Linking cell cycle to asymmetric division: Aurora-A phosphorylates the Par complex to regulate Numb localization

Drosophila neural precursor cells divide asymmetrically by segregating the Numb protein into one of the two daughter cells. Numb is uniformly cortical in interphase but assumes a polarized localization in mitosis. Here, we show that a phosphorylation cascade triggered by the activation of Aurora-A is responsible for the asymmetric localization of Numb in mitosis. Aurora-A phosphorylates Par-6, a regulatory subunit of atypical protein kinase C (aPKC). This activates aPKC, which initially phosphorylates Lethal (2) giant larvae (Lgl), a cytoskeletal protein that binds and inhibits aPKC during interphase. Phosphorylated Lgl is released from aPKC and thereby allows the PDZ domain protein Bazooka to enter the complex. This changes substrate specificity and allows aPKC to phosphorylate Numb and release the protein from one side of the cell cortex. Our data reveal a molecular mechanism for the asymmetric localization of Numb and show how cell polarity can be coupled to cell-cycle progression.     

Cell cycle regulation of pEg3, a new Xenopus protein kinase of the KIN1/PAR-1/MARK family.

We report the characterization of pEg3, a Xenopus protein kinase related to members of the KIN1/PAR-1/MARK family. The founding members of this newly emerging kinase family were shown to be involved in the establishment of cell polarity and both microtubule dynamic and cytoskeleton organization. Sequence analyses suggest that pEg3 and related protein kinases in human, mouse, and Caenorhabditis elegans might constitute a distinct group in this family. pEg3 is encoded by a maternal mRNA, polyadenylated in unfertilized eggs and specifically deadenylated in embryos. In addition to an increase in expression, we have shown that pEg3 is phosphorylated during oocyte maturation. Phosphorylation of pEg3 is cell cycle dependent during Xenopus early embryogenesis and in synchronized cultured XL2 cells. In embryos, the kinase activity of pEg3 is correlated to its phosphorylation state and is maximum during mitosis. Using Xenopus egg extracts we demonstrated that phosphorylation occurs at least in the noncatalytic domain of the kinase, suggesting that this domain might be important for pEg3 function.     

Localization of Aurora A and Aurora B kinases during interphase: Role of the N-Terminal domain

Aurora kinases possess a conserved catalytic domain (CD) and a N-terminal domain (ND) that varies in size and sequence. We have previously reported that the N-terminal domain of AuroraA (AurA) participates in the localization of the kinase to the centrosome in interphase. AuroraB (AurB) is a chromosome passenger protein and its N-terminal domain is not necessary for its localization or function during mitosis. Using various combinations of GFP-AurA and AurB protein domains we show that AurB N-terminal domain is required for nuclear localization in Xenopus XL2 cells in interphase. In human cells, however, we found both AurA and AurB kinases in the nucleus, AurA being mainly cytoplasmic and AurB mainly nuclear. Both proteins are actively excluded from the nucleus by a CRM1 dependent pathway. Interestingly, at a functional level, in interphase, every combination of Aurora kinase domains (ND-CD) rescues histone H3 Serine10 phosphorylation defect induced by AurB knockdown. This clearly indicates the presence of a functional AurA in the nucleus. However, the chimera ND-AurA/CD-AurB was much more efficient than the ND-AurB/CD-AurA to rescue multinucleation also induced by AurB knockdown. This indicates that the catalytic domain of AurB is required to fulfill specific functions during mitosis that cannot be fulfilled by the catalytic domain of AurA, probably for localization reasons during mitosis.     

Cytoskeleton organization during the cell cycle in two stramenopile microalgae, Ochromonas danica (Chrysophyceae) and Heterosigma akashiwo (Raphidophyceae), with special reference to F-actin organization and its role in cytokinesis

