Nanoparticle-Mediated Mechanical Destruction of

Jul 18, 2017 - accommodate the distortion caused by the rotating NP and therefore the natural motion of the lipids prevents membrane rupture. In other...
0 downloads 0 Views 2MB Size
Subscriber access provided by UNIV OF NEWCASTLE

Article

Nanoparticle-Mediated Mechanical Destruction of Cell Membranes: A Coarse-Grained Molecular Dynamics Study Liuyang Zhang, Yiping Zhao, and Xianqiao Wang ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b05741 • Publication Date (Web): 18 Jul 2017 Downloaded from http://pubs.acs.org on July 30, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 2 3 4 5 6 7

ACS Applied Materials & Interfaces

Nanoparticle-Mediated Mechanical Destruction of Cell Membranes: A Coarse-Grained Molecular Dynamics Study Liuyang Zhang1, Yiping Zhao2, and Xianqiao Wang1* 1 2

College of Engineering, University of Georgia, Athens, GA 30602, USA

Department of Physics and Astronomy, University of Georgia, Athens, GA 30602, USA * Corresponding author: [email protected]

8 9 10

Abstract

11

The effects of binding mode, shape, binding strength, and rotational speed of actively rotating

12

nanoparticles on the integrity of cell membranes have been systematically studied using

13

dissipative particle dynamics simulations. With theoretical analyses of lipid density, surface

14

tension, stress distribution, and water permeation, we demonstrate that the rotation of

15

nanoparticles can provide a strong driving force for membrane rupture. The results show that

16

nanoparticles embedded inside a cell membrane via endocytosis are more capable of producing

17

large membrane deformations under active rotation than nanoparticles attached on the cell

18

membrane surface. Nanoparticles with anisotropic shapes produce larger deformation and have a

19

higher rupture efficiency than those with symmetric shapes. Our findings provide useful design

20

guidelines for a general strategy based on utilizing mechanical forces to rupture cell membranes

21

and therefore destroy the integrity of cells.

22 23

Keywords: Dissipative Particle Dynamics, Rotating Nanoparticles, Membrane Rupture,

24

Nanoparticle-Membrane Interaction, Cell Apoptosis

25

1 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

Introduction

2

Recent years have witnessed the explosive growth of interest in cell membrane – nanoparticle

3

(NP) interaction, since nanoparticles are widely used as carriers to deliver drug molecules and/or

4

genes into cell interiors.1-3 With the modification of the geometry and physical properties of

5

these carrier nanoparticles, they can interact with a cell membrane in different ways: they can (i)

6

be adsorbed on a membrane surface,4 (ii) pass through the cell membrane either by passive

7

diffusion or endocytosis,1 or (iii) get trapped in the membrane bilayer interior.5-6 Cell membrane-

8

NP interactions can impose various degrees of membrane deformation (e.g., bending, swelling,

9

rupture) due to changes in membrane thickness, specific surface per lipid, and lipid tail

10

orientation.5,

7

11

delivered to cancer sites and rotated by applying a high-frequency oscillating magnetic field to

12

damage cell membranes.8-12 For magnetic NPs subjected to an oscillating magnetic field, their

13

non-equilibrium response results in a localized mechanical shear. 13-15 The imposed shear forces

14

due to the rotation of magnetic NPs on the cell membrane may lead to its rupture. Blood et al.

15

simulated solvent shear flow caused by NPs rotation, which produced a 2.91023 pN force on the

16

bilayer and increased stress in the membrane by an order of magnitude.16 The rupture of

17

biological cell membranes under mechanical stress is an integral component of cell behavior,

18

however, less attention has been paid to the fundamental mechanisms of the mechanical rupture

19

of cell membrane by NPs.

20

The intrinsic mechanical characteristics of cell membrane-NP interactions are closely related to

21

the NP size, shape, and surface properties.1 For NP size Lai et al. demonstrated that the presence

22

of small NPs enhances the probability of water penetration, thus decreasing the membrane

23

rupture tension, while large NPs have the opposite effect, which demonstrates the ability of NPs

Recent experimental studies have shown that magnetic NPs can be effectively

2 ACS Paragon Plus Environment

Page 2 of 31

Page 3 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

1

to modify the structural and dynamic properties of a lipid membrane.17 The shape anisotropy and

2

initial orientation of NPs are also crucial to their penetration capability across a lipid bilayer,

3

which is determined by the contact area and local curvature of NPs at the contact point.18-20

4

Matula et al. showed that during collision between NPs and prokaryotic or eukaryotic cells,

5

nanorods were more effective in damaging cell membranes than spherical NPs.21 In addition to

6

NP size and shape, NP surface modification can dramatically change the cell-NP interaction.

7

Therefore, the performance of NPs heavily depends on the type of coating material used and how

8

the material is attached to the NPs. For example, MDCK II cells exposed to

9

cetyltrimethylammonium bromide-coated gold NPs display a concentration dependent stiffening

10

that presumably results from the endocytosis of particles which removes excess membrane area

11

from the cell surface. Interestingly, gold NPs coated with biocompatible polyethylene glycol do

12

not demonstrate a significant change in the elastic properties of the cell membrane.22 Urs et al.

13

found that NPs coated with polyethylene oxide triblock were more favorable for uptake by cells

14

than uncoated NPs, however they also introduced toxicity into the cells.23 Such a mechanical

15

interaction of NPs with cell membranes provides a promising pathway towards causing rupture

16

in cell membranes. This interaction also heavily depends on NP characterization parameters as

17

size, shape, surface functionalization, rotational speed, etc. For instance, the range of potential

18

effects due to NP surface coatings allows for a diverse range of chemical materials to be used for

19

targeting, and enables modified NPs to have certain desired binding properties.24-25 However, the

20

fundamental mechanisms of the rupture process and the role of NP inherent properties on the

21

rupture process of cell membranes remain elusive.

