Trends in Cell Biology
Chromosome Segregation and Aneuploidy seriesAurora kinases, aneuploidy and cancer, a coincidence or a real link?
Introduction
The main goal of a mitotic cell is to equally segregate its chromosomes and centrosomes between two daughter cells (Figure 1). As early as 1902, Theodore Boveri predicted that mitotic abnormalities could lead to genetic instability and cancer [1]. Indeed, during mitosis, a spectacular reorganization of the cytoskeleton occurs that builds a bipolar microtubule spindle that assures proper segregation of chromosomes. Mitosis is progressively readied throughout cell-cycle progression: a series of precisely coordinated events must have occurred before the G2–M transition occurs. By the end of S-phase, the cell must have duplicated its centrosome and replicated its DNA. At the end of prophase, the duplicated and matured centrosomes must have become separated. During prometaphase, the two centrosomes and the chromosomes nucleate highly dynamic mitotic microtubules that assemble a bipolar spindle [2]. During progression from prometaphase to metaphase, the chromosomes must become bi-orientated and aligned at the metaphase plate. Bi-orientation is achieved by microtubule-organized attachment of kinetochore pairs to opposite centrosomes. During this process, the mitotic checkpoint is continuously activated; it controls microtubule attachment to the kinetochores and tension. When these two conditions are satisfied, the checkpoint signals are switched off, the chromatids separate and anaphase proceeds. In telophase, nuclear division occurs and the cell undergoes cytokinesis. Finally, each daughter cell receives one set of chromosomes and one centrosome (Box 1).
Considering the complexity of mitosis, it is not surprising that there are many mitotic defects that can lead to the formation of aneuploid daughter cells. An aneuploid cell possesses an altered content of DNA (abnormal number of chromosomes). To give some examples, during mitosis, one can easily imagine at least four ways to obtain an aneuploid cell: (1) incorrect centrosome duplication during interphase that can lead to the formation of multipolar spindles, (2) improper centrosome separation, (3) defective cytokinesis and (4) chromosome mis-orientation (Figure 1).
Progression through mitosis depends on three main regulatory mechanisms: protein localization, proteolysis and phosphorylation. Cdk1 (cyclin-dependent kinase 1 or p34cdc2), Plk1 (Polo-like kinase), Nek2 (NIMA-related kinase 2) and Aurora kinases are the main mitotic protein kinases known to date [3]. Aurora kinases are involved in centrosome separation and maturation, spindle assembly and stability, chromosome condensation, congression and segregation, and cytokinesis [4]. In mammalian cells, three aurora kinases have been identified, named Aurora-A, -B and -C [3]. They possess a very conserved catalytic domain and an N-terminal domain that varies in sequence and in length 5, 6. It is thought that the different Aurora kinases localize differentially and might perform different functions during mitosis. However, all three human kinases are overexpressed in many types of cancers, in which polyploid cells containing multiple centrosomes are observed. This review discusses how overexpression of aurora kinases provokes the development of such a phenotype, and whether this phenotype is key to oncogenesis. Although many studies of Aurora kinases in organisms such as yeasts, Drosophila, Caenorhabditis elegans and Xenopus have greatly contributed to our understanding of the functions of the kinases, we will only use here the nomenclature Aurora-A, -B and -C and the names of mammalian proteins whenever possible.
Section snippets
Aurora-A: a non residential centrosome kinase
Aurora-A kinase localizes on duplicated centrosomes from the end of S phase to the beginning of the following G1 phase 6, 7. The kinase is then degraded by the proteasome in a Cdh1-dependent manner 8, 9. During mitosis, the protein also localizes to the poles of the mitotic spindle (Box 1). Although these localizations are most commonly observed, it seems that the centrosomal Aurora-A kinase exchanges rapidly with an apparently large cytoplasmic pool of Aurora-A [10].
The catalytic activity of
Aurora-B: the chromosome passenger kinase
The Aurora-B kinase belongs to the chromosome passenger protein family (Box 1). Such proteins localize to the kinetochores from prophase to metaphase and to the central spindle and the midbody in cytokinesis [6]. Association of Aurora-B with the inner centromere protein INCENP led to the conclusion that Aurora-B participates in a ‘chromosome passenger complex’ 27, 28. In human cells, INCENP is required to localize Aurora-B at the kinetochores [27]. And, just like Aurora-A, Aurora-B activation
Aurora-C: only a male meiotic kinase or another chromosome passenger protein?
