Lab on a Chip
View Online / Journal Homepage / Table of Contents for this issue
C
Dynamic Article Links <
Cite this: Lab Chip, 2012, 12, 780
PAPER
www.rsc.org/loc
Sonolysis of Escherichia coli and Pichia pastoris in microfluidics
Downloaded by Nanyang Technological University on 06 June 2012
Published on 20 December 2011 on http://pubs.rsc.org | doi:10.1039/C2LC20861J
Tandiono Tandiono,†*a Dave Siak-Wei Ow,†b Leonie Driessen,c Cara Sze-Hui Chin,b Evert Klaseboer,a
Andre Boon-Hwa Choo,b Siew-Wan Ohla and Claus-Dieter Ohl*c
Received 9th September 2011, Accepted 28th November 2011
DOI: 10.1039/c2lc20861j
We report on an efficient ultrasound based technique for lysing Escherichia coli and Pichia pastoris with
oscillating cavitation bubbles in an integrated microfluidic system. The system consists of a meandering
microfluidic channel and four piezoelectric transducers mounted on a glass substrate, with the
ultrasound exposure and gas pressure regulated by an automatic control system. Controlled lysis of
bacterial and yeast cells expressing green fluorescence protein (GFP) is studied with high-speed
photography and fluorescence microscopy, and quantified with real-time polymerase chain reaction
(qRT-PCR) and fluorescence intensity. The effectiveness of cell lysis correlates with the duration of
ultrasound exposure. Complete lysis can be achieved within one second of ultrasound exposure with
a temperature increase of less than 3.3 C. The rod-shaped E. coli bacteria are disrupted into small
fragments in less than 0.4 seconds, while the more robust elliptical P. pastoris yeast cells require around
1.0 second for complete lysis. Fluorescence intensity measurements and qRT-PCR analysis show that
functionality of GFP and genomic DNA for downstream analytical assays is maintained.
Introduction
Micro-scale analysis of intracellular contents, such as nucleic
acids and proteins, is gaining importance in biology.1 Other than
enabling minimized analytical and cell biology profiling of
processes at the cellular level, microfluidics is also finding new
applications relating to micro-culturing of cells for high
throughput screening and biological research.2–5 The prokaryotic
Gram-negative Escherichia coli bacterium and the eukaryotic
Pichia pastoris yeast are microbial host cells extensively used for
screening of clones from genomic libraries and heterologous
protein expression.6,7 They both allow the functional expression
of multiple proteins in parallel in microplate assays, which can be
made amendable to micro-scale analysis. However, before the
micro-scale analysis can be carried out, an effective microfluidic
cell lysis for the release of active intracellular contents needs to be
achieved.
In microfluidics, cell lysis can be accomplished by means of
chemical,8 thermal,9 electrical,10 or mechanical lysis.11,12 Of these,
chemical lysis with lytic agents and thermal lysis with heat
frequently lead to the denaturation of proteins or interfere with
a
Institute of High Performance Computing, 1 Fusionopolis Way, #16-16
Connexis, Singapore, 138632, Singapore. E-mail: [email protected].
edu.sg
b
Bioprocessing Technology Institute, 20 Biopolis Way, #06-01 Centros,
Singapore, 138668, Singapore
c
Division of Physics and Applied Physics, School of Physical and
Mathematical Sciences, Nanyang Technological University, Singapore,
637371, Singapore. E-mail: [email protected]
† These authors contribute equally to this work.
780 | Lab Chip, 2012, 12, 780–786
subsequent assays. Furthermore, chemical lysis has the added
disadvantages of requiring wet chemical storage and intensive
mixing, which will add complexity in a microfluidic setting.
Although electrical cell lysis has the advantages of being reagentless and quick, the application of a direct current at elevated
voltage can lead to water hydrolysis, undesirable localized
heating, and denaturation of proteins.13 Mechanical lysis, which
involves the generation of high shear through the application of
high pressure, rapid agitation, or sonication, often needs intensive cooling to remove the heat produced by the dissipation of the
mechanical energy.
