Negative membrane curvature as a cue for... (PDF Download Available)

Negative membrane curvature as a cue for subcellular localization of a bacterial protein Kumaran S. Ramamurthi and Richard Losick1 Department of Molecular and Cellular Biology, Harvard University, 16 Divinity Avenue, Cambridge, MA 02138 Contributed by Richard Losick, June 19, 2009 (sent for review June 16, 2009)

geometric cue 兩 protein localization

P

roteins that localize to distinct regions inside a cell are guided by signals encoded within their primary amino acid sequence that recognize a feature or features of their intended destination (1). In bacterial cells, which lack vesicle-mediated sorting machinery that delivers cargo to particular destinations, proteins typically arrive at their correct subcellular address by diffusing to the target site and being captured by interaction with another protein that previously localized at that site (2). Thus, groups of proteins that form structures at particular subcellular sites arrive at their destination by recognizing a protein or proteins that arrived there before it, often in a highly ordered sequence. But how does the very earliest-arriving protein identify and mark the site for the recruitment of other protein components of the structure? We previously discovered that the localization of a small protein named SpoVM in the bacterium Bacillus subtilis is governed by recognition of a geometric cue (3). During spore formation, the rod-shaped bacterium produces a spherical internal organelle that eventually becomes a spore (4). Several scores of proteins localize on the surface of this organelle to form a supramolecular structure called the spore coat, and each protein is guided to this location by recognizing a protein that previously localized to the assembling structure (5, 6). Among the first, if not the first, coat protein that arrives at the surface of the developing spore is SpoVM (7, 8). We reported that SpoVM localizes to this surface by recognizing the convex geometry of the membrane surrounding the developing spore, the only convex, or positively curved, surface in the cytosol of the bacterium. The discovery that positive membrane curvature is a localization cue raised the possibility that negative membrane curvature might also be exploited for protein localization in bacteria. The inside surface of rod-shaped bacteria, such as B. subtilis, is concave but certain regions of the cell, such as division septa and the poles, are more negatively curved than the lateral inner surface of the cell. An attractive candidate for a protein that www.pnas.org兾cgi兾doi兾10.1073兾pnas.0906851106

might recognize negative membrane curvature is the cell division protein DivIVA in B. subtilis (9). In actively growing cells, DivIVA localizes to division septa and to the poles and recruits the cell division inhibitors MinC and MinD (via a bridge protein MinJ) to prevent the aberrant formation of a division septum close to these sites (10–12). DivIVA has a second function during spore formation (13). Cells that have entered the pathway to sporulate produce a chromosome-binding protein (RacA) that adheres to multiple sites in the replication origin-proximal region of the chromosome (14). DivIVA, which is located at the poles, is responsible for anchoring newly duplicated origin regions to the extreme opposite ends of the cell in a manner that depends on RacA, ensuring that each pole (and eventually each daughter cell) receives 1 copy of the origin of replication and, by extension, 1 chromosome. Herein, we present evidence consistent with the idea that DivIVA localizes by recognizing negative membrane curvature. We show that DivIVA displays a hierarchical preference for different degrees of concavity, such that it chiefly localizes to sharply concave division septa, secondarily to the hemispherical cell poles, and poorly or not at all to the inner lateral surface of the bacterium. We also show that localization is dynamic, with DivIVA redeploying itself from septa to poles in filamentous cells in which cytokinesis was blocked. Finally, we demonstrate that in cells that have a uniform concave inner surface, spherical protoplasts, DivIVA is unable to localize in a specific manner. Our current and previous findings raise the possibility that membrane curvature is exploited widely as a cue for protein localization in bacteria. Results Multiple Domains Mediate the Proper Localization of DivIVA. DivIVA localizes in two distinct fashions. One population of molecules localizes as a ring to nascent cell division septa, and a second population forms arcs at cell poles. Does DivIVA recognize different factors at each of these sites that drive its localization, or does it recognize a common feature that recruits it to both locations? A feature common to both sites is extreme concavity of the plasma membrane, compared with the relatively slight concavity found along the lateral edges of the rod-shaped B. subtilis cell (Fig. 1). Perhaps, then, the cell’s geometry (negative curvature) at these sites specifically recruits DivIVA. We therefore predicted that DivIVA would localize preferentially to the highly negatively curved cell division septa and to a lesser extent to the hemispherical, less negatively curved poles. To test this, we examined the localization of DivIVA fused to green fluorescent protein (DivIVA-GFP) in actively dividing cells. To prevent the saturation of different target sites, we designed DivIVA fusions to GFP that were produced under its relatively weak, native promoter. Furthermore, to compare more directly the localization of DivIVA to cell poles and division septa, we examined the

