From the Laboratory of Molecular Biophysics,
Department of Biochemistry, University of Oxford, South Parks Road
Oxford OX1 3QU, United Kingdom, ¶ Sir William Dunn School of
Pathology, University of Oxford, South Parks Road, Oxford OX1 3RE,
United Kingdom, ** Plate-Forme de Microscopie
Électronique, Insitute Pasteur, 25,28 rue du Docteur Roux, 75724 Paris Cedex 15, France, and §§ Department of
Biochemistry and Molecular Genetics, Box 800733, University of
Virginia, Charlottesville, Virginia 22908-0733
Received for publication, January 6, 2003, and in revised form, January 28, 2003
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Gram-negative bacteria commonly interact with
animal and plant hosts using type III secretion systems (TTSSs) for
translocation of proteins into eukaryotic cells during infection. 10 of
the 25 TTSS-encoding genes are homologous to components of the
bacterial flagellar basal body, which the TTSS needle complex
morphologically resembles. This indicates a common ancestry, although
no TTSS sequence homologues for the genes encoding the flagellum are
found. We here present an ~16-Å structure of the central component,
the needle, of the TTSS. Although the needle subunit is significantly smaller and shares no sequence homology with the flagellar hook and
filament, it shares a common helical architecture (~5.6
subunits/turn, 24-Å helical pitch). This common architecture implies
that there will be further mechanistic analogies in the functioning of
these two bacterial systems.
Gram-negative enteropathogenic bacteria cause a wide variety of
diseases ranging from relatively harmless infections to
life-threatening illnesses (1). They account for more than 3 million
deaths each year, mostly among children and immunocompromised adults in
developing countries (2). Infections usually occur where hygiene is
poor as the major route of infection with Salmonella, Escherichia, Yersinia, and Shigella spp. is the
consumption of contaminated food (3) and water (4). Despite
significant differences among the distantly related genera, a common
macromolecular system, the type III secretion system
(TTSS1 or secreton), is the
basis of each infectious cycle. This system consists of >20 proteins
that form a macromolecular assembly, which delivers the
bacterial virulence effectors not only across the bacterial inner and
outer membranes but also directly into the host cell. The TTSS is well
conserved among these bacteria, whereas the specific properties of the
effectors and hence the resulting symptomatic effects on the host
organism vary widely (5). TTSSs are not constitutively active but are
activated to secrete by signals that apparently vary among the
different genera, yet they seem to ultimately derive from
physical contact between the bacterium and its host cell. Understanding
how an activation signal is transmitted in such a complex
macromolecular assembly and how it results in secretion is one of
the major questions to be answered.
10 of the 25 TTS-encoding genes show a strong similarity to those that
encode the flagellar basal body, indicating a common ancestry (5).
Insights into the morphology of the system to date are derived from
electron microscopy that reveals a supramolecular structure grossly
resembling the flagellar hook-basal body complex (6-8). The major TTSS
structure seen is termed the needle complex and spans the inner
and outer bacterial membranes with a basal body into which a needle
(~70 Å in diameter traversed by a central channel 20-30-Å wide) is
inserted that protrudes ~500 Å into the extracellular space (8, 9).
The activation of the Shigella TTSSs for secretion seems to
require the contact of the needle tip with the host cell-limiting
membrane (7). It is particularly difficult to understand how such
physical contact leads to a change at the cytoplasmic face of the basal
body ~800 Å away and results in secretion of proteins through the
structure. Gross structural homologies can be added to the genetic
resemblance to the flagella noted above. Common ancestry has been used
to gain insights into the functioning of the TTSSs on the basis of the
well studied flagellar system (10). However, TTSS homologues for
several of the key flagellum-building blocks, notably hook protein and flagellin, are lacking. Their absence combined with a lack of any
detailed knowledge of the TTSS structures involved means that the
hypothesis of the common mode of biological function remains unproven.
