Cell, Vol. 120, 329–341, February 11, 2005, Copyright ©2005 by Elsevier Inc. DOI 10.1016/j.cell.2005.01.007
MreB Actin-Mediated Segregation of a
Specific Region of a Bacterial Chromosome
Zemer Gitai,1,3,* Natalie Anne Dye,2Ann Reisenauer,1
Masaaki Wachi,4and Lucy Shapiro1,*
1Department of Developmental Biology
2Department of Biochemistry
Beckman Center, School of Medicine
Stanford, California 94305
3Department of Molecular Biology
Princeton, New Jersey 08544
4Department of Bioengineering
Tokyo Institute of Technology
4259 Nagatsuta, Midori-ku
B. subtilis localize their origins of replication to the mid-
cell and quarter-cell positions at different stages of the
cell cycle (Li et al., 2002; Webb et al., 1997), whereas in
the crescent-shaped Caulobacter crescentus, chromo-
somal origins are at the cell poles (Jensen and Shapiro,
1999). Direct examination of chromosome segregation
in live cells revealed that the origin region (defined as
the site of the initiation of DNA replication) is the first
region of the chromosome to be segregated in all spe-
cies examined (Lau et al., 2003; Li et al., 2002; Viollier
et al., 2004; Webb et al., 1998).
The rapid, directed movement of the origins excludes
the cell growth-related mechanisms initially proposed
for chromosome segregation (Jacob et al., 1963) and is
suggestive of an active mechanism for origin transport.
Several candidates have been put forth as potential
contributors to the force that separates the chromo-
somes. The coincident timing of DNA replication and
segregation suggests that these processes are coupled
and that the act of DNA replication could provide the
motive force for chromosome segregation. The obser-
vation that the DNA replication machinery is stationary
in B. subtilis and E. coli led to the refinement of such
models into the “extrusion-capture” model (Lemon and
Grossman, 2001), wherein the extrusion of DNA from an
immobile replisome forces it poleward, where it is then
captured by as-yet unidentified factors. Coordinate
transcription of origin-proximal reading frames oriented
away from the origin has also been proposed to con-
tribute to the movement of bacterial chromosomes
(Dworkin and Losick, 2002). Though these models are
compelling and elegant, they have yet to be directly val-
In contrast, the mechanisms directing eukaryotic
chromosome segregation are well characterized. Eu-
karyotes use a microtubule-based cytoskeletal system
for chromosome partitioning. After replication, chromo-
some sisters are paired, and microtubules are attached
to specific, centromeric chromosomal loci through a ki-
netochore protein complex. Once attached, the micro-
tubules direct chromosome migration to opposite poles
through dynamic polymerization cycles and the action
of motor proteins (Kline-Smith and Walczak, 2004).
Could a similar mechanism also contribute to bacte-
rial chromosome segregation? The bacterial tubulin ho-
molog, FtsZ (Lowe and Amos, 1998), is an unlikely can-
didate, as ftsZ mutants do not display chromosome
segregation defects, and FtsZ’s midcell subcellular lo-
calization is inconsistent with a role in chromosome
segregation (Lutkenhaus and Addinall, 1997; Quardo-
kus et al., 1996). A more attractive possibility is the re-
cently identified bacterial actin homolog, MreB (Jones
et al., 2001; van den Ent et al., 2001). First, the origin of
replication is found in abnormal numbers and positions
in mreB mutants (Gitai et al., 2004; Kruse et al., 2003;
Soufo and Graumann, 2003); second, MreB localizes to
a spiral that traverses the length of the cell (Figge et
al., 2004; Gitai et al., 2004; Jones et al., 2001; Shih et
al., 2003), such that MreB could potentially be enlisted
to move chromosomes from one end of the cell to an-
Faithful chromosome segregation is an essential com-
ponent of cell division in all organisms. The eukary-
otic mitotic machinery uses the cytoskeleton to move
specific chromosomal regions. To investigate the po-
tential role of the actin-like MreB protein in bacterial
chromosome segregation, we first demonstrate that
MreB is the direct target of the small molecule A22.
We then demonstrate that A22 completely blocks the
movement of newly replicated loci near the origin of
replication but has no qualitative or quantitative ef-
fect on the segregation of other loci if added after ori-
gin segregation. MreB selectively interacts, directly or
indirectly, with origin-proximal regions of the chromo-
some, arguing that the origin-proximal region segre-
gates via an MreB-dependent mechanism not used by
the rest of the chromosome.
To maintain a consistent number of chromosomes,
cells must ensure that during cell division, the chromo-
some is reproducibly replicated and that each of the
new chromosomes is segregated to one of the two
daughter cells. Although the ability of bacteria to faith-
fully replicate and segregate their chromosomes has
been appreciated for over forty years (Jacob et al.,
1963), the molecular mechanism by which bacteria exe-
cute the segregation process has remained elusive.
Observing the placement and movement of chromo-
somal loci revealed that all bacteria examined have
highly reproducible, ordered chromosomal structure,
with loci exhibiting a linear correlation between their
genetic location on the chromosome and their place-
ment along the long axis of the cell (Niki et al., 2000;
Teleman et al., 1998; Viollier et al., 2004). Nonetheless,
the details of chromosomal topology vary among spe-
cies. For example, the rod-shaped bacteria E. coli and
Figure 1. A22 Affects Caulobacter Growth and Morphology
(A) The chemical structure of S-(3,4-dichlorobenzyl)isothiourea, A22.
(B) Growth curve of cells treated for up to 6 hr with 0, 0.1, 1, 10, and 100 ?g/ml of A22 or an equivalent volume of methanol (MeOH).
(C) Colony formation assay curve for cells treated with 0, 0.1, 1, 10, and 100 ?g/ml of A22 or an equivalent volume of methanol (MeOH).
(D) DIC images of Caulobacter grown for 6 hr in 0, 1, 10, or 100 ?g/ml of A22.
(E) Timecourse of cells grown in 10 ?g/ml of A22 for 0, 2, 4, 6, and 8 hr.
(F) Timecourse of cells depleted of mreB for 0, 6, 12, 18, and 24 hr. Scalebars represent 1 ?m.
other; finally, there is precedent for actin’s involvement
in bacterial DNA segregation, as another bacterial actin
homolog, the R1 plasmid-specific ParM protein, has
been directly implicated in R1 plasmid partitioning
(Moller-Jensen et al., 2002, 2003).
Though these results are tantalizing, demonstrating a
direct role for MreB in chromosome segregation has
proven difficult. MreB loss of function is lethal and pleio-
tropic (Figge et al., 2004; Gitai et al., 2004; Jones et
al., 2001), disrupting multiple cellular processes distinct
from chromosome dynamics, including cell shape de-
termination, polar protein localization, and cell division.
Thus, it remains unclear whether MreB plays a primary
role in chromosome segregation, or if MreB’s effect on
chromosome dynamics is a secondary consequence of
Here we present the first evidence that supports a
direct role for MreB in the segregation of a specific re-
gion of the chromosome. We reached these conclu-
sions by first characterizing a small molecule, A22, that
specifically, rapidly, and reversibly perturbs MreB func-
tion. Our studies focus on the differentiating bacterium,
Caulobacter crescentus, whose ability to be synchro-
nized renders it particularly powerful for the analysis
of chromosome segregation (Jensen et al., 2002). By
administering the MreB-perturbing compound at dif-
ferent stages of the Caulobacter cell cycle we show
that origin-proximal loci segregate through an MreB-
dependent mechanism, and that the rest of the chro-
mosome follows the origin using an MreB-independent
mechanism. The ability of A22 to block DNA segrega-
tion without affecting DNA replication also demon-
strates that the process of replication is not sufficient to
separate chromosomes, as was previously proposed.
