Lipid Tubules

Fabrication of Lipid Tubules with Embedded Quantum Dots by Membrane Tubulation Protein

Masayoshi Tanaka , * Kevin Critchley , Tadashi Matsunaga , Stephen D. Evans , and Sarah S. Staniland

The fi rst one-dimensional (1D) assembly of low-toxicity CuInS 2 /ZnS quantum dots (QDs) embedded in lipid nanotubules, formed from liposomes using the Amphiphysin-BAR (Bin-Amphiphysin-Rvs domain of human amphiphysin) protein to elongate the structure, is reported. The QD-containing lipid nanotubules display a high aspect ratio of ≈ 500:1 ( ≈ 40 nm diameter and 20 μ m length) and are stable for more than 20 h. Furthermore, this methodology is extended to the assembly of various nanoparticle species within 1D lipid nanotubules, and includes materials such as CdSe and Au. Encapsulation within the hydrophobic core of the bilayer makes these materials highly biocompatible. The developed methodology and materials with these unique characteristics could be useful for various applications in nanobiotechnology and nanomedicine.

1. Introduction

Over the past decade, nanoparticles (NPs) have been the sub-

ject of intense research due to their unique optical, catalytic,

electrical, and magnetic properties, which can be exploited

in nanotechnology. Among them, quantum dots (QDs) have

attracted signifi cant attention due to their promise for appli-

cations such as imaging and light harvesting. [ 1 ] Now the

challenge is to incorporate these NPs into functional, highly

controlled, nanostructured assemblies for the realization of

nanotechnological devices. One-dimensional (1D) structures

such as rods, wires, and tubules represent an important motif

required to create interconnects in devices and to build hier-

archical structures. There have been several reports demon-

strating the self-assembly of 1D structures using inorganic

© 2012 Wiley-VCH Verlag Gmb

DOI: 10.1002/smll.201102446

Dr. M. Tanaka , Dr. K. Critchley , Prof. S. D. Evans , Dr. S. S. Staniland School of Physics and AstronomyUniversity of LeedsLeeds LS2 9JT, UK E-mail: m.tanaka@leeds.ac.uk

Dr. M. Tanaka , Prof. T. Matsunaga Department of BiotechnologyTokyo University of Agriculture and Technology2-24-16 Naka-cho, Koganei, Tokyo 184-8588, Japan

small 2012, DOI: 10.1002/smll.201102446

and/or biological scaffold molecules such as silica, [ 2 ] carbon, [ 3 ]

peptides, [ 4 ] DNA, [ 5 ] and lipids. [ 6 ] In particular, lipid nano-

tubules provide structural fl exibility, high dispersity (due to

the electrostatic charges of phospholipids), and a biocom-

patible surface, which allows the encapsulation of biological

mole cules and their use as reaction vessels.

In biological cells from bacteria to human, the defor-

mation of membranous structures is highly regulated by

various proteins. [ 7 ] During endocytosis of eukaryotic cells,

proteins possessing a Bin-Amphiphysin-Rvs (BAR) domain

play an important role in the induction and maintenance of

high-curvature vesicle formation through plasma membrane

invagination. [ 8 ] Interestingly, the BAR protein superfamily

demonstrate a unique lipid membrane tubulation property

under in vitro experimental conditions. [ 9 ] When liposomes

are incubated with BAR proteins, the liposomes are dra-

matically modifi ed from spherical structures to tubules with

well-controlled diameters. [ 9b , 10 ] It is proposed that this defor-

mation mechanism is caused by the banana-shaped protein

domain binding to the membrane (by mainly electrostatic

interaction) and self-assembling to form a spiral-like struc-

ture surrounding the tubulated membrane. [ 10 ] This biological

phenomenon has inspired the development of novel tech-

nology to induce the 1D assembly of NPs within the lipid

bilayer using the BAR domain protein in ambient conditions.

Herein, we report the fi rst 1D assembly of embedded photo-

luminescent NPs in lipid nanotubules, formed from liposomes

and induced by the presence of Amphiphysin-BAR protein.

1H & Co. KGaA, Weinheim wileyonlinelibrary.com

M. Tanaka et al.full papers

Figure 1 . Schematic illustration of the fabrication of tubulated liposomes containing NPs within the lipid bilayer.