This hypothesis implies that the needle of the Shigella TTSS
will be made up by a helical arrangement of the 10-kDa protein MxiH
(previously identified to be the major needle component (9, 11)) with
an architecture similar to that shared by both the flagellar hook and
filament. However, MxiH shows no sequence homology to either the hook
or filament subunits and is in fact a much smaller protein being
approximately one-fifth the molecular weight of either of the flagellar
proteins. Homology between these two systems leads to the hypothesis
(10) that signal transduction will occur via changes in the
architecture of the needle because it is known that the flagellar
filament is able to switch between different helical forms (12, 13). To
test this predicted homology, we have used x-ray fiber diffraction and
electron microscopy-based three-dimensional reconstruction techniques
to determine the structure of the Shigella
flexneri needle at ~16 Å. We find that as the hypothesis
predicts, the MxiH subunits making up the needle are arranged in a
helical fashion to form an extended cylindrical structure with a
central channel. The parameters that describe the geometry of the helix
are very similar to those of the flagellar hook (14) and filament (15),
providing strong support for the hypothesis that these two systems
share a common architecture and thus common functional mechanisms.
Bacterial Strains and Growth
The Shigella flexneri serotype 2a non-polar null
mutant for mxiH Preparation of Purified Needle Samples
10% weight/volume polyethylene glycol 6000 (catalog number
443915V, BDH Ltd.) is added to the culture to precipitate free needles
in the media, and the culture is cooled on ice as soon as the
polyethylene glycol has dissolved. Cells (and precipitated needles) are
pelleted at 2000 × g (all centrifuge runs at 4 °C) in swing-out buckets. The supernatant is discarded, and the pellet is
resuspended in 1% the initial culture volume of
phosphate-buffered saline. The cell suspension is transferred to a
40-ml Dounce glass-glass tissue grinder ("tight" pestle,
manufactured by Wheaton, Millville, NJ) that shaves needles from the
cell surface through the exertion of sheer forces. An examination of
the sample in an electron microscope showed that 60 cycles of up and
down "grinding" were sufficient to remove the majority of needles
from the bacteria while keeping most cells intact.
The total volume of the suspension is adjusted with phosphate-buffered
saline to 2% the initial culture volume and centrifuged at 2000 × g. The pellet is discarded, and the supernatant is spun at 12,000 × g to remove remaining macromolecular
contaminants. Trichloroacetic acid protein precipitation
followed by SDS-PAGE analysis and EM confirmed that the supernatant
contained MxiH (in the form of long needles) free from major
contaminants. Finally, needles are precipitated by adding polyethylene
glycol 6000 and NaCl to a final concentration of 10% and 100 mM, respectively. Following 60-min incubation on ice,
needles are pelleted at 27,000 × g. After discarding
the supernatant, the small opaque needle pellet is resuspended in
~0.5 ml of buffer (roughly 0.001% of the initial culture
volume) at a physiological pH (10 mM Tris, pH 7.4) when it
becomes translucent and stored at 4 °C. This protocol allows ~5 mg
of MxiH (in the form of intact needles) to be purified from a culture
of 6 liters of Shigella
mxiH X-ray Fiber Diffraction Methods
Oriented samples of TTSS needles were prepared by a modification
of the method of Yamashita et al. (16, 17). 200 µl of needle sample (typically with a concentration ~8 mg/ml) is diluted approximately twice in a salt-free Tris buffer at pH 7.4 and needles concentrated by centrifugation overnight at 10,000 × g
in a swing-out rotor containing a long thin tube (Beckman Rotor
SW50.1/55 using 5 × 41-mm tube and suitable adapter). The
supernatant is then carefully removed, and 10-µl aliquots of the
viscous translucent pellet are transferred into quartz capillaries with
an internal diameter of 0.7 mm. The capillaries are centrifuged at
2000 × g for 8-24 h in a swing-out rotor to
concentrate the needles at the bottom of the capillary. The careful
removal of the liquid above this region reduces dilution effects and
increases sample stability. Samples were allowed to order in
capillaries for 1-3 weeks at room temperature within the bore of a 15 Tesla electromagnet (Magnex Scientific, Yarnton, Oxford, United Kingdom).
