NANO
LETTERS
Nanoscale Operation of a Living Cell
Using an Atomic Force Microscope with
a Nanoneedle
xxxx
Vol. 4 , No. 13
A-D
Ikuo Obataya,† Chikashi Nakamura,*,†,‡ SungWoong Han,‡
Noriyuki Nakamura,†,‡ and Jun Miyake†,‡
Research Institute for Cell Engineering (RICE),
National Institute of AdVanced Industrial Science and Technology (AIST),
3-11-46 Nakoji, Amagasaki, Hyogo 661-0974, Japan, and
DiVision of Biotechnology and Life Science, Tokyo UniVersity of Agriculture and
Technology (TUAT), 2-24-16 Naka-cho, Koganei, Tokyo 184-8588 Japan
Received September 7, 2004; Revised Manuscript Received November 11, 2004
ABSTRACT
We have developed a tool for performing surgical operations on living cells at nanoscale resolution using atomic force microscopy (AFM) and
a modified AFM tip. The AFM tips are sharpened to ultrathin needles of 200−300 nm in diameter using focused ion beam etching. Force−
distance curves obtained by AFM using the needles indicated that the needles penetrated the cell membrane following indentation to a depth
of 1−2 µm. The force increase during the indentation process was found to be consistent with application of the Hertz model. A threedimensional image generated by laser scanning confocal microscopy directly revealed that the needle penetrated both the cellular and nuclear
membranes to reach the nucleus. This technique enables the extended application of AFM to analyses and surgery of living cells.
There is great demand for single-cell manipulation in current
biotechnology because it is significantly important to investigate where and when molecules display their functions in
controlling the activity of cells. If the molecules of interest
such as nucleic acids, proteins and chemicals can be
transferred to the inside of a living cell at a known time, we
can modulate the cell activity while monitoring the reaction
of the cell in real-time. Such techniques will be applied not
only for investigation of cell activity but also in controlled
differentiation or therapy of living cells.
Microinjection of proteins, peptides, and genetic materials
into a living cell is widely practiced using microcapillaries.1,2
The technique plays an important role in the transplantation
of nuclei or introduction of nucleic acids for tracking the
protein expression. However, fatal damage from using
microcapillaries due to the shape of the capillaries and
inaccuracy of the displacement is often problematic for
manipulation of small cells. To overcome these problems,
fundamental changes in controlling devices for displacement
of inserting materials and optimization of the size and shape
of the inserting materials are necessary.
Recently, atomic force microscopy (AFM) has been used
for imaging cell surfaces because the apparatus allows
* Corresponding author. Phone: +81-6-6494-7858, Fax: +81-6-64947862, E-mail: chikashi-nakamura@aist.go.jp
† RICE/AIST.
‡ TUAT.
10.1021/nl0485399 CCC: $30.25
Published on Web 00/00/0000
© xxxx American Chemical Society
researchers to observe living cells in physiological conditions.3,4 Besides surface observation, the AFM probe can act
against a cell surface. For example, the elastic5-7 or viscoelastic8 properties of the cell surface have been estimated
for various cells by contacting and indenting the surface with
an AFM probe. Since the surface of an AFM probe can be
chemically modified, probes have been used for the chemical
modification of the surfaces of bacteria.9 Moreover, it is
considered that an atomic force microscope is suitable as a
controlling device for cell manipulation in order to access
the interior of living cells. Investigation of the localization
of mRNA by extraction of mRNA with an AFM tip from
living cells has been reported.10 The displacement of the
cantilever can be controlled at nanoscale resolution by piezo
devices while monitoring the exerting force on the cantilever.
Since the force should reflect the situation of cantilever
(contacting, indenting, or penetrating), the position of tip is
predictable. This feature is essentially superior to the
conventional micromanipulation techniques with respect to
precise displacement and monitoring. To insert a solid
material, the size and shape of the tip should be improved
for certainty of insertion and minimization of damage to a
living cell. Therefore, we have attempted to develop a system
and materials for a single-cell operation using AFM and have
fabricated AFM probes in the shape of an ultrathin needle,
called a nanoneedle.11,12 In this paper, we present the
PAGE EST: 3.9
Figure 1. (A) Schematic representation of a nanoneedle on the
AFM tip over a living cell. (B) Electron microscopic image of
nanoneedle with a scale bar of 3 µm.
mechanical response during insertion of a nanoneedle and
direct observation of a living cell and the nanoneedle inserted
by the system.
