Until recently, the standard therapy for hepatitis C treatment has been interferon and ribavirin. Such treatment has only 50% efficacy and is not well tolerated. The emergence of new drugs has increased the treatment efficacy to 90%. Despite such an achievement, the success is limited since the virus mutates rapidly, causing the emergence of drug resistant forms. In addition, most new drugs were developed to treat genotype 1 infections. Thus, development of new potent antivirals is needed and drug discovery against hepatitis C is continued. In this thesis, a FRET-based protease assay was used to evaluate new pyrazinone based NS3 protease inhibitors that are structurally different to the newly approved and currently developing drugs. Several compounds in this series showed good potencies in the nanomolar range against NS3 proteases from genotype 1, 3, and the drug resistance variant R155K. We assume that these compounds can be further developed into drug candidates that possess activity against above mentioned enzyme variants. By using SPR technology, we analyzed interaction mechanisms and characteristics of allosteric inhibitors targeting NS5B polymerases from genotypes 1 and 3. The compounds exhibited different binding mechanisms and displayed a low affinity against NS5B from genotype 3. In order to evaluate the activity and inhibitors of the NS5B polymerase, we established an SPR based assay, which enables the monitoring of polymerization and its inhibition in real time. This assay can readily be implemented for the discovery of inhibitors targeting HCV. An SPR based fragment screening approach has also been established. A screen of a fragment library has been performed in order to identify novel scaffolds that can be used as a starting point for development of new allosteric inhibitors against NS5B polymerase. Selected fragments will be further elaborated to generate a new potent allosteric drug candidate. Alternative approaches have successfully been developed and implemented to the discovery of potential lead compounds targeting two important HCV drug targets.

Deciphering the activity of chemical molecules against a pathogenic organism is an essential task in drug discovery process. Virtual screening, in which few plausible molecules are selected from a large set for further processing using computational methods, has become an integral part and complements the expensive and time-consuming in vivo and in vitro experiments. To this end, it is essential to extract certain features from molecules which in the one hand are relevant to the biological activity under consideration, and on the other are suitable for designing fast and robust algorithms. The features/representations are derived either from physicochemical properties or their structures in numerical form and are known as descriptors. In this work we develop two new molecular-fragment descriptors based on the critical analysis of existing descriptors. This development is primarily guided by the notion of coding degeneracy, and the ordering induced by the descriptor on the fragments. One of these descriptors is derived based on the simple graph representation of the molecule, and attempts to encode topological feature or the connectivity pattern in a hierarchical way without discriminating atom or bond types. Second descriptor extends the first one by weighing the atoms (vertices) in consideration with the bonding pattern, valence state and type of the atom. Further, the usefulness of these indices is tested by ranking and classifying molecules in two previously studied large heterogeneous data sets with regard to their anti-tubercular and other bacterial activity. This is achieved by developing a scoring function based on clustering using these new descriptors. Clusters are obtained by ordering the descriptors of training set molecules, and identifying the regions which are (almost) exclusively coming from active/inactive molecules. To test the activity of a new molecule, overlap of its descriptors in those cluster (interpolation) is weighted. Our results are found to be superior compared to previous studies: we obtained better classification performance by using only structural information while previous studies used both structural features and some physicochemical parameters. This makes our model simple, more interpretable and less vulnerable to statistical problems like chance correlation and over fitting. With focus on predictive modeling, we have carried out rigorous statistical validation. New descriptors utilize primarily the topological information in a hierarchical way. This can have significant implications in the design of new bioactive molecules (inverse QSAR, combinatorial library design) which is plagued by combinatorial explosion due to use of large number of descriptors. While the combinatorial generation of molecules with desirable properties is still a problem to be satisfactorily solved, our model has potential to reduce the number of degrees of freedom, thereby reducing the complexity.

