Tubulin isotypes optimize distinct spindle positioning mechanisms during yeast mitosis

Abstract

Microtubules are dynamic cytoskeleton filaments that are essential for a wide range of cellular processes. They are polymerized from tubulin, a heterodimer of α- and β-subunits. Most eukaryotic organisms express multiple isotypes of α- and β-tubulin, yet their functional relevance in any organism remains largely obscure. The two α-tubulin isotypes in budding yeast, Tub1 and Tub3, are proposed to be functionally interchangeable, yet their individual functions have not been rigorously interrogated. Here, we develop otherwise isogenic yeast strains expressing single tubulin isotypes at levels comparable to total tubulin in WT cells. Using genome-wide screening, we uncover unique interactions between the isotypes and the two major mitotic spindle positioning mechanisms. We further exploit these cells to demonstrate that Tub1 and Tub3 optimize spindle positioning by differentially recruiting key components of the Dyn1- and Kar9-dependent mechanisms, respectively. Our results provide novel mechanistic insights into how tubulin isotypes allow highly conserved microtubules to function in diverse cellular processes.

Figure 1.

Tubulin isotypes in budding yeast are not functionally equivalent. (a) Genetic organization and viability of yeast cells harboring deletions of endogenous α-tubulin isotypes. (b and c) Representative Western blot (b) and quantification from three independent experiments (c) of α-tubulin levels in WT and tub3Δ cells. Actin is loading control. Mean ± SEM; ***, P ≤ 0.001 by unpaired, two-tailed Student’s t test. Diamonds show individual values from each experiment. (d) Viability of cells with a single α-isotype gene under regulation of the other isotype locus. (e) Tetrad analyses from diploids with the indicated deletions and substitutions of TUB1 and TUB3 genes. ⁺, viable spores; ⁰, inviable spores. (f) Representative tetrad dissections of strains from panel e. Sister spores are arranged vertically. Green boxes mark representative spores of the indicated genotype to show growth is comparable to sister spores. No haploid spores of tub1Δ::TUB3 tub3Δ::URA3 genotype were viable (far right). Red boxes mark representative inviable spores that, based on genotype of their surviving sisters, harbor the tub1Δ::TUB3 tub3Δ::URA3 genotype. As reported, diploids heterozygous for tub1Δ or tub3Δ also display overall decreased spore viability relative to WT (Schatz et al., 1986b).

Figure 2.

Tubulin isotypes display functional differences. (a–c) Genetic organization of WT yeast (a) and those resulting from direct gene replacement to harbor Tub1 only (b) or Tub3 only (c) under endogenous TUB1 and TUB3 regulation. For each cell (left), the α/β-heterodimer combinations (center) and composition of cellular microtubules (right) are shown. (d and e) Representative Western blot (d) and quantification from three independent experiments (e) of α-tubulin levels in the indicated cell types. Actin is loading control. Mean ± SEM; ***, P ≤ 0.001 (n.s., not significant, WT vs. Tub1-only and Tub3-only, P = 0.13 and 0.16, respectively) by unpaired, two-tailed Student’s t test. Diamonds show individual values from each experiment. (f) Carbendazim (CBZ) sensitivity assay. Serially diluted cultures were spotted onto media with increasing concentrations of the microtubule-destabilizing drug and incubated at 24°C for 3 d.

Figure S1.

Growth rate of single isotype strains and TUB1 mRNA transcript levels measured by the predesigned TaqMan or custom assays. (a) Spotting growth assay. Log-phase cultures of the indicated cell type were serially diluted (fivefold), spotted onto rich media, and incubated at the indicated temperature for 3 d. (b–d) Growth rate of the indicated cell types in liquid media at 24°C (b), 30°C (c), and 37°C (d). Graphs represent mean ± SEM from two trials. (e) TUB1 mRNA levels in WT (black) and Tub1-only (gray) cells using either the predesigned TaqMan assay (Thermo Fisher Scientific) or a custom probe/primers assay. The experiment was repeated four times with the TaqMan assay and twice with the custom assay. Each repetition was performed with an independent culture and RNA/cDNA preparation. Results are shown as mean ± SEM; n.s., not significant by unpaired, two-tailed Student’s t test.

Figure 3.

