Measurement of the adhesive force between a single Klebsiella
pneumoniae type 3 fimbria and collagen IV using optical tweezers
Bo-Jui Chang
a,1, Ying-Jung Huang
b,1, Chia-Han Chan
c, Long Hsu
c,*, Hwei-Ling Peng
b,
Hwan-You Chang
d, Tri-Rung Yew
e, Cheng-Hsien Liu
f, Sien Chi
a,gaDepartment of Photonics and Institute of Electro-Optical Engineering, National Chiao Tung University, Hsinchu City 300, Taiwan bDepartment of Biological Science and Technology, National Chiao Tung University, Hsinchu City 300, Taiwan
cDepartment of Electrophysics, National Chiao Tung University, Hsinchu City 300, Taiwan
dInstitute of Molecular Medicine and Department of Life Sciences, National Tsing Hua University, Hsinchu City 300, Taiwan eDepartment of Materials Science and Engineering, National Tsing Hua University, Hsinchu City 300, Taiwan
fDepartment of Power Mechanical Engineering, National Tsing Hua University, Hsinchu City 300, Taiwan gDepartment of Electrical Engineering, Yuan Ze University, Zhongli City, Taoyuan County 32003, Taiwan
Received 22 August 2006 Available online 12 September 2006
Abstract
Type 3 fimbriae are important adhesive filaments that assist Klebsiella pneumoniae to establish an infection. Different MrkD adhesin
variants on the fimbriae are known to display distinct adherence capability for the bacteria to bind extracellular matrix proteins,
although the difference has not been determined physically. For this reason, the adhesive force between type 3 fimbriae and collagen
IV were measured using optical tweezers. The measured force data displayed a periodic histogram thus Fourier analysis was applied
to group it to extract the adhesive force of a single molecular pair. Specifically, we showed that grouping should begin with an offset
at the first half of the period. Finally, we first present the adhesive force between each mrkD
V2-, mrkD
V3-, and mrkD
V4-expressed fimbriae
and collagen IV is 2.03, 3.79, and 2.87 pN, respectively. This result can be referred to further research on mrkD allelic effect on bacteria
infection.
2006 Elsevier Inc. All rights reserved.
Keywords: Adhesive force; Type 3 fimbriae; Grouping offset; Klebsiella pneumoniae; Optical tweezers; Fourier analysis
Klebsiella pneumoniae is an opportunistic pathogen that
causes complicated urinary tract infection, pneumonia,
septicemia, and liver abscess in immunocompromised
patients
[1,2]
. The ability of many K. pneumoniae clinical
isolates to colonize respiratory and urinary epithelia has
been attributed to the presence of many adhesive
mole-cules, including type 3 fimbriae
[3]
. The fimbriae, encoded
by a multigene operon, contain the major subunit MrkA
that forms the shaft of the fimbriae and MrkD adhesin
[4]
. On the tip of the fimbriae, MrkD adhesin is responsible
for mediating erythrocyte agglutination in a mannose
inde-pendent way
[5]
and could affect biofilm formation of the
bacteria
[6]
. Although the receptor for the adhesin has
yet to be identified, in vitro research indicates that bacteria
expressing type 3 fimbriae are able to bind to type IV and V
collagen, which may be exposed on exfoliated and denuded
epithelial surfaces during infections
[7]
.
Our analysis of K. pneumoniae clinical isolates from
Tai-wan has revealed the presence of four mrkD alleles, namely
mrkD
v1(GenBank Accession No.
AY225462
), mrkD
v2(GenBank Accession No.
AY225463
), mrkD
v3(GenBank
Accession No.
AY225464
), and mrkD
v4(GenBank
Acces-sion No.