F-actin organization during the cell cycle was investigated in two stramenopile microalgae, Ochromonas danica (Chrysophyceae; UTEX LB1298) and Heterosigma akashiwo (Raphidophyceae; NIES-6) using FITC-phalloidin. In the interphase cell of O. danica, F-actin bundles were localized forming a network structure in the cortical region, which converged from the anterior region to the posterior, whereas in the interphase cell of H. akashiwo, F-actin bundles were observed forming a network structure in the cortical region without any polarity. In both O. danica and H. akashiwo, at the initial stage of mitosis the cortical F-actin disappeared, and then during cytokinesis assembly of an actin-based ring-like structure occurred in the cell cortex in the plane of cytokinesis. The ring-like structure initiated from aster-like structures was composed of F-actin in both O. danica and H. akashiwo. Different from animal cells, later stages of cytokinesis of O. danica seemed to be promoted by microtubules, although the early stages of cytokinesis progressed with a constriction of the ring-like structure, whereas cytokinesis of H. akashiwo was apparently completed by constriction of the cell mediated by the F-actin ring, as in animal cells.     

Subcellular localization of LGN during mitosis: evidence for its cortical localization in mitotic cell culture systems and its requirement for normal cell cycle progression

Mammalian LGN/AGS3 proteins and their Drosophila Pins orthologue are cytoplasmic regulators of G-protein signaling. In Drosophila, Pins localizes to the lateral cortex of polarized epithelial cells and to the apical cortex of neuroblasts where it plays important roles in their asymmetric division. Using overexpression studies in different cell line systems, we demonstrate here that, like Drosophila Pins, LGN can exhibit enriched localization at the cell cortex, depending on the cell cycle and the culture system used. We find that in WISH, PC12, and NRK but not COS cells, LGN is largely directed to the cell cortex during mitosis. Overexpression of truncated protein domains further identified the Galpha-binding C-terminal portion of LGN as a sufficient domain for cortical localization in cell culture. In mitotic COS cells that normally do not exhibit cortical LGN localization, LGN is redirected to the cell cortex upon overexpression of Galpha subunits of heterotrimeric G-proteins. The results also show that the cortical localization of LGN is dependent on microfilaments and that interfering with LGN function in cultured cell lines causes early disruption to cell cycle progression.     

The I/LWEQ Domain in RapGAP3 Required for Posterior Localization in Migrating Cells

Cell migration requires a defined cell polarity which is formed by diverse cytoskeletal components differentially localized to the poles of cells to extracellular signals. Rap-GAP3 transiently and rapidly translocates to the cell cortex in response to chemoattractant stimulation and localizes to the leading edge of migrating cells. Here, we examined localization of truncated RapGAP3 proteins and found that the I/LWEQ domain in the central region of RapGAP3 was sufficient for posterior localization in migrating cells, as opposed to leading-edge localization of full-length Rap-GAP3. All truncated proteins accumulated at the leading edge of migrating cells exhibited clear translocation to the cell cortex in response to stimulation, whereas proteins localized to the posterior in migrating cells displayed no translocation to the cortex. The I/LWEQ domain appears to passively accumulate at the posterior region in migrating cells due to exclusion from the extended front region in response to chemoattractant stimulation rather than actively being localized to the back of cells. Our results suggest that posterior localization of the I/LWEQ domain of RapGAP3 is likely related to F-actin, which has probably different properties compared to newly formed F-actin at the leading edge of migrating cells, at the lateral and posterior regions of the cell.     

Deletion analysis of the Drosophila Inscuteable protein reveals domains for cortical localization and asymmetric localization

The Drosophila Inscuteable protein acts as a key regulator of asymmetric cell division during the development of the nervous system [1] [2]. In neuroblasts, Inscuteable localizes into an apical cortical crescent during late interphase and most of mitosis. During mitosis, Inscuteable is required for the correct apical-basal orientation of the mitotic spindle and for the asymmetric segregation of the proteins Numb [3] [4] [5], Prospero [5] [6] [7] and Miranda [8] [9] into the basal daughter cell. When Inscuteable is ectopically expressed in epidermal cells, which normally orient their mitotic spindle parallel to the embryo surface, these cells reorient their mitotic spindle and divide perpendicularly to the surface [1]. Like the Inscuteable protein, the inscuteable RNA is asymmetrically localized [10]. We show here that inscuteable RNA localization is not required for Inscuteable protein localization. We found that a central 364 amino acid domain - the Inscuteable asymmetry domain - was necessary and sufficient for Inscuteable localization and function. Within this domain, a separate 100 amino acid region was required for asymmetric localization along the cortex, whereas a 158 amino acid region directed localization to the cell cortex. The same 158 amino acid fragment could localize asymmetrically when coexpressed with the full-length protein, however, and could bind to Inscuteable in vitro, suggesting that this domain may be involved in the self-association of Inscuteable in vivo.     

