Mid1p/anillin and the septation initiation network orchestrate contractile ring assembly for cytokinesis

Abstract

In both animal cells and fungi, cytokinesis proceeds via a contractile actomyosin ring (CAR). Many CAR components and regulators are evolutionarily conserved. In Schizosaccharomyces pombe, the spatial cue for cytokinesis is provided by Mid1p/Anillin, whereas temporal coordination is ensured by the septation initiation network (SIN). However, neither Mid1p nor the SIN is considered to be essential for CAR assembly per se. Here, using 4D imaging, we reveal an unanticipated, novel role for the SIN in CAR assembly. We demonstrate that CAR assembly involves three, genetically separable steps: establishment of a cortical network of CAR proteins, its lateral condensation, and finally, the formation of a homogeneous CAR. We show that SIN mutants fail to form a homogeneous CAR; we identify hypophosphorylation and recruitment of the conserved PCH-family protein Cdc15p to the CAR as critical steps requiring SIN function. Furthermore, we show that in the absence of Mid1p, CAR assembly proceeds via an actomyosin filament, rather than a cortical network of CAR proteins. This mode of assembly is totally dependent on SIN signaling, thereby demonstrating a direct role for the SIN in CAR formation. Taken together, these data establish that Mid1p and the SIN are the key regulators that orchestrate CAR assembly.

Figure 1.

Spg1p-induced CAR assembly in interphase depends on SIN proteins. (A) Expression of spg1 was induced in cdc2-17 cdc15-GFP (top) or cdc2-17 sid2-250 cdc15-GFP cells (bottom), as described in the Materials and Methods. Cells were fixed and stained. The panels are a merge of DAPI-Calcofluor staining (DNA and septum, respectively, shown in blue), rhodamine-conjugated phalloidin (F-actin, shown in red), and Cdc15-GFP (shown in green). Note that Cdc15p-GFP staining is not seen that cell tips, as it does not survive the fixation conditions used. (B) cdc16-116 cdc2-17 rlc1-GFP (top) or cdc16-116 cdc2-17 sid2-250 rlc1-GFP cells (bottom) were synchronized by centrifugal elutriation and shifted for 2 h 30 min to 36°C. Cells were fixed and stained as in A, except that green represent Rlc1p-GFP. (C) Expression of spg1 was induced in cdc15-GFP or cdc7-24 cdc15-GFP cells for 22 h at 19°C. HU was added to 12 mM for the last 5 h of the induction. This was followed by 2 h 30 min at 36°C. The color scheme is the same as for A. (D) Expression of spg1 was induced in cdc2-17 cdc15-GFP (top) orcdc2-17 bgs1-191 cdc15-GFP cells (bottom), as described in the Materials and Methods. Cells were fixed and stained. The color scheme is the same as for A. (E) Expression of spg1 was induced in cdc2-17 rlc1-GFP (top), cdc2-17 sid4-SA1 rlc1-GFP (middle), or cdc2-17 cdc11-123 rlc1-GFP cells (bottom), as described in the Materials and Methods. Cells were fixed and stained. The color scheme is the same as for B. (F) Expression of spg1 was induced in cdc2-17 plo1-35 cdc15-GFP cells as described in the Materials and Methods. Cells were fixed and stained. The top image shows uninduced cells, and the middle and bottom images show induced cells. The color scheme is the same as for A. (G) Western blot analysis showing that spg1 expression is unaffected in a sid2+ (left) compared with a sid2-250 background (right). Uninduced samples are shown in marked U lanes, while induced samples are shown in marked I lanes. Top and bottom panels show anti-Spg1p and anti-Tubulin blots, respectively. Bar, 10 μm.

Figure 2.

Time-course analysis of ring assembly in mutants for the SIN and ring components. (A–D) Cells of the indicated mutant genotype expressing Rlc1p-GFP (ring marker) and Cdc11p-GFP (SPB marker) were processed for live imaging at 36°C as described in the Materials and Methods. The frames shown were captured at 1-min intervals. The time courses have been aligned with respect to the frame showing maximal SPB distance (t16, gray box). Black bars at the bottom of the figure indicate the three phases of CAR assembly (see text for details). Bar, 10 μm.

Figure 3.

Analysis of CAR homogeneity in wild type and SIN mutants. (A,B) Ten cells from wild type (A) and spg1-B8 (B) were analyzed; rings were reconstructed at four time points as indicated; t16 represents the time of maxSPBd (see Fig. 2). (C) The reconstructed rings were divided into eight sectors as shown in the diagram for SDFi calculation. The red square shown in the center of the diagram was used to determine the background intensity to set the quantification threshold. (D) The graph represents the ratio of the SDFi to the mean fluorescence intensity of individual cells over time. Blue lines correspond to wild type cells, and red lines correspond to spg1-B8 mutant cells. (E) The graph represents the mean SDFi, the blue line corresponds to wild-type cells (n = 10), the red line corresponds to spg1-B8 mutant cells (n = 10), the green line corresponds to cdc15-140 mutant cells (n = 10), and the gray line corresponds to sid2-250 mutant cells (n = 5).