Inter-molecular interaction is at the heart of biological function. Proteins can interact with ligands, peptides, small molecules, and other proteins to serve their structural or functional purpose. With advances in combinatorial chemistry and the development of high throughput binding assays, the available inter-molecular interaction data is increasing exponentially. As the space of testable compounds increases, the complexity and cost of finding a suitable inhibitor for a protein interaction increases. Computational drug discovery plays an important role in minimizing the time and cost needed to study the space of testable compounds. This work focuses on the usage of various computational methods in identifying protein interaction inhibitors and demonstrates the ability of computational drug discovery to contribute to the ever growing field of molecular interaction. A program to predict the location of binding surfaces on proteins, STP (Mehio et al., Bioinformatics, 2010, in press), has been created based on calculating the propensity of triplet-patterns of surface protein atoms that occur in binding sites. The use of STP in predicting ligand binding sites, allosteric binding sites, enzyme classification numbers, and binding details in multi-unit complexes is demonstrated. STP has been integrated into the in-house high throughput drug discovery pipeline, allowing the identification of inhibitors for proteins whose binding sites are unknown. Another computational paradigm is introduced, creating a virtual library of -turn peptidomimetics, designed to mimic the interaction of the Baff-Receptor (Baff-R) with the B-Lymphocyte Stimulator (Blys). LIDAEUS (Taylor, et al., Br J Pharmacol, 2008; 153, p. S55-S67) is used to identify chemical groups with favorable binding to Blys. Natural and non-natural sidechains are then used to create a library of synthesizable cyclic hexapeptides that would mimic the Blys:Baff-R interaction. Finally, this work demonstrates the usage and synergy of various in-house computational resources in drug discovery. The ProPep database is a repository used to study trends, motifs, residue pairing frequencies, and aminoacid enrichment propensities in protein-peptide interaction. The LHRLL protein-peptide interaction motif is identified and used with UFSRAT (S. Shave, PhD Thesis, University of Edinburgh, 2010) to conduct ligand-based virtual screening and generate a list of possible antagonists from the EDULISS (K. Hsin, PhD Thesis, University of Edinburgh, 2010) compound repository. A high throughput version of AutoDock (Morris, et al., J Comput Chem, 1998; 19, p. 1639-62) was adapted and used for precision virtual screening of these molecules, resulting in a list of compounds that are likely to inhibit the binding of this motif to several Nuclear Receptors.

Methamphetamine abuse is a serious public health concern affecting millions of people worldwide, and there are currently no viable pharmacotherapies to treat methamphetamine abuse. Methamphetamine increases extracellular dopamine (DA) concentrations through an interaction with the DA transporter (DAT) and the vesicular monoamine transporter-2 (VMAT2), leading to reward and abuse. While numerous studies have focused on DAT as a target for the discovery of pharmacotherapies to treat psychostimulant abuse, these efforts have been met with limited success. Taking into account the fact that methamphetamine interacts with VMAT2 to increase DA extracellular concentrations; the focus of the current work was to develop novel compounds that interact with VMAT2 to inhibit the effects of methamphetamine. Lobeline, the principal alkaloid found in Lobelia inflata, inhibits VMAT2 binding and function. Inhibition of VMAT2 was hypothesized to be responsible for the observed lobeline-induced inhibition of methamphetamine-evoked DA release in striatal slices and decrease in methamphetamine self-administration in rats. Lobeline has recently completed Phase Ib clinical trials demonstrating safety in methamphetamine abusers. Lobeline is also a potent inhibitor of nicotinic acetylcholine receptors (nAChRs), limiting selectivity for VMAT2. Chemical defunctionalization of the lobeline molecule afforded analogs, meso-transdiene (MTD) and lobelane, which exhibited decreased affinity for nAChRs. MTD, an unsaturated analog of lobeline, exhibited similar affinity for VMAT2 and increased affinity for DAT compared to lobeline. Conformationally-restricted MTD analogs exhibited decreased affinity for DAT compared to MTD, while retaining affinity at VMAT2. One analog, UKMH-106 exhibited high affinity and selectivity for VMAT2 and inhibited METH-evoked DA release from striatal slices. Unfortunately, the MTD analogs exhibited poor water solubility which limited further investigation of these promising analogs. Importantly lobelane, a saturated analog of lobeline, exhibited increased affinity and selectivity for VMAT2 compared to lobeline. To improve water solubility, a N-1,2-dihydroxypropyl (diol) moiety was incorporated into the lobelane molecule. GZ-793A, an N-1,2-diol analog, potently and competitively inhibited VMAT2 function, exhibiting over 50-fold selectivity for VMAT2 over DAT, serotonin transporters and nAChRs. GZ-793A released DA from preloaded synaptic vesicles, fitting a two-site model with the high-affinity site inhibited by tetrabenazine and reserpine (classical VMAT2 inhibitors), suggesting a VMAT2-mediated mechanism of release. Further, low concentrations of GZ-793A that selectively interact with high-affinity sites on VMAT2 to evoke DA release, inhibit methamphetamine-evoked DA release from synaptic vesicles. Results showed that increasing concentrations of GZ-793A produced a rightward shift in the METH concentration response; however, the Schild regression revealed a slope different from unity, consistent with surmountable allosteric inhibition. In addition, GZ-793A specifically inhibited methamphetamine-evoked DA release in striatal slices and methamphetamine self-administration in rats. To examine the possibility that GZ-793A produced DA depletion, the effect of a behaviorally active dose of GZ-793A on DA content in striatal tissue and striatal vesicles was determined. GZ-793A administration did not alter DA content in striatal tissue or vesicles and pretreatment with GZ-793A prior to methamphetamine administration did not exacerbate the DA depleting effects of methamphetamine. Importantly, GZ-793A was shown to protect against methamphetamine-induced striatal DA depletions. Thus, GZ-793A represents an exciting new lead in the development of pharmacotherapies to treat methamphetamine abuse.

