The Min System Disassembles FtsZ Foci and Inhibits Polar Peptidoglycan Remodeling in Bacillus subtilis

Abstract

A microfluidic system coupled with fluorescence microscopy is a powerful approach for quantitative analysis of bacterial growth. Here, we measure parameters of growth and dynamic localization of the cell division initiation protein FtsZ in Bacillus subtilis Consistent with previous reports, we found that after division, FtsZ rings remain at the cell poles, and polar FtsZ ring disassembly coincides with rapid Z-ring accumulation at the midcell. In cells mutated for minD, however, the polar FtsZ rings persist indefinitely, suggesting that the primary function of the Min system is in Z-ring disassembly. The inability to recycle FtsZ monomers in the minD mutant results in the simultaneous maintenance of multiple Z-rings that are restricted by competition for newly synthesized FtsZ. Although the parameters of FtsZ dynamics change in the minD mutant, the overall cell division time remains the same, albeit with elongated cells necessary to accumulate a critical threshold amount of FtsZ for promoting medial division. Finally, the minD mutant characteristically produces minicells composed of polar peptidoglycan shown to be inert for remodeling in the wild type. Polar peptidoglycan, however, loses its inert character in the minD mutant, suggesting that the Min system not only is important for recycling FtsZ but also may have a secondary role in the spatiotemporal regulation of peptidoglycan remodeling.IMPORTANCE Many bacteria grow and divide by binary fission in which a mother cell divides into two identical daughter cells. To produce two equally sized daughters, the division machinery, guided by FtsZ, must dynamically localize to the midcell each cell cycle. Here, we quantitatively analyzed FtsZ dynamics during growth and found that the Min system of Bacillus subtilis is essential to disassemble FtsZ rings after division. Moreover, a failure to efficiently recycle FtsZ results in an increase in cell size. Finally, we show that the Min system has an additional role in inhibiting cell wall turnover and contributes to the "inert" property of cell walls at the poles.

Keywords: FtsZ; MinD; cell division; growth; microfluidics; peptidoglycan.

FIG 1

Microfluidic analysis of growth and division in the wild type and in min mutants of Bacillus subtilis. Snapshot fluorescence microscopy of a microfluidic channel was performed for the wild type (WT) (A and C) and minD mutants (B and D) growing at steady state. (A and B) Fluorescence microscopy of the wild type (DK5133) (A) and a minD mutant (DK5155) (B) in a microfluidic channel expressing cytoplasmic mCherry protein (falsely colored red; top) or mNeongreen-FtsZ (falsely colored green; middle) and an overlay of the two (bottom). Graphs present results from a quantitative analysis of mCherry fluorescence intensity (red line) and mNeongreen fluorescence intensity (green line) and match the fluorescence microscopy images immediately above. All images are reproduced at the same magnification. (C and D) Fluorescence microscopy of a microfluidic channel for the wild type (DK4394) (C) and a minD mutant (DK4407) (D) growing at steady state. Images represent cytoplasmic mCherry protein (falsely colored red; top) and peptidoglycan stained with BADA (falsely colored green; bottom). Graphs present results from a quantitative analysis of mCherry fluorescence intensity (red) and BADA fluorescence intensity (green) and match the fluorescence microscopy images immediately above. All images are reproduced at the same magnification.

FIG 2

Cells mutated for the Min system divide faster than wild-type cells. (A) A histogram of the division time of individual cells of the wild-type strain (gray) and a minD mutant (blue) measured by microscopic analysis. Division events were defined by a local 20% decrease in mCherry (cytoplasmic) fluorescence intensity below a threshold value. More than 3,000 division events were counted per data set. Minicells were excluded from the division time analysis as once they are formed, minicells never divide. (B) A histogram of division time of individual cells of the minD mutant from the data set represented in panel A to separately determine the time elapsed between the medial and polar divisions. The time between medial division events (cyan) was determined as the time between division events that gave rise to two mother cells. The time between polar divisions (magenta) was determined to be the time between the formation of a cell pole and the formation of a division plane at that pole to give rise to a minicell. (C) Cell elongation rates were measured as the rate at which the cell poles moved apart from one another in the wild-type strain (gray) and a minD mutant (blue). The growth rates of over 2,500 cells were measured for each strain. Minicells were excluded from the elongation rate analysis as once they are formed, minicells do not elongate. (D) Data from panel C were replotted as the instantaneous increase in cell length per total length of the cell observed. (E) Growth curve of the wild-type strain (gray) and a minD mutant (blue) growing in highly agitated LB broth at 37°C. Optical density at 600 nm (OD₆₀₀) was measured with a spectrophotometer (600-nm wavelength). (F) Frequency histogram of the division time that gives rise to minicells at either the old cell pole (yellow) or the new cell pole (violet). Old cell poles were defined as the poles that had last experienced a polar division event. New cell poles were defined as the poles that had not previously experienced a polar division event. The wild-type strain (DK5133) and a minD mutant (DK5155) were used to generate all of the data in this figure.

FIG 3

FtsZ foci remain at the poles indefinitely in the absence of Min. Data represent results of kymograph analysis of the wild-type strain (A) and a minD mutant (B) of cytoplasmic mCherry signal (falsely colored red; top) and mNeongreen-FtsZ intensity (falsely colored green; middle) and an overlay of the red and green channels (bottom). In each panel, a single microfluidic channel was followed in a series of stacked snapshots taken at 2-min intervals to assemble the kymograph. Polar Z-rings were persistent in the minD mutant (see Movie S4) but fluctuated in intensity. We note that decreases in minD polar Z-ring intensity often corresponded to polar cell division events that gave rise to minicells. All images are reproduced at the same magnification. The wild-type strain (DK5133) and a minD mutant (DK5155) were used to generate all of the data in this figure.

