Submitted by Noeme Timbo (noeme.timbo@metodista.br) on 2017-05-30T18:14:20Z No. of bitstreams: 1 Ricardo Kawai .pdf: 1449796 bytes, checksum: cc2bd093c4529296f07cab636da59c5a (MD5) / Made available in DSpace on 2017-05-30T18:14:20Z (GMT). No. of bitstreams: 1 Ricardo Kawai .pdf: 1449796 bytes, checksum: cc2bd093c4529296f07cab636da59c5a (MD5) Previous issue date: 2017-03-22 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / This study aimed to investigate the influence of corporate governance and three of its internal mechanisms: board of directors, concentration of ownership control and identity of the controlling shareholder, on the management of results in Brazilian listed companies, listed on BM&FBOVESPA. Results quality variables derived from the discretionary accruals determined by two models used as proxy for results management: Kang and Sivaramakrishnan (1995) and McNichols (2002), from 2009 to 2015. As proxy of corporate governance was the corporate governance index prepared by Leal, Carvalhal and Ievorlino (2015) adopted for the year 2015. The comparison between two groups made through non-parametric tests indicating a non-significant relationship between corporate governance and results management, which do not support the theory that corporate governance inhibits results management. In order to verify the relationship between the internal mechanisms of corporate governance and results management, multiple linear regressions using the Ordinary Least Squares (OLS) method. In which independent variables: board of directors, control concentration and controller identity; Together with the control variables: return on assets, leverage and company size did not present a statistically significant relationship with the corporate governance dependent variable. There was a negative relationship between company size and results management in both models and the variable return on assets presented a negative relation with the McNichols model. The study suggests that the internal mechanisms of corporate governance are not sufficient to inhibit the management of results measured by models that detect them through discretionary accruals. / Este estudo teve como objetivo investigar a influência da governança corporativa e três de seus mecanismos internos: conselho de administração, concentração de controle de propriedade e identidade do controlador, sobre o gerenciamento de resultados nas empresas brasileiras de capital aberto, não financeiras, listadas na BM&FBOVESPA. Foram utilizados como proxy de gerenciamento de resultados as variáveis de qualidade de resultados derivados dos accruals discricionários determinados por dois modelos: o Kang e Sivaramakrishnan (1995) e o McNichols (2002), no período de 2009 a 2015. Como proxy de governança corporativa foi adotado o índice de governança corporativa elaborado por Leal, Carvalhal e Ievorlino (2015), para o ano de 2015. A comparação entre dois grupos foi feita por meio de testes não paramétricos indicando uma relação não significativa entre governança corporativa e gerenciamento de resultados, que não confirmam a teoria de que a governança corporativa inibe o gerenciamento de resultados. Para verificação da relação dos mecanismos internos da governança corporativa com gerenciamento de resultados utilizou-se regressões lineares múltiplas pelo método dos Mínimos Quadrados Ordinários (MQO) em que variáveis independentes: conselho de administração, concentração de controle e identidade do controlador; juntamente com as variáveis de controle: retorno sobre ativo, alavancagem e tamanho da empresa não apresentaram relação estatisticamente significante com a variável dependente da governança corporativa. Houve uma relação negativa entre tamanho da empresa com gerenciamento de resultados nos dois modelos e a variável de retorno sobre ativo apresentou relação negativa com o modelo de McNichols. O estudo sugere que os mecanismos internos da governança corporativa não são suficientes para inibir o gerenciamento de resultados medidos por modelos que os detectam por meio de accruals discricionários.

Dry abrasive blasting is a surface preparation process used in shipyards for cleaning the surfaces of the metal plates to be used in various components of the ship. Commonly used abrasives include sand, steel grit, mineral abrasives, metallic abrasives, and synthetic abrasives. The basic objective of this study was to understand the environmental performance of two abrasives, Steel Grit and Specialty Sand. The project was funded by the Gulf Coast Region Maritime technology Center (GCRMTC) and USEPA. It simulated actual blasting operations conducted at shipyards under enclosed, controlled conditions on plates similar to steel plates commonly blasted at shipyards. The emissions were measured using EPA Source Test Method to quantify particulate emissions. Steel Grit was observed to be more productive, less consuming, and more environmentally friendly compared to Specialty Sand. The findings obtained in this study will be valuable in reducing costs, improving productivity, and protecting the environment.

Glioblastoma is the most common primary brain tumor in humans and has a poor prognosis. Current therapies are not curative. Oncolytic viruses (OVs) are being investigated as tools to induce immunogenic cell death (ICD), cell death capable of activating the immune system. Semliki Forest Virus (SFV) strain 4 is an OV being investigated to treat glioblastoma. Previous studies in our lab have shown that SFV4 can induce ICD in human osteosarcoma (HOS) cells and ongoing in vivo studies show that SFV4 infected GL261 cell vaccination providesprotective immunity in mouse models. This study aimed to characterize the ICD induced by SFV4in glioblastoma cell lines, namely GL261, SB28 and CT2A, and to explain some of our in vivoobservations, namely why vaccination with SFV infected GL261 provides protective immunity but vaccination with infected CT2A and SB28 does not. Our in vitro studies found that GL261 is resistant to SFV4 while SB28 and CT2A are susceptible. We show that the virus can replicate in all three cell lines as seen by the presence of dsRNA, but that viral translation is delayed or inhibited in GL261 cells as not all cells positive for dsRNA were positive for SFV4 protein. Additionally, the type I interferon (IFN) pathway, responsible for antiviral defense, was highly upregulated in CT2A and SB28 but not as much in GL261 after infectionas seen by surveying IFIT1, IFITM3 and IFN-beta genes. Interferon stimulated genes (ISG) like CXCL10, a chemoattractant, was highly upregulated in GL261 after infection and might account for the protective immunity seen in vivo after vaccination. PDL1, an interferon stimulated gene responsible for self-tolerance, was highly upregulated in CT2A after infection. The IFN-beta ELISA revealed that both infected and uninfected GL261 cells produce IFN-beta suggesting a constitutively active pathway. Our DC phagocytosis assay showed that SFV4 infection of CT2A and SB28 cells induced a significant increase in DC phagocytosis and SFV4 infection of all three cell lines significantly increased DC maturation. We conclude that SFV4 infection of GL261 cells may induce ICD in vivo through a persistent viral infection and increased expression of CXCL10.