INHIBITORS OF MYCOBACTERIUM TUBERCULOSIS DOSRST SIGNALING AND PERSISTENCE By Huiqing Zheng A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Microbiology and Molecular Genetics – Doctor of Philosophy 2017 ABSTRACT INHIBITORS OF MYCOBACTERIUM TUBERCULOSIS DOSRST SIGNALING AND PERSISTENCE By Huiqing Zheng Tuberculosis (TB) is one of the most deadly infectious diseases in human history and is caused by the bacterium Mycobacterium tuberculosis (Mtb). Current TB therapy requires 6-9 months of treatment with four different antibiotics, including isoniazid, rifampin, ethambutol and pyrazinamide. However, due to the long course of TB therapy and the evolution of drugresistant Mtb strains, first-line anti-mycobacterial drugs are not sufficient to control the TB epidemic. Therefore, it is urgent to develop new drugs with novel targets to shorten the course of therapy, control the spread of drug-resistant TB and eradicate this deadly disease. In response to host immune cues, Mtb modulates its metabolism to establish a state of low metabolic activity called non-replicating persistence (NRP). DosRST is a two-component regulatory system that plays an essential role to establish and maintain NRP in Mtb. It is induced by host immune stimuli, such as hypoxia, carbon monoxide and nitric oxide, through the histidine kinase sensors DosS and DosT. The response regulator DosR regulates about 50 genes in the dormancy regulon. NRP bacilli are problematic because 1) they are insensitive to several anti-mycobacterial agents and drive the long course of TB therapy; and, 2) they can resuscitate for growth once the immune system weakens for infectious transmission of the Mtb. Therefore, inhibiting the DosRST pathway may help reduce the population of NRP bacteria during infection and thus function to reduce drug tolerance and shorten TB treatment. This dissertation presents a whole-cell phenotypic high-throughput screen of a ~540,000 compound small-molecule library. The screen employed a DosR-dependent, hypoxia-inducible fluorescent reporter strain, CDC1551(hspX::GFP), and successfully identified six distinct, novel chemical inhibitors of DosRST signaling, named HC101A-106A. Physiological and mechanistic studies were performed to characterize HC101-104 and HC106A. All five inhibitors are shown to inhibit genes of the DosRST regulon and persistenceassociated physiologies, such as triacylglycerol accumulation. HC101A, HC102A, HC103A and HC106A also reduce Mtb survival when cultured under strongly hypoxic conditions. UVvisible spectroscopy studies show that HC101A (artemisinin) and HC106A target the heme group of sensor kinases DosS/T via distinct mechanisms. For example, artemisinin modulates the redox status of DosS/T and alkylates the heme to form artemisinin-heme adducts, whereas HC106A interacts with DosS heme. In contrast, HC102A and HC103A do not target the heme group, but instead inhibit sensor kinase autophosphorylation activity. Electrophoretic mobility assays suggest that HC104A functions by directly inhibiting DosR DNA binding activity. Overall, this dissertation provides proof-of-concept that multiple components of the DosRST pathway can be targeted by small molecules to inhibit Mtb persistence and antibiotic tolerance. Additionally, this dissertation presents the discovery of a new chemical inhibitor, HC2091, that kills Mtb by targeting the mycolic acid transporter MmpL3. MmpL3 is an essential protein that functions to transport trehalose monomycolate across mycomembranes for trehalose dimycolate biosynthesis. HC2091 is bactericidal against Mtb in a dose- and timedependent manner in vitro. It also has activity against Mtb inside of macrophages. Whole genome sequencing spontaneous mutants resistant to HC2091 identified five single nucleotide variants primarily located in the C-terminus of MmpL3, and HC2091-treated Mtb exhibits decreased mycolic acid synthesis, thus supporting that MmpL3 is the target of HC2091. I delicate this work to my parents Jiahe Zheng and Aiyu Huang, and to my wife Annie Yang iv ACKNOWLEDGEMENTS First, I would like to thank my mentor Robert B. Abramovitch for having me in the lab to work on such wonderful projects that have kept my interests for the duration of my studies. As well, he has provided me guidance and mentorship throughout my graduate career at Michigan State University and has been so supportive of me throughout my graduate career. His positive attitude has always encouraged me to be in the field of science and made me enthusiastic about science. I have been trained critically and scientifically through his time and effort. Second, I would like to thank Bob R. Hausinger, Lee Kroos, Linda S. Mansfield and Chris M. Waters for being on my guidance committee and providing helpful suggestions each time we had committee meetings. I am also grateful for Mark Farrugia for his assistance with UV-visible spectroscopy assay, Lijun Chen for her help with mass spectrometry analyses, and Chris Colvin and Ben Johnson for their support and data collection in the high-throughput screens, that identified the compounds studied in this dissertation. I would like to thank our collaborators, Scott D. Larsen, Paul D. Kirchhoff, Michael Wilson, and Edmund Ellsworth for their contributions to these studies. Third, I would like to thank my fellow lab mates, Jake Baker, Javiera Ortiz, Rajni Goyal, Garry Coulson, Uma Gautam, Shelby Dechow, and John Williams for their support and our regular scientific discussions that helped accelerate my research and created such a friendly and joyful work environment. Special thanks also go to undergraduate students I mentored, Katriana Jorgensen-Muga, Marilyn Werner, Sharon-Rose Nartey, and Ryan Borcher for their assistance with these projects. Our work together advanced these projects and really taught me how to be a good mentor. v Fourth, I would like to thank the Biomolecular Science program for bringing me to MSU, and the MMG department for always backing me up whenever I ran into problems. Staff in the MMG office have been always helpful to keep my studies on track. I would like to acknowledge the Bill and Melinda Gates Foundation, the National Institutes of Health, a G.D. Edith Hsiang and Margaret Everett Kimball Scholarship, and a Dissertation Completion Fellowship for financial support for my research, without which I would have been be unable to conduct this research. Last, but not least, my family. My heartfelt gratitude to my parents for their endless care throughout my life. Although they can’t fully understand my work, they are always with me wherever I go, and give me spiritual support for whatever I am doing. After we immigrated to the US, they worked tirelessly to support me and my brother to pursue the higher education that they couldn’t have. I am very proud of myself to be the first college student, and now the first Ph.D. in my entire family. Without their support, I wouldn’t be able to move to the stage where I am today. Also, I am grateful that I met and married my wife Annie Yang during graduate school. I am very lucky to have her standing side by side with me and aiding me through my graduate school life. vi TABLE OF CONTENTS LIST OF TABLES .......................................................................................................................... x LIST OF FIGURES ...................................................................................................................... xi KEY TO ABBREVIATIONS ..................................................................................................... xiv CHAPTER 1 - Strategies for controlling bacterial persistence ...................................................... 1 Introduction ................................................................................................................................ 1 Therapeutic strategies to control bacterial persistence .............................................................. 2 Modulation of metabolism reverts persisters to an antibiotic susceptible state .................. 2 Reactive oxygen species sensitize persister cells to antibiotic treatment ........................... 4 Enhancing antibiotic uptake promotes killing of persisters ................................................ 8 Targeting stationary phase respiration to inhibit persisters ................................................ 9 Targeting bacterial proteases to kill persisters .................................................................. 11 Modulating intracellular ATP levels to kill persisters ...................................................... 13 Disrupting membrane integrity to kill persisters ............................................................. 14 Inhibition of cell envelope biosynthesis potentiates killing of persisters by antibiotics... 16 Targeting two-component regulatory systems to inhibit bacterial persistence ................. 18 Concluding remarks ................................................................................................................ 23 REFERENCES ....................................................................................................................... 25 CHAPTER 2 - Inhibitors of Mycobacterium tuberculosis DosRST signaling and persistence ... 35 Summary .................................................................................................................................. 36 Introduction .............................................................................................................................. 37 Results ...................................................................................................................................... 38 Discovery of DosRST regulon inhibitors ......................................................................... 38 Artemisinin, HC102A and HC103A disrupt persistence .................................................. 46 Artemisinin targets DosS and DosT sensor kinase heme ................................................. 50 HC103A inhibits DosS and DosT autophosphorylation ................................................... 56 Discussion ................................................................................................................................ 58 Materials and methods ............................................................................................................. 60 Bacterial strains and growth conditions ............................................................................ 60 High-throughput screening assay and data analysis ......................................................... 61 Transcriptional profiling and data analysis ....................................................................... 63 Real-time PCR assays ....................................................................................................... 64 Triacylglycerol accumulation analysis ............................................................................. 64 NRP survival and antibiotic tolerance assays ................................................................... 65 DosS and DosT protein purification ................................................................................. 66 UV-visible spectroscopy assay and mass spectrometry.................................................... 67 DosS and DosT autophosphorylation assay...................................................................... 67 Structural modeling of DosT and the G85L and G115L substituted proteins .................. 68 Chemical synthesis of HC102A (CCG-232500) and HC103A (CCG-257424) ............... 69 vii Acknowledgments.................................................................................................................... 69 APPENDIX ................................................................................................................................... 70 REFERENCES ............................................................................................................................. 86 CHAPTER 3 - Chemical inhibition of the Mycobacterium tuberculosis DosRST two-component regulatory system by targeting sensor kinase heme and response regulator DNA binding ......... 93 Summary .................................................................................................................................. 94 Introduction .............................................................................................................................. 95 Results ...................................................................................................................................... 97 Inhibition of the DosR regulon by HC104A and HC106A............................................... 97 Inhibition of Mtb persistence-associated physiologies ................................................... 103 HC104A inhibits DosR DNA binding ............................................................................ 104 HC106A modulates DosS heme ..................................................................................... 108 Synergistic interactions between DosRST inhibitors ..................................................... 109 Structure-activity relationship study for HC104A and HC106....................................... 112 Discussion .............................................................................................................................. 114 Materials and methods ........................................................................................................... 118 Bacterial strains and growth conditions .......................................................................... 118 EC50 assays...................................................................................................................... 118 Transcriptional profiling and data analysis ..................................................................... 119 Real-time PCR assays ..................................................................................................... 119 TAG biosynthesis............................................................................................................ 120 NRP survival assays........................................................................................................ 120 DosR protein purification ............................................................................................... 121 Electrophoretic mobility assay ........................................................................................ 121 UV-visible spectroscopy assay ....................................................................................... 122 Checkboard synergy studies ........................................................................................... 122 Autophosphorylation assay ............................................................................................. 123 Kinetic solubility assay ................................................................................................... 123 Acknowledgments ................................................................................................................. 123 APPENDIX ................................................................................................................................ 125 REFERENCES .......................................................................................................................... 144 CHAPTER 4 - HC2091 kills Mycobacterium tuberculosis by targeting the MmpL3 mycolic acid transporter ................................................................................................................................... 149 Summary ................................................................................................................................ 150 Introduction ............................................................................................................................ 151 Results .................................................................................................................................... 152 HC2091 kills Mtb in vitro and in macrophages .............................................................. 152 HC2091 resistant strains have mutations in mmpL3 ....................................................... 157 Inhibition of mycolic acid biosynthesis by HC2091....................................................... 159 Transcriptional analysis of Mtb treated with HC2091 and SQ109 ................................. 159 HC2091 has no effect on Mtb membrane potential ........................................................ 160 Structure-activity relationship studies............................................................................. 163 Discussion .............................................................................................................................. 165 Materials and methods .......................................................................................................... 167 viii Bacterial strains and growth conditions .......................................................................... 167 EC50 determination and spectrum of activity in other mycobacteria and non-mycobacteria ......................................................................................................................................... 167 Bactericidal activity of HC2091 ..................................................................................... 168 Activity of HC2091 against intracellular Mtb ................................................................ 169 Whole genome sequencing to identify the target of HC2091 ......................................... 169 Mycolic acid accumulation analysis ............................................................................... 169 Transcriptional profiling and data analysis ..................................................................... 169 Membrane potential ........................................................................................................ 169 Structure activity relationship (SAR) and eukaryotic cytotoxicity of HC2091 .............. 170 Kinetic solubility assay ................................................................................................... 170 Acknowledgments....................................................................................................................... 172 APPENDIX ................................................................................................................................. 173 REFERENCES ........................................................................................................................... 181 CHAPTER 5 - Concluding Remarks .......................................................................................... 187 REFERENCES ........................................................................................................................... 194 ix LIST OF TABLES Supplementary Table 2.1. GFP fluorescence inhibition and eukaryotic cytotoxicity by DosR regulon inhibitors. ......................................................................................................................... 73 Supplementary Table 2.2. Small molecule screening data. ......................................................... 74 Supplementary Table 3.1. Structure and properties of HC104 series. ........................................ 127 Supplementary Table 3.2. Initial SAR studies of the HC106 series. .......................................... 128 Supplementary Table 3.3. Initial Topliss Tree evaluation of “A-ring” aniline of series HC106. ......................................................................................................................................................129 Table 4.1. Identification of HC2091 resistance mutations. ........................................................ 158 Table 4.2. HC2091 Structure activity relationship study. ........................................................... 164 Supplementary Table 4.1. Spectrum of activity of HC2091. ...................................................... 175 Supplementary Table 4.2. Cytotoxicity of HC2091 analogs. ..................................................... 176 Supplementary Table 4.3. Solubility of HC2091 analogs. ......................................................... 