1. Cell Biology
  2. Microbiology and Infectious Disease
Download icon

Polo-like kinase Cdc5 regulates Spc72 recruitment to spindle pole body in the methylotrophic yeast Ogataea polymorpha

  1. Hiromi Maekawa Is a corresponding author
  2. Annett Neuner
  3. Diana Rüthnick
  4. Elmar Schiebel
  5. Gislene Pereira
  6. Yoshinobu Kaneko
  1. Osaka University, Japan
  2. Kyushu University, Japan
  3. Zentrum für Molekulare Biologie der Universität Heidelberg, DKFZ-ZMBH Alliance, Germany
  4. University of Heidelberg, Germany
  5. German Cancer Research Centre (DKFZ), DKFZ-ZMBH Alliance, Germany
Research Article
Cited
0
Views
344
Comments
0
Cite as: eLife 2017;6:e24340 doi: 10.7554/eLife.24340

Abstract

Cytoplasmic microtubules (cMT) control mitotic spindle positioning in many organisms, and are therefore pivotal for successful cell division. Despite its importance, the temporal control of cMT formation remains poorly understood. Here we show that unlike the best-studied yeast Saccharomyces cerevisiae, position of pre-anaphase nucleus is not strongly biased toward bud neck in Ogataea polymorpha and the regulation of spindle positioning becomes active only shortly before anaphase. This is likely due to the unstable property of cMTs compared to those in S. cerevisiae. Furthermore, we show that cMT nucleation/anchoring is restricted at the level of recruitment of the γ-tubulin complex receptor, Spc72, to spindle pole body (SPB), which is regulated by the polo-like kinase Cdc5. Additionally, electron microscopy revealed that the cytoplasmic side of SPB is structurally different between G1 and anaphase. Thus, polo-like kinase dependent recruitment of γ-tubulin receptor to SPBs determines the timing of spindle orientation in O. polymorpha.

https://doi.org/10.7554/eLife.24340.001

eLife digest

Before a cell divides, it needs to duplicate its genetic material to provide the new daughter cell with a full set of genetic information. To do so, the cell forms a complex of proteins called the spindle apparatus, which is made up of string-like microtubules that divide the chromosomes evenly. In many organisms, the position of the spindle determines where in the cell this separation happens.

However, in baker’s yeast, the location where the cell will divide is determined well before the spindle is formed. Unlike many other eukaryotic cells, these yeast cells divide asymmetrically and create buds that will form the new daughter cells. The position of this bud determines where the spindle should be located and where the chromosomes separate.

The spindle itself is then organised by a structure called the spindle pole body, which connects to microtubules inside the cell nucleus and microtubules in the cell plasma. Several proteins control where and how the spindle forms, including a protein called the spindle pole component 72, or Spc72 for short, and an enzyme called Cdc5. However, until now it was unclear how spindle formation is timed and controlled in other yeast species.

Now, Maekawa et al. have used fluorescent markers and time lapse microscopy to examine how the spindle forms in the yeast species Ogataea polymorpha, an important industrial yeast used to produce medicines and alcohol. The results show that in O. polymorpha, the positioning and orientation of the spindle only occurred very late in the cell cycle and the microtubules in the cell plasma remained unstable until the chromosomes were about to separate. This was linked to changes in the level of Spc72, which increased at the spindle pole body before the chromosomes separated and then dropped again. This was controlled by Cdc5.

Understanding when and where microtubules are formed is an important step in understanding how cells divide. This is the first example of a budding yeast that creates new microtubules in the cell plasma every time the cell divides. Unravelling the molecular differences between yeast species could lead to new ways to optimise the use of industrial yeasts like O. polymorpha, or to combat disease-causing ones.

https://doi.org/10.7554/eLife.24340.002

Introduction

Segregation of sister chromatids into two daughter cells is pivotal to the proliferation of eukaryotic cells. Chromosome segregation is followed by cytokinesis, which results in physical separation of two daughter cells. In many organisms, the position of the mitotic spindle dictates the site of cytokinesis, which ensures the inheritance and maintenance of genomic information in the daughter cells. Astral microtubules or cytoplasmic microtubules (cMTs), which emanate from the spindle poles and extend to the cell cortex, have a principle role in positioning and orienting the spindle with respect to the polarity cues of the cell type. Mechanisms governing the spindle positioning/orientation have been studied in a number of systems. However, regulations that determine the timing of establishing the spindle orientation, or the position of the centrosome, the primary MT organizing centre (MTOC), in interphase, are not well understood (Kiyomitsu, 2015; Woyke et al., 2002).

Spindle positioning is of particular importance in the budding yeast Saccharomyces cerevisiae, where the cleavage site is determined at the start of the cell cycle independently of the position of the mitotic spindle. Therefore, cells position the pre-anaphase spindle close to the bud neck and orient it along the mother-bud axis. As the spindle elongates in anaphase, one spindle pole translocates into the bud to accomplish segregation of one set of chromosomes into the daughter cell (Pereira and Yamashita, 2011; Markus et al., 2012; Winey and Bloom, 2012).

In S. cerevisiae, the nuclear positioning and spindle orientation are regulated by two redundant pathways acting on cMTs, the Kar9 and dynein pathways (Li et al., 1993; Miller and Rose, 1998; Winey and Bloom, 2012). Concomitant deletions in components of both pathways result in lethality, whereas loss of one pathway can be compensated by the function of the other with moderate spindle orientation defects (Miller and Rose, 1998). Survival of single deletion mutants largely relies on the function of the spindle orientation checkpoint (SPOC) that retains cells in anaphase until the spindle orientation is corrected (Bardin et al., 2000; Pereira et al., 2000; Caydasi and Pereira, 2012).

Furthermore, MTs in S. cerevisiae are organized exclusively from the spindle pole body (SPB), which is the functional equivalent of animal centrosome. The SPB is a multilayered cylindrical organelle that is embedded in the nuclear envelope (NE) throughout the cell cycle (Byers and Goetsch, 1974Byers and Goetsch, 1975 )The outer plaque faces the cytoplasm and nucleates cMTs, whereas the inner plaque is inside the nucleus and organizes the nuclear MTs. The central plaque anchors and interconnects the outer and inner plaques (O'Toole et al., 1999Jaspersen and Winey, 2004). In G1 phase, some fractions of the cMTs are organized from a modified region of the NE associated with one side of the SPB known as the half-bridge (Byers and Goetsch, 1974; Byers and Goetsch, 1975). Spc72, a γ-tubulin complex (γ-TuSC) receptor, is required for nucleating MTs at both the outer plaque and the half-bridge (Chen et al., 1998; Knop and Schiebel, 1998; Wigge et al., 1998; Souès and Adams, 1998). Localisation of Spc72 at the outer plaque is mediated by binding to Nud1, whereas Kar1 serves as a G1 specific binding site of Spc72 at the half-bridge (Pereira et al., 1999; Gruneberg et al., 2000). Spc72 also has a structural role as an integral part of the outer layer and as such localisation of Spc72 to the SPB and the ability to nucleate cMTs persist through the entire cell cycle (Shaw et al., 1997; Pereira et al., 1999; Kosco et al., 2001). Importantly, Spc72, and hence cMTs, is not recruited for the formation of the SPB. New SPB acquires Spc72 and cMTs after the formation of a 1 µm long spindle (Shaw et al., 1997Segal et al., 2000; Juanes et al., 2013).

In addition to the γ-tubulin complexes, Spc72 exerts a role in recruiting several other proteins to SPBs including Stu2, a microtubule-associated protein (MAP) of the XMAP215/Dis1 family, the SPOC kinase Kin4, as well as polo-like kinase Cdc5 (Chen et al., 1998; Usui et al., 2003; Maekawa et al., 2007; Snead et al., 2007). Cdc5 regulates multiple cellular functions including SPB duplication, progression through G2/M phase, promoting mitotic exit, and cytokinesis (Shirayama et al., 1998; Hu et al., 2001; Song and Lee, 2001; Archambault and Glover, 2009; Elserafy et al., 2014). Cdc5 is also involved in the regulation of spindle orientation in pre-anaphase and migration of the anaphase spindle (Snead et al., 2007; Park et al., 2008). Although Spc72 becomes highly phosphorylated during mitosis in a Cdc5-dependent manner, it is unclear whether this phosphorylation has a regulatory effect on Spc72 and/or cMTs (Maekawa et al., 2007; Snead et al., 2007).

The molecular mechanisms that control spindle orientation in S. cerevisiae have been well established. However, other species that employ the budding mode of cell division may have adopted different strategies. In the pathogenic yeast Candida albicans, the nucleus is located away from the bud neck in pre-anaphase cells (Martin et al., 2004; Finley et al., 2008). C. albicans and probably some of other species in Saccharomycotena may therefore have different mechanisms and regulations in this fundamental biological process.

Ogataea polymorpha (previously Hansenula polymorpha) is extensively used in industrial biotechnology, for the production of various pharmaceuticals in particular for its advantageous characteristics including methylotrophy, nitrate assimilation, availability of strong promoters, and low amount of secreted proteins (Gellissen et al., 2005; Stöckmann et al., 2009 ). Another attractive property of O. polymorpha is its thermotolerant nature (up to approximately 50°C), which may reduce the cost of cooling in, for instance, bioethanol production that requires the treatment of raw materials at high temperature prior to fermentation. However, despite its importance, cell biology research on this organism remains limited. A better understanding of the molecular physiology of O. polymorpha is beneficial towards improving the abilities and characteristics of this yeast for a wide variety of applications.

Here, we describe cMT organization and its regulation during the cell cycle of the methylotrophic yeast O. polymorpha. Unlike S. cerevisiae, the pre-anaphase spindle is not readily positioned and oriented in O. polymorpha owing to the poorly organized cMTs at early cell cycle stages. The bottleneck of cMT nucleation/anchoring at SPBs occurs at the level of Spc72 recruitment to the SPBs, for which the polo-like kinase Cdc5 plays a crucial role. Consistent with the cell cycle dependent activity of cMTs, SPB structure also undergoes cell cycle dependent modification. Thus, our study shed light on the divergent nature of the temporal control of the cMT formation in yeast species.

Results

Nuclear positioning in O. polymorpha differs from that in S. cerevisiae and other budding yeast species

The nucleus is positioned close to the bud neck in large budded pre-anaphase cells of S. cerevisiae (Figure 1A). Similar organization was observed in other budding yeast species including Candida glabrata, Kluyveromyces lactis, Pichia pastoris, and Yarrowia lipolytica (Figure 1—figure supplement 1). Notably in O. polymorpha, however, nuclear position was not biased to the bud neck, although it remained in the mother cell body (Figure 1A, Figure 1—figure supplement 1). The phenotype resembled, but was more exaggerated than, that in C. albicans where the nucleus is located with a distance from the bud neck in pre-anaphase cells (Martin et al., 2004; Finley et al., 2008). A similar phenotype was observed in species closely related to O. polymorpha or C. albicans (Figure 1—figure supplement 1). Close examination of O. polymorpha revealed that the nucleus was located in the cell centre in 76.7% of G2/M cells whereas in the remainder of the cells it was off-centred with no bias towards the bud neck (Figure 1A, Figure 1—figure supplement 2). These results suggest that the nuclear position is not determined before anaphase onset in O. polymorpha. As a consequence, the early stages of anaphase occurred in the mother cell body (Figure 1B). This was further confirmed in cells expressing the α-tubulin gene (Tub1)-GFP and histone H3 gene (Hht1)-mRFP (monomeric red fluorescent protein) to visualize MTs and chromosomal DNA, respectively. Majority of the early anaphase spindles as judged according to their length (<5 µm) as well as a stretched DNA mass were located entirely in the mother cell body (97.4%, Figure 1C, Figure 1—figure supplement 3A). However, 94.7% of the early anaphase spindles were aligned along the bud-mother axis and almost all of the late anaphase spindles with two segregated DNAs were inserted into the bud (Figure 1—figure supplement 3), suggesting immediate and efficient orientation of the spindle during anaphase.

Figure 1 with 6 supplements see all
Nuclear positioning in O. polymorpha.

(A) Nucleus is positioned in the cell centre in pre-anaphase cells of O. polymorpha. S. cerevisiae strain YPH499 and O. polymorpha type strain CBS4732 were grown in YPDS at 30°C. DNA was stained with DAPI. The positions of nuclei were as outlined in the cartoon shown on the left side of the subfigure. Scale bar, 2 µm. N = 60 (YPH499), 55 (CBS4732). Result of a similar experiment using HHT1-GFP cells is shown in Figure 1—figure supplement 2. (B) Time-lapse microscopy of histone H3 gene (HHT1)-GFP cells (HPH31). Anaphase onset judged by Hht1-GFP was observed at the 3 min timepoint. Shown are a merged figure of bright field images and deconvolved and projected GFP images. Scale bar, 2 µm. (C) Early anaphase cells (HPH164) with a single DNA mass along the spindle grown in YPDS at 30°C. Microtubules and DNA are visualized by GFP-Tub1 and Hht1-mCherry fluorescence, respectively. Scale bar, 2 µm. (D) Early anaphase cells (HPH1678) grown in YPDS at 30°C. SPB and DNA are visualized by Mps3-GFP and Hht1-mCherry fluorescence, respectively. Scale bar, 2 µm. (E) Time lapse microscopy of cells expressing SPB-GFP maker. MPS3-GFP cells (HPH1681) were grown in SD complete medium at 30°C. Consecutive sections were taken every 60 s. Shown are representative images of cells with an inappropriate angled spindle against the polarity axis. Yellow arrows and orange arrowheads point SPB. Shown are deconvolved and projected GFP images merged with bright field image. Scale bar, 2 µm. Another example is shown in Figure 1—figure supplement 6.

https://doi.org/10.7554/eLife.24340.003

SPB position during the cell cycle was examined to clarify the spindle cycle relative to bud size in cells expressing the SPB marker (Mps3-GFP) (Figure 1D). Two SPB signals appeared in some of small/medium budded cells (Figure 1D, panel c, lower cell), suggesting that SPBs were duplicated at the timing of bud emergence or later, which is similar to that in S. cerevisiae although the precise cell cycle stage should be carefully determined. In the rest of small budded cells, one SPB signal was evident until large budded cells (Figure 1D, panel d). This may be because duplicated SPBs remained in a close proximity and could not be resolved by our fluorescence microscopy. Consistent with this, intensity of Mps3-GFP in medium/large-budded cells with a single SPB was much higher than that in unbudded G1 cells or cells with separated SPBs (Figure 1—figure supplement 4). Moreover, SPB in G1 cells as well as small budded cells was not in the defined position within the mother cell body (Figure 1D). Subsequent time lapse analysis revealed that after spindle assembly, ~1 µm long spindles remained at their central positions and loosely oriented toward the bud neck until shortly before anaphase onset (defined by the rapidly increase of pole-to-pole distance) (Figure 1—figure supplement 5). Anaphase initiated in the mother cell body (Figure 1E, 11 min, Figure 1—figure supplement 6, 3–4 min). These observation defined cells with 2SPBs in a < 2 µm distance as pre-anaphase cells. Spindle alignment was corrected around the time of (or shortly after) spindle elongation, followed by SPB insertion into the bud. After spindle breakdown, the SPB moved vigorously with no relationship to the polarity axis (Figure 1E, Figure 1—figure supplement 6).