AY225465
). To investigate the effect of the allelic
variation on binding activity of the fimbriae, a type 3
0006-291X/$ - see front matter 2006 Elsevier Inc. All rights reserved.doi:10.1016/j.bbrc.2006.08.190 *
Corresponding author. Fax: +886 3 5131241. E-mail address:[email protected](L. Hsu). 1 These authors contributed equally to this work.
www.elsevier.com/locate/ybbrc Biochemical and Biophysical Research Communications 350 (2006) 33–38
fimbriae display system carrying, respectively, each of the
MrkD adhesin variants was constructed in Escherichia coli
JM109. The availability of these type 3 fimbriae-producing
E. coli strains provides a useful mean to address many
fun-damental
questions
concerning
bacteria–extracellular
matrix protein interaction.
The adhesion strength of a fimbria to its target
mole-cules is conventionally determined by incubating the testing
bacteria to a substrate such as collagen and counting the
number of the bacteria remaining attached to the substrate
after several washes
[8]
. However, because the number of
fimbrial filaments varies considerably among different
bac-terial cells and in different growth conditions, this assay
could only determine the average binding activity between
a given bacterial population and the target molecules. To
determine precisely the direct interacting force between a
fimbria and its target molecules, more sophisticated
tech-niques and strategies must be adopted.
Among many techniques that may be used to investigate
biomolecule interaction, optical tweezers
[9]
have been
thought to be one of the most effective methods and has
been applied successfully in related questions
[10–14]
. By
detaching a biomolecule-coated bead from either another
biomolecule-coated surface or cell, the adhesive force of
the biomolecular pair is measured. A major problem that
is commonly encountered is that the adhesive force
magni-tude of each biomolecular pair may be different from
oth-ers, and some of the measured forces could be a result of
several molecular pairs. This creates the possibility that a
force data measured from a group of molecular pairs
dis-plays a continuous and periodic histogram. Thus, Fourier
analysis was applied to extract the period of the force
ele-ments in the histogram
[11]
. With the period, the force
ele-ments are grouped into several clusters that are related to
the number of biomolecular pairs. The mean force in each
cluster against the number of biomolecular pairs is plotted,
and the slope of the regression line estimates the adhesive
force of a single biomolecular pair. However, Fourier
anal-ysis provides no information of the grouping offset, which
is noticed to affect the estimation of the adhesive force in
our preliminary data analysis. Therefore, we modified the
method in the grouping of histogram with an offset at the
first half of the grouping period from a physics intuition.
In this paper, we report the result of using optical
twee-zers to measure the interaction force between different
MrkD adhesin variants and collagen IV for the first time.
We demonstrated that the grouping begins with an offset
at the first half of the period by examining the least mean
square error (MSE) of the averaged adhesive force as a
function of the offset. Finally, the adhesive force between
various fimbriae that expressed various mrkD alleles and
collagen IV is successfully presented.
Materials and methods
Preparation of recombinant E. coli displaying type 3 fimbriae of K. pneumoniae. The type 3 fimbriae of K. pneumoniae were displayed on the
surface of E. coli JM109, a non-type 3 fimbriae-producing strain, by transformation of a plasmid carrying the complete mrk operon including one of the four mrkD variants. The plasmids used in the study include pmrkABCDV1F, pmrkABCDV2F, pmrkABCDV3F, and pmrkABCDV4F. Plasmid pmrkABC, which was incapable of producing fimbriae, was used as a negative control. Prior to the adhesive force measurement, the E. coli transformants were grown in GCAA medium (with 100 lg/ml ampicillin antibiotics) at 37C for 20 h, for an optimal expression of type 3 fimbriae[15].
Preparation of collagen-coated beads. Polystyrene beads (Polysciences, Inc. Warrington, PA, USA) of 1 lm in diameter were added into 500 ll PBS (phosphate buffer saline: 137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, and 2 mM KH2PO4, pH 7.4) containing 10 lg collagen and then incubated at 4C for 16 h. The beads were then blocked with the blocking reagent (2% BSA in PBS) at room temperature for 1 h. After blocking, the beads were washed and re-suspended in 100 ml PBS. BSA-coated beads were also prepared similarly, with the replacement of the collagen to BSA solution.
