Neurite Outgrowth
DOI: 10.1002/anie.201106271
In Vitro Developmental Acceleration of Hippocampal Neurons on
Nanostructures of Self-Assembled Silica Beads in Filopodium-Size
Ranges**
Kyungtae Kang, Sung-Eun Choi, Hee Su Jang, Woo Kyung Cho, Yoonkey Nam,*
Insung S. Choi,* and Jin Seok Lee*
The neurite outgrowth and path-finding behaviors of neurons
are governed by two protrusive, actin-based molecular
structures, filopodia and lamellipodia, the diameter of which
is generally in the range of 100–300 nm.[1] The dynamic
stability of filopodia and lamellipodia allows neuronal cells to
recognize the surrounding environments at the nanometer
scale and to subsequently modify their cytoskeletal structures
in response to stimuli/cues.[2] For example, permissive cues
impede the retrograde flow of actin filaments and promote
their assembly by generating tensions to the filopodia,[3]
resulting in the overall advance of neurites toward the cues.
Neurons encounter nanotopographical distributions of
permissive and non-permissive cues in vivo, exemplified by
protein structures in the extracellular matrix and tissue
scaffolds or external morphologies of supporting cells.
Although the surface roughness is known to affect the
neuronal behaviors in vitro in various aspects, such as attach-
ment and survival,[4] axonal guidance,[5] and neurite out-
growth,[6] there have been few reports on developmental
responses of neurons to nanotopographies.[7,8] Because the
polarized morphology and specialized compartments (e.g.,
axons and dendrites) of neurons are crucial and decisive
characteristics for neuronal activities, it is much more
beneficial to the precise control of neuron/materials inter-
faces to manipulate neuronal development and neurite
outgrowth than to do general behaviors, such as attachment
and survival. The concrete understanding of nanotopograph-
ical effects also would contribute to the biomedical applica-
tions, which demand the controlled development or regener-
ation of nerve systems at a certain position towards a pre-
determined direction, as well as to fundamental studies on
nanotopographical role for neuronal development in vivo.[9]
To understand the function of filopodia as an antenna for
the environmental exploration,[2] it is necessary to investigate
the developmental responses of neurons to nanostructures,
the feature size of which is comparable to that of filopodia
(100–300 nm). Herein, we systematically varied the feature
sizes of nanostructures by organizing spherical nanoparticles
with different diameters on a glass substrate, and found that
the neuritogenetic acceleration of hippocampal neurons
occurred on the nanostructures the period of which was
larger than 200 nm.
For the generation of nanosurfaces with various feature
sizes, silica beads with the diameters ranging from 100 to
700 nm were synthesized by hydrolysis of tetraethyl orthosi-
licate.[10] The diameter was controlled to be 110, 190, 320, 480,
or 670 nm by changing the re-growth cycles of seeds and the
concentrations of reactants (See the Supporting Information
for the experimental details; Figure S1). The silica beads were
denoted as SB-110, SB-190, SB-320, SB-480, and SB-670,
respectively. Among the methods for organizing nano-sized
beads into two-dimensional arrays on a solid substrate,[11]
rubbing was reported to be a highly effective in themonolayer
assembly with the precise control of the interparticle distance,
and simple, fast, and reproducible on a large area (Fig-
ure 1a).[12] Figure 1b shows the SEM images of the bead
monolayers with the different diameters, formed by the
rubbing method. This simple method yielded high-quality
monolayers with uniform interparticle distance regardless of
the bead diameters. To prevent the undesired detachment of
silica beads from the surface during cell culture, we baked the
substrate at 350 8C for 3 h (see the Supporting Information,
Figure S2, for the experimental details). Figure 1c shows
a representative SEM image of a hippocampal neuron
cultured on a monolayer of SB-110, where the beads are
smaller than the filopodia.
