Influence of the bud neck on nuclear envelope fission in Saccharomyces cerevisiae

Abstract

Studies have shown that nuclear envelope fission (karyokinesis) in budding yeast depends on cytokinesis, but not distinguished whether this was a direct requirement, indirect, because of cell cycle arrest, or due to bud neck-localized proteins impacting both processes. To determine the requirements for karyokinesis, we examined mutants conditionally defective for bud emergence and/or nuclear migration. The common mutant phenotype was completion of the nuclear division cycle within the mother cell, but karyokinesis did not occur. In the cdc24 swe1 mutant, at the non-permissive temperature, multiple nuclei accumulated within the unbudded cell, with connected nuclear envelopes. Upon return to the permissive temperature, the cdc24 swe1 mutant initiated bud emergence, but only the nucleus spanning the neck underwent fission suggesting that the bud neck region is important for fission initiation. The neck may be critical for either mechanical reasons, as the contractile ring might facilitate fission, or for regulatory reasons, as the site of a protein network regulating nuclear envelope fission, mitotic exit, and cytokinesis. We also found that 77–85% of pairs of septin mutant nuclei completed nuclear envelope fission. In addition, 27% of myo1Δ mutant nuclei completed karyokinesis. These data suggested that fission is not dependent on mechanical contraction at the bud neck, but was instead controlled by regulatory proteins there.

INTRODUCTION:

In mammalian cells, nuclear envelope breakdown happens early in cell division, well separated from cytokinesis. In budding yeast and some other fungi, however, the nuclear envelope never breaks down, remaining intact throughout all stages of the life cycle including during cell fusion in mating and cell division (Byers 1981a; Wente et al. 1997). Therefore, in these fungi, nuclear envelope fission, also known as karyokinesis, is required for completion of cell division. Karyokinesis is likely to be carefully coordinated with other major cell division events, occurring prior to or concurrent with mitotic exit and cytokinesis. Moreover, karyokinesis may be spatially restricted to insure the even distribution nuclear envelope between the two daughter nuclei.

Unlike the growing body of data on the molecular control of mitotic exit and cytokinesis, the process of karyokinesis in budding yeast is poorly understood. Thought to be coupled to cytokinesis, karyokinesis has been shown to occur within minutes of contraction of the actomyosin ring (Lippincott and Li 2000). However, it was not known what aspects of cytokinesis were critical for karyokinesis. Some studies have documented defects or delays in karyokinesis in certain cytokinesis mutants, such as iqg1^- cells and Δhof1 cells (Lippincott and Li 2000). However, other studies have found, using anti-nuclear envelope protein antibodies in indirect immunofluorescence experiments, that karyokinesis can occur in four different septin mutants also defective in cytokinesis (Copeland and Snyder 1993). It is unclear whether or not the karyokinesis requirement for cytokinesis reflects a cell cycle link, a structural requirement for the bud neck, or a mechanical requirement for actomyosin ring contraction.

The bud neck between the mother and the daughter bud is a special place. In budding yeast, the future site of cytokinesis is determined early in the cell cycle, occurring at the site of bud formation, which becomes the mother-bud junction. The bud neck is required for the localization of many proteins including those regulating cytokinesis like cell cycle proteins, septins and actomyosin ring components. Therefore, the bud neck acts as a critical staging area for proteins contributing to the control of septum formation, plasma membrane separation and cytoplasmic division in Saccharomyces cerevisiae (Balasubramanian et al. 2004; Wolfe and Gould 2005; Wloka and Bi 2012). In addition to having a high concentration of proteins controlling cytokinesis, the bud neck also retains proteins directing the morphogenesis checkpoint and late mitotic events (Cid et al. 2002). The close spatial and temporal concurrence between karyokinesis and cytokinesis has suggested that they might be mechanistically coupled. However, the nature of the relationship between karyokinesis and cytokinesis has not yet been determined. It is possible that closure of the actin-myosin ring provides a mechanical role for cytokinesis in nuclear envelope separation, or cell signaling factors localized to the bud neck may serve as regulatory factors for karyokinesis. Moreover, it is possible that nuclear envelope fission may be directly regulated by the cell division cycle.