Optical tweezers. A diode pumped solid-state laser (1 W, 1064 nm, IRCL-1064-1000-L, CrystaLaser, USA) was used as the light source of the optical tweezers. Through a 12· beam expander, the laser beam was guided into an inverted microscope (DMIRB, Leica, Germany) and focused on the sample by an objective lens (100· oil, NA = 1.25, N PLAN, Leica, Germany). The specimen holder of the sample was mounted on a PZT stage (Tritor 100 CAP, piezosystem Jena, Germany) with a minimum step size of 1.2 nm and a maximum travel range of 80 lm during its closed operation. When a bead was trapped by the optical tweezers, the forward scattered laser light was collected by using a condenser (20·, NA = 0.5, Plan Fluor, Nikon, Japan). In addition, the forward scattering pattern of the collected light on the back focal plane of the condenser was imaged on a quadrant photodiode (QPD, G6849, Hamamatsu, Japan) by a lens (f = 50 mm). The voltage signals output from the QPD were processed by a preamplifier and then a main amplifier (O¨ ffner, MSR-Technik, Germany) with a maximum amplifica-tion factor of 500 and a cut-off frequency of 1 MHz. Subsequently, a DAQ card (NI-6115, National Instrument, USA) was used to record the voltage signals with a sampling rate up to 10 MHz for a simultaneous acquisition of four channels. A bright-field illumination was achieved by using a halogen lamp through a lens and the condenser. The images were taken by using a CCD camera (12v1E, Mintron, Taiwan) with a filter blocking the unwanted laser light.
The trapping force was calculated by multiplying the trapping stiffness and the displacement of a trapped bead. The stiffness of the optical tweezers was calibrated by using power spectrum method[16]. The dis-placement of the bead was derived from the signal of QPD. Since the unit of the QPD signal was in voltage, a conversion parameter of the QPD signal to the bead’s displacement was also calibrated by using the power spectrum method[16]. The stiffness and the conversion parameter of each bead were calibrated in advanced to each adhesive force measurement.
Force measurement using optical tweezers. The bacteria with type 3 fimbriae were adhered and fixed on a coverslip for 30 min. A collagen IV-coated bead was trapped still by the optical tweezers. Then, the coverslip was shifted by using the PZT stage so that one of the bacteria upon it can be moved toward the trapped bead. Once the bacterium contacted the collagen-coated bead, the coverslip was shifted immediately to detach the bacterium from the bead. During the detachment, the optical tweezers exerts a trapping force balance to the adhesive force and the position of the bead is displaced. The farther the displacement of the trapped bead, the larger the trapping force on the bead. Then, once the trapping force was larger than the adhesive force, the trapped bead was pulled away from the bacterium.
In our experiment, every single adhesive event was measured by moving a bacterium in contact with a trapped bead from the bacterium’s side. The contact time was about one second. The velocity of the stage was 1 lm/s during the entire measuring process. A typical record of the bead’s displacement during the measurement is illustrated inFig. 1. InFig. 1, the position of the bead along with time exhibits a triangle-like shape. The peak of the triangle indicates when the trapping force is just sufficient to
detach the bead from the bacterium. The adhesive force is obtained accordingly.
Data analysis. The measured force data were plotted in a histogram. It was thought that the adhesive force could be a result from the formation of multiple fimbria–collagen pairs, thus the force elements in the histo-gram displayed into a series of clusters. The series of clusters were dis-tributed periodically because the mean force associated with each cluster was an integer multiple of the mean force of the lowest cluster. Thus, Fourier analysis was applied to extract the period to group the force elements in the histogram. The analysis was presented via a periodogram as described[11]. A noticeable peak in the periodogram would directly indicate the initial guess of the period. However, it was noticed that Fourier analysis provided no information of the grouping offset. Without any available method, we thought from the nature of a periodic histogram. With a physics intuition, the mean forces of the first, second, . . . clusters in a periodic histogram should represent the adhesive forces of single, dou-ble, . . . pairs, respectively. For a histogram with a period of T, grouping with an offset of 1/2 T leaded to a grouping of 1/2 T 3/2 T, 3/2 T 5/2 T, and .etc. The mean force of each cluster thus approximated to T, 2 T, 3 T, . . ., which exactly represented the adhesive force of single, double, triple, . . . fimbria–collagen pairs. Therefore, the force elements were grouped with an offset at the first half of the period. Finally, the force elements were grouped according to a period and a resulting offset extracted from Fourier analysis. The sequence of the clusters corre-sponded with the number of fimbria–collagen pairs. The mean force in each cluster against the number of pairs was plotted, and a regression line
was fitted. The slope of the regression line then indicated the adhesive force of a single fimbria–collagen pair.
