LIMK1 et LIMK2 sont des sérines/thréonine kinases capables de phosphoryler et d’inactiver la cofiline, un facteur de dépolymérisation de l’actine. Elles sont régulées négativement par la neurofibromine, responsable de la neurofibromatose de type 1, et pourraient être impliquées à la fois dans les aspects tumoraux et cognitifs de cette maladie par leur rôle dans la dynamique de l’actine. Nous avons étudié l’isoforme LIMK2‐1 de LIMK2, spécifique des hominidés et précédemment associée à la déficience intellectuelle. Cette isoforme possède un domaine kinase tronqué et un domaine inhibiteur de la phosphatase 1 (PP1i) en C‐terminal. Nos résultats montrent, d’une part, que LIMK2‐1 existe sous forme de protéine et qu’elle est exprimée dans le système nerveux central chez l’homme, en particulier au cours du neurodéveloppement. D’autre part, il apparaît que cette isoforme favorise la polymérisation de l’actine. Cette action semble indépendant de l’activité kinase puisque LIMK2‐1 ne phosphoryle pas la cofiline. Nous avons également montré que le domaine PP1i interagissait spécifiquement avec la phosphatase 1 et des résultats complémentaires suggèrent un rôle de ce domaine dans l’inhibition de la dépolymérisation de l’actine. Ces données mettent en évidence un mécanisme moléculaire nouveau pour une protéine de la famille des LIMK et soulignent l’intérêt d’étudier ces protéines afin de mieux comprendre leur implication dans les troubles cognitifs et dans la neurofibromatose de type 1. / LIMK1 and LIMK2 are serine/threonine kinases that phosphorylate and subsequently inactivate cofilin, an actin-depolymerizing factor. Neurofibromin, the protein responsible for neurofibromatosis type 1, negatively regulates these proteins that may be involved in tumoral and cognitive aspects of the disease through their role in actin dynamics. We studied LIMK2-1, a hominidae-specific isoform previously involved in intellectual disability. This isoform possesses a truncated kinase domain and a protein phosphatase 1 inhibitory (PP1i) domain at its C-terminal extremity. Our results showed that LIMK2-1 exists at a protein level and that it is expressed in human central nervous system, especially during neurodevelopment. Moreover, LIMK2-1 promotes actin polymerization independently from a kinase activity, since this isoform does not phosphorylate cofiline. We also highlighted an interaction between the PP1i domain and protein phosphatase 1 and complementary results suggest a role of this domain in the inhibition of actin depolymerization. These data highlight a new molecular mechanism for a LIMK protein and emphasize the interest of studying these proteins to understand their involvement in cognitive disorders and in neurofibromatosis type 1.

LIMK2 is a serine/threonine/tyrosine kinase that promotes tumor cell invasion and metastasis by phosphorylating cell proteins and altering their functions. There is a need to find tumor-specific substrates for LIMK2 in order to understand the downstream pathway of these substrates, their function, and how they are regulated by LIMK2. Recently, our labrotory identified LIMK2 as an excellent target for curing castration-resistant prostate cancer (CRPC). In this study, we identify two novel substrates for LIMK2 in CRPC: speckle-type POZ protein (SPOP), and Y-box binding protein-1 (YBX1). While LIMK2 negatively regulates SPOP, it positively regulates YBX1 − both by phosphorylation using in-vitro kinase assays. A study in our labrotory also proved that LIMK2 regulates Aurora A kinase (AURKA), where AURKA directly phosphorylates LIMK2. AURKA is a serine/threonine kinase that regulates cell cycle during mitosis; it is known to be upregulated, with uncontrolled activity, in many types of cancer, including prostate cancer. It is therefore important to identify new substrates for AURKA, especially in light of reported lethality in early embryonic mice, in association with AURKA-knockout. In other words, targeting AURKA directly may cause severe toxicity, a finding that has prevented direct inhibitors from passing Phase II clinical trials. In this study, we also identified SPOP and YBX1 as direct substrates for AURKA. Our results confirm what we know about the LIMK2/AURKA relationship: that AURKA negatively regulates SPOP and positively regulates YBX1. Targeting LIMK2 and AURKA indirectly through SPOP, YBX1 and its other substrates holds tremendous therapeutic potential in treating prostate cancer. With this, we open the door for researches to investigate the direct phosphorylation of SPOP and YBX1 in other types of cancer cells known to have overexpression in SPOP and/or YBX1.