Migration of the nucleus through the bud neck is a critical event for completion of a normal cell cycle in budding yeast. Nuclear positioning and migration is dependent on astral microtubules, motor proteins such as Dhc1p/Dyn1p, and other proteins such as Kar9p (Palmer et al. 1992; Eshel et al. 1993; Li et al. 1993; Yeh et al. 1995; Carminati and Stearns 1997; Shaw et al. 1997b; Miller et al. 1998; Miller and Rose 1998; Beach et al. 2000; Yeh et al. 2000; Yamamoto and Hiraoka 2003). For a review, see (Kusch et al. 2003; Pringle 2006). Nuclear migration mutants provide a useful tool to examine nuclear envelope fission due to binucleate and multinucleate cells generated by impaired nuclear positioning.

Binucleate and/or multinucleate cells are not only found in nuclear migration mutants. Mutations affecting the morphogenesis checkpoint also generate multinucleate cells. Blocking bud formation and simultaneously preventing activation of the morphogenesis checkpoint, as occurs in the cdc24–1 swe1Δ mutant, generates cells with active nuclear division, but without bud formation, at non-permissive temperature (Sia et al. 1996). Cdc24p, a guanine nucleotide exchange factor, is critical for the establishment of polarized growth necessary for bud formation, and the cdc24–1 temperature sensitive mutant is defective in bud formation and nuclear division at non-permissive temperature (Adams et al. 1990). Swe1p, a homolog of Wee1, is a part of the morphogenesis checkpoint (Sia et al. 1996; Lew 2003; Keaton and Lew 2006; Pringle 2006; Howell and Lew 2012). Deletion of SWE1 results in cells that are defective in delaying the cell cycle in response to defects in bud formation or the actin cytoskeletal network (Sia et al. 1996).

The bud neck is marked by the presence of the septin scaffolding proteins. The septins were discovered as temperature-sensitive mutations in four genes (cdc3, 10, 11, and 12) that cause defects in cytokinesis, forming long chains of cells (Hartwell 1971; Byers and Goetsch 1976). Of the seven known budding yeast septins, five are expressed during vegetative growth (Cdc3p, Cdc10p, Cdc11p, Cdc12p, Shs1/Sep7p). The mitotic septins form a ring-like structure at the base of the bud neck, and act as a platform for proteins regulating cytokinesis and mitotic exit, as well as functioning as a diffusion barrier (Gladfelter et al. 2001; Versele and Thorner 2005; Pringle 2006)]. The septins colocalize at the future bud site from before the emergence of the bud until the completion of cytokinesis (Haarer and Pringle 1987; Ford and Pringle 1991; Kim et al. 1991). SHS1/SEP7 is not required for septin ring formation in wild type cells, although cytokinesis is impaired without it (Mino et al. 1998; Iwase et al. 2007). Septins also act as a diffusion barrier between the mother and daughter cell (Barral et al. 2000; Takizawa et al. 2000). Septins have been shown to help establish boundaries for key proteins required for cytokinesis and to restrict protein movement within the plasma membrane and endoplasmic reticulum between the mother and daughter (Dobbelaere and Barral 2004; Luedeke et al. 2005; Caudron and Barral 2009; McMurray and Thorner 2009b). Many cell cycle and cytokinesis regulatory proteins are localized to the septin rings, including mitotic kinases like Cdc5p, Cdc15p, and Dbf2p, that may have an overlapping role in controlling cytokinesis (Cid et al. 2002; McMurray and Thorner 2009b).

In this work, we used nuclear migration mutants, bud formation mutants and septin mutants to dissect out what parts of cytokinesis were necessary for karyokinesis. We also studied what role the bud neck played, mechanical or regulatory, in karyokinesis.

