Thesis (Ph. D.)--University of Delaware, 2006. / Principal faculty advisor: Eugene G. Mueller, Dept. of Chemistry & Biochemistry. Includes bibliographical references.

Iron-sulfur (Fe-S) clusters are critical protein cofactors found in all life forms. In eukaryotes, a well-conserved biosynthetic pathway located in the mitochondria is used to assemble Fe-S clusters. Although proteins required for Fe-S cluster biosynthesis have been identified, their precise function and mechanism remain elusive. In this study, biochemical and biophysical methods are applied to understand molecular details for the core components of the human Fe-S cluster biosynthesis: Nfs1, Isd11, Isu2, and frataxin (Fxn). Nfs1 is a cysteine desulfurase that converts cysteine into alanine and transfers the sulfur to a scaffold protein Isu2 for Fe-S clusters. Fxn depletion is associated with the neurodegenerative disease Friedreich’s ataxia (FRDA), and results in a complicated phenotype that includes loss of Fe-S clusters. The results presented here provide the first in vitro evidence for a stable protein complex that exists in at least two forms: an inactive complex with Nfs1, Isd11, and Isu2 (SDU) components and an active form that also includes Fxn (SDUF). Fxn binding dramatically changes the catalytic efficiency (kcat/KM) of Nfs1 from 25 to 10,100 M-1s-1 and enhances the rate of Fe-S cluster biosynthesis 25 fold. Oxidizing conditions diminish the levels of both complex formation and Fxn-based activation, whereas Fe2 further stimulates Nfs1 activity. Mutagenesis coupled to enzyme kinetics indicate that one of the three conserved cysteines (C104) on Isu2 accepts the sulfane sulfur from Nfs1 and that this transfer event likely requires prior binding of Fxn. In vitro interrogation of FRDA I154F and W155R and related Fxn variants revealed the binding affinity to SDU followed the trend Fxn ~ I154F > W155F > W155A ~ W155R. The Fxn variants also have diminished ability to facilitate both sulfur transfer and Fe-S cluster assembly. Fxn crystallographic structures reveal specific rearrangements associated with the loss of function. Importantly, the weaker binding and lower activity of the W155R variant compared to I154F explains the earlier onset and more severe disease progression. Finally, these experimental results coupled with computational docking studies suggest a model for how human Fxn functions as an allosteric activator and triggers sulfur transfer and Fe-S cluster assembly.

Le soufre est utilisé à des fins bioénergétiques par des micro-organismes tels que la bactérie hyperthermophile Aquifex aeolicus qui nécessite pour sa croissance de l'oxygène, de l'hydrogène et un composé soufré indispensable. Une soufre réductase réduisant des chaînes de soufre, une Sulfure Quinone Oxydoréductase (SQR) oxydant l'H2S et une Soufre Oxygénase Réductase (SOR) oxydant et réduisant simultanément des chaînes de soufre ont été caractérisées chez cette bactérie. L'organisation de certaines de ces enzymes dans des supercomplexes membranaires a également été démontrée.Nous avons montré qu'Aq_477, précédemment caractérisée comme une soufre transférase de la famille des rhodanèses, est capable (i) de « charger » des chaînes de soufre ; (ii) d'interagir avec deux partenaires (la soufre réductase et la SOR) ; (iii) de leur présenter ce substrat. Ceci conduit à une optimisation du métabolisme. Nous avons ainsi démontré l'implication directe d'Aq_477, rebaptisée SbdP pour Sulfur -binding -donating Protein, dans le métabolisme énergétique du soufre de la bactérie. Une analyse poussée du génome nous a permis de construire un nouveau modèle suggérant notamment un recyclage des composés soufrés entre différents systèmes enzymatiques. La recherche de l'existence d'un niveau d'organisation des complexes respiratoires supérieur aux supercomplexes chez Aquifex aeolicus nous a conduits à développer de nouvelles méthodes d'étude permettant de proposer plusieurs pistes de recherche. Enfin, nous avons montré l'existence d'un nanocompartiment protéique constitué de l'encapsuline Aq_1760, dans lequel vient s'ancrer la ferritine atypique à domaines en tandem Aq_331. / Sulfur can be used in energy metabolism by microorganisms as electron donor and acceptor. The hyperthermophilic bacterium Aquifex aeolicus, which need oxygen, hydrogen and an essential sulfur compound for its growth presents sulfur reduction and oxidation pathways linked to the energy synthesis. A sulfur reductase (reduction of sulfur chains), a Sulfide Quinone Oxidoreductase (SQR, oxidation of H2S) and a Sulfur Oxygenase Reductase (SOR, simultaneous oxidation and reduction of sulfur chains) have been characterized in this bacterium. It has also been shown that some of these enzymes are organized in membrane-bound supercomplexes.We have demonstrated that Aq_477, previously characterized as a sulfurtransferase belonging to the rhodanese superfamily, can load long sulfur chains and acts as a sulfur donor for its partners (sulfur reductase and SOR) which use these sulfur chains as substrate, thus optimizing the metabolism. These results show that Aq_477, renamed SbdP for Sulfur -binding -donating Protein, is involved in the sulfur energy metabolism of Aquifex aeolicus. The identification in the genome of some new proteins potentially involved in this metabolism permitted us to propose a new model which suggests a recycling of sulfur compounds between different enzymatic systems. We also looked for an organization level of respiratory complexes higher than supercomplexes, which led us to develop new study methods and propose several research trails. Finally, we have shown the existence of protein nanocompartment constituted by the encapsulin Aq_1760, in which the atypical tandem-domain ferritin Aq_331 is anchored.

