Protein-coated nanocapsules via multilevel surface modification. Controlled preparation and microscopic analysis at nanometer resolution

Article abstract of DOI:10.1039/c3cc40752g

We prepared self-assembled vesicles comprising layers of fullerene, alkyl chains, triazole, oligo(ethylene oxide) chains, biotin and avidin molecules. High-resolution scanning electron microscopy using a superhydrophilic indium-tin oxide substrate was found to be useful for nanometer-level structural analysis of the vesicles and individual avidin molecules whose structures are difficult to analyze.

Full text of DOI:10.1039/c3cc40752g

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Protein-coated nanocapsules via multilevel surface
modification. Controlled preparation and microscopic
analysis at nanometer resolution†
Cite this: Chem. Commun., 2013,
49, 3525
Received 29th January 2013,
Accepted 11th March 2013
Koji Harano,a Kosuke Minami,a Eisei Noiri,b Koji Okamotob and Eiichi Nakamura*a
DOI: 10.1039/c3cc40752g
www.rsc.org/chemcomm
We prepared self-assembled vesicles comprising layers of fullerene,
alkyl chains, triazole, oligo(ethylene oxide) chains, biotin and avidin
molecules. High-resolution scanning electron microscopy using a
superhydrophilic indium-tin oxide substrate was found to be useful
for nanometer-level structural analysis of the vesicles and individual
avidin molecules whose structures are difficult to analyze.
Sub-100 nm capsular objects such as virus capsids displaying
proteins on their surface serve as a model of nanometer-sized
tools for the detection, reporting and delivery of molecules in a
biological environment.1 Biological nanocapsules are formed
by controlled integration of functional parts through multistep
covalent and noncovalent chemical processes, which, however,
is difficult to achieve in a flask for several reasons: the paucity
of methods to prepare sub-100 nm capsular platforms under
mild conditions, to modify their structures after their formation
in water, and, most importantly, to analyze with molecular
precision the structure of such sub-100 nm objects that typically
lack structural uniformity and periodicity.2 We report here
that the ca. 8-nanometer-radius vesicles of the potassium salt
of penta(oct-7-yl)fullerene3 (C8(7Y)K) serve as a platform for
construction of protein-covered nanocapsules through multilevel
functionalization in a highly controlled manner (Fig. 1), and
that their structure can be studied with a state-of-the-art, sub-
nanometer spatial resolution, low-accelerating-voltage scanning
Fig. 1 Preparation of a biotinylated fullerene vesicle from fullerene amphiphile
C8(7Y)K via a click reaction and conjugation with avidin.
electron microscope (SEM) equipped with a monochromatic that cannot be obtained using dynamic light scattering (DLS) or
electron source.4 This SEM combined with the use of a super- conventional microscopy such as atomic force microscopy (AFM).
hydrophilic and conducting indium-tin oxide (ITO) substrate
To prepare the protein-covered vesicles, we employed a
permitted us to obtain subnanometer resolution images of the combination of four known sequences: quantitative organocopper
vesicles and the small protein molecules. This new SEM technology addition to form penta(oct-7-yl)fullerene,5 dissolution of its
provided structural information on nonconducting nanoparticles potassium salt as C8(7Y)K in water, installation of biotin
groups by a copper(I)-catalyzed click reaction,6–8 and biotin–
a Department of Chemistry, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku,
Tokyo 113-0033, Japan. E-mail: nakamura@chem.s.u-tokyo.ac.jp;
Fax: +81-3-5800-6889; Tel: +81-3-5800-6889
b Department of Hemodialysis and Apheresis, University Hospital,
The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8655, Japan.
E-mail: noiri-tky@umin.ac.jp; Fax: +81-3-5814-8696; Tel: +81-3-5800-8648
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c3cc40752g
avidin molecular recognition (Fig. 1). The fullerene amphiphile
is unique in its ability to form stable vesicles on which we can
display various functionalities in an aqueous phase.9 We prepared
a 2.0 mM (for fullerene) vesicle solution by slow injection of a THF
solution of C8(7Y)K into pure water or D2O followed by removal of
THF under vacuum. DLS analysis (cumulant analysis) showed a
very narrow size distribution (polydispersity index 0.05), and the
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 3525--3527 3525

View Article Online
Communication
ChemComm
hydrodynamic radius of the vesicle was determined to be 8.1 Æ
0.1 nm. Therefore, one vesicle exposes about 400 fullerene
molecules and 2000 alkyne termini on the surface.