MATERIALS AND METHODS:

Strains and yeast methods

Budding yeast strains used in this study can be found in Table 1. Strains were grown according to standard methods, using YPD media for strains with no markers or integrated markers, and minimal media supplemented with the necessary amino acids for strains carrying Sec63-GFP, a NE-ER marker, also known as pMR5029 (Melloy et al. 2007)(Melloy et al. 2009). The nuclear migration mutants, tub2–401, kar9Δ and dhc1Δ, carrying pMR5029, were grown overnight at room temperature and then incubated at 18–20° C for approximately four hours before observation to exacerbate the nuclear migration defect in these cells. The mutants were also studied at 25° C as a control. The wild type control (MS7617) with pMR5029 was also observed at 18° C and 25° C. The cdc24–1swe1Δ mutant, carrying pMR5029, was grown overnight at room temperature, then shifted to 37° C for 4–6 hours before observation. Wild type controls grown at either 25° C or 37° C were also studied. For shift down experiments, cdc24–1swe1Δ mutant cells incubated for 4–6 hours at 37 C were then shifted to room temperature for 1–1.5 hours before time-lapse microscopy was conducted. Time points were taken every 7 minutes.

Table 1:

Strains and Plasmids

Strain	Genotype	Source or Reference
MS22	MATa, trp1-Δ1, lys2–801, ade2–101, canS	Rose lab
MS7617	Same as MS22 except contains pMR5029	Rose lab
MS52	MATα, leu2–3, 112, trp1-Δ1, ura3–52	Rose lab
MS7619	Same as above except contains pMR5029	Rose lab
MY 3044 (CuY67–2)	MATa, ura3–52, lys2–801, his4–539, tub2–401	T. Huffaker
MY8488	MATa, ura3–52, lys2–801, his4–539, tub2–401, TRP1::URA3, contains pMR5029	Rose lab
MS4262	MATα, DHC1::URA3, leu2–3, 112, trp1-Δ1, ura3–52	Rose lab
MS7683	Same as above except contains pMR5029	Rose lab
MS4313	MATa, KAR9-Δ1::LEU2, leu2–3, 112, ura3–52, trp1-Δ1	Rose lab
MS8420	Same as above except contains pMR5029	Rose lab
MY8569 (DLY690)	MATa, ura3Δns, leu2–3, 112, trp1-Δ1, his2, bar1, cdc24–1, SWE1::LEU2	D. Lew
MY8644	MATa, ura3Δns, leu2–3, 112, trp1-Δ1, his2, bar1, cdc24–1, SWE1::LEU2 contains pMR5029	Rose lab
MY8566 (DLY1)	MATa, ura3Δns, leu2–3, 112, trp1-Δ1, his2, bar1	D. Lew
MY11675	Same as above except contains pMR5029	Rose lab
MY9161 (YEF743)	MATa,cdc12–6, leu2, ura3	E. Bi
MY9179	MATa,cdc12–6, leu2, ura3, TRP1::LEU2 contains pMR5029	Rose lab
MY9162 (YEF739)	MATa, cdc3–3, leu2, ura3	E. Bi
MY9175	MATa, cdc3–3, leu2, ura3, TRP1::URA3 contains pMR5029	Rose lab
MY9491	MATa, cdc3–3, leu2, ura3, TRP1::URA3, SHS-GFP-URA3	Rose lab
MY9492	MATa, cdc12–6, leu2, ura3, TRP1::LEU2, SHS1-GFP-LEU2	Rose lab
YEF473a	MATa, his3, leu2, lys2, trp1, ura3	E. Bi
yTM 231	Same as above except contains pMR5029	this work
YEF1804	MATa, his3, leu2, lys2, trp1, ura3, myo1Δ::KanMX6	E. Bi
yTM 232	Same as above except contains pMR5029	this work
Plasmid	Relevant Markers
pMR5029	pER-GFP marker, Sec63-GFP in pRS414 (TRP1)	Rose lab
pMR5487 (pM-6)	Integrating plasmid with SHS1-GFP (LEU2)	Iwase and Toh-e
pMR5488 (pM-7)	Integrating plasmid with SHS1-GFP (URA3)	Iwase and Toh-e
pMR5508	GFP-Tub1 in pRS305	Rose lab