Results
The histograms of successful fimbriae–collagen adhesion
events are shown in the left column of
Fig. 2
. The bin width
in the histograms was set at 0.1 pN because our force
mea-surement resolution was slightly below this level. The
max-imal force element was set at 20 pN to include all our data
points. A slight different setting of the bin width and the
maximal force element did not affect the result of Fourier
analysis. The periodograms are shown in the right column
of
Fig. 2
. From the periodogram, the period to group the
histogram was extracted. Accordingly, the force histogram
was grouped with an offset at the first half of the period. At
last, plots of the mean force of each cluster against the
number of fimbria–collagen pairs are shown in
Fig. 3
.
The adhesive force between each single fimbria and
colla-gen IV was obtained from the slopes of the regression lines
in
Fig. 3
. The data of mrkD
v1-expressed fimbriae is not
shown because they exhibited weak adhesion in both
opti-cal tweezers experiments and collagen-binding assay (data
not shown).
The control strain E. coli JM109 [pmrkABC] displayed
no fimbriae on its surface, and adhered poorly to the bead
coated with 2% bovine serum albumin (BSA; data shown in
supplemental material). JM109 [pmrkABC] also exhibited
poor adhesion with collagen IV in collagen-binding assay
(data shown in supplemental material). These results
illus-trate that our adhesive events were indeed come from the
interaction between the fimbriae and collagen IV.
Our periodograms, unlike those in Ref.
[11]
, reveal
multi-ple peaks that make it difficult to determine the period to
group the histogram. To select proper peaks in our
periodo-grams to group the histoperiodo-grams, there are two foregoing
con-straints. First, all adhesive elements should be included in
the grouping since they were all effective. Second, the
group-ing should begin with an offset at the first half of the period.
In
Fig. 2
(D), the periodogram of mrkD
v2displays a
noticeable peak at 2 pN, which leads to an offset at 1 pN.
The selected period and offset make the grouping includes
all the force elements in the data of mrkD
v2. In
Fig. 2
(E),
the periodogram of mrkD
v3shows two noticeable peaks
at 4 and 10 pN. However, the peak at 10 pN leads to an
off-set at 5 pN, in which these grouping parameters do not
make the grouping include all the force elements.
Conse-quently, the period and the offset to group the force
ele-ments in the data of mrkD
v3are selected as 4 and 2 pN,
respectively. In
Fig. 2
(F), the periodogram of mrkD
v4, the
noticeable peak at 5 pN is not selected due to the same
rea-son that it does not satisfy the constraints. The top at
2.9 pN thus becomes the selected grouping period with
and resulting offset at 1.45 pN.
Accordingly, the histograms of mrkD
v2, mrkD
v3, and
mrkD
v4were grouped into 6, 5, and 7 clusters, respectively.
