Research Article
15 August 2010

Changes in the Min Oscillation Pattern before and after Cell Birth


The Min system regulates the positioning of the cell division site in many bacteria. In Escherichia coli, MinD migrates rapidly from one cell pole to the other. In conjunction with MinC, MinD helps to prevent unwanted FtsZ rings from assembling at the poles and to stabilize their positioning at midcell. Using time-lapse microscopy of growing and dividing cells expressing a gfp-minD fusion, we show that green fluorescent protein (GFP)-MinD often paused at midcell in addition to at the poles, and the frequency of midcell pausing increased as cells grew longer and cell division approached. At later stages of septum formation, GFP-MinD often paused specifically on only one side of the septum, followed by migration to the other side of the septum or to a cell pole. About the time of septum closure, this irregular pattern often switched to a transient double pole-to-pole oscillation in the daughter cells, which ultimately became a stable double oscillation. The splitting of a single MinD zone into two depends on the developing septum and is a potential mechanism to explain how MinD is distributed equitably to both daughter cells. Septal pausing of GFP-MinD did not require MinC, suggesting that MinC-FtsZ interactions do not drive MinD-septal interactions, and instead MinD recognizes a specific geometric, lipid, and/or protein target at the developing septum. Finally, we observed regular end-to-end oscillation over very short distances along the long axes of minicells, supporting the importance of geometry in MinD localization.
Rod-shaped bacteria, such as Escherichia coli, divide by binary fission and thus assemble their cell division apparatus (the divisome) at the cell midpoint. Tubulin-like FtsZ is the major cytoskeletal protein of the divisome (17) and assembles into a polymeric ring on the inner surface of the cytoplasmic membrane (the Z ring). Assembly and eventual contraction of the Z ring are crucial for divisome function, and thus it is not surprising that many regulatory factors control FtsZ assembly (25). Notably, two negatively acting spatial regulatory systems, the Min system and nucleoid occlusion, ensure that the Z ring is located properly at the cell midpoint (18). Whereas a major component of the nucleoid occlusion system can be deleted with no major effects on cell division (2), inactivation of the Min system causes cells to divide either at midcell or aberrantly at cell poles (27). The result of polar cell division is the formation of chromosome-free minicells.
The Min system consists of three proteins, MinC, MinD, and MinE (7). MinC has two separate domains, each of which binds to FtsZ and promotes disassembly of FtsZ polymers and polymer bundles (6, 29, 30). MinC also binds to MinD, an ATPase with a carboxy-terminal amphipathic helix that binds to the membrane only when the protein is bound to ATP (11, 12). MinD also forms polymers (31). Finally, MinE is a small protein that binds to MinD and stimulates hydrolysis of its bound ATP in the presence of membranes. By doing so, MinE helps to dislodge MinD from the membrane, although MinE itself can bind to the membrane (10). The result is that MinD and MinE form zones that oscillate from one cell pole to the other, with an oscillation period of seconds to minutes, depending on a number of factors, including temperature (9, 23, 24, 34). In typical cells, MinD spends most of its time bound to the membrane at a cell pole, forming a U-shaped zone, and its transit to the opposite pole is rapid compared to its dwell time (23). MinE typically forms a ring at the edge of the MinD zone (22, 24). The direction of the oscillation is determined strongly by cell geometry (5, 35). Other factors, such as membrane phospholipid composition, also influence MinD oscillation; MinD-ATP preferentially binds anionic phospholipids, such as cardiolipin, which is enriched at cell poles (15, 21, 32).
Because MinC binds to MinD, MinC oscillates in concert with MinD and therefore is present at the cell poles for longer times than anywhere else in the cell (13, 22). This sets up a gradient of MinC, with the average smallest amount of MinC at midcell at any one time. The current model is that Z rings are most likely to assemble at the trough of the MinC gradient and are discouraged from assembling at cell poles at the peak of the gradient (14). This is supported by the observation that nonring FtsZ itself oscillates from pole to pole, presumably being chased back and forth by the alternating zones of high MinC concentration (33).
However, recent work in Bacillus subtilis has shed new light on the possible function of MinC on the Z ring and the divisome. B. subtilis lacks MinE and thus relies on a static MinC gradient. This is set up by the recruitment of MinC and MinD (MinCD) to the Z ring during formation of the division septum (19, 20). This seems paradoxical, as the presence of MinCD at the Z ring is predicted to destabilize it. However, in B. subtilis, Z rings containing MinCD remain functional. Therefore, MinCD seems to have an important role in preventing the immediate reassembly of Z rings at developing cell poles next to a recently used ring (4, 8).
This recruitment of MinCD to the Z ring of B. subtilis prompted us to examine in more detail Min oscillations in E. coli cells undergoing septation. We hypothesized that MinCD might bind to the Z ring at later stages of septation, perhaps helping the Z ring to function by stimulation of FtsZ disassembly. Previous results with green fluorescent protein (GFP)-MinC suggested that MinC could transiently localize to the Z ring during septation (13). Consequently, we tested if MinD, the driving force of the oscillation, could also localize to the Z ring and if this localization was dependent on MinC. We also hypothesized that a more central localization of MinCD during the time of septum formation might explain how Min proteins are partitioned equitably to both daughter cells.


