Letter
pubs.acs.org/NanoLett
Characterization of the Cell−Nanopillar Interface by Transmission
Electron Microscopy
Lindsey Hanson,† Ziliang Carter Lin,‡ Chong Xie,§,⊥ Yi Cui,*,§,∥ and Bianxiao Cui*,†
†
Department of Chemistry, ‡Department of Applied Physics, and §Department of Materials Science and Engineering, Stanford
University, Stanford, California 94305, United States
∥
SLAC National Accelerator Laboratory, Stanford Institute for Materials and Energy Sciences, 2575 Sand Hill Road, Menlo Park,
California 94025, United States
S Supporting Information
*
ABSTRACT: Vertically aligned nanopillars can serve as
excellent electrical, optical and mechanical platforms for
biological studies. However, revealing the nature of the interface
between the cell and the nanopillar is very challenging. In
particular, a matter of debate is whether the cell membrane
remains intact around the nanopillar. Here we present a detailed
characterization of the cell-nanopillar interface by transmission
electron microscopy. We examined cortical neurons growing on
nanopillars with diameter 50−500 nm and heights 0.5−2 μm.
We found that on nanopillars less than 300 nm in diameter, the
cell membrane wraps around the entirety of the nanopillar without the nanopillar penetrating into the interior of the cell. On the
other hand, the cell sits on top of arrays of larger, closely spaced nanopillars. We also observed that the membrane-surface gap of
both cell bodies and neurites is smaller for nanopillars than for a flat substrate. These results support a tight interaction between
the cell membrane and the nanopillars and previous findings of excellent sealing in electrophysiology recordings using nanopillar
electrodes.
KEYWORDS: Nanopillar, cell membrane, cell−nanostructure interface, transmission electron microscopy
V
intact cell membrane wraps around the nanopillar or the
nanopillar penetrates through the cell membrane to gain access
to the cell interior. With sharp, high aspect ratio structures, it
was first suggested4,9,11−14 that the membrane may be ruptured
at the site of the nanopillar. This view is supported by the
observation that nanopillars and similar structures facilitate
efficient delivery of a variety of biomolecules into mammalian
cells.9,14,15 It is also in line with the stronger cell adhesion6 and
limited cell motility2 observed on nanopillar substrates. On the
other hand, long-term survival of cells cultured on nanopillars
without apparent loss of cytosolic material indicates either that
the cell membrane remains intact around nanopillars or that the
cell membrane seals tightly around the base of nanopillars after
nanopillar penetration. Our recent electrical measurements by
nanopillar electrodes bear the signature of extracellular
recording before electroporation,10 suggesting that the cell
membrane remains intact around nanopillars. However, to add
to the debate, there have been reports of both intracellular and
extracellular recordings.3
In order to address the membrane conformation around the
entire structure, one requires a technique that has nanometer
resolution in both the radial and axial directions, relative to the
ertically aligned nanopillars and nanowires have demonstrated unique and at times surprising advantages over
planar or microstructured surfaces for biological studies. Cells
of many types, including primary animal cells such as mouse
embryonic stem cells1 and hippocampal neurons,2−4 have been
successfully cultured on nanopillar substrates, even those with
aspect ratios as high as 70.1 Cells were reported to retain
normal viabilities and proliferation.1−3,5 In conjunction with
those normal behaviors, many unusual behaviors have also been
observed. To begin with, cells have a particularly high affinity
for nanopillars. This leads to increased adhesion strength6 and
decreased mobility,2 as well as higher retention of cells from
suspension,7,8 than is observed on smooth planar surfaces.