177 x LIST OF FIGURES Figure 1.1. Summary of major pathways that can be targeted to kill persisters. ............................ 6 Figure 1.2. Structures of different classes of select inhibitors that kill persister cells. ................... 7 Figure 2.1. Discovery of inhibitors of the DosRST pathway. .......................................................44 Figure 2.2. Transcriptional profiling shows artemisinin, HC102A and HC103A inhibit the core genes of the DosRST regulon during hypoxia. ............................................................................. 45 Figure 2.3. Artemisinin, HC102A and HC103A inhibit TAG synthesis, survival and isoniazid tolerance. ....................................................................................................................................... 49 Figure 2.4. Artemisinin directly inhibits DosS and DosT by targeting sensor kinase heme. ....... 54 Figure 2.5. Amino acid substitutions in DosS or DosT promote resistance to artemisinin. ......... 55 Figure 2.6. HC103A inhibits DosS and DosT autophosphorylation....................................................... 57 Supplementary Figure 2.1. Statistical analysis of HTS controls. ................................................. 75 Supplementary Figure 2.2. Identification of inhibitors of the DosRST pathway. ........................ 76 Supplementary Figure 2.3. Transcriptional profiling shows artemisinin, HC102A and HC103A inhibit the core genes of the DosRST regulon during hypoxia. .....................................................77 Supplementary Figure 2.4. Artemisinin, HC102A and HC103A inhibit TAG synthesis, survival and isoniazid tolerance during NRP. ............................................................................................ 78 Supplementary Figure 2.5. Artemisinin directly modulates DosS and DosT (G85L) heme in a dosedependent manner. ........................................................................................................................ 79 Supplementary Figure 2.6. HC102A and HC103A do not modulate DosS redox. ...................... 80 Supplementary Figure 2.7. Substitutions in DosT provide resistance to artemisinin. .................. 81 Supplementary Figure 2.8. Modulation of DosS and DosT kinase activities by HC102A and HC103A. ....................................................................................................................................... 82 Figure 3.1. HC104A and HC106A inhibit DosRST reporter fluorescence. ............................... 100 Figure 3.2. Transcriptional profiling shows HC104A and HC106A inhibited DosR regulon during hypoxia........................................................................................................................................ 101 xi Figure 3.3. Inhibition of DosR regulon and persistence-associated physiologies by HC104A and HC106A. ......................................................................................................................................102 Figure 3.4. Inhibition of DosR DNA-binding by HC104A. ....................................................... 106 Figure 3.5. Interactions between HC106A and DosS heme. ...................................................... 107 Figure 3.6. Synergistic interactions between DosRST inhibitors. .............................................. 111 Supplementary Figure 3.1. Inhibition of the DosR regulon by HC104 and HC106A during hypoxia. ..................................................................................................................................................... 130 Supplementary Figure 3.2. TLC of TAG reduction in Mtb treated with HC104A and HC106A. ................................................................................................................................................…. 131 Supplementary Figure 3.3. Autoradiograph examining the impact of HC104A and HC106A on DosS autophosphorylation. ......................................................................................................... 132 Supplementary Figure 3.4. Investigation of interaction between HC104A and DosS. ...............133 Supplementary Figure 3.5. The impact of virstatin on DosR DNA-binding and DosRST signaling in Mtb. ......................................................................................................................................... 134 Supplementary Figure 3.6. Investigating the interaction between HC106A and DosS heme. ... 135 Supplementary Figure 3.7. Checkerboard assays examining paired interactions of DosRST inhibitors. .................................................................................................................................... 136 Supplementary Figure 3.8. Synthetic scheme for the HC106 analogs. ...................................... 138 Supplementary Figure 3.9. Comparison between artemisinin, HC103 and HC106 for interactions. ......................................................................................................................................................139 Figure 4.1. HC2091 is bactericidal to Mtb. ................................................................................ 155 Figure 4.2. HC2091 is active against intracellular Mtb and inactive against non-replicating Mtb ... ......................................................................................................................................................156 Figure 4.3. HC2091 inhibits TDM accumulation. ...................................................................... 161 Figure 4.4. Shared transcriptional profiles between HC2091 and SQ109. ................................. 162 Supplementary Figure 4.1. Dose-response curves for HC2091 resistant mutants. ..................... 178 Supplementary Figure 4.2. The map of HC2091 resistance substitutions on the MmpL3 protein.... xii ..................................................................................................................................................... 179 Supplementary Figure 4.3. Impact of HC2091 on Mtb membrane potential. ............................ 180 xiii KEY TO ABBREVIATIONS ADEP .................................................................................................................. Acyldepsipeptides Art .................................................................................................................................. Artemisinin CCCP ....................................................................... Carbonyl cyanide m-chlorophenyl hydrazone CFU ................................................................................................................. Colony forming units CI........................................................................................................................ Combination Index DCM ..................................................................................................................... Dichloromethane DHA ................................................................................................................... Dihydroartemisinin DMF ................................................................................................................. Dimethylformamide DMSO .................................................................................................................Dimethyl sulfoxide DTN .................................................................................................................................. Dithionite EC50 ....................................................................................... Half maximal effective concentration EMSA ...................................................................................... Electrophoretic mobility shift assay ETC ............................................................................................................. Electron transport chain FI ................................................................................................................. Fluorescence inhibition GFP ....................................................................................................... Green fluorescence protein GI ......................................................................................................................... Growth inhibition HTS ............................................................................................................. High-throughput screen INH ..................................................................................................................................... Isoniazid LC/MS .......................................................................... Liquid chromatography/mass spectrometry MRSA ......................................................................... Methicillin-resistant Staphylococcus aureus MSSA.......................................................................... Methicillin-sensitive Staphylococcus aureus xiv Mtb ....................................................................................................... Mycobacterium tuberculosis NRP ....................................................................................................... Non-replicating persistence OADC .................................................................................. Oleic acid, albumin, dextrose, catalase OH• ........................................................................................................................ Hydroxyl radical PMF ................................................................................................................. Proton motive force ROS ............................................................................................................ Reactive oxygen species Sar ..................................................................................................... structure-activity relationship TAG........................................................................................................................... Triacylglycerol TDM .............................................................................................................. Trehalose dimycolate THF ....................................................................................................................... Tetrahydrofuran TMM ....................................................................................................... Trehalose monomycolate TB ................................................................................................................................. Tuberculosis TLC ....................................................................................................... Thin-layer chromatography WT ............................................................................................................................... WILD TYPE xv CHAPTER 1 - Strategies for controlling bacterial persistence Introduction During the golden era of antibiotic discovery, between 1940 and 1960, many new antibiotics were discovered that are still currently in use to treat different infectious diseases (1). However, over time, antibiotics have become less effective at controlling bacterial infections. Treatment failure is due, in part, to the rise of drug resistant strains that have acquired genetic elements to inactivate or efflux antibiotics or strains with mutations in drug targets that modulate drug-target interactions. Another factor driving treatment failure is the phenomenon of bacterial persistence, which was first described by Hobby et al. in 1942 (2). At that time, scientists noticed that there was always a small fraction of a treated Staphylococcus aureus population that tolerated killing by penicillin. Pre-treatment of bacteria under stressful conditions, such as cold, or acidic pH, arrested bacterial replication and enhanced tolerance to antibiotics (2, 3). These populations of bacteria were referred to as persister cells. Biphasic killing is often observed for bacterial cultures exposed to antibiotics under growth permissive conditions (4). Treatment with antibiotics initiates rapid killing of the majority of bacteria in a population, however, a population of persister cells will often remain that are completely tolerant to the antibiotic. These persister cells are responsible for phenotypic tolerance to antibiotics. There are two different classes of bacterial persistence observed: Class I persistence is driven by a small sub-population of dormant bacteria (<0.1%) that naturally tolerates antibiotics (5); Class II persistence is induced when the bacterial population responds to stress conditions, such as hypoxia, acidic pH, or nutrient starvation (6). Class II bacterial persistence is also observed 1 during infections where bacteria in the host are exposed to stressful conditions that trigger a population of bacteria to cease replication and become persistent. Both classes of bacterial persistence cause antibiotic tolerance and treatment failure. For example, Mycobacterium tuberculosis (Mtb) requires a 6-month long course of therapy. This long course is driven, in part, by the difficulty of killing all of the persistent, drug-tolerant bacteria that are naturally associated with tuberculosis (TB) infection. Recent studies have advanced our understanding of bacterial persistence, and have highlighted different strategies to target each class. This introduction will survey recent approaches to controlling bacterial persistence with an emphasis on Mtb and other select Gram-negative and Gram-positive bacteria. The reader is directed to other recent reviews for additional information on mechanisms of controlling bacterial persistence (6, 7). Therapeutic strategies to control bacterial persistence Modulation of metabolism reverts persisters to an antibiotic susceptible state Bacteria have evolved to maintain a sub-population of dormant cells that are naturally tolerant to antibiotic treatment, known as class I persisters. One signature of these dormant bacteria is their partially inactive metabolism. Antibiotics usually can kill the majority of cells by targeting essential cellular functions required for growth, such as transcription, protein synthesis, and cell wall synthesis. The reduced activity of these essential cellular processes in non-growing cells can diminish the effectiveness of antibiotics and induce phenotypic antibiotic tolerance. Therefore, this subpopulation can function to prevent eradication of the population during drug treatment. One approach to target the class I persister population is by making them susceptible to antibiotics by re-initiating their metabolism. Allison et al. have reported that providing additional 2 metabolites is a strategy to eradicate persisters (8). Adding glucose, mannitol, or fructose, carbon sources that feed glycolysis, potentiates gentamicin killing of Escherichia coli and S. aureus persisters (Fig. 1.1). The killing is facilitated by the generation of a proton-motive force (PMF) by stimulation of metabolism and the electron transport chain (ETC). NADH produced from central metabolism is oxidized by the enzymes of the ETC, and thus activates the PMF required for the uptake of gentamycin. Interestingly, even though these exogenous metabolites stimulate metabolism, the growth of the persisters is not resumed. Therefore, other antibiotics that target cell-wall biosynthesis and DNA replication are still ineffective to kill persisters together with these metabolites. In a separate study glucose was also shown to potentiate class I S. aureus persister killing by daptomycin. However, this killing is not through a PMF-dependent mechanism (9). It is postulated that glucose triggers carbohydrate transport proteins to potentiate daptomycin activity or induces daptomycin-specific lysis of cells. However, further experiments are needed to determine the mode of action. Enhancement of antibiotic killing by addition of metabolites supports a model of overcoming persistence by “waking up” the dormant persister cells. Once the cells are replicating they may be more susceptible to antibiotics. For example, Marques et al. have shown that treating E. coli or Pseudomonas aeruginosa persisters with the fatty acid signaling molecule cis-2-decenoic acid stimulates dormant cells into a metabolically active state, and activates transcription and increases protein synthesis (10). In combination with antimicrobials, such as ciprofloxacin, tobramycin, or tetracycline, cis-2-decenoic acid causes decreased cell viability when treated with antimicrobials and a reduction of persister cells in E. coli and P. aeruginosa. Therefore, reprogramming the metabolism of persister cells to become active can function to control persisters. Stimuli that jump-start different metabolic activities, such as activating RNA 3 transcription, protein synthesis machinery, or cell wall synthesis, may selectively potentiate specific classes of antibiotics to eradicate persister cells. Reactive oxygen species sensitize persister cells to antibiotic treatment Persister cells that are present within a bacterial population have enhanced survival under a variety of stress conditions (11). However, numerous reports have shown that persisters are sensitive to specific intracellular stress signals, such as reactive oxygen species (ROS) (Fig. 1.1). MoronesRamirez et al. reported that ionic silver (Ag+) is not only bactericidal to a broad spectrum of Gramnegative and Gram-positive bacteria, but also kills class I bacterial persister cells when a nonbactericidal dose of Ag+ is used in combination with different classes of antibiotics, such as gentamicin, ofloxacin and ampicillin (Fig. 1.1) (12). Mechanistic studies show that Ag+ induces hydroxyl radical (OH•) production via Fenton chemistry. This is further supported by increasing Fe2+ release as a result of disturbance of iron hemostasis by disrupting Fe-S clusters. Furthermore, TCA cycle knockout mutants, including DicdA, DsucB, Dmdh, and DacnB, are all less susceptible to Ag+ treatment due to reduced production of OH•, suggesting generation of OH• is likely driven by disruption of metabolic pathways. By using a fluorescent dye, propidium iodide, impermeable to cell membrane, it was found that Ag+ treatment increases membrane permeability by destabilizing cell envelope to facilitate uptake of antibiotics. Another report by Grant et al. also shows that increasing ROS in Mycobacterium smegmatis, by incubating the culture under aerobic conditions with dissolved-oxygen saturation of 95%, potentiates killing of class I persisters by antibiotics (13). The oxygen-dependent phenotype is rescued by adding the OH• scavenger thiourea, suggesting that persisters are sensitive to ROS. Supporting this model, treating M. smegmatis persister cells with the OH• producing 4 antibiotic clofazimine eradicates persisters. A similar phenomenon is also observed in M. tuberculosis. Treating M. tuberculosis with high concentrations of vitamin C completely sterilizes the cultures (Fig. 1.1) (14). The killing is driven by increasing ROS production through HaberWeiss and Fenton reactions. A high concentration of vitamin C also increases the level of ferrous ion through reduction of ferric ion. Ferrous ion reacts with oxygen to produce ROS, and causes DNA damage, lipid alterations and redox stress. Isoniazid (INH) cannot sterilize a culture due to the presence of persisters, however, when INH is used in combination with vitamin C, at sub-lethal doses, INH can sterilize the culture. Furthermore, a recent report by Vilcheze et al. demonstrates that treating M. tuberculosis with cysteine and INH sterilizes the bacterial culture (Fig. 1.1) (15). Cysteine functions as a reductant like vitamin C that can quickly reduce ferric ion to ferrous ion, which leads to increased accumulation of ROS in the presence of oxygen. Addition of cation scavengers, such as the iron chelator deferoxamine, diminishes sterilization, and the combination of cysteine and rifampin causes no additional killing to Mtb cultures grown in anaerobic conditions. These findings indicate that potentiation of killing by addition of cysteine is dependent on cations and oxygen. Lastly, treating cultures with cysteine and INH increases oxygen consumption of non-respiring cells, as well as shifting the balance of menaquinol/menaquinone towards menaquinol, which has been shown to control the Mtb respiration rate, thus providing evidence that cysteine and INH can increase cellular respiration. Therefore, it is possible that cysteine can revert persister cells to metabolically active cells by stimulating their respiration. Together, these data support the hypothesis that modulating ROS production and respiration is a strategy to control persisters in mycobacterial species. 