TUB1 mRNA level is comparable to TUB3 in WT cells and unchanged in tub3Δ cells. (a) TUB1 (red) and TUB3 (blue) mRNA levels as a percentage of total expression in WT cells. (b) Transcript levels of TUB1, TUB2, and TUB3 in WT, tub3Δ, Tub1-only, and Tub3-only cells. (c) 3′ UTR–specific transcript levels originating from the TUB1 locus (red background) or TUB3 locus (blue background) as a percentage of total expression (cell type is indicated above the bars). For this assay, a forward primer and probe, both with 100% homology to TUB1 and TUB3, were coupled with reverse primers specific to the 3′ UTR of either locus. For panels a and b, n = 6 independent cultures and RNA/cDNA preps; for panel c, n = 4 independent cultures and RNA/cDNA preps. In panel b, TUB1 expression was repeated four times with the predesigned TaqMan assay and twice with the custom assay and the results combined, since expression was similar with either probe/assay. In all panels, graphs show mean ± SEM; n.s., not significant; **, P ≤ 0.01 by unpaired, two-tailed Student’s t test.

Figure 4.

Mapping genetic interactions of α-tubulin mutants. (a) Venn diagram of negative interactions identified in the SGA screens for loss of TUB1 (red; tub1Δ::TUB3) and loss of TUB3 (blue; tub3Δ::TUB1). (b) String network of negative synthetic interactions among the hits from panel a. Genes are clustered by GO terms as follows: 1–6 = depicted in panel c, 7 = protein degradation/proteosome, 8 = cell cycle progression/meiosis, 9 = cell polarity/morphogenesis, 10 = Golgi/endosome/vacuole/sorting, 11 = lipid/sterol/fatty acid biosynthesis, 12 = RNA processing, 13 = ribosome/translation, and 14 = miscellaneous/unknown function. (c) Hits in representative GO-term categories that show negative synthetic interaction with loss of TUB1 (red hashed; tub1Δ::TUB3) or TUB3 (blue hashed; tub3Δ::TUB1) specifically or both (gray). Significant enrichment marked by ***, P ≤ 0.001 with Fisher’s exact test. (d) Schematics of Kar9- and Dynein-dependent spindle positioning pathways. In the Kar9 pathway, in early mitosis, Bim1 links microtubule plus ends via Kar9 to the myosin Myo2, which transports the complex toward the bud along polarized actin cables at the cell cortex (McNally, 2013). This “sweeping” of microtubule plus ends moves the associated spindle pole close to the bud neck. In the Dynein pathway, the kinesin Kip2 in complex with Bik1 localizes Dyn1 to microtubule plus ends. Dyn1 together with dynactin (not shown) is activated by contact with Num1 on the cell cortex. Active Dyn1/dynactin “slides” microtubules along the cortex, which pulls the associated spindle pole into the bud (McNally, 2013). (e) Model of negative synthetic genetic relationship between hits from the loss of TUB1 (tub1Δ::TUB3) or loss of TUB3 (tub3Δ::TUB1) SGA screens and components of the Kar9 and Dynein pathways. If the SGA hits resulting from the loss of an isotype also display negative synthetic interactions with Dynein pathway components, it indicates that isotype may share functions with the Dynein pathway (A). Negative synthetic interactions between Kar9 pathway components and isotype SGA hits suggest that isotype and the Kar9 pathway may share functions (B). (f) Heatmaps of negative synthetic interactions between the microtubule-related genes identified in the loss of TUB3 (left; tub3Δ::TUB1) and loss of TUB1 (right; tub1Δ::TUB3) SGA screens (top row) and components of the Kar9 and Dynein pathways (left column). Each position reports the pairwise synthetic genetic interaction score between the indicated isotype screen hits and Kar9/Dynein pathway components, as previously reported (Usaj et al., 2017). Pairwise score < −0.12 (red) represents strong negative interaction (scale below heatmaps). MT, microtubule. (g) Strong negative synthetic interactions between the microtubule-related hits in the loss of TUB1 (red hashed; tub1Δ::TUB3) and loss of TUB3 (blue hashed; tub3Δ::TUB1) and Kar9 or Dynein pathway components as a percentage of total interactions possible from heatmaps in panel f.

Figure S2.

Genetic interactions for TUB1 and TUB3. (a) Venn diagram of negative synthetic genetic interactions identified in our loss of TUB3 (blue; tub3Δ::TUB1) SGA screen and nonessential genes reported for simple TUB3 deletion (green; tub3Δ) SGA (Costanzo et al. 2016). Both screens identified 59 stringent interactions, with 8 hits common between both screens. (b) CellMap diagram showing negative synthetic genetic interaction networks with associated GO-term categories for the loss of TUB1 (yellow; tub1Δ::TUB3) or TUB3 (blue; tub3Δ::TUB1; Usaj et al., 2017). MVB, multivesicular body pathway (involved in ubiquitin-dependent sorting into endosomes and lysosome); RIM, regulator of IME2 pathway (involved in cell wall synthesis, meiosis, spore formation).