Furthermore, the improved coupling has not been limited just
to cell attachment but has also improved communication with
the cell interior. Nanopillars have facilitated delivery of
biomolecules into cells at extraordinarily high efficiencies9
and recorded unexpectedly large electrical signals from
cardiomyocytes10 and neurons.3
In order to understand these phenomena and take full
advantage of the possibilities of nanopillar/nanowire devices, it
is imperative to understand the nature of the interface formed
between nanopillar structures and cells cultured on them. In the
case of vertically aligned nanopillars or nanowires, one of the
most important issues to address is how the cell membrane
interacts with the nanopillar surfacein particular, whether the
© 2012 American Chemical Society
Received: August 25, 2012
Revised: September 25, 2012
Published: October 3, 2012
5815
dx.doi.org/10.1021/nl303163y | Nano Lett. 2012, 12, 5815−5820
Nano Letters
Letter
Since one of the most promising applications for nanostructured devices is to the field of neurobiology,19,20 we
focused mainly on neurons for TEM studies. Cortical neurons
were dissected from E15 embryonic rats and plated on the
nanopillar substrate (see Supporting Information for more
details). In agreement with previous reports,2,3 cortical neurons
survived well on the nanopillar substrate, with similar survival
rate as those plated on a flat substrate. In fluorescence
microscopy, we observed that the neurons show similar
morphologies on both patterned and unpatterned areas (Figure
1b). The nanopillars can be seen as black dots embedded in the
cell bodies (Figure 1c). The neurons also grew out long
neurites after a few days in culture, as expected in dissociated
cultures. The soma and neurite morphologies were confirmed
by SEM (Figure 1d). We also imaged the cell membrane in
confocal microscopy (Supporting Information Figure 1), which
showed bright rings of membrane around each nanopillar.
However, as mentioned previously, fluorescence microscopy
does not provide enough resolution to address the question of
the membrane around the entire nanopillar, so we proceeded to
examine the membrane in TEM.
After 2−3 days in culture, the cells were prepared for TEM
imaging by fixing in 2% glutaraldehyde and 4% paraformaldehyde, postfixing in reduced osmium, and staining with uranyl
acetate (see Supporting Information for more details). The
sample was then dehydrated in a graded ethanol series and
infiltrated with Embed 812 resin. After the cells were fixed and
embedded in the resin, the quartz substrate was removed by
etching with 49% hydrofluoric acid. The exposed side, including
the space formerly occupied by the quartz nanopillars, was
refilled with fresh resin to provide support during the
sectioning step. When all of the resin had polymerized, 65
nm sections were taken on an ultramicrotome and deposited on
a 1 × 2 mm copper slot grid. A schematic of the whole process
is shown in Figure 2a. The grid was poststained with uranyl
acetate and lead citrate to enhance the contrast.
A zoomed-out TEM image in Figure 2b shows a vertical
cross section of a cell body on the nanopillar arrays. We then
zoomed in to the nanopillar areas to examine the cell
membrane at those locations. The cell membrane was more
darkly stained than the cytosol area, so it was clearly visible in
the TEM images of cross sections through the cell−nanopillar
interface (Figure 2c). The contrast to the membrane is not
entirely uniform because the TEM image is a projection
through a 65 nm thick section of the cell. If the membrane is
perpendicular to the plane of the section, that will result in a
thick line and maximal contrast, but if it passes through that
plane at any other angle, the result will be a blurred, less defined
line and lower contrast (yellow circled areas in Figure 2c).
Nonetheless, the membrane is visible around the entire
nanopillar. In agreement with previously reported SEM of
cells cultured on nanopillar structures,2 the nanopillar is fully
embedded in the cell, so the cell membrane reaches the flat
surface on either side of the nanopillar. As is the case on a flat
surface, the membrane is not a constant distance from the
surface but undulates between points of closer contact. It is
often closest at the corner of the nanopillar tip and gaps away at
the junction between the nanopillar and the flat surface. In spite
of the significant deformation imposed on the membrane by
this configuration, it is clear that the membrane conforms to the
shape of the nanopillar and wraps around the entirety of the tip.