22

Although experimental techniques have shed considerable light on the properties of NPs and

23

their interactions with cell membranes, the process of membrane rupture is rather fast and

3 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

involves subtle changes at the molecular level; subsequently it is very challenging to study these

2

molecular level mechanisms by experimental techniques alone. Molecular dynamics (MD)

3

simulation offers an alternative approach to study membrane rupture at the molecular level. For

4

example, simulation demonstrated that for a pure DPPC membrane, the application of a large

5

mechanical pressure of −200 bar led to pore formation and irreversible rupture.26 Taiki et al.

6

found that a transient recovery of the cholesterol and phospholipid molecular orientations was

7

more evident at lower stretching speeds than that of higher speeds, suggesting the formation of a

8

stretch-induced interdigitated gel-like phase and explaining the effect of stretching speed on

9

membrane pore formation.27 However, fully atomistic MD simulations of cell membrane systems

10

are computationally expensive, length-scale limited, and time consuming. Therefore, coarse-

11

grained molecular dynamics simulations are required to strike a balance between accuracy and

12

computational efficiency. Dissipative particle dynamics (DPD), as a mesoscopic coarse-grained

13

simulation method, is extensively used for soft materials and biomembrane systems.28-29 For

14

example, membrane rupture has been observed in simulations of different mixed bilayers of

15

phosphatidylethanolamine and C12E6, and can explain why dividing cells are more at risk for

16

rupture than static cells.28 The disruptive effect of continuous NPs rotation is ascribed to both the

17

enhanced membrane monolayer protrusion as well as shear force created.30 Coarse-grained

18

simulation works well for the investigation of the mechanisms behind dynamic cell membrane

19

rupture processes without compromising the molecular level details.

20

In this paper, by using DPD simulations, we have systematically studied cell membrane rupture

21

process due to NP rotation. We first estimate how both attachment and endocytosis of the NP

22

affect the cell membrane-NP interaction. The dynamic membrane structure is characterized by

23

lipid density, membrane lateral tension, and stress distribution. Second, we demonstrate how NP

4 ACS Paragon Plus Environment

Page 4 of 31

Page 5 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

1

shape affects the membrane structure during NP rotation. Membrane parameters such as lipid

2

density and water permeation will be used to predict whether a rotating NPs can induce

3

membrane rupture. A phase diagram demonstrating the combinatory effects of rotational speed

4

and membrane binding strength will also be studied, so as to provide insight into NP design for

5

mediating cell membrane rupture.

6

Computational Model and Methodology

7

In DPD models, a small group of atoms is treated as a single bead located at the center of mass

8 9 10

of the atoms. Beads  and  interact with each other via pairwise forces consisting of a conservative force  representing the excluded volume effect, a dissipative force 

representing the viscous drag between moving beads, and a random force  representing

11

stochastic impulse. Both  and  act together as a thermostat for the beads. Similar to

12

molecular dynamics simulations, the time evolution of beads is governed by Newton’s equation

13

of motion. The total force acting on bead  can be expressed as 

=    +   +   

=   ψ    − ψ     ∙     

14 15

(1)

+ ψ    ∆!"#/    where % is the maximum repulsive force, & is the distance between beads  and , &̂ is the unit vector pointing from bead  to bead , ( is the relative velocity between beads  and , and )

16

denotes a random number with zero mean and unit variance. The force weight function ψ& 

17

can be defined as a simple decaying function of distance as

  ,   < + + ψ   = * .,   > + #−

5 ACS Paragon Plus Environment

(2)

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 2

where &0 is the cutoff radius and 1 and 2 are parameters with the relationship 2 3 = 2156 7, with 56 being the Boltzmann constant and T the temperature of the system. The standard values

3

2 = 3.0 and 1 = 4.5 were used in our study.3, 31-33 The mass, length, time, and energy are all

4

normalized in our DPD simulations, with the unit of mass set to be equivalent to the mass of the

5

solvent bead, the unit of length taken to equal to the cutoff radius &0 , the unit of time to be equal

6

to the reduced simulation time =, and the unit of energy to be 56 7. The reduced DPD units can

7

be converted to SI units by examining the membrane thickness (4 >?) and the lipid diffusion

8

coefficient (5 @?3 A "B ) from experiment.34 The simulated thickness value of the bilayer is 5&0

9

and the effective time scale of simulation can be determined from the calculated lateral diffusion

10

constant of the lipid bilayer 0.0095 &03 ⁄=.3 By comparison with typical experimental values, it

11

can be shown that one DPD length unit corresponds to approximately 0.8 nm in physical units

12 13

and the time unit = = 24.32 ps. The time step is taken to be ∆E = 0.01=. All simulations are

performed using LAMMPS.35

14

The dimension of the simulation box is 60&0 (I) × 60&0 (L) × 25&0 (M). The system contains

15

54,600 lipid beads, 225,118 water beads, and 3,906 NPs beads, with a reduced particle density of

16

approximately 3.36 Solvent beads are not shown in figures for clarity. As shown in Figure 1, the

17

simulation system consists of the bilayer membrane with 4,200 lipid molecules (50% are set as

18

receptors), a rigid NP with coating ligand beads, and solvent particles. It should be noticed that

19

the size of the simulation box in the x-y plane is adjusted to maintain the zero-tension condition

20 21

of a bilayer membrane with the average area per lipid NO = 1 &03 . During the simulation, if the new average area per lipid N is greater than NO , the simulation box will be compressed in the x-y

22

plane until N = NO , while if N < NO the box will be stretched in the x-y plane until N = NO .

