The sirtuins are a family of NAD⁺-dependent deacetylase enzymes which are implicated in various illnesses including cancer and neurodegenerative diseases. Part I of this thesis describes the synthesis and biological evaluation of inhibitors of the SIRT1 and SIRT2 isoforms of this important family of enzymes. Chapter 1 gives an overview of sirtuin biology and the physiological roles of these enzymes. In particular the link between SIRT1 and cancer and SIRT2 and its role in the onset of neurodegenerative diseases is discussed. A review of the most potent and selective inhibitors of SIRT1 and SIRT2 is given including an introduction to the tenovin and cambinol classes of inhibitor. Chapter 2 describes various issues relating to the structure of the important chemical tool tenovin-6. The synthesis of analogues to improve the solubility, determine the preferred conformation and verify the products of metabolism of tenovin-6 is presented including their evaluation by in vitro and in cell methods. Part II of this chapter reports the design and use of a ¹H NMR method used to monitor the sirtuin-mediated deacetylation reaction. This was particularly relevant due to concerns raised about the possibility of false positive results obtained with the commercially available assay kit commonly used by the sirtuin community. This new ¹H NMR method was used to validate the inhibition of SIRT2 by tenovin-6. Chapter 3 describes the parallel synthesis and evaluation of tenovin analogues as inhibitors of SIRT1 and SIRT2. This study identified that replacement of the t-butyl substituent of tenovin-6 with the 3,5-dihalogen-4-alkoxy substitution pattern led to a variety of analogues having SIRT2 selectivity. As well as the collection of valuable SAR data, in cell data is also presented for the analogues. Chapter 4 provides attempts to rationalise the SAR data collected in Chapters 2 and 3 through a computational study. The molecular docking software GOLD was used to predict the binding site of the tenovin scaffold and hence rationalise the observed potencies of various analogues. Chapter 5 reports the synthesis and biological evaluation of triazole and cambinol analogues as SIRT1 and SIRT2 inhibitors. Part I details the synthesis and in vitro testing of a series of ring constrained tenovin analogues based on the 1,4-disubstituted triazole using click chemistry. A series of 1,5-disubstituted analogues were also synthesised. Part II describes the synthesis of S-alkylated cambinol analogues and the effect of N3-methylation upon activity and selectivity towards SIRT1. Part II of this thesis details the synthesis and biological testing of novel potent allosteric inhibitors of RmlA. RmlA is the first enzyme in the L-rhamnose biosynthetic pathway in bacteria. L-rhamnose is an important component of the bacterial cell wall and as such RmlA is therefore an important target in the discovery of novel anti-bacterial compounds. Chapter 7 provides an overview of the RmlA enzyme including its role in L-rhamnose biosynthesis and why it is an attractive target for anti-bacterial drug discovery. No small molecule inhibitors of RmlA have been reported previously. Chapter 8 describes the design and synthesis of pyrimidine-2,4-dione analogues as novel allosteric inhibitors of RmlA. SAR data is generated and rationalised by X-ray crystallographic techniques to study the structures of complexes of RmlA with various analogues. Analogues were also tested for their ability to inhibit the growth of the important human pathogen Mycobacterium tuberculosis.

