USING MULTI-FREQUENCY ELECTRICAL IMPEDANCE
SPECTROSCOPY TO MONITOR SINGLE BUDDING YEAST CELLS
Zhen Zhu*, Olivier Frey, Felix Franke, Niels Haandbæk and Andreas Hierlemann
ETH Zürich, SWITZERLAND
ABSTRACT
This work presents a microfluidic device that enables reliable immobilization and multi-frequency
electrical impedance spectroscopy (EIS, 10 kHz ‒ 10 MHz) of single budding-yeast cells, S. cerevisiae.
Immobilized cells with different shapes in 4 different orientations can be discerned by using multifrequency EIS data. The budding process can be continuously monitored in real-time through EIS and
distinguished from potential movements of immobilized cells by extracting representative vectors from the
multi-frequency EIS data.
KEYWORDS: Single-cell analysis, Electrical impedance spectroscopy, Cell trapping, Cell culturing
INTRODUCTION
Electrical impedance spectroscopy (EIS) enables the frequency-dependent multi-parametric readout of
cellular information. By using microfabrication techniques, EIS can be integrated in microfluidic devices
to detect single cells. Most of those devices, called electrical impedance cytometers, were used to
characterize suspended biological samples in a flow-through setup [1, 2]. However, EIS has not yet been
used for real-time monitoring of cellular dynamics, e.g., cell growth, at single-cell resolution.
The concept of single-cell immobilization and EIS has been demonstrated previously [3‒5]. In this
work, the device features capture and cultivation of single cells while performing real-time EIS of
immobilized budding yeast cells. Different orientations and different shapes of immobilized cells could be
discerned by using multi-frequency EIS data. Moreover, we have been able to continuously monitor the
budding process, and to detect potential movements of immobilized single yeast cells during measurements
by extracting representative vectors from the multi-frequency EIS data.
EXPERIMENTAL
The device, as schematically shown in Figure 1, included a glass substrate with patterned electrodes to
perform localized impedance measurements, a SiNx insulation layer to reduce electric crosstalk between
adjacent electrodes, and microchannels structured in SU-8 (30 µm high) and sealed with a PDMS cover for
optical access. Application of a slightly lower pressure to the suction channel enabled single-cell capturing
at the cell traps (4 µm wide) along the cell-culturing channel. EIS was performed by applying an AC signal
to the common stimulus electrode and by recording the resulting signal on the individual recording
electrodes at the traps. The orifices of the traps prevented cells from passing into the suction channel and
constrained the electric current to flow through the narrow opening, thereby improving the sensitivity of
EIS. After each measurement, cells were released, and EIS was performed for the empty trap, which was
then used to calculate relative magnitude 𝐴𝑟 and relative phase 𝜃𝑟 .
Glass
SiNx
Cell trap
SU-8
Stimulus electrode
Recording electrode
Cell
Figure 1: Schematic 3D view of the microfluidic device showing the cell traps and immobilized single cells. The
PDMS cover is not shown in the schematic for better illustration.
978-0-9798064-7-6/µTAS 2014/$20©14CBMS-0001
895
18th International Conference on Miniaturized
Systems for Chemistry and Life Sciences
October 26-30, 2014, San Antonio, Texas, USA
RESULTS AND DISCUSSION
As a consequence of the specific configuration of cell traps and of the cell morphologies (with buds
and without buds), single yeast cells were immobilized at the traps in 4 different orientations as shown in
the micrographs in Figure 2a: unbudded cells (UB); horizontally-immobilized cells with the bud inside the
trap (HBI); horizontally-immobilized cells with the bud outside the trap (HBO); and vertically-immobilized
cells with the bud and mother cell stacked (VB). VB cells were directly discriminated from other
orientations by using the relative magnitude at 1 MHz and the relative phase at 4 MHz (Figure 2b). By
analyzing the multi-frequency EIS data with principal component analysis, HBO, HBI and UB cells were
classified by means of linear discriminant analysis on the full multi-frequency EIS data set (Figure 2c).
b
HBO
UB
HBI
Ar at 1 MHz
0.95
HBI
VB
c
0.97
HBO
VB
Score on 2nd PC [a.u.]
a UB
0.93
VB
0.91
0.89
0.87
1
1.5
2
2.5
3
3.5
4
4.5
10
UB
HBI
HBO
9
UB
8
HBI
7
6
HBO
5
-3
-2
qr at 4 MHz [°]
-1
0
1
Score on 1st PC [a.u.]
2
Figure 2: Discrimination of immobilization orientations of budding yeast cells by using multi-frequency EIS. (a)
Micrographs of 4 orientations. Buds are marked with arrow heads. Scale bar is 5 µm. (b) Separation of VB cells.
