|Home | About | Journals | Submit | Contact Us | Français|
Mitochondrial inheritance, the transfer of mitochondria from mother to daughter cell during cell division, is essential for daughter cell viability. The mitochore, a mitochondrial protein complex containing Mdm10p, Mdm12p and Mmm1p, is required for mitochondrial motility leading to inheritance in budding yeast. We observe a defect in cytokinesis in mitochore mutants and another mutant (mmr1Δ gem1Δ) with impaired mitochondrial inheritance. This defect is not observed in yeast that have no mitochondrial DNA or defects in mitochondrial protein import or assembly of β-barrel proteins in the mitochondrial outer membrane. Deletion of MDM10 inhibits contractile ring closure, but does not inhibit contractile ring assembly, localization of a chromosomal passenger protein to the spindle during early anaphase, spindle alignment, nucleolar segregation or nuclear migration during anaphase. Release of the mitotic exit network (MEN) component, Cdc14p, from the nucleolus during anaphase is delayed in mdm10Δ cells. Finally, hyperactivation of the MEN by deletion of BUB2 restores defects in cytokinesis in mdm10Δ and mmr1Δ gem1Δ cells, and reduces the fidelity of mitochondrial segregation between mother and daughter cells in wild-type and mdm10Δ cells. Our studies identify a novel MEN-linked regulatory system that inhibits cytokinesis in response to defects in mitochondrial inheritance in budding yeast.
Equal segregation of mitochondria between mother and daughter cells during yeast cell division occurs as a result of bidirectional movement of mitochondria to the bud tip and mother cell tip and anchorage of the organelle at those sites (1). The mitochore, a mitochondrial membrane protein complex containing the proteins Mmm1p, Mdm10p and Mdm12p, is required for binding of mitochondria to actin filaments in vitro, actin cable-dependent bidirectional mitochondrial movement, and mitochondrial inheritance (1-3). In early characterizations of mitochondrial morphology and distribution mutants, Sogo and Yaffe (4) noted the presence of a multibudded phenotype in mdm10Δ cells. We find that multibudded clusters consisting of 3-5 buds are present during mid-log phase and accumulate with growth time in mdm10Δ cells. This multibudded phenotype is observed in mdm10Δ cells in three different genetic backgrounds: S288C, W303 and A264A (data not shown).
In wild-type yeast, mitochondria constitute a dynamic and tubular reticulum (Fig. 1A-B) (1). In mdm10Δ cells, mitochondria are large spherical structures that fail to move from mother cells to buds and undergo rapid loss of mitochondrial DNA (mtDNA) (2-3). The large spherical mitochondria typical of mdm10Δ cells are usually present in only one cell within a multibudded clump (Fig. 1E-F). Visualization of DNA confirmed that mdm10Δ cells have no mtDNA and revealed that each cell body in mdm10Δ clumps contains a nucleus (Fig. 1G-H). The viability of wild-type and mdm10Δ cells during mid-log phase growth, assessed using FUN-1 staining, is 93.5% and 76.5%, respectively. Thus, a mutation in MDM10 that results in severe defects in mitochondrial morphology and inheritance also produces defects in mother-daughter cell separation but does not inhibit nuclear inheritance or compromise cell viability.
Deletion of MDM10, MDM12 or MMM1 also results in defects in maintenance of mtDNA, mitochondrial morphology and assembly of β-barrel proteins in the mitochondrial outer membrane (OM) (2, 4-6). Therefore, we tested whether the multibudded phenotype of mdm10Δ cells is due to defects in these mitochondrial inheritance-independent processes by analysis of yeast bearing deletions in mtDNA, MAS37 or TOM7. rho0 cells have no mtDNA and severe defects in mitochondrial respiration (7). Mas37p is a subunit of the SAM/TOB complex, which mediates assembly of β-barrel proteins into the mitochondrial OM (8). Tom7p is a subunit of the protein-translocating pore in the mitochondrial OM (9). Deletion of TOM7 produces defects in mitochondrial morphology that are similar to those observed in mdm10Δ cells as well as defects in mitochondrial protein import (6). Tom7p also promotes the segregation of Mdm10p from the SAM/TOB complex (10).