Cytoskeletal Segregation of a Bacterial Origin
Table 1. mreB Mutations Confer A22 Resistance
Doubling in A22
when Replaced by
Doubling in A22
in WT Genetic
Mutation# Times Isolated
aAll doubling times refer to increased cell mass over a period of 6 hr as measured by optical density (A660). Wild-type (wt) cells do not divide
in the presence of A22.
Finally, we used chromatin immunoprecipitation assays
to demonstrate a specific physical association between
MreB and origin-proximal loci. Together, these results
suggest that Caulobacter chromosome segregation is
mediated by a cytoskeletally driven mechanism, with
regions near the origin functioning as a centromere.
teriostatic but not bacteriocidal. At a concentration of
1 ?g/ml, A22 partially slowed colony formation, and at
0.1 ?g/ml or methanol alone, colony formation was un-
perturbed (Figure 1C).
Consistent with previous observations in E. coli (Iwai
et al., 2002), A22 caused a dose-dependent alteration
of Caulobacter morphology, transforming normal, cres-
cent-shaped cells into rounded, lemon-shaped cells.
The progression of cell shape deformation by A22 treat-
ment was qualitatively similar to that caused by mreB
depletion (Figge et al., 2004; Gitai et al., 2004), though
the A22 effect was more rapid (Figures 1D–1F). At 0.1,
1, and 10 ug/ml, A22 both slowed cell growth and al-
tered cell shape, while treatment with 100 ?g/ml A22
Caulobacter to round up (Figure 1D). This result sug-
gests that the manifestation of A22’s effect on morphol-
ogy requires cell growth. For example, A22 could dis-
rupt the localization of new cell wall synthesis but have
no effect on the existing cell wall, such that without
additional growth and wall synthesis A22’s effects can-
not be perceived. A dependence on cell growth could
also explain why A22’s effect on morphology takes sev-
eral hours to manifest.
A22 Is a Small Molecule that Affects Caulobacter
Growth and Morphology
To dissect the cellular activities of Caulobacter MreB,
we sought a method to acutely perturb MreB function.
mreB mutants are lethal, and genetically depleting
mreB is slow, requiring multiple cell cycles before MreB
protein levels are significantly reduced (Figge et al.,
2004; Gitai et al., 2004). For a faster method of disrupt-
ing MreB function we explored the use of small mole-
cules, an approach that has proven highly successful
for dissecting the functions of the eukaryotic cytoskele-
ton (Fenteany and Zhu, 2003). Treatment of Caulobacter
with drugs, including cytochalasin D, latrunculin, jas-
plakinolide, and phalloidin, that affect eukaryotic actin
failed to perturb growth rate or morphology (data not
shown), suggesting that they either do not affect MreB
or cannot enter Caulobacter cells. One small molecule
that was known to affect bacteria, A22, showed more
promise. A22 (S-(3,4-dichlorobenzyl)isothiourea; Figure
1A) was identified through a screen for chemicals that
induce the formation of anucleate E. coli cells; A22 per-
turbs E. coli morphology in a manner reminiscent of
mreB mutants (Iwai et al., 2002).
Addition of A22 to Caulobacter cultures slowed
growth in a dose-dependent fashion (Figure 1B). Similar
volumes of solvent (methanol) alone had no effect on
growth. In cultures with a normal generation time of 87
min in rich PYE medium, a decline in growth rate was
observed as early as 30 min after A22 treatment, sug-
gesting that A22 rapidly exerts its effects on the cell.
This rapid effect on cell growth could be due to a dis-
ruption in cell wall deposition or could reflect a second-
ary response to some other cellular perturbation.
To assess the impact of A22 on cell division and sur-
vival, we examined the effect of A22 on colony forma-
tion. At 10 ?g/ml and higher, A22 completely halted the
increase in CFU, without causing it to diminish (Figure
1C), consistent with a block in cell division that is bac-
MreB Is the Direct Target of A22
To identify the cellular target(s) of A22, we carried out
a screen for A22-resistant mutants by growing large
numbers of Caulobacter on plates containing 10 ?g/ml
A22. We independently isolated 20 mutants that stably
and repeatedly formed colonies on plates containing 10
?g/ml A22 and grew in liquid culture containing 10 ?g/
ml A22. Since MreB was proposed as a possible target
of A22 (Iwai et al., 2002), we PCR-amplified and se-
quenced the mreB gene in the 20 A22-resistant strains
and a control wild-type strain. Each of the 20 A22-resis-
tant strains contained a single missense point mutation
in its mreB gene, whereas the wild-type A22-sensitive
strain had no mreB point mutations (Table 1). Though
isolated independently, the 20 strains had only 7 dif-
ferent point mutations: one specific point mutation was
found in 10 of the strains, another specific point muta-
tion was found in 5 of the strains, and 5 additional point
mutations were found in 1 strain each. Since all 7 of
these mutations affected amino acids that are con-
served between Caulobacter MreB and Thermotoga
maritima MreB, they could be mapped onto the solved
Figure 2. A22-Resistant mreB Mutations Map to the Nucleotide Binding Pocket of MreB
(A) A Clustal W alignment of the Caulobacter (C. cres), Thermotoga maritima (T. mari), and Saccharomyces cerevisiae (S. cer) MreB/Actin
proteins. Perfectly conserved residues are highlighted in gray and/or printed with white characters. The residues that form the active sites in
the crystal structures of the T. maritima and S. cerevisiae proteins are highlighted in red (van den Ent et al., 2001). The residues that are
mutated in the 20 A22-resistant Caulobacter strains are highlighted in green, and the corresponding mutation is denoted with an arrow. The
mutations that were identified multiple times are identified with the number of occurrences in parentheses.
(B) The T. maritima residues that correspond to those mutated in A22-resistant strains are highlighted in green in its crystal structure. The
AMPPNP ligand in this MreB structure is highlighted in red. Shown are two rotational views of the same structure.
(C) DIC (left) and fluorescence (right) images of cells expressing the T158A A22-resistant mreB allele fused to gfp.
(D) Fluorescence image of T158A GFP-MreB-expressing cells that have been treated with 10 ?g/ml A22 for 10 min. Scalebars represent 1 ?m.
three-dimensional structure of T. maritima MreB (van
den Ent et al., 2001). Interestingly, 5 of the 7 mutated
residues are located in MreB’s ATP binding site, and
the remaining 2 residues reside side-by-side in a helix
that also contacts the nucleotide, suggesting that A22
may interact with MreB’s nucleotide binding pocket
(Figures 2A and 2B). All of the resistant strains grow at
a similar rate in rich PYE medium (w115 min/doubling),
which is slower than wild-type cells without drug (87
min/doubling), but they are largely unaffected by the
presence of A22 (w115 min/doubling) (Table 1). All of
the resistant strains also exhibit a slightly abnormal
morphology, with cells that were uncharacteristically
straight and long, and exhibited occasional kinks (Fig-
ure 2C). The perturbation in cell shape is consistent
with an effect on MreB’s role in cell-shape determina-
tion, and the phenotypic similarity of all 20 strains sug-
gests a common mode of action, such as altering the
MreB nucleotide binding pocket.
