Polystyrene nanoparticles perturb lipid membranes

Polystyrene Nanoparticles Perturb Lipid Membranes Giulia Rossi1,2,3†, Jonathan Barnoud1,2,3, ¶ and Luca Monticelli1,2,3, ¶, *

1

2

INSERM, UMR-S665, Paris, F-75015, France

Université Paris Diderot, Sorbonne Paris Cité, UMR-S665, Paris, F-75013, France 3

INTS, Paris, France

AUTHOR INFORMATION Corresponding Author * email: [email protected] Present Addresses † Dept of Physics, University of Genoa; Via Dodecaneso 33, 16146 Genoa, Italy ¶ IBCP, CNRS UMR 5086, and University Lyon I; 7 Passage du Vercors, 69367, Lyon, France

1

ABSTRACT. Polystyrene is abundant in marine debris. Like most synthetic polymers, it degrades very slowly, producing smaller and smaller particles easily ingested by wildlife. The presence of plastic microscopic particles in fish and marine wildlife is massive and well documented, but its impact on cellular activity is not understood. Biological activity generally requires interaction with biological membranes, but this is difficult to study at the molecular scale in vivo.

Here we use coarse-grained molecular simulations to determine the effect of nano-sized polystyrene (PS) particles on the properties of model biological membranes. We find that PS nanoparticles permeate easily into lipid membranes. Dissolved in the membrane core, PS chains alter membrane structure, reduce significantly molecular diffusion, and soften the membrane. Moreover, PS severely affects membrane lateral organization by stabilizing raft-like domains. Changes in membrane properties and lateral organization can severely affect the activity of membrane proteins and thereby cellular function.

TOC GRAPHICS

KEYWORDS. Lipid membranes, polymers, nanoparticle, coarse-grained models, molecular dynamics, rafts

2

Worldwide annual production of plastics reached 280 million tons in 20121 and continues to increase. A significant fraction of the total production is used for disposable packaging materials, whose useful life cycle is typically short and ends in landfills and in the oceans. Plastics constitute 60-80% of marine litter2. Degradation of plastic materials is slow but inevitable, and plastic fragments are expected to persist in the marine environment for centuries3-4. Micrometer size particles are easily ingested by marine wildlife, such as mussels5, fish6-7, seabirds8, and whales9. Polystyrene (PS), one of the most common polymers found in marine plastic debris10, is commonly found in fish5. Ingestion of plastic microparticles can harm animals via the release of plastic monomers and toxic chemical additives, such as phthalates11. While the presence of micrometer-sized plastics in the oceans is well documented, smaller particles can be generated by further degradation of disposable items. Nanometer-size polymer particles are also produced industrially for specific research and technological applications, such as imaging, sensing and preparation of nanocomposites12. The presence of nano-sized plastics in research labs, industry and the environment raises questions on their potential toxicity. Biological membranes are the first barrier encountered by particles foreign to the cell. Micrometer sized plastic particles can enter cells probably via endocytosis-like mechanisms5, 13. In mussels, ingestion of polystyrene particles is followed by translocation from the gut to the circulatory system, and polymer particles are retained for more than a month5. Smaller particles accumulate in tissues more easily than larger ones5. Hemocytes can engulf polystyrene particles with a diameter of 800 nm14. The interaction of smaller plastic particles with cells is most likely mediated by cell membranes in a different way. As particle size decreases to the nanometer range, it approaches the thickness of a biological membrane (ca. 4 nm). Particles of such small

