Bni5p, a septin-interacting protein, is required for normal septin function and cytokinesis in Saccharomyces cerevisiae

Abstract

In the budding yeast Saccharomyces cerevisiae, the Cdc3p, Cdc10p, Cdc11p, Cdc12p, and Sep7p/Shs1p septins assemble early in the cell cycle in a ring that marks the future cytokinetic site. The septins appear to be major structural components of a set of filaments at the mother-bud neck and function as a scaffold for recruiting proteins involved in cytokinesis and other processes. We isolated a novel gene, BNI5, as a dosage suppressor of the cdc12-6 growth defect. Overexpression of BNI5 also suppressed the growth defects of cdc10-1, cdc11-6, and sep7Delta strains. Loss of BNI5 resulted in a cytokinesis defect, as evidenced by the formation of connected cells with shared cytoplasms, and deletion of BNI5 in a cdc3-6, cdc10-1, cdc11-6, cdc12-6, or sep7Delta mutant strain resulted in enhanced defects in septin localization and cytokinesis. Bni5p localizes to the mother-bud neck in a septin-dependent manner shortly after bud emergence and disappears from the neck approximately 2 to 3 min before spindle disassembly. Two-hybrid, in vitro binding, and protein-localization studies suggest that Bni5p interacts with the N-terminal domain of Cdc11p, which also appears to be sufficient for the localization of Cdc11p, its interaction with other septins, and other critical aspects of its function. Our data suggest that the Bni5p-septin interaction is important for septin ring stability and function, which is in turn critical for normal cytokinesis.

FIG. 1.

Suppression of the cdc11-6 and cdc12-6 growth defects by overexpression of BNI5. (A) Strain KLY1419 (cdc11-6) carrying a control vector (YCplac111) (no. 1), a low-copy-number CDC11 plasmid (pKL1064) (no. 2), or a multicopy BNI5 plasmid (pKL1119) (no. 3) was streaked onto YEP-glucose plates and grown at the indicated temperatures for 3 days. (B) Similar tests were performed with the cdc12-6 strain KLY1422 carrying plasmid YCplac111 (no. 1), pKL1072 (CDC12) (no. 2), or pKL1119 (BNI5) (no. 3).

FIG. 2.

Loss of BNI5 function leads to defects in septin localization and cytokinesis. (A) Strains KLY1350 (BNI5 YFP-CDC10) and KLY1831 (bni5Δ YFP-CDC10) grown in YEP-glucose at 23°C were shifted to 37°C for 12 h, fixed with formaldehyde, and examined by confocal microscopy. Representative morphologies are shown for each strain. Arrows indicate structures discussed in the text. WT, wild type. (B) Strain KLY1831 (bni5Δ YFP-CDC10), cultured as in panel A, was fixed with formaldehyde and subjected to immunostaining with anti-Cdc11p antibodies. (C) Strain KLY1940 (bni5Δ), cultured as in panel A, was fixed and stained with DiI to reveal membrane structures and stained with DAPI to visualize the DNA. The arrow indicates a neck with apparently connected cytoplasm. (D) Cells from the culture in panel C were stained with Calcofluor and subjected to confocal microscopy in 100-nm sections to investigate whether septa are formed. Arrows indicate seemingly incomplete primary septa as seen in the appropriate planes of focus. Bars, 5 μm.

FIG. 3.

Synthetic effects in bni5Δ septin and bni5Δ hof1Δ double mutants. (A) Septin single-mutant and septin bni5Δ double-mutant strains were cultured overnight at 23°C and shifted to 30°C (or 37°C for sep7Δ and sep7Δ bni5Δ) for 3 h before fixation and examination by DIC microscopy. Strains: cdc3-6, KLY1415; cdc10-1, KLY1798; cdc11-6, KLY1419; cdc12-6, KLY1422; sep7Δ, KLY3937; cdc3-6 bni5Δ, KLY2177; cdc10-1 bni5Δ, KLY1803; cdc11-6 bni5Δ, KLY2174; cdc12-6 bni5Δ, KLY2010; sep7Δ bni5Δ, KLY3941. Bar, 5 μm. (B) Strains KLY2214 (cdc12-6 YFP-CDC10) and KLY2216 (cdc12-6 bni5Δ YFP-CDC10) were cultured overnight at 23°C and then shifted to 30°C. Samples were fixed for microscopic analyses during growth at 23°C and 3 h after the shift to 30°C. Bar, 5 μm. (C) Strains KLY1546 (wild type) (lane 1), KLY2214 (lane 2), and KLY2216 (lane 3) were cultured overnight. These cultures were serially diluted, spotted on YEP-glucose (YEPD), and incubated at the indicated temperature for 3 days. (D) Quantitation of cells with YFP-Cdc10p localization at the neck as a function of time. The strains used in panel B were cultured overnight at 23°C and then shifted to 30°C. Samples were fixed for microscopic analyses during growth at 23°C and at the indicated times after the shift to 30°C. The fractions of cells with YFP-Cdc10p localization at the neck were determined by counting more than 200 cells for each sample. Cells with partially localized YFP-Cdc10p at the neck were scored as positive. (E) Synthetic lethality between bni5Δ and hof1Δ. Strains 1783 (wild type) (no. 1), KLY1940 (bni5Δ) (no. 2), KLY3022 (hof1Δ with a HOF1/CYK2 plasmid) (no. 3), and SKY2115 (bni5Δ hof1Δ with the HOF1/CYK2 plasmid) (no. 4) were streaked on YEPD plates with or without 5-FOA to select against the URA3-based plasmid. (F) Two examples of bni5Δ hof1Δ double-mutant segregants after tetrad dissection of a heterozygous diploid.

