Cooperation between Rho-GEF Gef2 and its binding partner Nod1 in the regulation of fission yeast cytokinesis.

Cytokinesis is the last step of the cell-division cycle, which requires precise spatial and temporal regulation to ensure genetic stability. Rho guanine nucleotide exchange factors (Rho GEFs) and Rho GTPases are among the key regulators of cytokinesis. We previously found that putative Rho-GEF Gef2 coordinates with Polo kinase Plo1 to control the medial cortical localization of anillin-like protein Mid1 in fission yeast. Here we show that an adaptor protein, Nod1, colocalizes with Gef2 in the contractile ring and its precursor cortical nodes. Like gef2, nod1 has strong genetic interactions with various cytokinesis mutants involved in division-site positioning, suggesting a role of Nod1 in early cytokinesis. We find that Nod1 and Gef2 interact through the C-termini, which is important for their localization. The contractile-ring localization of Nod1 and Gef2 also depends on the interaction between Nod1 and the F-BAR protein Cdc15, where the Nod1/Gef2 complex plays a role in contractile-ring maintenance and affects the septation initiation network. Moreover, Gef2 binds to purified GTPases Rho1, Rho4, and Rho5 in vitro. Taken together, our data indicate that Nod1 and Gef2 function cooperatively in a protein complex to regulate fission yeast cytokinesis.

INTRODUCTION

Cytokinesis is the last step of the cell cycle and is essential for cell proliferation and differentiation. Most proteins and key events in cytokinesis are evolutionarily conserved from fungal to human cells (Pollard and Wu, 2010; Green ; Lee ; Wloka and Bi, 2012). In the fission yeast Schizosaccharomyces pombe, anillin-related protein Mid1 plays a crucial role in early stages of cytokinesis (Chang ; Sohrmann ; Bähler ; Paoletti and Chang, 2000; Lee and Wu, 2012; Saha and Pollard, 2012a). Mid1 resides in the nucleus and in protein complexes called nodes at the medial cortex during interphase (Bähler ; Paoletti and Chang, 2000; Almonacid ). Together with the DYRK kinase Pom1, these medial nodes control cell size and mitotic entry (Martin and Berthelot-Grosjean, 2009; Moseley ; Hachet ). During G2/M transition, more Mid1 is released from the nucleus to the cortical nodes by Polo kinase Plo1 via phosphorylation of Mid1 (Bähler ; Almonacid ). These Mid1 nodes mature into cytokinesis nodes by recruiting other proteins, such as IQGAP Rng2, myosin-II, F-BAR protein Cdc15, and formin Cdc12 (Wu , 2006; Motegi ; Almonacid ; Laporte ; Padmanabhan ). Then the nodes and actin filaments condense into a compact ring through a search, capture, pull, and release mechanism (Vavylonis ; Chen and Pollard, 2011; Ojkic ; Laporte ). The compact ring matures and constricts, guiding the formation of a division septum (Pollard and Wu, 2010; Proctor ). The cell is then divided into two daughter cells with the degradation of primary septum.

The F-BAR protein Cdc15 is essential for cytokinesis (Fankhauser ; Carnahan and Gould, 2003; Roberts-Galbraith , 2010; Arasada and Pollard, 2011). In early cytokinesis, Mid1 recruits Cdc15 to cytokinesis nodes, which in turn recruits the formin Cdc12 to nucleate actin filaments (Carnahan and Gould, 2003; Kovar ; Laporte ). Cdc15 is also essential for contractile-ring maturation and assembly regulated by the septation initiation network (SIN) pathway (Wachtler ; Hachet and Simanis, 2008; Laporte ). During late cytokinesis, Cdc15 and another F-BAR protein, Imp2, recruit C2-domain protein Fic1 and paxillin Pxl1 to ensure the maintenance and integrity of the contractile ring (Pinar ; Roberts-Galbraith ).

The contractile ring and septation/septum formation are regulated by the SIN pathway, which is composed of a GTPase and a kinase cascade (Wachtler ; Hachet and Simanis, 2008; Krapp and Simanis, 2008; Johnson ). The SIN proteins locate at the spindle pole body (SPB) via scaffold proteins Cdc11 and Sid4 (Chang and Gould, 2000; Krapp ; Tomlin ; Morrell ). SIN pathway signaling is controlled by the activation of the GTPase Spg1 by Polo kinase, and the inactivation by the two component GTPase-activating proteins Cdc16 and Byr4 (Schmidt ; Furge , 1999; Jwa and Song, 1998; Tanaka ; Krapp ). The GTP-bound Spg1 interacts with kinase Cdc7 and causes its redistribution to the new SPB (Fankhauser and Simanis, 1994; Cerutti and Simanis, 1999; Mehta and Gould, 2006). The downstream kinases and their binding partners, including Sid1-Cdc14 and Sid2-Mob1, are then activated and recruited onto the SPB (Fankhauser and Simanis, 1993; Balasubramanian ; Sparks ; Guertin ; Hou ; Salimova ). Activated Sid2-Mob1 is then relocalized to the contractile ring to promote contractile-ring constriction and septum formation (Jin ; Chen )

Besides the equivalents of the SIN pathway, MEN and Hippo pathways, Rho GTPase Rho1/RhoA and its activators, the Rho guanine nucleotide exchange factor (GEF; Ect2, Pebble, etc.) are involved in division-site specification and contractile-ring formation by activating myosin-II and actin assembly in budding yeast and animal cells (Lehner, 1992; Imamura ; O'Keefe ; Tolliday ; Bement ; Yuce ; Nishimura and Yonemura, 2006; Yoshida ; Watanabe ; Su ). In contrast, Rho GTPases in S. pombe regulate only later stages of cytokinesis and cell polarity (García ; Pérez and Rincón, 2010). Fission yeast has six Rho GTPases (Cdc42 and Rho1-5) and seven Rho GEFs (Gef1-3, Rgf1-3, and Scd1). Cdc42, regulated by Gef1 and Scd1, is essential for cell polarity and morphology (Coll ; Hirota ; Rincón ). Rho-GEFs Rgf1-3 activate Rho1, which is essential for cell-wall synthesis, septum formation, and cell polarization (Tajadura ; Morrell-Falvey ; Mutoh ; García , 2009; Wu ). Rho2 is involved in cell morphology and septum formation by regulating cell wall α-glucan biosynthesis (Calonge ). Rho3 regulates exocytosis (Nakano ; Wang ; Kita ). Rho4 controls the secretion of lytic enzymes for septum degradation (Nakano ; Santos , 2005). Rho5 is a paralogue of Rho1 and shares similar functions (Nakano ; Rincón ). GEFs that regulate Rho2-5 GTPases are unknown, except that Rgf1 and Rgf2 might weakly interact with Rho5 (Mutoh ).

Recently we and others found that the putative Rho-GEF Gef2 localizes to cortical nodes and coordinates with Polo kinase Plo1 to regulate division-site selection (Moseley ; Ye ; Guzman-Vendrell ). In gef2∆ plo1 double mutants, Mid1 localization to the cortical nodes and the contractile ring is severely affected and the division site is misplaced. In addition, these studies showed that Gef2 interacts with Mid1 N-terminus (Ye ; Guzman-Vendrell ), which is essential for Mid1 function (Almonacid , 2011; Lee and Wu, 2012). The substrate GTPases for Gef2 and the regulation of Gef2, however, are largely unknown.

Here we show that Nod1 forms a complex with Gef2 to regulate cytokinesis. Nod1 and Gef2 are interdependent for their localization to cortical nodes and the contractile ring. Their localization at the contractile ring also depends on the physical interaction between Nod1 and the F-BAR protein Cdc15. Like gef2∆, nod1∆ suppresses SIN mutants by reducing cell lysis. In addition, the GEF domain of Gef2 interacts with GTPases Rho1, Rho4, and Rho5 in vitro. Thus it is possible that the Gef2/Nod1 complex may activate and function through Rho GTPases during cytokinesis.

RESULTS

Nod1 is a Gef2-related protein that localizes to cortical nodes and the contractile ring

We previously found that the putative Rho-GEF Gef2 plays a role in division-site positioning in cooperation with Polo kinase, Plo1 (Ye ). Concurrently, we identified a novel protein, Nod1 (SPAC12B10.10; Jourdain ), in the S. pombe protein database with sequence similarity to Gef2. Nod1 is annotated as a sequence orphan with 419 amino acids (aa; www.pombase.org/spombe/result/SPAC12B10.10). Although it has no GEF domain, Nod1 shares 18% identity and 34% similarity with Gef2 C-terminal aa 636–1101 (Figure 1A). The structure prediction program suggested that Nod1 is a helix-rich protein with no predicted domain (Jones, 1999; Wood ).