Fragment based drug discovery (FBDD) has been applied to two protease drug targets, MMP-12 and HIV-1 protease. The primary screening and characterization of hit fragments were performed with surface plasmon resonance -technology. Further evaluation of the interaction was done by inhibition studies and in one case with X-ray crystallography. The focus of the two projects was different. Many MMP inhibitors contain a strong zinc chelating group, hydroxamate, interacting with the catalytic zinc atom. This strategy may be the cause for the low specificity of MMP inhibitors. Using FBDD we found a fragment with an unusual strong affinity for MMP-12. An inhibition assay confirmed that it was an inhibitor but indicated a stoichiometry of 2:1. Crystallography data revealed that an adduct of the fragment was bound in the active site, with interactions both with the catalytic zinc and the S1’ pocket. This may present a new scaffold for MMP-12 inhibitors. For HIV-1 protease the focus was on identifying inhibitors not sensitive to current resistance mutations. A fragment library for screening with SPR-technology was designed and used for screening against wild type enzyme and three variants with resistance mutations. Many of the hits were promiscuous but a number of fragments with possible allosteric inhibition mechanism were identified. The temperature dependency of the dissociation rate and reported resistance mutations was studied with thermodynamics. A good, but not perfect correlation was found between resistance and both the dissociation data and the free energy for dissociation compared to data from wild type enzyme. However, the type of mutation also influenced the results. The flap mutation G48V displayed thermodynamic profiles not completely correlating with resistance. It was found that dissociation rate and thermodynamics may complement each other when studying resistance, but only one of them may not be enough.

Viral proteases are important targets for anti-viral drugs. Discovery of protease inhibitors as anti-viral drugs is aided by an understanding of the interactions between viral protease and inhibitors. This thesis addresses the characterization of protease-inhibitor interactions for application to drug discovery and model system development. The choice of a relevant target is essential to molecular interaction studies. Therefore, full-length NS3 protein of hepatitis C virus (HCV) was obtained, providing a more relevant target and a better model for the development of HCV protease inhibitors. In addition, resistance to anti-viral drugs, a serious problem in the treatment of AIDS, prompted the investigation of resistant variants of human immunodeficiency virus (HIV) protease. Drug resistance was initially explored by characterization of the interactions between a series of closely related inhibitors and resistant variants of HIV protease, using an inhibition assay to determine the inhibition dissociation constants (Ki). The relationship between structure, activity and resistance profiles was not clarified, indicating that the effect of structural changes in the inhibitors and the protease are not predictable and must be analyzed case wise. It was proposed that additional kinetic characterization of the interactions was required and a biosensor-based method allowing for determination of affinity, KD, and interaction rate constants, kon and koff, was adopted. The increased physiological relevance of this method was confirmed, and the affinity data have better correlation with cell culture data. In addition, interactions between clinical inhibitors of HIV protease and enzyme variants indicate that increased dissociation rates (koff) are associated with the development of resistance. Thermodynamic characterization of the interactions between HIV-1 protease and clinically relevant inhibitors revealed distinct energetic characteristics for inhibitors. The resolution of the energetics of association and dissociation identified an inhibitor with unique interaction characteristics and confirmed the validity of using this method for further characterization of molecular interactions. This work resulted in the development of model systems for the analysis of kinetics, resistance and thermodynamic characteristics of protein-inhibitor interactions. The results give increased understanding of the biomolecular interactions and can be applied to drug discovery.