rho0, mas37Δ, and tom7Δ cells exhibit significantly lower defects in mitochondrial inheritance and lower levels of multibudded cells compared to mitochore mutants (Fig. 1I-J). Thus, the multibudded phenotype observed in mdm10Δ cells is not a consequence of loss of mtDNA, or of defects in mitochondrial respiratory activity, protein import, or OM β-barrel protein assembly. Moreover, we observed a link between the extent of multibudded cells in late-log phase cultures and the severity of the mitochondrial inheritance defect in yeast carrying mutations in mitochore subunits: mdm10Δ = mmm1–1 mdm12Δ (Fig. 1I-J). Mdm12p coordinates mitochondrial inheritance and biogenesis through its direct interactions with the PUF family protein Puf3p (11). Thus, mdm12Δ cells may have less severe multibudded and inheritance phenotypes compared to mdm10Δ or mmm1-1 mutants because Mdm12p has regulatory effects on mitochondrial motility, while Mdm10p and Mmm1p have predominant roles in mediating mitochondrial motility. Overall, the multibudded phenotype observed in all mutants analyzed correlates with defects in mitochondrial inheritance.
mdm10Δ cells that enter the cell cycle cycle are in G2 phase 20 min later than wild-type cells (SFig. 1). Spindle assembly and disassembly as well as the appearance and disappearance of mitotic cyclin are delayed to a similar extent in mdm10Δ compared to wild-type cells (SFig. 2). Formation of the second bud (d2) in multibudded mdm10Δ cells occurs 150 min after release from pheromone-induced G1 arrest, 25 min after the first bud (d1) undergoes Clb2p degradation and spindle disassembly (SFig. 2).
rho0 cells undergo a delay in cell cycle progression similar to that observed in mdm10Δ, the decrease in cell cycle progression in mdm10Δ may be due to loss of mtDNA. However, the multibudded phenotype in mdm10Δ cells is not due to loss of mtDNA (Fig. 1J), or to defects in septation (degradation of the cell wall between mother and daughter cells), spindle alignment or nucleolar segregation (SFig. 3-4). Rather, it is due to defects in contractile ring closure. Actomyosin ring contraction was visualized in wild-type and mdm10Δ cells using a fully-functional fusion protein consisting of the type II myosin (Myo1p) fused to GFP (12), mitochondria-targeted DsRed, and 4-D imaging (time lapse imaging combined with 3-D reconstruction). Deletion of MDM10 has no effect on contractile ring assembly: Myo1p-GFP localizes to a ring at the mother-bud junction in both wild-type and mdm10Δ cells (Fig. 2A-D). Moreover, mdm10Δ cells have the capacity to undergo contractile ring closure (Fig. 2B), and to do so with kinetics (14.2 ± 3.5 min, n = 48) similar to that of wild-type cells (10.4 ± 2.1 min, n = 43). There is some loss of synchrony in mdm10Δ cells at the time of contractile ring closure. Nonetheless, mdm10Δ cells that undergo contractile ring closure do so 20-40 min later in the cell cycle compared to wild-type cells (n = 48).
However, mdm10Δ cells exhibit defects in contractile ring closure, which correlates with defects in mitochondrial inheritance (Fig. 2C). To quantitate the frequency of contractile ring closure, Myo1p-GFP and DsRed-labeled mitochondria were visualized in cells that bore large buds at the onset of imaging for 2 hrs. During this time, contractile ring closure occurred in 100% of the wild-type cells examined (n=19) and in only 29% of the mdm10Δ cell examined (n=38). To evaluate mitochondrial inheritance as a function of contractile ring closure, we measured the mitochondrial content in buds of mdm10Δ cells that undergo contractile ring closure (Fig. 2B) and in the first buds (d1) of multibudded mdm10Δ that failed to undergo contractile ring closure at the mother cell:d1 junction (Fig. 2E). In wild-type and mdm10Δ cells that undergo contractile ring closure 43±2% (n = 32), and 36.7±3.12% (n=37) of mitochondria are in the bud, respectively. In contrast, there are no detectable mitochondria in 87% of d1 cells within multibudded mdm10Δ cells (n = 100).