As illustrated in Figure 1 , the fabrication process consists of

two simple steps. Firstly, giant unilamellar vesicles (GUVs)

containing hydrophobic NPs were formed by a rehydration

method. [ 9a , 11 ] This study concentrates on the encapsulation

of CuInS 2 /ZnS (core/shell) QD NPs. Secondly, liposome

tubulation was performed using Amphiphysin-BAR protein.

Figure 2 . Fluorescence microscopy images of tubulated liposomes containing QDs. a) 1,2-Dihexadecanoyl- sn -glycero-3-phosphoethanolamine (DHPE) tagged with Oregon Green 0.4% (w/w), b) QDs (QD:lipid = 1:5000), and c) an overlay of the Oregon Green and QD fl uorescence. d,e) 1D assembled QDs in the lipid bilayer using Amphiphysin-BAR protein. Scale bars: 10 μ m.

Furthermore, the proposed technique was

tested for the production of various lipid

nanotubules incorporating NPs (CdSe/

ZnS QDs, Au, and Fe 3 O 4 ) under ambient

conditions, thereby demonstrating a highly

adaptable and important novel method-

ology for nanotechnology.

2. Results and Discussion

2.1. CuInS 2 /ZnS QD Synthesis and Embed-ding within the Lipid Bilayer

These cadmium-free CuInS 2 /ZnS QDs

are of interest due to their reduced tox-

icity compared to their CdSe- or CdTe-

based counterparts; they are chemically

stable, and have high photolumines-

cence quantum yield. [ 12 ] This study is

the fi rst account of liposome synthesis

2 www.small-journal.com © 2012 Wiley-VCH V

incorporating CuInS 2 /ZnS QDs within the lipid bilayer. Pre-

vious studies of QD incorporation into the liposome, within

the hydrophobic region, have highlighted the fact that incor-

poration is dependent on the QD size. [ 13 ] Recently, CdSe/ZnS

QDs capped with trioctylphosphine oxide were incorporated

within the lipid bilayer in liposomes. [ 13 ] In that study, the

CdSe QDs had an overall size of 4–5 nm including 1–2 nm

thickness of the organic coating (trioctylphosphine oxide and

ZnS). In this research, CuInS 2 /ZnS QDs (core/shell, 2–3 nm)

capped with 1-nonanethiol ( < 1.4 nm) [ 14 ] revealed NPs with a

diameter of 4.1 ± 0.8 nm and red emission ( λ em = 617 nm;

Figures S1 and S2 in the Supporting Information).

The liposomes containing QDs were successfully pro-

duced using the rehydration technique ( Figure 2 a–c). The

liposome size distributions were investigated with respect

to low and high concentrations of QDs (lower QD:lipid

ratio = 1:5000 and higher QD:lipid ratio = 1:1000) against

a control (liposomes without QDs). GUVs formed at lower

QD:lipid ratio had a diameter of 22.2 ± 9.9 μ m and were of

a similar size range to those prepared without QDs (25.2 ±

15.8 μ m). Increasing the QD loading led to a decrease in

GUV size to 11.7 ± 7.8 μ m (Figure S3 in the Supporting

Information). It should be noted that larger CuInS 2 /ZnS

QDs ( > 6 nm) were also synthesized and tested in the same

manner; however, the incorporation of the larger QDs was

rarely observed, which suggests that the small NP size is

critical.

2.2. CuInS 2 /ZnS QDs-Embedded Liposome Tubulation using Amphiphysin-BAR Protein

The addition of Amphiphysin-BAR protein to the solution

containing QD-embedded liposomes resulted in fl uorescent

1D lipid tubules (Figure 2 d,e). The morphological change

of the liposome by Amphiphysin-BAR protein did not alter

the optical properties of the QDs including photolumines-

cence and UV/Vis spectrum (Figure S4 in the Supporting

erlag GmbH & Co. KGaA, Weinheim small 2012, DOI: 10.1002/smll.201102446

Fabrication of Lipid Tubules with Embedded Quantum Dots

Figure 3 . a–c) TEM images of tubulated liposome containing QDs within the lipid bilayer: a) without QDs; b,c) with QDs (QD:lipid = 1:5000 and 1:1000) using Amphiphysin-BAR protein. d–f) Representative images of tubulated liposome d) without QDs, e) with QDs (QD:lipid = 1:5000), and f) magnifi ed image of the area indicated in (e). Scale bars: 50 nm.

