BioLabs-On-A-Chip: Monitoring Cells Using CMOS Biosensors
Somashekar B. Prakash, Nicole M. Nelson,
Alfred M. Haas, Victor Jeng, Pamela Abshire
Department of Electrical and Computer Engineering
University of Maryland
College Park, Maryland 20742
Abstract— Cell clinics, CMOS/MEMS hybrid microsystems for
capturing and in-situ investigation of living cells, aims at providing high-speed, automated, and economical cell monitoring.
Integrated sensors are being developed for extracellular signal
amplification, cell-substrate capacitance sensing, contact imaging,
and fluorescence detection. We describe the methodology for
characterizing the responses of these sensors to biological cells.
We also present results obtained from the long-term monitoring
of cells cultured on-chip using two of the sensors: (i) a bioamplifier, used for amplifying weak extracellular potentials from
electrically active cells, and (ii) a cell-substrate capacitance
sensor, used for tracking cell adhesion and assessing cell viability.
I. I NTRODUCTION
In order to gain a deeper understanding of the operation
of biological cells, and to learn how to exploit their sensitivity to environmental parameters for sensing applications,
we have developed integrated CMOS sensors to measure their
in vitro behavior and response to stimuli. More specifically,
we have developed CMOS sensor arrays to amplify extracellular potentials [1], monitor cellular capacitance [2], and
image the positions of biological cells [3]. The measurements
made using these sensors have shown strong correlations with
depolarization events, cell viability, and location, respectively.
Integrating these sensors into a CMOS/MEMS microsystem,
our measurement suite, or “cell clinic” [1], aims to perform
measurements rivaling those of conventional instrumentation,
but operating at power levels, size, and cost that are orders of
magnitude smaller.
II. C ELLS O N C HIPS – H OW IS IT B EING D ONE ?
The problem of packaging integrated biosensors is an obvious one: how to keep the electrical leads dry and insulated,
while exposing only the sensors (just tens of microns away), to
the aqueous cellular environment. This often overlooked problem can be a “show stopper.” One can conceive ways of solving this problem, but small tolerances, multiple length scales
(from µ m to cm), biocompatibility, and electrical requirements
are some of the core challenges involved. We describe our
approach for tackling the above mentioned challenges, which
allows us to characterize sensor responses to cells cultured
on-chip.
We thank the MOSIS service for providing chip fabrication; these chips
will be used to teach an undergraduate course in mixed signal VLSI design.
This research was supported by National Science Foundation through Awards
0238061 & 0515873, and by the Laboratory for Physical Sciences.
Mario Urdaneta, Elisabeth Smela
Department of Mechanical Engineering
University of Maryland
College Park, Maryland 20742
level 1
level 2
bond wires
CMOS/MEMS chip
Fig. 1. Left, photograph of a CMOS/MEMS chip after fabrication and before
Loctite patterning. Right, chip photograph after Loctite patterning.
A. Chip Fabrication and Packaging
The sensors were fabricated in a commercially available
0.5 µ m, 2-poly 3-metal CMOS process. The sensor chip was
packaged in a standard 40-pin DIP ceramic package. If the
sensing electrodes (fabricated using aluminum) are required to
be exposed to the electrolyte for direct contact with cells, they
are electrolessly gold plated to make the surface biocompatible
and electrochemically corrosion resistant. The chip is then
encapsulated using Loctite 3340, a photopatternable (365 nm)
and biocompatible polymer. The patterning process is quick
(on the order of minutes) and has been described elsewhere
[4]. The chips are encapsulated using one or two polymer
levels, depending upon the access requirements. Fig. 1 shows
photographs of a 3×3 mm2 CMOS/MEMS chip before and
after Loctite encapsulation. A well for containing the cell
culture is then glued over the encapsulation [2].
B. Sensor Testing with Cells Cultured On-Chip
All the sensing experiments were conducted with bovine
aortic smooth muscle cells (BAOSMCs). Fig. 2 shows
BAOSMCs adhered to an on-chip electrode. These cells exhibit spontaneous electrical activity that depends on their state
bioamplifier
module
sensing
electrodes
electrode
BAOSMC
Fig. 2. Left, photomicrograph of a bio-amplifier test chip comprising an
array of 10 bio-amplifier modules connected to an array of on-chip gold-plated
electrodes. Right, BAOSMC cultured on a chip surface. The long slanted cell
is healthy and viable. The spherical cell’s viability is compromised.
0.04
0.04
0.03
0.02
0.01
0
1
2
3
4
5
6
7
8
0.03
0.02
0.01
0
−0.01
−0.01
−0.02
−0.02
0.3
0.32
0.34
0.36
Time (Seconds)
0.38
25
electrode #
Computed Sensed Capacitance (fF)
0.05
Spike Voltage (Volts)
Spike Voltage (Volts)
0.05
0.503
0.5035
0.504
0.5045
Time (Seconds)
Fig. 3. Left, action potentials from BAOSMC cultured on a bio-amplifier
chip. Right, action potential propagation across the on-chip electrode array.
of health or age. BAOSMC loading into the sensor well is
performed under standard aseptic conditions. The chip is then
mounted onto a test board and placed inside a Faraday cage
for noise immunity. The assembled test fixture is maintained
inside the incubator at 37◦ C, 5% CO2 , throughout the monitoring period. Sensor outputs are continuously monitored using
a data acquisition system interfaced to the test board.