5 Boromycin 74a Artemisinin HC106A Hypoxia NO, CO H+ H H+ Pentobra NH125 + H+ H+ Ab DosS/T ATP Synthase Heme p p HC102A HC103A Cell lysis H+ H+ H+ HC104A H+ Bedaquiline DCCD DG70 ATP DosR Persistence genes Cell Death Glucose Mannitol Fructose ADEP4 TCA Cycle Proteases ROS Lassomycin Bortezomib Ag+ Vitamin C Cysteine Figure 1.1. Summary of major pathways that can be targeted to kill persisters. Pathways that can be targeted include: 1) reactivation of metabolic pathways with supplement of metabolites; 2) induction of intracellular ROS overproduction; 3) increase of cytoplasmic antibiotic accumulation by enhancing uptake and reducing the efflux activity; 4) activation of non-specific protease activity or inhibition of protease activity; 5) inhibition of ATP production; 6) disruption of membrane integrity to cause cell lysis; 7) inhibition of the DosRST persistence pathway. Ab, antibiotics. 6 NH 2 a OH OH O O O H 2N HO H 2N N N+ OH O H N WRRNQFWIKIQR I- NH 2 O NH 2 NH125 O Pentobra O b O O F N O NH O O O O N H L R N G O HN L D A F L Q G V R I N R OH B OH H N N H O N R O N N F CO2CH3 Bortezomib HN O ADEP4 Lassomycin N c N N (R) Br O H (S) N N N C N N N N S O Bedaquiline S GNF-NITD 46 DCCD GNF-NITD 82 O O N O N N H N H F H 2N Cl O DG70 GNF-NITD 101 H N WR H 2N d NH HN F O N H N H H N N HN Linear peptide NH 2 O H N RW NH O N H O NH H O 74a O H N RW HN N H WR Dendrimeric peptide O e O O O2N O S O O S N S N H S NH O NH 2 CD117 H N N S HN CD39 O NH N TCA1 H N O N H N H SQ109 F F F TBL-140 H f O O O H O H HO N H O O O S Cl N O HC103A Artemisinin HN N O H N O O N O NH O HC102A Cl N H N H Br HC104A HC106A Figure 1.2. Structures of different classes of select inhibitors that kill persister cells. (a) Targeting bacterial membrane permeability; (b) targeting bacterial proteases; (c) inhibiting ATP synthesis; (d) disrupting bacterial membrane; (e) inhibiting bacterial cell wall synthesis; (f) inhibiting DosRST. 7 Enhancing antibiotic uptake promotes killing of persisters One major reason that persister cells are tolerant to antibiotics is due to low cytoplasmic drug accumulation. Low metabolic and cellular transport activities of persister cells provide an excellent barrier for antibiotics to reach their cytoplasmic targets. The uptake of some classes of antibiotics, such as aminoglycosides and gentamycin, is an energy dependent process that requires a PMF. One approach to tackle this mechanism of antibiotic tolerance is to increase membrane permeability and promote entry of antibiotics into persister cells (Fig. 1.1). For example, a study by Cui et al. demonstrates that the membrane-active antibiotic colistin effectively eliminates class I E. coli persisters when used at higher concentrations (16). Moreover, using clinically relevant concentrations of colistin increases killing of persister cells in combination with gentamicin and ofloxacin. Enhanced accumulation of fluorescently-labeled gentamicin was observed following colistin treatment. Furthermore, Schmidt et al. have showed that conjugating tobramycin to a 12amino acid peptide transporter uptake sequence, called Pentobra (Fig. 1.2), enabled effective killing of class I E. coli and S. aureus persister cells (17). The transporter sequence functions to permeabilize the bacterial membrane, enabling tobramycin to cross the persister cell membrane. In another study, Kim et al. (2015) used a fluorescence-based phenotypic screen to identify inhibitors that enhance cell membrane permeability, by monitoring changes in SYTOX green dye staining (18). Several inhibitors, including lysostaphin, nisin, and NH125, kill methicillin-resistant S. aureus (MRSA) persister cells by inducing cell membrane permeabilization (Fig. 1.2). NH125 is an imidazolium cation with a long fatty acid tail. To further improve the potency of NH125, analogs were synthesized (19), with alkylated and benzylated NH125 analogs showing enhanced potency against S. aureus persisters. These analogs also enhance antimicrobial activities when used in combination with other membrane-active agents, such as BAC-16 and daptomycin (19). 8 These studies suggest that targeting membrane permeability represents an effective strategy to kill persisters. Another mechanism by which persister cells may develop antibiotic tolerance is by increasing the activity of efflux pumps. Recent research by Pu et al., using a fluorescently labeled antibiotic penicillin termed BOCILLIN, demonstrated that BOCILLIN accumulation inside of class I E. coli persister cells is much lower than for the actively growing cells (20). Transcriptional analysis shows that the b-lactam antibiotics strongly induce the expression of numerous multidrug efflux genes, such as tolC, in persister cells. This finding supports the notion that enhanced efflux pump activity in persister cells may be responsible for antibiotic tolerance. Supporting this model, a positive correlation is shown between tolC expression and persister cells. Deletion of tolC dramatically increases the accumulation of antibiotics inside the persister cells, suggesting TolC plays an important role to exclude the drugs from the persister cells. This finding supports the hypothesis that inhibiting the TolC efflux pump may function to enhance killing of class I persister cells by antibiotics. Indeed, it was observed that treating E. coli persister cells with TolC inhibitors, such as 1-(1-naphthylmethyl) piperazine and phenylalanine arginyl b-naphthylamide, decreased cell viability when bacteria were treated with carbencillin, cloxacillin, or nalidixic acid. Targeting stationary phase respiration to inhibit persisters When growing under aerobic conditions, persister cell populations increase as the bacterial cultures age. Studies conducted by Orman and Brynildsen (21) found a correlation between high redox activity, stationary phase respiration and class I persister formation. Redox Sensor Green (RSG) is a redox indicator that directly measures metabolic activity, with enhanced green fluorescence after reduction by bacterial reductases. Using RSG it was found that the population 9 of stationary phase cells with high redox activity is more likely to form class I persisters than the population with low redox activity. This high redox activity is also correlated with a high yield of non-growing cells, but inversely correlated with protein synthesis. For instance, persister cells with high redox activity have less activity of transcription and translation upon exposure to fresh medium. Therefore, it is suggested that the redox poise plays a decisive role in the increased formation of persister cells in stationary phase. It is further hypothesized that during stationary phase aerobic respiration is the major redox process, and inhibition of respiration decreases the population of persister cells. Indeed, treating the aerobically growing culture with potassium cyanide throughout stationary phase significantly reduces persister cells by up to 1,000-fold. A role of respiration for survival of stationary phase persisters is further supported by genetic studies, where deletions of E. coli genes associated with metabolism caused low levels of persister cells, including the strains mutated in sucB that encodes for dihydrolipoyltranssuccinase, mdh for malate dehydrogenase, and ubiF for oxygenase involved in ubiquinone biosynthesis (21). These mutants exhibit significant perturbation of stationary phase respiration. Their initial protein synthesis is increased after re-inoculating into fresh medium. The indicated proteins are involved in generating reducing equivalents to power respiration, such as ubiquinol and NADH (21). The reducing power during stationary phase is mainly derived from acetate or catabolism of endogenous cellular components, such as phospholipids, ribosomes and proteins. It was found that cells with a deficiency of acetate consumption (DackADpoxB) or phospholipid degradation (DfadDDfadK) have a similar level of persister cells as wild type (WT). However, the untreated stationary phase cells display reduced RNA integrity and protein level as compared to KCN-treated cells (21). This observation is consistent with previous observations that persisters have reduced transcription and translation. The mechanistic studies suggest that inhibiting stationary respiratory 10 activities prevents cells from digesting endogenous RNA and proteins, and leads cells to continue translation and replication. Additional studies have shown that inhibiting respiration of the stationary culture with nitric oxide also impairs persister formation (22). Like the cultures treated with KCN, the NO-treated culture also exhibits decreased RNA and protein degradation during stationary phase. Overall, these studies show that respiration is a critical process controlling persister formation, and they suggest this process can be targeted to inhibit bacterial persistence. Targeting bacterial proteases to kill persisters Recently, bacterial proteases, such as Clp system caseinolytic proteases (23), have generated interest as new targets for novel antibiotic development. The protease complex generally consists of a proteolytic core, ClpP, corresponding ATP-dependent AAA+ chaperone proteins and associated adaptor proteins (Fig. 1.1). The protease targets partially synthesized and misfolded proteins, as well as cleaving specific proteins with regulatory functions. Adaptor proteins assist with binding to protease substrates, which are unfolded by the chaperone proteins and then cleaved by ClpP. Several inhibitors of Clp have been reported to kill class I persisters. For instance, a newly discovered series of ClpP activators, named acyldepsipeptides (ADEPs), are potent against a broad spectrum of bacteria (24). ADEPs activate ClpP by keeping its catalytic chamber open, allowing peptides and proteins to enter, and promoting uncontrolled protein degradation in bacterial cells. Furthermore, recent research shows that ADEP4 (Fig. 1.2) also has an effect on non-replicating persister cells (25). Instead of triggering specific proteolytic activity, ADEP4-bound ClpP is a nonspecific protease that causes excess breakdown of proteins. S. aureus persisters that survive ciprofloxacin treatment are eradicated by addition of ADEP4 (25). ADEP4 is also effective in 11 sterilizing stationary cultures treated with rifampicin, linezolid or ciprofloxacin, suggesting the compound enhances the killing of persister cells (25). Triggering nonspecific proteolytic activity is lethal to bacteria and is effective at eliminating both replicating and persister cells, indicating the idea that some essential processes can be targeted to kill persister cells Several new protease inhibitors also have been reported to kill class II persisters. For instance, the new antibiotic lassomycin (Fig. 1.2) is reported to kill both growing and nonreplicating Mtb by targeting the caseinolytic protease ClpC1P1P2 (26). Lassomycin binds to the ATPase domain of ClpC1 and increases its ATP hydrolysis activity, but also inhibits substrate translocation from ClpC1 to the proteolytic subunits ClpP1P2 of the complex. It decreases degradation of target cell proteins and is toxic to Mtb during stressful conditions. Additionally, the well-known human proteasome inhibitor, bortezomib (Fig. 1.2), has been shown to inhibit ClpP1P2 by covalently attaching via its boron atom to the hydroxyl group of serine at the catalytic triad (27). Bortezomib only potentiates Mtb to aminoglycoside treatment, suggesting that induction of protein mistranslation by aminoglycosides is toxic to Mtb if the proteasome is inhibited to remove incomplete translation products. To increase selectivity of bortezomib to antimycobacterial activity over inhibition of the human proteasome, the boronic acid warhead was changed to chloromethyl ketone (28). The newly synthesized analogs only display activity against mycobacterial ClpP1P2, but lack activity to the human proteasome. Moreover, a series of dipeptidyl boronate derivatives of bortezomib has also been synthesized (29). Structurerelationship studies reveal several compounds with improved anti-Mtb activity and selectivity for the mycobacterial ClpP1P2 protein. Compound 58 has good pharmacokinetic criteria, such as aqueous solubility and low cytotoxicity. Overall, these inhibitors show that ClpP protease is an attractive target that can be modulated via different mechanisms to kill bacterial persister cells. 12 Modulating intracellular ATP levels to kill persisters Recent studies confirm a critical role of ATP in establishing persistence. For instance, Conlon et al. showed that S. aureus class I persister formation is directly linked to intracellular ATP levels (30), with stationary phase cultures showing significantly reduced ATP content (Fig. 1.1). Artificial reduction of ATP levels in an exponentially growing culture by adding arsenate increases antibiotic tolerance to rifampin, whereas increasing ATP concentration by supplementing tryptic soy broth medium with glucose causes a reduction in class I persister population during stationary phase (30). This phenomenon is also observed in another study in E. coli that shows a drop in intracellular ATP by treatment with arsenate led to antibiotic tolerance to ampicillin and ciprofloxacin (31). These two studies demonstrate that stochastic depletion of ATP intracellular concentration contributes significantly to persister cell formation. The drop in ATP level drives antibiotic tolerance because most antibiotic targets require ATP to function. Bacterial persister cells have thus evolved mechanisms to keep ATP at levels low enough to be viable, while still inducing persistence. One effective strategy to eliminate class II bacterial persistence is to inhibit bacterial respiration and ATP synthesis (Fig. 1.1). It is hypothesized that during dormancy or NRP, bacteria still require some energy for survival, and further depletion of ATP synthesis will be lethal to persister cells. For instance, Rao et al. originally reported the intracellular ATP levels of Mtb rapidly decreases when entering NRP and then is maintained constant over 25 d of incubation in the Wayne model of hypoxia-driven NRP (32). Treating non-replicating Mtb bacilli with the F0F1 ATP synthase inhibitor bedaquiline (diarylquinoline) or DCCD (N,N-dicyclohexylcarbo-diimide) dramatically decreases the ATP content, and lowers cell viability (Fig. 1.2). Bedaquiline is also effective in killing Mtb in a mouse model of TB infection, and sterilizes bacteria in 2 months when 13 included with isoniazid and pyrazinamide treatment (33). The F0F1 ATP synthase is involved in the last step of ATP synthesis, and its function is driven by PMF. Disrupting PMF by inhibiting the membrane potential with valinomycin or disrupting the transmembrane proton concentration gradient with nigericin, also kills non-replicating Mtb (34). This indicates that energy production is required for survival, and decreasing the ATP concentration is lethal to class II persisters. Several inhibitors of ATP homeostasis in non-replicating Mtb were identified from highthroughput screening, including imidazopyridines (GNF-NITD 46), benzimidazole (GNF-NITD 82), and thieno[2,3-b]pyridin-4-amine (GNF-NITD 101) (Fig. 1.2) (35). All three classes of inhibitors show decreased ATP levels and bacterial survival of non-replicating Mtb. In another study, the biphenyl amide compound DG70 (Fig. 1.2) was identified from a screen that targeted non-replicating Mtb under nutrient-deprived conditions (36). DG70 kills both replicating and nonreplicating Mtb and synergizes with other anti-mycobacterial drugs, including INH, bedaquiline and PA824. The compound was found to inhibit oxygen consumption and ATP biosynthesis. The resistant mutants show that this compound targets the MenG protein, which is involved in the last step of menaquinone biosynthesis. This finding is confirmed by biochemical studies showing that DG70 inhibited menaquinone biosynthesis. Overall, these findings suggest sal that DG70 inhibits replicating and non-replicating bacteria by inhibiting Mtb respiration. Disrupting membrane integrity to kill persisters Maintaining membrane integrity is essential for cell survival. Several studies have focused on discovering microbial membrane “disruptors” for controlling non-replicating microbes (Fig. 1.1). For instance, a recent screening conducted by Yang et al. has identified amphiphilic indole derivatives that target Mycobacterium bovis BCG membrane integrity (37). These novel 14 compounds cause BCG cell membrane depolarization and permeabilization in a time-dependent manner. The loss of membrane integrity by compound 74a is bactericidal and leads to a rapid killing of replicating cells, as well as effectively killing non-replicating bacilli in the Wayne model of hypoxia-driven NRP (Fig. 1.2). Boromycin is another example of how the cell membrane is a good target to kill persisters. This compound is a well-known polyether macrolide antibiotic that kills Gram-positive bacteria by functioning as a potassium ionophore. Studies demonstrate that boromycin has bactericidal activity against both replicating and non-replicating Mtb under the Wayne model of hypoxia-driven NRP (38). Like in Gram-positive bacteria, boromycin also causes a reduction of membrane potential in Mtb, as well as decreasing intracellular ATP level, suggesting it also acts as an iononphore in Mtb to collapse the potassium gradient across the membrane. A screen to target non-replicating S. aureus identified the quinolone-derived compound HT61 as being more active against stationary phase non-replicating S. aureus (both methicillin sensitive S. aureus (MSSA) and MRSA) as compared to replicating S. aureus (39). It is also bactericidal in other Gram-positive bacteria, including Streptococcus pyogenes, Streptococcus agalactiae, and Propionibacterium acnes. Non-replicating S. aureus treated with HT61 exhibits more rapid depolarization of the cytoplasmic membrane as compared to replicating exponentialphase bacteria, and causes the cells to burst. HT61 also effectively kills non-replicating S. aureus in a murine skin model. HT61 potentiates killing of S. aureus with neomycin, gentamicin and mupirocin both in vitro and in an in vivo mouse model (40), suggesting that membrane integrity is important for bacterial survival. Many new antibiotics that disrupt cell membranes have been discovered to eliminate persister populations. For instance, Chen et al. discovered that cationic peptides containing Trp/Arg are effective against both planktonic and biofilm-associated persister cell populations 15 (Fig. 1.2) (41). These peptides have been shown to penetrate and cause physical damage to the cell membrane as the mechanism of action in killing persister cells. Overall, these studies provide strong support for eradicating persister cell populations using agents that damage cellular membranes. Inhibition of cell envelope biosynthesis potentiates killing of persisters by antibiotics The Mtb cell envelope is composed of complex glycolipids that create a permeability barrier to the environment. The first-line antibiotic INH inhibits the enoyl-ACP reductase InhA after being activated by the catalase-peroxidase KatG to form an adduct with NAD. InhA is an essential protein involved in synthesis of mycolic acids in Mtb. Inhibition of InhA activity causes mycobacterial cell death. Efforts to discover novel InhA inhibitors that don’t require an activator identified two compounds CD39 ((Z)-2-imino-5-((E)-3-(5-nitrofuran-2-yl)allylidene)thiazolidin4-one) and CD117 (tetrahydrobenzothienopyrimidine) (Fig. 1.2) (42). They are not only bactericidal against the growing M. tuberculosis, but also active against non-replicating bacilli grown under anaerobic conditions. This finding suggests these inhibitors have different mechanisms of action from INH, because isoniazid only kills replicating bacilli (32). Both compounds enhance anti-mycobacterial activity with INH and rifampin, and inhibit mycolic acids and fatty acid biosynthesis. TCA1 is reported to inhibit DprE1, a decaprenylphosphoryl-b-D-ribofuranose oxidoreductase that is involved in cell wall biosynthesis, as well as MoeW, an enzyme involved in molybdenum cofactor synthesis (Fig. 1.2) (43). DprE1 and DprE2 (decaprenylphosphoryl-D-2keto erythro pentose reductase) together catalyze the epimerization of decaprenylphosphoryl ribose to decaprenylphosphoryl arabinose (44). Decaprenylphosphoryl arabinose is the only 16 precursor of arabinogalactan and lipoarabinomannan polysaccharides of the mycobacterial cell envelope. Chemical inhibition of DprE1 abolishes biosynthesis of decaprenylphosphoryl arabinose, and causes cell lysis and bacterial death. There are numerous DprE1 inhibitors identified, including benzothiazinones, dinitrobenzamides, azaindoles (45), pyrazolopyridones (46), 2-carboxyquinoxalines (47), and pyrido-benzimidazole (48). However, TCA1 is the only inhibitor reported to effectively kill both replicating and non-replicating Mtb in a nutrient starvation model. TCA1 sterilizes Mtb cultures treated with INH or rifampin in about 3 weeks, and shows efficacy in acute and chronic TB mouse models. Interestingly, transcriptional profiling also showed that TCA1 downregulates the persistence genes involved in Mtb dormancy, including rv3130c-rv3134c, fdxA and hspX, suggesting TCA1 potentiates Mtb killing by antibiotics by inhibiting Mtb dormancy. The multiple functions of TCA1 differentiate it from other inhibitors. The anti-mycobacterial compounds 1,2-ethylenediamine SQ109 and diphenylethermodified adamantyl 1,2-diamine TBL-140 have been shown to be bactericidal to replicating and non-replicating Mtb by targeting MmpL3 (Fig. 1.2) (49, 50). MmpL3 is a membrane-bound protein that acts as a flippase to translocate the mycolic acid precursor trehalose monomycolate (TMM) to the pseudoperiplasmic space, where TMM is further transported to the mycomembranes and modified into trehalose dimycolate (TDM) by Ag85 protein (51, 52). Treating Mtb cells with SQ109 and TBL-140 inhibits the transport activity of MmpL3, and results in accumulation of TMM in the cytoplasm and thus defective TDM biosynthesis. SQ109 also inhibits membrane potential and generation of the transmembrane proton gradient, while TBL-140 inhibits membrane potential (49, 50). SQ109 also potentiates killing of Mtb by first-line anti-tubercular drugs, including INH and rifampin (53). 