HHS Public Access Author manuscript Author Manuscript
Nat Microbiol. Author manuscript; available in PMC 2017 January 26. Published in final edited form as: Nat Microbiol. ; 1(8): 16121. doi:10.1038/nmicrobiol.2016.121.
Bacterial physiology: Life minus Z Piet A. J. de Boer* Department of Molecular Biology & Microbiology, School of Medicine, Case Western Reserve University, Cleveland, Ohio 44106-4960
Author Manuscript
The cytoplasmic membrane (CM) of almost all bacteria is surrounded by a peptidoglycan (PG) cell wall or sacculus, a cell-sized molecule of glycan strands that are crosslinked by peptide bridges1. One critical function of the sacculus is to withstand turgor pressure and cells typically burst when their PG meshwork is compromised. The sacculus also serves to maintain cell shape and as an anchor for additional cell envelope components, including the outer membrane (OM) in Gram-negative species. Growth and fission of the sacculus during the cell cycle requires trans-envelope complexes of enzymes that produce, modify, or degrade PG, and whose organization and subcellular activities are controlled by cytoskeletal elements on the cytoplasmic face of the CM. In many rod-shaped organisms, including E.coli (Gram negative) and Bacillus subtilis (Gram positive), dynamic filaments of actin-like MreB serve as primary guides for elongation of the sacculus during cell growth, while those of tubulin-like FtsZ orchestrate the whole process of cytokinesis (Fig.1a)1. E. coli cells lacking MreB form walled spheres that can still divide and survive provided the level of FtsZ is sufficiently high (Fig.1b)2.
Author Manuscript Author Manuscript
FtsZ is an ancient protein that directs cytokinesis in almost all Bacteria and Euryarchaeota, as well as plastidial and mitochondrial fission in many eukaryotic species3, 4. In preparation for cell fission, homopolymers of FtsZ accumulate at the future division site (usually midcell) together with protein partners that tether them to the CM. This Z-ring then recruits additional division proteins to form a mature septal ring (SR) organelle or divisome, which subsequently drives the coordinated invagination of the cell envelope layers and sister cell separation. Integral to the SR are specialized PG enzymes that have the onerous tasks of producing an inward growing annulus of septal peptidoglycan that remains continuous with the cylindrical portion of the mother sacculus, and then to cautiously split this layer from the outside inwards to generate the two new hemispherical caps of the nascent sister sacculi. When FtsZ function becomes compromised, the SR cannot form and cells fail to divide. Other cell cycle events are not immediately affected, however, and FtsZ-depleted cells of rod-shaped organisms typically elongate into long multiploid filamentous cells before eventually dying (Fig.1c)1, 3. FtsZ is essential to most bacteria, but not all. Streptomyces naturally grow as branching filamentous syncytia and only need the protein for spore formation5. More tellingly, some or
*
Mailing address: Department of Molecular Biology & Microbiology, School of Medicine, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH 44106-4960. Phone: (216) 368-1697. Fax: (216) 368-3055. [email protected]. Standfirst: The surprising discovery of viable mutants that retain a peptidoglycan cell wall but lack the essential director of normal cytokinesis, FtsZ, reveals that Escherichia coli can proliferate in a completely unexpected manner.
de Boer
Page 2
Author Manuscript Author Manuscript
all representatives of the phyla Tenericutes, Chlamydiae, and Planctomycetes manage life without FtsZ altogether3. Chlamydiae evolved an MreB-based mechanism6, but how other FtsZ-less species divide is largely unclear. Nevertheless, important information on FtsZ-less life has emerged from studies of B.subtilis and E.coli variants that completely lack a PG sacculus. Jeffery Errington and colleagues previously obtained such wall-less (also called Lform) variants by growing cells in isotonic medium, while genetically or chemically blocking the synthesis of PG7, 8, 9. Such cells lose their rod-shape, as expected, but can still be coaxed to propagate, especially after acquiring secondary mutations that promote proliferation in the PG-less state. Remarkably, neither growth nor fission of the L-forms of either organism requires MreB or FtsZ7, 8. Rather, these pleiomorphic cells produce progeny by CM tubulation and blebbing (Fig.1d), which results from the production of more phospholipid than is required to cover the cytoplasm. Such imbalance grows as cells grow larger, eventually inducing spontaneous fission to restore a membrane-surface to cell volume ratio that is energetically more favourable8, 9. This simple biophysical mechanism can also drive spontaneous fission of lipid vesicles in vitro, may well have operated in the first primitive cells prior to the invention of a cell wall, and might contribute to fission of extant Tenericutes that lost the wall again (e.g. Mycoplasma)10.