3

size could partition into cell membranes and diffuse through them, as observed in model systems. Radlinska et al. showed that polystyrene chains with a molecular weight of ~250,000 (about 2500 monomers and a radius of gyration of about 10 nanometers) can penetrate the hydrophobic core of synthetic nonionic surfactant bilayers and alter their structural and elastic properties15. In living systems, the interaction between plastic nanoparticles and biological membranes is difficult to study experimentally at the molecular level, mostly due to the limited resolution of optical techniques. The effects of polymer chains on the properties of biological membranes are unknown. Here we use molecular simulations to understand how polystyrene nano-sized particles alter the physical properties of model biological membranes. Biological membranes control a vast number of cellular functions by regulating the activity of membrane proteins; membrane properties such as thickness, elasticity, and lateral heterogeneity determine protein sorting and functioning16-18. We use the MARTINI coarse-grained (CG) force field19-20 for the description of the polymer21 and the membranes (Fig. 1a-b and Supplementary Fig. 1). Briefly, the model uses a 4:1 mapping of atoms onto coarse-grained interaction sites, and a building block approach. The potential energy function is based on free energies of transfer of each building block between polar and non-polar phases (see Supplementary Information and ref. 19 for details).

Validation of the polystyrene model We first carried out extensive validation of the polystyrene model to assess its reliability in terms of polymer-lipid interactions. We calculated the free energy of transfer of individual styrene monomers from octane to water using both the MARTINI coarse-grained model21 and the all-atom OPLS22 force field (see the Methods section and Supplementary Information for

4

!" details). Coarse-grained simulations yielded a free energy of transfer ∆𝐺!" = 21.0 ± 0.7 kJ/mol, !" in excellent agreement with the value of ∆𝐺!" = 20.6 ± 0.3 kJ/mol obtained from atomistic

calculations. Then we determined the free energy of transfer of a PS monomer and trimer from water to the interior of a 1-palmitoyl-2-oleoyl-phosphatidylcholine (POPC) membrane, both at the atomistic and the coarse-grained level. We found again very good agreement (Supplementary Fig. 2). Overall, our tests show that the MARTINI force field is very well suited for the purpose of the present study: it describes realistically the partitioning of different species; it provides information on molecular motions at near-atomistic level, overcoming resolution limits found in experiments; it allows simulations on time scales of tens of microseconds, relevant for partitioning of synthetic nanoparticles and for the investigation of membrane properties.

Polystyrene nanoparticles enter POPC membranes and dissolve We considered polystyrene chains of different length: PS10, PS20 and PS100, consisting of 10, 20, and 100 styrene monomers, respectively. In water, the highly hydrophobic PS chains formed compact aggregates, with a diameter of up to 7 nm (Fig. 1c). At physiological temperature, these PS nanoparticles were solid, with extremely slow internal dynamics, as expected. We simulated PS nanoparticles in the presence of homogeneous membranes of POPC, a phospholipid commonly found in cell membranes (see Supplementary Table 1 for a complete list of the simulations performed, and Fig. 1a and 1b for the CG representation of the lipid and the polymer). PS nanoparticles were initially placed in water. Once in contact with the lipids, the nanoparticles entered the membrane on a time scale of a few microseconds. Permeation in the hydrophobic core of the membrane (Fig. 1d) was followed by dissolution of the aggregate (Fig. 1e-f) on a time scale of 1-10 μs. Dissolution of PS nanoparticles proceeded through melting (i.e.,

5

increased mobility of the polymer chains) and disaggregation (i.e., separation of the chains) for all molecular weights considered. As expected, both permeation and disaggregation were faster for smaller particles.

Figure 1. PS nanoparticles enter lipid membranes and dissolve. a Coarse-grained model of PS10. Each monomer consists of one backbone bead and three ring beads. One PS10 monomer is highlighted in violet; b CG model of a POPC lipid. Two polar head beads and the glycerol beads are shown in spacefill representation; c Lateral view of an aggregate of 11 PS100 chains in the water phase (water is not shown), near a POPC membrane at t = 0; d Same system as in c, after 1 μs; e After 5 μs, top view; f After 10 μs, top view; the aggregate is completely dissolved.

At equilibrium, short PS chains (N = 10, 20) were dispersed throughout the hydrophobic part of the membrane (Fig. 2). Longer chains (N = 100), instead, were found preferably closer to the center of the membrane. PS100 chains, effectively confined between the membrane leaflets, were remarkably swollen with respect to their equilibrium configuration in theta solvents, and showed large aspect ratios. Strong confinement of polystyrene chains in bilayers is consistent with previous experimental results on non-ionic surfactant bilayers and PS derivatives15.