To determine if each of these mutations in mreB was
necessary for A22 resistance, an apramycin-resistance
cassette (Blondelet-Rouault et al., 1997) was integrated
immediately upstream of the mreB gene in wild-type
cells. Phage transduction and apramycin selection thus
enabled the replacement of the A22-resistant mreB mu-
tant loci with wild-type mreB. In all 20 cases, this ma-
nipulation caused the A22-resistant strains to become
A22 sensitive, suggesting that a mutation linked to
mreB is necessary for A22 resistance (Table 1). Con-
versely, the same apramycin-resistance cassette was
also integrated upstream of each of the seven different
A22-resistant mreB loci. Phage transduction was used
again, this time to replace the mreB loci of wild-type
strains with the mutant mreB. In each case, the result-
ing strains became A22 resistant, indicating that each
of these mreB mutations is sufficient to confer resis-
tance, and suggesting that there are no unlinked mut-
ations in the strain backgrounds that are necessary for
resistance (Table 1). Thus, mreB missense alleles ap-
pear to be necessary and sufficient for A22 resistance.
A22 Rapidly and Reversibly Delocalizes GFP-MreB
During the Caulobacter cell cycle, MreB forms a
lengthwise spiral in swarmer cells and stalked cells
(Figge et al., 2004; Gitai et al., 2004). It then appears to
condense into an increasingly tight ring positioned at
the future site of cell division. When cytokinesis initi-
ates, this ring is replaced by a lengthwise spiral, such
that upon cell division each daughter cell inherits a fully
expanded spiral (Figge et al., 2004; Gitai et al., 2004).
GFP-MreB expressing strains exhibit regularly spaced
puncta and bands that resolve into three-dimensional
Cytoskeletal Segregation of a Bacterial Origin
Figure 3. A22 Rapidly and Reversibly Delocalizes GFP-MreB
(A) DIC (left) and fluorescence (right) images of GFP-MreB-expressing Caulobacter grown for 1 hr in 0, 1, 10, or 100 ?g/ml of A22 imaged on
agarose pads containing the same concentration of A22.
(B) Fluorescence image of GFP-MreB-expressing Caulobacter imaged after being mounted on an agarose pad containing 10 ?g/ml of A22
for 1 min, with no previous incubation with A22.
(C) Fluorescence image of GFP-MreB-expressing Caulobacter that have been treated with 10 ?g/ml A22 for 1 hr and then mounted for 1 min
onto an agarose pad that does not contain A22.
(D–F) Timelapse images of individual Caulobacter cells expressing GFP-MreB at three different cell cycle stages: an early predivisional cell
with a tight medial MreB ring (D), a mid-predivisional cell with an expanding MreB ring (E), and a late predivisional cell with an expanded
MreB spiral (F). These cells were imaged in a flow chamber to which they were attached with polylysine. For each series, the image on the
left is taken before A22 treatment, the image in the middle is taken 1 min after washing in 10 ?g/ml A22, and the image on the right is taken
2 min after washing in medium lacking A22. Scalebars represent 1 ?m.
spirals and rings when examined by deconvolution mi-
croscopy (Gitai et al., 2004). Treatment with 10 or 100
?g/ml of A22 completely disrupted this GFP-MreB lo-
calization, resulting in cells with diffuse fluorescence
that contain no discernable puncta or bands (Figure
3A). MreB puncta and bands were largely maintained
when cells were treated with 1 ?g/ml or less of A22.
Since 10 ?g/ml of A22 strongly delocalizes GFP-MreB
and blocks cell division without completely inhibiting
cell growth, we focused most of our subsequent A22
studies on this concentration.
The onset of A22’s effect on GFP-MreB is very rapid:
the GFP-MreB delocalized within 1 min (Figure 3B).
A22’s effect on GFP-MreB is rapidly reversible. Diluting
the A22 concentration of cells treated with A22 for 1 hr
by mounting them on pads lacking A22 leads to a full
recovery of MreB localization within 1 min (Figure 3C).
GFP-MreB was monitored in the same cells before, dur-
ing, and after A22 treatment by timelapse imaging of
cells in a flow chamber that allows A22 to be washed
in and out (Figures 3D–3F). A22 delocalized both MreB
spirals (puncta) and rings (bands), though MreB rings
took longer to delocalize completely, suggesting that
they may be more stable or may consist of more MreB
filaments than the spirals. Cells generally retained a
memory of the MreB structure present before A22 treat-
ment (Figures 3D–3F). Cells with rings recovered rings,
cells with spirals recovered spirals, and cells with par-
tially compacted rings recovered partially compacted
rings. These findings are consistent with MreB’s local-
ization being determined by extrinsic factors regulated
by the cell cycle state, rather than an intrinsic MreB
We constructed a fusion of the most common A22-
resistant mreB missense mutation (T158A) to GFP and
asked if the cellular organization of this MreB mutant is
maintained in 10 ?g/ml of A22. This fusion was ex-
pressed in the corresponding A22-resistant strain, pro-
ducing merodiploid cells containing both GFP-tagged
and -untagged mutated mreB, but no wild-type mreB.
These cells still exhibited a punctate MreB localization
but did not appear to fully condense their spirals into
rings (Figure 2C). Since this mutation in the MreB ATP
binding pocket perturbs MreB dynamic localization, nu-
cleotide hydrolysis may play a role in regulating MreB
dynamics. The localization of the mutated GFP-MreB
was completely unaffected by treatment with A22 (Fig-
ure 2D). Together, all of these results strongly suggest
that MreB is the target of A22.
A22 Treatment Blocks Chromosome Segregation
Having established that A22 targets MreB and rapidly
disrupts its localization, we used A22 to assess the
acute role of MreB in chromosome segregation. To this
end we took advantage of strains previously generated
in our group that use the fluorescent repressor-operator
Figure 4. Treatment of Swarmer Cells with A22 Blocks DNA Replication and/or Chromosome Segregation
To determine A22’s effect on chromosome segregation, swarmer cells doubly labeled at their origins (green) and the CC2943 locus (red) were
isolated and incubated in the presence of A22.
(A–C) DIC (left), YFP + CFP overlayed fluorescence (middle), and schematic (right) images. LacI-CFP is bound to lacO arrays integrated at
the origin and is shown in green in the overlay. TetR-YFP is bound to tetO arrays integrated at CC2943 and is shown in red in the overlay: (A)
isolated swarmer cells, (B) cells grown synchronously for 150 min in medium lacking A22, and (C) containing 10 ?g/ml A22.
(D) Swarmer cells from T158A A22-resistant Caulobacter with labeled origins were isolated and incubated for 150 min in 10 ?g/ml A22. Shown
are DIC (left) and CFP fluorescence images of LacI-CFP bound to the origin.
(E) Schematic of the genetic positions of the origin and CC2943 in the Caulobacter genome.
(F) Percentage of doubly labeled origin and CC2943 FROS cells with two distinct foci at different times after synchrony and growth with or
without A22. At least 75 cells were analyzed for each time point, with standard deviations of 3%–5%. Below the graph is a schematic
illustration of what is observed without A22 (top line) and with A22 (bottom line). The green dots represent the origin and the red dots
represent CC2943. Scalebars represent 1 ?m.
system (FROS) to simultaneously label two specific
chromosomal loci in the same living cell (Lau et al.,
2003; Viollier et al., 2004). This system uses fusions of
two DNA binding proteins, the lac repressor (LacI) and
the Tet repressor (TetR), to CFP and YFP, respectively.
Two arrays, each containing multiple tandem repeats of
either the LacI or TetR binding sites, are integrated into
different chromosomal loci. Each fusion protein clus-
ters at its respective array, causing the cellular address
of that array to fluoresce in a detectable fashion.