A variety of HIV-1 protease inhibitors and their interactions with the enzyme have been characterized in order to identify novel and improved drugs against AIDS. The investigated inhibitors were represented by clinical and non-clinical inhibitors, active site and allosteric inhibitors, transition-state analogues and metal-ions. In addition, different enzyme variants were used to investigate the contribution of different amino acid residues to the interaction with different ligands. The problem of resistance has been addressed by exploring novel types of inhibitors, and resistant mutants of HIV-1 protease. A study resolving the inhibition of HIV-1 protease by Cu2+ showed that the enzyme can be allosterically inhibited and that copper inhibition is a result of an interaction with His-69 and a subsequent conformational change. Several types of transition-state analogues were analyzed with respect to their inhibition of wild-type and resistant mutants of HIV-1 protease. Unfortunately cyclic compounds were not found to be better than linear compounds. Moreover, it was not possible to identify structure-activity relationships that clearly correlated with efficacy towards mutants and a biosensor based method for more detailed kinetic studies was therefore adopted. By cross-linking the immobilized enzyme on the biosensor matrix, a stable surface was obtained and kinetic rate constants could be determined for the interaction between the enzyme and inhibitors. Additional improvements in the methodology involved identification of a more representative interaction model, allowing more detailed studies of interactions with resistant mutants and varying conditions. Finally, absorption to lipid membranes and interaction with human serum albumin and α1-glycoprotein by clinical drugs were studied in a simplified ADME model system for improvement of the earlier stages of drug development. These studies have revealed important characteristics of these drugs that can potentially be modeled into new compounds that have improved efficacy of both wild-type and resistant mutants of HIV-1 protease.

The present study has focused on characterising protease-ligand interactions in the context of drug discovery. The proteases that have been studied are human matrix metallopeptidase 12 (MMP-12), HIV-protease and Hepatitis C virus (HCV) NS3/NS4A protease. These studies have involved kinetic characterisation of protease-inhibitor interactions using biosensor technology, as well as determination of inhibition and activity regulation by using activity assays. The regulation of MMP-12 activity by calcium was proposed, based on the study of the calcium dependence of MMP-12 activity. Furthermore, it was shown that the high affinity of hydroxamate-based inhibitors of MMP-12 were due to slow dissociation of the enzyme-inhibitor complex by using a new biosensor assay for the study of interactions between MMP-12 and ligands. A study of the pH-dependency of protease-inhibitor interactions revealed that the interaction kinetics of HIV-protease inhibitors differed with pH in a way that could be related to the inhibitor structures. This suggested that the forces of interaction are different in the association and dissociation phases of an interaction. Furthermore, it demonstrated the usefulness of pH as a variable in characterising protein-ligand interactions. Results applicable in the discovery of drugs against Hepatitis C were obtained, with the analysis of structure-activity relationships of novel inhibitors. Furthermore, the mode of binding imposed by key functional groups of the inhibitors was explored by investigating the effect of pH on the interactions with NS3. The results show the importance of using appropriate model systems for drug discovery by selecting relevant targets and assay conditions. Furthermore, the usefulness of kinetic rate information in drug discovery is demonstrated. Thus, by contributing to the knowledge of protease-ligand interactions, applicable to both protease inhibitor interactions and protease activity regulation, this thesis is expected to have an impact on the field of protease inhibitor development and drug discovery in general.

