B and T cells comprise the two major arms of the adaptive immune response tasked with clearing and preventing infection; molecular recognition in these cells occurs through antibodies and T cell receptors (TCRs), respectively. Highly successful therapeutics, clinical diagnostics and laboratory tools have been engineered from fragments of these parent molecules. The binding specificity, affinity and biophysical characteristics of these fragments determine their potential applications and resulting efficacies. Thus engineering desired properties into antibody and TCR fragments is a major concern of the multi-billion dollar biopharmaceutical industry. Toward this goal, we (1) designed antibody specificity using a novel computational method, (2) engineered thermoresistant Fabs by phage-based selection and (3) modulated binding kinetics for a single-chain TCR. In the first study, de novo modeling was used to generate libraries of FLAG peptide-binding single-chain antibodies. Phage-based screening identified a dominant design, and activity was confirmed after conversion to soluble Fab format. Bioinformatics analysis revealed potential areas for design process improvement. We present the first experimental validation of this in silico design method, which can be used to guide future antibody specificity engineering efforts. In the second study, the variable heavy chain of a moderately stable EE peptide-binding Fab was subjected to random mutagenesis, and variants were selected for resistance to heat inactivation. Thermoresistant clones where biophysically characterized, and structural analysis of selected mutations suggested general mechanisms of stabilization. Framework mutations conferring thermoresistance can be grafted to other antibodies in future Fab stabilization work. In the third study, TCR fragment binding kinetics for a clonotypic antibody were modulated by varying valence during phage-based selection. Binding affinity and kinetics for representative variants depended on the display format used during selection, and all TCR fragments retained binding to native pMHC antigen. This work demonstrates a general engineering platform for tuning protein-protein interactions. Using a combination of computational design and phage-based screening, we have identified antibodies and TCR fragments with improved binding properties or biophysical characteristics. The optimized variants possess a wider range of potential applications compared to their parent molecules, and we detail engineering methods likely to be useful in the engineering of many other protein-based therapeutics. / text

The development of hybridoma technology nearly forty years ago set the foundation for the use of antibodies in the life sciences. Subsequent advances in recombinant DNA technology have allowed us to adapt antibody genes to various screening systems, greatly increasing the throughput and specialized applications for which these complex biomolecules can be adapted. While selection systems are a powerful tool for discovery and evolution, they can be slow and prone to unintended biases. We see computational approaches as an efficient process for rapid discovery and engineering of antibodies. This is particularly relevant for biodefense and emerging infectious disease applications, for which time is a valuable commodity. In the first chapter of this work, we examine computational protocols for ‘supercharging’ proteins. This process resurfaces the target protein, adding charged moieties to impart specialized functions such as thermoresistance and cell penetration. Current algorithms for resurfacing proteins are static, treating each mutation as an event within a vacuum. The net result is that while several variants can be created, each must be tested experimentally to ensure the resultant protein is functional. In many cases, the designed proteins were severely impaired or incapable of folding. We hypothesize that a more dynamic approach, keeping an eye on energetics and the consequences of mutations will yield a more efficient and robust method for supercharging, successfully adding charges to proteins while minimizing deleterious effects. We continue on this theme applying the successful algorithm to supercharging antibodies for increased function. Utilizing the MS2 model biosensor system, we rationally engineer charges onto the surface of an antibody fragment, increasing thermoresistance, minimizing destabilizing effects, and in some cases actually increasing affinity. Finally, we apply next-generation sequencing approaches to the rapid discovery of antibodies directed against the Zaire Ebolavirus species. We utilize a local immunization strategy to generate a polarized antibody repertoire that is then sequenced to provide a database of antigen-specific variants. This repertoire is probed in silico and individual antibodies selected for analysis, bypassing time- and resource-consuming selection experiments. / text