Cette thèse de doctorat traite de la conception et de l’évaluation de sondes magnétogènes et fluorogènes pour la détection in vivo d’activités enzymatiques.Des molécules capables d’acquérir un moment magnétique à partir d’un état diamagnétique et en réponse à l’action d’une enzyme seraient d’un grand intérêt pour l’imagerie moléculaire par résonance magnétique. Deux exemples de telles sondes magnétogéniques avaient été mis au point précédemment, l’un pouvant opérer en conditions physiologiques, l’autre nécessitant une acidification du milieu pour devenir paramagnétique. En préparant de nouveaux analogues du premier exemple, j’ai pu trouver une molécule dont la fragmentation a lieu trois fois plus rapidement que la molécule originale. J’ai ensuite travaillé à la conception de sondes dérivées du deuxième exemple et répondant à des activités enzymatiques ; de telles molécules permettraient de réaliser la quantification in vitro d’une activité enzymatique à des fins diagnostiques. À ce titre, j’ai participé à l’élaboration de deux preuves de concept de dispositifs dédiés à la mesure du temps de relaxation longitudinale de micro-volumes. J’ai enfin entamé le développement de nouveaux complexes s’inspirant du second exemple mais capables de fonctionner dans le milieu biologique.Le deuxième volet de mes travaux porte sur la réalisation de sondes fluorogènes précipitantes pour des activités glycosidases. Une sonde pour la leucine aminopeptidase profitant des propriétés exceptionnelles de stabilité, de luminescence et de solubilité du fluorophore ELF®-97 avaient démontré une grande efficacité pour marquer rapidement des cellules HeLa. J’ai mis au point une nouvelle architecture de sondes qui permet le ciblage de glycosidases via l’utilisation d’un tandem d’espaceurs cyclisants. Deux sondes ont été préparées, l’une pour la beta-galactosidase, l’autre pour la cellulase. La première a prouvé son bon fonctionnement pour marquer les cellules exprimant l’enzyme en bénéficiant d’une grande sensibilité. La seconde a pu être utilisée pour quantifier l’activité cellulase sécrétée par des levures, avec l’objectif d’obtenir un moyen économiquement intéressant de produire du bioéthanol à partir des déchets végétaux. / This PhD thesis deals with the design and evaluation of magnetogenic and fluorogenic probes for the in vivo detection of enzyme activities.Molecules capable of switching from a diamagnetic to a paramagnetic state in response to an enzyme stimulus would be of great interest for molecular magnetic resonance imaging. Two examples of such magnetogenic probes had been designed in a previous work : one can operate in physiological conditions, whereas the other needs an acidification of the water medium to become paramagnetic. I prepared new analogues of the first probe ; one molecule displayed fragmentation three times faster than the original compound. Then I designed and synthesized probes derived from the second example and responsive to enzyme activities ; such molecules are suitable for the in vitro quantification of enzyme biomarkers for diagnosis purposes. I participated to the conception of two proofs of concept of devices dedicated to the measurement of longitudinal relaxation times in micro-volumes. Finally, I started the development of a new family of molecules inspired by the second example but able to work at the physiological pH.I also worked on precipitating fluorogenic probes for the detection of glycosidase activities. A former probe for leucine aminopeptidase, based on the exceptional characteristics of the fluorophore ELF-97 in terms of solubility, luminescence and stability, had demonstrated great efficiency to label live HeLa cells. I designed a new architecture of probes responding to glycosidases via an original tandem of selfimmolative spacers. Two probes have been prepared, one targets beta-galactosidase and the second detects cellulase. The first probe performed a fast and sensitive labelling of beta-galactosidase-expressing cells. The second molecule was employed successfully to quantify the cellulase activity secreted by yeasts, which will be useful for the high-throughput screening of yeasts capable of producing bioethanol from vegetal waste.

Though antibodies as therapeutics are limited to extracellular targets, their repertoire of molecular interactions has particular relevance to the many intracellular cellular proteins for which small molecule screening has reached impasse. For such proteins there is little recourse to theory, since molecular recognition is, in practical terms, still not well understood. Here I apply antibody discovery to the lysine demthylases KDM4D and PHF8, two proteins difficult to inhibit selectively due to the similarity of their binding pockets to those of the larger family. With a selective, picomolar affinity antibody, dependent on residues distal to the KDM4D active site, I present what is likely the first example of allosteric inhibition of a KDM4 lysine demethylase, demonstrating that there is opportunity outside active sites oversubscribed with pan inhibitors. Antibody discovery for PHF8, however, was plagued by a familiar problem: antibodies that bound when their antigen was immobilised directly to a surface, but barely bound at all when it was free in solution. The common explanation is that the partial denaturation that accompanies immobilisation reveals epitopes unavailable in solution, but examining the problem in detail for the Plant Homeodomain of PHF8 revealed a connection to its rarely sampled conformations. The prominence these antibodies in the immune responses to PHF8, and to some extent KDM4D, motivates two hypotheses on their origin: either the states are very immunogenic or there is a connection between states of irreversible damage and those sampled reversibly, but rarely, by a protein in solution.