Author Manuscript Author Manuscript
In this issue of Nature Microbiology, Mercier and co-workers11 spring another surprise. When PG synthesis is restored in their wall-less variants of E.coli or B.subtilis, they readily revert back to life with a sacculus and this is accompanied by the expected changes from fragile L-forms to osmotically stable rods. Inquiring if FtsZ (and, hence, the SR) is required for this de novo generation of the normal rod-shape, they unblocked PG synthesis in wallless ΔftsZ mutant cells of E.coli 11. Interestingly, these cells morphed into non-dividing filaments, as if FtsZ were simply depleted from normal rods (Fig.1c). Thus, de novo formation of a cylindrical PG sacculus of normal diameter does not require the ability to form Z-rings11. Far more surprisingly, extended incubation of such PG-producing ΔftsZ cells allowed for secondary mutations to result in a few viable mutants that are both walled and lack FtsZ. The mere existence of these creatures, termed ‘coli-flower’ (CFL), and their mode of propagation, is wholly unexpected and bizarre. Instead of filamentous rods, timelapse imaging show bulbous CFL cells projecting protrusions that develop into branched syncytial structures of multiple bulbous compartments connected with thin bridges of cellular material. Over time, bulbous parts separate from this ramified structure, by fission or maybe breakage of connecting bridges, and go on to form similar branched structures (Fig. 1e)11. Interestingly, CFL growth minimally requires the absence of FtsZ as well as that of functional PBP1B, one of the PG synthases in the cell. Other factors appear to block the transition from rod-shaped or L-form E.coli to CFL growth, but may be less relevant once coli-flowering is established. These include anchoring of the OM to the PG sacculus by Braun's lipoprotein (Lpp), and biosynthetic pathways for colanic acid (CA) or other extracellular polysaccharides (EP)11. Strikingly, another group recently showed that functional PBP1B, Lpp, and a CA biosynthesis pathway that must be either intact or completely absent, are actually required for E.coli spheroplasts to reform viable rods after enzymatic removal of their original sacculi12, 13. Presumably, metabolic intermediates of EP production interfere with de novo generation and/or maintenance of an intact sacculus, whatever its (eventual) shape. More importantly, FtsZ, PBP1B, and Lpp proteins all appear
Nat Microbiol. Author manuscript; available in PMC 2017 January 26.
de Boer
Page 3
Author Manuscript
to help enforce a normal cylindrical shape on the PG sacculus of E.coli 11, 12, 13. Moreover, either FtsZ or PBP1B activity is sufficient to impose a cylindrical shape on CFL cells11, implying there are at least two pathways to do so. Can FtsZ and PBP1B independently direct the MreB-associated enzyme complexes to enforce a cylindrical shape? If so, how? If not, how do they dictate cell shape? And how is the OM lipoprotein Lpp involved?
Author Manuscript
Other pressing questions concern coli-flowering itself. Ultrastructural studies will be needed to understand the organization of cell components in the ramified CFL structures. Initial pharmacological experiments suggest that, in contrast to wall-less E.coli variants (fig.1d), CFL mutants do require functional MreB for survival (Fig.1e)11. Non-CFL walled cells that lack both MreB and FtsZ form giant non-dividing spheres before dying (Fig.1b)2. It will be interesting to learn if CFL mutants still do so as well, and to further define the role(s) of MreB and other cell shape proteins in their survival. Surprisingly, CFL mutants are also sensitive to drugs that target PBP311, an essential PG-synthase in the SR that is normally dedicated to fission1. Why PBP3 is still required in cells that cannot assemble normal SR's, and clearly do not divide like normal rods, is a new mystery. It also raises the question if other SR components play FtsZ-independent roles in coli-flowering and, perhaps, normal growth as well.