O. polymorpha cells contain only fewer cMTs

Lack of nuclear positioning and spindle orientation in pre-anaphase cells may indicate a very low number of cMTs at SPBs during this cell cycle window. To test this notion, we first investigated the MT organization during the cell cycle in GFP-TUB1 HHT1-mCherry cells (Figure 2A and B). Cell cycle stages were judged by the bud size and the number of DNA masses. cMTs were observed in all anaphase cells, whereas less than 50% of G1 and pre-anaphase cells carried cMTs. Furthermore, cMTs that apparently did not associate with the SPB were observed in 13.8% of cells prior to anaphase (Figure 2C). Time lapse analysis revealed that detached cMTs remained in the cytoplasm for only a short period of time before depolymerized (Figure 2—figure supplement 1). This situation is in stark contrast to that in S. cerevisiae where almost all cells exhibit cMTs that are stably associated with the SPB during the cell cycle (Shaw et al., 1997; Kosco et al., 2001).

Figure 2 with 3 supplements see all
O. polymorpha cells contain fewer cytoplasmic microtubules (cMTs).

(A) Wild-type cells (HPH164) were grown in YPDS medium at 30°C. Microtubules and DNA are visualized by GFP-Tub1 and Hht1-mCherry fluorescence, respectively. cMTs are marked by yellow arrows. a, b: unbudded G1 cell; c, d: preanaphase cell with monopolar nuclear MTs; e, f: preanaphase cell with bipolar spindle; g: anaphase cell. Scale bar, 2 µm. (B) Quantification of (A). G1, S/G2, G2/M, and anaphase represent unbudded cells, small budded cells with single unduplicated SPB, medium budded cells with 2 SPB, and large budded cells with an SPB in both the mother and the bud, and large budded cells with elongated spindle, respectively. n > 50 cells for each category. Ana, anaphase. (C) Cells containing the cMT detached from the SPB in (A). Scale bar, 2 µm. (D) Analysis of the duration of cMT persistence and cMT re-establishment time at SPBs in preanaphase cells with a bipolar spindle by time-lapse microscopy. GFP-TUB1 cells (HPH194) were grown in SD complete medium at 30°C. Consecutive sections were taken every 30 s. Duration of continuous cMT presence was scored as cMT persistance, while the time between the loss of cMT and the acquisition of a new cMT was scored as cMT re-establishment. Total recording time was 38460 s.

https://doi.org/10.7554/eLife.24340.010

The small number of observed cMTs might have arisen because of reduced cMT nucleation. Another and not mutually exclusive possibility considers that cMTs might not be stably anchored to the SPB and thus might not persist over long periods. To test these possibilities, we performed time lapse experiments with cells expressing GFP-TUB1, in which cMTs were observed during 20.6% of the recorded time points although the majority (>80%) did not persist longer than 30 s (Figure 2D, cMT persistence, Figure 2—figure supplement 2). These results suggest that cMTs are short-lived during early stages of the cell cycle. Once cMTs were lost, a relatively long time was required until new MTs appeared at the SPB (Figure 2D, cMT re-establishment; median value 90.0 s, average 173.8 ± 192.4 s). Thus, cMTs are less frequently nucleated and unstable at early stages of the cell cycle. Acquired cMTs efficiently corrected the spindle orientation in pre-anaphase cells, suggesting that the spindle orientation is regulated largely at the level of cMT acquisition (Figure 2—figure supplement 3).

Organization of the SPB structure on the cytoplasmic side is cell cycle dependent

Next, we examined the SPB structure in G1 and anaphase by electron microscopy. An electron dense SPB-like structure was evident in all cells examined, while a half-bridge-like structure—which plays an important role in cMT organization in G1 of S. cerevisiae—was not clearly observed (Figure 3). Anaphase SPBs had an additional thin layer in the cytoplasm that resembled the outer plaque of S. cerevisiae SPB (Figure 3A and B) (Byers and Goetsch, 1974; Byers and Goetsch, 1975). In contrast, there was no detectable outer plaque in G1 SPBs (Figure 3C and D). These results suggested that O. polymorpha SPBs undergo structural cycling in every cell cycle.

Figure 3 with 2 supplements see all
Cytoplasmic structure of SPB is regulated during the cell cycle.

Electron microscopy (EM) of thin serial sections of cells in G1 and anaphase. Wild-type cells (BY4329) were grown to log phase at 30°C in YPDS and then prepared for EM. Indicated are the cytoplasm (C), nucleus (N), nuclear envelope (NE), and nuclear microtubules (nMT). (A, B) Representative SPBs in anaphase cells (n = 10). The SPB in the mother does not appear in the section shown as the main image. The pale orange rectangle in the mother merely indicates the position of the lower inset which is the image of the section containing the SPB in the mother. (CD) representative SPBs in unbudded G1 cells (n = 10). Consecutive three sections are shown in (D). Scale bars of the main images in (A) and (C) represent 1 µm. Scale bars of the insets in (A) and (C), i.e., (B) and (D), represent 100 nm.

https://doi.org/10.7554/eLife.24340.014

Our attempt to arrest cells in late G1 by introducing cdc28-as allele, which arrest S. cerevisiae cells in late G1 with a single SPB and satellite, was failed probably because of insufficient inhibition of the kinase (Figure 3—figure supplement 1). However, inhibitor addition delayed cell cycle progression leading to the accumulation of cells with unseparated SPBs. This allowed us to examine the structure of side-by-side SPBs by EM (Figure 3—figure supplement 2). All six side-by-side SPBs had outer plaques which were similar to that in nocodazole arrested cells, albeit some of those were somewhat fuzzy. The result suggested that the SPB structure on cytoplasmic side is reconstructed before spindle formation. An additional electron dense cloud was observed on the cytoplasmic side of nuclear envelope between two SPB bodies, which resembled the half-bridge/bridge structure of S. cerevisiae SPB (Figure 3—figure supplement 2A, B and E; orange arrowheads). It was not clear whether this structure was present at other cell cycle stages. A better synchronization method is required to determine the fine structure and the precise timing of emergence/disappearance of the outer plaque during the cell cycle.

Spc72 associates with SPB in a cell cycle-dependent manner

Lack of the outer plaque in G1 prompted us to search for SPB components whose association with SPBs was cell cycle dependent. In S. cerevisiae, core SPB components are found at SPBs throughout the mitotic cell cycle including central plaque components (Spc42, Spc29), outer plaque components (Cnm67, Nud1), half-bridge components (Sfi1, Cdc31, Mps3, and Kar1), and membrane anchors (Ndc1, Nbp1, Mps2, and Bbp1) (Winey et al., 1991; Spang et al., 1995; Bullitt et al., 1997; Brachat et al., 1998; Wigge et al., 1998; Chial et al., 1998; Adams and Kilmartin, 1999; Elliott et al., 1999; Schramm et al., 2000; Kilmartin, 2003; Jaspersen et al., 2002; Araki et al., 2006). The γ-TuSC recruiting factors Spc110 and Spc72 also represent core components of SPB in the inner and outer plaques, respectively (Knop and Schiebel, 1997Knop and Schiebel, 1998). BLAST and HMMER searches have identified putative orthologues of genes for Mps3, Sfi1, Spc72, Spc110, and Nud1 as well as γ-TuSC components, Tub4, Spc97, and Spc98 in the O. polymorpha genome (Altschul et al., 1990; Sobel and Snyder, 1995; Geissler et al., 1996; Knop and Schiebel, 1997; Eddy, 1998; Maekawa and Kaneko, 2014; Riley et al., 2016). SPB-like localisation was verified by expressing GFP or mRFP-fused version of these proteins. GFP or mRFP dot-like signals of Tub4, Spc98, Mps3, Sfi1, and Nud1 were observed in most of the cells, suggesting that they represent constitutive components of the SPB throughout the cell cycle (Figure 4A, Figure 4—figure supplement 1). In contrast, the Spc72 signal was either weak or absent in cells at early cell cycle stages, whereas all anaphase cells carried two strong SPB signals (Figure 4A, Figure 4—figure supplement 1). Deletion of SPC72 in S. cerevisiae results in severe growth defects or lethality depending on the strain background. To evaluate the effect of SPC72 deletion in O. polymorpha, SPC72/spc72Δ::natNT2 heterozygous diploid cells were subjected to tetrad dissection analysis. Notably, 21 out of 29 tetrads yielded one or two viable colonies, all of which were sensitive to nourseothricin (Figure 4—figure supplement 2). Microscopic inspection revealed that 91.7% of spores with spc72Δ::natNT2 genotype derived from tetrads that gave two viable nourseothricin-sensitive colonies were germinated. These results suggested that SPC72 is essential for growth in O. polymorpha.

Figure 4 with 5 supplements see all
Accumulation of Spc72 at SPBs is cell cycle dependent.

(A) Cell cycle dependent localization of Spc72-GFP. SPC72-GFP MPS3-mRFP cells (HPH1394) were grown in SD complete medium at 30°C. Cell cycle stages are as shown in Figure 2B. Mps3-mRFP is a marker for SPB. Scale bar, 2 µm. (B) Quantification of the Spc72/Nud1/Sfi1 SPB signal of cells at different cell cycle stages. Yeast strains HPH972, HPH1396, and HPH1400 were used. Signal intensities were background-subtracted. Statistical significance of the difference between 2 SPBs and anaphase was determined by the student t-test and is indicated by an asterisk. Error bars indicate SD. n = 95, 127, and 124 for Spc72, Nud1, and Sfi1, respectively. Note that intensity of some of Sc72-GFP signals in G1 was high, because it decreased only gradually at SPB during the end of mitosis and the following G1 as shown in (C). (C) Time-lapse microscopy of SPC72-GFP MPS3-mRFP cells (HPH1394). Images were taken every 30 s. RFP signal was captured only before staring the time-lapse series. Anaphase onset judged by sudden spindle elongation was observed at the 5 min timepoint. Spindle orientation was corrected between 4.5 min and 5 min timepoints. Yellow arrowheads indicate the position of Mps3-mRFP before the image capture. Orange arrowheads indicate Spc72-GFP signals at SPBs. Shown are deconvolved and projected images. Scale bar, 2 µm. Another example is presented in Figure 4—figure supplement 1.

https://doi.org/10.7554/eLife.24340.017

To more precisely evaluate the amount of Spc72 at the SPB, images were obtained in logarithmically growing wild-type cells carrying SPC72-GFP MPS3-mRFP and the GFP intensity at the SPB was quantified (Figure 4B). The GFP signal was 2.5 times weaker in cells with a short spindle than in anaphase cells (p<0.001). In contrast, Nud1, which comprises the putative binding site of Spc72 on the outer plaque as suggested by the direct interaction between orthologues of these proteins in S. cerevisiae, did not show this trend, nor did Sfi1, a half-bridge component (Figure 4B) (Gruneberg et al., 2000; Kilmartin, 2003). High intensity of Sfi1-GFP signal in S/G2 cells most likely arose from SPBs that were duplicated but not yet separated. These results suggest that Spc72 is cell cycle regulated and the incorporation of Spc72 into SPBs may be the key step to stabilize cMTs. To further confirm this notion, time-lapse microscopy was carried out to determine the timing of Spc72 association with SPBs (Figure 4C, Figure 4—figure supplement 3). In all cells that progressed into anaphase, an Spc72-GFP signal became detectable <4 min prior to the initiation of anaphase (average 3.68 ± 1.74 min, n = 14) (Figure 4C, orange arrowheads). Within 5 min after appearance of the Spc72-GFP signal, spindle orientation was corrected when it had not done already (Figure 4—figure supplement 3, average 3.50 ± 1.61 min, n = 12); therefore, one half part of an anaphase nucleus was successfully inserted into the bud. Thus, Spc72 accumulates at SPB in early mitosis, most likely in metaphase, and remains high during anaphase. As cells exit from mitosis and entre the next cell cycle, Spc72-GFP signal was gradually decreased at SPBs with the timing that varied from cell to cell. This difference of timing may explain the relatively high and variable intensity of Spc72-GFP at SPB in G1 cells (Figure 4B). However, in all cases, Spc72-GFP levels reached a minimum well before short spindle was formed (Figure 4, Figure 4—figure supplement 4).

If low abundance of Spc72 at the SPB is the reason underlying cMT instability, higher expression of Spc72 might increase the level of Spc72 at the SPB and consequently raise the number of cMTs, and thereby promote positioning the nucleus close to the bud neck at early stages of the cell cycle. We expressed the SPC72-GFP gene from a strong constitutive TEF1 promoter in cells whose endogenous SPC72 was also fused to GFP and examined the position of the SPB relative to the bud neck in pre-anaphase cells as a readout of cMT function (Kiel et al., 2007). Overexpressed Spc72-GFP was efficiently targeted to SPB because a strong Spc72-GFP signal was observed in cells carrying the PTEF1-SPC72-GFP gene but not in wild-type cells during G1 and G2/M phases (Figure 5A, Figure 5—figure supplement 1). In cells overexpressing SPC72-GFP, SPB was positioned close to the bud neck, which is reminiscent of the SPB position in S. cerevisiae (Figures 1A and 5B), and cMTs were more often observed (Figure 5C and D). Together, these results strongly supported our hypothesis that cMT organization is regulated at the level of repeated Spc72 recruitment to the SPB in every cell cycle.

Figure 5 with 2 supplements see all
Overexpression of Spc72 converts the O. polymorpha type of nuclear position to the S. cerevisiae type.