6

z distance from membrane plane [nm]

4

a!

b!

Water

PS10! in POPC! 2

c! POPC

0

PS10

PS100! in POPC!

PS100

d! PS100!

-2

PO4 PS20! PS10!

-4

0

200

400

600

800

1000

Density [kg/m^3]

Figure 2. Distribution of PS within POPC membranes. a Density profiles from simulations of POPC membranes containing PS10 or PS100 at 10% PS/lipid mass ratio. PO4 indicates lipid phosphate groups; b and c Snapshots from the corresponding simulations. Water is not shown, lipids are shown in cyan, PS backbones in magenta; d Radius of gyration of PS chains in POPC, for different chain sizes and concentrations. The value reported for the 100% concentration refers to PS melts at T = 450 K (or, equivalently, to PS in very dilute conditions in a theta solvent, such as cyclohexane at room temperature). As for chain anisotropy, the average eigenvalues of the gyration tensor are written in parenthesis, for the lowest PS concentration. The chain anisotropy of our PS100 model in the melt is also reported for comparison.

Polystyrene chains alter the properties of homogeneous membranes

7

We determined the effect of PS chains on the properties of POPC membranes at polymer:lipid mass ratios up to 13% (Table 1). We observed a significant increase in the area of the membrane, reaching 10% at the highest polymer concentration. Membrane thickness remained approximately constant upon addition of PS10 and PS20, while we registered a few percent variation upon addition of PS100 at the large concentrations (Supplementary Fig. 3). The effect on lipid chain order parameter was minor for short polymer chains, and rather limited for PS100 (Supplementary Fig. 3), which is consistent with very minor variations in membrane thickness. For PS100, the slight increase in thickness can be ascribed to the presence of PS100 at the center of the membrane. The diffusion coefficient of the lipids, DL, was reduced by a factor of 3 at the highest polymer concentration (Table 1 and Supplementary Fig. 4). We also calculated the diffusion coefficient (DP) of a short helical transmembrane peptide, WALP2323. Peptide motion was significantly affected by the presence of PS100 chains: DP was reduced by almost an order of magnitude (Table 1). The polymer also altered the mechanical properties of the membrane. Both the area compressibility modulus (KA) and the bending modulus (Kb) decreased with increasing polymer concentration (see Table 1 and Supplementary Fig. 5), indicating mechanical softening of the membrane. Changes in the pressure profile of the membrane also point to membrane softening (see Supplementary Fig. 6). Membrane softening was also observed by Radlinska et al. in experiments with polystyrene derivatives in surfactant bilayers15. In our simulations, most changes (in area, diffusion constants, elastic moduli) were approximately linear with polymer mass concentration, and showed weak dependence on the length of polymer chains (Supplementary Fig. 3, 4, and 5).

8

Table 1. Structural, dynamic and mechanical properties of POPC membranes upon addition of PS100. A indicates the average area per lipid, DL the diffusion coefficient of the lipids, DP the diffusion coefficient of the protein, KA the area compressibility modulus, Kb the bending modulus. Composition

A

DL

DP

KA

Kb

(nm2)

(cm2/s)

(cm2/s)

(mN/m)

(kBT)

pure POPC

0.65

6.2 10-7

3.1 10-7

302

10.9

POPC + 2.7% PS100

0.66

4.6 10-7

NA

273

11.1

POPC + 7.6% PS100

0.68

3.0 10-7

NA

227

8.0

POPC + 10% PS100

0.70

2.1 10-7

NA

211

6.4

POPC + 13% PS100

0.71

1.9 10-7

4.4 10-8

216

6.4

Uncertainties are below 5%, except for Kb (see discussion in SI). NA: not available.