Author Manuscript
But, before we run into the lab, lets reflect a little and stand in awe of the plasticity of bacterial cells. Evidently, just a few mutations are sufficient to result in a completely new mode of propagation that also renders FtsZ non-essential and PBP1B toxic. Who would have thought plain E.coli was this malleable and resourceful? Suddenly, one imagines the natural evolution of walled FtsZ-less bacteria as perhaps a rather pedestrian affair. In fact, it is not so hard to imagine coli-flowering as a clumsy prelude to the budding-mode of propagation common amongst the walled FtsZ-less Planctomycetes14, 15. Allowing CFL mutants to further evolve under various conditions might be enlightening as well as entertaining.
References
Author Manuscript
1. Typas A, Banzhaf M, Gross CA, Vollmer W. Nat Rev Microbiol. 2012; 10:123–136. [PubMed: 22203377] 2. Bendezu FO, de Boer PA. J Bacteriol. 2008; 190:1792–1811. [PubMed: 17993535] 3. Adams DW, Errington J. Nat Rev Microbiol. 2009; 7:642–653. [PubMed: 19680248] 4. Leger MM, et al. Proc Natl Acad Sci USA. 2015; 112:10239–10246. [PubMed: 25831547] 5. McCormick JR, Su EP, Driks A, Losick R. Mol Microbiol. 1994; 14:243–254. [PubMed: 7830569] 6. Jacquier N, Viollier PH, Greub G. FEMS Microbiol Rev. 2015; 39:262–275. [PubMed: 25670734] 7. Leaver M, Dominguez-Cuevas P, Coxhead JM, Daniel RA, Errington J. Nature. 2009; 457:849–853. [PubMed: 19212404] 8. Mercier R, Kawai Y, Errington J. Elife. 2014; 3:e04629. 9. Mercier R, Kawai Y, Errington J. Cell. 2013; 152:997–1007. [PubMed: 23452849] 10. Errington J. Open Biol. 2013; 3:120143. [PubMed: 23303308] 11. Mercier R, Kawai Y, Errington J. Nature Microbiol. 2016 12. Ranjit DK, Young KD. J Bacteriol. 2013; 195:2452–2462. [PubMed: 23543719] 13. Ranjit DK, Young KD. J Bacteriol. 2016; 198:1230–1240. [PubMed: 26833417] 14. Kulichevskaya IS, et al. Front Microbiol. 2012; 3:146. [PubMed: 22529844] 15. Jeske O, et al. Nature Commun. 2015; 6:7116. [PubMed: 25964217]
Nat Microbiol. Author manuscript; available in PMC 2017 January 26.
de Boer
Page 4
Author Manuscript Author Manuscript
Figure 1. Plasticity in E. coli cell shape and proliferation
Author Manuscript
Wildtype cells (a) contain a peptidoglycan (PG) cell wall (red outline) and grow and divide as rods. The cytoskeletal proteins MreB (actin-like, blue) and FtsZ (tubulin-like, green) play critical roles in elongation and fission of cells, respectively. Walled cells lacking MreB (b) grow as spheres that divide in an FtsZ-dependent manner using the same cytokinetic machinery as rods. When FtsZ is additionally lacking, spheres grow very large and then die (✞). Cells lacking just FtsZ (c) form long filamentous rods that eventually die as well. Lform variants (d) lack PG and are osmotically unstable. On isotonic medium they propagate as pleomorphic cells that divide by membrane extrusion/blebbing, and both MreB and FtsZ are dispensable (and, hence, not indicated in the panel). The new E. coli lifestyle (e) called ‘coli-flower’ (CFL) requires the presence of a PG cell wall and functional MreB, as well as the absence of FtsZ and the PG synthase activity of PBP1B. The additional absence of Lpp and biosynthetic pathways for extracellular polysaccharides (EP's) promote the conversion from rod-shaped (a) or L-form (d) cells to CFL (e), but may be less critical once ‘coliflowering’ is established. It is attractive to speculate that forebears of extant FtsZ-less bacterial species that have retained a peptidoglycan cell wall survived in a CFL-like manner upon loss of the ftsZ gene.
Author Manuscript Nat Microbiol. Author manuscript; available in PMC 2017 January 26.
