Fabrics of Diverse Chemistries Promote the Formation of GiantVesicles from Phospholipids and Amphiphilic Block CopolymersVaishnavi Girish, Joseph Pazzi, Alexander Li, and Anand Bala Subramaniam*

Department of Bioengineering, University of California, Merced, Merced, California 95343, United States

*S Supporting Information

ABSTRACT: Giant vesicles composed of phospholipids and amphiphilic blockcopolymers are useful for biomimetic drug delivery, for biophysical experiments, and forcreating synthetic cells. Here, we report that large numbers of giant unilamellar vesicles(GUVs) can be formed on a broad range of fabrics composed of entangled cylindricalfibers. We show that fabrics woven from fibers of silk, wool, rayon, nylon, polyester, andfiberglass promote the formation of GUVs and giant polymer vesicles (polymersomes) inaqueous solutions. The result extends significantly previous reports on the formation ofGUVs on cellulose paper and cotton fabric. Giant vesicles formed on all the fabrics fromlipids with various headgroup charges, chains lengths, and chain saturations. Giant vesicles could be formed frommulticomponent lipid mixtures, from extracts of plasma membranes, and from amphiphilic diblock and triblock copolymers, inboth low ionic strength and high ionic strength solutions. Intriguingly, statistical characterization using a model lipid, 1,2-dioleoyl-sn-glycero-3-phosphocholine, revealed that the majority of the fabrics yielded similar average counts of vesicles.Additionally, the vesicle populations obtained from the different fabrics had similar distributions of sizes. Fabrics are ubiquitousin society in consumer, technical, and biomedical applications. The discovery herein that biomimetic GUVs grow on fabricsopens promising new avenues in vesicle-based smart materials design.

■ INTRODUCTIONWhen dispersed at low concentrations in aqueous solutions,amphiphiles with cylindrical packing shapes, such as certainphospholipids, amphiphilic block copolymers, and fatty acids,self-assemble into complex suspensions of nanotubes, nano-sheets, and vesicles of various sizes and lamellarities.1 Of thesestructures, giant unilamellar vesicles (GUVs), vesicles largerthan 1 μm and composed of a single bilayer, are of interestbecause they mimic the dimensions and compartmentalizationproperties of biological cell membranes.2 GUVs canencapsulate and protect hydrophilic cargo,3,4 and theirmembranes can host hydrophobic molecules and amphipathicproteins.5,6 Thus, GUVs are useful for biomimetic targeting7

and delivery of cargo.8,9 GUVs are also useful for biophysicalexperiments10−13 and for creating synthetic cells.3,4,14−16

Obtaining a large fraction of GUVs in aqueous solutionsrequires directed assembly.2 Procedures for directing theassembly of GUVs fall into two general categories,2 solvent-based methods and solvent-free methods. For solvent-basedmethods, an aqueous solution is first emulsified in acontinuous phase of oil or volatile organic solvent containingdissolved lipids. The amphiphilic lipids adsorb as a monolayeron the interfaces of the aqueous droplets. Next, the lipidmonolayer-stabilized droplets are transferred into a secondaqueous continuous phase. As the droplets cross the interface,they entrain a layer of solvent or oil along with a secondmonolayer of lipids. Slow dissolution of the solvent results inthinning of the solvent film,17−22 allowing the two monolayerleaflets to form bilayers. With appropriate solvent combina-tions, dewetting of the solvent film23,24 to form attached

solvent droplets also allows the formation of bilayermembranes. Manual agitation followed by centrifugation5,25−27

or microfluidic flows17−24 can accomplish emulsification andtransfer of droplets. Related procedures employ microfluidicjetting to disrupt solvent-stabilized black lipid membranes toform GUVs.28−31 Non-negligible solubility of most organicsolvents in the aqueous phase32 (e.g., toluene solubility inwater: 103 parts per million, 5.7 mM) and in the hydrophobiccore of the lipid bilayer,33 however, suggests that residualsolvent remains in the membrane and in the aqueous phases atequilibrium.33 This residual solvent effects properties such asthe bending rigidity and the phase transition temperatures ofthe membranes.33 Organic solvents are toxic,34 making theGUVs unsuitable for use in biomedical applications withoutadditional solvent-removal steps. For solvent-free methods, theprocess begins by spreading a film of lipids dissolved in avolatile organic solvent on a solid surface.2 Traces of solventthat remain in the film after ambient evaporation are driven-offin vacuo.2 When placed in an aqueous solution, the drysolvent-free film assembles into lamellar stacks of lipidbilayers.1,2 Various methods are known to direct thevesiculation of the lamellar stacks to form GUVs. In thegentle hydration method, up to 20 mol % of charged lipids isadded to the lipid mixture and hydrodynamic flows areminimized for several hours to several days.35−37 In theelectroformation method, an oscillating electric field is applied

Received: May 29, 2019Revised: June 19, 2019Published: July 5, 2019

Article

pubs.acs.org/LangmuirCite This: Langmuir 2019, 35, 9264−9273

© 2019 American Chemical Society 9264 DOI: 10.1021/acs.langmuir.9b01621Langmuir 2019, 35, 9264−9273

Dow

nloa

ded

via

UN

IV O

F C

AL

IFO

RN

IA M

ER

CE

D o

n Ju

ne 8

, 202

0 at

00:

09:3

0 (U

TC

).Se

e ht

tps:

//pub

s.ac

s.or

g/sh

arin

ggui

delin

es f

or o

ptio

ns o

n ho

w to

legi

timat

ely

shar

e pu

blis

hed

artic

les.

pubs.acs.org/Langmuirhttp://pubs.acs.org/action/showCitFormats?doi=10.1021/acs.langmuir.9b01621http://dx.doi.org/10.1021/acs.langmuir.9b01621

orthogonal to the plane of the lamellar stacks.38 In the gel-assisted method, the film of lipids is supported on hydrogelsurfaces.39−46

