outer leaflet of a polymer-tethered bilayer at low tethering density of 5 mol % lipopolymer. The FRAP behavior of the monolayer at Alipid ) 90 Å2 is almost identical with that of the unperturbed bilayer. On the other hand, it is noteworthy that DMPC has an average area per molecule within the bilayer of about Alipid ∼ 65 Å2,30 where we notice in our experiments a difference in the FRAP behavior between monolayer and bilayer. Such a behavior for an analogue area per molecule indicates different structural properties as a result of different interfacial properties. This is supported by neutron reflectivity experiments on phospholipid monolayers that determined hydration states for phospholipids at the air-water interface, which deviate from those within a bilayer.31 We did most of our experiments on mixed lipopolymer-phospholipid monolayers at Alipid ) 65 Å2 because the corresponding film pressures are located well above the low-film-pressure transitions, as shown in Figure 3. Note that PEG seems to penetrate into the phospholipid monolayer below this transition. Another interesting result in Figure 2 is the immobile fraction of about 20% found at Alipid ) 65 Å2. It is not expected because DMPC is under these conditions supposed to be in the liquid expanded phase (LE). Because IF is in our case a rather error-prone parameter (up to 20%), we do not want to speculate about possible explanations. Effect of Polymer-Polymer Interaction. To evaluate the effect of the obstacle mobility on the lateral diffusion of nontethered phospholipids within the monolayer, we performed FRAP experiments on mixed DSPE-EO45/DMPC monolayers at different lipopolymer molar concentrations, thereby keeping Alipid constant at Alipid ) 65 Å2. In Figure 6, three different regions are found, separated by transitions at Alipo* ) 650 Å2 and Alipo** ) 180 Å2. Note that in this graph we plot diffusion coefficients obtained from Figure 4 (different lipopolymer-phospholipid molar ratios) and Figure 5 (different values of Alipo for 30 mol % liopolymer). We should point out that the two sets of experiments represent qualitatively different experimental conditions. While Alipid is kept constant in the first case (Alipid ) 65 Å2), it is varied in the second one (change of molecular areas at 30mol % lipopolymer). At Alipo > Alipo* (region I in Figure 6), which is relevant for 0 (pure DMPC), 2, 5, and 10 mol % lipopolymer, there is no significant difference in D among different lipopolymer-phospholipid molar concentrations. These samples are characterized by D ) 7.2 ( 0.6 µm2/s and IF ) 23 ( 3%. This behavior is significantly different from findings on polymer-tethered phospholipid bilayers, which show a clear dependence of the phospholipid diffusion on the lipopolymer molar concentration.1 The reason for this discrepancy is probably related to the different obstacle mobility in both systems (immobile in the polymer-tethered bilayer and mobile in the polymer-tethered monolayer). If there is no

Biomacromolecules, Vol. 2, No. 4, 2001 1101

significant frictional coupling between phospholipid and polymer layer, the Saffmann-Delbrueck model predicts similar values of D for both phospholipids and lipopolymers in the case of mobile obstacles.32 But what could be the reason for the change at Alipo* ∼ 650 Å2, the borderline between regions I and II? To obtain a basic understanding, let us consider the case of a polymer chain squeezed into a tube, as calculated by de Gennes.33 In the case of a polymer in a thick tube, the relationship between tube diameter, d, and chain length, L, is as follows: L)

RF2 d

(3)

with the Flory radius, RF for a mushroom-like polymer configuration being RF ) aN3/5

(4)

From the segment length, a ) 3.5 Å, the number of monomer units, N ) 45, and d)2

x

Alipo* ) 28.8 Å π

(5)