We next installed biotin groups on the vesicle surface by the
click reaction of the C8(7Y)K vesicle with 5 equiv. (for one
C8(7Y)K molecule) of an azido-tagged biotin–oligo(ethylene
oxide) conjugate (biotin-N3; Fig. 1, top right) in the presence
of copper(II) sulfate (50 mol%, 10 mol% per alkynyl group) and
sodium ascorbate as a reducing agent.10 The size distribution
was rather narrow (polydispersity index 0.12) and the radius
increased by 3.5 nm from 8.1 nm to 11.6 Æ 0.1 nm (DLS)—an
increase in accordance with the length of the tethered biotin
unit (ca. 2.8 nm). This biotinylated fullerene vesicle is robust
enough to be purified by gel permeation chromatography on
Sephadex G50 for removal of unreacted biotin-N3. No DLS radius
change occurred during the purification. A separate set of
experiments using a fluorescent probe to estimate the efficiency
of the click reaction11 conducted at a micromolar concentration
indicated that the click reaction allowed ca. 400 biotin residues
per vesicle (ESI†)—a value much higher than the number of
60 required to fully cover the surface with avidin molecules in
the next step.
This biotinylation protocol can also be applied to a C8(7Y)K
vesicle that contains an anticancer drug (Fig. 1, middle).12
Thus, we added 5.0 mmol of C8(7Y)K in THF to an aqueous
solution containing 5.3 mmol of doxorubicin (DOX) to form the
DOX-containing vesicle, and then attached the biotin groups by
the click reaction followed by separation of the vesicles from
excess DOX and copper reagents, etc., by chromatography. We
incubated this vesicle solution, which contains 2.5 pmol of
DOX (25 nM if homogeneously dissolved in the medium),
together with human liver carcinoma cells (HepG2) for 48 h,
and observed a cell viability of 67 Æ 7%. In contrast, the biotin
vesicles free of DOX showed a cell viability of 93 Æ 3%, and the
25 nM solution of DOX without vesicle encapsulation showed a
cell viability of 93 Æ 3% (an IC50 value of 3.4 mM determined
under the same conditions; see ESI†). The experiments showed
Fig. 2 HR-SEM images of avidin-coated vesicles coated on a microcrystalline
ITO/glass substrate. The images were taken on an FEI Magellan 400L XHR SEM
equipped with a unicolor monochromatic electron source (o0.2 eV energy
spread; except in Fig. 2f). (a) The image taken as the first frame had an electron
beam current of 25 pA. Note a nearly homogeneous smooth surface. Scale bar is
100 nm. Inset: a single vesicle with a scale bar of 30 nm. (b) Representative
images of vesicles taken after scanning at 25 pA for 3 min. Note the development
of fine structures on the surface. The scale bar is 20 nm. (c) The same image as in
Fig. 2b, far right, where the bright areas are highlighted in red. (d) Histogram for
the 62 bright areas on 36 avidin-covered vesicles. (e) Vesicles and ITO micro-
crystals after deposition of 1 nm Pd/Pt coating. The scale bar is 100 nm. Inset:
single vesicles with a scale bar of 30 nm. (f) Image of the same sample taken with
an electron beam with >0.65 eV energy spread. The scale bar is 100 nm.
that the click reaction and the vesicle purification were remove two inherent problems of conventional SEM imaging
achieved without the loss of the encapsulated drug. that have prevented its application in the nanometer-resolution
Finally, we noncovalently bonded avidin molecules to the analysis of nonconducting objects: nanometer-thick Pt/Pd coating
vesicle surface. Upon mixing the vesicle with avidin in PBS of the specimen that greatly reduces the quality of the information
buffer (a fullerene/avidin mole ratio of 1 : 1), the DLS analysis on the surface structure and electron damage to the specimen.
of this solution showed a very broad peak centered at 41 nm.
Fig. 2 shows the images of the avidin-covered vesicle under
The very large polydispersity index (0.23) suggested vesicle some representative conditions. Fig. 2a shows the image under
aggregation, and hence gave little information on the structure the optimum conditions developed in this study, where the
of the individual vesicle particles. However, the SEM analysis vesicle solution was spin-coated (1500 rpm) on a superhydrophilic
described in the following paragraphs indicates that we ITO/glass substrate treated by UV/ozone just before use (a water
have successfully achieved the construction of protein-covered contact angle of 01). The coated substrate was dried for 1 h at room
nanocapsules through multilevel functionalization in a highly temperature before being placed in the SEM chamber kept at 5 Â
controlled manner.
