Báo cáo khoa học: F1 – Structure and Function of Motor Proteins pot

F1 – Structure and Function of Motor Proteins
F1-001
What single-molecule mechanics can tell us
about mitosis?
J. Howard
Max Planck Institute of Molecular Cell Biology & Genetics,
Dresden, Germany. E-mail:
Our laboratory is interested in the biochemical and biophysical
basis of cell structure. The structure of a cell is determined
primarily by its cytoskeleton, which serves as a scaffold to sup-
port the plasma membrane, and as a network of tracks along
which motor proteins transport sub cellular structures. Our
research is therefore focused on the mechanics of the cytoskele-
ton, with a particular emphasis on microtubules and microtu-
bule-based motors. On one hand, we are interested in the
mechanisms by which these proteins work: i.e. how do kinesins
and dyneins convert chemical energy derived from the hydrolysis
of ATP into mechanical work used to move along or to
Abstracts
332
depolymerize microtubules? And on the other hand, we are inter-
ested in the roles that microtubules and their motors play in cell
morphology and motility. In this regard we are particularly inter-
ested in how the dynamic properties of microtubules and motors
drive spindle and chromosome movements in mitosis. To address
these questions we are combining molecular biology techniques
with image processing, modeling, mechanical measurements and
single-molecule techniques.
F1-002
Control of actin assembly in cell motility
M F. Carlier

CNRS, Gif-sur-Yvette, France. E-mail:
Living cells change shape and move in response to environmen-
tal signals. These motile processes play a pivotal role in mor-
phogenesis, migration of embryonic and metastatic cells,
angiogenesis, synaptic plasticity, immune response and interac-
tion of the host cells with pathogens. They are generated by
polarized, spatially directed actin assembly. It is the treadmilling
(dissipative turnover) of actin filaments, regulated by specific
proteins, which is responsible for force and directional move-
ment. Two cellular machineries are responsible for spatially
directed initiation of actin filaments and they operate in distinct
processes, (i) the WASP-Arp2/3 system is at the origin of the
formation of a branched filament array; (ii) formins, in associ-
ation with profilin, catalyze the rapid processive assembly of
non-branched actin filaments. We have combined a biochemical
and a biomimetic approach to understand the molecular mecha-
nisms of these auto-organized processes. We have reconstituted
the sustained actin-based movement of a N-WASP- or formin-
functionalized particle in a biochemically controlled medium,
which enables measurement of force production in correlation
with structure and motility, and we can derive information on
the molecular mechanism of movement by single molecule
measurements.
F1-003
The structure of the myosin VI motor reveals
the mechanism of directionality reversal
J. Me
´
ne
´

trey
1
, A. Bahloul
1
, C. Yengo
2
, A. Wells
2
, C. Morris
2
,
H. L. Sweeney
2
and A. Houdusse
1
1
UMR144 - CNRS, Institut Curie, Paris, France,
2
Departement of
Physiology, Pennsylvania School of Medicine, Philadelphia, PA
19104–6085 United States of America.
E-mail:
We have solved a 2.4 A
˚
structure of a truncated version of the
reverse direction myosin motor, myosin VI that contains the
motor domain and binding sites for two calmodulins. Surpris-
ingly, the structure reveals only minor differences in the motor
domain as compared to plus-end directed myosins, with the
exception of two unique inserts. The first insert is near the nuc-

leotide-binding pocket, and alters the rates of nucleotide associ-
ation and dissociation. The second unique insert forms an
integral part of the myosin VI converter domain along with a
calmodulin bound to a previously unseen binding motif within
the insert. This serves to redirect the effective ‘‘lever arm’’ of
myosin VI, which includes a second myosin VI calmodulin bound
to an ‘‘IQ motif’’, towards the pointed (–) end of the actin fil-
ament. This repositioning largely accounts for the reverse direc-
tionality of this class of myosin motors. We propose a model
incorporating a kinesin-like uncoupling/docking mechanism to
fully explain the movements of myosin VI.
F1-004
Regulation and mechanics of myosin V
J. R. Sellers
Laboratory of Molecular Physiology, National Heart, Lung and
Blood Institute, National Institute of Health, Bethesda, MD United
States of America. E-mail:
Myosin V is vesicle motor that moves processively on actin filaments.
In melanocytes it functions in cooperation with microtubule motors
to localize melanosomes to the dendritic tips. Myosin V dimerizes
via a coiled coil motif in its tail region to produce a two-headed
structure. Each head is composed of a compact motor domain and a
long neck formed by the association of calmodulin residues with six
tandem IQ motifs. This structure allows the molecule to take 36 nm
steps along an actin filament corresponding to the helical pitch of the
actin and, in turn, allows the molecule to keep its cargo positioned
above the cytoskeleton. We study the mechanics of the processive
motor using optical trapping and total internal reflection fluores-
cence microscopy (TIRF). These data confirm that the neck of myo-
sin V acts as a rigid lever arm and that myosin V moves in a hand-

over-hand manner. The enzymatic activity of myosin V is regulated
by calcium in vitro. This is coupled to a large conformational change
in the molecule. In the absence of calcium myosin V folds into a
compact triangular-shaped structure in which the heads bend down
and contact the globular tail domain. Under these conditions the
molecule migrates with a sedimentation coefficient of 14S. In the
presence of calcium or at high ionic strength the molecule opens up
to form T- or Y-shaped structures which sediment at 11S. We pro-
pose that in the cell, the transition between the open, active state and
the folded, inactive state is regulated by binding of receptor or dock-
ing proteins to the myosin V globular tail domain. This would not
only target the myosin to the proper cargo, but would also lock the
myosin into an active form. We are currently studying the interac-
tions responsible for the folded, off state of this molecule.
F1-005
F-actin modulates myosin conformational
states by switch I loop movement
B. Kintses
1
, M. Gyimesi
1
, W. Zeng
2
, P. B. Conibear
2
,
C. R. Bagshaw
2
and A. Ma
´

lna
´
si-Csizmadia
1
1
Departement of Biochemistry, Eo
¨
tvo
¨
s Lora
´
nd University, Buda-
pest, Hungary,
2
Department of Biochemistry, University of Leices-
ter, Leicester, United Kingdom. E-mail:
Myosin converts chemical energy to mechanical work. It undergoes
a large conformational change during the ATPase cycle that results
a step on the actin filament. The main question is how the func-
tional regions communicate with each other. Another problem
concerns the development and relaxation of the mechanical strains
during the catalytic cycle and their perturbation by external force.
We have extensively characterized the energetic coupling between
nucleotide binding, the so-called open/closed transition and actin
binding using different single tryptophans located at the relay loop
(W501, Dictyostelium sequence), the nucleotide binding region
(W129) and the switch 1 region (W239 and W241). The fluores-
cence from native actin tryptophan residues is not significantly per-
turbed on binding to myosin, although a fluorescence signal is
detected as a consequence of a light scatter artifact. ATP, ATPa

˜
S
and ADP binding affected less than threefold by actin. The isome-
rization detected by W129 clearly precedes the dissociation of actin
in the case of ADP and ATPa
˜
S binding. The fluorescence from the
conserved W501 residue located at the distal end of the relay helix
is very sensitive to the lever arm disposition and the observed fluor-
escence emission intensity can be used to estimate the equilibrium
constant between the pre and post power-stroke conformations.
Actin modulates this equilibrium by no more than twofold. These
data suggest that actin activates another process in the mechanism,
Abstracts
333
such as switch 1 movement, rather than influencing the switch 2
equilibrium. Consequently, despite actin does not have major effect
on nucleotide binding and the equilibriums of the pre- and post-
power-stroke the fluxes of the kinetic routes change fundamentally
by actin binding: in the absence of actin phosphate release precedes
the closed-open transition while in ternary complex the main route
is that products releases follow the power stroke.
F1-006
Calmodulin and calmodulin-like protein as
light chains for myosin-10: Specificity and role
in protein stabilization leading to prolonged
function
E. E. Strehler, A. J. Caride, A. S. Mauer and R. D. Bennett
Biochemistry and Molecular Biology, Mayo Clinic College of
Medicine, Rochester, MN, United States of America.

E-mail:
Vertebrate myosin-10 (MYO10) is involved in filopodial motility,
phagocytosis, and microtubule-F-actin interactions. Each MYO10
heavy chain contains three IQ motifs, which are light chain binding
sites. Previous work has shown that the first two IQ domains of
MYO10 bind calmodulin (CaM) whereas IQ3 alternatively binds
calmodulin-like protein (CLP). We performed stopped-flow experi-
ments with fluorescent derivatives of CaM (TA-CaM) and CLP
(TA-CLP) to analyze their binding to the IQ3 peptide. TA-CaM
bound to IQ3 in the absence and in the presence of 100 lm Ca
2+
.
In the absence of Ca
2+
the binding was faster, albeit with lower
affinity, suggesting an even faster dissociation rate. In the presence
of Ca
2+
, the time course of the TA-CaM-IQ3 reaction was best
described by an exponential function. CLP was able to reduce the
amplitude of this exponential in the presence, but not in the
absence, of Ca
2+
. These results suggest a novel IQ domain interac-
tion, in which Ca
2+
regulates binding of CaM to IQ3 by modula-
ting competition with CLP. In HeLa cells over-expressing CLP,
endogenous MYO10 was strongly upregulated. Likewise, transfec-
tion with GFP-MYO10 resulted in increased fluorescence in cells

that co-expressed CLP or excess CaM. MYO10 upregulation resul-
ted in an increase in the size and number of filopodia. Cells expres-
sing CLP displayed increased motility as indicated by their
shortened wound-healing time compared to control cells lacking
CLP. CLP-dependent upregulation of MYO10 was due to
increased protein stability. CLP expression in specific epithelial
cells may thus prolong MYO10 function in conditions of elevated
intracellular Ca
2+
and limiting CaM. Acknowledgment: Suppor-
ted by grants from the Susan G. Komen Breast Cancer Foundation
(EES) and the American Heart Association (AJC).
F1-007P
Crystal structure of the portal protein from
bacteriophage SPP1 and model for DNA
translocation
A. A. Lebedev
1
, M. H. Krause
2
, A. Vagin
1
, E. V. Orlova
3
,
E. J. Dodson
1
, P. Tavares
4
and A. A. Antson

1
1
YSBL, Chemistry Department, York University, York, United
Kingdom,
2
Max-Planck Institut fu
¨
r Molekulare Genetik, Berlin,
Germany,
3
Birkbeck College, Department of Crystallography, Uni-
versity of London, London, United Kingdom,
4
Unite
´
de Virologie
Mole
´
culaire et Structurale, Gif-sur-Yvette, France.
E-mail:
The mechanism of DNA translocation into a viral procapsid
remains one of the most intriguing questions of viral particle
assembly [1]. Tailed bacteriophages and herpes viruses have a
specialized vertex for double-stranded DNA (dsDNA) entry
into the procapsid during viral chromosome packaging [2, 3].
The main component of this specific doorway is the portal pro-
tein, a circular oligomer with a central tunnel through which
the DNA transfer occurs. Together with viral ATPase (termi-
nase) the portal protein forms a molecular motor that is able
to translocate DNA against high internal pressure [4]. We

determined the X-ray structure of the SPP1 portal protein
(gp6) in its 13-subunit oligomeric form where the tunnel residue
segments, not visible in the previously determined structures of
bacteriophage phi-29 portal protein [5, 6], are well defined and
form a DNA-transfer arm. Our X-ray and electron microscopy
data suggest that DNA translocation is driven by a novel
mechanism involving mechanical movements of arms along the
inner walls of the tunnel. We propose a model for DNA trans-
location where such movements propagate around the double
helix of DNA similar to a ‘‘Mexican wave’’ moving across a
stadium.
References
1. Hendrix RW. Proc Natl Acad Sci USA 1978; 75: 4779–47783.
2. Bazinet C, King, J. Annu Rev Microbiol 1985; 39: 109–129.
3. Droge A, Tavares P. J Mol Biol 2000; 296: 103–115.
4. Smith DE, et al. Nature 2001; 413: 748–752.
5. Simpson AA, et al. Nature 2000; 408: 745–750.
6. Guasch A, et al. J Mol Biol 2002; 315: 663–676.
F1-008P
Unraveling the catalytic mechanism of the
bacteriophage T7 gene 4 helicase
D. J. Crampton, S. Mukherjee, A. van Oijen and
C. C. Richardson
Department of Biological Chemistry and Molecular Pharmacology,
Harvard Medical School, Boston, MA, USA.
E-mail:
The gene 4 helicase of bacteriophage T7 is a molecular motor
that unwinds double-stranded DNA using the energy derived
from the hydrolysis of deoxynucleoside 5-triphosphates. Here
we present data pertaining to four essential activities of the heli-

case that act cooperatively to unwind DNA. (i) Oligomer For-
mation. The T7 gene 4 helicase forms both heptamers and
hexamers dependent upon the presence of nucleoside di- or tri-
phosphates respectively. We find that the heptamer cannot bind
DNA but rather, in the presence of single-stranded DNA, hep-
tamer converts to hexamer when both nucleoside di- and tri-
phosphates are present together. This conversion between
oligomers is regulated by histidine 465 through differentiation
of the absence or presence of a c-phosphate on the bound nuc-
leotide. (ii) DNA Binding. The loop comprised of residues 466–
475 has been postulated to be the major site for the binding of
single-stranded DNA. Lysines 471 and 473 of this loop were
altered to determine their contribution to the overall binding of
single-stranded DNA. (iii) dTTP Hydrolysis. We find that all
subunits of the hexamer are active in the hydrolysis of dTTP.
Changing the catalytic base glutamate 343 to glutamine creates
a non-catalytic subunit that responds to dTTP in a manner sim-
ilar to the effect of non-hydrolyzable analog b,cc-methylene
dTTP on wild-type helicase. The presence of a single non-cata-
lytic subunit per hexameric unit abolishes all dTTPase activity.
(iv) Translocation. We are developing methods of observing
movement along DNA by the gene 4 helicase using single-mole-
cule fluorescence microscopy.
Abstracts
334
F1-009P
Myosin VI in chromaffin cells
M. Dominik
1
, L. Kiljanek

