3D-SIM super resolution microscopy reveals a bead-like arrangement for FtsZ and the division machinery: implications for triggering cytokinesis

Abstract

FtsZ is a tubulin-like GTPase that is the major cytoskeletal protein in bacterial cell division. It polymerizes into a ring, called the Z ring, at the division site and acts as a scaffold to recruit other division proteins to this site as well as providing a contractile force for cytokinesis. To understand how FtsZ performs these functions, the in vivo architecture of the Z ring needs to be established, as well as how this structure constricts to enable cytokinesis. Conventional wide-field fluorescence microscopy depicts the Z ring as a continuous structure of uniform density. Here we use a form of super resolution microscopy, known as 3D-structured illumination microscopy (3D-SIM), to examine the architecture of the Z ring in cells of two Gram-positive organisms that have different cell shapes: the rod-shaped Bacillus subtilis and the coccoid Staphylococcus aureus. We show that in both organisms the Z ring is composed of a heterogeneous distribution of FtsZ. In addition, gaps of fluorescence were evident, which suggest that it is a discontinuous structure. Time-lapse studies using an advanced form of fast live 3D-SIM (Blaze) support a model of FtsZ localization within the Z ring that is dynamic and remains distributed in a heterogeneous manner. However, FtsZ dynamics alone do not trigger the constriction of the Z ring to allow cytokinesis. Lastly, we visualize other components of the divisome and show that they also adopt a bead-like localization pattern at the future division site. Our data lead us to propose that FtsZ guides the divisome to adopt a similar localization pattern to ensure Z ring constriction only proceeds following the assembly of a mature divisome.

Figure 1. 3D-SIM images of FtsZ-GFP localization in live cells of B. subtilis.

(A) Conventional wide-field fluorescence microscopy (Zeiss) image of B. subtilis strain SU570 (ftsZ-gfp) stained with the membrane dye FM4-64 shows how FtsZ-GFP assembles into Z rings. Scale bar, 5 µm. (B) When the same strain is imaged using 3D-SIM (OMX V3), regions of interest from the image can be selected (dashed box) to zoom in and rotate the image around the z-axis to view 3D FtsZ structures in the axial plane. (C–D) The improved image resolution provided by 3D-SIM allows visualization of constricting Z rings and the inner the cell membrane during division (indicated by white arrows).

Figure 2. Examining the distribution of FtsZ inside the Z ring of B. subtilis in live SU570 (ftsZ-gfp) cells.

(A) When the Z ring is visualized in live B. subtilis rod-shaped (SU570) cells, it appears as a transverse band of fluorescence. However, 3D-SIM (OMX V3) allows images of the Z ring to be rotated around the z-axis to clearly see how FtsZ is distributed within the Z ring in live cells of this strain. Small regions of low fluorescence intensity (gaps) in the Z ring are indicated by white arrowheads and cannot be seen without rotating the image. (Bi–iii) Additional examples of heterogeneous Z rings with a typical Z ring diameter of ∼0.9 µm. (Biv) A constricting Z ring with a diameter of 0.65 µm. (Ci) A typical 3D intensity profile reveals differences in fluorescence intensity and thus concentration of FtsZ-GFP in the Z ring with a diameter of 0.85 µm. The amount of fluorescence emanating from gaps is minimal and almost approaches baseline levels of fluorescence (black arrow). (Cii) A similar 3D intensity profile shows FtsZ-GFP distribution remains heterogeneous in a constricting Z ring; diameter, 0.65 µm. (D) The total fluorescence intensity of 56 different Z rings was analyzed. The relative amount of FtsZ-GFP fluorescence stays constant even as the diameter of the Z ring decreases.

Figure 3. Z ring structure in S. aureus and B. subtilis.

(A and B) When the Z ring is examined closer to the x-y orientation in B. subtilis, it reveals that it is heterogeneous throughout the entire Z ring (white line represents microscope slide). Z ring diameter, 0.8 µm. (C and D) S. aureus expressing FtsZ-GFP cells (SA94) visualized using conventional wide-field fluorescence image (Zeiss) and 3D-SIM (acquired using OMX V3), respectively. (E) The appearance of the Z ring in SA94 cells imaged in the axial orientation with 3D-SIM. (F) A close up image of a Z ring in SA94 cells imaged in the lateral orientation with 3D-SIM (OMX V3). (G) A 3D intensity plot shows how the relative abundance of FtsZ-GFP in all regions of the Z ring shown in panels (F) and (H). A graphical representation of the heterogeneous Z ring in S. aureus. Red arrows indicate dimensions used for measuring bead size and width in S. aureus cells. (I) 3D-SIM (OMX V3) image of the Z ring visualized using 1∶100 dilution of anti-FtsZ antibodies. S. aureus SA94 cells were grown in L-broth induced with 0.05 mM IPTG.

Figure 4. Time-lapse analysis of FtsZ localization within the Z rings of B. subtilis using 3D-SIM Blaze.