We have described herein a convenient one-pot synthesis of lisulfides/diselenides from a-amino acids mediated by ammonium etrathiomolybdate in water. (Figure 1) (Figure) Figure 1 Transformation of α-amino acids into the corresponding tiiocyanates/selenocyanates/disulfides/diselenides Halo-de-amination of a-amino acids using HBr/NaNCte followed by treatment with ammonium tetrathiomolybdate (NH4)2]VloS4 jLb provided a general route for the the one-pot synthesis of chiral a,a' bis (dithio) carboxylic acids (Figure 1, 2b). The yields were moderate, limited mainly the moderate conversion of a-amino acids into the corresponding chiral a-bromides. It was possible to synthesize the 2-thiocyanto carboxylic acids from the corresponding a-amino acids by a similar strategy. Thus diazotization in the presence of KSCN yielded in the chiral 2-thiocyanto carboxylic acids in moderate yields (Figure 1, 3). Thiocyanato-de-amination thus afforded the thiocyanates which when treated with JJD provided the chiral disulfides (Figure 1, 4a). We could thus synthesize both enantiomers of the disulfide from a single enantiomer of the starting a-amino acid. (Figure 1, 4a,4b) Using a similar strategy we have also demonstrated an efficient method for the synthesis of chiral selenocyanates starting from a-amino acids, using selenocyanate anion as the nucleophile (Figure 1, 5). It is possible to demonstrate a one-pot synthesis of chiral diselenides by reductive coupling of selenocyanates using JJb. (Figure 1, 6) (for figure see the pdf file)

Les protéines à centre Fe-S sont impliquées dans de nombreux processus cellulaires. In vivo, la formation des centres Fe-S est réalisée par des machineries multi-protéiques dont ISC et SUF, conservées chez les eucaryotes et les procaryotes. D’autres composants participent à la formation des centres Fe-S chez les eucaryotes, comme la frataxine (FXN). La FXN est une protéine présente chez l’homme, les plantes, la levure ou encore les bactéries à Gram négatif. Chez les eucaryotes, l’absence de FXN conduit à des phénotypes drastiques comme une accumulation de fer dans la mitochondrie, une diminution drastique de l’activité d’enzymes à centre Fe-S ou encore des dommages oxydatifs. Chez l’homme, un déficit en FXN est responsable d’une maladie neurodégénérative, l’ataxie de Friedreich. A la différence des eucaryotes, chez les procaryotes comme Escherichia coli, l’absence de CyaY, homologue bactérien de la FXN, ne conduit à aucun des phénotypes évoqués ci-dessus.Durant ma thèse, je me suis intéressée au rôle de CyaY chez E. coli. J’ai montré que, in vivo, CyaY favorise la formation des centres Fe-S via la machinerie ISC. Un lien génétique entre CyaY et IscX a également pu être établi, montrant que ces deux protéines participent à la formation des centres Fe-S in vivo. Je me suis ensuite intéressée aux bases moléculaires pouvant expliquer la différence entre les phénotypes liés à l’absence de FXN chez les eucaryotes et les procaryotes. J’ai montré que le résidu 108 de IscU joue un rôle clé pour la dépendance de CyaY. Enfin, pour mieux comprendre le rôle de CyaY chez E. coli, j’ai réalisé une approche globale en caractérisant le transcriptome du mutant ∆cyaY. / Fe-S cluster containing proteins are involved in many cellular processes such as respiration, DNA repair or gene regulation. In vivo, Fe-S cluster biogenesis is catalysed by specific protein machineries, ISC and SUF, conserved in both eukaryotes and prokaryotes. Frataxin (FXN) is a small protein found in humans, plants, yeast and Gram negative bacteria. In eukaryotes, a defect in FXN leads to drastic phenotypes such as mitochondrial iron accumulation, drastic decrease of Fe-S cluster protein activity, sensitivity to oxidants. In humans, FXN deficiency is responsible for the neurodegenerative disease, Friedreich’s ataxia. In prokaryotes like E. coli, a defect in CyaY, the bacterial FXN homolog, does not lead to significant phenotypes compared to the wild-type strain. During my thesis, I investigated the role of the bacterial FXN CyaY in E. coli. I showed that, in vivo, CyaY assisted the ISC-catalyzed Fe-S cluster biogenesis. A genetic link was also observed between cyaY and iscX, demonstrating that these proteins participate in Fe-S cluster biogenesis. In a second part, I investigated the differences between the impact of the eukaryotic versus prokaryotic FXN. I showed that the IscU 108th residue is crucial for the CyaY-dependency. Finally, I used a transcriptomic approach to test whether CyaY has a global role in E. coli.

