Subscriber access provided by Umea University Library

Article

Interfacing Living Cells and Spherically Supported Bilayer Lipid Membranes Carolin Madwar, Gopakumar Gopalakrishnan, and R. Bruce Lennox Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.5b00862 • Publication Date (Web): 31 Mar 2015 Downloaded from http://pubs.acs.org on April 5, 2015

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.

Langmuir 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 40

Langmuir

1 2 3 4 6

5

Interfacing Living Cells and Spherically Supported 8

7 9 10 1

Bilayer Lipid Membranes 12 13 14 16

15

Carolin Madwar, Gopakumar Gopalakrishnan*†, and R. Bruce Lennox* 17 18 19

[] C. Madwar, Dr. G. Gopalakrishnan, Prof. Dr. R. Bruce Lennox 2

21

20

Department of Chemistry 24

23

McGill University 25 27

26

801 Sherbrooke Street West, Montréal (QC) H3A 0B8, Canada Fax: (+1) (514) 398-3797 29

28

E-mail: [email protected] 30 31 32

[†] Current address: 3 35

34

Institut Galien Paris-Sud (UMR CNRS 8612), Université Paris-Sud 37

36

5 rue Jean-Baptiste Clément, 92296 Châtenay-Malabry, France 38 39

E-mail: [email protected] 41

40 42 43 4 45 46 47 48 49 50 51 52 53 54 5 56 57 58 59

1 60 ACS Paragon Plus Environment

Langmuir

Page 2 of 40

1 3

2

KEYWORDS 4 6

5

lipid membranes • actin • lipid domains • rafts • supported bilayers. 7 9

8

ABSTRACT 10 1 12

Spherically supported bilayer lipid membranes (SS-BLMs) exhibiting co-existing membrane 15

14

13

microdomains were created on spherical silica substrates. These 5 µm SiO2-core SS-BLMs are 17

16

shown to interact dynamically when interfaced with living cells in culture, while keeping the 18 20

19

membrane structure and lipid domains on the SS-BLM surface intact. Interactions between the 2

21

SS-BLMs and cellular components could potentially be examined via correlating fluorescently 23 24

labeled co-existing microdomains on the SS-BLMs, their chemical composition and biophysical 25 27

26

properties with the consequent organization of cell membrane lipids, proteins and other cellular 29

28

components. This experimental approach is demonstrated in a proof-of-concept experiment 30 31

involving the dynamic organization of cellular cytoskeleton, monitored as a function of the lipid 34

3

32

domains of the SS-BLMs. The compositional versatility of SS-BLMs provides a means to 36

35

address the relationship between the phenomenon of lipid phase separation and the other 37 39

38

contributors to cell membrane lateral heterogeneity such as interactions with the cytoskeleton of 41

40

living cells. 42 43 4 45 46 47 48 49 50 51 52 53 54 5 56 57 58 59

2 60 ACS Paragon Plus Environment

Page 3 of 40

Langmuir

1 3

2

1. INTRODUCTION 4 6

5

Membrane heterogeneity is fundamental to many cellular events including signaling, 8

7

protein/receptor trafficking, and membrane fusion.1,2,3,4 Although the driving force(s) behind 1

10

9

these inhomogeneities are not fully understood, it is becoming increasingly evident that cell 13

12

membranes possess lateral domains or rafts that are constituted of lipids, proteins and other 14 15

membrane-associated entities.5,6 There is considerable evidence that the lateral distribution of 18

17

16

membrane components and their respective lipid-lipid and lipid-protein interactions are 20

19

important in membrane heterogeneity. The regulation of the formation and maintenance of 21 23

2

membrane heterogeneity at physiological conditions involves factors such as protein aggregation 25

24

on the membrane leaflet,7,8 lipid domains with distinct physical and mechanical properties,3,9,10 27

26

and cytoskeleton-induced asymmetric lipid distributions and protein domain stabilization.11,12 28 30

29

Particularly interesting in the context of the present study is the existence of thermodynamically 32

31

stable lipid microdomains that have been convincingly illustrated using model membrane 34

3

systems.13,14,15,16,17,18 The two most studied model systems used in lipid phase separation studies 37

36

35

are giant unilamellar vesicles (GUVs)14,16,19 and planar supported bilayer membranes (S39

38

BLMs).20,21,22,23 However, studies involving these membrane models have focused on the 40 42

41

physical/mechanical/dynamical properties of lipid domains rather than experiments involving 4

43

living cells in culture. This limitation is due in part to the physical instability (in the case of 46

45

GUVs)24,25 or the technical difficulties and restrictions in relation to planar geometries (in the 47 48

case of S-BLMs).26 The approach we introduce here circumvents these limitations and allows 51

50

49

one to explore how co-existing lipid microdomains of a well-characterized model system interact 52 53

with complex cellular components. 5

54 56 57 58 59

3 60 ACS Paragon Plus Environment

Langmuir

Page 4 of 40

1 3

2

We report here an experimental platform where co-existing lipid microdomains are formed on 5

4

a micron-scale solid, spherical substrate. The resulting lipid microdomains parallel both the 6 7

chemical and dynamical properties of those in cell membranes. The tailor-made character of 10

9

8

these synthetic membranes, in terms of both size and composition, along with the physical 12

1

stability they demonstrate under physiological conditions, allow for their use as an active system 13 14

capable of interacting and inducing a measurable response from living cells in culture. To assess 17

16

15

the versatility of this system, spherical supported bilayer membranes (SS-BLMs) are used as a 19

18

platform to examine the correlation between the lipid phase separation phenomenon and the 20 2

21

organization of cellular components such as the cytoskeletal networks. This is achieved via co24

23

culturing SS-BLMs which display lipid phase separation (i.e. coexisting lipid microdomains) 26

25

with living cells under physiological conditions.28 The SS-BLM system combines the versatility 27 29

28

of GUVs and the robustness of S-BLMs.26,27,29 The presence of the SiO2-core with a well-defined 31

30

diameter allows for facile observation of these rigid membrane structures using time-resolved 32 3

microscopy and spectroscopy techniques and also serves to differentiate them from other native 36

35

34

vesicle membranes present in the cell culture milieu. Furthermore, unlike planar S-BLM, the SS38

37

BLM system enables introduction of model lipid membranes to living cells in vitro at any time of 39 41

40

the cell culture. As established here, this experimental versatility allows use of this system in 43

42

experiments involving cell culture, long term live cell imaging, immunocytochemistry as well as 45

4

other experimental procedures involving the use of detergents, change of pH or osmotic pressure 46 48

47

- all without compromising the structural integrity of the membrane domains within the model 50

49

membranes. 51 52 53 5

54

2. EXPERIMENTAL SECTION 56 57 58 59

4 60 ACS Paragon Plus Environment

Page 5 of 40

Langmuir

1 3

2

2.1 Lipids. Cholesterol, 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), 1,2-dioleoyl-35

4

trimethylammonium 6

propane

chloride

salt

(DOTAP),

1,2-dipalmitoyl-sn-glycero-3-

7

phosphocholine (DPPC), 1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine (DPPE), 1,210

9

8

distearoyl-sn-glycero-3-phosphocholine 12

1

(DSPC),

1,2-distearoyl-sn-glycero-3-

phosphatidylethanolamine-N-biotinyl-(polyethylene glycol 2000)] ammonium salt (DSPE13 14

PEG2000-biotin), 17

16

15

and

1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine-N-(7-nitro-2-1,3-

benzoxadiazol-4-yl) (ammonium salt) (NBD PE) were purchased from Avanti Polar Lipids 19

18

(purity 20

>99%).

4,4-difluoro-5,7-dimethyl-4-bora-3a,4a,diaza-s-indacene-3-pentanoic

acid

2

21

(Bodipy-PC) was purchased from Molecular Probes, Invitrogen (NY, USA). 1,1’-dieicosanyl24

23

3,3,3’,3’-tetramethylindocarbocyanine perchlorate (DiI-C20) was purchased from Molecular 25 26

Targeting Technologies (Pennsylvania, USA). Lissamine™ Rhodamine B 1,2-dihexadecanoyl27 29

28

sn-glycero-3-phosphoethanolamine, triethylammonium salt (N-Rh-DHPE), secondary antibodies 31

30

and Alexa-488/Alexa-647−phalloidin were purchased from Molecular Probes, Invitrogen (NY, 32 3

USA). Poly-L-lysine hydrobromide was purchased from Sigma (NY, USA). GelTol (aqueous 36

35

34

mounting media) was purchased from Shandon Lipshaw Co., Lerner Labs (PA, USA). All other 38

37

culture media were purchased from Gibco, Invitrogen (NY, USA). 39 41

40

2.2 Preparation of Lipid Bilayer-Coated Silica Beads (SS-BLMs). All reported SS-BLMs 43

42

are tethered lipid bilayers supported on spherical silica substrates. SS-BLMs were prepared as 45

4

reported previously27, starting with a solution of 5 µm silica beads from Bangs Laboratories (IN, 46 48

