In order for haploid gametes to fuse during fertilization, microtubules (MTs) must generate forces that are sufficient to move the nuclei together. Nuclear movements during fertilization rely on microtubule-associated proteins (MAPs), many of which have been characterized extensively during mitosis. A useful model system to study MT-dependent forces before nuclear fusion, or karyogamy, is the mating pathway of budding yeast. Dynamic MTs are guided to the mating projection (shmoo tip) when plus-end-binding proteins interact with polarized actin microfilaments. If two shmoo tips are in proximity they may fuse, dissolving the MT-cortical interactions. Subsequently, oppositely oriented MT plus ends interact and draw the nuclei together. The plus-end-binding proteins in the yeast mating pathway are conserved in metazoan cells and may play a role in higher eukaryotic fertilizaton. Thus, understanding the mechanism of plus end orientation and karyogamy in budding yeast will reveal mechanisms of MT-dependent force generation conserved throughout evolution.
Eukaryotic fertilization requires microtubule (MT)-dependent nuclear movements for gamete fusion. MTs are polarized polymers nucleated at centrosomes or spindle pole bodies (SPBs) that are composed primarily of α- and β-tubulin dimers. MTs grow and shorten through rapid tubulin addition and subtraction at their plus ends, a process known as dynamic instability (Inoue and Salmon, 1995; Kirschner and Mitchison, 1986). Dynamic instability is not inhibited when MT plus ends interact with attachment sites or the plus ends of oppositely oriented MTs. Consequently, MTs can facilitate the movement of subcellular structures to which they attach. MT plus ends are linked to cortical sites or oppositely oriented MTs by plus-end-binding proteins, a subset of which are plus-end-tracking proteins (+TIPs) that preferentially accumulate on growing MTs (Akhmanova and Hoogenraad, 2005; Schuyler and Pellman, 2001). When a MT encounters an attachment site, the plus-end-binding proteins act as an adaptor to tether the dynamic plus end to the attachment site. After the MTs attach, growth and shrinkage continues, and because MTs are anchored at microtubule-organizing centers, the attached structures can oscillate, a behavior seen during chromosome directional instability in mitosis (Skibbens et al., 1993). Alternatively, directed movement may dominate as a consequence of persistent MT polymerization or depolymerization, such as during anaphase chromosome movement. A model system that allows us to examine MT attachment, MT-dependent oscillations, and MT-MT interactions is the mating pathway of Saccharomyces cerevisiae. In this pathway, nuclei are oriented towards the site of cell fusion by MTs and attach to the cell cortex. After the two cells fuse, the nuclei are then drawn together by MT-MT interactions. Thus, information about MT dynamics revealed by studies of the mating pathway is relevant to processes ranging from chromosome segregation to metazoan fertilization that are critical for cellular survival.
MT dynamics during S. cerevisiae mating
When budding yeast cells of opposite mating type are mixed, a signal transduction cascade is initiated to induce the mating pathway (Fig. 1A) (Bardwell, 2005). The mating response includes the polarization of the unbudded cell and formation of a mating projection known as the shmoo (MacKay and Manney, 1974; Tkacz and MacKay, 1979). Cells of opposite mating types fuse at the shmoo tips (Rose, 1996).
MT dynamics in the mating pathway were first examined in live cells by imaging of dynein or tubulin fused to GFP (Maddox et al., 1999). This revealed that MTs nucleated from the SPB are guided to the shmoo tip by the plus end, and this orients the nucleus (Maddox et al., 1999). The plus ends become oriented towards and attach persistently to the shmoo tip while growing and shrinking (Hasek et al., 1987; Maddox et al., 1999). The coordinated growth of MTs pushes the nucleus away from the shmoo tip, whereas shortening of MTs at the plus end draws the nucleus towards the shmoo tip, generating nuclear oscillations over time. Unlike higher eukaryotes, in which tubulin dimers are lost at the minus end, in budding yeast MT dynamics are regulated by proteins exclusively at the plus end (Maddox et al., 1999; Maddox et al., 2000; Tanaka et al., 2005).