At each division, Caulobacter divides asymmetrically
to produce daughters with distinct morphologies and
cell fates: a stalked cell that initiates DNA replication
and a G1-arrested swarmer cell (see diagram in Figure
4). The swarmer cell later differentiates into a stalked
cell, at which time DNA replication initiates (Marczynski
and Shapiro, 2002). Highly purified populations of
swarmer cells can be isolated by density centrifugation
and they then synchronously progress through the cell
cycle once resuspended in fresh medium. Swarmer
cells contain only one chromosome, oriented such that
the origin of replication is at the flagellated pole, the
terminus is at the other pole, and all other loci are lin-
early arrayed in between (Jensen and Shapiro, 1999;
Viollier et al., 2004). Soon after replication initiation, one
of the duplicated origins moves rapidly to the opposite
pole while the other origin remains at its original pole.
Other loci are then sequentially replicated and segre-
gated to their correct final cellular positions (Viollier et
Synchronized swarmer cells with lac arrays at the
origin locus and tet arrays at the midcell-positioned
CC2943 locus (Figure 4A; Viollier et al., 2004) were incu-
bated with 10 ?g/ml A22. After varying periods of time
in the presence of A22, the cellular positions of the ori-
gin and CC2943 were examined by fluorescence mi-
croscopy. The cultures continued to increase in cell
mass, as measured by an increase in optical density,
but cell division was blocked. Although two duplicated
and segregated loci were observed in untreated cells
at time points from 30–180 min, the A22-treated cells
continued to exhibit only one focus for each locus (Fig-
ures 4B, 4C, and 4F). This effect was reversible, since
upon washing out the A22 drug, the origins were rapidly
and normally segregated (data not shown). The origin
successfully segregated when this A22 treatment was
Cytoskeletal Segregation of a Bacterial Origin
repeated with cells whose wild-type mreB locus was
replaced with the T158A A22-resistant mreB allele (Fig-
ure 4D). This result confirms that A22 affects chromo-
some segregation by acting on MreB and not another
gated loci. A22 thus appears to block chromosome
segregation when added to cells prior to the replication
and polar movement of the origin. Moreover, the act of
DNA replication in and of itself seems to be insufficient
to move two loci apart.
A22 Treatment Does Not Perturb the Segregation
of Origin-Distal Loci
Treatment of swarmer cells with A22 before the origins
have segregated blocks the movement of chromosomal
loci. But what happens when A22 is delivered after the
origins have replicated and moved to opposite poles?
To address this question, cells double-labeled at the
origin and CC2943 were synchronized (Figures 6A and
6B). This time, however, we did not immediately treat
the swarmer cells with A22 but rather waited until the
duplicated origins had moved to the poles (30 min) be-
fore adding A22. At this point in the cell cycle, the
CC2943 locus has yet to replicate. Consequently, such
cells exhibit two origin foci but only one CC2943 focus.
Surprisingly, these cells proceeded to replicate and
separate their CC2943 locus into two foci (Figure 6).
The ability of the CC2943 locus to duplicate and sepa-
rate in the presence of A22 supports the conclusion
that A22 does not block replication. In addition, this
result suggests that the origin and CC2943 move apart
through separate sequential mechanisms, with the ori-
gin being dependent on, and CC2943 independent of,
Since the origin and CC2943 differ in their depen-
dence on MreB for segregation, similar experiments (in
which the MreB A22 inhibitor was added to synchro-
nized cultures after the origin had duplicated and
moved poleward) were performed on cells doubly la-
beled at their origins and sites in between CC2943 (840
kb to the left of the origin) and the origin. One of these
strains was fluorescently tagged at CC3300 (460 kb to
the left of the origin), and one at CC3656 (100 kb to the
left of the origin) (Figure 6A). These loci behaved just
like CC2943: they duplicated and separated in the pres-
ence of A22 so long as the drug was administered after
origin segregation (Figure 6E). An additional locus 100
kb to the right of the origin (CC0091) was also not de-
pendent on A22 for the separation of its loci following
duplication (Figure 6E). Thus, A22 appears to affect the
separation and movement of only a relatively small por-
tion of the genome located near the origin.
Though origin-distal loci can separate in the pres-
ence of A22, MreB could still partially contribute to their
cellular positioning and rate of movement across the
cell. Technical considerations prevent the quantitative
analysis of locus movement in doubly labeled cells.
Thus, to quantitatively assess the role of MreB in origin-
distal locus segregation, we examined two singly la-
beled strains, one with CC0786 marked at 860 kb to the
right of the origin, and one with CC3192 marked at 576
kb to the left of the origin (Viollier et al., 2004). These
cells were synchronized and grown in medium lacking
A22 for 50 min, a length of time that has been pre-
viously established as more than sufficient to allow ori-
gin duplication and segregation (Viollier et al., 2004).
After 50 min, A22 was added, the cells were mounted
onto agarose pads containing A22, and they were time-
A22 Does Not Block DNA Replication
Two scenarios could explain the A22-induced persis-
tence of single origin and CC2943 foci. Either the chro-
mosome never replicated such that the foci represent
the original single copy of each locus, or the chromo-
some was replicated but failed to segregate such that
both copies of each locus remain stuck together and
cannot be resolved. We thus used several independent
assays to distinguish between A22 impairment of DNA
replication or chromosome segregation.
We first examined the incorporation of radiolabeled
dGTP into DNA, which linearly correlates with the rate
of DNA replication (Marczynski and Shapiro, 1992).
Asynchronous cultures incorporated the same level of
radioactivity regardless of whether or not they were
treated with A22, directly demonstrating that A22 does
not interfere with bulk DNA synthesis (data not shown).
To assess the effect of A22 on DNA replication during
the cell cycle, the radiolabeled nucleotide incorporation
assay was repeated using synchronized cultures at dif-
ferent points in the cell cycle. Untreated cells display a
dramatic rise in radioactivity levels characteristic of
DNA replication initiation at the swarmer-to-stalked cell
transition (Figure 5A); 30–40 min. This rise in radioactiv-
ity levels was also observed in A22-treated cells. A22
treatment slightly delayed replication initiation, though
this delay is insufficient to explain A22’s dramatic inhi-
bition of chromosome movement. Cells treated with hy-
droxyurea (HU), a replication inhibitor, exhibited no
such increase in radioactivity levels (Figure 5A).
To assess the ability of these cells to assemble a re-
plisome at the replication origins, A22 was adminis-
tered to cells expressing a fusion of GFP to the HolB
subunit of DNA polymerase. Known inhibitors of DNA
replication, HU (Figure 5B) and novobiocin (data not
shown), can cause the rapid delocalization of HolB-
GFP (Jensen et al., 2001). One hour of A22 treatment,
however, had no effect on HolB-GFP localization (Fig-
ure 5B), demonstrating that A22 allows the replisome
Finally, we examined the DNA content of A22-treated
cells by FACS. Whereas untreated cultures contain cells
with a distribution of DNA content from one to two chro-
mosomes, A22-treated cultures predominantly con-
tained cells with two chromosomes (Figure 5C). Cells
incubated with A22 failed to divide, as indicated in the
diagram shown in Figure 5C. A22-treated cells thus ini-
tiate and finish replication but do not divide or reinitiate
replication, suggesting either the presence of a segre-
gation defect-induced cell cycle checkpoint or an inde-
pendent role for MreB in cell division.