New function of the proline rich domain in dynamin-2 to negatively regulate its interaction with microtubules in mammalian cells

Microtubule reorganization is necessary for many cellular functions such as cell migration, cell polarity and cell division. Dynamin was originally identified as a microtubule-binding protein. Previous limited digestion experiment revealed that C-terminal 100-amino acids proline rich domain (PRD) of dynamin is responsible for microtubule binding in vitro. However, as obvious localization of dynamin along microtubules is only observed at the spindle midzone during mitosis but not in interphase cells, it remains unclear how dynamin interacts with microtubules in vivo. Here, we report that GFP-dynamin-2-(1-786), a truncated mutant lacking a C-terminal portion of the PRD, localized along microtubules in interphase HeLa cells. GFP-dynamin-2-wild type (WT) and GFP-dynamin-2-(1-745), a construct that was further truncated to remove the entire PRD, localized in discrete punctate structures but not along microtubules. These data suggest that the N-terminal (residues 746-786) but not the entire PRD is necessary for the interaction of dynamin-2 with microtubules in the cell and that the C-terminus of PRD (787-870) negatively regulate this interaction. Microtubules in cells expressing GFP-dynamin-2-(1-786) were stabilized against exposure to cold. These results provide a first evidence for a regulated interaction of dynamin-2 with microtubules in cultured mammalian cells.     

Characterization and dynamics of cytoplasmic F-actin in higher plant endosperm cells during interphase, mitosis, and cytokinesis

We have identified an F-actin cytoskeletal network that remains throughout interphase, mitosis, and cytokinesis of higher plant endosperm cells. Fluorescent labeling was obtained using actin monoclonal antibodies and/or rhodamine-phalloidin. Video-enhanced microscopy and ultrastructural observations of immunogold-labeled preparations illustrated microfilament-microtubule co-distribution and interactions. Actin was also identified in cell crude extract with Western blotting. During interphase, microfilament and microtubule arrays formed two distinct networks that intermingled. At the onset of mitosis, when microtubules rearranged into the mitotic spindle, microfilaments were redistributed to the cell cortex, while few microfilaments remained in the spindle. During mitosis, the cortical actin network remained as an elastic cage around the mitotic apparatus and was stretched parallel to the spindle axis during poleward movement of chromosomes. This suggested the presence of dynamic cross-links that rearrange when they are submitted to slow and regular mitotic forces. At the poles, the regular network is maintained. After midanaphase, new, short microfilaments invaded the equator when interzonal vesicles were transported along the phragmoplast microtubules. Colchicine did not affect actin distribution, and cytochalasin B or D did not inhibit chromosome transport. Our data on endosperm cells suggested that plant cytoplasmic actin has an important role in the cell cortex integrity and in the structural dynamics of the poorly understood cytoplasm-mitotic spindle interface. F-actin may contribute to the regulatory mechanisms of microtubule-dependent or guided transport of vesicles during mitosis and cytokinesis in higher plant cells.     

An Elmo-like protein associated with myosin II restricts spurious F-actin events to coordinate phagocytosis and chemotaxis

Elmo proteins positively regulate actin polymerization during cell migration and phagocytosis through activation of the small G protein Rac. We identified an Elmo-like protein, ElmoA, in Dictyostelium discoideum that unexpectedly functions as a negative regulator of actin polymerization. Cells lacking ElmoA display an elevated rate of phagocytosis, increased pseudopod formation, and excessive F-actin localization within pseudopods. ElmoA associates with cortical actin and myosin II. TIRF microscopic observations of functional ElmoA-GFP reveal that a fraction of ElmoA localizes near the presumptive actin/myosin II cortex and the levels of ElmoA and myosin II negatively correlate with that of polymerizing F-actin. F-actin-regulated dynamic dispersions of ElmoA and myosin II are interdependent. Taken together, our data suggest that ElmoA modulates actin/myosin II at the cortex to prevent excessive F-actin polymerization around the cell periphery, thereby maintaining proper cell shape during phagocytosis and chemotaxis.     