47

USA) at a concentration of 9 million particles/mL in phosphate buffer saline (PBS), pH = 7.4. A 50

49

volume of 100 µL of this solution was mixed with 0.1 mg/mL avidin for 20 minutes and 51 52

incubated overnight at 4 ˚C. The avidin-coated beads were then washed by centrifugation (3× at 53 5

54

10 × 103 rpm for 10 minutes), and then re-suspended in the same buffer prior to incubation with 56 57 58 59

5 60 ACS Paragon Plus Environment

Langmuir

Page 6 of 40

1 3

2

the lipids. For the formation of small unilamellar vesicles (SUVs), chloroform solutions of 5

4

respective lipids (1 mg/mL, 95 μL), and DSPE PEG2000-biotin (0.1 mg/mL, 5 μL) were mixed 6 7

and dried overnight under vacuum. The film was then hydrated using PBS (warmed to 10

9

8

temperatures higher than the phase transition temperature (Tm) of lipids) through vortex mixing, 12

1

followed by sonication in a bath sonicator for 2 – 5 minutes. Giant unilamellar vesicles (GUVs) 13 14

were prepared via electroformation on the automated Vesicle Prep Pro (Nanion Technologies; 17

16

15

Munich, Germany) chamber. In this method, 10 µL of a lipid-chloroform solution was dried 19

18

overnight under vacuum on a glass slide coated with indium tin oxide (ITO). The lipid film was 20 2

21

then hydrated by incubating with approximately 150 µL of PBS (at temperatures higher than the 24

23

Tm of the constituent lipids). The formation chamber was assembled using two of the ITO 25 26

electrodes facing one another, spaced and sealed by a rubber O-ring. An electric field of 10 Hz 27 29

28

and 1.4 V was applied for 30 minutes to obtain vesicles with a diameter of 1 – 3 µm. A volume 31

30

of 100 μL of the vesicle solution (SUVs or GUVs) was mixed with 100 μL of the avidin-coated 32 3

silica beads dispersed in PBS, shaken gently, and incubated for 20 minutes. The bead-vesicle 36

35

34

solution was then sonicated for 1 minute, washed by centrifugation (3× at 7 × 10 3 rpm for 10 38

37

minutes) and the resulting pellet was re-suspended in PBS. The resulting bilayer-coated beads 39 41

40

(SS-BLMs) were then subjected to two successive heat/cool cycles at a controlled rate of 0.1 43

42

˚C/second starting at 4 ˚C and ending at 80 ˚C using a TProfessional Thermocycler (Biometra; 45

4

Göttingen, Germany). SS-BLMs were then incubated at 4 ˚C for 24 hrs prior to examination. 46 48

47

2.3 Primary Cultures of Rat Hippocampal Neurons. Cultures of dissociated rat 50

49

hippocampal neurons were prepared using a modified protocol described by Banker.30 51 52

Hippocampi were dissected from E17 embryos, treated with 0.25% (w/v) trypsin at 37 ˚C 53 5

54

followed by Dulbecco's modified Eagle medium (DMEM) supplemented with 10% horse serum, 56 57 58 59

6 60 ACS Paragon Plus Environment

Page 7 of 40

Langmuir

1 3

2

and mechanically dissociated with a plastic Pasteur pipette. The dissociated neurons were plated 5

4

at a density of (1.75 – 2.0) × 104 cm−1 on, poly-L-lysine coated glass coverslips (Ted Pella Inc., 6 7

CA, USA) in serum-free neurobasal medium supplemented with l-glutamine, penicillin, 10

9

8

streptomycin and B-27. The culture was kept in a humidified 5% CO2 atmosphere at 37 ˚C and 12

1

one-third of the medium was replaced every 2 − 3 days. All animal work was performed in 13 14

accordance with the Canadian Council of Animal Care Guidelines. 17

16

15

2.4 Co-cultures with SS-BLM Beads and Immunocytochemistry. Primary cells were 19

18

cultured to at least 9 days in vitro (DIV) before the addition of beads. SS-BLMs suspended in 20 21

sterile PBS, pH = 7.4, were added to the cells drop wise at a concentration of (1.0 − 1.5) × 105 24

23

2

beads/coverslip. The bead/cell co-culture was incubated for 24 hrs in a humidified 5% CO2 25 26

atmosphere at 37 ˚C. Cells were fixed with 4% (w/v) paraformaldehyde in phosphate buffer, pH 27 29

28

7.4, for 15 minutes and washed 3× in PBS. The cells were then incubated in blocking solution 31

30

(PBS, pH = 7.4, containing 4% normal donkey serum (NDS) and 0.1% (w/v) saponin) for 30 32 3

min, and then in primary antibody solution (rabbit anti--tubulin 1:100 in PBS containing 0.1% 36

35

34

(w/v) saponin and 0.5% (w/v) NDS), overnight at 4 ˚C. Cells were washed 3× in PBS, incubated 37 38

in Alexa-488/Alexa-647 (as appropriately) coupled secondary antibodies (rabbit-specific, highly 39 41

40

cross-adsorbed IgG, 1:200 in PBS-0.5% (w/v) NDS) for 30 minutes, washed 3× in PBS. For 43

42

actin labeling, Alexa-488/Alexa-647−phalloidin (as appropriately) were used (1:50 dilution) in 4 45

the secondary antibody buffer. The stained samples (on glass coverslips) were mounted on 48

47

46

microscopic slides using GelTol and sealed prior to imaging. 50

49

2.5 Confocal Microscopy. All fluorescence images were obtained using a Zeiss LSM 710 51 52

confocal microscope from Carl Zeiss AG, (Germany) with a 63x/1.4 oil-immersion objective 5

54

53

lens, using either one or a combination of the following optical settings (i) λex 488 nm/ λem LP > 56 57 58 59

7 60 ACS Paragon Plus Environment

Langmuir

Page 8 of 40

1 3

2

505 nm (single channel imaging) or λem BP 505 − 550 nm (multi-channel imaging), (ii) λex 543 5

4

nm/ λem LP > 565 nm, and (iii) λex 633 nm/ λem LP > 685 nm. The acquired intensity images 6 7

were checked to avoid detector saturation and loss of offsets by adjusting the laser power, the 10

9

8

detector gain and the detector offset. The 3D image stacks were acquired at a sampling rate that 12

1

satisfies the Nyquist frequency. The obtained confocal raw fluorescence image stacks were 13 14

deconvolved by AutoQuant X3 software using blind deconvolution algorithm. All raw confocal 17

16

15

images were processed using Imaris 7.4.0 software. 19

18

2.6 Fluorescence Recovery After Photobleaching (FRAP). Measurements were performed 20 2

21

using a Zeiss LSM-710 confocal laser-scanning microscope with a 63x/1.4 oil-immersion 24

23

objective lens and a 488 nm argon ion laser (25 mW power). The samples used for FRAP 25 26

experiments were either GUVs or SS-BLMs (prepared starting with SUVs) from DOPC lipids 27 29

28

labeled with 0.1 mol% Bodipy-PC. For tethering, 0.1 mol % DSPE-PEG2000-biotin was used in 31

30

the lipid mixture to allow formation of SS-BLMs on 5 µm avidin coated silica beads or 32 3

stabilization of GUVs on avidin coated glass coverslips. The FRAP experiment started by 36

35

34

choosing a single SS-BLM or GUV in the image field of view, followed by defining three 1.2 38

37

µm radius circular regions of interest (ROI) for subsequent imaging; a bleach ROI, a reference 39 41

40

ROI outside the bleach area and a third background ROI outside the field of view of the SS43

42

BLM. Five images were captured prior to bleaching in order to measure the initial pre-bleach 45

4

fluorescence intensity, followed by 10 consecutive bleach iterations using 100% laser intensity. 46 48

47

The laser power was reduced to 5% for collecting the following 50 post-bleaching images. The 50

49

total scan time was minimized by imaging only the three circular regions of interest rather than 51 52

the whole SS-BLM in the field of view. This allowed a reduction of the total experiment time to 53 5

54

ca. 14 s. For each experiment, the background signal (BG) was subtracted from the bleached 56 57 58 59

8 60 ACS Paragon Plus Environment

Page 9 of 40

Langmuir

1 3

2

ROI and then normalized to their initial pre-bleaching fluorescence intensity. In order to correct 5

4

for any photobleaching during the measurement, the normalized bleached ROI intensity was 6 7

divided by the normalized intensity of the reference region: 10

9

8

FROI  BG FROI ( prebleache d )  BG 12

1 13 14

FREF  BG FREF ( prebleache d )  BG

16

15

The corrected fluorescence curves ƒ (from 50 separate experiments) were used to construct an 18

17

average FRAP curve which was then fitted to a one component fit model describing one 19 20

diffusive species according to Equation 1 23

2

21

f (t )  A(1  et ) 24

Equation 1

25

where A is the ratio of mobile to immobile populations and  is the half-time of fluorescence 28

27

26

recovery (i.e. the diffusion time required to recover 50% of initial fluorescence intensity). 29 30