FIGURE 1:

Nod1 colocalizes with Gef2 in cortical nodes and the contractile ring and shares similar function with Gef2 in division-site selection. (A) Nod1 shares similarity with Gef2 C-terminus. Top, schematics of Gef2 and Nod1 domains or regions. The similar regions between Nod1 and Gef2 are marked with the same pattern. Bottom, sequence alignment between Gef2 aa 601–1101 (top row) and FL Nod1 (bottom row) using Vector NTI program. Identical and similar (D/E, I/L/V, K/R, N/Q, and S/T) aa are shaded in black and gray, respectively. (B–F) Cells were grown and imaged at 25°C. (B) Colocalization of Nod1 with Gef2 in cortical nodes and the contractile ring (strain JW4457). Top, maximum intensity projection. Bottom, single slice at cell bottom. (C) Molecule numbers of mECitrine-Gef2 (JW4912) and Nod1-mECitrine (JW4008) globally in whole cells and locally in the contractile ring and interphase nodes. (D) FRAP analysis of Nod1 (JW4008) and Gef2 (JW3825). Cells were bleached at time zero. Mean ± SEM. (E, F) Nod1 and Gef2 have similar function in division-site positioning. (E) Differential interference contrast (DIC) images and (F) quantification of the division-site positioning. The abnormal septa are defined as septa not placed within the central 20% of the cell or not within 80–100° angle to the long axis of the cell. Strains used: wt (JW81), nod1∆ (JW3773), gef2∆ (JW1826), nod1∆ gef2∆ (JW3814), plo1-ts18 (IH1600), nod1∆ plo1-ts18 (JW3815), gef2∆ plo1-ts18 (JW3078), and nod1∆ gef2∆ plo1-ts18 (JW3873). Bars, 5 μm.

To determine Nod1’s functions, we first tagged Nod1 with monomeric enhanced green fluorescent protein (GFP) at its C-terminus and examined its localization. Of interest, Nod1 colocalized with Gef2 throughout the cell cycle at interphase nodes, cytokinesis nodes, and the contractile ring (Figure 1B). We next counted Nod1 molecule numbers in cells by measuring its global and local fluorescence intensity (Wu and Pollard, 2005; Laporte ). In our previous study, we used strain kanMX6-Pgef2-mECitrine-gef2 (JW3825) to measure the intensity of Gef2 (Ye ). We found that the kanMX6 cassette in the strain affected Gef2 expression level, similar to N-terminal tagged F-BAR protein Cdc15 (Wu and Pollard, 2005). We therefore used the kanMX6 looped-out mECitrine‑gef2 strain (JW4912) to requantify Gef2 molecules globally and locally. The global Gef2 level was one-third in the kan-sensitive strain (JW4912), whereas the local Gef2 concentrations at the contractile ring and cortical nodes were similar to the original data (Ye ). Compared to Gef2 (1440 ± 660 molecules/cell, 570 ± 90 molecules at the contractile ring, and 16 ± 5 molecules/interphase node), Nod1 had 1520 ± 700 molecules/cell, 770 ± 150 molecules at the contractile ring, and 15 ± 5 molecules/interphase node (Figure 1C). Thus the ratio of Nod1 to Gef2 in interphase nodes and contractile ring is ∼1:1 and 1.35:1, respectively.

We performed fluorescence recovery after photobleaching (FRAP) assays on interphase nodes to determine Nod1 dynamics at the cell cortex. Nod1 fluorescence recovered with a half-time (t1/2) of 170 ± 77 s, and the mobile fraction was ∼40%, similar to Gef2 (t1/2 = 180 ± 85 s, 37% mobile fraction; Figure 1D). This indicates that both Nod1 and Gef2 are relatively stable on the plasma membrane compared with some other cytokinesis proteins (Laporte ). Together these data suggested that Nod1 might play a role in cytokinesis together with the putative Rho-GEF Gef2.

Nod1 regulates division-site positioning cooperatively with Polo kinase Plo1

Interphase nodes are important for cell-size control and mitotic entry in fission yeast (Martin and Berthelot-Grosjean, 2009; Moseley ; Hachet ; Deng and Moseley, 2013). As reported (Jourdain ), we found that similar to the length of dividing gef2∆ cells (Ye ), dividing nod1∆ cells were 16.2 ± 1.0 μm long (n = 148 septating cells), slightly but significantly longer than 14.4 ± 0.9 μm of wild-type (wt) cells (n = 117, p < 0.001). Thus Nod1 and Gef2 play a role in cell-size control.

Gef2 coordinates with Polo kinase, Plo1, to recruit anillin-related protein Mid1 to the cortical nodes for division-site specification (Ye ). Because of the sequence similarity between Nod1 and the C-terminus of Gef2 and their colocalization (Figure 1, A and B), we hypothesized that Nod1 has a function similar to Gef2 at early cytokinesis. To test this hypothesis, we crossed nod1∆ to the temperature-sensitive mutant of Polo kinase, plo1-ts18 (Figure 1E). Similar to gef2∆ plo1-ts18 (Ye ), 95% of nod1∆ plo1-ts18 cells had abnormal septa at 25°C (Figure 1, E and F). Moreover, nod1∆ and gef2∆ also had the same strong synthetic interactions with mutations known to affect early cytokinesis, such as mid1, rng2, and cdc4-8, but not with mutations in cell-size control such as cdr2∆ and blt1∆ (see Table 1 later in the article). Thus Nod1 shares a similar function with Gef2 in division-site specification and contractile-ring assembly (Ye ; Jourdain ).

TABLE 1:

Genetic interactions of nod1∆ with other mutations affecting cytokinesis and cell-size control.

	Temperature (°C)a
Strain	25	30	32	36	gef2∆b
plo1-ts18	+++	++	++	+	Same
plo1-ts18 nod1∆	+	+	+	+/-
mid1-6	+++	+++	+++	++	Same
mid1-6 nod1∆	+++	++	++	++
mid1-366	+++	+++	+++	++	Same
mid1-366 nod1∆	+	+	+	+
rng2-D5	+++	++	++	-	Same
rng2-D5 nod1∆	++	+	+	-
rng2-346	+++	++	++	-	Same
rng2-346 nod1∆	++	+	+	-
cdc4-8	+++	++	++	-	Same
cdc4-8 nod1∆	+++	+	+	-
cdc15-140	+++	+	-	-	Same
cdc15-140 nod1∆	+++	-	-	-
cdr2∆	+++	+++	+++	+++	Same
cdr2∆ nod1∆	+++	+++	+++	+++
blt1∆	+++	+++	+++	+++	Same
blt1∆ nod1∆	+++	+++	+++	+++
klp8∆	+++	+++	+++	+++	Same
klp8∆ nod1∆	+++	+++	+++	+++
cdc7-24	+++	+	-	-	Same
cdc7-24 nod1∆	+++	++	-	-
cdc11-136	+++	++	++	-	Same
cdc11-136 nod1∆	+++	+++	+++	+/-
sid2-250	++	-	-	-	Different
sid2-250 nod1∆	++	-	-	-
sid2-250 gef2∆	++	+/-	-	-

aGrowth and color of colonies on YE5S + phloxin B plates at various temperatures. +++, similar to wt; ++, mild defects or cell lysis; +, cell lysis with reduced growth rate; +/-, severe cell lysis and slow growth; -, inviable.

bThe genetic interactions of nod1∆ were compared with those of gef2∆ with corresponding mutants.

To examine whether Nod1 and Gef2 function in the same or parallel genetic pathways, we tested the genetic interactions among nod1∆, gef2∆, and plo1-ts18 (Figure 1, E and F). nod1∆ gef2∆ double-mutant cells resembled the single mutants. The nod1∆ gef2∆ plo1-ts18 triple mutant was still viable, with ∼96% cells displaying abnormal septa at 25°C, similar to nod1∆ plo1-ts18 and gef2∆ plo1-ts18. These results indicated that Nod1 and Gef2 are in the same genetic pathway.

Nod1 and Gef2 are interdependent on their C-termini for localization to cortical nodes

Because Gef2 and Nod1 are in the same genetic pathway, we tested whether they affect each other's localization. In wt cells, Gef2 localized to cortical nodes and the contractile ring (Figure 2A). Node localization was abolished, however, and contractile ring localization was greatly reduced in nod1∆ (Figure 2A). Gef2 was detected at the contractile ring with 115 ± 50 molecules, at ∼20% of wt levels, in nod1∆ cells (p < 0.001). Nod1 also failed to localize to cortical nodes in gef2∆, and the localization to the contractile ring was reduced to ∼60% of wt level, with 460 ± 130 molecules (p < 0.001; Figure 2A). The loss of localizations was not due to global protein concentration, since Nod1 and Gef2 protein levels were not significantly affected in the absence of one another (Figure 2B). Thus Gef2 and Nod1 are interdependent for localization to cortical nodes (Jourdain ) and partially interdependent for localization to the contractile ring.