Directionality of F-actin cables changes during the fission yeast cell cycle

Longitudinal F-actin cables are thought to be important for transporting materials for polarized cell growth in fission yeast. We show that most F-actin in the cables is oriented such that the barbed end faces the nearest cell tip during interphase; however, this directionality is reversed during mitosis. These orientations of F-actin ensure proper transport of materials to growing sites during these cell-cycle stages.     

Plant-specific mitotic targeting of RanGAP requires a functional WPP domain

The small GTPase Ran is involved in nucleocytoplasmic transport, spindle formation, nuclear envelope (NE) formation, and cell-cycle control. In vertebrates, these functions are controlled by a three-dimensional gradient of Ran-GTP to Ran-GDP, established by the spatial separation of Ran GTPase-activating protein (RanGAP) and the Ran guanine nucleotide exchange factor RCC1. While this spatial separation is established by the NE during interphase, it is orchestrated during mitosis by association of RCC1 with the chromosomes and RanGAP with the spindle and kinetochores. SUMOylation of vertebrate RanGAP1 is required for NE, spindle, and centromere association. Arabidopsis RanGAP1 (AtRanGAP1) lacks the SUMOylated C-terminal domain of vertebrate RanGAP, but contains a plant-specific N-terminal domain (WPP domain), which is necessary and sufficient for its targeting to the NE in interphase. Here we show that the human and plant RanGAP-targeting domains are kingdom specific. AtRanGAP1 has a mitotic trafficking pattern uniquely different from that of vertebrate RanGAP, which includes targeting to the outward-growing rim of the cell plate. The WPP domain is necessary and sufficient for this targeting. Point mutations in conserved residues of the WPP domain also abolish targeting to the nuclear rim and the cell plate, suggesting that the same mechanism is involved in both targeting events. These results indicate that plant and animal RanGAPs undergo different migration patterns during cell division, which require their kingdom-specific targeting domains.     

Localization of the 26S proteasome during mitosis and meiosis in fission yeast.

The 26S proteasome is a large multisubunit complex involved in degrading both cytoplasmic and nuclear proteins. We have investigated the localization of this complex in the fission yeast, Schizosaccharomyces pombe. Immunofluorescence microscopy shows a striking localization pattern whereby the proteasome is found predominantly at the nuclear periphery, both in interphase and throughout mitosis. Electron microscopic analysis revealed a concentration of label near the inner side of the nuclear envelope. The localization of green fluorescent protein (GFP)-tagged 26S proteasomes was analyzed in live cells during mitosis and meiosis. Throughout mitosis the proteasome remained predominantly at the nuclear periphery. During meiosis the proteasome was found to undergo dramatic changes in its localization. Throughout the first meiotic division, the signal is more dispersed over the nucleus. During meiosis II, there was a dramatic re-localization, and the signal became restricted to the area between the separating DNA until the end of meiosis when the signal dispersed before returning to the nuclear periphery during spore formation. These findings strongly imply that the nuclear periphery is a major site of protein degradation in fission yeast both in interphase and throughout mitosis. Furthermore they raise interesting questions as to the spatial organization of protein degradation during meiosis.     

In vivo dynamics and kinetics of pKi-67: transition from a mobile to an immobile form at the onset of anaphase

A cell proliferation marker protein, pKi-67, distributes to the chromosome periphery during mitosis and nucleolar heterochromatin in the interphase. We report here on the structural domains of pKi-67 that are required for its correct distribution. While both the LR domain and the conserved domain were involved in localization to the nucleolar heterochromatin, both the LR domain and the Ki-67 repeat domain were required for its distribution to the mitotic chromosome periphery. Using in vivo time-lapse microscopy, GFP-pKi-67 was dynamically tracked from the mitotic chromosome periphery to reforming nucleoli via prenucleolar bodies (PNBs). The signals in PNBs then moved towards and fused into the reforming nucleoli with a thin string-like fluorescence during early G1 phase. An analysis of the in vivo kinetics of pKi-67 using photobleaching indicated that the association of pKi-67 with chromatin was progressively altered from "loose" to "tight" after the onset of anaphase. These findings indicate that pKi-67 dynamically alters the nature of the interaction with chromatin structure during the cell cycle, which is closely related to the reformation process of the interphase nucleolar chromatin.     