Information). Figure 3 shows transmission electron micro-

scopy (TEM) images of lipid tubules at lower QD:lipid

ratio (1:5000; Figure 3 b) and higher QD:lipid ratio (1:1000;

Figure 3 c). The TEM images of the lipid nanotubules revealed

twisted structures, with many of them terminated by a spher-

ical liposome at each end of the long intermediate tubule. In

the case of lower QD concentrations, a more homogeneous

tubule similar to the control sample (without QDs) was suc-

cessfully obtained (Figure 3 a–d). The QDs within the lipid

bilayer showed a well-dispersed distribution within lipid

tubules (Figure 3 f). The diameters of tubulated liposomes

containing QDs at ratios 1:5000 and 1:1000 were approxi-

mately 40.0 ± 4.3 and 42.9 ± 4.3 nm, respectively, which shows

that embedding QDs within the liposome caused a slight

expansion of tubule diameter compared with control lipid

tubules (38.3 ± 4.1 nm). Once formed the lipid tubules con-

taining QDs within the lipid chains were stable for more than

20 h ( Figure 4 ). Interestingly, although the degree of tubule

formation was impeded by embedding QDs within the lipo-

some, the subsequent stability of the QD-embedded tubule

© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheimsmall 2012, DOI: 10.1002/smll.201102446

was similar to that of the control tubules

(without QDs).

For the control liposome samples, the

average tubulated region (the length of

tubule from liposome to liposome) was

3.3 ± 2.0 μ m and the average length of

entire tubules (from end to end including

intermediate liposomes) was 31.3 ± 12.8 μ m.

Tubules with embedded QDs within the

lipid bilayer had a reduced average region

and total length. At lower concentrations

(QD:lipid = 1:5000) the average tubule

and entire length was slightly reduced to

3.1 ± 2.0 and 19.7 ± 9.3 μ m, respectively.

At a higher QD:lipid ratio (1:1000) the

lengths were dramatically reduced: by half

for the average region (1.5 ± 1.1 μ m) and

by ten times for the entire length (3.1 ±

1.2 μ m). Therefore, the lipid nanotubule

containing QDs at the lower concentra-

tion (QD:lipid = 1:5000) had a high aspect

ratio of approximately 500. The fact that

the tubule length decreased when QDs

were embedded may be rationalized by

the decreased initial liposome size when

QDs are embedded before the tubulation

event, which offers less liposome material

to elongate to a length of fi xed diameter

(Figure S3 in the Supporting Information).

Although the protein effectively con-

trolled the diameter (40.0 ± 4.3 nm) of the

tubules, this protein-controlled diameter

selectability has not been a subject of this

study. Recently, wider lipid tubules (larger

than 100 nm) have been reported using

another BAR protein, FBP17. [ 10 ] There-

fore, we suggest this method can be devel-

oped to be more fl exible and tunable with

respect to specifi c diameter size control,

simply by using other BAR proteins and/or structurally mod-

ifi ed BAR protein. It is also noteworthy that no liposome size

fractionation step was performed in this study, but this could

be included to fabricate more homogeneous lipid tubules

with respect to length. Additionally, further optimization of

protein and lipid concentrations, lipid components, and lipo-

some size fractionation is required to develop completely

homogeneous lipid tubules without interspersed liposomes.

The ratio of tubulation events caused by Amphiphysin-

BAR protein was also affected by embedding the QDs within

the liposome (Figure 4 ). In a previous report, it was shown

that BAR proteins preferentially bind to larger liposomes,

due to their curvature-sensing properties, which arise from

their banana-shaped protein structure. [ 8b ] The fact that the

embedding of higher concentrations of QDs into the lipid

bilayer results in smaller liposomes will hence result in non-

optimal BAR protein binding, thus affecting the tubulation

activity for high concentrations of QD-embedded liposomes.