2
, and M. J. Redowicz
1
1
Laboratory of Cell Motility, Department of Muscle Biochemistry,
Nencki Institute of Experimental Biology, Warsaw, Poland,
2
Laboratory of Confocal Microscopy, Nencki Institute of Experi-
mental Biology, Warsaw, Poland.
E-mail:
Myosin VI (MVI), the member of one of at least 18 different
myosin families, is ubiquitously expressed in multicellular eukary-
otes. It is an unusual actin-based motor since it walks towards
the minus end of actin filaments, in the opposite direction to
other myosins. MVI is involved in subcellular transport of mem-
brane-containing structures, it has been found to be engaged in
clathrin-based endocytosis and cell spreading and migration.
Here, we aimed at testing the involvement of MVI in chromaffin
granule trafficking within chromaffin cells. MVI, but not myosin
V, has been found in vesicular fractions isolated from bovine
adrenal medulla and rat pheochromocytoma cells (PC12). MVI is
on the apical side of the vesicle and its association with the gran-
ule seems to be very tight as the stripping with high salt concen-
tration or high pH does not remove it from the granule surface;
MVI is stripped from the granules only after addition of Triton
X100. MVI colocalizes with dopamine-b-hydroxylase (DbH),
chromaffin granules marker, both in PC12 cells and primary cell
culture of adrenal medulla. It has been observed that MVI and
DbH were still found on the same granules after their isolation
from PC12 cells, confirming strong association of MVI with the

granules. Five minute stimulation of PC12 cells with 56 mm KCl
seems to affect MVI localization as the fluorescence intensity cor-
responding to MVI was enhanced within the perinuclear area
only in stimulated cells. These preliminary data seem to indicate
the important role of MVI in secretory cells.
F1-010P
Enzyme kinetics above denaturation
temperature
M. Gyimesi, Z. Simon and A. Ma
´
lna
´
si-Csizmadia
Department of Biochemistry, Eo
¨
tvo
¨
s Lora
´
nd University, Budapest,
Hungary. E-mail:
Wide range of temperature dependence of enzyme kinetics meas-
urements enables more precise determination of reaction energetics
and accurate separation of reaction steps. We have developed a
novel temperature-jump/stopped flow method to measure enzyme
transient kinetics at high temperature even above denaturation
temperature of enzymes. This method enables us to increase the
temperature in a millisecond time scale parallel with mixing of the
reactants. Heat denaturation normally occurs in a second time
scale. If the temperature is increased above the enzyme denatura-

tion temperature, essentially all of the events can be followed that
are faster than heat denaturation reaction. We have tested the new
method on myosin ATPase and GFP folding/unfolding reactions.
F1-011P
Changes of inter-subunit contacts in growing
filaments during salt-induced polymerization
of actin
A. Galinska-Rakoczy, B. Wawro and H. Strzelecka-Golaszewska
Laboratory of Structural Muscle Proteins, Department of Muscle
Biochemistry, Nencki Institute of Experimental Biology, Warsaw,
Poland. E-mail:
Polymerization/depolymerization of actin underlies numerous
motile events in eukaryotic cells. It is well established that
conformational changes in G-actin generated by exchange of its
tightly bound Ca for Mg accelerate the nucleation reaction but
do not influence substantially the rate of filament growth during
salt-induced polymerization of actin. We used N, N’’-1,4-phenyl-
enebismaleimide (PBM) to probe the conformation of F-actin at
various stages of filament growth during polymerization of
CaATP-, MgATP-, and MgADP-G-actin. Effects of the type of
polymerizing salt, filament stabilization by phalloidin, and pro-
teolytic cleavage of actin at Gly-42 within D-loop were also
investigated. The earlier described accumulation of ’upper dimer ’
(UD), a product of cross-linking neighboring protomers from
two long-pitch F-actin strands, concomitant with disappearance
of initially formed ‘‘lower dimmers’’ (LD) with subunits in a
non-filamentous, antiparallel orientation, was observed under all
conditions except for a polymer of cleaved actin that required
phalloidin binding to yield UD. The increase in yield of UD clo-
sely followed filament assembly from MgATP-G-actin, whereas it

lagged behind polymerization of CaATP- and MgADP-G-actin.
Phalloidin accelerated the filament assembly but did not eliminate
the delay in UD formation. These results, along with changes in
morphology of growing filaments visualized by electron micros-
copy, suggest that the tightly bound Mg and initially bound ATP
not only accelerate the nucleation reaction but also promote a
rearrangement of F-actin structure that eliminates LD’s from
growing filaments. They also confirm the suggested role of
D-loop in stabilization of both the longitudinal subdomain 2/1
contacts and of the cross-strand contacts in F-actin.
F1-012P
The dynein light chain binds to a non-coiled-
coil tail domain of myosin-Va that includes an
alternatively spliced exon coding for three
amino acid residues
Z. Ho
´
di
1
,A.Ne
´
meth
1
, E. Kova
´
cs
1
, C. Hete
´
nyi

1
, A. Bodor
2
,
A. Perczel
2
and L. Nyitray
1
1
Deptartment of Biochemistry, Eo
¨
tvo
¨
s Lora
´
nd University, Buda-
pest, Hungary,
2
Department of Organic Chemistry, Eo
¨
tvo
¨
s Lora
´
nd
University, Budapest, Hungary. E-mail:
Class V myosins are involved in short-range intracellular trans-
port along actin filaments. Of the three mammalian myosin-V
heavy chain genes, mutations of MYO5A are responsible for the
dilute phenotype and Griscelli syndrome type1 in mice and

humans respectively. A dynein light chain (DLC) has been identi-
fied as a tail domain light chain of myosin-Va (myo5a). DLC
may function as a cargo-binding and/or regulatory subunit of
both motor proteins. Our goal was to identify and characterize
the binding site of DLC on myo5a. Various fragments of myo5a
tail and DLC were expressed in E. coli and human cells. Forma-
tion of their complex was analyzed by pull-down assays, gel fil-
tration, and spectroscopic methods. DLC was found to bind as a
homodimer to a 12 residues segment (Pro1282-Thr1293) locali-
zed between the medial and distal coiled-coil predicted domains
of the tail. The binding region contains 3 residues, coded by the
alternatively spliced exon B that is essential for DLC binding.
Using CD spectroscopy, we demonstrate that binding of DLC to
the intrinsically unstructured DLC binding domain (DBD) stabil-
izes the neighboring coiled-coil domain. NMR spectroscopy and
molecular docking simulations show that a short synthetic pep-
tide of DBD binds to a surface groove on DLC, as has been
found with other known binding partners of DLC. We hypothes-
ize that the second binding site of the DLC dimer is either occu-
pied by the DBD of the other heavy chain, or by a cargo, or
interacts with other regulatory proteins.
Acknowledgment: Supported by OTKA T43746
Abstracts
335
F1-013P
Mutagenic analysis of the HsdR motor subunit
of type IC restriction modification enzyme
EcoR124I
E. Sisakova and M. Weiserova
Laboratory of Molecular Genetics of Bacteria, Division of Cell

and Molecular Microbiology, Institute of Microbiology,
Czech Academy of Sciences, Prague, Czech Republic.
E-mail:
Enzyme EcoR124I belongs to the IC family of restriction modifi-
cation enzymes, intelligent molecular motors. It is able to detect
the methylation status of its DNA target sequence and respond
with alternative activities, methylation or translocation of DNA.
While bound to its target site, it translocates DNA towards itself
simultaneously in both directions (500bp/sec). It uses the free
energy associated with ATP hydrolysis to translocate DNA so
that DNA cleavage occurs remote from the asymmetric recogni-
tion site. The enzyme EcoR124I is multifunctional, multi-subunit
enzyme, composed of three different subunits, which are encoded
by the genes hsdR, hsdM and hsdS. Products of all three genes
are required for DNA cleavage, producing the endonuclease.
HsdR subunit is a multifunctional motor protein, which has been
shown to posses ATPase, helicase and restriction activity. To
provide a fully functional molecular motor, which can never
cleave DNA, the amino acid motif X, the active site of the endo-
nuclease domain of the HsdR subunit, was subjected to site-
directed mutagenesis. The complementation analysis proved that
the substitutions D151A, E165A, E165D, E165H, K167A in the
HsdR subunit fully removed the restriction activity in vivo of the
EcoR124I enzyme. The mutant subunits were separately overpro-
duced, purified and mixed with purified methylase to reconstitute
the EcoR124I endonuclease in vitro. As a substrate for DNA
cleavage in vitro we used the plasmid pCFD30 containing a sin-
gle site for EcoR124I. The test of restriction activity showed that
reconstituted endonucleases were not able to cleave covalently
closed plasmid DNA to linear DNA in contrast to the wild-type

enzyme.
F1-014P
Dielectric and fluoroscopic study on the
dynamic effects of myosin-S1 with/without
ATP on the hyper-mobile water around actin
filaments
M. Suzuki
1
, M. P. Siddique
1
, T. Miyazaki
1
, J. Mogami
1
,
E. Katayama
2
, T. Kodama
3
and T. Q. P. Uyeda
4
1
Laboratory of Physicochemistry of Biomolecular Systems, Depart-
ment of Materials Science and Engineering, Tohoku University,
Sendai, Japan,
2
Division of Fine Morphology, The Institute of
Medical Science, The University of Tokyo, Tokyo, Japan,
3
Molecular Enzymology, Department of Bioscience and Bioinfor-

matics, Kyushu Institute of Technology, Ihzuka, Japan,
4
Gene
Function Research Center, AIST, Tsukuba, Japan.
E-mail:
Using high resolution microwave dielectric spectroscopy, we have
recently shown hyper-mobile water molecules around actin fila-
ments (F-actin) which have a much higher rotational mobility
than that of bulk water [1] and its marked volume increase with-
out significant change in the ordinary hydration shell when the
myosin motor-domain (S1) binds to F-actin [2]. Here, we report
that (i) hyper-mobile water around actin filaments has been
detected with the fluoroscopic technique based on the fact that
the fluorescence intensity of fluorophore molecules, such as ribo-
flavin and pyrene-derivatives in caotropic salt aqueous solutions
was in clear correlation with those viscosity B-coefficients, (ii) the
fluorescence intensity from pyrenyl actin markedly increased
when myosin-S1 binds to actin filaments, which is consistent with
the dielectric result (BBRC, 2004), (iii) in the solution of acto-S1
fused chimera protein and skeletal actin copolymers the fluores-
cent intensity from pyrenyl-group (linked at Cys374 of actin
SD-I) increased in the presence of ATP indicating a marked
decrease of hyper-mobile water around actin filaments. An
important implication of the present result is that the S1/F-actin
interaction forms an asymmetric field of viscosity-gradient along
the filament axis, which drives the unidirectional sliding of S1
hydrolyzing ATP by means of thermal brownian movements.
References
1. Kabir SR. et al. Biophys J 2003; 85: 3154–3161.
2. Suzuki M. et al. Biochem Biophys Res Comm 2004; 322: 340–