The surface of silica beads was made to be neuron-
adhesive by 2 min-O2-plasma treatment, followed by 1 h
dipping in aqueous poly-d-lysine solution (0.1 mgmL1 in
[*] K. Kang,[+] Dr. W. K. Cho, Prof. Dr. I. S. Choi
Molecular-Level Interface Research Center, Department of Chemis-
try, KAIST
Daejeon 305-701 (Korea)
E-mail: ischoi@kaist.ac.kr
Homepage: http://cisgroup.kaist.ac.kr
Prof. Dr. Y. Nam, Prof. Dr. I. S. Choi
Department of Bio and Brain Engineering, KAIST
Daejeon 305-701 (Korea)
E-mail: ynam@kaist.ac.kr
Homepage: http://neuros.kaist.ac.kr
S.-E. Choi,[+] H. S. Jang, Prof. Dr. J. S. Lee
Department of Chemistry, Sookmyung Women’s University
Seoul 140-742 (Korea)
E-mail: jinslee@sookmyung.ac.kr
Homepage: http://sookmyung.ac.kr/~FETNS
[+] These authors contributed equally to this work.
[**] This work was supported by the Korea Research Foundation Grant
funded by the Korean Government (MOEHRD, KRF-2008-313-
d00614) and the Basic Science Research Programs through the
National Research Foundation of Korea (NRF) funded by the
Ministry of Education, Science and Technology (2011-0001318,
2009-0077751, 2009-0080081, and 2010-0025065) and the Brain
Research Center of the 21st Century Frontier Research Program.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201106271.
Angewandte
Chemie
2855Angew. Chem. Int. Ed. 2012, 51, 2855 –2858  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

deionized water). Hippocampal neurons were then plated at
the density of 50 cells/mm2. Figure 2a shows phase-contrast
optical images of the neurons at 1 DIV (day in vitro) and
2 DIV. Based on neuronal morphologies, the substrates were
categorized into two groups: Group-I (SB-110 and SB-190)
and Group-II (SB-320, SB-480, and SB-670). We observed
noticeable differences in neuronal development between the
two groups even at 1 DIV. In Group-I, most of neurons had
either lamellipodia or short minor neurites at 1 DIV, and the
ratio between neurons with and without neurites was about
4:6. This ratio was similar to neurons grown on a polylysine-
treated glass coverslip at 1 DIV.[13] In a stark contrast, more
than 80% of Group-II neurons had neurites at 1 DIV, and the
length of the neurites was significantly longer than that of
Group-I neurons. In terms of axonal development, more
neurons in Group-II had a major neurite (“putative” axon)
than those in Group-I, which implied that axonal polarization
processes were more active on Group-II substrates than
Group-I substrates. At 2 DIV, Group-II neurons had one long
branching axon and a few minor neurites, which was
a phenotype of cultured hippocampal neurons.[13] The immu-
nostaining showed that the microtubules and filopodial/
lamellipodial structures were well-developed on all the bead
substrates (See the Supporting Information; Figure S3).
These results were in a good agreement with our previous
report: the developmental acceleration occurred on a 400 nm-
pitched anodized aluminum oxide substrate.[8]
Scanning electron microscopy (SEM) was used to exam-
ine the physical interactions of neurons with silica beads. The
bead size seemed to affect the filopodial contact with the
substrate in a different fashion: filopodia of the neurons in
Group-I extended straight regardless of the bead curvature,
while filopodia in Group II did not grow straight and seemed
to have many interactions with surface structures such as
trenches, bead surfaces, and gaps between the beads (Fig-
ure 2b, insets). The SEM micrographs showed that different
bead sizes caused the different spatial distribution of inter-
action spots between the filopodia and the substrate. These
interactions were believed to be phenomenologically differ-
ent from the ones with microstructures. Microtopographies,
such as microchannels and micropillar array with feature sizes
larger than 1 mm,[4–6] were reported to have influences on
neuronal growth directly at cellular scale; entire neurites were
trapped within the microstructures, which physically
restricted the direction of neurite growth or generated
cellular-scale tensions.[5,6] In contrast, nanotopographies
would have localized effects on subcellular structures such
as filopodia, amplified to the cellular-scale responses such as
developmental acceleration. Our results supports the idea
that nanotopographical features trigger intracellular signaling
pathways through mechanotransductions of cytoskeletal
structures, and the discontinuous or jumping stimuli of
Figure 1. a) A schematic illustration of the formation of two-dimen-
sional nanostructures of silica beads on a glass substrate. b) SEM
images of the nanostructures assembled with SB-110, SB-190, SB-320,
SB-480, and SB-670 (from left to right). The scale bar is 1 mm. c) SEM
image of a neuron at 1 DIV cultured on SB-110. The scale bar is 5 mm.