Taking into account that the reported half-time of recovery corresponds to the fastest recovery 3

32

31

time that can be measured with the confocal set up and experimental parameters described 35

34

above, a lower limit of the diffusion constant D was calculated according to Equation 2.31 36 38

37

D  0.224.w2 /  40

39

Equation 2

where w is the radius of the circular ROI. 41 42

All data processing and fitting were performed using Kaleidagraph (Synergy software). 45

4

43

2.7 Quantification of % Co-localization of Cytoskeletal Network with SS-BLM Co47

46

existing Lipid Domains. The cytoskeletal co-localization with co-existing lipid phases is 48 50

49

quantified for at least 50 SS-BLMs for each lipid system. The SS-BLMs are co-cultured with 52

51

hippocampal neurons (DIV 9) for 24 hrs and are then fixed and immunostained for both 54

53

microtubules and actin. The more ordered lipid phases (either gel phase or liquid ordered (Lo)) 5 57

56

are labeled using N-Rh-DHPE or DiI-C20, respectively. The co-localization between the 58 59

9 60 ACS Paragon Plus Environment

Langmuir

Page 10 of 40

1 3

2

cytoskeletal filament and the ordered lipid phases was quantified by measuring the overlap of 5

4

their respective fluorescence intensity signals across the surface of the SS-BLM. On the other 6 7

hand, co-localization with the fluid domains is measured by the absence of fluorescence signals 10

9

8

overlap since the fluid domains are not fluorescently-labeled in this experimental set up (see 12

1

Supporting Information, Section 1) 13 14 15 17

16

3. RESULTS AND DISCUSSION 19

18

3.1 Co-existence of Lipid Microdomains on SS-BLMs. Lipid phase separation on the 20 2

21

spherical solid support is established by varying the composition of the lipid mixture and tuning 24

23

the procedure for preparing the SS-BLMs. Figure 1 shows the co-existence of these lipid 25 26

microdomains on SS-BLMs. Visualization of lipid phase separation, using confocal microscopy, 27 29

28

is achieved using fluorescent lipid markers which preferentially partition into different 31

30

microdomains.32 Different combinations of synthetic lipids as well as lipid dyes confirm phase 32 3

separation in the SS-BLM. As seen in Figures 1a and 1c, the images are representative of the 36

35

34

sample population. In these experiments, a binary lipid mixture was used with the appropriate 38

37

combination of fluorescent lipids. Bodipy-PC (Figure 1, green) and DiI-C20 (Figure 1, red) are 39 40

used to identify co-existing microdomains in a DOPC/DSPC lipid mixture.33,34 42

41 43 4 45 46 47 48 49 50 51 52 53 54 5 56 57 58 59

10 60 ACS Paragon Plus Environment

Page 11 of 40

Langmuir

1 3

2

Figure 1. Visualization of phase separated lipid microdomains on SS-BLMs using confocal 5

4

fluorescence microscopy. Representative confocal cross-sectional images (a & b) of the binary 6 7

lipid mixture DOPC/DSPC (70:30), where the ordered domains (DSPC-rich) are labeled using 10

9

8

0.1 mol% DiI-C20 (red) and the fluid domains (DOPC-rich) are labeled using 0.1 mol% Bodipy12

1

PC (green). Representative confocal 3D-reconstruction images (c & d) of the same lipid mixture. 13 14

In panel (c) only the fluid domains (DOPC-rich) are shown in white. In all preparations, 5 µm 17

16

15

silica beads, coated with avidin, were used as the solid spherical support and 0.1 mol% DSPE19

18

PEG2000-biotin was used in the lipid mixture for tethering purposes. It is important to note that 20 21

although the extent to which each dye occupies a distinct phase is not definitively known,13 the 24

23

2

observed contrast is consistent with preferential partitioning. This figure thus establishes that 25 26

lipid domains are present in the SS-BLMs studied. 27 28 30

29

3.2 Domain Shape and Organization. Cholesterol is an important component of membrane 32

31

rafts and studies using cholesterol-containing lipid model systems have shown that net changes 34

3

in cholesterol content considerably influence lipid organization and diffusion within the 35 36

membrane.35 The effect of cholesterol on the stability and organization of the co-existing lipid 39

38

37

microdomains in SS-BLMs is illustrated in Figure 2 a-b. Inclusion of 30 mol% cholesterol into a 40 41

phase-separated binary lipid mixture causes a noticeable change in the ordering of fluid phase 4

43

42

(DOPC-rich) domains within the gel phase (DSPC-rich) domains. Comparison of Figures 2 a and 46

45

2 b reveals that the fluid domains (DOPC-rich, labeled) are more connected and branched within 47 48

the ordered domains (DSPC-rich, unlabeled) when cholesterol is present in the mixture, resulting 51

50

49

in a decrease in the net area of the ordered phase. The effect of cholesterol starts to appear at 53

52

concentrations of 10 and 20 mol% cholesterol (images not shown), becoming more prominent at 54 5

30 mol%.16 By modulating the lipid composition and lipid/cholesterol ratio, the SS-BLM can 57

56 58 59

11 60 ACS Paragon Plus Environment

Langmuir

Page 12 of 40

1 3

2

serve as a stable bilayer model system which mimics native cellular membranes and incorporates 5

4

lipid domains and raft components. 6 7

It has also been suggested that temperature of the lipid film formation and the temperature at 10

9

8

which the film is hydrated should be kept above the characteristic lipid phase transitions (Tm) of 12

1

the constituent lipids.36 In order to test if these conditions contribute to the quality of the 13 14

resulting membrane domains, the SS-BLMs are subjected to multiple temperature cycles through 17

16

15

the Tm after assembling the lipid bilayer on the spherical support as previously reported. 27 The 19

18

application of heat/cool/heat cycles at a relatively slow rate (0.1 ˚C/second) promotes the 20 2

21

segregation of lipid phases into well-defined co-existing two-phase milieu (Figure 2 e-f), similar 24

23

to that reported in repetitive freeze-thaw cycling of small unilamellar vesicles (SUVs).37 These 25 26

samples were examined using confocal microscopy through a time series study of 24 hrs after 27 29

28

preparation, which was found to be a period of time sufficient for the lipid microdomains to 31

30

achieve a steady state structure on the micron scale (Figure 2 g). 32 3

Similar to the effects of cholesterol, temperature and time, the size of the starting lipid vesicles 36

35

34

(used in the preparation of the SS-BLMs) influences the organization of lipids and resulting 38

37

shape of co-existing domains. For example, in a 70:30 DOPC/DSPC lipid mixture, relatively 39 41

40

smaller, disconnected fluid domains (Figure 2 a) are reproducibly formed when SUVs of 43

42

diameter < 200 nm were used to prepare the SS-BLMs. On the other hand, when 1-3 µm GUVs 45

4

are used to prepare the SS-BLMs (Figure 2 c), a more connected network of fluid domains 46 48

47

(Figure 2 d) results. Using GUVs of sizes ≥ 10 µm (data not shown) led to a wide variation in the 50

49

domain shape and size. These observations are likely related to vesicle fusion and rupture 52

51

processes taking place during the formation of SS-BLMs.38 This might also contribute to a 53 5

54

rearrangement of lipid components that takes place as vesicles fuse onto the solid substrate39 and 56 57 58 59

12 60 ACS Paragon Plus Environment

Page 13 of 40

Langmuir

1 3

2

these factors altogether influence the individual shape, size and distribution of c-existing lipid 5

4

domains on SS-BLMs. When using larger GUVs in the preparation of SS-BLMs, only a small 6 7

number fuse to the spherical solid support, resulting in a noticeable variation between different 10

9

8

sample preparations. This however is not the case when using SUVs. 1 12 13 14 15 16 17 18 19 20 21 2 23 24 25 26 27 28 29 30

Figure 2. Visualization of the shape and organization of phase separated lipid microdomains on 31 3

32

SS-BLMs and GUVs using confocal fluorescence microscopy as a function of multiple factors: 35

34

confocal 3D-reconstruction images of the binary lipid mixture DOPC/DSPC (70:30) with no 36 37

cholesterol (a) and with 30% cholesterol (b). Representative GUVs (c) and corresponding SS40

39

38

BLMs (d) formed from the same lipid mixture. SS-BLMs of the ternary lipid mixture 42

41

DOPC/DPPE/DOTAP (25:50:25) with ordered domains (DPPE-rich) prior to (e) and after (f) the 43 45

4

application of two heat/cool/heat cycles starting at 4 ˚C, passing through the Tm of DPPE at 63 47

46

˚C and ending at 80 ˚C. Panel g displays a time series collected following the temperature cycles 49

48

(scale bars are 1 µm). (a-d) The fluid domains (DOPC-rich) are labeled using 0.1 mol% Bodipy50 51

PC and (e-g) the ordered domains (DPPE-rich) are labeled using 0.1 mol% N-Rh-DHPE.40 In all 54

53

52

preparations, 5 µm silica beads, coated with avidin, were used as the solid spherical support and 5 56

