MICROSCOPY RESEARCH AND TECHNIQUE 70:26–33 (2007)
Combining Optical Tweezers and Scanning Probe
Microscopy to Study DNA–Protein Interactions
JURGEN H.G. HUISSTEDE, VINOD SUBRAMANIAM, AND MARTIN L. BENNINK*
Biophysical Engineering Group and MESA þ Institute for Nanotechnology, Faculty of Science and Technology,
University of Twente, AE Enschede 7500, The Netherlands
KEY WORDS
optical tweezers; scanning probe microscopy; DNA–protein interactions; digoxygenin; single molecule force spectroscopy
ABSTRACT
We present the first results obtained with a new instrument designed and built to
study DNA–protein interactions at the single molecule level. This microscope combines optical
tweezers with scanning probe microscopy and allows us to locate DNA-binding proteins on a single
suspended DNA molecule. A single DNA molecule is stretched taut using the optical tweezers,
while a probe is scanned along the molecule. Interaction forces between the probe and the sample
are measured with the optical tweezers. The instrument thus enables us to correlate mechanical
and functional properties of bound proteins with the tension within the DNA molecule. The typical
friction force between a micropipette used as probe and a naked DNA molecule was found to be
<1 pN. A 16 lm DNA molecule with 10–15 digoxygenin (DIG) molecules located over a 90 nm
range in the middle of the DNA was used as a model system. By scanning with an antidigoxygenin
(a-DIG) antibody-coated pipette we were able to localize these sites by exploiting the high binding
affinity between this antibody–antigen pair. The estimated experimental resolution assuming an
infinitesimally thin and rigid probe and a single a-DIG/DIG bond was 15 nm. Microsc. Res. Tech.
70:26–33, 2007. V 2006 Wiley-Liss, Inc.
C
INTRODUCTION
Single-molecule approaches yield new information
about mechanical and dynamic properties of biomolecules. This powerful approach for studying molecular
systems allows the user to assess properties that, using
most other techniques, are not accessible, or that would
be masked in the ensemble average.
Several techniques have been developed to measure
or manipulate these single molecules, including optical
tweezers and atomic force microscopy. Optical tweezers
(OT) is a technique that is widely used for single molecule force spectroscopy capable of measuring forces up
to 100 pN with subpiconewton resolution, which makes
the technique excellent for measuring forces involved
in stretching individual molecules such as DNA, and
proteins or complexes thereof, (Bennink et al., 2001;
Cui and Bustamante, 2000; Kellermayer et al., 1997)
and measuring forces involved in the action of molecular motors (Abbondanzieri et al., 2005; Mehta et al.,
1999).
Scanning probe microscopy (SPM), in particular the
atomic force microscopy (AFM), is capable of obtaining
high resolution images of single biomolecules (Bahatyrova et al., 2004; Czajkowsky and Shao, 1998; Nikova
et al., 2004; Viani et al., 2000). AFM is also used for single molecule force spectroscopy studies of various biomolecular complexes (Hinterdorfer et al., 1996; Lee
et al., 1994; Muller et al., 1999; Oesterhelt et al., 2000)
with a typical force resolution of >5 pN.
In many OT and AFM force spectroscopy experiments the change in elastic and mechanical properties
of DNA as a result of protein binding is studied. In
C
V
2006 WILEY-LISS, INC.
these studies, however, it is not possible to accurately
localize these proteins in order to correlate their position with the force data obtained. To localize proteins
on a single DNA molecule we have developed a new
instrument based on OT combined with scanning probe
microscopy (SPM).
Figure 1a presents the concept of the scanning probe
optical tweezers (SPOT) microscope. A single DNA molecule is stretched taut using the optical tweezers, while
a sharp probe is scanned along the molecule. Interactions between the probe and the DNA and proteins are
sensed by the optical trap in contrast to conventional
SPM techniques, where the probe acts as a sensor. One
of the controllable parameters influencing the magnitude of the interactions between the probe and the
sample is the indentation d of the DNA molecule as
indicated in Figure 1b, which is typically in the range
of 0–200 nm for localization experiments. An important
advantage of this microscope with respect to current
SPM techniques is that the DNA molecule is not influenced by any interfering surface except for the two
beads to which the DNA ends are attached. This configuration allows the study of DNA–protein interactions
in conditions where the proteins can move freely. Fur*Correspondence to: Dr. Martin L. Bennink or Prof. Vinod Subramaniam, Biophysical Engineering Group and MESA þ Institute for Nanotechnology, Faculty
of Science and Technology, University of Twente, PO Box 217, 7500 AE Enschede,
The Netherlands; E-mail: [email protected] or [email protected]
Received 5 April 2006; accepted 14 July 2006
DOI 10.1002/jemt.20382
Published online 1 November 2006 in Wiley InterScience (www.interscience.
wiley.com).