III. O N -C HIP E XTRACELLULAR S IGNAL A MPLIFICATION
A low noise, low power bio-amplifier architecture has been
implemented and tested to record extra-cellular potentials
corresponding to cellular depolarization events. The circuit
comprises a CMOS transconductance amplifier with capacitive
feedback [1]. The amplifier has been designed for a supply
voltage of +/-1.5 V, a gain of 100, and a bandwidth of 3 kHz.
Fig. 2 shows a photograph of the bio-amplifier test-chip.
The chip was tested with BAOSMCs. Fig. 3 shows a trace
acquired from a single bio-amplifier channel that features several action potentials. The pre-amplification spike amplitudes
measure approximately 700 µ V peak-to-peak. The figure also
shows an action potential propagating across the cultured cell
monolayer from electrode 7 to electrode 2 (distance of 500
µ m) over a time frame of approximately 1 ms.
IV. C ELL -S UBSTRATE C APACITANCE S ENSING
Living cells growing on a supporting substrate behave
capacitively when exposed to weak, low-frequency electric
fields. The capacitance results primarily from the insulating
nature of cells and the polarization of the surrounding solution
under these excitation conditions [2]. Cell-substrate capacitance sensing relies on the fact that healthy cells possess
well-formed plasma membranes and adhere strongly to their
substrates. In contrast, unhealthy cells possess compromised
membrane structures and are weakly adherent. Consequently,
viable cells offer higher electrical capacitance and show more
activity compared to cells with compromised viability. This
forms the basis for employing on-chip capacitance sensing for
cell viability monitoring.
CMOS capacitance sensors with different sensing electrode
areas were designed, and tested on bench and in vitro with
BAOSMCs cultured on the chip surface [2], [5]. The sensor
20
cells active & healthy
15
adhesion
cells compromised
10
5
sedimentation
0
0
20
40
60
80
100
Time (Hours)
Fig. 4.
Long term measurement of cell-substrate capacitance.
circuit employs charge sharing for mapping the sensed capacitances to measured voltages. Cell monitoring experiments
with BAOSMCs have demonstrated that on-chip capacitance
measurements can track the cell-substrate interaction process
and respond to changes in cell viability [2]. Fig. 4 shows
a four day plot of the sensed capacitance as recorded by a
sensor with a sensing electrode area of 40×40 µ m2 . The
cells were continuously monitored in a closed, undisturbed
environment on top of the sensor chip, without growth medium
replenishment. As shown in the figure, the capacitance tracks
the initial sedimentation and adhesion phases over the first
few hours. Then the capacitance exhibits many fluctuations,
indicating ongoing cell activity. Over the last two days the
time averaged value of the measured capacitance levels out,
indicating compromised viability and inactivity due to starvation and lack of oxygen.
V. F UTURE W ORK
Validation experiments employing traditional cell biology
techniques are currently in progress for characterizing these
sensors. Extracellular potentials acquired from the bioamplifier
will be compared with those obtained from standard electrophysiology techniques such as “patch clamp.” Capacitance
sensor responses to cell adhesion and viability will be correlated with spectrophotometric measurements using standard
cell viability dyes such as Alamar blue and MTT assays. In
addition to the above, other on-chip cell sensing techniques,
such as contact imaging and fluorescence detection, are also
being investigated and developed.
R EFERENCES
[1] N. Reeves, Y. Liu, N.M. Nelson, S. Malhotra, et al., “Integrated MEMS
structures and CMOS circuits for bioelectronic interface with single
cells,” Proceedings of IEEE ISCAS, vol. 3, pp. 673-676, 2004.
[2] S.B. Prakash and P. Abshire, “A CMOS capacitance sensor that monitors
cell viability,” Proceedings of IEEE Sensors, pp. 1177 - 1180, 2005.
[3] J. Honghao, M. Urdaneta, E. Smela and P. Abshire, “CMOS contact
imager for monitoring cultured cells,” Proceedings of IEEE ISCAS, vol.
4, pp. 3491 - 3494 , 2005.
[4] R. Delille, M. Urdaneta, S. Moseley, and E. Smela, “Benchtop polymer
MEMS,” Journal of Microelectromechanical Systems, in press, 2006.
[5] S.B. Prakash, M. Urdaneta, E. Smela and P. Abshire, “A CMOS
capacitance sensor for cell adhesion characterization,” Proceedings of
IEEE ISCAS, vol. 4, pp. 3495 - 3498, 2005.