0.1 mol% DSPE-PEG2000-biotin was used in the lipid mixture for tethering purposes. 58

57 59

13 60 ACS Paragon Plus Environment

Langmuir

Page 14 of 40

1 3

2

3.3 SS-BLM Fluidity and Lipid Diffusion Characteristics. The dynamics of the lipids 5

4

within the SS-BLMs were studied in order to evaluate their ability to self organize within their 6 7

respective microdomains as well as to further re-organize when interfaced with cellular 10

9

8

membranes. Despite the use of biotin-avidin tethering for their preparation, the SS-BLMs retain 12

1

their fluidity27 as confirmed using comparative fluorescence recovery after photobleaching 13 14

(FRAP).41 This involves evaluating the diffusion of fluorescent lipids included in the SS-BLM 17

16

15

lipid mixture. Figure 3 summarizes a FRAP study for SS-BLMs from DOPC lipids labeled using 19

18

0.1 mol% of the fluorescent lipid Bodipy-PC. The apparent diffusion coefficients ( D ), half-life 20 2

21

of fluorescence recovery (), and the ratio of mobile to immobile lipid molecules are measured 24

23

and compared to those of non-supported bilayers, i.e. GUVs of equivalent size (Table 1). 25 26 27 28 29 30 31 32 3 34 35 36 37 38 39 40 41 42 43 4 45 46 48

47

Figure 3. Diffusion properties of DOPC SS-BLMs labelled using 0.1 mol% Bodipy-PC and 50

49

tethered on 5 µm avidin coated silica beads using 0.1 mol % DSPE-PEG2000-biotin: (a) time 51 52

series displaying fluorescence recovery after photobleaching of a circular ROI (1 µm, shown in 5

54

53

red). Fluorescence intensity data were collected from an additional non-bleached reference 56 57 58 59

14 60 ACS Paragon Plus Environment

Page 15 of 40

Langmuir

1 3

2

circular ROI of the same diameter (1 µm, not shown) and used in subsequent data analysis to 5

4

correct for any bleaching occurring during imaging. (b) Averaged fluorescence data and 6 7

corresponding standard error for bleached (shown in black) and reference (shown in green) 10

9

8

regions. The data correspond to 50 bleaching experiments on different SS-BLMs and collected in 12

1

a single experimental set up. After normalization to pre-bleaching fluorescence levels, the 13 14

averaged FRAP data is fit (curve shown in grey) to a one diffusing component model (R value of 17

16

15

0.989). (c) Histogram displaying the frequency of different % mobile fractions measured for the 19

18

50 SS-BLMs from the same sample preparation. 20 21 2

Lipid System Diffusion Half-time  (s) Diffusion constant D (µm2/s)[b] Mobile fraction (%)

29

28

27

26

25

24

23

GUVs[a]

0.315

1.02

91.9

SS-BLMs[a]

0.358

0.901

75.1

[a] lipid mixture composed of DOPC labeled using 0.1 mol% Bodipy-PC and tethered using 0.1 mol% DSPE-PEG2000-biotin 32

31

30

[b] Due to the spherical nature of the lipid bilayers, the equations which have been derived for planar systems become unsuitable to analyze the FRAP data. Therefore an apparent diffusion coefficient was estimated from the fluorescence recovery half time measured on these curves using the equation D  0.224.w2 / (see Section 2.6) and used for comparative purposes rather than to report an absolute value. 39

38

37

36

35

34

3

Table 1. Summary of diffusion parameters 40 42

41

Recovery of fluorescence in the SS-BLM confirms the presence of a continuous and fluid lipid 43 45

4

bilayer membrane coating the solid support, as opposed to a layer of adhered vesicles. The 47

46

diffusivity of Bodipy-PC fluorescent lipids in homogenous DOPC membranes that were either 48 49

supported (SS-BLMs) or free standing (GUVs) are very similar ( ca. 0.3 s), suggesting that the 52

51

50

tethering caused by the biotinylated lipid (DSPE-PEG2000-biotin at 0.1 mol%) binding to 54

53

avidin-coated silica bead does not significantly hinder the diffusion of supported lipids. The 5 57

56

estimated diffusion constant is in agreement with previous reports for DOPC lipid bilayers 58 59

15 60 ACS Paragon Plus Environment

Langmuir

Page 16 of 40

1 2

supported on planar glass substrates (values ca. 1 – 2.5 µm2/s).42,43,44,45,46 However, the measured 5

4

3

fractions of mobile fluorescent lipid molecules in SS-BLMs vary significantly (Figure 3c). This 6 7

variability may be related to the actual quantity of lipopolymers (i.e., DSPE-PEG2000-biotin) 10

9

8

incorporated in the SS-BLMs and possibly their uneven distribution between the inner and outer 12

1

membrane leaflets. 13 14

3.4 Domain-Specific Cytoskeletal Organization. Scheme 1 depicts the steps involved in a 17

16

15

typical experiment involving SS-BLMs and their interactions with living cells: (i) SS-BLMs with 19

18

co-existing microdomains are prepared on 5 µm silica beads, (ii) SS-BLMs are added to cells in 20 2

21

culture (in this case rat embryonic hippocampal neural culture) and are allowed to interact with 24

23

the living cells for up to 24 hrs,28 and (iii) immunofluorescence and confocal microscopy are 25 26

used to examine the organization of the cellular proteins (in this case the cytoskeletal network) in 27 29

28

relation to the lipid microdomains in the SS-BLMs. As depicted in part (iii) of Scheme 1, actin 31

30

filaments and microtubules preferentially extend and assemble around the fluid lipid domains 32 3

rather than the gel or solid-like lipid domains. Actin and microtubules are major cytoskeletal 36

35

34

components known to be involved in important cellular processes including the maintenance of 38

37

cell shape, providing mechanical support, signal transduction, axon path-finding, and synaptic 39 40

vesicle trafficking.47,48 The actin cytoskeleton has also been shown to be critical in establishing 43

42

41

raft formation in cell membranes.11,12,49 To our knowledge, this is the first report of the 45

4

relationship between lipid microdomains on a model membrane and a raft-influenced component 46 47

in living cells, such as the cytoskeleton organization.11,12,49 The SS-BLM thus provides a 50

49

48

platform for mechanistic studies involving different contributors to membrane heterogeneity. 51 52 53 54 5 56 57 58 59

16 60 ACS Paragon Plus Environment

Page 17 of 40

Langmuir

1 2 3 4 5 6 7 8 9 10 1 13

12

Scheme 1. Scheme (not to scale) illustrating the preparation of SS-BLMs displaying co-existing 14 16

15

lipid microdomains and their subsequent interaction with living cells. 17 19

18

Cytoskeleton-induced domain formation11 is known to be one of the factors which influence 21

20

membrane heterogeneity in biological membranes. Studies have shown that the cytoskeletal 2 23

networks are important in establishing and maintaining membrane organization.11,12,49 For 26

25

24

example, Liu et. al. have shown that actin networks can control the spatial and temporal 28

27

organization of lipid domains.50 This was demonstrated by allowing dendritic actin monomers to 29 30

polymerize on model membranes (GUVs) which exhibit lipid domains.50 The importance of 3

32

31

obtaining new insights into the coupling of the model membrane bilayer and native membrane 35

34

skeleton was stressed.51 38

37

36

The SS-BLM platform described here provides for facile access to BLMs with robust yet fluid 40

39

lipid microdomains. Two different lipid phase domain situations were used to explore the spatial 41 42

correlation between lipid microdomains and the cytoskeleton of living cells. One involves co45

4

43

existing “liquid disordered (Ld)  liquid ordered (Lo)” phases (Figure 4) and the other involves 46 47

co-existing “fluid  gel” phases (Figure 5). Figure 4 shows a representative primary neuron/SS50

49

48

BLM co-culture, where SS-BLMs consisting of DOPC/DPPC/Chol (50:30:20) were incubated 52

51

with living cells in culture up to 24 hrs. The addition of 0.1 mol% DiI-C20 was used to label (red) 53 5

54

the Lo phase (DPPC-rich). In this study, cellular microtubules, an important dynamic cytoskeletal 57

56

element that is responsible for intracellular transport are labeled (Figure 4 a and c; green) using 58 59

17 60 ACS Paragon Plus Environment

Langmuir

Page 18 of 40

1 3

2

β-tubulin primary antibodies via immunocytochemistry and filamentous actin, another important 5

4

cytoskeletal element that is involved in cell motility and in cell signaling are labeled (Figure 4 b 6 7

and e; green) using Alexa-488−phalloidin. The magnified views in Figures 4c and 4e show the 10

9

8

close association between the cytoskeleton filaments and the lipid domains of the SS-BLMs, in 12

1

this case the unlabeled region (Ld phase; DOPC-rich). Figures 4d and 4f provide additional views 13 14

of the domain organization on the SS-BLM. It is important to note that the fluorescence 17

16

15

visualized in the SS-BLMs derives solely from the proximal surface, because the excitation light 19