Cell cycle-dependent expression and centrosome localization of a third human aurora/Ipl1-related protein kinase, AIK3

We earlier isolated cDNAs encoding novel human protein kinases AIK and AIK2 sharing high amino acid sequence identities with Drosophila Aurora and Saccharomyces cerevisiae Ipl1 kinases whose mutations cause abnormal chromosome segregation. In the present study, a third human cDNA (AIK3) highly homologous to aurora/IPL1 was isolated, and the nucleotide sequence was determined. This cDNA encodes 309 amino acids with a predicted molecular mass of 35.9 kDa. C-terminal kinase domain of AIK3 protein shares high amino acid sequence identities with those of Aurora/Ipl1 family protein kinases including human AIK, human AIK2, Xenopus pEg2, Drosophila Aurora, and yeast Ipl1, whereas the N-terminal domain of AIK3 protein shares little homology with any other Aurora/Ipl1 family members. AIK3 gene was assigned to human chromosome 19q13.43, which is a frequently deleted or rearranged region in several tumor tissues, by fluorescence in situ hybridization, somatic cell hybrid panel, and radiation hybrid cell panel. Northern blot analyses revealed that AIK3 expression was limited to testis. The expression levels of AIK3 in several cancer cell lines were elevated severalfold compared with normal fibroblasts. In HeLa cells, the endogenous AIK3 protein level is low in G1/S, accumulates during G2/M, and reduces after mitosis. Immunofluorescence studies using a specific antibody have shown that AIK3 is localized to centrosome during mitosis from anaphase to cytokinesis. These results suggest that AIK3 may play a role(s) in centrosome function at later stages of mitosis.     

Distinct functions of the unique C terminus of LAP2alpha in cell proliferation and nuclear assembly

The non-membrane-bound lamina-associated polypeptide 2 isoform, LAP2alpha, forms nucleoskeletal structures with A-type lamins and interacts with chromosomes in a cell cycle-dependent manner. LAP2alpha contains a LEM (LAP2, emerin, and MAN1) domain in the constant N terminus that binds to chromosomal barrier-to-autointegration factor, and a C-terminal unique region that is essential for chromosome binding. Here we show that C-terminal LAP2alpha fragment efficiently bound to mitotic chromosomes and inhibited assembly of endogenous LAP2alpha, nuclear membranes, and lamins A/C in in vitro nuclear assembly assays. Full-length recombinant LAP2alpha, which bound to chromosomes, and N-terminal fragment, which did not bind, had no effect on assembly. This suggested an essential role for the LAP2alpha C terminus in chromosome association and for the N-terminal LEM domain in subsequent assembly stages. In vivo analysis upon transient expression of GFP-tagged LAP2alpha fragments confirmed that, unlike the N-terminal fragment, the C-terminal fragment was able to bind to chromosomes during mitosis, if expressed weakly. At higher expression levels, C-terminal LAP2alpha fragment and full-length protein led to cell cycle arrest in interphase and apoptosis, as shown by fluorescence-activated cell sorter analysis, time lapse microscopy, and BrdUrd incorporation assays. These data indicated distinct functions of LAP2alpha in cell cycle progression during interphase and in nuclear reassembly during mitosis.     

Membrane targeting and asymmetric localization of Drosophila partner of inscuteable are discrete steps controlled by distinct regions of the protein

