Dynamic Binary Modification (DBM) is a technique for modifying applications at runtime, working at the level of native code. It has numerous applications, including instrumentation, translation and optimisation. However, DBM introduces a performance overhead, which in some cases can dominate execution time, making many uses impractical. While avenues for reducing this overhead have been widely explored on x86, ARM, an architecture gaining widespread adoption, has received little attention. Consequently, the overhead of DBM on ARM, as reported in the literature and measured using the available DBM systems, has fallen behind the state-of-the-art by one or two orders of magnitude. The research questions addressed in this thesis are: 1) how to develop low overhead DBM systems for the ARM architecture, and 2) whether new optimisations are plausible and needed. Towards that end, a number of novel optimisations were developed and evaluated specifically to address the sources of overhead for DBM on various ARM microarchitectures. Furthermore, many of the optimisations in the literature were ported to ARM and evaluated. This work was enabled by a new DBM system, named MAMBO, created specifically for this purpose. MAMBO, using the optimisations presented in this thesis, is able to achieve an overhead an order of magnitude smaller than that of the most efficient DBM system for ARM available at the start of this PhD.

Throughout this thesis, we studied the biochemical and structural impact of the essential residues on the activity of class D beta-lactamases. The production of these enzymes plays a major role in the bacterial resistance. Our work is subdivided in two parts : the study of the post-translational modification of lysine 70 and the screening of new potential inhibitors for the class D β-lactamases. The first part concerns the impact of the residues tryptophan 154 and valine 117 located in the hydrophobic core. Our data indicate that the mutation of tryptophan 154 in alanine or glycine lead to a large decrease of the catalytic efficiencies of the beta-lactamase. The apo-enzyme structures of these mutants show that the lysine 70 is not carboxylated. This absence of carboxylate group induces a modification of the hydrogen network of the active site. The analysis of the complex structure of W154A-benzylpenicillin demonstrates that the deacylation step is clearly the most affected by the mutation. The mutation of tryptophan 154 in histidine leads to a slight decrease of catalytic efficiencies because the imidazol group of histidine mimics the indole group of tryptophan 154. The apo-enzyme structure reveals that lysine 70 is partially carboxylated and stabilized by an hydrogen bond between the carboxylate group and the imidazol group. In the case of the V117T mutant, a strong increase of the catalytic constant values is observed at 50 mM in NaHCO3. The structure of this mutant at pH 8.0 shows that the lysine 70 is partially carboxylated in the monomer A. The determination of individual rate constants of acylation and deacylation steps indicates that the deacylation is the limiting step for the class D beta-lactamase. The k2/k3 ratio is similar between the V117T mutant and the wild-type enzyme. The mutation of lysine 70 in alanine or cysteine leads to a large decrease of the deacylation constants inducing a poorly efficient enzyme. The obtaining of the K70C-Ampicillin complex by X-ray cristallography and the trapping of acyl-enzyme by reaction with fluorescent ampicillin are supplemental proofs that the deacylation step is the limiting rate. By crystallographic and kinetic studies, we demonstrate that the chloride inhibition of the class D beta-lactamases is due to a competition between the carboxylate group of lysine 70 and the chloride ions. At high concentration in bicarbonate, this inhibition is abolished for the wild-type enzyme. The second part of this work concerns the screening of the citrate and aminophosphonate derivated molecules for the class D beta-lactamases. In the case of OXA-10, a citrate molecule is strongly stabilized by hydrogen bonds in the active site. The benzyl esters derivatives of citrate inhibits OXA-10(KI = 20 µM) but the hydrophobic substituents are necessary to obtain a good inhibition.

published_or_final_version / abstract / toc / Sociology / Master / Master of Philosophy