The development of anti-drug antibodies (ADA) against therapeutic proteins can impact upon drug safety and efficacy. This is a major challenge in the development of biotherapeutics. Various factors have the potential to contribute to protein immunogenicity and the production of ADA. Protein aggregation is one of these factors, though the mechanisms underlying aggregate immunogenicity are poorly understood. In this thesis the effect of protein aggregation on immunogenicity has been investigated. The thermal and/or mechanical stresses required in order to achieve subvisible aggregates of three test proteins were determined. Stressed preparations of proteins were characterised using a suite of biophysical techniques, including dynamic light scattering and circular dichroism. The immunogenic potential of subvisible aggregates of a humanised single chain variable fragment (scFv) and ovalbumin (OVA) was studied following intraperitoneal exposure in BALB/c strain mice. Monomeric proteins induced a T helper (Th) 2 dominant immune response, but when aggregated, the responses gained a Th1 phenotype, with a significant increase in the antigen-specific IgG2a antibody response. Cytokine profiles in supernatants taken from splenocyte-dendritic cell co-cultures were also consistent with aggregated preparations of OVA inducing a Th1-type response. Host cell protein (HCP) impurities can also contribute to immunogenicity. Mass spectrometry analysis of an scFv preparation identified the presence of the Escherichia coli (E.coli) heat shock protein DnaK, amongst other HCP, as an impurity. Protein preparations free from DnaK were spiked with recombinant E.coli DnaK to mimic the HCP impurity. The effect of DnaK on the immunogenicity of aggregated and monomeric scFv preparations was then investigated. BALB/c mice were immunised with monomeric and aggregated preparations, with and without E.coli DnaK at 0.1% by mass. Aggregation alone resulted in an enhanced IgG2a antibody response, and the presence of DnaK increased this further. Comparable investigations were also conducted using mouse albumin; here an increase in immunogenicity was observed with protein aggregation, and the presence of DnaK was found to increase the IgG2a response. Collectively, the evidence presented in this thesis shows that aggregation can impact upon the magnitude and character of induced immune responses, and that subvisible aggregation promotes a Th1 immune skewing. Additionally, E.coli HCP DnaK enhances protein aggregate immunogenicity, which indicates that heat shock proteins, as a class of HCP, could have an adjuvant-like effect on biotherapeutic aggregates.

Protein aggregation is a hallmark of a number of neurodegenerative disorders including Alzheimer???s Disease, Huntington???s Disease, and Amyotrophic Lateral Sclerosis. Despite the common occurrence of protein aggregation in disease, the fundamental mechanisms controlling the propensity of a protein to aggregate are not well understood. Over the past decade, one of the most significant advancements in the field of understanding protein aggregation has been the development of several aggregation prediction algorithms. In this study, two separate approaches were used to investigate the detailed molecular mechanisms of protein aggregation. First, a thorough investigation that compared nine protein aggregation prediction techniques was performed. Protein aggregation propensity calculations were performed on wild type and mutant sequences of three diverse proteins including Superoxide Dismutase (SOD), human Acylphosphatase (AcP), and the amyloid beta peptide (A??42). This study presents the first wide-scale comparison of such a large number of prediction algorithms, and additionally provides new information on the ability of the algorithms to successfully predict the experimentally observed aggregation of several mutations of diverse proteins. The algorithms were predominantly developed based on a set of known amyloid-forming proteins and peptides, however, are quite diverse in the way they were designed and the proteins on which they were tested. Interestingly, significant variation was observed when predicting the aggregation propensity of identical sequences by multiple techniques, indicating that the algorithms do not possess a consensus on the primary factors that govern aggregation. Further analyses compared predicted and observed aggregation data for several mutants of the test proteins. The aggregation prediction algorithms predominantly demonstrated poor to moderate correlations with observed aggregation, and the strongest correlations occurred in instances where the test data was used in the development of the algorithms. The general lack of ability of the algorithms to predict the aggregation patterns of more than one test protein suggests that aggregation may be a much more specific process that it is generally attributed to be in that there may be inherently different properties modulating the aggregation mechanisms of different proteins towards varying aggregate structures. The second component of this project was to experimentally examine the role of salt in influencing protein aggregation as a method to elucidate the specific molecular mechanisms controlling protein aggregation pathways. The ALS-causing SOD1 mutation, A4V, in both the oxidized and reduced apo form, was used as a model protein. The role of NaCl and Na2SO4 in mediating protein aggregation was studied using several techniques. While oxidized apo A4V showed very little evidence of aggregation even in the presence of salt, for reduced apo, aggregates readily formed and were promoted by the addition of salt. This finding correlated with the increasing kosmotropic nature of the salt as described by the Hofmeister series. The aggregates formed in the presence of salt contained intermolecular disulphide bonds and demonstrated ANS and ThT binding, indicating aggregates are likely to be largely hydrophobic and possess beta-sheet morphology. Salt promotes protein aggregation in two ways: 1) electrostatic interactions shield protein charges and reduce repulsion between proteins, and 2) specific interactions stabilize various aggregation-prone conformations of the protein. This work is evidence of the important role of salt in influencing protein aggregation and provides a framework for future studies into the complex effects of solution conditions in modulating protein aggregation pathways. Both aspects of this study contribute greatly to furthering the understanding of the molecular mechanisms governing protein aggregation. This is of particular importance to neurodegenerative diseases, where uncovering the factors that modulate the formation of toxic aggregate species is important for disease treatment and prevention. The potential aggregation mechanisms of SOD1, and the contributions it may play in ALS pathogenesis, will be discussed throughout this study.