The MEN regulates cell cycle progression in response to spindle alignment and elongation, and to the transfer of the nucleus from mother to daughter cell during the anaphase-to-telophase transition. Cdc14p activation and localization of the active protein to its sites of action are essential for degradation of a mitotic cyclin (Clb2p), inactivation of a mitotic cyclin-dependent kinase (CDK; Cdc28p/Clb2p), dephosphorylation of CDK substrates, and exit from mitosis (13). However, several studies indicate that the MEN also has a direct role in regulating contractile ring closure during cytokinesis in budding yeast (14-19).
mdm10Δ cells undergoes mitotic exit, as assessed by degradation of Clb2p and spindle disassembly (SFig. 2). To evaluate the role of the MEN in the observed cytokinesis defect, we studied the localization of Cdc14p-GFP in mdm10Δ and wild-type cells. Cdc14p is released from its inhibitor Cfi1p/Net1p in the nucleolus during two stages in the cell division cycle. In early anaphase, separase, as part of the Cdc fourteen early-anaphase release (FEAR) pathway, promotes a transient and partial release of Cdc14p from the nucleolus. In a second phase, signal transduction through the MEN releases the remaining Cdc14p, which facilitates mitotic exit and cytokinesis (20).
We confirmed that Cdc14p-GFP in wild-type cells localizes to the nucleolus through early stages of the cell division cycle, and is released from the nucleolus and localizes to the spindle pole bodies and bud neck as the spindle apparatus elongates (Fig. 3A). When the spindle is at its maximum length (6-8 μm), 100% of the Cdc14p-GFP is released from the nucleolus (Fig. 3C). In mdm10Δ cells, some cytosolic Cdc14p localizes to the spindle pole body in mdm10Δ cells bearing fully elongated spindles. However, release of Cdc14p-GFP from the nucleolus is inhibited by 50% in mdm10Δ cells bearing 4-6 μm spindles, and to a lesser extent in cells with 6-8 μm spindles compared to wild type cells (Fig. 3B-C). Thus, deletion of MDM10 results in a delay in release of Cdc14p from the nucleolus.
Sli15p, a chromosomal passenger protein and substrate for Cdc14p that is released from the nucleolus during early anaphase (21), localizes to the spindle apparatus to the same extent in mdm10Δ and in wild-type cells (SFig. 5). Thus, mislocalization of Cdc14p in mdm10Δ cells is due to an alteration in MEN-mediated control of Cdc14p and not the FEAR pathway. In light of these findings and our observation that release of Cdc14p from the nucleolus is partially inhibited in mdm10Δ cells, it is possible that the level of MEN-mediated Cdc14p activation in mdm10Δ cells is sufficient to support mitotic exit but insufficient to support cytokinesis.
Consistent with this, conditions that hyperactivate the MEN promote cytokinesis in mdm10Δ cells. Deletion of BUB2 suppresses the subtle mitotic exit defect observed in mdm10Δ cells, but has no effect on the time of entry of mdm10Δ cells into anaphase (SFig. 6). Deletion of BUB2 or overexpression of CDC5 in mdm10Δ cells results in a 67% decrease in the number of multibudded cells in late-log phase cell cultures compared to mdm10Δ cells (Fig. 4A-B). Thus, conditions that bypass MEN regulation bypass the cytokinesis defects observed in mdm10Δ cells.