Glycosidases are carbohydrate processing essential enzymes necessary for the growth and development of all organisms such as intestinal digestion, post-translational processing of glycoproteins and the lysosomal catabolism of glycoconjugates. The function of these glycosidases is limited and studies are still in progress to understand their function at cellular level. In recent years, biological role of carbohydrates has resulted in various carbohydrate-based therapeutics2. These carbohydrates serve as a tool to study the function of glycosidases by inhibiting their active site. The concept of inhibition is yet another approach for the discovery of drugs. Glycosidase inhibitors studied are often sugar analogs and a wide range of such inhibitors are reported in the literature.3, 4 Thiosugars, in particular, have gained new perspectives owing to their electronic, geometric, conformational and flexibility differences, as sulfide moiety being less electronegative and more polarizable than the oxygen counter-part.5 These differences make the thiosugars distinct from their oxygen analogs and hence can mimic the active site of the enzyme. Many molecules are reported to be promising glycosidase inhibitors but are not easily accessible due to difficulties in their synthesis. Hence, the chemical synthesis of thio-analogs of carbohydrates, by synthetic routes, remains a major challenge. To address the complexity of synthesis and to make available new strategies, we envisioned the use of benzyltriethylammonium tetrathiomolybdate [BnEt3N]2MoS4, a versatile and efficient sulfur transfer reagent. Objectives of the study: a. Design novel thiosugars as glycosidase inhibitors. b. Devise strategy for the synthesis of novel thiosugars through a simple, practical approach. c. Evaluate the synthesized molecules as glycosidase and HIV-1 protease inhibitors, in silico. d. Study miscellaneous applications of the novel thiosugar-derived thialactones. The thesis is divided into five sections: Section A entitled “Synthesis of deoxythiosugars and thiosugar-based lactones” is divided into two parts, Part A and Part B. Part A – “An introduction and background on thiosugars and sulfur transfer reagents” has been provided. A brief discussion of sulfur transfer reagents in carbohydrate synthesis and earlier work related to the use of benzyltriethylammonium tetrathiomolybdate, [BnEt3N]2MoS4, as an efficient sulfur transfer reagent have been provided. Part B –“Design of inhibitors of glycosidases and HIV-1 protease” deals with the design of inhibitors of glycosidase and HIV-1 protease. The designed thiosugar molecules exhibit the characteristics of sugars and will act as planar molecules to mimic the active site conformation of a good inhibitor. Synthetic methodologies devised and adopted for the synthesis of constrained sugar-derived thialactones include: (a) Double displacement, (b) Displacement-cum-intramolecular thia-Michael addition, (c) Epoxide ring-opening-cum-intramolecular thia-Michael addition, and (d) Displacement-cum-epoxide ring opening in an intramolecular fashion. In all the above mentioned strategies, sulfur transfer step is the crucial step which was achieved by the use of benzyltriethylammonium tetrathiomolybdate [BnEt3N]2MoS46 as the key reagent. (a) Various constrained thialactones synthesized by double displacement strategy using tetrathiomolybdate as the sulfur transfer reagent are shown in Scheme – 1. (b) A number of constrained thialactones were synthesized following nucleophilic displacement-cum-intramolecular thia-Michael addition strategy as shown in Scheme – 2. (c) Synthesis of bicyclic thiolactones was achieved using the strategy of epoxide ring-opening-cum-intramolecular thia-Michael addition. (Scheme – 3) (d) A few bicyclic thialactones were synthesized through