Many cancers have long been known to display an absolute requirement for the amino acid methionine (L-Met). Studies have shown that in the absence of L-Met, sensitive neoplasms experience cell cycle arrest and perish. Without the metabolic deviations that characterize L-Met auxotrophs, normal cells are able to grow on precursors such as homocysteine and tolerate periods of L-Met starvation. The differential requirement for this amino acid between normal and tumor cells has been exploited through enzymatic serum degradation of L-Met by a bacterial methionine-γ-lyase (MGL). Though MGL was able to deplete L-Met to therapeutically useful levels in animal models and exert a significant cytotoxic effect on malignant cell lines in vitro and on tumor xenografts in vivo, the clinical implementation of this enzyme is hampered by its short serum half-life and potential for catastrophic immune response. In the chapters that follow, we describe the engineering of a novel human methionine degrading enzyme (hMGL) that overcomes the limitations of the bacterial therapeutic. We have shown that hMGL is capable of degrading methionine at a therapeutically useful rate and inducing extensive cell killing in a variety of neoplasms. This enzyme is expected to have low immunogenicity in patients and a high therapeutic index. We have developed a high throughput screen for methionine degrading activity that we can utilize to further engineer the enzyme based on the results of additional preclinical development. We have found that hMGL is also capable of degrading cystine to operate as a dual amino acid depletion treatment that is expected to be more potent than methionine depletion alone. Due to the wide array of neoplasms sensitive to methionine and cystine starvation, the engineered enzyme holds a great deal of promise as a unique and powerful cancer therapeutic. / text

<p>QC 20150922</p>

Ultra-stable cystine knotted microproteins, in which two disulfides and their connecting backbones form a circle that is penetrated by the third disulfide bonds, have attracted high interest due to their resistance to degradation in vitro and potential for the development of peptide drugs. This thesis gives new insights into engineering of that framework of microproteins, including approaches to their chemical synthesis, backbone engineering, structural and biological evaluations. Synthetic and oxidative folding approaches for bracelet cyclotides, a family of cyclic cystine knotted microproteins, was developed using a model peptide, cycloviolacin O2. Following assembly of the peptide chain, protected peptide was generated by mild cleavage that was subsequently thioesterified and cyclized in solution. The cyclic peptide was oxidatively folded under optimized conditions containing co-solvent and non-ionic detergent affording native cycloviolacin O2 as a major product. To gain further insights into the heterogeneity, efficiency and kinetics of cyclotides’ oxidative folding, the intermediates that accumulate in oxidative refolding pathways of all cyclotide subfamilies: Möbius, bracelet and the hybrid cyclotides were quantitatively determined under four different folding conditions. The results were used for defining major folding pathways, which indicated that Möbius cyclotides might accumulate heterogeneous folding intermediates with one-, two- and three-disulfides, whereas bracelet tend to accumulate a homogenous intermediate with three-disulfides, depending on the buffer systems used. Furthermore, to probe the internal factors contributing to inefficiency of oxidative folding, as well as undesired bioactivities of bracelet cyclotides (e.g., cytotoxic activity), polymer-hybridized cyclotides were designed by replacing non-conserved residues with small isosteric polymers. The designed hybrid analogs in which hybridization involved replacement of loop 3 with isosteric polymers showed improved synthetic and oxidative folding properties. The cytoxicity of a model hybrid designed with replacement of loop 3 and 5 exhibited no cytotoxic activity at concentration of 128-fold relative to that of native peptide. Furthermore, 1D and 2D 1H NMR analysis of this hybrid showed that it had well structured fold.