Strains and plasmids.

Strain WM1264 (5) is wild-type W3110 containing P lac , driving expression of a gfp-minD fusion, with minE downstream of minD in its native genetic context relative to minE. The entire construct is on a lambda InCh vector at the lambda attachment site (3). To make strain WM3149, WM1264 was transduced to Kanr with P1 phage carrying ΔminCDE::kan; the resulting strain is functionally Min, despite the expression of gfp-minD and minE, because of the lack of minC. To shorten the longer cells of WM3149, we introduced plasmid pZAQ, a Tetr pBR322 derivative carrying the ftsQAZ region from the E. coli chromosome (36). The medium copy number of this plasmid results in severalfold-excess expression of the ftsQ, ftsA, and ftsZ genes. In WM3149, and especially in WM1264, where it was introduced as a control, pZAQ induces polar Z rings and minicells at a high frequency, as expected. Importantly, this plasmid also induces more frequent midcell divisions (1), thus increasing the number of short cells in the population of WM3149.

Growth conditions and microscopy.

A single colony was inoculated in Luria-Bertani (LB) medium with 50 to 100 μM IPTG (isopropyl-β-d-thiogalactopyranoside) and grown at 30°C or 37°C until mid-logarithmic phase. Three microliters of culture was then placed on a cover glass on a glass-bottomed culture dish (MatTek Corp.) and overlaid with a thin strip of 1.5% LB agarose. Time-lapse microscopy was done with a fully automated Olympus IX-71 microscope outfitted with a 100× objective (numerical aperture, 1.4) and filter wheels, with exposure intervals from 1.5 to 2 s and exposure times of 0.5 to 0.8 s with a 488-nm excitation light. The growth temperature on the slide was maintained at 32 to 33°C with a WeatherStation enclosure. Identical intervals and exposure times were used during any given time course. At these temperatures, the GFP-MinD oscillation period was often as fast as 10 s (34), and cells grew significantly (0.5 to 1 μm) and sometimes divided during the time courses, which were mostly 6 to 8 min in length.

Image processing.

Images were captured and analyzed with SlideBook 5.0 software (Intelligent Imaging Innovations). Kymographs were obtained by drawing a linear mask region along the length of a cell and then performing a smooth-curve analysis of that mask. Color intensity graphs were obtained by drawing four contiguous squares of identical sizes that encompassed the pole, left midcell, right midcell, and opposite pole and by quantitating total fluorescence in those squares over time. Midcell pausing data were compiled for each time course and analyzed in Microsoft Excel. To measure the frequency of GFP-MinD pausing at the midpoint of the cell, the number of midcell pauses for each cell was counted for the duration of the time course and then the number of pauses per 5 min was plotted versus cell length. Some cells divided during the time course, becoming two new cells. Therefore, only the length of time prior to cell division was used to normalize the frequency of GFP-MinD midcell pausing. Cell lengths were measured at the beginning and end of each time course; for simplicity, only the starting lengths were used in the analysis. For comparison of partitionings into sister cells, the total intensities of GFP-MinD in 37 pairs of newborn cells were calculated with ImageJ software and analyzed with Microsoft Excel.


MinD pauses often and asymmetrically at the septum in dividing cells.