18

does not pass through the silica core of the SS-BLM. 20 21 2 23 24 25 26 27 28 29 30 31 32 3 34 35 36 38

37

Figure 4. Representative confocal 3D-reconstruction images showing the co-localization of 40

39

cytoskeletal networks with lipid microdomains on SS-BLMs presenting an Ld  Lo phase 43

42

41

separation. Assembly of (a) microtubules (β-tubulin, green), and (b) actin (phalloidin, green) 45

4

around the fluid phase on DOPC/DPPC/ Chol (50:30:20) SS-BLMs. Panels c & e are magnified 46 48

47

views of images a & b respectively, showing the specific organization of the cytoskeletal 50

49

networks around the unlabeled regions on the SS-BLMs, representing the fluid phase (DOPC51 52

rich). Panels d & f are single channel images of c & e, showing the exact location of ordered 53 5

54

(DPPC-rich) domains on SS-BLMs that are labeled using 0.1 mol% DiI-C20 (red). The SS-BLMs 56 57 58 59

18 60 ACS Paragon Plus Environment

Page 19 of 40

Langmuir

1 3

2

are co-cultured with hippocampal neurons (DIV 9) for 24 hrs and immunostained for either 5

4

microtubules or actin filaments. 6 7 8 9 10 1 12 13 14 15 16 17 18 19 20 21 2 23 24 25 26 27 28 29 30 31 32 34

3

Figure 5. A representative 3-channel confocal 3D-recontruction image showing the co36

35

localization of cytoskeletal networks with lipid microdomains on SS-BLMs presenting a fluid  37 38

gel phase separation: (a) assembly of microtubules (β-tubulin, green) and actin (phalloidin, blue) 41

40

39

around the fluid phase on DOPC/DPPE/DOTAP (25:50:25) SS-BLMs. The ordered domains 43

42

(DPPE-rich) are labeled using 0.1 mol% N-Rh-DHPE (red). Magnified views of image “a” show 4 46

45

microtubule (b) and actin (c) co-localization with the lipid microdomains. 47 49

48

The association between the cellular cytoskeleton and the lipid microdomains derived from a 51

50

ternary lipid mixture composed of DOPC/DPPE/DOTAP (25:50:25) was also studied. This lipid 52 53

mixture exhibits fluid  gel phase domain co-existence at 37 ˚C and has also been shown to 56

5

54

induce interesting cellular responses when model membranes containing these lipids interact 57 58 59

19 60 ACS Paragon Plus Environment

Langmuir

Page 20 of 40

1 2

with biological membranes.28,52 We recently reported the ability of such model membranes to 5

4

3

induce artificial synapse formation when interfaced with hippocampal neurons.28 Similar to the 6 7

DOPC/DPPC/Chol mixture discussed in Figure 4, the association of cytoskeletal networks with 10

9

8

the SS-BLM lipid domains is observed. Figure 5 is a representative 3D confocal image showing 12

1

co-staining of F-actin (blue) and microtubules (green) when primary neurons were co-cultured 13 14

with SS-BLMs consisting of DOPC/DPPE/DOTAP. In this case, 0.1 mol% N-Rh-DHPE was 17

16

15

used to label (red) the gel phase (DPPE-rich).40 Figures 5 b and 5 c are magnified views of 19

18

individual channels of the selected area (white box) in Figure 5a. Both channels show that the 20 2

21

cytoskeletal labeling is preferentially co-localized with the disordered fluid phase (unlabeled 24

23

dark region) of the SS-BLMs. 26

25

This study, as well as that of Liu et. al.,50 uses DOPC as the fluid phase lipid. The question 27 29

28

arises as to whether a domain-specific interaction correlates with cytoskeletal organization, or if 31

30

molecular specificity is also a determinant. The adaptability of the SS-BLM system to such 32 3

experimental questions is exemplified in the ability to modulate the lipid compositions and 36

35

34

functionalities in order to assess each of these possibilities. For example, SS-BLMs consisting of 38

37

a DOPE-rich fluid phase and a DPPC-rich gel phase were examined (see Supporting Information 39 41

40

Figure S1). Thus, unlike the lipid composition used in Figures 4 and 5, in this experiment the 43

42

fluid phase involves phosphatidylethanolamine (PE) headgroups and the gel phase involves 45

4

phosphatidylcholine (PC) headgroups. The observed cytoskeletal organization (favoring the fluid 46 48

47

phase) however remains unchanged with this change in molecular specification. In addition, the 50

49

inclusion of a cationic lipid (such as DOTAP) acts to promote the adhesion of the cells to SS52

51

BLMs.28 This is concluded by the closer assembly of the cellular membranes around them. 53 5

54

However, DOTAP does not influence the phase-specific cytoskeleton co-localization in the lipid 56 57 58 59

20 60 ACS Paragon Plus Environment

Page 21 of 40

Langmuir

1 3

2

mixtures examined (DOPC/DPPC/Chol or DOPC/DPPE). Moreover, it appears that the 5

4

cytoskeletal filaments direct the cells away from associating with the surface of those SS-BLMs 6 7

which do not display phase-separated co-existing lipid microdomains. Control studies conducted 10

9

8

using SS-BLMs with uniform compositions that do not exhibit lipid phase separation (for 12

1

example 100% DOPC or 100% DPPE) revealed non-specific cellular organization and no 13 14

significant interactions between the living cells with the co-cultured SS-BLMS (see Supporting 17

16

15

Information Figure S2). 19

18

3.5 Comparison to GUVs. The stability of model membranes is a critical factor when 20 2

21

considering their applications. It is important to note that we observed that the fragility of GUVs 24

23

precludes their use when the experimental protocol involves either or both in vitro cell culture 25 26

and immunostaining methods (see Supporting Information Table S1). This is consistent with 27 29

28

reported limitations of GUVs25,26 and is due to a combination of factors such as: (i) shrinkage and 31

30

rupture when exposed to detergents53,54,55,56,57 during immunostaining procedure, (ii) structural 32 3

fluctuations and deformation due to osmotic stress resulting from using multiple solutions of 36

35

34

different salt concentrations58,59 (iii) general sensitivity to solution conditions and environmental 38

37

changes (i.e. pH, temperature, extensive washings and so on)60 and (iv) unexpected topological 39 41

40

transformations, vesicle budding or fusion occurring during long incubation periods especially in 43

42

the presence of non-liposomal components in cell culture medium. 45

4

3.6 Quantification of Domain-Specific Cytoskeletal Organization. The co-localization of 46 48

47

the fluorescently-labeled microtubules and actin filaments with either one of co-existing SS50

49

BLM domains was quantified. As described in Section 2.7, this was measured by the presence or 51 52

absence of a spatial overlap of their respective fluorescence signals (Figure 6). 53 54 5 56 57 58 59

21 60 ACS Paragon Plus Environment

Langmuir

Page 22 of 40

1 2 3 4 5 6 7 8 9 10 1 12 13 14 15 17

16

Figure 6. Quantification of co-localization between cytoskeletal filaments and SS-BLM co18 19

existing domains from the lipid mixture DOPC/DPPE/DOTAP (25:50:25), where the ordered 2

21

20

phase (DPPE-rich) is labeled using 0.1 mol% N-Rh-DHPE (red). Panels (c, d) displaying the 24

23

fluorescence intensity profiles across an area of the SS-BLM (indicated by white lines in images 25 27

26

a and b), using the same color codes for the fluorescence channels where microtubules are 29

28

labeled in green (a) and actin is labeled in blue (b). (Scale bars = 2 µm). 30 32

31

The % preferential co-localization with the more fluid phases from different lipid systems is 3 35

34

summarized in Figure 7. In the case of microtubules and F-actin filaments, it was found that 37

36

more than 50% fluorescence co-localization occurs with the more fluid phases for SS-BLM 39

38

populations from different lipid mixtures. It is important to note that experiments involving no 40 42

41

phase separation (e.g. only fluid phase or solid phase present) did not show comparatively high 4

43

promotion of cytoskeletal network assembly. Phase separated lipid state clearly promotes 45 46

cytoskeleton preferential assembly. 47 48 49 50 51 52 53 54 5 56 57 58 59

22 60 ACS Paragon Plus Environment

Page 23 of 40

Langmuir

1 2 3 4 5 6 7 8 9 10 1 12 13 14 15 16 17 18

Figure 7. Preferential co-localization of cytoskeletal filaments with lipid phase domains in SS19 21

20

BLMs. % co-localization is calculated with respect to the single lipid phase present or with the 23

2

disordered phase when co-existing lipid phases are present. For quantification details see 24 25

Experimental Section 2.7 and Figure 6. 27

26

29

28

These observations are consistent with suggestions that components of the cellular 31

30

cytoskeleton regulate and/or favor membrane heterogeneity in lipid membranes and, in 3

32

particular, in biological membranes.11 Although the mechanism through which this correlated 36

35

34

action is regulated still not understood, the system and experimental approach presented here 38

37

offers a novel platform for investigating the collective role of multiple raft components in 39 40

regulating membrane heterogeneity. 43

42

41

It is important to note that although primary neuronal cultures were used in the experiments 45