A modified yeast one-hybrid (MY1H) system has been developed for in vivo investigation of simultaneous protein-protein and protein:DNA interactions. The traditional yeast one-hybrid assay (Y1H) permits examination of one expressed protein targeting one DNA site, whereas our MY1H allows coexpression of two different proteins and examination of their activity at the DNA target. This single-plasmid based MY1H was validated by use of the DNA-binding protein p53 and its inhibitory partners, large T antigen (LTAg) and 53BP2. The MY1H system could be used to examine proteins that contribute inhibitory, repressive, coactivational or bridging functions to the protein under investigation, as well as potential extension toward library screening for identification of novel accessory proteins. After development and validation of the MY1H with the p53/LTAg/53BP2 system, we applied the MY1H system to investigate the DNA binding activities of heterodimeric proteins, the bHLH/PAS domains of AhR and Arnt that target the xenobiotic response element (XRE). The AhR/Arnt:XRE interaction, which served as our positive control for heterodimeric protein binding of the XRE DNA site, showed negative signals in initial MY1H experiments. These false negative observations were turned into true positives by increasing the number of DNA target sites upstream of the reporter genes and increasing the number of activator domains fused to the two monomers. This methodology may help trouble-shooting false negatives stemming from unproductive transcription in yeast genetic assays, which can be a common problem. In the study of XRE-binding proteins, two bHLHZ-like hybrid proteins, AhRJunD and ArntFos were designed and coexpressed in the MY1H and yeast two-hybrid (Y2H) systems; these proteins comprise the bHLH domains of AhR and Arnt fused to the leucine zipper (LZ) elements from bZIP proteins JunD and Fos, respectively. The in vivo assays revealed that in the absence of the XRE DNA site, heterodimers and homodimers formed, but in the presence of the nonpalindromic XRE, only heterodimers bound to the XRE and activated reporter transcription. The present results provide valuable information on DNA-mediated protein heterodimerization and specific DNA binding, as well as the relationship between protein structure and DNA-binding function.

A modified yeast one-hybrid (MY1H) system has been developed for in vivo investigation of simultaneous protein-protein and protein:DNA interactions. The traditional yeast one-hybrid assay (Y1H) permits examination of one expressed protein targeting one DNA site, whereas our MY1H allows coexpression of two different proteins and examination of their activity at the DNA target. This single-plasmid based MY1H was validated by use of the DNA-binding protein p53 and its inhibitory partners, large T antigen (LTAg) and 53BP2. The MY1H system could be used to examine proteins that contribute inhibitory, repressive, coactivational or bridging functions to the protein under investigation, as well as potential extension toward library screening for identification of novel accessory proteins. After development and validation of the MY1H with the p53/LTAg/53BP2 system, we applied the MY1H system to investigate the DNA binding activities of heterodimeric proteins, the bHLH/PAS domains of AhR and Arnt that target the xenobiotic response element (XRE). The AhR/Arnt:XRE interaction, which served as our positive control for heterodimeric protein binding of the XRE DNA site, showed negative signals in initial MY1H experiments. These false negative observations were turned into true positives by increasing the number of DNA target sites upstream of the reporter genes and increasing the number of activator domains fused to the two monomers. This methodology may help trouble-shooting false negatives stemming from unproductive transcription in yeast genetic assays, which can be a common problem. In the study of XRE-binding proteins, two bHLHZ-like hybrid proteins, AhRJunD and ArntFos were designed and coexpressed in the MY1H and yeast two-hybrid (Y2H) systems; these proteins comprise the bHLH domains of AhR and Arnt fused to the leucine zipper (LZ) elements from bZIP proteins JunD and Fos, respectively. The in vivo assays revealed that in the absence of the XRE DNA site, heterodimers and homodimers formed, but in the presence of the nonpalindromic XRE, only heterodimers bound to the XRE and activated reporter transcription. The present results provide valuable information on DNA-mediated protein heterodimerization and specific DNA binding, as well as the relationship between protein structure and DNA-binding function.

