Logo: to the web site of Uppsala University

As antibiotic resistance of various pathogens emerged globally, the need for new effective drugs with novel modes of action became urgent. In this thesis, we focus on infectious diseases, e.g. tuberculosis, malaria, and nosocomial infections, and the corresponding causative pathogens, Mycobacterium tuberculosis, Plasmodium falciparum, and the Gram-negative ESKAPE pathogens that underlie so many healthcare-acquired diseases. Following the same-target-other-pathogen (STOP) strategy, we attempted to comprehensively explore the properties of three promising drug targets.

Signal peptidase I (SPase I), existing both in Gram-negative and Gram-positive bacteria, as well as in parasites, is vital for cell viability, due to its critical role in signal peptide cleavage, thus, protein maturation, and secreted protein transport. Three factors, comprising essentiality, a unique mode of action, and easy accessibility, make it an attractive drug target. We have established a platform, investigating the protein purification, enzymatic kinetics, and inhibition. A full-length SPase I from E. coli, including two transmembrane segments, was produced and purified in the presence of 0.5 % Triton X-100. In the in vitro biochemical assay, it exhibits proteolytic activity on antigen 85A from M. tuberculosis, with a K_m of 20 µM and a k_cat of 135 s^-1_­. A series of macrocyclic oligopeptides that have been proven inhibitory to E. coli SPase I also showed potency against a panel of Gram-negative bacteria. 

1-Deoxy-D-xylulose 5-phosphate reductoisomerase (DXR) is responsible for the production of methylerythritol phosphate (MEP) in the non-mevalonate pathway of isoprenoid biosynthesis, and is thus essential for cell growth. DXRs from M. tuberculosis and P. falciparum have been under investigation in our lab for years. I addressed structural and biochemical characterizations of PfDXR with analogs of 3-(N-formyl-N-hydroxyamino)propyl- phosphonate (fosmidomycin) and 3-(N-acetyl-N-hydroxyamino)propyl- phosphonate (FR-9000098), two natural products showing potency against P. falciparum. Chemical modifications, methylation at Cg, and double bond formation between Ca and Cb, were investigated to increase the pathogenicidal activity. Crystallographic complex structures of PfDXR and four novel compounds inhibitory to PfDXR in a dose-dependent manner were solved, and ligand binding will be discussed in detail.

Type II NADH dehydrogenase (NDH-2) is an essential component in the respiratory chain, playing an important role in electron transfer. Biomembrane-bound NDH-2 from M. tuberculosis was over-expressed in E. coli, as well as the homolog from M. smegmatis. The purified NDH-2s were kinetically characterized, and showed a similar affinity to previously reported NDH-2s expressed M. smegmatis. A collection of novel inhibitors in the scaffold of quinolinyl pyrimidines were synthesized and tested for inhibition in a biochemical assay.

The integration of computational methods into the drug discovery process provides valuable tools to help advance new and improved drugs into the clinic. As medicinal chemists explore novel targets and new areas of chemical space, our computational toolkit must also evolve.

Macrocycles are high-value scaffolds in medicinal chemistry due to their attractive physiochemical properties and intricate interactions with biomolecules. However, given the significant challenges associated with their synthesis, improved computational tools are required to both understand macrocycle conformation and binding preferences and to also efficiently guide medicinal chemistry efforts. Therefore, this thesis focuses on the evaluation, optimization and application of computational methods for macrocycle drug discovery.

Our initial work, Paper I, investigated both rigid and flexible macrocycle docking techniques. This showed that rigid docking of conformational ensembles generated using a range of sampling methods could result in significant differences in docking accuracy. Furthermore, we showed that either rigid docking of MD/LLMOD generated conformers or flexible docking could be applied.

In Paper II, we conducted further investigations of macrocycle conformational sampling by comparing more general sampling methods to those specialized towards macrocyclic scaffolds. The study showed that the general conformational sampling methods perform well compared with the more specialized methods. Our work also shows that the general methods can themselves be modified for improved macrocycle sampling.

Building on these findings, Paper III compares the conformational preferences of linear ligands and their closely related macrocyclic analogs. Interestingly, our analysis showed that for many of the macrocyclic ensembles they were not significantly more focused towards the bioactive conformation than their linear analogs.

In Paper IV, our computational toolkit was used to design novel antibacterial macrocycles targeting signal peptidase I. Here we developed macrocyclic compounds with nanomolar inhibitory activity against signal peptidase I, which also showed antibacterial activity.

The discovery and implementation of antibiotics for clinical use was unquestionably the greatest medical breakthrough of the 20th century. However, the widespread misuse and overuse of these antibiotics, has led to the rapid emergence and spread of antibiotic resistance. The 'ESKAPE' pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species) represent a critical threat in multidrug-resistant infections. The Gram-negative species (such as E. coli, K. pneumoniae, P. aeruginosa, and A. baumannii) are especially difficult to combat due to their dual-membrane and efficient efflux pumps, which limit the efficacy of many antibiotics. Despite significant efforts, no new antibiotic class with a new mechanism of action has been approved for Gram-negative pathogens in over five decades. New chemical classes of antibacterial compounds targeting distinct mechanisms within Gram-negative bacteria are therefore urgently called for. The studies outlined in this thesis addresses these challenges by designing and synthesising new antibacterial compounds of three distinct chemical classes, which interact with two unrealized targets, LepB and LpxH, in Gram-negative bacteria, including E. coli and K. pneumoniae.  This thesis investigates the effect of macrocyclization of type I signal peptidase (LepB) inhibitors by optimizing previously studied linear lipopeptide boronic acids and esters to address their cytotoxic and hemolytic liabilities while retaining activity. This resulted in the synthesis of first-in-class P2-P1' boronic ester-linked macrocycles with modest improvement of cytotoxicity but at the cost of reduced antibacterial activity (paper I). In another optimization attempt, isosteric modification of LepB inhibitors was explored by introducing the sulfonimidamide motif into oligopeptide boronic esters, displaying potent LepB inhibitors. Prior to the synthesis of these pseudopeptides novel methods were developed to introduce sulfonimidamides into peptides on solid-phase (paper II and paper III). These studies demonstrated the potential of sulfonimidamides to alter the drug properties and which was herein compared to a corresponding sulfonamide. Additionally, the thesis describes how a new series of LpxH inhibitors, meta-sulfonamidobenzamide-based sulfonyl piperazine derivatives, were identified and prepared. This resulted in inhibitors with wild-type activity without causing hemolysis, cell toxicity or inhibition of hERG ion channel (paper IV). In summary, strategies suitable for the synthesis and optimization of new antibacterial compounds targeting two distinct Gram-negative bacterial targets, LepB and LpxH, are described in this thesis. While there was no success in separating toxicity from antibacterial activity of the LepB inhibitors, these results highlight the challenge in this task and contribute to a better understanding of the structure-activity and toxicity relationships of such inhibitors and provide strategies that could be of use in antibacterial drug discovery. The identification of the meta-sulfonamidobenzamide derivatives offer promising LpxH-targeting hits with the potential for further development in future hit-to-lead antibacterial programs.