Capillaries were then mounted in the beam at station 14.2 of the SRS
Daresbury, and images were collected on a ADSC scanner using
Electron Microscopy Data Collection
Negative Staining of Isolated Needles for Electron
Microscopy--
Purified needles were diluted 1:250 in 10 mM Tris, pH 7.5. An aliquot of 5 µl was deposited for 1 min on a glow-discharged carbon-coated copper grid. After the removal
of the excess liquid with filter paper, the sample was stained with a
drop of 2% uranyl acetate, pH 7.5.
Observation and Image Acquisition--
The needles were observed
with a defocus value of Digitization and Image Processing Preparation--
Images were
scanned on a Eurocore Hi-SCAN rotating drum scanner at a 1600 dots per
inch resolution. Individual filaments were selected using EmTool
(ncmi.bcm.tmc.edu/ncmi/software), and image analysis was carried out
using SPIDER/WEB (18) software package. Three-dimensional
reconstructions were done within SPIDER using the iterated helical real
space reconstruction method (IHRSR) (19).
Image Processing and Electron Microscopy
Reconstruction--
Approximately, 10,000 segments (each = 100 pixels or ~265 Å in length) were extracted from micrographs of the
mutant filaments, and roughly 5000 segments were extracted from
wild-type needle complexes (9). The averaged power spectrum was
computed for each group of images after only allowing the rotations of
the individual segments to make the filament axis vertical. This is because the power spectrum is invariant under translations surmounting the need to align particles. The intensities of the power spectra from
each individual segment were simply added to produce an averaged power
spectrum. Resolution was determined by generating two reconstructions, each from half of the image segments, and comparing the correlation of
Fourier coefficients from these two reconstructions within resolution
shells. The coefficient of correlation dropped below 0.5 at a
resolution of 12.5 Å, but a significantly more conservative value of
16 Å was used for filtering the final volume.
Purity of the Needle Preparation Allows Biophysical
Studies--
The prerequisite for the majority of biophysical
techniques is a highly purified and concentrated sample of the
structure to be studied. To facilitate such studies of the
Shigella flexneri needles, we used a Shigella
mxiH X-ray Fiber Diffraction Analysis--
X-ray fiber diffraction
(XRFD) can be a useful method for determining helical parameters of
filamentous assemblies as well as for obtaining constraints in building
pseudoatomic models of helical polymers when only the atomic structure
of a monomer is known (20). Unfortunately, no atomic structure exists
yet for MxiH. However, XRFD has been used on the needles to provide
information that is complementary to that obtainable by electron
microscopy (see below). The key to recording high resolution XRFD data
is the preparation of a highly ordered, highly concentrated sample of
the macromolecular complex to be studied, i.e. the
preparation of the biological molecule in the form of a liquid crystal.
We used a variation on the method of Yamashita et al. (16,
17), which uses repeated centrifugation and equilibration of the
samples in a high magnetic field to produce highly ordered needle
samples (see "Experimental Procedures"). The exposure of these
samples to high intensity x-rays allowed for the collection of patterns such as that shown in Fig. 2. All of the
patterns showed the diffraction expected of a helical assembly,
demonstrating that MxiH assembles in a helical fashion to produce the
Shigella needle. The parameters needed to completely define
the helical symmetry are the axial rise (along the filament axis) per
subunit and the azimuthal rotation (around the filament axis) per
subunit. The axial rise per subunit gives rise to a reflection in the
XRFD pattern that is on the meridian, and the first meridional
reflection was observed at a spacing of ~1/4.3 Å (Fig.
2C). Strong near-meridional reflections were also observed
at a spacing of ~1/24 and 1/5.2 Å. The full interpretation of this
pattern comes from using electron microscopy.