OPTICAL TWEEZERS, SPM, AND DNA–PROTEIN BINDING
27
Fig. 1. (a) Principle of measurement. A single DNA molecule is
stretched taut while a lm-sized probe is scanning along the molecule
in order to \feel" the individual proteins. Interactions between the
probe and the proteins are detected by the optical tweezers and in
combination with the probe position, allows the accurate localization
of these proteins on the DNA molecule. (b) When a molecule stretched
to a length L is indented with the probe over a distance d an additional force Fdna is generated in the molecule and is a function of the
probe position x along the molecule. The magnitude of the interactions between the probe and the sample is dependent on the indentation d. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
thermore this microscope enables us to study the effect
of tension in the DNA molecule on the functional properties of proteins.
Here we demonstrate the feasibility of the SPOTmicroscope by using an antidigoxygenin (a-DIG) coated
micropipette to localize digoxygenin (DIG) molecules
bound over a range of 90 nm in the middle of a 16 lm
long DNA molecule. The high binding affinity of this
antibody–antigen linkage results in strong local interactions between the probe and the DNA molecule. In
addition we use this configuration to determine the
friction force between the probe and the naked DNA.
between a precleaned microscope glass and a coverslip.
This creates a flow channel with dimensions of 200 lm
3 5 mm 3 50 mm. In the microscope glass three holes
were powder-blasted (Wensink et al., 2000). Two holes
(2 mm in diameter) at the end of the flow channel act
as entry and exit points for the flow channel. Inlet and
outlet tubes necessary to flow the desired solutions in
and out were attached on top of these. An additional
entry-hole in the middle of the cell was created to enable the injection of the scanning probe. The powderblasting procedure used created an entry-hole that has
a conical shape with a smallest diameter of 200 lm at
the bottom side as indicated in Figure 3. The other two
holes at either end of the flow channel were drilled after powder-blasting with a diamond drill to provide a
cylindrical hole.
A square glass capillary with an inner diameter of
50 3 50 lm2 (VitroCom, Mountain Lakes, NJ) was
aligned in between the coverslip and the microscope
glass such that a suction pipette for the bead immobilization can be injected into the flow cell with the pipette
end located below the entry-hole. This complete sandwich was heated up shortly to 608C using a heating
plate to provide a waterproof sealing. To prevent leakage through the center hole a small reservoir was
placed on top of the outlet hole. Because the diameter
of the outlet was 10 times larger than the center entryhole, the fluid resistance is highest at the center hole
and fluid first leaves the flow cell through the outlet. A
small tube placed in this reservoir actively sucks away
fluid whenever it reaches the tube.
MATERIALS AND METHODS
Experimental Setup
To create a single beam gradient optical trap, a high
NA objective was used (Leica N PLAN, NA 1.20,
Wetzlar, Germany), resulting in a strongly focused
laser beam. The proximity of the scanning probe to the
DNA molecule stretched taut by the OT (as schematically depicted in Fig. 1) interferes with the light transmitted through the bead as a force signal, as is common
in a conventional OT. Therefore the basis of the SPOTmicroscope is a reflection-based OT instrument as
described elsewhere (Huisstede et al., 2005). The complete configuration of the scanning probe optical tweezers microscope is depicted in Figure 2.
The Flow Cell
A flow cell is required to build up a bead–DNA–bead
construct (Fig. 3). A single k-DNA molecule (16.4 lm)
end-labeled with biotin is suspended between two
streptavidin-coated 2.67 lm polystyrene beads (Polysciences, Warrington, PA) according to a procedure previously described in literature (Bennink et al., 1999).
One bead was immobilized on a suction pipette integrated in the flow cell and one bead was held by the optical trap. Moving the micropipette with respect to the
optical trap using a piezo-controlled XYZ-stage (P-509,
Physik Instrumente, Karlsruhe, Germany) allowed
stretching of the DNA molecule.