The adhesive force shows a satisfactory linear relation to
-6
-5
-4
-3
-2
-1
0
1
2
0
1
2
3
4
5
6
7
8
9
10
Time (sec)
Displacement (V)
-14.2
-11.9
-9.5
-7.1
-4.7
-2.4
0.0
2.4
4.7
For
ce (
p
N)
A
B
C
D
E
Detached
displacement
A
B
C
E
D
Fig. 1. A schematic presentation describing the adhesive force measure-ment. (A) The trapped bead is under a confined Brownian motion. (B) The bead is adhered on a single fimbria. The rise in force is due to over push of the bacterium on the bead. (C) and (D) The bead is being pulled away from the trapping center. The farther the displacement of the trapped bead to the trapping center, the larger the trapping force on the bead. (E) The trapping force is larger than the adhesive force so that the bead is detached from the bacterium and trapped back to the trapping center. The QPD signal is the same as the one in (A). The data is sampled with 10 kHz, and averaged by each 100 sample points. The left vertical axis represents the bead’s displacement recorded by the QPD (units in voltage). The right vertical axis represents the resulting adhesive force (units in pico-Newton). The adhesive force is calculated by multiplying the displacement when the bead is detached from the bacterium (in V), the conversion parameter (in nm/V), and the stiffness of the optical tweezers (in N/m).
the number of fimbriae–collagen pairs, as indicated in
Fig. 3
. In
Fig. 3
(A), the means ± standard deviations of
the data points are 2.33 ± 0.38, 4.06 ± 0.39, 6.16 ± 0.5,
7.90 ± 0.5, 14.10 ± 0.58, and 16.41 ± 0.2 pN, respectively.
In
Fig. 3
(B) the means ± standard deviations of the data
points
are
4.06 ± 0.95,
7.38 ± 1.04,
12.01 ± 1.11,
14.92 ± 0.64, and 18.75 ± 0.66 pN. In
Fig. 3
(c), the
means ± standard deviations of the data points are
2.64 ± 0.27,
5.92 ± 0.90,
8.39 ± 0.69,
12.22 ± 0.04,
13.78 ± 0.73, 17.37, and 19.95 pN. The estimated adhesive
force ± error is the slope ± the mean squared error (MSE)
of the regression line. As a result, we present for the first
time the adhesive force between collagen IV and each of
the type 3 fimbriae, which expressed with various mrkD
alleles, mrkD
V2, mrkD
V3, or mrkD
V4are 2.03 ± 0.03 pN,
3.79 ± 0.12 pN, and 2.87 ± 0.15 pN, respectively. It is
shown that the mrkD
V3-containing fimbriae exhibited the
largest adhesive force on collagen IV whereas mrkD
V1-con-taining fimbriae showed almost no adhesion.
Discussions
In this study, we proposed that the grouping of a
histogram begins with an offset of 1/2 T in data analysis.
To verify this, the MSE of the regression line against offset
is presented as shown in
Fig. 4
. The most reasonable offset
in terms of precise force estimation should occur at the
least MSE. As shown in
Fig. 4
, the least MSE occurs at
the offset of 1/2 T in both curves for mrkD
v2- and
mrkD
v3-carried fimbriae. As for the curve of the mrkD
v4-carried fimbriae, although the least MSE occurs at an offset
of 3/10 T, the MSE at an offset of 1/2 T is still relative low.
We thus conclude that the most reasonable offset occurs
1/2 T. Further, the largest difference of the adhesive force
was about 27% in terms of different applying offset in our
experiment (data shown in supplementary material).
In our experience, longer contact periods increase the
overall adhesive force, presumably due to the formation
of multiple fimbria–collagen pairs. In this experiment, the
contact period of the fimbriae and collagen was kept short
to ensure fewer fimbria–collagen pairs formation. Thus, the
histograms shown in
Fig. 2
were grouped only into several
clusters. Otherwise a peak at short period in terms of
grouping into more clusters would be selected.
Corre-spondingly, our data analysis processes as well as the
selected peaks shown in
Fig. 2
were reasonable.