We developed a cell manipulation apparatus using a
molecular force probe (MFP-1D, Asylum Research, Santa
Barbara) as a force detecting instrument. The MFP is placed
on the stage of a confocal fluorescence microscope FV-300/
IX71 (Olympus, Tokyo). This combination allows us to
obtain phase contrast, fluorescent and confocal fluorescent
images of cells, and a cantilever simultaneously. Ultrathin
needles were prepared by focused ion beam (FIB) etching
from pyramidal Si AFM tips (Nanosensors, Neuchatel) that
have spring constants from 0.1 to 0.2 N/m. There are
commercially available AFM tips with high aspect ratio,
however, they are not long enough to reach inside the cell.
Since the heights of living cells are more than 3 µm, the
length of the tip body must be more than this height,
otherwise the base of the tip bumps the cell surface and
disturbs the force analyses. Furthermore, the tapered shapes
of such etched tips are not favorable with respect to the
probability of penetration and indentation depth before
penetration as we reported previously.11 To fabricate the
ultrathin needle, the FIB was sequentially focused on two
areas of the pyramid tip so as to leave a thin area on the tip.
Then, the resulting triangular plate on the tip was etched in
the same manner after rotating the cantilever perpendicular
to the original direction. The needle was tilted about 11°
with respect to the center axis of the tip in order to
compensate for the mounting angle of its holder (Figure 1A).
The tips were etched into needles of 200-300 nm in diameter
and 6-8 µm in length (Figure 1B).
Force measurements were conducted on a human epidermal melanocyte (NHEM, Kurabo, Osaka) in a medium at
37 °C under 5% CO2 atmosphere. The tip was lowered over
the center of the cell. Figure 2A shows a typical forcedistance curve during indenting and pulling of a normal
pyramidal AFM tip. The force curve shows a continuous
increase as the cantilever was lowered, implying that the tip
indented the cell surface without penetration. It has been
reported that an indenting force of no more than 10 nN is
required when using a normal pyramidal AFM tip in order
to penetrate the cell surface.13-15 But such a large indentation
depth relative to the height of the cell may cause serious
damage. Even if the cell could survive after indentation, the
large deformation may cause undesirable mechanotransducB
Figure 2. Force-distance curves during approach and retraction
from a melanocyte using a normal pyramidal AFM tip (A) and the
nanoneedle (B). The cantilever approached (blue arrow) and then
retracted (red arrow).
tions.16 On the other hand, the force curve using a nanoneedle
was quite different from that using a pyramidal tip (Figure
2B). The force increased until about 1 µm in the approaching
process, but the force did not show a continuous increase as
seen in the plot with a pyramidal tip. This force dropping
during the approach process was considered to be the point
of the penetration of the needle tip into the cell. Such force
dropping or relaxation has been reported when erythrocytes
were indented to penetrate the membrane using a pyramidal
tip.13 However, both the force and the indentation depth for
penetration were quite small when using the nanoneedle. It
is important to maintain the condition or the activity of the
cell for a living cell operation; therefore, in these respects
we believe that the nanoneedle is superior to a normal
pyramidal tip for penetration into a cell. After the indentation,
the force fluctuated by 1 to 2 nN, suggesting that the needle
was inserted through the cell membrane. The force might
contain friction force and viscous resistance between the
surface of the needle and the cell constituents. After this
inserting section, the force increased continuously producing
a steeper slope, suggesting that the base of needle made
contact with the cell surface over a larger area.
Since the first increase of the force curve was expected to
be due to indentation of the cell surface, we analyzed the
force curve by applying the Hertz model that describes the
indentation of homogeneous elastic material by a stiff
material with a defined geometry.6,17 In fact, living cells are
not ideal samples for applying the model because of their
anisotropy and heterogeneity of the surface; however, it
should be worth rationalizing the force curve and comparing
with reported values estimated by applying this model. We
prepared two types of nanoneedles, varying the shapes of
their tip (flat-ended cylinder and cone shape), for comparison.
Figure 3 shows the typical force curves when using the
needles. For the needle with a cone-shaped tip, the force
curve showed a quadratic force increase after contacting the
cell surface. Then the force increased with a linear function.
The point of this functional change in the force increase
corresponded to the point of the geometrical change from a
cone to a cylinder. On the other hand, the force increase
showed only a linear function with the cylindrical needle
Nano Lett.
Figure 3. Typical force-distance curves for a melanocyte using
nanoneedles that have cone-like shapes (A) and cylindrical shapes
(B) at the edges. Insets show magnified SEM images around the
tip of the needle. Proposed situations of the nanoneedle tip and the
cell surface are depicted under the plots.