(A) Overexpressed Spc72-GFP were recruited to SPB at all stages of the cell cycle. SPC72-GFP was expressed from a strong promoter of the TEF1 gene (HPH1393). Enrichment of Spc72 to SPB was evident in G1 cells and cells with short spindles (SPB distance <1 µm, yellow arrows) compared with images of wild-type cells (HPH1394). Mps3-mRFP marks SPB. Scale bar, 5 µm. (B) SPB is positioned close to the bud neck in G2/M cells carrying PTEF1-SPC72-GFP. The position of the SPB closer to the bud was as outlined in the cartoon shown on the left side of the subfigure. Strains used were HPH1393 (n = 110) and HPH1394 (n = 117). (C) Overexpression of Spc72 stimulated cMT acquisition. MTs were visualized with CFP-TUB1 in wild type and cells overexpressing SPC72-GFP (HPH1653 and HPH1652, respectively) were grown in YPDS medium at 30°C. Images were captured only for CFP and brightfield. Scale bar, 5 µm. (D) Quantification of (C). Presence/absence of cMTs was scored in cells with short spindle. Shown is the average of three independent experiments. Error bars indicate SD. n > 100. Average of three independent experiments.

https://doi.org/10.7554/eLife.24340.023

cMT play important roles in yeast mating and karyogamy, which are initiated in G1. Because mating is triggered by nutrient starvation in O. polymorpha, we examined cMTs and Spc72 in nutrient starved cells. Interestingly, while Spc72 was accumulated at SPBs, cMTs were not observed (Figure 5—figure supplement 2). Thus, specific mechanism may regulate Spc72 and cMT organization under such conditions.

SPB association of Spc72 is dependent on the polo-like kinase Cdc5

Spc72 might be regulated at the level of protein expression. To synchronize cells for monitoring changes in protein levels during the cell cycle, we transferred the recently developed auxin-inducible degradation (AID)-degron system to O. polymorpha (Nishimura and Kanemaki, 2014). CDC5 encodes the only polo-like kinase in yeast. It is thus essential for growth in S. cerevisiae, and its inactivation causes cell cycle arrest in late anaphase (Kitada et al., 1993). Similarly, a single CDC5 orthologue was identified in the O. polymorpha genome (OpCDC5). Logarithmically growing cells carrying a 3mAID-tagged version of CDC5 were arrested as large budded cells by incubation in the presence of auxin and then released into fresh medium without auxin to resume the cell cycle (Figure 6A). Spc72 protein abundance did not fluctuate as cells entered into anaphase and proceeded into the following cell cycle (Figure 6A, Figure 6—figure supplement 1A). Furthermore, the Spc72 band migrated slower in nocodazole-arrested cells than that in asynchronous cells (Figure 6—figure supplement 1B). These results suggested that either post-translational modification of Spc72 or regulation of Spc72 binding proteins might be utilized to achieve cell cycle dependency of SPB localisation.

Figure 6 with 3 supplements see all
The recruitment of Spc72 to SPBs is dependent on the polo-like Cdc5 kinase.

(A) Abundance of Spc72 does not fluctuate during the cell cycle. Logarithmically growing SPC72-GFP CDC5-3mAID PCDC28-OsTIR cells (HPH1380) were synchronized with IAA followed by release. Samples were analysed by immunoblotting for Spc72-GFP. Comparable intensities of unspecific signal in immunoblotting (asterisk) indicate equal loading of samples. Budding index and mitotic index (DAPI) were determined over time. n > 100 cells per time point. (B) SPC72-GFP CDC5-3mAID PCDC28-OsTIR cells (HPH1380) were synchronized and released as in (A). Images were captured without fixation. Shown are deconvolved and projected images. Time after release is indicated on the left. Scale bar, 5 µm. (C) Quantification of Spc72-GFP intensity at SPBs in (B). Signal intensities were background-subtracted. Statistical significance of the difference between 2 SPBs and anaphase was determined by the student t-test. – Cdc5: before the release;+Cdc5: 20 min after the release. Error bars indicate SD. n > 50 cells per time point. (D) Accumulation of Spc72 at SPBs in metaphase depends on Cdc5 function. SPC72-GFP CDC5-3mAID PCDC28-OsTIR cells (HPH1380) were arrested with nocodazole in the presence (− Cdc5) or absence (+Cdc5) of IAA and Spc72-GFP signal at SPBs was quantified. Signal intensities were background-subtracted. Error bars indicate SD. n > 50 cells.

https://doi.org/10.7554/eLife.24340.026

Furthermore, we noticed that the GFP intensity of Spc72-GFP at the SPB was significantly lower in cells arrested by Cdc5-depletion than that in cells after re-accumulation of Cdc5 in both metaphase (SPB distance <2 µm, p<0.0001) and anaphase (SPB distance >4 µm, p<0.0001) (Figure 6B and C). Strong dependency of SPB binding of Spc72 on Cdc5 was further confirmed in metaphase-arrested cells by nocodazole (Figure 6D, Figure 6—figure supplement 2). These results suggested that the stable association of Spc72 requires Cdc5 kinase, the activity of which is likely cell cycle-regulated.

Spc72 is phosphorylated by Cdc5 kinase in S. cerevisiae (Maekawa et al., 2007; Snead et al., 2007), which prompted us to investigate whether Spc72 is subjected to a Cdc5-dependent phosphorylation in O. polymorpha. Our gel electrophoresis analyses of nocodazole-arrested cells suggested that Spc72 of O. polymorpha is subjected to post-translational modifications in a Cdc5-dependent manner because the Spc72 band was more smeared and migrated slightly slower in wild type and CDC5-overexpressing cells as compared to that in Cdc5-depleted cells (Figure 6—figure supplement 3A). Although this difference was largely lost during preparation of cell extract, the λ-phosphatase treatment revealed in vivo phosphorylation of Spc72 that was independent of Cdc5 (Figure 6—figure supplement 3B), suggesting that Cdc5 contributes to a subset of phosphorylations of Spc72. Thus, Spc72 is phosphorylated at multiple sites, only some of which depend on Cdc5.

Cdc5 localises to NE and SPB during mitosis

We next examined the localisation of Cdc5 during the cell cycle. Distinctive localization of Cdc5-GFP became apparent after S phase and was lost prior to or during the following G1 phase (Figure 7A and B, Figure 7—figure supplement 1). Nuclear and NE localisation appeared at early stages of the cell cycle and persisted until the end of mitosis. In addition, a fraction of the GFP signals appeared to overlap with those of SPBs during mitosis. Neither NE nor SPB-associated GFP signals were detected in unbudded or small budded cells. Notably, SPB signal may arise from the SPB itself, NE surrounding the SPB, or kinetochores that cluster close to the SPB during interphase and anaphase in yeast (Jin et al., 1998). Therefore, in order to clarify on which side of SPB the Cdc5-GFP signal resided, we employed structured illumination microscopy (SIM). Localisation was investigated in metaphase-arrested cells with nocodazole where Cdc5-GFP signal was observed at SPB as well as in nucleus (Figure 7—figure supplements 1B and 2), and Spc72, and Spc110 were used as references for the cytoplasmic and nuclear side of SPB, respectively. Spc72-GFP and Spc110-tdTomato signals were clearly distinguished in 58% of the cells, which verified that our method could discriminate signals in the cytoplasmic and the nuclear side of SPB in >50% of cells (Figure 7C). The resolution of both signals probably depends on the orientation of the SPB (top versus side view). Only the SPB side view will resolve Spc72-GFP and Spc110-tdTomato signals at SPBs. Cdc5-GFP overlapped with Spc110-tdTomato in 53% of cells, which is similar to the degree of co-localisation observed between Spc72 and Spc110. In contrast, the Cdc5-GFP signal overlapped with Spc72-tdTomato in 87% of cells (Figure 7D). These results suggest that Cdc5-GFP locates at the position on SPBs closer to Spc72 than to Spc110, indicating that Cdc5-GFP signal arises from the cytoplasmic side of SPB. Thus, Cdc5 likely becomes first localised to the nucleus and the NE in G2, and then in mitosis to the cytoplasmic side of SPBs. The timing of Cdc5 binding to SPBs coincides well with the recruitment of Spc72 to SPBs.

Figure 7 with 2 supplements see all
Localization of polo-like Cdc5 kinase in the nucleus, the nuclear envelope, and at the SPB is cell cycle dependent.

(A) CDC5-GFP MPS3-mRFP cells (HPH1562) were grown in YPDS at 30°C. Blue, red, green, and purple cell contours mark G1, S/G2, G2/M, and anaphase cells, respectively. Shown is a projected image after deconvolution. Scale bar, 5 µm. (B) Quantification of Cdc5-GFP localization in (A). The position of SPB (magenta) was as outlined in the cartoon shown at the bottom of the subfigure. n > 30 cells for each cell cycle stage. Ana, anaphase. (C) SIM images of Spc72-GFP and Spc110-tdTomato in nocodazole-arrested cells. SPC72-GFP SPC110-tdTomato cells (HPH1581) grown in SC medium at 30°C were arrested in metaphase with nocodazole. Scale bars, 1 µm and 0.2 µm in the large and small images, respectively. n = 78. (D) SIM images of Cdc5-GFP together with either Spc110-tdTomato or Spc72-tdTomato in nocodazole-arrested cells. CDC5-GFP SPC110-tdTomato and CDC5-GFP SPC72-tdTomato cells grown in SC medium at 30°C were arrested in metaphase with nocodazole. Scale bars, 1 µm and 0.2 µm in the large and small images, respectively. Strains used were HPH1583 (n = 75) and HPH1575 (n = 55). Diffused nuclear signal of Cdc5-GFP were also observed in all cells (Figure 7—figure supplement 2).

https://doi.org/10.7554/eLife.24340.030

CDC5 overexpression accelerates the Spc72 recruitment to SPB

To further confirm the significance of Cdc5 kinase in the recruitment of Spc72 to SPBs, we constitutively expressed the CDC5 gene at high level. While Cdc5 expression showed no effect on the protein level of Spc72-GFP (Figure 8—figure supplement 1), Spc72-GFP intensity at the SPB was higher in metaphase-arrested cells following ectopic expression of CDC5 from the TEF1 promoter than in wild-type cells (p<0.0001) (Figure 8A and B). In the similar analysis performed in asynchronously growing cells, accumulation of Spc72-GFP at SPB was significantly higher at all stages of the cell cycle in cells overexpressing CDC5 than in wild type cells (p<0.01 for G1 cells; p<0.0001 for S/G2, G2/M, and anaphase cells), with the strongest effect observed in G2/M phase (Figure 8C and D), and cMTs were more prevalent (Figure 8E and F). As a consequence, the SPB positioned closer to the bud neck (Figure 8G) and the spindle was at an angle within 30° with respect to the mother-bud axis in 83% of pre-anaphase cells overexpressing CDC5 compared with 48% in wild type cells (Figure 8—figure supplement 2). These results suggest that the stable SPB association of Spc72 is restricted to the time period where Cdc5 kinase activity is sufficiently high.

Figure 8 with 2 supplements see all
Overexpression of Cdc5 kinase promotes early association of Spc72 to SPBs.

(A) Overexpression of CDC5 enhances SPB binding of Spc72. SPC72-GFP MPS3-mRFP (HPH1394) and SPC72-GFP MPS3-mRFP PTEF1-CDC5∆53 cells (HPH1542) were arrested in metaphase with nocodazole. Deconvolved and projected images are shown. Scale bar, 5 µm. (B) Spc72-GFP signal at SPBs in (A) was quantified. Signal intensities were background-subtracted. Error bars indicate SD. n > 50 cells. (C) Spc72-GFP was recruited to SPB at earlier stages of the cell cycle when CDC5 gene is overexpressed. SPC72-GFP MPS3-mRFP (HPH1394, -) and SPC72-GFP MPS3-mRFP PTEF1-CDC5∆53 cells (HPH1542, CDC5 OP) were grown in YPDS medium. Enrichment of Spc72 to SPB and alignment of the spindle along the mother-bud axis were evident in preanaphase cells (yellow arrows) overexpressing CDC5, compared with wild-type cells (white arrow). Shown are deconvolved and projected images. Mps3-mRFP marks SPB. Scale bar, 2 µm. (D) Quantification of Spc72-GFP intensity at SPBs in (C). –: wild-type; OP: CDC5 overexpression. Signal intensities were background-subtracted. Statistical significances of the difference between wild-type (-) and CDC5 overexpressing cells (CDC5 OP) were determined by the student t-test. Error bars indicate SD. (E) Overexpression of Spc72 stimulated cMT acquisition. MTs were visualized with CFP-TUB1 in wild type and CDC5 overexpressing cells (HPH1680 and HPH1673, respectively) were grown in YPDS medium at 30°C. Images were captured only for CFP and brightfield. Scale bar, 5 µm. (F) Quantification of (E). Presence/absence of cMTs was scored in cells with short spindle. Shown is the average of three independent experiments. Error bars indicate SD. n > 100. (G) SPB is positioned close to the bud neck in G2/M cells overexpressing CDC5. The position of the SPB closer to the bud was as outlined in the cartoon shown in Figure 5B. Strains used were HPH1394 (n = 125) and HPH1542 (n = 101).

https://doi.org/10.7554/eLife.24340.033

Discussion

The mode of cell division by budding represents a type of asymmetric cell division. The mechanism to achieve a high-fidelity of chromosome segregation in such a situation has been a focus of interest and has been investigated intensively in S. cerevisiae. These process was recently studied in the other ascomycetous yeast C. albicans as well as in the yeasts Cryptococcus neoformans and Ustilago maydis in another phylum of fungi, Basidiomycota (McCully and Robinow, 1972aMcCully and Robinow, 1972b; Kopecká et al., 2001; Steinberg et al., 2001; Woyke et al., 2002; Straube et al., 2003; Martin et al., 2004; Finley et al., 2008). Although the movement of the nucleus during the cell cycle differs between ascomycetous and basidiomycetous yeasts, it is commonly positioned close to the bud neck in both phyla prior to chromosome segregation. We report here that the ascomycetous yeast O. polymorpha does not follow the same strategy. In O. polymorpha, the nucleus generally locates centrally within the mother cell body and the spindle is not aligned properly along mother-bud axis until anaphase onset. Consequently, spindle elongation in early anaphase occurs entirely in the mother often with an inappropriate angle against the polarity axis. Despite this potential complication, one nucleus penetrates successfully into the bud during anaphase, which may largely rely on an immediate correction of the orientation of the spindle and on SPOC activity. Those SPB movements are in contrast to S. cerevisiae in which spindle is aligned during metaphase and therefore SPB translocation into the bud coincides with spindle elongation. Currently molecular mechanism(s) that regulate spindle orientation is unknown. However, although the timing of spindle orientation relative to cell cycle progression appears to be different from that of other yeasts, two redundant molecular mechanisms of spindle orientation, one requiring dynein and the other Kar9, may be conserved in O. polymorpha, because putative orthologs of KAR9 and dynein were identified in O. polymorpha genome sequences (Li et al., 1993; Miller and Rose, 1998; Maekawa and Kaneko, 2014; Nordberg et al., 2014).