To facilitate the approach of the stretched DNA molecule by a scanning probe some modifications were
made to the existing flow cell design (Bennink et al.,
1999). Two layers of thermoplastic laboratory film (Parafilm1) with a channel cut out were sandwiched
Microscopy Research and Technique DOI 10.1002/jemt
Scanning Probe
In the SPOT-microscope configuration, and in contrast to SPM techniques, the probe is not used as a sensor, but as an actuator, which demands a thin and rigid
probe. The optical tweezers in this instrument are the
force-measuring element. In the experiments presented here micropipettes were used as scanning
probes (pipette probe in Fig. 3). Borosilicate micropipettes were pulled from 1.2 mm outer diameter and
0.94 mm inner diameter capillaries (Harvard Apparatus GC120TF-15, Holliston, MA) using a Sutter P-87
micropipette puller (Novato, CA) to obtain end diameters in the range of 1–2 lm.
28
J.H.G. HUISSTEDE ET AL.
Fig. 2. Schematic layout of the scanning probe optical tweezers
(SPOT) microscope. The inset shows the bead–DNA–bead construct
and the scanning probe. A beam expander (BE) creates a laser beam
with a diameter of 1 cm that overfills the back-aperture of a 1003
high NA objective. The laser power at this aperture can be tuned by a
half-wave plate (k/2) and a polarizing beam splitter cube (PBS). A
beam splitter (BS1) directs the backscattered light onto a position
sensitive detector (PSD) where a second beam splitter in the detection
path (BS2) enables visualization of the reflection pattern on a CCD
camera (CCD2). A quarter-wave plate (k/4) placed in front of the
objective converts the incident p-polarized laser light into circularly
polarized light, providing an equal trap stiffness in both lateral directions (Worland et al., 1996). A halogen lamp provides white light illumination for optical microscopy imaging of the trapped bead via a dia-
chronic mirror (DM) on a second CCD camera (CCD1). A short-pass
filter (SPF) in front of the camera blocks the 1064 nm laser light. The
probe for scanning along the DNA molecule is fixed to a piezo tube
that on its turn is mounted on a translation stage (TS2) to control the
probe position in the Y and Z direction and a separate translation
stage (TS1) with an integrated piezo stack to accurately control the
position in the X direction and thus the indentation of the DNA molecule by the probe. The integrated piezo stack was used to keep hysteresis of the piezo-tube constant during the scanning experiments. The
high NA objective and a 103 objective were mounted on a sliding
mechanism to be able to switch objectives where the 103 objective
was used to provide a larger field of view required to be able to inject
the scanning probe into the flow cell.
Hysteresis of the Piezo-Tube
A piezo-tube was used to scan the probe back and
forth along the DNA molecule while a piezo stack independently controlled the indentation. Piezo tubes exhibit hysteresis, which makes independent calibration
of the probe position necessary.
The actual position of the tube was determined with
a LED mounted to the end of the piezo tube in combination with a position sensitive detector (PSD, DL100-7KER, Pacific Silicon Sensor, Westlake Village, CA)
mounted to the optical table. To convert the PSD signal
from volts to microns, optical microscope images of a
micropipette clamped to the piezo tube, as for the scanning experiments, were obtained with the high NA
objective as a function of the applied voltage in a direct
current (DC) measurement. A centroid algorithm
(Wuite et al., 2000) deduced the position of the probe in
microns. Since the hysteresis is a function of the scan
speed and the scan range, for each experiment the hysteresis was determined with the settings used. These
curves were used to correct the obtained data in the
scanning experiments.
Silanization of Pipettes
Freshly pulled pipettes were placed upright in a
small glass beaker (25 mL), which was placed on a
glass Petri dish. The small glass beaker with the pipettes was covered under a 100 mL glass beaker and
baked in an oven at 2008C for at least 4 h. All glassware (except the pipettes) was presilanized by placing
them in a 2% APTES (3-aminopropyltriethoxysilane,
Sigma, St. Louis, MO) solution in 95% aqueous acetone
for 1 h. If the pipettes were not freshly pulled they
were cleaned by placing them in a solution of 65%
HNO3 (nitric acid) for at least a few hours and rinsed
several times with acetone.
After the oven was cooled TMSDMA (N,N-dimethyltrimethylsilyamine, Fluka, Seelze, Germany) was injected under the 100 mL beaker through a syringe. The
TMSDMA almost immediately vaporized and the vapor
produced a homogeneous hydrophobic silane coating
on the pipettes. After 30 min the beaker was opened for
a short period of time to allow any residual vapor to
escape. The silanized pipettes were baked at 2008C
overnight.