Recently, we reported that GUVs formed spontaneously oncellulose fibers when dry lipid-coated cellulose filter papers andcotton fabric are hydrated in aqueous solutions (PAPYRUS,paper-abbetted amphiphile hydration in aqueous solu-tions).47−49 GUVs formed from a wide variety of glycer-ophospholipids and sphingolipids, including those withcharged headgroups and those with saturated and unsaturatedalkyl chains.47 GUVs can be formed in water, sugar solutions,and physiologically relevant ionic buffers, such as phosphatebuffered saline (PBS).47 Cellulose, even at elevated temper-atures,50 is insoluble in aqueous solutions and in organicsolvents typically used to dissolve lipids, such as chloroform,methanol, and toluene.50 The insolubility of cellulose allowsthe formation of GUVs with single and multicomponentmembranes containing lipids with high chain-melting temper-atures that are free from contamination from cellulose.47

Cellulose paper also promotes the formation of giant polymervesicles (polymersomes) and fatty acid vesicles fromamphiphilic block copolymers and fatty acids.48 Comparedto gentle hydration that requires hours to days to formGUVs,35−37 the formation of GUVs on cellulose paper wasrapid (typical time

used a magnetic stirrer and Teflon stir bar to agitate the fabric inchloroform for 30 min. We repeated the process twice with freshchloroform each time. After the final chloroform wash, we removedthe fabric and allowed the solvent to evaporate from the fabric. Wethen placed the fabric in a 1000 mL glass media bottle and alternatedbetween soaking and rinsing the fabric in fresh batches of 1000 mL ofultrapure water over the course of 3 h. Ambient drying and storage ina clean Petri dish completed the process.Scanning Electron Microscopy. To prepare the fabrics for imaging,

we cut a small 5 mm × 5 mm piece of the cleaned fabrics andmounted them on aluminum scanning electron microscopy (SEM)stubs using double-sided copper tape. We used a field emissionscanning electron microscope (GeminiSEM 500, Zeiss, Germany) toobtain images of the surfaces of the fabrics. The fabrics were exposedto a beam accelerating voltage of 1 kV, and the secondary electronsthat were scattered from the surface were collected using an Everhart-Thornley detector. We collected images at a pixel resolution of 1333nm/pixel for the low-magnification images and a pixel resolution of147 nm/pixel for the high-magnification images.Deposition of Lipids and Growth of DOPC GUVs for

Quantitative Characterization of Yields and Size Distributions.

We standardized growth conditions to allow comparison between thedifferent substrates. We prepared a solution of 99.5:0.5 mol %DOPC/TopFluor-PC in neat chloroform at a concentration of 2 mg/mL. We cut each piece of fabric into 9.5 mm diameter circular diskswith a pair of scissors and deposited appropriate volumes of the lipidsolution to obtain 3 μg of lipid per 1 mg of the substrate. For wool, weused 1.5 μg of lipid per 1 mg of substrate. After the solvent hadevaporated ambiently, we placed the lipid-coated fabric in a vacuumchamber for 1 h to drive off residual solvent. The substrates were thenplaced into 2.0 mL Eppendorf tubes. We added 500 μL of a 100 mMsolution of sucrose to hydrate the lipid-coated substrates andincubated the substrates for 60 min.

Growth of GUVs with Varying Membrane Compositions on theFabrics. All amphiphile-coated substrates were placed under vacuumfor a minimum of 1 h to drive off residual solvent and then incubatedin an aqueous solution for 60 min before imaging. Growth of giantvesicles from the various amphiphiles required the use of varyinggrowth temperatures and nominal surface concentrations. For theamphiphilic triblock Pluronic L121 polymersomes, we deposited 14μg of 99.5:0.5 mol % Pluronic L121/TopFluor-PC per 1 mg of thesubstrate. Growth was performed at room temperature. For the

Table 1. Properties of the Fabricsa

aStructural formula diagrams were obtained from the refs 62−68. Water contact angles were obtained from refs 60, 69 and 70. The fiber diameterswere measured from SEM images.

Langmuir Article

DOI: 10.1021/acs.langmuir.9b01621Langmuir 2019, 35, 9264−9273

9266

http://dx.doi.org/10.1021/acs.langmuir.9b01621

amphiphilic diblock PBD46PEO30 polymersomes, we deposited 7 μgof 99.5:0.5 mol % PBD46PEO30/TopFluor-PC per 1 mg of thesubstrate. Growth was performed at 80 °C. For 89.5:10:0.5 mol %DOPC/DOPG/TopFluor-PC, we deposited 3 μg of the lipid mixtureper 1 mg of the substrate. Growth was at room temperature. For89.5:10:0.5 mol % DOPC/DOTAP/TopFluor-PC, we deposited 7 μgof the lipid mixture per 1 mg of the substrate. Growth was at roomtemperature. For 99.5:0.5 mol % DOPC/TopFluor-PC, we deposited3 μg of the lipid mixture per 1 mg of the substrate. Growth was atroom temperature. For 99.5:0.5 mol % DPPC, we deposited 3 μg ofthe lipid mixture per 1 mg of the substrate. Growth was at 65 °C. For35.5:36:28:0.5:0.5 mol % DOPC/DPPC/cholesterol/TopFluor-PC/Rhod-PE, we deposited 8 μg of the lipid mixture per 1 mg of thesubstrate. Growth was at 65 °C. For the E. coli membrane extract wedeposited 3 μg of the lipid mixture per 1 mg of the substrate. Growthwas at 37 °C. To obtain GUVs in standard PBS, we first incubated thelipid-coated fabrics in ultrapure water for 10 min. We then added aconcentrated stock of the buffer salts (Amresco phosphate bufferedsaline (PBS) 20× concentrate, 2.74 M sodium chloride, 0.54 Mpotassium chloride, 0.2 M sodium phosphate, pH 7.5) to obtainstandard PBS and allowed the vesicles to grow for an additional 50min before imaging.Harvesting of the GUVs.We placed a 100 μL droplet of a 100 mM

solution of sucrose on a clean glass coverslip. We removed thesubstrates from the Eppendorf tubes using forceps and quicklyimmersed the wet substrate into the droplet on the coverslip. Weharvested the GUVs by gently aspirating the solution into a 1000 μLpipette tip while moving the tip systematically over the surface of thefabric. We enlarged the opening of the pipette tip by cutting off theend of the tip. The larger opening minimizes damage to the GUVs byreducing fluid shear forces during aspiration.Confocal Imaging. We fabricated square imaging chambers (width

