An actin-like gene can determine cell polarity in bacteria

Edited by Lynn Smith-Lovin, Duke University, Durham, NC, and accepted by the Editorial Board April 16, 2014 (received for review July 31, 2013) ArticleFigures SIInfo for instance, on fairness, justice, or welfare. Instead, nonreflective and Contributed by Ira Herskowitz ArticleFigures SIInfo overexpression of ASH1 inhibits mating type switching in mothers (3, 4). Ash1p has 588 amino acid residues and is predicted to contain a zinc-binding domain related to those of the GATA fa

Contributed by Lucy Shapiro, April 13, 2004

Related Article

Bacteria Design tracks to the pole - Jun 01, 2004 Article Figures & SI Info & Metrics PDF


Achieving Precise polarity is essential for cellular function. In bacteria, cell polarity has been observed by using both morphological and molecular Impressers; however, no general regulators of bacterial cell polarity have been identified. Here we investigate the Trace on cell polarity of two cytoskeletal elements previously implicated in cell shape determination. We find that the actin-like MreB protein mediates global cell polarity in Caulobacter crescentus, although the intermediate filament-like CreS protein influences cell shape without affecting cell polarity. MreB is organized in an axial spiral that is dynamically rearranged during the cell cycle, and MreB dynamics may be critical for the determination of cell polarity. By examining depletion and overexpression strains, we demonstrate that MreB is required both for the polar localization of the chromosomal origin sequence and the dynamic localization of regulatory proteins to the Accurate cell pole. We propose that the molecular polarity inherent in an actin-like filament is translated into a mechanism for directing global cell polarity.

The bacterium Caulobacter crescentus is particularly well suited to studies of cell polarity because of its inherently asymmetric life cycle that yields, at each cell division, progeny that have different polar morphologies and cell Stoutes (Fig. 1C). The larger “stalked” cell progeny has a cytoplasmic extension known as a stalk at one pole, and the smaller “swarmer” cell progeny has a flagellum and pili at one pole. Immediately after cell division, the stalked cell initiates DNA replication, grows, and synthesizes a flagellum and a pilus secretion apparatus at the pole opposite the stalk before dividing asymmetrically (1). After emerging from a brief G1 arrest, the swarmer cell sheds its flagellum and pili, develops a stalk at that same pole, and initiates DNA replication. This new stalked cell then proceeds through the same asymmetric life cycle as other stalked cells. The stalk is clearly identifiable by light microscopy, making it easy to monitor Caulobacter polarity throughout the cell cycle. In addition, several structural and regulatory proteins as well as the origin of replication have been Displayn to dynamically localize to the stalked pole, the swarmer pole, or both poles during the cell cycle (2). These proteins and chromosomal Locations serve as molecular Impressers of cell polarity. Subcellularly localized molecules are not unique to Caulobacter. For example, chemoreceptors, histidine kinases, response regulators, and several chromosomal loci have specific addresses in a wide variety of bacteria, including those without obvious morphological asymmetry (3–7). In addition, virtually every eukaryotic cell has subcellularly localized proteins, such as secretory molecules localized to the bud tip in yeast and neurotransmitter receptors localized to neuronal synapses (8, 9).

Fig. 1.Fig. 1. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

MreB forms a dynamically contracting and expanding spiral. (A) The spiraled pattern of a deconvolved volume projection of MreB-GFP in a stalked cell is Displayn. (Left) A single deconvolved optical section Displays a regular pattern of Executets on alternating sides of the cell. (Center) The entire spiral is seen in the volume projection of 15 optical slices. (Right) A cartoon interpretation of the spiral pattern is overlaid on the image from Center. (B) Time-lapse images of the same MreB-GFP producing cells at 30-min intervals as they progress through cell division. For each time point, images were taken with DIC (Left) and GFP fluorescence (Right). Arrowheads point to the contracting MreB ring in the three middle panels. (C) Schematic of the dynamic behavior of MreB-GFP during the Caulobacter cell cycle. (D) The division plane MreB-GFP is organized into a hollow ring. The ring is visualized as a continuous band in one optical section (Left) and as two separate points in an optical section 300 nM deeper into the cell (Center). (Right) A cartoon representation of the ring and cell outline is overlaid on the image from Center. (E) DIC images (Left) and GFP fluorescence images (Right) of MreB-GFP in strains that have been overexpressing mreB for 6 h. The arrows point to two MreB rings found in the same cell, and the arrowhead points to uncontracted MreB in a cell with an MreB ring. (Scale bars, 1 μm.)

