(理学療法基礎系31)

Les protéines subissent des modifications post-traductionnelles tout au long de leur vieillissement mais aussi dans certains états pathologiques. Parmi ces modifications progressives des protéines, on retrouve notamment la formation des résidus aspartates isomérisés. La protéine L-isoaspartyl méthyltransférase (PIMT) reconnaît et répare des résidus L-isoaspartates anormaux retrouvés dans des protéines. Il a été démontré que la PIMT joue un rôle crucial dans la formation et le maintien du système nerveux central. Ainsi, son fonctionnement s'est avéré important dans le désordre neurologique qu'est l'épilepsie, mais aussi potentiellement dans certaines maladies neurodégénératives comme l'Alzheimer. Notre laboratoire a déjà démontré que l'acide valproïque (VPA), un médicament anticonvulsant, induit l'expression de la PIMT via la voie de signalisation de ERK et également la voie de signalisation glycogènen-synthase-kinase-3 (GSK-3)/β-caténine. Le but de notre recherche a été d'identifier des nouvelles voies de signalisation qui contrôlent l'expression de la PIMT lorsque celle-ci est stimulée par le VPA. Comme modèle cellulaire nous avons utilisé des neuroblastomes SH-SY5Y qui ont été traités avec le VPA pour étudier l'implication de la voie de signalisation phosphatidylinositol 3-kinase (PI3K)/Akt dans l'induction de la PIMT. Nos résultats montrent que lors de l'induction de la PIMT par VPA, la protéine Akt est phosphorylée (Thr 308). De plus, lorsque la protéine PI3K est inhibée par des inhibiteurs pharmacologiques, wortmannin et le LY294002, la phosphorylation de Akt est bloquée et l'induction de la PIMT par VPA est arrêtée. L'inhibition de Akt par un siRNA spécifique produit le même effet. Également, lorsque la voie de signalisation PI3K/Akt est stimulée par le VPA on observe une phosphorylation de la protéine GSK-3 (Ser 21) qui est également observable lorsque les cellules sont traitées avec le lithium, un inhibiteur directe de GSK-3. Finalement, l'inhibition du facteur de transcription CREB avec un siRNA spécifique n'a pas affecté l'induction de la PIMT par VPA. En conclusion, notre étude a démontrée que j'induction de la PIMT par VPA est dépendante de la voie de signalisation PI3K/Akt. L'activation de cette voie de signalisation permet la phosphorylation et donc l'inhibition de la kinase GSK-3, mais l'induction de la PIMT par VPA est indépendante de facteur de transcription CREB. Ces résultats suggèrent plutôt que VPA en inhibant la kinase GSK-3 stabilise la β-caténine permettant ainsi l'expression de la PIMT. ______________________________________________________________________________ MOTS-CLÉS DE L’AUTEUR : PIMT, épilepsie, VPA, PI3K/Akt