To study in detail the behavior of MinD protein during cell division, we used WM1264, a derivative of W3110 containing a lac promoter driving single-copy expression of gfp-minD along with the minE gene in its natural context downstream of minD. When expressed at low levels at 32 to 34°C on the surface of LB agarose, mobile fluorescent zones of MinD were detectable at exposures of several hundred milliseconds. Most cells grew and divided normally over the time courses, which were typically 6 to 10 min, with time intervals of 1.5 to 2 s between exposures.
As expected, GFP-MinD alternately and repeatedly assembled into a U-shaped zone at each pole in nearly all short (1.5 μm to ∼3 μm in length) WM1264 cells that were not in the process of dividing. This regular oscillation is depicted as a kymograph, which plots fluorescence localization versus time (Fig. 1A). Whereas about half of the cells in this size class displayed pole-to-pole oscillations only, the other half displayed a transient, membrane-bound collection, or pause, of GFP-MinD molecules at least once at midcell during the time course in addition to the polar zones. The time course of a cell displaying midcell pausing is shown as a kymograph (Fig. 1B) and, for another cell undergoing septation, as a series of representative micrographs from a time course experiment (Fig. 2, horizontal arrows).
As cells increased in length (to >3 μm) and developed visible constrictions at the division sites, the proportion of cells exhibiting at least one midcell pause of GFP-MinD during a time course increased until all cells examined showed some midcell pausing during the time course (Fig. 3A). This midcell pausing usually involved a polar U-shaped zone that alternated with assembly of a U-shaped zone at the developing septum (Fig. 1B, arrowheads, and 2; see also the movie in the supplemental material). Like the polar zones, these septal zones usually persisted for only a few seconds, but interestingly, they were usually asymmetric, appearing only at one side of the septum at a time.
While analyzing kymographs of numerous time-lapse movies of cells at late stages of septation, we asked if this septal localization of GFP-MinD was always followed by polar localization and whether consecutive localizations favored one side of the dividing cell over the other. One prediction was that if a GFP-MinD zone started at the left cell pole, then its next zone would more likely be the left side of the septum. However, we often observed zones that migrated from one side of the septum immediately to the opposite cell pole or to roughly the middle of the opposite cell, as well as zones that migrated from one side of the septum directly to the other side of the septum, bypassing an immediate return to a pole (Fig. 1B and 2).
Importantly, the frequency of pausing at the constricting septum steadily increased in the population as the cells elongated (Fig. 3B). With an average of 6 total pauses per minute in constricting cells, the percentage of total midcell pauses reached ∼50% (15 pauses per 5 min, or 3 pauses per min). Likewise, as cells approached division, the percentage of time that GFP-MinD dwelled at septal versus polar locations increased to about half of the total dwell time (data not shown).

GFP-MinD switches its oscillation pattern near the time of septum closure.

To confirm the above observations, we tracked GFP-MinD movement in cells that seemed to complete their division septa during the several-minute time course. In cells with pronounced constrictions, there was frequent pausing of GFP-MinD at one side of the septum and occasional localization to other nonpolar positions. Often, this irregular pattern switched to a transient, symmetric, doubled septum-to-pole pattern that then reverted to the previous single, irregular oscillation before switching again into a regular, stable, doubled septum-to-pole oscillation in each of the daughter cells. Two representative cells are shown in Fig. 4, either as kymographs (Fig. 4A and C) or as fluorescence intensity profiles (Fig. 4B and D). Cells at this point had very deep constrictions, as viewed by differential interference contrast (DIC) microscopy, consistent with full or almost complete daughter cell separation, and the cells could be seen pushing apart (Fig. 4A and C). The transient splitting of septal MinD into two separate polar zones can be clearly seen as an overlap between the two polar intensity peaks (green and blue waves, indicated by arrowheads in Fig. 4B and D).
These results strongly suggest that in constricting cells, irregular oscillation with frequent midcell pauses is a precursor to a transient doubled oscillation between the two new and two old cell poles, which can revert to the irregular pattern but soon leads to regular, doubled oscillation around the time of septum closure. As described above, this pole-to-pole oscillation persisted for a while in short cells; GFP-MinD pausing at midcell was not observed in newborn cells (Fig. 2, vertical arrows, and 3A and B). As newborn cells grew, GFP-MinD began to pause occasionally at the cell midpoint (Fig. 4E). This switch from midcell pausing to a double septum-to-pole oscillation may prove to be a useful assay in future studies of protein localization and distribution after daughter cell separation.

Midcell pausing as a mechanism to distribute MinD to daughter cells.

These data also address the mechanism by which daughter cells receive equivalent amounts of the Min proteins. If cell separation were to occur when MinD is oscillating solely between poles of the mother cell, then the probability that one daughter cell would receive most or all of the Min proteins, at least MinD, would be high. For example, a MinD zone at the top cell pole at the moment of septum closure would result in the top daughter cell receiving most of the MinD, while the bottom daughter cell would theoretically receive little or none.
However, the absence of minicells in wild-type strains suggests that distribution of MinD must be generally symmetric. Consequently, if MinD oscillation doubled in the mother cell during the final seconds of septum closure, it would significantly increase the probability of both cells receiving MinD after septum closure. Intensity plots of the top poles, top septa, bottom septa, and bottom poles of the cells depicted in kymographs in Fig. 4A and C indicate that equivalent amounts of GFP-MinD were indeed transferred to both daughter cells at about the time of septum closure (Fig. 4B and D). We also determined that GFP-MinD fluorescence was partitioned roughly equally to daughter cells in a number of other cells that divided during time course experiments (Fig. 3C).
If MinD is equally partitioned to daughter cells because of a divided population of MinD undergoing doubled oscillations during the moment of septum closure, then MinE would also be distributed and oscillations would precede the formation of daughter cells. Moreover, these oscillations should be in synchrony, at least temporarily. The overlapping peaks to the right of the arrow in Fig. 4B show that doubled oscillations were synchronous and remained so for the rest of the time course. We observed many other instances of the switch to stable pole-to-pole oscillations in daughter cells that continued after cell separation. One such cell is depicted in Fig. 4D, although in this case synchrony was lost after a few oscillations. Figure 2 (horizontal arrows) shows another cell with similar MinD behavior that clearly divided during the time course (compare DIC images before and after the time course). These results demonstrate that MinD is distributed fairly equally to daughter cells upon septum closure and suggest that the same must be true for MinE.