Since A22 blocks neither the initiation nor the pro-
gression of DNA replication, we conclude that cells la-
beled at the origin and CC2943 and treated with A22
actually contained two copies of each of these loci.
Since only one focus was observed for each, these foci
must have contained two tightly associated, unsegre-
Figure 5. A22 Does Not Block DNA Repli-
(A) DNA synthesis measured by the incorpo-
ration of32P-radiolabeled dGTP relative to
OD660in synchronized cells grown for vary-
ing lengths of time in no drug, 10 ?g/ml A22,
or 5 mg/ml HU.
(B) Fluorescence images
expressing Caulobacter grown for 1 hr in no
drug (left), 10 ?g/ml A22 (middle), and 5 mg/ml
(C) A schematic illustrating the proposed ef-
fect of A22 treatment on origin (green) dupli-
cation and movement. To the right of each
schematic is the FACS analysis of cells treated
for 8 hr with or without 10 ?g/ml A22. The
dotted lines represent the one chromosome
and two chromosome peaks. Scale bars re-
present 1 ?m.
lapse imaged for 60 min, with pictures taken every 2
min. The timelapses were analyzed using automated
software that provides the absolute position of each
locus in each cell (Viollier et al., 2004). Timelapse series
for 15 cells were averaged (Figure 6F) to obtain a rate
of segregation. For both loci examined, the locus posi-
tion and rate of segregation were unaffected by the
presence of A22 (Figure 6F). Thus, MreB is both qualita-
tively and quantitatively dispensible for the movement
of origin-distal loci, suggesting that it does not partici-
pate in this process.
dent experiments, the origin of replication, the gidA
promoter region (8 kb to the left of the origin), and the
CC3752 gene (12 kb to the left of the origin) robustly
amplified, indicating that they biochemically associate
with MreB (Figure 7). We detected a faint signal with
primers to the dnaA promoter located 6 kb to the right
of the origin (Figure 7), possibly reflecting the fact that
dnaA is at the edge of the MreB binding region. When
regions farther from the origin (64 kb, 760 kb, and 2 Mb
to the left and 50 kb and 1.17 Mb to the right) were
amplified from the MreB ChIPs, no signal was detected
(Figure 7), indicating that MreB does not associate with
these sites and that the association of MreB with the
chromosome exhibits specificity. The origin coimmuno-
precipitation with MreB was significantly reduced,
though not abolished, in the presence of A22 (data
As a control, we performed ChIP experiments with
an anti-CtrA antibody. CtrA is a DNA binding response
regulator whose binding sites throughout the genome
have been systematically characterized (Laub et al.,
2002), such that we know the identities both of loci that
bind CtrA as well as those that do not bind CtrA. In
multiple independent experiments, five known CtrA
targets (PctrA, PcreS, PgidA, ori, and PCC1035) were
MreB Associates with Origin-Proximal Loci
The specific inhibition of origin movement by A22
raises the possibility that the duplicated origin moves
to the opposite pole through a specific physical associ-
ation with MreB, much as eukaryotic centromeres are
segregated by association with microtubules. We sought
to detect such an association in vivo by chromatin im-
munoprecipitation (ChIP). Following previously estab-
lished protocols (Holtzendorff et al., 2004; Lin and Gross-
man, 1998), Caulobacter extracts were ChIP’ed with
polyclonal antisera raised against purified Caulobacter
MreB. PCR was then used to amplify specific segments
of the Caulobacter chromosome. In multiple indepen-
Cytoskeletal Segregation of a Bacterial Origin
Figure 6. A22 Does Not Block the Movement of Origin-Distal Loci
(A) Schematic of the genetic positions of the origin and the other loci examined.
(B–D) YFP + CFP overlayed fluorescence images (left) and schematic illustrations (right). LacI-CFP is bound to lacO arrays integrated at the
origin and is shown in green in the overlay. TetR-YFP is bound to tetO arrays integrated at CC2943 and is shown in red in the overlay. Cells
were synchronized and grown for 30 min without drug (B) and an additional 120 min with either no drug (C) or 10 ?g/ml A22 (D). Scale bars
represent 1 ?m.
(E) Percentage of cells that are doubly labeled at the origin and CC2943, CC3300, CC3656, or CC0091 with two distinct foci at different times
after synchrony, growth for 30 min without drug, and subsequent growth in 10 mg/ml A22. At least 75 cells were analyzed for each time point,
with standard deviations of 3%–5%. Below the graph is a schematic illustration of what is observed without A22 (top line) and with A22
(bottom line). The green dots represent the origin and the red dots represent C2943.
(F) Graphs depicting the average cellular position (distance from the stalked pole) at 2 min time intervals of loci labeled with lacO at CC0768,
862 kb to the right of the origin (top) and CC3192, 576 kb to the left of the origin. Loci from cells incubated without A22 after origin segregation
are shown in brown, and loci from cells incubated with 10 ?g/ml A22 after origin segregation are shown in blue. Position and rate analysis
was performed as previously described (Viollier et al., 2004).
consistently amplified from lysates precipitated with
anti-CtrA antibodies, while three loci known to not in-
teract with CtrA (PdnaA, CC1807, CC3752, and PCC0048)
never amplified (Figure 7). Moreover, no loci whatsoever
were amplified when the ChIP experiments were re-
peated with MreB preimmune serum or antibodies to
FliF, a flagellar motor known to not bind DNA (Jenal and
Shapiro, 1996) (Figure 7).
Thus, the ChIP results indicate that MreB exhibits
specificity of interaction with the chromosome and rec-
ognizes, at minimum, the origin of replication and se-
quences 8 kb and 12 kb to the left of the origin. Al-
though these experiments do not differentiate between
direct binding of MreB to DNA or to a DNA binding pro-
tein complex, MreB’s association with an origin-proxi-
mal region is consistent with MreB playing a direct role
in the movement of the origin region. MreB’s lack of
association with origin-distal regions also supports our
A22 findings, suggesting that MreB directs the move-
ment of the origin region, but that the rest of the chro-
mosome segregates through an MreB-independent
Eukaryotic cells have long been known to use cytoskel-
etal proteins for a wide variety of cellular activities in-
cluding cell shape determination, division, polarization,
protein localization, and chromosome segregation. The
identification of bacterial cytoskeletal proteins has
raised the possibility that these proteins carry out sim-
ilar functions in bacteria. What is known about the bac-
terial cytoskeleton is consistent with this notion, with
the MreB actin homolog representing a particularly
attractive candidate for playing a role in chromosome
segregation. MreB perturbations lead to defective num-
bers and positioning of origin sequences (Gitai et al.,
2004; Kruse et al., 2003; Soufo and Graumann, 2003).
However, these defects take several cell cycles to man-
ifest and could be an indirect result of other MreB func-
tions or general cellular deterioration. The dissection of
acute activity is required to understand the direct ac-
tions of the multifunctional MreB actin. For example,
the use of rapidly acting drugs such as cytochalasins
and nocodazole has revealed the many roles of actin
Figure 7. MreB Binds a Specific Chromosomal Region near the Origin
ChIPs were performed using anti-MreB, MreB preimmune, anti-CtrA, and anti-FliF antisera. The origin, PdnaA, CC0048, CC1035, CC1807,
PctrA, PcreS, CC3752, and PgidA chromosomal loci were then PCR amplified.