Author contributions: K.R. and R.L. designed research; K.R. performed research; K.R. and R.L. analyzed data; and K.R. and R.L. wrote the paper. The authors declare no conflict of interest. 1To

whom correspondence should be addressed. E-mail: [email protected].

PNAS 兩 August 11, 2009 兩 vol. 106 兩 no. 32 兩 13541–13545

MICROBIOLOGY

Bacterial proteins often localize to distinct sites within the cell, but the primary cues that dictate localization are largely unknown. Recent evidence has shown that positive membrane curvature can serve as a cue for localization of a peripheral membrane protein. Here we report that localization of the peripheral membrane protein DivIVA is determined in whole or in part by recognition of negative membrane curvature and that regions of the protein near the N and C terminus are important for localization. DivIVA, which is a cell division protein in Bacillus subtilis, localizes principally as a ring at nascent septa and secondarily to the less negatively curved, inside surface of the hemispherical poles of the cell. When cytokinesis is prevented, DivIVA redistributes itself to, and becomes markedly enriched at, the poles. When the rod-shaped cells are converted into spheres (protoplasts) by treatment with lysozyme, DivIVA adopts a largely uniform distribution around the cell. Recognition of membrane curvature by peripheral membrane proteins could be a general strategy for protein localization in bacteria.

Table 1. Strains used in this study Strain PY79 KR528 KR534 KR533 KR536 KR541 KR571 KR570 KR572 KR568 Fig. 1. Geometric cues for bacterial protein localization. (A) A cross-section of a sporulating B. subtilis cell is represented in which the spherical developing spore displays the only convex, or positively curved, surface (yellow) in the cytosol of the surrounding, rod-shaped cell (blue) which SpoVM recognizes as a localization cue. (B) Schematic cross section of an actively dividing B. subtilis cell, which harbors two areas of extreme concave, or negative, curvature (yellow) that recruits DivIVA. The most sharply concave region is formed along the edge where the division septum (center of the cell) meets the periphery of the cell. A second, less extremely concave site is the inside surface of the hemispherical cell poles.

localization of DivIVA-GFP in cells that harbored a deletion of the sinI gene, which results in the production of single cells rather than unseparated chains (15). In single cells, DivIVA-GFP localized as arcs at both poles; in cells that had elaborated a division septum, DivIVA-GFP preferentially localized to the septum, and to a lesser extent, to the extreme poles (Figs. 1 and 2A). The results so far are consistent with the idea that DivIVA localizes to regions of negative curvature, preferring the more extreme curvature of division septa (when present) over the hemispherical curvature of the poles. To identify regions of the protein required for proper localization, we made successive deletions at either the 3⬘ or the 5⬘ ends of divIVA and examined the localization of the resulting truncated fusion proteins in dividing cells. Removal of 14

A

B

C

D

E

F

G

H

I

J

K

L

Fig. 2. DivIVA-GFP localizes to septa and poles. (A) Localization of full-length DivIVA-GFP (strain KR528), and truncated versions (B) DivIVA⌬(151–164)-GFP (strain KR534), (C) DivIVA⌬(126 –164)-GFP (strain KR533), and (D) DivIVA⌬(2–50)GFP (strain KR536) in a sinI mutant in which cell chaining is prevented. Fusions were expressed from the native divIVA promoter. A–D and E–H are, respectively, fluorescence and phase contrast images of the same cells in each column. I–L are overlays of the corresponding fluorescence and phase contrast images. 13542 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0906851106

KR569

Genotype/Description

Ref.

