Engineering an improved adenovirus packaging cell line

Recombinant human adenovirus (Ad) vectors are being extensively explored for their use in gene therapy and recombinant vaccines. Ad vectors are attractive for many reasons, including the fact that (1) they are relatively safe, based on their use as live oral vaccines, (2) they can accept large transgene inserts, (3) they can infect dividing and postmitotic cells, and (4) they can be produced to high titers. However, there are also a number of major problems associated with Ad vectors, including transient foreign gene expression due to host cellular immune responses, problems with humoral immunity, and the creation of replication competent adenoviruses (RCA). Most Ad vectors contain deletions in the E1 region that allow for insertion of a transgene. However, the E1 gene products are required for replication and thus must be supplied in trans by a helper ceillille that will allow for the growth and packaging of the defective virus. For this purpose the 293 cell line (Graham et al., 1977) is used most often; however, homologous recombination between the vector and the cell line often results in the generation of RCA. The presence of RCA in batches of adenoviral vectors for clinical use is a safety risk because tlley . may result in the mobilization and spread of the replication-defective vector viruses, and in significant tissue damage and pathogenicity. The present research focused on the alteration of the 293 cell line such that RCA formation can be eliminated. The strategy to modify the 293 cells involved the removal of the first 380 bp of the adenovirus genome through the process of homologous recombination. The first step towards this goal involved identifying and cloning the left-end cellular-viral jUl1ction from 293 cells to assemble sequences required for homologous recombination. Polymerase chain reaction (PCR) was performed to clone the junction, and the clone was verified through sequencing. The plasn1id PAM2 was then constructed, which served as the targeting cassette used to modify the 293 cells. The cassette consisted of (1) the cellular-viral junction as the left-end region of homology, (2) the neo gene to use for positive selection upon tranfection into 293 cells, (3) the adenoviral genome from bp 380 to bp 3438 as the right-end region of homology, and (4) the HSV-tk gene to use for negative selection. The plasmid PAM2 was linearized to produce a double strand break outside the region of homology, and transfected into 293 cells using the calcium-phosphate technique. Cells were first selected for their resistance to the drug G418, and subsequently for their resistance to the drug Gancyclovir (GANC). From 17 transfections, 100 pools of G418f and GANCf cells were picked using cloning lings and expanded for screening. Genomic DNA was isolated from the pools and screened for the presence of the 380 bps using PCR. Ten of the most promising pools were diluted to single cells and expanded in order to isolate homogeneous cell lines. From these, an additional 100 G41Sf and GANef foci were screened. These preliminary screening results appear promising for the detection of the desired cell line. Future work would include further cloning and purification of the promising cell lines that have potentially undergone homologous recombination, in order to isolate a homogeneous cell line of interest.

Hyperspectral Mineral Identification using SVM and SOM

Remote sensing techniques involving hyperspectral imagery have applications in a number of sciences that study some aspects of the surface of the planet. The analysis of hyperspectral images is complex because of the large amount of information involved and the noise within that data. Investigating images with regard to identify minerals, rocks, vegetation and other materials is an application of hyperspectral remote sensing in the earth sciences. This thesis evaluates the performance of two classification and clustering techniques on hyperspectral images for mineral identification. Support Vector Machines (SVM) and Self-Organizing Maps (SOM) are applied as classification and clustering techniques, respectively. Principal Component Analysis (PCA) is used to prepare the data to be analyzed. The purpose of using PCA is to reduce the amount of data that needs to be processed by identifying the most important components within the data. A well-studied dataset from Cuprite, Nevada and a dataset of more complex data from Baffin Island were used to assess the performance of these techniques. The main goal of this research study is to evaluate the advantage of training a classifier based on a small amount of data compared to an unsupervised method. Determining the effect of feature extraction on the accuracy of the clustering and classification method is another goal of this research. This thesis concludes that using PCA increases the learning accuracy, and especially so in classification. SVM classifies Cuprite data with a high precision and the SOM challenges SVM on datasets with high level of noise (like Baffin Island).