Despite the undesirable heating, sonication has been widely
used in the lab-scale to attain mechanical lysis of cells.14–16 The
basic principle of sonication is to generate mechanical shear
stress by oscillating cavitation bubbles using an ultrasound field.
In bulk medium, this typically involves the application of
a bench-top vibrating probe directly into the liquid, where the
rapid movement of the probe tip creates a series of rapidly
collapsing cavitation bubbles that break apart cells. In general,
this process is inefficient and some energy is lost as heat.
The ability to introduce strongly oscillating cavitation bubbles
in a microfluidic setting without localized heating offers an
unparalleled potential for reagent-less cell lysis without protein
denaturation, hence facilitating lab-on-chip analysis. Taylor
et al.15 reported on a microfluidic cell lysis technique by sonication which uses glass beads mixed with the cell sample, and the
ultrasound vibration is subsequently used to create a rapid
agitation in the micro chamber. Recently, we successfully created
intense cavitation in a microfluidic channel by exciting gas–liquid
interfaces with ultrasound vibration.17 The cavitation is initiated
This journal is ª The Royal Society of Chemistry 2012
Downloaded by Nanyang Technological University on 06 June 2012
Published on 20 December 2011 on http://pubs.rsc.org | doi:10.1039/C2LC20861J
View Online
from a nonlinear interface instability that entraps small gas
bubbles. The bubbles later serve as cavitation nuclei. Their
oscillations create regions of high stress, which can rupture the
cells membrane if they are sufficiently close to the cells.18
In this work, we build on this technology to mechanically
disrupt the bacterial and yeast cells for harvesting their intracellular contents. Escherichia coli and Pichia pastoris were chosen
because they are widely used for screening of cDNA genomic
libraries and functional protein expression.19 The method allows
the sonication of small volumes of samples without chemical
reagents or direct contact between transducer and samples, hence
reducing the possibility of assay interference or sample crosscontamination. Furthermore, the ultrasound exposure is performed for a brief duration with minimal heating, hence facilitating functional downstream characterization of both nucleic
acids and proteins in microfluidics.
Experimental details
The experiments were conducted on an integrated setup (Fig. 1A)
consisting of a microfluidic system, a control system, and off-line
assay systems, which include: a quantitative real-time polymerase
chain reaction (qRT-PCR) analysis and a fluorescence intensity
measurement.
The microfluidic system comprises of a meandering microfluidic channel and four piezoelectric transducers (PZT material,
disc transducer, 20 mm diameter with 2.1 mm thickness, Steiner
& Martins) attached on a glass substrate. The microfluidic
channel is made from polydimethylsiloxane. It consists of two
inlets and one outlet. The inlets are connected through a T
junction such that the gas can be injected into the main (liquid)
channel to create gas–liquid interfaces within the channel. The
width of the gas and the main channels are 50 and 100 mm,
respectively. The main channel expands to 500 mm downstream
to achieve shorter gas/liquid slugs and thus longer interfaces in
the channel. The height of the channel is 20 mm. Details of the
microfluidic fabrication and assembly are described in Tandiono
et al.17 The outlet of the channel is connected to the collection
tube. The tube is specially designed to minimize evaporation of
the supernatant.