× length × height = 6 mm × 6 mm × 1 mm) frompoly(dimethylsiloxane) bonded onto glass slides. We passivated thesurface with casein to prevent rupture of GUVs onto the bare glass.We chose aliquot sizes that resulted in approximately 100 000 GUVsin the imaging chamber while minimizing vesicle overlap: 4 μLaliquots for all the fabrics, except for rayon, where we used 2 μLaliquots. Isomolar solutions of glucose (100 mM) were added to bringthe total volume in the chamber to 60 μL. The sucrose-filled vesicles,which had a higher density, sedimented to the bottom of the chamber.After 3 h, we used an upright confocal laser scanning microscope(LSM 880 with Airyscan+FAST, Axio Imager.Z2m, Zeiss, Germany)to capture single-plane confocal images of the entire area of thechamber using an automated tile scan routine (850.19 μm × 850.19μm, 2140 pixels × 2140 pixels per image, 64 images). We used a 10×EC PlanNeofluar objective with a numerical aperture, NA = 0.3 toimage. The TopFluor dye was excited using a 488 nm argon laser setat 4% power. For images of GUVs growing on the fibers of the fabric,we collected confocal z-stacks with an axial-spacing of 0.67 μm using a20× W Plan-Apochromat water immersion objective with an NA =1.0.Image Analysis. We used a custom routine written in MATLAB

(Mathworks Inc., Natick, MA) to analyze the images. The routineused an intensity threshold followed by watershed segmentation toidentify fluorescent objects.48,49 The native regionprops routinetabulated the equivalent diameters and mean fluorescence intensitiesof the objects. GUVs were distinguished from other lipid structures onthe basis of their mean intensities and their size (>1 μm).48,49 Toobtain the counts of GUVs normalized per unit mass of lipid, we use

the following: =GUVs per mass lipid VV M

GUVchamber harvestedaliquot lipid

. In this equa-

tion, GUVchamber is the number of GUVs counted in the chamber,Valiquot is the aliquot volume, Vharvested is the volume of the harvestedsolution, and Mlipid is the mass of lipid that was deposited on thefabrics. Two-sample t-tests were performed in MATLAB to test forstatistical significance.

■ RESULTS AND DISCUSSIONProperties of the Fabrics. We tested fabrics woven from

pure fibers of silk, wool, rayon, nylon, polyester, or fiberglass.Table 1 shows the structural formulas, the average diameter ofthe fibers, the weave pattern, and the single-fiber water contactangle of the fabrics. The structural formulas62−68 and watercontact angle of these fibers have been characterized in theliterature.61,69,70 Fiberglass, cotton, and rayon have smallerwater contact angles compared with silk, wool, rayon, nylon,and polyester. All fibers have water contact angles of

fiberglass fibers had smooth surfaces. Wool had a scaly surface.Furthermore, the rayon, nylon, polyester, and fiberglass fibersappeared very uniform, reflective of their synthetic nature. Thefiber sizes, measured from SEM images, ranged from anaverage diameter of 4.7 ± 0.4 μm for fiberglass to 22.9 ± 1.3μm for nylon. The radii of curvature of the fibers on thedifferent fabrics were thus within a relatively narrow range of0.1−0.4 μm−1.All Fabrics Promoted Vesiculation of GUVs on Their

Fibers. Figure 2 shows confocal microscope images of the

lipid-coated fabrics 1 h after incubation in an aqueous buffer.All fabrics tested had GUVs growing from the surfaces of theirfibers. Most of the GUVs had tethers to the lipid layer coatingthe fibers.48 These results show that the spontaneousformation of GUVs is not limited to cellulose and is generalto fibers of differing surface chemistries.Silk and rayon had spherical GUVs that appeared

qualitatively similar to those seen previously on cellulosepaper and cotton fabric49 (Figure S1a). The GUVs were highlyabundant on these fabrics. GUVs coated the fibers in stackedlayers that filled the pores between the fibers of the fabric.Wool had noticeably fewer GUVs than the other fabrics.Nylon, polyester, and fiberglass fabrics had an intermediatenumber of GUVs, with some regions having a higherabundance of GUVs and others having a lower abundance ofGUVs. We observed that GUVs formed only from the lipidlayer that coated the fibers. Lipid deposits that spanned thegaps between the fibers gave rise to multilamellar lipidstructures (Figure S2). These lipid deposits were moreprevalent on wool, nylon, polyester, and fiberglass than onrayon, silk, or cotton. Thus, along with GUVs, other lamellarstructures were present on the wool, nylon, polyester, andfiberglass fabrics.All fabrics promoted the growth of GUVs from a wide

variety of lamellar-phase-forming amphiphiles. GUVs could beobtained in both low ionic strength solutions, such as ultrapurewater and sugar solutions, and in high ionic strength solutions,such as standard phosphate buffered saline (standard PBS: 137mM sodium chloride, 2.7 mM potassium chloride, 10 mMsodium phosphate, pH 7.4−7.6). GUVs composed of lipidswith long saturated alkyl chains that require growth at elevatedtemperatures could also be obtained since all fibers are

thermally stable to 100 °C (the boiling point of water) orhigher.51,52 Figure 3 shows representative confocal images of a

giant polymer and lipid vesicles of various membranecompositions growing both in low ionic strength solutions(100 mM sucrose, Figure 3a−c) and high ionic strengthsolutions (standard PBS, Figure 3d−f) on silk fibers. Forgrowth in PBS, samples were incubated in ultrapure water for10 min before a concentrated stock of the buffer salts wasadded to obtain standard PBS. We allowed the vesicles tocontinue to grow in standard PBS for a further 50 min beforeimaging. Gradients in salt concentration between thecontinuous phase and the lumens of the GUVs were rapidlyequilibrated through the nanotube tethers that connected theGUVs to the external aqueous phase.48 Figure 3a,b shows giantpolymersomes composed of the amphiphilic triblock copoly-mer Pluronic L121 and polymersomes composed of theamphiphilic diblock copolymer PBD46PEO30. Figure 3c showsGUVs composed of negatively charged membranes(89.5:10:0.5 mol % DOPC/DOPG/TopFluor-PC), and Figure3d shows GUVs composed of positively charged membranes(89.5:10:0.5 mol % DOPC/DOTAP/TopFluor-PC). Figure3e shows membranes of the zwitterionic lipid DOPC in PBS,and Figure 3f shows membranes of the zwitterionic lipidDPPC in PBS grown at 65 °C. DPPC has a main transitiontemperature of 45 °C.11 Thus, DPPC is in the gel phase atroom temperature. Figure 3g shows GUVs with a ternarymembrane composition of 35.5:36:28:0.5:0.5 mol % DOPC/DPPC/cholesterol/TopFluor-PC/Rhod-PE that exhibitsliquid−liquid phase coexistence at room temperature.11 The