346.
F1-015P
Drosophila myosin V: solution kinetics and
motile properties
J. To
´
th
1,2
, M. Kova
´
cs
1
, F. Wang
1
, L. Nyitray
2
and J. R. Sellers
1
1
Laboratory of Molecular Physiology, NHLBI, National Institutes
of Health, Bethesda, MD, United States of America,
2
Bioke
´
miai
Tansze
´
k, ELTE, Budapest, Hungary. E-mail:
Myosin V is the best characterized vesicle transporter myosin
motor in vertebrates but it is unknown whether all members of

the myosin V family share a common, evolutionarily conserved,
mechanism. To address this question, we performed steady state
and transient kinetic measurements on a recombinant Drosophila
myosin V S1 fragment. Interestingly, none of the product release
steps can be identified as a single rate-limiting step. The ADP
release rate constant from acto-S1, which is the bottleneck in the
vertebrate myosin V ATPase cycle, appears to be 5-fold higher
than the maximal steady-state ATPase activity. The Pi release
rate constant from the acto-S1-products complex is even faster,
implying that Drosophila myosin V does not follow a myosin
II-type kinetic pattern, either. The key features of the ATPase
cycle are the low ATP hydrolysis equilibrium constant (0.33) and
the high steady-state actin attachment (96% at 5 m actin). The
behavior of double-headed Drosophila myosin V in an in vitro
motility assay suggests that a single molecule of myosin V cannot
take multiple steps on the actin filament while being attached to
it. We propose a mechanism whereby the ATP hydrolysis and
the so-called weak actin-binding equilibria define the steady-state
ATPase rate rather than the product release process. This mech-
anism suggests that Drosophila myosin V will not act as a single
molecule vesicle transporter, in contrast to vertebrate myosin Vs.
However, it may be able to perform processive transport if pre-
sent in small ensembles on the surface of its cargo.
F1-016P
A novel centrosomal protein, Cep57 stabilizes
microtubules in vivo and binds FGF-2 in vitro.
Y. Yoshitake, N. Tsuruoka and A. Iwai
Department of Biochemistry, Kanazawa Medical University, Uchi-
nada, Ishikawa Japan. E-mail:
We have identified a gene encoding a novel protein that binds

microtubules (MT) and fibroblast growth factor-2 (FGF-2). It
was found as FGF-2 associated protein in the course of studying
the function of the 24kDa isoform of FGF-2 in nuclei by yeast
two-hybrid screening. This gene shares a sequence similarity with
an uncharacterized cDNA reported in GenBank as KIAA0092
and the encoded protein is identical to Cep57, a coiled-coil pro-
Abstracts
336
tein in the centrosomes. The occurrence of Cep57 in human cen-
trosomes has been recently reported but its function is unknown.
GFP-fused protein over-expressed in HeLa cells was predomin-
antly observed as filamentous structures and co-localized with
MT. Weekly-expressed tagged proteins were co-localized with
MT as spindle, spindle pole and MT in the midbody of the mito-
tic cells and also localized in the centrosomes of interphase cells.
Treatments disturbing the MT network did not have any effect
on this protein distribution in the cells. Over-expressed Cep 57
protected the MT structure against those treatments. Therefore,
Cep 57 stabilizes MT. Recombinant Cep57 protein bound both
isoforms of FGF-2 (24kDa and 18kDa) in vitro. However, the
localization patterns of endogenous Cep57 observed by immuno-
fluorescent staining using the antibodies differed in both patterns
of FGF-2 isoforms. siRNA to Cep57 induced a morphological
change (spindle-like shape with a poor cytoskeletal MT network)
and decreased cell growth in HeLa cells. Cep57 may work as a
MT stabilizing protein in the centrosomes and spindle pole, and
then play a role on MT network formation and cell division.
F1-017P
A novel dynamic model for actin
polymerization based on the idea of

self-catalysis
J. Zhang
1,2
and Y. L. Wang
2
1
Department Biology, Chongqing University of Medical Sciences,
Chongqing City, PR China,
2
Key Lab of Biomechanics & Tissue
Engineering, Institute of Bioengineering, Chongqing University,
Chongqing City, PR China. E-mail:
More and more researches have recently demonstrated that there
exist multiple pathways in actin polymerization. In order to eluci-
date the macroscopic dynamics of actin assembly, a novel phe-
nomenological theory for actin polymerizing under simple
thermodynamic conditions has been put forward. The theoretical
model, called self-catalytic dynamics, emphasizes that the process
of actin polymerizing can be thought of an autocatalytic chain
reactions. To set up the quantified kinetic equations group, three
basic postulations is employed, which are in good agreement with
the fundamental physical rules and published literatures: (i) the
actin monomers driven by Brownian motion in solution can
interact with each other, and further form actin dimer via non-
covalent association under certain molecular colliding conditions;
(ii) the formed actin dimer and higher polymer spontaneously
own catalyzing ability and can catalyze free actin monomer or
dimer changing into actin residues (that refer to the G-actin
monomers existing in actin filaments); (iii) the polymerization/
depolymerization processes can go on synchronously at the both

ends (generally, the barbed and pointed ends, sometimes, also at
the side of the elongating filaments) with different polymeriza-
tion/depolymerization rates. According to the analyses and basic
enzyme catalysis principles, a scheme of kinetic differential equa-
tions can be established with definite initial conditions. By strug-
gling out the integral curves of the equations group, we get a
series of dynamic functions, via which we can accurately calculate
the length distribution of filaments, the concentration of free
G-actin and polymerized filaments versus the polymerizing time,
and also quantificationally explain the treadmilling phenomenon
of actin turnover.
F2–Single Molecule Biochemistry and Mechanics
F2-001
Recent Advances in Single Molecule
Biophysics
C. Bustamante
Department of Physics, University of California, Berkeley, CA,
USA. E-mail:
I will present our recent results on the packaging of DNA by the
connector motor at the base of the head of bacteriophage Ø29.
As part of their infection cycle, many viruses must package their
newly replicated genomes inside a protein capsid to insure its
proper transport and delivery to other host cells. Bacteriophage
Ø29 packages its 6.6 mm long double-stranded DNA into a
42 nm dia · 54 nm high capsid via a portal complex that hydro-
lyses ATP. This process is remarkable because entropic, electro-
static, and bending energies of the DNA must be overcome to
package the DNA to near-crystalline density. We have used opti-
cal tweezers to pull on single DNA molecules as they are pack-
aged, thus demonstrating that the portal complex is a force

generating motor. We find that this motor can work against
loads of up to 57 picoNewtons on average, making it one of the
strongest molecular motors ever reported. Movements of over
5 mm are observed, indicating high processivity. Pauses and slips
also occur, particularly at higher forces. We establish the force–
velocity relationship of the motor and find that the rate-limiting
step of the motor’s cycle is force dependent even at low loads.
Interestingly, the packaging rate decreases as the prohead is
filled, indicating that an internal pressure builds up due to DNA
compression. We estimate that at the end of the packaging the
capsid pressure is 6 MegaPascals, corresponding to an internal
force of 50 pN acting on the motor. The biological implications
of this internal pressure and the mechano-chemical efficiency of
the engine are discussed.
F2-002
The complete folding/unfolding trajectory
of a protein captured with single molecule
force-clamp spectroscopy
J. M. Fernandez
Biological Sciences, Columbia University, New York, NY 10027,
USA. E-mail:
A dense network of interconnected proteins and carbohydrates
forms the complex mechanical scaffold of living tissues. The
recently developed technique of single molecule force spectro-
scopy has enabled a detailed analysis of the force-induced confor-
mations of these molecules and the determinants of their
mechanical stability. These studies provide some of the basic
knowledge required to understand the mechanical interactions
that define all biological organisms. The application of mechanical
force to biological polymers produces conformations that are dif-

ferent than those that have been investigated by chemical or ther-
mal denaturation, and are inaccessible to conventional methods of
measurement such as NMR spectroscopy and X-ray crystallogra-
phy. Force-induced conformational transitions are physiologically
Abstracts
337
important, and offer novel perspectives on the structure of bio-
molecules. Recent developments in single molecule force spectros-
copy have enabled study of the full unfolding and refolding
pathways of a protein under force-clamp conditions. Mechanical
unfolding of a protein placed under a constant stretching force
allows us to obtain precise information about the pathway and
kinetics of unfolding. Force-quench experiments capture for the
first time the full folding pathway of a protein. Due to the force
of gravity and the need of living organisms to perform mechan-
ical work, mechanical stretching is most likely to have played a
role in the evolution of proteins. By contrast, the large changes
in temperature or chemical denaturants commonly employed in
protein folding studies are not found in living cells. Hence, the
mechanical unfolding/folding trajectories captured by force-clamp
spectroscopy reflect much more closely the conformations and
pathways of proteins in vivo, compared to those obtained by
means of thermal or chemical manipulations. This lecture will
focus mostly on the use of these novel techniques to study the
dynamic changes that proteins undergo in response to a mechan-
ical stretching force.
References
1. Fernandez JM, Li HB. Force-clamp spectroscopy monitors the
folding trajectory of a single protein. Science 2004; 303: 1674–
1678.

2. Carrion-Vazquez M, Li H, Lu H, Marszalek PE, Oberhauser
AF, Fernandez JM. The mechanical stability of ubiquitin is
linkage dependent. Nature Structural Biology 2003; 10(9): 738–
743.
3. Li HB, Linke WA, Oberhauser AF, Carrion-Vazquez M,
Kerkvliet JG, Lu H, Marszalek PE, Fernandez JM. Reverse
engineering of the giant muscle protein titin. Nature 2002; 418:
998–1002.
F2-003
Tethering single protein concatamers to
enable measurement of thermal noise
response by Atomic Force Microscopy (AFM)
K. Byrne, M. Kawakami, B. Khatri and A. Smith
Laboratory of Molecular and Nanoscale Physics, Department of
Physics and Astronomy, University of Leeds, Leeds,
West Yorkshire, UK. E-mail:
The thermal noise response of single molecules under tension can
be measured using AFM and used to reveal both the conserva-
tive and dissipative response of the molecule. This recently devel-
oped method has been applied successfully to single sugar
molecules tethered via non-specific interactions between the
AFM cantilever and the base substrate. In order to apply the
method to single protein concatamers we have found that a more
durable means of tethering the molecules is required. Single pro-
tein molecules can be suspended between a cantilever and sub-
strate using non-specific tethering but the tether breaks before
sufficient data can be captured to give an accurate noise spec-
trum in all cases observed thus far. Tethering molecules through
covalent rather than non-specific bonds should enable longer dur-
ation experiments to be carried out with ease and has the added

advantage that the exact location of the tether point is known. A
protocol to tether a modified 5-domain concatamer of I27
between a gold coated cantilever and gold substrate via distinct
covalent links is being developed. The gold binding functionality
of the C-terminal cysteine residue and the ability of N-terminal
histadine residues to bind succinamide will be exploited. The
lysine residues have been removed from the concatamer so that
succinamide-amine bonds should only form at the N-terminal
end of the concatamer. A self assembled monolayer of the short
linker molecule Di-thio-bis-succinamidyl-propionate (DSP) on
gold will be used to tether the N-terminus. The short linker will
interfere only minimally in the thermal noise spectrum molecule
of interest.
F2-004
Single molecule and single paritcle imaging
in solution and in live cells
X. Zhuang
Department of Chemistry and Chemical Biology,
Harvard University, Cambridge, MA, USA.
E-mail:
Understanding the molecular mechanisms of complex biological
processes is one of the major goals in modern biology. As
molecular and cell biology get increasingly quantitative, a com-
prehensive understanding of biological processes at the molecular
level is becoming more readily accessible. However, roadblocks
still exist, among which is the challenge that we face in character-
izing the complex dynamics of biological processes. The existence
of multiple kinetic paths and transient intermediate states often
makes these processes difficult to dissect, as individual steps of a
multi-step process are typically not synchronized among mole-

cules. To tackle this problem, we are exploring optical imaging
techniques to monitor, in real-time, the behavior of individual
biological molecules and complexes, in vitro and in live cells. In
this talk, I will report our recent progress in the following two
areas. (i) Molecular mechanisms of viral infection: our single-
virus tracking experiments allow us to visualize the viral infection
process in real time, dissect individual stages of the viral entry
pathway, and obtain a better understanding of the molecular
mechanisms governing the influenza infection. (ii) Structural
dynamics of RNA and ribonucleoprotein enzymes: our single-
molecule studies provide critical insights into the molecular
mechanisms governing RNA structural dynamics, and the effects
of proteins on the structural dynamics of RNA enzymes.
F2-005
Single molecule analysis of protein
aggregation and prions by SIFT
U. Bertsch
1
, K. F. Winklhofer
2
, T. Hirschberger
3
, J. Bieschke
1
,
P. Weber
1
, F. U. Hartl
2
, P. Tavan