For the cdc3–3 and cdc12–6 septin mutants, carrying pMR5029, cultures were grown overnight at room temperature and then shifted to 37° C for approximately 4 hours before being imaged at room temperature. Time points were taken every 7 minutes. For longer time-lapse experiments, when pER-GFP was observed in combination with Tub1-GFP (pMR5508), time points were taken every 20 minutes and observed during the temperature shift to 37° C. Cells were maintained at 37° C using the DeltaT microscope stage heater and objective heater (Bioptechs, Butler, PA). Myo1Δ mutant cells were observed at room temperature after an overnight incubation. The wild type control YEF473A was used in karyokinesis counts for the septin and myo1Δ mutant strains.

To observe nuclei, cells were stained with Hoechst 33342 (H3570; Invitrogen, Carlsbad, CA) for 15 minutes before microscopic observation. Some DNA staining was also done using DAPI (32670; Sigma-Aldrich, St. Louis, MO).

Fluorescence Microscopy

Budding yeast cells were imaged using either the Delta Vision Fluorescence Microscopy System (Applied Precision, Issaquah, WA) fitted with a Cool SNAP HQ CCD camera (Photometrics, Tucson, AZ) or a Leica Microsystems DM5500A microscope (Bannockburn, IL) linked to an ORCA-ER CCD camera (Hamamatsu, Japan).

For both systems, a 100X oil immersion objective was used. At various time points, images were collected using a GFP (FITC) filter set and a DNA (DAPI/A4) filter set. Reference images were also taken using DIC. For z stacks, images were captured at 0.5 micrometer intervals over 5 micrometers. Images captured on the Delta Vision system were analyzed and deconvolved using softWoRx (Applied Precision), while the images captured using the Leica Microsystems apparatus were analyzed using Leica Application Suite, Advanced Fluorescence software.

RESULTS:

In the first part of our study, we wanted to examine nuclear envelope fission in the context of nuclear migration mutants in which division of the nuclear contents occurs in the mother cell, due to the migration failure. We used several nuclear migration mutants, including kar9Δ, dhc1Δ, and tub2–401, by transforming them with a nuclear envelope-endoplasmic reticulum marker (Melloy et al. 2007). All of the mutants formed binucleate or multinucleate cells under non-permissive temperature conditions (Sullivan and Huffaker 1992; Miller and Rose 1998).

All three mutants were grown at permissive temperature overnight, and then shifted to non-permissive temperature to enrich for the population of binucleate cells arising from nuclear migration failure. The cells contained a GFP-tagged nuclear envelope-ER marker (green in images) and the cells were stained with a DNA dye (blue). In the tub2–401, kar9Δ, and dhc1Δ mutants, binucleate cells were detected by the DNA stain, but the nuclear envelopes were still connected, as seen with the GFP marker (indicated by arrows in Figure 1). We counted cells of each genotype, including wild type, at both permissive and non-permissive temperature. In the cells observed with two or more nuclei in the mother cell, 88–100% of the cells had connected nuclear envelopes (Table 2). In the remaining cells, the orientation of the cells in the image prevented a conclusive statement on any connection between the nuclei. These results suggested that karyokinesis required either cell cycle progression, cleavage by the cytokinetic furrow, or the regulatory influence of the bud neck.

Figure 1: Karyokinesis in defective in nuclear migration mutants.

