Refine
Has Fulltext
- yes (7)
Is part of the Bibliography
- yes (7)
Document Type
- Doctoral Thesis (7)
Language
- English (7) (remove)
Keywords
- Tuberkelbakterium (7) (remove)
Mycobacterium tuberculosis is the causative agent of tuberculosis and responsible for more than eight million new infections and about two million deaths each year. Novel chemotherapeutics are urgently needed to treat the emerging threat of multi drug resistant and extensively drug resistant strains. Cell wall biosynthesis is a widely used target for chemotherapeutic intervention in bacterial infections. In mycobacteria, the cell wall is comprised of mycolic acids, very long chain fatty acids that provide protection and allow the bacteria to persist in the human macrophage. The type II fatty acid biosynthesis pathway in Mycobacterium tuberculosis synthesizes fatty acids with a length of up to 56 carbon atoms that are the precursors of the critical mycobacterial cell wall components mycolic acids. KasA, the mycobacterial ß-ketoacyl synthase and InhA, the mycobacterial enoyl reductase, are essential enzymes in the fatty acid biosynthesis pathway and validated drug targets. In this work, KasA was expressed in Mycobacterium smegmatis, purified and co-crystallized in complex with the natural thiolactone antibiotic thiolactomycin (TLM). High-resolution crystal structures of KasA and the C171Q KasA variant, which mimics the acyl enzyme intermediate of the enzyme, were solved in absence and presence of bound TLM. The crystal structures reveal how the inhibitor is coordinated by the enzyme and thus specifically pinpoint towards possible modifications to increase the affinity of the compound and develop potent new drugs against tuberculosis. Comparisons between the TLM bound crystal structures explain the preferential binding of TLM to the acylated form of KasA. Furthermore, long polyethylene glycol molecules are bound to KasA that mimic a fatty acid substrate of approximately 40 carbon atoms length. These structures thus provide the first insights into the molecular mechanism of substrate recognition and reveal how a wax-like substance can be accommodated in a cytosolic environment. InhA was purified and co-crystallized in complex with the slow, tight binding inhibitor 2-(o-tolyloxy)-5-hexylphenol (PT70). Two crystal structures of the ternary InhA-NAD+-PT70 were solved and reveal how the inhibitor is bound to the substrate binding pocket. Both structures display an ordered substrate binding loop and corroborate the hypothesis that slow onset inhibition is coupled to loop ordering. Upon loop ordering, the active site entrance is more restricted and the inhibitor is kept inside more tightly. These studies provide additional information on the mechanistic imperatives for slow onset inhibition of enoyl ACP reductases.
With 9.6 million new cases and 1.5 million deaths in 2014, tuberculosis (TB) is alongside with AIDS the most deadly infection. Foremost, the increased prevalence of resistant strains of M. tuberculosis among the TB-infected population represents a serious thread. Hence, in the last decades, novel drug targets have been investigated worldwide. So far a relatively unexplored target is the cell wall enzyme β-ketoacyl-ACP-synthase “KasA”, which plays a crucial role in maintaining the membrane impermeability and hence the cell ability to resist to the immune response and drug therapy. KasA is a key enzyme in the fatty acid synthase “FAS-II” elongation cycle, responsible for the extension of the growing acyl chain within the biosynthesis of precursors for the most hydrophobic constituents of the cell wall – mycolic acids. Design of the novel KasA inhibitors, performed in the research group of Prof. Sotriffer by C. Topf and B. Schaefer, was based on the recently published crystal structure of KasA in complex with its known inhibitor thiolactomycin (TLM). Considering the essential ligand-enzyme interactions, a pharmacophore model was built and applied in the virtual screening of a modified ZINC database. Selected hits with the best in silico affinity data have been reported by Topf and Schaefer.