The ultrasound exposures and the gas pressure injected into
the microchannel are run by an automatic control system. The
timings of the exposure and the controlled gas pressure are
shown in Fig. 1B. A syringe pump pushes the sample liquid at
a constant flow rate such that the sample is exposed to six bursts
of ultrasound every 5 seconds by an amplifier (AG1021, LF
Amplifier/Generator, T&C Power Conversion). This time
interval of 5 seconds allows the sample to cool down. Each burst
is composed of a harmonic driving of 500 to 50 000 ultrasound
cycles of 200 V amplitude at the resonance frequency of the
microfluidic system. As we focus on the cell lysis due to cavitation bubbles, the driving amplitude was chosen such that intense
cavitation always occurs when the microfluidic system is exposed
to the ultrasound. The amplitudes of the acoustic pressure inside
the microchannel and displacement of the glass substrate at this
driving amplitude were measured to be approximately 10 bars
Fig. 1 Design of the integrated microfluidic system for controlled cell lysis experiments. (A) Schematic of the experimental setup. The microfluidic
system consists of a meandering microfluidic channel and four piezoelectric transducers attached on a glass substrate. Two inlet ports for gas and liquid
samples respectively are connected through a T junction to create gas–liquid interfaces within the channel. Samples are collected after ultrasound
treatment in a collection tube for qRT-PCR analysis and fluorescence intensity measurement. The graph at the bottom left is a typical plot obtained from
qRT-PCR analysis for samples with and without ultrasound exposure. (B) A timing diagram of the ultrasound exposure and flushing of the liquid in the
microchannel. For each cycle, the sample is exposed to six bursts of ultrasound. At the end of the cycle, the liquid sample in the channel is flushed out by
applying higher gas pressure. The total duration of a full cycle tcyc varies between 32 and 45 seconds, depending on the flushing and feeding duration. (C)
Images of the microchannel taken by a high-speed camera with an exposure time of 1 ms. The lower left of each frame indicates the time in microseconds.
From top to bottom: a channel filled with yeast cells, cavitation bubbles during expansion phase, and the collapse of the bubbles.
This journal is ª The Royal Society of Chemistry 2012
Lab Chip, 2012, 12, 780–786 | 781
Downloaded by Nanyang Technological University on 06 June 2012
Published on 20 December 2011 on http://pubs.rsc.org | doi:10.1039/C2LC20861J
View Online
and 0.4 mm, respectively.17 The resonance frequency was determined prior to the experiments by adjusting the driving
frequency of the amplifier such that the RF power delivered to
the system is at its maximum. The frequency alters slightly with
each device by a few kilohertz; on average it is around 130 kHz.
The total duration of the ultrasound exposure is varied between
tens of milliseconds to seconds. The exposed sample is subsequently flushed out to the collection tube by applying a high
pressure to the gas inlet for 5–10 seconds, followed by 5–20
seconds of reduced pressure for the flow in the channel to
stabilize and fresh sample to be injected into the channel. The
automatic control system then starts a new cycle. The amplifier
and pressure controller (VSO-BT Benchtop Controller, Parker,
USA) are timed by function generators (33220A 20MHz Function/Arbitrary Waveform Generator, Agilent Technologies,
USA) which are triggered by a digital delay generator (Model
575, BNC, USA).
The microbial strains used were green fluorescence protein
(GFP) expressing Escherichia coli BL21(DE3) and Pichia pastoris
GS990 strains harbouring the pAcGFP1 (Clontech) and pGAPEGFPd vectors respectively. The cell concentration of the
samples was maintained constant at an OD600 of approximately
4.0 and 7.0 for E. coli and P. pastoris, respectively. Cell suspension at a volume of 60–100 ml was fed into the microchannel for
ultrasound treatment. The treated samples were collected in
a collection tube, and centrifuged (13 000 rpm, 1 min) to collect
the supernatant. As the cells lysed, intracellular proteins and
nucleic acids would be released into the supernatant and can be
subsequently measured based on GFP fluorescence and qRTPCR analysis.
The GFP fluorescence intensity measurement of the E. coli was
performed on a microplate reader (Infinite 200, Tecan, Switzerland). Treated and untreated samples were diluted two times in
tris-buffered saline buffer and placed on a microplate (Greiner 96
Flat Bottom Black Polystyrol, Germany) according to the
manufacturer’s protocol. The excitation and emission wavelengths used in the measurement were 475 nm and 509 nm,
respectively, which correspond to the maximum excitation and
emission peaks of the wildtype GFP variant expressed in the
experiments.