L'autisme et la déficience mentale (DM) sont des pathologies neurodéveloppementales fréquentes qui partagent des facteurs génétiques communs. Afin de mieux comprendre leur étiologie, nous avons étudié les mécanismes moléculaires de la neurofibromatose de type 1 (NF1), qui est souvent associée à l'autisme et à la DM. La neurofibromine, dont le gène est muté dans la NF1 interagit avec LIMK2. Cette protéine fait partie de la voie des Rho-GTPases dont des mutations de plusieurs membres ont été trouvés mutés dans des cas d'autisme et de DM. Chez le rat, nous avons montré que l’expression de Limk2d, une isoforme sans domaine kinase, augmente la croissance des neurites des cellules neuronales NSC-34. Chez l'homme, LIMK2-1 est la seule isoforme qui comporte un domaine inhibiteur de la phosphatase 1 (PP1i). Nous avons montré que l’expression de cette protéine diminue la longueur des neurites des cellules NSC-34 in vitro. Nous avons observé l'association de la variation située dans le domaine PP1i à la DM (p.S668P, rs151191437) (p=0,04, test de Fisher, OR = 3,29). Elle abolit l’effet inhibiteur de croissance des neurites de l'isoforme LIMK2-1 diminue l'interaction de LIMK2-1 avec la neurofibromine. La fréquence de l'autisme est plus élevée chez les patients atteints ayant des délétions de 14 gènes du locus NF1. Nous avons observé une association entre la variation R115K (rs111902263) du gène RNF135 de ce locus et l'autisme (p=0,00014, test de Fisher) ainsi qu’une anomalie du nombre de copies située dans l'intron 2 de ce gène chez un d’entre eux. Ce travail souligne la spécificité de deux isoformes de LIMK2 sur la croissance des neurites. Il renforce l’intérêt d’étudier l’implication du gène RNF135 dans l’autisme. Des études fonctionnelles seront entreprises afin de confirmer le rôle de LIMK2 et de RNF135 dans l'étiologie de l'autisme et de la DM. / Autism and mental deficiency (MD) are two neurodevelopemental diseases which share genetic factors in common. To better understand their etiologies, we studied the molecular mechanisms of neurofibromatosis type 1, a pathology frequently associated with autism and MD. Neurofibromatosis type 1 is due to deletions or mutations of the NF1 gene which encodes neurofibromin. This protein interacts with several proteins such as LIMK2. This protein belongs to the Rho-GTPases pathway in wich mutations of numerous members have been associated with autism and MD. In our study, we showed that LIMK2 isoforms do not only have important structural differencies but have also functional specificities. Limk2d, which lacks the kinase domain, promotes neurite outgrowth of NSC-34 cells. On the contrary, LIMK2-1, which is primate specific and has a C-terminal PP1i domain, inhibits neurite outgrowth. Analysis of the LIMK2-1 coding sequence, revealed the association between MD and a variation located in the PP1i domain, S668P (rs151191437) (p=0.04, Fisher test, OR = 3.29). This variation abrogated the LIMK2-1 effect on neurite outgrowth and inhibited LIMK2-1 interaction with neurofibromin. Deletions occuring in neurofibromatosis type 1 which include the NF1 gene and 13 others are associated with a higher frequency of autism. Mutations of one of them, RNF135, have been identified in patients with MD and overgrowth syndrome. Two of these patients also presented autistic features. By analysing RNF135 gene in autistic patients, we showed the association of the variation R115K (rs111902263) with autism. We also identified a duplication of a region located in RNF135 gene intron 2 in one patient presenting autism and MD. Our results highlight the importance and specificity of LIMK2 isoforms on neurite outgrowth and strengthen the importance to analyze both the sequence and copy-number of RNF135 gene. Further functional experiments will be undertaken to confirm the implication of LIMK2 and RNF135 in autism and MD etiology.

Sulfonamide derivatives have been designed and synthesized for LIMK2 inhibitors. The inhibition activities of the sulfonamide derivatives were tested against LIMK2. The best sulfonamide derivative was three times more potent than the best sulfonamide inhibitors previously published based on MTT assay results.<br>Two LIMK2 single mutants, LIMK2-365 and LIMK2-549, and a LIMK2 double mutant (LIMK2-DM) were cloned, and their phosphorylation activities were measured and compared against normal LIMK2. LIMK2-365, LIMK2-549, and LIMK2-DM all showing lower activities than the normal LIMK2, with the LIMK2-DM showing even lower activity than each of the single mutant LIMK2, suggesting the importance of these amino acid residues in phosphorylation and in the formation of dimer LIMK2.<p></p>

<p dir="ltr">UBE2C is upregulated in castration-resistant prostate cancer and shows strong correlation with high tumor grade. Currently, the scarcity of UBE2C inhibitors is alarming. This study is the first to report UBE2C post-translational modulation mediated by LIMK2 kinase. A proteome-wide screen previously conducted in the Shah lab has identified UBE2C as a direct substrate of LIMK2 using an innovative chemical genetic approach. LIMK2 regulates UBE2C in a variety of ways. First, LIMK2 directly associates with UBE2C in cells. Second, LIMK2 phosphorylates UBE2C at S123 and increases its stability at the protein level. Third, LIMK2 upregulates UBE2C mRNA and protein expression levels in cells. Contrary to its well-established function as an enzyme involved in the ubiquitin-proteosome pathway, UBE2C stabilizes LIMK2 protein expression in a reciprocal loop. This study is the first to show UBE2C stabilizing its substrate. Likewise, UBE2C increases LIMK2 mRNA and protein levels; however, the mechanism is to be elucidated. LIMK2-UBE2C loop is extremely oncogenic creating CRPC pathogenesis <i>in vivo</i>. Targeting LIMK2 is a suitable approach to effectively degrade both UBE2C and LIMK2 which leads to significant inhibition of tumor formation, cancer stem cell phenotype and epithelial to mesenchymal transition <i>in vivo</i>. Additionally, CDK1 for the longest time was thought to be the only protein of the cyclin dependent kinase family which binds to and is activated by cyclin B1 to regulate cell cycle progression. We first showed CDK5 activity in cell division and its importance in maintaining mitotic fidelity. We first established the activation of CDK5 by cyclin B1 <i>in vitro</i>. The phospho-mimetic CDK5 was observed to be less active when bound to cyclin B1 than its wild-type counterpart.</p>