The plant homeodomain (PHD) is a protein domain of ~45�100 residues characterised by a Cys4-His-Cys3 zinc-binding motif. When we commenced our study of the PHD in 2000, it was clear that the domain was commonly found in proteins involved in transcription. Sequence alignments indicate that while the cysteines, histidine and a few other key residues are strictly conserved, the rest of the domain varies greatly in terms of both amino acid composition and length. However, no structural information was available on the PHD and little was known about its function. We were therefore interested in determining the structure of a PHD in the hope that this might shed some light on its function and molecular mechanism of action. Our work began with the structure determination of a representative PHD, Mi2b-P2, and this work is presented in Chapter 3. Through comparison of this structure with the two other PHD structures that were determined during the course of our work, it became clear that PHDs adopt a well-defined globular fold with a superimposable core region. In addition, PHDs contain two loop regions (termed L1 and L3) that display increased flexibility and overlay less well between the three PHD structures available. These L1 and L3 regions correspond to variable regions identified earlier in PHD sequence alignments, indicating that L1 and L3 are probably not crucial for the PHD fold, but are instead likely to be responsible for imparting function(s) to the PHD. Indeed, numerous recent functional studies of PHDs from different proteins have since demonstrated their ability in binding a range of other proteins. In order to ascertain whether or not L1 and L3 were in fact dispensable for folding, we made extensive mutations (including both insertions and substitutions) in the loop regions of Mi2b-P2 and showed that the structure was maintained. We then went on to illustrate that a new function could be imparted to Mi2b-P2 by inserting a five-residue CtBP-binding motif into the L1 region and showed this chimera could fold and bind CtBP. Having established that the PHD could adopt a new binding function, we next sought to use combinatorial methods to introduce other novel functions into the PHD scaffold. Phage display was selected for this purpose, because it is a well-established technique and has been used successfully to engineer zinc-binding domains by other researchers. However, in order to establish this technique in our laboratory, we first chose a control system in which two partner proteins were already known to interact in vitro. We chose the protein complex formed between the transcriptional regulators LMO2 and ldb1 as a test case. We have examined this interaction in detail in our laboratory, and determined its three-dimensional structure. Furthermore, inappropriate formation of this complex is implicated in the onset of T-cell acute lymphoblastic leukemia. We therefore sought to use phage display to engineer ldb1 mimics that could potentially compete against wild-type ldb1 for LMO2, and this work is described in Chapter 4. Using a phage library containing ~3 x 10 7 variants of the LMO2-binding region of ldb1, we isolated mutants that were able to interact with LMO2 with higher affinity and specificity than wild-type ldb1. These ldb1 mutants represent a first step towards finding potential therapeutics for treating LMO-associated diseases. Having established phage display in our laboratory, we went on to search for PHD mutants that could bind selected target proteins. This work is described in Chapter 5. We created three PHD libraries with eight randomized residues in each of L1, L3 or in both loops of the PHD. These PHD libraries were then screened against four target proteins. After four rounds of selection, we were able to isolate a PHD mutant (dubbed L13-FH6) that could bind our test protein Fli-ets. This result demonstrates that a novel function can be imparted to the PHD using combinatorial methods and opens the way for further work in applying the PHD scaffold to other protein design work. In summary, the work detailed in Chapters 3 and 5 demonstrates that the PHD possesses many of the properties that are desirable for a protein scaffold for molecular recognition, including small size, stability, and a well-characterised structure. Moreover, the PHD motif possesses two loops (L1 and L3) of substantial size that can be remodeled for target binding. This may lead to an enhancement of binding affinities and specificities over other small scaffolds that have only one variable loop. In light of the fact that PHDs are mainly found in nuclear proteins, it is reasonable to expect that engineered PHDs could be expressed and function in an intracellular environment, unlike many other scaffolds that can only function in an oxidizing environment. Therefore, our results together with other currently available genomic and functional information indicate PHD is an excellent candidate for a scaffold that could be used to modify cellular processes. Appendices 1 and 2 describe completed bodies of work on unrelated projects that I have carried out during the course of my PhD candidature. The first comprises the invention and application of DNA sequences that contain all N-base sequences in the minimum possible length. This work is presented as a reprint of our recently published paper in Nucleic Acids Research. The second Appendix describes our structural analysis of an antifreeze protein from the shorthorn sculpin, a fish that lives in the Arctic and Antarctic oceans. This work is presented as a manuscript that is currently under review at the Journal of the American Chemical Society.

