![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Journal of Bacteriology, October 2005, p. 6867-6869, Vol. 187, No. 20
0021-9193/05/$08.00+0 doi:10.1128/JB.187.20.6867-6869.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
| GUEST COMMENTARY |
Department of Molecular, Cellular and Developmental Biology, University of Michigan, Ann Arbor, Michigan 48109
The morphological progression of cell division in bacteria takes one of two paths: via a clearly defined division septum or by successive pinching of the cell envelope at the division site. The general features of both types of cell division have been established for decades, primarily through electron microscopic techniques. Unknown, however, were the molecular details, particularly late in cell division. In this issue of the Journal of Bacteriology, McAdams and colleagues present the molecular architecture of late cell division in Caulobacter crescentus using cryoelectron tomography (9). These three-dimensional reconstructions offer a high-resolution view of cell division unattainable using conventional approaches.
Unlike that of Escherichia coli, cell division in C. crescentus results in two distinct cell types, each with its own developmental program. The stalked sessile cell immediately initiates a new round of cell division. The swarmer cell, however, assumes a marauding lifestyle until a cellular cue triggers a developmental program that results in differentiation into a stalked cell. The establishment of the distinct progeny is initiated during growth of the predivisional cell and includes the localization of proteins to specific cell poles, differential transcription from one cell half, and cell-type-specific proteolysis (2, 16-18). A predivision compartmentalization of the daughter cells would ensure that these spatial mechanisms would have an impact on the appropriate cell type. Previous studies have revealed that the two daughter cytoplasms are compartmentalized well before cell division (10). Using a fluorescence loss in photobleaching (FLIP) assay, Judd et al. (9) now show that diffusion of both inner membrane and periplasmic proteins is hindered in a late predivisional cell. Taken together, these studies provide a powerful view of the late stages of C. crescentus cell division.
Cell constriction occurs by progressive pinching.
In order to divide, all bacteria use a similar set of core cell division proteins, yet formation of a septum in E. coli has little resemblance to the successive pinching of the cell envelope seen in C. crescentus. To date, morphological characterization of cell division in C. crescentus has been limited to whole-mount electron micrographs, such as those shown in Fig. 1, or thin-section electron microscopic analysis (15). Although powerful, these approaches are limited in that significant damage to the sample occurs during sample fixation, resin embedding, and staining. Furthermore, images obtained from standard electron microscopy are essentially two-dimensional projections in which information from the entire sample is collapsed into a single plane. Thus, although the general features of cell division have been described, the molecular details were lacking.
|
Judd et al. (9) show that cell division in C. crescentus occurs by successive pinching of the inner membrane, periplasm, and outer membrane. Inner membrane constriction proceeds ahead of outer membrane constriction, resulting in an increase in the spacing between the outer and inner membranes. Remarkably, constriction continues until the two cells are connected by a tubular section on the order of 60 nm in diameter. The proportion of cells visualized with connecting tubes was quite small, indicating that this phase of cell division is rapid. Visualization of these high-resolution "snapshots" demonstrates the power of using a rapid freezing technique to capture infrequent events. Early constriction and division of the inner membrane result in separation of the cytoplasms, with daughter cells remaining tethered by only outer membrane connections for many minutes late in the cell division process.
The divisome poses no insurmountable barrier to diffusion
Cell division in all prokaryotes is mediated by the divisome (also called the Z-ring), a multiprotein complex composed of at least 10 proteins, including cytoplasmic, inner membrane, and periplasmic proteins (8). Division is initiated by the polymerization of the tubulin-like GTPase, FtsZ, on the periphery of the inner membrane at the division site. Additional cell division proteins are recruited, in an interdependent order, to the site of constriction to form the divisome (1). In E. coli, cell division is achieved by inward growth of the cytoplasmic membrane, peptidoglycan, and outer membrane to ultimately generate a septum. In contrast to the formation of a specific septum, cell division in C. crescentus is achieved by a gradual pinching of the cell envelope. Interestingly, however, many of the key divisome proteins are conserved between E. coli and C. crescentus, indicating that despite the apparent differences, the core mechanisms of cell division are probably similar in all bacteria.
In E. coli, with the exception of the abundant FtsZ and ZipA proteins, most of the remaining divisome proteins are present at modest levels (40 to 300 molecules per cell) (8). Although it is unlikely that these membrane and periplasmic proteins form a ring, it is possible that the divisome complex as a whole could effectively serve as a barrier to the diffusion of periplasmic and/or inner membrane proteins during cell division. To determine whether the formation of the divisome impedes diffusion in the C. crescentus predivisional cell, Judd et al. (9) examined the movement of membrane-bound proteins during cell division, using a FLIP assay. This assay relies on the photobleaching of fluorescent molecules that are exposed to a laser pulse. Since diffusion occurs rapidly, the focused application of a laser pulse at one end of an uncompartmentalized C. crescentus cell expressing tdimer2 (a tandem dimer mutant of DsRed) (3) and/or enhanced green fluorescent protein results in the bleaching of all of the fluorescent protein. In a compartmentalized cell, however, only those fluorescent proteins contained within the compartment subjected to the laser pulse are photobleached. In a late predivisional cell, for example, a laser pulse to the distal pole results in loss of fluorescent signal from the cytoplasm of the distal cell compartment but not from the proximal cell cytoplasm (10).