The interactions of tumours with normal host tissue are key determinants of cancer growth and progression. The ability or inability of the patient’s immune system to mount a response against the tumour is tightly correlated with prognosis. One of the ways tumours avoid detection and elimination by the immune system is by expressing programmed death ligand 1 (PD-L1). PD-L1 binds to its receptor programmed death 1 (PD1) on T cells, inhibiting T cell responsiveness to antigenic stimuli. Blockade of the PD1/PD-L1 pathway removes this negative signal and restores anti-tumour immunity. While this blockade of PD1/PD-L1 is well established through the use of antibodies, small molecule inhibitors of PD1/PD-L1 are relatively unknown. We employed in silico docking in order to find small molecules capable of binding to either PD1 or PD-L1, and the highest-ranked compounds were tested in biophysical assays for their ability to inhibit PD1/PD-L1 binding. A thermal shift assay identified a pyrazole compound as a possible binding partner for PD-L1, but follow-up assays showed that it had no effect on the PD1/PD-L1 interaction and that its apparent binding was probably due to aggregation. An ELISA assay identified a tryptophan diamine compound as an apparent stabilizer of the PD1/PD-L1 interaction. However this compound, too, was later identified to be inactive in orthogonal assays. We identified a family of salicylic acid derivatives that interfered with TR-FRET measurements – an unusual observation, given that TR-FRET is touted as being insensitive to most mechanisms of compound interference. This discovery should help other fragment- screening groups identify false positives more easily. We also probed the mechanism of inhibition of a recently disclosed family of small molecule PD1/PD-L1 inhibitors from Bristol-Myers Squibb. Concurrently with other groups, we used protein NMR, size exclusion chromatography, and SPR to determine that the compounds were inducing homodimerization through the PD1-binding face of PD-L1. Furthermore, using cellular crosslinking and live cell imaging, we showed that these first generation inhibitors are fairly ineffective at inhibiting this interaction on the cell surface. More potent compounds will be needed to see any cellular effect from this mechanism of action. / Graduate / 2019-02-15

Carnitine is an essential metabolite in the human body. It carries out several roles in human metabolism, including that in fatty acid metabolism. γ-Butyrobetaine hydroxylase (BBOX) is an Fe(II) and 2-oxoglutarate dependent oxygenase, which catalyses the final step of carnitine biosynthesis, i.e. hydroxylation of γ-butyrobetaine (GBB) to carnitine. Inhibition of BBOX has potential in the treatment for cardiovascular diseases. The work described in this thesis focussed on mechanistic and inhibition aspects of BBOX catalysis. Firstly, a set of analytical tools for BBOX activity measurements was developed. The synthesis of fluorinated substrate analogues provided the basis for development of two assays for use in vitro with the isolated protein and in lysates, with detection by fluorescence or ¹⁹F NMR, respectively. Furthermore, the use of ¹⁹F NMR to monitor protein-ligand interactions was exemplified with the work on metallo-β-lactamases. The developed fluoride-release assay was then used to screen a library of small molecules and led to recognition of scaffolds with potential applications as inhibitors. Further structure-activity relationship studies led to the identification of potent BBOX inhibitors, which were then evaluated for their activity in cells. The crystal structure of human BBOX with one of the lead inhibitors revealed that BBOX can undergo significant conformational changes, involving a movement of an active site loop. BBOX conformational flexibility may have a role in the GBB mediated substrate inhibition observed both with isolated protein and in cells. In addition to the mechanistic and functional studies, the potential of BBOX as a biocatalytic tool was examined. BBOX has been shown to catalyse a hydroxylation of the symmetrical dialkyl piperidine carboxylic acids, leading to formation of up to three stereocentres in one reaction. In the last part of this work properties of human BBOX were compared to BBOX from Pseudomonas sp. AK1, revealing differences in kinetic behaviour and substrate specificity. Novel substrates for bacterial BBOX were identified. Pseudomonas sp AK1 BBOX was shown to hydroxylate amino acid analogues leading to formation of 1,2-amino alcohols.