we finally obtain RF ) 34.3 Å, L ) 40.9 Å, and L/RF ) 1.36/1. The L/RF ratio tells us that the polymer chains are only slightly stretched. In contrast to ”free” polymers in solution with L/RF ) 1, polymer chains tethered to a surface tend to stretch slightly even in their equilibrium configuration because of space restrictions.34 Therefore, we assume at Alipo* ∼ 650 Å2 that tethered polymer chains start to interact with each other even though they are still not far away from their equilibrium chain configuration found in the case of noninteracting polymer chains. Between Alipo**< Alipo < Alipo* (region II in Figure 6), we observe a decrease of the lateral mobility among DMPC molecules with increasing tethering density. Obviously, there is a correlation between the strength of PEG-PEG interaction and the lateral mobility of DMPC molecules. While the immobile fraction remains rather unchanged with IF ∼ 2025% (only 30 mol % lipopolymer shows a higher value of IF ∼ 39%), the lateral diffusion of nontethered DMPC molecules is significantly changed with D ) 7.2 ( 0.6 µm2/s at 10 mol %, to D ) 4.5 ( 0.6 µm2/s at 20 mol %, and to D ) 3.4 ( 0.6 µm2/s at 30 mol %. One might speculate that the dependence of D on the tethering density is related to the mobility of lipopolymer obstacles or that it is due to a higher frictional coupling between polymer and phospholipid layer. The latter might simply be related to a higher interfacial roughness. On the basis of neutron reflectivity data, Israelachvili and co-workers reported a significant roughening of a PEG lipopolymer monolayer at the airwater interface with smaller Alipo.35 On the other hand, Saxton has theoretically predicted that the lateral mobility of twodimensional liquids, such as phospholipid membranes, is affected not only by the size and the density of obstacles within the fluid matrix but also by their mobility.28,29 But what happens at Alipo**, the borderline between regions II

1102

Biomacromolecules, Vol. 2, No. 4, 2001

Naumann et al.

and III? If we consider again the situation of a polymer squeezed into a tube (in this case a thin tube), the relationship between L and d for a thin tube is33 L ) Na

(da)

2/3

(4)

By using eq 4 for Alipo** ) 180 Å2, which provides d ) 15.1 Å, we obtain from eq 6 L ) 59.3 Å and L/d ) 3.9/1. Thus, the polymer chain is now much more stretched. Still this does not explain the different diffusion behavior between regions II and III. We have already indicated that open and close circles in Figure 6 represent different experimental conditions regarding Alipid. Thus, a qualitative change between both data sets around Alipo** is not surprising. Another, at least, partial explanation could come from recent surface rheology experiments on these monolayers. They revealed a physical network formation at Alipo ∼ 170 Å2, where Alipo is independent of the lipopolymer concentration.20 Because the physical network formation leads to an immobilization of lipopolymer obstacles within the phospholipid-lipopolymer monolayer, the change in diffusion around Alipo** could be related to a gelation-induced change in obstacle mobility. Effect of Lipopolymer Microphase Separation. So far, we have discussed how the phospholipid mobility is affected by polymer-polymer interactions. Recent film balance and surface rheology experiments on DMPC/DSPE-EO45 mixed monolayers at the air-water interface indicated microphase separations among both molecules above the high-filmpressure transition of their π-A isotherms.20 This transition is related to a first-order-like alkyl chain condensation among hydrophobic lipopolymer moieties.26 Lipopolymer aggregation can be seen as a change of the obstacle size within the phospholipid monolayer, which is another critical parameter affecting diffusion properties.28,29 As shown in the insert in Figure 3, FRAP measurements at a lipopolymer molar concentration of 30 mol % represent a monolayer slightly below and above the high-film-pressure transition. While microphase separation seems not to affect the diffusion properties in Figure 6sa linear D-Alipo relationship can be found for 10, 20, and 30 mol % lipopolymersthe immobile fraction is significantly increased between 20 mol % lipopolymer (no microphase separation) and 30 mol % lipopolymer (microphase separation). This is in good agreement with our data obtained for 30 mol % at Alipo ) 230 Å2. In this case, we also observe a significant increase of the immobile fraction to IF ) 30%. We are aware that both D and IF should be coupled parameters. The diffusion behavior at Alipo < Alipo* is, therefore, also caused by the significant immobile fraction. Model of Hindered Diffusion. From our previous discussion, we can suggest the following model about the diffusion properties of phospholipids in polymer-tethered monolayers at the air-water interface, which is shown in Figure 7. Three different situations should be considered: (I) no significant polymer-polymer interaction; (II) weak-moderate polymerpolymer interaction, (III) strong polymer-polymer interaction (physical gelation). (I) At Alipo > Alipo*, adjacent polymer chains show no significant interaction. The area available per polymer chain is greater than the space requirements for the equilibrium