In S. cerevisiae, Spc72 is stably incorporated into SPBs once it is recruited and organise cMTs throughout the cell cycle. In O. polymorpha, the strong SPB association of OpSpc72 in anaphase becomes weakened as cells enter into the following G1 phase, whereas they re-accumulate later in the cell cycle. The timing of OpSpc72 recruitment to SPBs during early mitosis appears to primarily dictate the organization of cMTs and hence nuclear position (Figure 9). As the polo-like kinase Cdc5 protein plays an important role in this regulation, a question arises regarding the substrates of Cdc5 kinase in this process (Archambault and Glover, 2009). Because ScSpc72 binds to and is phosphorylated by Cdc5 in S. cerevisiae, OpSpc72 represents an obvious candidate (Maekawa et al., 2007; Snead et al., 2007). Our electrophoresis analyses indeed detected Cdc5-dependent phosphorylation of Spc72. However, OpCdc5 failed to phosphorylate recombinant Spc72 in vitro. Further analyses are required to verify whether Cdc5 directly phosphorylates Spc72 and the potential effects of such modifications on the regulation of cMT. Additionally, it may also be important to identify other kinase(s) that are responsible for the Cdc5-independent phosphorylation of Spc72 observed in our analyses. Cdc5 may have another important substrates other than Spc72. Notably, a polo box binding site present in ScSpc72 is missing in OpSpc72. Cdc5 might therefore phosphorylate other SPB proteins such as Nud1, which has one site matching the consensus sequences of polo box binding site (S-Sp/Tp-P), and thereby indirectly influence the affinity of Spc72 towards the SPB.

Model of the SPB cycle in O. polymorpha.

See Discussion for details.

https://doi.org/10.7554/eLife.24340.036

It is unclear what brought highly expressed Spc72-GFP to SPBs at early cell cycle stages when Cdc5 activity was low. Weak Spc72-GFP signal in the absence of Cdc5 may be because of low affinity or unstable association to SPBs. Increased Spc72 protein levels may simply result in the increased number of Spc72 protein at SPBs at any given time. Alternatively, overexpression may overcome a negative regulation that normally maintains the low level of Spc72 at SPBs during early stages of the cell cycle (Figure 4). Such Cdc5-independent regulation is consistent with the observation that Spc72 was gradually lost from SPBs at early stages of the cell cycle. Examining properties of Spc72 protein when overexpressed such as post-translational modifications, protein stability, and molecular dynamics at SPBs would clarify this point.

In S. cerevisiae, SPB duplication initiates in late G1 phase by forming a satellite at the distal end of an extended half-bridge of the pre-existing ‘old’ SPB, which is then inserted into the nuclear envelope. In contrast to the old SPB which maintains cMTs from the previous mitosis, this ‘new’ SPB acquires the MT nucleation activity in the inner plaque prior to spindle assembly while it is still connected to the old SPB (side-by-side SPBs). On the other hand, a recent report suggested that the acquisition of ScSpc72 to the outer plaque, and hence of cMTs, occurs only after SPB separation and spindle assembly (Juanes et al., 2013). In O. polymorpha, the cMTs acquisition is also regulated at the level of Spc72 recruitment to SPBs. OpSpc72 dissociate from SPBs at the end of mitosis and recruited to both old and new SPBs shortly before anaphase of the following cell cycle. This suggests that the cell cycle regulation of Spc72 recruitment may be applied to both old and new SPBs in O. polymorpha. Even though the Spc72 recruitment is cell cycle regulated in both species, its timing seems to be different: while it occurs in early G2 phase in S. cerevisiae, it does in metaphase in O. polymorpha. Moreover, their regulatory mechanisms are likely different because their acquisition is Cdc5-dependent in O. polymorpha, but not in S. cerevisiae (Juanes et al., 2013). Inhibition of Cdc5 kinase in S. cerevisiae causes the misaligned spindle phenotype, which indicates a role of Cdc5 in cMT functions (Snead et al., 2007). However, it is unlikely that Cdc5 acts at the level of Spc72 recruitment since both SPBs of the misaligned spindle caused by Cdc5 inhibition carried cMTs (Snead et al., 2007). Alternatively, it is possible that the Cdc5-dependent regulation of ScSpc72 function on cMT nucleation/anchoring has been overlooked. Growth suppression of cnm67∆ cells by CDC5 overexpression might point towards this possibility (Park et al., 2004). It might also be because of the redundancy with the SPB-anchored Stu2 function (Usui et al., 2003). The Stu2 binding domain of Spc72 (aa176–230 in ScSpc72, Figure 4—figure supplement 5) is conserved in ascomycetous yeasts including O. polymorpha but it differs somewhat in OpSpc72. In particular, the detached cMTs observed in approximately 14% of O. polymorpha pre-anaphase cells are reminiscent of the spc72∆Stu2 phenotype in S. cerevisiae (Usui et al., 2003). Thus, the potential binding of Stu2 to Spc72 in O. polymorpha should be investigated.

In O. polymorpha, structure of the cytoplasmic side of SPB, along with cMT nucleation competence, is modified as the cell cycle progresses (Figure 9). Notably, our BLAST search failed to identify orthologs of several SPB core components identified in S. cerevisiae in genomes of outside of Saccharomycetaceae: the central plaque components Spc42 and Spc29, the membrane anchors Nbp1 and Bbp1, and the half-bridge component Kar1. The failure of identification may be due to the highly divergent nature of amino acid sequences of these coiled-coil proteins, or alternatively the function of these SPB components is not required in O. polymorpha. The absence of these proteins could be one of the reasons underlying the poor appearance of the central plaque and the half-bridge in our electron microscopy anlayses. Our observation that the outer plaque in the cytoplasm was evident in anaphase SPBs but not in G1 SPBs may suggest the outer plaque from the previous mitosis is removed in the following G1 phase. However, given that Nud1, a putative outer plaque component, is present at the SPB in G1, it is more likely that the outer plaque is only partially disassembled. Furthermore, the appearance of the outer plaque in EM analysis proceeded that of Spc72 in fluorescent microscopy (Figure 9). Electron microscopy analysis using another fixation method that better preserves the SPB structure might clarify this point. Equally important is the analysis of the half-bridge which organizes cMTs in G1 in S. cerevisiae. If the half-bridge plays the same role in O. polymorpha SPB as that in S. cerevisiae, the loss of Spc72 from outer plaque alone should not cause the loss of cMTs. In addition, in S. cerevisiae, SPBs are segregated in a defined mode where the old SPB normally migrates into the daughter cell while the new SPB remains in the mother cell (Pereira et al., 2001). The SPB history in the outer plaque was proposed to primarily determine the destination of SPBs and to bias spindle asymmetry via Nud1 (Hotz et al., 2012; Juanes et al., 2013). Whether the mode of SPB inheritance is conserved in O. polymorpha and other yeast species is also worthy of further study.

It is unlikely that the mode of nuclear positioning in O. polymorpha is ancestral, given that Y. lipolytica and P. pastoris, who are diverted from the common ancestor earlier than O. polymorph, share the same nuclear organization with S. cerevisiae. However, several yeast species relatively close to O. polymorpha or C. albicans exhibited a similar nuclear position in pre-anaphase cells (Figure 1—figure supplement 1). Thus, the Spc72 recruitment mechanism described in O. polymorpha may be widely utilized in several Clades in Saccharomycotina, including Ogataea, Ambrosiozyma, and Nakazawaea.

Materials and methods

Yeast strains and plasmids

Yeast strains and plasmids used in this study are listed in Table 1. Unless otherwise indicated, all O. polymorpha strains were derived from NCYC 495 and were generated by PCR-based methods (Lu et al., 2000; Janke et al., 2004; Saraya et al., 2012). GFP, mRFP, and 5flag tagged alleles were generated in ku80Δ or ku70Δ cells and then crossed with auxotrophic wild-type strains to obtain KU80+ or KU70+ cells carrying the tagged allele (Maekawa and Kaneko, 2014). O. polymorpha cells were transformed by electroporation (Faber et al., 1994). The 500 bp sequences up- and downstream of the OpTEF1 open reading frame (ORF) were used as the OpTEF1 promoter and terminator, respectively and those of the OpCDC28 ORF were used as the OpCDC28 promoter and terminator (Kiel et al., 2007). For overexpression of CDC5, we expressed an N-terminal-truncated version of CDC5 (CDC5∆53) that is equivalent to the ScCDC5∆N70 allele which is resistant to APC-dependent ubiquitination (Shirayama et al., 1998). The CDC5∆53 ORF was amplified by PCR and inserted into pHM949. The resulting plasmid pHM950 was digested and a zeocin resistance marker was inserted (pHM956). To obtain the dyn1∆ strain, tetrad dissection plates were incubated at room temperature for >5 days until colonies were formed because the dyn1∆ cells grew extremely slowly. Glycerol stocks were prepared from the initial master plate of tetrad analysis. YPDS liquid medium was inoculated with either the initial colonies from tetrad dissection or glycerol stocks, and the resulting cells were subjected to analyses.

Table 1
Yeast strains and plasmids
https://doi.org/10.7554/eLife.24340.037
NameGenotype/species/construction Source or reference
O. polymorpha strains
BY4329leu1-1NBRP
BY21401CBS4732 Type strainNBRP
HPH31HHT1::pHM713 ura3-1this study
HPH41ura3-1 pHM719this study
HPH164HHT1::pHM726 TUB1:: pHM737 leu1-1this study
HPH194TUB1::pHM737 leu1-1this study
HPH206Δdyn1::natNT2 leu1-1this study
HPH207Δdyn1::natNT2 leu1-1this study
HPH221wild typethis study
HPH222Δbub2::hphNT1this study
HPH223Δkar9::natNT2this study
HPH224Δkar9::natNT2 Δbub2::hphNT1this study
HPH225leu1-1this study
HPH399SPC72-GFP-hphNT1this study
HPH449SPC72-mRFP-hphNT1 TUB1:: pHM737 leu1-1this study
HPH466wild typethis study
HPH475hphNT1-PSPC98-GFP-SPC98this study
HPH972SPC72-GFP-hphNT1this study
HPH1150Δcdc28::natNT2::pHM878 ura3-1this study
HPH1210Δcdc28::natNT2::pHM878 TUB1:: pHM737 HHT1::pHM713 ura3-1this study
HPH1380SPC72-GFP-hphNT1 CDC5-3mAID-natNT2 ura3-1::pHM922this study
HPH1393SPC72-GFP-hphNT1 MPS3-mRFP-kanMX6 ura3-1::pHM859this study
HPH1394SPC72-GFP-hphNT1 MPS3-mRFP-kanMX6this study
HPH1396hphNT1-PNUD1-GFP-NUD1this study
HPH1400SFI1-GFP-hphNT1this study
HPH1405SPC110-GFP-hphNT1this study
HPH1430SPC72-5flag-hphNT1 CDC5-3mAID-natNT2 ura3-1::pHM922this study
HPH1542SPC72-GFP-hphNT1 MPS3-mRFP-kanMX6 TEF1::pHM950::pHM956this study
HPH1562CDC5-GFP-hphNT1 MPS3-mRFP-kanMX6this study
HPH1564CDC5-GFP-hphNT1 MPS3-mRFP-kanMX6 HHT1::pHM726this study
HPH1575CDC5-GFP-hphNT1 SPC72-tdTomato-hphMXthis study
HPH1581SPC72-tdTomato-hphMX SPC110-GFP-hphNT1this study
HPH1583CDC5-GFP-hphNT1 SPC110-tdTomato-natNT2this study
HPH1678MPS3-GFP-hphNT1 HHT1::pHM713this study
other yeast strains
BY21467S. cerevisiae YPH499NBRP
BY21165Kluyveromyces lactis NH27NBRP
BY21167Yarrowia lipolytica T22NBRP
BY23876Candida glabrata YAT3377NBRP
BY5243Ogataea parapolymorpha DL-1NBRP
JCM9829Candida peltataJCM
JCM 10237Ogataea methanolicaJCM
JCM15019Ambrosiozyma kashinagacolaJCM
plasmids
pHM713pREMI-Z carrying HHT1-GFP and HHF1(histoneH4)this study
pHM719pKS144 carrying TUB4-GFPthis study
pHM726pREMI-Z carrying HHT1-mCherry and HHF1(histoneH4)this study
pHM737pRS305 carrying PTUB1-GFP-TUB1this study
pHM859pBSII carring HpURA3 and PTEF1-HpSPC72-GFPthis study
pHM878pBSII carring HpURA3 and cdc28-asthis study
pHM922pBSII carring HpURA3 and PCDC28-OsTIRthis study
pHM950pFA6a-natNT2 carrying PTEF1-HpCDC5Δ53this study
pHM956pFA6a-natNT2 carrying PTEF1-HpCDC5Δ53 and zeothis study

Yeast growth conditions and general methods

Yeast strains were grown either in YPD medium containing 200 mg/l adenine, leucine, and uracil (YPDS) or in synthetic/defined (SD) medium supplemented with appropriate amino acids and nucleotides (Sherman, 1991). Cells were grown at 30°C unless otherwise indicated. To depolymerize MTs, cells were incubated in either YPDS medium or SD medium containing 1.5 μg/ml nocodazole at 30°C for 1.5 hr.

Microscopy

Yeast cells carrying GFP-TUB1, HHT1-mCherry, SPC72-GFP, GFP-NUD1, SFI1-GFP, SPC98-GFP, TUB4-GFP, SPC110-GFP, CDC5-GFP, or MPS3-mRFP were immediately analysed by fluorescence microscopy without washing or fixation in Figures 1A, 2, 4, 5, 6B–E, 7 and 8A, Figure 4—figure supplement 1, Figure 5—figure supplement 1. For the visualisation of DNA with 4'6,-diamidino-2-phenylindole (DAPI), cells were fixed with 70% ethanol, washed with phosphate buffered saline (PBS), and incubated in PBS containing 1 µg/ml DAPI.