Polystyrene chains affect lateral organization in multi-component membranes We then determined the effect of PS chains (PS10 and PS100) on the properties of more realistic models of biological membranes, consisting of different lipid species. In experiments, multi-component lipid bilayers including saturated and unsaturated phospholipids and cholesterol show lateral segregation between liquid-ordered domains (Lo), which are cholesterolrich, thicker and stiffer, and liquid disordered (Ld) domains, depleted in cholesterol and richer in unsaturated lipids24. The separation of Lo and Ld phases has been reproduced in coarse-grained simulations of ternary mixtures25. We considered membranes consisting of unsaturated phosphatidylcholine (PC) lipids (dilinoleyl-PC, DUPC), saturated PC lipids (dipalmitoyl-PC, DPPC) and cholesterol (CHOL), with composition DUPC:DPPC:CHOL 0.28:0.42:0.3. Mixtures with similar composition show phase separation in experiments at room temperature26. In our

9

simulations, at room temperature, the mixture formed Ld and Lo domains with well defined composition (Fig. 3), consistent with previous results25, 27. We then added PS chains to the phaseseparated membranes, using polymer:lipid mass ratios of 1% and 5%. We placed the PS chains randomly in the hydrophobic core of the membrane, in contact with both the Lo and the Ld domains. PS showed a striking tendency to partition to the Ld domain, independently of chain length and polymer concentration. Partitioning was observed on a time scale of a few microseconds (Fig. 3). a!

b!

T = 295 K!

T = 305 K!

c!

d!

T = 315 K!

T = 325 K!

Figure 3. PS chains partition to the Ld domain. a Lateral and top view of a DPPCDUPC-CHOL membrane (top), and of the same membrane upon addition of 6 chains (5% PS/lipid mass ratio) of PS100 (bottom), at T = 295 K. Only lipid head beads are shown. DUPC is cyan, DPPC is grey, CHOL is black. PS backbone beads are magenta. In the bottom snapshots, only one lipid leaflet is shown to better appreciate the position of the polymer chains; b Top view only, at T = 305 K; c at T = 315 K; d At T = 325 K. All snapshots, with and without PS100, are taken after 10 µs of MD simulation.

10

By partitioning into the Ld phase, PS affected the lipid composition of both phases. We quantified the changes in domain composition by calculating the number of contacts between different lipids at equilibrium, before and after the addition of polymer chains (Table 2). Remarkably, adding PS resulted in almost complete exclusion of cholesterol from the Ld phase, and in a significant reduction of DUPC-DPPC contacts. The effect of PS did not depend on the length of the polymer chains (Supplementary Table 3). To verify if the changes in domain composition affect the thermal stability of the domains, we performed simulations of the same membrane systems, with and without polymer, at higher temperatures, up to 325 K. In the absence of PS, the interface between the ordered and disordered domains appeared increasingly blurred as the temperature increased, and disappeared completely at 325 K (Fig. 2d). In sharp contrast, phase separation was very stable up to 325 K in the presence of PS (Fig. 2d). Changes in the composition of both phases were even more dramatic at higher temperature, with reductions of DPPC-DUPC and DUPC-CHOL contacts by over 50% (Table 2). All results on polymer partitioning, compositional changes and domain stabilization were highly reproducible and robust with respect to minor changes in the parameterization of the polystyrene model (See Supporting Information). We conclude that polystyrene chains strongly promote domain segregation in ternary lipid mixtures.

Table 2. Changes in the number of contacts between different lipid species, in phaseseparated membranes, upon addition of polystyrene (PS100). Differences are calculated between simulations at the same temperature, with and without polymer. = indicates no change, i.e., the number of contacts is the same with and without the polymer. Composition

T [K]