The qRT-PCR assay was conducted on an ABI PRISM 7500
instrument (Applied Biosystems, California, USA) to quantify
the release of intracellular DNA from lysed cells according to
a modified protocol from Lee et al.20 The assay was performed
with the following cycling conditions: 50 C for 2 min, 95 C for
10 min, and followed by 40 cycles of 95 C for 15 s and 60 C for 1
min each. Each 25 ml PCR reaction contains 2 ml of standard or
sample DNA, 12.5 ml of SYBRR Green PCR Master Mix
(Applied Biosystems), 9.25 ml water, and 12.5 pmol each of
forward and reverse primers (Pichia 18SrDNA primer-1
ATTACGTCCCTGCCCTTTGTAC, Pichia 18SrDNA primer2 CCAAAGCCTCACTAAACCATTCA, E. coli 16SrDNA
primer-1 TCGTGTTGTGAAATGTTGGGTTA, E. coli
16SrDNA primer-2 CCGCTGGCAACAAAGGATA). Serial
dilutions of E. coli or P. pastoris genomic DNA standards were
run in duplicate to establish the standard curves of PCR
threshold cycle (Ct) versus log DNA concentration (Co). From
these, the interpolation of the Ct value against the standard curve
would allow the quantification of total DNA presented in
782 | Lab Chip, 2012, 12, 780–786
samples. The theoretical DNA yield was calculated based on
a DNA content of 3.1% for cells with a single copy of chromosome and an average 2–4 copies of chromosome per dividing
cell.21,22 As an OD 4.0 culture corresponds to approximately 2 g
cells per litre (1 OD ¼ 0.5 g dry cell mass per litre), the theoretical
DNA yield is estimated to be around 124–248 ng ml1.
A camera (Fastcam SA1.1, Photron, Japan) connected to an
inverted microscope (IX-71, Olympus, Japan) is used for highspeed imaging at up to 300 000 frames per second. Fig. 1C shows
selected frames from a sequence showing yeast cells in a microchannel with an exposure time of 1 ms. The first frame depicts
a microchannel with yeast cells prior to the ultrasound exposure
and the lower two frames capture the scene with bubble activity
at the expansion and collapse time respectively.
Results and discussion
Green fluorescence protein (GFP) from jellyfish Aequorea
Victoria23 is a well-established reporter protein for functional
gene expression studies in bacteria, yeast, and several other
organisms.24,25 For quantifying the effectiveness of our cell lysis
method and for imaging purposes, the GFP gene was cloned
under the control of a constitutive promoter and introduced into
E. coli and P. pastoris.
Bacteria lysis: Escherichia coli
The cell wall of the Gram-negative bacterium E. coli comprises of
a single layer of peptidoglycan surrounded by an outer
membrane.26 The outer membrane is composed of lipopolysaccharides, lipoproteins, and phospholipids. The disruption of the
cell wall requires the destruction of the thin layer of the peptidoglycan network with an approximately 10 nm thickness. The
diameter of the cavitation bubbles at maximum expansion (see,
for example, Fig. 1C) is much larger than the typical length of E.
coli, i.e. 4–6 mm. However, during the collapse phase, the bubbles
shrink below the resolution limit of the camera, thus the fluid
mechanic cavitational force on the bacteria may occur on scales
comparable to the bacterial size.
We demonstrate the effect of cavitation on the integrity and
fluorescence emission of the cells. Fig. 2A depicts the GFP
emission before (left), during (middle), and after (right) ultrasound exposure using fluorescence microscopy, while Fig. 2B
shows bright field microscopy images of the bacteria before (left)
and after (right) exposure. The image was captured with
a camera connected to a microscope and a long working distance
objective lens with 100 magnification and 1.0 mm correction
for the glass substrate (LCPLFL-LCD 100, Olympus). The
depth of focus of the objective is 0.79 mm, which is much shorter
than the microchannel height of 20 mm. Ultrasound (US) is
applied for 389 ms at a frequency of 128.7 kHz using a driving
voltage of 200 V; the duration corresponds to 50 000 US cycles.