Prototrophic derivative of B. subtilis 168 ⌬sinI::spec amyE::divIVA-gfp cat ⌬sinI::spec amyE::divIVA⌬(151–164)-gfp cat ⌬sinI::spec amyE::divIVA⌬(126–164)-gfp cat ⌬sinI::spec amyE::divIVA⌬{2–50)-gfp cat amyE::Phyperspank-divIVA-gfp spec amyE::Phyperspank-divIVA⌬(2–50)-gfp spec amyE::Phyperspank-divIVA-gfp spec thrC::Phyperspank-sirA erm amyE::Phyperspank-divIVA⌬(2–50)-gfp spec thrC::Phyperspank-sirA erm bkdB::Tn917::amyE::cat::Phyperspank-divIVA-gfp spec bkdB::Tn917::amyE::cat::Phyperspank-divIVA-gfp spec amyE::Pxyl-mciZ cat

24

residues from the C terminus (DivIVA⌬(151–164)-GFP) did not impair localization (Fig. 2B). However, further truncation of the protein by 25 residues abolished both septal localization and arc-like polar localization of the fluorescent signal. Instead, DivIVA⌬(126 –164)-GFP localized as foci (typically a single focus) at various positions near the cell’s periphery (Fig. 2C). Removal of residues 2–50 from the N terminus (DivIVA⌬(2–50)-GFP) abolished localization and resulted in the uniform cytosolic distribution of the truncated fusion protein (Fig. 2D). Thus, regions near both ends of the protein are critical for localization. An N-terminal region of 49 residues seems to be important for DivIVA molecules to assemble into higher order clusters, and, of particular interest, a stretch of 25 aa from the C-terminal region (residues 126 to 151) is implicated in the capacity of DivIVA to localize to concave surfaces. Nucleoid Occlusion Does Not Restrict DivIVA to the Poles. An alter-

native hypothesis is that DivIVA is restricted to the septum and the poles by the phenomenon of nucleoid exclusion. Nucleoid exclusion is a mechanism for preventing the assembly of the cytokinetic protein FtsZ into a Z-ring in nucleoid-containing regions the cell (16). Conceivably, therefore, DivIVA is restricted to the septum and the poles because the nucleoid prevents it from assembling at all sites in the cell except the gap between the two nucleoids at the center of the cell and the gaps between the ends of nucleoids and the cell poles. To test this hypothesis, we exploited the capacity of cells that overexpress a gene encoding a DNA replication inhibitor called SirA (YneE) to produce rod-shaped cells that are devoid of chromosomes (17). If the positioning of DivIVA is dictated by nucleoid occlusion, then in anucleate cells DivIVA should localize indiscriminately. We therefore examined the localization of DivIVA-GFP in cells overexpressing sirA. C and R of Fig. 3 show a representative chain of 3 cells in which the central cell was devoid of DNA. Despite the absence of a nucleoid, DivIVAGFP was restricted to the division septa at both ends of the DNA-free cell (C and H) as was seen for nucleoid-containing, wild-type cells (Fig. 3A and F). The images of Fig. 3C are representative of the results obtained with greater than 91% of anucleate cells (n ⫽ 23) generated by overexpression of sirA. It might be argued that the fluorescent signal at the ends of the anucleate cell of Fig. 3C was from DivIVA-GFP in the adjacent nucleoid-containing cells. However, 3D and S show a (less common) chain in which 2 adjacent cells are nucleoid free. Here, too, DivIVA-GFP was seen at the ends of the cells, including the septum between the 2 anucleate cells. Therefore, DivIVA-GFP must have been present in the anucleate cells in which it was Ramamurthi and Losick

A

B

C

D

E

A

B

C

D

F

G

H

I

J

E

F

G

H

K

L

M

N

O

P

Q

R

S

T

Fig. 3. DivIVA-GFP localizes to septa in anucleate cells. A and B show, respectively, localization of DivIVA-GFP and DivIVA⌬(2–50)-GFP in the wild type (strains KR541 and KR571). C–E show localization of DivIVA-GFP (C and D) and DivIVA⌬(2–50)-GFP (E) in cells overexpressing the DNA replication inhibition gene sirA (strains KR570 and KR572). D shows an example of adjacent anucleate cells that arise from SirA overproduction. A–E are composite images showing GFP fluorescence (green) and DNA (blue) overlaid on membrane (red). F–J show fluorescence from GFP, K–O fluorescence from membrane staining with the dye FM4 – 64, and P–T DNA staining with DAPI. gfp fusions and sirA was expressed from the inducible promoter hyperspank by the addition of IPTG. The same cells are shown from top to bottom in each column.