To determine whether other mutations that inhibit mitochondrial inheritance also affect cytokinesis, we studied GEM1, a member of the rho (Miro) family of GTPases and MMR1, a protein that localizes to mitochondria, binds to the type V myosin Myo2p and is required for anchorage of mitochondria in the bud tip (22-23). mmr1Δ or gem1Δ mutants exhibit subtle defects in mitochondrial inheritance, and low but detectable defects in cytokinesis. However, gem1Δ mmr1Δ double mutants exhibit mitochondrial distribution and inheritance defects that are significantly greater than those observed in either single mutant (23) and a cytokinesis defect that is more severe than that observed in either single mutants and similar to that observed in the mdm10Δ mutant. In addition, deletion of BUB2 suppresses the cytokinesis defect observed in the gem1Δ mmr1Δ double mutant (Fig. 4B). These findings provide additional evidence for the existence of a mechanism to inhibit cell cycle progression at cytokinesis when there are severe defects in mitochondrial inheritance.
Finally, the primary function of a checkpoint is to insure that critical cell division processes occur with high fidelity and at the correct time as cells divide. Thus, if the MEN regulates cell cycle progression in response to mitochondrial inheritance, then hyperactivation of the MEN should reduce the fidelity of mitochondrial inheritance. Indeed, we find that conditions that bypass MEN regulation, deletion of BUB2 or overexpression of CDC5, result in defects in partitioning of mitochondria between mother cells and buds (Fig. 5). Deletion of BUB2 reduces the amount of mitochondria in daughter cells. Deletion of MDM10 produces more severe defects in the fidelity of mitochondrial inheritance. Finally, mdm10Δ mutants bearing a deletion in BUB2 or overexpressing CDC5 exhibit defects in mitochondrial partitioning that are more severe than that in mdm10Δ mutants.
Overall, there are numerous cell cycle checkpoints to monitor events associated with nuclear inheritance, including replication of nuclear DNA and segregation of chromosomes and nuclei. Here, we provide evidence for a mitochondrial inheritance checkpoint that inhibits cytokinesis when there are defects in mitochondrial inheritance in budding yeast, and for a role for the MEN in this process. In Drosophila melanogaster, mitochondrial second messengers, either ROS or ATP, can function as two independent signals to enforce checkpoints at G1/S that are not due to metabolic restriction (23). Our findings indicate that a checkpoint for mitochondrial inheritance, that is also independent to metabolic restriction, exist in budding yeast. Finally, since there are mechanisms to insure the inheritance of many organelles and the MEN is a conserved pathway, our findings also raise the possibility that there are similar checkpoints for organelle inheritance in yeast and other cell types.
A summary of the materials and methods used for this study is included. Please refer to Supplemental Information for more detailed description.
Yeast strains used in this work are listed in Table S1. rho0 derivatives were generated from wild-type cells expressing plasmid-borne mitochondria-targeted DsRed (ISY001), as described by Goldring et al. (7). Other yeast methods were performed according to Sherman (24). Yeast cell viability was measured using FUN-1 (25).
The carboxy terminus of Myo1p and Cdc14p were tagged with GFP using PCR-based insertion into the chromosomal copies of the MYO1 or CDC14 loci (26). Table S2 lists primers used to tag these genes. Standard molecular techniques for cloning procedures were used (27).
Mitochondria, tubulin and Sli15p were visualized using plasmid borne GFP fusion proteins. Chitin in bud scars and DNA were visualized using Calcofluor White and DAPI. Acquisition, manipulation and analysis of fluorescence images was carried out as described previously (4).
We are grateful to Drs. A. Amon, J. Shaw, K. Bloom and E. Schiebel of for plasmids; to Drs. J. Aris, A. Amon, T. Davis, D. Kellogg, J. Kitajewski and G. Schatz for antibodies; to the members of the Pon laboratory for support and critical evaluation; to Jessica Lui for data analysis and interpretation, to K. Gordon in the Flow Cytometry Shared Resource of the Herbert Irving Comprehensive Cancer Center (Columbia U.) for assistance with flow cytometry experiments, and to Drs. F. Luca and J. Gautier for enlightening discussions. This work was supported by research grants to L. A. Pon from the NIH (GM45735), and to L.J.G-R. from the Ramón Areces Foundation (Spain).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.