Reconstruction of the Needle Structure from Electron
Micrographs--
To obtain more detailed structural information
regarding the needle, we collected a series of electron micrographs
from negatively stained specimens. The attempts to use conventional
methods of helical image analysis and three-dimensional reconstruction
(21) were not successful in part because of the very weak contrast provided by these thin polymers. Long segments could not readily be
found that consistently showed the multiple layer lines needed for
Fourier-Bessel methods. Instead, a single particle approach was used
(19) based on the analyses of overlapping segments extracted from the
micrographs (where each segment was ~265-Å long for the mutant and
~220-Å long in length for the wild type). The averaged power
spectrum derived from these segments (Fig. 2B) is the
computational equivalent of the XRFD pattern, because the diffraction
intensities have been added from specimens that have been aligned only
to have the filament axes oriented in the same direction. Three layer
lines with spacings of ~1/(58 Å), 1/(37 Å), and 1/(24 Å) were seen
in both the average spectrum and in the XRFD pattern (Fig. 2). The
1/(24 Å) layer line had peak intensities that were near-meridional,
suggesting that it arose from a one-start helix. A reference-free
average (22) of these segments (data not shown) gave rise to a power
spectrum showing these same three layer lines. The 1/(24 Å) layer line was determined to be odd and must correspond to a Bessel order of ±1.
The strongest layer line was at 1/(37 Å). An analysis of this layer
line showed that it was also odd and that it most likely corresponded
to a Bessel order of ±5. Similarly, the layer line at 1/(58 Å) was
determined to be even and most likely corresponded to a Bessel order of
±6. The only indexing scheme possible would involve the layer lines at
1/(58 Å) and 1/(24 Å) arising from helices with the same hand and the
layer line at 1/(37 Å) arising from a helix with the opposite
hand. We are unable to determine the handedness of this
structure from these data and have therefore assigned a hand based on
earlier reconstructions of the flagellar hook and major filament forms,
right-handed one-start helix and left-handed five-start helix (14,
15).
A crude three-dimensional reconstruction was generated by
Fourier-Bessel methods from the reference-free average, and this was
used as an initial model for the IHRSR method (19). This method allows
for the precise determination of helical symmetry during the course of
many cycles where the symmetry is freely allowed to change and, in this
case, converged to a structure with ~5.6 subunits/turn of a 24-Å
pitch helix after starting from the preliminary model. A confirmation
of this reconstruction was that the trans-form of the image generated
from the projection of the structure was found to closely match both
the averaged power spectrum and the XRFD data. Since generating either
the averaged power spectrum or the XRFD patterns involved no
assumptions regarding symmetry, this is a strong check on the
reconstruction. The use of very different indexing schemes to generate
the starting model either led the IHRSR cycles to diverge with the
program terminating when the axial rise per subunit became less than
zero (when starting with 7.5 subunits/turn) or led to a
three-dimensional reconstruction whose power spectrum did not match the
averaged power spectrum or XRFD pattern (when starting with 3.5 subunits/turn). Fig. 3 shows the helical
symmetry parameters as a function of cycle number during the course of
iterations starting from ~5.2 subunits/turn. The parameters converge
back to approximately the same value obtained when the procedure is
started with ~5.6 subunits/turn. Most importantly, the results of the
IHRSR method predict an axial rise per subunit of ~4.2 Å, well
beyond the resolution of the EM images. This is the actual rise (within
the level of error introduced by uncertainties in the magnification of
the EM images (~2%)) seen in the XRFD pattern as judged by the
meridional intensity corresponding to this spacing (Fig.
2C). This can completely exclude alternate indexing schemes
that would have ~3.6 or ~7.6 subunits/per turn of the 24-Å pitch
helix, predicting an axial rise of 6.7 or 3.2 Å, respectively.
Reconstructions were made from long needles (Fig.