10À5 Pa at room temperature. Imaging was achieved with an
Knowing that aggregation is a ubiquitous phenomenon and electron beam current of 25 pA and an acceleration voltage of ca.
represents a bottleneck for structural analysis in studies on self- 1–2 kV. Note that the high affinity of the ITO surface for the avidin
assembled organic materials and protein, we consider it useful molecules entirely deaggregated the vesicles, allowing us to study
to explore the utility of a state-of-the-art SEM that features low individual vesicles and their surface (Fig. 2a and Fig. S7 in ESI†).
acceleration voltage, a monochromatic electron beam (o0.2 eV Note that, on the hydrophobic Si/SiO2 surface, the vesicles
energy spread), and a subnanometer spatial resolution. Not just aggregated and fused extensively (contact angle = ca. 701) and
allowing for imaging of fine surface details, these features could not be coated on gold or graphite. The vesicle radius
c
3526 Chem. Commun., 2013, 49, 3525--3527
This journal is The Royal Society of Chemistry 2013

View Article Online
ChemComm
Communication
measured for 112 vesicles was 15.5 Æ 2.7 nm (Fig. 2b), and we see This sub-nm-resolution SEM technique provides a new single
hints of surface morphology. molecule-level approach to the analysis of nm-sized organic
The first few scans (ca. 3 min at a beam current of 25 pA) molecules and materials,15 supplementing the AFM technology
caused a change in the surface structure (e.g. dehydration of that gives much lower point resolution and the conventional
protein), which now shows bright and protruded areas, and optics-based microscopy that is useful for molecular ensemble
dark areas (Fig. 2b and c). The histogram of the bright areas analysis. The subnanometer-resolution electron microscopic
(circled in red in Fig. 2c) showed two peaks at 30–45 nm2 and imaging of small molecules and proteins is an emerging
65–75 nm2 (analyzed for those located nearly in the center of methodology,16 and the method reported here will provide a
the vesicle image, Fig. 2d). The former is close to the projection useful platform for future exploration of the potential of SEM.
area of a single molecule of avidin (25–30 nm2),13 and the latter
This work was supported by KAKENHI on Specially Promoted
to that of two molecules. Therefore, we consider that the bright Research (22000008) to E.N. and Innovative Areas ‘‘Coordination
areas are due to individual avidin molecules. From these data, Programming’’ (Area 2107, 24108710) to K.H. from MEXT, Japan.
we can estimate that at least ca. 20 avidin molecules are K.M. thanks the Japan Society for Promotion of Science for a
attached to the vesicle surface.
predoctoral fellowship.
Fig. 2e illustrates the detrimental effects on the image quality
of the conventional Pt/Pd coating (1 nm) of the specimen taken
under the same conditions as those used for Fig. 2a. The metal
coating covered the avidin molecules and the ITO microcrystals.
The coating increased the average vesicle radius by 4 nm to
19.7 Æ 2.2 nm (average of 93 vesicles). The average radius of
vesicles after 5 nm Pt/Pd coating increased further to 24.6 Æ
3.4 nm (ESI†). Therefore, nonconducting substrates such as
mica are unsuitable for nanometer-resolution SEM imaging
because they inevitably need a metal coating. An SEM using
a nonmonochromatic electron source (electron beam with
>0.65 eV energy spread) produced the image in Fig. 2f that has
an image contrast too low to be useful.
The SEM measurement of the biotinylated vesicle showed a
radius of 11.7 Æ 1.6 nm, which agrees very well with the
hydrodynamic radius determined by DLS (11.6 Æ 0.1 nm, see
above). Thus, we consider that the SEM measurement provides
a reliable alternative for the determination of the size of
nanoparticles. In contrast, AFM imaging of the vesicles showed
a radius larger by the size of the probe (i.e., 2 nm), and the
conventional SEM conditions using Pt/Pd coating also gave a
size larger than the real size (see above).