Z-series images of 0.4 μm steps were captured with DeltaVision (Applied Precision, Issaquah, WA, USA) equipped with GFP and TRITC filters (Chroma Technology Corp., Bellows Falls, VT, USA), a 100 × NA 1.4 UPlanSApo oil immersion objective (IX71; Olympus, Tokyo, Japan), and a camera (CoolSNAP HQ; Roper Scientific, Trenton, NJ, USA) and were quantified/processed with SoftWoRx 3.5.0 (Applied Precision, Issaquah, WA, USA) or Prism4.3.0 software (Chen et al., 1992; Chen et al., 1996). Deconvolved and projected images are shown. The fluorescence intensity of Spc72-GFP was measured on a plane that has an SPB in focus. Time-lapse experiments of Figures 1A and 2D and that of Figure 4C were carried out in YPDS and SD complete medium respectively on a glass-bottom dish (MatTek, Ashland, MA, USA) coated with concanavalin A (037–08771, Wako, Japan) at room temperature. Z series at 0.4 μm steps were acquired every 3 min for Figures 1A and 4C, or every 30 s for Figure 2D.

For Figure 1B, cells were fixed with 70% ethanol, washed with PBS, and incubated in PBS containing 1 mg/ml DAPI to visualize the DNA (Maekawa and Kaneko, 2014). ImageJ 1.47 (NIH, Bethesda, MD, USA) and Photoshop (Adobe Systems, San Jose, CA, USA) were used to mount the images and to produce merged colour images. No manipulations other than contrast and brightness adjustments were used.

To exclude cells that were non-proliferating from the GFP intensity measurements in Figure 4B, cells were first incubated in YPDS containing Alexa 594 conjugated concanavalin A (Thermo Fisher Scientific, Waltham, MA, USA) until all cells were labelled and then washed once with YPDS and incubated in YPDS for 1 hr prior to image capture. Cells that had lost the label or had a bud with no label were subject to the analyses.

Electron microscopy

Cells were mounted on a glass-bottom dish (MatTek) coated with concanavalin A and covered with fixative [2% glutaraldehyde in 0.1 M sodium phosphate buffer (pH 7.2)]. After 1 min, cells were further fixed with fresh fixative for 2 hr at 4°C. After washing with buffer, low melting agarose was applied onto the cells to prevent loss of cells during subsequent procedures. Zymolyase solution (0.4 mg⁄ml zymolyase 100T, Seikagaku Co., Tokyo, Japan) was applied on top of the agarose for 60 min at 37°C, postfixed with 2% OsO4 for 2 hr at room temperature, stained with 1% uranyl acetate for 1 hr, dehydrated with acetone in an ascending series from 50% to 100%, and embedded in epoxy resin. Serial sections of 80 nm thicknesses were obtained, poststained with uranyl acetate and lead citrate, and analysed using a Zeiss EM900 Transmission Electron Microscope at Central Unit Electron Microscopy in the German Cancer Research Center (DKFZ) (ZEISS, Oberkochen, Germany) or a Hitachi H-7500 Transmission Electron Microscope at Research Centre for Ultra-High Voltage Electron Microscopy at Osaka University (Hitachi, Tokyo, Japan).

Cell cycle analysis and growth conditions

For synchronization, CDC5-3mAID cells were incubated in YPDS containing indole-3-acetic acid (IAA) (45533, Sigma-Aldrich, St. Louis, MO, USA)) for 2.5 hr at 30°C until >80% of cells had a large sized bud to deplete Cdc5 (Nishimura and Kanemaki, 2014). Cells were then washed with pre-warmed growth medium to remove IAA and re-suspended in YPAD medium at 30°C.

Yeast cell extract and immunoblotting

Whole cell extracts were prepared for SDS-PAGE and immunoblotting (Knop et al., 1999; Janke et al., 2004; Meitinger et al., 2016). Samples representing 1–2 OD600 of liquid culture were resuspended in 950 μl 0.29 M NaOH and incubated on ice for 10 min. Then, 150 μl 55% (w/v) trichloroacetic acid was added and the solutions were mixed and incubated for 10 min on ice. After centrifugation the supernatant was removed. The protein pellet was resuspended in high urea buffer (8 M urea, 5% SDS, 200 mM NaPO3 pH 6.8, 0.1 mM EDTA, 100 mM dithiothreitol, and bromophenol blue) and heated at 65°C for 10 min. A sample comprising one-fifth of the total sample amount was loaded for SDS-PAGE (Figure 6A) and western blotting was performed using a standard protocol. For immunoprecipitation, total cell extracts were prepared from logarithmically growing cells in immunoprecipitation buffer (100 mM Tris, pH 8.0, 10 mM EDTA, 150 mM NaCl, 5% glycerol, 0.2 mM NaVO3100 mM β-glycerophosphate, 50 mM NaF, 1 mM PMSF, 1 mM DTT, 1% NP-40, and Complete EDTA-free protease inhibitor cocktail [Roche]). 10 mg of total cell extract was incubated with M2-bound magnetic beads (M8823, Sigma) for 2 hr at 4°C. The beads were washed three times with immunoprecipitation buffer. The bound proteins were subjected to λ phosphatase treatment and then eluted in 30 μl of SDS-PAGE sample buffer by incubated at 37°C for 30 min. 3 μl of eluates were loaded on a Mini-PROTEAN TGX Precast Gels (4561021, BIO-RAD Laboratories, Hercules, CA, USA) and western blotting was performed using a standard protocol. Monoclonal antibodies JL-8 (632381, TaKaRa Bio Clontech, Shiga, Japan) and M2 (F1804, Sigma) were used to detect GFP- and flag-tagged proteins respectively. Plot profile function of ImageJ was used to plot intensity value across a line in Figure 6—figure supplement 2.

Structured illumination microscopy (SIM)

Cells were arrested for 2.5 hr with 1.5 µg/ml nocadazole and fixed for 15 min in 4% paraformaldehyde/2% sucrose in phosphate-buffered saline (PBS) solution followed by extensive washing in PBS. The cells were immobilized on a concanavalin A (Sigma-Aldrich, MO, USA)- coated 35 mm glass bottom dish (MatTek, P35G-1.5–14C) and maintained in PBS for the duration of the imaging process in PBS. The samples were imaged in the 2D-SIM mode on a Nikon N-SIM system (Tokyo, Japan) equipped with a TIRF Apochromat 100x/1.49 NA oil immersion objective and a single photon detection EM-CCD camera (Andor iXon3 DU-897E; Belfast, UK). The 488 nm and 561 nm laser lines were used for excitation of yeGFP and tdTomato, respectively, combined with emission band pass filter 520/45 and 610/60. Images were taken sequentially within a small z-stack and in consideration of imaging SPBs close to the coverslip to minimise spherical aberrations. Subsequently the reconstruction and channel alignment was performed using the NIS imaging and image analysis software (Nikon). For the xyz chromatic shift correction we used in a reference sample tetraspeck beads in a reference sample. All images show a single stack of the z-slices.

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
    Behavior of spindles and spindle plaques in the cell cycle and conjugation of Saccharomyces cerevisiae
    1. B Byers
    2. L Goetsch
    (1975)
    Journal of Bacteriology 124:511–523.
  10. 10
  11. 11
    PRIISM: an integrated system for display and analysis of 3-D microscope images
    1. H Chen
    2. WK Clyborne
    3. JW Sedat
    4. DA Agard
    (1992)
    SPIE/IS&T 1992 Symposium on Electronic Imaging: Science and Technology. pp. 784–790.
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
    The spindle pole body component Spc98p interacts with the gamma-tubulin-like Tub4p of Saccharomyces cerevisiae at the sites of microtubule attachment
    1. S Geissler
    2. G Pereira
    3. A Spang
    4. M Knop
    5. S Souès
    6. J Kilmartin
    7. E Schiebel
    (1996)
    The EMBO Journal 15:3899–3911.
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
    The Yeasts
    1. C Kurtzman
    2. JW Fell
    3. T Boekhout
    (2011)
    Elsevier.
  41. 41
    Disruption of mitotic spindle orientation in a yeast dynein mutant
    1. YY Li
    2. E Yeh
    3. T Hays
    4. K Bloom
    (1993)
    PNAS 90:10096–10100.
  42. 42
    Cloning, sequencing, and functional analysis of H-OLE1 gene encoding delta9-fatty acid desaturase in Hansenula polymorpha
    1. SF Lu
    2. II Tolstorukov
    3. S Anamnart
    4. Y Kaneko
    5. S Harashima
    (2000)
    Applied Microbiology and Biotechnology 54:499–509.
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
    Mitosis in Heterobasidiomycetous Yeasts. 1. Leucosporidium-Scottii (Candida-Scottii)
    1. EK McCully
    2. CF Robinow
    (1972a)
    Journal of Cell Science 10:857–881.
  48. 48
    Mitosis in Heterobasidiomycetous Yeasts. 2. Rhodosporidium Sp (Rhodotorula-Gultinis) and Aessosporon-Salmonicolor (Sporobolomyces-Salmonicolor)
    1. EK McCully
    2. CF Robinow
    (1972b)
    Journal of Cell Science 11:1–31.
  49. 49
  50. 50
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
  57. 57
  58. 58
  59. 59
  60. 60
  61. 61
  62. 62
  63. 63
  64. 64
  65. 65
    Getting started with yeast
    1. F Sherman
    (1991)
    Methods in Enzymology 194:3–21.
  66. 66
  67. 67
  68. 68
  69. 69
    A novel function of Saccharomyces cerevisiae CDC5 in cytokinesis
    1. S Song
    2. KS Lee
    (2001)
    The Journal of Cell Biology 152:451–469.
  70. 70
    SPC72: a spindle pole component required for spindle orientation in the yeast Saccharomyces cerevisiae
    1. S Souès
    2. IR Adams
    (1998)
    Journal of Cell Science 111:2809–2818.
  71. 71
  72. 72
    Microtubules in the fungal pathogen Ustilago maydis are highly dynamic and determine cell polarity
    1. G Steinberg
    2. R Wedlich-Söldner
    3. M Brill
    4. I Schulz
    (2001)
    Journal of Cell Science 114:609–622.
  73. 73
  74. 74
  75. 75
  76. 76
  77. 77
  78. 78
  79. 79

Decision letter

  1. Anna Akhmanova
    Reviewing Editor; Utrecht University, Netherlands

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

Thank you for submitting your article "Polo-like kinase Cdc5 regulates Spc72 recruitment to spindle pole body in the methylotrophic yeast Ogataea polymorpha" for consideration by eLife. Your article has been reviewed by three peer reviewers, one of whom is a Guest Reviewing Editor for eLife, and the evaluation has been overseen by Anna Akhmanova as the Senior Editor. The reviewers have opted to remain anonymous.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

The orientation of the mitotic spindle axis determines the direction of chromosome segregation in a spatiotemporal manner. In many eukaryotic cells undergoing close mitosis such as yeast, intranuclear spindle positioning is controlled by cytoplasmic microtubules (cMTs). A similar role is played by astral microtubules in open mitosis.

In their manuscript, Maekawa et al. undertake the first characterisation of spindle positioning in the budding yeast O. polymorpha. The work reveals outstanding differences in the sequence of events previously established in S. cerevisiae, a well-studied paradigm. These idiosyncrasies are invaluable for dissecting cell cycle controls that are more apparent in one particular system while nevertheless conserved, as they converge on the spindle pole body or SPB, the yeast functional equivalent of the centrosome. Maekawa et al. present evidence for unbiased nuclear position (and the spindle within) in O. polymorpha at early stages of the cell cycle. However, the elongating spindle aligns along the mother-bud axis and inserts into the bud neck during anaphase. By contrast, the spindle in S. cerevisiae adopts a polarised orientation much earlier via prominent cMTs. Accordingly, Maekawa et al. observed a very low number of cMTs from G1 to anaphase onset in O. polymorpha. The authors linked this distinction to the late addition of Spc72 (a γ-tubulin complex receptor) to the SPB cytoplasmic face, under the control of the polo-like kinase Cdc5. Supporting this view, depletion or overexpression of Cdc5 abolished or anticipated Spc72 recruitment, respectively. Furthermore, cells overexpressing Cdc5 preempted nuclear migration close to the bud neck, pointing to Spc72 recruitment as a possible trigger for spindle positioning.

The reviewers have concurred that these correlations constitute important and interesting findings. However, the key claim that cell cycle-dependent recruitment of Spc72 controls cMT organisation and nuclear positioning in O. polymorpha should be supported by direct evidence and not solely left to be inferred from those correlations.

Therefore, as detailed below, the reviewers recommend that:

– direct observation of cMTs should be carried out to verify their active involvement in nuclear/spindle positioning along the cell cycle.

– conclusive evidence supporting the nature of the link between Cdc5 and Spc72 should be provided.

Essential revisions:

1) Since this is a first characterisation, an overview of the spindle pathway should be documented by providing:

– A set of representative images for an SPB marker and the nucleus at the different stages along the entire cell cycle.

– Time lapse data of cells expressing an SPB marker to document the kinetics of SPB separation in order to define discrete stages by length along the spindle pathway, as previously done in S. cerevisiae. Stages such as "large budded preanaphase cells" (e.g. Results section first paragraph) should be correctly defined in quantitative terms.

2) The authors should report the impact of deleting SPC72.

3) Experiments were performed exclusively in vegetative cells. Are Spc72 recruitment and cMT organisation affected by nutrient conditions that induce mating? Are cMTs expected to have a role during mating?

There might be key differences regarding the (half)bridge with respect to S. cerevisiae (Figure 3). This should be stated clearly for its implications in cMT organisation along the cell cycle. Could the authors provide more detailed EM analysis? In particular, does SPB duplication proceed along the lines shown for S. cerevisiae? Is the (half)-bridge visible at any point before or after duplication? Is there a satellite in late G1? Could you provide corresponding images for S. cerevisiae SPBs obtained under identical conditions to appreciate differences in outer plaque appearance?

4) Nuclear position in O. polymorpha is described as random in the Abstract and main text. Yet, the nucleus tends to position centrally (Results section first paragraph and Figure 1A). Time lapse must be presented to cover the events leading to this central position using cells expressing GFP-Tub1 for the interval between mitotic exit (when nuclei are positioned at opposite ends by the spindle) and the emergence of the bud in the following cell cycle to draw conclusions regarding any relationship between the axis of cell polarity and SPB/nuclear position. Please clarify and amend the text accordingly.

5) Subsection “O. polymorpha cells contain only fewer cMTs”, concerning cMT detachment. It is unclear how cMTs that dissociate from the SPB arise. Are cMTs indeed "unstable"? Time-lapse analysis should be implemented to document in more detail dynamic instability parameters as well as the timeline from cMT emergence to detachment.

