Microtubule Cytoskeleton: Navigating the Intracellular Landscape Dispatch

The intracellular cytoskeleton in embryos and single-cell eukaryotes is extremely dynamic and constantly remodeled during growth and differentiation. Microtubules and actin filaments comprise the filamentous cytoskeletal network. We are starting to understand how microtubules and actin filaments collaborate to ‘read’ intracellular cues. Microtubules are nucleated from microtubule organizing centers, called centrosomes or spindle pole bodies. Proteins bound to dynamic microtubule plus ends mediate interactions with polarized actin filaments or asymmetrically distributed cell polarity cues in the cell cortex [1,2]. Actin is required for polarized growth and directed membrane secretion. In budding yeast, the direction of bud growth dictates the axis of mitotic spindle alignment and microtubules, guided by actin cables, align the spindle apparatus. In embryos of the nematode Caenorhabditis elegans, polarity is established through asymmetric segregation of several Par proteins, including Par-2 and Par-3, which specify the anterior–posterior axis of the embryo. After the first division, Par-2 and Par-3 are asymmetrically localized in one daughter cell (P1), while Par-3 is present throughout the cortex of the second daughter cell (AB). The spindle rotates 90° to align along the anterior–posterior axis in P1 cells with asymmetrically localized Par-2 and Par-3. As in the case of spindle orientation in budding yeast, actin is involved in spindle positioning by directing Par-2 and Par-3 localization [3]. Understanding the mechanism of asymmetric protein accumulation that facilitates mitotic spindle orientation will provide new insights into basic processes that underlie development. In budding yeast, Kar9p links microtubules to the actin cytoskeleton. Kar9p binds microtubules via the yeast EB1 homolog, Bim1p [4–6]. EB1/Bim1p is highly conserved and functions at microtubule plus ends [4]. Mammalian EB1 interacts with the adenomatous polyposis coli protein (APC) to guide microtubule plus ends to specific cortical sites. Kar9p interacts with actin through the type V myosin Myo2p, thereby linking the microtubules to polarized actin [1,2]. The old and newly duplicated spindle poles are distinct in budding yeast: the old pole is oriented toward the bud [7], and microtubules emanating from this pole lead the mitotic spindle into the bud [8]. One of the outstanding problems in the field has been to understand the genetic control of spindle pole ‘fate’ and how Kar9p directs the old pole toward the bud. Two recent papers [9,10] report new insights into the mechanistic basis of spindle pole differentiation. Protein binding and release from the spindle pole is a highly regulated event. The cell-division kinase Cdc28p and cyclin are central to spindle pole differentiation. Cdc28p–Clb5p is required for proper spindle assembly and orientation in yeast [11]. Maekawa et al. [10] screened for proteins that interact with Cdc28p and Clb5p, and found Kar9p, among others. Kar9p has fifteen consensus sites for phosphorylation by Cdc28p and is phosphorylated in a cell-cycle-dependent fashion. Kar9p binds spindle poles, microtubule plus ends and the neck of budded cells in G1/S phase of the cell cycle. Interestingly, Kar9p is lost from the tip of microtubule plus ends in cells with reduced Cdc28p–Clb5p activity, and spreads along the entire cytoplasmic microtubule. Cdc28p–Clb5p thus confers tight spatial control of Kar9p. Microtubules grow prematurely to the tip of the bud in cdc28 clb5 mutants, resulting in net migration of the entire spindle into the bud [10,11]. How, then, does Clb5p promote microtubule tip binding by Kar9p? Phosphorylated Kar9p is transported by the microtubule-based motor protein Kip2p from spindle poles to microtubule plus-ends [10]. But the question remained as to how Kar9p binds to the old pole but not to the new pole which is destined to remain in the mother. Liakopoulos et al. [9] found that Kar9p phosphorylation is significantly reduced in clb4 mutants and, importantly, Kar9p localized symmetrically to both spindle poles in these cells. The regulation of Kar9p by phosphorylation is thus required for its spatial distribution to one and only one spindle pole. Furthermore, the phosphorylated form of Kar9p has reduced affinity for Bim1p. Clb4p was found to bind to the mother pole, thereby restricting ‘active’ Kar9p to the old daughter bound pole. Kar9p loading to the daughter pole promotes microtubule penetration into the bud by binding to Myo2p and, subsequently, to actin cables. The spindle pole is thus the regulatory center for spindle positioning, and Kar9p delivery to specific microtubule plus ends from one pole guides these structures to their destination (Figure 1). These data reflect a spatial feature inherent in differential regulatory cascades. Cdc28p–Clb4p prevents Kar9p from binding to the new pole in the mother cell, while a different Cdk–cyclin, perhaps Cdc28p–Clb5p, promotes transport of Kar9p from the old pole to cytoplasmic microtubule plus ends destined for the bud. In addition to this programmed genetic control of its positioning, the spindle is a dynamic structure that can reposition or realign within cells experiencing a variety of mutational or external perturbations [12]. Even in the Current Biology, Vol. 13, R430–R432, May 27, 2003, ©2003 Elsevier Science Ltd. All rights reserved. DOI 10.1016/S0960-9822(03)00362-2