17 Additionally, a new class of antibiotic called Art-175, which consists of a bacteriophageencoded endolysin covalently attached to a transporting peptide, functions to kill class I persisters (54). The endolysin degrades the peptidoglycan, and is transported through the outer membrane of Gram-negative bacteria with assistance of the transporting peptide. It was shown that Art-175 is able to effectively kill both replicating P. aeruginosa and persister cells by penetrating the peptidoglycan layers of P. aeruginosa and causing rapid cell lysis. Overall, these findings support the idea that compounds inhibiting cell wall biosynthesis are a good approach for developing antipersistence therapy. Targeting two-component regulatory systems to inhibit bacterial persistence Transcriptional regulators in bacterial pathogens often play essential roles to establish infection and evade host immunity. Anti-virulence therapies target virulence proteins that are not required for growth in vitro, but are necessary during infection. Targeting bacterial virulence is different from conventional antibiotics that directly kill bacteria, and may reduce selection for antibiotic resistant strains (55). The DosRST signaling pathway in Mtb has been implicated in Mtb persistence (Fig. 1.1). The primary focus of this dissertation was to determine if chemical inhibitors of the DosRST pathway could be discovered, and if so, could they function to inhibit Mtb persistence. DosRST is a well-characterized two-component regulatory systems that functions to establish and maintain non-replicating persistence. It is composed of two histidine kinase sensors, DosS and DosT, and the cognate response regulator DosR (56), which regulates expression of approximately 50 genes in the DosR regulon (57). Mtb can sense a variety of environmental cues, including nitric oxide, carbon monoxide and oxygen, through DosS and DosT (57-61). All three 18 ligands of DosST are relevant for Mtb persistence. For instance, NO is often produced by NO synthase as an antibacterial agent to kill bacteria (62). CO is mainly produced by heme oxygenase 1 during the host immune response to influence anti-inflammatory and anti-apoptotic activities (63). Hypoxia is an important environmental stimulus to facilitate Mtb transition to the latent form. The center of caseous granulomas where Mtb resides during latent infection is highly hypoxic (64). Therefore, Mtb has evolved immune evasion strategies through its ability to sense environmental signals and then orchestrate appropriate genetic, metabolic, and physiologic responses to survive within the host. Upon receiving the environmental signals, DosS/T catalyzes an ATP-dependent autophosphorylation at a conserved histidine and transmits the signal by phosphorylating DosR at an aspartate residue (61). Phosphorylated DosR is active and able to regulate transcription of the regulon (56, 65). DosS and DosT share extensive amino acid sequence similarity, ~61%, and have structural homology. Each contains two N-terminal GAF domains (cGMP, adenylyl cyclase, FhlA), a HisKA domain (histidine kinase phospho-acceptor) where a phosphoryl group-accepting histidine is located, and a C-terminal HATPase domain (histidine kinase-like ATPase) that is the ATP binding site (66, 67). The GAF domain functions as a small molecule binding regulatory domain that is present throughout prokaryotes and eukaryotes. Specifically, a b-type heme is embedded in the hydrophobic cavity of the GAF-A domain of DosS and DosT (68, 69). Even though DosS and DosT are heme-based sensor proteins sharing a similar structure, their biochemical mechanism of signal sensing is different. For instance, DosS is a redox and oxygen sensor and autoxidizes quickly under aerobic conditions (70). In contrast, DosT is a hypoxia sensor and has high affinity and sensitivity to O2 (71). Both kinases sense O2 or other ligands via heme, and are inactive when the heme group exists as either the Met (Fe3+) form (DosS) or the oxy (Fe2+- 19 O2) form (DosT) in the presence of O2. However, hypoxic conditions induce the conversion of DosS to the ferrous form and DosT to the deoxy form, activating the kinases (59, 68-70, 72). Therefore, DosS/T plays different roles in sensing the redox and oxygen status of the environment to turn on the DosR pathway. In responding to the environmental cues, phosphorylated DosS/T triggers the induction of the DosR regulon and transitions Mtb into NRP. The non-replicating bacilli are problematic for two reasons: 1) they are naturally tolerant to most anti-mycobacterial drugs and are partially responsible for the required long course of TB therapy (32); and 2) they serve as a potential reservoir for transmitting infection once growth is resumed in favorable environmental conditions. Therefore, targeting DosRST can potentially inhibit Mtb from establishing and maintaining bacterial persistence. The in vitro Wayne model has been employed to study hypoxia-driven NRP in Mtb. Bacteria are grown in a sealed screw cap tube with stirring. Oxygen levels diminish over time as bacteria grow in the culture. In the Wayne model, two distinct phases are observed in a timedependent manner, including NRP1 (initiated at ~1% oxygen) and NRP2 (initiated at <0.06% oxygen). At NRP1, the bacilli cease replication and thicken their cell wall, while at NRP2 the bacilli become anaerobic-adapted bacteria and no longer increase in optical density (73). The DosRST pathway is a potential therapeutic target because in vitro and in vivo studies have shown that Mtb is highly attenuated for survival during NRP in the absence of DosRST. For instance, in an in vitro rapid anaerobic dormancy model, a slightly modified version of the Wayne model with a higher stirring speed, deletion of dosR causes a significant loss of viability as bacilli fail to enter the NRP, and there is also a delay in recovery after re-inoculating the dormant cells into aerobic conditions (74). This finding suggests that DosR plays a functional role for Mtb to 20 enter and exit NRP, as well as survive during NRP. The ΔdosR mutant during NRP shows sharply reduced intracellular ATP and NAD, ~10-fold and ~50%-fold lower than WT, respectively (74). The ΔdosRS mutant also exhibits a growth defect in several animal models, including rabbits, guinea pigs, C3HeB/FeJ mice, and non-human primates (75-77). For instance, rabbits infected with the ΔdosRS mutant display significantly lower lung bacterial burdens at 8-week infection than those infected with WT. WT Mtb also stimulates formation of numerous intense and organized lesions in the lungs at week 5, whereas the mutant exhibits fewer lesions (75). In the guinea pig model of TB infection, ΔdosRS shows a 3-log reduction lung CFUs as compared to the WT at week 6. The mutant also causes less virulence, such as lack of formation of mature granulomas, while WT induces massive inflammation and lesions with central necrosis (75). Furthermore, in the C3HeB/FeJ mouse model of TB infection, infection with ΔdosS leads to 0.5- and 1.5-log lower lung CFUs than infection with WT at week 8 and 12, respectively (76). ΔdosS also results in 11.5 log less liver CFUs at week 12 and 15, respectively, as compared to WT. In the Naïve Indian rhesus macaques model of TB infection, all mutants, including ΔdosR, ΔdosS and ΔdosST, exhibit lower bacterial burden than WT (77). Specifically, the mutant strains have at least a 100-fold lower lung bacterial burden as compared to WT. Little or no TB-related pathology and granuloma formation are observed in the animals infected the mutant, while WT or complemented strains initiate formation of numerous granulomas with necrosis in the central area (77). All these animal models form well-organized hypoxia-associated granulomas when infected with Mtb, which mimics the pathology observed in humans quite well. The DosR-regulated genes also play a role in drug tolerance. For example, tgs1 is a DosR-regulated gene that is highly induced during hypoxic conditions (56, 78, 79). It encodes a triacylglycerol (TAG) synthase, which is involved in 21 the last step of TAG biosynthesis. An Mtb tgs1 mutant strain exhibits reduced tolerance to several antibiotics, such as isoniazid and ciprofloxacin, in vitro and in vivo (80, 81). The in vitro and in vivo data confirm that DosRST is required for full virulence during Mtb infection and for survival during extended periods of hypoxia. Therefore, inhibiting DosRST may help reduce Mtb persistence and drug tolerance and shorten TB treatment. Based on the homology modeling of DosR, a DosR regulon inhibitor, a phenylcoumarin derivative, was found to inhibit DosR binding to DNA (82). Screening a phage display library allowed the discovery of short DosR mimetic peptides that inhibit the autophosphorylation activity of DosS (83). This dissertation describes the work toward the discovery and characterization of small molecules HC101AHC106A (Fig. 1.2) that inhibit the DosRST signaling pathway, and shows that chemical inhibition of the DosRST pathway reduces Mtb persistence-associated physiologies, including TAG biosynthesis, survival and INH antibiotic tolerance. Mechanistic studies show a specific component of the DosRST pathway is targeted by each inhibitor through a distinct mechanisms. Both HC101A and HC106 target the sensor domain of DosS/T via different mechanisms, where HC101A modulates the redox status of DosS/T and alkylates the heme, and HC106A likely binds directly to the heme. HC102A and HC103A are proposed to target the kinase function of DosS/T by inhibiting their authophosphorylation activities. Biochemical assays indicates that HC104A functions by inhibiting DosR binding to promoter DNA. The long-term goal of these studies is to use these inhibitors, or other DosRST inhibitors, in combination with current TB drugs to eradicate persister populations and thus shorten the course of TB therapy. 22 Concluding remarks Experimental evolution studies show that bacteria have the capacity to become tolerant to antibiotics as they evolve in the presence of these substances (84, 85). Phenotypic drug tolerance, driven by persisters, may facilitate the evolution of heritable drug resistance and represents a threat to the durable, long-term deployment of antibiotics. For instance, in in vitro evolution experiments, it was found that continued passaging of the surviving E. coli persisters in fresh medium with a high dose of ampicillin produces strains with high drug tolerance (84). It requires an extended time, measured by minimum duration for killing 99% of the population (86), to kill the evolved bacteria as compared to the WT strain. Imaging the division of single cells of stationary phase culture after passaging into fresh medium shows that evolved bacteria requires a longer time to initiate replication. This finding supports the hypothesis that an extended lag time or dormancy may protect the evolved bacteria from antibiotics. However, the evolved bacteria in the exponential phase of growth are more susceptible to antibiotic treatment than the stationary phase cultures. Thus drug tolerance is lag-time dependent, and this phenomenon is referred to as tolerance by lag (tbl) (84). Additionally, these high-tolerance bacteria also exhibit increased tolerance to other antibiotics, although they were evolved for ampicillin resistance, showing that they are not ampicillin specific. Genetic mutations were identified in the evolved strains, and introducing a WT copy of some genes decreased the lag times to be similar to WT, suggesting that mutant proteins have a role in the high tolerance, thus the encoded genes are referred to as tbl genes (84). These findings are clinically relevant because treatment of infectious disease with antibiotics exerts selection for high-tolerance bacteria, which may eventually reduce treatment effectiveness. 23 Persister cells are also associated with chronic infections and may drive drug resistance in chronic diseases, such as cystic fibrosis and tuberculosis. For example, it was found that strains isolated from cystic fibrosis patients infected with P. aeruginosa exhibit 100-fold greater numbers of persister cells (87). Moreover, studies have suggested that drug tolerance promotes the emergence of drug resistance. For instance, a recent in vitro experimental evolution experiment has demonstrated that passaging the surviving phenotypically drug-tolerant persisters in fresh medium with ampicillin promotes the evolution of genetically encoded drug resistance in mutant strains (88). Three different E. coli strains independently evolve cultures with increased minimum inhibitory concentration as compared with the parental strains, and each carries a mutation in the promoter of ampC that encodes b-lactamase. Further analysis shows that cultures become more phenotypically tolerant over time before evolving heritable drug resistance. The phenotype of high tolerance is not due to slow growth. Additional mutations in genes belonging to the “tolerome” are also present in the mutants with high tolerance. Without the platform of tolerance, mathematical models reveal a lower probability of establishing heritable resistance, suggesting that persistence and drug tolerance contribute significantly to the evolution of resistance. Therefore, bacterial persistence is a fundamental problem, and controlling bacterial persistence is an essential strategy to combat infectious disease and control drug resistance. 24 REFERENCES 25 REFERENCES 1. Lewis K. 2017. New approaches to antimicrobial discovery. Biochem Pharmacol 134:8798. 2. Hobby GL, Meyer K, Chaffee E. 1942. Observations on the mechanism of action of penicillin. Proc Soc Exp Biol Med 50:281-285. 3. Bigger J. 1944. Treatment of Staphylococcal infections with penicillin by intermittent sterilisation. The Lancet 244:497-500. 4. Balaban NQ, Merrin J, Chait R, Kowalik L, Leibler S. 2004. Bacterial persistence as a phenotypic switch. Science 305:1622-1625. 5. Kint CI, Verstraeten N, Fauvart M, Michiels J. 2012. New-found fundamentals of bacterial persistence. Trends Microbiol 20:577-585. 6. Nathan C. 2012. Fresh approaches to anti-infective therapies. Sci Transl Med 4:140sr142. 7. Gold B, Nathan C. 2017. Targeting phenotypically tolerant Mycobacterium tuberculosis. Microbiol Spectr 5:TBTB2-0031- 2016. 8. Allison KR, Brynildsen MP, Collins JJ. 2011. Metabolite-enabled eradication of bacterial persisters by aminoglycosides. Nature 473:216-220. 9. Prax M, Mechler L, Weidenmaier C, Bertram R. 2016. Glucose augments killing efficiency of daptomycin challenged Staphylococcus aureus persisters. PLoS One 11:e0150907. 10. Marques CN, Morozov A, Planzos P, Zelaya HM. 2014. The fatty acid signaling molecule cis-2-decenoic acid increases metabolic activity and reverts persister cells to an antimicrobial-susceptible state. Appl Environ Microbiol 80:6976-6991. 11. Lewis K. 2007. Persister cells, dormancy and infectious disease. Nat Rev Microbiol 5:4856. 12. Morones-Ramirez JR, Winkler JA, Spina CS, Collins JJ. 2013. Silver enhances antibiotic activity against gram-negative bacteria. Sci Transl Med 5:190ra181. 13. Grant SS, Kaufmann BB, Chand NS, Haseley N, Hung DT. 2012. Eradication of bacterial persisters with antibiotic-generated hydroxyl radicals. Proc Natl Acad Sci U S A 109:12147-12152. 26 14. Vilcheze C, Hartman T, Weinrick B, Jacobs WR, Jr. 2013. Mycobacterium tuberculosis is extraordinarily sensitive to killing by a vitamin C-induced Fenton reaction. Nat Commun 4:1-10. 15. Vilcheze C, Hartman T, Weinrick B, Jain P, Weisbrod TR, Leung LW, Freundlich JS, Jacobs WR, Jr. 2017. Enhanced respiration prevents drug tolerance and drug resistance in Mycobacterium tuberculosis. Proc Natl Acad Sci U S A 114:4495-4500. 16. Cui P, Niu H, Shi W, Zhang S, Zhang H, Margolick J, Zhang W, Zhang Y. 2016. Disruption of membrane by colistin kills uropathogenic Escherichia coli persisters and enhances killing of other antibiotics. Antimicrob Agents Chemother 60:6867-6871. 17. Schmidt NW, Deshayes S, Hawker S, Blacker A, Kasko AM, Wong GC. 2014. Engineering persister-specific antibiotics with synergistic antimicrobial functions. ACS Nano 8:8786-8793. 18. Kim W, Conery AL, Rajamuthiah R, Fuchs BB, Ausubel FM, Mylonakis E. 2015. Identification of an antimicrobial agent effective against methicillin-resistant Staphylococcus aureus persisters using a fluorescence-based screening strategy. PLoS One 10:e0127640. 19. Basak A, Abouelhassan Y, Zuo R, Yousaf H, Ding Y, Huigens RW. 2017. Antimicrobial peptide-inspired NH125 analogues: bacterial and fungal biofilm-eradicating agents and rapid killers of MRSA persisters. Org Biomol Chem 15:5503-5512. 20. Pu Y, Zhao Z, Li Y, Zou J, Ma Q, Zhao Y, Ke Y, Zhu Y, Chen H, Baker MA, Ge H, Sun Y, Xie XS, Bai F. 2016. Enhanced efflux activity facilitates drug tolerance in dormant bacterial cells. Mol Cell 62:284-294. 21. Orman MA, Brynildsen MP. 2015. Inhibition of stationary phase respiration impairs persister formation in E. coli. Nat Commun 6:7983. 22. Orman MA, Brynildsen MP. 2016. Persister formation in Escherichia coli can be inhibited by treatment with nitric oxide. Free Radic Biol Med 93:145-154. 23. Brotz-Oesterhelt H, Sass P. 2014. Bacterial caseinolytic proteases as novel targets for antibacterial treatment. Int J Med Microbiol 304:23-30. 24. Brotz-Oesterhelt H, Beyer D, Kroll HP, Endermann R, Ladel C, Schroeder W, Hinzen B, Raddatz S, Paulsen H, Henninger K, Bandow JE, Sahl HG, Labischinski H. 2005. Dysregulation of bacterial proteolytic machinery by a new class of antibiotics. Nat Med 11:1082-1087. 25. Conlon BP, Nakayasu ES, Fleck LE, LaFleur MD, Isabella VM, Coleman K, Leonard SN, Smith RD, Adkins JN, Lewis K. 2013. Activated ClpP kills persisters and eradicates a chronic biofilm infection. Nature 503:365-370. 27 26. Gavrish E, Sit CS, Cao S, Kandror O, Spoering A, Peoples A, Ling L, Fetterman A, Hughes D, Bissell A, Torrey H, Akopian T, Mueller A, Epstein S, Goldberg A, Clardy J, Lewis K. 2014. Lassomycin, a ribosomally synthesized cyclic peptide, kills Mycobacterium tuberculosis by targeting the ATP-dependent protease ClpC1P1P2. Chem Biol 21:509-518. 