6) Figure 2D and Figure 2—figure supplement 1. The cMTs in the time-lapse series are very hard to visualize. Arrows pointing at cMTs should be added and the basis for quantification of cMT persistence and cMT re-establishment should be fully described so that the significance of these data is made clear.

7) Figure 4. It is unclear when Spc72 level changes along the preanaphase interval. Is label lost after a short spindle forms? Furthermore, there are issues of consistency throughout the manuscript regarding images and quantification of Spc72-GFP signals. In Figure 4A, G1 cells do not show Spc72-GFP signal. Yet, the quantification in Figure 4B points to a higher signal in G1 than in S/G2. Again, in Figure 4C at time 0 signal is absent despite the quantification shown before. Later on, in Figure 8D, quantification of Spc72-GFP at different stages of the cell cycle indicates lower signals than in Figure 4B. It is important to explain this variability and show the quantification of several replicates of the experiments.

In Figure 4B, the legend describes "signal intensities were background-subtracted". This may not be a proper method. Instead, Mps3-mRFP signals should be used for internal controls. This is also the case for Figure 6C.

With regards to Figure 4C, unfortunately the interval of live imaging, i.e. 3 min, is too long to visualise the timing of the increase in Spc72-GFP signals during the cell cycle.

Time lapse analysis of cells expressing Spc72-GFP should be extended to document how the label is lost at the short spindle stage and regained close to anaphase as hinted by still image data, using suitable temporal resolution (at least 1 min intervals).

8) On overproduction of Spc72 (Figure 5). The authors claim that by overproducing Spc72, SPBs become localised to the vicinity of the bud neck in preanaphase cells. It is critical to observe cMTs. Please show MT structures, possibly by time-lapse live imaging.

9) In Figure 6, cells were synchronised using an inducible degron to deplete Cdc5. However, in support of the authors' claims, time course analysis following release from a nocodazole block should be carried out to monitor Spc72 levels, localisation and phosphorylation in the next cell cycle in otherwise wild type cells.

Regarding phosphorylation of Spc72 (Figure 6—figure supplement 2). The difference between the lanes is too subtle to conclude on the specific contribution of Cdc5. The experiment should be performed using Phos-tag gels or changing the acrylamide/bis ratio to better display Spc72 phosphoisoforms. An alternative mode of synchronisation could be implemented, since nocodazole arrest is likely to further enhance phosphorylation thus masking the specific contribution of Cdc5. Finally, the authors should provide data to address the physical interaction between Cdc5 and Spc72 and in vitro kinase assays to demonstrate whether Spc72 is a Cdc5 substrate.

10) On Cdc5 localisation (Figure 7). Using SIM, the authors claim that Cdc5 is localised exclusively to the cytoplasmic side of the SPB during metaphase. Are the authors claiming that Cdc5-polo kinase plays no roles inside the nucleus during mitosis in O. polymorpha? Please clarify this issue.

Along the same lines, most SPB markers in S. cerevisiae contribute cytoplasmic label by a free pool, the same holds true for Cdc5. The images in this manuscript virtually lack any whole cell background – this suggests over-processing. Diffuse nuclear label might have been lost as well here. The authors should provide sample raw images to assess whether loss of whole-cell and nuclear signals might be due to processing.

11) Regarding Cdc5 overproduction (Figure 8). The impact of manipulating Cdc5 on cMT presence should be assessed directly in cells expressing GFP-Tub1 to score cMTs along the cell cycle. In wild type or Cdc5 overexpressors, when do cMT plus ends access the bud?

Figure 8D presents an increase in Spc72-GFP signal rather than advanced recruitment of Spc72 along the cell cycle. This rise might be due to increased Spc72 protein level and/or phosphorylation. Controls should be added to report on Spc72 levels and modification upon Cdc5 overexpression. Moreover, Cdc5 localisation upon overexpression should be analysed to determine whether Cdc5 association to the SPB is also affected.

Finally, it remains unclear whether Cdc5 kinase activity or only Cdc5 association to the SPB is the key determinant for Spc72 recruitment. While SPB tethering experiments might help resolve this point, the authors should at least address whether the ability of Cdc5 to recruit Spc72 to the SPB may involve their physical interaction. To this end, co-immunoprecipitation assays or protein pull downs with recombinant baits should be provided.

12) Could the authors propose how Spc72 acquisition brings about spindle alignment? At least they should provide time lapse data of cells expressing GFP-Tub1 at sufficient temporal resolution to visualise cMT acquisition and the dynamics of spindle alignment and insertion across the bud neck to judge the involvement of cMTs in response to Spc72 loading.

13) Figure legends should be edited for clarity and plots should be described in full for what is being scored in every case. Arrows or arrowheads should be added in every image to point at relevant structures that are difficult to discern without this aid as requested above.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Polo-like kinase Cdc5 regulates Spc72 recruitment to spindle pole body in the methylotrophic yeast Ogataea polymorpha" for further consideration at eLife. Your revised article has been favorably evaluated by Anna Akhmanova (Senior editor), and a Guest Reviewing editor.

The manuscript has been improved but there are some remaining issues that need to be addressed by editing the text of the manuscript before acceptance, as outlined below.

The authors have addressed the bulk of the original review but left some specific points unanswered, mainly on the grounds of technical limitations in the use of O. polymorpha as a model. Given this problem, those points should be qualified correctly in the final report.

Taken together, the manuscript portrays substantial novelty if focused on the behaviour of Spc72, Cdc5 and cMTs. Thus, it would be important to carry out a number of revisions in the text for the narrative in the Results and Discussion to be brought in line with the level of proof. Following is a list of action points and we would like to encourage the authors to revise accordingly so that data are not mis-represented here.

1) Concerning the path for spindle orientation in O. polymorpha

Although nuclear position is not biased toward the bud neck, SPB movement depicted in the time lapse series for SPB localisation is inconsistent with the statement that position is entirely random (e.g. Discussion first paragraph), as it is steadily maintained at the mother cell centre. The temporal resolution used in this study may not reveal the underlying basis for this position but it does not convey proof of random location either. Thus, the text should be amended here to reflect the actual data.

Without detracting from the key conclusion on the role of Spc72 recruitment, we note that the representative time lapse series presented as supplemental data to Figure 1 showed the following: a) In Figure 1—figure supplement 4, after assembly 1 µm long spindles remain centrally positioned and loosely oriented toward the bud neck b) According to Figure 1—figure supplement 5, spindle alignment is apparent at onset of spindle elongation followed later by SPB translocation. Instead, in S. cerevisiae, SPB translocation (of spindles held aligned during metaphase) coincides with spindle elongation. The analysis and narration of the current Figure 1 mis-represents this dynamic behaviour. We would recommend moving the time lapse series to be part of the main Figure 1 and, if space is limiting, move supporting quantitation of still images to the supplemental section of Figure 1. Furthermore, the narration should be edited to reflect these data faithfully.

2) On Dyn1 and Kar9 pathways and SPOC function in O. polymorpha (final paragraph in subsection “Nuclear positioning in O. polymorpha differs from that in S. cerevisiae and other budding yeast species” and second paragraph of the Discussion section) the data are not sufficiently developed to make these claims and should be substantially edited for the following reasons:

– In the new sample image of dyn1∆ cells (Figure 1—figure supplement 6) it is very difficult to tell the nuclear stain from the cell background and hard to see how the quantitation was possible based on such data. Furthermore the stain is inconsistent with the live image of nuclear label by tagging. These data do not support conclusions regarding nuclear size as a differential factor between S. cerevisiae and O. polymorpha. Finally, the defect in the dyn1∆ mutant may transcend cMT function per se in line with the severe growth defect. Note that dyn1∆ mutant cells appeared to re-bud inappropriately and accumulate supernumerary nuclei, arguing that the SPOC may not operate as robustly (if at all) compared to S. cerevisiae, resulting in low viability of dyn1∆ cells. Furthermore, the interaction between kar9∆ and bub2∆ appears much stronger than what is apparent in S. cerevisiae (in which kar9∆ bub2∆ is not lethal at 30°C)

Under the circumstances, the authors are merely inferring cellular defects that they have not proven satisfactorily, for inclusion in this report. The same holds true for their claim that the Kar9 and Dynein pathways are functionally conserved between the two yeasts. cMT images failed to show plus ends reaching the distal bud cell cortex to talk about Kar9-dependent positioning here.

In conclusion, this section should be deleted from a revised manuscript to maintain the focus on the findings linking cMTs and Spc72. Also note that the authors could not answer here a number of essential queries from the original review on technical grounds.

– Discussion paragraph two. Please revise to simply mention synthetic lethality between kar9 and dyn1 mutants, with dyn1 alone showing a severe growth defect at best. Whether this is due to the demands of aligning the spindle during anaphase (as opposed to S. cerevisiae in which spindle alignment is achieved prior to anaphase) remains highly speculative. The severe growth defect due to dyn1∆ in O. polymorpha is consistent with weak SPOC-dependent delay. As stated above, the authors report excess of fragmented nuclei and re-budding in dyn1∆ cells, supporting this idea.

3) The definition of metaphase spindle is not kept consistent throughout. At one point, metaphase is defined by SPB distance < 2 µm but earlier in the text, it was defined as < 1 µm. Please amend accordingly.

4) The progressive and gradual loss of Spc72 from SPBs at early stages of the cell cycle is inconsistent with Cdc5 acting as the sole switch in Spc72 recruitment (with reversal triggered by mitotic exit). The authors should make this point clearer when discussing Spc72 detection in G1 cells in Figure 4. Please revise to bring the narration in line with the data.

5) In the phosphorylation analysis, the change of tags on Spc72 should be explained in full and justified.

6) The rational for testing Spc72 upon starvation could be made clearer – mating is triggered by starvation in O. polymorpha.

7) Discussion paragraph three, on Nud1 being a possible Cdc5 target – please clarify if you meant that Nud1 contains a consensus site for binding the Cdc5 polo box (rather than a polo box consensus sequence?).

8) The authors should discuss why Spc72 overexpression is sufficient to force recruitment to SPBs in their model (data in Figure 5).

9) The Introduction is incorrect regarding Spc72 and cMTs in S. cerevisiae. Simply put, cMTs at the new SPB are formed after a 1 µm long spindle has formed (certainly not G1/S!, Shaw et al., 1997). Please revise to state correctly that the new SPB acquires Spc72 and cMTs after onset of spindle assembly (also see Juanes et al., 2013).

https://doi.org/10.7554/eLife.24340.043

Author response

Essential revisions:

1) Since this is a first characterisation, an overview of the spindle pathway should be documented by providing:

– A set of representative images for an SPB marker and the nucleus at the different stages along the entire cell cycle.

– Time lapse data of cells expressing an SPB marker to document the kinetics of SPB separation in order to define discrete stages by length along the spindle pathway, as previously done in S. cerevisiae. Stages such as "large budded preanaphase cells" (e.g. Results section first paragraph) should be correctly defined in quantitative terms.

According to the reviewer’s comment, we now analyzed SPB behavior in respect to bud formation and/or chromosome segregation. Representative images of cells with SPB marker is shown in Figure 1F. Time lapse analysis is presented in Figure 1—figure supplement 4 and 5.

Comparison of Mps3-GFP intensity in cells with different bud size suggested SPB duplicates in small budded cells as shown in S. cerevisiae. Duplicated and separated SPBs upon the short spindle formation maintain close proximity until anaphase onset, as shown in Figure 1—figure supplement 4. Anaphase was initiated in cells with large bud and the event occurred within the mother cell body (Figure 1—figure supplement 5). These results are described in paragraph two of subsection “Nuclear positioning in O. polymorpha differs from that in S. cerevisiae and other budding yeast species” of the revised manuscript, and “pre-anaphase cells” are now defined at the end of the paragraph.

2) The authors should report the impact of deleting SPC72.

We analyzed the essentiality of SPC72 for growth by tetrad analysis using a SPC72/spc72Δ::natNT2 heterozygous diploid strain. We found that SPC72 is essential for growth. The result is shown in Figure 4—figure supplement 2 and described in subsection “Spc72 associates with SPB in a cell cycle-dependent manner”.

3) Experiments were performed exclusively in vegetative cells. Are Spc72 recruitment and cMT organisation affected by nutrient conditions that induce mating? Are cMTs expected to have a role during mating?

This is a very interesting point. We now analysed Spc72 localisation in starved conditions. We found that while cMTs were not observed, Spc72-mRFP or Spc72-GFP accumulated at SPBs under starvation conditions (representative images of SPC72-mRFP GFP-TUB1 cells are shown in Figure 5—figure supplement 2). This may suggest that Spc72 recruited to SPBs does not immediately organize cMTs in such conditions. Although we assume that cMTs are involved in karyogamy, we were unable to analyse cMTs in zygotes because mating frequency was low and zygotes failed to proceed karyogamy once they were placed on a glass-bottom dish. We are now discussing these data in the final paragraph of “Spc72 associates with SPB in a cell cycle-dependent manner”.

There might be key differences regarding the (half)bridge with respect to S. cerevisiae (Figure 3). This should be stated clearly for its implications in cMT organisation along the cell cycle. Could the authors provide more detailed EM analysis? In particular, does SPB duplication proceed along the lines shown for S. cerevisiae? Is the (half)-bridge visible at any point before or after duplication? Is there a satellite in late G1? Could you provide corresponding images for S. cerevisiae SPBs obtained under identical conditions to appreciate differences in outer plaque appearance?

We agree that clarifying the structure and cell cycle-dependency of half-bridge and bridge formation is an important issue. In S. cerevisiae, EM studies of satellite formation at the half-bridge have been facilitated by the use of synchronised cultures. Pheromone induced G1 arrest (similar to α-factor synchronisation in S. cerevisiae) is not yet established in O. polymorpha. Thus, we tried alternative strategies. We constructed cdc28-as strain in order to observe SPB in late G1. But cdc28-as cells did not arrest at the same cell cycle stage as in S. cerevisiae. We think that this is because the inhibitor cannot completely inhibit Cdc28-as kinase. Nevertheless, we observed a delay in cell cycle progression (accumulation of cells with a single SPB signal by fluorescence microscopy) upon Cdc28-as inhibition. We could show by EM that some of these cells had side-by-side SPBs, which suggested that SPB duplication proceed along the lines shown for S. cerevisiae. We could detect an electron dense cloud between the two SPBs on the cytoplasmic side of the nuclear envelope (Figure 3—figure supplement 2ABE). Therefore, we think that a bridge-like structure is also formed in O. polymorpha. However, it was not clear whether the half-bridge/bridge like structure was present at other cell cycle stages. These data is presented in Figure 3—figure supplement 2, and the description of the result is in paragraph two of subsection “Organization of the SPB structure on the cytoplasmic side is cell cycle dependent”.