23

Meanwhile, the box length in the direction normal to the membrane (z) will change

6 ACS Paragon Plus Environment

Page 6 of 31

Page 7 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

1

correspondingly to keep the box volume fixed. The lipid molecule is represented by the coarse-

2

grained model proposed by Groot and Rabone

3

head beads and 10 lipid hydrophobic tail beads. As shown in Table 1, the repulsive interaction

4

28

as shown in Figure 1, with 3 lipid hydrophilic

parameter between the same types of beads within NPs and the membrane is taken as % = 25,

5

and for two beads of different types it is set as % = 100. Within each lipid molecule, an elastic

6

harmonic force,

7 8

  )  (3) R  is assumed to exist between two neighboring beads, where 5S = 100.0 and &S = 0.7 &0 are the

P  = QR (# −

spring constant and equilibrium bond length used for the lipid molecule respectively.37 The

9

ligands are anchored to the NP surface by a harmonic force with 5S = 200.0 and &S = 0.25&0 to

10

prevent the detachment of coating ligands from the NP during the simulation. The bending

11

resistance of a lipid chain is represented as an additional force due to a harmonic constraint on

12

two consecutive bonds,

U = −VWXYZ[ = −V\QU (U − U. ) ]

(4)

13

where 5^ , _ and _O are the bending constant, inclination angle, and equilibrium angle,

14

respectively. For three consecutive lipid tail beads or three consecutive lipid head beads in a lipid

15

molecule, we take 5^B = 6 and _OB = 180° ; for the last head-bead and the top two tail-beads,

16

5^3 = 3 and _O3 = 120° ; for the bottom two consecutive head beads and the first bead in each

17

tail, 5^b = 4.5 and _Ob = 120° .38

18

Results and Discussions

19

Binding Modes

20

The surface coating of a NP can dramatically affect its binding affinity with the cell membrane.39

21

The strength of the ligand-receptor binding allows the NP to have two different binding modes, 7 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

surface attachment and endocytosis into the interior of cell membrane, during the cell membrane-

2

NP interaction.12, 40 Here we set the ligand coating density on the NP surface to be 100%, the

3

binding strength between receptor and ligand % = 5, the rotational speed of the NP c =

4

0.0219 °/=, and the diameter of the NP d = 15 &0 . Figure 2 shows the dynamic cell membrane

5

rupture process by a spherical NP when the system is set up to be in attachment mode. Before

6

interacting with the NP, the receptors in the cell membrane are assumed to be uniformly

7

distributed. When the NP attaches to the cell membrane, the receptors on the surface of the cell

8

membrane travel towards the NP and eventually bind to the ligands on the NP surface.1 After an

9

equilibrium process (E ≥ 100 =), the NP starts to actively rotate (see the arrow in Figure 2(a))

10

and the ligand-receptor interaction causes the local lipid molecules in the membrane to follow

11

the rotation of the NP, which generates a shear force inside the lipid bilayer. When the cell

12

membrane cannot sustain the shear force exerted from the rotating NP, it starts to rupture. It can

13

be seen that an actively rotating NP can accelerate the wrapping process and promote the cellular

14

uptake of the NP.30 In addition, active rotation of the NP significantly shortens the traveling time

15

of cell membrane receptors during the cellular uptake process. The observed phenomenon is in

16

good agreement with the membrane response to the rotating NPs by Yue et al.30 Figure 3 shows

17

the rupture process of the cell membrane with a rotating NP under endocytosis mode. The NP

18

starts to rotate after the NP is fully endocytosed by the cell membrane shown at Figure 3(a).

19

Similar to the attachment mode, the membrane undergoes severe distortion and out-of-plane

20

deformation in a very short time. Therefore, we remove the zero-tension condition for the cell

21

membrane. Compared with the attachment mode, the endocytosis mode takes 25.7 ± 0.2% less

22

time to reach the ruptured status for the cell membrane. Our results suggest that a rotating NP

8 ACS Paragon Plus Environment

Page 8 of 31

Page 9 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

1

which has undergone endocytosis would have a higher possibility to cause cell membrane

2

rupture.

3

The area per lipid molecule N is a good metric to evaluate the deformation of the cell membrane

4

under mechanical stimulus. In order to demonstrate the dynamic effect of the rotating NP on the

5 6 7 8 9 10

mechanical response of the cell membrane, the time evolution of N under the two binding modes with and without the rotation of the NP is shown in Figure 4. Without the rotation of NP, N keeps

fluctuating around 1 &03 , which is in good agreement with the zero-tension status of a normal cell membrane.28 Accordingly, no membrane rupture is observed. With the rotation of the NP under

both attachment and endocytosis modes, N first increases with the shear force exerted on the membrane from the rotation of the NP. When N reaches ~1.8 &03 , an abrupt decline of N at

11

E = 795= for the endocytosis mode and E = 995= for the attachment mode indicates the

12

mechanical rupture of the membrane, which is in good agreement with the theoretical prediction

13

for N as the membrane undergoes pore formation and fracture.41 Specifically, during the time

14

period E = 145=~520=, N shows a more dramatic increase under endocytosis mode than it does

15

under attachment mode. From the molecular viewpoint, the active rotational motion of the NP is

16

accompanied by breaking and rebinding of ligand-receptor pairs between the NP and the cell

17

membrane during the rotation process. With ligand-receptor pairwise binding, the rotation of the

18

NP can generate a shear force inside the cell membrane. For endocytosis mode, the NP has a

19

greater initial contact area with the cell membrane than it does under attachment mode; under

20

attachment mode the ligand-receptor bound pairs increase substantially during the wrapping

21

process. In addition, the time delay of rupture between the two modes further indicates that cell

22

membrane rupture under endocytosis mode takes less time to occur than it does under attachment

23

mode. Figure 5 depicts the initial and final probability distributions of lipid molecules in terms of

9 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 31

1

von Mises stress with the NP under both attachment and endocytosis modes as shown in Figure 2

2

and Figure 3, respectively. For clarity, the initial state for both modes represents the beginning of

3

the binding mode; the final state indicates the stage when cell membrane rupture occurs. It can

4

be seen that rotation of the NP generates a heterogeneous stress distribution on the cell

5

membrane. At the initial state the stress has a lower magnitude when compared with the final

6

state, indicating a severe membrane distortion after rotation. Therefore, through the average lipid

7

molecule area and stress distribution we can further verify that NP rotation provides a driving

8

force for membrane rupture.