The Bcl-2 family of proteins contains both anti-apoptotic (e.g.Bcl-2, Bcl-xL) and pro-apoptotic (e.g.Bad, Bcl-xS) proteins. The Bcl-xL and Bcl-xS are splice variants, but have different functions during apoptosis. The pro-survival kinase Akt can phosphorylate certain Bcl-2-related proteins, specifically on serine residues, to regulate their function and localization. This is an extension of the work from our laboratorys finding that haloperidol induces PC12 cell death by inducing Bcl-xS which then translocates from cytosol to mitochondria where it facilitates the release of cytochrome c. The toxicity induced by Bcl-xS is reversed by expression of constitutively active Akt. I hypothesized that Akt-mediated post-translational modification may be important for regulating the function of Bcl-xS and Bcl-xL.<p> Three specific serine residues were ultimately chosen for the characterization of Bcl-xS/xL function: Ser62 (inactivation mutant), Ser106 (putative Akt phosphorylation motif), and Ser165 in Bcl-xS (and the corresponding Ser228 in Bcl-xL) (immediately upstream of hydrophobic tail). The individual substitution of all three Serines with Alanines (which precludes phosphorylation at that site) in Bcl-xS did not affect the expression of the protein, but they did induce varying degrees of cytotoxicity in both PC12 and HEK cultures. I focused on the Ser106 substitution mutant given my hypothesis that Akt targeted this site. Overexpression of Bcl-xS(S106A) was toxic in both PC12 and HEK cultures, as expected, and this coincided with the appearance of the Bcl-xS(S106A) protein in the mitochondrial fraction. The release of cytochrome c from PC12 cell mitochondria coincided with the co-immunoprecipitation of the Bcl-xS protein with VDAC (voltage-dependent anion channel), a channel-forming protein that is known to mediate cytochrome c release, and with the initiation of caspase-dependent events. This was not the case in HEK cells, where the mitochondrial VDAC seemed to be diminished and the toxicity was cytochrome c-independent as well as caspase-independent. In addition, I was able to demonstrate that the S106A substituted protein was not able to co-immunoprecipitate with Akt, supporting Ser106 as a potential target for the Akt protein. I then studied the effects of the homologous substitutions in Bcl-xL on cell function. I chose to use treatment with the potent inducer of apoptosis, staurosporine, as a model of cytotoxicity. Again, substituted proteins exerted toxicity, but they did not potentiate the effects of staurosporine, at least not on MTT conversion. I did notice, however, that there was a clear morphological change with certain concentrations of staurosporine, and subsequently demonstrated that the Bcl-xL(S106A) protein potentiated PC12 cell differentiation induced by staurosporine. This protein also co-immunoprecipitated better with Akt, which was unexpected given my results with the Bcl-xS(S106A) protein described above. Perhaps the extra amino acids in Bcl-xL account for this.<p> It is clear that the phosphorylation of Bcl-xS and Bcl-xL proteins is an important means of regulating their function and localization within the cell. These data support the S106 residues in both Bcl-xS and Bcl-xL as novel targets for the pro-survival Akt kinase, and indicate a role for this/these residue(s) in cellular functions as diverse as apoptosis and differentiation.

Activated receptor tyrosine kinase (RTK), such as the epidermal growth factor (EGF) receptor (EGFR) and the platelet-derived growth factor (PDGF) receptor (PDGFR), recruit downstream signalling proteins, including phosphatidylinositol 3-kinase (PI3K). PI3K, composed of a regulatory p85 subunit and a catalytic p110 subunit, phosphorylates phosphatidylinositol 4,5-bisphosphate at the 3 position to generate phosphatidylinositol 3,4,5-trisphosphate. This lipid second messenger activates Akt, which promotes cell growth, cell cycle entry and progression, as well as cell survival and cellular migration. PTEN, a tumor suppressor protein, dephosphorylates phosphatidylinositol 3,4,5-trisphosphate at the 3 position, turning off Akt signalling. PTEN contains a C-terminal PDZ binding motif that binds to the PDZ2 domain of MAGI-2, a scaffolding protein that localizes signalling molecules to the plasma membrane. MAGI-2 has ten domains that potentially mediate multiple protein-protein interactions simultaneously. A PTEN associated-complex (PAC) has been described and may contain MAGI-2, PTEN and p85. The PAC is hypothesized to form at the plasma membrane at appropriate sites for PTEN to gain access to its lipid substrates, since the binding of PTEN to MAGI-2 has been shown to enhance the suppression of PI3K-mediated Akt signalling. In order to better understand the role of the PAC in attenuation of the Akt signalling pathway, regions of the MAGI-2 scaffolding protein were mapped to identify the interactions taking place in the PAC. MAGI-2, and its individual domains, were expressed as GST fusion proteins. These were immobilized onto beads and allowed to bind to cellular proteins including PTEN, p85, PDGFR and EGFR using a GST pull-down experiment. The proteins bound to GST-MAGI-2 were identified using an immunoblot analysis. In vitro pull-down experiments revealed that MAGI-2 PDZ2 and PDZ5 domains bind to PTEN, and both MAGI-2 WW domains were shown to bind to p85. EGFR and PDGFR did not bind to the PDZ domains of MAGI-2 under the conditions studied. In order to study protein-protein interactions in cells, immunoprecipitation assays were also performed. Full length MAGI-2 was expressed tagged to a Myc epitope. This was used in immunoprecipitation assays to determine if MAGI-2 could co-immunoprecipitate with proteins involved in the Akt signalling pathway, such as PTEN, p85, PDGFR and EGFR. MAGI-2 can co-immunoprecipitate with PTEN upon 5 min EGF stimulation however, this result was inconclusive because replicate experiments did not verify this initial observation. MAGI-2 does not co-immunoprecipitate with the EGFR nor p85, under the conditions tested. We examined for these interactions after 5 min of growth factor stimulation and more experiments that test different time points after growth factor stimulation would reveal if these interactions are present at shorter time points. MAGI-2 has been shown to bind to PTEN and p85 in vitro and therefore has the potential to regulate the attenuation of the PI3K/Akt signalling pathway in response to activated EGFR and/or PDGFR.