In this work, two of the obtained hits were synthesized and their structure was systematically varied. First, a virtual screening hit, chromone-2-carboxamide derivative GS-71, was modified in the amide part. Since the most of the products possessed a very low solubility in the aqueous buffer medium used in biological assays, polar groups (nitro, succinamidyl and trimethyl-amino substituent in position 6 of the chromone ring or hydroxyl group on the benzene ring in the amide part have been inserted to the molecule. Further variations yielded diaryl ketones, diaryl ketone bearing a succinamidyl substituent, carboxamide bearing a methylpiperazinyl-4-oxobutanamido group and methyl-malonyl ester amides. Basically, the essential structural features necessary for the ligand-enzyme interactions have been maintained. The latter virtual screening hit, a pyrimidinone derivative VS-8 was synthesized and the structure was modified by substitution in positions 2, 4, 5 and 6 of the pyrimidine ring. Due to autofluorescence, detected in most of the products, this model structure was not further varied.
Simultaneously, experiments on solubilization of the first chromone-2-carboxamides with cyclodextrins, cyclic oligosacharides known to form water-soluble inclusion complexes, were performed. Although the assessed solubility of the chromone 3b/DIMEB (1:3) mixture exceeded 14-fold the intrinsic one, the achieved 100 µM solubility was still not sufficient to be used as a stock solution in the binding assay. The experiments with cyclodextrin in combination with DMSO were ineffective. Owing to high material costs necessary for the appropriate cyclodextrin amounts, the aim focused on structural modification of the hydrophobic products.
Precise structural data have been obtained from the solved crystal structures of three chromone derivatives: the screening hit GS-71 (3b), its trimethylammonium salt (18) and 6-nitro-substituted N-benzyl-N-methyl-chromone-2-carboxamide (9i). The first two compounds are nearly planar with an anti-/trans-rotamer configuration. In the latter structure, the carboxamide bridge is bent out of the chromone plane, showing an anti-rotamer, too. Considering the relatively low partition coefficient of compound 3b (cLogP = 2.32), the compound planarity and correlating tight molecular packing might be the factors significantly affecting its poor solubility.
Regarding the biological results of the chromone-based compounds, similar structure-activity correlations could be drawn from the binding assay and the whole cell activity testing on M. tuberculosis. In both cases, the introduction of a nitro group to position 6 of the chromone ring and the presence of a flexible substituent in the amide part showed a positive effect. In the binding study, the nitro group at position 4 on the N-benzyl residue was of advantage, too. The highest enzyme affinity was observed for N-(4-nitrobenzyl)-chromone-2-carboxamide 4c (KD = 34 µM), 6-nitro substituted N-benzyl-chromone-2-carboxamide 9g (KD = 40 µM) and 6‑nitro-substituted N-(4-nitrobenzyl)-chromone-2-carboxamide 9j (KD = 31 µM), which could not be attributed to the fluorescence quenching potential of the nitro group. The assay interference potential of chromones, due to a covalent binding on the enzyme sulfhydryl groups, was found to be negligible at the assay conditions. Moderate in vivo activity was detected for 6‑nitro-substituted N-benzyl-chromone-2-carboxamide 9g and its N-benzyl-N-methyl-, N‑furylmethyl-, N-cyclohexyl- and N-cyclohexylmethyl derivatives 9i, 9d, 9e, 9f, for which MIC values 20 – 40 µM were assessed. Cytotoxicity was increased in the N‑cyclohexylmethyl derivative only. None of the pyrimidine-based compounds showed activity in vivo. The affinity of the model structure, VS-8, surpassed with KD = 97 µM the assessed affinity of TLM (KD = 142 µM).
Since for the model chromone compound GS-71 no reliable KasA binding data could be obtained, a newly synthesized chromone derivative 9i was docked into the KasA binding site, in order to derive correlation between the in silico and in vitro assessed affinity. For the 6‑nitro-derivative 9i a moderate in vivo activity on M. tuberculosis was obtained. The in silico predicted pKi values for TLM and 9i were higher than the corresponding in vitro results, maintaining though a similar tendency, i.e., the both affinity values for compound 9i (pKi predicted = 6.64, pKD experimental = 4.02) surpassed those obtained for TLM (pKi predicted = 5.27, pKD experimental = 3.84). Nevertheless, the experimental pKD values are considered preliminary results.