9

NP Shape Effect

10

NP shape anisotropy is crucial to the cellular uptake process, which is determined by the contact

11

area(s) and local curvature(s) of the NP at the contact point(s) with cell membrane. Prior studies

12

have shown that NP shape significantly affects its interaction with the cell membrane. For

13

example, elongated particles can more efficiently interact with lipid bilayer membranes than

14

spherical particles, which can be explained by the lower free energy required for the membrane

15

to wrap elongated particles as compared to that needed for spherical particles.42 A rod-shaped NP

16

prefers firmly adhering to the vessel wall through ligand-receptor interactions and settles down at

17

equilibrium state with full contact.43 Thus, it is expected that different shaped NPs may induce

18

cell membrane rupture to different extents. Here we investigate the effect of two anisotropic NPs,

19

the cylindrical and L-shaped NPs on the membrane rupture process. These two kinds of NPs with

20

the same volume (and thus a larger surface area) as the spherical NP can increase the contact

21

area with the cell membrane and therefore create more ligand-receptor pairs between the NP and

22

cell membrane. Figure 6 shows the dynamic process of membrane rupture by rotating a

23

cylindrical NP under both attachment and endocytosis modes. The rotational axis is

10 ACS Paragon Plus Environment

Page 11 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

1

perpendicular to the cross-section of the NP and passes through the center of mass of the

2

cylindrical NP as shown in Figure 6. Regardless of binding mode, the membrane undergoes more

3

severe local distortion for the cylindrical NP than it does for the spherical NP, which confirms

4

that the shape anisotropy of NP plays an important role in mechanically deforming the cell

5

membrane. When the NP starts to rotate, the local membrane distortion develops into a

6

membrane rupture adjacent to the NP, and the membrane rupture then further develops into an

7

even larger membrane pore. In order to vividly capture the membrane disruption caused by the

8

rotating NP, we divide the membrane into 20 × 20 sub-domains (systematically represented in

9

Figure 6(e)) and calculate the density of lipid molecules for each sub-domain (3&0 × 3&0 ). As

10

shown in Figure 6(c), it can be clearly observed that the membrane far away from the NP is

11

trivially affected by the rotation of the NP. The density stays around a normal value of ~10/&03 .

12

A high concentration of lipid molecules with density ~50/&03 occurs at the contacting site,

13

indicating that lipid molecules may experience a compressive deformation or even an out-of-

14

plane deformation with overlapped lipid layers. Similarly, the dynamic process of membrane

15

rupture by the rotation of an L-shaped NP under both attachment (Supplemental Materials:

16

Video 1) and endocytosis modes (Supplemental Materials: Video 2) is shown in Figure 7(a) and

17

Figure 7(b), respectively. In experiments, it is difficult to control the rotational axis of the L-

18

shaped NP as the rotation of NP under the external magnetic field may be dynamically random.

19

For simplicity, in our simulation, the L-shaped NP rotates along the axis that passes through the

20

crossing joint of the L-shaped NP. The simulation demonstrates the robust capability of the L-

21

shaped NP to cause reliable damage to the cell membrane. Additionally, it takes less time for the

22

L-shaped NP to break the integrity of the cell membrane when compared with the spherical and

23

cylindrical NPs. From the discussion of NP shape effects, we can draw a conclusion that the

11 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 31

1

magnitude of the shear force exerted on the membrane determines whether or not the membrane

2

undergoes mechanical rupture, while the rupture location and occurrence are strongly affected by

3

NP’s shape.

4

Figure 8 shows the surface tension distribution inside the cell membrane for spherical,

5

cylindrical, and L-shaped NPs, respectively, under attachment mode. The surface tension with a

6

surface normal in the z-direction inside a cell membrane can be characterized by the pressure

7

tensor i via the following relation 44,

m

 = j \kl (m) − kn (m)] [m

8 9

10

m#

(5)

with 1 being the surface tension, io the pressure normal to the bilayer, and ip the tangential

pressure. For MD simulations, this equation can be further simplified as 45,  = qm \km − (kr + ks )/]

(6)

where tu is the size of simulation cell in the z-direction normal to the cell membrane and i is

11

the pressure in the  = I, L, M direction. For the case with a spherical NP, high surface tension

12

inside the membrane occurs at the contact area between the NP and the cell membrane as

13

predicted while the tension on the non-contact area remains at zero. For a cylindrical NP, the

14

highest surface tension occurs at the sharp end when the NP rotates with the specified axis as

15

shown in Figure 6(a). For an L-shaped NP, multiple sites with high surface tension can be

16

observed if the rotation axis is chosen along the axis as shown in Figure 7(a). The dramatic

17

change of surface tension at the contact site by the rotation of the NP implies that the bilayer

18

membrane experiences a severe heterogeneous deformation, which can lead to mechanical

19

rupture. For all these cases, the rotation of the NP causes an increase of the average distance

20

between lipid molecules (its deformation), leading to an increase in the potential energy (strain

21

energy) of the membrane. When the distance between intermolecular lipid molecules reaches a 12 ACS Paragon Plus Environment

Page 13 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

1

critical threshold, the membrane will snap to an energetically favorable state by releasing the

2

strain energy via rupture. These theoretical findings on the surface tension distributions suggest

3

that utilizing differently shaped actively rotating NPs could change the cell rupture process.