After cell fusion, MTs that were originally attached to the shmoo tip interact with MTs from the other cell to form a bridge between the two nuclei (Hasek et al., 1987; Read et al., 1992). The cross-linked MTs then coordinately depolymerize, drawing the nuclei together for karyogamy (Maddox et al., 1999; Molk et al., 2006; Rose, 1996). Inhibitor studies and genetic analysis of tubulin mutants have demonstrated that MTs are required for karyogamy (Delgado and Conde, 1984; Huffaker et al., 1988). [Class I karyogamy mutants exhibit defective nuclear congression whereas class II karyogamy mutants do not undergo nuclear fusion (Kurihara et al., 1994).] Genetic analyses have similarly demonstrated that plus-end-binding proteins are required for nuclear congression and for karyogamy (Berlin et al., 1990; Meluh and Rose, 1990; Schwartz et al., 1997) (Fig. 1B).
MT dynamics in the mating pathway are thus crucial for orientation of the nucleus to the shmoo tip and nuclear congression. They are also required at mitosis in the first zygotic division. Here, we review what is known about these processes and highlight some outstanding questions.
Formation of the MT-shmoo tip attachment
There are three steps in the formation of the MT-shmoo tip attachment (Fig. 2 and Table 1): (1) polarization of the actin cytoskeleton and formation of the shmoo tip; (2) nuclear orientation - guidance of MT plus ends toward the shmoo tip along actin cables; (3) attachment of MT plus ends to the shmoo tip cell cortex.
Cell polarization is initiated when pheromones bind to cell surface receptors that are coupled to a heterotrimeric G-protein signaling cascade (reviewed by Bardwell, 2005). Disassociation of the Gα subunit, Gpa1p, from the Gβγ subunits activates Cdc42p, a key component of the cell polarization pathway, and induces the formation of the polarisome, a protein complex that includes Pea2p, Spa2p, Bud6p, and Bni1p (Bardwell, 2005; Etienne-Manneville, 2004; Pruyne and Bretscher, 2000). Gpa1p is required for localization of Bni1p to the future shmoo site, and this event is coupled to Cdc42p activation to polarize the cell (Matheos et al., 2004). As the cell polarizes, new cell-surface materials are synthesized to form the shmoo (Tkacz and MacKay, 1979). After shmoos form, cells polarize towards and mate with the partner expressing the highest pheromone concentration - this process is termed courtship (Jackson and Hartwell, 1990). If cells are treated with synthetic mating pheromone, a default mating pathway is initiated and the shmoo forms next to the bud scar from previous divisions (Dorer et al., 1995). In both cases, the major cytoskeletal component required for cell polarization is actin (Hasek et al., 1987; Read et al., 1992).
After the cell polarizes, MT plus ends are transported along actin cables to the shmoo tip; because the minus ends are attached to the SPB, the result is orientation of the nucleus to the shmoo tip (Fig. 2B). Nuclear orientation requires the +TIP Bim1p and Kar9p. Bim1p, the budding yeast EB1 (end binding protein) ortholog, binds to MTs and preferentially localizes to growing plus ends (Maddox et al., 2003; Tirnauer et al., 1999). Bim1p interacts with Kar9p to link MTs to the actin network. Kar9p is transported to MT plus ends and serves as an adaptor protein between Bim1p and the actin-associated type V myosin motor Myo2p (Hwang et al., 2003; Korinek et al., 2000; Lee et al., 2000; Maekawa et al., 2003; Yin et al., 2000). Once Kar9p links the plus end to the actin network, Myo2p moves along polarized actin pulling the nucleus to the shmoo tip. When both mating partners have defects in nuclear orientation, such as kar9Δ cells, karyogamy can still occur when the plus ends of MTs stochastically interact (Molk et al., 2006). When karyogamy mutants, such as kar9Δ, are mated to a wild-type cell, no defects in nuclear congression occur (Berlin et al., 1990; Kurihara et al., 1994). In this unilateral condition, it is likely that the wild-type protein diffuses quickly into the other cell body after fusion and binds MT plus ends, driving nuclear congression. Therefore, nuclear orientation is not necessary for karyogamy, but proteins such as Kar9p and Myo2p that guide plus ends to the shmoo tip are required for high-fidelity nuclear congression. Because the plus end is the site of MT-MT cross-linking for nuclear congression, guidance of plus ends to the shmoo tip significantly increases the probability that MTs from the different nuclei will interact.