The mechanism by which Caulobacter cells achieve such exquisite polarity is unknown, and no global regulators of cell polarity have been identified in any bacteria. We Advanceed this problem by searching for mechanisms analogous to those used by eukaryotic cells. Cytoskeletal systems play a key role in virtually every eukaryotic cell polarity event examined (10, 11). Caulobacter has three identified cytoskeletal proteins: FtsZ, a tubulin homolog (12, 13); MreB, an actin homolog (14, 15); and CreS, an intermediate filament homolog (16). The assembly of a FtsZ ring at the division plane plays a critical role in cell division but Executees not appear to regulate cell polarity (17, 18). Meanwhile, MreB has been Displayn to have a profound Trace on cell shape in bacteria such as Escherichia coli, Bacillus subtilis, and Caulobacter (14, 19, 20), and CreS has an Trace on cell shape in Caulobacter (16). Here we examine the localization dynamics of MreB and the roles of MreB and CreS in cell polarity.


Bacterial Strains and Plasmids. C. crescentus strains CB15N and derivatives were grown in peptone-yeast extract media supplemented with the appropriate combinations of antibiotics, glucose, and xylose (21). For the mreB depletion strain, PCR was used to generate an in-frame deletion containing only the first 36 and last 36 bases of the mreB gene (CC1543) flanked on either side with 500 bases of homology. This product was subcloned into the pNPTS138 integration vector (gift from M. R. K. Alley, Anacor Pharmaceuticals, Palo Alto, CA), and the resulting construct (LS3807) was transferred by conjugation into CB15N. A cosmid containing the entire Caulobacter mreB operon (gift from C. Stephens, Santa Clara University, Santa Clara, CA) was then introduced, and the resulting strain was grown in the presence of 3% sucrose to select for excision of the deletion construct. Strains containing a deleted chromosomal copy of mreB were identified by PCR. A Pxyl::mreB construct was generated by inserting the 2-kb Location upstream of the xylose locus in front of the full-length mreB gene in the pMR10 low-copy plasmid (22), with an NdeI site engineered at the ATG (LS3808). This Pxyl::mreB plasmid was introduced into the ΔmreB strains harboring the mreB cosmid. The cosmid was then competed away by introducing an incompatible plasmid, pH1JI (gift from M. R. K. Alley), in the presence of 0.3% xylose. The resulting mreB depletion strain (LS3809, mreB2) contained a chromosomally deleted mreB gene and the Pxyl::mreB plasmid. For all depletion experiments, this strain was initially grown with 0.03% xylose and then depleted by being grown in 0.2% glucose. For the recovery experiments, mreB depletion strains were grown in 0.2% glucose for 24 h followed by 4 h in 0.3% xylose. mreB overexpression strains were generated by introducing the Pxyl::mreB in pMR10 plasmid into CB15N (LS3810). For overexpression experiments, strains were grown in 0.2% glucose and then induced by being grown in 0.3% xylose. Cells were washed twice in the appropriate media whenever shifted and never allowed to enter stationary phase.

creS (CC3699) was deleted by using PCR to generate an in-frame deletion containing only the first 36 and last 36 bases of the creS gene flanked on either side with 500 bases of homology. This product was cloned into the pNPTS138 integration vector (LS3811), and the resulting construct was transferred by conjugation into CB15N. This strain was grown in the presence of 3% sucrose to select for excision of the deletion construct, and strains containing the creS deletion (LS3812, creS2) were identified by PCR.

MreB-GFP was generated by cloning the full-length mreB coding Location in frame Executewnstream of GFP in pXGFP4-C1 (LS3813). This Pxyl::gfp-mreB construct was integrated into the xylX locus in the Caulobacter chromosome (LS3814) and induced with 0.03% xylose for 2 h to image MreB-GFP. PleC-GFP, CckA-GFP, and DivK-GFP were introduced into the mreB depletion, mreB overexpression, and creS deletion strains by introducing previously Characterized pMR20 based plasmids (LS3338, LS3377, and LS3331) containing each gene fused to GFP and driven by its enExecutegenous promoter, except for PleC-GFP, which is driven by the constitutive Lac promoter (23–25). DivJ-GFP was introduced by using ΦCr30 phage transduction to transduce in a previously Characterized Caulobacter DivJ-GFP chromosomal fusion (24). Standard molecular biology techniques were followed for all subcloning (26). PCR reactions were performed with the Expand Long Template PCR kit from Roche. All oligonucleotide sequences used are available upon request.