In subsection “Organization of the SPB structure on the cytoplasmic side is cell cycle dependent”, we also added following sentence “while half-bridge-like structure, which plays an important role in cMT organization in G1 of S. ccerevisiae, was not clearly observed”.

In the Discussion section, possible difference of the half-bridge/bridge is described: “Equally important is the analysis of the half-bridge which, in S. cerevisiae, organizes cMTs in G1. If the half-bridge plays the same role in O. polymorpha SPB as that in S. cerevisiae, the loss of Spc72 from outer plaque alone should not cause the loss of cMTs”.

Also, earlier timing of outer plaque appearance than Spc72-GFP is stated in the Discussion: “Furthermore, the appearance of the outer plaque in EM analysis proceeded that of Spc72 in fluorescent microscopy (Figure 9).”

4) Nuclear position in O. polymorpha is described as random in the Abstract and main text. Yet, the nucleus tends to position centrally (Results section first paragraph and Figure 1A). Time lapse must be presented to cover the events leading to this central position using cells expressing GFP-Tub1 for the interval between mitotic exit (when nuclei are positioned at opposite ends by the spindle) and the emergence of the bud in the following cell cycle to draw conclusions regarding any relationship between the axis of cell polarity and SPB/nuclear position. Please clarify and amend the text accordingly.

We think that because nucleus is a large structure relative to the cell size, it is more likely to be located near the cell centre than at the periphery in the absence of external forces. Thus, we do not think that there is a specific regulation for positioning the nucleus centrally. To clarify this point, time lapse movies of SPB movement during anaphase to G1 of the next cell cycle are presented as Figure 1—figure supplement 5. In these movie series, we could see the rapid and continuous movement of SPBs in the entire cell body. Thus, we concluded that SPB position is random in G1. The result is described in the text: “After spindle breakdown, the SPB moved vigorously in no relationship to the polarity axis (Figure 1—figure supplement 5).”

5) Subsection “O. polymorpha cells contain only fewer cMTs”, concerning cMT detachment. It is unclear how cMTs that dissociate from the SPB arise. Are cMTs indeed "unstable"? Time-lapse analysis should be implemented to document in more detail dynamic instability parameters as well as the timeline from cMT emergence to detachment.

We have now performed Time lapse microscopy using cells expressing GFP-TUB1 and HHT1-mCherry. We see that cMTs dettached from SPBs and remained in the cytoplasm for only a short period of time before disappearing. This data thus indicate that detached cMT are short-lived. We could not determine dynamic instability parameters due to fast disappearance of detached cMTs. Therefore, to avoid confusion, we replaced “unstable” with “short-lived”.

Representative images of time lapse series showing cMT detachment are shown in Figure 2—figure supplement 1, and the result is described in the text: “Time lapse analysis revealed that detached cMTs remained in the cytoplasm only for a short period of time before depolymerized (Figure 2—figure supplement 1).”

6) Figure 2D and Figure 2—figure supplement 1. The cMTs in the time-lapse series are very hard to visualize. Arrows pointing at cMTs should be added and the basis for quantification of cMT persistence and cMT re-establishment should be fully described so that the significance of these data is made clear.

Thanks for pointing this out. For better visualization of cMTs, we have replaced one of the time lapse sequences with another example in which cMTs are easily visible in Figure 2—figure supplement 3. Yellow arrows are added to point cMTs.

7) Figure 4. It is unclear when Spc72 level changes along the preanaphase interval. Is label lost after a short spindle forms? Furthermore, there are issues of consistency throughout the manuscript regarding images and quantification of Spc72-GFP signals. In Figure 4A, G1 cells do not show Spc72-GFP signal. Yet, the quantification in Figure 4B points to a higher signal in G1 than in S/G2. Again, in Figure 4C at time 0 signal is absent despite the quantification shown before. Later on, in Figure 8D, quantification of Spc72-GFP at different stages of the cell cycle indicates lower signals than in Figure 4B. It is important to explain this variability and show the quantification of several replicates of the experiments.

We are sorry for this confusion. Decrease of Spc72-GFP SPB signal after mitotic exit is slow and the timing varied from cell to cell. We think this may be the reason why some of G1 cells carry strong Spc72-GFP signal. Time-lapse series from anaphase to the next cell cycle are now shown in Figure 4—figure supplement 4, and the results are described in the text: “As cells exit from mitosis and entre into the next cell cycle, Spc72-GFP signal was gradually decreased at SPBs with the timing that varied from cell to cell. However, in all cases, Spc72-GFP levels reached a minimum well before short spindle was formed (Figure 4 and Figure 4—figure supplement 4).”

In Figure 4B, the legend describes "signal intensities were background-subtracted". This may not be a proper method. Instead, Mps3-mRFP signals should be used for internal controls. This is also the case for Figure 6C.

In all quantification, signal intensities were background-subtracted. We do not think that Mps3-mRFP is a suitable internal control in O. polymorpha because Mps3-mRFP signal bleached so fast that its intensity decline even during z-series acquisition. There is no other SPB components that gives stronger RFP signal than Mps3. Therefore we think that the background-subtraction is currently the best method available.

With regards to Figure 4C, unfortunately the interval of live imaging, i.e. 3 min, is too long to visualise the timing of the increase in Spc72-GFP signals during the cell cycle.

Time lapse analysis of cells expressing Spc72-GFP should be extended to document how the label is lost at the short spindle stage and regained close to anaphase as hinted by still image data, using suitable temporal resolution (at least 1 min intervals).

We now show new time lapse series of Spc72-GFP with 30 sec or 1 min intervals are shown in Figure 4C and Figure 4—figure supplement 3 and 4.

The text is changed to “In all cells that progressed into anaphase, an Spc72-GFP signal became detectable < 4 min prior to the initiation of anaphase (average 3.68 ± 1.74 min, n=14) (Figure 4C yellow arrowhead). Within 5 min after appearance of the Spc72-GFP signal, spindle orientation was corrected when it had not done already (Figure 4—figure supplement 3, average 3.50 ± 1.61 min, n=12); therefore, one half part of an anaphase nucleus was successfully inserted into the bud.”

8) On overproduction of Spc72 (Figure 5). The authors claim that by overproducing Spc72, SPBs become localised to the vicinity of the bud neck in preanaphase cells. It is critical to observe cMTs. Please show MT structures, possibly by time-lapse live imaging.

According to the reviewer’s comment, we visualized cMTs by introducing CFP-TUB1 in SPC72-overexpressing cells (Figure 5B and 5C). cMTs are now observed more frequently and the following sentence was added; “cMTs are more often observed (Figure 5B and 5C)”.

9) In Figure 6, cells were synchronised using an inducible degron to deplete Cdc5. However, in support of the authors' claims, time course analysis following release from a nocodazole block should be carried out to monitor Spc72 levels, localisation and phosphorylation in the next cell cycle in otherwise wild type cells.

In our hands, the only synchronization protocol that arrest cells and then allow them to resume the cell cycle was the Cdc5-depletion/re-expression. Other methods did not work. For examples, although O. polymorpha cells can be arrested by nocodazole, they cannot be easily released. Even at the lowest concentration that causes cell cycle arrest, cells required more than two hours to resume cell cycle and did so in an unsynchronized fashion. We generated degron strains for CDC4, MCM4, MCM10 as well as cdc28-as, but none of them worked well enough for synchronization.

In order to address the fluctuation of Spc72 protein level in cell cycle, we have compared Spc72 protein levels in cycling cells, nocodazole arrested cells (metaphase), cdc28-as cells (delayed G1/S phase), CDC5-3mAID cells (late anaphase). We found that the amount of Spc72 protein was at the comparable levels in all cells. The result is presented in Figure 6—figure supplement 1A and described in the text; “Spc72 protein abundance did not fluctuate as cells entered into anaphase and proceeded into the following cell cycle (Figure 6A and Figure 6—figure supplement 1A).”

As for the localization of Spc72 from anaphase to the next cell cycle, we performed time lapse analysis in asynchronized SPC72-GFP cells. Although we could not obtain time lapse series that covers the entire cell cycle due to bleaching of GFP signal, we could show that Spc72-GFP disappeared as cells exit mitosis and before the bud emergence of the following cell cycle. The results are presented in Figure 4C (4.5 min and 17 min) and Figure 4—figure supplement 4. Appearance of Spc72-GFP prior to anaphase onset is presented in Figure 4C (1 min) and Figure 4—figure supplement 3 (5.5 min). The results are described in subsection “Spc72 associates with SPB in a cell cycle-dependent manner”

We have compared phosphorylation status of Spc72 in asynchronous and nocodazole-arrested cells and found that it migrated slower in nocodazole-arrested cells. This result suggests Spc72 is subjected to cell cycle dependent phosphorylation. The result is now presented in Figure 6—figure supplement 1B and described in the text: “Furthermore, the Spc72 band migrated slower in nocodazole-arrested cells than that in asynchronous cells (Figure 6—figure supplement 1B).”

Regarding phosphorylation of Spc72 (Figure 6—figure supplement 2). The difference between the lanes is too subtle to conclude on the specific contribution of Cdc5. The experiment should be performed using Phos-tag gels or changing the acrylamide/bis ratio to better display Spc72 phosphoisoforms. An alternative mode of synchronisation could be implemented, since nocodazole arrest is likely to further enhance phosphorylation thus masking the specific contribution of Cdc5. Finally, the authors should provide data to address the physical interaction between Cdc5 and Spc72 and in vitro kinase assays to demonstrate whether Spc72 is a Cdc5 substrate.

We performed western blotting analysis to better show Cdc5-dependent phosphorylation of Spc72 using total protein prepared under a denatured condition (Figure 6—figure supplement 3A). In our hand, Phos-tag gels did not help resolving Cdc5-dependent phosphorylation.

As for physical interaction between Cdc5 and Spc72, we did in vitro pull-down assay using recombinant Spc72 as a bait, but the result was inconclusive because of non-specific binding between Cdc5 and sepharose beads (data not shown). We also performed in vitro kinase assay with negative results. However, we cannot completely exclude that Spc72 is not a direct substrate of Cdc5. Previously, we have shown that S. cerevisiae Cdc5 can phosphorylate recombinant Spc72, but incorporation of 32P was much stronger for the degradation products of the recombinant Spc72 (Maekawa et al., 2007). This may suggest that a specific conformation or SPB binding is necessary for the reaction. Whether Spc72 is directly phosphorylated by Cdc5, is therefore still an open question. This point is further discussed in the text: “Cdc5 may have another important substrates other than Spc72. Notably, a polo box binding site present in ScSpc72 is missing in OpSpc72. Cdc5 might therefore phosphorylate other SPB proteins such as Nud1, which has one site matching the polo box consensus sequences (S-Sp/Tp-P), and thereby indirectly influence the affinity of Spc72 towards the SPB.”

10) On Cdc5 localisation (Figure 7). Using SIM, the authors claim that Cdc5 is localised exclusively to the cytoplasmic side of the SPB during metaphase. Are the authors claiming that Cdc5-polo kinase plays no roles inside the nucleus during mitosis in O. polymorpha? Please clarify this issue.

We thank the reviewer to point it out. We did observe nuclear signal of Cdc5-GFP and assume that it has nuclear functions during the cell cycle. To clarify this, we added the following sentence: “Nuclear and NE localisation appeared at early stages of the cell cycle and persisted until the end of mitosis”. Also we modified the text; “Thus, Cdc5 likely becomes first localised to the nucleus and the NE in G2, and then in mitosis to the cytoplasmic side of SPBs.”

The legend of Figure 7 now clearly states the nuclear localization of Cdc5-GFP.

Along the same lines, most SPB markers in S. cerevisiae contribute cytoplasmic label by a free pool, the same holds true for Cdc5. The images in this manuscript virtually lack any whole cell background – this suggests over-processing. Diffuse nuclear label might have been lost as well here. The authors should provide sample raw images to assess whether loss of whole-cell and nuclear signals might be due to processing.

We thank the reviewer to point it out. We prepared the high intensity version of SIM images of Cdc5-GFP, in which diffuse nuclear signal is easily visible. It is present in Figure 7—figure supplement 2.

11) Regarding Cdc5 overproduction (Figure 8). The impact of manipulating Cdc5 on cMT presence should be assessed directly in cells expressing GFP-Tub1 to score cMTs along the cell cycle. In wild type or Cdc5 overexpressors, when do cMT plus ends access the bud?

In order to address this point, cMTs were visualized by introducing CFP-TUB1 in CDC5-overexpressing cells (Figure 8E and 8F). CFP-Tub1 signal was not strong enough to unambiguously determine the timing of cMT contact with the bud cortex. However, cMTs were more frequently observed. This is now discussed in the text: “and cMTs were more prevalent (Figure 8E and 8F).”

Figure 8D presents an increase in Spc72-GFP signal rather than advanced recruitment of Spc72 along the cell cycle. This rise might be due to increased Spc72 protein level and/or phosphorylation. Controls should be added to report on Spc72 levels and modification upon Cdc5 overexpression. Moreover, Cdc5 localisation upon overexpression should be analysed to determine whether Cdc5 association to the SPB is also affected.

We have added data in Figure 8—figure supplement 1 showing that Spc72-GFP protein levels were not affected by Cdc5 depletion or CDC5-overexpression. The result is described in subsection “CDC5 overexpression accelerates the Spc72 recruitment to SPB”; “While Cdc5 expression showed no effect on the protein level of Spc72-GFP (Figure 8—figure supplement 1),”

Finally, it remains unclear whether Cdc5 kinase activity or only Cdc5 association to the SPB is the key determinant for Spc72 recruitment. While SPB tethering experiments might help resolve this point, the authors should at least address whether the ability of Cdc5 to recruit Spc72 to the SPB may involve their physical interaction. To this end, co-immunoprecipitation assays or protein pull downs with recombinant baits should be provided.

In order to address this point, we have overexpressed kinase-dead version of CDC5 to see whether physical presence of the protein stimulate the recruitment of Spc72 to SPB. However, this experiment did not bring a conclusion because overexpressed kinase-dead Cdc5 did not localize to SPB at any stage of the cell cycle.

We attempted another experiments to address this issue, as previously discussed in comment 9. However, what determines Spc72 recruitment to SPB is still an open question.