Contacts

11

DPPC DPPC

DPPC DUPC

DPPC CHOL

DUPC DUPC

DUPC CHOL

295

=

-20%

+10%

-5%

-41%

DUPC:DPPC:CHOL 305 0.28:0.42:0.3 315 + 5% mass PS100

=

-48%

+10%

=

-61%

+3%

-55%

+10%

+5%

-60%

+3%

-61%

+10%

+18%

-63%

325

Our findings on polystyrene partitioning and the alteration of membrane properties have general relevance for the interaction of polymer nanoparticles with biological membranes. Partitioning of polystyrene into lipid membranes and subsequent softening of the membrane are consistent with experimental observations in non-ionic surfactant membranes15. This behavior is to be expected only for polymer sizes comparable to membrane thickness, while different mechanisms of interaction with and transport across membranes are relevant for larger particle sizes (hundreds of nanometers or larger), both in model systems and in living cells. All the effects on membrane properties observed here are therefore peculiar to nano-sized polymer particles. We find large effects of the polymer on all membrane properties, although only at relatively high concentrations (8% in weight or more). Alterations in membrane structural properties and elasticity can affect important cellular processes, such as membrane protein sorting17 and functioning18. The effect of PS on lipid and protein diffusion is similar to the effect of cholesterol28, but the mechanism is different: here the dynamics is slowed down by the configurational rearrangements of the long PS chains, occurring on time scales longer than the time scale of lipid diffusion. In contrast with cholesterol, PS has little effect on the degree of ordering of the CG lipid tails, and thus on membrane thickness. Moreover, PS softens the

12

membrane, as quantified by a decrease in both the bending modulus and the area compressibility modulus, while cholesterol increases the rigidity of mono-unsaturated lipid membranes, and has little effect on the rigidity of unsaturated lipid membranes29. The most striking effects of polymer nanoparticles are observed on phase-separated membranes. Polystyrene chains partition strongly to the Ld phase, increasing polymer concentration locally (in our case, by a factor of 4). The effect of polymer chains on the properties of laterally heterogeneous membranes will therefore be significant even at relatively low polymer concentrations. Most importantly, the polymer significantly stabilizes lipid domains and alters their chemical composition. Increasing experimental evidence indicates that cell membranes in vivo are laterally heterogeneous, and functioning of cell membranes depends on membrane lateral organization16,

30

. So-called rafts – membrane domains enriched with

sphingolipids, sterols and specific proteins – are involved in membrane sorting and trafficking, cell polarization, and signal transduction16, 30, and may play a role in various diseases31. Although it is very difficult to predict the effect of PS nanoparticles in vivo, alterations of physical properties and lateral organization in model membranes suggest that similar effects are likely in cell membranes. Results presented here call for in-depth experimental investigations on the biological effects of nano-sized plastics.

Methods Setup of coarse-grained (CG) simulations. All CG simulations were carried out using the MARTINI force field19-21 (see Supplementary Information for details). Molecular dynamics (MD)

simulations

of

homogeneous

1-palmytoyl-2-oleoyl-phosphatidylcholine

(POPC)

membranes contained either 512 or 2048 lipids. The polymer:lipid mass ratio was between 0 and

13

13%. Simulation time was between 2 and 20 µs. Supplementary Table 1 lists all the unbiased MD simulations performed in POPC membranes. MD simulations of PS and POPC membranes started from two alternative configurations: (a) PS chains dispersed in the water phase on top of an equilibrated membrane, or (b) PS chains randomly dispersed within the hydrophobic core of an equilibrated membrane. MD simulations of ternary lipid mixtures included unsaturated phosphatidylcholine (PC) lipids (dilinoleyl-PC, DUPC), saturated PC lipids (dipalmitoyl-PC, DPPC) and cholesterol (CHOL). Membrane composition and size were identical in all simulations performed, and consisted of 828 DPPC lipids, 540 DUPC and 576 cholesterol molecules (ratio: 0.42:0.28:0.3). All simulations started with PS chains randomly dispersed within the membrane hydrophobic core. In calculation of membrane properties, sampling started when the number of contacts between the different lipid species reached convergence, typically within the first microsecond. A list of all simulations of PS in lipid mixtures is reported in the Supporting Information (Supplementary Table 2). CG simulation parameters. CG simulations were performed in the NpT ensemble, with periodic boundary conditions in all directions. We used a time step of 20 fs, except for the runs where cholesterol was present, when the time step was reduced to 15 fs. The neighbor list for all non-bonded interactions was updated every 10 steps. The time scales indicated throughout the paper correspond to the formal simulated CG time, with no rescaling based on the comparison with atomistic dynamics. The temperature was set to T = 310 K using the Bussi-Donadio-Parrinello thermostat32 (with a time constant t = 2 ps). The pressure was set to the atmospheric value and controlled by the Parrinello-Rahman barostat33 (with a 4 ps relaxation time and a compressibility of 4.5 × 10-5