MinC is not required for midcell pausing of GFP-MinD.

A previous study of MinC localization in E. coli showed that GFP-MinC was sometimes observed oscillating from a cell pole to one side of the division septum during late stages of septation (13), consistent with our observations for GFP-MinD. As MinC binds to FtsZ and to MinD, increased interactions between MinC and FtsZ at the constricting Z ring might be a mechanism to explain the visible pausing of MinC or MinD at the developing septum. In support of this idea, B. subtilis MinC, which does not oscillate, is recruited to closing cell division septa by DivIVA and MinJ (4, 8). The presence of MinC at the septum in B. subtilis is thought to prevent immediate reassembly of adjacent Z rings, potentially acting as a licensing factor for new divisome formation.
It is not yet clear if MinC has a similar role in E. coli. However, if it does, then we might expect the pausing of GFP-MinD at invaginating septal membranes to be dependent on a transient population of MinC localized there. To test this idea, we repeated our time-lapse studies with GFP-MinD in WM3149, a derivative of WM1264 carrying a complete deletion of the minCDE locus. WM3149 cells were a mixture of short filaments and minicells, typical of a Δmin mutant. This pattern results from the limited use of Z rings either for polar divisions, which generate chromosome-free minicells, or for nonpolar divisions, which generate two viable daughter cells.
The consequence of infrequent nonpolar division is that fewer cells actively divide at any given time, making it much more difficult to find cells at the proper late stage of septum formation. Furthermore, many of these cells were significantly longer than those of WM1264 because of deficiencies in nonpolar division. Therefore, it became much less likely that we would observe a cell of normal length with a simple pole-to-pole oscillation, because the oscillation wavelength of GFP-MinD (∼7 μm) was often shorter than the cell length (23). Indeed, many Δmin cells had three nodes of peak MinD accumulation—the two poles and midcell. Nonetheless, we were able to analyze a few dividing WM3149 cells, and GFP-MinD still paused at invaginating division septa in these cells, with similar asymmetric patterns (data not shown). As MinC is completely lacking in these cells, this suggested that MinC is not required for midcell pausing of GFP-MinD.
To confirm this idea in a larger number of cells, we introduced pZAQ, a multicopy plasmid that expresses the native ftsQAZ gene cluster in levels severalfold above native levels. The overproduction of FtsA and FtsZ from this plasmid partially overcomes the nonpolar division block in Δmin cells (1). As a result, cells of WM3149 plus pZAQ (WM3454) were on average shorter than Δmin cells, although they still produced polar minicells at high frequency and were longer than min + WM1264 cells.
Importantly, WM3454 cells displayed the same types of oscillation patterns as WM1264 cells, including exclusively pole-to-pole oscillations in the shortest cells (Fig. 5A) and midcell pausing in longer cells (Fig. 5B). As with WM1264, the proportion of cells with GFP-MinD pausing at future midcell division sites in WM3454 increased as cell length increased (Fig. 5C), and the frequency of midcell pausing steadily increased as cells elongated and approached cytokinesis (Fig. 5D). The pausing was not solely a result of the higher average lengths of these cells, because many cells in the normal range of 4 to 5 μm also exhibited midcell pausing. From these data, we can conclude that pausing of GFP-MinD at midcell does not require MinC.

MinD oscillation in minicells.

The changes in the GFP-MinD oscillation pattern with cell length prompted us to ask if cells with very short lengths also exhibited pole-to-pole oscillations. In general, minicells formed by WM3149 were quite symmetrical, and we did not detect a concerted motion of GFP-MinD in these minicells (data not shown). However, WM3149 plus pZAQ and especially WM1264 plus pZAQ tended to generate minicells with diverse sizes and shapes, including double minicells. Although the mechanism behind this shape diversity is not known, it is well established that severalfold overproduction of FtsZ increases polar divisions, both in min mutant and min + cells, because the excess FtsZ antagonizes the action of the Min proteins (36).
Remarkably, many of these oblong or double minicells exhibited stable bidirectional oscillation (Fig. 6). The direction of the oscillation was always along the long axis, consistent with our previous geometrical model for Min oscillation (5). Long axes were generally between 1.5 and 2 μm in length. Therefore, the oscillation pattern in these minicells is similar to that in newborn cells with a normal rod shape. These results indicate not only that these minicells have sufficient ATP to drive the oscillation but also that pole-to-pole oscillation can be robust even over very short distances, provided there is a long axis.


