01 Esseling
01 Esseling
01 Esseling
org
BIOPHOTONICS
N T
R I
E P
R
J. Biophotonics 5, No. 1, 913 (2012) / DOI 10.1002/jbio.201100052
Journal of
BIOPHOTONICS
LETTER
Multimodal biophotonic workstation
for live cell analysis
Michael Esseling*; 1, Bjorn Kemper 2, Maciej Antkowiak 3; 4 , David J. Stevenson 3; 4 ,
Lionel Chaudet 5 , Mark A. A. Neil 5 , Paul W. French 5 , Gert von Bally 2, Kishan Dholakia 3; 4 ,
and Cornelia Denz 1
1
Institute of Applied Physics, University of Muenster, Corrensstrae 2/4 Muenster 48149, Germany
2
Center for Biomedical Optics and Photonics (CeBOP), University of Muenster, Germany
3
SUPA, School of Physics & Astronomy, University of St Andrews, North Haugh, St Andrews, UK
4
SULSA, School of Biology, Medical and Biological Sciences Building, University of St Andrews, St Andrews, UK
5
Department of Physics, Imperial College, London, UK
Key words: biophotonics, holographic optical tweezers, micromanipulation, digital holographic microscopy,
dynamic phase-contrast microscopy, fluorescence lifetime imaging, light emitting diode
BIOPHOTONICS
10 M. Esseling et al.: Multimodal biophotonic workstation for live cell analysis
sight into cellular activity. In this letter, we present a Dynamic phase-contrast microscopy consists of
multimodal approach to the measurement of both the combination of an all-optical novelty filter with a
biophysical and biochemical properties to determine commercial microscope. The setup that is incorpo-
the influence of the micromanipulation of cells. The rated here uses two beam-coupling in a photorefrac-
setup includes four different techniques combined tive barium titanate crystal to achieve background
in a single biophotonic workstation integrated in a suppression and quantitative phase-measurements in
commercially available microscope. Holographic a mapping area of 0 j p [12]. Due to these
optical tweezers (HOT) are used for micromanipu- working characteristics, DynPCM provides a high
lation, while dynamic phase-contrast microscopy Signal-to-Noise-Ratio, even for pure phase-objects
(DynPCM) provides real-time monitoring of cell [13]. In this experiment, the focus is on the novelty
changes and movement. Digital holographic micro- filtering capability of the setup. It complements
scopy (DHM) interferograms recorded during the DHM in the sense that it does not need any image
manipulation of cells provide a high-resolution de- reconstruction. Hence, it enables real-time obser-
tection of deformations and refractive index proper- vation of dynamic processes at camera-dependent
ties of the sample, whilst fluorescence lifetime ima- frame rates. This allows researchers to instantly de-
ging (FLIM) monitors the fluorescence lifetimes of tect movement in a biological sample or a microflow
dyes deposited in the membrane of cells. In order to [14].
test our system in a real biological experiment, Chi- Fluorescence Lifetime Imaging is an attractive
nese Hamster Ovary (CHO) cells, which have been method for contrasting different molecular dyes and
treated to contain a membrane-localized dye, are their environment [15]. Compared to fluorescence
stretched using optical tweezers and observed by all intensity imaging, FLIM is independent of fluoro-
three imaging methods (DynPCM, DHM, FLIM) si- phore concentration and excitation intensity and can
multaneously. provide quantitative readouts of local fluorophore
environment [16]. Therefore, this technique provides
complementary information on local physical and
chemical molecular parameters. FLIM is usually im-
plemented using a pulsed or externally modulated
2. Workstation technologies laser for the excitation, however, it has recently been
shown that inexpensive light emitting diodes (LED)
Since their invention in 1986 [6], optical tweezers can also be an effective excitation source for FLIM
have been used in the life sciences for a wide range [17]. In this project, we use FLIM to read out the
of experimentation, and have been combined with a fluorescence lifetime of the membrane-localized dye,
multitude of imaging modalities including confocal di-4-ANEPPDHQ [18], as a reporter of membrane
laser scanning microscopy, epi-fluorescence, multi- lipid order in order to track the effect of trapping on
photon microscopy, and Raman spectroscopy [7]. the cell membrane over the course of the experi-
Their functionality has been greatly increased with ment.