During lipid self-assembly in aqueous solution to form

liposomes, most of the small ( < 6 nm) hydrophobic NPs at low

3www.small-journal.com

M. Tanaka et al.full papers

Figure 4 . Stability of tubulated liposome containing QDs within the lipid bilayer. After liposome tubulation by Amphiphysin-BAR for 15 min at 37 ° C, the samples were incubated at 4 ° C. The time point was plotted at 0 h, and the percentage of tubulated liposome was directly counted by TEM. The stability of lipid tubules was evaluated from the different concentrations of QDs within the lipid bilayer; QD:lipid = 1:5000 ( � ), 1:1000 ( � ), and without QDs ( � ). The average values were obtained from more than 100 liposomes based on the three independent experiments. Error bars show standard deviations.

0

20

40

60

80

100

0 20 40 60 80

0

20

40

60

80

00

0 20 40 60 80

Times (Hours)

Per

cen

t o

f tu

bu

late

d l

ipo

som

e

Figure 5 . TEM images of lipid nanotubules with various embedded NPs: a) CdSe/ZnS QDs-embedded lipid nanotubules, b) Au NPs-embedded lipid nanotubules. Scale bars: 50 nm.

concentration are easily integrated within the lipid bilayer

based on the hydrophobic interaction between acyl chains

in the lipid and surface ligands (such as 1-nonanethiol) on

the QDs. The BAR domain within Amphiphysin consists of

a three-helix bundle and forms extended, banana-shaped

dimers. The concave surface of the arc presents a positive

electrostatic potential that allows for a strong interaction

with the head groups of lipids. [ 8b , 9b ] In addition to the inter-

action, the self-assembly of the Amphiphysin dimers to form

the spiral structure around the liposome surface induces the

elongation of the spherical liposome forming the lipid tubule,

even with NPs present within the lipid bilayer.

2.3. Fabrication of Lipid Nanotubules with Various Embedded NPs

This approach was also extended, at the lower particle con-

centration (NP:lipid = 1:5000), to the fabrication of lipid

tubules containing various NPs, including CdSe/ZnS, Au,

and magnetite (Fe 3 O 4 ). Hydrophobic CdSe/ZnS QDs (4.1 ±

1.2 nm), Au (4.4 ± 1.4 nm), and magnetite NPs (3.8 ± 1.8 nm)

were embedded using the same method to fabricate the

lipid nanotubules. Lipid tubules containing CdSe/ZnS QDs

and Au NPs were successfully obtained with similar sizes

to the CuInS 2 /ZnS QD-embedded lipid tubule (for CdSe/

ZnS QDs, the average entire length was 19.7 ± 8.9 μ m and

diameter 42.3 ± 6.2 nm; for Au NPs, the average entire length

4 www.small-journal.com © 2012 Wiley-VCH Verlag GmbH & Co. KGa

was 16.2 ± 7.3 μ m and diameter 41.8 ±

5.6 nm; Figure 5 ). However, lipid tubules

containing magnetite crystals were not

obtained even though spherical liposomes

containing magnetite within the lipid

bilayer were formed. This phenomenon is

curious because the magnetite NPs were

small enough to be encapsulated in the

lipid bilayer. With the exception of the NP

size, other factors such as solubility in the

lipid matrix might be important in devel-

oping the lipid tubule containing NPs.

In this study, 1-nonanethiol was used as

the surface ligand molecule for CuInS 2 /

ZnS, hexadecylamine for CdSe/ZnS,

1-octanethiol for Au, and oleic acid for

Fe 3 O 4 . While all of the ligands are hydro-

phobic, they are all straight alkane chains

in the hydrophobic region (with the excep-

tion of oleic acid). It is thus interesting to

note that it is only the magnetite coating

in oleic acid that did not embed within the

lipid bilayer. This may be a result of the

steric hindrance due to the double bond

within oleic acid that may destabilize the

incorporation process. With this in mind,

further work is being conducted to synthe-

size magnetite NP-embedded lipid tubules

with this method using a different ligand.

3. Conclusion

In summary, lipid tubules with QDs embedded within the

lipid bilayer were successfully assembled under ambient

conditions based on a simple two-step method using

A, Weinheim small 2012, DOI: 10.1002/smll.201102446

Fabrication of Lipid Tubules with Embedded Quantum Dots

Amphiphysin-BAR protein. By choosing various NPs, the

lipid tubules could be designed to have a range of proper-

ties. Additionally, the optically observable liposome/lipid

tubule (low/high-curvature membrane) with high photo-

stability could be a useful tool for elucidation of biological

reactions onto/within highly curved membranous structures.