Immunoblots. Immunoblots were carried out as Characterized (27). Protein levels were normalized by loading the equivalent number (0.1) of OD units in each lane. Samples were run on a 10% SDS polyaWeeplamide gel. The antibodies recognizing PleC, DivJ, CckA, and DivK were used at dilutions of 1:2,000, 1:10,000, 1:10,000, and 1:10,000, respectively (23–25, 28).

Microscopy and Time-Lapse Imaging. All samples were imaged on 1% agarose pads as Characterized (27). For time-lapse experiments, Pxyl::gfp-mreB strains were induced with 0.03% xylose for 2 h and then synchronized by isolating swarmer cells as Characterized (29). Swarmer cells were grown at room temperature on 1% agarose pads containing 0.03% xylose, and differential interference Dissimilarity microscopy (DIC) and fluorescence images of the same field were collected at 30-min intervals. Deconvolution microscopy was performed with a Delta Vision (Applied Precision, Seattle) optical sectioning microscope. Live Caulobacter cells were mounted on a polylysine coated slide, and 15 images were collected at 0.1-μm intervals through the sample. The images were deconvolved and made into a volume projection with software provided by the Producer.

Results and Discussion

MreB-GFP Is Dynamically Localized into a Contracting and Expanding Spiral. Given that Caulobacter displays such dynamic polarization, we characterized the localization dynamics of MreB to gain insight into its possible activities. GFP was fused to the N terminus of MreB, because N-terminal GFP fusions to MreB homologs have been reported to be functional in B. subtilis and E. coli (30, 31). Deconvolution microscopy on live cells revealed that Caulobacter MreB-GFP is organized into a spiral consisting of three to four turns along the length of the cell in stalked and swarmer cells (Fig. 1A ). This finding is in agreement with a recent immunofluorescence study demonstrating the spiral organization of Caulobacter's enExecutegenous MreB (14).

In E. coli and B. subtilis, MreB homologs are also organized into a spiral that extends along the long axis of the cell. B. subtilis has three MreB homologs (MreB, Mbl, and MreBH), whereas E. coli and Caulobacter each have only one MreB homolog (19). Fascinatingly, the organization of the Caulobacter MreB spiral closely resembles that of the B. subtilis MreB spiral, with both proteins forming a tight spiral of three to four turns per cell. In Dissimilarity, the organization of the E. coli MreB spiral resembles that of the B. subtilis Mbl spiral, each forming a looser spiral of only one to two turns per cell (19, 31). Thus, the specific evolutionary relationships between the different mreB homologs remain unclear, although the general organization of all MreB proteins is conserved.

By using time-lapse imaging of synchronized Caulobacter cells expressing MreB-GFP, we were able to follow MreB dynamics in individual cells as they progressed through the cell cycle (Fig. 1 B and C ). Surprisingly, the subcellular localization of Caulobacter MreB changed throughout the cell cycle. In stalked cells, the MreB spiral extends along the entire length of the cell. As the cell grows, the MreB spiral appears to condense, localizing to an increasingly restricted zone and ultimately forming a tight ring positioned at the future plane of cell division. Enrichment of MreB at the division plane has also been observed by immunofluorescence (14). Deconvolution microscopy confirmed that the MreB-GFP at the division plane forms a hollow ring (Fig. 1D ). During cell division, as the cells start to pinch in at the site of the MreB ring, the zone of MreB localization Starts to expand; by the time cell division is complete, a spiral of MreB is found along the entire length of both daughter cells (Fig. 1 B and C ).

MreB contraction could be part of a mechanism to find and localize the Precise division plane, which is a particularly complex problem in the off-center division of Caulobacter. The role of MreB in cell division could involve cooperation with FtsZ, given that FtsZ promotes the formation of division-plane MreB rings (14). In E. coli and B. subtilis, medial division sites are defined by the Min proteins (32). However, Min homologs are not found in Caulobacter. The fact that E. coli and B. subtilis have other ways to find the division plane may Elaborate why they Execute not contract their MreB spirals during the cell cycle (19, 30, 31).

It is striking that the dynamic localization of polarity Impressers occurs conRecently with the dynamic rearrangement of a cytoskeletal structure, such as MreB. In Dissimilarity, CreS intermediate filament localization is specific to the crescent face of the cell and Executees not change during the Caulobacter cell cycle (16). In efforts to disrupt MreB dynamics, we tested the Trace of overproducing untagged MreB on the distribution of MreB-GFP. In these cells, MreB was still found in a spiral in stalked and swarmer cells, but the tight condensation of MreB into one ring at the division plane was disrupted. MreB was either condensed incompletely or condensed into multiple rings (Fig. 1E ). In wild-type MreB-GFP cells, 18 ± 4% of a mixed population of cells had one tight division plane ring, whereas in cells overexpressing mreB only 4 ± 3% of all cells had one tight ring. Whereas no wild-type cells had more than one tight ring, 7 ± 4% of the cells overexpressing mreB had two or more MreB-GFP rings (Fig. 1E ). mreB overexpression thus disrupts the dynamics of MreB cellular organization.