Shown are three different nuclear migration mutants: dhc1Δ (a), kar9Δ (b), and tub2–401 (c) with more than one nucleus in the mother cell. All three have connected nuclear envelopes as seen using a fluorescent marker for the nuclear envelope, pMR5029 (indicated by arrows). Green, Sec63-GFP, Blue, DNA

Table 2:

Nuclear envelope fission defects in nuclear migration mutants

Genotype	Temperature (° C)	≥ 2 nuclei in unbudded to medium budded cell connected inconclusive*		Total # of binucleate cells observed	Total # of cells observed
tub2–401	18–20	16 (94%)	1	17	265
tub2–401	25	3 (100%)	0	3	53
dhc1Δ	18–20	68 (88%)	9	77	933
dhc1Δ	25	2 (100%)	0	2	34
kar9Δ	18–20	10 (91%)	1	11	2202
WT	18	1 (100%)	0	1	122
WT	23	1 (100%)	0	1	128

^*may still be connected but unclear from Z stack

To distinguish among the possible mechanisms for karyokinesis, we examined nuclear envelope fission in a mutant defective in both bud emergence and the morphogenesis checkpoint, cdc24–1 swe1Δ (Sia et al. 1996). At the non-permissive temperature, these mutants cannot form a bud because of a lack of functional Cdc24p. However, the nuclear cycle continues in the mother cell because of the lack of Swe1p, generating multinucleate cells after several hours at non-permissive temperature. The bud emergence defect is reversible; after return to lower temperatures, the cdc24–1 swe1Δ mutant begins budding (Sia et al. 1996). By studying the double mutant at non-permissive temperature and upon shifting down to permissive temperature, cdc24–1 swe1Δ, one can follow the nuclear cycle and nuclear envelope fission independent of any cell cycle delay triggered through the morphogenesis checkpoint.

As shown in Figure 2A and 2B, binucleate and multinucleate cells formed in the cdc24–1 swe1Δ mutant at the non-permissive temperature. Among the cells examined, 98% of the cells with two or more nuclei had connected nuclear envelopes; no wild type cells were found at a comparable stage (Table 3). We then asked whether nuclear fission occurred after the cells were shifted back to permissive temperature, to allow cell cycle and bud emergence to occur. If cell cycle progression alone regulates karyokinesis, then it should occur for all nuclei regardless of their cellular position. Alternatively, if the bud neck is required then, karyokinesis should only occur on nuclei that pass through the neck.

Figure 2: Karyokinesis failure in cdc24–1swe1Δ mutants at non-permissive temperature is reversed upon return to permissive temperature after passage of a nucleus through the bud neck.

Shown in (A) and (B) are two examples of multinucleate cdc24–1swe1Δ cells at non-permissive temperature for 4–6 hours displaying connected nuclear envelopes (arrows). After the same incubation period at non-permissive temperature, shift down experiments were performed to watch the nuclear envelope during bud emergence. Shown in (C) is a time-lapse experiment (each time point is 7 minutes apart) following the nuclear envelope using pMR5029. Note that the nucleus that passes through the bud neck undergoes fission (arrows), while the nuclei still in the mother cell do not undergo fission (black and white arrowheads). Green, Sec63-GFP, Blue, DNA

Table 3:

Nuclear envelope fission defects in cdc24swe1 mutants

Genotype	Temperature (° C)	≥ 2 nuclei in unbudded to medium budded cell connected inconclusive*		Total # of binucleate cells observed	Total # of cells observed
cdc24swe1 Sec63-GFP	4–6 hrs, 37	28 (100%)	0	28	60
cdc24swe1 Sec63-GFP with Hoechst staining	4–6 hrs, 37	16 (94%)	1	17	94
Wild type (DLY1) Sec63-GFP	5 hrs, 37	0 (0%)	0	0	22

^*May still be connected but unclear from Z stack or static images

By time-lapse microscopy (n= 4 movies), we observed that the nucleus that passed through the bud neck underwent fission, while the nuclei remaining in the cell body did not (Supplemental literature). A series of eight images in one of the time-lapse analyses are shown in Figure 2C. The position of the nuclear envelope as it passes through the bud neck is noted with an arrow. After several time points, one can see the occurrence of nuclear envelope fission in the part of the nuclear envelope present in the bud neck. However, for the other nuclei retained in the cell body karyokinesis had not occurred (arrows in the bottom half of the images). These results suggested that cell cycle progression is not sufficient for karyokinesis, but requires the presence of the bud neck. In principal, the bud neck may act either mechanically to facilitate karyokinesis, or proteins localized to the bud neck may regulate fission. When the bud neck is formed, once the nuclear envelope passed through the neck, it can undergo karyokinesis.