Microscopy Research and Technique DOI 10.1002/jemt
OPTICAL TWEEZERS, SPM, AND DNA–PROTEIN BINDING
29
Fig. 3. Schematic representation of the flow cell. An additional
entry-hole was drilled at the position of the optical trap and the suction pipette used to immobilize a
2.6 lm bead. This hole allows
injection of a probe used for scanning along a stretched dsDNA
molecule to localize DNA-bound
molecules. [Color figure can be
viewed in the online issue, which
is available at www.interscience.
wiley.com.]
a-DIG Coated Pipette
For the preparation of a-DIG coated pipettes small
volumes are essential to prevent the use of large
amounts of antibody. Glass capillaries with an inner diameter of 1.5 mm were cleaned by placing them for 1 h
in a 65% HNO3 solution. Subsequently they were
rinsed several times with Milli-Q and finally dried
using nitrogen. Freshly pulled micropipettes (1–2 lm
tip diameter) were silanized with TMSDMA and
inserted into the capillaries. A small rubber cap placed
at the end of the capillary held the pipette at its backend.
From a 1 mL syringe (BD Plastipak, Franklin
Lakes, NJ) the rubber cap was taken off from the
plunger. Subsequently we placed the rubber cap
upside down in the syringe tube, creating a small but
long container. 200 lL TE-buffer with 100 lg/mL aDIG was injected in this container. Upon inserting the
capillary with the micropipette in this container the
capillary was filled with the a-DIG solution due to
capillary forces. After filling, the capillaries were
taken out and closed at their open ends with another
rubber cap and stored at 48C. Shortly before the
experiment an a-DIG coated pipette was taken out of
the capillary and rinsed several times with TE-buffer.
Next it was mounted into the microscope and directly
injected into the flow cell.
Digoxygenin-Functionalized DNA
Preparation
Bacteriophage k-DNA was end-labeled with biotin as
described (Bennink et al., 1999). This k-DNA was
digested with two enzymes, SacI (NEB, Ipswich, UK)
and XbaI (NEB) to create two long fragments, one with
a SacI overhang (22.6 kbp) and one with an XbaI overMicroscopy Research and Technique DOI 10.1002/jemt
Fig. 4. Schematic representation of a DNA molecule that is biotinylated at each end. In approximately the middle of the molecule there
is a 268 bp fragment including DIG molecules in a ratio 1:20, resulting in 10–15 DIG molecules available over a range of 90 nm.
hang (24.5 kbp). These fragments were purified by electro-elution (Bio-trap, Chromtech, Cheshire, UK).
A 391 bp double-stranded DNA fragment labeled
with digoxygenin was created by PCR using digoxygenin labeled dUTPs (Roche, Basel, Switzerland). Since
both dUTP and dTTP binds to dATP some of the dTTPs
can be replaced for DIG-dUTP. The 391 bp fragment
was incubated with DIG-dUTP and dTTP (DIGdUTP:dTTP ratio: 1:20) and dATP, dGTP, dCTP, and
Klenow DNA polymerase for PCR. One in 20 dTTPs
was replaced by a labeled dUTP. The 391 bp fragment
was digested with SacI and XbaI to create a 268 bp
fragment with a SacI overhang at one end and an XbaI
overhang at the other end. Finally the fragment was
purified by electro-elution. The two long fragments
(22.6 and 24.5 kbp) were mixed together with the 268
bp fragment at a ratio of 1:1:20 for annealing. The mixture was heated and cooled down slowly and finally
ligated with T4 DNA ligase (NEB) at 168C. A schematic
structure of the resulting DNA molecules is shown in
Figure 4, where the lengths of the different regions are
indicated. The contour length of the complete construct
is 16.1 lm.
30
J.H.G. HUISSTEDE ET AL.
Fig. 5. (a) A typical force-extension curve of a single dsDNA molecule (16.4 lm). After acquiring a force-extension curve to ensure one
molecule is suspended it was stretched to the enthalpic linear force regime (10–50 pN) indicated by the arced area for scanning experiments. (b) Histogram of the friction force values between a 170 nm
pipette and a single DNA molecule that is stretched to 30 pN. At each
indentation the friction forces were determined for 50 line traces over
a range of 1.25 lm containing 256 points around the middle of the
DNA molecule by subtracting the traces and corresponding retraces.
Each histogram therefore contains 12,800 points from which the
mean friction force was calculated.