The arrowheads indicate filopodia.
Figure 2. a) Phase-contrast optical micrographs of hippocampal neu-
rons cultured on bead-packed substrates. The scale bar is 50 mm.
b) SEM images of the hippocampal neurons on SB-110 (left) and SB-
480 (right). The scale bar is 5 mm. Insets: Magnified images of
filopodial tips. The scale bar is 1 mm. The original gray images were
rendered with a monochromatic color for the comparison between the
groups.
.Angewandte
Communications
2856 www.angewandte.org  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2012, 51, 2855 –2858

adhesion bigger than the size of filopodia would change the
developmental characteristics of neurons.
To investigate neuritogenesis between the groups (Group-
I and Group-II) further, we cultured neurons on coverslips
and classified them into three developmental stages:[8,13]
lamellipodia form around the soma (stage 1); the lamellipo-
dia coalesce at several discrete sites around the cell periphery,
where minor neurites begin to extend with growth cones at
their ends (stage 2); one neurite (major neurite) grows two- or
three-times longer than the others, and cell morphology is
polarized (stage 3). Figure 3a clearly shows the differences in
distribution of neuronal populations in the two groups. In
Group-I, more than 50% of the neurons belonged to stage 1
at 1 DIV, while more than 75% of the neurons in Group-II
were already at stage 2 or 3. At 2 DIV, few neurons in Group-
II remained in stage 1 (4, 0, and 0% for SB-320, SB-480, and
SB-670, respectively), and many neurons populated more at
stage 3 (58, 54, and 66%). The neurons in Group-I were also
developed, but the percentage of the neurons in stage 1 was
still 21–25% at 2 DIV. The length of the longest major neurite
also showed similar bimodal distribution (Figure 3b). At
1 DIV, the average length for Group-II was 44–49 mm, while
that for Group-I was 23–25 mm. The length difference became
much larger at 2 DIV: 151.8 16.5 mm for SB-670, 129.2
17.6 mm for SB-480, 74.3 4.9 mm for SB-320, 61.2 4.8 mm
for SB-190, and 59.3 5.1 mm for SB-110. Although the length
for SB-320 was relatively short, it was still longer than that for
Group-I. Taken together, the results indicated that neurons
sprouted neurites faster, and the polarization process was also
accelerated on the substrates with beads larger than 200 nm in
diameter. The neurons sensed the nanostructures differently
and behaved differently with a threshold of about 200–
300 nm.
To examine the importance of filopodia in sensing nano-
topographical differences, we treated neurons with cytocha-
lasin D, an F-actin-depolymerization agent, while they were
cultured on silica beads (Figure 4a). Cytochalasin D has been
shown to be effective in disrupting F-actin-based structures of
hippocampal neurons, such as filopodia, generating elongated
Figure 3. a) Percentages of neurons in each stage at 1 DIV and 2 DIV.
The results from the substrates were compared with SB-110 by the chi-
square test. There was a significant difference (*p<0.001). b) Average
length (standard error) of major neurites. All substrates were
compared with SB-110 by one-way ANOVA at the significant level of
99%, followed by the Bonferroni’s multiple comparison test
(*p<0.001). The numbers indicate data points for the statistics.
CTRL: poly-d-lysine-coated coverslips.
Figure 4. a) Chemical structure of cytochalasin D. b) Phase-contrast
micrographs of hippocampal neurons cultured on bead-packed sub-
strates (SB-110 and SB-670) with the treatment of cytochalasin D. The
scale bar is 50 mm. c) Quantitative analyses of the longest-neurite
lengths and the average number of neurites. There were no significant
differences between any pair of them for both graphs (N=26, 60, 36,
25, and 30 for the number of neurites, and 27, 66, 33, 29, and 40 mm
for the neurite lengths; mean s).