Figure 7. Model describing the relationship between lipopolymer molar concentration and lateral mobility within the monolayer. (I) Situation of noninteracting lipopolymers. The diffusion properties are independent of the lipopolymer molar concentration. (II) Lipopolymers interacting via frictional coupling of their polymer moieties. This leads to a decreased lateral mobility of both lipopolymers and phospholipids. (III) Lipopolymers forming a two-dimensional physical network. The resulting immobilization of lipopolymers triggers a sharp mobility decrease among phospholipids. A further slowing down might be caused by microphase separation within the monolayer and increased interfacial roughness at the polymer-lipid interface. Also shown are the parameters L (average length of the polymer chain) and d (tube diameter in which the polymer chains are squeezed in) which are calculated using eqs 2-6. The circles within the polymer layer indicate the existence of physical junction points necessary for the formation of a physical polymer network.

polymer configuration. Lipopolymers can diffuse freely within the fluid phospholipid matrix. The diffusion properties are independent of the lipopolymer concentration. (II) At Alipo** < Alipo < Alipo*, adjacent polymer chains are able to interact with each other via frictional coupling. Polymer configurations may be either mushroom or brushlike. The frictional coupling among lipopolymer molecules leads to a significant slowing down of the lipopolymer mobility within the monolayer, which leads to a decrease of the lateral diffusion of phospholipids. Above the high-filmpressure transition, where microphase separation occurs, an increased immobile fraction can be observed. (III) At Alipo< Alipo**, adjacent polymers form a physical network, which leads to the immobilization of lipopolymer molecules within the monolayer. This leads to a significant decrease of the phospholipid diffusion. Additional contributions, which might result in a slower phospholipid diffusion are (1) increased interfacial roughness at the polymer-lipid interface and (2) microphase separation between lipopolymers and phospholipids.

Polymer-Tethered Membranes

Biomacromolecules, Vol. 2, No. 4, 2001 1103

erties of the membrane. On the other hand, membrane properties can be manipulated based on properties of the underlying polymer cushion. Acknowledgment. We want to thank S. Boxer for use of his microscopy setup and for helpful discussions. This work was supported by the Deutsche Forschungsgemeinschaft (DFG) and the NSF Materials Research Science and Engineering Center Program under DMR 94-00354 and DMR 98-08677 through the Center on Polymer Interfaces and Macromolecular Assemblies (CPIMA). References and Notes

Figure 8. Comparison of the D-Alipo relationship between polymertethered monolayer and polymer-tethered bilayer. The diffusion is much more affected in the case of the bilayer system as the more pronounced slope indicates. While the bilayer system with its immobile obstacles shows a linear relationship over the whole area range, the monolayer bahaves much more complex depending on the different obstacle mobilities.

Comparison between Monolayer and Bilayer. Can we now compare our monolayer results to earlier findings on polymer-tethered phospholipid bilayers? It was already mentioned above that monolayer and bilayer data should be compared with great care. Regarding absolute diffusion data, a direct comparison is rather problematic, if the area per phospholipid is kept constant in both cases at Alipid ) 65 Å2. This is shown in Figure 8 where diffusion data of a polymertethered monolayer and a polymer-tethered bilayer are plotted together for Alipid ) 65 Å2. While the polymer-tethered bilayer exhibits a linear relationship over the concentration range shown, the polymer-tethered monolayer reveals a more complex behavior dependent on the different mobility of obstacles. In this case, different D-Alipo relationships can be related to different obstacle mobilities. Concluding Remarks On the basis of our experimental findings, we argue that polymer-tethered phospholipid monolayer experiments at the air-water interface are very useful additions to corresponding experiments of polymer-tethered bilayers. This is not only because the existing results regarding the diffusion behavior of phospholipids in polymer-tethered bilayers are contrary1,17 but also because monolayer experiments allow a straightforward variation of both the lipopolymer molar concentration and the molecular areas Alipo and Alipid over a wide concentration range without disturbance from an underlying substrate. The experiments presented above have led to new insights into the problem of controlled covalent tethering of biomimicking composite materials, which go beyond experimental results obtained recently on polymer-tethered bilayers. A very interesting aspect of these experiments is our finding that mobile tethers do not affect the lateral mobility of tethered phospholipid membranes. This means that complex composite architectures of biomembranes can be designed in a way that stability can be improved via covalent tethering without limiting important dynamic prop-