4

The diffusion of water perpendicular to the membrane layer provides another useful way of

5

characterizing the porosity of the bilayer membrane. Such a diffusivity D can be estimated by

6

solving the diffusion equation for the mean square displacement (MSD) of water in the z-

7

direction before and after the rotation of a NP. The vwd x= 〈& (E) − & (0) 〉{ of water

8 9 10 11

3

molecule i versus diffusion time E is governed by the Einstein relation, vwd = 2d>E, where > is

the allowed dimensionality of the diffusion process. Figure 9 plots d at the initial and final stages of membrane rupture processes for different shaped NPs. For the attachment mode, d

increases from 0.013 &03 ⁄= at the initial stage to 0.014 &03 ⁄= for the spherical NP, to 0.161 &03 ⁄=

12

for the cylindrical NP, and to 0.2966 &03 ⁄= for the L-shaped NP. Such an increment in D

13

indicates that the cell membrane becomes porous after its interaction with rotating NPs and

14

confirms our previous observation shown in Figure 6(c) and Figure 7(c). The presence of these

15

pores could accelerate the passive transport of water molecules across the membrane as well as

16

being initiation sites for structural defects associated with phase transitions, cell fusion, lysis, etc.

17

In addition, the final stage D value for the L-shaped NP is almost an order of magnitude larger

18

than that for the spherical NP, demonstrating that the L-shaped NP could generate far more

19

damage in a cell membrane than the spherical and cylindrical NPs. Similarly, under endocytosis

20

mode, D increases from 0.013 &03 ⁄= at the initial stage to 0.3402 &03 ⁄= for cylindrical NP, and to

21

0.5305 &03 ⁄= for L-shaped NP, which reflects severe membrane rupture, in agreement with the

22

phenomenon observed in previous sections. The NPs embedded inside the cell membrane are

13 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

more capable of producing membrane rupture under active rotation than nanoparticles attached

2

on the cell membrane surface.

3

Phase Diagram

4

In previous sections, we have demonstrated that various shapes of rotating NPs can exert

5

different shear forces on the membrane and are able to rupture it under both attachment and

6

endocytosis modes. The mechanical rupture of the cell membrane by NPs depends on a number

7

of factors, such as particle size, shape, and surface chemistry. Effects of particle size have been

8

investigated in detail by Yue et al.

9

Experimentally, NPs can bind to the cell surface through sophisticated chemical modification,

10

and the binding strength and modes can be tuned as well. Here we perform a large number of

11 12 13 14 15 16 17

30

and Li et al.

38

Page 14 of 31

and are not considered in our study.

DPD simulations with spherical NPs to determine the effect of the binding strength % between

the ligand and receptor as well as the effect of rotational speed c of NPs on the integrity of the cell membrane. For simplicity, the diameter of the spherical NP is fixed to be d = 12&0 . Figure 10(a) shows the phase diagram of the integrity of the cell membrane with respect to both % and

c of the NP under attachment mode, and Figure 10(b) shows the phase diagram under

endocytosis mode. Under the attachment mode, when % is relatively small (for example, % = 15), the rotation of NP with c ≤ 0.0061 °/= barely affects the cell membrane as the

18

membrane can be gently pushed away by the flow created by the rotation of the NP.31 This

19

observation may be explained by the fact that a low rotational speed gives the cell membrane

20

enough time to accommodate the distortion caused by the rotating NP and therefore the natural

21

motion of the lipids prevents membrane rupture. In other words, a high rotational speed of NP is

22

required to generate sufficient disruption to exceed the critical value of the area per lipid

23

molecule (N~2&03), thus inducing rupture. With respect to cases with a high % (for 14 ACS Paragon Plus Environment

Page 15 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

ACS Applied Materials & Interfaces

example, % = 5), the membrane undergoes mechanical rupture with a relatively low NP c such

2

as 0.00061 °/=. Increasing % can also decrease the possibility of the detachment of the NP

3

from the cell membrane, thus facilitating cell membrane rupture.

4

Under endocytosis mode, the combined effect from the % and c is more pronounced than it is

5

under attachment mode, as during endocytosis mode the contact area between the NP and

6

membrane is increased and so is the shear force inside the membrane. Therefore, a relatively low

7 8 9

c or % may lead to mechanical rupture of the cell membrane. For example, the membrane is

expected to rupture when c is as low as 0.00061 °/= if % is taken to be 10. When c is larger

than 0.0109°/=, mechanical rupture of the cell membrane always occurs regardless of % value.

10

Within the range of c and % tested for the spherical NP, we also find that mechanical

11

membrane rupture always occurs under the rotation of the cylindrical and L-shaped NPs. The

12

phase diagram provides a direct insight into the cooperative effects between the binding strength

13

between the NP and the cell membrane and the critical rotational speed required for the

14

mechanical membrane rupture caused by the active rotational motion of NPs.

15

Concluding remarks

16

In summary, we have used dissipative particle dynamics simulations to investigate how the

17

rotation of NPs affect their local interaction with the cell membrane and can trigger mechanical

18

rupture of the membrane. We find that with a ligand-receptor binding interaction, the membrane

19

receptor will follow the rotation of the NP and deteriorate the membrane integrity. The rotation

20

of a NP can generate shear force through the ligand-receptor pairs to create pores in the cell

21

membrane. Both the rotational speed of the NPs and the ligand-receptor binding strength

22

contribute to the mechanism of NP-induced cell membrane rupture. In addition, the rotation of

23

NPs under endocytosis mode can cause a larger degree of cell membrane disruption than 15 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 31

1

demonstrated under attachment mode. Furthermore, through the investigation of the time

2

evolution of the cell membrane tension, lipid density, and water diffusion coefficient for different

3

shaped NPs, we find that the rotation of complex anisotropic shaped NPs possesses more

4

destructive power for the mechanical rupture of cell membrane. Our simulation results show a

5

detailed molecular-level rupture process of the cell membrane and suggest that actively rotating

6

NPs can act as an efficient mechanism for membrane poration and cell lysing.