<p>Pathogenic bacteria present a critical threat to modern medicine. Therapeutic strategies to target and eliminate resilient bacteria are not advancing at the same rate as the emergence of bacterial resistance. An associated urgent concern regarding antibiotic resistance is the existence and proliferation of intracellular bacteria, which find refuge from bactericidal mechanisms by hiding within mammalian cells. Therefore, many once-successful antibiotics become ineffective through the development of resistance, or through failure to reach intracellular locations in therapeutic concentration. To overcome these challenges, the covalent combination of a conventional antibiotic with an antibiotic, cell-penetrating peptide was explored to develop dual-action antibiotic conjugates. </p> <p>Herein, we utilized a strategy in conjugating the antibiotics by a cleavable linkage to cationic amphiphilic polyproline helices (CAPHs) to improve vancomycin and linezolid antibiotics. This approach enables the conjugate to penetrate cells and deliver two potent monomeric antimicrobial drugs. The vancomycin-CAPH conjugate, <b>VanP14S</b>, showed enhanced mammalian cell uptake compared to vancomycin, a poor mammalian cell-penetrating agent; and <b>VanP14S</b> was capable of cleaving and releasing two antibiotics under mimicked physiological conditions. Enhanced antibacterial activity was observed against a spectrum of Gram-positive and Gram-negative pathogens, including drug-resistant strains. Further investigation revealed that this conjugate’s bactericidal activity was not entirely the result of significant membrane perturbation such as a lytic mode of action. Mammalian cell toxicity and red blood cell lysis were insignificant at relevant bactericidal concentrations below 20 µM. The current results suggest an enhanced binding to the peptidoglycan of bacteria, the target of vancomycin, although more work is needed to justify this claim. Preliminary results on <b>VanP14GAPS</b>, a conjugate with a more rigid CAPH, convey similar activity to <b>VanP14S; </b>however,<b> </b>moderate increases in red blood cell lysis and cytotoxicity were observed. </p> <p>Regarding the <b>LnzP14</b> conjugate, preliminary data reveal that the conjugate has Gram-negative activity against <i>Escherichia coli</i>, whereas linezolid is ineffective in killing Gram-negative bacteria. This conjugate showed significant enhancement in cellular uptake compared to the CAPH, and the release of linezolid and CAPH in physiological conditions was confirmed. Overall, arming a conventional antibiotic with an antimicrobial, cell-penetrating peptide appears to be a powerful strategy in providing novel antibiotic conjugates with the propensity to overcome the limitations in treating challenging pathogens.</p>

京都大学 / 0048 / 新制・課程博士 / 博士(理学) / 甲第19260号 / 理博第4115号 / 新制||理||1592(附属図書館) / 32262 / 京都大学大学院理学研究科化学専攻 / (主査)教授 杉山 弘, 教授 三木 邦夫, 教授 藤井 紀子 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM

The alarming rise of antibacterial drug resistance and the dwindling supply of novel antibiotics highlight the need for innovative approaches in combating bacterial infections. Traditionally, antibacterial drug discovery campaigns have largely been conducted in rich media. Such growth conditions are not representative of the host environment and render many metabolic pathways, otherwise needed for survival and infection, dispensable. Such pathways have been overlooked in conventional drug discovery campaigns despite their validity as potential antibacterial targets. The work presented in this thesis focuses on the development and validation of a screening strategy for the identification and mechanism of action determination of novel inhibitors of metabolic pathways in bacteria under nutrient-limited conditions. This screen led to the identification of MAC168425, MAC173979 and MAC13772 as inhibitors that target glycine metabolism, p-aminobenzoic acid biosynthesis and biotin biosynthesis, respectively. Moreover, it established this approach as a general platform that can be applied for different organisms with synthetic or natural product libraries. Additional mechanistic studies of the biotin biosynthesis inhibitor, MAC13772, resulted in solving the crystal structure of BioA in complex with MAC13772. Analysis of the co-structure confirmed our proposed mode of inhibition and provided information for strategies for rational drug design. Investigation of the antibacterial activity of MAC13772 revealed its potency against a number of pathogens. Furthermore, we show how MAC13772 acts synergistically with rifampicin in clearing growing mycobacterial cultures. The potential of this inhibitor as a lead for preclinical pharmacokinetic studies and for antibacterial drug development is discussed. We also discuss our current efforts to develop a metabolomic platform for the characterization of novel antibacterials that can be used in concert with our current approach to chart the metabolic response of bacteria to chemical perturbants and to generate testable hypotheses regarding the mode of action of novel inhibitors of bacterial metabolism. / Thesis / Doctor of Philosophy (PhD)

