Browsing by Author "Lewis, John D."
Now showing 1 - 3 of 3
Results Per Page
Sort Options
- ItemOpen AccessREPORT ON AAAI-91(1991-08-01) Lewis, John D.This is a report on the aspects of AAAI-91 which were particularly applicable to my research. Included is a brief description of the problem my thesis attacks--understanding instruction during task learning--and a summary of a variety of papers which appear in the proceedings, as well as one of the invited talks. The papers which will be summarized are from the following sections of the proceedings: generation and interaction, syntax and semantics, formalisms for coordination, temporal constraints, tractable inference, and an invited talk on discourse understanding. A discussion of how these ideas relate to the problem of understanding instruction is incorporated into each summary, as well as forming a conclusion for the report.
- ItemEmbargoThe Role and Regulation of Proteinase Activated Receptors (PARs) in Urothelial Carcinoma(2020-01-24) de Lima, Stacy Grace; Hollenberg, Morley D.; Hyndman, Matthew Eric; Zijlstra, Andries; Morris, Don G.; Lewis, John D.; Black, Peter McL.; Riabowol, Karl T.Introduction: Urothelial carcinoma (UC) is the most common type of bladder cancer in the North American population. UC is the most expensive cancer to treat per patient owing to the absence of noninvasive diagnostic and prognostic tests complemented by ineffective treatments and high recurrence rates necessitating long-term follow-up. One phenomenon that has been well demonstrated in UC is that the expression of proteinases and their inhibitors changes in the UC tumor microenvironment as the tissue progresses from healthy urothelium to low grade and then to high grade tumors. This information has not advanced clinical care likely due to the complexity of the proteolytic cascade in both healthy and tumor microenvironments. Proteinase activated receptors (PAR) are specific cell surface receptors that integrate complex proteinase signals into downstream cellular signaling pathways that have been shown to promote survival, proliferation, migration and invasion in carcinomas, but have yet to be explored in UC. Working hypothesis: UC cells produce proteinases that drive an oncogenic (pro-migratory and invasive) phenotype through proteinase activated receptor (PAR) 1 and 2 induced cell signaling which changes with grade, stage and prognosis of disease. Methods: The hypothesis was tested in 5 bladder and UC-derived cell lines of increasing aggressiveness (PC3-420-010, RT4, TCCSUP, 5637, and T24) with the postulation cell aggressiveness (i.e. migration and invasion capacity) would correlated with proteinase and PAR expression and signaling sensitivity. In vitro testing included: PAR expression (qPCR, IHC), PAR signaling (calcium and western blotting for MAP kinase), PAR cleavage (using a variety of published techniques using transfected PAR-indicator constructs in the cell of interest as well as “bystander” responding cells), proteinase expression (proteomics) and activity (fluorescent substrates, PAR-cleavage modeling) testing complemented by organotypic urothelium-fibroblast modeling, as well as immunohistochemical profiling of the expression and cleavage status of PAR1 in patient tumour samples. Results: PAR 1, 2 are expressed by most cell lines tested, but PAR 3, and 4 are not. The expression and sensitivity of these receptors was found to be lowest in healthy urothelium and low-grade papilloma-derived cell lines, and highest in UC cells with increased capacity to migrate and invade. All cell lines produce proteinases and inhibitors, but only high-grade-derived cells produced proteinases that could cleave PAR1 in an autocrine manner. Agonism of PAR1 or PAR2 had little effect on proliferation, but enhanced migration, and invasion in the cell lines that already had this capacity. Only PAR1 antagonism not PAR2 was able to inhibit migration and invasion. Interestingly, in an organotypic model there seemed to be little effect on cellular behaviour (i.e. proliferation, migration, invasion) but robust effects on phenotype with PAR1 targeting upregulating cellular markers associated with a basal phenotype. Patient-derived invasive UC tissue was found to express PAR1. Conclusions: UC-produced proteinases act through PAR1 and promote migration and invasion in UC cell lines in vitro but not in an organotypic model which supported PAR agonism in initiating cell differentiation. UC patient tissue was found to express PAR1 but more data will be necessary to understand the implications of this expression.
- ItemOpen AccessXINTERFACE: A DYNAMICALLY CONFIGURABLE PROCESS LEVEL INTERFACE TO X(1990-09-01) MacDonald, Bruce A.; Lewis, John D.This paper describes XINTERFACE, a process level interface to the X window system. It allows programmers with a minimum knowledge of X to create an interface for new or existing applications, as well as allowing more freedom in choosing a language in which to implement the functional portion of the application. XINTERFACE is a skeleton client that is invoked interactively by a user or by an application, to form a shell process dedicated to maintaining the interface. The initial appearance and behaviour of the interface are specified by a compiled User Interface Language (UIL) file, augmented by command line arguments and default resource files. A separate process is responsible for implementing the functionality of the application. Inter-process communication allows the interface to communicate user interactions and the functional process to reconfigure the interface as needed. This approach extends the advantages of UIL as well as exploiting the superior language support for process communications, over direct window system interaction.