localizing. As a further demonstration that fusion protein was present in anucleate cells, E and J show that the non-localizing truncated fusion protein, DivIVA⌬(2–50)-GFP was present in, and uniformly dispersed throughout, the cytosol of an anucleate cell. We conclude that DivIVA remains capable of localizing normally even if a chromosome is not present. DivIVA Redistributes from Septa to Poles When Cytokinesis is Blocked.

As noted above, DivIVA localizes more intensely at septa than at poles. Our explanation for this is that the edges of septa have more extreme negative curvature than do the hemispherical poles. As test of this idea, we asked whether the protein would redistribute itself from septa to the poles when cytokinesis is blocked. To accomplish this, we examined the localization of DivIVA in filamentous cells in which cytokinesis was prevented by overproduction of the FtsZ inhibitor MciZ using an xyloseinducible copy of the mciZ gene (18). Before the addition of inducer, DivIVA-GFP localized preferentially to regularlyspaced division septa, with relatively little signal to be seen at the poles, a pattern similar to that seen in wild type cells (Fig. 4A and B). Filamentation commenced by 30 min after the addition of inducer, and by 1 h the level of fluorescence at the cell poles had increased markedly (Fig. 4C and D). These results are consistent with the idea that DivIVA preferentially coalesces at septa but in their absence redistributes itself to the somewhat less negatively curved inner surface of the poles.

Fig. 4. DivIVA-GFP redistributes from septa to poles upon inhibition of cytokinesis. (A) DivIVA-GFP localization in wild type strain (KR568). (B–D) DivIVA-GFP localization in cells overexpressing mciZ under the control of xylose-inducible promoter at the indicated times after addition of xylose (strain KR569). (E–H) Membranes of cells in A–D visualized with the dye FM4 – 64. GFP fusions were expressed from the IPTG-inducible hyperspank promoter.

ited a largely uniform distribution in protoplasts, with occasional small foci seen at various, apparently random, positions along the periphery (Fig. 5B). We conclude that DivIVA preferentially localizes to areas of extreme negative membrane curvature, and that in its absence, DivIVA localizes indiscriminately. It is not excluded, however, that lysozyme treatment removed a mark from septation that contributed to DivIVA localization. Discussion A major challenge in the field of bacterial cytology is to elucidate the signals that govern protein subcellular localization. The traditional solution to this problem is to demonstrate that the localization of a particular protein is determined by its interaction with another protein, which, in turn, might depend for its localization on yet another protein in a hierarchy of interactions. Eventually, however, we are faced with the challenge of identifying the primary cue upon which the hierarchy depends. The recent discovery that the assembly of the spore coat in sporulating cells of B. subtilis is ultimately dictated by a geometric cue, positive membrane curvature, recognized by the protein SpoVM raised the possibility that negative membrane curvature might be also be exploited as a geometric landmark for protein localization (3). Here we have presented evidence consistent with this possibility for the cell division protein DivIVA. Multiple lines of evidence support this view. First, DivIVA favors sites of high negative curvatures over sites with less concavity. Thus, in cells with both division septa and poles, DivIVA localized chiefly to the extremely negatively curved surface of division septa, secondarily to the hemispherical, less

A

B

C

D

concavity to preferentially localize to division septa rather than poles, then DivIVA should lose its ability to localize to a particular site if membrane curvature were uniform. To test this prediction, we created spherical cells by enzymatic removal of the cell wall with lysozyme and stabilization of the resulting protoplasts by the addition of sucrose to the buffer. As compared to rod-shaped cells undergoing cell division, the inner surface of protoplasts is more or less uniform. Whereas DivIVA-GFP localized to division septa, and to a lesser extent the poles, in mock-treated, rod-shaped cells (Fig. 5A), DivIVA-GFP exhibRamamurthi and Losick