(A) Changes in the distribution of FtsZ-GFP were clearly evident in the top and bottom regions of the Z ring when using 3D-SIM (OMX Blaze) in the B. subtilis strain SU570 grown in PAB at 30°C. Time (seconds) is indicated on the upper left corner of each image. Z ring diameter, 0.89 µm. (B) 3D intensity plots clearly show that the distribution of FtsZ in the Z ring remains heterogeneous and dynamic. Each graph represents an image for each time point as shown above. (C) To analyze FtsZ dynamics in more detail, two regions of interest were monitored over time showing how FtsZ-GFP fluorescence fluctuates in the Z ring.

Figure 5. Z ring dynamics in S. aureus.

(A) 3D-SIM (OMX Blaze) time-lapse images show how FtsZ localization changes within the Z ring in S. aureus RN4220 cells (SA89). A white arrowhead marks the position of a gap when it initially forms inside the Z ring. The subsequent position of the arrowhead in each time point does not change and indicates how FtsZ is redistributed to a region of the Z ring, which previously had very little FtsZ present. White arrows indicate the formation of additional gaps in the Z ring. (B) Deconvolution time-lapse microscopy of SA94 cells expressing FtsZ-GFP. Time (s) is indicated on the upper left-hand side for each image. Cells were grown in L-broth at 37°C in the presence of 0.05 mM IPTG.

Figure 6. Monitoring FtsZ dynamics in non-dividing B. subtilis cells.

(A) At the permissive temperature SU744 (div-355) cells are able to divide, but at the non-permissive temperature division is inhibited as seen by the increased cell length. FtsZ-GFP was induced with 0.005 mM IPTG and grown in PAB at 30°C until mid-exponential growth before dilution and further growth at 45°C for 30 min to inhibit division. Scale bar, 5 µm. (B) To monitor FtsZ dynamics in non-dividing cells we visualized FtsZ-GFP localization in cells with the div-355 mutation at non-permissive temperature using conventional deconvolved microscopy (images acquired on the OMX Blaze system). FtsZ remains dynamic even when division is inhibited and Z ring constriction does not occur. A white arrowhead marks the position of a gap when it initially forms inside the Z ring. White arrows indicate the formation of additional gaps in the Z ring. Time (s) is indicated in the upper left corner. Z ring diameter, ∼0.9 µm.

Figure 7. EzrA rings in S. aureus are also heterogeneous and show dynamic movement.

(A) To examine EzrA-GFP localization using 3D-SIM (OMX V3), S. aureus strain SA126 was grown in L-broth at 37°C. The EzrA-GFP fusion protein displays a similar type of localization at the division site as FtsZ. (B) A close-up image of an EzrA ring in SA126 cells imaged in the lateral orientation with 3D-SIM (OMX V3). (C) A 3D intensity plot of EzrA rings orientated in the lateral plane shows a similar profile to FtsZ-GFP. (D) Deconvolved images were obtained through conventional wide-field fluorescence to monitor the localization of EzrA-GFP over time (acquired using OMX Blaze system). EzrA-GFP localization is dynamic and similar to that observed using FtsZ-GFP in S. aureus. White arrowheads show areas where the EzrA concentration is reduced. Arrowheads indicate the formation of additional gaps in the EzrA-GFP rings. Time (s) is indicated on the upper left-hand side for each image.

Figure 8. PBP2 rings in S. aureus are also heterogeneous and show dynamic movement.

(A) To examine GFP-PBP2 localization using 3D-SIM (OMX V3), S. aureus strain SA136 was grown in L-broth at 37°C. (B) A 3D intensity plot of a GFP-PBP2 ring orientated in the lateral plane shows a similar profile to FtsZ-GFP (left) and a close-up image of a PBP2 ring (right) in SA136 cells imaged in the lateral orientation with 3D-SIM (OMX V3) (C) Deconvolved images were obtained through conventional wide-field fluorescence microscopy; the localization of GFP-PBP2 over time is dynamic and similar to that observed using FtsZ-GFP. Arrowheads indicate the formation of additional gaps in the PBP2 rings. Images were acquired on the OMX Blaze system. Time (s) is indicated on the upper left-hand side for each image.

Figure 9. Models for the arrangement of FtsZ protofilaments inside the Z ring.

(A) We propose two models to describe the architecture of the Z ring. The overlapping model predicts that short FtsZ protofilaments begin to bundle in both lateral (circumferential) and radial directions of the cell. Alternatively, FtsZ protofilaments could be arranged into a single layer in the radial direction and only bundle in the lateral direction. (B) The amount of FtsZ inside the Z ring is constant throughout constriction, but the distribution of FtsZ inside the rings fluctuates over time. As constriction begins the Z ring exerts a localized pinching force on the membrane where increased levels of FtsZ are found. Gaps (indicated by red brackets) in the Z ring allow constriction to occur continuously as the circumference of this structure becomes reduced (indicated by the decrease in height of the blue rectangle).