Fig. 2A (left) depicts a population of intact GFP-fluorescent
bacterial cells with a dark non-fluorescing background prior to
US exposure. During the exposure, the cavitation bubbles create
such intense mixing that images of cells are motion blurred (see
Fig. 2A middle). The intense mixing also causes the cells move
out of the focal plane, making it difficult to resolve the history of
cell translation during the US exposure. Shortly after the US
This journal is ª The Royal Society of Chemistry 2012
Downloaded by Nanyang Technological University on 06 June 2012
Published on 20 December 2011 on http://pubs.rsc.org | doi:10.1039/C2LC20861J
View Online
Fig. 2 Images of the GFP Escherichia coli captured before and after ultrasound exposure. (A) Fluorescence microscopy images of the bacteria. The
distorted image at the middle is taken during the ultrasound exposure, where the cells were motion-blurred. After the exposure, the supernatant becomes
greenish, indicating that the intracellular contents of the cells are released into the liquid media as the cells are effectively disrupted. (B) Zoom-in images
of the bacterial cells using bright field microscopy. After the ultrasound treatment, the rod-shape E. coli cells are broken into fragments.
exposure, when the flow has ceased, the effect on the cells
becomes apparent (Fig. 2A right). Besides a greenish supernatant, no intact bacteria remains, indicating that the US exposure
has completely lysed the cells thus releasing the intracellular GFP
into the medium. Further checks along the microchannel do not
reveal any intact GFP-fluorescent cells within the channels. To
further examine the integrity of the treated cells, high
Fig. 3 Fluorescence intensity of the supernatant of the GFP Escherichia
coli after ultrasound exposures. The mean values and error bars in all
data points were obtained from multiple runs (3 or 4 runs). The excitation
and emission wavelengths are 475 nm and 509 nm, respectively. The
dashed-dotted line is the exponentially fitted curve. The fluorescence
intensity increases with the increase of ultrasound exposure duration,
indicating an increase in the number of cells lysed. The dashed line
indicates the negative control, which was obtained from the untreated
samples.
This journal is ª The Royal Society of Chemistry 2012
magnification images were also taken before and after the US
exposures under bright field illumination. Fig. 2B right depicts
the complete fragmentation of the rod-shaped cells into small
fragments. We attribute the fragmentation to the fluid mechanic
forcing, i.e. shear stress generated from the oscillating cavitation
bubbles.
The amount of intracellular proteins released to the supernatant can be determined by measuring the fluorescence of the GFP
protein at its corresponding emission wavelength of 509 nm.
Fig. 3 shows the measured fluorescence intensity in the clarified
supernatant from E. coli samples after the ultrasound treatment
at different exposure durations. The dashed negative control line
corresponds to the fluorescence intensity of the untreated sample.
A significant increase in the fluorescence is observed after only
a short US exposure of approximately 20 milliseconds. The
fluorescence increases with exposure duration reaching a plateau
in less than 1 second of US exposure.
As cells undergo ultrasound-mediated lysis, nucleic acids and
other cytoplasmic molecules will also be released into the
medium. Hence, another alternative method to assess for efficient cell lysis is by performing qRT-PCR to quantify the amount
of genomic DNA released into the supernatant. Fig. 4 shows the
results of qRT-PCR analysis of the same samples shown in Fig. 3
using a pair of primers targeting E. coli 16SrDNA. As expected,
a significantly higher amount of DNA was detected in the treated
samples compared to the untreated samples (negative control).
The trend of increasing DNA concentration over the exposure
duration is similar to that of fluorescence intensity (Fig. 3). As
more cells are lysed due to the longer ultrasound exposure, more
nucleic acids are released to the supernatant, resulting in higher
DNA concentration. As before, the DNA concentration also
reaches a plateau approximating the maximum theoretical
Lab Chip, 2012, 12, 780–786 | 783
Downloaded by Nanyang Technological University on 06 June 2012
Published on 20 December 2011 on http://pubs.rsc.org | doi:10.1039/C2LC20861J
View Online
Fig. 4 DNA concentration of the treated Escherichia coli cell suspension
with increasing ultrasound exposure duration. The mean values and error
bars in all data points were obtained from multiple runs (3 or 4 runs). The
DNA concentration increases as the exposure duration increases, indicating that more cells are disrupted. The negative control is obtained
from the average DNA concentration of the untreated samples from
various batches of cells.
genomic DNA yield within 1 second of exposure. The plateau is
believed to relate to an extensive lysis of nearly all cells in the
sample, which was also visually confirmed by the fluorescence
microscopy images shown in Fig. 2. It may be argued that a built
up of cavitation activity leads to a change of the acoustic
impedance which reduces the transmission of acoustic energy.