Erythrocyte pyruvate kinase (R-PK) is a key glycolytic enzyme catalyzing the transphosphorylation of phosphoenolpyruvate (PEP) and ADP to pyruvate and ATP respectively3,4. The substrate PEP and product pyruvate of this reaction are involved in a number of energetic and biosynthetic pathways; hence a tight regulation of R-PK activity is crucial not only for glycolysis, but also for the entire cellular metabolism. Deficiency of R-PK is one of the most common enzymatic defects of RBC, and may be caused by mutations of the PK-LR (pyruvate kinase liver red blood cell) gene31, 32. Clinically, R-PK deficiency manifests itself as a chronic life-long hemolysis ranging from very mild or fully compensated anemia to life-threatening neonatal anemia and pronounced jaundice. Current treatment options are limited to continuous blood transfusions and splenectomy. Thus, there is an urgent need for medications to counter R-PK deficiency without resorting to these complicated procedures. Our aim is to identify novel allosteric modifiers of R-PK using a combination of computational studies and enzyme activity assays. Such compounds could be of medical interest. Human R-PK was expressed in DH-5α cells and was purified by the procedure reported by Wang et al57. However, this method gave a very low yield of R-PIS (5mg/L). In an attempt to increase the yield, we expressed R-PK in Rosetta strain cells. Further, addition of His-tag to the protein's N-terminus simplified purification to a one step Ni-NTA (Nickel- nitrilotriacetic acid) column resulting in a 6-fold increase in the yield. Computational methods were applied to identify small molecules that bind to the allosteric activator fructose 1,6-bisphosphate (FBP) binding site of R-PK to identify compounds that could interact with the protein. The software UNITY, as present in the molecular modeling software Sybyl was used to perform 3D searches against the National Cancer Institute (NCI) chemical database. From these searches we obtained 29 hits that were subjected to further computational analysis. The small molecules were docked into the FBP binding site of R-PK with different docking methods includingFlexX, GOLD and energy minimization. The energy scoring function of HINT was then applied to analyze the interactions between the docked molecules and R-PK. Compounds with highest HINT score were requested fiom NCI and were subjected to further kinetic analysis to identify possible allosteric effectors of R-PK.In the kinetic analysis, we employed a lactate dehydrogenase (LDH) coupled spectrophotometric assay to determine the activity of R-PK in the presence of these compounds. The steady state kinetics of R-PK gave a typical S-shaped curve that fitted a signloidal function indicative of allosterism. All the kinetic parameters of our enzyme were in excellent agreement with native R-PK activity as previously reported5, 57. R-PK activity in the presence of the analyzed compounds revealed both activators and inhibitors of R-PK. X-ray crystallographic analysis of R-PK in the presence of FBP and the identified small molecule effectors are currently in progress. These experiments were initiated to reveal the binding site of the compounds in R-PK, allowing for further optimization of the starting phannacophores and syntheses of new molecular entities for enhanced allosteric activity.In conclusion, we have developed a simple and efficient method for the expression and purification of R-PK. Using computational screening and HINT analysis we have also identified several compounds that interact with R-PK and kinetic analysis revealed both activators and inhibitors of the protein. Crystals of R-PK in the presence of effectors have been obtained and identification of the binding site on R-PK is under investigation. R-PK effectors discovered in this study could prove to be lead compounds for developing medications for the treatment of anemia and other disorders arising from R-PK malfunction.

The gene for methylthioadenosine phosphorylase (MTAP) is absent in almost 30% of cancers, opening a door for selective chemotherapy. One strategy to target the absence of MTAP involves the design of a cytotoxic methylthioadenosine (MTA) analogue. Non-cancerous cells would break down the cytotoxic analogue, since they contain MTAP, but cancerous cells would not, since they do not have MTAP. However, before such a compound can be made, we need to better understand the types of substrates accommodated by MTAP. The purpose of this thesis was therefore to explore a series of MTA analogues, probing the structure-function relationships between MTAP and specific structural modifications of MTA. Nine phenylthioadenosine (PTA) derivatives bearing ortho-, meta-, or para- methyl carboxylate, carboxylate, and hydroxymethyl substituents were synthesized and tested for cytotoxicity and as substrates for MTAP. The biological results of these nine compounds suggested that addition of substituents to the ortho-position was not tolerated by MTAP, and substituents similar to the hydroxymethyl might be accommodated by MTAP. None of the compounds were cytotoxic. This informed the design of ten more PTA derivatives, most of which were synthesized and tested for cytotoxicity and as substrates for MTAP. The range of functionalities included an amine, an acetamide, a urea, an isovaleramide, and an N-nitrosourea group inspired by the known anticancer agent lomustine. The amine derivatives of PTA were the best substrates of all MTA analogues tested (including PTA). The meta-amine derivative and the meta-isovaleramide showed minor cytotoxicity. Finally, the urea derivatives were moderate substrates of MTAP, and this pointed towards the future creation of other nitrosoureas as potential cytotoxic MTAP substrates. / Graduate / 2017-08-25