4, A and B) and
also from wild-type needles (segments extracted from micrographs of
intact purified needle complexes prepared as described by Blocker et al. (9); data not shown). The helical parameters defining both reconstructions converged to similar values (Table
I), indicating that the architecture of
the needle is essentially unaltered by the overexpression of MxiH. We
have conservatively estimated the resolution of the reconstruction to
be ~16 Å, even though a power spectrum generated from this shows a
layer line at 1/14 Å that is in good agreement with the layer line
observed at this spacing in the XRFD pattern (Fig. 2A). The
needle is revealed to be a cylinder of an approximate outer diameter of
70 Å traversed by a central canal of ~20 Å in diameter. The
three-dimensional model shows distinct subunits connected by weaker
density in the helix. Assuming that each subunit corresponds to a
single copy of MxiH, we can determine an expected molecular volume
assuming a partial specific volume of protein of 0.75 cm3 g Common ancestry of the TTS and flagellar systems led to the
proposal (10) that there must be TTSS homologues of either flagellin and/or the hook proteins. Our data show that the needle has the same
architecture (~5.6 subunits/turn, ~24-Å helical pitch) (Table I)
as the flagellar filament (12, 15) and hook (14), implying that MxiH
and the flagellar hook protein and flagellin must be structural but not
sequence homologues of one another. This is somewhat surprising given
the huge difference in the size of these two proteins (MxiH is 83 amino
acids long, flagellin is 494 amino acids long, and the hook protein 402 amino acids long). However, the three-dimensional reconstruction of the
flagellar filament (13, 15) shows homologies other than similarities of
helical parameters to our needle structure. The flagellar filament
structure can be divided into a series of different domains: D0, an
inner tube of ~70 Å in diameter with a central channel of ~20 Å;
D1, the outer tube that surrounds D0 and has an outer radius of ~130 Å; and D2 and D3 that project outwards from the helical axis. The
fitting of an x-ray crystallographic structure for flagellin into this
EM reconstruction allowed Yamashita et al. (13) to determine
that the D0 domain was made up of the N and C termini of flagellin,
which were not present in the x-ray model as they needed to be
truncated to prevent the protein from polymerizing. The size and
helical architecture of the inner tube, D0, is very similar to the
structure we present here for the Shigella needle. Therefore, the question arises as to whether the Shigella
protein MxiH shares any homology with the termini of flagellin.
Although we cannot detect any sequence homology, there is some degree
of homology at the level of predicted secondary structure (using programs such as PredictProtein/PHDsec (23-25) and JPred (26)) because
we find that the N and C termini of MxiH and flagellin as well as those
of the periplasmic rod, hook, and filament junction proteins give
strong predictions for By extending the homology with the flagellar system, it was proposed
that the assembly of MxiH into the Shigella needle would require the presence of a cap at the needle tip (10). Located at the
end of the assembling structure, the cap would aid construction by
forcing consecutive subunits into the correct position within the helix
(27). This cap protein would be present in the needle complex at such a
low molar ratio to MxiH (and is likely to be of a similar size) that it
is not surprising that we do not biochemically detect it in our
purified needles. Further experiments will be required to address this issue.
The hypothesis of common ancestry is primarily attractive in providing
a mechanism by which secretion may be switched on by the tip of the
Shigella needle contacting a host cell (10). The mechanism
by which a signal is propagated from the needle tip to the ATPase,
putatively located at the base of the basal body at an approximate
800-Å distance, is difficult to understand on a molecular
level. The flagellar filament is able to switch helical forms depending
on the direction of motor rotation or in response to the chemical
environment (28). Switching is thought to be linked to small changes in
the conformation of the flagellin subunit, leading to gross
architectural changes in the filament as a whole (12). This in turn
leads to changes in the supercoiling of the filaments within the
flagellum and thus transmits the change in motor direction to the
rotating flagellum. The idea that this helical architecture might be
poised to switch between different conformational states suggests that
the effect of contact of the needle with the host cell may be to switch
the helical state of the needle, thus producing a signal that is almost
instantly transmitted to the basal end of the needle and the proteins
associated with the basal body.
Although this functional hypothesis remains to be proven, our structure
in demonstrating that the Shigella needle equates to at
least the D0 domain of the flagellar filament provides strong support
for the hypothesis that contact-activated secretion by Shigella is sensed by changes in the architecture of the
needle itself. This remains to be confirmed by studies of activated
needles; however, these are likely to be difficult because the
preparation of homogenous, highly concentrated samples of activated
needles presents many experimental challenges that will need to be overcome.