The binding assay method has been improved in order to acquire more accurate data. Owing to the method development, limited enzyme batches and solubility issues, only selected compounds could be evaluated. The best hits, together with the compounds active on the whole cells of M. tuberculosis, will be submitted to the kinetic enzyme assay, in order to confirm the TLM-like binding mechanism. Regarding the in vivo testing results, no correlations could be drawn between the predicted membrane permeability values and the experimental data, as for the most active compounds 9e and 9f, a very low permeability was anticipated (0.4 and 0.7 %, respectively). Further biological tests would be required to investigate the action- or transport mode.
The high infection rates and recent emergence of extremely drug resistant forms of
Mycobacterium tuberculosis pose a significant challenge for global health. The NADH-
dependent enoyl-ACP-reductase InhA of the type II mycobacterial fatty acid biosynthesis
pathway is a well-validated target for inhibiting mycobacterial growth. InhA has been
shown to be inhibited by a variety of compound series. Prominent classes of InhA
inhibitors from literature include diaryl ethers, pyrrolidine carboxamides and arylamides
which can be subjected to further development. Despite the progress in this area, very
few compounds are in clinical development phase. The present work involves a detailed
computational investigation of the binding modes and structure-based optimisation of
pyrrolidine carboxamides as InhA inhibitors.
With substituents of widely varying bulkiness, the pyrrolidine carboxamide dataset
presented a challenge for prediction of binding mode as well as affinity. Using advanced
docking protocols and in-house developed pose selection procedures, the binding modes
of 44 compounds were predicted. The poses from docking were used in short molecular
dynamics (MD) simulations to ascertain the dominant binding conformations for the
bulkier members of the series. Subsequently, an activity-based classification strategy
could be developed to circumvent the affinity prediction problems observed with this
dataset. The prominent motions of the bound ligand and the active site residues were
then ascertained using Essential Dynamics (ED). The information from ED and literature
was subsequently used to design a total of 20 compounds that were subjected to extensive
in-silico evaluations. Finally, the molecular determinants of rapid-reversible binding of
pyrrolidine carboxamides were investigated using long MD simulations.
The work deals with the synthesis and characterization of cerulenin analogues as inhibitors of efflux pump mediated resistance of Candida albicans isolates and as inhibitors of the fatty acid synthesis enzyme KasA of Mycobacterium tuberculosis. Cerulenin was chosen as the lead structure, being a substrate of the efflux pumps in Candida albicans on one hand and therefore variations on the structure could lead to a blocking of the efflux pumps as in the case of tetracycline and inhibitor 13-CPTC of the TetB efflux pump. On the other hand, cerulenin is a known inhibitor of the FAS system but inhibition is unselective in type I and II FAS. Therefore, analogues could result in increased selectivity towards the type II FAS system in M. tuberculosis. The first cerulenin derivatives were prepared by coupling 2,3-dihydrofuran to the before synthesized 1-octaniodide, followed by ring opening and oxidation in one step by chromic acid and transfer of the resulting 4-keto acid to amides to give analogues 4a-d, 4e was prepared in analogy. To include the epoxide function especially with regard to the mechanism of action of cerulenin in the FAS system (considering known crystal structures of cerulenin and the KasA analogue of E. coli) tetrahydro- and dihydrocerulenin analogues were synthesized. Starting from the corresponding aldehyde, lactone 5 (tetrahydrocerulenin analogues) was obtained via two different routes A and B. Route A included the coupling of the aldehyde 1-nonanal to propiolic acid via a Grignard reaction with subsequent hydrogenation with the Lindlar catalyst under hydrogen pressure to give 5. Via Route B 1-nonanal was coupled to methyl propiolate by n-BuLi with subsequent hydrogenation under reflux with the catalytic system Lindlar cat./NH4HCO2 to yield 5. These hydrogenations were also executed in a microwave oven resulting in better yields and/or reaction times. The lactone 5 was then epoxidized, the ring opened by amidation and the remaining alcohol was oxidized via Collins oxidation to result in tetrahydrocerulenin analogues 8a-e. The same procedure was used for dihydrocerulenin analogues 10a-c except that to obtain the corresponding lactone 9a only route A was used and a further step had to be executed for ring closure. To obtain analogues with all structural features of cerulenin including two double bonds and the epoxide function, a third pathway was chosen. To obtain the future side chain, aldehyde 12 was synthesized by coupling protected 4-pentyn-1-ol to either crotyl bromide or crotyl chloride, which then was deprotected, hydrogenated with Lindlar catalyst under hydrogen pressure and oxidized via a Swern oxidation. The following synthesis sequence starting from 12 was executed similar to that of dihydrocerulenins via the corresponding lactone (51) with the major exception of the oxidation procedure in the last step via TPAP/NMO to result in (4Z,7E)-cerulenin