Fig. 5. DivIVA-GFP localizes uniformly in protoplasts. (A) DivIVA-GFP localization in sinI mutant cells that were not treated with lysozyme or (B) that were treated with lysozyme. (C and D) Phase contrast images of cells in A and B. PNAS 兩 August 11, 2009 兩 vol. 106 兩 no. 32 兩 13543

MICROBIOLOGY

Regions of Preferential Concavity Are Required for Distinct Localization by DivIVA. If DivIVA can discriminate between degrees of

concave inner surface of the cell poles, and poorly or not at all along the lateral inner surface of the rod-shaped cell. This localization pattern is consistent with a model in which DivIVA discriminates among degrees of concavity. Second, when the formation of division septa was blocked by induction of the synthesis of an inhibitor of cytokinesis, DivIVA became greatly enriched at the cell poles. This observation is consistent with the view that position of DivIVA is dynamic, and when it preferred localization site (the septum) is absent it accumulates instead at its secondary localization site, the less negatively curved inner surface of the pole. Third, when both the primary and secondary sites of preferred localization are removed by treatment of cells with lysozyme, which converts the rod-shaped cell into a uniformly curved sphere, DivIVA exhibits little or no preferential sites of localization and instead distributes itself in a more or less uniform fashion around the cell. We also considered the alternative hypothesis that DivIVA is restricted to the septum and the poles by nucleoid occlusion. But when DNA replication was blocked by the induction of the synthesis of a chromosome replication inhibitor, DivIVA exhibited a normal pattern of septal localization in the resulting anucleate cells. Our results dovetail with the recent report that DivIVA preferentially localized to aberrant regions of high negative curvature in misshapen cells generated by the loss of a cytoskeletal protein (19). Thus, our findings and the complementary observation of Lenarcic et al. (19) are most easily explained by positing that DivIVA relies on negative curvature to guide its localization in the cell. Negative curvature also neatly explains the old observation that DivIVA localizes to the poles in heterologous rod-shaped cells (E. coli and S. pombe), which at the time was not interpreted in terms of membrane curvature (20). It is not known how DivIVA recognizes negative curvature but our dissection of the protein and the work of Lenarcic et al. (19) reveal a region near the N terminus needed for higher order assembly and membrane localization and a region near the C teminus required for association with negatively curved surfaces. Interestingly, striking parallels can be drawn between DivIVA in B. subtilis and the unrelated protein PopZ of the bacterium Caulobacter crescentus (21, 22). Like DivIVA, PopZ oligomerizes at, and localizes to, the extreme poles of the cell. Also, like DivIVA, PopZ recruits other proteins involved in anchoring the origin-region of the chromosome to the cell pole and in controlling the timing and placement of Z-ring formation. It will be interesting to see whether the polar localization of PopZ is also governed by its recognition of extreme membrane concavity at the inside surface of the end of the cell. SpoVM and now DivIVA are two bacterial proteins whose subcellular localization has been attributed to geometric cues, and we note several provocative parallels between them. First, both proteins are relatively small, SpoVM and DivIVA being 26 and 164 amino acids in length, respectively. Second, both SpoVM and DivIVA are among the earliest, if not the earliest, proteins to arrive at their respective subcellular target sites and are critical for the direct or indirect recruitment of other proteins to that site. For example, SpoVM recruits the basement protein of the spore coat SpoIVA, which in turn recruits other coat proteins (8, 23). Likewise, DivIVA recruits the chromosome-anchoring protein RacA and the division inhibitor MinC/MinD (11, 12, 14). As 1. Blobel G (1980) Intracellular protein topogenesis. Proc Natl Acad Sci USA 77:1496 – 1500. 2. Shapiro L, McAdams HH, Losick R (2002) Generating and exploiting polarity in bacteria. Science 298:1942–1946. 3. Ramamurthi KS, Lecuyer S, Stone HA, Losick R (2009) Geometric cue for protein localization in a bacterium. Science 323:1354 –1357. 4. Stragier P, Losick R (1996) Molecular genetics of sporulation in Bacillus subtilis. Annu Rev Genet 30:297–241.