3
, J. Tatzelt
2
,
H. A. Kretzschmar
1
and A. Giese
1
1
ZNP, University of Munich, Munich, Germany,
2
Department of
Cellular Biochemistry, Max-Planck-Institute for Biochemistry,
Martinsried, Germany,
3
Department of BioMolecular Optics,
University of Munich, Munich, Germany.
E-mail:
Protein aggregation is a key event in a number of diseases such
as Alzheimer’s disease, Parkinson’s disease and prion diseases.
We present a general method to quantify and characterize pro-
tein aggregates by dual-colour scanning for intensely fluorescent
targets (SIFT). In addition to high sensitivity, this approach
offers a unique opportunity to study co-aggregation processes at
the single particle level in complex mixtures containing different
types of aggregates. In the case of prion diseases the prion pro-
tein (PrPC), a neuronal glycoprotein, undergoes a conformational
change from the normal, mainly alpha-helical conformation to a
disease-associated, mainly beta-sheeted scrapie-isoform (PrPSc),
which forms amyloid aggregates. This conversion, which is cru-

cial for disease progression, depends on direct PrPC/PrPSc inter-
action. We developed a high-throughput SIFT assay for the
identification of drugs, which interfere with this interaction at the
molecular level. Screening a library of 10 000 drug-like com-
pounds yielded 256 primary hits, 80 of which were confirmed by
Abstracts
338
dose–response curves with half-maximal inhibitory effects ranging
from 0.3 to 60 lm. Among these, six compounds displayed an
inhibitory effect on PrPSc propagation in scrapie-infected N2a
cells. Four of these candidate drugs share a N
´
-benzylidene-ben-
zohydrazide (NBB) core microstructure. Thus the combination of
high-throughput in vitro assay with the established cell culture
system provides a rapid and efficient method to identify new
anti-prion drugs, which corroborates that interaction of PrPC
and PrPSc is a crucial molecular step in the propagation of pri-
ons. Moreover, SIFT-based screening may facilitate the search
for drugs against other diseases linked to protein aggregation.
F2-006
Studying protein reaction kinetics by
fluorescence correlation spectroscopy in
microfluidic mixers
P. Galajda
1
, J. Puchalla
2
, R. Riehn
1

and R. H. Austin
1
1
Department of Physics, Princeton University, Princeton, NJ,
USA, ,
2
Department of Molecular Biology, Princeton University,
Princeton, NJ, USA. E-mail:
Fluorescence correlation spectroscopy is a powerful method to
gather information about the motion and interactions of mole-
cules and particles in solution. The autocorrelation of the fluores-
cence intensity fluctuation emerging from fluorophores traversing
a confined excitation/observation volume is used for analysis.
Usually either diffusion or convection dominates on the time-
scales of the fluctuations. Measurements lasts for seconds making
it inadequate for kinetic studies of most biomolecular reactions.
Here we present a method to overcome this limitation. Reactants
meet in a microfluidic diffusional mixer. A steady-state flow
establishes a direct mapping of the temporal evolution of the
reaction to the spatial position along the channel opening a way
for kinetic studies. The experimental conditions are set so that
we can extract both the characteristics of diffusion and flow from
the data. A variety of biological reactions might be studied by
the above technique from protein folding to DNA-protein inter-
actions.
F2-007P
Changes in cell morphology, viability,
proliferation and cytotoxicity associated with
helium–neon laser irradiation of diabetic
wounded human skin fibroblasts

H. Abrahamse, N. Houreld and D. Hawkins
Laser Research Unit, Faculty of Health, University of
Johannesburg, Johannesburg, Gauteng South Africa.
E-mail:
Low Level Laser Therapy (LLLT) is a form of phototherapy
used to promote wound healing in different clinical conditions.
Laser radiation has the capability to inhibit or stimulate cellular
activity in the absence of significant heating. Currently, no uni-
versally accepted theory has explained the mechanism of laser
biostimulation. Although not established, an alternative treat-
ment modality for diabetic wound healing includes LLLT. Laser
biostimulation of such wounds may be of benefit to patients by
reducing healing time. Structural, cellular and genetic events in
normal, wounded and unwounded diabetic induced human skin
fibroblasts was evaluated after exposing cells to increasing doses
of helium–neon (632.8 nm) laser irradiation. Structural changes
were evaluated by assessing colony formation, haptotaxis and
chemotaxis. Cellular changes were evaluated using cell viability,
(ATP), proliferation, (ALP), cytotoxicity (LDH) while the Comet
assay evaluated genetic integrity. Morphologically, wounded
diabetic fibroblasts responded to single exposure of 5 J/cm
2
with
the highest rate of haptotaxis and chemotaxis indicating a stimu-
latory effect while 16 J/cm
2
showed the lowest rate indicating an
inhibitory effect. Cell viability and proliferation changes indicated
a decrease at higher doses while a dose of 5.0 J/cm
2

appeared to
stimulate mitochondrial activity, which leads to normalization of
cell function with an increase in cell viability. A dose of 5 J/cm
2
showed the highest rate of cell viability and proliferation with a
minimal amount of damage to the cell membrane or DNA, sup-
porting morphological evidence. In conclusion, LLLT of 5 J/cm
2
stimulates migration, proliferation, and metabolism of wounded
fibroblasts to accelerate wound closure.
F2-008P
Methionine sulfoxide reductase from class A:
aminoacids involved in catalysis of the
reductase step and in substrate binding
S. Boschi-Muller, M. Antoine, A. Gand and G. Branlant
Laboratoire MAEM UMR 7567 CNRS-UHP, Universite
´
Henri
Poincare
´
Nancy I, vandoeuvre-le
`
s-Nancy, France.
E-mail:
Methionine sulfoxide reductases (Msrs) are ubiquitous enzymes
that reduce protein-bound methionine sulfoxide back to Met in
the presence of thioredoxine. In vivo, the role of Msrs is des-
cribed as essential in protecting cells against oxidative damages,
and to play a role in infection of cells by pathogenic bacteria.
There exist two structurally-unrelated classes of Msrs, called

MsrA and MsrB. MsrA is specific for the S epimer of MetSO,
whereas MsrB reduces the R epimer. Both Msrs present a similar
catalytic mechanism of sulfoxide reduction by thiols via the sulf-
enic acid chemistry. The rate of the reductase step that leads to
the formation of the sulfenic acid intermediate is fast. This sup-
ports an activation of the catalytic Cys and an increase of the
electrophilic character of the sulfur atom of the sulfoxide sub-
strate via an acid catalyst from the active site. Both activations
should favor the efficiency of the reductase process. Data will be
presented which illustrate the role of some amino acids involved
in the catalytic mechanism of the reductase step and in the sub-
strate recognition of MsrAs, specifically in the formation of the
hydrophobic pocket responsible for the methyl group recogni-
tion.
F2-009P
Role of Induced cAMP Early Repressors in B/K
Gene Transcription
M H. Choi, Y M. Jang, Y S. Jang and O J. Kwon
Department of Biochemistry, The Catholic University of Korea,
College of Medicine, Seoul, South Korea.
E-mail:
B/K protein, structurally classified as a member of double C2
domain proteins, is highly expressed in PC12 cells. We previously
found that forskolin decreased the expression of both B/K
mRNA and protein by PKA-dependent mechanism, and that
inducible cAMP early repressor (ICER) might be involved in that
process via the interaction with a CRE-like domain that is
located in the promoter region of B/K gene. Here, we showed
the direct involvement of ICER in the mechanism of forskolin-
induced down-regulation of B/K gene expression in PC12 cells.

To investigate the role of ICER, we’ve successfully cloned four
types of rat ICER (I, Ig, II and IIg). Transient transfection
(24 h) of each construct resulted in the successful expression of
ICERs, and the expression level was even higher than that
induced by 3-h treatment of forskolin, the time when the expres-
Abstracts
339
sion of ICER increased to the highest level. In promoter assay,
ICER suppressed the promoter activity of B/K in CHO-K1 cells
that were cotransfected with the construct containing B/K CRE-
like sequence. Inhibition level might not be related with the iso-
types of ICER but with the expression level in the cells, suggest-
ing the absence of isotype specificity of ICER in the repression of
B/K gene transcription. Effect of ICER did not observed in the
cells cotransfected with the B/K construct having mutated CRE-
like sequence (AC:TG). Direct binding of B/K CRE-like
sequence to ICER was also observed in chromatin immunopre-
cipitation experiment. These results strongly suggest that ICER
may be the principal suppressor molecule in the process of B/K
gene expression.
F2-010P
Comparative structural and conformational
studies on two forms of beta lactoglobulin
(A and B) upon interaction with lead ion
A. Divsalar and A. A. Saboury
Laboratory of Biophysical Chemistry, Institute of Biochemistry
and Biophysics, University of Tehran, Tehran, Iran.
E-mail:
One of the most widely studied and fascinating proteins of animal
origin is Beta lactoglobulin (blg). This polypeptide, with 150–170

amino acids (depending on the species), is an abundant compo-
nent of the milk of most animals, although it has not been detec-
ted in human milk. Bovine blg consists of 162 amino acid residues
and contains two disulfide bonds and a free thiol. Its interaction
with a great variety of hydrophobic ligands has led to its inclusion
in the lipocalin (transport protein) superfamily. blg is a dimeric
native conformation at neutral pH, while the conformation at pH
2 is a monomeric but still native. Thermodynamic studies have
been made on the effect of Pb
2+
on the structure of two forms of
blg (A and B) in 50 mm sodium chloride at 27 °C using UV spec-
trophotometry, circular dichroism (CD) and fluorescence spectros-
copy. UV spectrophotometry studies show that the protein
absorbance changes due to the increase concentration of ion lead
are a biphasic curve for blg-A form and a monophasic sigmoid
curve for blg-B form. Far CD studies do not show any change on
the secondary structure of blg-A upon interaction with concentra-
tions of 15.54 and 35.8 mm of Pb
2+
, while far CD studies show
considerable change in the secondary structure of blg-B by
increasing the concentration of Pb
2+
. Near CD and fluorescence
spectroscopy studies show considerable change in the tertiary
structure of Blg-A related to the interaction with lead ion, but
these studies do not show any change in the tertiary structure of
blg-B. Different behaviors of blg (A and B) upon interaction with
lead ion, related to different native structure of both proteins.

F2-011P
Strong specific binding forces between
adhesive extracellular proteoglycan
carbohydrates
I. Bucior
1
, S. Garcia-Manyes
2
, R. Ros
3
, F. Sanz
2
,
D. Anselmetti
3
, M. M. Burger
1
and X. Ferna
`
ndez-Busquets
4
1
Friedrich Miescher-Institut, Basel, Switzerland,
2
Department of
Physical Chemistry, University of Barcelona, Barcelona, Spain,
3
Experimental Biophysics, University of Bielefeld, Bielefeld,
Germany,
4

Laboratory of Nanobioengineering, Barcelona Science
Park, University of Barcelona, Barcelona, Spain.
E-mail:
Specific carbohydrate-carbohydrate interactions are rarely repor-
ted in biologically relevant situations such as cell recognition.
However, carbohydrate structures have immense structural diver-
sity, a ubiquitous distribution in vertebrate and invertebrate tis-
sues, and are associated with the cell surface, as required of cell
recognition molecules. Carbohydrate–carbohydrate interactions
are characterized by relatively weak forces which, when multi-
merized, can be easily potentiated by orders of magnitude, repre-
senting a highly versatile form of cell recognition and adhesion
given the extraordinary plasticity of their structures. Sponge cells
associate in a species-specific process through multivalent interac-
tions of carbohydrate structures on a type of extracellular bifunc-
tional proteoglycan molecules termed spongicans. In the marine
sponge Microciona prolifera the spongican molecule, Microciona
aggregation factor (MAF), self-interacts via a Ca
2+
-dependent
interaction. We have dissected MAF into its components, track-
ing the individual self-binding units down to the circular core of
the sunburst-like molecule and further down to a 200-kDa glycan
(g200). Force spectroscopy data and surface plasmon resonance
measurements reveal that the carbohydrate self adhesion is highly
specific when compared with the binding to other sulfated carbo-
hydrates such as chondroitin sulfate. The strength of the binding
per surface area between two spongican molecules is comparable
to that of focal contacts between vertebrate cells. The results
obtained reveal surprisingly high forces and selectivity for this

most ancient cell adhesion system, and the existence of poly-
morphism between g200 purified from different individuals, thus
representing the first indication of specific carbohydrate–carbohy-
drate interactions implicated in the discrimination between cells
of different individual origin within a single species.
Acknowledgments: This work was supported by grant
BIO2002-00128 from the Ministerio de Ciencia y Tecnologı
´
a,
Spain, that included FEDER funds.
F2-012P
Scanning probe microscopy as tool to study
the stability of azurin in air
V. Frascerra, G. Maruccio, V. Arima, L. del Mercato,
P. P. Pompa, F. Calabi, R. Cingolani and R. Rinaldi
National Nanotechnology Laboratory of INFM, Ingegneria dell’
Innovazione, University of Lecce, Lecce, Italy.
E-mail:
Scanning probe microscopy (SPM) comprises a family of tech-
niques that can be used as tool to study biomolecules morphol-
ogy and properties on the atomic scale. We focus on the use of
the SPM to study the azurin stability in air. Azurin is a metallo-
protein involved in electron transfer during denitrification path-
way of the bacterium Pseudomonas aeruginosa. Thanks to its
electron transfer properties, azurin and its physiological partners
could be employed for the implementation of planar biosensors,
with enhanced sensitivity and selectivity, exploiting a monolayer
of proteins that should act as an optimal surface for the analyte
molecules to react with. For this purpose it is fundamental that
the immobilized proteins preserve their native conformation and

properties, especially in ambient conditions. First, Scanning
Tunneling Microscopy (STM) allows us to achieve a very high
resolution and to probe the electronic properties of Au (111)-
immobilized Azurin, both in buffer solution and in air. Our
results demonstrates that immobilized proteins preserve their
electron transfer function. Secondly, the force distance mode of
atomic force microscopy has been used to measure the interac-
tion between tip and SiO2-immobilized azurin by mercaptosi-
lanes. The force curves show that the adhesion of the tip to the
azurin is weaker than to the SiO2 and silanized surface. The
mechanical properties of azurin could be deduced by analyzing
these curves and correlated with the native conformation of the
protein.
Abstracts
340
F2-013P
Hydrogen peroxide mediates Rac1 activation
of S6K1
J W. Han
1
, Y G. Chung
1
, S. J. Paek
1
, S. H. Park
1
, J S. You
1
,
J K. Kang