Phonon spectra and temperature variation of thermodynamic properties of fcc metals via Finnis-Sinclair type many body potentials: Sutton-Chen and improved Sutton-Chen models

Volume(density)-independent pair-potentials cannot describe metallic cohesion adequately as the presence of the free electron gas renders the total energy strongly dependent on the electron density. The embedded atom method (EAM) addresses this issue by replacing part of the total energy with an explicitly density-dependent term called the embedding function. Finnis and Sinclair proposed a model where the embedding function is taken to be proportional to the square root of the electron density. Models of this type are known as Finnis-Sinclair many body potentials. In this work we study a particular parametrization of the Finnis-Sinclair type potential, called the "Sutton-Chen" model, and a later version, called the "Quantum Sutton-Chen" model, to study the phonon spectra and the temperature variation thermodynamic properties of fcc metals. Both models give poor results for thermal expansion, which can be traced to rapid softening of transverse phonon frequencies with increasing lattice parameter. We identify the power law decay of the electron density with distance assumed by the model as the main cause of this behaviour and show that an exponentially decaying form of charge density improves the results significantly. Results for Sutton-Chen and our improved version of Sutton-Chen models are compared for four fcc metals: Cu, Ag, Au and Pt. The calculated properties are the phonon spectra, thermal expansion coefficient, isobaric heat capacity, adiabatic and isothermal bulk moduli, atomic root-mean-square displacement and Gr\"{u}neisen parameter. For the sake of comparison we have also considered two other models where the distance-dependence of the charge density is an exponential multiplied by polynomials. None of these models exhibits the instability against thermal expansion (premature melting) as shown by the Sutton-Chen model. We also present results obtained via pure pair potential models, in order to identify advantages and disadvantages of methods used to obtain the parameters of these potentials.

Isoflavone exposure throughout suckling results in improved adult bone health in mice

Exposure to isoflavones (ISO), abundant in soy protein infant formula, for the first 5 days of life results in higher bone mineral density (BMD),greater trabecular connectivity and higher fracture load of lumbar vertebrae (LV) at adulthood. The effect of lengthening the duration of exposure to ISO on bone development has not been studied. This study determined if providing ISO for the first 21 days of life, which more closely mimics the duration that infants are fed soy protein formula, results in higher BMD, improved bone structure and greater strength in femurs and LV than a 5-day protocol. Female CD-1 mice were randomized to subcutaneous injections of ISO (7 Q1 mg kg/body weight/day) or corn oil from postnatal day 1 to 21. BMD, structure and strength were measured at the femur and LV at 4 months of age, representing young Q2 adulthood. At the LV, exposure to ISO resulted in higher (P,0.05) BMD, trabecular connectivity and fracture load compared with control (CON). Exposure to ISO also resulted in higher (P,0.05) whole femur BMD, higher (P,0.05) bone volume/total volume and Q3 lower (P,0.05) trabecular separation at the femur neck, as well as greater (P,0.05) fracture load at femur midpoint and femur neck compared with the CON group. Exposure to ISO throughout suckling has favorable effects on LV outcomes, and, unlike previous studies using 5-day exposure to ISO, femur outcomes are also improved. Duration of exposure should be considered when using the CD-1 mouse to model the effect of early life exposure of infants to ISO.