Figure 2. Confocal microscope images of the surface of the hydratedfabrics after 1 h. (a) Silk, (b) wool, (c) rayon, (d) polyester, (e)nylon, and (f) fiberglass. GUVS are the bright green circles with darkinteriors. Although not fluorescently labeled, the fibers are visible dueto the fluorescent lipid coating. (a−f) Scale bars: 25 μm.

Figure 3. Confocal microscope images of giant polymersomes andgiant lipid vesicles on the surface of silk fibers after 1 h in aqueousbuffer. (a) Giant polymersomes of the amphiphilic triblock copolymerPluronic L121 in a solution of sucrose. (b) Giant polymersomes of theamphiphilic diblock copolymer PBD46PEO30 in a solution of sucrose.(c) Giant vesicles with negatively charged membranes (89.5:10:0.5mol % DOPC/DOPG/TopFluor-PC) in a solution of sucrose. (d)Giant vesicles with positively charged membranes (89.5:10:0.5 mol %DOPC/DOTAP/TopFluor-PC) in standard PBS. (e) Giant vesicleswith zwitterionic membranes composed of two unsaturated alkylchains (DOPC, chain-melting temperature −20 °C) in standard PBS.(f) Giant vesicles with zwitterionic membranes composed of saturatedalkyl chains (DPPC, chain-melting temperature −45 °C) in standardPBS. (g) Giant vesicles with membranes exhibiting liquid−liquidphase coexistence composed of 35.5:36:28:0.5:0.5 mol % DOPC/DPPC/cholesterol/TopFluor-PC/Rhod-PE in standard PBS. (h)Giant vesicles composed of membrane extracts from E. coli instandard PBS. (a, b, d, e) Scale bars 15 μm. (c, f, g, h) Scale bars 25μm.

Langmuir Article

DOI: 10.1021/acs.langmuir.9b01621Langmuir 2019, 35, 9264−9273

9268

http://pubs.acs.org/doi/suppl/10.1021/acs.langmuir.9b01621/suppl_file/la9b01621_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acs.langmuir.9b01621/suppl_file/la9b01621_si_001.pdfhttp://dx.doi.org/10.1021/acs.langmuir.9b01621

GUVs were grown at 65 °C to ensure that the membraneswere fully mixed. Upon cooling to room temperature, themembranes phase-separated into liquid ordered and liquiddisordered domains, as evidenced by the partitioning of Rhod-PE (false-colored red) and TopFluor-PC (false-colored green)into distinct compartments in the membrane. Rhod-PEpartitions strongly into the liquid disordered phase, whereasTopFluor-PC partitions equally between the liquid disorderedphase and liquid ordered phase.11,47 Figure 3f shows GUVsobtained from the polar fraction of membrane extracts of thebacteria E. coli. The membranes, composed of phosphatidyle-thanolamine/phosphatidylglycerol/cardiolipin 67.0:23.2:9.8wt/wt % (manufacturer’s specifications), are highly negativelycharged.Quantification of Effective Yield of GUVs from the

Fabrics. Having shown that vesicle growth on the fabrics isgeneral to a wide range of solution ionic strengths, temper-atures, and lamellar-phase-forming-amphiphile types, we nextused the model lipid DOPC to conduct statistical experimentsto quantify the distribution of sizes and effective yields ofGUVs harvested from the fabrics. We performed threeindependent experiments for each type of fabric. We harvestedthe lipid structures from the fabrics and imaged aliquots fromthe suspensions after allowing the structures to sediment for 3h in a custom imaging chamber. A fraction of GUVs and otherstructures remain trapped in the fabric after the harvestingprocess. We were careful to reproduce similar harvestingconditions for all fabrics to allow a comparison of the effectiveyields. We analyzed our images using a custom MATLABroutine. Each experiment had large sample sizes ranging from

=n (10 )4 to (10 )5 GUVs. GUVs were identified on thebasis of their fluorescence intensity48,49 and their size (>1 μm).Control experiments using coupled sedimentation and dyeleakage confirmed that membrane fluorescence intensitydistinguishes GUVs from other structures (SupportingInformation, Figures S3 and S4). Furthermore, measurementsof lipid mobility using fluorescence recovery after photo-bleaching (FRAP) yielded a diffusion coefficient, D = 7.9 ± 1.7μm2/s for the lipids in the membranes of vesicles (SupportingInformation, Figure S5). This value of the diffusion coefficientis consistent with those reported in the literature for GUVscomposed of DOPC.72−74

Figure 4a−f shows a typical field of view obtained from a 4μL aliquot of the harvested solution diluted in 56 μL ofisomolar glucose buffer. Structures harvested from silk andrayon, similar to those from cotton fabric (Figure S1b) andcellulose paper,49 appeared predominantly to be GUVs.Consistent with our direct observations of the structures onthe fabrics, a larger fraction of lipid aggregates and lipidnanotubes were present in the samples harvested from wool,nylon, polyester, and fiberglass. On the basis of typicalnumbers reported in the literature (ten to hundreds ofvesicles),10−13 the number of GUVs in a characteristic field ofview obtained from all fabrics was sufficient to performbiophysical experiments.Figure 4g shows a bar and scatter plot of the number of