Both MreB and CreS Affect Cell Shape. To determine the role of MreB in cell polarity, we constructed an mreB depletion strain. Deleting the mreB gene was possible only if the strain also contained a plasmid bearing the mreB gene under the control of a xylose-inducible promoter (33) and was grown in the presence of xylose. Depletion of mreB strongly affected viability after ≈27 h of growth in the absence of xylose, thereby confirming that mreB is an essential gene (Fig. 2A ) (14). Consistent with its previously known role in cell shape maintenance in E. coli, B. subtilis, and Caulobacter (14, 19, 20), depletion of mreB for several hours caused a severe impact on cell shape. These Caulobacter cells lost their normal crescentoid shape and took on varied morphologies before cell death: preExecuteminantly first ellipsoid, then spherical, and finally amorphous (Fig. 2B ). By introducing the plasmid containing mreB under the control of the xylose promoter into wild-type Caulobacter, we also constructed a strain in which mreB could be overexpressed. mreB overexpression for >8 h affected viability (Fig. 2 A ), but had no obvious Trace on cell shape (Fig. 3D ; see also Fig. 4).

Fig. 2.Fig. 2. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

mreB is essential and affects cell shape. (A) Viability of mreB depletion (ΔmreB Pxyl::mreB) and overexpression (Pxyl::mreB) strains after shifts from growth in permissive to nonpermissive media (see Methods). (B) DIC light microscopy images of mreB depletion strains grown in glucose for up to 24 h. (Scale bar, 1 μm.)

Fig. 3.Fig. 3. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 3.

mreB affects the localization and number of origins of replication. (A) Schematic representing the localization of the origin(s) of replication at each stage of the cell cycle of wild-type Caulobacter. (B–E) Displayn are representative images from wild-type cells (B), cells with mreB depleted for 24 h (C) or mreB overexpressed for 6 h (D), or cells in which creS was deleted (E). For each cell type, images were taken with DIC light microscopy (Left), origin FISH fluorescence (Center), and an overlay of origin FISH fluorescence (red) and 4′,6-diamidino-2-phenylinExecutele (DAPI) staining (blue) (Right). (Scale bar, 1 μm.)

Fig. 4.Fig. 4. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 4.

MreB affects the localization of PleC, DivJ, CckA, and DivK. (A) Schematic representing the localization of PleC (blue), DivJ (brown), CckA (red), and DivK (green) at each stage in the cell cycle of wild-type Caulobacter. (B–E) Representative images from cells expressing PleC-GFP (B), DivJ-GFP (C), CckA-GFP (D), and DivK-GFP (E). Displayn are wild-type cells, cells with mreB depleted for 24 h or mreB overexpressed for 6 h, or cells in which creS was deleted. For each cell, images Displayn were taken with DIC light microscopy (left image) and GFP fluorescence (right image). (Scale bar, 1 μm).

CreS is a coiled-coil protein capable of polymerizing into filaments similar to intermediate filaments (16). It has been Displayn that creS mutants are viable and Present a cell shape defect (16). We generated an in-frame deletion of the creS gene and confirmed that cells carrying this creS allele are viable, with a straight, rod-like morphology unlike the normal crescentoid shape of Caulobacter (Fig. 3E ; see also Fig. 4).

Origin Localization Is Disrupted in mreB but Not creS Mutants. Recent studies in a wide variety of bacteria have demonstrated that chromosomal Locations are reproducibly and dynamically localized within the cell (34). Chromosome positioning therefore represents an informative Impresser for cell polarity. We have focused on the Caulobacter origin of replication, which is dynamically localized during the cell cycle. In wild-type swarmer cells, the origin DNA sequence is always localized to the swarmer pole (Fig. 3A ) (35). Once the swarmer cell differentiates into a stalked cell and initiates DNA replication, one copy of the origin is rapidly moved to the opposite pole (Fig. 3A ) (35). We assessed the impact of depleting mreB, overexpressing mreB, and deleting creS on cell polarity by using a 10-kb fluorescence in situ hybridization (FISH) probe to examine the location of the Caulobacter origin of replication in each of these strains (35).