We can now combine all of the structural information regarding the
Shigella needle to present the model for the
Shigella TTSS shown in Fig. 4C. This figure
illustrates the global assembly of the TTSS and reinforces the
structural similarities to the flagellar hook-basal body complex,
raising interesting questions on how substrates are recruited,
partially unfolded, and secreted. In particular, understanding the
nature of the interfaces between the basal body and the needle and
understanding the nature of the interfaces between the needle and its
putative cap, which may allow a signal to be transduced to produce
secretion, remain an interesting challenge.
INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
EXPERIMENTAL PROCEDURES
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
strain harboring pKT001, which
encodes the cloned mxiH gene under an
isopropyl-1-thio-
-D-galactopyranoside-inducible
promoter (a gift of Prof. Chihiro Sasakawa, Tokyo University, Tokyo,
Japan) (11) hereafter termed
mxiH
/mxiH+++, is plated on a
Congo Red culture plate containing kanamycin (50 µg/ml), ampicillin
(100 µg/ml), and trimethoprim (5 µg/ml) as selecting
antibiotics. Within a week of plating the strain, a single Congo
Red-positive colony is selected and grown at 37 °C for 8-10 h in 60 ml of trypticase soy broth, plus antibiotics as above. Using this
culture as inoculate, it is diluted 1:100 into the main culture volume
of trypticase soy broth (plus antibiotics). After induction with 1 mM isopropyl-1-thio-
-D-galactopyranoside at
the time of inoculation, the culture is incubated overnight at 37 °C
to obtain maximum bacterial density.
/mxiH+++ grown overnight.
= 0.96 Å, d = 407 mm,
= 0°, and
exposure time = 300 s.
700 nm in low dose electron beam conditions
(
10 e
/Å2) on a Philips CM12 TEM at 120 kV. Micrographs were recorded at a ×60,000 magnification on Kodak
S0163 films.
RESULTS
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
strain harboring a plasmid carrying mxiH
under an inducible promoter (hereafter called
mxiH
/mxiH+++). This strain can
be induced to hugely overexpress the needle subunit MxiH. The
overexpression of MxiH produces a phenotypic change in the TTSS with
the mean needle length changing from ~500 Å in the wild-type
bacteria to ~3000 Å in the mutant (11). Using the protocol described
above, the needles (3000 Å in length on average) could be shaved from
the mutant bacteria and purified to greater than 95% homogeneity as
demonstrated by SDS-PAGE and electron microscopy analysis (Fig.
1).
View larger version (106K):
[in a new window]
Fig. 1.
Analysis of purified needles.
A, electron micrograph of a preparation of needles in
negative stain. B, SDS-PAGE analysis of needles shaved from
the surface of the mxiH /mxiH+++
Shigella flexneri mutant. Lane 1 shows a
preparation of Shigella long needles; lane 2 shows markers with masses as indicated.
View larger version (50K):
[in a new window]
Fig. 2.
Comparison of XRFD image from oriented needle
samples and EM power spectrum. A, half of the central region
of x-ray fiber diffraction pattern from aligned
mxiH /mxiH+++ long needles taken
at station 14.2 of the SRS Daresbury (see "Experimental
Procedures"). B, half of the averaged power spectrum
generated from 5402 partially overlapping segments extracted from
images of 108 negatively stained
mxiH
/mxiH+++ long needles. For
both images, the peaks corresponding to specific layer lines are
indicated with the layer line spacings and indices (as determined as in
text) shown. C, XRFD pattern as shown in A but
displaying the high resolution layer lines discussed in the text.
View larger version (17K):
[in a new window]
Fig. 3.
Convergence of IHRSR helical parameters.
Convergence of the helical pitch, axial rise, and azimuthal rotation as
a function of iteration is shown for one test of images from the
mxiH /mxiH+++ long needles. In
this example, the helical parameters were initially chosen to be
significantly different from the final value determined by other such
tests.
1. The reconstruction is reasonable
when a threshold is chosen that encloses this volume. If a threshold is
chosen that encloses a much smaller or larger volume, the
reconstruction is no longer reasonable (e.g. connectivity is
lost, the stain filled volume in the center is eliminated, and so on),
suggesting that there is a single copy of MxiH per asymmetric unit.
View larger version (71K):
[in a new window]
Fig. 4.