1
, Y. M. Kim
2
, H. Y. Lee
3
, Y. K. Kim
4
and
H. W. Lee
1
1
Lab of Biochem & Mol Biol, College of Pharmacy, Sungkyunk-
wan University, Suwon, Gyeonggi-do South Korea,
2
College of
Pharmacy, Duksungsung Women’s University, Seoul, South Korea,
3
College of Medicine, Konyang University, Nonsan, Chungcheong-
nam-do South Korea,
4
College of Medicine, Kwandong University,
Gangneung, Gangwon-do South Korea.
E-mail:
We previously reported that hydrogen peroxide (H
2
O
2
) mediates
mitogen activation of ribosomal protein S6 kinase 1 (S6K1)
which plays an important role in cell proliferation and growth.

In this study, we investigated a possible role of hydrogen perox-
ide as a molecular linker in Rac1 activation of S6K1. Overexpres-
sion of recombinant catalase in NIH-3T3 cells led to the drastic
inhibition of H
2
O
2
production by PDGF, which was accompan-
ied by a decrease in S6K1 activity. Similarly, PDGF activation of
S6K1 was significantly inhibited by transient transfection or sta-
ble transfection of the cells with a dominant-negative Rac1
(Rac1N17), while overexpression of constitutively active Rac1
(Rac1V12) in the cells led to an increase in basal activity of
S6K1. In addition, stable transfection of Rat2 cells with
Rac1N17 dramatically attenuated the H
2
O
2
production by
PDGF as compared with that in the control cells. In contrast,
Rat2 cells stably transfected with Rac1V12 produced high level
of H
2
O
2
in the absence of PDGF, comparable to that in the con-
trol cells stimulated with PDGF. More importantly, elimination
of H
2
O

2
produced in Rat2 cells overexpressing Rac1V12 inhib-
ited the Rac1V12 activation of S6K1, indicating the possible role
of H
2
O
2
as a mediator in the activation of S6K1 by Rac1. How-
ever, H
2
O
2
could be also produced via other pathway which is
independent of Rac1 or PI3K, because in Rat2 cells stably trans-
fected with Rac1N17, H
2
O
2
could be produced by arsenite which
has been shown to be a stimulator of H
2
O
2
production. Taken
together, these results suggest that H
2
O
2
plays a pivotal role as a
mediator in Rac1 activation of S6K1.

F2-014P
Reaction of hydroxymethylarginine and
endogenous formaldehyde with
tetrahydrofolate producing N
5
,N
10
-methylene
tetrahydrofolate coenzyme
L. Hulla
´
n
1
, L. Tre
´
zl
2
,Z.M.Ja
´
szay
3
, M. Tejeda
4
, A. Csiba
5
,
J. Bariska
6
and T. Szarvas
7

1
Department of Biochemistry, National Institute of Oncology,
Budapest, Hungary,
2
Department of Organic Chemical Technol-
ogy, Budapest University of Technology and Economics, Budapest,
Hungary,
3
Organic Chemical Technological Research Group, Hun-
garian Academy of Science, Budapest, Hungary,
4
Department of
Experimental Pharmacology, National Institute of Oncology,
Budapest, Hungary,
5
Veterinary and Food Control Station, Buda-
pest, Hungary,
6
Central Laboratory, National Institute of Rheuma-
tology and Physiotherapy, Budapest, Hungary,
7
Institute of
Isotopes Co. Ltd., Budapest, Hungary. E-mail:
The endogenous formaldehyde (E-CH
2
O) means the always exist-
ing CH
2
O content of various biological samples. E-CH
2

O is sup-
posed to be bound to biomolecules, because it has no toxic
effect. N
G
-hydroxymethyl derivatives of arginine (HMA) synthes-
ized by Tre
´
zl and co-workers from CH
2
O and L-arginine may
belong to that type of molecules. HMA was found in human
blood and urine, and in plants (leaves, fruits and vegetables) in
higher concentration than in human samples since HMA is one
of the products of photosynthesis. We have proved that the reac-
tion of HMA and THF produces – under physiological condi-
tions – N
5
,N
10
-methylene tetrahydrofolate (CH
2
THF), the
coenzyme of thymidylate synthase (TS). In recent work
13
C-HMA was synthesized and the reaction of
13
C-HMA and
THF was compared to that of
13
C-CH

2
O and THF by
13
C-NMR
in the reaction mixture for producing of CH
2
-THF according to
the Roberts-method which is used for the determination of TS
activities. The two reactions were also compared to the reaction
between E-CH
2
O and THF. To test the CH
2
-THF product meas-
urement of TS activities was applied using either cytosol of P388
lymphoid leukemia tumor, or TS purified from Lactobacillus ca-
sei. Both the NMR spectra and the enzyme assays proved that
the reactions produced CH
2
-THF, however, the rate of reaction
by free CH
2
O is higher than those of the two other reactions.
The results suggest that TS and probably other proteins are the
most important molecules in keeping of E-CH
2
O in bound form.
Acknowledgment: This work was supported by National Sci-
ence Foundation OTKA Grant No. T034245.
F2-015P

Co-operation of H
2
O
2
-mediated ERK activation
with Smad pathway in TGF-b1 induction of
p21
WAF1/Cip1
Y. K. Kim
1
, E. K. Lee
2
, J. W. Park
2
, J. Y. Lee
2
, H. Y. Jung
2
,
H. J. Kim
2
, J. H. Park
2
, S N. Kim
3
, H. Y. Lee
4
, H. W. Lee
2
and J W. Han

2
1
College of Medicine, Kwandong University, Gangneung, South
Korea,
2
College of Pharmacy, Sungkyunkwan University, Suwon,
South Korea,
3
Skin Research Institute, Amore-Pacific Corporation,
Yong-In, South Korea,
4
College of Medicine, Konyang University,
Nonsan, South Korea. E-mail:
Although it has been demonstrated that p21
WAF1/Cip1
could be
induced by transforming growth factor-b1 (TGF-b1) in a Smad-
dependent manner, the cross-talk of Smad signaling pathway
with other signaling pathways still remains poorly understood. In
this study, we investigated a possible role of hydrogen peroxide
(H
2
O
2
)-ERK pathway in TGF-b1 induction of p21
WAF1/Cip1
in
human keratinocytes HaCaT cells. Using pharmacological inhibi-
tors specific for MAP kinase family members, we found that
ERK, but not JNK or p38, is required for TGF-b1 induction of

p21
WAF1/Cip1
. ERK activation by TGF-b1 was significantly
attenuated by treatment with N-acetyl-L-cysteine or catalase,
indicating that reactive oxygen species (ROS) generated by TGF-
b1, mainly H
2
O
2
, stimulates ERK signaling pathway to induce
the p21
WAF1/Cip1
expression. In support of this, TGF-b1 stimula-
tion caused an increase in intracellular ROS level, which was
completely abolished by pre-treatment with catalase. ERK activa-
tion does not appear to be associated with nuclear translocation
of Smad-3, because ERK inhibition did not affect nuclear trans-
location of Smads by TGF-b1, and H
2
O
2
treatment alone did
not cause nuclear translocation of Smad-3. On the other hand,
ERK inhibition led to the disruption of interaction between
Smad-3 and Sp1 induced by TGF-b1, indicating that ERK signa-
ling pathway might be necessary for their interaction. Taken
together, these results suggest that activation of H
2
O
2

-mediated
ERK signaling pathway is required for p21
WAF1/Cip1
expression
by TGF-b1 and led us to propose a co-operative model whereby
TGF-b1-induced receptor activation stimulates not only a Smad
pathway but also a parallel H
2
O
2
-mediated ERK pathway that
acts as a key determinant for association between Smads and
Sp1 transcription factor.
Abstracts
341
F2-016P
Structural dynamics of amyloid explored by
manipulating individual fibrils
M. S. Kellermayer
1
, L. Grama
1
,A
´
. Karsai
1
, A. Nagy
1
, A. Kahn
1

,
Z. Datki
2
and B. Penke
2
1
Nanobiotechnology and Single Molecule Biophysics Group,
Department of Biophysics, University of Pe
´
cs, Pe
´
cs, Hungary,
2
Protein Research Group of Hungarian Academy of Sciences,
Department of Medicinal Chemistry, University of Szeged, Szeged,
Hungary. E-mail:
Amyloid fibrils are self-associating filamentous structures, the
deposition of which is considered to be one of the most import-
ant factors in the pathogenesis of Alzheimer’s disease and other
disorders. We used single-molecule manipulation methods to
explore the mechanics and structural dynamics of amyloid fibrils.
In mechanically-manipulated amyloid fibrils, formed from either
Ab1–40 or Ab25–35 peptides, elastic structures can be
‘‘unzipped’’ from the fibril with constant forces in a process that
resembles mechanical polymer desorption from a surface. The
constant-force plateaus were often superimposed to form hierar-
chical force staircases. The fundamental unzipping forces were
different for Ab1–40 (33 pN ± 7 SD) and Ab25–35 (41 pN ± 7
SD) fibrils, indicating that different interactions contribute to
their mechanical stability. Unzipping was fully reversible across a

wide range of stretch rates provided that elastic coupling between
bound and dissociated states was maintained. In situ force spectr-
oscopy, in which amyloid fibrils were imaged pior to and follow-
ing mechanical stretch, revealed that the filamentous structures
unzipped from the surface were likely to be b-sheets rather than
protofilaments or entire fibrils. Amyloid mechanics were simula-
ted well with a simple two-state model in which the mechanically
loaded, elastic b-sheet rapidly fluctuates between the surface-
bound and dissociated states. The rapid, cooperative zipping
together of b-sheets could be an important mechanism behind
the self-assembly of amyloid fibrils. The repetitive force staircase
patterns contribute to a mechanical fingerprint that could be util-
ized in the characterization different amyloid fibrils.
F2-017P
Incorporation of the mechanosensitive channel
(MscL) and the amyloid Abeta[1–40] peptide
by vesicle fusion into planar lipid bilayers
G. P. Mendes
1,2
, M. R. de Planque
1
, S. L. Grage
3
, R. Berry
2
and
A. Watts
1
1
Biomembrane Structure Unit, Department of Biochemistry,

University of Oxford, Oxford, UK,
2
Bionanotechnology IRC,
Department of Physics, University of Oxford, Oxford, UK,
3
Forschungszentrum Karlsruhe, Karlsruhe, Germany.
E-mail:
Monitoring the dynamics of fluorescently labeled single ion chan-
nels and other biomolecules in real-time and at nanometer resolu-
tion can provide valuable kinetical, pharmacological and
functional information on biomembranes. However, the interdis-
ciplinary resources involved in such a project have not yet
reached a point where the methodology can be routinely used,
and further advances in instrumentation and in the biochemical
methods are required. For example, single molecule imaging of
membrane proteins requires incorporation of protein in con-
trolled amounts into planar bilayer lipid membranes (BLMs).
Here we present a development in protein incorporation into
BLMs using vesicle fusion. The amount of protein or peptide
delivered to BLMs is controlled using proteoliposomes containing
ergosterol and nystatin, where each vesicle fusion event is signa-
led by the incorporation of a transient nystatin-ergosterol chan-
nel in the BLM, observed as a decaying ion current spike. We
have optimized this method to control the extent of vesicle fusion
and the time scale of the fusion marker. The ion channel activity
of the mechanosensitive channel (MscL) and the amyloid peptide
Abeta[1–40] (implicated in Alzheimer’s disease) are being investi-
gated under well-defined protein-to-lipid ratios, lipid composition
and salt gradients.
F2-018P