Figure 5.

Tub1 optimizes Dyn1 pathway function. (a) Schematics of GFP- and mRUBY3-tubulin fusion constructs used to visualize microtubules in the indicated single isotype and control cells. (b) Percentage of kar9Δ cells with the indicated tubulin isotypes that display properly aligned anaphase spindles. (c) Percentage of cells displaying Dynein-mediated spindle movements during 10-min observation. (d) Frequency of Dynein-mediated spindle movements in individual cells. (e) Frequency of astral microtubule contact with the bud cell cortex in individual cells. (f) Dynein-mediated spindle movements normalized by microtubule–cortex interactions in individual cells. (g) Spindle translocation per Dynein-mediated event. (h) Duration of Dynein-mediated spindle movements. (i) Astral microtubule length at initiation of Dynein-mediated spindle movements. (j) Spindle translocation per Dynein-mediated event binned by astral microtubule length. (k) Dynein-mediated spindle movements normalized by microtubule–cortex interactions in epothilone B (EpoB)–treated cells. (l) Spindle translocation per Dynein-mediated event binned by astral microtubule length in EpoB-treated cells. For panels b and c, graphs show mean ± SEM from three trials. In panel b, for Tub1 only, n = 86, 92, and 150; for WT-Tub1, n = 83, 107, and 117; for WT-Tub3, n = 98, 76, and 163; for Tub3 only, n = 66, 113, and 158. In panel c, n = 40, 30, and 30 cells for each genotype. For panels d–i, graphs show mean ± SEM from 90–100 cells observed over three trials, where n = 40, 20–30, and 30 for each genotype. For panels g–i, the total number of events recorded is 128, 92, 65, and 36 for Tub1 only, WT-Tub1, WT-Tub3, and Tub3 only, respectively. In panels j–l, graphs show mean ± SEM; panel j includes a total of 33 and 10 Dynein events from three trials for Tub1-only and Tub3-only cells, and panel l includes a total of 56 and 22 Dynein events from two trials for Tub1-only and Tub3-only cells, respectively; for panel k, n = 40, 30, 30, and 40 cells analyzed from two trials for Tub1 only, WT-Tub1, WT-Tub3, and Tub3 only, respectively. *, P ≤ 0.05; **, P ≤ 0.01; and ***, P ≤ 0.001 by unpaired, two-tailed Student’s t test. MT, microtubule.

Figure S3.

Anaphase spindle orientation, Dynein localization at the bud cortex and SPB, and the proportion of cells with plus end foci of Bik1. (a) Orientation of anaphase spindle in mother cell of kar9Δ cells with the indicated tubulin genotype. (b) Length of astral microtubules during anaphase. (c and d) Dyn1-3GFP fluorescence intensity at the bud cell cortex (c) and daughter-bound SPB (d) in the indicated cell types. (e) Percentage of cells with visible Bik1-3GFP foci at astral microtubule plus ends. Each panel shows mean ± SEM from three trials; in panel a, for Tub1 only, n = 20, 20, and 21; for WT-Tub1, n = 20, 23, and 23; for WT-Tub3, n = 20, 20, and 19; for Tub3 only, n = 20, 20, and 21. In panel b, plot shows microtubules observed from three trials with a total of 128, 124, 126, and 127 for Tub1 only, WT-Tub1, WT-Tub3, and Tub3 only, respectively. In panel c, for Tub1 only, n = 14, 17, and 16; for WT-Tub1, n = 12, 12, and 9; for WT-Tub3, n = 9, 11, and 11; for Tub3 only, n = 4, 15, and 17. In panel d, for Tub1 only, n = 34, 33, and 28; for WT-Tub1, n = 29, 30, and 33; for WT-Tub3, n = 42, 34, and 31; for Tub3 only, n = 30, 30, and 35. In panel e, for Tub1 only, n = 65, 62, and 63; for WT-Tub1, n = 73, 61, and 76; for WT-Tub3, n = 67, 63, and 70; for Tub3 only, n = 68, 63, and 81. *, P ≤ 0.05; **, P ≤ 0.01; n.s., not significant for Tub1 only versus Tub3 only (P = 0.94 in panel b, P = 0.479 in panel c, P = 0.721 in panel d, and P = 0.075 in panel e) using the unpaired, two-tailed Student’s t test. MT, microtubule.