4

described here, the generality of this approach was demonstrated by also performing the cell 46 48

47

interaction experiments using COS-7 cell lines. As shown in Supporting Information Figure S3, 50

49

the cytoskeletal networks follow the same trend observed in primary neuronal cultures. 51 52

4. Conclusions 53 54 5 56 57 58 59

23 60 ACS Paragon Plus Environment

Langmuir

Page 24 of 40

1 3

2

The results presented here establish that the lipid microdomains in SS-BLMs interact with 5

4

living cells in culture. Because they are both robust and dynamic, SS-BLM domains can 6 7

withstand cell culture conditions and the experimental manipulations necessary to investigate 10

9

8

their interactions with cellular components of living cells. As a demonstration of their versatility, 12

1

experiments presented here follow the organization of cytoskeleton networks as a function of the 13 14

specific lipid domain present. The interactions with living cells explored here are in very good 17

16

15

agreement with those performed on GUVs in combination with purified and/or synthetic actin 19

18

monomers.50 Since both lipid domains and the cellular cytoskeleton clearly contribute to cell 20 2

21

membrane heterogeneity, the SS-BLM system provides a means to further address the 24

23

fundamental relationship between membrane heterogeneity and membrane-mediated functions. 25 26

Future experiments will help establish if certain types of lipids, adhesion molecules and/or actin27 29

28

binding proteins associated with the cell membranes also take part in the observed lipid domain 31

30

preferential organization of cellular components. Finally, because of its simplicity, robustness 32 3

and experimental versatility, the SS-BLM platform is an attractive complement to GUVs and 36

35

34

planar S-BLMs in membrane biophysical studies, especially in experiments involving live cell 38

37

cultures as well as in guiding the development of new materials for bioengineering applications. 39 40 41 43

42

ASSOCIATED CONTENT 4 46

45

Supporting Information. 47 48 50

49

Experimental details of control experiments. This material is available free of charge via the 52

51

Internet at http://pubs.acs.org. 53 54 5

AUTHOR INFORMATION 57

56 58 59

24 60 ACS Paragon Plus Environment

Page 25 of 40

Langmuir

1 3

2

Corresponding Authors 5

4

*† Gopakumar Gopalakrishnan. * R. Bruce Lennox. 6 8

7

* Department of Chemistry, McGill University, 801 Sherbrooke Street West, Montreal (QC) 9 1

10

H3A 0B8, Canada Fax: (+1) (514) 398-3797, E-mail: [email protected] 12 14

13

† Current address: 16

15

Institut Galien Paris-Sud (UMR CNRS 8612), Université Paris-Sud 17 19

18

5 rue Jean-Baptiste Clément, 92296 Châtenay-Malabry, France 21

20

E-mail: [email protected]. 2 23 24

Funding Sources 25 27

26

This work was funded by the NSERC CREATE Neuroengineering training grant and an NSERC 28 29

Discovery Grant (RBL). CM is a recipient of an NSERC doctoral scholarship. 30 31 3

32

ACKNOWLEDGMENT 35

34

Image processing was performed in the McGill Life Sciences Complex Imaging Facility. Dr. 36 37

Claire Brown (Director, LSC Imaging Facility) is acknowledged for the introduction into image 40

39

38

processing. Drs. Patricia T. Yam and Dalinda Liazoghli are thanked for their assistance with 42

41

primary cell culture and immunocytochemistry protocols. Dr. Asmahan Abu Arish is thanked for 43 45

4

her help with FRAP experiments data collection and analysis. 47

46

REFERENCES 49

48

(1) 50

Brown, D. A.; London, E. Functions of lipid rafts in biological membranes. Annu Rev

52

51

Cell Dev Bi 1998, 14, 111-136. 54

53

(2) 5

Mueller, N.S.; Wedlich-Soldner, R.; Spira, F. From mosaic to patchwork: Matching lipids

56

and proteins in membrane organization. Mol Membr Biol 2012, 29, 186-196. 57 58 59

25 60 ACS Paragon Plus Environment

Langmuir

Page 26 of 40

1 3

2

(3) 5

4

Rietveld, A.; Simons, K. The differential miscibility of lipids as the basis for the

formation of functional membrane rafts. Biochim Biophys Acta Rev Biomembr 1998, 1376, 4676 7

479.

13

12

1

10

9

8

(4)

Simons, K.; Ikonen, E. Functional rafts in cell membranes. Nature 1997, 387, 569-572.

(5)

Lingwood, D.; Simons, K. Lipid rafts as a membrane-organizing principle. Science 2010,

14

327, 46-50. 17

16

15

(6) 19

18

Coskun, U.; Simons, K. Cell membranes: the lipid perspective. Structure 2011, 19, 1543-

1548. 20 2

21

(7) 24

23

Brown, D. A.; London, E. Structure and origin of ordered lipid domains in biological

membranes. J Membrane Biol 1998, 164, 103-114. 25 26

(8) 27

Thomas, J. L.; Holowka, D.; Baird, B.; Webb, W. W. Large-scale coaggregation of

29

28

fluorescent lipid probes with cell-surface proteins. J Cell Biol 1994, 125, 795-802. 31

30

(9) 32

van Meer, G.; Voelker, D. R.; Feigenson, G. W. Membrane lipids: where they are and

3

how they behave. Nat Rev Mol Cell Bio 2008, 9, 112-124. 36

35

34

(10) 38

37

Johnston, I.; Johnston, L. J. Ceramide promotes restructuring of model raft membranes.

Langmuir 2006, 22, 11284-11289. 39 41

40

(11) 43

42

Chichili, G. R.; Rodgers, W. Cytoskeleton-membrane interactions in membrane raft

structure. Cell Mol Life Sci 2009, 66, 2319-2328. 45

4

(12) 46

Holowka, D.; Sheets, E. D.; Baird, B. Interactions between Fc epsilon RI and lipid raft

48

47

components are regulated by the actin cytoskeleton. J Cell Sci 2000, 113, 1009-1019. 50

49

(13) 51

Bagatolli, L. A. To see or not to see: Lateral organization of biological membranes and

52

fluorescence microscopy. Biochim Biophys Acta Biomembr 2006, 1758, 1541-1556. 53 54 5 56 57 58 59

26 60 ACS Paragon Plus Environment

Page 27 of 40

Langmuir

1 3

2

(14) 5

4

Hac, A. E.; Seeger, H. M.; Fidorra, M.; Heimburg, T. Diffusion in two-component lipid

membranes - A fluorescence correlation spectroscopy and Monte Carlo simulation study. 6 7

Biophys J 2005, 88, 317-333. 10

9

8

(15) 12

1

Honerkamp-Smith, A. R.; Veatch, S. L.; Keller, S. L. An introduction to critical points

for biophysicists; observations of compositional heterogeneity in lipid membranes. Biochim 13 14

Biophys Acta Biomembr 2009, 1788 , 53-63. 17

16

15

(16) 19

18

Korlach, J.; Schwille, P.; Webb, W. W.; Feigenson, G. W. Characterization of lipid

bilayer phases by confocal microscopy and fluorescence correlation spectroscopy. P Natl Acad 20 2

21

Sci USA 1999, 96, 8461-8466. 24

23

(17) 25

McConnell, H. M.; Vrljic, M. Liquid-liquid immiscibility in membranes. Annu Rev Bioph

26

Biom 2003, 32, 469-492. 27 29

28

(18) 31

30

Mohwald, H. Phospholipid and phospholipid-protein monolayers at the air/water

interface. Annu Rev Phys Chem 1990, 41, 441-476. 32 3

(19) 36

35

34

Bagatolli, L. A.; Gratton, E. Two-photon fluorescence microscopy observation of shape

changes at the phase transition in phospholipid giant unilamellar vesicles. Biophys J 1999, 77, 38

37

2090-2101. 39 41

40

(20) 43

42

Groves J. T.; Ulman, N.; Cremer, P. S.; Boxer, S. G. Substrate-membrane interactions:

mechanisms for imposing patterns on a fluid bilayer membrane. Langmuir 1998, 12, 3347-3350. 45

4

(21) 46

Tamm, L. K.; McConnell, H. M. Supported phospholipid-bilayers. Biophys J 1985, 47,

48

47

105-113. 50

49

(22) 51

Merkel, R.; Sackmann, E.; Evans, E. Molecular friction and epitactic coupling between

52

monolayers in supported bilayers. J. Phys. (Paris) 1989, 50 , 1535-1555. 53 54 5 56 57 58 59

27 60 ACS Paragon Plus Environment

Langmuir

Page 28 of 40

1 3

2

(23) 5

4

Anderson, T. H.; Min, Y. J.; Weirich, K. L.; Zeng, H. B.; Fygenson, D.; Israelachvili, J.