Minimum energy paths for substrate binding
to heme proteins
K. W. Olsen, P. L. Kramer and S. D. Golden
Department of Chemistry, Loyola University Chicago, Chicago,
IL, USA. E-mail:
The conjugate peak refinement (CPR) method (S. Fischer & M.
Karplus, Chem Phys Let 1992; 194: 252) determines the path
between two known conformations. We have used CPR to obtain
paths for ligand binding to heme proteins. The truncated hemo-
globin, trHbN, from Mycobacterium tuberculosis belongs to the
hemoglobin superfamily and may have functional roles in NO
detoxification, long-term ligand storage, O
2
/NO chemistry, and
O
2
delivery under hypoxic conditions. Its tertiary structure is
based on a 2-on-2 a helical sandwich rather than the 3-on-3 a
helical sandwich of the classical hemoglobin fold. A tunnel sys-
tem through the protein matrix may be used for ligand diffusion
to and from the heme distal site. Seven different ligand binding
pathways were examined by CPR. One new branch of the exist-
ing tunnel system was discovered, two non-tunnel pathways were
found, and the AB-turn/GH-turn and G-helix/H-helix tunnel sys-
tems were confirmed. CPR has also been used to identify a path
between the unbound and bound states of nitrophorin 4 (NP4)
to understand how NO binds. NP4 is a small heme protein in the
insect Rhodnius prolixus. The calculated path identifies an order
of movements for the conformational change including the two
major shifts seen in loops A–B and G–H and the packing of

hydrophobic groups around the NO ligand. The hydrophobic
residues surrounding the heme move almost continuously during
the conformational change. The ruffled configuration of the heme
in the bound state is partially initiated by movement of the prop-
ionate groups caused by neighboring residues.
F2-019P
Evaluation of c-Kit and PDGF-R signaling in
breast cancer cells
A. E. Roussidis, A. D. Theocharis and N. K. Karamanos
Biochemistry, Chemistry, Patras, Patras, Greece.
E-mail:
Signal transduction through tyrosine kinase receptors (RTKs)
regulates most of cell functions as a number of growth factors
and cytocines transmit their signal using these receptors. More-
over, aberrant tyrosine kinase activity (overexpression and/or
mutation) is associated with various hyper-proliferative diseases
and cancer. RTKs are therefore, considered as molecular targets
for ideal therapy. Several tyrosine kinase inhibitors are already
examined for the treatment of different type of cancers and some
of them exhibit quite hopeful results. Current bibliographic
results on the expression of c-Kit in breast cancer are quite con-
troversial and thus SCF-mediated effects have been poorly stud-
ied. Already a number of molecular inhibitors are directed to
RTK in a number of cancers. In this study we examine whether
c-kit and PDGF-R are potent molecular targets in breast cancer.
Five cell lines, two estrogen receptor positive, one with low inva-
sion potential (MCF-7) one of high invasiveness (ZR-75-1) and
Abstracts
342
two estrogen receptor negative with high invasion potential

(MDA-MB-231) and one with low invasion potential (BT-20),
and also normal cells MCF-12A were used. The expression of
PDGF-R and c-Kit, their ligands, MMP-9 and signal transduc-
tion pathways PI3/Akt and MAPK were studied by RT-PCT and
western blotting and immunocytochemistry. The three receptors
are expressed in all the breast cancer cell lines that were studied.
Exogenous SCF and PDGF-BB induced cell proliferation enhan-
cing PI3/Akt and MAPK pathways which were naturally activa-
ted probably through autocrine regulation. More over SCF
induced the expression of MMP-9 and enhanced the invasive
potential of breast cancer cell lines. Thus it is believed that signa-
ling through these RTKs may be potent molecular target for the
therapy of estrogen receptor positive and negative breast cancer.
F2-020P
Towards in vivo dynamical observations of
individual protein membrane using X-rays
Y. C. Sasaki
1,2
1
Biomedical Group, SPring-8/JASRI, Sayou-gun, Japan,
2
SASAKI-team, JST/CREST, Sayou-gun, Japan.
E-mail:
Recently, we succeeded in vitro time-resolved X-ray observations
of picometer-scale slow Brownian motions of individual protein
molecules in aqueous solutions [1]. In this new single molecular
detection system, which we call Diffracted X-ray Tracking
(DXT), we observed the rotating motions of an individual nano-
crystal, which is labeled to the specific site in individual protein
molecules, using time-resolved Laue diffraction. DXT can mon-

itor dynamics of the individual molecules or specific sites in indi-
vidual single protein molecules. Now, we try to observe
individual DNA molecules, myosin head molecules, denatured
proteins, protein membranes (Bacteriorhodopsin (BR) [2] and
KcsA) and actin filament. Especially, in the case of BR, we
observed Brownian motions and momentarily structural change
of single BR, which is a light-driven proton pumping found in
the purple membrane, in the light irradiation (560 nm). We have
consequently confirmed Brownian motions, and that the average
size of the momentarily structural changes by light irradiation in
35th residue of BR was 76.0(+/–)48.2 pm. Additionally, we
observed many mutants of BR. In order to achieve higher sensi-
tive detection of Laue spots from nanocrystals, both crystalliza-
tion and sizes of nanocrystals, which are labeled with single
molecules, are one of important factors in DXT. We have
recently succeeded in fabricating gold nanocrystals (20 nm) by
vacuum evaporations.
References
1. Sasaki YC et al. Picometre-scale dynamical X-ray imaging of
single DNA molecules. Phys Rev Lett 2001; 87: 248102.
2. Okumura Y, Oka T, Kataoka M, Taniguchi Y, Sasaki YC.
Picometer-scale dynamical observations of individual mem-
brane proteins: the case of bacteriorhodopsin. Phys Rev E
2004; 70: 021917.
F2-021P
Differentiation induced by DADS on human
gastric cancer cell
H. Ling, L Y. Zhang, Q. Su and Y. Song
Institute of Oncology, Medical College, Nanhua University,
Hengyang, Hunan PR China. E-mail:

Abstract DADS, naturally occurring organosulfide from garlic,
was very effective in inhibiting the in vitro growth of colon, lung,
breast and leukemia tumor cell lines. Previous study reported
that DADS had inhibitory effect on human gastric cancer
MGC803 cell line, and the effect of inhibition is significant and
exhibits a dose-dependent manner, and also induced G2/M arrest
which involved activation of p38 MAP kinase pathways; Further-
more, DADS had the differentiation induction effect on HL-60
cells by long time treatment. In this study, MTT assay, Electron
microscope, ConA-mediated cell agglutination, Alkaline phos-
phatase (ALP) specific activity detection and Scrape-loading and
dye transfer technique (SLDT ª) were analyzed to elucidate pro-
liferation inhibition and differentiation induction of DADS-
induced MGC803 cell line. Results demonstrated that exposure
to DADS from 20 to 35 mg/l MGC803 cells had less absorbance
value than control, which exhibited a dose-dependent model
(P < 0.05), whereas DADS induced-growth inhibition ratio
enhanced from 11.3, 35.2, 50.7 to 60.6% as assessed by the MTT
test (P < 0.05); it was revealed by transmission electron micro-
scope that the nucleo-cytoplasmic ratio of MGC803 cells was rel-
atively large – nucleus was irregular – many heterochromatins in
the nucleus and nucleolus were large-varied in shape and with
some nucleolus vacuoles. Organelles in cytoplasm were few; how-
ever, after 30 mg/l DADS treatment, the ultrastructure of
MGC803 cells had undergone a significant change. Cells took on
appearances as the number of microvilli on cell surface
decreased, and cells tendency to form gland and intercellular con-
junction structures also be showed, and cellular apparatus were
abundant in plasma maturely, nuclei, endoreclure and mitochon-
dria faded; the ribosal body existed as single and multi-forms,

when secreted mucoid granules and fatty dropt was also co-exis-
ted. The nucleo-cytoplasmic ratio lessened, the nuclear shape
became regular, most being round or oval, heterochromatin in
nucleus decreased while euchromatin increased – the volume of
nucleolus lessened. MGC803 cells showed some ultrastructural
characteristics of their normal relevant cells after being treated
with DADS. The ConA-mediated MGC803 cell agglutination
ratio indirectly reacted amount of microvilli on the face of cellu-
lar. ConA-mediated cell agglutination ratio lowered from 37.50
to 26.98% vs. controls 79.10% when MGC803 cells were treated
by DADS from 20 to 35 mg/l (P < 0.05), which showed a dose-
dependent manner (F = 29.78, P < 0.05). Alkaline phosphatase
(ALP) specific activity of MGC803 cells was 2.00 nkat/g. After
treated by DADS from 20 to 35 mg/l ALP specific activity
decreased from 1.70 to 0.67 nkat/g and decreasing ratio from
15.7 to 66.1% (F = 207.6, P < 0.05), which demonstrated that
DADS inhibited ALP specific activity of MGC803 cells, and
showed a dose-dependent manner. MGC803 cells showed nega-
tive of dye transfer while positive transfer of LY dye from the
load cells at the scraped-line to neighboring cells induced by
DADS. All these showed above implied that MGC803 cells were
induced differentiation into normal gastric cells by DADS. The
above results suggest that DADS may induce differentiation of
human gastric cancer MGC803 cells.
F2-022P
D-Aspartic acid: a putative neurotransmiter
in the marine mollusk Aplysia limacina
P. Spinelli
1
, E. D’Aniello

2
, G. Ferrandino
1
and A. D’Aniello
1
1
Department of Neurobiology, Zoological Station, Naples, Italy,
2
Department of Biochemistry and Molecular Biology, Zoological
Station, Naples, Italy. E-mail:
Using a chromatographic HPLC method associated to the use of
a D-aspartate oxidase (D-AspO; EC 1.4.3.3) purified from beef
kidney, we obtained evidence for a high concentration of D-Asp
(D-aspartic acid) in the nervous system of the mollusk Aplysia
limacina and a set of experimental results which prove that this
amino acid could be a candidate neurotransmitter in this animal.
Abstracts
343
The evidence supporting this hypothesis was based on the follow-
ing results: (i) D-Asp was found at relative high concentration in
all nervous ganglia (cerebral, pedal, buccal, visceral, abdominal)
of the animal at a concentration of 3.0–4.0 lmol/g tissue. (ii) In
purified synaptic vesicles prepared from the cerebral or abdom-
inal ganglia we found D-Asp at a hygh concentration which con-
stituted the 8–12% of total amount of amino acids present in
these vesicles. (iii) Immunocytochemistry study conducted by
using light microscopy with an anti-D-aspartate, shown the pres-
ence of D-Asp in the soma and in dendrites of cultured neurons.
(iv) Electron immunocytochemistry by using the above antibody,
demonstrated that D-Asp was present in the synaptic vesicles. (v)

Using an antibody against D-aspartate oxidase (the enzyme
which specifically oxidize D-Asp in oxaloacetate), we shown that
this enzyme is present in the post-synaptic membranes. (vi) As-
partate racemase, an enzyme capable of synthesizing D-Asp by
conversion of L-Asp in D-Asp, was found in the nervous system.
(vii) Potassium ions induced the release of D-Asp from the syna-
ptosomes; and (viii) In vivo experiments (injection of D-Asp into
the whole animal) or in vitro experiments (incubating isolated
neurons with D-Asp) induced a significant increase of cAMP.
F2-023P
Experimental study of differentiation
of leukemia HL-60 cells induced by diallyl
disulfide
Q. Su, W g. Huang, M H. Wu, Q j. Liao, H l. Xie, J. He,
J. Zhao and H. Tan
Institute of Cancer Research, medical college, Nanhua University,
hengyang, hunan province PR China. E-mail:
Abstract Diallyl Disulfide (DADS) has been known to have
potent chemopreventative activity against cancers. In this study,
MTT assay demonstrated that DADS has significantly antiprolif-
erative effect on HL-60 cells in a concentration-dependent and
time-dependent manner; morphologic observation showed that
HL-60 cells were induced differentiation toward granulocytic lin-
eage after treated by DADS from 0.625–20 lg/ml and its best
differentiational concentration is at 1.25 lg/ml level. The reduc-
tion ability of NBT had significantly increased (P < 0.01) and
cell surface differentiation antigen CD11b increased from 12%
(untreated) to 44% (treated by 1.25 lg/ml DADS for 3 days)
(P < 0.01). G0/G1 cells were added to 44.6% (P < 0.01), but S
and G2/M phase cells descended to 48.5 and 6.9% (P < 0.01)