Figure S4.

Dynein pathway function is enhanced in Tub1-only versus Tub3-only cells with epothilone-stabilized microtubules. Drug-sensitized cells (pdr1Δ pdr3Δ) in mid-log phase were arrested in HU for 90 min, followed by addition of 10 µM epothilone B (EpoB) for another 60 min (a–e). Parameters of Dynein-mediated sliding activity in epothilone B– or mock (DMSO)-treated cells of the indicated genotypes. Note: Dynein deletion cells (far right; dyn1Δ) are not drug sensitized and were not treated with epothilone B or DMSO. Absence of a bar indicates no events were observed in dyn1Δ cells. (a) Maximum length attained by astral microtubules during Dynein-mediated spindle sliding. Maximum length during sliding is generally comparable in mock and epothilone B–treated cells. (b) Percentage of cells that display Dynein-mediated spindle sliding movements. (c) Frequency of Dynein-mediated spindle sliding movements in individual cells. (d) Frequency of astral microtubule contact with the bud cell cortex in individual cells. (e) Distance of spindle translocation by Dynein-mediated sliding events. For panels a–e, graphs show mean ± SEM from two trials; 30–40 cells were analyzed per genotype for each treatment condition; for Dynein-mediated sliding events, in Tub1-only epothilone B, n = 12 and 43; WT-Tub1 epothilone B, n = 13 and 12; WT-Tub3 epothilone B, n = 5 and 14; Tub3-only epothilone B, n = 6 and 16; Tub1-only DMSO, n = 27 and 19; WT-Tub1 DMSO, n = 8 and 13; WT-Tub3 DMSO, n = 12 and 10; and Tub3-only DMSO, n = 6 and 12. For dyn1Δ cells, n = 0 and 0 for observed sliding events. n.s., not significant; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001 by unpaired, two-tailed Student’s t test. MT, microtubule.

Figure S5.

Representative lifetime history plots of astral microtubules in HU-arrested, mock-treated, and epothilone-treated cells. Epothilone treatment significantly stabilizes microtubules in WT, Tub1-only, and Tub3-only cells. Two microtubules are shown for each of the indicated cell types in untreated (left), mock (DMSO; middle), and 10 µM epothilone B treatment (EpoB; right). Tub1-only cells contain the TUB1 ORF in both the TUB1 and TUB3 loci, while Tub3-only cells harbor the TUB3 ORF in both. Microtubules in Tub1-only and Tub3-only cells were visualized with an exogenous copy of GFP-Tub1 and GFP-Tub3, respectively. WT-Tub1 and WT-Tub3 are WT cells harboring an exogenous copy GFP-Tub1 and GFP-Tub3, respectively. MT, microtubule.

Figure 6.

Tub1 preferentially directs Dynein pathway proteins to astral microtubule plus ends. (a) Schematic of Dyn1-3GFP localization (green) in WT cells. (b) Dyn1-3GFP localization to microtubule plus ends. Left: Representative wide-field and DIC images showing Dyn1-3GFP localization to SPB and astral microtubule plus end. Right: Dyn1-3GFP signal at microtubule plus ends. (c) Percentage of cells with visible Dyn1-3GFP foci at astral microtubule plus ends. (d) Dyn1-3GFP localization to mother cell cortex. Left: Cell images showing Dyn1-3GFP localization at mother cortex. Right: Dyn1-3GFP signal at mother cortex foci. (e) Bik1 localization to microtubule plus ends. Left: Cell images showing Bik1-3GFP localization to the spindle, SPBs and astral microtubule plus ends. Right: Bik1-3GFP signal at microtubule plus ends. (f) Bik1 localization along astral microtubules. Left: Cell images showing Bik1-3GFP localization along astral microtubules. Right: Number of Bik1-3GFP foci along bud-directed astral microtubules. For Tub1 only versus Tub3 only, P = 0.002, 0.04, and 0.004 for one, two, and three or more foci; for Tub3 only versus WT-Tub3, P = 0.012 and 0.024 for one and two foci, respectively. (g) Kip2 localization to microtubule plus ends. Left: Cell images showing Kip2-3YFP localization along and at plus ends of astral microtubules. Right: Kip2-3YFP signal at microtubule plus ends. For panels b–g, microtubules are visualized by mRUBY3-Tub1 or mRUBY3-Tub3; merge shows microtubules in red and GFP-tagged proteins in green. Graphs show mean ± SEM from three trials. In panel b, for Tub1 only, n = 47, 58, and 60; for WT-Tub1, n = 54, 57, and 70; for WT-Tub3, n = 54, 68, and 52; for Tub3 only, n = 61, 60, and 51. In panel c, for Tub1 only, n = 69, 59, and 58; for WT-Tub1, n = 82, 55, and 61; for WT-Tub3, n = 63, 60, and 87; for Tub3 only, n = 53, 63, and 85. In panel d, for Tub1 only, n = 26, 48, and 42; for WT-Tub1, n = 38, 25, and 21; for WT-Tub3, n = 40, 54, and 54; for Tub3 only, n = 55, 42, and 25. In panel e, for Tub1 only, n = 49, 58, and 57; for WT-Tub1, n = 56, 50, and 60; for WT-Tub3, n = 56, 62, and 60; for Tub3 only, n = 50, 61, and 55. In panel f, for Tub1 only, n = 65, 62, and 63; for WT-Tub1, n = 76, 73, and 61; for WT-Tub3, n = 67, 63, and 70; for Tub3 only, n = 68, 63, and 81. In panel g, for Tub1 only, n = 59, 60, and 60; for WT-Tub1, n = 60, 60, and 67; for WT-Tub3, n = 60, 60, and 60; for Tub3 only, n = 63, 60, and 60. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001 by unpaired, two-tailed Student’s t test. Scale bars = 2 µm. MT, microtubule.