Angewandte
Chemie
2857Angew. Chem. Int. Ed. 2012, 51, 2855 –2858  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.angewandte.org

neurites and multiple axons.[14] We added cytochalasin D to
the culture media (final concentration: 1 mm) immediately
after seeding, and cultured neurons for 1 day. Interestingly, we
did observe the same phenotype for all the substrates
including the glass control: treated with cytochalasin D,
neurons exhibited elongated, multiple, thin, and curved
neurites at 1 DIV (Figure 4b for representative images for
SB-110 and SB-670). In addition, there was no significant
difference in both the length of the major neurite and the
number of the neurites (Figure 4c). The results implied that
filopodia were essential for the observed, different responses
of hippocampal neurons toward nanotopographical surface
features.
In summary, we showed that the developmental acceler-
ation of hippocampal neurons occurred on the well-packed
structures of silica beads bigger than 200 nm in diameter. The
biochemical inhibition of filopodia formation suggested that
neurons sensed nanotopographies through filopodial activ-
ities, which in turn modulated intracellular cytoskeletal
dynamics, just as they sense the biochemical cues. Our results
also imply that nanotopographical cues are an important
feature for guiding neurites during the neural developments
in vivo. We believe that this work would provide fundamental
but crucial information for studying nanotopographical
manipulation of neuronal development, and also be useful
for designing sophisticated neural interfaces in neural tissue
engineering and others.
Experimental Section
Cell Culture: Primary hippocampal neurons were cultured in serum-
free condition. Hippocampus from E-18 Sprague-Dawley rat was
triturated in 1 mL of Hanks Balanced Salt Solution (HBSS) using
a fire-polished Pasteur pipette. The cell suspension was centrifuged
for 2 min at 1000 rpm, and a cell pallet was extracted. The cell pallet
was suspended in Neurobasal media supplemented with B-27, 2 mml-
glutamine, 12.5 mm l-glutamic acid, and Penicillin-Streptomycin.
Dissociated cells were seeded at the density of 50 cellsmm2 on
a silica bead-assembled substrate. Cultures were maintained in an
incubator (5% CO2 and 37 8C), and a half of media was replaced with
fresh culture media without l-glutamic acid supplement every 3–
4 days. This study was approved by IACUC (Institutional Animal
Care and Use Committee) of KAIST.
Instruments and Characterizations: The surface topography of
the prepared substrates was investigated by field-emission scanning
electron microscopy (FE-SEM; Hitachi S-4800). Before FE-SEM
imaging, the cultured substrates were coated with platinum (30 mA,
360 s). Fluorescence micrographs of neuron cultures were obtained
using Olympus BX51m (Olympus Corp.) equipped with a CCD
camera (DP71, Olympus Corp.). From the images, the lengths of
major neurites were measured with Neuron J plugin in Image J
software (NIH).
Received: September 5, 2011
Revised: October 7, 2011
Published online: November 25, 2011
.Keywords: axon outgrowth · cytoskeletal proteins ·
nanostructures · neurochemistry · silica bead
[1] P. K. Mattila, P. Lappalainen, Nat. Rev. Mol. Cell Biol. 2008, 9,
446 – 454.
[2] a) F. Polleux,W. Snider,Cold SpringHarbor Perspect. Biol. 2010,
2, a001925; b) D. A. Fletcher, R. D. Mullins, Nature 2010, 463,
485 – 492; c) C. Conde, A. Caceres, Nat. Rev. Neurosci. 2009, 10,
319 – 332.
[3] L. A. Lowery, D. Van Vactor, Nat. Rev. Mol. Cell Biol. 2009, 10,
332 – 343.
[4] a) V. Brunetti, G. Maiorano, L. Rizzello, B. Sorce, S. Sabella, R.
Cingolani, P. P. Pompa, Proc. Natl. Acad. Sci. USA 2010, 107,
6264 – 6269; b) W. Hllstrçm, T. Mrtensson, C. Prinz, P.
Gustavsson, L. Montelius, L. Samuelson, M. Kanje, Nano Lett.