14

bar-1, z coupled independently of x and y for all simulations of bilayer systems). Some simulations of ternary lipid mixtures were carried out using the Berendsen weak coupling algorithm34 (with the same reference pressure and relaxation time), as indicated in Supplementary Table 2. All simulations were carried out with the Gromacs (v4.5.3) software package35.

Contact analysis. Two molecules were considered to be in contact whenever the distance between them, d, was shorter than a threshold distance, dt. Only one bead per molecule has been used to calculate d: the PO4 bead for lipids, and the ROH bead for cholesterol (see reference

19

for the definition of the particles). The results reported in the paper were obtained with a threshold dt = 1 nm. In order to compare runs at different temperatures, with and without polystyrene, the number of contacts was normalized to the total number of contacts in each system at each time t, and then averaged over time. The calculation was performed with the Gromacs tool g_mindist. Diffusion coefficient. Diffusion coefficients D in the membrane plane were obtained by fitting lipid (or peptide) mean square displacements (MSD) to the linear time function 𝑓 𝑡 = 4𝑡𝐷, for 𝑡 > 100 ns. Mechanical properties. Area compressibility modulus KA is calculated using the following equation: 𝐾! = 𝑘! 𝑇 𝐴!

𝐴 − 𝐴!

!

where the average indicates a time average, A0 is the

average in-plane area of the membrane, A is the in-plane area of the membrane at time t, kB is the Boltzmann constant and T is the absolute temperature. The bending modulus was calculated from the undulation spectrum [see references

36

and

37

]. In-house software for this calculation is

available free of charge through the web at http://www.dsimb.inserm.fr/~luca/downloads. The bending modulus was calculated only from simulations containing 2048 POPC lipids (lateral size

15

of about 26 nm) for which sampling was 10 µs or longer. Both large size and long time scale are necessary to sample slow undulation modes.

Supporting Information. Details of the MARTINI models, methods and results for the validation of the models, list of simulations performed in POPC lipid membranes, figures with structural, dynamic and mechanical properties of POPC membranes in the presence of the polymer, list of simulations performed in heterogeneous membranes, contact analysis of PS10 in heterogeneous membranes, detailed explanation of the tests to verify the robustness of the model. This material is available free of charge via the Internet at http://pubs.acs.org.

Notes The authors declare no competing financial interests.

AUTHOR INFORMATION Corresponding Author * email: [email protected]

ACKNOWLEDGMENT

16

This work was performed using HPC resources from GENCI-CINES (Grant 2011-076353 and 2012-076353). G.R. acknowledges funding from the FP7 Marie Curie IEF program. L.M. acknowledges D.P. Tieleman, O.H.S. Ollila, and S.J. Marrink for fruitful discussions.