Since our fabrication process allows for independent control
of the spacing, diameter, and height of the nanopillars, we were
nanopillar. The cell membrane is less than 5 nm thick, while the
nanopillars it interacts with are between 50 and 500 nm in
diameter and 1−5 μm tall.2,3,9,10,16 By comparison, optical
microscopy is constrained by the diffraction limit to resolutions
of 200−300 nm in x and y and 500−800 nm in the z
direction.17 Recent progress in super-resolution microscopy has
allowed a typical resolution18 of 20 nm in x and y and 50 nm in
z, which is still not sufficient for the required resolution. This
leaves us with electron microscopy for the requisite resolution.
Since scanning electron microscopy (SEM) does not achieve
sufficient contrast in biological tissues, we employed transmission electron microscopy (TEM) to provide the subnanometer resolution and contrast necessary for direct observation
of the cell membrane around the nanopillar. In the following
studies, we carefully examined the nanopillar−cell interface by
TEM. Our results indicate a continuous membrane around the
entirety of nanopillars with diameters 50−500 nm and heights
0.5−2 μm and that the nanopillar plays an important role in
determining the extracellular gap between the cell and the
substrate surface.
We first fabricated vertical nanopillars on a quartz coverslip.
Electron beam lithography followed by metal evaporation was
used to define arrays of circular metal dots, between 50 and 500
nm in diameter, which served as protection masks for the
following etching step. Then, the nanopillars were formed by
anisotropic reactive ion etching. Finally, the metal mask was
removed with metal chemical etchant (see Supporting
Information for a detailed description of the fabrication
procedures). The diameter and spacing of the nanopillars
were determined by the metal mask pattern. The duration of
the reactive ion etching step determined the height of the
pillars, which ranged between 0.5 and 2 μm. An SEM image of
a finished substrate is displayed in Figure 1a, showing an array
of vertically aligned SiO2 nanopillars 200 nm in diameter and
800 nm in height. Before cell plating, the nanopillar substrate
was cleaned and sterilized in oxygen plasma and then coated
with 0.2 mg/mL poly(L-lysine) overnight.
Figure 1. (a) An SEM image of SiO2 nanopillar array. (b)
Fluorescence images of cortical neurons cultured on large squares of
SiO2 nanopillar arrays, which are outlined in white. Cells have been
loaded with calcein-AM for visualization. The dark rectangles in the
upper left corner of each array are alignment marks. (c) A zoomed-in
fluorescent image showing more than 10 nanopillars, which appear as
black dots, interfacing with the cell body of a cortical neuron. (d) An
SEM image of a cortical neuron cultured on a SiO2 nanopillar array,
showing the cell body and the extension of several neurites.
5816
dx.doi.org/10.1021/nl303163y | Nano Lett. 2012, 12, 5815−5820
Nano Letters
Letter
Figure 2. The cell membrane is visible wrapping around the nanopillar. (a) Schematic of the preparation of the nanopillar−cell interface for TEM
imaging. (b) Low-magnification TEM image showing a cell body engulfing four nanopillars. Black arrows indicate the first and last nanopillars. (c)
Three high-magnification TEM images of vertical cross sections of the cell−nanopillar interface. Yellow circles surround points where the membrane
is not perpendicular to the section and is less contrasted. The same TEM images are shown below with the nanopillar surface (in blue) and cell
membrane (in red) outlined for visual guidance.
Figure 3. The interface between the cell and nanopillar substrate depends on the diameter of the nanopillars and the spacing between them. (a)
Nanopillars with a diameter of 500 nm spaced 1 μm center-to-center cause the cell to rest atop the nanopillar array. From left to right: An SEM
image of the array of nanopillars, a low-magnification TEM image of a cell body growing on top of the 500 nm nanopillars, and a high-magnification
TEM image showing the cell membrane suspended between two 500 nm nanopillars. (b) With the same interpillar distance as in (a), cells engulf
nanopillars with diameters of less than 200 nm. From left to right: An SEM image of the array of nanopillars of 200 nm diameter and 1 μm spacing, a
low-magnification TEM image of a cell body engulfing the nanopillars, and a high-magnification TEM image of the cell membrane wrapping around
two adjacent nanopillars.