7

Acknowledgements

8

The authors acknowledge support from the National Science Foundation (Grant. No. CMMI-

9

1306065 and ECCS-1303134). The facility support for modeling and simulations from the UGA

10

Advanced Computing Resource Center is also greatly appreciated.

11

Additional information

12

Competing financial interests: The authors declare no competing financial interests.

13

16 ACS Paragon Plus Environment

Page 17 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

ACS Applied Materials & Interfaces

References 1. Zhang, S.; Gao, H.; Bao, G. Physical Principles of Nanoparticle Cellular Endocytosis. ACS Nano 2015, 9 (9), 8655-8671. 2. Blanco, E.; Shen, H.; Ferrari, M. Principles of nanoparticle design for overcoming biological barriers to drug delivery. Nat. Biotech. 2015, 33 (9), 941-951. 3. Zhang, L.; Becton, M.; Wang, X. Designing Nanoparticle Translocation through Cell Membranes by Varying Amphiphilic Polymer Coatings. J. Phys. Chem. B 2015, 119 (9), 3786-3794. 4. Lesniak, A.; Salvati, A.; Santos-Martinez, M. J.; Radomski, M. W.; Dawson, K. A.; Åberg, C. Nanoparticle Adhesion to the Cell Membrane and Its Effect on Nanoparticle Uptake Efficiency. J. Am. Chem. Soc. 2013, 135 (4), 1438-1444. 5. Gao, H. Probing mechanical principles of cell–nanomaterial interactions. J. Mech. Phys. Solids 2014, 62 (0), 312-339. 6. Zhang, S.; Li, J.; Lykotrafitis, G.; Bao, G.; Suresh, S. Size-Dependent Endocytosis of Nanoparticles. Adv. Mater. 2009, 21 (4), 419-424. 7. Westerhausen, C.; Strobl, F. G.; Herrmann, R.; Bauer, A. T.; Schneider, S. W.; Reller, A.; Wixforth, A.; Schneider, M. F. Chemical and Mechanical Impact of Silica Nanoparticles on the Phase Transition Behavior of Phospholipid Membranes in Theory and Experiment. Biophys. J. 2012, 102 (5), 1032-1038. 8. Tietze, R.; Zaloga, J.; Unterweger, H.; Lyer, S.; Friedrich, R. P.; Janko, C.; Pöttler, M.; Dürr, S.; Alexiou, C. Magnetic nanoparticle-based drug delivery for cancer therapy. Biochem. Biophys. Res. Commun. 2015, 468 (3), 463-470. 9. Fathi Karkan, S.; Mohammadhosseini, M.; Panahi, Y.; Milani, M.; Zarghami, N.; Akbarzadeh, A.; Abasi, E.; Hosseini, A.; Davaran, S. Magnetic nanoparticles in cancer diagnosis and treatment: a review. Artif. Cells Nanomed. Biotechnol. 2017, 45 (1), 1-5. 10. Halbreich, A.; Groman, E. V.; Raison, D.; Bouchaud, C.; Paturance, S. Damage to the protein synthesizing apparatus in mouse liver in vivo by magnetocytolysis in the presence of hepatospecific magnetic nanoparticles. J. Magn. Magn. Mater. 2002, 248 (2), 276-285. 11. Shen, Y.; Wu, C.; Uyeda, T. Q. P.; Plaza, G. R.; Liu, B.; Han, Y.; Lesniak, M. S.; Cheng, Y. Elongated Nanoparticle Aggregates in Cancer Cells for Mechanical Destruction with Low Frequency Rotating Magnetic Field. Theranostics 2017, 7 (6), 1735-1748. 12. Marcos-Campos, I.; Asín, L.; Torres, T. E.; Marquina, C.; Tres, A.; Ibarra, M. R.; Goya, G. F. Cell death induced by the application of alternating magnetic fields to nanoparticle-loaded dendritic cells. Nanotechnology 2011, 22 (20), 205101. 13. Zhang, E.; Kircher, M. F.; Koch, M.; Eliasson, L.; Goldberg, S. N.; Renström, E. Dynamic Magnetic Fields Remote-Control Apoptosis via Nanoparticle Rotation. ACS Nano 2014, 8 (4), 31923201. 14. Reddy, L. H.; Arias, J. L.; Nicolas, J.; Couvreur, P. Magnetic Nanoparticles: Design and Characterization, Toxicity and Biocompatibility, Pharmaceutical and Biomedical Applications. Chem. Rev. 2012, 112 (11), 5818-5878. 15. Zhao, Z.; Huang, D.; Yin, Z.; Chi, X.; Wang, X.; Gao, J. Magnetite Nanoparticles as Smart Carriers to Manipulate the Cytotoxicity of Anticancer Drugs: Magnetic Control and Ph-Responsive Release. J. Mater. Chem. 2012, 22 (31), 15717-15725.