REFERENCES 1. Rochman, C. M., et al., Policy: Classify Plastic Waste as Hazardous. Nature 2013, 494 (7436), 169-171. 2. Moore, C. J., Synthetic Polymers in the Marine Environment: A Rapidly Increasing, Long-Term Threat. Environ. Res. 2008, 108 (2), 131-139. 3. Barnes, D. K. A.; Galgani, F.; Thompson, R. C.; Barlaz, M., Accumulation and Fragmentation of Plastic Debris in Global Environments. Philos. Trans. R. Soc. B 2009, 364 (1526), 1985-1998. 4. Thompson, R. C.; Moore, C. J.; vom Saal, F. S.; Swan, S. H., Plastics, the Environment and Human Health: Current Consensus and Future Trends. Philos. Trans. R. Soc. B 2009, 364 (1526), 21532166. 5. Browne, M. A.; Dissanayake, A.; Galloway, T. S.; Lowe, D. M.; Thompson, R. C., Ingested Microscopic Plastic Translocates to the Circulatory System of the Mussel, Mytilus Edulis (L.). Environ. Sci. Technol. 2008, 42 (13), 5026-5031. 6. Boerger, C. M.; Lattin, G. L.; Moore, S. L.; Moore, C. J., Plastic Ingestion by Planktivorous Fishes in the North Pacific Central Gyre. Mar. Pollut. Bull. 2010, 60 (12), 2275-2278. 7. Carpenter, E. J.; Anderson, S. J.; Harvey, G. R.; Miklas, H. P.; Peck, B. B., Polystyrene Spherules in Coastal Waters. Science 1972, 178 (4062), 749-750. 8. Andrady, A. L., Microplastics in the Marine Environment. Mar. Pollut. Bull. 2011, 62, 1596-1605. 9. Fossi, M. C.; Panti, C.; Guerranti, C.; Coppola, D.; Giannetti, M.; Marsili, L.; Minutoli, R., Are Baleen Whales Exposed to the Threat of Microplastics? A Case Study of the Mediterranean Fin Whale (Balaenoptera Physalus). Mar. Pollut. Bull. 2012, 64 (11), 2374-2379. 10. Hidalgo-Ruz, V.; Gutow, L.; Thompson, R. C.; Thiel, M., Microplastics in the Marine Environment: A Review of the Methods Used for Identification and Quantification. Environ. Sci. Technol. 2012, 46 (6), 3060-3075. 11. Cole, M.; Lindeque, P.; Halsband, C.; Galloway, T. S., Microplastics as Contaminants in the Marine Environment: A Review. Mar. Pollut. Bull. 2011, 62 (12), 2588-2597. 12. Lu, Y.; Mei, Y.; Walker, R.; Ballauff, M.; Drechsler, M., ‘Nano-Tree’—Type Spherical Polymer Brush Particles as Templates for Metallic Nanoparticles. Polymer 2006, 47 (14), 4985-4995. 13. Pappo, J.; Ermak, T. H., Uptake and Translocation of Fluorescent Latex Particles by Rabbit Peyers Patch Follicle Epithelium - a Quantitative Model for M Cell Uptake. Clin. Exp. Immunol. 1989, 76 (1), 144-148. 14. Cajaraville, M. P.; Pal, S. G., Morphofunctional Study of the Hematocytes of Bivalve Mollusk Mytilus-Galloprovincialis with Emphasis on the Endolysosomal Compartment. Cell Struct. Funct. 1995, 20 (5), 355-367. 15. Radlinska, E. Z.; Gulik-Krzywicki, T.; Lafuma, F.; Langevin, D.; Urbach, W.; Williams, C. E.; Ober, R., Polymer Confinement in Surfactant Bilayers of a Lyotropic Lamellar Phase. Phys. Rev. Lett. 1995, 74 (21), 4237-4240. 16. Simons, K.; Ikonen, E., Functional Rafts in Cell Membranes. Nature 1997, 387 (6633), 569-572. 17. Killian, J. A., Hydrophobic Mismatch between Proteins and Lipids in Membranes. Biochim. Biophys. Acta - Reviews Biomembranes 1998, 1376 (3), 401-416.