During its intraerythrocytic stages (IE), Plasmodium falciparum, the causative agent of the deadliest human malaria, remodels the host red cell membrane with a poorly defined assortment of parasite-­encoded proteins that undergo antigenic variation. Despite the requirement for immunologic stealth, exported parasite proteins also mediate strain-independent functions such as endothelial sequestration that are critical for parasite survival and pathogenesis. This thesis explores the hypothesis that P. falciparum displays novel structurally conserved proteins on the IE surface and these proteins may serve as useful antigens for a broadly effective anti-­malarial vaccine. In order to test this hypothesis, we developed an in vitro selection technique that sequentially incorporates unique P. falciparum isolates as the targets for Systematic Evolution of Ligands by EXponential enrichment (Serial-SELEX) to generate nucleic acid molecular probes, aptamers, capable of recognizing conserved cell surface determinants. Ten of 11 enriched aptamers were -parasite selective and three of these aptamers demonstrated strain-independent binding to P. falciparum. Aptamer recognition extended beyond the parasites used in Serial-SELEX to other laboratory and recent field isolates. Surprisingly the same three broadly binding aptamer selected against P. falciparum also recognized all laboratory-adapted and clinical isolates of P. vivax and P. knowlesi tested, strongly supporting our hypothesis that structurally conserved molecules are present on the surface IEs. Competition studies showed that the aptamers bound a single target which was confirmed as an IE membrane protein. Aptamer­‐mediated affinity purification and tandem mass spectrometry enabled identification of the aptamer target as parasite-encoded protein. Discovery of a protein conserved between the major human malarias may have implications for vaccine development and validates the Serial‑SELEX technique as a powerful tool for antigen discovery.

The IgG Fc domain mediates a range of antibody effector functions, including antibody dependent cell-mediated cytotoxicity (ADCC), complement activation, phagocytosis, and the recently emerged general anti-inflammatory effect of immunoglobulin therapy (IVIg). The conserved N-glycan attached to Fc N297 maintains the Fc structural integrity for the effector functions, while its glycoform is known to modulate the affinity for the Fc γ-receptors (FcγRs), complement, and the C-type lectin DC-SIGN. IgG Fc exhibits protein-directed glycosylation characterized by a series of biantennary complex type glycoforms, with a small population of sialylated species. The sialylated Fc has been proposed to bind DC-SIGN and initiate an anti-inflammatory signalling pathway. The restricted Fc glycan processing is partially attributed to the hydrophobic interaction between Fc glycan and the hydrophobic Fc protein backbone. Mutations within the hydrophobic Fc protein-glycan interface dramatically increases Fc glycan processing, while concomitantly decreases Fc affinity for the FcγRs. However, it is unclear whether this disrupted Fc-FcγR interaction was due to the increased terminal glycan processing, or the perturbed Fc protein-glycan interface. Here, the integrity of the Fc protein-glycan interface was demonstrated to be important in maintaining the productive Fc-FcγR interaction independently of glycoform. This glycoform-independent effect was exploited to generate novel inhibitory Fc variants. In addition, the interaction between sialylated IgG and the putative IVIg receptor DC-SIGN was re-evaluated. Analysis shows that IVIg binds DC-SIGN in a glycan-independent, Fab-mediated manner. Furthermore, the effect of IVIg sialylation on human antigen presenting cells was examined; evidence presented here indicate that IVIg deglycosylation, not desialylation, has an anti-inflammatory effect on human dendritic cells (DCs). These data suggest the need for a general re-evaluation of the current mechanistic model of anti-inflammatory IVIg.