Nat Microbiol. Author manuscript; available in PMC 2017 January 26. Published in final edited form as: Nat Microbiol. ; 1(8): 16121. doi:10.1038/nmicrobiol.2016.121.
Bacterial physiology: Life minus Z Piet A. J. de Boer* Department of Molecular Biology & Microbiology, School of Medicine, Case Western Reserve University, Cleveland, Ohio 44106-4960
Author Manuscript
The cytoplasmic membrane (CM) of almost all bacteria is surrounded by a peptidoglycan (PG) cell wall or sacculus, a cell-sized molecule of glycan strands that are crosslinked by peptide bridges1. One critical function of the sacculus is to withstand turgor pressure and cells typically burst when their PG meshwork is compromised. The sacculus also serves to maintain cell shape and as an anchor for additional cell envelope components, including the outer membrane (OM) in Gram-negative species. Growth and fission of the sacculus during the cell cycle requires trans-envelope complexes of enzymes that produce, modify, or degrade PG, and whose organization and subcellular activities are controlled by cytoskeletal elements on the cytoplasmic face of the CM. In many rod-shaped organisms, including E.coli (Gram negative) and Bacillus subtilis (Gram positive), dynamic filaments of actin-like MreB serve as primary guides for elongation of the sacculus during cell growth, while those of tubulin-like FtsZ orchestrate the whole process of cytokinesis (Fig.1a)1. E. coli cells lacking MreB form walled spheres that can still divide and survive provided the level of FtsZ is sufficiently high (Fig.1b)2.
Author Manuscript Author Manuscript
FtsZ is an ancient protein that directs cytokinesis in almost all Bacteria and Euryarchaeota, as well as plastidial and mitochondrial fission in many eukaryotic species3, 4. In preparation for cell fission, homopolymers of FtsZ accumulate at the future division site (usually midcell) together with protein partners that tether them to the CM. This Z-ring then recruits additional division proteins to form a mature septal ring (SR) organelle or divisome, which subsequently drives the coordinated invagination of the cell envelope layers and sister cell separation. Integral to the SR are specialized PG enzymes that have the onerous tasks of producing an inward growing annulus of septal peptidoglycan that remains continuous with the cylindrical portion of the mother sacculus, and then to cautiously split this layer from the outside inwards to generate the two new hemispherical caps of the nascent sister sacculi. When FtsZ function becomes compromised, the SR cannot form and cells fail to divide. Other cell cycle events are not immediately affected, however, and FtsZ-depleted cells of rod-shaped organisms typically elongate into long multiploid filamentous cells before eventually dying (Fig.1c)1, 3. FtsZ is essential to most bacteria, but not all. Streptomyces naturally grow as branching filamentous syncytia and only need the protein for spore formation5. More tellingly, some or
*
Mailing address: Department of Molecular Biology & Microbiology, School of Medicine, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH 44106-4960. Phone: (216) 368-1697. Fax: (216) 368-3055. [email protected]. Standfirst: The surprising discovery of viable mutants that retain a peptidoglycan cell wall but lack the essential director of normal cytokinesis, FtsZ, reveals that Escherichia coli can proliferate in a completely unexpected manner.
de Boer
Page 2
Author Manuscript Author Manuscript
all representatives of the phyla Tenericutes, Chlamydiae, and Planctomycetes manage life without FtsZ altogether3. Chlamydiae evolved an MreB-based mechanism6, but how other FtsZ-less species divide is largely unclear. Nevertheless, important information on FtsZ-less life has emerged from studies of B.subtilis and E.coli variants that completely lack a PG sacculus. Jeffery Errington and colleagues previously obtained such wall-less (also called Lform) variants by growing cells in isotonic medium, while genetically or chemically blocking the synthesis of PG7, 8, 9. Such cells lose their rod-shape, as expected, but can still be coaxed to propagate, especially after acquiring secondary mutations that promote proliferation in the PG-less state. Remarkably, neither growth nor fission of the L-forms of either organism requires MreB or FtsZ7, 8. Rather, these pleiomorphic cells produce progeny by CM tubulation and blebbing (Fig.1d), which results from the production of more phospholipid than is required to cover the cytoplasm. Such imbalance grows as cells grow larger, eventually inducing spontaneous fission to restore a membrane-surface to cell volume ratio that is energetically more favourable8, 9. This simple biophysical mechanism can also drive spontaneous fission of lipid vesicles in vitro, may well have operated in the first primitive cells prior to the invention of a cell wall, and might contribute to fission of extant Tenericutes that lost the wall again (e.g. Mycoplasma)10.