FIGURE 2:

Nod1 and Gef2 are interdependent on their C-termini for cortical-node localization and partially interdependent for contractile-ring localization. (A) Micrographs of Nod1 and Gef2 localization in wt and deletion mutants (left). Molecules in the contractile ring were counted (right). Cells expressing mECitrine-Gef2 (JW3825 and JW4014) and Nod1-mECitrine (JW4008 and JW4038) were used. (B) Nod1 and Gef2 protein levels in wt (+) and the deletion (−) mutants. Cells extracts from the strains used in A were loaded in triplicate in Western blotting (top). Tubulin was used as the loading control. The graphs show the quantification of the protein levels (bottom). (C) Micrographs of Nod1 localization in cells expressing mECitrine-tagged FL Nod1 (JW4750 and JW4008) or Nod1 truncations (JW5065, JW4856, JW4325, and JW4326). (D) Micrographs of localization of Nod1 and Gef2 (strains JW4226, JW5107, JW4359, JW4010, JW4256, and JW4355). (E) Summary of Nod1 and Gef2 localization to cortical nodes in different truncation mutants. +, localized to cortical nodes; −, not localized to cortical nodes. Bars, 5 μm.

Gef2 C-terminal aa 957–1101 are necessary and sufficient for its cellular localization (Ye ). To test which region of Nod1 is important for its localization, we truncated Nod1 at its native chromosomal locus under the control of nod1 promoter based on the sequence alignment between Gef2(601–1101) and Nod1 (Figures 1A and 2E). N-terminal truncations of Nod1 still localized to the cortical nodes and contractile ring (Figure 2C, top). When the last 91 aa of Nod1 from the C-terminus were truncated, however, Nod1 failed to localize to cortical nodes, but it still localized to the contractile ring with lower intensity (Figure 2C, bottom). We conclude that Nod1 C-terminal aa 329–419 are both essential and sufficient for Nod1 node localization.

Next we studied how the Nod1 and Gef2 truncations affect each other's localization (Figure 2, D and E). Gef2 localized to both cortical nodes and the contractile ring in nod1(∆1-328) but only localized to the contractile ring weakly when the last 91 aa of Nod1 were truncated in nod1(∆329-419) (Figure 2D), which is similar to Gef2 localization in nod1∆ (Figure 2A). Similarly, Nod1 localized normally in gef2(∆1-956) but failed to localize to cortical nodes when Gef2 C-terminal aa 957–1101 were truncated (Figure 2D). Together Nod1 and Gef2 are interdependent on their C-termini for cortical node localization and partially interdependent on their C-termini for localization to the contractile ring (Figure 2E).

Nod1 physically interacts with Gef2 through their C-termini

Based on the interdependence between Nod1 and Gef2 for localization, we hypothesized that the two proteins interact with each other. Indeed, monomeric enhanced Citrine (mECitrine)–Gef2 pulled down Nod1-13Myc in the coimmunoprecipitation (co-IP) assay (Figure 3A). In reciprocal co-IP, mECitrine–Nod1 also pulled down Gef2-13Myc (Figure 3B). Because Nod1 and Gef2 C-termini are important for their localization, we tested whether they interact through their C-termini. As expected, Nod1 interacted with Gef2(957–1101) (Figure 3A) and Gef2 with Nod1(329–419) (Figure 3B) in co-IP assays. These data suggested that Nod1 and Gef2 interact with each other in vivo through their C-termini.

FIGURE 3:

Nod1 and Gef2 physically interact through their C-termini. (A, B) Antibodies against mECitrine were used in IP. Monoclonal antibodies against mECitrine and Myc were used in Western blotting. (A) Nod1 co-IP with Gef2 C-terminus. IPs were carried out from cell extracts of nod1-13Myc (JW4013), mECitrine-gef2 (JW3825), mECitrine-gef2 nod1-13Myc (JW4330), mECitrine-gef2(957-1101) (JW3826), nod1-13Myc (JW4013), and mECitrine-gef2(957-1101) nod1-13Myc (JW4331). (B) Gef2 co-IP with Nod1 C-terminus. Strains JW3622, JW4453, JW5093, JW4455, and JW5095 were used. Asterisks mark the expected bands. (C) Nod1 and Gef2 interact via their C-termini in yeast two-hybrid assays. β‑Galactosidase activities (mean ± SD, n = 2) are shown as fold changes over the highest negative control.

We tested whether the interaction might be direct between Nod1 and Gef2 through yeast two-hybrid assays (Figure 3C). Full length (FL) Nod1 displayed positive interaction with Gef2 and Gef2(957–1101) but not with Gef2(1–956), whereas FL Gef2 bound to Nod1 and Nod1(210–419) but not to Nod1(1–209). Moreover, Nod1(210–419) interacted with Gef2(957–1101). In summary, Nod1 and Gef2 physically interact with each other through their C-termini, and the interaction is critical for their localization.

The F-BAR protein Cdc15 recruits Nod1 and Gef2 to the contractile ring through its interaction with the Nod1 N-terminus

Gef2 localizes to cytokinesis nodes and the contractile ring in blt1∆, although interphase-node localization is abolished (Ye ). The timings of appearance at cytokinesis nodes for Gef2 in blt1∆ and the F-BAR protein Cdc15 in wt cells are similar (Laporte ; Ye ). Thus we observed Gef2 and Nod1 localization in the temperature-sensitive mutant cdc15-140 at the restrictive temperature (Figure 4, A and B). After 2 h at 36°C, both Gef2 and Nod1 formed some aggregates, and signals were weaker in cdc15 than at 25°C (Figures 2A and 4, A and B). Gef2 and Nod1 still localized to cortical nodes with low intensity, but their contractile-ring localizations were greatly reduced in cdc15-140 cells (Figure 4, A and B). Unlike in gef2∆ cells, the contractile-ring localization of Nod1 was completely abolished in gef2∆ cdc15-140 cells (Figure 4B). Together our data indicate that the contractile-ring localizations of Nod1 and Gef2 depend on each other and on the F-BAR protein Cdc15.

FIGURE 4:

The F-BAR protein Cdc15 recruits or stabilizes Gef2 and Nod1 localization to the contractile ring by interaction with the Nod1 N-terminus. (A, B) Involvement of Cdc15 in Gef2 and Nod1 localization at the contractile ring. Cells expressing mECitrine-Gef2 (A) and Nod1-mECitrine (B) were cultured at 25°C and shifted to 36°C for 2 h before imaging at 36°C. Myo2 was used to mark the contractile ring in A. Strains used were JW4008, JW4038, JW5027, JW5028, JW5582, and JW5583. (C, D) Cdc15 interacts with Gef2 and Nod1 in co-IPs (similar to Figure 3A). Strains used were JW1063, JW5120, JW4013, JW3325, and JW3204. (E) Cdc15 interacts with Nod1 N-terminus in yeast two-hybrid assays. β‑Galactosidase activities (mean ± SD, n = 2) as fold changes over the highest negative control are shown. Bars, 5 μm.

We next investigated whether Cdc15 physically interacts with Gef2 and Nod1. Monomeric yellow fluorescent protein (mYFP)–Cdc15 pulled down both Gef2-13Myc and Nod1-13Myc from cell lysates in co-IP assays (Figure 4, C and D), suggesting that the three proteins were in a protein complex. Yeast two-hybrid assays revealed no positive interactions between Cdc15 and Gef2, whereas Cdc15 bound to Nod1 and Nod1(1–328) (Figure 4E). This is consistent with our data that Nod1 N-terminal truncations still localize to the contractile ring (Figure 2C). Thus we conclude that the F‑BAR protein Cdc15 recruits or stabilizes the Nod1/Gef2 complex to the contractile ring through the N-terminus of Nod1 during mitosis.

Nod1 and Gef2 affect contractile-ring stability during late cytokinesis

The F-BAR protein Cdc15 is an essential component of the contractile ring, which plays multiple roles during early and late cytokinesis (Fankhauser ; Roberts-Galbraith , 2010; Laporte ). The fact that Cdc15 recruits the Nod1/Gef2 complex to the contractile ring indicated that Nod1 and Gef2 might have additional functions during late cytokinesis besides their role in division-site positioning. Indeed, we found that nod1∆ and gef2∆ had synthetic genetic interactions with cdc15-140. The double mutants nod1∆ cdc15-140 and gef2∆ cdc15-140 failed to form colonies, whereas cdc15-140 mutant still grew at 30°C (Figure 5A; see Table 1 later in the paper). At 25°C, both cdc15-140 single mutant and the double mutants resembled wt (Figure 5B, top). After 6 h at 30°C, cells proliferated with a mean cell length of 11.9 μm for wt and 17.3 μm for cdc15-140 cells. In contrast, most nod1∆ cdc15-140 and gef2∆ cdc15-140 cells were significantly longer, with mean cell length of 26.5 and 28.3 μm, respectively (Figure 5, B and C). We next quantified the number of nuclei per cell in these mutants at 30°C (Figure 5D). Wild type had ∼13% binucleated cells, whereas cdc15-140 had 24% binucleated cells, and <1% of cells had more than two nuclei. The majority of nod1Δ cdc15-140 and gef2∆ cdc15-140 mutants, however, were binucleated (62 and 58%, respectively), and ∼13 and 7% of cells contained more than two nuclei. These results indicated that the synthetic lethality in nod1∆ cdc15-140 and gef2∆ cdc15-140 cells was due to defects in cytokinesis.