Apprentissage de représentations sur-complètes par entraînement d’auto-encodeurs

Les avancés dans le domaine de l’intelligence artificielle, permettent à des systèmes informatiques de résoudre des tâches de plus en plus complexes liées par exemple à la vision, à la compréhension de signaux sonores ou au traitement de la langue. Parmi les modèles existants, on retrouve les Réseaux de Neurones Artificiels (RNA), dont la popularité a fait un grand bond en avant avec la découverte de Hinton et al. [22], soit l’utilisation de Machines de Boltzmann Restreintes (RBM) pour un pré-entraînement non-supervisé couche après couche, facilitant grandement l’entraînement supervisé du réseau à plusieurs couches cachées (DBN), entraînement qui s’avérait jusqu’alors très difficile à réussir. Depuis cette découverte, des chercheurs ont étudié l’efficacité de nouvelles stratégies de pré-entraînement, telles que l’empilement d’auto-encodeurs traditionnels(SAE) [5, 38], et l’empilement d’auto-encodeur débruiteur (SDAE) [44]. C’est dans ce contexte qu’a débuté la présente étude. Après un bref passage en revue des notions de base du domaine de l’apprentissage machine et des méthodes de pré-entraînement employées jusqu’à présent avec les modules RBM, AE et DAE, nous avons approfondi notre compréhension du pré-entraînement de type SDAE, exploré ses différentes propriétés et étudié des variantes de SDAE comme stratégie d’initialisation d’architecture profonde. Nous avons ainsi pu, entre autres choses, mettre en lumière l’influence du niveau de bruit, du nombre de couches et du nombre d’unités cachées sur l’erreur de généralisation du SDAE. Nous avons constaté une amélioration de la performance sur la tâche supervisée avec l’utilisation des bruits poivre et sel (PS) et gaussien (GS), bruits s’avérant mieux justifiés que celui utilisé jusqu’à présent, soit le masque à zéro (MN). De plus, nous avons démontré que la performance profitait d’une emphase imposée sur la reconstruction des données corrompues durant l’entraînement des différents DAE. Nos travaux ont aussi permis de révéler que le DAE était en mesure d’apprendre, sur des images naturelles, des filtres semblables à ceux retrouvés dans les cellules V1 du cortex visuel, soit des filtres détecteurs de bordures. Nous aurons par ailleurs pu montrer que les représentations apprises du SDAE, composées des caractéristiques ainsi extraites, s’avéraient fort utiles à l’apprentissage d’une machine à vecteurs de support (SVM) linéaire ou à noyau gaussien, améliorant grandement sa performance de généralisation. Aussi, nous aurons observé que similairement au DBN, et contrairement au SAE, le SDAE possédait une bonne capacité en tant que modèle générateur. Nous avons également ouvert la porte à de nouvelles stratégies de pré-entraînement et découvert le potentiel de l’une d’entre elles, soit l’empilement d’auto-encodeurs rebruiteurs (SRAE).

A study of chymotrypsin immobilization conditions for improved peptide mapping

La cartographie peptidique est une technique de grande importance utilisée lors de l’identification des protéines et la caractérisation des modifications post-traductionnelles des protéines. Deux méthodes sont utilisées afin de couper les protéines en peptides pour la cartographie : les méthodes chimiques et les méthodes enzymatiques. Dans ce projet, l’enzyme chymotrypsine a été utilisée pour l’hydrolyse (la digestion) des liens peptidiques. Cependant, l’autoprotéolyse des enzymes peut augmenter la complexité des échantillons, rendant ainsi ardue l’obtention de pics résolus suite à l’apparition de pics non-désirés dans la carte peptidique. Par conséquent, nous avons utilisé la réticulation des enzymes protéolytiques par réaction avec le glutaraldéhyde (GA) donnant une enzyme insoluble afin de réduire l’autoprotéolyse. L’immobilisation de la chymotrypsine par GA a été effectuée selon une méthode rapportée précédemment par le groupe Waldron. L’électrophorèse capillaire (CE) couplée à l’absorption UV-visible a été utilisée pour la séparation et la détection de peptides et pour obtenir ainsi une cartographie peptidique. Deux tampons différents ont été évalués afin d’obtenir les meilleures conditions pour la digestion de substrats protéiques par la chymotrypsine libre (soluble) ou la GAchymotrypsine et l’analyse par CE. Les cartes des peptides autoprotéolytiques ont été comparées entre les deux formats de chymotrypsine. Afin d’améliorer la cartographie peptidique, nous avons évalué trois méthodes de conditionnement du capillaire CE et deux méthodes pour stopper la digestion. Le bicarbonate d’ammonium s’est avéré être le tampon optimal pour la digestion en solution et l’utilisation d’un bain d’acétone et de glace sèche s’est avérée être la méthode optimale pour stopper la digestion. Une solution de SDS, 25 mM, dans l’étape de rinçage a été utilisée après chaque analyse CE et a permis d’améliorer la résolution des cartes peptidiques. La comparaison entre l’autoprotéolyse de la chymotrypsine libre et de celle immobilisé par GA a été effectuée par des tests utilisant une gamme de six différentes combinaisons de conditions afin d’évaluer le temps (30 et 240 min) et la température de digestion (4, 24 et 37°C). Dans ces conditions, nos résultats ont confirmé que le GA-chymotrypsine réduit l’autoprotéolyse par rapport à l’enzyme libre. La digestion (à 37°C/240 min) de deux substrats modèles par la chymotrypsine libre et immobilisée en fonction de la température de dénaturation du substrat a été étudiée. iii Avant la digestion, les substrats (l’albumine de sérum bovine, BSA, et la myoglobine) ont été dénaturés par chauffage pendant 45 min à trois températures différentes (60, 75 et 90°C). Les résultats ont démontré que la dénaturation par chauffage du BSA et de la myoglobine n’a pas amélioré la cartographie peptidique pour la GA-chymotrypsine, tandis que la digestion de ceux-ci en présence de la chymotrypsine libre a amélioré de façon quantifiable à des températures élevées. Ainsi, le chauffage du substrat à 90°C avec l’enzyme soluble facilite le dépliement partiel du substrat et sa digestion limitée, ce qui a été mieux pour la myoglobine que pour la BSA.