The gene encoding Aurora-C was first isolated from a testis cDNA library and the gene localized to Chr19q13 [54]. Northern blot analyses of Aurora-C then confirmed that the kinase was indeed only detectable in testis 55, 56. During spermatogenesis, high expression of Aurora-C mRNA was detected mainly in pachytene spermatocytes, concomitant with meiotic spindle formation [57]. Although the protein level was not measured, the data suggested that Aurora-C might be involved in meiotic spindle
Surviving without Aurora kinase
Cells usually do not tolerate a lack of Aurora kinase. In the absence of Aurora-A, centrosomes fail to assemble bipolar spindles, inducing the formation of several abnormal mitotic spindles such as spindles with short astral microtubules, monopolar spindles (one pole), monoastral bipolar spindles (bipolar, but with one pole lacking a centrosome), bipolar spindles with multiple centrosomes at the poles, and multipolar spindles (Box 2) 11, 60. A decrease in Aurora-A protein level induced by RNA
To be or not to be an oncogene – that is the question
The protein encoded by the Aurora-A gene was first named BTAK for ‘breast tumor activated kinase’ because its gene, located at 20q13, was found amplified and its mRNA overexpressed in breast tumors. The presence of the amplicon 20q13 in breast tumors is an indicator of poor prognosis [71]. But a breakthrough came from two different groups that demonstrated that ectopic overexpression of Aurora-A was sufficient to transform NIH3T3 and rat1 cells and that transformed cells induced tumors when
Which one of the aurora kinases is the best anti-cancer therapeutic target?
It has been established that, first, overexpression of Aurora-A can transform cells carrying genetic defects and, second, that Aurora-B overexpression induces tumor metastasis. In both cases, the catalytic activity of the kinase is required. So would it be a good idea to develop inhibitors of Aurora as chemotherapeutic drugs? It is always difficult to answer such questions when the functions of the kinases are not yet fully understood. But usually one takes a chance – and indeed a breakthrough
Concluding remarks
The similarities and differences between the three Aurora kinases seem to fit with the proposed evolutionary origin of the kinases [88]. Aurora kinases evolved from a common chordate ancestor, then two Aurora kinases appeared in cold-blooded vertebrates: a true Aurora-A kinase and an Aurora-B/C ancestor. In mammals, the Aurora-B/C ancestor gave rise to two closely related kinases – Aurora-B and Aurora-C. It is not surprising that these last two kinases colocalize (to kinetochore and midbody)
Acknowledgements
I express my apologies to all the authors who have published work related to this review but have not been cited here owing to space constraints. Research in the Claude Prigent laboratory is financed by the Ligue Nationale Contre le Cancer (équipe labellisée) and by the CNRS. Clotilde Petretti is a fellow of the region Bretagne.
References (89)
- et al.
Chromosomes take the lead in spindle assembly
Trends Cell Biol.
(1995) Aurora-A kinase maintains the fidelity of early and late mitotic events in HeLa cells
J. Biol. Chem.
(2003)Identification of a functional destruction box in the Xenopus laevis aurora-A kinase pEg2
FEBS Lett.
(2001)Degradation of human Aurora-A protein kinase is mediated by hCdh1
FEBS Lett.
(2002)- et al.
Drosophila Aurora-A is required for centrosome maturation and actin-dependent asymmetric protein localization during mitosis
Curr. Biol.
(2002) Centrosome maturation and duplication in C. elegans require the coiled-coil protein SPD-2
Dev. Cell
(2004)A novel mechanism for activation of the protein kinase Aurora A
Curr. Biol.
(2003)Interaction and feedback regulation between STK15/BTAK/Aurora-A kinase and protein phosphatase 1 through mitotic cell division cycle
J. Biol. Chem.
(2001)Aurora-A and an interacting activator, the LIM protein Ajuba, are required for mitotic commitment in human cells
Cell
(2003)The Xenopus laevis aurora-related protein kinase pEg2 associates with and phosphorylates the kinesin-related protein XlEg5
J. Biol. Chem.
(1999)