Yet, only a small area fraction of the radiating glass plate is
covered with cavitation bubbles, and the cavitation activity is still
observed even after many cycles of ultrasound exposure. Thus,
the plateau may not be explained by mechanical causes, for
example, due to a drift of the system’s resonance frequency.
Yeast lysis: Pichia pastoris
Pichia pastoris is a species of yeast cells with diameters of approximately 4 mm (comparable to the length of E. coli bacteria), but their
shape is elliptical or oval in contrast to the rod-shaped bacteria.
Yeast cells have a rigid extracellular cell wall consisting of a layered
mesh of embedded glucans, chitin and mannoproteins27 which
provides physical protection and gives structural strength to the
cells. The bursting strength of yeasts measured using micromanipulation methods was found to be at least an order of magnitude
higher than typical animal cells.28 Hence, Pichia pastoris can be
regarded as comparatively tough cells to lyse.
A high speed sequence of images showing the deformation and
disruption of the yeast cells when they are exposed to the US
vibration is presented in Fig. 5. The camera was focused on
a group of cells near the bubble prior to the ultrasound exposure
to capture the bubble–cells interaction. Soon after the ultrasound
was applied, cavitation was observed in the microchannel. The
frames in the figure display images of the cells when the cavitation bubbles are close to their minimum size (thus no bubbles
were visible in the frames). This allows the capturing of a clearer
view of the cells as they are moved and deformed (stretched) by
a straining flow. The arrows in Fig. 5 (labelled cell 2) point to
a cell that experiences a particularly strong deformation. Initially
(t ¼ 0 ms), this cell has a round shape with a diameter of 3–4 mm.
Later, the cell is stretched (at t ¼ 20 ms) and splits into two
fragments (at t ¼ 40 ms) as a result of high shear stress of the
oscillating bubbles. Presumably, cells in the microchannels can
become stretched above their yield strength and then rupture.
This may consequently result in leakage of intracellular content
into the supernatant.
The DNA released from P. pastoris was quantified using qRTPCR analysis, as shown in Fig. 6. The amount of DNA is plotted
as a function of US exposure duration double-logarithmically.
The DNA concentration, thus the number of disrupted cells,
increases with exposure duration and levels off at an exposure
duration of about 1 second. This plateau indicates that majority
of the cells have lysed. Any further increase in the exposure
duration in our experiments consistently leads to a slight
decrease in the DNA concentration. This may be explained by
mechanical or chemical damage of the harvested DNA, e.g.,
formation of OH radicals from long ultrasound exposure.29
Fig. 5 Sequential images of the deformation of Pichia pastoris yeast cells during ultrasound exposure. The driving frequency and the driving voltage of
the ultrasound exposure are 99 kHz and 230 V, respectively. The dashed circle shows a group of cells near the collapsing bubbles. The bubbles are not
visible because the frames are taken during the bubble collapse phase. The arrows point to a cell labelled 2, which experiences the largest deformation due
to shear stress generated by cavitation bubbles. The lower left number is the time in microseconds. At t ¼ 20 ms, cell 2 becomes stretched and splits into
two at t ¼ 40 ms. The video is recorded at a frame rate of 300 000 frames per second and an exposure time of 1.76 ms. The width of each frame is 21.5 mm.
784 | Lab Chip, 2012, 12, 780–786
This journal is ª The Royal Society of Chemistry 2012
View Online
frequency of 130 kHz. As shown earlier, this exposure parameter
is sufficient for complete lysis of both bacteria and yeast. Each
burst of ultrasound gives rise to a temperature increase of only 1–
2 C. The time delay between the bursts allows the temperature
of the sample to cool down by 1.0–1.5 C. Hence, for the full
ultrasound exposure treatment, the maximum temperature
increase of the sample remains below 3.3 C.