Author Manuscript Author Manuscript
In this issue of Nature Microbiology, Mercier and co-workers11 spring another surprise. When PG synthesis is restored in their wall-less variants of E.coli or B.subtilis, they readily revert back to life with a sacculus and this is accompanied by the expected changes from fragile L-forms to osmotically stable rods. Inquiring if FtsZ (and, hence, the SR) is required for this de novo generation of the normal rod-shape, they unblocked PG synthesis in wallless ΔftsZ mutant cells of E.coli 11. Interestingly, these cells morphed into non-dividing filaments, as if FtsZ were simply depleted from normal rods (Fig.1c). Thus, de novo formation of a cylindrical PG sacculus of normal diameter does not require the ability to form Z-rings11. Far more surprisingly, extended incubation of such PG-producing ΔftsZ cells allowed for secondary mutations to result in a few viable mutants that are both walled and lack FtsZ. The mere existence of these creatures, termed ‘coli-flower’ (CFL), and their mode of propagation, is wholly unexpected and bizarre. Instead of filamentous rods, timelapse imaging show bulbous CFL cells projecting protrusions that develop into branched syncytial structures of multiple bulbous compartments connected with thin bridges of cellular material. Over time, bulbous parts separate from this ramified structure, by fission or maybe breakage of connecting bridges, and go on to form similar branched structures (Fig. 1e)11. Interestingly, CFL growth minimally requires the absence of FtsZ as well as that of functional PBP1B, one of the PG synthases in the cell. Other factors appear to block the transition from rod-shaped or L-form E.coli to CFL growth, but may be less relevant once coli-flowering is established. These include anchoring of the OM to the PG sacculus by Braun's lipoprotein (Lpp), and biosynthetic pathways for colanic acid (CA) or other extracellular polysaccharides (EP)11. Strikingly, another group recently showed that functional PBP1B, Lpp, and a CA biosynthesis pathway that must be either intact or completely absent, are actually required for E.coli spheroplasts to reform viable rods after enzymatic removal of their original sacculi12, 13. Presumably, metabolic intermediates of EP production interfere with de novo generation and/or maintenance of an intact sacculus, whatever its (eventual) shape. More importantly, FtsZ, PBP1B, and Lpp proteins all appear
Nat Microbiol. Author manuscript; available in PMC 2017 January 26.
de Boer
Page 3
Author Manuscript
to help enforce a normal cylindrical shape on the PG sacculus of E.coli 11, 12, 13. Moreover, either FtsZ or PBP1B activity is sufficient to impose a cylindrical shape on CFL cells11, implying there are at least two pathways to do so. Can FtsZ and PBP1B independently direct the MreB-associated enzyme complexes to enforce a cylindrical shape? If so, how? If not, how do they dictate cell shape? And how is the OM lipoprotein Lpp involved?
Author Manuscript
Other pressing questions concern coli-flowering itself. Ultrastructural studies will be needed to understand the organization of cell components in the ramified CFL structures. Initial pharmacological experiments suggest that, in contrast to wall-less E.coli variants (fig.1d), CFL mutants do require functional MreB for survival (Fig.1e)11. Non-CFL walled cells that lack both MreB and FtsZ form giant non-dividing spheres before dying (Fig.1b)2. It will be interesting to learn if CFL mutants still do so as well, and to further define the role(s) of MreB and other cell shape proteins in their survival. Surprisingly, CFL mutants are also sensitive to drugs that target PBP311, an essential PG-synthase in the SR that is normally dedicated to fission1. Why PBP3 is still required in cells that cannot assemble normal SR's, and clearly do not divide like normal rods, is a new mystery. It also raises the question if other SR components play FtsZ-independent roles in coli-flowering and, perhaps, normal growth as well.