Neuroblastoma is one of the most aggressive cancers and has a complex form of differentiation. We hypothesized that the advanced cellular differentiation may alter the susceptibility of neuroblastoma to photodynamic therapy (PDT) and have a selective survival advantage. We compared the photocytotoxicity treated by Hematoporphyrin (Hp) for PDT on human neuroblastoma SH-SY5Y cells with retinoic acid (RA)-differentiated SH-SY5Y cells. The undifferentiated neuroblastoma cells were shown to cause elevated photocytotoxic effect by MTT assay and also confirmed by Annexin V-FITC/PI staining. In undifferentiated cells, Hp-PDT increased the generation of intracellular reactive oxygen species (ROS), the loss of mitochondrial membrane potential, characteristic chromatin condensation displaying, PARP cleavage, the downregulated expression of Bcl-2, and the activation of caspase-9, -3 was more significant than that of the differentiated cells. In undifferentiated SH-SY5Y cells, cell cycle arrest at G2/M phase was accompanied by the decrease in cyclin B1 level, and could be reversed by the disruption of intracellular ROS caused by PDT. Furthermore, the ROS scavenger markedly inhibited Hp-PDT induced activation of caspase-3, a sustained phosphorylation of Akt/GSK-3£] and ERK, and cytotoxicity in undifferentiated SH-SY5Y cells, but not in differentiated SH-SY5Y cells. Blockage of p38 and JNK activation can significantly attenuate PDT-induced viability loss in both SH-SY5Y cells, but the less significant activation of p38 and JNK, as well as more significant phosphorylation of Akt and GSK-3£], and a prolonged ERK activation appeared to make differentiated SH-SY5Y cells more resistant to photocytotoxicity. Collectively, these data suggested that differentiated SH-SY5Y cells were more resistant to PDT induced apoptosis than undifferentiated SH-SY5Y cells, and ROS played the most important regulatory role on the susceptibility to Hp-PDT between undifferentiated and differentiated neuroblastoma cells. These results may have important implications for neuroblastoma patients undergoing PDT.