Non invasive classification system of scoliosis curve types using least-squares support vector machines

Objective To determine scoliosis curve types using non invasive surface acquisition, without prior knowledge from X-ray data. Methods Classification of scoliosis deformities according to curve type is used in the clinical management of scoliotic patients. In this work, we propose a robust system that can determine the scoliosis curve type from non invasive acquisition of the 3D back surface of the patients. The 3D image of the surface of the trunk is divided into patches and local geometric descriptors characterizing the back surface are computed from each patch and constitute the features. We reduce the dimensionality by using principal component analysis and retain 53 components using an overlap criterion combined with the total variance in the observed variables. In this work, a multi-class classifier is built with least-squares support vector machines (LS-SVM). The original LS-SVM formulation was modified by weighting the positive and negative samples differently and a new kernel was designed in order to achieve a robust classifier. The proposed system is validated using data from 165 patients with different scoliosis curve types. The results of our non invasive classification were compared with those obtained by an expert using X-ray images. Results The average rate of successful classification was computed using a leave-one-out cross-validation procedure. The overall accuracy of the system was 95%. As for the correct classification rates per class, we obtained 96%, 84% and 97% for the thoracic, double major and lumbar/thoracolumbar curve types, respectively. Conclusion This study shows that it is possible to find a relationship between the internal deformity and the back surface deformity in scoliosis with machine learning methods. The proposed system uses non invasive surface acquisition, which is safe for the patient as it involves no radiation. Also, the design of a specific kernel improved classification performance.

Latex Stage Blending of Natural Rubber and Poly(Vinyl Chloride ) for Improved Mechanical Properties

A novel method of blending natural rubber with polyvinylchloride in the latex stage was developed, Dioctyl phthalate (DOP) and Amine terminated natural rubber (ATNR) were used as plasticisers, for improving the mechanical properties of these blends. Properties of the latex stage blends were compared with those of dry blends. Latex stage blends showed superior mechanical properties compared to the blends prepared in the dry state. The ageing resistance, oil resistance and processability were found to be improved by latex stage blending.

Improved Mechanical Properties of NR/EPDM and NR/Butyl Blends by Precuring EPDM and Butyl

Ethylene-propylene-diene rubber (EPDM) and isobutylene-isoprene rubber (IIR) were compounded, precured to a low degree, and then were blended with natural rubber (NR). The compounding ingredients for NR were then added and the final curing was done. NR/ EPDM and NR/IIR blends, prepared using this method, were found to possess much improved mechanical properties as compared to their conventional counterparts. The optimum precuring crosslink density that has to be given to the EPDM and IIR phases has been determined.

New oligomer-bound antioxidants for improved flex crack resistance and ozone resistance

New oligomer-bound antioxidants have been prepared by condensation reaction. The efficiency and permanence of these oligomer- bound paraphenylene diamines as antioxidants has been compared with conventional amine type antioxidants in NR, SBR, IIR and NBR and in elastomer blends like NR/BR and NR/SBR. The oligomer-bound antioxidants are found to impart improved ozone, flex resistance and mechanical properties to the vulcanizates of NR, SBR, IIR and NBR and to blends of NR/BR and NR/ SBR in comparison with those containing conventional antioxidants.

Studies on development of MgB2 superconductor with improved in-field critical current density

National Institute for Interdisciplinary Science and Technology