FIGURE 5:

Nod1 and Gef2 affect contractile-ring stability. (A) nod1∆ and gef2∆ display synthetic interaction with cdc15-140. Serial dilutions (3×) of indicated strains (JW81, JW1743, JW4259, JW4016, JW2854, and JW2937) on YE5S plates at 25, 30, and 36°C, respectively. (B–D) nod1∆ cdc15-140 and gef2∆ cdc15-140 cells display typical cytokinesis defects with elongated and multinucleated cells. Relevant strains used in A were cultured in YE5S liquid at 25° (top) or 30°C (bottom) for 6 h before imaging. (B) Before imaging at 30°C, cells were stained with Hoechst for 10 min at 30°C to visualize DNA (green). DIC in gray. (C) Cell length and (D) number of nuclei in cells grown at 30°C for 6 h. (E–G) Nod1 and Gef2 affect contractile-ring stability during cytokinesis at 30°C. Rlc1 and Bgs1 were used to monitor the contractile ring and septum formation. Cells were grown at 30°C for 6 h before imaging at 30°C. Strains used: JW5357, JW5329, and JW5330. (E) Time courses of selected images from a movie with 1-min delay. The entire series can be viewed in Supplemental Videos S1 and S2. (F) Quantification of cells that fail to maintain the contractile ring (CR) after ring assembly. (G) Mean intensity of Rlc1-tdTomato at CR. Rlc1 intensity is significantly reduced in nod1∆ cdc15-140 (p < 0.001) and gef2∆ cdc15-140 (p < 0.001) cells vs. cdc15-140 cells. Bars, 5 μm.

To further determine the nature of the defects in nod1∆ cdc15-140 and gef2∆ cdc15-140 cells, we visualized contractile-ring and septum formation in the mutant cells using markers myosin regulatory light chain Rlc1-tandem dimer Tomato (tdTomato) and (1,3)β-d-glucan synthase GFP-Bgs1 (Figure 5E, Supplemental Figure S1, and Supplemental Videos S1–S3). At 36°C, most cdc15-140 mutant cells cannot maintain the contractile ring and form multinucleated cells (Balasubramanian ; Wachtler ). At a semipermissive temperature of 30°C, Rlc1 localized to the cytokinesis nodes, which coalesced into the contractile ring in most cells. Then Bgs1 left the growing cell tips and accumulated at the contractile ring. The contractile ring constricted, and septum formed (Figure 5E). Approximately 30% cdc15-140 cells were defective in contractile-ring assembly and stability, however, and the ring eventually collapsed into aggregates (Figure 5, F and G). Consequently, Bgs1 dispersed around the cell cortex, and the cells became elongated and swollen. These defects were more pronounced in nod1∆ cdc15-140 and gef2∆ cdc15-140 cells, for which Rlc1-tdTomato levels at the division site were significantly reduced to ∼30% of those in cdc15-140 single mutant (Figure 5, E and G). Approximately 52% of nod1∆ cdc15-140 and gef2∆ cdc15-140 cells failed to maintain the contractile ring (Figure 5, E and F, Supplemental Figure S1, and Supplemental Videos S1–S3). Thus our data suggest that Nod1 and Gef2 help to stabilize the contractile ring.

Nod1 and Gef2 suppress mutants in the SIN pathway and affect Sid2 kinase localization

The SIN pathway regulates contractile-ring maturation, stability, and septum formation (Krapp and Simanis, 2008; Roberts-Galbraith and Gould, 2008). We reported that gef2∆ suppresses cdc11-136 and sid2-250 mutants in the SIN pathway, but the mechanism is unknown (Ye ). We tested whether nod1∆ suppressed SIN mutants, using gef2∆ as a control (see Table 1 later in the paper). Both gef2∆ and nod1∆ partially restored cell growth of cdc7-24 at 30°C (Figure 6A, top) and of cdc11-136 at both 30 and 36°C (Figure 6A, middle). Surprisingly, unlike gef2∆, nod1∆ did not suppress sid2-250 (Figure 6A, bottom; see Discussion). To explore the mechanism of the suppression of SIN mutants by gef2∆ and nod1∆, we examined cell morphology of SIN single mutants and SIN gef2∆ or SIN nod1∆ double-mutant cells. cdc7-24 and sid2-250 displayed cell lysis (Figure 6B and Supplemental Figure S2). Except for nod1∆ sid2-250, all double mutants partially restored cell viability by reducing cell lysis. Approximately 60% gef2∆ sid2-250 cells survived at a semipermissive temperature of 30°C, whereas only ∼20% sid2-250 and nod1∆ sid2-250 cells were viable (Figure 6C). On the other hand, cells overexpressing Gef2 from 3nmt1 or 41nmt1 promoter under inducing conditions were synthetic lethal with sid2-250 at 30°C and synthetic sick with sid2-1 from 30 to 36°C (Figure 6D). Taken together, our data suggest that both Nod1 and Gef2 negatively affect the SIN pathway or the process regulated by the pathway.

FIGURE 6:

nod1∆ and gef2∆ suppress SIN mutants by reducing cell lysis. (A) Serial dilutions (3×) of indicated strains on YE5S or YE5S + phloxin B (red dye accumulated in dead cells) plates at 25, 30, and 36°C. Strains used: wt (JW81), cdc7-24 (TP34), nod1∆ (JW4259), nod1∆ cdc7-24 (JW4304), gef2∆ (JW2854), gef2∆ cdc7-24 (JW3021), cdc11-136 (TP47), nod1∆ cdc11-136 (JW4306), gef2∆ cdc11-136 (JW2972), sid2-250 (YDM429), nod1∆ sid2-250 (JW4294), and gef2∆ sid2-250 (JW3009). (B, C) gef2∆ but not nod1∆ partially rescued cell lysis in sid2-250. Cells were grown in liquid culture at 25°C and then shifted 30°C for 6 h. (B) DIC images of sid2 mutant strains used in A. (C) Percentage of viable cells. Dead or lysed cells were identified as those that failed to maintain their cytoplasm. (D) Overexpression of Gef2 is synthetic lethal with sid2 mutants. Serial dilutions (3×) of indicated strains on YE5S or YE5S + phloxin B plates at 25, 30, and 36°C. Strains used: JW81, JW3561, JW3562, YDM429, JW5360, JW5361, VS2367, JW5405, and JW5406. (E, F) Sid2 localization at the division site is compromised in nod1∆ and gef2∆. Time 0 marks the end of anaphase B. (E) Time courses of representative cells expressing Sid2-GFP in wt (YDM415), gef2∆ (JW5580), and nod1∆ (JW5581). (F) Quantification of the intensity (mean ± SEM) of Sid2-GFP at the division site for strains in E. Black arrow and dashed line mark time 0. Bars, 5 μm.

We next tested whether Sid2 localization is affected in gef2∆ and nod1∆. Sid2 localizes to the SPB, the contractile ring, and the septum during cytokinesis (Sparks ). Sid2 appeared at the contractile ring at the beginning of anaphase B, and the level gradually increased until the contractile ring started to constrict (Figure 6, E, top row, and F) as reported (Sparks ; Tebbs and Pollard, 2013). In gef2∆ and nod1∆, Sid2 appeared at the contractile ring at a similar timing as in wt. Recruitment of Sid2 to the division site, however, was defective. By the end of anaphase B, Sid2 intensity at the division site in gef2∆ and nod1∆ was only ∼20% of that in wt (Figure 6, E, middle and lower rows, and F; p < 0.001 for both gef2∆ and nod1∆ vs. wt). Moreover, the peak level of Sid2 at the division site in gef2∆ and nod1∆ was reduced to 57 and 46% that of wt (Figure 6F; p < 0.005 for both gef2∆ and nod1∆ vs. wt). Both wt and mutant cells expressing Sid2-GFP spent more time in mitosis. Because Sid2 regulates proper spindle elongation during anaphase (Mana-Capelli ), it seems that Sid2-GFP may not be fully functional. Together these data suggest that Gef2 and Nod1 play a role in recruiting Sid2 to the contractile ring.

Gef2 interacts with Rho GTPases in vitro and is involved in Rho4 localization

Rho GTPases regulate contractile-ring formation, septum formation, and degradation during cytokinesis (Arellano ; Nakano , 2003, 2005; Tolliday ; Santos ; Tajadura ; Mutoh ; Yoshida ). To further dissect the role of Gef2, we tested the interactions between the GEF domain of Gef2 and all six Rho GTPases from S. pombe. The hexahistidine (6His)-tagged GEF domain (aa 211–600) of Gef2 consisting of the DBL homology (DH) and pleckstrin homology (PH) domains was purified from Escherichia coli. The purified GEF domain was then pulled down by purified glutathione S-transferase (GST)–tagged Rho proteins. We found that Gef2 interacted with Rho1, Rho4, and Rho5 but not with Rho2, Rho3, and Cdc42 in the pull-down assays (Figure 7, A and B).