Downloaded by Nanyang Technological University on 06 June 2012
Published on 20 December 2011 on http://pubs.rsc.org | doi:10.1039/C2LC20861J
Concluding remarks
Fig. 6 DNA concentration of the treated Pichia pastoris cell suspension
with increasing ultrasound exposure duration. The data points with error
bar were obtained from multiple runs (2–4 runs). As the exposure
duration increases, the DNA concentration also increases, thus more cells
are disrupted. The negative control is obtained from the average DNA
concentration of the untreated samples from various batches.
Minimising heat during cell lysis
One ever present concern of ultrasound usage for cell lysis is the
heating of the cells and their subsequent denaturation of proteins
which may affect downstream bio-assays. In the above
mentioned protocol we applied six short bursts of US with some
time-interval between the bursts to allow cooling of the sample.
Fig. 7 shows the temperature of the sample over a cycle of
ultrasound exposure measured with a type K thermocouple
introduced through small holes into the microchannel. The
measurements were carried out for ultrasound exposures with
a total duration of 0.92 seconds (6 bursts of 0.154 seconds),
which corresponds to 6 20 000 ultrasound cycles at a driving
The main advantage of our microfluidic system lies in the ability
to lyse microlitre volume of cells without generation of excess
heat, hence maintaining the quality of the harvested intracellular
contents. This is achieved by exposing the cells to a controlled
amount of cavitation for sufficiently short times. The large
surface area due to the microfluidic environment enhances the
transport of heat quickly away from the liquid.
In conclusion, we have presented a gentle yet efficient technique for lysis of Escherichia coli and Pichia pastoris in microfluidics. The technique is based on ultrasound driven cavitation
bubbles which successfully disrupt bacterial and yeast cells. Both
cell lines constitutively express green fluorescence protein to
facilitate visualization of intact cells and quantification of the
released cellular content from US-treated cells. We attribute the
efficient cell lysis to the high shear produced by the fast-moving
cavitation bubbles. The rod-shaped bacteria, E. coli, are broken
into small fragments in less than 0.4 seconds, while more robust
elliptical-shaped yeast, P. pastoris, requires only about 1.0
second to be disrupted. The temperature increase of the samples
during ultrasound exposures is minimal. Fluorescence intensity
measurements and qRT-PCR analysis provide evidence that the
functional integrity of the harvested protein and DNA for
downstream analytical assays is maintained.
Acknowledgements
This work is supported by the Agency for Science, Technology
and Research (A*STAR) through the JCO Grant no. 10/03/FG/
05/02. We would like to thank Victor Wong for his insightful
discussions and Lisa Heng for her administrative support.
References
Fig. 7 Temperature of the sample during a cycle of ultrasound exposure.
The values were obtained from the average temperature of two different
locations on the microchannel (near the piezoelectric transducer and at
the middle of microfluidic device). The sample was exposed to 6 bursts of
0.154 seconds ultrasound. The dashed line refers to the initial temperature of the sample. During the measurement, the same sample was kept in
the channel to monitor the changes in temperature.
This journal is ª The Royal Society of Chemistry 2012
1 M. Chudy, I. Grabowska, P. Ciosek, A. Filipowicz-Szymanska,
D. Stadnik, I. Wyzkiewicz, E. Jedrych, M. Juchniewicz,
M. Skolimowski, K. Ziolkowska and R. Kwapiszewski, Anal.
Bioanal. Chem., 2009, 395, 647–668.
2 A. Arora, G. Simone, G. B. Salieb-Beugelaar, J. T. Kim and A. Manz,
Anal. Chem., 2010, 82, 4830–4847.
3 G. B. Salieb-Beugelaar, G. Simone, A. Arora, A. Philippi and
A. Manz, Anal. Chem., 2010, 82, 4848–4864.
4 A. Edlich, V. Magdanz, D. Rasch, S. Demming, S. A. Zadeh,
R. Segura, C. K€ahler, R. Radespiel, S. B€
uttgenbach, E. FrancoLara and R. Krull, Biotechnol. Prog., 2010, 26, 1259–1270.