analogues 15a-b. A fourth class of cerulenin analogues was synthesized with the aromatic analogues 17a-e. This synthesis pathway started with the formation of the benzoyl acrylamides 16a-e from benzoylacrylic acid via a mixed anhydride which was prepared with isobutylchloroformate followed by the addition of the corresponding amine. Subsequent epoxidation with H2O2 in basic EtOH gave the aromatic cerulenin analogues 17a-e. Pharmacological testings for the synthesized substances were executed on efflux pump-resistant and -sensitive Candida albicans isolates, on the fatty acid synthesis enzyme KasA of Mycobacterium tuberculosis and on other organisms such as Leishmania major, Trypanosoma brucei brucei, Staphylococcus aureus, Staphylococcus epidermidis, Escherichia coli and Pseudomonas aeruginosa within the Sonderforschungsbereich 630.
KasA is a key enzyme which plays an essential part in the biosynthetic pathway of mycolic acids, the building block of cell wall in Mycobacterium tuberculosis. Its importance was demonstrated by the finding that the depletion of KasA leads to the cell lysis of Mycobacterium tuberculosis. Since Mycobacterium tuberculosis is a pathogen of tuberculosis, the second leading cause of death from an infectious disease worldwide, KasA has drawn attention as one of the attractive drug targets against tuberculosis. Due to the emergence of extensively drug-resistant strains which make most of the known antibiotics for treating tuberculosis ineffective, it became an urgent issue to develop new drugs against tuberculosis. In chapter 3.1, the protonation state of the catalytic residues in the resting state was mainly addressed. The FEP computation and MD simulations were employed for this investigation, and the results showed that the zwitterionic state is most probable. To underpin this conclusion with more solid data, The PESs for the proton transfer between the neutral and zwitterionic state were computed in the context of QM/MM. However, due to the strong dependency of the QM/MM optimization on the initial structure, it was not possible to obtain consistent results from these computations. To circumvent this problem, QM/MM based umbrella sampling was carried out with a semi-empirical method (RM1), and the resulting PMF surface indicated that the zwitterionic state is more stable than the neutral state. In chapter 3.2, the protonation state of significant residues in the acyl-enzyme state was investigated. Unlike other catalytic residues, the protonation state of His311 is ambiguous in the acyl-enzyme state, and different decarboxylation mechanisms can be derived depending on the protonation state of His311 in the acyl-enzyme state. Therefore, FEP computations were carried out to find most probable protonation state of His311 in terms of free energy, and the results showed that the pKa value at Nδ is considerably lowered by the enzyme environment while that of Nε is not. Additionally, the PMF profiles for the proton transfer between Lys340 and Glu354 were computed using QM/MM based umbrellas sampling method, and the results showed that the property of the Lys340/Glu354 pair is neutral rather than ionic when His311 is protonated at Nε. Moreover, a relatively larger ionic character of the Lys340/Glu354 pair when His311 is doubly protonated provides a valuable insight into how the Lys340/Glu354 pair plays a role in shifting the protonated state from Nδ to Nε in His311 after the acyl-transfer step. Overall, the results demonstrated that His311 is neutral and protonated at Nε, and the Lys340/Glu354 pair is also neutral in the acyl-enzyme state. Those computational results lead to the conclusion that the decarboxylation reaction is facilitated by an oxyanion hole which is comprised of two catalytic histidines. In chapter 3.3, the protonation state of catalytic residues in the resting state was revisited because a recent benchmark study showed that the employed semi-empirical method (RM1) in chapter 3.1 tends to overestimate the stabilization of the zwitterionic state. Furthermore, the Lys340/Glu354 pair was considered as purely ionic in chapter 3.1, while it actually has a mixed neutral and ionic character as demonstrated in chapter 3.2. The new investigations employed a larger QM region including the Lys340/Glu354 pair with the BLYP/6-31G** approach, which was proven to be accurate enough for the present purpose by benchmark computations. The new results from the QM/MM MD and FEP computations indicated the catalytic residues to be neutral most probably in the resting state, and this in turn brought up the question how KasA can be activated to initiate the catalytic reaction. On the basis of the results from the MD simulations and FEP computations for the His311Ala mutant in chapter 3.1, we hypothesized that the open conformation of Phe404 would trigger the activation of the catalytic residues by the formation of a strong hydrogen bond. The QM/MM MD simulation proved that the activation of the catalytic residues can indeed be accomplished by the open conformation of Phe404 we suggested, and the corresponding force field based PMF profile also indicated that this conformational change is energetically feasible. The distribution of hydrophilic and hydrophobic residues in the malonyl binding pocket in conjunction with our computational results further provided a valuable insight into the detailed process how the catalytic residues is activated upon the substrate entering.