The developed material with unique characteristics, such as

protein-induced morphological change, biocompatibility, high

dispersity, and the potential of diameter tuning with a further

molecular engineering technique, could be useful for various

applications in nanobiotechnology.

4. Experimental Section

CuInS 2 /ZnS QD Synthesis and Incorporation within the Lipid Bilayer in Liposomes : CuInS 2 /ZnS QDs were synthesized based on the method described previously with slight modifi cation. Briefl y, indium acetate (1 mmol) was mixed with copper(I) iodide (1 mmol) and 1-nonanethiol (5 mL) in a three-necked fl ask. The reaction mixture was purged with nitrogen gas for 20 min, then heated to 100 ° C until a clear yellow solution was produced. The temperature was raised to 200 ° C. When the color changed to red, the reaction was quenched by cooling the fl ask in a water bath. To passivate the core particle surface, a ZnS shell was over-coated as described previously. [ 15 ] Briefl y, this was conducted by the reac-tion of Zn stearate (25 mmol) in octadecane solution (2 mL) for 10 min at 200 ° C. The obtained QDs were washed three times with an excess amount of acetone-based solution (hexane/methanol/chloroform/acetone = 2:1:1:20). The size distribution of CuInS 2 /ZnS QDs was evaluated by high-resolution TEM analysis based on more than 200 QD images. The QD concentration after purifi cation was estimated to be 4.7 ± 0.4 μ M by measuring the absorbance at the fi rst excitation and applying the Beer–Lambert law. The molar extinction coeffi cient of the CuInS 2 /ZnS QDs was determined to be 11 000 ± 1000 cm − 1 M − 1 (methodology described by Yu et al. [ 16 ] ).

The liposome based on bovine brain lipid extract was prepared by the standard rehydration method with slight modifi cations. [ 9a , 11 ] A lipid extract (bovine brain lipid extract, Folch fraction I, 75 μ g) and QDs (18.5 or 92.3 pmol) in chloroform/methanol (4:1, v/v) were dried into a 2 mL glass vial under nitrogen gas. To avoid phase separation between different organic solvents and to achieve a homogeneous dispersion of QDs within the lipid fi lm, the lipid and QD dried precipitates were resuspended in chloroform (200 μ L) and dried under nitrogen again. The dried lipid fi lm with QDs was hydrated with nitrogen-saturated MilliQ water (50 μ L) for 20 min. After gentle addition of degassed 0.3 M sucrose (200 μ L), the glass vial was fl ushed with nitrogen, sealed, and the contents were incu-bated for more than 16 h at 37 ° C to allow spontaneous formation of GUVs. The liposome size distribution was evaluated from more than 200 liposomes in TEM images (Philips CM10 instrument, operated at an accelerating voltage of 100 kV).

Amphiphysin-BAR Protein Expression, Purifi cation, and Lipo-some Tubulation : The gene of Amphiphysin-BAR protein (human amphiphysin aa1-239) was cloned into pGEX6p3. [ 17 ] Recombinant protein was expressed in Escherichia coli as GST fusions, and puri-fi ed by glutathione–sepharose and cationic ion-exchange chro-matography (Figure S5 in the Supporting Information). The buffer was replaced with liposome buffer (20 m M 4-(2-hydroxyethyl)-1-

© 2012 Wiley-VCH Verlag GmbHsmall 2012, DOI: 10.1002/smll.201102446

[ 1 ] a) B. Dubertret , P. Skourides , D. J. Norris , V. Noireaux , A. H. Brivanlou , A. Libchaber , Science 2002 , 298 , 1759 – 1762 ; b) Y. F. Kong , J. Chen , F. Gao , W. T. Li , X. Xu , O. Pandoli , H. Yang , J. J. Ji , D. X. Cui , Small 2010 , 6 , 2367 – 2373 .