Various views of the three-dimensional
reconstruction. All of the images were produced using AESOP
(a program developed by M. E. M. Noble). A and
B, show a top view and side view,
respectively, of the three-dimensional reconstruction of the
Shigella flexneri needles obtained as described under
"Results." The electron density is thresholded as described in
the text to generate the solid surface. C, composite image
giving a model for the structure of the needle complex in the bacterial
membranes using the needle reconstruction presented here in combination
with the earlier reconstruction of the basal body of the needle complex
(9).
Helical parameters determined from the Shigella needle reconstructions
and comparison with those obtained for the flagellar filament (13) and
straightened hook (14)
DISCUSSION
TOP
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
REFERENCES
-helices at their termini. Therefore, our
data suggest that the needle represents the minimum core required to
build a helix of this size with this architecture. In contrast, the
flagellin subunit contains a large inserted domain between the N and C
termini that is not required for helix formation but presumably relates
to functions strictly required of flagellar filaments.
![]() |
ACKNOWLEDGEMENTS |
---|
We thank Shin-Ichi Aizawa (Utsunomiya, Japan) for suggesting and testing with us the needle purification protocol based on experience in purifying bacterial flagella. We thank Keiichi Namba for discussions on the structure of bacterial flagellar and advice on EM and x-ray fiber diffraction techniques, Daniel DeRosier for suggesting that we should try a single particle approach, and M. E. M. Noble (Oxford, United Kingdom) for repeated extensions to program AESOP to allow the construction of these figures.
![]() |
FOOTNOTES |
---|
* This work was also supported in part by National Institutes of Health GM66771 (to E. H. E.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ Funded by a Wellcome Trust Studentship.
Recipient of an Edward P. Abrahams Cephalosporin Trust D. Phil. studentship.
Present address: Laboratoire de Minéralogie
Cristallographie Paris, Université Paris 6, CNRS, 4 place Jussieu
F-75252 Paris Cedex 05, France.
¶¶ Laboratory is supported by the Guy G. F. Newton Cephalosporin Trust Senior Research Fellowship, the National Institutes of Health Career Transition Grant K22 AI01847, and the Sasakawa Fund at the Oriental Institute (Oxford, United Kingdom).
To whom correspondence should be addressed. Tel.:
44-1865-275181; Fax: 44-1865-275182; E-mail:
susan@biop.ox.ac.uk.
Published, JBC Papers in Press, February 5, 2003, DOI 10.1074/jbc.M300091200
![]() |
ABBREVIATIONS |
---|
The abbreviations used are: TTSS, type III secretion systems; IHRSR, iterated helical real space reconstruction method; XRFD, x-ray fiber diffraction.
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
1. | Dixon, M. F. (2000) in General and Systematic Pathology (Underwood, J. C. E., ed), 3rd Ed. , p. 383, Churchill Livingstone, London |
2. | Donnenberg, M. S. (2000) Nature 406, 768-774[CrossRef][Medline] [Order article via Infotrieve] |
3. | N. N. (2000) in The Merck Manual of Diagnosis and Therapy (Berkow, R., Burs, M., and Beers, M. H., eds) Centenial Edition, pp. 1164-1166, Merck Publishing Group, London, United Kingdom |
4. | Szewzyk, U., Szewzyk, R., Manz, W., and Schleifer, K. H. (2000) Annu. Rev. Microbiol. 54, 81-127[CrossRef][Medline] [Order article via Infotrieve] |
5. |
Hueck, C. J.
(1998)
Microbiol. Mol. Biol. Rev.
62,
379-433 |
6. |
Thomas, D.,
Morgan, D. G.,
and DeRosier, D. J.
(2001)
J. Bacteriol.
183,
6404-6412 |
7. |
Blocker, A.,
Gounon, P.,
Larquet, E.,
Niebuhr, K.,
Cabiaux, V.,
Parsot, C.,
and Sansonetti, P.
(1999)
J. Cell Biol.
147,
683-693 |
8. |
Kubori, T.,
Matsushima, Y.,
Nakamura, D.,
Uralil, J.,
Lara-Tejero, M.,
Sukhan, A.,
Galan, J. E.,
and Aizawa, S.