(c) Classification of UB, HBI and HBO cells (projections on first two principal components shown).
Among the 4 orientations, VB cells were chosen to perform the bud-growth monitoring through EIS,
as EIS was most sensitive to growth in this orientation. Figure 3a shows the real-time EIS recording of the
budding process of immobilized VB cells. The size increment of the bud, which is difficult to determine
optically within short time periods (inserts in Figure 3a), could be clearly observed through EIS signals.
Figure 3b shows the real-time EIS recording of cell motion of a VB cell. The impedance measurements
also reflected changes in position or motion of single yeast cells in the trap.
a
0
x 10-3
Rec 2 0 min
Rec 2 5 min
Mother cell
Mother cell
-4
Rec 1
Rec 2
Rec 3
Rec 4
Rec 5
-6
-8
0
2
0.95
0 min
0.925
Ar at 1 MHz
ΔAr at 1 MHz
-2
b
Mother cell
18 min
9 min
5 min
0.9
3 min
0.875
4
6
Time [min]
8
0.85
10
0
5
10
Time [min]
15
20
Figure 3: Real-time EIS recording of the budding process and cell motion of immobilized single yeast cells. (a)
Budding process of 5 immobilized cells. Inserts show images of the monitored cell at the beginning and the end of
Rec 2. (b) Cell motion of a cell with inserted images showing its mother cell moving towards the outside of the trap.
Buds are marked with arrow heads. Scale bar is 5 µm.
By analyzing the multi-frequency EIS data of bud-growth and cell-motion recordings with principal
component analysis, two principal vectors were extracted that represent bud growth in Figure 3a and cell
motion in Figure 3b. The multi-frequency EIS data were then projected to the two representative vectors,
as shown in Figure 4. Cell movement (trajectory along vertical axis) during the real-time recordings could
then be qualitatively discerned from bud growth (trajectories horizontal axis). Therefore, this type of
projection enables the discrimination of cell activities during the overall recording duration.
896
Rec 1
Rec 2
Rec 3
Rec 4
Rec 5
Motion
Cell motion [a.u.]
6
4
2
0
0
0.5
1
1.5
Cell growth [a.u.]
Figure 4: Projection of multi-frequency EIS data with respect to cell growth and cell motion vectors. Rec 1 to Rec 5
are from Figure 3a. Motion represents the cell-motion recording in Figure 3b. For each recording, the first data
point was set to (0, 0) for better comparison.
CONCLUSION
A microfluidic device that combines immobilization and localized multi-frequency electrical
impedance measurements of single cells has been presented in this work. The experiments using budding
yeast cells have validated the functionality and sensitivity of the EIS-integrated microfluidic device. The
results demonstrate that single-cell multi-frequency EIS can be used to monitor cell growth, while also
detecting potential cell motion in real-time and label-free, and that EIS constitutes a sensitive tool for singlecell analysis.
ACKNOWLEDGEMENTS
The yeast cells used in this study were kindly supplied by Diana Ottoz and Dr. Fabian Rudolf, ETH
Zurich, D-BSSE, CSB Group. This work was financially supported through the Swiss SystemX.ch program
within the RTD project “CINA” and the FP7 ERC Advanced Grant “NeuroCMOS”.
REFERENCES
[1] T. Sun and H. Morgan, “Single-cell microfluidic impedance cytometry: a review,” Microfluid.
Nanofluid. 8, 423, 2010.
[2] N. Haandbaek, S. C. Burgel, F. Heer and A. Hierlemann, “Characterization of subcellular morphology
of single yeast cells using high frequency microfluidic impedance cytometer,” Lab Chip 14, 369, 2014.
[3] Z. Zhu, O. Frey, D. S. Ottoz, F. Rudolf, and A. Hierlemann, “Microfluidic single-cell cultivation chip
with controllable immobilization and selective release of yeast cells,” Lab Chip 12, 906, 2012.
[4] Z. Zhu, O. Frey, N. Haandbaek, D. S. Ottoz, F. Rudolf, and A. Hierlemann, “Real-time multiparameter monitoring of immobilized single yeast cells via electrical impedance spectroscopy,” in
Proceedings of Transducers’13, Barcelona, Spain, 1527, 2013.
[5] Z. Zhu, O. Frey, F. Franke, N. Haandbaek, and A. Hierlemann, “Real-time monitoring of immobilized
single yeast cells through multifrequency electrical impedance spectroscopy,” Anal. Bioanal. Chem.
DOI 10.1007/s00216-014-7955-9, 2014.
CONTACT
* Zhen Zhu; phone: +41-61-387-3296; [email protected]
897