27. Moreira W, Ngan GJ, Low JL, Poulsen A, Chia BC, Ang MJ, Yap A, Fulwood J, Lakshmanan U, Lim J, Khoo AY, Flotow H, Hill J, Raju RM, Rubin EJ, Dick T. 2015. Target mechanism-based whole-cell screening identifies bortezomib as an inhibitor of caseinolytic protease in mycobacteria. MBio 6:e00253-00215. 28. Moreira W, Santhanakrishnan S, Dymock BW, Dick T. 2017. Bortezomib warheadswitch confers dual activity against mycobacterial caseinolytic protease and proteasome and selectivity against human proteasome. Front Microbiol 8:1-6. 29. Moreira W, Santhanakrishnan S, Ngan GJY, Low CB, Sangthongpitag K, Poulsen A, Dymock BW, Dick T. 2017. Towards selective mycobacterial ClpP1P2 inhibitors with reduced activity against the human proteasome. Antimicrob Agents Chemother 61:e0230702316. 30. Conlon BP, Rowe SE, Gandt AB, Nuxoll AS, Donegan NP, Zalis EA, Clair G, Adkins JN, Cheung AL, Lewis K. 2016. Persister formation in Staphylococcus aureus is associated with ATP depletion. Nat Microbiol 1:16051. 31. Shan Y, Brown Gandt A, Rowe SE, Deisinger JP, Conlon BP, Lewis K. 2017. ATPdependent persister formation in Escherichia coli. MBio 8:e02267-02216. 32. Rao SP, Alonso S, Rand L, Dick T, Pethe K. 2008. The protonmotive force is required for maintaining ATP homeostasis and viability of hypoxic, nonreplicating Mycobacterium tuberculosis. Proc Natl Acad Sci U S A 105:11945-11950. 33. Andries K, Verhasselt P, Guillemont J, Gohlmann HW, Neefs JM, Winkler H, Van Gestel J, Timmerman P, Zhu M, Lee E, Williams P, de Chaffoy D, Huitric E, Hoffner S, Cambau E, Truffot-Pernot C, Lounis N, Jarlier V. 2005. A diarylquinoline drug active on the ATP synthase of Mycobacterium tuberculosis. Science 307:223-227. 34. Pethe K, Bifani P, Jang J, Kang S, Park S, Ahn S, Jiricek J, Jung J, Jeon HK, Cechetto J, Christophe T, Lee H, Kempf M, Jackson M, Lenaerts AJ, Pham H, Jones V, Seo MJ, Kim YM, Seo M, Seo JJ, Park D, Ko Y, Choi I, Kim R, Kim SY, Lim S, Yim SA, Nam J, Kang H, Kwon H, Oh CT, Cho Y, Jang Y, Kim J, Chua A, Tan BH, Nanjundappa MB, Rao SP, Barnes WS, Wintjens R, Walker JR, Alonso S, Lee S, Kim J, Oh S, Oh T, Nehrbass U, Han SJ, No Z, et al. 2013. Discovery of Q203, a potent clinical candidate for the treatment of tuberculosis. Nat Med 19:1157-1160. 35. Mak PA, Rao SP, Ping Tan M, Lin X, Chyba J, Tay J, Ng SH, Tan BH, Cherian J, Duraiswamy J, Bifani P, Lim V, Lee BH, Ling Ma N, Beer D, Thayalan P, Kuhen K, 28 Chatterjee A, Supek F, Glynne R, Zheng J, Boshoff HI, Barry CE, 3rd, Dick T, Pethe K, Camacho LR. 2012. A high-throughput screen to identify inhibitors of ATP homeostasis in non-replicating Mycobacterium tuberculosis. ACS Chem Biol 7:11901197. 36. Sukheja P, Kumar P, Mittal N, Li SG, Singleton E, Russo R, Perryman AL, Shrestha R, Awasthi D, Husain S, Soteropoulos P, Brukh R, Connell N, Freundlich JS, Alland D. 2017. A novel small-molecule inhibitor of the Mycobacterium tuberculosis demethylmenaquinone methyltransferase MenG is bactericidal to both growing and nutritionally deprived persister cells. MBio 8:e02022-02016. 37. Yang T, Moreira W, Nyantakyi SA, Chen H, Aziz DB, Go ML, Dick T. 2017. Amphiphilic indole derivatives as antimycobacterial agents: structure-activity relationships and membrane targeting properties. J Med Chem 60:2745-2763. 38. Moreira W, Aziz DB, Dick T. 2016. Boromycin kills mycobacterial persisters without detectable resistance. Front Microbiol 7:1-7. 39. Hu Y, Shamaei-Tousi A, Liu Y, Coates A. 2010. A new approach for the discovery of antibiotics by targeting non-multiplying bacteria: a novel topical antibiotic for staphylococcal infections. PLoS One 5:e11818. 40. Hu Y, Coates AR. 2013. Enhancement by novel anti-methicillin-resistant Staphylococcus aureus compound HT61 of the activity of neomycin, gentamicin, mupirocin and chlorhexidine: in vitro and in vivo studies. J Antimicrob Chemother 68:374-384. 41. Chen X, Zhang M, Zhou C, Kallenbach NR, Ren D. 2011. Control of bacterial persister cells by Trp/Arg-containing antimicrobial peptides. Appl Environ Microbiol 77:48784885. 42. Vilcheze C, Baughn AD, Tufariello J, Leung LW, Kuo M, Basler CF, Alland D, Sacchettini JC, Freundlich JS, Jacobs WR, Jr. 2011. Novel inhibitors of InhA efficiently kill Mycobacterium tuberculosis under aerobic and anaerobic conditions. Antimicrob Agents Chemother 55:3889-3898. 43. Wang F, Sambandan D, Halder R, Wang J, Batt SM, Weinrick B, Ahmad I, Yang P, Zhang Y, Kim J, Hassani M, Huszar S, Trefzer C, Ma Z, Kaneko T, Mdluli KE, Franzblau S, Chatterjee AK, Johnsson K, Mikusova K, Besra GS, Futterer K, Robbins SH, Barnes SW, Walker JR, Jacobs WR, Jr., Schultz PG. 2013. Identification of a small molecule with activity against drug-resistant and persistent tuberculosis. Proc Natl Acad Sci U S A 110:E2510-2517. 44. Mikusova K, Huang H, Yagi T, Holsters M, Vereecke D, D'Haeze W, Scherman MS, Brennan PJ, McNeil MR, Crick DC. 2005. Decaprenylphosphoryl arabinofuranose, the donor of the D-arabinofuranosyl residues of mycobacterial arabinan, is formed via a twostep epimerization of decaprenylphosphoryl ribose. J Bacteriol 187:8020-8025. 29 45. Shirude PS, Shandil R, Sadler C, Naik M, Hosagrahara V, Hameed S, Shinde V, Bathula C, Humnabadkar V, Kumar N, Reddy J, Panduga V, Sharma S, Ambady A, Hegde N, Whiteaker J, McLaughlin RE, Gardner H, Madhavapeddi P, Ramachandran V, Kaur P, Narayan A, Guptha S, Awasthy D, Narayan C, Mahadevaswamy J, Vishwas KG, Ahuja V, Srivastava A, Prabhakar KR, Bharath S, Kale R, Ramaiah M, Choudhury NR, Sambandamurthy VK, Solapure S, Iyer PS, Narayanan S, Chatterji M. 2013. Azaindoles: noncovalent DprE1 inhibitors from scaffold morphing efforts, kill Mycobacterium tuberculosis and are efficacious in vivo. J Med Chem 56:9701-9708. 46. Panda M, Ramachandran S, Ramachandran V, Shirude PS, Humnabadkar V, Nagalapur K, Sharma S, Kaur P, Guptha S, Narayan A, Mahadevaswamy J, Ambady A, Hegde N, Rudrapatna SS, Hosagrahara VP, Sambandamurthy VK, Raichurkar A. 2014. Discovery of pyrazolopyridones as a novel class of noncovalent DprE1 inhibitor with potent anti-mycobacterial activity. J Med Chem 57:4761-4771. 47. Neres J, Hartkoorn RC, Chiarelli LR, Gadupudi R, Pasca MR, Mori G, Venturelli A, Savina S, Makarov V, Kolly GS, Molteni E, Binda C, Dhar N, Ferrari S, Brodin P, Delorme V, Landry V, de Jesus Lopes Ribeiro AL, Farina D, Saxena P, Pojer F, Carta A, Luciani R, Porta A, Zanoni G, De Rossi E, Costi MP, Riccardi G, Cole ST. 2015. 2-Carboxyquinoxalines kill Mycobacterium tuberculosis through noncovalent inhibition of DprE1. ACS Chem Biol 10:705-714. 48. Warrier T, Kapilashrami K, Argyrou A, Ioerger TR, Little D, Murphy KC, Nandakumar M, Park S, Gold B, Mi J, Zhang T, Meiler E, Rees M, SomersanKarakaya S, Porras-De Francisco E, Martinez-Hoyos M, Burns-Huang K, Roberts J, Ling Y, Rhee KY, Mendoza-Losana A, Luo M, Nathan CF. 2016. N-methylation of a bactericidal compound as a resistance mechanism in Mycobacterium tuberculosis. Proc Natl Acad Sci U S A 113:E4523-4530. 49. Li W, Upadhyay A, Fontes FL, North EJ, Wang Y, Crans DC, Grzegorzewicz AE, Jones V, Franzblau SG, Lee RE, Crick DC, Jackson M. 2014. Novel insights into the mechanism of inhibition of MmpL3, a target of multiple pharmacophores in Mycobacterium tuberculosis. Antimicrob Agents Chemother 58:6413-6423. 50. Foss MH, Pou S, Davidson PM, Dunaj JL, Winter RW, Pou S, Licon MH, Doh JK, Li Y, Kelly JX, Dodean RA, Koop DR, Riscoe MK, Purdy GE. 2016. DiphenyletherModified 1,2-Diamines with Improved Drug Properties for Development against Mycobacterium tuberculosis. ACS Infect Dis 2:500-508. 51. Tahlan K, Wilson R, Kastrinsky DB, Arora K, Nair V, Fischer E, Barnes SW, Walker JR, Alland D, Barry CE, 3rd, Boshoff HI. 2012. SQ109 targets MmpL3, a membrane transporter of trehalose monomycolate involved in mycolic acid donation to the cell wall core of Mycobacterium tuberculosis. Antimicrob Agents Chemother 56:1797-1809. 30 52. Grzegorzewicz AE, Pham H, Gundi VA, Scherman MS, North EJ, Hess T, Jones V, Gruppo V, Born SE, Kordulakova J, Chavadi SS, Morisseau C, Lenaerts AJ, Lee RE, McNeil MR, Jackson M. 2012. Inhibition of mycolic acid transport across the Mycobacterium tuberculosis plasma membrane. Nat Chem Biol 8:334-341. 53. Chen P, Gearhart J, Protopopova M, Einck L, Nacy CA. 2006. Synergistic interactions of SQ109, a new ethylene diamine, with front-line antitubercular drugs in vitro. J Antimicrob Chemother 58:332-337. 54. Briers Y, Walmagh M, Grymonprez B, Biebl M, Pirnay JP, Defraine V, Michiels J, Cenens W, Aertsen A, Miller S, Lavigne R. 2014. Art-175 is a highly efficient antibacterial against multidrug-resistant strains and persisters of Pseudomonas aeruginosa. Antimicrob Agents Chemother 58:3774-3784. 55. Johnson BK, Abramovitch RB. 2017. Small molecules that sabotage bacterial virulence. Trends Pharmacol Sci 38:339-362. 56. Park HD, Guinn KM, Harrell MI, Liao R, Voskuil MI, Tompa M, Schoolnik GK, Sherman DR. 2003. Rv3133c/dosR is a transcription factor that mediates the hypoxic response of Mycobacterium tuberculosis. Mol Microbiol 48:833-843. 57. Voskuil MI, Schnappinger D, Visconti KC, Harrell MI, Dolganov GM, Sherman DR, Schoolnik GK. 2003. Inhibition of respiration by nitric oxide induces a Mycobacterium tuberculosis dormancy program. J Exp Med 198:705-713. 58. Vos MH, Bouzhir-Sima L, Lambry JC, Luo H, Eaton-Rye JJ, Ioanoviciu A, Ortiz de Montellano PR, Liebl U. 2012. Ultrafast ligand dynamics in the heme-based GAF sensor domains of the histidine kinases DosS and DosT from Mycobacterium tuberculosis. Biochemistry 51:159-166. 59. Kumar A, Toledo JC, Patel RP, Lancaster JR, Jr., Steyn AJ. 2007. Mycobacterium tuberculosis DosS is a redox sensor and DosT is a hypoxia sensor. Proc Natl Acad Sci U S A 104:11568-11573. 60. Ohno H, Zhu G, Mohan VP, Chu D, Kohno S, Jacobs WR, Jr., Chan J. 2003. The effects of reactive nitrogen intermediates on gene expression in Mycobacterium tuberculosis. Cell Microbiol 5:637-648. 61. Roberts DM, Liao RP, Wisedchaisri G, Hol WG, Sherman DR. 2004. Two sensor kinases contribute to the hypoxic response of Mycobacterium tuberculosis. J Biol Chem 279:23082-23087. 62. Gomez JE, McKinney JD. 2004. M. tuberculosis persistence, latency, and drug tolerance. Tuberculosis (Edinb) 84:29-44. 31 63. Shiloh MU, Manzanillo P, Cox JS. 2008. Mycobacterium tuberculosis senses hostderived carbon monoxide during macrophage infection. Cell Host Microbe 3:323-330. 64. Via LE, Lin PL, Ray SM, Carrillo J, Allen SS, Eum SY, Taylor K, Klein E, Manjunatha U, Gonzales J, Lee EG, Park SK, Raleigh JA, Cho SN, McMurray DN, Flynn JL, Barry CE, 3rd. 2008. Tuberculous granulomas are hypoxic in guinea pigs, rabbits, and nonhuman primates. Infect Immun 76:2333-2340. 65. Wisedchaisri G, Wu M, Sherman DR, Hol WG. 2008. Crystal structures of the response regulator DosR from Mycobacterium tuberculosis suggest a helix rearrangement mechanism for phosphorylation activation. J Mol Biol 378:227-242. 66. Sardiwal S, Kendall SL, Movahedzadeh F, Rison SC, Stoker NG, Djordjevic S. 2005. A GAF domain in the hypoxia/NO-inducible Mycobacterium tuberculosis DosS protein binds haem. J Mol Biol 353:929-936. 67. Sivaramakrishnan S, de Montellano PR. 2013. The DosS-DosT/DosR mycobacterial sensor system. Biosensors 3:259-282. 68. Cho HY, Cho HJ, Kim YM, Oh JI, Kang BS. 2009. Structural insight into the hemebased redox sensing by DosS from Mycobacterium tuberculosis. J Biol Chem 284:1305713067. 69. Podust LM, Ioanoviciu A, Ortiz de Montellano PR. 2008. 2.3 A X-ray structure of the heme-bound GAF domain of sensory histidine kinase DosT of Mycobacterium tuberculosis. Biochemistry 47:12523-12531. 70. Ioanoviciu A, Yukl ET, Moenne-Loccoz P, de Montellano PR. 2007. DevS, a hemecontaining two-component oxygen sensor of Mycobacterium tuberculosis. Biochemistry 46:4250-4260. 71. Cho HY, Cho HJ, Kim MH, Kang BS. 2011. Blockage of the channel to heme by the E87 side chain in the GAF domain of Mycobacterium tuberculosis DosS confers the unique sensitivity of DosS to oxygen. FEBS Lett 585:1873-1878. 72. Sousa EH, Tuckerman JR, Gonzalez G, Gilles-Gonzalez MA. 2007. DosT and DevS are oxygen-switched kinases in Mycobacterium tuberculosis. Protein Sci 16:1708-1719. 73. Chao MC, Rubin EJ. 2010. Letting sleeping dos lie: does dormancy play a role in tuberculosis? Annu Rev Microbiol 64:293-311. 74. Leistikow RL, Morton RA, Bartek IL, Frimpong I, Wagner K, Voskuil MI. 2010. The Mycobacterium tuberculosis DosR regulon assists in metabolic homeostasis and enables rapid recovery from nonrespiring dormancy. J Bacteriol 192:1662-1670. 32 75. Converse PJ, Karakousis PC, Klinkenberg LG, Kesavan AK, Ly LH, Allen SS, Grosset JH, Jain SK, Lamichhane G, Manabe YC, McMurray DN, Nuermberger EL, Bishai WR. 2009. Role of the dosR-dosS two-component regulatory system in Mycobacterium tuberculosis virulence in three animal models. Infect Immun 77:12301237. 76. Gautam US, McGillivray A, Mehra S, Didier PJ, Midkiff CC, Kissee RS, Golden NA, Alvarez X, Niu T, Rengarajan J, Sherman DR, Kaushal D. 2015. DosS is required for the complete virulence of Mycobacterium tuberculosis in mice with classical granulomatous lesions. Am J Respir Cell Mol Biol 52:708-716. 77. Mehra S, Foreman TW, Didier PJ, Ahsan MH, Hudock TA, Kissee R, Golden NA, Gautam US, Johnson AM, Alvarez X, Russell-Lodrigue KE, Doyle LA, Roy CJ, Niu T, Blanchard JL, Khader SA, Lackner AA, Sherman DR, Kaushal D. 2015. The DosR Regulon Modulates Adaptive Immunity and Is Essential for Mycobacterium tuberculosis Persistence. Am J Respir Crit Care Med 191:1185-1196. 78. Zheng H, Colvin CJ, Johnson BK, Kirchhoff PD, Wilson M, Jorgensen-Muga K, Larsen SD, Abramovitch RB. 2017. Inhibitors of Mycobacterium tuberculosis DosRST signaling and persistence. Nat Chem Biol 13:218-225. 79. Rustad TR, Harrell MI, Liao R, Sherman DR. 2008. The enduring hypoxic response of Mycobacterium tuberculosis. PLoS One 3:e1502. 80. Deb C, Lee CM, Dubey VS, Daniel J, Abomoelak B, Sirakova TD, Pawar S, Rogers L, Kolattukudy PE. 2009. A novel in vitro multiple-stress dormancy model for Mycobacterium tuberculosis generates a lipid-loaded, drug-tolerant, dormant pathogen. PLoS One 4:e6077. 81. Baek SH, Li AH, Sassetti CM. 2011. Metabolic regulation of mycobacterial growth and antibiotic sensitivity. PLoS Biol 9:e1001065. 82. Gupta RK, Thakur TS, Desiraju GR, Tyagi JS. 2009. Structure-based design of DevR inhibitor active against nonreplicating Mycobacterium tuberculosis. J Med Chem 52:63246334. 83. Kaur K, Taneja NK, Dhingra S, Tyagi JS. 2014. DevR (DosR) mimetic peptides impair transcriptional regulation and survival of Mycobacterium tuberculosis under hypoxia by inhibiting the autokinase activity of DevS sensor kinase. BMC Microbiol 14:1-9. 84. Fridman O, Goldberg A, Ronin I, Shoresh N, Balaban NQ. 2014. Optimization of lag time underlies antibiotic tolerance in evolved bacterial populations. Nature 513:418-421. 85. Van den Bergh B, Michiels JE, Wenseleers T, Windels EM, Boer PV, Kestemont D, De Meester L, Verstrepen KJ, Verstraeten N, Fauvart M, Michiels J. 2016. Frequency 33 of antibiotic application drives rapid evolutionary adaptation of Escherichia coli persistence. Nat Microbiol 1:16020. 86. Brauner A, Shoresh N, Fridman O, Balaban NQ. 2017. An experimental framework for quantifying bacterial tolerance. Biophys J 112:2664-2671. 87. Mulcahy LR, Burns JL, Lory S, Lewis K. 2010. Emergence of Pseudomonas aeruginosa strains producing high levels of persister cells in patients with cystic fibrosis. J Bacteriol 192:6191-6199. 88. Levin-Reisman I, Ronin I, Gefen O, Braniss I, Shoresh N, Balaban NQ. 2017. Antibiotic tolerance facilitates the evolution of resistance. Science 355:826-830. 34 CHAPTER 2 - Inhibitors of Mycobacterium tuberculosis DosRST signaling and persistence This discovery and initial characterization of HC101A-HC103A presented in this chapter has been previously published: Zheng H., Colvin C.J., Johnson B.K., Kirchhoff P.D., Wilson M., Jorgensen-Muga K., Larsen S.D., and Abramovitch R.B. (2017). Inhibitors of Mycobacterium tuberculosis DosRST signaling and persistence. Nature Chemical Biology, 13, 218-225. Author contributions for the study: H.Z., C.J.C., B.K.J. and R.B.A. conducted high-throughput screen. H.Z. performed the Mtb physiology and biochemistry experiments. M.W. and S.D.L. synthesized chemical compounds; K.J.-M. contributed reagents. P.D.K. performed structural modeling. 35 Summary The Mycobacterium tuberculosis (Mtb) DosRST two-component regulatory system promotes the survival of Mtb during non-replicating persistence (NRP). NRP bacteria help drive the long course of tuberculosis therapy, therefore, chemical inhibition of DosRST may inhibit the ability of Mtb to establish persistence and thus shorten treatment. Using a DosRST-dependent fluorescent Mtb reporter strain, a whole-cell phenotypic high-throughput screen of a ~540,000 compound small molecule library was conducted. The screen discovered novel inhibitors of the DosRST regulon, including three compounds that were subject to follow-up studies: artemisinin, HC102A and HC103A. Under hypoxia, all three compounds inhibit Mtb persistence associated physiologies, including triacylglycerol synthesis, survival and antibiotic tolerance. Artemisinin functions by disabling the heme-based DosS/T sensor kinases by oxidizing ferrous heme and generating hemeartemisinin adducts. In contrast, HC103A inhibits DosS and DosT autophosphorylation activity without targeting the sensor kinase heme. 36 Introduction Mycobacterium tuberculosis (Mtb), the causative agent of tuberculosis (TB), can persist in the host for decades without causing disease symptoms (1). Mtb non-replicating persistence (NRP) is characterized by a gradual slowing of metabolic activity upon encountering pressures from the host immune system, including hypoxia, acidic pH or starvation (2-4). The DosRST twocomponent regulatory system plays an important role in Mtb NRP physiology (5). It is composed of two heme-based histidine sensor kinases, DosS and DosT, and the response regulator DosR, and strongly regulates the expression of approximately 50 genes known as the DosRST regulon (6-8). Mtb can sense host stimuli, including nitric oxide (NO), carbon monoxide (CO) and oxygen (O2), through DosS and DosT (9), with DosS acting as an oxygen and redox sensor and DosT acting as an oxygen sensor (7, 9-11). During hypoxia-driven NRP, DosT is associated with initiating expression of the DosR regulon in response to hypoxia and DosS promotes sustained expression of the DosR regulon (12). A fundamental challenge of current TB therapy is the long course of treatment. New drugs that shorten the course of therapy could revolutionize TB control. dosR mutants have reduced survival during hypoxia in vitro (13) and reduced virulence in rabbits, guinea pigs, non-human primates, and C3HeB/FeJ mice (14-16), animal models that generate hypoxic granulomas where DosR-dependent persistence is predicted to be required for survival. Additionally, disruption of a DosR regulated gene, tgs1, results in enhanced sensitivity of Mtb to antibiotics in vitro and during mouse infection (17). Therefore, chemical inhibition of the DosR regulon may stop the establishment and survival of persistent, drug-tolerant Mtb in the granuloma. Whole-cell, phenotypic high-throughput screening (HTS) is a powerful tool in Mtb drug discovery. However, many important virulence targets are not accessible to whole-cell HTS 37 because the targets are only required during infection and not required in vitro. One method to address this shortcoming is to conduct a HTS of Mtb growing inside macrophages, which has been successfully employed to discover inhibitors specifically targeting physiology required during infection (18). An alternative approach is to use reporter strains that indicate the stimulation of a virulence pathway and then screening for inhibitors of a reporter signal, an approach that has been successfully employed to discover antivirulence compounds (19-23). The goal of this study was to conduct a reporter-based phenotypic HTS to discover chemical inhibitors of the DosRST regulon and Mtb persistence. Here we report the discovery and characterization of new chemical compounds that reduce expression of DosRST regulon genes, inhibit Mtb persistence-associated physiologies and directly inhibit the DosS/T sensor kinases. Results Discovery of DosRST regulon inhibitors A whole-cell phenotypic high-throughput screen was conducted to identify small molecule inhibitors of DosRST. The CDC1551(hspX’::GFP) fluorescent reporter strain was previously reported to exhibit DosR-dependent GFP fluorescence