N. formation of supported bilayers on silica substrates. Langmuir 2009, 25, 6997-7005. 6 7

(24) 10

9

8

Prashar, J.; Sharp, P.; Scarffe, M.; Cornell, B. Making lipid membranes even tougher. J

Mater Res 2007, 22, 2189-2194. 12

1

(25) 13

Schwille, P.; Diez, S. Synthetic biology of minimal systems. Crit Rev Biochem Mol 2009,

14

44, 223-242. 17

16

15

(26) 19

18

Chemburu, S.; Fenton, K.; Lopez, G. P.; Zeineldin, R. Biomimetic silica microspheres in

biosensing. Molecules 2010, 15, 1932-1957. 20 2

21

(27) 24

23

Gopalakrishnan, G.; Rouiller, I.; Colman, D. R.; Lennox, R. B. Supported bilayers

Formed from different phospholipids on spherical silica substrates. Langmuir 2009, 25, 545525 26

5458. 27 29

28

(28) 31

30

Gopalakrishnan, G.; Thostrup, P.; Rouiller, I.; Lucido, A. L.; Belkaid, W.; Colman, D.

R.; Lennox, R. B. Lipid bilayer membrane-triggered presynaptic vesicle assembly. Acs Chem 32 3

Neurosci 2010, 1, 86-94. 36

35

34

(29) Bayerl, T.M.; Bloom, M. Physical properties of single phospholipid bilayers adsorbed to 38

37

micro glass beads. A new vesicular model system studied by 2H-nuclear magnetic resonance. 39 41

40

Biophys J 1990, 58, 357-362. 43

42

(30) 45

4

Goslin, K.; Asmussen, H.; Banker, G. In Culturing Nerve Cells. Banker G., Goslin K,

Eds.; 2nd ed.; MIT Press: Cambridge, Massachusetts, 1998; p 666. 46 48

47

(31) 50

49

Soumpasis, D. M. Theoretical analysis of fluorescence photobleaching recovery

experiments. Biophys J 1983, 41, 95-97. 51 52

(32) 53

Davenport, L. Fluorescence probes for studying membrane heterogeneity. Method

5

54

Enzymol 1997, 278, 487-512. 56 57 58 59

28 60 ACS Paragon Plus Environment

Page 29 of 40

Langmuir

1 3

2

(33) 5

4

Huang, N. N.; Florinecasteel, K.; Feigenson, G. W.; Spink, C. Effect of fluorophore

linkage position of normal-(9-anthroyloxy) fatty-acids on probe distribution between coexisting 6 7

gel and fluid phospholipid phases. Biochim Biophys Acta 1988, 939, 124-130. 10

9

8

(34) 12

1

Spink, C. H.; Yeager, M. D.; Feigenson, G. W. Partitioning behavior of indocarbocyanine

probes between coexisting gel and fluid phases in model membranes. Biochim Biophys Acta 13 14

1990, 1023, 25-33. 17

16

15

(35) 19

18

Silvius, J. R. Role of cholesterol in lipid raft formation: lessons from lipid model systems.

Biochim Biophys Acta Biomembr 2003, 1610, 174-183. 20 2

21

(36) 24

23

Gennis, R. B. The Structures and Properties of Membrane Lipids. In Biomembranes:

molecular structure and function; Cantor, C. R., Ed.; Springer: New York, 1989;. pp. 64-83 25 26

(37) 27

MacDonald, R. C.; MacDonald, R. I. Applications of freezing and thawing in liposome

29

28

technology. 3rd ed.; CRC Press: Boca Raton, 1984; Vol. 1, p 209-228. 31

30

(38) 32

Richter, R. P.; Berat, R.; Brisson, A. R. Formation of solid-supported lipid bilayers: An

3

integrated view. Langmuir 2006, 22, 3497-3505. 36

35

34

(39) 38

37

Giocondi, M. C.; Vie, V.; Lesniewska, E.; Milhiet, P. E.; Zinke-Allmang, M.; Le

Grimellec, C. Phase topology and growth of single domains in lipid bilayers. Langmuir 2001, 17, 39 41

40

1653-1659. 43

42

(40) 45

4

Bagatolli, L. A.; Gratton, E. Two photon fluorescence microscopy of coexisting lipid

domains in giant unilamellar vesicles of binary phospholipid mixtures. Biophys J 2000, 78, 29046 48

47

305. 50

49

(41) 51

Vaz, W. L. C.; Stumpel, J.; Hallmann, D.; Gambacorta, A.; Derosa, M. Bounding fluid

52

viscosity and translational diffusion in a fluid lipid bilayer. Eur Biophys J Biophy 1987, 15, 11153 5

54

115. 56 57 58 59

29 60 ACS Paragon Plus Environment

Langmuir

Page 30 of 40

1 3

2

(42) 5

4

Berquand, A.; Mazeran, P. E.; Pantigny, J.; Proux-Delrouyre, V.; Laval, J. M.;

Bourdillon, C. Two-step formation of streptavidin-supported lipid bilayers by PEG-triggered 6 7

vesicle fusion. Fluorescence and atomic force microscopy characterization. Langmuir 2003, 19, 10

9

8

1700-1707. 12

1

(43) 13

Cambrea, L. R.; Hovis, J. S. Formation of three-dimensional structures in supported lipid

14

bilayers. Biophys J 2007, 92, 3587-3594. 17

16

15

(44) 19

18

Guo, L.; Har, J. Y.; Sankaran, J.; Hong, Y. M.; Kannan, B.; Wohland, T. Molecular

diffusion measurement in lipid Bilayers over wide concentration ranges: A comparative study. 20 2

21

Chemphyschem 2008, 9, 721-728. 24

23

(45) 25

Seu, K. J.; Cambrea, L. R.; Everly, R. M.; Hovis, J. S. Influence of lipid chemistry on

26

membrane fluidity: Tail and headgroup interactions. Biophys J 2006, 91, 3727-3735. 27 29

28

(46) 31

30

Shenoy, S.; Moldovan, R.; Fitzpatrick, J.; Vanderah, D. J.; Deserno, M.; Losche, M. In-

plane homogeneity and lipid dynamics in tethered bilayer lipid membranes (tBLMs). Soft Matter 32 3

2010, 6, 1263-1274. 36

35

34

(47) 38

37

Doherty, G. J.; McMahon, H. T. Mediation, modulation, and consequences of membrane-

cytoskeleton interactions. Annu Rev Biophys 2008, 37, 65-95. 39 41

40

(48) 43

42

Pollard, T. D.; Cooper, J. A. Actin, a central player in cell shape and movement. Science

2009, 326, 1208-1212. 45

4

(49) 46

Lenne, P. F.; Wawrezinieck, L.; Conchonaud, F.; Wurtz, O.; Boned, A.; Guo, X. J.;

48

47

Rigneault, H.; He, H. T.; Marguet, D. Dynamic molecular confinement in the plasma membrane 50

49

by microdomains and the cytoskeleton meshwork. Embo J 2006, 25, 3245-3256. 51 52

(50) 53

Liu, A. P.; Fletcher, D. A. Actin polymerization serves as a membrane domain switch in

5

54

model lipid bilayers. Biophys J 2006, 91, 4064-4070. 56 57 58 59

30 60 ACS Paragon Plus Environment

Page 31 of 40

Langmuir

1 3

2

(51) 5

4

Edidin, M. Switching sides: The actin/membrane lipid connection. Biophys J 2006, 91,

3963-3963. 6 7

(52) 10

9

8

Gopalakrishnan, G.; Danelon, C.; Izewska, P.; Prummer, M.; Bolinger, P. Y.;

Geissbuhler, I.; Demurtas, D.; Dubochet, J.; Vogel, H. Multifunctional lipid/quantum dot hybrid 12

1

nanocontainers for controlled targeting of live cells. Angew Chem Int Edit 2006, 45, 5478-5483. 13 14

(53) 17

16

15

Ahyayauch, H.; Bennouna, M.; Alonso, A.; Goni, F. M. Detergent effects on membranes

at subsolubilizing concentrations: transmembrane lipid motion, bilayer permeabilization, and 19

18

vesicle lysis/reassembly are independent phenomena. Langmuir 2010, 26, 7307-7313. 20 2

21

(54) 24

23

Delamaza, A.; Parra, J. L. Vesicle-micelle structural transition of phosphatidylcholine

bilayers and triton X-100. Biochem J 1994, 303, 907-914. 25 26

(55) 27

Goni, F. M.; Alonso, A. Spectroscopic techniques in the study of membrane

29

28

solubilization, reconstitution and permeabilization by detergents. Biochim Biophys Acta 31

30

Biomembr 2000, 1508, 51-68. 32 3

(56) 36

35

34

Heerklotz, H.; Seelig, J. Titration calorimetry of surfactant-membrane partitioning and

membrane solubilization. Biochim. Biophys. Acta, Biomembr.2000, 1508, 69-85. 38

37

(57) 39

Sudbrack, T. P.; Archilha, N. L.; Itri, R.; Riske, K. A. Observing the solubilization of

41

40

lipid bilayers by detergents with optical microscopy of GUVs. J Phys Chem B 2011, 115, 26943