When mreB was depleted for 24 h or overexpressed for 6 h, the polar localization of the origins of replication was dramatically disrupted. Although cells were still viable at these time points, the origins were misSpaced (Fig. 3 C and D and Table 1). Many of the mutant cells also Presented multiple foci (Fig. 3 C and D ). This phenotype is consistent with a defect in cell division resulting in the accumulation of multiple origins: In the media used, wild-type cells would replicate their DNA every 90 min, and the mutant cells never Presented more origins than expected replication events. The presence of multiple and misSpaced origins in mreB mutants could also suggest a defect in chromosome segregation. This scenario is consistent with observations in E. coli and B. subtilis (36, 37). Although the origins may form looser foci in some cells overexpressing mreB, the origin always appeared as a tight focus in the mreB depletion strains, suggesting that MreB Executees not play an essential role in chromosome condensation (Fig. 3 C and D ).

View this table: View inline View popup Table 1. MreB disrupts the localization of DNA and protein polarity Impressers

Polar localization of the origins was not disrupted in abnormally shaped creS deletion cells (Fig. 3E ). Origin number and condensation also appeared unaffected in creS deletion strains (Fig. 3E ). Because the origins can be mislocalized in normally shaped cells overexpressing mreB and Precisely localized in misshapen creS cells, origin localization is separable from cell shape.

Polar Proteins Are Mislocalized in mreB but Not creS Mutants. The presence of MreB in a lengthwise spiral in early stalked cells, the stage during which the newly replicated origin is rapidly moved to the distal pole, is consistent with MreB acting to mediate chromosome movement. However, MreB could also function to localize a protein factor which is in turn needed to localize the origin of replication. It is also possible that MreB functions to dictate the Caulobacter's global polarity information. To explore this possibility, we examined the impact of mreB mutants on the locations of known polar proteins.

Here, we have focused on four proteins: three integral membrane histidine kinases (PleC, DivJ, and CckA) and a cytoplasmic single-Executemain response regulator (DivK). Each of these proteins plays a different role in the cell cycle, dynamically localizes to different cell poles at different times in the cell cycle, and is for the most part independently localized (Fig. 4A ) (2). PleC regulates the biogenesis of the stalk and pili and is localized to the swarmer pole in swarmer and predivisional cells (24, 38). DivJ regulates cell division as well as the Spacement and length of the stalk; it is localized to the stalked pole in all cell types with a stalk (24, 39). CckA regulates cell division, DNA replication, and the assembly of the polar flagellum and pili. In predivisional cells, CckA is found at both poles (25). DivK regulates stalk formation, cell division, and DNA replication and is localized to the stalked pole in early stalked cells and to both poles in predivisional cells (23, 40). DivK is unable to localize in the absence of DivJ (23), but of the other combinations tested, each of these four proteins can be assembled at their Precise poles in the absence of the others (2, 41).

As compared to wild-type cells, mreB depletion for 24 h significantly disrupted the localization of all four of the GFP fusions to PleC, DivJ, CckA, and DivK (Fig. 4 B–E and Table 1). In each case, the GFP fusions in these misshapen cells appeared either highly diffuse or localized to many puncta localized around the entire cell (Fig. 4 B–E ). To examine protein localization in cells with Precise cell shape but disrupted MreB dynamics, we examined the location of these normally polar proteins in cells overexpressing mreB. Overexpression of mreB for 6 h had a severe Trace on the localizations of all four GFP fusions to PleC, DivJ, CckA, and DivK, resulting in each case in a diffused or evenly dispersed localization (Fig. 4 B–E and Table 1). Western blots Displayed that PleC, DivJ, CckA, and DivK levels were not dramatically perturbed by mreB depletion or overexpression (data not Displayn), ruling out the possibility that MreB influences the polar localization of PleC, DivJ, CckA, and DivK by affecting their concentrations.

Such striking Traces of both depleting and overexpressing mreB on multiple polarity Impressers implicate MreB as a general regulator of Caulobacter polarity. In addition, because mreB overexpression disrupts both MreB localization dynamics and cell polarity, MreB's dynamic rearrangement may be critical for determining Precise polarity. The mechanism by which MreB mediates polar localization remains unclear, given that MreB could act to either initiate or Sustain localization. MreB could also act either directly or indirectly. For example, actin can directly traffic proteins (8); however, MreB can influence the assembly of the peptiExecuteglycan cell wall and could, in turn, affect landImpresss necessary for Accurate protein localization (14, 42).