RESULTS AND DISCUSSION
Friction Forces Between the Probe
and Naked DNA
In Figure 5a, a typical force-extension curve is shown
for dsDNA, always acquired before starting a scanning
experiment to ensure that a single molecule was suspended in between the two polystyrene beads. For
scanning experiments the molecule was stretched to a
force of 10–50 pN, where it behaves as a linear Hookean spring (enthalpic regime) with a typical stiffness
of 66 pN/lm for k-DNA (Smith et al., 1996). An
uncoated pipette with a tip diameter of 170 nm
(TIP01TW1F-L, WPI, Sarasota, FL) was scanned back
and forth along a single stretched DNA molecule that
was prestretched at a force of 30 pN. The total force
within the DNA molecule was recorded for the probe
moving away (trace) from the trapped bead and for the
probe moving towards the bead (retrace). Subtracting
the traces and retraces yields the friction force multiplied by a factor 2 (2Ffric ¼ FtraceFretrace).
Macroscopically the friction force between two solids
moving with respect to each other is dependent on the
normal force N and the dynamic friction coefficient ls
according to Coulomb’s law (Bowden and Tabor, 1950)
indentation 50 (re)traces were analyzed. After subtracting the trace and retraces the resulting force values
(2Ffric) were plotted in a histogram as shown in Figure 5b
and also averaged and divided by 2 to calculate the mean
friction force at the corresponding indentation.
The friction force values found were corrected for an
offset attributed to the drag force induced by the moving
pipette. The offset values at the three indentations were
determined for the beads without a suspended DNA
molecule and found to be nearly constant with a value of
0.08 pN. For no indentation the mean friction force
was 0.0 pN. At an indentation of 150 nm the mean friction force was 0.24 pN and at 500 nm 0.55 pN.
F ¼ ls N:
The normal force is a function of the indentation d
(Fig. 1b) and the position of the probe with respect to
the trapped bead. Therefore recordings were obtained
for three different indentations, 0, 150, and 500 nm
with a total scan range of 5 lm and a scan speed of
50 lm/s, which is the typical speed used in the scanning experiments. The probe was manually positioned
such that its center position approximately coincides
with the center of the DNA molecule. To minimize the
influence of the probe position on the normal force, a
small range of 1.25 lm from the obtained data around
the center position of the DNA was analyzed. For each
Localization of DIG Molecules on a
Single DNA Molecule
As a next step towards single protein detection a
DNA construct was prepared with 10–15 DIG molecules spread over a DNA length of 90 nm around the
middle of the molecule. To localize these DIG-molecules
a micropipette with a tip diameter of 2 lm was coated
with a-DIG to achieve a large interaction between the
probe and the sample, since DIG forms a strong noncovalent bond with the antibody a-DIG mediated by a
combination of van der Waals, hydrogen and ionic
bonds.
A single DNA molecule with DIG sites was suspended between two beads and stretched to 25 pN. It
was scanned with an a-DIG coated pipette over a range
of 6.0 lm around the middle of the molecule at a speed
of 50 lm/s. Upon passing the DIG-sites an increase in
the force measured by the OT was observed, a result of
the interactions between DIG and a-DIG molecules.
Depending on the binding strength and loading rate
(in this case in the order of 3–4 3 103 pN/s) the interactions will be disrupted at a certain force (Evans and
Ritchie, 1997).
Figure 6a shows a three-dimensional representation
where consecutive force traces are plotted next to each
Microscopy Research and Technique DOI 10.1002/jemt
OPTICAL TWEEZERS, SPM, AND DNA–PROTEIN BINDING
31
Fig. 6. (a) A 3D representation where consecutive force traces are
plotted next to each other that were acquired by scanning along a
DNA molecule with 10–15 DIG-sites located over a range of 90 nm in
the middle of the molecule with an a-DIG coated micropipette. In one
axis the position of the probe along the DNA is plotted, while in the
orthogonal axis the number of the line trace is plotted. In the z direction the force added as a result of scanning is plotted. (b) A few indi-
vidual line traces and retraces showing the additional force generated
in the DNA molecule by scanning with the a-DIG coated pipette. It
clearly shows the stick-slip behavior in a range of 1–2 lm presumably
caused by the binding of a-DIG molecules on the probe pipette to
DIG-sites on the DNA molecule. [Color figure can be viewed in the
online issue, which is available at www.interscience.wiley.com.]
other. On one axis the position of the probe along the
DNA is plotted, while on the orthogonal axis the number of the line trace is plotted. In the z direction the
force is plotted. Clear peaks in force are visible in the
area where the DIG-sites are located. The signal
visualized here was DC-filtered to enable electronic
amplification to prevent limitations in the resolution
due to digitization of the analog detection signals. Consequently the offset force of 25 pN is not observed. In
Figure 6b several individual traces and retraces are
shown, indicating the stick-slip motion within an area
Dx1 % Dx2 % 1–1.5 lm caused by a-DIG/DIG binding.