piperazineethanesulfonic acid (HEPES), 150 m M NaCl, 1 m M dithio-threitol (DTT), pH 7.4). The liposome tubulation was performed as previously described; [ 8b , 9a ] see also the McMahon’s lab proto-cols: http://www.endocytosis.org/techniqs/Liposome.html (last accessed March 2012). The obtained liposomes (300 μ g mL − 1 , 50 μ L) containing QDs were mixed with Amphiphysin-BAR protein (160 μ g mL − 1 fi nal concentration) and incubated for 15 min at 37 ° C. The samples were spotted onto 300-mesh Formvar/carbon-coated copper grids (Agar Scientifi c Ltd.), stained with uranyl acetate (1%), and analyzed by TEM. All the experiments were per-formed in triplicate and were reproducible. The diameter, length of lipid tubule, and the tubules from end to end including inter-mediate liposomes (the length of entire tubules) were evaluated by measuring at least 100 lipid tubules for each condition. Fluores-cence microscopy was carried out on an Olympus IX70 instrument. The images were acquired with the × 100 oil objective, overlaid, and cropped in the ImageJ software.

Fabrication of Lipid Nanotubules with Various Embedded NPs : The novel technique for fabrication of lipid tubules containing NPs was also performed with various NPs embedded (CdSe/ZnS QDs, Au, and magnetite (Fe 3 O 4 )) under lower particle concentration conditions (QD:lipid = 1:5000). Hydrophobic CdSe/ZnS QDs and Au NPs were purchased (hexadecylamine-coated CdSe/ZnS QDs (Sigma), 4.1 ± 1.2 nm; octanethiol-coated Au NPs (Nanoprobes), 4.4 ± 1.4 nm). Hydrophobic oleic acid-coated magnetite particles (3.8 ± 1.8 nm) were kindly provided by Dr. Puerto Morales. The magnetite particles were synthesized according to the previously reported method. [ 18 ] Using these particles, the system developed in this study was applied to the fabrication of various lipid nano-tubules with the same procedure as that for CuInS 2 /ZnS QDs lipid tubule fabrication. Fabricated lipid tubules were observed with TEM as described above.

Supporting Information

Supporting Information is available from the Wiley Online Library or from the author.

Acknowledgements

M.T. thanks the Royal Society UK for funds under the Newton inter-national fellowships scheme and is grateful for support from the Leeds EPSRC Nanoscience and Nanotechnology Facility (LENNF). K.C. thanks the Royal Society for research funding. We would like to thank Prof. Stephen Baldwin for supporting bacterial culture works and Nicole Hondow (Leeds) for support with TEM analysis. We thank Prof. Naoki Mochizuki and Dr. Michitaka Masuda for kindly providing plasmid pGEX6p3-AmphiphysinBAR (aa1-239). We thank Dr. Puerto Morales for kindly providing magnetite particles.

5www.small-journal.com & Co. KGaA, Weinheim

M. Tanaka et al.

6

full papers

[ 2 ] a) L. R. Ren , L. Guo , M. Wark , Y. L. Hou , Appl. Phys. Lett. 2005 ,

87 ; b) M. S. Hu , H. L. Chen , C. H. Shen , L. S. Hong , B. R. Huang , K. H. Chen , L. C. Chen , Nat. Mater. 2006 , 5 , 102 – 106 .

[ 3 ] a) Y. Y. Ou , M. H. Huang , J. Phys. Chem. B 2006 , 110 , 2031 – 2036 ; b) S. Sahoo , S. Husale , S. Karna , S. K. Nayak , P. M. Ajayan , J. Am. Chem. Soc. 2011 , 133 , 4005 – 4009 .

[ 4 ] I. A. Banerjee , L. Y. M. Shima , T. Yoshino , H. Takeyama , T. Matsunaga , H. Matsui , Adv. Mater. 2005 , 17 , 1128 – 1131 .

[ 5 ] C. J. Loweth , W. B. Caldwell , X. G. Peng , A. P. Alivisatos , P. G. Schultz , Angew. Chem. Int. Ed. 1999 , 38 , 1808 – 1812 .

[ 6 ] a) Y. Zhou , Q. M. Ji , M. Masuda , S. Kamiya , T. Shimizu , Chem. Mater. 2006 , 18 , 403 – 406 ; b) T. Shimizu , M. Masuda , H. Minami-kawa , Chem. Rev. 2005 , 105 , 1401 – 1443 ; c) Y. Zhou , M. Kogiso , C. He , Y. Shimizu , N. Koshizaki , T. Shimizu , Adv. Mater. 2007 , 19 , 1055 – 1058 .