the advent of multiple dynamically addressable twee-
zers, made possible with acousto-optic deflectors or
spatial light modulators [8]. One exciting application
of optical forces is that of optical stretching. Here, a 3. Technical assembly of workstation
cell is pulled in order to characterise its plasticity, al-
lowing the discrimination between highly elastic can- The novel biophotonics workstation was built on a
cer cells and less compliant healthy cell phenotypes Nikon TE2000 platform (Figure 1). For the holo-
[9]. graphic optical tweezers, a Coherent MIRA900 laser
Digital holographic microscopy enables label- (800 nm CW) pumped by a Verdi-V5 was coupled
free, quantitative phase contrast imaging for high re- into the upper turret of the microscope. A spatial
solution technical inspection and minimal invasive light modulator (SLM, Hamamatsu X8267-13) was
live cell analysis and provides information about the relayed to the back focal plane of a 60X 1.4 NA Ni-
intracellular refractive index ([10] and references kon microscope objective (MO). Multiple tweezers
therein). Compared to other phase-contrast meth- were created by a superposition of lens and prisms
ods, interferometry-based techniques, and optical co- elements on the SLM, allowing full 3D control of ar-
herence tomography or optical coherence micro- bitrary numbers of objects. In the experiment, a dou-
scopy, DHM provides quantitative phase-contrast ble-trap was used to stretch the cell. The lateral dis-
with subsequent numerical focus correction (multi- tance between the traps was increased in steps of
focus imaging) from a single recorded hologram. In 100 nm at a rate of approx. 100 nm/s.
combination with algorithms for the quantification For the combination of DHM and DynPCM, the
of the image sharpness, numerical autofocusing with- respective components were designed to operate at
out mechanical focus realignment is possible [11]. the same wavelength of 532 nm to simplify illumina-
tion. A fiber coupling box (not shown in Figure 1) In the FLIM setup, which was configured to oper-
with three fiber outputs as well as an illumination ate in the wide-field frequency domain, an inexpen-
adapter were developed in close collaboration be- sive blue Light Emitting Diode (LED) (AcuLED
tween the Institute of Applied Physics and the DYO 455 nm) was used as the excitation source and
CeBOP in Munster. A frequency-doubled Nd : YAG was driven electrically at f 80 MHz with a DC bias
laser (Coherent Verdi) was split up in the fiber cou- [18]. A dichroic mirror set (DM2) with 450480 nm
pler using two beamsplitters. For the connection of excitation filter and >515 nm emission filter was
the coupler box and the actual setup of DHM and used to direct the excitation light into the micro-
DynPCM, polarization-maintaining fibers (Thorlabs scope and separate the fluorescence emission onto a
488 PM) were used. The illumination adapter fits gated optical image intensifier (GOI). The GOI is a
onto the microscope condenser mount and intro- high trigger rate gated intensifier (Kentech Instru-
duced the laser light into the illumination path by a ments Ltd. model HRI) that was sinusoidally modu-
polarizing beamsplitter (PBS). In this configuration, lated at the same frequency as the LED drive source
simultaneous illumination at all desired wavelengths with its output phosphor screen being imaged onto a
as well as white light observation was possible. A di- CCD camera (Hamamatsu Orca ER). For each
chroic mirror (DM1) separated the 532 nm from the FLIM image, five separate images were recorded
fluorescence light for FLIM, a 50/50 beamsplitter from the CCD for five different relative phases of
(BS) divided the image information for DHM and the HRI and LED drive signals that were tem-
DynPCM. For DHM, the intermediate image plane porally separated by 2.5 ns. A fluorescent plastic
was imaged onto a CCD (The Imaging Source DMK slide (CHROMA Red) was used as a known stand-
41BF02, Bremen, Germany), where off-axis holo- ard in order to provide calibration for the system.
grams were recorded. The numerical reconstruction The phase lifetime was obtained from the phase Dj
of the quantitative DHM phase contrast images from of the signal in the measured fluorescence images
the digital holograms was performed by spatial relative to the excitation using the calculation t
phase shifting-based reconstruction as reported pre- Dj=2pf .
viously ([10] and references therein). In case of un-
focused imaging numerical auto-focusing was applied
as described in [19]. The refractive index determi-
nation of suspended cells was performed as de- 4. Sample preparation
scribed in [19]. For DynPCM, signal and reference
beam interfered within a photorefractive barium tita- Chinese Hamster Ovary (CHO) cells were routinely
nate crystal that was placed shortly after the Fourier cultured in a humidified atmosphere of 5% CO2/
plane of the imaging lens [14]. Due to the fact that 95% air at 37 C in 25 cm2 Nuncleon culture flasks
DynPCM requires long-time stability for the grating (Fischer, UK) containing 5 ml of modified Eagles
formation in the photorefractive crystal, the imaging medium with 10% foetal calf serum (Sera Labs,
path from the microscope to the DynPCM setup was UK), 20 mg/ml streptomycin, and 20 mg/ml penicillin.
encapsulated with protective tubing. Immediately prior to experimentation, cells were
BIOPHOTONICS
12 M. Esseling et al.: Multimodal biophotonic workstation for live cell analysis
trypsinised, treated with 5 mM di-4-ANEPPDHQ DynPCM highlights the parts of the cell under mo-
(Invitrogen, UK), centrifuged for 10 minutes at tion, e.g. due to the optical forces, for real-time in-
200 rcf, re-suspended in Hanks balanced salt solution spection of the movement and refractive index
(Mg2Ca2) and plated onto 170 mm thick glass changes. In this experiment, only the cell membrane
bottomed dishes (WORLD PRECISION INSTRU- (Figure 2(a)) is visible since it is the only part that
MENTS) which were previously coated with poly-2- moves significantly during the stretching process.
hydroxyethyl-methacrylate (20 mg/ml in ethanol) to The rest of the cell is suppressed due to the stable
prevent adhesion. All solutions were obtained from position in the optical trap. The evaluated DHM im-
Sigma, UK unless otherwise stated. age maps the quantitative phase retardation of the
light inside the cell (Figure 2(b)). Within the meas-
urement accuracy, no variation of the optical thick-
ness during the stretching process could be found.