The size of the area within which the stick-slip motion
is visible is a result of the convolution of the contact
area of the pipette and the area in which the DIG-sites
are located. Depending on the orientation of the DNA
molecule suspended between the beads, there is a different DNA length in between the DIG-sites and the
trapped bead, which can be 7.7 lm or 8.3 lm (Fig. 4).
The relatively large area of Dx1 % Dx2 caused by the
large size of the micropipette makes it impossible to
draw any conclusions on the orientation of the DNA
molecule.
The mean value of the maximum amplitude for all the
force peaks was 11 pN. With an offset of 25 pN the average rupture force is 36 pN where no conclusions can be
drawn about whether we have single or multiple bonds
that are disrupted. Several groups have measured the
binding forces of ligand–receptor and antibody–antigen
linkers using atomic force microscopy (AFM) and OT.
The streptavidin-biotin linker has been most frequently
investigated, and is one of the strongest noncovalent
bonds known (Evans and Ritchie, 1997; Florin et al.,
1994; Hinterdorfer et al., 1996; Lee et al., 1994; Lo et al.,
2001; Merkel et al., 1999; Ota et al., 2005). In these studies it was found that the rupture force was slightly dependent on the loading rate. For streptavidin-biotin the
individual rupture force was found to be 150 pN for a
loading rate of 104 pN/s. For a lower loading rate the
rupture force was found to be lower. For a-DIG/DIG a
rupture force of 30–40 pN was found for a loading rate of
103 pN/s (Neuert et al., 2006), which corresponds well to
our data, and would suggest that we indeed observe single a-DIG/DIG binding events.
The experiments discussed here are not sufficient to
extract statistically relevant conclusions on the localization accuracy that can be achieved and its dependence on the interactions between the probe and the
sample. However, to give an indication of the localization accuracy we provide an estimation where we
assumed an infinitesimally thin and rigid probe.
For the detection of DIG molecules the additional
force in the DNA molecule caused by the interaction
between the probe and the DIG molecules has to overcome the (thermal) force fluctuations of the trapped
bead. For the experiments shown the detection bandwidth B was 32 kHz, which is far beyond the cornerfrequency of the Lorentzian describing the response of
the trapped bead (Svoboda and Block, 1994). Therefore,
according to the equipartition theorem, the force
fluctuations
can be expressed as Fsd ¼ ktr xsd ¼
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ktr kb T=ðktr þ kdna Þ with xsd the standard deviation of
the trapped bead position, kb Boltzmann’s constant, T
the temperature in Kelvin, ktr the trap stiffness, and
kdna the stiffness of the DNA molecule (Gates and
Schmidt, 1998). Fsd is the standard deviation of the
force signal. With a trap stiffness ktr of 378 pN/lm and
a stiffness of the DNA molecule of 59 pN/lm for forces
between 10–50 pN in our experiment the theoretical
force resolution is 1.2 pN. Experimentally these force
fluctuations are in the order of 1.5 pN as derived from
the individual line traces plotted in Figure 6b, where
low-frequency noise as a result of interference of the
backscattered light from the probe and the trapped
bead was ignored.
The force increase in the DNA molecule caused by
the interaction between the probe and the DIG molecules has to overcome the fluctuations of 1.5 pN. This
minimum required force increase can be expressed as a
distance over which we have to displace the DIG-DNA
Microscopy Research and Technique DOI 10.1002/jemt
32
J.H.G. HUISSTEDE ET AL.
complex denoted as the localization accuracy using the
local slope of the force-distance traces at the position
where a-DIG binds to the DIG in the DNA molecule.
For the local slope an average value of 100 6 20 pN/lm
was found (based on 19 line traces) and thus an estimated experimental localization accuracy of 1.5 pN/
(100 pN/lm) ¼ 15 nm assuming an infinitesimally thin
and rigid probe.
Assuming the bond between DIG and a-DIG to be
noncompliant, the slope represents the stiffness of the
DNA between the trapped bead and the point where
the DNA is attached to the probe, which is either
22.6 kbp (7.7 lm; Fig. 4) or 24.5 kbp (8.3 lm), depending on the orientation of the DNA molecule. Note that
the stiffness of DNA molecule in the enthalpic force regime (10–50 pN) is a function of its length. From the
force-extension curve of the individual molecule
(47.4 kbp (16.1 lm) the slope at a tension of 25 pN was
59 6 2 pN/lm. For a rigid connection between the
probe and the DIG molecules the effective stiffness was
expected to be (16.1/7.7) 3 59 ¼ 123 pN/lm or (16.1/
8.3) 3 59 ¼ 114 pN/lm. The experimentally observed
stiffness of 100 6 20 pN/lm is thus slightly lower than
the expected values, and the relatively large error indicates the variance in compliance of the a-DIG/DIG
bonds, where the number of bonds per interaction is
unknown.