(Figure 3B). The Hertz model mechanics describes force F
as a function of indentation depth I with a defined shape of
the indenter with constant values co and cy.
F ) coI2 (for cone-shaped indenter)
F ) cyI (for cylindrical indenter)
The force curves using the two types of the tip were
reasonable with respect to the tip geometry by applying this
mechanics. Estimated elastic moduli were 2 to 10 kPa for
melanocytes in the experiments. The values were comparable
to the elasticity for HUVEC cells (7.2 kPa)18 and fibroblast
cells (3-5 kPa)19 using the same parameters for estimation,
suggesting that the first force increase after contacting with
the nanoneedles was also explained by the Hertz model.
Furthermore, laser scanning confocal microscopy (LSM)
was employed to observe directly the cell and the inserted
needle. We used a nanoneedle fabricated from another type
of AFM probe. Since the probe has a tetrahedral tip that
protrudes from the very end of the cantilever, the needle can
be aimed precisely at the living cell of interest. The needles
were modified with 3-(aminopropyl)triethoxysilane followed
by fluorescein isothiocyanate (FITC). Human embryonic
kidney (HEK293) cells expressing red-fluorescent protein
DsRed2-NES were used in the experiments in order to
distinguish the FITC-needle from cell body by LSM. The
DsRed2-NES has a nuclear exporting signal (NES) at the
C-terminus of DsRed2. The FITC nanoneedle and the cells
were simultaneously excited at 488 nm by an Ar laser, and
the emission signal was divided by a dichroic mirror into
green (510-530 nm) and red (565-660 nm) signals through
emission filters.
The needle was lowered toward the HEK293 cell by
controlling the piezo device while the force on the cantilever
was monitored. Since the heights of HEK293 cells were 1020 µm, the needle was lowered 8 µm after the force drop
was observed after indentation of the cell surface in order
to reach the needle tip into the nucleus. Figure 4A shows a
stack of all confocal images. Figure 4B shows vertical section
images at the positions indicated by the white line in Figure
4A. The cross-section indicated clearly that the nanoneedle
was inserted accurately in the nucleus. The piezo device for
displacement and the force detection used here were accurate
enough for needle insertion into the cell. When the force
curve without a force drop after indentation was observed,
the needle failed to insert into the cell body through the cell
membrane (Figure 4C). Therefore, the results suggested that
the force drop or force relaxation could be regarded as
indication of successful insertion.
In addition, the shape of the plasma membrane and nucleus
remained the same as before insertion. If the needle just
indented or dragged into the cell, the shape of red fluorescence of DsRed2-NES protein should be deformed. Therefore, the needle did not indent the membrane, but it
penetrated through the membrane. Meanwhile, the pyramidal
tip caused great deformation by indenting both the cell
Figure 4. (A) Stack of confocal slices of HEK293 cell expressing DsRed2-NES and FITC-labeled nanoneedle excited at 488 nm when the
nanoneedle was inserted. (B) Cross-section images for green and red emission processed from confocal slices. (C) Images are from when
the needles are failing to penetrate the cell surface. (D) Images when a pyramidal tip is indenting the cell surface. Scale bars in all images
indicate 10 µm.
Nano Lett.
C
did penetrate the cell and nucleus membranes vertically as
we expected. Intrinsic advantages of this system are the
capability of accurate displacement and sensing the force
on the needle, as does a surgeon’s finger during surgery. By
modifying the surface of a needle, it can load various
molecules such as nucleic acids, proteins or chemicals
through standard immobilization chemistries. The system
should be one of a number of versatile apparatus for cell
surgery and observation in the future.
Figure 5. A sequence of three-dimensional images reconstructed
from confocal slices of the HEK293 cell and inserted nanoneedle
with a schematic representation for each.
surface and the nucleus despite its high aspect ratio of 5:1
(Figure 4D). It has been reported that mechanical stress
causes various changes in cell activity.16,20 Minimal deformation of cells is essential for cell manipulation because
undesired mechanical responses may interfere with the result
of manipulation. The results suggested that the deformation
from using the nanoneedle was small because the distance
of indentation before penetration was less than 1 µm.
Figure 5 shows a sequence of schematic representations
and reconstructed three-dimensional fluorescent images by
processing data from all confocal slices with FLUOVIEW
software (Olympus). The signal intensity of cell body is
reduced for clarity purposes. To the best of our knowledge,
the results demonstrated for the first time that solid material
was inserted into a nucleus of such a small living cell with
highly accurate positioning.