Figure 7.

Tub3 optimizes spindle positioning by the Kar9 pathway. (a) Small-budded cells of the indicated genotype with properly positioned early mitotic spindles (category 1 on the left). (b) Preanaphase spindle positioning relative to the bud neck. Micrograph shows representative GFP-tubulin image with cell outline in white. (c) Small-budded cells with early mitotic spindle that display an astral microtubule oriented toward, or away from, the bud. Micrographs show cell (left) and GFP-tubulin image with cell outline in white (right) that are representative of categories shown in cartoons. (d) Astral microtubule length in small-budded cells. In panels a and c, bars show mean ± SEM from three trials; in panel a, for kar9Δ, n = 67, 64, and 83; for Tub1 only, n = 84, 82, and 94; for WT-Tub1, n = 57, 66, and 134; for WT-Tub3, n = 52, 79, and 100; for Tub3 only, n = 102, 116, and 116. In panel c, for Tub1 only, n = 65, 65, and 70; for WT-Tub1, n = 60, 60, and 58; for WT-Tub3, n = 62, 55, and 59; for Tub3 only, n = 66, 64, and 55. For panels b and d, lines and error bars show mean ± SD from three trials. In panel b, scatterplot shows individual cells observed in three trials with a total of 169, 205, 187, 180, and 165 for kar9Δ, Tub1 only, WT-Tub1, WT-Tub3, and Tub3 only, respectively. In panel d, scatterplot shows microtubules observed in three trials with a total of 193, 186, 193, and 207 for Tub1 only, WT-Tub1, WT-Tub3, and Tub3 only, respectively. **, P ≤ 0.01; and ***, P ≤ 0.001 by unpaired, two-tailed Student’s t test. Scale bars = 2 µm. aMT, astral microtubule.

Figure S6.

Tub3 optimizes Kar9-dependent spindle positioning as observed in cells expressing nontagged tubulins solely from the two endogenous loci. Spindle positioning was monitored by visualizing SPBs (yeast spindle poles) with fluorescently tagged Cnm67-mRUBY2. (a) Number of small-budded cells of the indicated genotype with properly positioned early mitotic spindles (category 1 on left). Micrographs show cell (far left) and Cnm67-mRUBY2 image with cell outline in white (second from left) that are representative of categories shown in cartoons. Early mitotic spindles were classified as small-budded cells with duplicated but unseparated SPBs. (b) Preanaphase spindle position relative to the bud neck. Micrographs under cartoon show representative cell (lower) and Cnm67-mRUBY2 image with cell outline in white (upper). For panel a, bars show mean ± SEM from two trials; for kar9Δ, n = 67 and 92; Tub1 only, n = 86 and 40; WT, n = 118 and 132; and Tub3 only, n = 120 and 90. For panel b, line and error bars represent mean ± SEM, and scatterplot shows individual cells observed in two trials with a total of 101, 111, 175, and 167 for kar9Δ, Tub1 only, WT, and Tub3 only, respectively. *, P ≤ 0.05; ***, P ≤ 0.001 by unpaired, two-tailed Student’s t test. Scale bars = 2 µm.

