Target-guided synthesis is an approach to drug discovery that uses the biological target as a template to direct synthesis of its own best inhibitors from small molecule fragments. The process bridges the gap between chemical synthesis of drug candidates and their biological binding assay, merging the two operations into a single process whereby the active site or a binding pocket within the structure of the biological target directly controls the assembly of the best inhibitor in situ. Two different approaches to target-guided synthesis, the thermodynamic approach, making use of reversible reactions, and the kinetic approach, which uses an irreversible reaction, have been employed to discover novel, isoform selective inhibitors of the glutathione transferase (GST) enzyme family – possible drug targets in cancer and parasitic disease treatments. The thermodynamic approach described in this thesis uses the aniline-catalysed reversible acyl hydrazone formation reaction to create a dynamic covalent library of bivalent ligands designed to bind the dimeric structure of GST. In the presence of GST one of the bivalent ligands was selectively amplified at the expense of the other library members. This ligand was shown, via biological assays, to be a specific inhibitor for one isoform of GST, the mu isoform mGSTM1-1. A kinetic approach has also been investigated as a way to identify novel bivalent GST inhibitors utilising the Huisgen 1, 3 dipolar cycloaddition reaction. An azide and alkyne fragment library was designed to bind across the dimeric GST structure. The inhibitor structures are therefore bivalent, containing two anchoring fragments known to bind to the GST active site, linked by a triazolopeptide spacer. The triazole provides the click chemistry disconnection, enabling rapid in situ screening of candidate alkyne and azide fragments for inhibitor discovery. Whilst the in situ reaction with GST yielded inconclusive results, a number of the triazole products were found to have low nanomolar inhibitory activity towards GST.

Cytosolic glutathione transferases (GSTs) are a large family of enzymes that play an important role in detoxification of xenobiotics. They catalyse the conjugation of the glutathione tripeptide (GSH) to a wide range of toxic electrophilic acceptors. The overall 3D folds and architectures of the catalytic sites of many GSTs are conserved. They are composed of a well conserved glutathione binding site (G-site) and a promiscuous hydrophobic binding site (H-site). The 3D structure and ligand specificity has allowed the sub-classification of the multiple isoforms within the soluble GST superfamily. GSTs are involved in the drug detoxification and so are the target of medicinal chemistry programmes but it has proven difficult to generate isoform-specific inhibitors due to their inherent promiscuity. In this project, Venughopal Bhat (University of Edinburgh, laboratory of Dr. Mike Greaney) and I have explored a new platform to probe enzyme specificity. Protein-directed dynamic combinatorial chemistry (DCC) allows the assembly and amplification of a ligand within the confines of a binding site. DCC was used as a tool to explore the promiscuous H-site of four eukaryotic GSTs. I purified recombinant forms of SjGST, hGST P1-1, mGST M1-1 and mGST A4-4 from E. coli and assayed them with the universal, synthetic GST substrate 1-chloro-2,4-dinitrobenzene (CDNB). Venughopal Bhat prepared a ten-member, thermodynamically-controlled, dynamic combinatorial library (DCL) of acyl hydrazones from a 1-chloro-2-nitrobenzene aldehyde and ten acylhydrazides. This DCL was incubated with each of the four GST isozymes (spanning diverse classes) and distinct amplification effects were observed for SjGST and hGST P1-1. I subsequently carried out several biophysical experiments in an attempt to rank each of the ligands. These experiements, coupled with molecular modelling, provided insight into the basis of the observed selectivity. Bacterial GSTs are thought to play a role in primary metabolism and display a different GSH-conjugation mechanism compared to the eukaryotic GSTs. A recombinant form of the beta-class GST from the pathogenic bacterium Burkholderia cenocepacia was isolated, purified and biochemically characterised. The same ten-member acylhydrazone DCL was interfaced with the bacterial GST which was shown to amplify a hydrophobic library member that shared structural features with the known substrate 2-hydroxy-6-oxo-6-phenyl-2,4-dienoate (HOPDA). With the collaboration of Venughopal Bhat, I attempted to explore the putative active site of a GST-like protein with an unknown function using the same DCL. Although no amplification was observed, a new aldehyde template was suggested for future DCC experiments on this protein. GSTs are widely employed in biotechnology as protein fusion tags to enhance target protein solubility coupled with a facile enzyme assay. Manish Gupta and Juan Mareque-Rivas (University of Edinburgh) used the N-terminal, hexahistidine-tagged SjGST to demonstrate that quantum dots (QDs) coated with nitrilotriacetic acid (NTA) bound to Ni2+ ions can be used to reversibly and selectively bind, purify, and fluorescently label a His6-tagged GST in one step with retention of enzymatic activity. For this prupose, I purified and characterized both the untagged and hexahistidinetagged – SjGST prior to their experiments.

The advantage of one-bead-one-compound combinatorial libraries is that hundreds of thousands to millions of compounds can be rapidly synthesized and screened simultaneously. The beads supporting the compounds of interest are then isolated and analyzed to decipher the structure of the desired compound. Many methods are currently used to allow deconvolution of the compound on the individual beads. Herein is described a novel method to encode TentaGel beads using absorption of different ratios of lanthanide ions. The encoding process is completed in parallel with the synthesis of the library of compounds. Once the desired beads are identified, the lanthanide ions can be released from the bead and analyzed using Inductively Coupled Plasma-Mass Spectrometry (ICP-MS).

Peptide nucleic acids (PNAs) were used as encoding tags to enable the analysis of peptide libraries by PNA/DNA hybridisation onto DNA microarrays. This allowed entire peptide libraries to be organised and sorted in a two dimensional format whereby all library members could be interrogated and analysed on a one-byone basis. In this thesis, PNA-encoded peptide libraries, generated by split-and-mix library synthesis, were screened for a variety of functions. Peptide sequences identified from the screening of a PNA-encoded library were analysed in detail as the first specific substrates for chymopapain. A new PNAencoded library consisting of D-amino acids was synthesised and screened with a number of proteases in attempts to identify novel/unusual substrates. PNA-encoded libraries were also used in the screening of peptide libraries for other activities. Thus substrates for catalyst-free Hüisgen cycloaddition were identified following the reaction between an alkyne modified peptide library and azidofluorescein, while cell-penetrating peptides were identified by hybridization of an internalized encoded library onto a DNA microarray.

The advantage of one-bead-one-compound combinatorial libraries is that hundreds of thousands to millions of compounds can be rapidly synthesized and screened simultaneously. The beads supporting the compounds of interest are then isolated and analyzed to decipher the structure of the desired compound. Many methods are currently used to allow deconvolution of the compound on the individual beads. Herein is described a novel method to encode TentaGel beads using absorption of different ratios of lanthanide ions. The encoding process is completed in parallel with the synthesis of the library of compounds. Once the desired beads are identified, the lanthanide ions can be released from the bead and analyzed using Inductively Coupled Plasma-Mass Spectrometry (ICP-MS).

Diss. (sammanfattning) Stockholm : Karol. inst., 2004. / Härtill 5 uppsatser.

Thesis (Ph. D. in Chemistry)--Vanderbilt University, Dec. 2004. / Title from title screen. Includes bibliographical references.

Thesis (Ph.D.) - University of St Andrews, May 2009. / Restricted until 29th May 2010.

Thesis (Ph.D.) -- University of Texas Southwestern Medical Center at Dallas, 2006. / Not embargoed. Vita. Bibliography: 262-264.

Ph.D.