We show here that GFP-MinD, and by inference MinD itself, forms transient assemblies not only at cell poles but also at the developing division septum. The frequency of pausing at the septum steadily increased as the cells elongated and the septum matured. Although septal pausing of GFP-MinD rarely occurred in newborn cells, over the course of many oscillations, we detected at least one septal pause in approximately half of cells shorter than 3 μm in length. This argues against the idea that the septal pausing is simply a result of longer distances that allow a doubling of the oscillation wave, because the wavelength interval between MinD segments in filamentous cells lacking division septa generally exceeds 7 μm (23).
Moreover, the morphology of the septal pausing suggests that MinD recognizes a specific cue at the septum. In cells with deep constrictions, GFP-MinD often localized at a U-shaped zone corresponding to one side of the division septum. This was followed either by reassembly at a cell pole or by migration to the other side of the septum, forming a corresponding inverse U shape. This is consistent with GFP-MinD recognizing the developing septum as a new cell pole. The ability of GFP-MinD to move from one side of the septum to the other instead of always back to an old pole implies that the new cell poles become competent targets for MinD assembly and compete effectively with the old cell poles. This also indicates that once its bound ATP is hydrolyzed by MinE, MinD does not always have to move several microns away from its previous assembly point to assemble a new zone.
We also documented the transition of GFP-MinD oscillation from an undivided to a divided cell. The data indicate that the irregular oscillation pattern, which includes extensive septal pausing, ultimately switches to a stable double oscillation around the time of septum closure (Fig. 7). This can occur in several ways, as illustrated in the model. Importantly, the affinity of GFP-MinD for the developing septum is crucial for any of these pathways to doubling and therefore for the symmetric partitioning of MinD into daughter cells. Our data are consistent with mathematical simulations of this process (28), in which the highest probability of correct partitioning of Min proteins into daughter cells correlates with frequent localization to the developing division septum. It will also be interesting to understand how the distribution of MinE is coordinated with MinD.
As MinC interacts with FtsZ and MinD, one explanation for septal pausing is that interaction between MinC and the Z ring might cause MinD to stall transiently at the Z ring during its rapid transit from pole to pole. Indeed, MinC was also previously shown to pause at the septum (13), and the idea that MinC helps the Z ring to disassemble is attractive. However, we found that GFP-MinD often pauses midcell in cells lacking MinC. This suggests that MinC-FtsZ interactions are not required for the equal distribution of MinD to daughter cells, arguing against a model in which MinC is the sole trigger of septal closure (28), although MinC has a role in activating divisome function in general, as many Z rings are not used right away in the absence of Min (37).
Our data suggest that MinD itself may recognize a membrane determinant for developing a new pole that competes with its old-pole-to-old-pole regime. We propose that this determinant may be membrane curvature induced by the invaginating Z ring. This explains the high frequency of midcell pauses in cells with deep constrictions and is consistent with the behavior of proteins, such as DivIVA, that localize to areas of high membrane curvature (16). However, cells with no visible constrictions also exhibited midcell pausing. One possibility is that once the Z ring is assembled and activated, it can induce small shape changes in the cytoplasmic membrane, possibly in association with penicillin-insensitive peptidoglycan synthesis (26), and MinD can sense these changes. Another possibility is that MinD may target anionic phospholipids such as cardiolipin, which are enriched at cell poles and division septa. If this is true, then MinD still can oscillate robustly in minicells, which are enriched for these lipids, if the cell geometry is asymmetric. A third possibility is that MinD may bind directly to a divisome protein. Further studies will be needed to distinguish among these models, although it will be difficult to separate the effects of phospholipid composition from membrane curvature, as one is correlated with the other. In any case, we postulate that MinD pauses at the developing division septum for two purposes—to stimulate Z ring contraction and to ensure its equal distribution to daughter cells. Work is in progress to confirm these ideas.
FIG. 1.
FIG. 1. GFP-MinD pauses at midcell. Space-time kymographs and DIC images of representative cells of WM1264 are shown, with scale bars indicating cell length (1 μm, horizontal bar) and time (1 min, vertical bar). (A) Kymograph of a nonconstricting cell exhibiting pole-to-pole oscillation of GFP-MinD without midcell pausing. The start time is at the top. (B) Kymograph of a constricting cell exhibiting GFP-MinD midcell pausing. Arrowheads indicate two consecutive GFP-MinD midcell pausing events.