The measurement of adhesion force in several other
bacteria have been reported previously
[10–14]
and the
Histogram -
mrkDv2
0 1 2 3 4 5 0 2 4 6 8 10 12 14 16 18 20 Force (pN) Cou nts Histogram -mrkDv3
0 1 2 3 4 0 2 4 6 8 10 12 14 16 18 20 Force (pN) Cou nts Histogram -mrkDv4
0 1 2 3 4 5 0 2 4 6 8 10 12 14 16 18 20 Force (pN) Cou nts 2.9 Periodogram -mrkD
v4 0 10 20 30 40 0 5 10 15 20 25 Force (pN/cycle) A.U. 4 Periodogram -mrkD
v3 0 5 10 15 20 25 0 5 10 15 20 25 Force (pN/cycle) A.U. 2 Periodogram -mrkD
v2 0 10 20 30 40 0 5 10 15 20 25 Force (pN/cycle) A.U.A
D
B
E
C
F
Fig. 2. The histograms (A–C) and the periodograms (D–F) of mrkDv2-, mrkDv3-, and mrkDv4- expressed fimbriae adhered to collagen IV. The numbers of experimental events are 51, 48 and 53, respectively.
magnitude of the interacting force varies significantly
depending on the type of molecular pairs. In the case of
Staphylococcus sp.
[11–13]
, the adhesive force measured
between extracellular matrix (ECM) molecules fibronectin
and a series of Staphylococcus strains is in the range of 16–
25 pN. Unlike fimbriae in Gram-negative bacteria that
extend out from the cell, the protein mediating the adhesion
of the staphylococcal cells to fibronectin is covalently
anchored on the bacterial cell wall and spread all over the
sur-face. Thus, the adhesion forces measured in these studies are
surely the sum of multiple components on the bacterial
sur-face and therefore are much larger than that of our results.
Another study that is more related to ours was the
measure-ment of interacting force between a single type 1 pilus and an
a-C-mannoside ligand
[10]
. The adhesive force obtained in
the study was 1.7 pN, which is close to our data of the
adhe-sive force between type 3 fimbriae and collagen IV.
In summary, by using a set of high resolution optical
tweezers, we have demonstrated that the sequence
varia-tion of the adhesins could affect their binding force to the
receptor. Although the biological meaning of the
phenom-enon remains to be investigated, the information can be
referred to further research on the connection between
adhesion and infection of bacteria such as how bacteria
sustain the shearing force in gastric and urinary systems.
In data analysis, we verified that the grouping should begin
with an offset at the middle of the first period by selecting
the least MSE of the regression line. The adhesive force
measurements are reliable because the errors of the
mea-sured forces are in the range of the system resolution.
Finally, as slight differences between the collagen binding
forces exerted by different MrkD adhesin variants could
be distinguished, we believe that our force measurements
are reliable and have a high resolution. Thus, the optical
tweezers and the analytical process set-up in our laboratory
can also be applied to study other biomechanical processes
that involve subtle differences.
Acknowledgments
This work is supported by research grants from the
National Science Council, Taiwan, 2005 National Research
Program for Nanoscience and Technology,
#94-2120-M-009-015.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found,
in
the
online
version,
at
doi:10.1016/
j.bbrc.2006.08.190
.
References
[1] S.H. Han, An association with diabetes mellitus and septic endoph-thalmitis, Review of hepatic abscess from Klebsiella pneumoniae, West. J. Med. 162 (1995) 220–224.
[2] R. Podschun, U. Ulmann, Klebsiella spp. as nosocomial pathogens: epidemiology, taxonomy, typing methods, and pathogenicity factors, Clin. Microbiol. Rev. 11 (1998) 589–603.
[3] A.M. Tarkkanen, R. Virkola, S. Clegg, T.K. Korhonen, Binding of the type 3 fimbriae of Klebsiella pneumoniae to human endothelial and urinary bladder cells, Infect. Immun. 65 (1997) 1546–1549.
[4] A.M. Tarkkanen, B.L. Allen, B. Westerlund, H. Holthofer, P. Kuusela, L. Risteli, S. Clegg, T.K. Korhonen, Type V collagen as the target for type-3 fimbriae, enterobacterial adherence organelles, Mol. Microbiol. 4 (1990) 1353–1361.