(A) Schematic of the genetic positions of the amplified loci. The underlined loci are known CtrA binding targets. Red arrows point to loci that
(B) Ethidiumbromide-stained agarose gel electrophoresis of the PCR reactions for each amplified locus from each ChIP. The distance of each
locus from the origin is shown, color coded to correspond to the schematic in (A).
and tubulin in eukaryotic cells (Fenteany and Zhu,
2003). We thus pursued a pharmacological approach to
quickly and specifically perturb MreB and assess its
direct involvement in chromosome segregation.
MreB helical polymers upon A22 treatment. The resi-
dues mutated in the A22-resistant strains are generally
not conserved with eukaryotic actin. Consistent with
this observation, A22 does not appear to perturb
eukaryotic actin polymerization (data not shown).
A22 Is a New Tool for Studying MreB Function
A small molecule, A22, discovered in a screen for anu-
cleate E. coli cells and shown to convert rod-shaped
E. coli to round cells (Iwai et al., 2002) mimics MreB
depletion and rapidly and reversibly delocalizes MreB
in Caulobacter. We have shown that seven missense
mutations in mreB are each sufficient to render cells
completely insensitive to A22, and no mutations in any
other genes were found to cause resistance to this
drug. Thus, A22 is highly specific for MreB. The combi-
nation of A22’s high specificity and rapid perturbation
of MreB localization make it particularly useful for the
temporal analysis of MreB function.
In addition to demonstrating specificity, the A22-
resistant mreB alleles shed light on how A22 might in-
teract with MreB. The mutated residues are dispersed
throughout the protein, but when mapped onto the
crystal structure of a close homolog (van den Ent et al.,
2001), they all cluster near MreB’s nucleotide binding
site. The actin inhibitor latrunculin binds the active site
of eukaryotic actin and displaces ATP, preventing
monomers from polymerizing (Morton et al., 2000), per-
haps indicating a parallel with the loss of Caulobacter
The MreB Actin Directs the Segregation
of Origin-Proximal, but Not Origin-Distal Loci
As Caulobacter transition from swarmer cells to stalked
cells, replication initiates, one of the replicated origins
rapidly moves across the cell to the opposite pole, and
other loci are sequentially replicated and segregated to
their proper cellular destinations (Viollier et al., 2004).
When A22 is added to swarmer cells in which the origin
and an origin-distal locus are tagged with fluorescent
markers, neither locus separates into discernable foci.
Since A22 does not block replication initiation and a full
round of replication occurs in the presence of A22, we
conclude that these loci are duplicated, but that the
replicated loci do not move apart. As a result, the fluo-
rescence foci from each duplicated locus cannot be re-
solved, causing them to appear as a single dot. MreB
thus appears to be essential for Caulobacter origin seg-
regation. MreB-depleted E. coli cells also segregate
their chromosomes in cohesive pairs (Kruse et al., 2003),
suggesting that this may be a general mechanism for
bacterial chromosome segregation.
The observation that loci can be replicated without
Cytoskeletal Segregation of a Bacterial Origin
being segregated demonstrates that replication and
segregation can be uncoupled. This uncoupling is con-
sistent with the findings from E. coli in which migS was
identified as a locus responsible for segregation but not
replication (Yamaichi and Niki, 2004). In addition, alter-
ing the order of chromosomal replication in E. coli by
debilitating the normal origin and inserting an exoge-
nous origin of replication at another position in the
chromosome does not alter the order of chromosomal
segregation (Gordon et al., 2002). This has significant
implications for the prevailing “extrusion-capture”
model of chromosome segregation (Lemon and Gross-
man, 2001). While the force generated by extruding
DNA from the replisome may still contribute to chromo-
some dynamics, it alone is clearly insufficient to drive
chromosomes apart, and a replication-independent
segregation machinery must exist.
In contrast to the origin, loci in origin-distal regions
of the chromosome are able to segregate in the pres-
ence of A22, so long as A22 is administered after the
origins are replicated and moved to the cell poles. Not
only are the origin-distal loci separated in the presence
of A22, but A22 has no effect on their cellular position
or the kinetics of their movement. These results sug-
gest a two-step mechanism for Caulobacter chromo-
some segregation: first an MreB-dependent polar seg-
regation of the origin region occurs, and then the rest
of the chromosome segregates independently of MreB.
Since treatment of cells with A22 prior to the movement
of the origin to the cell poles blocks the movement of
loci throughout the chromosome, it appears that the
second step can only occur if the first is completed.
This would be consistent with a “harpoon” model for
Caulobacter chromosome segregation wherein origin
movement to its proper location is mediated by MreB,
followed by the rest of the chromosome being pulled
along behind the origin. The pulling force could come
from DNA replication or DNA compaction proteins such
as SMC and the histone-like HU, potentially explaining
the chromosome partitioning defects observed in these
mutants (Britton et al., 1998; Jaffe et al., 1997; Jensen
and Shapiro, 1999).
this region (a possible bacterial kinetochore), and de-
termine the mechanism by which MreB facilitates origin
movement and subsequent chromosome segregation.
At the time of Caulobacter origin segregation, MreB is
assembled in a spiral that corkscrews along the length
of the cell (Figge et al., 2004; Gitai et al., 2004). This
MreB spiral could serve as a track for a motor protein
to pull one origin to the other pole. Alternatively, the
dynamics of MreB monomers within a polarized MreB
filament (Defeu Soufo and Graumann, 2004) could be
harnessed for origin movement.
Understanding the link between MreB and the
Caulobacter chromosome should prove critical to un-
derstanding chromosome segregation. The essential
Caulobacter ParA and ParB proteins may be candi-
dates for this link: parA and parB mutants exhibit strong
chromosome segregation defects (Mohl and Gober,
1997), and ParB and MreB bind similar regions of the
Caulobacter chromosome, slightly to the left of the ori-
gin (Mohl et al., 2001). However, the chromosome seg-
regation defects of the Caulobacter par mutants have
been largely attributed to their triggering of a cell cycle
checkpoint (Easter and Gober, 2002; Mohl et al., 2001),
Par protein homologs are not found in E. coli (Gerdes
et al., 2000), and B. subtilis soj (parA) and spo0J (parB)
mutants are viable and have very subtle chromosome
defects (Lin and Grossman, 1998).
In addition to identifying the putative kinetochore
complex, many other questions must be addressed, in-
cluding the role of MreB in chromosome segregation in
other bacteria, and the mechanism by which specific
loci are delivered to reproducible cellular locations. Our
work provides a framework for directing these future
studies by demonstrating that MreB physically associ-
ates with an origin-proximal region of the chromosome
and is required to direct origin segregation, which in
turn mediates the movement of the rest of the chro-
Bacterial Strains and Growth Conditions
Caulobacter crescentus strains CB15N and derivatives were grown
in PYE or M2G medium or agar plates supplemented with the ap-
propriate combinations of antibiotics, glucose, xylose, A22, and the
A22 diluent, methanol (Ely, 1991). The FROS strains labeled at both
the origin and CC2943, CC3300, CC3656, and CC0091, or singly
labeled at CC0786, and CC3192 were generated and characterized
in a previous study, where they were referred to as MPO40, MPO53,
MPO7, MPO133, MP156, and MP82, respectively (Viollier et al.,
2004). Growth and analysis conditions for the mreB depletion (Gitai
et al., 2004), holB:gfp (Jensen et al., 2001), and FROS strains (Viol-
lier et al., 2004) have been previously described. The methods for
fCr30-mediated generalized phage transduction (Ely, 1991) and
swarmer cell isolation by Percoll density gradient centrifugation (Al-
ley, 2001) have also been described. Washes were performed by
micro centrifugation at 8000 g.