17 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44

Page 18 of 31

16. Blood, P. D.; Ayton, G. S.; Voth, G. A. Probing the Molecular-Scale Lipid Bilayer Response to Shear Flow Using Nonequilibrium Molecular Dynamics. J. Phys. Chem. B 2005, 109 (39), 18673-18679. 17. Lai, K.; Wang, B.; Zhang, Y.; Zheng, Y. Computer simulation study of nanoparticle interaction with a lipid membrane under mechanical stress. Phys. Chem. Chem. Phys. 2013, 15 (1), 270-278. 18. Yang, K.; Ma, Y.-Q. Computer Simulation of the Translocation of Nanoparticles with Different Shapes across a Lipid Bilayer. Nat. Nanotechnol. 2010, 5 (8), 579-583. 19. Nangia, S.; Sureshkumar, R. Effects of Nanoparticle Charge and Shape Anisotropy on Translocation through Cell Membranes. Langmuir 2012, 28 (51), 17666-17671. 20. Li, Y.; Yue, T.; Yang, K.; Zhang, X. Molecular modeling of the relationship between nanoparticle shape anisotropy and endocytosis kinetics. Biomaterials 2012, 33 (19), 4965-4973. 21. Matula, K.; Richter, L.; Adamkiewicz, W.; Akerstrom, B.; Paczesny, J.; Holyst, R. Influence of nanomechanical stress induced by ZnO nanoparticles of different shapes on the viability of cells. Soft Matter 2016, 12 (18), 4162-4169. 22. Pietuch, A.; Brückner, B. R.; Schneider, D.; Tarantola, M.; Rosman, C.; Sönnichsen, C.; Janshoff, A. Mechanical properties of MDCK II cells exposed to gold nanorods. Beilstein J. Nanotechnol. 2015, 6, 223-231. 23. Häfeli, U. O.; Riffle, J. S.; Harris-Shekhawat, L.; Carmichael-Baranauskas, A.; Mark, F.; Dailey, J. P.; Bardenstein, D. Cell Uptake and in Vitro Toxicity of Magnetic Nanoparticles Suitable for Drug Delivery. Mol. Pharmaceutics 2009, 6 (5), 1417-1428. 24. Mornet, S.; Vasseur, S.; Grasset, F.; Veverka, P.; Goglio, G.; Demourgues, A.; Portier, J.; Pollert, E.; Duguet, E. Magnetic nanoparticle design for medical applications. Prog. Solid State Chem. 2006, 34 (2–4), 237-247. 25. Scarberry, K. E.; Dickerson, E. B.; McDonald, J. F.; Zhang, Z. J. Magnetic Nanoparticle−Peptide Conjugates for in Vitro and in Vivo Targeting and Extraction of Cancer Cells. J. Am. Chem. Soc. 2008, 130 (31), 10258-10262. 26. Tieleman, D. P.; Leontiadou, H.; Mark, A. E.; Marrink, S.-J. Simulation of Pore Formation in Lipid Bilayers by Mechanical Stress and Electric Fields. J. Am. Chem. Soc. 2003, 125 (21), 6382-6383. 27. Shigematsu, T.; Koshiyama, K.; Wada, S. Effects of Stretching Speed on Mechanical Rupture of Phospholipid/Cholesterol Bilayers: Molecular Dynamics Simulation. Sci. Rep. 2015, 5, 15369. 28. Groot, R. D.; Rabone, K. L. Mesoscopic Simulation of Cell Membrane Damage, Morphology Change and Rupture by Nonionic Surfactants. Biophys. J. 2001, 81 (2), 725-736. 29. Groot, R. D.; Warren, P. B. Dissipative Particle Dynamics: Bridging the Gap between Atomistic and Mesoscopic Simulation. J. Chem. Phys. 1997, 107 (11), 4423-4435. 30. Yue, T.; Zhang, X.; Huang, F. Molecular Modeling of Membrane Responses to the Adsorption of Rotating Nanoparticles: Promoted Cell Uptake and Mechanical Membrane Rupture. Soft Matter 2015, 11 (3), 456-465. 31. Zhang, L.; Wang, X. Coarse-Grained Modeling of Vesicle Responses to Active Rotational Nanoparticles. Nanoscale 2015, 7 (32), 13458-13467. 32. Zhang, L.; Wang, X. Nanotube-Enabled Vesicle–Vesicle Communication: A Computational Model. J. Phys. Chem. Lett. 2015, 6 (13), 2530-2537. 33. Zhang, L.; Xu, B.; Wang, X. Cholesterol Extraction from Cell Membrane by Graphene Nanosheets: A Computational Study. J. Phys. Chem. B 2016, 120 (5), 957-64. 34. Shillcock, J. C.; Lipowsky, R. Tension-Induced Fusion of Bilayer Membranes and Vesicles. Nat. Mater. 2005, 4 (3), 225-228. 18 ACS Paragon Plus Environment

Page 19 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

ACS Applied Materials & Interfaces

35. Plimpton, S. Fast Parallel Algorithms for Short-Range Molecular Dynamics. J. Comput. Phys. 1995, 117 (1), 1-19. 36. Nielsen, S. O.; Ensing, B.; Ortiz, V.; Moore, P. B.; Klein, M. L. Lipid Bilayer Perturbations around a Transmembrane Nanotube: A Coarse Grain Molecular Dynamics Study. Biophys. J. 2005, 88 (6), 3822-3828. 37. Venturoli, M.; Smit, B.; Sperotto, M. M. Simulation Studies of Protein-Induced Bilayer Deformations, and Lipid-Induced Protein Tilting, on a Mesoscopic Model for Lipid Bilayers with Embedded Proteins. Biophys. J. 2005, 88 (3), 1778-1798. 38. Li, Y.; Li, X.; Li, Z.; Gao, H. Surface-Structure-Regulated Penetration of Nanoparticles Across a Cell Membrane. Nanoscale 2012, 4 (12), 3768-3775. 39. Domenech, M.; Marrero-Berrios, I.; Torres-Lugo, M.; Rinaldi, C. Lysosomal Membrane Permeabilization by Targeted Magnetic Nanoparticles in Alternating Magnetic Fields. ACS Nano 2013, 7 (6), 5091-5101. 40. Hervault, A.; Thanh, N. n. T. K. Magnetic nanoparticle-based therapeutic agents for thermochemotherapy treatment of cancer. Nanoscale 2014, 6 (20), 11553-11573. 41. Grafmüller, A.; Shillcock, J.; Lipowsky, R. Dissipative particle dynamics of tension-induced membrane fusion. Mol. Simul. 2009, 35 (7), 554-560. 42. Tree-Udom, T.; Seemork, J.; Shigyou, K.; Hamada, T.; Sangphech, N.; Palaga, T.; Insin, N.; PanIn, P.; Wanichwecharungruang, S. Shape Effect on Particle-Lipid Bilayer Membrane Association, Cellular Uptake, and Cytotoxicity. ACS Appl. Mater. Interfaces 2015, 7 (43), 23993-24000. 43. Shah, S.; Liu, Y.; Hu, W.; Gao, J. Modeling Particle Shape-Dependent Dynamics in Nanomedicine. J. Nanosci. Nanotechnol. 2011, 11 (2), 919-928. 44. Simunovic, M.; Voth, G. A. Membrane tension controls the assembly of curvature-generating proteins. Nat. Commun. 2015, 6, 7219. 45. Chiu, S. W.; Clark, M.; Balaji, V.; Subramaniam, S.; Scott, H. L.; Jakobsson, E. Incorporation of surface tension into molecular dynamics simulation of an interface: a fluid phase lipid bilayer membrane. Biophys. J. 1995, 69 (4), 1230-1245.