42

277. 45

4

(58). 46

Oglecka, K.; Sanborn, J.; Parikh, A.; Kraut, R. Osmotic gradients induce bio-reminiscent

48

47

morphological transformations in giant unilamellar vesicles. Front Physiol 2012, 3, 120. 50

49

(59) 51

Mathivet, L.; Cribier, S.; Devaux, P. F. Shape change and physical properties of giant

52

phospholipid vesicles prepared in the presence of an AC electric field. Biophys J 1996, 70, 111253 5

54

1121. 56 57 58 59

31 60 ACS Paragon Plus Environment

Langmuir

Page 32 of 40

1 3

2

(60) 5

4

Wesolowska, O.; Michalak, K.; Maniewska, J.; Hendrich, A. B. Giant unilamellar

vesicles - a perfect tool to visualize phase separation and lipid rafts in model systems. Acta 6 7

Biochim Pol 2009, 56, 33-39. 9

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

32 60 ACS Paragon Plus Environment

Page 33 of 40

Langmuir

1 2 3 4 5 6 7 8 9 10 1 12 13 14 15 16 17 18 19 20 21 2 23 24 25 26 27 28 29 30 32

31 Visualization of phase separated lipid microdomains on SS-BLMs using confocal fluorescence microscopy. Representative confocal cross-sectional images (a & b) of the binary lipid mixture DOPC/DSPC (70:30), where the ordered domains (DSPC-rich) are labeled using 0.1 mol% DiI-C20 (red) and the fluid domains (DOPC-rich) are labeled using 0.1 mol% Bodipy-PC (green). Representative confocal 3D-reconstruction images (c & d) of the same lipid mixture. In panel (c) only the fluid domains (DOPC-rich) are shown in white. In all preparations, 5 µm silica beads, coated with avidin, were used as the solid spherical support and 0.1 mol% DSPE-PEG2000-biotin was used in the lipid mixture for tethering purposes. It is important to note that although the extent to which each dye occupies a distinct phase is not definitively known, the observed contrast is consistent with preferential partitioning. This figure thus establishes that lipid domains are present in the SS-BLMs studied. 73x55mm (300 x 300 DPI)

41

40

39

38

37

36

35

34

3

42 43 4 45 46 47 48 49 50 51 52 53 54 5 56 57 58 59 60 ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 1 12 13 14 15 16 17 18 19 20 21 2 23 24 25 26 27 28 29 30 31 32 3 34 35 Visualization of the shape and organization of phase separated lipid microdomains on SS-BLMs and GUVs using confocal fluorescence microscopy as a function of multiple factors: confocal 3D-reconstruction images of the binary lipid mixture DOPC/DSPC (70:30) with no cholesterol (a) and with 30% cholesterol (b). Representative GUVs (c) and corresponding SS-BLMs (d) formed from the same lipid mixture. SS-BLMs of the ternary lipid mixture DOPC/DPPE/DOTAP (25:50:25) with ordered domains (DPPE-rich) prior to (e) and after (f) the application of two heat/cool/heat cycles starting at 4 ˚C, passing through the Tm of DPPE at 63 ˚C and ending at 80 ˚C. Panel g displays a time series collected following the temperature cycles (scale bars are 1 µm). (a-d) The fluid domains (DOPC-rich) are labeled using 0.1 mol% Bodipy-PC and (e-g) the ordered domains (DPPE-rich) are labeled using 0.1 mol% N-Rh-DHPE.40 In all preparations, 5 µm silica beads, coated with avidin, were used as the solid spherical support and 0.1 mol% DSPE-PEG2000-biotin was used in the lipid mixture for tethering purposes. 76x66mm (300 x 300 DPI)

46

45

4

43

42

41

40

39

38

37

36

47 48 49 50 51 52 53 54 5 56 57 58 59 60 ACS Paragon Plus Environment

Page 34 of 40

Page 35 of 40

Langmuir

1 2 3 4 5 6 7 8 9 10 1 12 13 14 15 16 17 18 19 20 21 2 23 24 25 26 27 28 29 30 31 32 3 34 35 36 37 38 39 40 41 Diffusion properties of DOPC SS-BLMs labelled using 0.1 mol% Bodipy-PC and tethered on 5 µm avidin coated silica beads using 0.1 mol % DSPE-PEG2000-biotin: (a) time series displaying fluorescence recovery after photobleaching of a circular ROI (1 µm, shown in red). Fluorescence intensity data were collected from an additional non-bleached reference circular ROI of the same diameter (1 µm, not shown) and used in subsequent data analysis to correct for any bleaching occurring during imaging. (b) Averaged fluorescence data and corresponding standard error for bleached (shown in black) and reference (shown in green) regions. The data correspond to 50 bleaching experiments on different SS-BLMs and collected in a single experimental set up. After normalization to pre-bleaching fluorescence levels, the averaged FRAP data is fit (curve shown in grey) to a one diffusing component model (R value of 0.989). (c) Histogram displaying the frequency of different % mobile fractions measured for the 50 SS-BLMs from the same sample preparation. 74x78mm (300 x 300 DPI)

51

50

49

48

47

46

45

4

43

42

52 53 54 5 56 57 58 59 60 ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 1 12 13 14 15 16 17 18 19 20 21 Scheme (not to scale) illustrating the preparation of SS-BLMs displaying co-existing lipid microdomains and their subsequent interaction with living cells. 81x35mm (300 x 300 DPI)

24

23

2

25 26 27 28 29 30 31 32 3 34 35 36 37 38 39 40 41 42 43 4 45 46 47 48 49 50 51 52 53 54 5 56 57 58 59 60 ACS Paragon Plus Environment

Page 36 of 40

Page 37 of 40

Langmuir

1 2 3 4 5 6 7 8 9 10 1 12 13 14 15 16 17 18 19 20 21 2 23 24 25 26 27 28 29 30 31 Representative confocal 3D-reconstruction images showing the co-localization of cytoskeletal networks with lipid microdomains on SS-BLMs presenting an Ld - Lo phase separation. Assembly of (a) microtubules (βtubulin, green), and (b) actin (phalloidin, green) around the fluid phase on DOPC/DPPC/ Chol (50:30:20) SS-BLMs. Panels c & e are magnified views of images a & b respectively, showing the specific organization of the cytoskeletal networks around the unlabeled regions on the SS-BLMs, representing the fluid phase (DOPC-rich). Panels d & f are single channel images of c & e, showing the exact location of ordered (DPPCrich) domains on SS-BLMs that are labeled using 0.1 mol% DiI-C20 (red). The SS-BLMs are co-cultured with hippocampal neurons (DIV 9) for 24 hrs and immunostained for either microtubules or actin filaments. 58x43mm (600 x 600 DPI)

39

38

37

36

35

34

3

32

40 41 42 43 4 45 46 47 48 49 50 51 52 53 54 5 56 57 58 59 60 ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 1 12 13 14 15 16 17 18 19 20 21 2 23 24 25 26 27 28 29 30 31 32 3 34 35 36 37 38 39 40 41 42 43 4 46

45 A representative 3-channel confocal 3D-recontruction image showing the co-localization of cytoskeletal networks with lipid microdomains on SS-BLMs presenting a fluid - gel phase separation: (a) assembly of microtubules (β-tubulin, green) and actin (phalloidin, blue) around the fluid phase on DOPC/DPPE/DOTAP (25:50:25) SS-BLMs. The ordered domains (DPPE-rich) are labeled using 0.1 mol% N-Rh-DHPE (red). Magnified views of image “a” show microtubule (b) and actin (c) co-localization with the lipid microdomains. 80x95mm (300 x 300 DPI)

51

50

49

48

47

52 53 54 5 56 57 58 59 60 ACS Paragon Plus Environment

Page 38 of 40

Page 39 of 40

Langmuir

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

41 Quantification of co-localization between cytoskeletal filaments and SS-BLM co-existing domains from the lipid mixture DOPC/DPPE/DOTAP (25:50:25), where the ordered phase (DPPE-rich) is labeled using 0.1 mol% N-Rh-DHPE (red). Panels (c, d) displaying the fluorescence intensity profiles across an area of the SSBLM (indicated by white lines in images a and b), using the same color codes for the fluorescence channels where microtubules are labeled in green (a) and actin is labeled in blue (b). (Scale bars = 2 µm). 46x50mm (300 x 300 DPI)

47

46

45

4

43

48 49 50 51 52 53 54 5 56 57 58 59 60 ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 1 12 13 14 15 16 17 18 19 20 21 2 23 24 25 26 27 29

28 Figure 7. Preferential co-localization of cytoskeletal filaments with lipid phase domains in SS-BLMs. % colocalization is calculated with respect to the single lipid phase present or with the disordered phase when coexisting lipid phases are present. For quantification details see Experimental Section 2.7 and Figure 6. 83x56mm (300 x 300 DPI)

3

32

31

30

34 35 36 37 38 39 40 41 42 43 4 45 46 47 48 49 50 51 52 53 54 5 56 57 58 59 60 ACS Paragon Plus Environment

Page 40 of 40