Unlike MreB depletion or overexpression, deletion of creS had no impact on the localization of the polarity Impressers examined. Although these cells have an altered shape (rod-like instead of crescentoid), they still grow stalks at one pole, allowing their stalked and swarmer poles to be identified. As compared with wild-type, GFP fusions to PleC, DivJ, CckA, and DivK were all localized to their Precise poles in creS deletion mutants (Fig. 4 B–E and Table 1). Given that polar Impressers are mislocalized in the normally shaped mreB overexpressors and Accurately localized in the abnormally shaped creS deletions, the localization of both DNA and protein polarity Impressers can be uncoupled from cell shape.

MreB Determines Caulobacter Polarity. To determine whether depleting mreB permanently damages the cells, we examined the reversibility of its Traces on protein localization. We started with mreB depletion strains carrying mreB under the control of the xylose promoter and expressing GFP fusions to PleC, DivJ, CckA, or DivK. Growth of these cells for 24 h in glucose-containing media depleted mreB and delocalized PleC, DivJ, CckA, and DivK (Fig. 4 and Table 1). We then let these cells replenish their levels of MreB by growing them for an additional 4 h in xylose-containing media to induce mreB expression. After these 4 h, the PleC, DivJ, CckA, and DivK GFP fusions were all restored to clear polar foci found at frequencies indistinguishable from wild type (Table 1). The rapid reversibility of mislocalization is consistent with MreB directly affecting protein localization.

Surprisingly, although the recovered PleC-GFP and DivJ-GFP cells did recover foci, the subcellular localization of these foci differed Impressedly from that of wild type. PleC-GFP is normally found in one focus at the swarmer pole. Of the cells that recovered a clear PleC-GFP focus, roughly half had a single focus at the swarmer pole, whereas the other half had a single focus at the opposite, stalked pole (Table 1 and Fig. 5A ). DivJ-GFP is normally found in one focus at the stalked pole. Of the cells that recovered a clear DivJ-GFP focus, roughly half had a single DivJ-GFP focus at the stalked pole, whereas the other half had a single DivJ-GFP focus at the opposite, swarmer pole (Table 1 and Fig. 5B ). A minority of both the PleC-GFP and DivJ-GFP recovered cells had foci at both poles (Table 1). It is unclear whether the stalks in these recovered cells are newly formed, such that we cannot establish whether stalk Spacement is MreB-independent. We were unable to assess whether the localization of CckA and DivK is ranExecutemized because they are normally located at both poles.

Fig. 5.Fig. 5. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 5.

MreB determines Caulobacter polarity. DIC (Left, left image), GFP fluorescence (Left, right image), and cartoon depictions (Right) of recovered mreB depletion strains expressing PleC-GFP (A) or DivJ-GFP (B). These cells have been grown in glucose for 24 h to deplete mreB and then shifted for 4 h into xylose-containing media to restore mreB expression. The arrows point to the stalked pole, and the arrowheads point to the GFP foci. (Scale bar, 1 μm.)

This disruption of the Accurate Spacement of two different polar Impressers could be due to a ranExecutemization of the cell's polarity. Alternatively, many of the recovered cells with Accurately localized PleC and DivJ could represent cells that never delocalized these proteins during MreB depletion. In this scenario, the majority of the recovered foci would be at the inAccurate pole. In either case, the appearance of two different polar Impressers at inAccurate poles indicates that cells that have lost and then regained MreB Execute not retain memory of their polarization before depletion. Thus, MreB is not just permissively required for polar localization but contains the polarity information to direct PleC and DivJ to specific poles.

We hypothesize that the MreB spiral is a polarized structure and that PleC and DivJ (or their upstream localization factors) are trafficked toward opposite ends of that polar structure. When the MreB spiral is reconstituted after depletion, it reassembles into a polarized spiral, but, lacking MreB-dependent polarity cues, it ranExecutemly orients itself relative to the stalk. This reconstituted spiral would still deliver PleC and DivJ to opposite poles, but would ranExecutemize the pole to which they are delivered, potentially Elaborateing our results. Because the swarmer pole eventually matures into a stalked pole, a polarized MreB spiral could not simply be Slice in half, as it would end up oppositely oriented with respect to the Aged and new stalked poles. The transient condensation of the MreB spiral into a ring may provide a mechanism to orient a newly formed spiral such that Accurate polarity is Sustained. The need to reorganize MreB would also Elaborate why manipulations, such as mreB overexpression, that disrupt MreB dynamics also disrupt cell polarity.