In summary, we have coated the surface of a nanocapsule
with avidin molecules through stepwise noncovalent and
covalent modifications with good control and precise analysis
of each functionalization step. The alkynylated fullerene
formed a vesicle, which withstood the conditions of chemical
reactions and chromatographic purification, and served as a
carrier for drugs. We have found that the combined use of a
superhydrophilic ITO/glass as a substrate and the low-accelerating-
voltage SEM provides a viable method for imaging nonconducting
nanometer-scale objects and proteins with high precision and
surface sensitivity. ITO/glass is frequently used in organic electro-
Notes and references
1 (a) D. Peer, J. M. Karp, S. Hong, O. C. Farokhzad, R. Margalit and
R. Langer, Nat. Nanotechnol., 2007, 2, 751; (b) S. Ganta,
H. Devalapally, A. Shahiwala and M. Amiji, J. Controlled Release,
2008, 126, 187; (c) J. Voskuhl and B. J. Ravoo, Chem. Soc. Rev., 2009,
38, 495; (d) B. Worsdorfer, K. J. Woycechowsky and D. Hilvert,
Science, 2011, 331, 589.
2 T. Ogura, PLoS One, 2012, 7, e46904.
3 S. Q. Zhou, C. Burger, B. Chu, M. Sawamura, N. Nagahama,
M. Toganoh, U. E. Hackler, H. Isobe and E. Nakamura, Science,
2001, 291, 1944.
4 R. Young, S. Henstra, J. Chmelik, T. Dingle, A. Mangnus, G. van Veen
and I. Gestmann, Proc. SPIE, 2009, 737803.
5 Y. Matsuo and E. Nakamura, Chem. Rev., 2008, 108, 3016.
6 (a) V. V. Rostovtsev, L. G. Green, V. V. Fokin and K. B. Sharpless,
Angew. Chem., Int. Ed., 2002, 41, 259; (b) C. W. Tornøe,
C. Christensen and M. Meldal, J. Org. Chem., 2002, 67, 3057;
(c) M. Meldal and C. W. Tornøe, Chem. Rev., 2008, 108, 2952.
7 (a) S. Cavalli, A. R. Tipton, M. Overhand and A. Kros, Chem.
Commun., 2006, 3193; (b) F. S. Hassane, B. Frisch and F. Schuber,
Bioconjugate Chem., 2006, 17, 849; (c) C. Schatz, S. Louguet, J.-F.
Le Meins and S. Lecommandoux, Angew. Chem., Int. Ed., 2009, 48, 2572.
8 H. Isobe, K. Cho, N. Solin, D. B. Werz, P. H. Seeberger and
E. Nakamura, Org. Lett., 2007, 9, 4611.
9 (a) T. Homma, K. Harano, H. Isobe and E. Nakamura, Angew. Chem.,
Int. Ed., 2010, 49, 1665; (b) T. Homma, K. Harano, H. Isobe and
E. Nakamura, J. Am. Chem. Soc., 2011, 133, 6364.
10 C. M. Santos, A. Kumar, W. Zhang and C. Cai, Chem. Commun., 2009,
2854.
11 K. Sivakumar, F. Xie, B. M. Cash, S. Long, H. N. Barnhill and
Q. Wang, Org. Lett., 2004, 6, 4603.
12 (a) G. Russell-Jones, K. McTavish, J. McEwan, J. Rice and
D. Nowotnik, J. Inorg. Biochem., 2004, 98, 1625; (b) J. Chen,
S. Chen, X. Zhao, L. V. Kuznetsova, S. S. Wong and I. Ojima,
J. Am. Chem. Soc., 2008, 130, 16778.
13 B. Steiger, C. Padeste, A. Grubelnik and L. Tiefenauer, Electrochim.
Acta, 2003, 48, 761.
14 H. Pluk, D. J. Stokes, B. Lich, B. Wieringa and J. Fransen, J. Microsc.,
2009, 233, 353.
15 H. Yabu, M. Kanahara, M. Shimomura, T. Arita, K. Harano,
E. Nakamura, T. Higuchi and H. Jinnai, ACS Appl. Mater. Interfaces,
2013, DOI: 10.1021/am4003149.
nics research, but it has seldom been used for biological imaging 16 (a) E. Nakamura, Angew. Chem., Int. Ed., 2013, 52, 236;
(b) M. Koshino, T. Tanaka, N. Solin, K. Suenaga, H. Isobe and
E. Nakamura, Science, 2007, 316, 853; (c) K. Harano, T. Homma,
Y. Niimi, M. Koshino, K. Suenaga, L. Leibler and E. Nakamura, Nat.
except for one report on micrometer-resolution SEM imaging of
cultured cells.14 We found that the ITO substrate is useful
for imaging individual objects that tend to aggregate in water.
Mater., 2012, 11, 877.
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 3525--3527 3527