GUVs obtained from the fabrics arranged in a descendingorder of counts per microgram of lipid from left to right. Theheight of the bar is the average of the three experiments foreach type of fabric. The points are the counts from the threeindependent experiments. To allow comparison between thefabrics, we normalized the data to the mass of lipid depositedon the fabrics. Rayon appeared to produce the highest average

number of GUVs per unit mass of deposited lipid at 2.6 × 105

GUVs per μg of lipid and wool had the lowest average numberof GUVs at 6.2 × 104 GUVs per μg of lipid. Two-sample t-testsbetween rayon and all fabrics showed that the higher yieldfrom rayon was statistically significant (p < 0.05), except whencompared to nylon, where p = 0.1373. Two-sample t-testsshowed no statistical difference between yields obtained fromcotton, silk, nylon, polyester, and fiberglass with each other.The lower yield of wool was statistically significant (p < 0.05)when compared with the other fabrics, except when comparedwith polyester (p = 0.2032). A matrix of the pairwise p-valuesbetween the fabrics is shown in Table S1. These results showthat when measured at large sample sizes, yields fromchemically and structurally distinct fabrics, such as silk, cotton,fiberglass, nylon, and polyester, were statistically indistinguish-able, whereas the higher yield of GUVs from rayon and loweryield of GUVs from wool were statistically distinguishable.Note that even the lowest yielding fabric, wool, produced alarger number of GUVs than electroformation on stainless steelelectrodes.75 Furthermore, unlike electroformation, directingthe assembly of GUVs on fabric does not require access toelectrical power, electronic equipment, or specialized cham-bers.Similar to cotton fabric,49 the process of GUV growth did

not alter the fibers of the fabrics. All fabrics supported at leastfive cycles of GUV growth and harvest with no measurablechanges in the characteristics of the GUVs (data not shown).

Distribution of Sizes of the GUVs Obtained from theFabrics. We next analyzed the sizes of the DOPC GUVs thatwe obtained from the different fabrics. Figure 5a shows ahistogram of the diameters of the GUVs harvested from silk.The bin width is 1 μm, and the sample size, n = 113 461. Thedistribution of diameters was unimodal with a prominent righttail. Smaller GUVs were more abundant than larger GUVs.Populations of GUVs harvested from the other fabrics alsoshowed similar unimodal right-skewed distribution of diame-ters (Figure S6). These results further add to the growing body

Figure 4. (a−f) Representative images of the harvested GUVs in theimaging chamber. (g) Scatter and bar plot showing the number ofGUVs per microgram of lipid from each of the fabrics. The blackcircles are the data points for the three experiments for each fabric.The blue bar is the average number of GUVs from the threeexperiments for each fabric. p-Values from two-sample t-tests betweenrayon and the other fabrics are indicated in the plot. p-Values forother fabric pairs are shown in Supporting Information Table 1. p <0.05 is statistically significant. (a−f) Scale bars: 25 μm.

Langmuir Article

DOI: 10.1021/acs.langmuir.9b01621Langmuir 2019, 35, 9264−9273

9269

http://pubs.acs.org/doi/suppl/10.1021/acs.langmuir.9b01621/suppl_file/la9b01621_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acs.langmuir.9b01621/suppl_file/la9b01621_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acs.langmuir.9b01621/suppl_file/la9b01621_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acs.langmuir.9b01621/suppl_file/la9b01621_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acs.langmuir.9b01621/suppl_file/la9b01621_si_001.pdfhttp://pubs.acs.org/doi/suppl/10.1021/acs.langmuir.9b01621/suppl_file/la9b01621_si_001.pdfhttp://dx.doi.org/10.1021/acs.langmuir.9b01621

of evidence that populations of GUVs obtained from thevesiculation of lamellar stacks of phospholipids on disparatesurfaces, such as solid glass,36 hydrogels,41,43,44,46 andcylindrical fibers47−49 have broad and right-skewed distribu-tions in diameters.Figure 5b shows empirical cumulative distribution functions

of the populations of GUVs obtained from each fabric. Thepoints are the mean cumulative probability from the threeexperiments. The error bars are the standard deviation fromthe mean. Overall, 98% of the GUVs harvested from all fabricshad diameters

gram of sizes of giant vesicles obtained from differentfabrics; p-values for the effective counts of GUVs fromthe fabrics; p-values for the median diameter of GUVsfrom the fabrics (PDF)

■ AUTHOR INFORMATIONCorresponding Author*E-mail: asubramaniam@ucmerced.edu.ORCIDAnand Bala Subramaniam: 0000-0002-1998-9299Author ContributionsThe manuscript was written through contributions of allauthors. All authors have given approval to the final version ofthe manuscript.FundingThis work was funded by NSF Award CBET-1512686 andNSF-CREST: Center for Cellular and Biomolecular Machinesat the University of California, Merced (NSF-HRD-1547848).The data in this work was collected, in part, with a confocalmicroscope acquired through the National Science FoundationMRI Award Number DMR-1625733 and a scanning electronmicroscope acquired through NASA Grant NNX15AQ01A.NotesThe authors declare no competing financial interest.