ROGDI, a novel gene, locates on human¡¦s chromosome 16p13.3. According to Gene Ontology Annotation database, ROGDI is related to hemopoiesis and positive regulation of cell proliferation. In order to investigate the function of this novel gene in pulmonary fibrosis, fibrotic models in vivo and in vitro were created. Mice which received single intra-tracheal bleomycin injection were sacrificed on various intervals. Rogdi and other pro-fibrotic mediators, including CCL2 and TGF-£]1, were up-regulated in the early phase(< 10 days). On contrary, the anti-fibrotic mediators IL-10, IFN-£^ and heme oxygenase(HO)-1 were up-regulated in the late phase(> 10 days). The precursor microRNA 21 (miR-21) was up-regulated as the fibrotic severity increased. The human embryonic fibroblasts(WI-38 cells) showed fibrogenic phenotype and up-regulation of precursor miR-21 and ROGDI after bleomycin treatment. Human embryonic fibroblasts transfected by coding sequence of ROGDI showed up-regulated precursor miR-21 and £\-SMA compared to those transfected by empty vectors after bleomycin treatment. Two signaling molecules related to positive regulation of cell proliferation, Akt and Erk, showed over-expressed after ROGDI transfection and bleomycin treatment compared to those with empty vector transfection. Our results imply that ROGDI is up-regulated in pulmonary fibrosis and turns fibroblasts into fibrogenic phenotype through positive regulation of miR-21. The increase of precursor, but not primary miR-21, after ROGDI transfection and bleomycin treatment indicates that ROGDI may regulate the TGF-£] signaling pathway in human embryonic fibroblasts. Our results support that ROGDI is a novel gene for pulmonary fibrosis and warrants for further investigation. £[

名古屋大学博士学位論文 学位の種類:博士(リハビリテーション療法学) (課程) 学位授与年月日　平成22年3月25日

LRBA (LPS Responsive Beige-like Protein Kinase A anchor) gene expression is induced by the mitogen LPS and is a member of the WBW gene family member which is comprised of genes that are involved in cellular proliferation and differentiation. This work provides evidence for the over-expression of LRBA in certain cancers, and that LRBA promoter activity and endogenous LRBA mRNA levels are negatively regulated by the tumor suppressor p53 and positively regulated by E2F transactivators. Furthermore, we demonstrate that inhibition of LRBA expression or function leads to decreased proliferation of cancer cells and that LRBA plays a role in the EGFR signal transduction pathway. In addition to the findings of LRBA's role in carcinogenesis, this work also shows evidence of the knockdown of the SH2-containing Inositol 5' Phosphatase (SHIP) in both mouse and human cells. Furthermore, we provide evidence that SHIP-1 is involved in the AKT signal transduction pathway in human Natural Killer cells.

PKC βII is alternatively spliced during acute insulin stimulation in L6 skeletal muscle cells. This PKC βII isoform is critical in propagating GLUT4 translocation. PKC β protein and promoter dysfunction correlate with human insulin resistance. TZD treatment ameliorates whole-body insulin-resistance. Its primary target is adipocyte PPAR γ, which it activates upon binding. This causes both altered circulating serum FFA concentrations and adipokine secretion profile. How TZDs affect the intracellular signaling of skeletal muscle cells is unknown. RT-PCR and Western blot analysis showed that TZDs elevated PKC βII by a process that involves co-transcriptional splicing. PGC1 α overexpression most closely resembled TZD treatment by increasing PKCβII protein levels and keeping PKC βI levels relatively constant. Use of a heterologous PKCβ promoter driven PKC β minigene demonstrated that PPARγ could regulate the PKCβ promoter, but whether this is direct or indirect is unclear. SRp40 splicing factor has been shown to dock onto the PGC1 α CTD and influence splicing. SRp40, through overexpression and silencing, appears to play a part in PKC β promoter regulation. PKC β promoter regulation was also studied in 3T3-L1 cells. TZDs were experimentally shown to have no role in PKC β promoter regulation despite PPARγ activation. Chromatin immunoprecipitation assays revealed PU.1 as a putative PKC β transcription factor that can cross-talk with the spliceosome, possibly through SRp40 which was also associated with the PKC β promoter. 3T3-L1 adipocyte differentiation revealed a novel developmentally-regulated switch from PKC βI to PKCβ II, using western blot and Real-Time PCR analysis. Pharmacological inhibition of PKC β II using CGP53353 and LY379196 blocked [ 3 H]2-deoxyglucose uptake and revealed a functional role for PKC β II in adipocyte ISGT. CGP53353 specifically inhibited phosphorylation of PKC β II Serine 660 and not other critical upstream components of the insulin signaling pathway. Subcellular fractionation and PM sheet assay pointed to PKC β II-mediated regulation of GLUT4 translocation to the PM. Co-immunoprecipitation between PKC β II and GLUT4 allude to possible direct interaction. Western blot and immunofluorescence assays show PKC β II activity is linked with Akt Serine 473 phosphorylation, thus full Akt activity. Western blot and co-immunoprecipitation suggested that insulin caused active mTORC2 to directly activate PKC βII. Data support a model whereby PKCβ II is downstream of mTORC2 yet upstream of Akt, thereby regulating GLUT4 translocation.