Twenty independent A22-resistant Caulobacter mutants were
isolated by growing 20 single CB15N colonies overnight and plat-
ing 109cells per culture on PYE plates containing 10 ?g/ml A22.
One A22-resistant colony from each plate was isolated, retested,
and characterized. To sequence the mreB gene from these strains,
mreB was PCR-amplified and the resulting band was gel-purified
(Qiagen) and sequenced. Overlay PCR mutagenesis was used to
introduce the T158A point mutation (A509G) into the previously de-
scribed Pxyl:gfp:mreB construct (Gitai et al., 2004).
A Putative Bacterial Centromere
The A22 experiments demonstrate that the movement
of the origin depends on MreB but do not address
whether MreB acts on the chromosome itself or acts
indirectly by binding to other proteins required for chro-
mosome segregation. ChIP assays demonstrated that
MreB biochemically associates either directly or indi-
rectly with chromosomal DNA, and that this association
exhibits specificity. The origin and loci −8 and −12 kb
away from the origin bound MreB, while sites at −64
kb, +50 kb, and farther from the origin did not. The fact
that the origin-proximal region both depends on MreB
for segregation and binds MreB suggests that this re-
gion could bind to MreB to mediate its segregation,
which in turn promotes the segregation of the rest of
the chromosome. In this manner, the origin-proximal
DNA would function as a bacterial centromere.
The challenges now are to finely delineate this puta-
tive centromere region in the Caulobacter chromo-
some, define the protein complex that interacts with
Microscopy and Timelapse Imaging
All experiments other than the A22 flow-cell analysis were per-
formed by mounting the samples on 1% agarose pads as de-
scribed (Ryan et al., 2002), with the appropriate concentration of
A22 added to the agarose. The rates of origin-distal locus segrega-
tion were determined by inducing Pxyl:lacI:cfp expression with
0.03% xylose for 75 min, isolating swarmer cells, and growing the
swarmer cells in M2G liquid medium for 50 min. After 50 min, 10
?g/ml of A22 was added to these liquid cultures, they were
mounted onto 1% M2G agarose pads containing 10 ?g/ml A22,
and a phase and CFP fluorescence image was captured every 2
min for 60 min. The images were analyzed by previously described
software (Viollier et al., 2004) to determine the average position and
rate of movement of the observed foci.
The A22 flow-cell treatment timelapse images were performed
on Pxyl:gfp-mreB cells induced for 2 hr with 0.03% xylose. These
cells were adhered to polylysine-coated cover slips. The cover
slips were then adhered to microscope slides with two parallel
strips of double-sided tape (3M), leaving a central channel. An ex-
cess of medium containing 0 or 10 ?g/ml A22 was pipetted onto
one opening of this channel and wicked through by placing a tissue
(VWR) on the other end. Images were collected every 30 s.
Defeu Soufo, H.J., and Graumann, P.L. (2004). Dynamic movement
of actin-like proteins within bacterial cells. EMBO Rep. 5, 789–794.
Dworkin, J., and Losick, R. (2002). Does RNA polymerase help drive
chromosome segregation in bacteria? Proc. Natl. Acad. Sci. USA
Easter, J., Jr., and Gober, J.W. (2002). ParB-stimulated nucleotide
exchange regulates a switch in functionally distinct ParA activities.
Mol. Cell 10, 427–434.
Ely, B. (1991). Genetics of Caulobacter crescentus. Methods Enzy-
mol. 204, 372–384.
Fenteany, G., and Zhu, S. (2003). Small-molecule inhibitors of actin
dynamics and cell motility. Curr. Top. Med. Chem. 3, 593–616.
Figge, R.M., Divakaruni, A.V., and Gober, J.W. (2004). MreB, the cell
shape-determining bacterial actin homologue, co-ordinates cell
wall morphogenesis in Caulobacter crescentus. Mol. Microbiol. 51,
Gerdes, K., Moller-Jensen, J., and Bugge Jensen, R. (2000). Plas-
mid and chromosome partitioning: surprises from phylogeny. Mol.
Microbiol. 37, 455–466.
Gitai, Z., Dye, N., and Shapiro, L. (2004). An actin-like gene can
determine cell polarity in bacteria. Proc. Natl. Acad. Sci. USA 101,
Gordon, G.S., Shivers, R.P., and Wright, A. (2002). Polar localization
of the Escherichia coli oriC region is independent of the site of
replication initiation. Mol. Microbiol. 44, 501–507.
Holtzendorff, J., Hung, D., Brende, P., Reisenauer, A., Viollier, P.H.,
McAdams, H.H., and Shapiro, L. (2004). Oscillating global regula-
tors control the genetic circuit driving a bacterial cell cycle. Science
Iwai, N., Nagai, K., and Wachi, M. (2002). Novel S-benzylisothiourea
compound that induces spherical cells in Escherichia coli probably
by acting on a rod-shape-determining protein(s) other than penicil-
lin-binding protein 2. Biosci. Biotechnol. Biochem. 66, 2658–2662.
Jacob, F., Brenner, S., and Cuzin, F. (1963). On the regulation of
DNA replication in bacteria. Cold Spring Harb. Symp. Quant. Biol.
Jaffe, A., Vinella, D., and D’Ari, R. (1997). The Escherichia coli his-
tone-like protein HU affects DNA initiation, chromosome partition-
ing via MukB, and cell division via MinCDE. J. Bacteriol. 179,
Jenal, U., and Shapiro, L. (1996). Cell cycle-controlled proteolysis
of a flagellar motor protein that is asymmetrically distributed in the
Caulobacter predivisional cell. EMBO J. 15, 2393–2406.
Jensen, R.B., and Shapiro, L. (1999). The Caulobacter crescentus
smc gene is required for cell cycle progression and chromosome
segregation. Proc. Natl. Acad. Sci. USA 96, 10661–10666.
Jensen, R.B., Wang, S.C., and Shapiro, L. (2001). A moving DNA
replication factory in Caulobacter crescentus. EMBO J. 20, 4952–
Jensen, R.B., Wang, S.C., and Shapiro, L. (2002). Dynamic localiza-
tion of proteins and DNA during a bacterial cell cycle. Nat. Rev.
Mol. Cell Biol. 3, 167–176.
Jones, L.J., Carballido-Lopez, R., and Errington, J. (2001). Control
of cell shape in bacteria: helical, actin-like filaments in Bacillus sub-
tilis. Cell 104, 913–922.
Kline-Smith, S.L., and Walczak, C.E. (2004). Mitotic spindle assem-
bly and chromosome segregation: refocusing on microtubule dy-
namics. Mol. Cell 15, 317–327.
Kruse, T., Moller-Jensen, J., Lobner-Olesen, A., and Gerdes, K.
(2003). Dysfunctional MreB inhibits chromosome segregation in
Escherichia coli. EMBO J. 22, 5283–5292.
Lau, I.F., Filipe, S.R., Soballe, B., Okstad, O.A., Barre, F.X., and
Sherratt, D.J. (2003). Spatial and temporal organization of replicat-
ing Escherichia coli chromosomes. Mol. Microbiol. 49, 731–743.
Laub, M.T., Chen, S.L., Shapiro, L., and McAdams, H.H. (2002).
Genes directly controlled by CtrA, a master regulator of the
Caulobacter cell cycle. Proc. Natl. Acad. Sci. USA 99, 4632–4637.