2 μm in height, and from 1 to 3 μm center-to-center spacing.
To begin with, we found that how cells interface with
nanopillars of 1 μm height strongly depends on the ratio of
able to probe whether and how the cell−nanopillar interface
depends on these parameters. We examined how cells interact
with nanopillars ranging from 100 to 500 nm in diameter, 0.5−
5817
dx.doi.org/10.1021/nl303163y | Nano Lett. 2012, 12, 5815−5820
Nano Letters
Letter
the nanopillar diameter to the spacing. Closely spaced
nanopillars of large (>300 nm) diameter are not engulfed by
the cellrather, the cell rests atop the nanopillar array. This
behavior is exhibited in Figure 3a, where a soma is suspended
between posts of 500 nm diameter. All of the nanopillars in this
area were the same sizethe appearance of diminishing size
from right to left is due to the fact that the plane of sectioning
was not perfectly parallel to the row of nanopillars. When the
spacing between nanopillars is kept at a constant of 1 μm, the
transition between suspension and engulfment can be observed
as the diameter decreases from 500 to 200 nm. The cell body
engulfs most of the 200 nm diameter nanopillars, even with
only 800 nm space in between (Figure 3b). With more widely
spaced nanopillars, the cell engulfs larger nanopillars, even
those up to 500 nm in diameter. We also observed that cells
engulf nanopillars down to 50 nm in diameter and up to 2 μm
in height with similar behaviors to those seen on shorter
nanopillars (see Supporting Information Figures 1 and 2). The
knowledge of these parameters allows the design of future
nanopillar devices to include control over cell engulfment,
which can drastically alter the functionality of these devices.
In addition to the cell bodies, we also examined how neurites
interact with nanopillar structures. Neurons have unique shapes
in that they grow out long and thin neurites >1 mm in length
but only about 1 μm in diameter. In order to observe the
interactions between neurites and nanopillars more thoroughly,
we took both vertical cross sections, perpendicular to the plane
of the coverslip, and horizontal cross sections, parallel to the
plane of the coverslip. The vertical cross sections were prepared
as previously described. For the horizontal cross sections, we
sectioned parallel to the face of the coverslip immediately after
the coverslip was removed by hydrofluoric acid, rather than
refilling the vacated space with resin. Thus, the nanopillars
appear as empty holes that appear white in TEM.
The interaction between neurites and nanopillars depends on
the size of the neurites, and it differs from that between cell
body and nanopillars. We found that thin neurites, of less than
1 μm in diameter, do not deform to a large extent around the
nanopillar. They generally held their cylindrical shapes and
grew between the nanopillars, attaching tightly to one side of
each nanopillar (Figure 4a). From sections in both horizontal
(Figure 4b) and vertical planes (Figure 4a), we were able to
observe the microtubules in the small neurites and see that they
are not significantly deformed around the nanopillars. On the
other hand, larger neurites often fully engulf the nanopillars
(Figure 4c,d). This engulfment differs from that observed in the
cell body. For the cell body, the cell membrane is usually closer
to the top of the nanopillar than to the bottom of the
nanopillar. For large neurites, a tight interaction is maintained
around the bottom of the nanopillar (Figure 4e), while we
sometimes observed the membrane curving away on one side of
the nanopillar to form a cavity around the top of the nanopillar
(Figure 4f). This phenomenon is likely due to the combined
effect of the increased cytoskeletal rigidity and the different
membrane composition in the neurites.21
To elaborate on our observations regarding the effect of the
nanopillars on the cell−substrate interface, we quantitatively
measured the gap distance between the cell membrane and the
substrate surface on both flat and nanopillar areas. We further
classified the interaction according to different locations on the
nanopillar (Figure 5a). The average gap distances for each
location are summarized in Figure 5b. All of these studies were
done on nanopillars with diameters of 200 nm, heights of 1 μm,
Figure 4. Neurites interacting with nanopillars. (a, b) Small neurites
do not engulf nanopillars but keep their cylindrical shape and interact
with the side or top of the nanopillar. Vertical cross section is shown in
(a) and horizontal cross section in (b). (c, d) Vertical cross sections of
larger neurites engulfing nanopillars. Red arrows mark cavities formed
on the upper half of the nanopillar, while blue arrows point out where
the neurite seals tightly at the bottom of the nanopillar. (e) Horizontal
cross section near the bottom of a nanopillar, where a large neurite
engulfs the nanopillar and seals tightly. (f) Horizontal cross section
near the top of the nanopillar, where the engulfing neurite leaves a
larger space around the top. A diagram indicating placement of cross
sections in (e) and (f) is shown to the left. All scale bars are 500 nm in
length.