(1) Naumann, C. A.; Prucker, O.; Lehmann, T.; Ruehe, J.; Knoll, W.; Frank, C. W. Biomacromolecules, in press. (2) Sackmann, E. Science 1996, 271, 43. (3) Loefas, M. M. S.; Roenneberg, I.; Senberg, E.; Liedberg, E.; Lundtsroem, I. Sens. Actuators B 1991, 5, 79. (4) Stelzle, M.; Weissmueller, G.; Sackmann, E. J. Phys. Chem. 1993, 97, 2974. (5) Tamm, L. K.; McConnell, H. M. Biophys. J. 1985, 47, 105. (6) Johnson, S. J.; Bayerl, T. M.; McDermott, D. C.; Adam, G. W.; Rennie, A. R.; Thomas, R. K.; Sackmann, E. Biophys. J. 1991, 59, 289. (7) Naumann, C. A.; Brumm, T.; Bayerl, T. M. Biophys. J. 1992, 63, 1314. (8) Merkel, R.; Sackmann, E.; Evans, E. J. Phys. Fr. 1989, 50, 1535. (9) McConnell, H. M.; Watts, T. H.; Weis, R. M.; Brian, A. A. Biochim. Biophys. Acta 1986, 864, 95. (10) Hinterdorfer, P.; Baber, G.; Tamm, L. K. J. Biol. Chem. 1994, 269, 20360. (11) Salafsky, J.; Groves, J. T.; Boxer, S. Biochemistry 1996, 35, 14773. (12) Spinke, J.; Yang, J.; Wolf, J.; Liley, M.; Ringsdorf, H.; Knoll, W. Biophys. J. 1992, 63, 167. (13) Kuehner, M.; Tampe´, R.; Sackmann, E. Biophys. J. 1994, 67, 217. (14) Kuehner, M. Ph.D. Thesis, TU-Munich, 1995. (15) Wong, J. Y.; Majewski, J.; Seitz, M.; Park, C. K.; Israelachvili, J. N.; Smith, G. S. Biophys. J. 1999, 77, 1445. (16) Wong, J. Y.; Park, C. K.; Seitz, M.; Israelachvili, J. N. Biophys. J. 1999, 77, 1458. (17) Wagner, M. L.; Tamm, L. K. Biophys. J. 2000, 79, 1400. (18) Prucker, O.; Naumann, C. A.; Ruehe, J.; Knoll, W.; Frank, C. W. J. Am. Chem. Soc. 1999, 121, 8766. (19) Naumann, C. A.; Brooks, C. F.; Fuller, G. G.; Knoll, W.; Frank, C. W. Langmuir 1999, 15, 7752. (20) Naumann, C. A.; Brooks, C. F.; Wiyatno, W.; Knoll, W.; Fuller, G. G.; Frank, C. W. Macromolecules 2001, 34, 3024. (21) Naumann, C. A.; Brooks, C. F.; Fuller, G. G.; Lehmann, T.; Ruehe, J.; Knoll, W.; Kuhn, P.; Nuyken, O.; Frank, C. W. Langmuir 2001, 17, 2801. (22) Axelrod, D.; Koppel, D. E.; Schlessinger, J.; Elson, E.; Webb, W. W. Biophys. J. 1976, 16, 1055. (23) Kim, S.; Yu, H. J. Phys. Chem. 1992, 96, 4034. (24) Sonnleitner, A.; Schuetz, G. J.; Schmidt, T. Biophys. J. 1999, 77, 2638. (25) Baekmark, T. R.; Elender, G.; Lasics, D. D.; Sackmann, E. Langmuir 1995, 11, 3975. (26) Wiesenthal, T.; Baekmark, T. R.; Merkel, R. Langmuir 1999, 15, 6837. (27) Evans, E.; Sackmann, E. J. Fluid Mech. 1988, 194, 553. (28) Saxton, M. J. Biophys. J. 1987, 52, 989. (29) Saxton, M. J. Biophys. J. 1989, 56, 615. (30) Lis, L. J.; McAlister, M.; Fuller, M.; Rand, R. P.; Parsegian, V. A. Biophys. J. 1982, 37, 1352. (31) Naumann, C. A.; Brumm, T.; Rennie, A. R.; Penfold, J.; Bayerl, T. M. Langmuir 1995, 11, 3948. (32) Saffmann, P. G.; Delbrueck, M. Proc. Natl. Acad. Sci. U.S.A. 1975, 72, 3111. (33) deGennes P. G. Scaling Concepts in Polymer Physics; Cornell University Press: Ithaca, NY, and London, 1993. (34) Elias, H.-G. An Introduction to Polymer Science; VCH: Weinheim, Germany, 1997. (35) Majewski, J.; Kuhl, T. L.; Gerstenberg, M. C.; Israelachvili, J. N.; Smith, G. S. J. Phys. Chem. B 1997, 101, 3122.

BM010022T