that is induced by hypoxia and NO (24). Notably, the reporter is strongly induced under conditions of mild hypoxia (e.g. 2% O2) where Mtb is capable of robust growth (24). dosR mutant strains grow well in rich medium until oxygen is almost fully consumed (5, 13), therefore, discovery of compounds that inhibit hypoxia-inducible reporter fluorescence, but leave growth unaffected, may be inhibitors of the DosRST pathway. To discover inhibitors of the DosRST regulon, the CDC1551(hspX’::GFP) reporter strain was used to screen a 540,288 compound library. The reporter strain was grown in rich medium with individual compounds (at a screening concentration of ~10 µM) in 384-well plates and incubated 38 for 6 d. Growth of Mtb causes the consumption of oxygen and promotes hypoxic conditions at the bottom of the well. GFP fluorescence and growth (as measured by optical density) were measured after 6 d incubation. For analysis of hits, fluorescence and growth inhibition were normalized at 100% or 0% inhibition based on rifampin and dimethyl sulfoxide (DMSO) controls, respectively. The Z-factor for the screen was 0.9 and the variation of controls was limited (Supplementary Fig. 2.1ab) confirming that the screen was robust. Hits were then distinguished based on their ability to specifically inhibit reporter fluorescence or as general inhibitors of Mtb growth. Putative DosRST pathway inhibitors were defined as compounds that exhibit >1.5-fold greater fluorescence inhibition as compared to growth inhibition (p<0.0003, Supplementary Fig. 2.1c) with at least 35% fluorescence inhibition (Class 1 inhibitors, Fig. 2.1a). Fresh powders of several putative DosRST inhibitors were obtained and tested in secondary assays to confirm activity, and exclude compounds with GFP quenching activity and eukaryotic cytotoxicity. Six distinct scaffolds named HC101-HC106 (Fig. 2.1b) were confirmed as inhibitors of reporter fluorescence, while exhibiting no GFP quenching activity and limited eukaryotic cytotoxicity (e.g. half maximal effective concentration (EC50) >70 µM for eukaryotic cytotoxicity in murine bone marrow derived macrophages, Supplementary Table 2.1). The most frequently identified scaffold from the primary screen was artemisinin (HC101A) and its analogs artemether, artesunate and dihydroartemisinin (DHA, Fig. 2.1a). This scaffold was identified as nine independent hits in the screen. Artemisinin and its analogs inhibit CDC1551(hspX’::GFP) reporter fluorescence with an EC50 ranging from 1.2-3.7 µM (Fig. 2.1c, Supplementary Fig. 2.2a, Supplementary Table 2.1), while the growth inhibition EC50 is >80 µM, indicating a limited impact on growth. HC102A (1, (+/-) (5S,9R)-7,7,9-trimethyl-1,3- diazaspiro[4.5]decane-2,4-dione) was isolated as a singleton and inhibits dosR-dependent GFP 39 fluorescence with an EC50 of 12.4 µM, while not inhibiting Mtb growth (e.g. a growth inhibition EC50 >80 µM). HC103A (2, N-[3-[(3-hydroxyphenyl)carbamoyl]phenyl]thiophene-2- carboxamide) and the analog HC103B inhibit dosR-dependent GFP fluorescence with an EC50 of 2.7 µM and 5.0 µM, respectively (Fig. 2.1c, Supplementary Fig. 2.2b) while not inhibiting growth (e.g. a growth inhibition EC50 >80 µM). (dimethylamino)propyl]benzo[de]isoquinoline-1,3-dione), HC104A (3, 6-bromo-2-[3- HC105A (4, 9-ethyl-3-((4- (propylsulfonyl)piperazin-1-ium-1-yl)methyl)-9H-carbazol-9-ium oxalate) and HC106A (5, 1(2,4-dichlorophenyl)-3-(1,2-oxazol-5-yl)urea) inhibit dosR-dependent GFP fluorescence with EC50 of 2.8 µM, 12.7 µM and 6.9 µM respectively, while not inhibiting Mtb growth (e.g. a growth inhibition EC50 >80 µM; Supplementary Fig. 2.2b and Supplementary Table 2.1). Artemisinin, HC102A and HC103A were selected for proof-of-concept follow-up experiments characterizing their ability to inhibit dosRST signaling. Both HC102A and HC103A were regenerated by organic synthesis and confirmed to have the activity of the commercially sourced compounds, thus confirming the assigned structures as the active structures (Supplementary Fig. 2.2cd). The DosRST regulon is strongly induced by hypoxia and nitric oxide and composed of ~50 genes that are directly regulated by DosR (6). An additional >100 genes are also differentially expressed in a dosR mutant, possibly due to weak binding by DosR or indirect consequences of misregulated DosR regulon genes (8). To investigate the inhibitory mechanism of the compounds, RNAseq-based transcriptional profiling was undertaken on CDC1551 treated with 40 µM artemisinin, HC102A, HC103A or a DMSO control. The cultures were grown in standing flasks where growth causes the consumption of oxygen, and following 6 d of treatment RNA was isolated, sequenced and analyzed (25) (Supplementary Datasets 2.1-3). Artemisinin caused the strong downregulation of well-characterized DosR regulon genes, including hspX, fdxA, tgs1, and 40 dosRS (6) (Fig. 2.2a). Real-time PCR confirms the RNA-seq data with hspX, tgs1 and dosRS showing 51-, 166-, and 37-fold inhibition by artemisinin, respectively (Supplementary Fig. 2.3a). Artemisinin inhibited 85 genes (>2-fold, p<0.05) that are also repressed in the CDC1551(DdosR) mutant, accounting for greater than two-thirds of the 125 downregulated genes in the CDC1551(DdosR) mutant (Fig. 2.2b, Supplementary Dataset 2.1). Notably, artemisinin also inhibited 157 genes that are not modulated in the CDC1551(DdosR) mutant, suggesting the drug is also impacting DosRST-independent targets. HC102A and HC103A also inhibited DosRST regulon genes, however, in contrast to artemisinin, HC102A and HC103A showed greater specificity for inhibition of the DosRST regulon; for example, 48 out of 55 genes downregulated by HC102A and 76 out of 90 genes downregulated by HC103A are also downregulated in the CDC1551(DdosR) mutant (Fig. 2.2bc, Supplementary Fig. 2.3cd, Supplementary Dataset 2.1). These transcriptional profiles demonstrate that artemisinin, HC102A and HC103A inhibit induction of the core DosRST regulon. To further assess the specificity of the compounds for inhibition of the DosRST pathway, a CDC1551(DdosR) mutant was treated with the compounds of interest, with the hypothesis that compounds specific for the DosRST pathway will not modulate gene expression in the CDC1551(DdosR) mutant. The CDC1551(DdosR) mutant treated with HC102A or HC103A, exhibited only 0 and 13 downregulated genes, respectively, confirming the on-target specificity of HC102A and HC103A (Fig. 2.2c, Supplementary Dataset 2.2). In contrast, the CDC1551(DdosR) mutant treated with artemisinin exhibited 69 downregulated genes (Fig. 2.2c, Supplementary Dataset 2.2), confirming substantial off-target activity for artemisinin. Overall, these data further support the hypothesis that artemisinin, HC102A and HC103A function to inhibit the core DosRST 41 regulon, with HC102A and HC103A showing strong specificity for the intended target of the DosRST regulon. The DosRST pathway is also induced by NO and vitamin C (7, 11, 26) and it was examined if the inhibitors could suppress induction of the DosR pathway by these stimuli. CDC1551 was pre-treated with DHA, HC102A or HC103A for 1 d prior to induction with NO or vitamin C. As markers for the DosR regulon, the expression of three DosR regulated genes (dosR, tgs1 and hspX) was monitored by real-time PCR. dosR, hspX and tgs1, were strongly up-regulated when Mtb was treated with vitamin C or NO-donor reagent, DETA-NONOate (Supplementary Fig. 2.3b). For example, vitamin C caused a 4-, 14- and 52-fold induction of dosR, hspX and tgs1, respectively, and NO caused an 491-, 373-, and 47-fold induction of dosR, hspX and tgs1, respectively. Pretreatment with HC102A or HC103A strongly inhibited the induction of dosR, hspX and tgs1 transcripts in response to both vitamin C and DETA-NONOate. For example, in HC102A and HC103A pretreated cells the tgs1 transcript following treatment with DETA-NONOate is repressed 3-fold and 50-fold, respectively, whereas tgs1 is induced >47-fold in the DMSO pretreated cells. Similarly, in vitamin C treated cells, the tgs1 transcript is repressed 2-fold and 3fold and in HC102A and HC103A pretreated cells, respectively, while induced >50 fold in the DMSO treated cells. DHA only weakly inhibited the induction of the DosR regulated genes by NO or vitamin C. In DHA pretreated cells, the dosR, hspX and tgs1 transcripts remain induced by both vitamin C and DETA-NONOate treatments. Notably, the magnitude of the induction of the transcripts in response to DETA-NONOate is significantly reduced ~2-fold in DHA pretreated cells compared to the DMSO treatment, demonstrating partial inhibition of NO-dependent DosRST signaling by DHA. These data support the view that HC102A and HC103A act as broad inhibitors of the DosRST regulon in response to both hypoxia and redox environmental cues. In 42 contrast, artemisinin likely acts by a mechanism that is distinct from HC102A and HC103A, given its limited effectiveness to inhibit redox-mediated stimulation of the DosRST regulon. 43 Figure 2.1. Discovery of inhibitors of the DosRST pathway. (a) Scatter plot of primary screening data showing compounds that inhibit CDC1551(hspX’::GFP) reporter GFP fluorescence with limited impact on Mtb growth. Six distinct classes of compounds (HC101-HC106) are highlighted. (b) Structures of compounds confirmed to selectively inhibit CDC1551(hspX’::GFP) reporter fluorescence. (c) Dose response curves for artemisinin (Art, HC101A), HC102A and HC103A inhibition of GFP fluorescence. Dose response curves for other characterized molecules are presented in Supplementary Fig. 2.1. 44 Figure 2.2. Transcriptional profiling shows artemisinin, HC102A and HC103A inhibit the core genes of the DosRST regulon during hypoxia. (a) Mtb differential gene expression in response to artemisinin. Genes in red have a p-value <0.05. Indicated gene names include characterized DosR regulated genes. (b) Venn diagram showing genes that are downregulated (>2-fold, p<0.05) in CDC1551 treated with artemisinin, HC102A or HC103A relative to a DMSO treated CDC1551 control. Also shown are genes downregulated (>2-fold, p<0.05) in a DMSO treated CDC1551(DdosR) mutant strain relative to a DMSO treated CDC1551 control. (c) Venn diagram showing genes that are downregulated (>2-fold, p<0.05) in a CDC1551 DdosR) mutant strain treated with artemisinin, HC102A or HC103A relative to a DMSO treated CDC1551(DdosR) control. The limited genes modulated by HC102A and HC103A confirm that these compounds are highly specific for the DosR regulon. 45 Artemisinin, HC102A and HC103A disrupt persistence DosRST is required for several persistence-associated physiologies during hypoxia, including triacylglycerol (TAG) synthesis (3, 27) and survival (13). We hypothesized that artemisinin, HC102A or HC103A may target these physiologies and compared the activity of the inhibitors to a CDC1551(DdosR) mutant. Transcriptional profiling data in CDC1551 showed that the tgs1 gene, which encodes for the TAG synthase involved in last step of TAG synthesis, is downregulated ~100 fold by artemisinin and ~20 and ~180 fold by HC102A and HC103A, respectively (Fig. 2.2 and Supplementary Fig. 2.3a, Supplementary Dataset 2.1). Therefore, we hypothesized that CDC1551 treated with artemisinin, HC102A or HC103A would be defective in TAG accumulation. To test this hypothesis, 14 C-labeled lipids were isolated from CDC1551(DdosR) mutant and WT CDC1551 treated with inhibitors or an equal volume of DMSO. The lipids were analyzed by thin layer chromatography and quantified. TAG accumulated in the DMSO-treated cells, whereas it was reduced 82% in the CDC1551(ΔdosR) mutant (Fig. 2.3a, Supplementary Fig. 2.4a). Treatment with artemisinin, HC102A and HC103A caused a 74%, 67% and 56% reduction in TAG accumulation, respectively (Fig. 2.3a, Supplementary Fig. 2.4a), thus providing functional evidence that the inhibitors are impacting persistence-associated lipid metabolism. During NRP, the ΔdosR mutant has previously been shown to exhibit reduced intracellular survival as compared to WT bacteria (13). Using the hypoxic shift down model of NRP (27), we examined the impact of DHA, HC102A or HC103A on survival during NRP. Following 10 d of incubation in the hypoxic shift down assay, CDC1551 treated with 40 µM DHA, HC102A or HC103A, exhibited significantly reduced survival (70-80% reduction) as compared to the DMSO control (Fig. 2.3b). This reduction in survival is comparable to the CDC1551(DdosR) mutant relative to the DMSO treated WT control. The survival defect of the CDC1551(DdosR) mutant 46 was partially complemented in the CDC1551(DdosR) complemented strain, indicating that the observed survival defect is dosR-dependent. We also examined the impact of the inhibitors in the Mtb Erdman strain, to ensure the observed physiologies are not unique to the CDC1551 strain. Although minor differences between CDC1551 and Erdman existed, we observed that the inhibitors also significantly inhibited survival during NRP in the Erdman strain (Fig. 2.3b). The function of the inhibitors was examined in the hypoxic shift down assay in an 8 point dose response covering 1 µM -100 µM and percent viability relative to the DMSO treated control was examined following 10 d and 15 d treatment. Artemisinin, HC102A and HC103A exhibited dose dependent inhibition of viability in the hypoxic shift down model with all three compounds causing an ~50% reduction of viability at 10 µM following 10 or 15 d of incubation (Fig. 2.3c and Supplementary Fig. 2.4b). The DosR regulated gene tgs1 has previously been shown to be required for Mtb tolerance to antibiotics during hypoxia (17, 28). Because the inhibitors strongly inhibit tgs1 gene expression, we hypothesized these compounds may sensitize Mtb to isoniazid (INH). Using the hypoxic shift down assay, Mtb Erdman was pretreated with either 20 µM or 40 µM of artemisinin, HC102A or HC103A and following 2 d, the cells were treated again with the experimental inhibitors (for a combined treatment of 40 µM or 80 µM) in addition to INH over a dose response (1 µM, 5 µM, 25 µM INH or a DMSO control). Following 10 and 15 d of INH treatment surviving bacteria were enumerated (Fig. 2.3d and Supplementary Fig. 2.4c). At day 10, treatment with artemisinin, HC102A or HC103A alone at 40 µM or 80 µM causes a significant reduction of survival ranging from a 3-fold to 30-fold decrease in surviving bacteria (Fig. 2.3d). Ten days post INH treatment, INH alone had minimal impact on Mtb survival, with ~100% viability at 1 µM and 5 µM INH and ~75% viability at 25 µM INH (Fig. 2.3e), confirming that Mtb is tolerant to INH in the hypoxic 47 shift down assay. Treatment with 40 µM artemisinin, HC102A or HC103A caused a significant ~30% reduction of Mtb viability in the presence of 5 µM INH (Fig. 2.3e) relative to cultures not treated with INH, suggesting that the inhibitors inhibit INH tolerance. At day 15, similar trends were observed in reduction of Mtb survival and antibiotic tolerance in cultures treated with artemisinin, HC102A or HC103A (Supplementary Fig. 2.4cd). Notably, at day 15, treatment with 40 µM artemisinin or HC103A caused a ~50% reduction of Mtb viability when treated with 5 µM INH (Supplementary Fig. 2.4d), suggesting that the function of artemisinin and HC103A may be enhanced during longer periods of NRP. 48 Figure 2.3. Artemisinin, HC102A and HC103A inhibit TAG synthesis, survival and isoniazid tolerance. (a) CDC1551 treated with compounds of interest (at 40 µM) exhibits a 60-70% reduction of TAG accumulation, similar to the CDC1551(DdosR) mutant control. The position of TAG is indicated by the asterisk. (b) CDC1551 and Erdman strains treated with 40 µM of the compounds of interest exhibit reduced survival following 10 d in the hypoxic shift down assay. CFUs were counted and percent survival calculated relative to the WT control at day 0. Differences between treated samples as compared to the DMSO control are significant (P-value <0.05 based on a t-test, marked with an asterisk). (c) Dose dependent inhibition of Mtb survival in the hypoxic shift down assay. Percent viability was calculated relative to viable bacteria in the DMSO control at day 10. (d) Following 10 d in the hypoxic shift down assay, treatment with artemisinin, HC102A or HC103A, with or without INH, significantly reduces bacterial survival relative to the respective DMSO controls (P-value <0.05 based on a t-test). (e) Artemisinin, HC102A and HC103A reduce INH tolerance in the hypoxic shift down assay. Percent viability at 1, 5 and 25 µM INH was calculated relative to the 0 µM INH control (using data in panel d). Significant inhibition of INH tolerance (marked with an asterisk, P-value <0.05 based on a t-test) is observed relative to the respective DMSO controls. For each panel, error bars represent the standard deviation and experiments were repeated at least twice with similar results. 49 Artemisinin targets DosS and DosT sensor kinase heme Artemisinin is a first-line drug for treating malaria (29) and the mechanism of action has been extensively studied (29, 30). Evidence suggests that reductive cleavage of the artemisinin endoperoxide bridge is initiated by ferrous iron (Fe2+) under reduced conditions, and generates a C4-centered radical. The radical form of artemisinin can alkylate heme and results in artemisininheme adduct formation (30, 31). Because DosS/T are also heme-containing proteins, we hypothesized that artemisinin interacts similarly with the heme in DosS/T leading to artemisininmediated inhibition of the DosRST regulon. Biochemical data suggest DosS is a redox sensor that autoxidizes quickly under aerobic conditions (32), whereas, DosT is a hypoxia sensor and has high affinity and sensitivity to O2 (33). Both kinases sense environmental cues via heme, and are inactive when the heme group exists as either the Met (Fe3+) form (in the case of DosS) or the oxy (Fe2+-O2) form (in the case of DosT) in the presence of O2; the kinases are activated when DosS is in the ferrous form and DosT is in the deoxy form (11, 32, 34-36). A UV-visible spectroscopy assay was employed to determine the interaction between DosS/T and artemisinin (11). DosS and DosT, purified under aerobic conditions, have Soret peaks at 409 nm and 412 nm, respectively (11) (Fig. 2.4ab). Reduction of the heme by dithionite (DTN) shifts the DosS/T Soret peaks to 430 nm. Treatment of DosS with artemisinin (purged of O2) causes the Soret peak of DosS to gradually shift back to the original oxidized Soret peak. This indicates that artemisinin can function to modulate DosS redox status. Notably, treatment of DosT with artemisinin reduces the amplitude of the Soret peak, a response that has previously been shown to be associated with artemisinin-mediated degradation of heme (37, 38). The position of the peaks did not shift to the oxidized state in the DMSO treated proteins. Dose-response studies further show that artemisinin inhibits DosT at 50 µM (Fig. 2.5d), whereas artemisinin only causes the shift 50 of the reduced Soret peak of DosS at a much higher concentration of 400 µM (Supplementary Fig. 2.5a). This finding suggests that DosT is more sensitive to artemisinin than DosS and may explain why the artemisinins had weaker activity for the inhibition of the DosR pathway when stimulated by NO or vitamin C. Because the DosS/T kinases are active in the reduced form and inactive in the oxidized form (11), these data are consistent with artemisinin inhibiting DosS/T kinases by modulating their redox status (e.g. DosS at high concentrations) or causing degradation of the heme (e.g. DosT). To test the hypothesis that artemisinin can alkylate heme carried by the sensor kinase, DosS reaction samples treated with artemisinin or DMSO were subjected to LC-MS analysis. Molecules with masses of ~898 Da were identified in the artemisinin treated sample that are absent in the DMSO treated sample (Fig. 2.4c). This molecular weight corresponds to the sum of the masses of artemisinin (282 Da) and heme (616 Da), supporting the hypothesis that artemisinin alkylates the sensor kinase heme to form heme-artemisinin adducts (39). In a previous study, a structure of the artemisinin-alkylated heme with a mass of 898.3 was shown to involve alkylation of the heme at the a, b or d positions with the iron in the ferric state (39). Notably, adducts of DosS peptides were not observed. Together, these findings support a mechanism of action where artemisinin directly targets the heme to inactivate the DosS and DosT sensor kinases. The