FIGURE 7:

Gef2 GEF domain binds to GTPases Rho1, Rho4, and Rho5 in vitro. (A, B) Purified GST-Rho GTPases and GST control were bound to the beads and then incubated with purified His‑GEF domain (aa 211–600) of Gef2. The amount of pulled down Gef2 was detected by Western blotting (A) and quantified (B). The intensities of His-Gef2(GEF) bands were measured, background subtracted, corrected for Rho GTPase amount, and normalized by setting the intensity of His-Gef2(GEF) in GST control as 1. The experiment was repeated, and mean ± SD is shown in B. (C–E) rho4∆ suppresses SIN mutants. Strains used: JW81, JW3041, YDM429, JW5505, TP34, JW5503, TP47, and JW5504. (C) Serial dilutions (3×) of indicated strains on YE5S or YE5S + phloxin B plates at 25, 30, 32, and 36°C for 3 d. (D, E) rho4∆ rescues the cell-lysis phenotype of sid2-250. (D) DIC images of cells grown in liquid culture at 25°C or after 6 h at 30°C. (E) Quantification of viable (not lysed or dead) cells after 6 h at 30°C. (F, G) Gef2 and Nod1 play a role in Rho4 localization. (F) Micrographs of GFP-Rho4 in wt (PPG1580) and the deletion mutants (JW4909 and JW4910). (G) Quantification of Rho4 intensity at the division site for strains in (F). Bars, 5 μm.

To investigate whether Gef2 might function through a Rho GTPase in vivo, we crossed rho4∆ to mutants in the SIN pathway, since rho4∆, like gef2∆, has been shown to suppress sid2-250 (Jin ). We found that in addition to rescuing sid2-250 at both 25 and 30°C, rho4∆ also partially rescued cdc7-24 at 30°C and cdc11-136 at 30–36°C (Figure 7C). We next observed the cell morphology of rho4∆ sid2-250 at 25°C or after 6 h at 30°C (Figure 7D). At 25°C, both rho4∆ and rho4∆ sid2-250 resembled wt, whereas sid2-250 displayed slight cell lysis. At 30°C, only ∼20% sid2-250 cells were viable, whereas ∼85% cells survived in rho4∆ sid2-250 double mutant (Figure 7E). Thus rho4∆ resembled gef2∆ (Figure 6, A–C) in the suppression of the SIN mutants. Together these data suggest that Gef2 functions through Rho4 GTPase to regulate late cytokinesis.

We next determined whether Gef2 or Nod1 affect Rho4 localization. GFP-Rho4 localized to the cell-division site, as well as to the cell periphery, in wt cells (Nakano ; Santos ). Although its localization was not abolished, Rho4 intensity at the division site was reduced to 82 and 75% of wt level in gef2∆ and nod1∆, respectively (Figure 7, F and G; p < 0.005 for both gef2∆ and nod1∆ vs. wt). Thus Gef2 and Nod1 are involved in concentrating Rho4 GTPase to the division site during cytokinesis.

DISCUSSION

In this study we found that Nod1, a new player in cytokinesis, regulates division-site positioning and contractile-ring stability together with the putative Rho-GEF Gef2 (Figure 8). In addition, we identified the potential Rho GTPase substrates for Gef2, suggesting the possible involvement of Gef2 GEF activity and Rho GTPases in the regulation of cytokinesis.

FIGURE 8:

Model of Nod1 and Gef2 localization and interactions with other proteins on the cytoplasmic side of the plasma membrane during the cell cycle. i) During interphase, Nod1 and Gef2 localize to interphase nodes via Blt1 or other interphase-node proteins, ii) where they help to recruit and stabilize anillin-related protein Mid1. iii) The nodes mature into cytokinesis nodes and coalesce into the contractile ring as more Mid1 and other cytokinesis proteins like F-BAR protein Cdc15 arrive at the division site. iv) Cdc15 continuously recruits or stabilize the Nod1/Gef2 complex during ring maturation, which helps to maintain the contractile-ring integrity and stability. v) Mid1 disappears from the ring at the onset of its constriction. For clarity, the potential interactions between Gef2 and Rho GTPases are not shown.

The roles of Rho GTPases during cytokinesis

Among the seven Rho GEFs in S. pombe, Gef2 and Gef3 have no identified Rho substrates. We find that Gef2 interacts with Rho1, Rho4, and Rho5 in vitro (Figure 7, A and B). It is unclear whether Gef2 interacts with and activates these Rho GTPases in vivo, but these data provide insight into Gef2’s functions as a potential Rho GEF. In previous study, we reported that deletion of Gef2 DH domain causes defects in division-site positioning in ∼50% plo1-ts18 mutant cells at 25°C (Ye ). Therefore it is possible that the GEF activity of Gef2 is involved in division-site placement. Rho1 regulates cell integrity and septum formation during late cytokinesis in fission yeast (Nakano ; Mutoh ). Its homologues RhoA or Rho1 in animal cells and budding yeast, however, are active in early cytokinesis for division-site selection and contractile-ring assembly (Imamura ; Tolliday ; Bement ; Piekny ; Yoshida ; Watanabe ). The presence and function of Gef2 in the cortical nodes might suggest a role of Rho1 during early cytokinesis if Gef2 indeed activates Rho1 in vivo. One difficulty in studying RhoA or Rho1, however, is that its native concentration is low, and therefore it is difficult to detect Rho1 at the division site during early cytokinesis by fluorescence microscopy. Whether Rho1 participates in division-site positioning in fission yeast remains to be tested, and we cannot rule out the possibility that other Rho candidates are also involved.

Of the six Rho GTPases in fission yeast, Rho1 and Cdc42 are relatively well studied, whereas our knowledge on Rho2-5 is limited. For example, no Rho GEFs have been assigned to Rho2, Rho3, and Rho4. Rho4 affects the localization and activity of β-glucanase Eng1 and α-glucanase Agn1, which results in cell separation defects (Nakano ; Santos , 2005). Rho5 is a Rho1 paralogue that shares similar functions (Nakano ). How Rho4 and Rho5 are regulated and localized is unknown. Our data suggest that Gef2 might be a GEF for Rho4 or Rho5 and help recruit Rho4 to the division site. Rho4 localization, however, is only partially dependent on Gef2 (Figure 7, F and G). More efforts are needed to investigate whether and how Gef2 works with these Rho GTPases.

Localization of Nod1 and Gef2 during the cell cycle

We and others found that Gef2 coordinates with Polo kinase, Plo1, to recruit anillin-like protein Mid1 to the cortical nodes during G2/M transition (Ye ; Guzman-Vendrell ; Jourdain ). During the course of that study, we identified Nod1 as a Gef2-related protein and binding partner. We found that Gef2 and Nod1 form a complex, which is important for their cortical node localization and functions. These results are consistent with a recent report on Nod1 (Jourdain ). Gef2 and Nod1 are stable in interphase nodes, as revealed by FRAP assays. Besides a GEF (DH-PH) domain, Gef2 has no other known structures or motifs (Figure 1A; Iwaki ). Blt1 was reported to recruit Gef2 to the interphase nodes (Ye ; Guzman-Vendrell ; Jourdain ). It is likely that Blt1 interacts with Nod1 and Gef2 through their C-termini (Figure 8). Both Nod1 and Gef2 have enriched α-helix structures at C-termini (Jones, 1999). Gef2 still localizes to cytokinesis nodes in blt1∆, so Gef2 must have other binding partners during early mitosis. We previously showed that Gef2 interacts with Mid1(300–350) in vivo (Ye ). Although we found that Mid1(1–580), which includes the Gef2-binding region, depended on Gef2 C-terminus for node localization, no positive interactions were observed between Mid1(300–350) and several regions of Gef2 or Nod1 in yeast two-hybrid assays (unpublished data). Thus the interactions between Gef2 and Mid1 may be indirect.

Although the majority of Gef2 is recruited to the contractile ring through the cortical nodes, our localization dependence data reveal that both Nod1 and Gef2 are capable of localizing to the contractile ring without each other. We find that F-BAR protein Cdc15 physically interacts with Nod1 and recruits Nod1 to the contractile ring (Figures 4 and 8). Cdc15 appears at cytokinesis nodes ∼5 min before SPB separation and is continuously recruited to the contractile ring during mitosis (Wu and Pollard, 2005; Laporte ). Consistently, the contractile ring contains ∼40% more molecules of Nod1 than Gef2 (Figure 1C). Nod1 intensity at the contractile ring in gef2∆ also increases during ring maturation at late mitosis. Without Nod1, Gef2 can still localize to the division site during later stages of cytokinesis (Figure 2A), although Gef2 does not interact with Cdc15 in yeast two-hybrid assays. It is possible that Gef2 depends on alternative mechanisms to localize. One attractive candidate is a Rho GTPase. We found that Gef2 can interact with Rho1, Rho4, and Rho5, and all of them localize to the division site at late cytokinesis (Nakano , 2005; Santos ; Mutoh ). In budding yeast, activated Cdc42 recruits the Rho-GEF Cdc24 and scaffold protein Bem1 to activate more Cdc42 and establish cell polarity (Butty ; Slaughter ; Bi and Park, 2012). It is possible that Gef2 and its Rho substrates are involved in a similar positive feedback loop to regulate cytokinesis.