■ REFERENCES(1) Israelachvili, J. N. Intermolecular and Surface Forces, 3rd ed.;Elsevier Academic Press Inc., 2011.(2) Walde, P.; Cosentino, K.; Engel, H.; Stano, P. Giant Vesicles:Preparations and Applications. ChemBioChem 2010, 848−865.(3) Schmitt, C.; Lippert, A. H.; Bonakdar, N.; Sandoghdar, V.; Voll,L. M. Compartmentalization and Transport in Synthetic Vesicles.Front. Bioeng. Biotechnol. 2016, 4, No. e25529.(4) York-Duran, M. J.; Godoy-Gallardo, M.; Labay, C.; Urquhart, A.J.; Andresen, T. L.; Hosta-Rigau, L. Recent advances incompartmentalized synthetic architectures as drug carriers, cellmimics and artificial organelles. Colloids Surf., B 2017, 152, 199−213.(5) Yanagisawa, M.; Iwamoto, M.; Kato, A.; Yoshikawa, K.; Oiki, S.Oriented reconstitution of a membrane protein in a giant unilamellarvesicle: Experimental verification with the potassium channel KcsA. J.Am. Chem. Soc. 2011, 133, 11774−11779.(6) Girard, P.; Pećreáux, J.; Lenoir, G.; Falson, P.; Rigaud, J.-L.;Bassereau, P. A new method for the reconstitution of membraneproteins into giant unilamellar vesicles. Biophys. J. 2004, 87, 419−429.(7) De Oliveira, H.; Thevenot, J.; Lecommandoux, S. Smartpolymersomes for therapy and diagnosis: Fast progress towardmultifunctional biomimetic nanomedicines. Wiley Interdiscip. Rev.:Nanomed. Nanobiotechnol. 2012, 4, 525−546.(8) Nieth, A.; Verseux, C.; Barnert, S.; Süss, R.; Römer, W. A firststep toward liposome-mediated intracellular bacteriophage therapy.Expert Opin. Drug Delivery 2015, 12, 1411−1424.(9) Singla, S.; Harjai, K.; Raza, K.; Wadhwa, S.; Katare, O. P.;Chhibber, S. Phospholipid vesicles encapsulated bacteriophage: Anovel approach to enhance phage biodistribution. J. Virol. Methods2016, 236, 68−76.(10) Steer, D.; Leung, S. S. W.; Meiselman, H.; Topgaard, D.; Leal,C. Structure of Lung-Mimetic Multilamellar Bodies with LipidCompositions Relevant in Pneumonia. Langmuir 2018, 34, 7561−7574.(11) Veatch, S. L.; Keller, S. L. Separation of liquid phases in giantvesicles of ternary mixtures of phospholipids and cholesterol. Biophys.J. 2003, 85, 3074−3083.(12) Baumgart, T.; Hammond, A. T.; Sengupta, P.; Hess, S. T.;Holowka, D. A.; Baird, B. A.; Webb, W. W. Large-scale fluid/fluid

phase separation of proteins and lipids in giant plasma membranevesicles. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 3165−3170.(13) Dimova, R.; Aranda, S.; Bezlyepkina, N.; Nikolov, V.; Riske, K.A.; Lipowsky, R. A practical guide to giant vesicles. Probing themembrane nanoregime via optical microscopy. J. Phys.: Condens.Matter 2006, 18, S1151−S1176.(14) Nomura, S.-i. M.; Tsumoto, K.; Hamada, T.; Akiyoshi, K.;Nakatani, Y.; Yoshikawa, K. Gene Expression within Cell-Sized LipidVesicles. ChemBioChem 2003, 4, 1172−1175.(15) Blain, J. C.; Szostak, J. W. Progress Toward Synthetic Cells.Annu. Rev. Biochem. 2014, 83, 615−640.(16) Walde, P. Building artificial cells and protocell models:Experimental approaches with lipid vesicles. BioEssays 2010, 296−303.(17) Abkarian, M.; Loiseau, E.; Massiera, G. Continuous dropletinterface crossing encapsulation (cDICE) for high throughputmonodisperse vesicle design. Soft Matter 2011, 7, 4610.(18) Arriaga, L. R.; Datta, S. S.; Kim, S. H.; Amstad, E.; Kodger, T.E.; Monroy, F.; Weitz, D. A. Ultrathin shell double emulsiontemplated giant unilamellar lipid vesicles with controlled micro-domain formation. Small 2014, 10, 950−956.(19) do Nascimento, D. F.; Arriaga, L. R.; Eggersdorfer, M.; Ziblat,R.; Marques, M. d. F. V.; Reynaud, F.; Koehler, S. A.; Weitz, D. A.Microfluidic Fabrication of Pluronic Vesicles with ControlledPermeability. Langmuir 2016, 32, 5350−5355.(20) Kim, S.-H.; Kim, J. W.; Cho, J.-C.; Weitz, D. A. Double-emulsion drops with ultra-thin shells for capsule templates. Lab Chip2011, 11, 3162−3166.(21) Hu, P. C.; Li, S.; Malmstadt, N. Microfluidic fabrication ofasymmetric giant lipid vesicles. ACS Appl. Mater. Interfaces 2011, 3,1434−1440.(22) Blosser, M. C.; Horst, B. G.; Keller, S. L. cDICE methodproduces giant lipid vesicles under physiological conditions of chargedlipids and ionic solutions. Soft Matter 2016, 12, 7364−7371.(23) Deshpande, S.; Caspi, Y.; Meijering, A.; Dekker, C. Octanol-assisted liposome assembly on chip. Nat. Commun. 2016, No. 10447.(24) Deng, N.-N.; Yelleswarapu, M.; Zheng, L.; Huck, W. T. S.Microfluidic Assembly of Monodisperse Vesosomes as Artificial CellModels. J. Am. Chem. Soc. 2017, 139, 587−590.(25) Peyret, A.; Ibarboure, E.; Le Meins, J. F.; Lecommandoux, S.Asymmetric Hybrid Polymer−Lipid Giant Vesicles as Cell MembraneMimics. Adv. Sci. 2018, 5, No. 1700453.(26) Morita, M.; Onoe, H.; Yanagisawa, M.; Ito, H.; Ichikawa, M.;Fujiwara, K.; Saito, H.; Takinoue, M. Droplet-Shooting and Size-Filtration (DSSF) Method for Synthesis of Cell-Sized Liposomes withControlled Lipid Compositions. ChemBioChem 2015, 16, 2029−2035.(27) Marguet, M.; Edembe, L.; Lecommandoux, S. Polymersomes inpolymersomes: Multiple loading and permeability control. Angew.Chem., Int. Ed. 2012, 51, 1173−1176.(28) Stachowiak, J. C.; Richmond, D. L.; Li, T. H.; Brochard-Wyart,F.; Fletcher, D. A. Inkjet formation of unilamellar lipid vesicles forcell-like encapsulation. Lab Chip 2009, 9, 2003−2009.(29) Richmond, D. L.; Schmid, E. M.; Martens, S.; Stachowiak, J. C.;Liska, N.; Fletcher, D. Forming giant vesicles with controlledmembrane composition, asymmetry, and contents. Proc. Natl. Acad.Sci. U.S.A. 2011, 108, 9431−9436.(30) Coyne, C. W.; Patel, K.; Heureaux, J.; Stachowiak, J.; Fletcher,D. A.; Liu, A. P. Lipid Bilayer Vesicle Generation Using MicrofluidicJetting. J. Visualized Exp. 2014, 84, No. e51510.(31) Kamiya, K.; Kawano, R.; Osaki, T.; Akiyoshi, K.; Takeuchi, S.Cell-sized asymmetric lipid vesicles facilitate the investigation ofasymmetric membranes. Nat. Chem. 2016, 8, 881−889.(32) Yalkowsky, S. H.; He, Y.; Jain, P. Handbook of AqueousSolubility, 2nd ed.; CRC Press: Boca Raton, 2010.(33) Kirchner, S. R.; Ohlinger, A.; Pfeiffer, T.; Urban, A. S.; Stefani,F. D.; Deak, A.; Lutich, A. A.; Feldmann, J. Membrane composition ofjetted lipid vesicles: A Raman spectroscopy study. J. Biophotonics2012, 5, 40−46.