[ 7 ] a) M. Tanaka , A. Arakaki , T. Matsunaga , Mol. Microbiol. 2010 , 76 , 480 – 488 ; b) T. Itoh , K. S. Erdmann , A. Roux , B. Habermann , H. Werner , P. De Camilli , Dev. Cell 2005 , 9 , 791 – 804 ; c) P. Sens , L. Johannes , P. Bassereau , Curr. Opin. Cell Biol. 2008 , 20 , 476 – 482 .

[ 8 ] a) H. T. McMahon , J. L. Gallop , Nature 2005 , 438 , 590 – 596 ; b) B. J. Peter , H. M. Kent , I. G. Mills , Y. Vallis , P. J. Butler , P. R. Evans , H. T. McMahon , Science 2004 , 303 , 495 – 499 .

[ 9 ] a) K. Takei , V. I. Slepnev , V. Haucke , P. De Camilli , Nat. Cell Biol. 1999 , 1 , 33 – 39 ; b) Q. Wang , M. V. Navarro , G. Peng , E. Molinelli , S. Lin Goh , B. L. Judson , K. R. Rajashankar , H. Sondermann , Proc. Natl. Acad. Sci. USA 2009 , 106 , 12700 – 12705 .

www.small-journal.com © 2012 Wiley-VCH V

[ 10 ] a) A. Shimada , H. Niwa , K. Tsujita , S. Suetsugu , K. Nitta , K. Hanawa-Suetsugu , R. Akasaka , Y. Nishino , M. Toyama , L. Chen , Z. J. Liu , B. C. Wang , M. Yamamoto , T. Terada , A. Miyazawa , A. Tanaka , S. Sugano , M. Shirouzu , K. Nagayama , T. Takenawa , S. Yokoyama , Cell 2007 , 129 , 761 – 772 .

[ 11 ] C. Kirby , G. Gregoriadis , Nat. Biotechnol. 1984 , 2 , 979 – 984 . [ 12 ] a) T. Pons , E. Pic , N. Lequeux , E. Cassette , L. Bezdetnaya ,

F. Guillemin , F. Marchal , B. Dubertret , ACS Nano 2010 , 4 , 2531 – 2538 ; b) H. Z. Zhong , S. S. Lo , T. Mirkovic , Y. C. Li , Y. Q. Ding , Y. F. Li , G. D. Scholes , ACS Nano 2010 , 4 , 5253 – 5262 ; c) D. E. Nam , W. S. Song , H. Yang , J. Colloid Interface Sci. 2011 , 361 , 491 – 496 .

[ 13 ] W. T. Al-Jamal , K. T. Al-Jamal , B. Tian , L. Lacerda , P. H. Bomans , P. M. Frederik , K. Kostarelos , ACS Nano 2008 , 2 , 408 – 418 .

[ 14 ] G. Z. Xu , D. P. Woodruff , N. Bennett , M. Elliott , J. E. Macdonald , Langmuir 2010 , 26 , 8174 – 8179 .

[ 15 ] L. Li , A. Pandey , D. J. Werder , B. P. Khanal , J. M. Pietryga , V. I. Klimov , J. Am. Chem. Soc. 2011 , 133 , 1176 – 1179 .

[ 16 ] W. W. Yu , L. H. Qu , W. Z. Guo , X. G. Peng , Chem. Mater. 2003 , 15 , 2854 – 2860 .

[ 17 ] M. Masuda , S. Takeda , M. Sone , T. Ohki , H. Mori , Y. Kamioka , N. Mochizuki , Embo J. 2006 , 25 , 2889 – 2897 .

[ 18 ] A. G. Roca , S. Veintemillas-Verdaguer , M. Port , C. Robic , C. J. Serna , M. P. Morales , J. Phys. Chem. B 2009 , 113 , 7033 – 7039 .

Received: November 22, 2011 Revised: December 13, 2011Published online:

erlag GmbH & Co. KGaA, Weinheim small 2012, DOI: 10.1002/smll.201102446