We have developed a novel computational approach to functional de novo protein design using gradient based continuous optimization techniques. Motivated by many engineering applications in which a cost function is optimized subject to a set of constraints, we pose a functional protein design task as a continuous optimization problem to search sequence and conformation spaces simultaneously. The methods used in sequence-space search are analogous to the material-design formulations in the topology optimization of structures, whereas the conformation search techniques are similar to mechanical-link like models and modal analysis of structures. Computationally efficient techniques such as the nonlinear conjugate gradient and interior point optimization are used to solve optimization problems. Both the sequence and conformation search techniques are individually validated with real proteins. Coarse-grained as well as atomistic potentials are used to model the energy. Finally, we combine the sequence and conformation search methods and propose a new strategy for a simultaneous search in sequence and conformation spaces for designing functional de novo proteins. In view of the lack of experimental resources, the proposed computational scheme is validated by re-designing an existing protein, the hen-egg white lysozyme. Since the thrust of this method is on developing computationally efficient models, we developed an amino acid grouping scheme based on metric multi-dimensional scaling. Some structure-prediction problems are also solved using Graphics Processing Unit (GPU) based Compute Unified Device Architecture (CUDA) programming.

Directing groups (DGs) are moieties installed onto organic molecules to confer regioselectivity in subsequent reactions. DGs have found utility in selective CH activations catalyzed by transition metal (TM) catalysis on starting materials with multiple CH bonds. Despite their utility, DGs are scarcely used in industrial applications due to the generally wasteful nature of conventional DG strategies and their associated increase in step-count. Transient directing groups (TDGs) have been developed to overcome these limitations, with additives reversibly forming adducts with compounds of interest prior to the DG-mediated CH activation, in one-pot processes. However, the use of TDGs still requires harsh conditions to achieve significant yields, hindering broad applications. Chemoenzymatic catalytic cascades have attracted attention due to the mild and environmentally friendly nature of biocatalysis, with the greatest challenge being compatibility issues between biocatalytic and traditional chemical transformations. Here we propose a concurrent chemoenzymatic catalytic cascade that would enable TM-catalyzed DG chemistry via flanking biocatalytic reductive amination to install, and oxidative deamination to remove, a TDG. Preliminary efforts have identified some incompatibilities arising from the biocatalytic portion of the cascade, namely substrate specificity and organic co-solvent tolerance, that need to be addressed to achieve the proposed chemoenzymatic cascade in a one-pot concurrent protocol.

No description available.

No description available.

Ces dernières années, plusieurs modèles d'évolution moléculaire, basés sur l'hypothèse que les séquences des protéines évoluent sous la contrainte d'une structure bien définie et constante au cours de l'évolution, ont été développés. Cependant, un tel modèle repose sur l'expression de la fonction représentant le lien entre la structure et sa séquence. Les potentiels statistiques proposent une solution intéressante, mais parmi l'ensemble des potentiels statistiques existants, lequel serait le plus approprié pour ces modèles d'évolution ? Dans cette thèse est développé un cadre probabiliste d'optimisation de potentiels statistiques, dans le contexte du maximum de vraisemblance, et dans une optique de protein design. Ce cadre intègre différentes méthodes d'optimisation, incluant la prise en compte de structures alternatives pour l'optimisation des potentiels, et fournit un cadre robuste et des tests statistiques (à la fois dans le contexte de l'optimisation des potentiels mais aussi dans le contexte de l'évolution moléculaire) permettant de comparer différentes méthodes d'optimisation de potentiels statistiques pour les modèles soumis à des contraintes structurales. / In the field of molecular evolution, so called Structurally constrained (SC) models have been developped. Expressed at the codon level, they explicitely separe the mutation (applied to the nucleotide sequence) and the selection (applied to the encoded protein sequence) factors. The selection factor is described as a function between the structure and the sequence of the protein, via the use of a statistical potential. However, the whole evolutionary model depends on the expression of this potential, and one can ask wether a potential would be better than another. In this thesis, is developped a probabilistic framework to optimize statistical potentials especially meant for protein design, using a maximum likelihood approach. The statistical potential used in this thesis is composed by a contact potential and a solvent accessibility potential, but the probabilistic framework can easily be generalized to more complex statistical potentials. In a first part, the framework is defined, and then an algorithmical enhancement is proposed, and finally, the framework is modified in order to take into account misfolded structures (decoys). The framework defined in this thesis and in other works allows to compare different optimization methods of statistical potentials for SC models, using cross-validation and Bayes factor comparisons.