After MTs have been transported to the shmoo tip along microfilaments, their plus ends interact with the cell cortex (Maddox et al., 1999; Maddox et al., 2003). Plus-end-binding proteins keep polymerizing and depolymerizing plus ends at the shmoo tip. Thus far, only Bim1p has been proposed to link polymerizing plus ends to the shmoo tip (Maddox et al., 2003). Depolymerizing plus ends attach to the shmoo tip via Kar3p and Bik1p. Kar3p is a kinesin 14 motor protein that has minus-end-directed motility (Endow et al., 1994; Maddox et al., 2003; Meluh and Rose, 1990). Kar3p forms a heterodimer with the light chain Cik1p, and this heterodimer is targeted to the plus end, where Kar3p promotes depolymerization (Barrett et al., 2000; Maddox et al., 2003; Sproul et al., 2005). In kar3Δ cells, MTs lose their persistent attachment to the shmoo tip when switching to depolymerization and shorten back to the SPB (Maddox et al., 2003). Therefore, Kar3p is proposed to anchor depolymerizing plus ends to the shmoo tip, preventing their detachment and shortening.
Bik1p, the CLIP-170 ortholog in budding yeast, helps maintain depolymerizing MTs at the shmoo tip and is required for nuclear congression (Berlin et al., 1990; Lin et al., 2001; Molk et al., 2006). CLIP-170 is a MT-binding protein that was originally characterized as a linker between MTs and membranes in metazoan cells (Vaughan, 2005). In budding yeast, Bik1p may directly interact with the plasma membrane, anchoring MTs to the shmoo tip. In pheromone-treated kar3Δ cells, Bik1p localizes to the plus end, which suggests Bik1p localization does not depend on Kar3p (Molk et al., 2006). However, in kar3Δ cells MTs detach from the shmoo tip when they switch to shortening (Maddox et al., 2003), which demonstrates that Bik1p is necessary but not sufficient for MT interactions with the shmoo tip.
In addition to plus-end-binding proteins, other proteins involved in cell polarization could play a role in MT-shmoo-tip attachments. Kip2p is a plus-end-directed kinesin-like protein that transports both Kar9p and Bik1p along MTs in mitotic cells but kip2Δ cells do not exhibit karyogamy defects (Carvalho et al., 2004; Maekawa et al., 2003; Miller et al., 1998). It is unknown whether, after nuclear orientation, actin cables are required to maintain MT attachments. In myo2-17, myo2-18 and myo2-20 mutants that lack functional Myo2p myosin motors, MTs appear detached from the shmoo tip but the SPB remains near the base of the mating projection (Hwang et al., 2003). This phenotype is reminiscent of kar3Δ cells, and Myo2p may play a direct role in attachment of MTs to the shmoo tip. Furthermore, proteins that establish cell polarity and are found at the plasma membrane, such as Gpa1p or polarisome components, may link the MT to the shmoo tip. Thus, it is likely that additional proteins are required to couple MTs to the shmoo tip.
Models for MT attachment to the shmoo tip
How do MTs attach to the cell cortex at the shmoo tip? The leading hypothesis is the cortical anchorage model in which polymerizing MTs have Bim1p on their plus ends and are kept at the shmoo tip by Kar9p that is linked to the actin network or polarity proteins at the cell cortex (Fig. 3A) (Maddox et al., 2003; Miller et al., 1999). When MTs switch to depolymerization, the plus ends are held at the shmoo tip by Bik1p and Kar3p. Kar3p is anchored to the cell cortex by an unknown attachment factor. Candidates for this include a component of the heterotrimeric G-protein complex or the polarisome.
An alternative, but not mutually exclusive, model is the plus end cycling hypothesis (Fig. 3B). In this model, MT plus ends are not anchored to the cell cortex; instead they are transported by Myo2p towards the shmoo tip and are maintained at the ends of actin cables. In this hypothesis, Myo2p does not disassociate immediately from the end of the actin cable. By remaining bound, the Myo2p-Kar9p-Bim1p complex holds the MT at the shmoo tip while it grows or shortens. Kar9p could reinforce this linkage by interacting with actin or polarisome components (Miller et al., 1999). This would result in `attachment' where the affinity of a MT for the shmoo tip is defined by the ability of Myo2p and/or Kar9p to persistently associate with actin. Since Bim1p preferentially associates with growing MTs in the shmoo tip, Bim1p may maintain the attachment of polymerizing MT plus ends by recruiting Kar9p to strengthen the MT-actin interactions. As MTs shorten, Kar3p and Bik1p would keep plus ends at the shmoo tip, but at some frequency, MTs could be released (Molk et al., 2006). A released plus end could then be replaced by newly nucleated MTs guided by Myo2p into the shmoo tip. The movement of the new plus end into the shmoo tip would pull the nucleus toward the mating projection. The cycle of MT attachment, growth, detachment, and replacement would define a single nuclear oscillation that is repeated until cell fusion.