Asymmetric division of neural progenitors is a key mechanism by which neuronal diversity in the Drosophila central nervous system is generated. The distinct fates of the daughter cells derived from these divisions are achieved through preferential segregation of the cell fate determinants Prospero and Numb to one of the two daughters. This is achieved by coordinating apical and basal mitotic spindle orientation with the basal cortical localization of the cell fate determinants during mitosis. A complex of apically localized proteins, including Inscuteable (Insc), Partner of Inscuteable (Pins), Bazooka (Baz), DmPar-6, DaPKC, and G alpha i, is required to mediate and coordinate basal protein localization with mitotic spindle orientation. Pins, a molecule which directly interacts with Insc, is a key component required for the integrity of this complex; in the absence of Pins, other components become mislocalized or destabilized, and basal protein localization and mitotic spindle orientation are defective. Here we define the functional domains of Pins. We show that the C-terminal region containing the G alpha i binding GoLoco motifs is necessary and sufficient for targeting to the neuroblast cortex, which appears to be a prerequisite for apical localization of Pins. The N-terminal tetratricopeptide repeat-containing region of Pins is required for two processes; TPR repeats 1 to 3 plus the C-terminal region are required for apical localization but are insufficient to recruit Insc to the apical cortex, whereas TPR repeats 1 to 7 plus C-terminal Pins can perform both functions. Hence, the abilities of Pins to cortically localize, to apically localize, and to restore Insc apical localization are all separable, and all three capabilities are necessary to mediate asymmetric division. Moreover, the need for N-terminal Pins can be obviated by fusing a minimal Insc functional domain with the C-terminal region of Pins; this chimeric molecule is apically localized and can fulfill the functions of both Insc and Pins.     

Drosophila Aurora-A is required for centrosome maturation and actin-dependent asymmetric protein localization during mitosis

BACKGROUND: During asymmetric cell division in the Drosophila nervous system, Numb segregates into one of two daughter cells where it is required for the establishment of the correct cell fate. Numb is uniformly cortical in interphase, but in late prophase, the protein concentrates in the cortical area overlying one of two centrosomes in an actin/myosin-dependent manner. What triggers the asymmetric localization of Numb at the onset of mitosis is unclear. RESULTS: We show here that the mitotic kinase Aurora-A is required for the asymmetric localization of Numb. In Drosophila sensory organ precursor (SOP) cells mutant for the aurora-A allele aurA(37), Numb is uniformly localized around the cell cortex during mitosis and segregates into both daughter cells, leading to cell fate transformations in the SOP lineage. aurA(37) mutant cells also fail to recruit Centrosomin (Cnn) and gamma-Tubulin to centrosomes during mitosis, leading to spindle morphology defects. However, Numb still localizes asymmetrically in cnn mutants or after disruption of microtubules, indicating that there are two independent functions for Aurora-A in centrosome maturation and asymmetric protein localization during mitosis. Using photobleaching of a GFP-Aurora fusion protein, we show that two rapidly exchanging pools of Aurora-A are present in the cytoplasm and at the centrosome and might carry out these two functions. CONCLUSIONS: Our results suggest that activation of the Aurora-A kinase at the onset of mitosis is required for the actin-dependent asymmetric localization of Numb. Aurora-A is also involved in centrosome maturation and spindle assembly, indicating that it regulates both actin- and microtubule-dependent processes in mitotic cells.     

Rearrangements of F-actin during Stomatogenesis Visualised by Confocal Microscopy in Fixed and Permeabilised Tradescantia Leaf Epidermis

Abstract: New details of F-actin organisation in leaf epidermal and stomatal cells were revealed by rhodamine — and fluorescein — phalloidin staining of fixed epidermal peels of Tradescantia virginiana and visualisation by confocal microscopy. Non-specialised epidermal cells contain highly organised arrays of fine cortical actin filaments aligned in transverse or oblique orientations. In interphase guard mother cells (GMCs), the arrangement of cortical F-actin changes on the periclinal and anticlinal cell walls at different times during differentiation. Initially, cortical F-actin on the periclinal surfaces is oriented transversely and F-actin is evenly distributed around the anticlinal walls. Following polarisation of the adjacent subsidiary mother cells (SMCs), actin in GMCs concentrates on the lateral anticlinal walls, but not on the transverse walls. Subsequently, F-actin on the periclinal walls reorients to radial and then longitudinal. Organisation of F-actin in SMCs appears to be influenced by the adjacent GMCs and co-ordination in F-actin arrangements in cells of the stomatal complex continues through to the formation of the guard cell pair. Our studies indicate that actin bands marking the division site in prophase cells, and detected in microinjected living material, are a particularly labile subset of F-actin. Actin bands were difficult to preserve, even when aldehyde fixation was avoided, in contrast to all interphase and mitotic F-actin.