In this century new experimental and computational techniques are adding an enormous amount of information, revealing many biological mysteries. The complexities of biological systems still broach new questions. Till now the main approach to understand a system has been to divide it in components that can be studied. The upcoming new paradigm is to combine the pieces of information in order to understand it at a global level. In the present thesis we have tried to study infectious diseases with such a global ‘Systems Biology’ approach. In the first part the apoptosis pathway is analyzed. Apoptosis (Programmed cell death) is used as a counter measure in different infections, for example viral infections. The interactions between death domain containing proteins are studied to address the following questions: i) How specificity is maintained - showing that it is induced through adaptors, ii) how proliferation/ survival signals are induced during activation of apoptosis – suggesting the pivotal role of RIP. The model also allowed us to detect new possible interacting surfaces. The pathway is then studied at a global level in a time step simulation to understand the evolution of the topology of activators and inhibitors of the pathway. Signal processing is further modeled in detail for the apoptosis pathway in M. musculus to predict the concentration time course of effector caspases. Further, experimental measurements of caspase-3 and viability of cells validate the model. The second part focuses on the phagosome, an organelle which plays an essential role in removal of pathogens as exemplified by M. tuberculosis. Again the problem is addressed in two main sections: i) To understanding the processes that are inhibited by M. tuberculosis; we focused on the phospholipid network applying a time step simulation in section one, which plays an important role in inhibition or activation of actin polymerization on the phagosome membrane. ii) Furthermore, actin polymers are suggested to play a role in the fusion of the phagosome with lysosome. To check this hypothesis an in silico model was developed; we find that the search time is reduced by 5 fold in the presence of actin polymers. Further the effect of length of actin polymers, dimensions of lysosome, phagosome and other model parameter is analyzed. After studying a pathway and then an organelle, the next step was to move to the system. This was exemplified by the host pathogen interactions between Bordetella pertussis and Bordetella bronchiseptica. The limited availability of quantitative information was the crucial factor behind the choice of the model type. A Boolean model was developed which was used for a dynamic simulation. The results predict important factors playing a role in Bordetella pathology especially the importance of Th1 related responses and not Th2 related responses in the clearance of the pathogen. Some of the quantitative predictions have been counterchecked by experimental results such as the time course of infection in different mutants and wild type mice. All these computational models have been developed in presence of limited kinetic data. The success of these models has been validated by comparison with experimental observations. Comparative models studied in chapters 6 and 9 can be used to explore new host pathogen interactions. For example in chapter 6, the analysis of inhibitors and inhibitory paths in three organism leads to the identification of regulatory hotspots in complex organisms and in chapter 9 the identification of three phases in B. bronchiseptica and inhibition of IFN-γ by TTSS lead us to explore similar phases and inhibition of IFN-γ in B. pertussis. Further an important significance of these models is to identify new components playing an essential role in host-pathogen interactions. In silico deletions can point out such components which can be further analyzed by experimental mutations.