UV-visible spectra and mass spectrometry data support differing mechanisms for DosS and DosT inactivation by artemisinin, with DosS alkylated-heme remaining intact but in the ferric state (39) and the DosT heme being degraded by artemisinin (37, 38). In both cases, these hemeartemisinin interactions would result in disabling the sensor functions of DosS/T. Molecular modeling of DosS and DosT structures (34, 35) shows that the kinases have a channel through which the artemisinin may access and dock to the heme (Fig. 2.5a). To test this model, amino acid substitutions were generated along the channel in DosS/T, including E87L and 51 G117L in DosS or G85L and G115L in DosT, that are predicted to limit the ability of artemisinin to access the heme based on modeling conducted in this study and published studies (33). In the UV-visible spectroscopy assay, DosS (E87L) and DosS (G117L) exhibited similar overall spectra as WT DosS under aerobic conditions and treatment with DTN caused the Soret peak to shift to the reduced position (Fig. 2.5bc); thus, the heme in both mutant proteins retains the ability to respond to reduction by DTN. Notably, the DosS (E87L) and DosS (G117L) proteins were resistant to oxidation by 400 µM artemisinin and the major Soret peak (430 nm) did not shift to the oxidized position following 60 min of treatment (Fig. 2.5bc). The analogous mutations of DosT, G85L and G115L, also exhibited similar overall spectrum as WT DosT under aerobic conditions, as well as in responding to DTN treatment (Fig. 2.5de and Supplementary Fig. 2.5b). However, the lower peak at 560 nm of deoxy-DosT (G115L) was maintained when treated with 100 µM artemisinin as compared to WT DosT and DosT (G85L) (Fig. 2.5de and Supplementary Fig. 2.5b). The 560 nm peak is generated by merging two lower peaks at 538 nm and 575 nm together after DTN treatment (40). This peak is highly sensitive to oxygen or artemisinin treatment and disappears immediately upon exposure to oxygen or artemisinin (11, 37, 38). Additionally, the Soret peak in the G115L mutant is not reduced in a dose-dependent manner, as compared to the WT and G85L mutants (Fig 2.5de and Supplementary Fig. 2.5b), further supporting the DosT (G115L) protein exhibits artemisinin resistance. The UV-visible spectroscopy data collectively support the notion that DosS E87L and G117L and DosT G115L substitutions may limit artemisinin from fully accessing the heme, thereby providing resistance to artemisinin. To test this hypothesis in Mtb, we transformed CDC1551 with a plasmid overexpressing the WT dosT, dosT (G85L) or dosT (G115L) genes and determined if artemisinin resistance is observed. The strains were grown in standing flasks to stimulate the 52 DosRST regulon and expression of DosRST regulon genes (dosR, hspX, tgs1) was examined by real-time PCR following 6 d of treatment with artemisinin over a dose response curve (Supplementary Fig. 2.6). Strains expressing WT dosT or dosT (G85L) exhibited EC50 for artemisinin-mediated inhibition of DosR regulon genes of 0.2-0.3 µM whereas, dosT (G115L) exhibited EC50 of 1.0-1.6 µM. (Fig. 2.5f and Supplementary Fig. 2.6). Therefore, the DosT (G115L) protein provides ~5 fold resistance to artemisinin and nearly full resistance at 1 µM artemisinin (Fig. 2.5f). These biochemical and biological data confirm that artemisinin modulates the DosRST signaling by directly targeting the heme sensor carried by DosS and DosT histidine kinases. 53 Figure 2.4. Artemisinin directly inhibits DosS and DosT by targeting sensor kinase heme. UV-visible spectra of DosS (panel a) and DosT (panel b) showing treatment with dithionite (DTN) reduces the heme (the “on” state for the kinases) and that artemisinin oxidizes or degrades the heme (“off” states of the kinases). (c) MS spectra showing the presence of peaks at ~898 Da that are present in artemisinin treated DosS samples, but absent in DMSO treated samples. This mass is the approximate combined mass of heme (616.487 Da) and artemisinin (282.332 Da) and is consistent with the formation of heme-artemisinin adducts. Experiments were repeated at least twice with similar results. 54 Figure 2.5. Amino acid substitutions in DosS or DosT promote resistance to artemisinin. (a) Molecular modeling indicates a channel exists in DosS and DosT through which artemisinin may access the heme. In WT DosT, the heme (colored yellow) and iron (green ball) are accessible to artemisinin via a channel. G85L and G115L substitutions are predicted to block this channel and access to the heme. (b and c) UV visible spectra show that DosS (G87L) and DosS (G117L) proteins can be reduced by the addition of dithionite (DTN) but are resistant to oxidation by artemisinin (Art). (d and e) WT DosT exhibits a dose-dependent decrease in the amplitude of the Soret peak at 430 nm (left side of panel d) and a loss of the peak at 560 nm (magnified in right side of panel d). In contrast, DosT (G115L) exhibits resistance to artemisinin because it does not exhibit a dose-dependent decrease in the 430 nm peak (left side of panel e) and the 560 nm peak is maintained at treatments of 50 and 100 µM artemisinin (right side of panel e). DosT (G85L) exhibits spectra similar to WT DosT (Supplementary Fig. 2.5B). (f) Overexpression of DosT (G115L) in CDC1551 provides artemisinin resistance. Mtb overexpressing WT DosT or DosT (G85L) exhibit strongly inhibited expression of DosR regulated genes (dosR, hspX and tgs1) in the presence of 1 µM artemisinin, while the strain overexpressing DosT (G115L) is resistant. Dose response curves are presented in Supplementary Fig. 2.6. Experiments were repeated at least twice with similar results. 55 HC103A inhibits DosS and DosT autophosphorylation UV-visible spectroscopy studies showed that HC102A and HC103A have no impact on the redox status of DosS or DosT heme (Supplementary Fig. 2.7ab), suggesting these compounds function by a mechanism that is distinct from artemisinin. Given the strong specificity with which these compounds inhibit the DosRST regulon, we hypothesized that HC102A and HC103A may directly inhibit DosS/T autophosphorylation activity. To test this hypothesis, in vitro phosphorylation assays were performed as previously described (41). DosS and DosT were quickly phosphorylated within 30 s of initiating the assay by adding [γ-32P] ATP. The amount of phosphorylated protein increased over time consistent with previous reports (41). DosS treated with HC102A and HC103A showed decreased autophosphorylation activity (Fig. 2.6, Supplementary Fig. 2.8). Inhibition of DosS autophosphorylation activity increased in a dose-dependent manner with IC50s of 1.9 µM and 0.5 µM for HC102A and HC103A, respectively. Notably, differences in maximal inhibition were observed, with ~60% and ~90% for HC102A and HC103A, respectively. DosT treated with HC103A also showed a dose-dependent inhibition of autophosphorylation and an IC50 of ~5 µM (Fig. 2.6). These findings support the proposal that HC103A functions by directly modulating DosS and DosT kinase activity. 56 Figure 2.6. HC103A inhibits DosS and DosT autophosphorylation. Recombinant DosS or DosT was treated with HC103A (a) or HC102A (b) across a dose response curve. The autophosphorylation assay was incubated for 1 h, the proteins were blotted and the protein autophosphorylation was assessed by following exposure of the blot to a phosphor screen. HC102A and HC103A inhibit DosS autophosphorylation with IC50 of 1.9 µM and 0.5 µM, respectively. HC102A had limited impact on DosT autophosphorylation, whereas HC103A inhibited DosT autophosphorylation with an IC50 of ~5 µM. Experiments were repeated at least twice with similar results. 57 Discussion In this study a CDC1551(hspX’::GFP) fluorescent reporter strain was used as a synthetic phenotype for the targeted discovery of several compounds that inhibit the DosRST pathway. Treatment of Mtb with these compounds copies several phenotypes of a CDC1551(DdosR) mutant, including: downregulation of the core DosRST regulon, reduced TAG synthesis, and decreased survival during NRP. These findings provide proof-of-concept data that the high-throughput screen successfully identified inhibitors of the DosRST regulon and support further studies characterizing additional putative DosRST regulon inhibitors, including HC104-HC106, as well as uncharacterized putative hits from the primary screen. Several new chemical inhibitors of Mtb during NRP have been recently described. One promising target is the direct inhibition of ATP homeostasis by targeting the components of the electron transport chain or ATP synthase (27, 4244). However, the approach of targeting the DosRST top-level regulators has the potential to inhibit multiple physiologies required for establishing or maintaining NRP and this multifactorial approach may broadly limit persistence in heterogeneous NRP-inducing environments. Using homology modeling, the discovery of a DosR regulon inhibitor that specifically inhibits DosR binding to target DNA (45) has been reported. Additionally, a screen for small molecules that modulate Esx-1 export, identified inhibitors that indirectly modulate the DosR regulon (46). However, to the best of our best knowledge, the newly discovered compounds represent novel inhibitors of the DosRST regulon with distinct mechanisms of action. Both HC102A and HC103A appear to be remarkably specific for inhibiting the DosRST pathway. Indeed, no genes were downregulated (>2-fold, p<0.05) in the CDC1551(DdosR) mutant treated with HC102A, supporting the idea that only DosR controlled pathways are targeted by HC102A. HC103A treatment of the CDC1551(DdosR) caused downregulation (>2-fold, p<0.05) 58 of 13 genes. Several genes fell into related classes including: four arginine biosynthesis genes (argC, argB, argJ and argF), two acyl-(ACP) desaturases (desA1 and desA2), two ferredoxins (fdxC and fprB), three PE-PPE genes (Rv0160c, Rv1386 and Rv1387), and an orphan response regulator (Rv0260c). This finding suggests that although HC103A is highly specific for the DosRST pathway other targets likely exist. In contrast to HC102A and HC103A, artemisinin exhibits significant off-target activities, with 69 genes differentially regulated in the CDC1551(DdosR) mutant treated with this compound. Given that artemisinin has a reactive endoperoxide bridge, it is perhaps surprising that there are not a greater number of genes that are differentially regulated, as one might expect artemisinin to react with any proteins carrying reduced iron. The downregulated genes do not match genes regulated by ROS (47), suggesting that artemisinin is not promoting Fenton reactions and acting as an indiscriminate oxidant. Notably, it has been shown that when artemisinin is delivered into Mtb as a mycobactin−artemisinin conjugate it causes an intracellular burst of reactive oxygen via Fenton reactions that kills Mtb (48). Based on this finding, it is tempting to speculate that artemisinin alone cannot fully access the Mtb cytoplasm and is thus modulating membraneassociated proteins, such as DosS/T. Other membrane proteins that may interact with artemisinin (e.g. heme-bearing cytochromes) may be resistant to artemisinin if the heme is buried in the protein and not accessible to artemisinin. Structural analysis of the heme-embedded GAF domain of DosST shows that DosT has a wider channel than DosS that could potentially provide greater accessibility to artemisinin (34, 35). This suggestion is supported by our data that DosT is more sensitive to artemisinin than DosS and is consistent with artemisinins having limited ability to inhibit NO-mediated stimulation of the DosR regulon. Alternatively, DosS may be less sensitive to artemisinin than DosT due to differences in autooxidation rates, where DosS is more quickly 59 oxidized to the ferric state in the presence of oxygen (49). Autooxidized DosS in the ferric form would not react with artemisinin, resulting in the observed insensitivity. Collectively, we have shown this channel is susceptible to drugs with artemisinin as a proof-of-concept, therefore structure-based synthesis of compounds that block the heme-bearing channel of DosS and DosT may promote the development of additional novel inhibitors of Mtb persistence. The discovery that artemisinin inhibits Mtb persistence raises interesting questions about the use of artemisinin to treat malaria in individuals co-infected with malaria and Mtb, however, further studies will be required to determine the clinical significance of artemisinin-Mtb interactions. Accession Codes: The transcriptional profiling data have been submitted to the NCBI GEO database (accession no. GSE76566). Materials and methods Bacterial strains and growth conditions. Mtb strains CDC1551 and Erdman were used as indicated. CDC1551(ΔdosR) and Erdman(ΔdosR) mutants were constructed using gene replacement by homologous recombination (50) using methods as previously described (51). Deletions were confirmed by PCR and transcriptional profiling. All strains were cultured at 37oC and 5% CO2 in standing, vented tissue culture flasks in 7H9 Middlebrook medium supplemented with 10% OADC (oleic acid albumin dextrose catalase) and 0.05% Tween-80. For dosT overexpression strains, the dosT gene was cloned under the control of the strong hsp60 promoter in the pVV16 vector and transformed into CDC1551. dosT (G85L) and and dosT(G115L) mutants 60 were generated using the QuickChange site directed mutagenesis approach (Agilent) in pVV16 and confirmed by sequencing. High-throughput screening assay and data analysis. The HTS was conducted against two compound collections, the 211,655 compound ICCB-Longwood collection and the 328,633 compound NIH Molecular Libraries Program (MLP) collection, both provided by the ICCB at Harvard Medical School. The compounds were arrayed in 384-well clear bottom, black sided microtiter plates (Corning) at a final screening concentration of ~ 10 µM. Two columns of each plate were left blank for positive and negative controls of 0.3 µM rifampicin and DMSO alone, respectively. The M. tuberculosis CDC1551 (hspX’::GFP) fluorescent reporter was grown to mid- to late-exponential phase in vented T-150 standing flasks in Middlebrook 7H9 (OADC) medium (buffered to pH 7.0 with 100 mM MOPS (3-morpholinopropanesulfonic acid)). The cultures were then re-suspended in 7H9 (OADC) pH 7.0 medium and dispensed into the 384well assay plates utilizing a Cy-Bio Selma liquid handler robot to an OD595 of 0.05. The plates were then placed in a Ziploc bag with a moistened paper towel (to limit evaporation) and incubated for 6 d at 37ºC. Fluorescence and optical density (OD) readings were made on an EnSpire plate reader (Perkin Elmer, Inc.) at excitation and emission wavelengths of 488 and 509 nm as a top read, with the OD being taken at 595 nm as a bottom read. Data analysis was performed utilizing an in-house developed computational tool written in Python. Raw fluorescence and optical density measurements were exported from the EnSpire plate reader (Perkin Elmer, Inc.) in plate format as comma-separated files. Measurements were then normalized as a function of percent inhibition compared to the negative (DMSO) control (see equation below). 61 The normalized percent inhibition (NPI) for fluorescence or optical density was calculated by subtracting the overall mean of the negative controls within the run (𝜇"# ) from the measured value (β), divided by the dynamic range and multiplied by 100. The overall means for the positive and negative controls within the run (𝜇"# , 𝜇%# ) were determined if there existed at least one plate in the run with a Z’ greater than or equal to 0.5(52). ijk represents the ith value in the jth plate within the kth run. σ represents the standard deviation. Potential inhibitors of the DosRST regulon were defined as compounds with greater than 35% fluorescence inhibition, limited growth inhibition, and at least 1.5-fold selectivity in the fluorescence to growth inhibition ratio. To determine the statistical significance of the 1.5 fold selectivity cutoff, Z-scores were calculated for each experimental compound fluorescence inhibition:growth inhibition ratio relative to the negative controls and P-values were derived by testing against the null distribution. Due to the high number of tests, each P-value was false-discovery rate corrected (Supplementary Fig. 2.1c). These “class 1” compounds may be directly or indirectly inhibiting DosRST signaling. The Z-factors of the screens were 0.90 and 0.89 for the ICCB-L and MLP library screens, respectively (52). For GFP quenching assays, the CDC1551 (hspX’::GFP) reporter was grown under GFPinducing conditions, aliquoted into 96 well plates, treated with a dose response of HC101AHC106A and then the plates were immediately read for GFP fluorescence. GFP quenchers cause an inhibition of GFP fluorescence and none of the compounds exhibited GFP quenching activity. Cytotoxicity assays were conducted against three eukaryotic cells, primary C57Bl/6 murine derived macrophages (BMDMs), THP-1 and J774 cells. Macrophages were prepared as previously described (53) and seeded in white, opaque, 96 well plates (Corning) and treated for 3 62 d with the compounds treated with a 8-point dose response curve ranging from 400 µM to 0.65 µM. Following 3 d, viability was determined using the CellTiter-glo luminescent cell viability assay (Promega). Percent inhibition was normalized to a Triton X-100 positive control and a DMSO negative control. EC50s were calculated using the GraphPad Prism software package (version 6). Each experiment included two technical replicates per plate and two biological replicates and error bars represent the standard deviation of the biological replicates. The experiment was repeated at least twice. EC50 determinations for HC101A-HC106 compounds were performed in clear bottom, black, 96 well plates (Corning), following methods similar to those described above for the HTS. Briefly, 200 µL of the CDC1551(hspX’::GFP) reporter was inoculated into each well at an initial OD of 0.05. The cells were treated for 6 d with compounds using an 8-point dose response curve ranging from 400 µM to 0.65 µM. The plates were then read for GFP fluorescence and optical density and percent inhibition was normalized to a rifampin positive control and DMSO negative control. EC50s were calculated using the GraphPad Prism software package (version 6). Each experiment included two technical replicates per plate and two biological replicates and error bars represent the standard deviation of the biological replicates. The experiment was repeated at least twice. Transcriptional profiling and data analysis. CDC1551 or CDC1551(ΔdosR) cultures were treated with 40 µM artemisinin, HC102A, HC103A or DMSO (as a negative control) and grown at 37°C without shaking in T-25 vented, standing tissue culture flasks in 8 mL of 7H9 medium seeded at an initial OD of 0.1. The experiments were performed with two biological replicates. Following 6 d of incubation, total bacterial RNA was extracted and sequenced as described by Baker, Johnson 63 and Abramovitch (2). RNA-seq data were analyzed using the SPARTA software package (25). Real-time PCR assays. For the NO and vitamin C sensitivity assays, CDC1551 was seeded at an initial density of 0.6 OD and treated with 80 µM DHA, HC102A, or HC103A for 24 h, and then induced with 50 µM DETA-NONOate or 2 mM vitamin C for 2 h. After treatment, total bacterial RNA was extracted as previously described (54). Transcripts of representative genes from the dosR regulon, including dosR, hspX, and tgs1, were quantified by RT-PCR using gene-specific primers as previously described (51). The experiment included three biological replicates and error bars represent the standard deviation from the mean. The experiment was repeated twice with similar results. For the artemisinin resistance assays, CDC1551 was seeded at an initial density of 0.1 and treated with 0.025 µM - 20 µM artemisinin for 6 d at 37 °C. Total RNA was extracted and RT-PCR quantification of DosR-regulated genes (dosR, hspX and tgs1) was conducted as described above. The experiment was repeated with three biological replicates with similar results. EC50s were calculated using the GraphPad Prism software package (version 6). Triacylglycerol accumulation analysis. CDC1551 cultures were seeded at a density of 0.1 OD in 8 mL of 7H9 medium and treated with 40 µM of artemisinin, HC102A, HC103A or DMSO. The cultures were radiolabeled