Langmuir Article

DOI: 10.1021/acs.langmuir.9b01621Langmuir 2019, 35, 9264−9273

9271

http://pubs.acs.org/doi/suppl/10.1021/acs.langmuir.9b01621/suppl_file/la9b01621_si_001.pdfmailto:asubramaniam@ucmerced.eduhttp://orcid.org/0000-0002-1998-9299http://dx.doi.org/10.1021/acs.langmuir.9b01621

(34) Spector, M. S.; Zasadzinski, J. A.; Sankaram, M. B. Topology ofMultivesicular Liposomes, a Model Biliquid Foam. Langmuir 1996,12, 4704−4708.(35) Bangham, A. D.; Standish, M. M.; Watkins, J. C. Diffusion ofunivalent ions across the lamellae of swollen phospholipids. J. Mol.Biol. 1965, 13, 238−252.(36) Reeves, J. P.; Dowben, R. M. Formation and properties of thin-walled phospholipid vesicles. J. Cell Physiol. 1969, 73, 49−60.(37) Akashi, K.; Miyata, H.; Itoh, H.; Kinosita, K. Preparation ofgiant liposomes in physiological conditions and their characterizationunder an optical microscope. Biophys. J. 1996, 71, 3242−3250.(38) Angelova, M. I.; Dimitrov, D. S. Liposome electroformation.Faraday Discuss. Chem. Soc. 1986, 81, 303−311.(39) Horger, K. S.; Estes, D. J.; Capone, R.; Mayer, M. Films ofagarose enable rapid formation of giant liposomes in solutions ofphysiologic ionic strength. J. Am. Chem. Soc. 2009, 131, 1810−1819.(40) Weinberger, A.; Tsai, F. C.; Koenderink, G. H.; Schmidt, T. F.;Itri, R.; Meier, W.; Schmatko, T.; Schröder, A.; Marques, C. Gel-assisted formation of giant unilamellar vesicles. Biophys. J. 2013, 105,154−164.(41) Loṕez Mora, N.; Hansen, J. S.; Gao, Y.; Ronald, A. A.; Kieltyka,R.; Malmstadt, N.; Kros, A. Preparation of size tunable giant vesiclesfrom cross-linked dextran(ethylene glycol) hydrogels. Chem. Commun.2014, 50, 1953−1955.(42) Mora, N. L.; Gao, Y.; Gutierrez, M. G.; Peruzzi, J.; Bakker, I.;Peters, R. J. R. W.; Siewert, B.; Bonnet, S.; Kieltyka, R. E.; van Hest, J.C. M.; et al. Evaluation of dextran(ethylene glycol) hydrogel films forgiant unilamellar lipid vesicle production and their application for theencapsulation of polymersomes. Soft Matter 2017, 13, 5580−5588.(43) Movsesian, N.; Tittensor, M.; Dianat, G.; Gupta, M.;Malmstadt, N. Giant Lipid Vesicle Formation Using Vapor-DepositedCharged Porous Polymers. Langmuir 2018, 34, 9025−9035.(44) Peruzzi, J.; Gutierrez, M. G.; Mansfield, K.; Malmstadt, N.Dynamics of Hydrogel-Assisted Giant Unilamellar Vesicle Formationfrom Unsaturated Lipid Systems. Langmuir 2016, 32, 12702−12709.(45) Wang, M.; Liu, Z.; Zhan, W. Janus Liposomes: Gel-AssistedFormation and Bioaffinity-Directed Clustering. Langmuir 2018, 34,7509−7518.(46) Greene, A. C.; Henderson, I. M.; Gomez, A.; Paxton, W. F.;VanDelinder, V.; Bachand, G. D. The Role of Membrane Fluidizationin the Gel-Assisted Formation of Giant Polymersomes. PLoS One2016, 11, No. e0158729.(47) Kresse, K. M.; Xu, M.; Pazzi, J.; Garcia-Ojeda, M.;Subramaniam, A. B. Novel Application of Cellulose Paper As aPlatform for the Macromolecular Self-Assembly of Biomimetic GiantLiposomes. ACS Appl. Mater. Interfaces 2016, 8, 32102−32107.(48) Li, A.; Pazzi, J.; Xu, M.; Subramaniam, A. B. Cellulose abettedassembly and temporally-decoupled loading of cargo into vesiclessynthesized from functionally diverse lamellar phase formingamphiphiles. Biomacromolecules 2018, 19, 849−859.(49) Pazzi, J.; Xu, M.; Subramaniam, A. B. Size distributions andyields of giant vesicles assembled on cellulose papers and cottonfabric. Langmuir 2019, 35, 7798−7804.(50) Marsh, J. T.; Wood, F. C. An Introduction to the Chemistry ofCellulose; Ossler Press: Great Britain, 1939.(51) Cook, J. G. Handbook of Textile Fibres: Natural Fibres, 5th ed.;Woodhead Publishing: Cambridge, U.K., 1984.(52) Cook, J. G. Handbook of Textile Fibres: Man-Made Fibres, 5thed.; Woodhead Publishing: Cambridge, U.K., 1984.(53) Sandin, G.; Peters, G. M. Environmental impact of textile reuseand recycling − A review. J. Cleaner Prod. 2018, 184, 353−365.(54) Chen, H.; Burns, L. D. Environmental Analysis of TextileProducts. Clothing Text. Res. J. 2006, 24, 248−261.(55) Leal-Egaña, A.; Scheibel, T. Silk-based materials for biomedicalapplications. Biotechnol. Appl. Biochem. 2010, 55, 155−167.(56) Kundu, B.; Kundu, S. C. Osteogenesis of human stem cells insilk biomaterial for regenerative therapy. Prog. Polym. Sci. 2010, 35,1116−1127.