Nod1 and Gef2 coordinate with F-BAR protein Cdc15 to maintain contractile-ring stability

Cdc15 has multiple functions during cytokinesis. During early cytokinesis, Cdc15 recruits the formin Cdc12 to promote contractile-ring assembly (Carnahan and Gould, 2003; Kovar ; Laporte ). During ring maturation at anaphase, Cdc15, together with the SIN pathway and the F-BAR protein Imp2, is believed to be important for maintaining contractile-ring stability and integrity (Wachtler ; Hachet and Simanis, 2008; Huang ; Roberts-Galbraith ). The exact mechanism remains elusive.

Here we add another layer of complexity to the function of Cdc15 during late cytokinesis. In nod1∆ cdc15-140 and gef2∆ cdc15-140, most cells form a fragile contractile ring and become elongated and multinucleated (Figure 5). The severely reduced level of the myosin regulatory light chain Rlc1 suggests loss of proteins from the contractile ring (Figure 5, E–G, Supplemental Figure S1, and Supplemental Videos S1–S3). One possible explanation could be related to the scaffolding protein Mid1. Mid1 is anchored to the equatorial cortex through the cooperation of its own lipid-binding domains and other cytokinesis proteins, including Cdr2, Gef2, and Blt1 (Almonacid ; Lee and Wu, 2012; Ye ; Guzman-Vendrell ). Mid1 is more dynamic and mobile at the division site without Gef2 (Ye ). As a result, the recruitment and maintenance of the contractile-ring components might be less effective during late mitosis, which aggravates the cdc15-mutant phenotype. It is also possible that Rho1 and/or Rho5 GTPases are also involved in contractile-ring stability and their activities are compromised in nod1∆ and gef2∆ cells. Further experiments are needed to distinguish these possibilities.

Nod1 and Gef2 suppress the SIN pathway

The SIN pathway includes a small GTPase and several protein kinases and their adaptors, which form a kinase cascade on the SPB (Fankhauser and Simanis, 1993, 1994; Furge , 1999; Sparks ; Chang and Gould, 2000; Guertin ; Hou ; Salimova ; Tomlin ). The activation of SIN pathway leads to contractile-ring constriction and septum formation (Wachtler ; Hachet and Simanis, 2008; Krapp and Simanis, 2008; Johnson ). This is executed by translocation of kinase Sid2 and its adaptor Mob1 from the SPB to the contractile ring (Sparks ; Hou ; Chen ). Discoveries of suppressors of SIN pathway mutants, especially those of sid2, have helped us understand how SIN pathway regulates cytokinesis (Jiang and Hallberg, 2001; Jin and McCollum, 2003; Jin ; Goyal and Simanis, 2012). Here we found that nod1∆ and gef2∆ suppress the SIN mutants by improving cell survival at the semipermissive temperature, whereas single-SIN-mutant cells lyse when trying to separate with defective septa (Figure 6, A–C, and Table 1). We also observed that Sid2 accumulation at the division site is delayed and compromised in nod1∆ and gef2∆ cells (Figure 6, E and F). Similar results were observed in IQGAP rng2 without the IQ motifs (Tebbs and Pollard, 2013), suggesting a requirement of intact contractile ring for Sid2 stable localization. Therefore the contractile ring components, including Gef2 and Nod1, may regulate the SIN pathway through direct or indirect influence on contractile-ring localization of Sid2. It is still possible, however, that the defects caused by nod1∆ and gef2∆ affect the rates of contractile-ring maturation and constriction, allowing more time for septum synthesis. Consistently, increasing the amount and activity of β-glucan synthase Bgs1 by overexpressing Rho1 GTPase or its GEF Rgf3 can rescue sid2 mutants (Jin ).

Rho4 GTPase, however, might be also involved in the suppression of sid2-250 by gef2∆. We found that Gef2 binds to Rho4 in vitro. Of interest, deletion of rho4 or its effector eng1 or agn1 partially suppresses sid2-250 (Jin ), which is consistent with our results (Figure 7C–E). Thus it is likely that suppression of SIN mutants by gef2∆ is due to a reduced function of Rho4 and its effectors. Consistently, we found that Rho4 localization to the division site was slightly but significantly reduced in both nod1∆ and gef2∆ cells (Figure 7, F and G). This suggests that Gef2 and Nod1 contribute to Rho4 localization besides the undefined role of Rho4 activation. The cell-separation defect of rho4∆ is mild even at 36°C (Santos ), suggesting that other mechanisms and pathways are involved in septum degradation. Further studies are needed to identify the redundant pathways.

In conclusion, we find that the Nod1/Gef2 complex functions in division-site positioning, contractile-ring maintenance, and septation besides its role in cell-size control. We also discover the potential Rho GTPase substrates for Gef2. It will be very informative to investigate whether Gef2 has GEF activity toward the Rho GTPase candidates and whether Nod1 affects Gef2 activity in addition to its localization.

MATERIALS AND METHODS

Strains and genetic, molecular, and cellular methods

Table 2 lists the strains used in this study. We used PCR-based gene targeting and standard yeast genetics to construct strains (Moreno ; Bähler ). All tagged and truncation strains are regulated under endogenous promoters or 5′ untranslated region (UTR) and integrated into native chromosomal loci, except for the overexpression strains that are integrated at native loci under the control of 3nmt1 or 41nmt1 promoter, which is repressed by thiamine (Maundrell, 1990).

TABLE 2:

S. pombe strains used in this study.

Strain	Genotype	Source/reference
JW81	h− ade6-210 ura4-D18 leu1-32	Wu et al. (2003)
JW1063	h+ mYFP-cdc15 ade6-M216 leu1-32 ura4-D18	Wu and Pollard (2005)
JW1636	h+ mid1-6 ade6-M210 leu1-32 ura4-D18	Coffman et al. (2013)
JW1743	cdc15-140 ade6-M210 leu1-32 ura4-D18	Coffman et al. (2013)
JW1824	h+ klp8∆::kanMX4 ade6 leu1-32 ura4-D18	Kim et al. (2010)
JW1825	h+ blt1∆::kanMX4 ade6-M216 leu1-32 ura4-D18	Ye et al. (2012)
JW1826	h+ gef2∆::kanMX4 ade6 leu1-32 ura4-D18	Ye et al. (2012)
JW2249	rng2-346 ade6-M210 leu1-32 ura4-D18	This study
JW2255	h+ mid1-366 ade6-M210 leu1-32 ura4-D18	Ye et al. (2012)
JW2854	h+ gef2∆::hphMX6 ade6 leu1-32 ura4-D18	This study
JW2937	cdc15-140 gef2∆::kanMX4 ade6 leu1-32 ura4-D18	This study
JW2972	h+ cdc11-136 gef2∆::hphMX6 ade6 leu1-32 ura4-D18	Ye et al. (2012)
JW3009	gef2∆::hphMX6 sid2-250 ade6 leu1-32 ura4-D18	Ye et al. (2012)
JW3021	gef2∆::hphMX6 cdc7-24 ade6 leu1-32 ura4-D18	This study
JW3041	h+ rho4∆::kanMX4 ade6 leu1-32 ura4-D18	Kim et al. (2010)
JW3078	h− gef2∆::hphMX6 plo1.ts18::ura4+ ade6 leu1-32 ura4-D18	Ye et al. (2012)
JW3204	h− gef2-13Myc-hphMX6 ade6-M210 leu1-32 ura4-D18	Ye et al. (2012)
JW3325	gef2-13Myc-hphMX6 mYFP-cdc15 ade6-M210 leu1-32 ura4-D18	This study
JW3561	h− kanMX6-3nmt1-gef2 ade6-M216 leu1-32 ura4-D18	This study
JW3562	h− kanMX6-41nmt1-gef2 ade6-M216 leu1-32 ura4-D18	This study
JW3622	h+ gef2-13Myc-hphMX6 ade6-M210 leu1-32 ura4-D18	This study
JW3773	h− nod1∆::kanMX6 ade6-M210 leu1-32 ura4-D18	This study
JW3814	h+ nod1∆::kanMX6 gef2∆::kanMX4 ade6 leu1-32 ura4-D18	This study
JW3815	nod1∆::kanMX6 plo1.ts18::ura4+ ade6-M210 ura4-D18 leu1-32	This study
JW3825	h− kanMX6-Pgef2-mECitrine-4Gly-gef2 ade6-M216 leu1-32 ura4-D18	Ye et al. (2012)
JW3826	h− kanMX6-Pgef2-mECitrine-4Gly-gef2-(957-1101) ade6-M216 leu1-32 ura4-D18	Ye et al. (2012)
JW3861	h+ nod1∆::kanMX6 mid1-6 ade6-M210 leu1-32 ura4-D18	This study
JW3873	nod1∆::kanMX6 gef2∆::kanMX4 plo1.ts18::ura4+ade6 leu1-32 ura4-D18	This study
JW3875	h− nod1∆::kanMX6 mid1-366 ade6-M210 leu1-32 ura4-D18	This study
JW4008	h− nod1-mECitrine-kanMX6 ade6-M210 leu1-32 ura4-D18	This study
JW4010	h− nod1-tdTomato-hphMX6 ade6-M210 leu1-32 ura4-D18	This study
JW4013	h− nod1-13Myc-hphMX6 ade6-M210 leu1-32 ura4-D18	This study
JW4014	nod1∆::kanMX6 kanMX6-Pgef2-mECitrine-4Gly-gef2 ade6 leu1-32 ura4-D18	This study
JW4015	h− nod1Δ::kanMX6 cdc4-8 ade6 leu1-32 ura4-D18	This study
JW4016	h− nod1∆::kanMX6 cdc15-140 ade6-M210 leu1-32 ura4-D18	This study
JW4038	nod1-mECitrine-kanMX6 gef2∆::hphMX6 ade6 leu1-32 ura4-D18	This study
JW4042	nod1∆::kanMX6 rng2-D5 ade6-M210 leu1-32 ura4-D18	This study
JW4043	h+ nod1∆::kanMX6 rng2-346 ade6-M210 leu1-32 ura4-D18	This study
JW4098	nod1∆::kanMX6 cdr2∆::kanMX6 ade6 leu1-32 ura4-D18	This study
JW4099	h+ nod1∆::kanMX6 blt1∆::kanMX4 ade6 leu1-32 ura4-D18	This study
JW4226	h+ kanMX6-Pgef2-tdTomato-4Gly-gef2 ade6-M210 leu1-32 ura4-D18	Ye et al. (2012)
JW4256	nod1-tdTomato-hphMX6 kanMX6-Pgef2-mECitrine-4Gly-gef2-(957-1101) ade6 leu1-32 ura4-D18	This study
JW4259	h− nod1∆::hphMX6 ade6-M210 leu1-32 ura4-D18	This study
JW4294	nod1∆::hphMX6 sid2-250 ade6-M210 leu1-32 ura4-D18	This study
JW4295	klp8∆::kanMX4 nod1∆::hphMX6 ade6 leu1-32 ura4-D18	This study
JW4304	nod1∆::hphMX6 cdc7-24 ade6 leu1-32 ura4-D18 his2 or his7	This study
JW4306	nod1∆::hphMX6 cdc11-136 ade6-M210 leu1-32 ura4-D18 his2 or his7	This study
JW4325	h− nod1(1-209)-mECitrine-kanMX6 ade6-M210 leu1-32 ura4-D18	This study
JW4326	h− nod1(1-328)-mECitrine-kanMX6 ade6-M210 leu1-32 ura4-D18	This study
JW4330	nod1-13Myc-hphMX6 kanMX6-Pgef2-mECitrine-4Gly-gef2 ade6 leu1-32 ura4-D18	This study
JW4331	nod1-13Myc-hphMX6 kanMX6-Pgef2-mECitrine-4Gly-gef2-(957-1101) ade6 leu1-32 ura4-D18	This study
JW4355	nod1-tdTomato-hphMX6 kanMX6-Pgef2-mECitrine-4Gly-gef2-(1-956)-TADH1-hphMX6 ade6 leu1-32 ura4-D18	This study
JW4359	h− nod1(1-328)-mECitrine-kanMX6 kanMX6-Pgef2-tdTomato-4Gly-gef2 ade6 leu1-32 ura4-D18	This study
JW4453	h− kanMX6-Pnod1-mECitrine-nod1 ade6-M210 leu1-32 ura4-D18	This study
JW4455	h− kanMX6-Pnod1-mECitrine-nod1(329-419) ade6-M210 leu1-32 ura4-D18	This study
JW4457	nod1-mEGFP-hphMX6 kanMX6-Pgef2-tdTomato-4Gly-gef2 ade6-M210 leu1-32 ura4-D18	This study
JW4750	Pnod1-mECitrine-nod1 ade6-M210 leu1-32 ura4-D18	This study
JW4856	h+ Pnod1-mECitrine-nod1(329-419) ade6-M210 leu1-32 ura4-D18	This study
JW4909	rho4∆::kanMX6 leu1::GFP-rho4 gef2∆::kanMX4 leu1-32 ura4-D18 ade6	This study
JW4910	h− rho4∆::kanMX6 leu1::GFP-rho4 nod1∆::kanMX6 leu1-32 ura4-D18	This study
JW4912	Pgef2-mECitrine-4Gly-gef2 ade6 leu1-32 ura4-D18	This study
JW5027	cdc15-140 nod1-mECitrine-kanMX6 gef2∆::hphMX6 ade6 leu1-32 ura4-D18	This study
JW5028	cdc15-140 nod1-mECitrine-kanMX6 ade6-M210 leu1-32 ura4-D18	This study
JW5065	h+ Pnod1-mECitrine-nod1(210-419) ade6-M210 leu1-32 ura4-D18	This study
JW5093	kanMX6-Pnod1-mECitrine-nod1 gef2-13Myc-hphMX6 ade6-M210 leu1-32 ura4-D18	This study
JW5095	kanMX6-Pnod1-mECitrine-nod1(329-419) gef2-13Myc-hphMX6 ade6-M210 leu1-32 ura4-D18	This study
JW5107	kanMX6-Pgef2-tdTomato-4Gly-gef2 kanMX6-Pnod1-mECitrine-nod1(329-419) ade6-M210 leu1-32 ura4-D18	This study
JW5120	nod1-13Myc-hphMX6 mYFP-cdc15 ade6 leu1-32 ura4-D18	This study
JW5329	h+ gef2∆::kanMX4 cdc15-140 GFP-bgs1-leu1+ bgs1∆::ura4+ rlc1-tdTomato-natMX6 ade6 leu1-32 ura4-D18	This study
JW5330	nod1∆::kanMX6 cdc15-140 GFP-bgs1-leu1+ bgs1∆::ura4+ rlc1-tdTomato-natMX6 ade6-M210 leu1-32 ura4-D18	This study
JW5357	h− cdc15-140 GFP-bgs1-leu1+ bgs1∆::ura4+ rlc1-tdTomato-natMX6 ade6-M210 leu1-32 ura4-D18	This study
JW5360	sid2-250 kanMX6-3nmt1-gef2 ade6 leu1-32 ura4-D18	This study
JW5361	sid2-250 kanMX6-41nmt1-gef2 ade6 leu1-32 ura4-D18	This study
JW5405	sid2-1 kanMX6-3nmt1-gef2 ade6 leu1-32 ura4-D18	This study
JW5406	sid2-1 kanMX6-41nmt1-gef2 ade6 leu1-32 ura4-D18	This study
JW5503	rho4∆::kanMX4 cdc7-24 ade6 leu1-32 ura4-D18 his7-366	This study
JW5504	rho4∆::kanMX4 cdc11-136 ade6 leu1-32 ura4-D18	This study
JW5505	rho4∆::kanMX4 sid2-250 ade6 leu1-32 ura4-D18	This study
JW5580	gef2∆::kanMX4 sid2-GFP-ura4+ade6-M210 leu1-32 ura4-D18	This study
JW5581	nod1∆::kanMX6 sid2-GFP-ura4+ade6-M210 leu1-32 ura4-D18	This study
JW5582	Pgef2-mECitrine-4Gly-gef2 Pmyo2-mCFP-myo2 ade6 leu1-32 ura4-D18	This study
JW5583	cdc15-140 Pgef2-mECitrine-4Gly-gef2 Pmyo2-mCFP-myo2 ade6 leu1-32 ura4-D18	This study
IH1600	h+ plo1.ts18::ura4+ ura4-D18 leu1-32 ade6-M210 his2	MacIver et al. (2003)
JM578	h+ cdr2∆::kanMX6 ade6 leu1-32 ura4-D18	Moseley et al. (2009)
PPG1580	h− rho4∆::kanMX6 leu1::GFP-rho4 leu1-32 ura4-D18	Santos et al. (2003)
TP7	h− cdc4-8 his7-366 leu1-32 ura4-D18 ade6-M216	Thomas Pollard (Yale University, New Haven, CT)
TP34	h− cdc7-24 his7-366 leu1-32 ade6-M216 ura4-D18	Thomas Pollard
TP47	h− cdc11-136 ura4-D18 leu1-32 his7-366	Bezanilla et al. (1997)
VS2367	h+ sid2-1 ade6-M210 leu1-32 ura4-D18	Salimova et al. (2000)
YDM26	h− rng2-D5 ade6-210 ura4-D18 leu1-32	Eng et al. (1998)
YDM415	h− sid2-GFP-ura4+ ade6-M210 leu1-32 ura4-D18	Sparks et al. (1999)
YDM429	h+ sid2-250 ade6-M210 leu1-32 ura4-D18	Sparks et al. (1999)

Nod1 C-terminal truncations and Nod1 overexpression were constructed as previously described (Bähler ). For N-terminal truncations, nod1 5′ UTR −300 to +3 base pairs was cloned into pFA6a-kanMX6-P3nmt1-mECitrine at BglII and PacI sites to replace the 3nmt1 promoter. The resulting plasmid (JQW560) was then used as the template for PCR amplification and gene targeting. Primers were designed according to desired truncation sites, and the PCR products were transformed into wt cells. The resulting strains were sequenced. Some kanMX6 marker at 5′ end of nod1 or gef2 gene was looped out by crossing the strains to wt cells.