\textbf{Molecular Determinants of Drug-Target Residence Times of Bacterial Enoyl-ACP Reductases.} Whereas optimization processes of early drug discovery campaigns are often affinity-driven, the drug-target residence time $t_R$ should also be considered due to an often strong correlation with \textit{in vivo} efficacy of compounds. However, rational optimization of $t_R$ is not straightforward and generally hampered by the lack of structural information about the transition states of ligand association and dissociation. The enoyl-ACP reductase FabI of the fatty acid synthesis (FAS) type II is an important drug-target in antibiotic research. InhA is the FabI enzyme of \textit{Mycobacterium tuberculosis}, which is known to be inhibited by various compound classes. Slow-onset inhibition of InhA is assumed to be associated with the ordering of the most flexible protein region, the substrate binding loop (SBL). Diphenylethers are one class of InhA inhibitors that can promote such SBL ordering, resulting in long drug-target residence times. Although these inhibitors are energetically and kinetically well characterized, it is still unclear how the structural features of a ligand affect $t_R$.
Using classical molecular dynamics (MD) simulations, recurring conformational families of InhA protein-ligand complexes were detected and structural determinants of drug-target residence time of diphenyl\-ethers with different kinetic profiles were described. This information was used to deduce guidelines for efficacy improvement of InhA inhibitors, including 5'-substitution on the diphenylether B-ring. The validity of this suggestion was then analyzed by means of MD simulations.
Moreover, Steered MD (SMD) simulations were employed to analyze ligand dissociation of diphenylethers from the FabI enzyme of \textit{Staphylococcus aureus}. This approach resulted in a very accurate and quantitative linear regression model of the experimental $ln(t_R)$ of these inhibitors as a function of the calculated maximum free energy change of induced ligand extraction. This model can be used to predict the residence times of new potential inhibitors from crystal structures or valid docking poses.
Since correct structural characterization of the intermediate enzyme-inhibitor state (EI) and the final state (EI*) of two-step slow-onset inhibition is crucial for rational residence time optimization, the current view of the EI and EI* states of InhA was revisited by means of crystal structure analysis, MD and SMD simulations. Overall, the analyses affirmed that the EI* state is a conformation resembling the 2X23 crystal structure (with slow-onset inhibitor \textbf{PT70}), whereas a twist of residues Ile202 and Val203 with a further opened helix $\alpha 6$ corresponds to the EI state. Furthermore, MD simulations emphasized the influence of close contacts to symmetry mates in the SBL region on SBL stability, underlined by the observation that an MD simulation of \textbf{PT155} chain A with chain B' of a symmetry mate in close proximity of the SBL region showed significantly more stable loops, than a simulation of the tetrameric assembly. Closing Part I, SMD simulations were employed which allow the delimitation of slow-onset InhA inhibitors from rapid reversible ligands.
\textbf{Prediction of \textit{Mycobacterium tuberculosis} Cell Wall Permeability.} The cell wall of \textit{M. tuberculosis} hampers antimycobacterial drug design due to its unique composition, providing intrinsic antibiotic resistance against lipophilic and hydrophilic compounds. To assess the druggability space of this pathogen, a large-scale data mining endeavor was conducted, based on multivariate statistical analysis of differences in the physico-chemical composition of a normally distributed drug-like chemical space and a database of antimycobacterial--and thus very likely permeable--compounds. The approach resulted in the logistic regression model MycPermCheck, which is able to predict the permeability probability of small organic molecules based on their physico-chemical properties. Evaluation of MycPermCheck suggests a high predictive power. The model was implemented as a freely accessible online service and as a local stand-alone command-line version.
Methodologies and findings from both parts of this thesis were combined to conduct a virtual screening for antimycobacterial substances. MycPermCheck was employed to screen the chemical permeability space of \textit{M. tuberculosis} from the entire ZINC12 drug-like database. After subsequent filtering steps regarding ADMET properties, InhA was chosen as an exemplary target. Docking to InhA led to a principal hit compound, which was further optimized. The quality of the interaction of selected derivatives with InhA was subsequently evaluated using MD and SMD simulations in terms of protein and ligand stability, as well as maximum free energy change of induced ligand egress. The results of the presented computational experiments suggest that compounds with an indole-3-acethydrazide scaffold might constitute a novel class of InhA inhibitors, worthwhile of further investigation.