The features of this technique are advantageous compared
to microcapillary-based techniques. First, the displacement
of the needle is accurate, therefore this technique is not
particular about the size of cells. Second, the position can
be estimated by monitoring the exerting force of the needle.
The microcapillary-based techniques have a limit for thinner
capillaries because of the limit of optical observation of the
capillary tip as well as difficulty in the vertical observation
of it. Third, the ultrathin needle does not cause fatal damage
to several types of living cells (Han et al., manuscript in
preparation). In contrast, capillary insertion caused 50% of
the plated cells to die in our preliminary experiments. Low
invasiveness of the inserting materials is an essential feature
for the operation in living cells.
In summary, we developed a system for cell operation by
using AFM and fabricated probes shaped as ultrathin needles.
The force analysis suggested that the nanoneedle indented
the cell’s surface, and then it penetrated through the cell
surface, accompanied by a drop in the exerting force on the
needle. LSM observation confirmed directly that the needle
D
Acknowledgment. We thank Industrial Technology Research Grant Program in ′03 from New Energy and Industrial
Technology Development Organization (NEDO) of Japan.
I.O. acknowledges Industrial Technology Fellowship Program from NEDO.
Supporting Information Available: Construction of
pDsRed2-NES plasmid. Description of the Hertz model.
SEM image of nanoneedle fabricated from ATEC-CONT.
Sequential reconstituted LSM images as a movie file. This
information is available free of charge via the Internet at
http://pubs.acs.org.
References
(1) Tsulaia, T. V.; Prokopishyn, N. L.; Yao, A.; Carsrud, N. D. V.; Carou,
M. C.; Brown, D. B.; Davis, B. R.; Yannariello-Brown, J. J. Biomed.
Sci. 2003, 10, 328-336.
(2) King, R. Methods Mol. Biol. 2004, 245, 167-174.
(3) Fotiadis, D.; Scheuring, S.; Muller, S. A.; Engel, A.; Muller, D. J.
Micron. 2002, 33, 385-397.
(4) Wojcikiewicz, E. P.; Zhang, X.; Moy, V. T. Biol. Proc. Online 2004,
6, 1-9.
(5) Schaer-Zammaretti, P.; Ubbink, J. Ultramicroscopy 2003, 97, 199208.
(6) Radmacher, M. IEEE Eng. Med. Biol. Mag. 1997, 16, 47-57.
(7) Rotsch, C.; Braet, F.; Wisse, E.; Radmacher, M. Cell Biol. Int. 1997,
21, 685-696.
(8) Haga, H.; Nagayama, M.; Kawabata, K.; Ito, E.; Ushiki, T.;
Sambongi, T. J. Electron. Microsc. (Tokyo) 2000, 49, 473-481.
(9) Firtel, M.; Henderson, G.; Sokolov, I. Ultramicroscopy 2004, in press.
(10) Uehara, H.; Osada, T.; Ikai, A. Ultramicroscopy 2004, 100, 197201.
(11) Obataya, I.; Nakamura, C.; Han, S.; Nakamura, N.; Miyake, J.
Biosens. Bioelectron. 2004, in press.
(12) Han, S.; Nakamura, C.; Obataya, I.; Nakamura, N.; Miyake, J.
Biosens. Bioelectron. 2004, in press.
(13) Hategan, A.; Law, R.; Kahn, S.; Discher, D. E. Biophys. J. 2003,
85, 2746-2759.
(14) Le Grimellec, C.; Lesniewska, E.; Giocondi, M. C.; Finot, E.; Vie,
V.; Goudonnet, J. P. Biophys. J. 1998, 75, 695-703.
(15) Takeuchi, M.; Miyamoto, H.; Sako, Y.; Komizu, H.; Kusumi, A.
Biophys. J. 1998, 74, 2171-2183.
(16) Charras, G. T.; Horton, M. A. Biophys. J. 2002, 82, 2970-2981.
(17) Weisenhorn, A. L.; Khorsandi, M.; Kasas, S.; Gotzos, V.; Butt,
H.-J. Nanotechnology 1993, 4, 106-113.
(18) Mathur, A. B.; Truskey, G. A.; Reichert, W. M. Crit. ReV. Biomed.
Eng. 2000, 28, 197-202.
(19) Rotsch, C.; Jacobson, K.; Radmacher, M. Proc. Natl. Acad. Sci. U.S.A.
1999, 96, 921-926.
(20) Hochmuth, R. M. J. Biomech. 2000, 33, 15-22.
NL0485399
PAGE EST: 3.9
Nano Lett.