(1998)
Science
280,
602-605 |
9. | Blocker, A., Jouihri, N., Larquet, E., Gounon, P., Ebel, F., Parsot, C., Sansonetti, P., and Allaoui, A. (2001) Mol. Microbiol. 39, 652-663[CrossRef][Medline] [Order article via Infotrieve] |
10. | Blocker, A., Komoriya, K., and Aizawa, S. (2003) Proc. Natl. Acad. Sci. U. S. A., in press |
11. |
Tamano, K.,
Aizawa, S.,
Katayama, E.,
Nonaka, T.,
Imajoh-Ohmi, S.,
Kuwae, A.,
Nagai, S.,
and Sasakawa, C.
(2000)
EMBO J.
19,
3876-3887 |
12. | Samatey, F. A., Imada, K., Nagashima, S., Vonderviszt, F., Kumasaka, T., Yamamoto, M., and Namba, K. (2001) Nature 410, 331-337[CrossRef][Medline] [Order article via Infotrieve] |
13. | Yamashita, I., Hasegawa, K., Suzuki, H., Vonderviszt, F., Mimori-Kiyosue, Y., and Namba, K. (1998) Nat. Struct. Biol. 5, 125-132[Medline] [Order article via Infotrieve] |
14. | Morgan, D. G., Macnab, R. M., Francis, N. R., and Derosier, D. J. (1993) J. Mol. Biol. 229, 79-84[CrossRef][Medline] [Order article via Infotrieve] |
15. | Mimori, Y., Yamashita, I., Murata, K., Fujiyoshi, Y., Yonekura, K., Toyoshima, C., and Namba, K. (1995) J. Mol. Biol. 249, 69-87[CrossRef][Medline] [Order article via Infotrieve] |
16. | Yamashita, I., Vonderviszt, F., Noguchi, T., and Namba, K. (1991) J. Mol. Biol. 217, 293-302[Medline] [Order article via Infotrieve] |
17. | Yamashita, I., Suzuki, H., and Namba, K. (1998) J. Mol. Biol. 278, 609-615[CrossRef][Medline] [Order article via Infotrieve] |
18. | Frank, J., Radermacher, M., Penczek, P., Zhu, J., Li, Y. H., Ladjadj, M., and Leith, A. (1996) J. Struct. Biol. 116, 190-199[CrossRef][Medline] [Order article via Infotrieve] |
19. | Egelman, E. H. (2000) Ultramicroscopy 85, 225-234[CrossRef][Medline] [Order article via Infotrieve] |
20. | Holmes, K. C., Popp, D., Gebhard, W., and Kabsch, W. (1990) Nature 347, 44-49[CrossRef][Medline] [Order article via Infotrieve] |
21. | DeRosier, D. J., and Klug, A. (1968) Nature 217, 130-134 |
22. | Penczek, P., Radermacher, M., and Frank, J. (1992) Ultramicroscopy 40, 33-53[CrossRef][Medline] [Order article via Infotrieve] |
23. | Rost, B. (1996) in Computer Methods for Macromolecular Sequence Analysis (Doolittle, R. F. , Abelson, J. N. , and Simon, M. I., eds), Vol. 266 , pp. 525-539, Academic Press, Orlando, FL |
24. | Rost, B., and Sander, C. (1994) Proteins Struct. Funct. Genet. 19, 55-72[Medline] [Order article via Infotrieve] |
25. | Rost, B., and Sander, C. (1993) J. Mol. Biol. 232, 584-599[CrossRef][Medline] [Order article via Infotrieve] |
26. | Cuff, J. A., Clamp, M. E., Siddiqui, A. S., Finlay, M., and Barton, G. J. (1998) Bioinformatics 14, 892-893[Abstract] |
27. | Yonekura, K., Maki-Yonekura, S., and Namba, K. (2001) J. Struct. Biol. 133, 246-253[CrossRef][Medline] [Order article via Infotrieve] |
28. | Calladine, C. R. (1975) Nature 255, 121-124[Medline] [Order article via Infotrieve] |