5 K. S. Lee, P. Boccazzi, A. J. Sinskey and R. J. Ram, Lab Chip, 2011,
11, 1730–1739.
6 M. Boettner, B. Prinz, C. Holz, U. Stahl and C. Lang, J. Biotechnol.,
2002, 99, 51–62.
7 R. Daly and M. T. W. Hearn, J. Mol. Recognit., 2005, 18, 119–138.
8 E. A. Schilling, A. E. Kamholz and P. Yager, Anal. Chem., 2002, 74,
1798–1804.
9 L. C. Waters, S. C. Jacobson, N. Kroutchinina, J. Khandurina,
R. S. Foote and J. M. Ramsey, Anal. Chem., 1998, 70, 158–162.
Lab Chip, 2012, 12, 780–786 | 785
Downloaded by Nanyang Technological University on 06 June 2012
Published on 20 December 2011 on http://pubs.rsc.org | doi:10.1039/C2LC20861J
View Online
10 H. Y. Wang, A. K. Bhunia and C. Lu, Biosens. Bioelectron., 2006, 22,
582–588.
11 D. D. Carlo, K. H. Jeong and L. P. Lee, Lab Chip, 2003, 3, 287–291.
12 J. Kim, S. H. Jang, G. Jia, J. V. Zoval, N. A. Da Silva and
M. J. Madou, Lab Chip, 2004, 4, 516–522.
13 S. K. Jha, G. S. Ra, G. S. Joo and Y. S. Kim, Curr. Appl. Phys., 2009,
9, e301–e303.
14 D. E. Hughes and W. L. Nyborg, Science, 1962, 138, 108–114.
15 M. T. Taylor, P. Belgrader, B. J. Furman, F. Pourahmadi,
G. T. A. Kovacs and M. A. Northrup, Anal. Chem., 2001, 73, 492–
496.
16 T. A. Bigelow, T. Northagen, T. M. Hill and F. C. Sailer, Ultrasound
Med. Biol., 2009, 35, 1026–1031.
17 Tandiono, S. W. Ohl, D. S. W. Ow, E. Klaseboer, V. V. T. Wong,
A. Camattari and C. D. Ohl, Lab Chip, 2010, 10, 1848–1855.
18 S. Le Gac, E. Zwaan, A. Berg and C. D. Ohl, Lab Chip, 2007, 7, 1666–
1672.
786 | Lab Chip, 2012, 12, 780–786
19 M. B€
ottner and C. Lang, Methods Mol. Biol., 2002, 267, 277–286.
20 C. L. Lee, D. S. W. Ow and S. K. W. Oh, J. Microbiol. Methods, 2006,
65, 258–267.
21 Z. Wang, L. Xiang, J. Shao, A. We˛grzyn and G. We˛grzyn, Microb.
Cell Fact., 2006, 5(34), 18.
22 K. Nordstr€
om and S. Dasgupta, EMBO Rep., 2006, 7, 484–489.
23 O. Shimomura, F. H. Johnson and Y. Saiga, J. Cell. Comp. Physiol.,
1962, 59, 223–239.
24 M. Chalfie, Y. Tu, G. Euskirchen, W. W. Ward and D. C. Prasher,
Science, 1994, 263, 802–805.
25 R. Y. Tsien, Annu. Rev. Biochem., 1998, 67, 509–544.
26 V. Braun, J. Infect. Dis., 1973, 128, S9–S16.
27 F. M. Klis, Yeast, 1994, 10, 851–869.
28 T. Srinorakutara, J. Ferment. Bioeng., 1998, 86, 253–260.
29 Tandiono, S. W. Ohl, D. S. W. Ow, E. Klaseboer, V. V. T. Wong,
R. Dumke and C. D. Ohl, Proc. Natl. Acad. Sci. U. S. A., 2011,
108, 5996–5998.
This journal is ª The Royal Society of Chemistry 2012