28 29

19 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 31

1 2 3

Figure 1. Schematic representation of the simulation model. The lipids and receptors are represented by

4

H3(T5)2 models with tail beads in dark blue. Lipid head beads are shown in red and receptor head beads

5

are colored green. A representative spherical NP is shown in yellow with hydrophilic coating ligand beads

6

shown in light blue.

7

20 ACS Paragon Plus Environment

Page 21 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

1 2 3

Table 1. Soft-repulsive interaction parameters   for head and tail beads of lipid molecules (q} and qn ),

head and tail beads of receptor molecules (} and n ), ligand beads (G), and nanoparticle (NP) beads.

4

q}

qn

}

n

~

NP

25

100

25

100

25

100

100

25

100

25

100

25

25

100

25

100

5

100

100

25

100

25

100

25

~

25

100

5

100

25

100

NP

100

25

100

25

100

25

q} qn

} n

5 6

21 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 31

1 2

3 4

Figure 2. Dynamic process of cell membrane rupture at different time by a spherical NP under

5

attachment mode.

6 7

22 ACS Paragon Plus Environment

Page 23 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

1

2 3

Figure 3. Dynamic process of cell membrane rupture at different time by a spherical NP under

4

endocytosis mode.

5

23 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2

2.2 Area per lipid molecule A ( rc2)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 31

2 1.8 1.6 1.4 1.2 1 0.8 0

3

Free NP Attachment mode Endocytosis mode

500

1000 Time t (τ )

1500

4

Figure 4. Time evolution of the average area per lipid  versus spherical NP rotation time ! for different

5

binding modes.

6 7

24 ACS Paragon Plus Environment

Page 25 of 31

1.4 Attachment mode: initial Attachment mode: rupture

1.2

Endocytosis mode: initial

Probability (%)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

1

Endocytosis mode: rupture

0.8 0.6 0.4 0.2 0 0

5

10

15

20

25

30

35

3

1 2

von Mises stress (ε /σ ) Figure 5. The von Mises stress distribution in the membrane during the spherical NP rotation process.

3 4

25 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 31

1

2 3

Figure 6. Dynamic membrane rupture process by the rotation of a cylindrical NP. (a) Attachment mode;

4

(b) the density distribution of lipid beads at the final stage under attachment mode (the inset shows the

5

method to divide the membrane into sub-domains); (c) endocytosis mode; (d) the density distribution of

6

lipid beads at the final stage under endocytosis mode.

7 8

26 ACS Paragon Plus Environment

Page 27 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

1 2

3 4

Figure 7. Dynamic membrane rupture process by the rotation of a L-shaped NP. (a) Attachment mode;

5

(b) endocytosis mode; (c) the density distribution of lipid beads at the final stage under attachment mode;

6

(d) the density distribution of lipid beads at the final stage under endocytosis mode.

7

27 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

(a)

(b) 20

20

10

10

10

-10

-10

-20

-20

2 3 4

-10

0 X-axis

10

20

‚ = 18&0

0

Y-axis

 = 12 &0

0

-20

1

(c)

20

Y-axis

Y-axis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 31

-10

0 X-axis

10

‚

0

‚ = 15&0

-10

 = 8 &0 -20

150

-20 20

-20

-10

0 X-axis

10

 = 6 &0

100

50

20

0

Figure 8. Initial surface tension distribution for different shaped NPs: (a) spherical, (b) cylindrical, and (c) L-shaped. The dashed curves outline the shape of each NP. The x- and y- axes in units of + represent

the dimensionless membrane size. The color bar represents the surface tension value in unit of Q€ n⁄+ .

5

28 ACS Paragon Plus Environment

Page 29 of 31

1 2 3

(b)

(a)

0.5

Attachment mode

Endocytosis mode

c

Diffusion Coefficient D (r 2/τ )

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Initial Final

0.4

0.3

0.2

0.1

0

Sphere

Cylinder L-shape

Sphere Cylinder L-shape

4 5

Figure 9. Water diffusion coefficient D perpendicular to the membrane surface when (a) the NP is

6

attached to the membrane or (b) endocytosed by the membrane.

7

29 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 31

1 2

3 4

Figure 10. Phase diagram of membrane rapture by a rotational spherical NP as a function of binding

5

strength and rotational speed when the NP is (a) attached to or (b) endocytosed by the membrane. The

6

symbol

represents a case with cell rupture and

represents a case resulting in an intact cell membrane.

7

30 ACS Paragon Plus Environment

Page 31 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

ACS Applied Materials & Interfaces

TOC

2

31 ACS Paragon Plus Environment