13544 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0906851106

such, SpoVM and DivIVA are ultimate anchors whose localization does not directly depend on other previously localized proteins at the target site. Finally, neither SpoVM nor DivIVA is an integral membrane protein. SpoVM is an amphipathic helical peptide that inserts along its long axis into the phospholipid bilayer and buries its hydrophobic face into the membrane (8, 24), and Lenarcic et al. recently reported that DivIVA also harbors an amphipathic helix near its N terminus, which mediates its association with the plasma membrane (19). This last parallel prompts us to suggest that peripherally-bound membrane proteins, like those harboring amphipathic helices, might be ideally suited for sensing very slight degrees of membrane curvature, like those found at the surface of the spore and at cell poles. Unlike integral membrane proteins, peripherally-bound proteins insert into the membrane spontaneously, do not require prelocalized machinery for membrane insertion, and could sample various membranes before localizing at equilibrium (partitioning) at a favorable site. It will be interesting in future work to see whether recognition of membrane curvature is widespread in bacteria, and, if so, whether it is generally mediated by peripherally membrane proteins. Materials and Methods Strain Construction. Strains are otherwise isogenic derivatives of B. subtilis PY79 (25). B. subtilis competent cells were prepared as described previously (26). To place divIVA-gfp at the amyE locus, 500 nucleotides upstream of the divIVA ORF harboring the divIVA promoter were PCR amplified using primers containing abutted 5⬘ EcoRI and 3⬘ NheI restriction sites. The divIVA ORF, excluding the stop codon, was PCR amplified with primers containing abutted 5⬘ NheI sites and 3⬘ HindIII sites. PCR products were cloned into plasmid pKC13 (harboring gfp mut2) digested with EcoRI and HindIII to create plasmid pKR179. Truncations of divIVA were created similarly, with PCR primers harboring 5⬘ NheI and 3⬘ HindIII sites. IPTG-inducible divIVA-gfp at amyE was created by PCR amplifying the divIVA ORF including the upstream ribosome binding site and excluding the stop codon, with primers containing abutted 5⬘ HindII and 3⬘ NheI sites; and PCR amplifying gfp with 5⬘ NheI and 3⬘ SphI sites. PCR products were cloned into integration vector pDR111 containing the hyperspank promoter to create plasmids pKR196 (full-length divIVA) or pKR207 (divIVA⌬(2–50)). Microscopy. Overnight cultures were diluted 1:20 into CH medium (27) and grown for 2.5 h at 37 °C. IPTG was added (1 mM final concentration), if necessary, to induce expression of divIVA-gfp. Expression of mciZ in strain KR569 was induced with 20 mM xylose at t ⫽ 1.5 h after dilution of overnight culture. Protoplasts of strain KR528 were formed by growing cells as detailed above, removing 200 ␮L of the culture, harvesting cells by centrifugation, and resuspending in 10 ␮L SM buffer (0.5 M sucrose, 20 mM MgCl2, and 10 mM potassium phosphate at pH 6.8) either with or without 0.1 mg/mL lysozyme and incubating at 37 °C for 10 min. To observe DivIVA-GFP localization in anucleate cells resulting from sirA overexpression, single colonies were inoculated into 3 mL LB medium containing 1 mM IPTG and 1.5 ␮g/mL FM4 – 64 membrane dye and grown at 37 °C for 3 h. Cells were harvested in PBS containing 1 ␮g/mL FM4 – 64 and/or 2 ␮g/mL DAPI to visualize DNA, as necessary (protoplasts were viewed directly, without resuspending in PBS). Cell suspensions were placed on 1% agarose pads made with distilled water. Glass coverslips were not treated with poly(L-lysine). Cells were viewed with an Olympus BX61 microscope using a UplanFI 100⫻ phase contrast oil immersion objective. Images were captured with a Hamamatsu Orca-R2 digital CCD camera using Simple PCI software and adjusted with Adobe Photoshop CS. ACKNOWLEDGMENTS. We thank Lilah Rahn-Lee and Joo Eun Lim for strains, Renate Hellmiss for artwork, and Frederico J. Gueiros-Filho and members of the laboratory for helpful discussions. This work was supported by National Institutes of Health grant GM18568 (to R.L.).