17

18. Cantor, R. S., Lateral Pressures in Cell Membranes: A Mechanism for Modulation of Protein Function. J. Phys. Chem. B 1997, 101 (10), 1723-1725. 19. Marrink, S. J.; Risselada, H. J.; Yefimov, S.; Tieleman, D. P.; de Vries, A. H., The Martini Force Field: Coarse Grained Model for Biomolecular Simulations. J. Phys. Chem. B 2007, 111 (27), 7812-7824. 20. Monticelli, L.; Kandasamy, S. K.; Periole, X.; Larson, R. G.; Tieleman, D. P.; Marrink, S. J., The Martini Coarse-Grained Force Field: Extension to Proteins. J. Chem. Theory Comput. 2008, 4 (5), 819834. 21. Rossi, G.; Monticelli, L.; Puisto, S. R.; Vattulainen, I.; Ala-Nissila, T., Coarse-Graining Polymers with the Martini Force-Field: Polystyrene as a Benchmark Case. Soft Matter 2011, 7 (2), 698-708. 22. Jorgensen, W. L.; Maxwell, D. S.; TiradoRives, J., Development and Testing of the Opls All-Atom Force Field on Conformational Energetics and Properties of Organic Liquids. J. Am. Chem. Soc. 1996, 118 (45), 11225-11236. 23. de Planque, M. R. R.; Greathouse, D. V.; Koeppe, R. E.; Schafer, H.; Marsh, D.; Killian, J. A., Influence of Lipid/Peptide Hydrophobic Mismatch on the Thickness of Diacylphosphatidylcholine Bilayers. A H-2 Nmr and Esr Study Using Designed Transmembrane Alpha-Helical Peptides and Gramicidin A. Biochemistry (Mosc.) 1998, 37 (26), 9333-9345. 24. Simons, K.; Vaz, W. L. C., Model Systems, Lipid Rafts, and Cell Membranes. Annu. Rev. Biophys. Biomol. Struct. 2004, 33 (1), 269-295. 25. Risselada, H. J.; Marrink, S. J., The Molecular Face of Lipid Rafts in Model Membranes. Proc. Natl. Acad. Sci. U. S. A. 2008, 105 (45), 17367-17372. 26. Filippov, A.; Orädd, G.; Lindblom, G., Domain Formation in Model Membranes Studied by PulsedField Gradient-Nmr: The Role of Lipid Polyunsaturation. Biophys. J. 2007, 93 (9), 3182-3190. 27. Baoukina, S.; Mendez-Villuendas, E.; Bennett, W. F. D.; Tieleman, D. P., Computer Simulations of the Phase Separation in Model Membranes. Faraday Discuss. 2013, 161, 63-75. 28. Filippov, A.; Oradd, G.; Lindblom, G., The Effect of Cholesterol on the Lateral Diffusion of Phospholipids in Oriented Bilayers. Biophys. J. 2003, 84 (5), 3079-3086. 29. Pan, J. J.; Tristram-Nagle, S.; Nagle, J. F., Effect of Cholesterol on Structural and Mechanical Properties of Membranes Depends on Lipid Chain Saturation. Phys. Rev. E 2009, 80 (2). 30. Lingwood, D.; Simons, K., Lipid Rafts as a Membrane-Organizing Principle. Science 2010, 327 (5961), 46-50. 31. Simons, K.; Ehehalt, R., Cholesterol, Lipid Rafts, and Disease. The Journal of Clinical Investigation 2002, 110 (5), 597-603. 32. Bussi, G.; Donadio, D.; Parrinello, M., Canonical Sampling through Velocity Rescaling. J. Chem. Phys. 2007, 126 (1). 33. Parrinello, M.; Rahman, A., Polymorphic Transitions in Single Crystals - a New Molecular Dynamics Method. J. Appl. Phys. 1981, 52 (12), 7182-7190. 34. Berendsen, H. J. C.; Postma, J. P. M.; van Gunsteren, W. F.; di Nola, A.; Haak, J. R., Molecular Dynamics with Coupling to an External Bath. J. Chem. Phys. 1984, 81, 3684-3690. 35. Hess, B.; Kutzner, C.; van der Spoel, D.; Lindahl, E., Gromacs 4: Algorithms for Highly Efficient, Load-Balanced, and Scalable Molecular Simulation. J. Chem. Theory Comput. 2008, 4 (3), 435-447. 36. Helfrich, W., Elastic Properties of Lipid Bilayers - Theory and Possible Experiments. Zeitschrift Fur Naturforschung C - a Journal of Biosciences 1973, C 28 (11-1), 693-703. 37. Lindahl, E.; Edholm, O., Mesoscopic Undulations and Thickness Fluctuations in Lipid Bilayers from Molecular Dynamics Simulations. Biophys. J. 2000, 79 (1), 426-433.

18