5. Cell analysis This observation is supported by the fact that most
parts of the cell stay suppressed in phase sensitive
As proof-of-concept experiment to demonstrate the DynPCM imaging. Figure 2(c) displays the map of
workstations capability to combine optical microma- phase lifetimes obtained by FLIM as described
nipulation with all three imaging methods simul- above. In our experiment, the average fluorescence
taneously, cell compartments of a suspended CHO lifetime was 2.27 ns (/0.33 ns). While the ob-
cell, treated with a membrane-staining fluorescent served lifetime did vary over the course of the expe-
dye (di-4-ANEPPDHQ), are trapped with HOT and riment, there was no systematic correlation between
moved to apply stress to the cell as described above. the optical stretching of the cell and the fluorescence
The membrane staining dye is reported to change its lifetime of the membrane dye for the observed cells.
fluorescence lifetime according to the degree of or- We further note that under these degrees of optical
der in the lipid bilayer [18]. Stretching cells has been stretching, the phase-sensitive techniques DynPCM
demonstrated theoretically to reduce the order of li- and DHM did not show significant changes of mor-
pid tails [20], which produces weaker lipid packing phology and integral cellular refractive index proper-
and higher diffusion rates [21] such that the transla- ties.
tional freedom, or lateral mobility, of liquid ordered
membranes is 23 times lower than that of liquid
disordered membranes [22]. Thus a decrease in
fluorescent lifetime of di-4-ANEPPDHQ might be 6. Conclusion
expected following a decrease in lipid order gener-
ated by the process of stretching. Figure 2 sum-
marizes the observations from the first experiment. We have successfully demonstrated the assembly of
a biophotonic workstation, which combines for the
first time four different biophotonic techniques that
have so far only been used in separate setups. Due
to the fact that the wavelengths were carefully sepa-
rated, the imaging modes could all be used at the
same time. The feasibility of this combined approach
was demonstrated with a multimodal observation
of a CHO cell, where DynPCM provided real-time
monitoring of a cell, while FLIM and DHM re-
construction was used to investigate changes in re-
fractive index and membrane lipid order due to
stretching with HOT. Our simple proof-of-principle
experiment shows the potential of the biophotonic
workstation, and we believe that the combined avail-
ability of these complementary techniques will enable
biologists to develop new experiments, for example
spatially probing the biophysical or biochemical
properties with HOT actuated microprobes [23], for
which our biophotonic setup offers all capabilities.
Figure 2 (online color at: www.biophotonics-journal.org)
Simultaneous multimodal imaging of suspended CHO dur- Acknowledgements This project was funded by the Euro-
ing tweezer stretching: (a) DynPCM, (b) DHM, (c) FLIM pean Union in the frame of Photonics4Life. Further funding
image showing the phase lifetime calculated from five from COST Action MP604 is gratefully acknowledged.
phase-delayed acquisitions. KD is a Royal Society-Wolfson Merit Award Holder.
+++ NEW +++ NEW +++ NEW +++ NEW +++ NEW +++ NEW +++ NEW +++
This new handbook covers the world of citing the various tools to solve the
biophotonics not only geographically scientific task, making it of particular
with the editors coming from different value to medical doctors.
continents but also in terms of Divided into several sections, the first
content, since the authors come from part offers introductory chapters on
the whole spectrum of biophotonic the different fields of research, with
basic and applied research. Designed to subsequent parts focusing on the
set the standard for the scientific applications and techniques in various
community, these three volumes break fields of industry and research. The
new ground by providing readers with result is a handy source for scientists
the physics basics as well as the seeking the basics in a condensed
2011. LII, 1132 pages biological and medical background, form, and equally a reference for
390 figures 311 in color, 46 tables. together with detailed reports on recent quickly gathering the knowledge from
Hardcover. technical advances. The Handbook also neighboring disciplines.
ISBN: 978-3-527-41048-4 adopts an application-related approach, Absolutely invaluable for biophotonic
starting with the application and then scientists in their daily work.
Register now for the free WILEY-VCH P.O. Box 10 11 61 69451 Weinheim, Germany
WILEY-VCH Newsletter! Fax: +49 (0) 62 01 - 60 61 84
www.wiley-vch.de/home/pas e-mail: [email protected] http://www.wiley-vch.de