FIG. 2.
FIG. 2. Micrographs of individual cells expressing GFP-MinD. The results of a 10-min time-lapse experiment are shown, with DIC images at the beginning and end. Horizontal arrows indicate a predivisional cell showing midcell pausing with eventual doubling and cell division, and vertical arrows indicate a postdivisional cell undergoing regular pole-to-pole oscillations. Images are shown at 12-s intervals (except between 1 min 27 s and 1 min 33 s) until 4 min 57 s, at which point the intervals are every 6 s until 5 min 51 s to highlight the initiation of GFP-MinD doubling in the dividing cell, and irregular intervals are shown afterward to maximize the visibility of the oscillation. Scale bar, 1 μm.
FIG. 3.
FIG. 3. Increased GFP-MinD midcell pausing as septation approaches and equal distributions of GFP-MinD in daughter cells. (A) Frequencies of nonconstricting or constricting cells displaying no midcell pausing or some midcell pausing are shown. (B) The frequencies of GFP-MinD pausing at midcell in either nonconstricting or constricting cells are plotted. The frequencies were calculated from time-lapse experiments and normalized to midcell pauses at 5-min intervals. Open symbols represent cells that had visible constrictions in DIC images. Filled symbols represent cells that had no visible constrictions in DIC images. (C) Graph of the ratio of GFP-MinD intensities in each pair of recently divided sister cells.
FIG. 4.
FIG. 4. Switch in GFP-MinD oscillation patterns near the moment of septum closure. (A and C) Kymographs show GFP-MinD oscillation patterns in dividing cells. DIC images of cells before and after each time course are shown on each side of the kymographs. About the time of septation, the irregular oscillation pattern of GFP-MinD pausing at midcell switches to a double septum-to-pole pattern and then back to an irregular pattern with midcell pauses and again to a regular, doubled septum-to-pole pattern. Scale bars represent 1 μm (vertical) and 1 min (horizontal). (B and D) Corresponding graphs of GFP intensities in four regions of the cell versus time. Green, region 1, bottom cell pole; orange, region 2, lower side of the septum; purple, region 3, top side of the septum; blue, region 4, top cell pole. Near the time of complete septation, orange and purple waves synchronize, as do blue and green waves. Arrows indicate the switch to a regular, doubled pattern. Arrowheads indicate a transient doubled pattern before it becomes stable. (E) Kymograph and DIC images of a recently born cell that shows some GFP-MinD midcell pausing, indicated by the arrows. Scale bars are the same as in panels A and C.
FIG. 5.
FIG. 5. GFP-MinD oscillation in Δmin cells expressing extra FtsQAZ. (A and B) Kymographs of representative cells of WM3454 are shown. Scale bars indicate 1 μm (horizontal) and 1 min (vertical). (A) Kymograph of a nonconstricting cell exhibiting pole-to-pole oscillation of GFP-MinD without midcell pausing. (B) Kymograph of a constricting cell exhibiting GFP-MinD midcell pausing. Arrowheads indicate two consecutive GFP-MinD midcell pausing events. (C) Frequencies of nonconstricting or constricting cells displaying no midcell pausing or some midcell pausing are shown. (D) The frequencies of GFP-MinD pausing at midcell in either nonconstricting or constricting cells are plotted. The frequencies were calculated from time-lapse experiments and normalized per 5-min intervals. Open symbols represent cells that had visible constrictions in DIC images. Filled symbols represent cells that had no visible constrictions in DIC images.
FIG. 6.
FIG. 6. MinD oscillates along a long axis in minicells. Kymographs show GFP-MinD oscillation over time in representative minicells of WM1264 containing pZAQ, featuring either a double minicell (A) or an oblong minicell (B). DIC images of the minicells are shown to the left. The scale bar under the micrograph represents 1 μm, and the scale bar under the kymograph represents 1 min.
FIG. 7.
FIG. 7. Model of MinD oscillation patterns before and after cell division. The regular pole-to-pole oscillation of a newborn cell (A) switches to an irregular pattern of a single MinD zone, with pausing not only at cell poles but also at either side of the developing septum (B). In cells approaching cytokinesis, a single MinD zone at the septum (C) or cell pole (D) will split into two zones and initiate a stable, doubled oscillation in both daughter cell compartments. Sometimes the doubled oscillation pattern reverts to the irregular pattern with only a single MinD zone per mother cell (E), although a stable doubling pattern will ultimately emerge (F). Each panel represents different pathways toward stable doubling observed over the course of many time-lapse experiments. The affinity of MinD for the developing septum is crucial for each pathway, which leads to symmetric partitioning of MinD into daughter cells.