Surprisingly, these studies reveal that in late predivisional cells both the periplasm and inner membrane are also somewhat compartmentalized, as cells were observed in which only the periplasmic or inner membrane proteins within the cell half exposed to the laser were significantly photobleached. Reexamination of these predivisional cells at later times, however, revealed that in a subset of cells, both membrane and periplasmic fluorescent protein migrated through the septal barrier into the adjoining compartment. Thus, it appears that although early in cell division, the divisome apparatus does not impede the movement of either inner membrane or periplasmic proteins. Late in cell division, the divisome acts as a leaky barrier, hindering rapid exchange of proteins between the two cell halves. In E. coli, assembly of the divisome appears to occur in two distinct phases with the formation of the cytoplasmic Z-ring followed by the recruitment of proteins downstream of FtsK (1). One possibility is that a similar two-stage assembly occurs in C. crescentus and the partial diffusion barrier is not established until after the second assembly step is complete.
A few unresolved questions
It is likely that there is a fitness advantage for C. crescentus progeny swarmer cells' maturation prior to their release after cell division. It is well established that the developmental program is initiated in the predivisional cell (2, 16-18). The studies from Judd et al. (9) reveal that late in cell division, the cytoplasms of the daughter cells are completely segregated, whereas the inner membrane and periplasm are only restricted to movement between compartments. How the incomplete compartmentalization of the inner membrane and periplasm can constrain development of the swarmer cell is unknown, particularly since several of the known developmental proteins, such as PleC (21) and PodJ (19), are inner membrane proteins.
Establishment of polarity in C. crescentus is also a poorly understood process. One long-held model proposes that critical cell division proteins are left at the newly formed pole and act as a polar scaffold for subsequent polar localizations (1). The cryoelectron tomographic images of cell division, however, reveal that completely separated cell poles do not posses a gross morphological tag, such as a birth scar (9). Although this observation does not a priori eliminate the possibility that one or more components of the divisome could be retained as a polar tag, clearly there is no structure analogous to the bud scar in yeast (4) that could be utilized as a polar signal. In addition to a cell division polar marker, it is possible that the inert polar peptidoglycan caps generated during division play a role in polarity, as has been shown in E. coli (14). Additionally, the establishment of polarity is complicated by the contributions of the actin-like filament MreB. Recent studies reveal that polarization in C. crescentus could be due to the localized deposition of cargo by actin-like MreB filaments (6, 7, 20). Therefore, although the cryoelectron microscopy studies exclude the existence and contribution of a birth scar, the mechanism by which polarity is established remains unknown. Most likely, C. crescentus uses a number of hierarchical approaches to ensure temporal and spatial localization of polar proteins and complexes.
Having high-resolution images of dividing cells allows for refined models of late cell division events. It is curious that there appear to be two distinct constriction mechanisms: one for the inner membrane and one for the outer membrane (9). Inner membrane constriction likely is a consequence of FtsZ depolymerization. Currently, the mechanism to ensure medial constriction of the outer membrane and subsequent separation is undefined. The presence of temporally distinct inner and outer membrane constriction is particularly interesting as all bacteria utilize the same core cell division proteins, and yet, morphologically, septation has little resemblance to progressive pinching. Differences between the division processes are likely mediated through division-type-specific accessory proteins. Finding these proteins and determining their mechanism of action are clearly a daunting challenge. The clean separation of the two constrictive mechanisms—one for the inner membrane and one for the outer membrane—may permit identification of the molecular composition of each contractile complex. It is clear that optimization of methods for labeling of individual molecular species and spatial visualization by cryoelectron tomography will provide a powerful technique for addressing these and other structural/mechanical questions.
ACKNOWLEDGMENTS
A warm thank-you goes to Susan Sullivan for help with Fig. 1 and for helpful suggestions on the manuscript.
 |
FOOTNOTES |
|---|
FOOTNOTES
The views expressed in this Commentary do not necessarily reflect the views of the journal or of ASM.
REFERENCES
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Appl. Environ. Microbiol. | Infect. Immun. | Eukaryot. Cell |
|---|---|---|
| Mol. Cell. Biol. | J. Virol. | Microbiol. Mol. Biol. Rev. |
| ALL ASM JOURNALS |
