Functional study of the NIMA protein kinases Nek9, Nek6 and Nek7 at the onset of mitosis. Control of the kinesin Eg5 and prophase centrosome separation

[eng] Mitosis is a tightly regulated process that aims to ensure the correct distribution of the chromosomes between the two newly generated cells. Many protein kinases have been defined as essential for this process: cyclin- dependent kinases, Aurora family and Polo family kinases are some of the most relevant players. The objective of this thesis is to characterize one of the less studied kinase pathways involved in this process, which is constituted by the NIMA-related kinases Nek9, Nek6 and Nek7. Nek9 is activated at the onset of mitosis by a double step mechanism mediated by CDK1 and Plk1. Once Nek9 is activated it can bind to Nek6 and Nek7 and phosphorylate them, promoting their activation. Finally, Nek6 and Nek7 are responsible for the phosphorylation of the kinesin Eg5, promoting Eg5 accumulation at centrosome, and consequently, centrosome separation. The kinesin eg5 motor protein is considered as one of the major players for centrosome separation and formation of the bipolar spindle. The tetramer configuration allows Eg5 to bind antiparallel microtubules and slide them apart, exerting a force that promotes centrosome separation and the maintenance of the bipolar spindle. Centrosome separation, however, is a highly intricate process that involves several pathways, including Eg5 activity. Dynein presents a directed activity towards the minus ends of microtubules, which has a redundant role to Eg5 in centrosome separation. Dynein accumulation at the cell cortex and the nuclear membrane, through its adaptor BicD2, is also involved in centrosome tethering at the nuclear envelope, a necessary step prior to separation. Furthermore, dynein can control the position of Eg5 at the spindle via TPX2, an event that could also happen before nuclear envelope breakdown (NEB). Here we describe the conditions required for Eg5 accumulation at the centrosmes after Ser1033 phosphorylation. During the development of this project we have explored the essential circumstances for correct Eg5 localization in cells. By using protein-protein interaction techniques and shRNA depletion of protein candidates we have determined that another motor protein, dynein, together with the adaptor BicD2 and the protein TPX2 are responsible for Eg5 accumulation around centrosomes. Additionally, we proposed TPX2 as a novel Nek9 substrate and we have investigated the role of this phosphorylation, which affects TPX2 localization during prophase, before NEB. We present with this thesis a model for Eg5 accumulation at microtubule minus ends and centrosome separation during prophase summarized in the following points: 1) Dynein complex transports Eg5 towards the centrosome. Dynein interacts with Eg5 independently of the Ser1033 phosphorylation. The adaptor BicD2, which interacts directly with Eg5 tail domain, mediates the interaction. Dynein motility towards microtubule minus ends and the presence of BicD2 on the complex are required for Eg5 localization at centrosomes. Thus, the dynein complex is required for Eg5 transport to the centrosomes during G2-M transition. 2) TPX2 inhibits Eg5 motility in response to Ser1033 phosphorylation. TPX2 is necessary for the correct localization of Eg5 at centrosomes during prophase. TPX2 mislocalization at centrosomes without altering its overall levels leads to failed Eg5 localization, therefore the presence of TPX2 at centrosomes during prophase is required for Eg5 localization. TPX2 interacts with Eg5 during mitosis and the interaction is abolished when the Ser1033 can’t be phosphorylated. Thus, TPX2 is able to respond to Eg5 Ser1033 phosphorylation, which we propose is promoting the interaction between these two proteins, and consequently inhibiting Eg5 motility at centrosomal levels. 3) TPX2 phosphorylation by Nek9 promotes its centrosomal localization. Nek9 phosphorylation of TPX2 is responsible for TPX2 localization at the spindle poles during prophase. Nek9 phosphorylates TPX2 at residues that are proximal to a NLS, making TPX2 localization more cytoplasmic and promoting its accumulation to the area where Nek9 is more active, the centrosome.[spa] La mitosis es un proceso altamente regulado cuyo objetivo es asegurar la correcta distribución de los cromosomas entre las dos células nuevamente generadas. Diferentes proteínas quinasas han sido definidas como esenciales en este proceso pero el objetivo de esta tesis es caracterizar una de las rutas de señalización menos estudiada, la cual la componen las NIMA quinasas Nek9, Nek6 y Nek7. Nek9 es activada al inicio de mitosis por un doble mecanismo mediado por CDK1 y Plk1. Una vez activada, se puede unir a Nek6 y Nek7 y fosforilarlas, promoviendo su activación. Finalmente, Nek6 y Nek7 son responsables de la fosforilación de la quinesina Eg5, promoviendo la acumulación de Eg5 en los centrosomas, y en consecuencia, la separación de los mismos en profase. Aquí describimos las condiciones necesarias para la acumulación de Eg5 en los centrosomas después de la fosforilación en la Ser1033. Durante el desarrollo de este trabajo hemos explorado las circunstancias esenciales para una correcta localización de Eg5 en las células. Usando técnicas de interacción proteína-proteína y técnicas de silenciamiento proteico de candidatos con shRNA hemos determinado que otra proteína motora, dineína, junto con el adaptador BicD2 y la proteína TPX2, son responsables de la acumulación de Eg5 alrededor de los centrosomas. Además, hemos propuesto a TPX2 como un nuevo substrato regulado por Nek9 y hemos investigado el papel de esta fosforilación, la cual afecta la localización de TPX2 durante profase, antes de la rotura de la membrana nuclear. Con esta tesis presentamos un modelo para la acumulación de Eg5 y la separación de los centrosomas en profase que puede ser resumido en los siguientes puntos: - El complejo de dineína transporta Eg5 hacia el centrosoma independientemente de la fosforilación en la Ser1033. El adaptador BicD2 media esta interacción uniéndose directamente al dominio C terminal de Eg5. -TPX2 inhibe movilidad de Eg5 en respuesta a la fosforilación en la Ser1033. - La presencia de TPX2 en los centrosomas es necesaria para la localización de Eg5. La fosforilación de TPX2 por Nek9 promueve la localización de TPX2 en los centrosomas durante la profase