Figure 4.

The SIN is required for recruitment of Cdc15p to the CAR. (A,B) Live imaging was performed at 36°C on cells expressing Cdc15p-GFP and Cdc11p-GFP in wild type (A) and the sid2-250 mutant (B). (C) The graph represents the cytoplasmic fluorescence relative to the extracellular background in wild-type (black squares) and sid2-250 mutant cells (white circles). (D) Western blot analysis of Cdc15p-HA (top panel) and Tubulin (bottom panel). The left panel shows extracts from cdc16-116 cells at the indicated temperature arrested with HU. The right panel shows extracts from cdc2-17 cells and cdc16-116 cdc2-17 double-mutant cells arrested in G2 at 36°C. Bar, 10 μm.

Figure 5.

Spg1p activation in interphase triggers Mid1p-independent CAR assembly. cdc2-17 rlc1-GFP cells were arrested in G2, and expression of spg1 was induced as described in Materials and Methods. Cells were then fixed and analyzed. (A,B) The image shows the Rlc1p-GFP protein (green), DNA stained with DAPI (blue), the septum stained with Calcofluor (blue), and F-actin stained with rhodamine-conjugated phalloidin (red). The arrowheads indicate mispositioned and misoriented or tilted rings. (C) Graphical representation of ectopic ring distribution, where A, B, and C correspond to the cell areas described in the diagram in the top right corner. The proportion of misoriented rings is represented in the column entitled “Tilted.” (D,E) spg1 expression was induced as described above, and cells were fixed and stained with an anti-Mid1p antibody (red), and DAPI (blue); Rlc1p-GFP is shown in green. Bar, 10 μm.

Figure 6.

Live imaging analysis of SIN-induced CAR assembly and in the absence of Mid1p activity. (A–D) Cells expressing Rlc1p-GFP were filmed as they progressed through mitosis and cytokinesis at 36°C. A Shows a wild-type cell going through normal mitosis, B shows a cdc2-17 interphase-arrested cell overexpressing spg1, C shows a mid1-6 mutant cell in mitosis, and D shows a mid1Δ-null mutant cell in mitosis. Bar, 10 μm. (E) Synchronous cultures of wild-type, cdc2-17 cdc16-116, and mid1-6 cells expressing Rlc1p-GFP were generated by centrifugal elutriation and filmed in yeast extract medium at 36°C to calculate the average CAR contraction time represented in this graph.

Figure 7.

Ring assembly depends on SIN signaling in the absence of Mid1p activity. (A–C) Cells of the indicated genotypes were fixed and stained after shifting an exponentially growing population of cells for 3 h from 25°C to 36°C. F-actin is shown in red, Rcl1p-GFP is shown in green, and DNA is shown in blue in the merge; the corresponding grayscale panels for each channel are also shown. A shows mid1-6 mutant cells, B shows sid2-250 mutant cells, and C shows mid1-6 sid2-250 double-mutant cells. (D–F) Graphical representations of quantification of the phenotypes observed are shown; blue bars represent the percentage of cells with a normal ring or septum, red bars represent the percentage of cells with a defective nonhomogeneous ring, green bars represent the percentage of cells with an actomyosin filament, and gray bars represent the percentage of cells with an Rlc1p-GFP dot. (D,E) Cells of the indicated genotype were shifted for 3 h at 36°C. (F) Overexpression of plo1-con was induced for 16 h at 25°C followed by a 2 h 30 min shift at 36°C in wild-type or sid2-250 cells expressing Rlc1p-GFP. Bar, 10 μm.

Figure 8.

Models for contractile ring assembly. Models recapitulating the three phases of contractile ring assembly in situations described in this study. (A) In wild-type mitosis, ring assembly is initiated by the formation of a cortical network of ring components that undergoes lateral condensation, giving rise to a nonhomogeneous ring precursor. This structure eventually matures by recruitment of additional factors such as Cdc15p, displaying the uniform distribution of ring components distribution that characterizes a functional ring. (B) When the SIN is activated in interphase, ring assembly is initiated by the formation of a Mid1p-independent filamentous actomyosin structure that eventually forms a ring structure. This structure is then competent for contraction, although it contracts more slowly. This mode of assembly is also observed in mid1 mutants in mitosis. (C) In a cdc12 mutant, ring assembly is initiated correctly in that it allows for the formation of a cortical network; however, this network fails to coalesce by lateral condensation and instead remains loose. (D) In SIN and cdc15 mutants, ring assembly proceeds correctly until lateral condensation of the cortical network. Thereafter, the ring precursor fails to mature, remains nonuniform, showing gaps and clumps of ring components, and, in the case of SIN mutants, fails to recruit Cdc15p. Finally, the defective rings seem to collapse, perhaps by a process involving contraction. (E) Model for the coordination of CAR assembly by plo1 controlling the two key regulators mid1 and the SIN. The SIN and Mid1p are both important for assembly of a functional CAR and cooperate to effect this. Regulation of both by Plo1p provides an attractive means of orchestrating CAR assembly during mitosis.