FIG 4

The cytokinetic period is longer than the cell division time of a minD mutant. (Top) Sample kymograph analysis of a microfluidic channel in which either the wild-type strain or a minD mutant was grown. Cytoplasmic mCherry signal is falsely colored red (left) and overlaid with mNeongreen-FtsZ (falsely colored green) (right). Events necessary for defining division parameters are indicated and labeled as follows: a, septation; b, appearance of a nascent Z-ring; c, disappearance of a Z-ring; d, FtsZ peak intensity achieved. Each event designation was given a number as follows: 0, preceding generation; 1, current generation; 2, subsequent generation. Thin white lines are included to indicate cell tracking and lineage analysis. (Middle) Graphs of 100 manually tracked wild-type (gray) and minD mutant (blue) mother cell division cycles presented as bars of average values and whiskers of standard deviations for the following parameters: division time (the time between cell septation events [i.e., between consecutive “a” events]); Z-ring appearance period (the time between the appearance of one Z-ring and another [i.e., between consecutive “b” events]), Z-ring persistence period (the time between the appearance of a Z-ring and the disappearance of that Z-ring [i.e., between consecutive “b” and “c” events]), Z-ring polar duration (the time between a septation event and the disappearance of the Z-ring resulting from that septation events [i.e., between consecutive “a” and “c” events]), Z-ring medial delay (the time between a septation event and the appearance of a Z-ring that will eventually give rise to the next medial division event [i.e., between an “a” event and a “b” event that will give rise to the next round of septation]), Z-ring maturation period (the time between Z-ring formation and when that Z-ring achieves peak local intensity [i.e., between consecutive “a” and “d” events]), and cytokinetic period (time between Z-ring formation and septation directed by that Z-ring [i.e., between a “b” event and the “a” event that is caused by that particular Z-ring]). The raw data representing the histogram for each bar are presented in Fig. S2. (Bottom) Time line representations of the various events indicated in the bar graph depicted in cartoon form, color coded to match the indicated parameter of like color above, and annotated with relevant events marked by the defining letters.

FIG 5

FtsZ density is constant, but FtsZ accumulations in Z-rings differ in the wild type and the minD mutant. (A) A frequency histogram of the location of snapshot peak FtsZ intensity plotted relative to total cell length for wild type. The poles of the cell have relative position values of 0 and 1.0, whereas the midcell has a value of 0.5. (B) A frequency histogram of cell length distribution of the wild-type strain (gray) and a minD mutant (blue). The minD mutant has two peaks, namely, a shorter peak corresponding to “minicells” and a longer peak corresponding to “mother” cells that have chromosomes and are capable of division. (C) A frequency histogram of the location of snapshot peak FtsZ intensity plotted relative to total cell length for minD. (D) A frequency histogram of total FtsZ fluorescence intensity per cell. For each frame, multiple line scans through the longitudinal axis of the cell was performed, and total FtsZ fluorescence intensity was measured by integrating the area under the line scans. (E) A frequency histogram of FtsZ density was produced by dividing the total fluorescence intensity by cell length for each individual. (F) A frequency histogram of peak FtsZ fluorescence intensity per cell. For each frame, multiple line scans through the longitudinal axis of the cell determined the location of peak fluorescence intensity, and peak fluorescence magnitude was recorded. Data represent wild-type strain distribution (gray) and minD strain distribution (blue). Measurements were taken for 9,000 cells of the growing wild-type strain (DK5133), amounting to over 30,000 measurements, and for 6,000 cells of the growing minD strain (DK5155), amounting to over 20,000 measurements, to generate the data in this figure.

FIG 6

Some cells of a minD mutant experience multiple simultaneous divisions. Kymograph analysis of the cytoplasmic mCherry fluorescence intensity of single channels of the wild-type strain (left) or minD mutant (middle and right) was performed. Wild-type cells experience regular division once per cell cycle (arrowhead, left panel), but multiple simultaneous (arrowheads, middle panel) or slightly offset (arrowheads, right panel) division events were observed occasionally in the minD mutant. The wild-type strain (DK5133) and a minD mutant (DK5155) were used to generate all of the data presented in this figure.

FIG 7

minD mutants lose BADA staining fluorescence intensity faster than the wild type. Cells were grown in the presence of the peptidoglycan synthesis/remodeling indicator stain BADA for 4 min, the stain was washed out of the microfluidic device for 8 min, and then fluorescence intensity was tracked in time. (A) A graph of the total BADA intensity per cell per the length of the cell measured in the first frame of the experiment. Each dot represents an individual cell of the wild-type strain (gray) or the minD mutant (blue). (B) A graph of the total BADA intensity per cell per unit time after washout of fluorescent d-amino acid. Gray, wild type; blue, minD mutant; cyan, mother cells of the minD mutant; magenta, minicells of the minD mutant. (C) A graph of the rate of decrease in BADA intensity loss as measured by the slope of the lines shown in panel B. (D) A representative kymograph of the wild type after BADA washout. (E) A representative kymograph of the minD mutant after BADA washout. The wild-type (DK4393) and minD mutant (DK4407) strains were used for the experiments whose results are shown in all panels in this figure. Over 500 measurements were taken for each strain.

FIG 8

Recently formed minicells are proficient in remodeling of the polar peptidoglycan. (Top panels) Microfluidic analysis of a minD mutant (DK4407) expressing cytoplasmic mCherry (falsely colored red; left) and stained for 20 min with the fluorescent d-amino acid BADA (falsely colored green; right). Individual areas of the channels are highlighted by boxes and numbered. (Bottom) Images of the boxes (enlarged to increase detail) with corresponding numbers.