2007, 7, 2960 – 2965; c) J. M. Bruder, A. P. Lee, D. Hoffman-Kim,
J. Biomater. Sci. Polym. Ed. 2007, 18, 967 – 982.
[5] a) L. Yao, S. Wang, W. Cui, R. Sherlock, C. OConnell, G.
Damodaran, A. Gorman, A. Windebank, A. Pandit, Acta
Biomater. 2009, 5, 580 – 588; b) A. Ferrari, M. Cecchini, A.
Dhawan, S. Micera, I. Tonazzini, R. Stabile, D. Pisignano, F.
Beltram, Nano Lett. 2011, 11, 505 – 511; c) F. Johansson, P.
Carlberg, N. Danielsen, L. Montelius, M. Kanje, Biomaterials
2006, 27, 1251 – 1258; d) D. Y. Fozdar, J. Y. Lee, C. E. Schmidt, S.
Chen, Biofabrication 2010, 2, 035005; e) A. Rajnicek, S. Brit-
land, C. McCaig, J. Cell Sci. 1997, 110(Pt 23), 2905 – 2913;
f) N. M. Dowell-Mesfin, M. A. Abdul-Karim, A. M. Turner, S.
Schanz, H. G. Craighead, B. Roysam, J. N. Turner, W. Shain, J.
Neural Eng. 2004, 1, 78 – 90.
[6] N. Gomez, Y. Lu, S. Chen, C. E. Schmidt, Biomaterials 2007, 28,
271 – 284.
[7] a) C. C. Gertz, M. K. Leach, L. K. Birrell, D. C. Martin, E. L.
Feldman, J. M. Corey, Dev. Neurobiol. 2010, 70, 589 – 603; b) H.
Hu, Y. Ni, V. Montana, R. C. Haddon, V. Parpura, Nano Lett.
2004, 4, 507 – 511; c) M. J. Jang, S. Namgung, S. Hong, Y. Nam,
Nanotechnology 2010, 21, 235102.
[8] W. K. Cho, K. Kang, G. Kang, M. J. Jang, Y. Nam, I. S. Choi,
Angew. Chem. 2010, 122, 10312 – 10316; Angew. Chem. Int. Ed.
2010, 49, 10114 – 10118.
[9] a) N. Y. Harel, S. M. Strittmatter, Nat. Rev. Neurosci. 2006, 7,
603 – 616; b) D. Hoffman-Kim, J. A. Mitchel, R. V. Bellam-
konda, Annu. Rev. Biomed. Eng. 2010, 12, 203 – 231.
[10] a) W. Stçber, A. Fink, E. Bohn, J. Colloid Interface Sci. 1998, 26,
62 – 69; b) K. D. Hartlen, A. P. T. Athanasopoulos, V. Kitaev,
Langmuir 2008, 24, 1714 – 1720; c) T. Yokoi, Y. Sakamoto, O.
Terasaki, Y. Kubota, T. Okubo, T. Tatsumi, J. Am. Chem. Soc.
2006, 128, 13664 – 13665.
[11] a) P. Jiang,M. J. McFarland, J. Am. Chem. Soc. 2004, 126, 13778 –
13786; b) S. Wong, V. Kitaev, G. A. Ozin, J. Am. Chem. Soc.
2003, 125, 15589 – 15598; c) R. Micheletto, H. Fukuda, M.
Ohtsut,Langmuir 1996, 11, 3333 – 3336; d) P. Jiang, J. F. Bertone,
K. S. Hwang, V. L. Colvin, Chem. Mater. 1999, 11, 2132 – 2140.
[12] a) J. S. Lee, J. H. Kim, Y. J. Lee, N. C. Jeong, K. B. Yoon, Angew.
Chem. 2007, 119, 3147 – 3150; Angew. Chem. Int. Ed. 2007, 46,
3087 – 3090; b) N. N. Khanh, K. B. Yoon, J. Am. Chem. Soc. 2009,
131, 14228 – 14230.
[13] S. Kaech, G. Banker, Nat. Protoc. 2006, 1, 2406 – 2415.
[14] F. Bradke, C. G. Dotti, Science 1999, 283, 1931 – 1934.
.Angewandte
Communications
2858 www.angewandte.org  2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2012, 51, 2855 –2858