The range over which single interactions took place
(from the initiation point to the sudden drop in force)
for a line trace was typically in the order of 110 nm.
This range is not to be confused with the total range
over which all the interactions took place, which is a
convolution of the probe diameter and the 90 nm range
with DIG-sites.
PERSPECTIVES
We have presented the first results obtained with the
scanning probe optical tweezers microscope where we
have shown that we are able to localize DIG molecules
in a range of 90 nm on the 16 lm long DNA using an aDIG coated pipette. Future experiments have to point
out the detection limits such as the minimum required
pretension in the molecule, which is important for the
initial DNA–protein complex configuration, the achievable localization accuracy and its dependence on trap
stiffness, polymer stiffness, probe configuration, and
the interaction forces.
In the instrumental design of the SPOT-microscope
there are possible ways to improve the detection. The
positional fluctuations of the trapped bead are a function of the effective stiffness of the DNA molecule.
Using shorter DNA molecules the accuracy with which
a protein can be detected is improved, but a trade-off
has to be made between this improvement and the possibility of the use of shorter molecules from a practical
point of view. If the pipette becomes too close to the
trapped bead, the bead can be pushed out of the trap or
the pipette influences the force measurement. A safe
distance between the probe and each of the two beads
is in the order of 3 lm. Shorter molecules will therefore
strongly limit the scan range.
The implementation of a second optical trap instead
of the glass micropipette to hold the bead has some
advantages. This provides the ability to cross-correlate
both detector signals from the two beads. By tracking
the statistical correlation between movement of both
beads, events that may not be visible in the position
data of each bead can be detected (Mehta et al., 1997).
In addition, two traps allow the bead–DNA–bead system to be moved around freely. In this case an integrated probe is an option, which will greatly reduce the
time needed for setting up the experiment and allows
the possibility of a transmission-based detection
scheme.
We expect that this microscope will have broad
applicability to the study of DNA–protein interactions
by providing a direct way to correlate force spectroscopy data with the location of DNA-binding proteins or
protein complexes and therefore allow the study of the
influence of tension in the DNA molecule on the functionality of proteins. An advantage of the microscope is
that the motion of proteins is not obstructed by nearby
surfaces, which more closely resembles the in vivo situation in comparison with conventional scanning probe
microscopy, where a supporting surface is needed. A
drawback is that a minimum pretension is required for
the detection, which can prohibit the investigation of
proteins whose functional properties are altered at this
minimum tension.
The application of the microscope will not be restricted only to DNA–protein interactions, but can be
applied also to other types of (bio)polymers such as
RNA, microtubules, and actin interacting with proteins. Important parameters for deciding whether this
is feasible are the rigidity and the length of the polymer, and the ability to provide biochemical linkers to
the ends of the polymer required for suspending the
molecule between the two beads.
REFERENCES
Abbondanzieri EA, Greenleaf WJ, Shaevitz JW, Landick R, Block SM.
2005. Direct observation of base-pair stepping by RNA polymerase.
Nature 438:460–465.
Bahatyrova S, Frese RN, van der Werf KO, Otto C, Hunter CN, Olsen
JD. 2004. Flexibility and size heterogeneity of the LH1 light harvesting complex revealed by atomic force microscopy: Functional
significance for bacterial photosynthesis. J Biol Chem 279:21327–
21333.
Bennink ML, Scharer OD, Kanaar R, Sakata-Sogawa K, Schins JM,
Kanger JS, de Grooth BG, Greve J. 1999. Single-molecule manipulation of double-stranded DNA using optical tweezers: Interaction
studies of DNA with RecA and YOYO-1. Cytometry 36:200–208.
Bennink ML, Leuba SH, Leno GH, Zlatanova J, de Grooth BG, Greve J.
2001. Unfolding individual nucleosomes by stretching single chromatin fibers with optical tweezers. Nat Struct Biol 87:606–610.
Bowden FP, Tabor D. 1950. Friction and lubrication of solids, Part 1.
New York: Oxford University Press.