and spacing 1 μm. On a flat area, neurons are not very
adherent, and the cell bodies have a space of about 50 nm
between the cell membrane and the solid surface.22 We also
measured an average gap of 50 nm on a completely flat
substrate, which confirms the accurate preservation of this
ultrastructural feature through the lengthy TEM preparation. In
contrast, at the top and the top side of the nanopillar there was
on average a gap of less than 15 nm, with many areas having
less than 10 nm between the membrane and nanopillar surface.
Along the bottom side of the nanopillar, the gap distance
averages 40 nm, intermediate between the tip of the nanopillar
and the flat surface. The difference consists mostly of a decrease
in the variability of the gap distance as shown in Figure 5c. On
the flat surface and the bottom half of the nanopillar sidewall,
there are many places where the gap between the membrane
and surface is larger than 100 nm, while around the tip of the
nanopillar the gap is consistently less than 50 nm.
The neurites showed tighter attachment to the flat area
between nanopillars than a completely flat substrate (25 nm vs
45 nm), while the cell body show similar gap distance in the
two locations (50 nm vs 50 nm). This tight interaction is
maintained along the bottom half of the nanopillar, where
larger gaps are much rarer than on a flat substrate (Figure 5d).
The membrane−surface gap in neurites is also larger along the
top half of the nanopillar than it is in cell bodies (27 nm vs 13
5818
dx.doi.org/10.1021/nl303163y | Nano Lett. 2012, 12, 5815−5820
Nano Letters
Letter
nm), a reflection of the tendency of neurites to form cavities
around the top of the nanopillar. However, the tightest
interaction in neurites is still at the very tip of the nanopillar,
where the gap averages only 18 nm.
The effect of surface structure on the membrane−substrate
gap also provides quantitative insight into the use of
nanostructures as electrodes to measure action potentials in
previous studies3,10 and in the future. Given the membrane−
surface distance measured here, the seal resistance between the
tip of a nanopillar engulfed by a cell body and the extracellular
medium can be estimated to be around 80 MΩ, or at least 10
times the seal resistance that would be achieved with a planar
electrode23 (see Supporting Information for details of the
calculation). This is in agreement with previous reports that
nanopillar electrodes were able to record comparable
extracellular signal to those achieved with commercial multielectrode arrays, in spite of the higher electrode resistance that
results from the smaller surface area of the nanopillar
electrodes. For neurites that engulf nanopillars, we calculated
the average leak resistance to be ∼60 MΩ, over 15 times
greater than the sealing with a planar electrode as derived from
the gap distances. Therefore, nanopillars offer promising
prospects for electrophysiology measurements of axons,
which are difficult to access by traditional methods like patch
clamp or multielectrode arrays.
In summary, vertical nanopillars alter the interface between
cells and inorganic substrates, which may account for many of
the previously reported advantages of nanopillars for biological
studies. By imaging the interface with TEM, we confirmed that
neurons maintain the integrity of the cell membrane around
nanopillars with diameters 50−500 nm and heights 0.5−2 μm.