This work was supported by National Institutes of Health grant R01GM61074.

Supplemental Material

File (
ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.


Begg, K., Y. Nikolaichik, N. Crossland, and W. D. Donachie. 1998. Roles of FtsA and FtsZ in activation of division sites. J. Bacteriol. 180 : 881-884.
Bernhardt, T. G., and P. A. de Boer. 2005. SlmA, a nucleoid-associated, FtsZ-binding protein required for blocking septal ring assembly over chromosomes in E. coli. Mol. Cell 18 : 555-564.
Boyd, D., D. S. Weiss, J. C. Chen, and J. Beckwith. 2000. Towards single-copy gene expression systems making gene cloning physiologically relevant: lambda InCh, a simple Escherichia coli plasmid-chromosome shuttle system. J. Bacteriol. 182 : 842-847.
Bramkamp, M., R. Emmins, L. Weston, C. Donovan, R. A. Daniel, and J. Errington. 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.
Corbin, B. D., X.-C. Yu, and W. Margolin. 2002. Exploring intracellular space: function of the Min system in round-shaped Escherichia coli. EMBO J. 21 : 1988-2008.
Dajkovic, A., G. Lan, S. X. Sun, D. Wirtz, and J. Lutkenhaus. 2008. MinC spatially controls bacterial cytokinesis by antagonizing the scaffolding function of FtsZ. Curr. Biol. 18 : 235-244.
de Boer, P. A. J., R. E. Crossley, and L. I. Rothfield. 1989. A division inhibitor and a topological specificity factor coded for by the minicell locus determine the proper placement of the division site in Escherichia coli. Cell 56 : 641-649.
Gregory, J. A., E. C. Becker, and K. Pogliano. 2008. Bacillus subtilis MinC destabilizes FtsZ-rings at new cell poles and contributes to the timing of cell division. Genes Dev. 22 : 3475-3488.
Hale, C. A., H. Meinhardt, and P. A. de Boer. 2001. Dynamic localization cycle of the cell division regulator MinE in Escherichia coli. EMBO J. 20 : 1563-1572.
Hsieh, C. W., T. Y. Lin, H. M. Lai, C. C. Lin, T. S. Hsieh, and Y. L. Shih. 2010. Direct MinE-membrane interaction contributes to the proper localization of MinDE in E. coli. Mol. Microbiol. 75 : 499-512.
Hu, Z., E. P. Gogol, and J. Lutkenhaus. 2002. Dynamic assembly of MinD on phospholipid vesicles regulated by ATP and MinE. Proc. Natl. Acad. Sci. U. S. A. 99 : 6761-6766.
Hu, Z., and J. Lutkenhaus. 2003. A conserved sequence at the C-terminus of MinD is required for binding to the membrane and targeting MinC to the septum. Mol. Microbiol. 47 : 345-355.
Hu, Z., and J. Lutkenhaus. 1999. Topological regulation of cell division in Escherichia coli involves rapid pole to pole oscillation of the division inhibitor MinC under the control of MinD and MinE. Mol. Microbiol. 34 : 82-90.
Huang, K. C., Y. Meir, and N. S. Wingreen. 2003. Dynamic structures in Escherichia coli: spontaneous formation of MinE rings and MinD polar zones. Proc. Natl. Acad. Sci. U. S. A. 100 : 12724-12728.
Koppelman, C. M., T. Den Blaauwen, M. C. Duursma, R. M. Heeren, and N. Nanninga. 2001. Escherichia coli minicell membranes are enriched in cardiolipin. J. Bacteriol. 183 : 6144-6147.
Lenarcic, R., S. Halbedel, L. Visser, M. Shaw, L. J. Wu, J. Errington, D. Marenduzzo, and L. W. Hamoen. 2009. Localisation of DivIVA by targeting to negatively curved membranes. EMBO J. 28 : 2272-2282.
Margolin, W. 2005. FtsZ and the division of prokaryotic cells and organelles. Nat. Rev. Mol. Cell Biol. 6 : 862-871.
Margolin, W. 2001. Spatial regulation of cytokinesis in bacteria. Curr. Opin. Microbiol. 4 : 647-652.
Marston, A. L., and J. Errington. 1999. Selection of the midcell division site in Bacillus subtilis through MinD-dependent polar localization and activation of MinC. Mol. Microbiol. 33 : 84-96.
Marston, A. L., H. B. Thomaides, D. H. Edwards, M. E. Sharpe, and J. Errington. 1998. Polar localization of the MinD protein of Bacillus subtilis and its role in selection of the mid-cell division site. Genes Dev. 12 : 3419-3430.
Mileykovskaya, E., I. Fishov, X. Fu, B. D. Corbin, W. Margolin, and W. Dowhan. 2003. Effects of phospholipid composition on MinD-membrane interactions in vitro and in vivo. J. Biol. Chem. 278 : 22193-22198.
Raskin, D. M., and P. A. de Boer. 1999. MinDE-dependent pole-to-pole oscillation of division inhibitor MinC in Escherichia coli. J. Bacteriol. 181 : 6419-6424.
Raskin, D. M., and P. A. de Boer. 1999. Rapid pole-to-pole oscillation of a protein required for directing division to the middle of Escherichia coli. Proc. Natl. Acad. Sci. U. S. A. 96 : 4971-4976.
Raskin, D. M., and P. A. de Boer. 1997. The MinE ring: an FtsZ-independent cell structure required for selection of the correct division site in E. coli. Cell 91 : 685-694.
Romberg, L., and P. A. Levin. 2003. Assembly dynamics of the bacterial cell division protein FtsZ: poised at the edge of stability. Annu. Rev. Microbiol. 57 : 125-154.
Rothfield, L. 2003. New insights into the developmental history of the bacterial cell division site. J. Bacteriol. 185 : 1125-1127.
Rothfield, L., A. Taghbalout, and Y. L. Shih. 2005. Spatial control of bacterial division-site placement. Nat. Rev. Microbiol. 3 : 959-968.
Sengupta, S., and A. Rutenberg. 2007. Modeling partitioning of Min proteins between daughter cells after septation in Escherichia coli. Phys. Biol. 4 : 145-153.
Shen, B., and J. Lutkenhaus. 2009. The conserved C-terminal tail of FtsZ is required for the septal localization and division inhibitory activity of MinC(C)/MinD. Mol. Microbiol. 72 : 410-424.
Shiomi, D., and W. Margolin. 2007. The C-terminal domain of MinC inhibits assembly of the Z ring in Escherichia coli. J. Bacteriol. 189 : 236-243.
Suefuji, K., R. Valluzzi, and D. RayChaudhuri. 2002. Dynamic assembly of MinD into filament bundles modulated by ATP, phospholipids, and MinE. Proc. Natl. Acad. Sci. U. S. A. 99 : 16776-16781.
Szeto, T. H., S. L. Rowland, C. L. Habrukowich, and G. F. King. 2003. The MinD membrane targeting sequence is a transplantable lipid-binding helix. J. Biol. Chem. 278 : 40050-40056.
Thanedar, S., and W. Margolin. 2004. FtsZ exhibits rapid movement and oscillation waves in helix-like patterns in Escherichia coli. Curr. Biol. 14 : 1167-1173.
Touhami, A., M. Jericho, and A. D. Rutenberg. 2006. Temperature dependence of MinD oscillation in Escherichia coli: running hot and fast. J. Bacteriol. 188 : 7661-7667.
Varma, A., K. C. Huang, and K. D. Young. 2008. The Min system as a general cell geometry detection mechanism: branch lengths in Y-shaped Escherichia coli cells affect Min oscillation patterns and division dynamics. J. Bacteriol. 190 : 2106-2117.
Ward, J. E., and J. Lutkenhaus. 1985. Overproduction of FtsZ induces minicells in E. coli. Cell 42 : 941-949.
Yu, X.-C., and W. Margolin. 1999. FtsZ ring clusters in min and partition mutants: role of both the Min system and the nucleoid in regulating FtsZ ring localization. Mol. Microbiol. 32 : 315-326.