Upon identifying the bud neck region as important for nuclear envelope fission, we next tested whether nuclear envelope fission can occur in mutants that are defective for constructing the septin scaffold necessary for protein localization required for mitotic exit and cytokinesis. Passage through the bud neck may be needed in a mechanical or regulatory way. Disrupting the septin scaffold may reveal if there is a mechanical requirement for bud neck constriction to facilitate karyokinesis. However, if karyokinesis proceeds in the septin mutant, then this would indicate that regulatory proteins are still present and are sufficient for karyokinesis to occur.

To test the requirement for the bud neck, we used two temperature-sensitive septin mutants, cdc12–6 and cdc3–3. Using the GFP nuclear envelope marker and DNA stain, we observed distinct nuclei surrounded by their own nuclear envelopes, even at the non-permissive temperature where cytokinesis failed. Time-lapse analysis of nuclear envelope fission in a cdc12–6 mutant is shown in Figure 3A. Karyokinesis occurred between time-points 2 and 3, as evidenced by the break in the nuclear envelope. We also studied karyokinesis using Tub1-GFP as a marker for the mitotic spindle in longer time-lapse studies (Figure 3B). We observed the formation and elongation of the mitotic spindle within the nuclear envelope, followed by karyokinesis. Both telophase and karyokinesis occurred despite the presence of a functional bud neck. Examination of static images of live cells containing the pER-GFP marker, showed that at the non-permissive temperature, 77–85% of pairs of nuclei underwent nuclear envelope fission in septin mutants (Table 4). After longer incubations (e.g. 6 hours) at the non-permissive temperature, nuclear fission rose to 100% in the septin mutants (data not shown). These results demonstrated that the structure of the bud neck, per se, is not required for nuclear fission.

Figure 3: Karyokinesis occurs in septin mutants failing to undergo cytokinesis.

Shown in (A) is a time-lapse experiment using cdc12–6 carrying the pMR5029 marker. After four hours at non-permissive temperature, long chains of cells have formed with multiple nuclei. Time points were taken every 7 minutes. One can see that the nucleus passing through the bud neck undergoes karyokinesis during the time-lapse (indicated by arrow). Green, Sec63-GFP, Blue, DNA. Shown in (B) is the same strain but also carrying the Tub1-GFP marker. Time points were taken every 20 minutes. One can see spindle elongation happening between time points 3 and 4 as indicated by the arrows, and a karyokinesis event happening by time point 5. In addition, one can also see spindle elongation happening between time points 9 and 10 as indicated by arrows, followed by a karyokinesis event by time point 11. In (C) are representative images of the myo1Δ strain carrying the pMR5029 marker. Arrows indicate the nuclear envelope. Note one cell has completed karyokinesis in the top panel. GFP (left panel), DNA (middle panel), merge (right panel). Analysis of the distance between nuclei in relation to fission or no fission in the mutants is shown in (D).

Table 4:

Nuclear envelope fission in septin and myo1Δ mutants

Genotype	Temperature (° C)	≥ 2 nuclei * connected not connected		Number of nuclei pairs
cdc3–3	25	25 (47%)	28 (53%)	53
cdc12–6	25	17 (52%)	16 (48%)	33
WT	25	19 (46%)	22 (54%)	41
cdc3–3	4hrs, 37	27 (15%)	148 (85%)	175
cdc12–6	4hrs, 37	50 (23%)	170 (77%)	220
WT	4hrs, 37	11 (44%)	14 (56%)	25
myo1Δ	25	37 (73%)	14 (27%)	51

^*In wildtype or mutants at room temperature, cells counted were medium to large-budded cells that had completed anaphase with typically one pair of nuclei. In the septin mutants at non-permissive temperature, more than one pair of nuclei might be detected in the mutant cell body.