respectively (the ratio of G0/G1, S and G2/M phase in untreated
HL-60 cells were 23, 66, 11% respectively) and cell cycle was
arrested at G1 stage. Kinase activity of JAKs/STATs was tested
by western blotting and results demonstrated that the activity of
phosphative of Jak1, stat3 kinase were inhibited. STAT3, c-myc
gene expression decreased and c-fos, c-jun gene expression
increased in HL-60 cells treated by DADS. The above results
suggested that HL-60 cells could be induced differentiation
toward granulocytic lineage by DADS. And inhibition of phos-
phative Jak1, stat3 were involved in HL-60 cells differentiation
induced by DADS, its possible molecular mechanism might relate
to modulation of proliferation associated gene expression and
inhibition of DNA synthesis.
F2-024P
Heme peroxidase encrusting and inhibition
with polyphenols revealed by molecular
modeling
A. Ziemys
1,2
and J. Kulys
2
1
Department of Biology, Vytautas Magnus University, Kaunas,
Lithuania,
2
Department of Enzymatic Chemistry, Institute of
Biochemistry, Vilnius, Lithuania. E-mail:
Molecular modeling techniques were applied to study the interac-
tions of polyphenols with Arthromyces ramosus peroxidase.
Polyphenols are produced during peroxidase-catalyzed oxidation

of phenolic compounds. The interaction of substrates and poly-
mers with Arthromyces ramosus peroxidase was analyzed by
docking and molecular dynamics methods. The calculation
showed that the most possible interaction site of polymers is the
active center of the peroxidase. The affinity of polymers increases
with increasing length of oligomers. However, the complexed pol-
ymers produce non-productive complexes with the peroxidase.
Molecular dynamics studies showed that polymer-peroxidase
complexes are stable. It seems likely that strong and stable, but
non-productive interaction of the polymers determinates peroxi-
dase inhibition during the reaction by preventing the access of
regular substrates to the active center of the enzyme.
F3–Rotary Motor Complexes
F3–001
The rotary motors in F- and V-type ATPases
D. Stock and R. A. Bernal
MRC Laboratory of Molecular Biology, Cambridge, UK.
E-mail:
ATP synthase, or F-type ATPase, is central to biological energy
conversion. Photosynthetic and respiratory systems of all living
organisms convert energy derived from light or from nutrients
into transmembrane electrochemical proton gradients. ATP syn-
thase uses the energy stored in these gradients to synthesize the
universal biological energy carrier ATP from ADP and inorganic
phosphate. The transmembrane F
o
sector of the enzyme contains
a rotary motor that is fuelled by the proton gradient in reminis-
cence of a turbine. The rotary torque induced by the passage of
protons through F

o
is transmitted via a central stalk to the cata-
lytic subunits in the soluble F
1
domain, where the rotation
induces conformational changes that enable the synthesis and
subsequent release of ATP. Eukaryotic vacuolar (V-type)
ATPases operate in reverse: they utilize energy derived from ATP
hydrolysis to build up transmembrane ion gradients. V-type
ATPases play an important role in pH homeostasis and enable
transport processes across membranes. While most eubacterial
ATPases are of the F-type, some eubacteria and all known arch-
aea have ATPases of the A-type, which are close homologues of
V-type ATPases, but are used for ATP synthesis. Although
V- and A-type ATPases are similar in size and shape to F-type
ATPases, only the catalytic subunits and the core of the trans-
membrane motor share significant sequence homology. In my
talk I will discuss the X-ray structure of the F
1c10
complex from
yeast mitochondria, which represents the largest structurally
known portion of any type of ATPase to date. I will compare
this structure to a lower resolution 3D reconstruction of an intact
eubacterial A-type ATPase obtained by electron microscopy.
Abstracts
344
F3-002
Brakes and engine of ATP synthase
M. Yoshida
Chemical Resources Laboratory, Tokyo Institute of Technology,

Yokohama, Kanagawa, Japan. E-mail:
The ATP synthase can be thought of as a complex of two motors
– the ATP-driven F1 motor and the proton-driven Fo motor –
that rotate in opposite directions. Our recent findings are follows.
1. Direct observation of reaction sequence: Single molecule ima-
ging of F1 motor has provided direct evidence that all of the
three catalytic beta subunits cooperate in sequence during a sin-
gle 120-degree rotation; the first beta, ATP binding (80 degree
substep); the second ATP hydrolysis (1 ms); the third, probably
release of Pi or ADP (1 ms, 40 degree substep). 2. Brakes: At
least two kinds of brakes of ATPase activity of ATP synthase
have been known; ADPMg inhibition and epsilon inhibition.
ADPMg inhibition stops the rotation at a 80 degree substep
once in every 30 s, in average, and resumes rotation that contin-
ues next 30 s. 3. Conformational transition of epsilon subunit:
epsilon can adopt either inhibitory extended form or non-inhibi-
tory folded form. In ADP and with proton motive force, it
adopts the former, and in ATP the latter. 4. Epsilon binds ATP:
The epsilon subunit is an ATP-binding protein, which may mon-
itor the cellular ATP concentration. 5. Non-integer proton/ATP
ratio: The rotor c-ring of Fo was proved to be a c-decamer (pro-
ton/ATP=3.333 !). The non-integer unit rotation ratio (36 deg/
Fo:120 deg/F1) could be buffered by protein elasticity.
F3-003
Control of the bacterial flagellar motor
H. C. Berg
Molecular & Cellular Biology & Physics, Harvard University,
Cambridge, MA 02138, USA. E-mail:
Flagellated bacteria swim by rotating long, thin, helical filaments
that arise at different points on the cell surface. Each filament is

driven at its base by a rotary motor only 45 nm in diameter
made from about 20 different kinds of parts. Control of the
direction of rotation of these motors is the basis for the chemo-
tactic response, i.e. for the ability of cells, such as E. coli,to
swim up spatial gradients of chemical attractants. I will tell you
about motor rotation and its modulation by the cell’s signal
transduction pathway.
F3-004
Three-dimensional structures of the bacterial
flagellar rotary motor
D. R. Thomas, N. R. Francis and D. J. DeRosier
Rosenstiel Basic Medical Sciences Research Center, Biology
Department, Brandeis University, Waltham, MA, USA.
E-mail:
The bacterial flagellum used to swim by many motile bacteria,
has a reversible rotary motor powered by the proton gradient
across the cell’s plasma membrane. The motor is coupled to the
flagellar filament, and transmits torque to the filament via the
rod and the hook. The rotor contains four proteins, FliF, FliG
FliM and FliN, forming a complex of two rings, the C ring
(FliM and FliN), located in the cytoplasm and the M ring (FliF
and FliG) found in the membrane. The transmembrane proteins
MotA and MotB form the proton channel and are anchored to
the peptidoglycan layer thus forming the stator of the motor.
There are many models for how the motor functions but they
have been devised in the absence of structural information. 3D
reconstructions from cryoEM images of bacterial flagellar motors
of S. typhimurium will be presented. Edge on views C rings have
been into symmetry classes and reconstructions have been gener-
ated from three of these symmetry classes (33-, 34- and 35-fold)

which are nearly identical except for differing numbers of sub-
units and diameter. We find the M-ring symmetry also varies
having 24-, 25- and 26-fold symmetry. Images have been sorted
based on the combination of C ring and M ring symmetry and
motors reconstructed with specific symmetry combinations such
as those having 34-fold C ring symmetry and 25-fold M ring
symmetry. The inner domain of the C ring which lies nearest the
M ring has the same symmetry as the M ring. This suggests that
FliG is part of both the C and M rings.
F3-005
Mechanism of flagellar hook bending flexibility
revealed by molecular dynamic simulation of
the protofilament extension and compression
F. A. Samatey
1,2
, H. Matsunami
1,2
, K. Imada
1,2
, S. Nagashima
1
,
A. Kitao
3
and K. Namba
1,2
1
Dynamic Nanomachine Proj., ICORP, Japan Science and
Technology, Suita, Osaka Japan,
2

Grad Sch. of Front. Biosc.,
Osaka University, Suita, Osaka Japan,
3
Inst of Mol & Cell Biosci,
University of Tokyo, Tokyo, Tokyo Japan.
E-mail:
The bacterial flagellar hook is a short tubular segment that con-
nects the flagellar motor to the long helical filament. Its function
as a universal joint to transmit the motor torque to the helical
propeller is essential for the chemotactic behaviours of bacteria.
The hook is a helical assembly of bout 120 copies of a single pro-
tein FlgE (42 kDa, 402 residues). Like the flagellar filament, the
hook is a supercoiled tube that transforms into several different
types of helical structures. However, in contrast to a relatively
rigid structure of the filament to work as a propeller, the hook
has a much higher bending flexibility to function as a universal
joint. We reported the X-ray crystal structure of a 31 kDa core
fragment of FlgE named FlgE31 (residues 70–363) and an atomic
model of the straight hook built by docking the FlgE31 model to
a density map obtained by electron cryomicroscopy and image
analysis. Since then, we have been studying the structure and
intermolecular interactions of the coiled hook model as to how
the hook protofilament can go through a large amount of exten-
sion and compression by changing its axial repeat distance by
about 15 (around that of the straight hook, which is 46). We pre-
dicted that bonding partners must be able to switch at the inter-
subunit interface between the triangular-shaped loop of domain
D1 and the inner surface of domain D2 to allow the two
domains slide each other axially. Molecular dynamic simulation
of protofilament extension and compression actually demonstra-

ted that it is really the case.
F3-006
Intracellular sodium concentration of chimera
Escherichia coli
C J. Lo, M. C. Leake and R. M. Berry
Clarendon Laboratory, Department of Physics, University of
Oxford, Oxford, UK. E-mail:
Many species of bacteria have flagellar motors that convert free
energy into controlled motion. The bacterial flagellar motor is a
tiny molecular machine that couples ion flow to flagellar rotation.
The coupling ion is either proton or sodium ion. The energy-
transducing cytoplasmic membrane of bacteria contains pumps
Abstracts
345
and antiports maintaining the membrane potential and ion gradi-
ents. We demonstrate the use of fluorescence sodium indicators,
Sodium Green and CoroNa Green to measure the single cell
intracellular sodium concentration. We use a hybrid flagellar
motor with proton- and sodium-ion driven components that can
function well with sodium ions in Escherichia coli. We measure
the dependence of the motor speed and internal sodium concen-
tration on external sodium concentration and pH. These meas-
urements will investigate the proton and sodium energetics and
the mechanism of coupling in E. coli.
F3-007P
Replication termination in Escherichia coli:
the end of the story?
M. D. Mulcair, P. M. Schaeffer, A. J. Oakley and N. E. Dixon
Research School of Chemistry, Australian National University,
Canberra, A.C.T., Australia. E-mail:

The mechanism of termination of DNA replication in E. coli has
been debated for nearly 20 years. The terminator protein (Tus)
binds Ter site DNA as a monomer and requires no auxiliary pro-
teins, yet replication arrest is a polar (unidirectional) process. A
replisome approaching from the permissive face of the Tus-Ter
complex can continue its progress, apparently by displacing Tus,
whereas a replisome approaching from the non-permissive face is
arrested [1]. How can this be? Two models have been proposed.
One assumes that Tus acts as a simple roadblock that prevents
passage of the replisome in only one direction [2]. The other pro-
poses a specific protein–protein interaction between one face of
Tus and the replicative DnaB helicase that halts its progress [3].
Here we describe a series of BiaCore experiments that show that
neither of these models is correct. From these data, we propose a
third model: the progress of the replisome generates structures in
the DNA to which Tus binds more weakly or tightly depending
on whether approach is from the permissive or the non-permis-
sive face. In particular, the data are consistent with a new mech-
anism for fork arrest whereby strand separation by the helicase
at the non-permissive face leads to flipping of a specific base out
of the Ter DNA duplex into a pocket in the surface of Tus, to
form a tighter-binding ’locked’ structure. This model can explain
the data previously obtained both by us [4] and others, and
finally brings an end to the debate.
Reference
1. Hill T.M., et al. P.N.A.S. 1987; 84, 1754–1758.
2. Lee, E.H., et al. P.N.A.S. 1989; 86, 9104–9108.
3. Mulugu S., et al. P.N.A.S. 2001; 98, 9569–9574.
4. Neylon C., et al. Biochemistry 2000; 39, 11989–11999.
F3-008P