Cui Y, Bustamante C. 2000. Pulling a single chromatin fiber reveals
the forces that maintain its higher-order structure. Proc Natl Acad
Sci USA 97:127–132.
Czajkowsky DM, Shao Z. 1998. Submolecular resolution of single
macromolecules with atomic force microscopy. FEBS Lett 430:51–
54.
Evans E, Ritchie K. 1997. Dynamic strength of molecular adhesion
bonds. Biophys J 72:1541–1555.
Florin EL, Moy VT, Gaub HE. 1994. Adhesion forces between individual ligand–receptor pairs. Science 264:415–417.
Gittes F, Schmidt CF. 1998. Thermal noise limitations on micromechanical experiments. Eur Biophys J 27:75–81.
Hinterdorfer P, Baumgartner W, Gruber HJ, Schilcher K, Schindler H.
1996. Detection and localization of individual antibody–antigen recognition events by atomic force microscopy. Proc Natl Acad Sci USA
93:3477–3481.
Huisstede JHG, van der Werf KO, Bennink ML, Subramaniam V.
2005. Force detection in optical tweezers using backscattered light.
Opt Express 13:1113–1123.
Microscopy Research and Technique DOI 10.1002/jemt
OPTICAL TWEEZERS, SPM, AND DNA–PROTEIN BINDING
Kellermayer MSZ, Smith SB, Granzier HL, Bustamante C. 1997.
Folding-unfolding transitions in single titin molecules characterized with laser tweezers. Science 276:1112–1116.
Lee GU, Kidwell DA, Colton RJ. 1994. Sensing discrete streptavidin biotin interactions with atomic-force microscopy. Langmuir 10:354–357.
Lo YS, Zhu YJ, Beebe TP. 2001. Loading-rate dependence of individual ligand–receptor bond-rupture forces studied by atomic force microscopy. Langmuir 17:3741–3748.
Mehta AD, Finer JT, Spudich JA. 1997. Detection of single-molecule
interactions using correlated thermal diffusion. Proc Natl Acad Sci
USA 94:7927–7931.
Mehta AD, Rief M, Spudich JA, Smith DA, Simmons RM. 1999. Singlemolecule biomechanics with optical methods. Science 283:1689–1695.
Merkel R, Nassoy P, Leung A, Ritchie K, Evans E. 1999. Energy landscapes of receptor-ligand bonds explored with dynamic force spectroscopy. Nature 397:50–53.
Muller DJ, Baumeister W, Engel A. 1999. Controlled unzipping of a
bacterial surface layer with atomic force microscopy. Proc Natl Acad
Sci USA 96:13170–13174.
Neuert G, Albrecht C, Pamir E, Gaub HE. 2006. Dynamic force spectroscopy of the digoxigenin-antibody complex. FEBS Lett 580:505–509.
Nikova DN, Pope LH, Bennink ML, van Leijenhorst-Groener KA, van
der Werf K, Greve J. 2004. Unexpected binding motifs for subnu-
Microscopy Research and Technique DOI 10.1002/jemt
33
cleosomal particles revealed by atomic force microscopy. Biophys J
87:4135–4145.
Oesterhelt F, Oesterhelt D, Pfeiffer M, Engel A, Gaub HE, Muller DJ.
2000. Unfolding pathways of individual bacteriorhodopsins. Science
288:143–146.
Ota T, Sugiura T, Kawata S. 2005. Rupture force measurement of biotin–streptavidin bonds using optical trapping. Appl Phys Lett 87.
Smith SB, Cui YJ, Bustamante C. 1996. Stretching individual DNA
molecules with force-measuring laser tweezers. Abstr Pap Am
Chem Soc 212:40-Phys.
Svoboda K, Block SM. 1994. Biological applications of optical forces.
Annu Rev Biophys Biomol Struct 23:247–285.
Viani MB, Pietrasanta LI, Thompson JB, Chand A, Gebeshuber IC,
Kindt JH, Richter M, Hansma HG, Hansma PK. 2000. Probing protein–protein interactions in real time. Nat Struct Biol 7:644–647.
Wensink H, Jansen HV, Berenschot JW, Elwenspoek MC. 2000. Mask
materials for powder blasting. J Micromech Microeng 10:175–180.
Wohland T, Rosin A, Stelzer EHK. 1996. Theoretical determination of
the influence of the polarization on forces exerted by optical tweezers. Optik 102:181–190.
Wuite GJL, Davenport RJ, Rappaport A, Bustamante C. 2000. An
integrated laser trap/flow control video microscope for the study of
single biomolecules. Biophys J 79:1155–1167.