Since recent studies indicated that a partially functional actomyosin ring may be still present in a septin mutant, we next observed karyokinesis in a myo1Δ mutant lacking the budding yeast Type II myosin (Bi et al. 1998; Feng et al. 2015). Karyokinesis counts were taken on four different occasions. Individual cells were difficult to observe because mutant cells in this strain background tended to clump together as noted previously (Bi et al. 1998) Cells were not sonicated to separate so as not to disturb the karyokinesis observation. However, karyokinesis was observed in at least 27% of the medium to large-budded cells (Table 4).

One possibility for the observation that karyokinesis occurs in the septin mutants, but not in nuclear migration or bud emergence mutants, is that there is not sufficient space to fully elongate the mitotic spindle within a single cell. Karyokinesis might then occur in septin mutants because they form long chains of cells with sufficient space for full spindle elongation. Nuclear fission might then result from the rupture of hyper-elongated nuclei. If so, then we would expect that the occurrence of karyokinesis in the septin mutants would be correlated with nuclear length. We saw no major difference in the distance between nuclei for nuclei that had undergone fission and those that had not (Figure 3D). For example, the distance between septin mutant nuclei that had not undergone fission had a range of 5–11 micrometers. The distance between septin mutant nuclei that had undergone fission had a range of 5–15 micrometers, with over 75% of the nuclei falling in the 7–11 micrometer range (Figure 3D). In addition, our time-lapse analysis using Tub1-GFP in the septin mutants indicated that spindle elongation can still occur in these cells.

Karyokinesis occurring in the septin mutants suggests that the bud neck is not mechanically required for nuclear fission. Presumably proteins normally required at the bud neck may be present but dispersed in the cells. We therefore examined the integrity of the septin ring in the cdc12–6 and cdc3–3 mutants using a Shs1/SEP7-GFP marker (Iwase et al. 2007). Our results indicated that the septin ring was indeed dispersed, as evidenced by a diffuse Shs-GFP signal (arrows in Figure 4A). We studied the Shs1/Sep7-GFP signal in wild type, cdc3–3 mutant, and cdc12–6 mutant cells and saw abnormal Shs1/Sep7-GFP localization along the cortex of the cell or at the end of a cell projection in both septin mutants, but not in the wild type cells (Figure 4B). This supported previous work indicating that the septin ring is not intact in cdc12 mutants (Kim et al. 1991; Barral et al. 2000; Dobbelaere et al. 2003). These data indicated that without the diffusion barrier or protein scaffold of the septins present, nuclear envelope fission could still be successfully triggered by proteins presented in the cell, but not concentrated at the bud neck by an intact septin ring. Also, the bud neck does not fulfill a mechanical requirement for karyokinesis. Cell cycle progression involving division of the nuclear contents is not sufficient to trigger karyokinesis, and failure of cytokinesis is not enough to stop karyokinesis.

Figure 4: No intact septin ring is detected in temperature-sensitive septin mutants.

Shown in (A) are cdc12–6 cells at non-permissive temperature expressing a SHS-GFP marker. In (B), the SHS-GFP localization is quantified, with the majority of cdc12–6 mutants either displaying dispersed SHS-GFP signal or an end cap-like structure. (indicated by arrows).

DISCUSSION:

Karyokinesis can be uncoupled from cytokinesis

Ordinarily, nuclear envelope fission is connected to passage through the bud neck. Using more than one way to generate binucleate/multinucleate cells, we detected connected nuclear envelopes, suggesting a defect common to any type of budding yeast cell mutant lacking a bud neck. However, septin mutants, with a failure of cytokinesis, still undergo nuclear envelope fission most of the time. In our septin mutant studies, we used live cell imaging monitored over longer periods of time, in contrast to previous studies noting a lack of karyokinesis in a higher percentage of fixed septin mutant cells (Copeland and Snyder 1993). In addition, even in a myo1Δ mutant, karyokinesis can still occur.