The cell membrane appears to interact more tightly with the
nanopillar substrate than it typically does with a flat surface. In
addition to accounting for the good sealing of nanopillar
electrodes,10 the close interaction provides insight into the
origin of stronger adhesion6 and lower mobility2 by cells on
nanopillar devices and the ability of nanowire surfaces to
capture cells from suspension.7,8 On top of explaining previous
studies, knowledge of the cell−nanopillar interface can be used
to guide the design of future investigations and applications of
this versatile platform. One remaining mystery is how the
nanopillars facilitate delivery of biomolecules into cells9 if not
by direct penetration. It is likely that the interaction between
the cell membrane and the nanopillar is dependent on the
chemical nature of the nanopillar surface. In this study, the
nanopillar surface was unmodified, while some other studies
have nanopillars modified with molecules that facilitate cell
penetration.24,25 It is also possible that the cell−nanopillar
interaction is time-dependent, and there is some transient
penetration shortly after plating which is later resealed. This
would be consistent with both the successful delivery shortly
after plating9 and the continuous membrane we observed 2−3
days later. The surface dependence and the time dependence of
cell−nanopillar interactions certainly warrant further investigation.
Figure 5. The gap distance between the cell membrane and different
locations on the nanopillars. (a) Diagram of the classification of
different locations: the flat area near a nanopillar, the bottom half of
the nanopillar, the top half of the nanopillar, and the top of the
nanopillar. These locations are compared with a plain flat substrate in
the following analyses. (b) Average gap distance at different locations.
Averages for the soma are shown in blue and for neurites are shown in
red. Error bars show a 95% confidence interval for the average and in
some cases are smaller than the marker. The membrane appear to
attach more tightly to the top of the nanopillar than to the bottom side
of the nanopillar or on a completely flat substrate. (c) The distribution
of gap distances for the cell body at the top of the nanopillar, the
bottom side of the nanopillar, and on a completely flat substrate. (d)
The distribution of gap distances for the neurite at the top of the
nanopillar, the bottom side of the nanopillar, and on a completely flat
substrate.
■
ASSOCIATED CONTENT
S Supporting Information
*
Confocal microscopy of the cell membrane, TEM images of
additional nanopillars, and details of seal resistance calculation;
additional fabrication and experimental details. This material is
available free of charge via the Internet at http://pubs.acs.org.
5819
dx.doi.org/10.1021/nl303163y | Nano Lett. 2012, 12, 5815−5820
Nano Letters
■
Letter
(23) Buitenweg, J. R.; Rutten, W. L.; Willems, W. P.; van
Nieuwkasteele, J. W. Med. Biol. Eng. Comput. 1998, 36, 630−637.
(24) Almquist, B. D.; Melosh, N. A. Proc. Natl. Acad. Sci. U. S. A.
2010, 107, 5815−5820.
(25) Hai, A.; Shappir, J.; Spira, M. E. Nat. Methods 2010, 7, 200−2.
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected] (B.C.); [email protected] (Y.C.).
■
Present Address
⊥
Department of Chemistry and Chemical Biology, Harvard
University, Cambridge, MA 02138.
NOTE ADDED AFTER ASAP PUBLICATION
Figure 1 and Figure 4 were duplicated in the paper published
ASAP on October 8, 2012. Figure 4 was replaced and the
revised version was re-re-posted on October 22, 2012.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
This work was supported by the NSF (CAREER award no.
1055112), the NIH (grant no. NS057906), a Searle Scholar
award, a Packard Science and Engineering Fellowship (to B.C.),
and a National Defense Science and Engineering Graduate
Fellowship (to Z.L.).
■
REFERENCES
(1) Kim, W.; Ng, J. K.; Kunitake, M. E.; Conklin, B. R.; Yang, P. J.
Am. Chem. Soc. 2007, 129, 7228−7229.
(2) Xie, C.; Hanson, L.; Xie, W.; Lin, Z.; Cui, B.; Cui, Y. Nano Lett.
2010, 10, 4020−4024.