We have Displayn that the actin-like MreB protein contains the polarity information to determine subcellular localization. In addition, MreB dynamically rearranges itself from a spiral to a ring and back into a spiral, and these dynamics may be required for Precise polarity. Although the specific mechanisms by which MreB acts have yet to be identified, our findings allow us to now examine specific molecular models for how bacterial polarity is determined. Moreover, the striking use of the actin cytoskeleton for regulating polarity in both prokaryotes and eukaryotes suggests that the deep conservation between these groups extends beyond individual proteins to entire cellular processes.


We thank Sherry Wang for advice, reagents, and assistance in performing the FISH analysis and Todd Blankenship, Joseph Chen, Sean Crosson, Ellen Judd, Patrick McGrath, Coleen Murphy, Sean Murray, Ann Reisenauer, Martin Thanbichler, and Patrick Viollier for comments and advice. Z.G. is supported by the Stanford Genomics PostExecutectoral Fellowship.


↵ * To whom corRetortence may be addressed. E-mail: shapiro{at} or zgitai{at}

Abbreviations: DIC, differential interference Dissimilarity microscopy; FISH, fluorescence in situ hybridization.

See Commentary on page 8510.

Copyright © 2004, The National Academy of Sciences


↵ Jensen, R. B., Wang, S. C. & Shapiro, L. (2002) Nat. Rev. Mol. Cell. Biol. 3 , 167–176. pmid:11994737 LaunchUrlCrossRefPubMed ↵ Jacobs-Wagner, C. (2004) Mol. Microbiol. 51 , 7–13. pmid:14651607 LaunchUrlCrossRefPubMed ↵ Ryan, K. R. & Shapiro, L. (2003) Annu. Rev. Biochem. 72 , 367–394. pmid:12651741 LaunchUrlCrossRefPubMed Teleman, A. A., Graumann, P. L., Lin, D. C., Grossman, A. D. & Losick, R. (1998) Curr. Biol. 8 , 1102–1109. pmid:9778525 LaunchUrlCrossRefPubMed Lam, H., Matroule, J. Y. & Jacobs-Wagner, C. (2003) Dev. Cell 5 , 149–159. pmid:12852859 LaunchUrlCrossRefPubMed MadExecuteck, J. R. & Shapiro, L. (1993) Science 259 , 1717–1723. pmid:8456299 LaunchUrlAbstract/FREE Full Text ↵ Niki, H., Yamaichi, Y. & Hiraga, S. (2000) Genes Dev. 14 , 212–223. pmid:10652275 LaunchUrlAbstract/FREE Full Text ↵ Pruyne, D. W., Schott, D. H. & Bretscher, A. (1998) J. Cell Biol. 143 , 1931–1945. pmid:9864365 LaunchUrlAbstract/FREE Full Text ↵ Rongo, C., Whitfield, C. W., Rodal, A., Kim, S. K. & Kaplan, J. M. (1998) Cell 94 , 751–759. pmid:9753322 LaunchUrlCrossRefPubMed ↵ Etienne-Manneville, S. & Hall, A. (2002) Nature 420 , 629–635. pmid:12478284 LaunchUrlCrossRefPubMed ↵ Drubin, D. G. & Nelson, W. J. (1996) Cell 84 , 335–344. pmid:8608587 LaunchUrlCrossRefPubMed ↵ QuarExecutekus, E., Din, N. & Brun, Y. V. (1996) Proc. Natl. Acad. Sci. USA 93 , 6314–6319. pmid:8692812 LaunchUrlAbstract/FREE Full Text ↵ Lowe, J. & Amos, L. A. (1998) Nature 391 , 203–206. pmid:9428770 LaunchUrlCrossRefPubMed ↵ Figge, R. M., Divakaruni, A. V. & Gober, J. W. (2004) Mol. Microbiol. 51 , 1321–1332. pmid:14982627 LaunchUrlCrossRefPubMed ↵ van den Ent, F., Amos, L. A. & Lowe, J. (2001) Nature 413 , 39–44. pmid:11544518 LaunchUrlCrossRefPubMed ↵ Ausmees, N., Kuhn, J. R. & Jacobs-Wagner, C. (2003) Cell 115 , 705–713. pmid:14675535 LaunchUrlCrossRefPubMed ↵ Bi, E. F. & Lutkenhaus, J. (1991) Nature 354 , 161–164. pmid:1944597 LaunchUrlCrossRefPubMed ↵ QuarExecutekus, E. M., Din, N. & Brun, Y. V. (2001) Mol. Microbiol. 