To test the functionalities of tagged FL Nod1, both N- and C-terminally tagged Nod1 strains were crossed to plo1-ts18. Double mutants had <10% abnormal septa at 25°C, which is similar to plo1-ts18 single mutant but different from the ∼95% abnormal septa in plo1-ts18 nod1∆. Thus both N- and C-terminally tagged Nod1 are functional.

For DNA staining, cells were incubated with 10 μg/ml Hoechst 33258 for 10 min in the dark before imaging in the 4′,6-diamidino-2-phenylindole (DAPI) channel as described (Wu ).

Microscopy and data analysis

Strains were restreaked from −80°C stock and grown 1–2 d on yeast extract plus five supplements (YE5S) plates at 25°C. Cells were then inoculated and kept in exponential phase for ∼48 h at 25°C except where noted. Before microscopy, cells were washed in Edinburgh minimal medium plus five supplements (EMM5S) twice to reduce autofluorescence and imaged on EMM5S with 20% gelatin pad with 5 μM n-propyl-gallate as described (Laporte ; Ye ). For long movies, cells were washed in YE5S and resuspended in YE5S with 5 μM n-propyl-gallate. Then 2-μl concentrated cells were spotted onto a coverglass-bottom dish (Delta TPG Dish; Biotechs, Butler, PA) and covered with a layer of YE5S agar before imaging at 23.5ºC or in a preheated climate chamber (stage top incubator INUB-PPZI2-F1 equipped with UNIV2-D35 dish holder; Tokai Hit, Shizuoka-ken, Japan) for imaging at the restrictive temperatures for certain mutants.

Microscopy was performed at 23.5–25ºC except where noted. To visualize cell morphology, DNA, and septum, Hoechst-stained cells were imaged with a 100×/1.4 numeral aperture (NA) Plan-Apo objective lens on a Nikon Eclipse Ti inverted microscope (Nikon, Melville, NY) equipped with a Nikon cooled digital camera DS-Ql1 and a DAPI filter. Other experiments were performed using 100×/1.4 NA Plan-Apo objective lenses (Nikon) on a spinning disk confocal microscope (UltraVIEW ERS; PerkinElmer Life and Analytical Sciences, Waltham, MA) with 440- and 568-nm solid state lasers and 488- and 514-nm argon ion lasers and an ORCA-AG camera (Hamamatsu, Bridgewater, NJ) with 2 × 2 binning, or on a spinning disk confocal microscope (UltraVIEW Vox CSUX1 system, PerkinElmer Life and Analytical Sciences) with 440-, 488-, 515-, and 561-nm solid-state lasers and a back-thinned, electron-multiplying charge-coupled device camera (Hamamatsu C9100-13) without binning.

Images were analyzed using ImageJ (National Institutes of Health, Bethesda, MD), UltraVIEW, or Volocity (PerkinElmer) software. Fluorescence images shown in figures and movies are maximum projections of images stacks at 0.4- to 0.6-μm spacing except where noted. Nod1 and Gef2 molecules in cells were counted globally or locally by measuring fluorescence intensity as described (Laporte ). Briefly, tagged Nod1 or Gef2 cells were mixed with wt cells and imaged with 11 z-sections with 0.4-μm spacing on the UltraVIEW ERS confocal system. The offset was subtracted from images that were then corrected for uneven illumination. Mean intensity in whole cells was measured in sum intensity projections and subtracted by that of wt cells as background. Mean intensity in the mature contractile ring was measured using the polygon region of interest (ROI) tool in ImageJ on a sum intensity projection. A ≥3× larger ROI that included the contractile ring was chosen for calculation of background intensity after subtracting ring intensity. For nodes, the fluorescence intensity was measured using a circular ROI with a diameter of five pixels that covered the whole node at the best focal plan. The intensity near the plasma membrane outside of the broad band of nodes was used for background subtraction to avoid overlapping with other nodes. The global and local intensities of Nod1 and Gef2 were then normalized to molecule numbers using previous Gef2 data as a reference (Wu and Pollard, 2005; Wu ; Ye )

FRAP analysis

FRAP assays were performed using the photokinesis unit on the UltraVIEW Vox confocal system, similar to the assays described before (Coffman ; Laporte ). The best focal plane for bleaching was chosen from z-stacks. Selected ROIs were bleached to <50% of the original fluorescence intensity after five prebleach images were collected. One hundred postbleach images with 10-s delay were collected. The images were then corrected for background and photobleaching during image acquisition at nonbleached sites. We normalized prebleach intensity of the ROI to 100%, the intensity just after bleaching to 0%, and the end of the bleach time as time 0. Intensity of every three consecutive postbleaching time points was averaged to reduce noise. The data were then plotted and fitted using the exponential equation y = m1 + m2 exp(−m3x), where m3 is the off-rate (KaleidaGraph; Synergy Software, Reading, PA). The half-time of recovery was calculated as t1/2 = (ln 2)/m3. The p values in this study were calculated using two-tailed Student's t tests.

IP and Western blotting

IP assay and Western blotting were carried out as previously described (Laporte ; Lee and Wu, 2012). Briefly, mECitrine-tagged proteins were pulled down from fission yeast cell extract by protein G covalently coupled magnetic Dynabeads (100.04D; Invitrogen, Carlsbad, CA) with polyclonal anti-GFP antibodies (NB600-308; Novus Biologicals, Littleton, CO). The bead samples were then boiled in sample buffer after washing three times. The protein samples were then separated in SDS–PAGE, and Western blotting was performed using monoclonal anti-GFP antibody (11814460001, 1:2000 dilution; Roche, Mannheim, Germany) or monoclonal anti-Myc antibody (9E10, 1:5000 dilution; Santa Cruz Biotechnology, Santa Cruz, CA). The anti-tubulin monoclonal TAT1 antibody was used at 1:20,000 dilution (Woods ). Anti-mouse secondary antibody was used at 1:5000 dilution.

Yeast two-hybrid assays

β-Galactosidase activity assays were performed to semiquantitatively detect protein interactions in yeast two-hybrid assays (Laporte ). DNAs or cDNAs of interest were constructed into vectors with either VP16 activation domain or GBT9 DNA-binding domain. The pairs of plasmids were then cotransformed into Saccharomyces cerevisiae strain MAV203 (11281-011; Invitrogen) and plated on solid medium lacking leucine and tryptophan (SD-L-W). The transformants were selected and used for β-galactosidase activity measurements in the o-nitrophenyl β-d-galactopyranoside assay (Sigma-Aldrich, St. Louis, MO). The results are displayed as fold changes over the highest negative control value.

Protein purification and the interaction between Gef2 and Rho GTPases

Pull-down assays between recombinant 6His-Gef2 (GEF) and GST-Rho proteins were adapted from a previous study (Iwaki ). Expression of 6His-tagged GEF domain of Gef2 (aa 211–600) was induced when ArcticExpress RIL cells (230193; Agilent Technologies, Santa Clara, CA) carrying the plasmid were grown at 10ºC for 18 h after adding 1 mM isopropyl-β-d-thiogalactoside (IPTG; Saha and Pollard, 2012b). After sonication (output 9, 50% duty cycle, 4× 20 pulses) and ultracentrifugation (25,000 rpm for 15 min, then 38,000 rpm for 30 min), 6His-Gef2 (GEF) was purified on Talon Metal Affinity Resin (635501; Clontech, Mountain View, CA) followed by gel filtration with a HiLoad 16×60 Superdex 200 (17-5175-01; GE Healthcare, Buckinghamshire, United Kingdom) in phosphate buffer (50 mM sodium phosphate, pH 6.2, 0.3 M NaCl, 1 mM dithiothreitol [DTT]). The purified His-Gef2 (GEF) was then dialyzed into the final binding buffer (25 mM 3-(N-morpholino)propanesulfonic acid, pH 7.2, 60 mM β-glycerophosphate, 15 mM p-nitrophenyl phosphate, 1 mM DTT, 1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, and protease inhibitor tablets). GST and GST-Rho1 to Rho5 and Cdc42 were purified from BL21(DE3)pLysS cells (69451; Novagen, EMD Chemicals, Darmstadt, Germany; induced with 0.5 mM IPTG at 15°C for 6 h) using glutathione–Sepharose beads (17-5132-01; GE Healthcare). The beads with Rho proteins were then incubated at 30°C for 10 min with buffer containing 50 mM Tris (pH 7.5), 1 mM DTT, and 5 mM EDTA to deplete nucleotides. Then 500 μl of 0.25 μM 6His-Gef2 (GEF) in binding buffer was added to 30 μl of beads with each nucleotide-depleted Rho protein and incubated at 4°C for 1 h. After incubation, glutathione beads were washed with 1 ml of binding buffer three times, and the bound proteins were detect by Western blotting. Rho GTPases were detected by monoclonal anti-GST antibody (3G10/1B3, 1:5000 dilution; NB600-446, Novus Biologicals), and bound 6His-Gef2 (GEF) was detected by anti-His antibody (631212, 1:10,000 dilution; Clontech). Secondary anti-mouse antibody was used at 1:5000 dilution.