(3) Robinson, J. T.; Jorgolli, M.; Shalek, A. K.; Yoon, M.-H.; Gertner,
R. S.; Park, H. Nat. Nanotechnol. 2012, 7, 180−184.
(4) Hällström, W.; Mårtensson, T.; Prinz, C.; Gustavsson, P.;
Montelius, L.; Samuelson, L.; Kanje, M. Nano Lett. 2007, 7, 2960−
2965.
(5) Berthing, T.; Bonde, S.; Sørensen, C. B.; Utko, P.; Nygård, J.;
Martinez, K. L. Small 2011, 7, 640−647.
(6) Qi, S.; Yi, C.; Ji, S.; Fong, C.-C.; Yang, M. ACS Appl. Mater.
Interfaces 2009, 1, 30−34.
(7) Wang, S.; Wang, H.; Jiao, J.; Chen, K.-J.; Owens, G. E.; Kamei,
K.; Sun, J.; Sherman, D. J.; Behrenbruch, C. P.; Wu, H.; Tseng, H.-R.
Angew. Chem., Int. Ed. 2009, 48, 8970−8973.
(8) Wang, S.; Liu, K.; Liu, J.; Yu, Z. T.-F.; Xu, X.; Zhao, L.; Lee, T.;
Lee, E. K.; Reiss, J.; Lee, Y.-K.; Chung, L. W. K.; Huang, J.; Rettig, M.;
Seligson, D.; Duraiswamy, K. N.; Shen, C. K.-F.; Tseng, H.-R. Angew.
Chem., Int. Ed. 2011, 50, 3084−3088.
(9) Shalek, A. K.; Robinson, J. T.; Karp, E. S.; Lee, J. S.; Ahn, D.-R.;
Yoon, M.-H.; Sutton, A.; Jorgolli, M.; Gertner, R. S.; Gujral, T. S.;
MacBeath, G.; Yang, E. G.; Park, H. Proc. Natl. Acad. Sci. U. S. A. 2010,
107, 1870−1875.
(10) Xie, C.; Lin, Z.; Hanson, L.; Cui, Y.; Cui, B. Nat. Nanotechnol.
2012, 7, 185−190.
(11) Friedmann, A.; Hoess, A.; Cismak, A.; Heilmann, A. Acta
Biomater. 2011, 7, 2499−2507.
(12) McKnight, T. E.; Melechko, A. V.; Hensley, D. K.; Mann, D. G.
J.; Griffin, G. D.; Simpson, M. L. Nano Lett. 2004, 4, 1213−1219.
(13) McKnight, T. E.; Melechko, A. V.; Griffin, G. D.; Guillorn, M.
A.; Merkulov, V. I.; Serna, F.; Hensley, D. K.; Doktycz, M. J.; Lowndes,
D. H.; Simpson, M. L. Nanotechnology 2003, 14, 551−556.
(14) Vandersarl, J. J.; Xu, A. M.; Melosh, N. A. Nano Lett. 2012, 12,
3881−3886.
(15) Mann, D. G. J.; McKnight, T. E.; McPherson, J. T.; Hoyt, P. R.;
Melechko, A. V.; Simpson, M. L.; Sayler, G. S. ACS Nano 2008, 2, 69−
76.
(16) Xie, C.; Hanson, L.; Cui, Y.; Cui, B. Proc. Natl. Acad. Sci. U. S. A.
2011, 108, 3894−3899.
(17) Zhuang, X. Nat. Photonics 2009, 3, 365−367.
(18) Huang, B.; Wang, W.; Bates, M.; Zhuang, X. Science 2008, 319,
810−813.
(19) Silva, G. A. Nat. Rev. Neurosci. 2006, 7, 65−74.
(20) Pancrazio, J. J. Nanomedicine 2008, 3, 823−830.
(21) Calderon, R. O.; Attema, B.; DeVries, G. H. J. Neurochem. 1995,
64, 424−9.
(22) Gleixner, R.; Fromherz, P. Biophys. J. 2006, 90, 2600−2611.
5820
dx.doi.org/10.1021/nl303163y | Nano Lett. 2012, 12, 5815−5820