FIG. 3.

Synthetic effects in bni5Δ septin and bni5Δ hof1Δ double mutants. (A) Septin single-mutant and septin bni5Δ double-mutant strains were cultured overnight at 23°C and shifted to 30°C (or 37°C for sep7Δ and sep7Δ bni5Δ) for 3 h before fixation and examination by DIC microscopy. Strains: cdc3-6, KLY1415; cdc10-1, KLY1798; cdc11-6, KLY1419; cdc12-6, KLY1422; sep7Δ, KLY3937; cdc3-6 bni5Δ, KLY2177; cdc10-1 bni5Δ, KLY1803; cdc11-6 bni5Δ, KLY2174; cdc12-6 bni5Δ, KLY2010; sep7Δ bni5Δ, KLY3941. Bar, 5 μm. (B) Strains KLY2214 (cdc12-6 YFP-CDC10) and KLY2216 (cdc12-6 bni5Δ YFP-CDC10) were cultured overnight at 23°C and then shifted to 30°C. Samples were fixed for microscopic analyses during growth at 23°C and 3 h after the shift to 30°C. Bar, 5 μm. (C) Strains KLY1546 (wild type) (lane 1), KLY2214 (lane 2), and KLY2216 (lane 3) were cultured overnight. These cultures were serially diluted, spotted on YEP-glucose (YEPD), and incubated at the indicated temperature for 3 days. (D) Quantitation of cells with YFP-Cdc10p localization at the neck as a function of time. The strains used in panel B were cultured overnight at 23°C and then shifted to 30°C. Samples were fixed for microscopic analyses during growth at 23°C and at the indicated times after the shift to 30°C. The fractions of cells with YFP-Cdc10p localization at the neck were determined by counting more than 200 cells for each sample. Cells with partially localized YFP-Cdc10p at the neck were scored as positive. (E) Synthetic lethality between bni5Δ and hof1Δ. Strains 1783 (wild type) (no. 1), KLY1940 (bni5Δ) (no. 2), KLY3022 (hof1Δ with a HOF1/CYK2 plasmid) (no. 3), and SKY2115 (bni5Δ hof1Δ with the HOF1/CYK2 plasmid) (no. 4) were streaked on YEPD plates with or without 5-FOA to select against the URA3-based plasmid. (F) Two examples of bni5Δ hof1Δ double-mutant segregants after tetrad dissection of a heterozygous diploid.

FIG. 4.

Expression and localization of Bni5p through the cell cycle. (A) Localization of Bni5p to the neck of budded cells. Exponentially growing cells of strain KLY1737 (BNI5-GFP₂) were fixed and examined for Bni5p-GFP localization and cell morphology. Arrows, unbudded cells or cells with nascent buds without detectable Bni5p-GFP signal; barbed arrows, large-budded cells without detectable Bni5p-GFP signal. Bar, 5 μm. (B) Colocalization of Bni5p and Cdc11p. Bni5p was visualized by the GFP signal, whereas Cdc11p was detected by immunofluorescence. Barbed arrows, unbudded or large-budded cells with Cdc11p signal but no Bni5p signal; arrows, localization of Bni5p within the Cdc11p bands. Bar, 5 μm. (C and D) Levels of Bni5p and Bni5p localization in a synchronous culture. Strain KLY1737 was arrested in G₁ by α-factor treatment for 3 h and then released. At the indicated times, samples were taken to prepare total cellular proteins and to determine Bni5p localization and the presence or absence of buds after fixation. (C) Proteins were separated by 10% SDS-PAGE and analyzed by Western blotting with antibodies against GFP, Clb2p, or Cdc28p. Both short and long exposures of the Bni5p-GFP immunoblot are shown for clarity. The Cdc28p levels provide a loading control. No tag, wild-type strain KLY1546; G, asynchronously growing KLY1737 cells. (D) The percentages of budded cells and of cells showing detectable Bni5p-GFP localization were determined by counting more than 200 cells for each time point.

FIG. 5.