[5] B.L. Allen, G.F. Gerlach, S. Clegg, Nucleotide sequence and functions of mrk determinants necessary for expression of type 3 fimbriae in Klebsiella pneumoniae, J. Bacteriol. 173 (1991) 916–920.
mrkD
v2- expressed fimbriae
y = 2.0312x
R
2= 0.9992
0
5
10
15
20
25
0
2
4
6
8
10
Number of pairs
Force (pN)
mrkD
v4- expressed fimbriae
y = 2.8662x
R
2= 0.997
0
5
10
15
20
25
0
2
4
6
8
10
Number of pairs
Fo
rce (pN)
mrkD
v3- expressed fimbriae
y = 3.7865x
R
2= 0.9975
0
5
10
15
20
25
0
2
4
6
8
10
Number of pairs
Force (pN)
A
B
C
Fig. 3. The adhesive force against the number of (A) mrkDV2-, (B) mrkDV3-, (C) mrkDV4- expressed fimbriae adhered to collagen IV.
MSE of Estimated Force vs Grouping Offset
0 0.2 0.8 1 1.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 Offset (T) MS E (pN) mrkDv2 mrkDv3 mrkDv4
[6] J. Jagnow, S. Clegg, Klebsiella pneumoniae MrkD-mediated biofilm formation on extracellular matrix- and collagen-coated surfaces, Microbiology 149 (2003) 2397–2405.
[7] D.C. Old, Inhibition of the interaction between fimbrial haemagglu-tinin and erythrocytes by D-mannose and other carbohydrates, J.
Gen. Microbiol. 71 (1972) 149–157.
[8] T.A. Sebghati, T.K. Korhonen, D.B. Kornich, S. Clegg, Character-ization of the type 3 fimbrial adhesins of Klebsiella strains, Infect. Immun. 66 (1998) 2887–2894.
[9] A. Ashkin, J.M. Dziedzic, J.E. Bjorkholm, S. Chu, Observation of a single-beam gradient force optical trap for dielectric particles, Opt. Lett. 111 (1986) 288–290.
[10] M.N. Liang, S.P. Smith, S.J. Metallo, I.S. Choi, M. Prentiss, G.M. Whitesides, Measuring the forces involved in polyvalent adhesion of uropathogenic Escherichia coli to mannose-present-ing surfaces, Proc. Natl. Acad. Sci. USA 97 (2000) 13092– 13096.
[11] K.H. Simpson, G. Bowden, M. Ho¨o¨k, B. Anvari, Measurement of adhesive forces between individual Staphylococcus aureus
MSCRAMMs and protein-coated Surfaces by use of optical tweezers, J. Bacteriol. 185 (2003) 2031–2035.
[12] K.H. Simpson, M.G. Bowden, M. Ho¨o¨k, B. Anvari, Measurement of adhesive forces between S. epidermidis and fibronectin-coated surfac-es using optical tweezers, Laser. Surg. Med. 31 (2002) 45–52. [13] K.H. Simpson, M.G. Bowden, S.J. Peacock, M. Arya, M. Ho¨o¨k, B.
Anvaria, Adherence of Staphylococcus aureus fibronectin binding protein A mutants: an investigation using optical tweezers, Biomol. Eng. 21 (2004) 105–111.
[14] E. Fa¨llman, S. Schedin, J. Jass, M. Andersson, B.E. Uhlin, O. Axner, Optical tweezers based force measurement system for quantitating binding interactions: system design and application for the study of bacterial adhesion, Biosens. Bioelectron. 19 (2004) 1429–1437. [15] G.F. Gerlach, B.L. Allen, S. Clegg, Type 3 fimbriae among
enterobacteria and the ability of spermidine to inhibit MR/K hemagglutination, Infect. Immun. 57 (1989) 219–224.
[16] F. Gittles, C.F. Schmidt, Signals and Noise in Micromechanical Measurements, in: M.P. Sheetz (Ed.), Laser Tweezers in Cell Biology, Academic Press, San Diego, 1998, pp. 129–156.