Author Manuscript
But, before we run into the lab, lets reflect a little and stand in awe of the plasticity of bacterial cells. Evidently, just a few mutations are sufficient to result in a completely new mode of propagation that also renders FtsZ non-essential and PBP1B toxic. Who would have thought plain E.coli was this malleable and resourceful? Suddenly, one imagines the natural evolution of walled FtsZ-less bacteria as perhaps a rather pedestrian affair. In fact, it is not so hard to imagine coli-flowering as a clumsy prelude to the budding-mode of propagation common amongst the walled FtsZ-less Planctomycetes14, 15. Allowing CFL mutants to further evolve under various conditions might be enlightening as well as entertaining.
References
Author Manuscript
1. Typas A, Banzhaf M, Gross CA, Vollmer W. Nat Rev Microbiol. 2012; 10:123–136. [PubMed: 22203377] 2. Bendezu FO, de Boer PA. J Bacteriol. 2008; 190:1792–1811. [PubMed: 17993535] 3. Adams DW, Errington J. Nat Rev Microbiol. 2009; 7:642–653. [PubMed: 19680248] 4. Leger MM, et al. Proc Natl Acad Sci USA. 2015; 112:10239–10246. [PubMed: 25831547] 5. McCormick JR, Su EP, Driks A, Losick R. Mol Microbiol. 1994; 14:243–254. [PubMed: 7830569] 6. Jacquier N, Viollier PH, Greub G. FEMS Microbiol Rev. 2015; 39:262–275. [PubMed: 25670734] 7. Leaver M, Dominguez-Cuevas P, Coxhead JM, Daniel RA, Errington J. Nature. 2009; 457:849–853. [PubMed: 19212404] 8. Mercier R, Kawai Y, Errington J. Elife. 2014; 3:e04629. 9. Mercier R, Kawai Y, Errington J. Cell. 2013; 152:997–1007. [PubMed: 23452849] 10. Errington J. Open Biol. 2013; 3:120143. [PubMed: 23303308] 11. Mercier R, Kawai Y, Errington J. Nature Microbiol. 2016 12. Ranjit DK, Young KD. J Bacteriol. 2013; 195:2452–2462. [PubMed: 23543719] 13. Ranjit DK, Young KD. J Bacteriol. 2016; 198:1230–1240. [PubMed: 26833417] 14. Kulichevskaya IS, et al. Front Microbiol. 2012; 3:146. [PubMed: 22529844] 15. Jeske O, et al. Nature Commun. 2015; 6:7116. [PubMed: 25964217]
Nat Microbiol. Author manuscript; available in PMC 2017 January 26.
de Boer
Page 4
Author Manuscript Author Manuscript
Figure 1. Plasticity in E. coli cell shape and proliferation
Author Manuscript
Wildtype cells (a) contain a peptidoglycan (PG) cell wall (red outline) and grow and divide as rods. The cytoskeletal proteins MreB (actin-like, blue) and FtsZ (tubulin-like, green) play critical roles in elongation and fission of cells, respectively. Walled cells lacking MreB (b) grow as spheres that divide in an FtsZ-dependent manner using the same cytokinetic machinery as rods. When FtsZ is additionally lacking, spheres grow very large and then die (✞). Cells lacking just FtsZ (c) form long filamentous rods that eventually die as well. Lform variants (d) lack PG and are osmotically unstable. On isotonic medium they propagate as pleomorphic cells that divide by membrane extrusion/blebbing, and both MreB and FtsZ are dispensable (and, hence, not indicated in the panel). The new E. coli lifestyle (e) called ‘coli-flower’ (CFL) requires the presence of a PG cell wall and functional MreB, as well as the absence of FtsZ and the PG synthase activity of PBP1B. The additional absence of Lpp and biosynthetic pathways for extracellular polysaccharides (EP's) promote the conversion from rod-shaped (a) or L-form (d) cells to CFL (e), but may be less critical once ‘coliflowering’ is established. It is attractive to speculate that forebears of extant FtsZ-less bacterial species that have retained a peptidoglycan cell wall survived in a CFL-like manner upon loss of the ftsZ gene.
Author Manuscript Nat Microbiol. Author manuscript; available in PMC 2017 January 26.