Information & Contributors


Published In

cover image Journal of Bacteriology
Journal of Bacteriology
Volume 192Number 1615 August 2010
Pages: 4134 - 4142
PubMed: 20543068


Received: 1 April 2010
Accepted: 3 June 2010
Published online: 15 August 2010


Request permissions for this article.



Jennifer R. Juarez
Department of Microbiology and Molecular Genetics, University of Texas Medical School at Houston, 6431 Fannin St., Houston, Texas 77030
William Margolin [email protected]
Department of Microbiology and Molecular Genetics, University of Texas Medical School at Houston, 6431 Fannin St., Houston, Texas 77030


Supplemental material for this article may be found at .

Metrics & Citations



  • For recently published articles, the TOTAL download count will appear as zero until a new month starts.
  • There is a 3- to 4-day delay in article usage, so article usage will not appear immediately after publication.
  • Citation counts come from the Crossref Cited by service.


If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. For an editable text file, please select Medlars format which will download as a .txt file. Simply select your manager software from the list below and click Download.

View Options

Figures and Media






Share the article link

Share with email

Email a colleague

Share on social media

American Society for Microbiology ("ASM") is committed to maintaining your confidence and trust with respect to the information we collect from you on websites owned and operated by ASM ("ASM Web Sites") and other sources. This Privacy Policy sets forth the information we collect about you, how we use this information and the choices you have about how we use such information.
FIND OUT MORE about the privacy policy