(57) Kundu, B.; Kurland, N. E.; Bano, S.; Patra, C.; Engel, F. B.;Yadavalli, V. K.; Kundu, S. C. Silk proteins for biomedicalapplications: Bioengineering perspectives. Prog. Polym. Sci. 2014, 39,251−267.(58) Klemm, D.; Heublein, B.; Fink, H. P.; Bohn, A. Cellulose:Fascinating biopolymer and sustainable raw material. Angew. Chem.,Int. Ed. 2005, 44, 3358−3393.(59) Maitz, M. F. Applications of synthetic polymers in clinicalmedicine. Biosurf. Biotribology 2015, 1, 161−176.(60) Gupta, B. S. Manufacture, Types and Properties of Biotextiles forMedical Applications; Woodhead Publishing Limited, 2013.(61) Singh, C.; Wong, C.; Wang, X. Medical Textiles as VascularImplants and Their Success to Mimic Natural Arteries. J. Funct.Biomater. 2015, 6, 500−525.(62) Shen, G.; Hu, X.; Guan, G.; Wang, L. Surface Modification andCharacterisation of Silk Fibroin Fabric Produced by the Layer-by-Layer Self-Assembly of Multilayer Alginate/Regenerated Silk Fibroin.PLoS One 2015, 10, No. e0124811.(63) Hutchinson, S.; Evans, D.; Corino, G.; Kattenbelt, J. Anevaluation of the action of thioesterases on the surface of wool.Enzyme Microb. Technol. 2007, 40, 1794−1800.(64) Nsor-Atindana, J.; Chen, M.; Goff, H. D.; Zhong, F.; Sharif, H.R.; Li, Y. Functionality and nutritional aspects of microcrystallinecellulose in food. Carbohydr. Polym. 2017, 172, 159−174.(65) Lubin, G.; Dastin, S. J. Raw MaterialsHandbook ofComposites; Lubin, G., Ed.; Springer: Boston, MA, 1982; pp 19−272.(66) Ortega, R. A.; Carter, E. S.; Ortega, A. E. Nylon 6,6 NonwovenFabric Separates Oil Contaminates from Oil-in-Water Emulsions.PLoS One 2016, 11, No. e0158493.(67) Militky, J. 9 - The Chemistry, Manufacture and TensileBehaviour of Polyester Fibers. In Woodhead Publishing Series inTextiles; Bunsell, A. R., Ed.; Woodhead Publishing, 2009; pp 223−314.(68) Cao, C.; Li, Z.-B.; Wang, X.-L.; Zhao, X.-B.; Han, W.-Q. RecentAdvances in Inorganic Solid Electrolytes for Lithium Batteries. Front.Energy Res. 2014, 2, No. 25.(69) Le, C. V.; Ly, N. G.; Stevens, M. G. Measuring the ContactAngles of Liquid Droplets on Wool Fibers and Determining SurfaceEnergy Components. Text. Res. J. 1996, 66, 389−397.(70) Dutta, S.; Talukdar, B.; Bharali, R.; Rajkhowa, R.; Devi, D.Fabrication and characterization of biomaterial film from gland silk ofmuga and eri silkworms. Biopolymers 2013, 99, 326−333.(71) Arthur, A. P. G.; Adamson, W. Physical Chemistry of Surfaces,6th ed.; John Wiley & Sons Inc.: New York, 1997.(72) Pincet, F.; Adrien, V.; Yang, R.; Delacotte, J.; Rothman, E.;Urbach, W.; Tareste, D. FRAP to Characterize Molecular Diffusionand Interaction in Various Membrane Environments. PLoS One 2016,11, No. e0158457.(73) Heinemann, F.; Betaneli, V.; Thomas, F. A.; Schwille, P.Quantifying Lipid Diffusion by Fluorescence Correlation Spectrosco-py: A Critical Treatise. Langmuir 2012, 28, 13395−13404.(74) Przybylo, M.; Sy, J.; Humpolı, J.; Zan, A.; Hof, M.; Heyro, J. V.Lipid Diffusion in Giant Unilamellar Vesicles Is More than 2 TimesFaster than in Supported Phospholipid Bilayers under IdenticalConditions. Langmuir 2006, 6, 9096−9099.(75) Pereno, V.; Carugo, D.; Bau, L.; Sezgin, E.; Bernardino De LaSerna, J.; Eggeling, C.; Stride, E. Electroformation of GiantUnilamellar Vesicles on Stainless Steel Electrodes. ACS Omega2017, 2, 994−1002.(76) Parthasarathy, R.; Yu, C.; Groves, J. T. Curvature-modulatedphase separation in lipid bilayer membranes. Langmuir 2006, 22,5095−5099.(77) Subramaniam, A. B.; Lecuyer, S.; Ramamurthi, K. S.; Losick, R.;Stone, H. A. Particle/fluid interface replication as a means ofproducing topographically patterned polydimethylsiloxane surfaces fordeposition of lipid bilayers. Adv. Mater. 2010, 22, 2142−2147.(78) Ramamurthi, K. S.; Losick, R. Negative membrane curvature asa cue for subcellular localization of a bacterial protein. Proc. Natl.Acad. Sci. U.S.A. 2009, 106, 13541−13545.

Langmuir Article

DOI: 10.1021/acs.langmuir.9b01621Langmuir 2019, 35, 9264−9273

9272

http://dx.doi.org/10.1021/acs.langmuir.9b01621

(79) Ramamurthi, K. S.; Lecuyer, S.; Stone, H. A.; Losick, R.Geometric cue for protein localization in a bacterium. Science 2009,323, 1354−1357.

Langmuir Article

DOI: 10.1021/acs.langmuir.9b01621Langmuir 2019, 35, 9264−9273

9273

http://dx.doi.org/10.1021/acs.langmuir.9b01621