Figure 8.

Tub3 preferentially directs Kar9 pathway proteins to astral microtubule plus ends. (a) Bim1 localization to microtubule plus ends. Left: Representative wide-field and DIC images of a preanaphase cell showing Bim1-GFP localization to the spindle and astral microtubule plus end. Right: Percentage of cells with visible Bim1-GFP foci at astral microtubule plus ends. (b) Bim1-GFP signal at microtubule plus ends. (c) Kar9 localization to microtubule plus ends. Left: Images of a preanaphase cell showing Kar9-GFP localization to astral microtubule plus end. Right: Percentage of cells with visible Kar9-GFP foci at microtubule plus ends. (d) Kar9-GFP signal at microtubule plus ends. (e) Exogenous BIM1-GFP on a centromeric plasmid rescues Bim1 localization to astral microtubules in Tub1-only cells. (f) Exogenous BIM1 restores Kar9-dependent astral microtubule transport to the bud in Tub1-only cells. Number of small-budded cells with early mitotic spindle that display an astral microtubule oriented toward the bud. Micrographs show representative GFP-Tub1 images from each category with cell outline in white. (g) Exogenous BIM1 rescues deficient Kar9-dependent spindle positioning in Tub1-only cells. Micrographs show representative cell (left) and GFP-Tub1 images with cell outline in white (right) from categories shown in cartoons. Bars represent mean ± SEM from two trials (a) or three trials (b–g); in panel a, for Tub1 only, n = 73 and 61; for WT-Tub1, n = 82 and 56; for WT-Tub3, n = 88 and 60; for Tub3 only, n = 82 and 69. In panel b, for Tub1 only, n = 50, 50, and 65; for WT-Tub1, n = 50, 65, and 57; for WT-Tub3, n = 50, 56, and 50; for Tub3 only, n = 50, 59, and 50. In panel c, for Tub1 only, n = 61, 51, and 103; for WT-Tub1, n = 101, 57, and 79; for WT-Tub3, n = 75, 61, and 79; for Tub3 only, n = 65, 62, and 95. In panel d, for Tub1 only, n = 50, 52, and 59; for WT-Tub1, n = 61, 60, and 60; for WT-Tub3, n = 55, 57, and 60; for Tub3 only, n = 59, 59, and 59. In panel e, without Bim1 rescue plasmid (left), for Tub1 only, n = 48, 44, and 51; for WT-Tub1, n = 39, 39, and 48; with Bim1 rescue plasmid (right), for Tub1 only, n = 40, 40, and 32; for WT-Tub1, n = 38, 40, and 34. In panel f, without Bim1 rescue plasmid (left), for bim1Δ, n = 49, 70, and 59; for Tub1 only, n = 46, 55, and 47; for WT-Tub1, n = 37, 48, and 49; with Bim1 rescue plasmid (right), for bim1Δ, n = 49, 46, and 49; for Tub1 only, n = 38, 43, and 46; for WT-Tub1, n = 38, 50, and 48. In panel g, without Bim1 rescue plasmid (left), for bim1Δ, n = 54, 60, and 65; for Tub1 only, n = 61, 47, and 81; for WT-Tub1, n = 53, 48, and 57; with Bim1 rescue plasmid (right), for bim1Δ, n = 63, 69, and 65; for Tub1 only, n = 61, 48, and 54; for WT-Tub1, n = 47, 62, and 61. For panels a and c, merge shows microtubules in red and GFP-tagged proteins in green. n.s., not significant; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001 by unpaired, two-tailed Student’s t test. Scale bars = 2 µm; aMT, astral microtubule.

Figure 9.

Model for the role of tubulin isotypes in optimizing diverse microtubule functions. (a) Yeast microtubules polymerized from only the α-isotype Tub1 (red) optimize the Dynein-mediated spindle positioning mechanism at the expense of the Kar9 mechanism. (b) Those made from only Tub3 (blue) optimize the Kar9-mediated mechanism at the expense of the Dynein mechanism. (c) In WT cells, microtubules containing both Tub1 and Tub3 support both mechanisms sufficiently. Tub1 and Tub3 are likely to differentially support other microtubule-dependent functions as well. (d) Utilization of multiple α- and β-isotypes (e.g., in higher eukaryotes) would allow microtubules to be optimized for an increased range of molecular functions.