No description available.

No description available.

No description available.

No description available.

No description available.

Video streaming is an increasingly popular Internet application. However, despite its popularity, real-time video streaming still remains a challenge in many scenarios. Limited home broadband bandwidth and mobile phone 3G bandwidth means many users stream videos at low quality and compromise on their user experience. To overcome this problem, we propose CStream, a system that aggregates bandwidth from multiple co-operating users in a neighborhood environment for better video streaming. CStream exploits the fact that wireless devices have multiple network interfaces and connects co-operating users with a wireless ad-hoc network to aggregate their unused downlink Internet bandwidth to improve video quality. CStream dynamically generates a streaming plan to stream a single video using multiple connections and continuously adapts to changes in the neighborhood and variations in the available bandwidth. We have built CStream and evaluated it on a controlled test-bed of computers with various performance measures. The results show linear increase in throughput and improved video streaming quality as the number of cooperating users in a neighborhood increase.

Neuronal pentraxin with chromo domain (NPCD) belongs to a family of neuronally-expressed pentraxin proteins thought to be involved in synaptic refinement and plasticity. One isoform of Npcd, neuronal pentraxin receptor (NPR), is a type-II transmembrane protein responsible for the clustering of related neuronal pentraxins 1 and 2. However, recently identified cytosolic NPCD isoforms with no known function were discovered through their interaction with the intracellular domain of a receptor protein tyrosine phosphatase PTPRO. PTPRO is a signaling molecule known to be involved in the development of the nervous system. Additionally, upregulated expression of neuronal pentraxins has been implicated in neuronal cytotoxicity and associated with neurodegenerative diseases. Here, we demonstrate that a novel cytosolic NPCD isoform interacts with the BTB-Kelch protein Mayven/KLHL2. This interaction was identified through a yeast two-hybrid screen using the C-terminal pentraxin domain region of NPCD and confirmed through mammalian cell colocalization and co-immunoprecipitation studies. Domain truncation analysis suggests that the kelch domains of Mayven/KLHL2 are responsible for this interaction with NPCD. We also show that Mayven/KLHL2 is capable of interacting with Cullin 3, an integral protein in the Cullin-RING ubiquitin ligase complex. An in-vivo ubiquitylation assay demonstrates that overexpression of Mayven/KLHL2 increases NPCD ubiquitylation, and suggests a novel E3 ubiquitin ligase function of Mayven/KLHL2 with NPCD as its substrate. Furthermore, we observed an increased propensity of overexpressed NPCD to form aggresomes with coexpression of Mayven/KLHL2. As the formation of aggresomes is often associated with protein aggregation and deposition diseases, including a multitude of neurodegenerative diseases, we tested NPCD overexpression and the effects of Mayven/KLHL2 coexpression on neuronal cytotoxicity and apoptosis. Overexpressed NPCD in hippocampal neuron cultures resulted in increased cytotoxicity and apoptosis, further exacerbated by Mayven/KLHL2 coexpression. Our findings report an interaction between NPCD and Mayven/KLHL2, demonstrate a novel role of Mayven/KLHL2 as an E3 ubiquitin ligase, and explore a possible intersection between the ubiquitin-proteasome degradation pathway, neuronal pentraxins, and neurodegenerative disease.

We propose dynamic aggregation of virtual tags in TLB to increase its coverage and improve the overall miss ratio during address translation. Dynamic aggregation exploits both the spatial and temporal locality inherent in most application programs. To support dynamic aggregation, we introduce the use of ternary-CAM (TCAM) cells at the second-level TLB. The modified TLB architecture results in an increase of TLB reach without additional CAM entries. We also adopt bulk prefetching concurrently with aggregation technique to enhance the benefits due to spatial locality. The performance of the proposed TLB architecture is evaluated using SPEC2000 benchmarks concentrating on those that show high data TLB miss ratios. Simulation results indicate a reduction in miss ratios between 59% and 99.99% for all the considered bench-marks except for one benchmark, which has a reduction of 10%. We show that the L2 TLB when enhanced using TCAM cells is an attractive solution to high miss ratios exhibited by applications.