3D structure of a mammalian VO-ATPase by
cryo-electron microscopy
M. Gregorini
1
, R. Milligan
2
, G. Chang
2
, X S. Xie
3
and
A. Engel
1
1
Department of Structural Biology, University of Basel, Basel,
Switzerland,
2
Department of Cell Biology, The Scripps Research
Institute, La Jolla, United States of America,
3
Department of
Internal Medicine, University of Texas Southwest, Dallas, USA.
E-mail:
The eukaryotic V-ATPase is an ATP-hydrolizing proton pump
that functions both to energize biological membranes by creating
a proton gradient and to acidify intracellular compartments. The
enzyme is found in the plasma membrane of specialized cells and
in organelles such as endosomes, lysosomes and clathrin coated
vesicles where it is crucial for processes such as coupled trans-
port, receptor-mediated endocytosis, intracellular trafficking and,

in osteoclasts, bone resorption. The bovine V-ATPase is made
up of 13 subunits and has a total mass of 900 kDa. It is com-
posed of a cytosolic moiety V1 with a stoichiometry of
A3B3CDEFG2H2 and a transmembrane portion VO with a pre-
dicted stoichiometry of a(cc’)4–5c’’d. The two subcomplexes V1
and V0 are linked by a central, and at least one other peripheral
stalk. ATP-hydrolysis takes place at and alternates between the
three catalytic A subunits of V1. This circular motion drives
rotation of the asymmetric central stalk, which is firmly linked
to the c-ring of VO. The protons are believed to be transported
through the membrane at the interface of the c-ring and a sub-
unit, which is connected to the A3B3 complex via peripheral
stalk. There is evidence that the mechanism by which this motor
acts is similar to that of the better-known F1FO-ATPase under
hydrolizing conditions. Both negative stain and cryo-electron
microscopy were used to investigate how the V1VO-ATPase
functions with emphasis on the membrane embedded VO por-
tion. The electron micrographs were processed by single particle
image analysis. We present a preliminary volume of the V-AT-
Pase holoenzyme computed from micrographs of heavy metal
stained samples and a medium-resolution 3D map of the trans-
membrane VO complex obtained by cryo-microscopy of frozen-
hydrated samples.
F3-009P
Characterization and expression of the
thermophilic PS3
J. Hayakawa, Y. Takahashi, F. Nakazawa, M. Kwata and
M. Ishizuka
Applied Chemistry, Chuo, Tokyo, Japan.
E-mail:

Many species of bacteria swim by means of flagellar rotation.
Energy for rotation comes from H
+
(or Na
+
) electrochemical
potential gradient. The bacterium flagellum is composed of three
structurally defined parts: the basal body, the hook and the fil-
ament. The P and L ring located around the rod, since Bacillus
(B. PS3) is the gram-positive thermophilic bacterium studied by
this work, has no these structure. The rotor of flagellar motor
consisted of three proteins (switch complex) that regulate rota-
tional direction. However, until now rotation mechanism could
not be understandable. Flagellum component, in vitro, reconstitu-
tion could be clarified these mechanism and functions. In order
to analyze fla/che operon (large flagellar gene cluster >26 kbp)
was isolated from PS3 genomic library, and the sequences of all
32 ORFs were determined. These deduced amino acid sequences
correspond to switch complex, MS ring, rod, hook, export appar-
atus, chemotaxis, sigma factor, and, flagellar biosynthesis protein
were compared with those of another microorganisms. C-termi-
nus of FliY is homologous to E. coli FliN, but the N-terminus
share homology with B.PS3 FliM. In this work, MS-ring (FliF),
C-ring (FliM, FliN, FliG), rod (FlgB FlgC), and CheY, were
over-expressed as CBD (chitin binding domain) or MBP (maltose
binding protein)-tagged protein. FliF-CBD (MS-ring) expressed
in insoluble fraction, so pellet was solubilized in 8 m urea solu-
tion. For confirming interaction between FliM and CheY, we
constructed MBP-FliM and D 1–16 FliM fusion protein, since
N-zterminus region of FliM has described as CheY binding area

for E. coli. The result of these test showed that FliM N-terminus
residue is important for CheY binding.
Abstracts
346
F3-010P
Mutations in yeast 18S rRNA affect
translational fidelity and oxidative stress
T. C. Konstantinidis
1
, N. Patsoukis
2
, C. D. Georgiou
2
and
D. Synetos
1
1
Laboratory of Biochemistry, School of Medicine, University of
Patras, Patras, Greece,
2
Laboratory of Biochemistry, Department
of Biology, University of Patras, Patras, Greece.
E-mail:
The function of mutations rdn1A, rdn1T and rdn2 in 18S rRNA of
Saccharomyces cerevisiae has been examined. They correspond to
C1054A, C1054U in helix 34 and G517A in loop 530 of helix 18 of
16S rRNA in Escherichia coli, respectively, in which they behave as
nonsense suppressors and reduce translational fidelity. In yeast,
mutation rdn1A caused severe loss of translational fidelity accom-
panied by an increase in ribosomal A-site binding, extreme sensitiv-

ity to paromomycin, and higher initial rates of protein synthesis
activity, all compatible with its error-prone character. Unlike in E.
coli, however, mutations rdn1T and rdn2 caused hyperaccuracy fol-
lowed by resistance to paromomycin. It is concluded that loop 530
and helix 34 of 18S rRNA participate in the accurate decoding of
the genetic information. Moreover, mutational changes in con-
served rRNA regions may affect the same functions in the various
species but in opposite directions. Translational fidelity and oxida-
tive stress are two of several parameters affected during a cell’s life.
Mutants carrying error-prone ribosomes exhibited decreased oxi-
dative stress compared to wild type. In contrast, hyperaccurate ri-
bosomes, such as those from rdn2 or rdn1T strains seem more
susceptible to oxidative modifications, i.e. lipid peroxidation and
non-protein thiol oxidation. The notion is investigated that the
cells consume more energy in order to achieve hyperaccuracy lead-
ing to increased production of free radicals.
Acknowledgment: We thank the European Social Fund (EKT),
the Operational Program for Educational and Vocational Train-
ing II (EPEAEK II), and particularly the Program HERAKLEI-
TOS for funding the above work.
F3-011P
Polyamines stimulate the function of the
rotary motor complex implicated in the A- to
P-site translocation of tRNA substrates during
protein synthesis
A. D. Petropoulos and D. L. Kalpaxis
Laboratory of Biochemistry, School of Medicine, University of
Patras, Patras, GR-26500, Greece. E-mail:
Spermine, one of the naturally occurring polyamines, stimulates
at 6 mm Mg2+ both the spontaneous and enzymatic transloca-

tion of AcPhe-tRNA from the A- to the P-site of poly(U)-pro-
grammed Escherichia coli ribosomes. To unveil the molecular
basis of the spermine effect, the locations of spermine bound to
rRNA were characterized by covalent binding of a photoreactive
analogue of spermine, N1-azidobenzamidino (ABA)-spermine, to
E. coli ribosomes under mild irradiation conditions. Cross-linking
sites were identified by RNase H digestion and primer extension
analysis at C1400 and C1411 of the 3
´
-minor domain of 16S
rRNA, at positions located in the central loop of domain V, and
in helices H42-H44 and H95 of 23S rRNA. Specifically, in the
central loop of domain V of 23S rRNA, ABA-spermine labels
nucleosides U2584 and A2602. Nucleoside A2602 has been sug-
gested to propel the spiral rotation of the tRNA-3
´
end from the
A- to P-site in concert with the tip of helix H69, while U2585,
lying next to U2584, has been proposed to anchor the rotary
motion by direct interaction with the aminoacyl end of the tRNA
and assure the proper positioning of the P-site substrate. Interest-
ingly, pre-translocation ribosomal complex totally labelled or
partially modified in its 50S subunit by ABA-spermine exhibited
higher efficiency in translocation, compared with untreated com-
plex or complex labelled only in its 30S subunits. Our results sug-
gest that polyamines activate the rotary machine that governs the
translocation function in ribosomes, by linking in close proximity
to certain nucleosides of rRNA.
Acknowledgment: Supported by a grant from the Research Com-
mittee of the University of Patras (Programme K. Karatheodori).

F3-012P
The FliH component of the flagellar export
apparatus is a multi-zinc enzyme with
phospholipase activity
A. Sebestye
´
n
1
, Z. Gugolya
1
, G. Jakab
1
, Z. Dio
´
szeghy
2
,
P. Za
´
vodszky
2
and F. Vonderviszt
1,2,3
1
Department of Nanotechnology, University of Veszpre
´
m,
Veszpre
´
m, Hungary,

2
Institute of Enzymology, Budapest, Hungary,
3
Research Institute for Technical Physics and Materials Sciences,
Budapest, Hungary. E-mail:
FliH and FliI are cytoplasmic components of the type III flagellar
protein export apparatus. FliI is an ATPase whose enzymatic activ-
ity is necessary to drive the export process, while FliH is believed
to function as a negative regulator of FliI. Our isothermal titration
calorimetric experiments, however, question this regulatory role
because the FliH–FliI interaction is too weak to allow effective
binding in the physiological concentration range. Characterization
of the metal binding ability of FliH by surface plasmon resonance
spectroscopy revealed that FliH can specifically bind three zinc
atoms. The binding appears to be quite strong with an overall dis-
sociation constant of about 2 · 10
–6
m, as obtained by titration
calorimetric studies. Temperature induced unfolding experiments
demonstrated that FliH is well folded even under zinc-free condi-
tions suggesting that zinc is required for functional rather than
structural reasons. FliH was found to exhibit structural character-
istics similar to those of bacterial phospholipase C, a multi-zinc
enzyme with phosphodiesterase activity. Indeed, our experiments
demonstrated that FliH has a phosphatidylcholine-hydrolyzing
capability. Further studies are needed to clarify how the phospholi-
pase C activity of FliH is related to the functioning of the flagellar-
specific export machinery.
F3-013P
Catenanes- simpler analogues of biomolecular

rotors
Y. Tourleigh and K. Shaitan
Molecular Modelling Group, Dept. of Bioengineering, Lomonosov
Moscow State University, Moscow, Russia.
E-mail: ,
Molecular machines described in this paper are meant to be such
molecular systems that make use of conformational mobility (i.e.
hindered rotation around chemical bonds as well as molecular con-
struction deformations with formation and breakage of non-valent
bonds). Components of molecular machines move mainly by
means of restricted diffusion. Whereas it is true to consider biologi-
cal molecules-motors (various ATPases, bacterial flagellar com-
plexes, membrane transporters) as machines, one may facilitate
their studying by applying simpler models from macromolecular
chemistry. As an example of molecular machines of a non-biologi-
cal nature catenanes (compounds with two interlocked molecular
rings) can be proposed. Thus, for example, model catenane [(2)-
(cyclo-bis(paraquat-p-phenylene))-(1(2,6)-tetrathiafulvalena-16(1,5)
naphtalena-3, 6, 9, 12, 15, 17, 20, 23, 26, 29-decaoxatnaconta-
Abstracts
347
phane)-catenane] changes its redox status when an electric field is
applied, and rotation of the rings takes place. It occurs with fix-
ation at certain moments of the influence. To find out characteristic
properties of rings movements under various external conditions
molecular dynamics (MD) simulations were carried out. Three cat-
ionic forms of the catenane were first subjected to geometrical opti-
mization and quantum chemical calculation. Then at different
temperature conditions and under varying magnitudes of external
electrical field these catenane states were run in MD calculations.

Summarizing of the results allowed to develop an idea about poss-
ible mechanism of the catenane’s functioning.
Acknowledgment: The authors acknowledge the financial sup-
port of RF Ministry of Education and Science (grants I0431,
01.106.11.0001 and 01.165.11.0001), RFBR (grant
1
04-04-49645).
F3-014P
The flagellum-specific type III secretion signal
is confined to a 22-amino-acid sequence in the
disordered N-terminal region of flagellin
B. M. Ve
´
gh
1
,P.Ga
´
l
1
, J. Dobo
´
1
,P.Za
´
vodszky
1
and
F. Vonderviszt
1,2
1

Institute of Enzymology, Hungarian Academy of Sciences, Buda-
pest, Hungary,
2
Department of Nanotechnology Faculty of Infor-
mation Technology, University of Veszpre
´
m, Veszpre
´
m, Hungary.
E-mail:
Bacteria swim by rotating their flagellar filaments, each of which
has a helical shape and works as a propeller. External flagellar
proteins, lying beyond the cytoplasmic membrane, are synthesized
in the cell and exported by the flagellum-specific export apparatus
to the site of assembly through the central channel of the filament.
The flagellum-specific export system is a specialized, signal-peptide
independent type III export machinery. Earlier studies suggested
that the N-terminal region of flagellar proteins is essential for
recognition for export, however, comparison of their amino acid
sequences failed to identify a common recognition signal at the
primary structure level. In this work we aimed at identifying the
secretion signal in Salmonella flagellin. Two approaches were fol-
lowed: fragments of flagellin from the N-terminal part of the
molecule were expressed and their secretion efficiency was charac-
terized. Also, various segments of the N-terminal region were
fused to the 74-amino-acid long CCP2 module of human comple-
ment C1r protein and secretion of these fusion constructs was
investigated by Western blot assays of the culture media. We
found that residues 26–47 of Salmonella flagellin are sufficient and
essential for secretion. This region lies within the disordered

N-terminal part of flagellin. Amino acid sequence alignment of
flagellins from various sources revealed that the region containing
the export signal is highly conserved. Our experiments demon-
strate that the flagellum-specific export system can be utilized to
secret over-expressed recombinant proteins into the culture media
by generating fusions to the secretion signal.
Abstracts
348