12) Could the authors propose how Spc72 acquisition brings about spindle alignment? At least they should provide time lapse data of cells expressing GFP-Tub1 at sufficient temporal resolution to visualise cMT acquisition and the dynamics of spindle alignment and insertion across the bud neck to judge the involvement of cMTs in response to Spc72 loading.

In order to address this point, we performed timelapse analysis of GFP-TUB1 cells with short intervals. We could show the correlation between cMT acquisition and the spindle orientation. The result is presented in Figure 2—figure supplement 3 and described in the text; “Acquired cMTs efficiently corrected the spindle orientation in pre-anaphase cells, suggesting that the spindle orientation is regulated largely at the level of cMT acquisition (Figure 2—figure supplement 3).”

Time lapse data showing the relative timing between Spc72 loading and the spindle orientation is presented in Figure 4C and Figure 4—figure supplement 3 and the result is discussed in the text; “Within 5 min after appearance of the Spc72-GFP signal, spindle orientation was corrected when it had not done already (Figure 4—figure supplement 3, average 3.50 ± 1.61 min, n=12); therefore, one half part of an anaphase nucleus was successfully inserted into the bud.”

Time lapse with double labelling of Spc72 and MTs was technically not feasible in O. polymorpha. Spc72-RFP signal beached out too quickly for time lapse and mCherry-Tub1 was not efficiently incorporated into microtubule filaments.

13) Figure legends should be edited for clarity and plots should be described in full for what is being scored in every case. Arrows or arrowheads should be added in every image to point at relevant structures that are difficult to discern without this aid as requested above.

Figure legends are modified to include all required information. Relevant structures and fluorescent signals are now marked with arrows/arrowheads.

[Editors' note: further revisions were requested prior to acceptance, as described below.]

Taken together, the manuscript portrays substantial novelty if focused on the behaviour of Spc72, Cdc5 and cMTs. Thus, it would be important to carry out a number of revisions in the text for the narrative in the Results and Discussion to be brought in line with the level of proof. Following is a list of action points and we would like to encourage the authors to revise accordingly so that data are not mis-represented here.

1) Concerning the path for spindle orientation in O. polymorpha

Although nuclear position is not biased toward the bud neck, SPB movement depicted in the time lapse series for SPB localisation is inconsistent with the statement that position is entirely random (e.g. Discussion first paragraph), as it is steadily maintained at the mother cell centre. The temporal resolution used in this study may not reveal the underlying basis for this position but it does not convey proof of random location either. Thus, the text should be amended here to reflect the actual data.

The text is now modified to “the nucleus generally locates centrally”.

Without detracting from the key conclusion on the role of Spc72 recruitment, we note that the representative time lapse series presented as supplemental data to Figure 1 showed the following: a) In Figure 1—figure supplement 4, after assembly 1 µm long spindles remain centrally positioned and loosely oriented toward the bud neck b) According to Figure 1—figure supplement 5, spindle alignment is apparent at onset of spindle elongation followed later by SPB translocation. Instead, in S. cerevisiae, SPB translocation (of spindles held aligned during metaphase) coincides with spindle elongation. The analysis and narration of the current Figure 1 mis-represents this dynamic behaviour. We would recommend moving the time lapse series to be part of the main Figure 1 and, if space is limiting, move supporting quantitation of still images to the supplemental section of Figure 1. Furthermore, the narration should be edited to reflect these data faithfully.

One of time lapse series in the former Figure 1—figure supplement 5 is now moved to Figure 1 as Figure 1E. To accommodate the time lapse in Figure 1, quantification data (the former Figure 1D and 1E) have been moved to a supplement (Figure 1—figure supplement 3).

a) Passage on the still images of SPB marker and the time lapse in Figure 1DE and Figure 1—figure supplement 6 have been modified in the text; “Moreover, SPB in G1 cells as well as small budded cells was not in the defined position within the mother cell body (Figures 1D). Subsequent time lapse analysis revealed that after spindle assembly, ~ 1 µm long spindles remained at their central positions and loosely oriented toward the bud neck until shortly before anaphase onset.”

b) Time lapse images in Figure 1E and Figure 1—figure supplement 6 represent cells with mis-aligned spindle. However, there are cells with loosely positioned/aligned spindle as shown in Figure 5B (wild type category I and II), and it is not possible to narrow down the precise timing of spindle alignment. Therefore, we left the time window for spindle alignment wider than “anaphase onset” as suggested by the reviewer, and added the following passage to the Results; “Spindle alignment was corrected around the time of (or shortly after) spindle elongation, followed by SPB insertion into the bud.”

We have also added the following texts; “after spindle assembly, ~ 1 µm long spindles remain at their central positions and loosely oriented toward the bud neck”, and “Those SPB movements are in contrast to S. cerevisiae in which spindle is aligned during metaphase and therefore SPB translocation into the bud coincides with spindle elongation.”

2) On Dyn1 and Kar9 pathways and SPOC function in O. polymorpha (final paragraph in subsection “Nuclear positioning in O. polymorpha differs from that in S. cerevisiae and other budding yeast species” and second paragraph of the Discussion section) the data are not sufficiently developed to make these claims and should be substantially edited for the following reasons:

– In the new sample image of dyn1∆ cells (Figure 1—figure supplement 6) it is very difficult to tell the nuclear stain from the cell background and hard to see how the quantitation was possible based on such data. Furthermore the stain is inconsistent with the live image of nuclear label by tagging. These data do not support conclusions regarding nuclear size as a differential factor between S. cerevisiae and O. polymorpha. Finally, the defect in the dyn1∆ mutant may transcend cMT function per se in line with the severe growth defect. Note that dyn1∆ mutant cells appeared to re-bud inappropriately and accumulate supernumerary nuclei, arguing that the SPOC may not operate as robustly (if at all) compared to S. cerevisiae, resulting in low viability of dyn1∆ cells. Furthermore, the interaction between kar9∆ and bub2∆ appears much stronger than what is apparent in S. cerevisiae (in which kar9∆ bub2∆ is not lethal at 30°C)

Under the circumstances, the authors are merely inferring cellular defects that they have not proven satisfactorily, for inclusion in this report. The same holds true for their claim that the Kar9 and Dynein pathways are functionally conserved between the two yeasts. cMT images failed to show plus ends reaching the distal bud cell cortex to talk about Kar9-dependent positioning here.

In conclusion, this section should be deleted from a revised manuscript to maintain the focus on the findings linking cMTs and Spc72. Also note that the authors could not answer here a number of essential queries from the original review on technical grounds.

– Discussion paragraph two. Please revise to simply mention synthetic lethality between kar9 and dyn1 mutants, with dyn1 alone showing a severe growth defect at best. Whether this is due to the demands of aligning the spindle during anaphase (as opposed to S. cerevisiae in which spindle alignment is achieved prior to anaphase) remains highly speculative. The severe growth defect due to dyn1∆ in O polymorpha is consistent with weak SPOC-dependent delay. As stated above, the authors report excess of fragmented nuclei and re-budding in dyn1∆ cells, supporting this idea.

According to the reviewer’s suggestion, we have deleted the former Figure 1—figure supplement 6 that showed data on Kar9 and Dynein pathways as well as SPOC from the Result section. The passage on those in the Introduction has been simplified, and the presence of probable orthologs of Kar9 and Dyn1 is now discussed in the Discussion; “Currently molecular mechanism(s) that regulate spindle orientation is unknown. However, although the timing of spindle orientation relative to cell cycle progression appears to be different from that of other yeasts, two redundant molecular mechanisms of spindle orientation, one requiring dynein and the other Kar9, may be conserved in O. polymorpha, because putative orthologs of KAR9 and dynein were identified in O. polymorpha genome sequences (Li et al., 1993; Miller and Rose, 1998; Maekawa and Kaneko, 2014; Nordberg et al., 2014).” The text was also edited; “…which may largely rely on an immediate correction of the orientation of the spindle and on SPOC activity.”

3) The definition of metaphase spindle is not kept consistent throughout. At one point, metaphase is defined by SPB distance < 2 µm but earlier in the text, it was defined as < 1 µm. Please amend accordingly.

We thank the reviewer to point out this inconsistency. It is now amended to be 2 µm.

4) The progressive and gradual loss of Spc72 from SPBs at early stages of the cell cycle is inconsistent with Cdc5 acting as the sole switch in Spc72 recruitment (with reversal triggered by mitotic exit). The authors should make this point clearer when discussing Spc72 detection in G1 cells in Figure 4. Please revise to bring the narration in line with the data.

According to the reviewer’s comment, Spc72 signal in G1 was narrated in the text; “This difference of timing may explain the relatively high and variable intensity of Spc72-GFP at SPB in G1 cells (Figure 4B).”

5) In the phosphorylation analysis, the change of tags on Spc72 should be explained in full and justified.

The tag on Spc72 was changed to flag tag for higher efficiency in immunoprecipitation and having less unspecific signals in immunoblotting. These reasons are now described in the figure legend of Figure 6—figure supplement 3; “The flag tag was used for the efficiency of immunoprecipitation and lower unspecific signals in immunoblotting.”

6) The rational for testing Spc72 upon starvation could be made clearer – mating is triggered by starvation in O polymorpha.

We thank the reviewer to point this out. We have added a passage which explains the rational for testing Spc72 in starved cells; “cMT play important roles in yeast mating and karyogamy, which are initiated in G1. Because mating is triggered by nutrient starvation in O. olymorpha, we examined cMTs and Spc72 in nutrient starved cells.”

7) Discussion paragraph three, on Nud1 being a possible Cdc5 target – please clarify if you meant that Nud1 contains a consensus site for binding the Cdc5 polo box (rather than a polo box consensus sequence?).

We thank the reviewer to point this out. It meant a polo box binding motif. The text has been modified to state clearer; “Nud1, which has one site matching the consensus sequences of polo box binding site”

8) The authors should discuss why Spc72 overexpression is sufficient to force recruitment to SPBs in their model (data in Figure 5).

According to the reviewer’s comment, discussion on how Spc72 was recruitmented to SPB when overexpressed is now added to the Discussion.

9) The Introduction is incorrect regarding Spc72 and cMTs in S. cerevisiae. Simply put, cMTs at the new SPB are formed after a 1 µm long spindle has formed (certainly not G1/S!, Shaw et al., 1997). Please revise to state correctly that the new SPB acquires Spc72 and cMTs after onset of spindle assembly (also see Juanes et al., 2013).

According to the reviewer’s comment, the passage in the Introduction is revised; “New SPB acquires Spc72 and cMTs after the formation of a 1 µm long spindle”

https://doi.org/10.7554/eLife.24340.044

Article and author information

Author details

  1. Hiromi Maekawa

    1. Graduate School of Engineering, Osaka University, Suita, Japan
    2. Faculty of Agriculture, Kyushu University, Fukuoka, Japan
    Contribution
    Conceptualization, Resources, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Writing—original draft, Project administration, Writing—review and editing
    For correspondence
    hmaekawa@agr.kyushu-u.ac.jp
    Competing interests
    No competing interests declared
    ORCID icon 0000-0002-0175-1610
  2. Annett Neuner

    Zentrum für Molekulare Biologie der Universität Heidelberg, DKFZ-ZMBH Alliance, Heidelberg, Germany
    Contribution
    Investigation, Writing—review and editing
    Competing interests
    No competing interests declared
  3. Diana Rüthnick

    Zentrum für Molekulare Biologie der Universität Heidelberg, DKFZ-ZMBH Alliance, Heidelberg, Germany
    Contribution
    Investigation, Writing—review and editing
    Competing interests
    No competing interests declared
  4. Elmar Schiebel

    Zentrum für Molekulare Biologie der Universität Heidelberg, DKFZ-ZMBH Alliance, Heidelberg, Germany
    Contribution
    Supervision, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon 0000-0002-3683-247X
  5. Gislene Pereira

    1. Centre for Organismal Studies, University of Heidelberg, Heidelberg, Germany
    2. Division of Centrosomes and Cilia, German Cancer Research Centre (DKFZ), DKFZ-ZMBH Alliance, Heidelberg, Germany
    Contribution
    Supervision, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon 0000-0002-6519-4737
  6. Yoshinobu Kaneko

    Graduate School of Engineering, Osaka University, Suita, Japan
    Contribution
    Resources, Funding acquisition
    Competing interests
    No competing interests declared
    ORCID icon 0000-0002-7379-9373

Funding

Japan Society for the Promotion of Science (JP24570214)

  • Hiromi Maekawa

Deutsche Forschungsgemeinschaft (Schi 295/4-3)

  • Elmar Schiebel

Deutsche Forschungsgemeinschaft (PE1883)

  • Gislene Pereira

Deutsche Forschungsgemeinschaft (SFB873)

  • Gislene Pereira

Deutsche Forschungsgemeinschaft (SFB1036)

  • Gislene Pereira

Institute for Fermentation, Osaka (the Endowed Chair Program)

  • Yoshinobu Kaneko

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

We thank Dr. K Richter and the Central Unit Electron Microscopy at the German Cancer Research Center (DKFZ) for their kind support for the electron microscopy (EM) analyses. The Nikon Imaging Facility Heidelberg is acknowledged for the SIM analyses. We thank Research Center for Ultra-High Voltage Electron Microscopy at Osaka University for supporting our EM analyses. We thank Japan Collection of Microorganisms (JCM) and Dr T Endo for yeast strains, information, and advice on phylogenetic relationships among yeast species. We thank Dr. T Lin for assistance on bioinformatics. Dr. H Takuma, Dr. Yamaguchi, and Integrated Imaging Research Support (IISR) Japan are acknowledged for technical advice and the primary EM analysis, respectively. HM is grateful to Dr. Takegawa for generously allowing access to facilities in the lab. This work was supported by the Endowed Chair Program of the Institute for Fermentation, Osaka (IFO), Japan (YK) and JSPS KAKENHI Grant Number JP24570214 (HM). GP acknowledges funding from the German Research Council (DFG, PE1883) and DFG collaborative programs SFB873 and SFB1036. ES acknowledges funding from the German Research Council (DFG, Schi 295/4-3).

Reviewing Editor

  1. Anna Akhmanova, Reviewing Editor, Utrecht University, Netherlands

Publication history

  1. Received: December 19, 2016
  2. Accepted: August 17, 2017
  3. Accepted Manuscript published: August 30, 2017 (version 1)
  4. Version of Record published: October 3, 2017 (version 2)

Copyright

© 2017, Maekawa et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 344
    Page views
  • 96
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: PubMed Central, Scopus, Crossref.

Comments

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)