One way to distinguish between the two models would be an analysis of shmoo tip attachment in tropomyosin mutants in which actin cables are defective. The cortical anchorage model predicts that actin cables are not required once the MT interacts with the shmoo tip. The plus end cycling model requires actin cables for MT-shmoo-tip attachment.
MT dynamics during nuclear congression
After cell fusion, MTs rapidly associate during nuclear congression (Maddox et al., 1999). Nuclear congression requires Bik1p and Kar3p (Fig. 4A) (Berlin et al., 1990; Meluh and Rose, 1990). In early models for nuclear congression based on genetic analysis, Kar3p was proposed to cross-link and slide oppositely oriented MTs, which would result in extensive MT overlap (Fig. 4B) (Rose, 1996). This model predicts that plus ends would be located near the SPBs. However, live cell fluorescence microscopy has demonstrated that MTs do not slide past one another over long distances and that cross-linking occurs only near the plus ends (Molk et al., 2006). Therefore, we favor a new model for nuclear congression that focuses on plus end interactions (Fig. 4B). In this model, the opposing plus ends at the shmoo tip are juxtaposed and cross-linking occurs with a minimum of MT-MT sliding. The cross-linking is initiated by the Kar3p-Cik1p heterodimer and Bik1p could also play a role. Once plus ends interact, we propose that the MTs switch to persistent shortening to drive both nuclei together.
What other proteins are required for nuclear congression? Bim1p is necessary for efficient karyogamy (Schwartz et al., 1997) and may play a role in MT-MT interactions during nuclear congression. It is puzzling that bim1Δ cells have a severe karyogamy defect but Bim1p is proposed to function on polymerizing MTs. Bim1p could be required for MT growth after the cell wall breaks down to link the plus ends. After the polymerizing plus ends are linked by Bim1p, Kar3p may maintain the linkage after the switch to shortening occurs. Other plus-end-binding proteins that regulate MT dynamics could also be required for nuclear congression. Additionally, we do not know how the cell regulates the switch from the dynamic instability that occurs before cell fusion to persistent depolymerization after plus ends interact. Since Kar3p has been shown to depolymerize MT plus ends in vitro (Sproul et al., 2005), local regulation of Kar3p, either through protein-protein interactions or post-translational modification, could trigger the switch.
Mitosis during the first zygotic division
Little is known about mitosis during the first zygotic division. Thus far, it appears spindle elongation during the first zygotic division is similar to vegetative divisions (Maddox et al., 1999). A failure to perform karyogamy produces a single cell that has two haploid nuclei, each of which generates its own mitotic spindle. In experiments using kar1 mutants with two spindles, the response to DNA damage was determined to inhibit anaphase onset locally within a single nucleus (Demeter et al., 2000). A similar analysis of other karyogamy mutants may uncover characteristics about how cell division occurs in the presence of multiple spindles.
Conclusions and future perspectives
MT plus ends and plus-end-binding proteins are required for efficient mating in budding yeast. Plus ends are crucial for attachment to the shmoo tip and MT-MT interactions during nuclear congression. Plus-end-binding proteins required for MT dynamics in mating include Bim1p, Bik1p and Kar3p. Bik1p and Kar3p function at the kinetochore in budding yeast (Lin et al., 2001; Middleton and Carbon, 1994; Tanaka et al., 2005), which suggests MT linkages could be similar at the shmoo tip and centromere. Thus, the shmoo tip attachment site could also serve as a model for kinetochore-MT attachments. Future studies of MT-binding proteins may reveal further karyogamy factors that are necessary for nuclear congression. There are some differences between yeast karyogamy, in which dynein is not required, and metazoan fertilization, which does require cytoplasmic dynein (Molk et al., 2006; Payne et al., 2003). However, plus-end-binding proteins may well have conserved functions in yeast and metazoans that facilitate nuclear movements. Therefore, understanding how MT dynamics affect shmoo tip attachment and nuclear congression could reveal underlying mechanisms that apply to a variety of problems in cell biology.
We thank David Bouck, Richard Cheney, Julian Haase, Ted Salmon, David Stone and members of the Bloom and Salmon laboratories for helpful discussions and critical reading of the manuscript. We would like to acknowledge Paul Maddox for initiating the live cell studies of karyogamy and his insight into the mechanisms that govern shmoo tip attachment. This work was supported by the National Institutes of Health grant GM-32238 (to K.B.).
- Accepted July 26, 2006.
- © The Company of Biologists Limited 2006