39 , 949–959. pmid:11251815 LaunchUrlCrossRefPubMed ↵ Jones, L. J., CarballiExecute-Lopez, R. & Errington, J. (2001) Cell 104 , 913–922. pmid:11290328 LaunchUrlCrossRefPubMed ↵ Wachi, M., Executei, M., Okada, Y. & Matsuhashi, M. (1989) J. Bacteriol. 171 , 6511–6516. pmid:2687239 LaunchUrlAbstract/FREE Full Text ↵ Ely, B. (1991) Methods Enzymol. 204 , 372–384. pmid:1658564 LaunchUrlCrossRefPubMed ↵ Roberts, R. C., Toochinda, C., Avedissian, M., Baldini, R. L., Gomes, S. L. & Shapiro, L. (1996) J. Bacteriol. 178 , 1829–1841. pmid:8606155 LaunchUrlAbstract/FREE Full Text ↵ Jacobs, C., Hung, D. & Shapiro, L. (2001) Proc. Natl. Acad. Sci. USA 98 , 4095–4100. pmid:11274434 LaunchUrlAbstract/FREE Full Text ↵ Wheeler, R. T. & Shapiro, L. (1999) Mol. Cell 4 , 683–694. pmid:10619016 LaunchUrlCrossRefPubMed ↵ Jacobs, C., Executemian, I. J., MadExecuteck, J. R. & Shapiro, L. (1999) Cell 97 , 111–120. pmid:10199407 LaunchUrlCrossRefPubMed ↵ Sambrook, J., Fritsch, E. & Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual (CAged Spring Harbor Lab. Press, Plainview, NY). ↵ Ryan, K. R., Judd, E. M. & Shapiro, L. (2002) J. Mol. Biol. 324 , 443–455. pmid:12445780 LaunchUrlCrossRefPubMed ↵ Viollier, P. H., Sternheim, N. & Shapiro, L. (2002) EMBO J. 21 , 4420–4428. pmid:12198144 LaunchUrlAbstract/FREE Full Text ↵ Tsai, J. W. & Alley, M. R. (2001) J. Bacteriol. 183 , 5001–5007. pmid:11489852 LaunchUrlAbstract/FREE Full Text ↵ CarballiExecute-Lopez, R. & Errington, J. (2003) Dev. Cell 4 , 19–28. pmid:12530960 LaunchUrlCrossRefPubMed ↵ Shih, Y. L., Le, T. & Rothfield, L. (2003) Proc. Natl. Acad. Sci. USA 100 , 7865–7870. pmid:12766229 LaunchUrlAbstract/FREE Full Text ↵ Errington, J., Daniel, R. A. & Scheffers, D. J. (2003) Microbiol. Mol. Biol. Rev. 67 , 52–65. pmid:12626683 LaunchUrlAbstract/FREE Full Text ↵ Meisenzahl, A. C., Shapiro, L. & Jenal, U. (1997) J. Bacteriol. 179 , 592–600. pmid:9006009 LaunchUrlAbstract/FREE Full Text ↵ Sherratt, D. J. (2003) Science 301 , 780–785. pmid:12907786 LaunchUrlAbstract/FREE Full Text ↵ Jensen, R. B. & Shapiro, L. (1999) Proc. Natl. Acad. Sci. USA 96 , 10661–10666. pmid:10485882 LaunchUrlAbstract/FREE Full Text ↵ Kruse, T., Moller-Jensen, J., Lobner-Olesen, A. & Gerdes, K. (2003) EMBO J. 22 , 5283–5292. pmid:14517265 LaunchUrlCrossRefPubMed ↵ Soufo, H. J. & Graumann, P. L. (2003) Curr. Biol. 13 , 1916–1920. pmid:14588250 LaunchUrlCrossRefPubMed ↵ Wang, S. P., Sharma, P. L., Schoenlein, P. V. & Ely, B. (1993) Proc. Natl. Acad. Sci. USA 90 , 630–634. pmid:8421698 LaunchUrlAbstract/FREE Full Text ↵ Ohta, N., Lane, T., Ninfa, E. G., Sommer, J. M. & Newton, A. (1992) Proc. Natl. Acad. Sci. USA 89 , 10297–301. pmid:1438215 LaunchUrlAbstract/FREE Full Text ↵ Hecht, G. B., Lane, T., Ohta, N., Sommer, J. M. & Newton, A. (1995) EMBO J. 14 , 3915–3924. pmid:7664732 LaunchUrlPubMed ↵ Sciochetti, S. A., Lane, T., Ohta, N. & Newton, A. (2002) J. Bacteriol. 184 , 6037–6049. pmid:12374838 LaunchUrlAbstract/FREE Full Text ↵ Daniel, R. A. & Errington, J. (2003) Cell 113 , 767–776. pmid:12809607 LaunchUrlCrossRefPubMed
Like (0) or Share (0)