<p>This thesis focuses on the development of algorithms that will allow protein design calculations to incorporate more realistic modeling assumptions. Protein design algorithms search large sequence spaces for protein sequences that are biologically and medically useful. Better modeling could improve the chance of success in designs and expand the range of problems to which these algorithms are applied. I have developed algorithms to improve modeling of backbone flexibility (DEEPer) and of more extensive continuous flexibility in general (EPIC and LUTE). I’ve also developed algorithms to perform multistate designs, which account for effects like specificity, with provable guarantees of accuracy (COMETS), and to accommodate a wider range of energy functions in design (EPIC and LUTE).</p> / Dissertation

In protein engineering, rational design and selection from combinatorial libraries are methods used to develop proteins with new or improved features. A very important protein for the biological sciences is the antibody that is used as a detecting agent in numerous laboratory assays. Antibodies used for these purposes are often ”man-made”, by immunising animals with the desired target, or by selections from combinatorial libraries. Naturally, antibodies are part of the immune defence protecting us from foreign attacks from e.g. bacteria or viruses. Some bacteria have evolved surface proteins that can bind to proteins abundant in the blood, like antibodies and serum albumin. By doing so, the bacteria can cover themselves in the host’s own proteins and through that evade being detected by the immune system. Two such proteins are Protein A from Staphylococcus aureus and Protein G from group C and G Streptococci. Both these proteins contain domains that bind to antibodies, one of which is denoted C2 (from Protein G) and another B (from Protein A). The B domain have been further engineered to the Z domain. In this thesis protein engineering has been used to develop variants of the C2 and Z domains for site-specific labelling of antibodies and for antibody purification with mild elution. By taking advantage of the domains’ inherent affinity for antibodies, engineering and design of certain amino acids or protein motifs of the domains have resulted in proteins with new properties. A photo crosslinking amino acid, p-benzoylphenylalanine, have been introduced at different positions to the C2 domain, rendering three new protein domains that can be used for site-specific labelling of antibodies at the Fc or Fab fragment. These domains were used for labelling antibodies with lanthanides and used for detection in a multiplex immunoassay. Moreover, a library of calcium-binding loops was grafted onto the Z domain and used for selection of a domain that binds antibodies in a calcium dependent manner. This engineered protein domain can be used for the purification of antibodies using milder elution conditions, by calcium removal, as compared to traditional antibody purification.

<p>Proteins are essential components of cells and are crucial for catalyzing reactions, signaling, recognition, motility, recycling, and structural stability. This diversity of function suggests that nature is only scratching the surface of protein functional space. Protein function is determined by structure, which in turn is determined predominantly by amino acid sequence. Protein design aims to explore protein sequence and conformational space to design novel proteins with new or improved function. The vast number of possible protein sequences makes exploring the space a challenging problem. </p><p>Computational structure-based protein design (CSPD) allows for the rational design of proteins. Because of the large search space, CSPD methods must balance search accuracy and modeling simplifications. We have developed algorithms that allow for the accurate and efficient search of protein conformational space. Specifically, we focus on algorithms that maintain provability, account for protein flexibility, and use ensemble-based rankings. We present several novel algorithms for incorporating improved flexibility into CSPD with continuous rotamers. We applied these algorithms to two biomedically important design problems. We designed peptide inhibitors of the cystic fibrosis agonist CAL that were able to restore function of the vital cystic fibrosis protein CFTR. We also designed improved HIV antibodies and nanobodies to combat HIV infections.</p> / Dissertation

PSD-95/Dlg/ZO-1 (PDZ) domain - PDZ binding motif (PBM) interactions have been one of the most well studied protein-protein interaction systems through biochemical, biophysical and high-throughput screening (HTS) strategies. This has allowed us to understand the mechanism of individual PDZ-PBM interactions and the re-engineering of PBMs to bind tighter or to gain or lose certain specificity. However, there are several thousand native PDZ domains whose biological ligands remain unknown. Because of the low sequence identity among PDZ domain homologues, promiscuous binding profiles (defined as a PDZ domain that can accommodate a set of PBMs or a PBM that can be recognized by many PDZ domains), and context-dependent interaction mechanism, we have an inadequate understanding of the general molecular mechanisms that determine the PDZ-PBM specificity. Therefore, predicting PDZ specificity has been elusive. In addition, no de novo PBM ligand or artificial non-native PDZ domain have been successfully designed. This reflects the general challenges in understanding the general principles of PDZ-ligand interactions, namely that they are context-dependent, exhibit weak binding affinity, narrow binding energy range, and larger interaction surface than other protein-ligand interactions. Together, PDZ domains make good model systems to investigate the fundamental principles of protein-protein interactions with a wide spectrum of biomedical implications. My studies suggest that understanding PBM specificity with the set of structural positions forming the binding pocket can connect sequence, structure and function of a PDZ domain in a general context. They also suggest that this way of understanding the specificity will shed light on prediction and engineering of specificity rationally. Structural analysis on most of the available PDZ domain structures was established to support the principle (Chapter I). The principle was tested against two different types of PBM; C-terminal PBM (Chapter II) and internal PBM (Chapter III), and shown to support better understanding and design of PDZ domain specificity. We further applied the principle to design de novo PDZ domains, and the preliminary data hints that it is optimistic to engineer PDZ domain specificity (Appendix A and B).