Thesis (M.Tech.: Biotechnology)-Dept. of Biotechnology, Durban University of Technology, 2007 vi, 92 leaves / Interest in xylanases from different microbial sources has increased markedly in the past decade, in part because of the application of these enzymes in a number of industries, the main area being the pulp and paper industry. While conventional methods will continue to be applied to enzyme production from micro-organisms, the application of recombinant DNA techniques is beginning to reveal important information on the molecular basis and this knowledge is now being applied both in the laboratory and commercially. In this study, a directed evolution strategy was used to select an enzyme variant with high thermostability. This study describes the use of error-prone PCR to modify the xylanase gene from Thermomyces lanuginosus DSM 5826, rendering it tolerant to temperatures in excess of 80°C. Mutagenesis comprised of different concentrations of nucleotides and manganese ions. The variants were generated in iterative steps and subsequent screening for the best mutant was evaluated using RBB-xylan agar plates. The optimum temperature for the activity of xylanases amongst all the enzyme variants was 72°C whilst the temperature optimum for the wild type enzyme was 70°C. Long term thermostability screening was therefore carried out at 80°C and 90°C. The screen yielded a variant which had a 38% improvement in thermostability compared to the wild type xylanase from pX3 (the unmutated gene). Successive rounds of error-prone PCR were carried out and in each round the progeny mutant displayed better thermostability than the parent. The most stable variant exhibited 71% residual activity after 90 minutes at 80˚C. Sequence analysis revealed four single amino acid residue changes that possibly enhanced their thermostabilities. This in vitro enzyme evolution technique therefore served as an effective tool in improving the thermostable property of this xylanase which is an important requirement in industry and has considerable potential for many industrial applications.

Thesis (D.Tech.: Biotechnology)-Dept. of Biotechnology, Durban University of Technology, 2007 xi, 209 leaves / Protein engineering technologies, such as directed evolution and DNA recombination, are often used to modify enzymes on a genetic level for the creation of useful industrial catalysts. Pre-treatment of paper pulps with xylanases have been shown to decrease the amounts of toxic chlorine dioxide used to bleach pulp. This study was undertaken to improve the thermal and alkaline stabilities of the xylanase from the fungus Thermomyces lanuginosus using ep-PCR and DNA shuffling.

Submitted in fulfillment for the requirement of a Degree of Master of Technology: Biotechnology, in the Department of Biotechnology and Food Technology, Faculty of Applied Sciences, Durban University of Technology, Durban, South Africa, 2009. / Protein engineering has provided a key for adapting naturally-occurring enzymes for industrial processes. However, several obstacles have to be overcome after these proteins have been adapted, the main one being finding a suitable host to over-express these recombinant protein. This study investigated Saccharomyces cerevisiae, Pichia pastoris and Escherichia coli as suitable expression hosts for a previously modified fungal xylanase, which is naturally produced by the filamentous fungus, Thermomyces lanuginosus. A xylanase variant, NC38, that was made alkaline-stable using directed evolution was cloned into four different vectors: pDLG1 with an ADH2 promoter and pJC1 with a PGK promoter for expression in S. Cerevisiae, pBGP1 with a GAP promoter for expression in P. pastoris and pET22b(+) for expression in E. Coli BL21 (DE3). S. Cerevisiae clones with the p DLG1-NC38 combination showed very low activity on the plate assay and were not used for expression in liquid media as the promoter was easily repressed by reducing sugars used during production experiments. S. cerevisiae clones carrying pJC1-NC38 were grown in media without uracil while P. Pastoris clones were grown in YPD containing the antibiotic, zeocin and E. Coli clones were grown in LB with ampicillin. The levels of xylanase expression were then compared between P. Pastoris, S. cerevisiae and E. coli. The highest recombinant xylanase expression was observed in P. Pastoris with 261.7U/ml, followed by E.coli with 47.9 U/ml and lastly S. cerevisiae with 13.2 U/ml. The localization of the enzyme was also determined. In the methylotrophic yeast, P. Pastoris, the enzyme was secreted into the culture media with little or no contamination from the host proteins, while the in other hosts, the xylanase was located intracellularly. Therefore in this study, a mutated alkaline stable xylanase was successfully expressed in P. Pastoris and was also secreted into the culture medium with little or no contamination by host proteins, which favours the application of this enzyme in the pulp and paper industry.