5. Kim H, et al. (2006) The Bacillus subtilis spore coat protein interaction network. Mol Microbiol 59:487–502. 6. Henriques AO, Moran CP, Jr (2007) Assembly and function of the spore surface layers. Ann Rev Microbiol. 7. van Ooij C, Losick R (2003) Subcellular localization of a small sporulation protein in Bacillus subtilis. J Bacteriol 185:1391–1398. 8. Ramamurthi KS, Clapham KR, Losick R (2006) Peptide anchoring spore coat assembly to the outer forespore membrane in Bacillus subtilis. Mol Microbiol 62:1547–1557.

Ramamurthi and Losick

19. Lenarcic R, et al. (2009) Localization of DivIVA by targeting to negatively curved membranes. EMBO J, epub ahead of print. 20. Edwards DH, Thomaides HB, Errington J (2000) Promiscuous targeting of Bacillus subtilis cell division protein DivIVA to division sites in Escherichia coli and fission yeast. EMBO J 19:2719 –2727. 21. Ebersbach G, Briegel A, Jensen GJ, Jacobs-Wagner C (2008) A self-associating protein critical for chromosome attachment, division, and polar organization in Caulobacter. Cell 134:956 –968. 22. Bowman GR, et al. (2008) A polymeric protein anchors the chromosomal origin/ParB complex at a bacterial cell pole. Cell 134:945–955. 23. Roels S, Driks A, Losick R (1992) Characterization of spoIVA, a sporulation gene involved in coat morphogenesis in Bacillus subtilis. J Bacteriol 174:575–585. 24. Prajapati RS, Ogura T, Cutting SM (2000) Structural and functional studies on an FtsH inhibitor from Bacillus subtilis. Biochim Biophys Acta 1475:353–359. 25. Youngman P, Perkins JB, Losick R (1984) Construction of a cloning site near one end of Tn917 into which foreign DNA may be inserted without affecting transposition in Bacillus subtilis or expression of the transposon-borne erm gene. Plasmid 12:1–9. 26. Wilson GA, Bott KF (1968) Nutritional factors influencing the development of competence in the Bacillus subtilis transformation system. J Bacteriol 95:1439 –1449. 27. Sterlini JM, Mandelstam J (1969) Commitment to sporulation in Bacillus subtilis and its relationship to development of actinomycin resistance. Biochem J 113:29 –37.

MICROBIOLOGY

9. Cha JH, Stewart GC (1997) The divIVA minicell locus of Bacillus subtilis. J Bacteriol 179:1671–1683. 10. Edwards DH, Errington J (1997) The Bacillus subtilis DivIVA protein targets to the division septum and controls the site specificity of cell division. Mol Microbiol 24:905–915. 11. Patrick JE, Kearns DB (2008) MinJ (YvjD) is a topological determinant of cell division in Bacillus subtilis. Mol Microbiol 70:1166 –1179. 12. Bramkamp M, Emmins R, Weston L, Donovan C, Daniel RA, Errington J (2008) A novel component of the division-site selection system of Bacillus subtilis and a new mode of action for the division inhibitor MinCD. Mol Microbiol 70:1556 –1569. 13. Thomaides HB, Freeman M, El Karoui M, Errington J (2001) Division site selection protein DivIVA of Bacillus subtilis has a second distinct function in chromosome segregation during sporulation. Genes Dev 15:1662–1673. 14. Ben-Yehuda S, Rudner DZ, Losick R (2003) RacA, a bacterial protein that anchors chromosomes to the cell poles. Science 299:532–536. 15. Kearns DB, Chu F, Branda SS, Kolter R, Losick R (2005) A master regulator for biofilm formation by Bacillus subtilis. Mol Microbiol 55:739 –749. 16. Woldringh CL, Mulder E, Huls PG, Vischer N (1991) Toporegulation of bacterial division according to the nucleoid occlusion model. Res Microbiol 142:309 –320. 17. Rahn-Lee L, Gorbatyuk B, Skovgaard O, Losick R (2009) The conserved sporulation protein YneE inhibits DNA replication in Bacillus subtilis. J Bacteriol 191:3736 –3739. 18. Handler AA, Lim JE, Losick R (2008) Peptide inhibitor of cytokinesis during sporulation in Bacillus subtilis. Mol Microbiol 68:588 –599.

Ramamurthi and Losick

PNAS 兩 August 11, 2009 兩 vol. 106 兩 no. 32 兩 13545