Dynamics of Bni5p neck localization in relation to spindle structure. Strain KLY1737 (BNI5-GFP₂) was transformed with TUB1-GFP (see Materials and Methods) and then examined by time-lapse video microscopy as described in Materials and Methods. In the two cells imaged here (outlined in the first panel), Bni5p-GFP is visible as a band crossing the elongated spindle. Spindle disassembly began at 10 min in the cell on the left or 12 min in the cell on the right.

FIG. 6.

Dependence of Bni5p localization and Cdc11p function on the N-terminal portion of Cdc11p. (A) Localization of YFP-Bni5p in wild-type and viable septin deletion mutants. Wild-type strain KLY1546 and the cdc10Δ mutant (KLY1715), both expressing YFP-Bni5p from plasmid pKL1900, were cultured in YEP-glucose (YEPD) medium. Strain KLY1718 (cdc11Δ + YCp111-GST/CDC11) expressing YFP-Bni5p from plasmid pKL1900 was cultured in YEP-galactose overnight and then shifted to YEPD to repress GAL1-GST-CDC11 expression for 1 h prior to fixation. Bar, 5 μm. (B) Structures of the Cdc11p truncations used in these analyses (see Materials and Methods). Solid box, the conserved P-loop motif; shaded box, the predicted coiled-coil domain. (C to E) Strain KLY1718 (cdc11Δ + YCp111-GST/CDC11) derivatives harboring various C-terminally HA-tagged Cdc11p constructs and a plasmid expressing either YFP-CDC10 or ADH1-YFP-BNI5 were cultured in YEP-galactose (YEPG) overnight and then shifted to YEPD medium to repress GAL1-GST-CDC11 expression for 1 h. For each culture, total cellular lysates were prepared for Western analysis (C), and an aliquot was fixed with formaldehyde to assess Cdc10p and Bni5p localization by YFP fluorescence (D) or Cdc11p localization by immunostaining with an anti-HA antibody (D and E). Cdc10p, YFP-Cdc10p; Bni5p, YFP-Bni5p; Cdc11p, wild-type and mutant Cdc11p-HA proteins; vector, strain KLY3404 or KLY3412; 1-415, strain KLY3405 or KLY3413; 1-385, strain KLY3406 or KLY3414; 31-385, strain KLY3410 or KLY3418; 1-200, strain KLY3411 or KLY3419. For panel D, more than 200 cells with clear buds were counted for each sample. (F) Wild-type strain KLY1546 and the KLY1718 derivatives KLY3404, KLY3405, KLY3406, KLY3410, and KLY3411 were cultured overnight, serially diluted, and spotted onto either YEPG or YEPD plates at 30°C. Strains expressing either Cdc11p^1-415 or Cdc11p^1-385 grow better than the others even on YEPG, because GAL-GST-CDC11 does not fully complement the cdc11Δ mutation. (G) Strains expressing myc-tagged Cdc3p together with no tagged Cdc11p (SKY1601), HA-tagged full-length Cdc11p (SKY1824), or HA-tagged Cdc11p^1-385 (SKY1825) were grown on YEPD at 30°C, and extracts were prepared as described in Materials and Methods. Proteins present in the total lysates or in the immunoprecipitates (IP) prepared with anti-HA antibody were evaluated by Western blotting with anti-myc (above) and anti-HA (below) antibodies.

FIG. 7.

Physical interactions between Bni5p and septins. (A) Two-hybrid assays were conducted as described in Materials and Methods with plasmids that expressed the full-length genes as AD or DBD fusions. Numbers indicate the Miller units of β-galactosidase activity averaged from two independent experiments. Immunoblotting (lower panels) indicated that the four septin fusions were expressed equally. (B) In vitro binding studies were carried out with ³⁵S-labeled, in vitro-translated T7-Bni5p-His₆ and bacterially expressed, bead-bound GST and GST-septin fusion proteins (see Materials and Methods). After pull-down, proteins were separated by SDS-PAGE. The gel was stained with Coomassie to detect GST and GST-septin fusion proteins (lower panel) and then subjected to autoradiography to detect bound Bni5p (upper panel). Note that although there was considerable degradation of the fusion proteins (particularly GST-Cdc3p), approximately equal amounts of the full-length fusion proteins were still present (arrows). Input, 2% of the ³⁵S-labeled T7-Bni5p-His₆ that was added to each binding reaction. (C) In vitro binding studies were carried out with bacterially expressed T7-Bni5p-His₆ and bead-bound GST-Cdc11p or GST (see Materials and Methods). After SDS-PAGE, the amounts of bound T7-Bni5p-His₆ were determined by immunoblotting with an anti-T7 antibody (upper panel), while the amounts of GST or GST-Cdc11p were determined by immunoblotting with an anti-GST antibody (lower panel). Input (upper panel), 5% of the T7-Bni5p-His₆ that was incubated with GST or GST-Cdc11p.