Lemon, K.P., and Grossman, A.D. (2001). The extrusion-capture
Analysis of DNA Replication
DNA replication was assessed by incorporation of
dGTP (Marczynski et al., 1990) or chromomycin-stained DNA-content
FACS analysis (Winzeler and Shapiro, 1995) as described pre-
Formaldehyde-crosslinked chromatin immunoprecipitation was
performed as described previously (Holtzendorff et al., 2004; Laub
et al., 2002; Lin and Grossman, 1998). The anti-MreB antibody was
raised by cloning the mreB coding sequence into the pET28a vec-
tor (Novagen). The resulting His6-tagged MreB was overexpressed
in BL21(DE3)/pLysS E. coli and purified using Qiagen Ni2+-NTA
agarose as described previously (Viollier et al., 2002). Two milli-
grams of purified His6-MreB was used to immunize two rabbits
(Josman Laboratories, Napa, California). Oligonucleotide primer
sequences for the regions PCR amplified are available upon re-
We would like to thank Tom Wandless (Stanford University) for syn-
thesizing additional A22 compound based on its published struc-
ture (Iwai et al., 2002), Patrick McGrath for assistance with the
quantitative analysis of locus movement and positioning, and
members of the Shapiro and McAdams labs and Coleen Murphy
for critical discussions. Z.G. is supported by a Ruth Kirschstein NIH
postdoctoral fellowship. This work was supported by NIH grant
Received: November 5, 2004
Revised: December 22, 2004
Accepted: January 6, 2005
Published: February 10, 2005
Alley, M.R. (2001). The highly conserved domain of the Caulobacter
McpA chemoreceptor is required for its polar localization. Mol.
Microbiol. 40, 1335–1343.
Blondelet-Rouault, M.H., Weiser, J., Lebrihi, A., Branny, P., and Per-
nodet, J.L. (1997). Antibiotic resistance gene cassettes derived
from the omega interposon for use in E. coli and Streptomyces.
Gene 190, 315–317.
Britton, R.A., Lin, D.C., and Grossman, A.D. (1998). Characterization
of a prokaryotic SMC protein involved in chromosome partitioning.
Genes Dev. 12, 1254–1259.
Cytoskeletal Segregation of a Bacterial Origin
model for chromosome partitioning in bacteria. Genes Dev. 15,
Li, Y., Sergueev, K., and Austin, S. (2002). The segregation of the
Escherichia coli origin and terminus of replication. Mol. Microbiol.
Lin, D.C., and Grossman, A.D. (1998). Identification and character-
ization of a bacterial chromosome partitioning site. Cell 92, 675–
Lowe, J., and Amos, L.A. (1998). Crystal structure of the bacterial
cell-division protein FtsZ. Nature 391, 203–206.
Lutkenhaus, J., and Addinall, S.G. (1997). Bacterial cell division and
the Z ring. Annu. Rev. Biochem. 66, 93–116.
Marczynski, G.T., and Shapiro, L. (1992). Cell-cycle control of a
cloned chromosomal origin of replication from Caulobacter cres-
centus. J. Mol. Biol. 226, 959–977.
Marczynski, G.T., and Shapiro, L. (2002). Control of chromosome
replication in Caulobacter crescentus. Annu. Rev. Microbiol. 56,
Marczynski, G.T., Dingwall, A., and Shapiro, L. (1990). Plasmid and
Caulobacter crescentus cell cycle. J. Mol. Biol. 212, 709–722.
Mohl, D.A., and Gober, J.W. (1997). Cell cycle-dependent polar lo-
calization of chromosome partitioning proteins in Caulobacter
crescentus. Cell 88, 675–684.
Mohl, D.A., Easter, J., Jr., and Gober, J.W. (2001). The chromosome
partitioning protein, ParB, is required for cytokinesis in Caulobacter
crescentus. Mol. Microbiol. 42, 741–755.
Moller-Jensen, J., Jensen, R.B., Lowe, J., and Gerdes, K. (2002).
Prokaryotic DNA segregation by an actin-like filament. EMBO J. 21,
Moller-Jensen, J., Borch, J., Dam, M., Jensen, R.B., Roepstorff, P.,
and Gerdes, K. (2003). Bacterial mitosis: ParM of plasmid R1 moves
plasmid DNA by an actin-like insertional polymerization mecha-
nism. Mol. Cell 12, 1477–1487.
Morton, W.M., Ayscough, K.R., and McLaughlin, P.J. (2000). Latrun-
culin alters the actin-monomer subunit interface to prevent poly-
merization. Nat. Cell Biol. 2, 376–378.
Niki, H., Yamaichi, Y., and Hiraga, S. (2000). Dynamic organization
of chromosomal DNA in Escherichia coli. Genes Dev. 14, 212–223.
Quardokus, E., Din, N., and Brun, Y.V. (1996). Cell cycle regulation
and cell type-specific localization of the FtsZ division initiation pro-
tein in Caulobacter. Proc. Natl. Acad. Sci. USA 93, 6314–6319.
Ryan, K.R., Judd, E.M., and Shapiro, L. (2002). The CtrA response
regulator essential for Caulobacter crescentus cell-cycle pro-
gression requires a bipartite degradation signal for temporally con-
trolled proteolysis. J. Mol. Biol. 324, 443–455.
Shih, Y.L., Le, T., and Rothfield, L. (2003). Division site selection in
Escherichia coli involves dynamic redistribution of Min proteins
within coiled structures that extend between the two cell poles.
Proc. Natl. Acad. Sci. USA 100, 7865–7870.
Soufo, H.J., and Graumann, P.L. (2003). Actin-like proteins MreB
and Mbl from Bacillus subtilis are required for bipolar positioning
of replication origins. Curr. Biol. 13, 1916–1920.
Teleman, A.A., Graumann, P.L., Lin, D.C., Grossman, A.D., and Los-
ick, R. (1998). Chromosome arrangement within a bacterium. Curr.
Biol. 8, 1102–1109.
van den Ent, F., Amos, L.A., and Lowe, J. (2001). Prokaryotic origin
of the actin cytoskeleton. Nature 413, 39–44.
Viollier, P.H., Sternheim, N., and Shapiro, L. (2002). A dynamically
localized histidine kinase controls the asymmetric distribution of
polar pili proteins. EMBO J. 21, 4420–4428.
Viollier, P.H., Thanbichler, M., McGrath, P.T., West, L., Meewan, M.,
McAdams, H.H., and Shapiro, L. (2004). Rapid and sequential
movement of individual chromosomal loci to specific subcellular
locations during bacterial DNA replication. Proc. Natl. Acad. Sci.
USA 101, 9257–9262.
Webb, C.D., Graumann, P.L., Kahana, J.A., Teleman, A.A., Silver,
P.A., and Losick, R. (1998). Use of time-lapse microscopy to visual-
ize rapid movement of the replication origin region of the chromo-
some during the cell cycle in Bacillus subtilis. Mol. Microbiol. 28,
Webb, C.D., Teleman, A., Gordon, S., Straight, A., Belmont, A., Lin,
D.C., Grossman, A.D., Wright, A., and Losick, R. (1997). Bipolar lo-
calization of the replication origin regions of chromosomes in vege-
tative and sporulating cells of B. subtilis. Cell 88, 667–674.
Winzeler, E., and Shapiro, L. (1995). Use of flow cytometry to iden-
tify a Caulobacter 4.5 S RNA temperature-sensitive mutant de-
fective in the cell cycle. J. Mol. Biol. 251, 346–365.
Yamaichi, Y., and Niki, H. (2004). migS, a cis-acting site that affects
bipolar positioning of oriC on the Escherichia coli chromosome.
EMBO J. 23, 221–233.