Melanoma manifests itself from the malignant transformation of melanocytes and represents the deadliest form of skin cancer, being responsible for the disproportionate majority of all skin cancer deaths. The 2002 discovery that 50% of all melanoma patients possess activating BRAF mutations ignited a significant paradigm shift in the way the melanoma field approached research and how patients were treated [1]. The era of targeted therapy had begun and with it came successful targeted BRAF inhibitor therapy regimens, which have accomplished improved clinical benefit (response rate, progression free survival, and overall survival) compared with treatment with chemotherapy in three phase III clinical trials [2]. Although there has been much success in the subgroup of patients whose melanomas harbor activating BRAF mutations, approximately 50% of all melanoma patients do not harbor BRAF mutations. This subgroup of melanoma is composed of ~15-20% of all patients with NRAS mutations and another ~25-30% of patients with neither BRAF nor NRAS mutations. Successful targeted treatment strategies are currently lacking for this subgroup of BRAF-wild type melanomas and therefore novel targeted therapeutic modalities are urgently needed. The work described in this dissertation sheds light on potential approaches for the treatment of BRAF wild type melanoma and will be split into three separate strategies. The first will focus upon the treatment of melanomas without BRAF or NRAS mutations (BRAF/NRAS wild type melanoma) and will expand upon a clinical observation where two melanoma patients were treated with an experimental combination of carboplatin and paclitaxel, with the addition of the AKT inhibitor MK-2206. We demonstrate that the inhibition of AKT significantly enhances the efficacy of chemotherapy in a reactive oxygen species (ROS) mediated fashion, and an induction of autophagy plays a cyto-protective role. The second story focuses upon the treatment of NRAS mutant melanomas by investigating resistance mechanisms to MEK inhibitor treatment. We discovered a MEKi-mediated induction of receptor tyrosine kinase (RTK) signaling to serve as a significant mechanism of escape for NRAS mutant melanomas treated chronically with the MEK inhibitor AZD6244, as well as the recently U.S. Food and Drug Administration (FDA) approved MEK inhibitor trametinib. Novel targeted therapy combinations were then added to overcome the escape from MEK inhibitor therapy. Co-targeting of the receptor tyrosine kinases AXL, PDGFR-β and c-MET with a pan-RTK inhibitor, as well as the mitogen-activated protein kinase (MAPK) pathway with a MEK inhibitor greatly enhanced treatment-induced apoptosis and inhibition of proliferation. The final strategy builds upon the observation that single agent MEK-inhibition is largely ineffective in the treatment of NRAS mutant melanomas. A recovery of MAPK pathway activity in response to MEK inhibition was established to play a significant role in escape of NRAS mutant cells from cell cycle arrest and apoptosis. The combination of a MEK inhibitor with the novel ERK inhibitor VTX-11e prevents the onset of resistant clones and enhances cytotoxicity of the NRAS mutant melanoma cells. This body of work establishes original targeted therapy combinations for the treatment of both NRAS mutant melanomas and BRAF/NRAS wild type melanomas. We propose future clinical investigation with these strategies in the treatment of BRAF wild type melanoma patients in hopes to further extend overall survival.