displacement-epoxide ring opening-cyclization as shown in Scheme – 4. The methodology was also utilized for the synthesis of thiosugar derivatives and azido-thialactones. (Fig. 1) Figure 1 Synthesis of deoxythiosugars: The bicyclic thialactones (designed as inhibitors) on reduction with borohydride exchange resin (BER) easily furnished the deoxythiosugars (Fig. 2). It is worth mentioning that the synthesis of these thiosugars as reported in the literature involved lengthy procedures whereas the present methodology turns out to be short and concise. Figure 2 Section B entitled “Synthesis of amines, β-amino acids and novel thiosugar-based dehydroamino acids” comprises a brief introduction on the importance of amines, β-amino acids and dehyroamino acids. In this section the effective utilization of benzyltriethylammonium tetrathiomolybdate as a key reagent for reductive transformations and its application in the synthesis of amines, β-amino acids and dehyroamino acids have been presented. A one pot reduction of azides to amines followed by intermolecular aza-Michael addition employing tetrathiomolybdate was achieved to furnish a number of different β-amino esters as shown in Scheme -4: Scheme 4 The study was further extended to the reduction of a few anomeric azides to afford the corresponding anomeric amines and derivatives. (Fig. 3) Figure 3 A one-pot thia-Michael addition-vinyl azide reduction in a tandem fashion employing benzyltriethylammonium tetrathiomolybdate was studied and was shown to be effective for the synthesis of thiosugar derived dehydroamino acid derivatives. (Scheme – 5) Scheme 5 Section C entitled “Molecular docking studies of deoxythiosugar probes” gives an overview of different glycosidases, HIV-1 protease and their inhibitors. This section also deals with a brief introduction on active site conformations of potent inhibitors. In this connection we have studied the crystal conformations of the synthesized molecules whose conformations were the same as that of the existing inhibitors in the active site. (Fig. 4) With this background in silico study of the synthesized deoxythiosugar probes was conducted on human glycosidases: α-mannosidase, α-galactosidase, β-glucosidase and HIV-1 protease, respectively. Figure 4 Molecular docking was carried out using Autodock suite, molecular modeling simulation. Separate docking procedures were employed for the four different receptors. The PDBs representing the four enzyme targets were 2V3D, 3H53, 1X9D and 3I8W for β–glucosidase, α–galactosidase, α–mannosidase and HIV–1 protease respectively. The control compounds used for α–mannosidase were mannostatin and kifunensine. NMB, THK, and BED were the positive controls for HIV–1 protease. Similarly, NBV and cyclophellitol were the controls used for β–glucosidase and NOJ, N–methyl calystegine B2 for α–galactosidase. (Fig. 5) Ligands TGSB68 and TGSB482 had the energy value of –6.49 kcal/mol comparable to that of the average reference value of the positive control, and thus, the potent candidate as identified by molecular docking to HIV-1 protease. (Fig. 6a) The control compounds used for α–mannosidase were mannostatin and kifunensine, which bind with mean binding energy of -9.11 and -5.56. In the case of α–mannosidase, the same compounds TGSB68 and TGSB482 were selected due to comparable energy and a good cluster size with that of positive control. (Fig. 6b) For β– glucosidase, ligands TGSC108 and TGSC236, which had comparable values to that of positive control was identified as the Figure 5 Figure 6 potent candidate. (Fig. 6c) In the case of α–galactosidase, again the ligands TGSB68 and TGSB482 were selected based on binding energies. (Fig. 6d) In conclusion, the concept analogy (deoxy nature, planarity, thiosugar framework, lactone moiety) for the design of inhibitors indeed worked positively. The results are really encouraging. An in vivo study of the synthesized novel thiosugar probes will certainly provide a potent inhibitor. Section D entitled “Research methodology” provides experimental procedures adopted with details of synthesis. Section E entitled “Bibliography” provides the references cited in this work.