Previous studies indicated a temporal relationship between karyokinesis and cytokinesis (Lippincott and Li 2000). Our studies indicate that although the timing of the two processes is very close, cytokinesis may fail and yet karyokinesis goes forward in the right regulatory environment. Many proteins can be found at the bud neck, and could potentially regulate karyokinesis independently of cytokinesis. In the case of a septin mutant where bud neck proteins are dispersed, karyokinesis can still occur without a functional bud neck. In contrast, in elongated cell types such as in pseudohyphal growth, the nucleus still divides at the bud neck (Kim and Rose 2015).

Candidate regulatory proteins localized to the bud neck region by septins that may control karyokinesis

Many classes of regulatory proteins may play a role in nuclear envelope fission. Some candidate proteins are cell cycle checkpoint proteins already known to control at least one event near the end of cell division, including cytokinesis or mitotic exit. One class includes the mitotic exit network proteins. The mitotic exit network has an overlapping role in both mitotic exit and cytokinesis. The MEN proteins themselves localize to the bud neck via the septins near the end of cell division (Jimenez et al. 1998; Lippincott et al. 2001; Luca et al. 2001; Menssen et al. 2001; Song and Lee 2001; Surana et al. 2002; Seshan and Amon 2004). In the basidiomycete, Ustilago maydis, which undergoes an “open” mitosis unlike budding yeast, a Tem1p-like protein is involved in nuclear envelope breakdown during mitosis (Straube et al. 2005). In addition, the anaphase-promoting complex (APC/C) is thought to play a role in a “semi-open” mitosis in Schizosaccharomyces japonicus by marking Oar2 for destruction. Oar2 helps regulate the production of new membrane phospholipids (Aoki et al. 2013). Ibd1p (also known as Bfa1p) has been shown to promote binucleate and anucleate cell formation when overexpressed (Lee et al. 1999). Bfa1p is involved in the spindle checkpoint pathway and may work in conjunction with the checkpoint protein Bub2p (Li 1999). Bub2p may also be important in regulating nuclear envelope fission, given its many roles in regulating the end of the cell cycle. Overexpression of Bub2p has also been linked to cytokinesis defects (Park et al. 2009).

A second class of candidate proteins includes the septin-dependent kinases such as Hsl1p and Gin4p. Defects in ER compartmentalization in an hsl1Δgin4Δ double mutant suggest that the septin-dependent kinases play a role in directing other membrane movements just prior to cytokinesis (Luedeke et al. 2005). Select septin mutants (cdc10Δ and sep7/shs1Δ) display perturbed mitotic exit and mislocalization of Lte1p (Hofken and Schiebel 2002; Seshan et al. 2002; Castillon et al. 2003). Recent studies also suggest that Hsl7 and the Elm1 kinase may actually sense bud emergence and interact with septins differently after bud emergence (Kang et al. 2016). However, the function of these kinases would presumably be impaired in the septin mutants, where karyokinesis still occurs.

Mechanics of karyokinesis

The mechanism of karyokinesis remains unclear. Proteins may act to catalyze fission by effecting membrane curvature, either through generating mechanical force, participating in a protein scaffold assembly that could force curvature of the outer and inner envelopes, or by mediating insertion of additional phospholipids in the membrane, or some combination of these functions (Shibata et al. 2009). Septin filaments have been shown to affect membrane curvature in liposomes (Tanaka-Takiguchi et al. 2009). Managing membrane dynamics as the nuclear envelope goes through the major shape changes of a “closed” mitosis might involve the conserved proteins participating in homotypic ER membrane fusion such as Sey1p, Dsl1p, Lnp1, and reticulons (Chen et al. 2013b; Rogers et al. 2014).