by addition of 80 µCi of [1,2-14C] sodium acetate to the culture, which was then grown at 37°C in vented, standing, T-25 tissue culture flasks. Total lipid was extracted after 6 d incubation and analyzed in thin-layer chromatography (TLC), as previously described (51). Total extractable lipid 14C incorporation was determined using a scintillation counter, and 20,000 cpm was loaded for analysis in a 100-cm2 high-performance TLC silica gel 60 aluminum sheet. To analyze TAG, the lipids were resolved in hexane-diethyl ether-acetic acid (80:20:1 64 [vol/vol/vol]) solvent system. The TLC was exposed to a phosphor screen for 3 d, imaged on a Typhoon imager and quantified by ImageJ (55). The experiment was repeated with two biological replicates with similar results. NRP survival and antibiotic tolerance assays. The hypoxic shift down assay was used as a model for NRP and performed as previously described (27). CDC1551 or Erdman cultures were pelleted and resuspended in Dubos medium at OD600 of 0.25, and inoculated in 24-well plates (1 mL/well). In the experiments presented in Figure 2.3b, cells were treated with 40 µM artemisinin, HC102A or HC103A or equal volume of DMSO, and incubated in an anaerobic chamber (BD GasPakTM) for 12 d. Cultures become anaerobic within 48 h incubation as indicated by methylene blue turning to colorless, and consequently day 0 is considered after 48 h of incubation. Bacteria were plated on solid medium to enumerate CFUs at day 0 and day 10. Percent viability was determined by comparing surviving bacteria at day 10 relative to day 0. Experimental conditions were examined with three biological replicates and error bars represent the standard deviation from the mean. The dose response experiments (Fig. 2.3c and Supplementary Fig. 2.4b) were performed as described above with Mtb CDC1551 using an 8-point dose response covering 1-100 µM and a DMSO control. CFUs were enumerated at day 10 and day 15 and percent viability was determined relative to the DMSO control at day 10 or day 15. The INH tolerance assays (Fig. 2.3d and Supplementary Fig. 2.4c) were performed as described above with the following modifications. Mtb Erdman was pretreated with 20 or 40 µM artemisinin, HC102A or HC103A for 2 d in the hypoxic shift down assay and then the anaerobic chamber was opened and the cells were treated again with 20 or 40 µM artemisinin, HC102A or HC103A (for a total treatment of 40 or 80 µM). The cells were also treated with 1, 5 or 25 µM INH or a DMSO control. The cells were incubated in the anaerobic 65 chamber for 10 or 15 d and CFUs were enumerated by plating on solid medium. To quantify INH tolerance, percent viability at 1, 5 and 25 µM INH was measured relative to the 0 µM INH control (DMSO control). These experiments were repeated at least twice with similar results. DosS and DosT protein purification. The dosS (Rv3132c) and dosT (Rv2027c) genes were amplified from Mtb genomic DNA by PCR and cloned into the expression vector pET15b (Novagen Darmstadt, Germany). The DosS E87L and G117L substitutions and DosT G85L and G115L mutants were generated using the QuickChange site directed mutagenesis approach (Agilent) and confirmed by sequencing. The resulting constructs were confirmed by DNA sequencing. DosS/T protein expression in E. coli BL21(DE3) and purification via Co2+ column were performed as previously described (34). Briefly, the His6-DosS or His6-DosT were expressed in E. coli BL21(DE3) supplemented with hemin (30 mg/L) and induced by isopropyl 1-thio-β-D-galactopyranoside (IPTG, 1 mM) at 18°C for 20 h. Cell pellet was suspended in lysis buffer (50 mM sodium phosphate (pH 7.6), 10% glycerol, 200 mM sodium chloride, 1% Triton X-100, 0.5 mg/mL lysozyme, 0.1 mg/mL phenylmethylsulfonyl fluoride). The cell suspension was incubated with shaking at 37°C for 0.5 h and then sonicated. Soluble extract was applied to a Co2+ column (Clontech) and washed with washing buffers (with or without 20 mM imidazole in 50 mM sodium phosphate (pH 7.6), 10% glycerol, 500 mM sodium chloride). The recombinant proteins were eluted with 200 mM imidazole in the same buffer. The fractions containing the purified proteins were pooled together and dialyzed in 20 mM Tris-HCl, pH 7.5. 66 UV-visible spectroscopy assay and mass spectrometry. The absorption spectra of DosS (7.5 µM) and DosT (7.5 µM or 16.9 µM) were analyzed in a stoppered quartz cuvette by a DU800 spectrophotometer (Beckman Coulter). All reagents were degassed with argon in a sealed cuvette or vial prior to use. Proteins were also degassed, and then treated with 400 µM DTN. The UVvisible spectra were recorded before and after DTN treatment. Lastly, different concentration of artemisinin or an equal volume of DMSO was added to the reaction. The kinetic changes in the absorption spectra were recorded for 2 h. For mass spectroscopy (MS) analysis, the proteins were subjected to pepsin digestion at 37°C for 30 min after the assay, and then analyzed by liquid chromatography MS (LC-MS). Sample analysis was carried out on Waters Xevo G2-XS QTof mass spectrometer (Milford, MA, USA) with an electrospray ionization positive mode. The parameters were: capillary voltage, 3 kV; sampling cone, 40 V; source temperature, 100 °C; desolvation temperature, 350 °C; cone gas flow, 25 L/h; desolvation gas flow, 600 L/h. Chromatographic separation was done in Waters ultra-performance liquid chromatograph (ACQUITY UPLC) system equipped with a Waters BEH C18 column (1.7 µM, 100x2.1 mm). The column temperature was kept at 30 °C. Solvents were (A) 0.1% (vol) folic acid in water, and (B) acetonitrile. The flow rate was 0.2 mL/min with following gradient: A/B = 99/1 to A/B = 70/30 in 8 m, then A/B = 1/99 for 10 min, and A/B = 99/1 for last 2 min. The acquisition mass range was 200-2,000 Da. The retention time for heme and heme-artemisinin adducts was between 10-11 min. The experiment was repeated with at least two biological replicates with similar results. DosS and DosT autophosphorylation assay. The in vitro phosphorylation assays were performed as previously described (41). Briefly, 4 µL reaction contained 0.2 µg/µL purified DosS or DosT 67 protein, 100 mM Tris-HCl, pH 8.0, 50 µM KCl2, 5 µM MgCl2 and 2.5 µCi/µL [γ-32P] ATP (6000 Ci/mmol, PerkinElmer Life Science). The proteins were treated with a 9 or 10 point dose response curve of HC102A or HC103A and the reaction was incubated at RT for 1 h. All aliquots were analyzed in 4-12% gradient SDS-PAGE (Bio-Rad), and blotted to PVDF membrane. The blot was exposed to phosphor screen overnight and quantified by ImageJ (55). The experiment was repeated with three biological replicates with similar results. Structural modeling of DosT and the G85L and G115L substituted proteins. Modeling of DosT was performed using the Molecular Operating Environment (MOE) software (Montreal, Canada). The structure for DosT (2VZW) was downloaded from the RCSB protein data bank. For DosT images, all of Chain B and its associated water molecules were deleted. For Chain A, all water molecules and the acetic acid were deleted. Heavy atoms were fixed and hydrogen atoms relaxed with energy minimization to a gradient of 0.001. Good parameters do not exist for the heme group. Sets were defined for the protein, heme, iron, and oxygen (O2 ligand). Carbon atoms making up the heme were colored yellow. The iron atom of the heme was colored green as a large sphere. The oxygen atoms of the oxygen group were hidden from view in all of the remaining MOE files and pictures. Residues 85 and 115 were labeled and colored purple with heavy bonds. A “Molecular Surface” was created on the protein only. The surface was colored by electrostatics using Poisson-Boltzmann to compute the electrostatic field. The iron atom of the heme can be seen down the narrow gorge. Chemical synthesis of HC102A (CCG-232500) and HC103A (CCG-257424). Please refer to Supplementary Notes. 68 Acknowledgements High-throughput screening libraries and their preparation was supported by the New England Regional Center of Excellence (U54 AI057159) and the Institute of Chemistry and Cell Biology (ICCB) at Harvard Medical School. Mark Farrugia and Bob Hausinger provided support with the UV-visible spectroscopy experiments. The MSU RTSF provided technical support for the RNAseq library preparation and sequencing. The Vahlteich Medicinal Chemistry Core is grateful for ongoing support from the Ella and Hans Vahlteich Fund and Beverly Vahlteich Delaney. We thank members of the Abramovitch lab for critical reading of the manuscript. This project was supported by start-up funding from Michigan State University and AgBioResearch, a grant from the NIHNIAID (R21AI105867), and Grand Challenges Explorations awards (OPP1059227 and OPP1119065) from the Bill & Melinda Gates Foundation. 69 APPENDIX 70 APPENDIX Supplementary Information Inhibitors of Mycobacterium tuberculosis DosRST signaling and persistence Supplementary Datasets Supplementary Dataset 2.1. Differential gene expression data of WT Mtb treated with inhibitors and the DMSO treated DosR mutant Supplementary Dataset 2.2. Differential gene expression data of the DosR mutant treated with the inhibitors. Supplementary Dataset 2.3. Complete gene expression tables for transcriptional profiling experiments. 71 Supplementary Table 2.1. EC50 calculation of GFP flourescence inhition and eukaryotic cytotoxicity by DosR regulon inhibitors Name Structure GFP FI (EC50 , μM) Eukaryotic cytotoxicity (EC50 , μM) CH3 BMDM THP-1 J774 H O H3C O O H H O CH3 O HC101A (Artemisinin) CH3 H O H3C 1.2 >400 >400 >400 1.9 160.0 33.9 0.7 3.7 100 18.8 0.4 12.4 >400um >400 >400 2.7 >160 112 152 5.0 179.8 91.1 204.6 2.8 72.7 16.3 84.8 12.7 235.8 90.5 57.2 6.9 86 86 90 O O H H O O CH3 O OH O HC101B (Artesunate) CH3 H O H3C O O H H O CH3 OH HC101C (Dihydroartemisinin) H3C H N H3C O NH H3C O HC102A O H N HO S N H O HC103A O HO OH O N H N H HC103B CH3 N O N CH3 O Br HC104A H3C N O N N S CH3 O O HO OH O HC105A HN HN O N O Cl Cl HC106A Supplementary Table 2.1. GFP fluorescence inhibition and eukaryotic cytotoxicity by DosR regulon inhibitors. For reporter fluorescence EC50 determination, CDC1551(hspX’::GFP) 72 Supplementary Table 2.1. (cont’d) reporter was treated with compounds over an 8 point dose response curve ranging from 400 µM to 0.65 µM. Following 6 d of treatment, GFP fluorescence and optical density were measured. In all cases, no inhibition of growth was observed at the reported EC50 for inhibition of GFP fluorescence. For eukaryotic cytotoxicity, macrophage cells including bone marrow derived macrophage (BMDM), THP-1 and J774, were tested with serial dilutions of DosR regulon inhibitors over an 8 point dose response curve ranging from 400 µM to 0.65 µM. Macrophages were incubated for 3 d and viability was determined using CellTiter-Glo (Promega) luminescent cell viability assay. 73 Supplementary Table 2. Small molecule screening data Category Parameter Description Assay Type of assay Whole Cell Phenotypic Screen against a M. tuberculosis hypoxia-inducible fluorescent reporter strain Inhibition of DosRST regulated reporter fluorescence. The intended targets were DosR (P95193), DosS (P95194) or DosT (O53473). Inhibition of hypoxia-inducible reporter GFP fluorescence and with limited impact on reporter growth (measured by optical density) CDC1551 (aprA’::GFP) reporter strain Target Primary measurement Key reagents Assay protocol Library size Described in methods section “High throughput screening and data analysis” Validation of reporter strain is published in Tan et al., Plos Pathogens 2013, ref 26). 540,288 Library composition Small molecules and known bioactives Source Format Harvard Medical Schools Institute of Chemistry and Cell Biology’s (ICCB) in house library (211,655 compounds) and the NIH MLP collection (328,633 compounds) Compound libraries described at: http://iccb.med.harvard.edu/libraries/compoundlibraries/ 384-well optical plate Concentration(s) tested ~10 µM in DMSO Plate controls DMSO negative control, 0.3 µM rifampin positive control Compounds were arrayed using Harvard’s in house robotics platform. Reporter strain was added to the plates using a CyBio Selma Liquid handler. Plates were read using a Perkin-Elmer Enspire plate reader. Z’-Factors are calculated for each plate. Z’-factor for whole screen averaged 0.9. None Additional comments Library Additional comments Screen Reagent/ compound dispensing system Detection instrument and software Assay validation/QC Correction factors Normalization Additional comments Post-HTS analysis Hit criteria Hit rate Additional assay(s) Confirmation of hit purity and structure Percent inhibition relative to the positive and negative controls. See equation of page 17 in the methods section. The screen was conducted with a clinical strain of M. tuberculosis at Michigan State University in a BSL-3 HTS facility. >35% inhibition of reporter GFP fluorescence, and >1.5 fold inhibition of reporter fluorescence relative to optical density Remains to be fully determined, but below 0.01%. GFP quenching; eukaryotic cytotoxicity in primary bone marrow macrophages and two macrophage like cell lines; EC50 determinations; RNA-seq; inhibition of Mtb survival during persistence; biochemical assays showing targeting of DosS or DosT using UV-visible spectroscopy or autophosphorylation assays. HC101 (artemisinin and analogs) are well characterized and were purchased from Sigma. For HC102A and HC103A,the commercial compounds were purchased from Enamine and Chembridge, respectively, and verified by chemists (see methods). HC102A and HC103A were also synthesized by organic chemists and confirmed in the whole cell and biochemical assays. Additional comments Supplementary Table 2.2. Small molecule screening data 74 MLP Collection Normalized Percent Inhibition Normalized Percent Inhibition ICCB-Longwood collection Normalized Optical Density Normalized Fluorescence c Fluorescence Inhibition to Growth Inhibition ratio a Normalized Optical Density Normalized Fluorescence FDR P-Values b Screening collec3on ICCB-Longwood MLP Measure Op3cal Density Fluoresence Op3cal Density Fluoresence Normalized Percent Inhibi3on Average Standard Devia3on 0.28 8.60 0.40 7.39 0.02 5.91 -0.01 6.72 Supplementary Figure 2.1. Statistical analysis of HTS controls. (a) Box plots showing the variation of DMSO control wells (negative control) from the screens of the ICCB- Longwood and Molecular Libraries Program (MLP) collections. Boxes show the 25 and 75% quartiles and the whiskers are 1.5x the interquartile range (approximately 3 standard deviations away from the mean). The dots are considered “outliers”. (b) Table with the means and standard deviations from the DMSO controls. Along with the Z- factor of 0.9, tight clustering of the control wells around 0% inhibition for both fluorescence and optical density support robustness of the screen. (c) FDR p-values plotted vs. the fluorescence inhibition to growth inhibition ratios shows that the chosen 1.5-fold cut-off ratio (red line) is significant (p<0.0003). 75 % Inhibition 100 c Artesunate GFP Artesunate OD DHA GFP DHA OD 100 % Inhibition a 50 0 0.01 CGC-232500 GFP CGC-232500 OD 50 0 0.01 0.1 1 10 100 Concentration (µM) 0.1 1 10 100 Concentration (µM) b d HC103B-GFP HC103B-OD HC104A-GFP HC104A-OD HC105A-GFP HC105A-OD HC106A-GFP HC106A-OD 50 0 0.01 0.1 1 10 Concentration (µM) 100 % Inhibition % Inhibition 100 50 0 0.01 100 CGC-257424 GFP CGC-257424 OD 0.1 1 10 Concentration (µM) 100 Supplementary Figure 2.2. Identification of inhibitors of the DosRST pathway. Dose response curves of GFP fluorescence inhibition of CDC1551(hspX’::GFP) reporter treated with DosRST regulon inhibitor compounds. (a) Artemisinin analogs, artesunate and dihydroartemisinin (DHA). (b) HC103B and HC104A-HC106A. (c) HC102A was generated by organic synthesis (CCG2323500) and exhibited expected activity. (d) HC103A was generated by organic synthesis (CCG257424) and exhibited expected activity. Error bars represent the standard deviation and experiments were repeated at least twice with similar results. 76 c ● 0.01 2 ● ●● ● ●● ● ● ●● ●● ● ● ● ●● ● ●●● ●● ●● ● ● ● ● ● ● ● ● ● ●●● ●●●● ● ● ● ● ●● ● ●● ● ●● ● ●● ● ● ●● ●●●● ●● ●● ● ● ●●● ●●● ● ● ● ●●● ● ●●● ● ● ● ● ● ●● ● ●●● ● ● ●● ● ●● ● ● ●●●● ● ●● ● ●● ● ●● ● ● ● ● ● ●● ● ● ● ● ● ●● ●● ● ●● ● ● ● ● ● ●● ● ● ● ● ● ● ● ●● ●● ● ● ● ● ● ● ● ● ● ● ●●●●● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●●●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●●●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●●●● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●●●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●●●●●● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ●●●●●● ● ● ●● ● ●● ●● ●● ● ●● ● ●● ● ●●● ● ●●● ● ●● ●●●● ●●● ●● ●●●● ●●● ●● ●● ● ●●● ●●●● ●●● ● ●● ● ● ● ● ● ●● ●● ●●●●●● ● ● ●●●● ● ●●● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ●● ● ●● ● ● ●● ● ● ● ● ● ● ● ● ● ● ●● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ●●● ● ● ● ● ● ● dosS ● fdxA Rv3134c dosR -8 0.001 0 0.1 -2 dosR hspX tgs1 ● -4 HC102A HC103A -6 Art −8 −6 Fold −4 change −2 (log 0 2) 2 1 Fold−change (log2) Fold change (relative to DMSO control) 4 4 a ● ● hspX ● ● ● tgs1 Rv3131 HC102A ● !!!2!!!!!!!!!!!!!4!!!!!!!!!!!!!!6!!!!!!!!!!!!!8!!!!!!!!!!!!10!!!!!!!!!!!12!!!!!!!!!!!14!! 4 ● n.s HC102A DHA DMSO 2 0 -2 14 Average Counts (logCPM) ● ● ●● ● ● ● ● ● ● ● ●● ● ● ●● ● ●● ● ●● ● ●● ● ● ● ● ● ● ● ● ● ● ● ●● ●●●● ● ● ● ● ●●● ●●● ● ● ●● ● ● ● ● ● ●● ● ● ●● ● ●●● ●● ● ●● ● ● ● ●● ● ●● ● ●● ● ● ●●●● ●● ●● ●● ●● ●●● ● ● ● ● ●●● ● ● ●●●●●●● ●● ● ●●● ● ●● ● ● ● ● ●●● ●● ● ● ● ● ●●● ● ● ● ● ● ●● ● ● ● ● ● ●● ● ● ● ● ●● ●● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●●●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ●● ● ● ● ● ● ● ●● ● ●● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ●●●●● ● ●● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●●● ●● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●●●● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●●● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ●● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ●● ● ●●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ●●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●●● ●●● ● ● ●●●● ● ● ● ● ●●●● ● ●● ● ●● ●●● ● ● ● ● ● ● ●●● ●● ● ● ● ● ● ● ●● ●●● ●● ● ● ● ● ●● ● ● ●●● ● ● ●● ●● ●●●● ● ●● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ●● ● ● ● ● ● ● ● ● ●● ● ●● ● ● ● ● ● ● ●● ● ●● ● ● ●●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ●● ● ● ● ● ●● ● ● ● ●● ● ● ● ● ● ● ●● ●● ● ●● ● ● ● ● ● ●●● ●●● ● ●● ● ● ● ● ●● ●● ● ●●●● ● ● ● ● ●●● ● ●● ● ●● ● ● ● ● ●● ● ●● ● ● ● ●●● ●● ● ● ●●● ●● ●● ●● ● ● ● ● ● ● ● ●● ● ● ●● ●●● ●● ●● ●● ● ● ● ● ●●● ●●● ●● ● ● ●● ● ● ● ● ● ● ●● ● ● ● ●● ● ● ● ●● ●● ● ●● ● ● ● ● ● ● ● ● ● ●●● ●● ● ● ● ● ● ● ●● ● ● ● ● ●●●● ● ●● ● ● ● ● ●● ● ●● ● ● ● ●● ● ●● ● ●●● ●●● ● ●● ●● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ●● ● ●● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ●● ● ● ●● ● ● ● ● ● ● ●● ● ● -6 -4 dosR dosS -8 Fold−change (log2) Vitamin C HC103A HC102A DHA DMSO HC103A 0.1 HC103A 1 HC102A Average counts (logCPM) 6 8 10 12 ● 10 0.01 4 4 dosR hspX tgs1 100 2 d −8 −6 Fold −4change −2 (log 0 2) 2 1000 DHA Fold change (relative to DMSO control) b hspX fdxA ● Rv3131 Rv3134c ● HC103A ● ● ● ● tgs1 !!!2!!!!!!!!!!!!!4!!!!!!!!!!!!!!6!!!!!!!!!!!!!8!!!!!!!!!!!!10!!!!!!!!!!!12!!!!!!!!!!!14!! 2 NO 4 6 8 10 12 Average counts (logCPM) 14 Supplementary Figure 2.3. Transcriptional profiling shows artemisinin, HC102A and HC103A inhibit Average Counts (logCPM) the core genes of the DosRST regulon during hypoxia. (a) Inhibition of DosR regulon under hypoxic conditions by the DosR regulon inhibitors. Mtb treated with compounds of interest was grown at 37 °C without shaking for 6 d, and total RNA was extracted for RT-PCR quantification. RT-PCR shows three highly induced DosR regulon genes (dosR, hspX and tgs1) under hypoxia were repressed by the DosR regulon inhibitors. Error bars represent the standard deviation. (b) NO and Vitamin C assays. Mtb cells were pre-treated with compounds of interest for 24 h, and total RNA was extracted after inducing with NO or vitamin C for 2 h. HC102A and HC103A inhibited the induction of DosR regulon by NO and vitamin C, but DHA had little effect. In all cases, the difference in the drug treated samples compared to DMSO treated samples in response to vitamin C or NO is significant with a P-value <0.001 based on t-test, except those marked as nonsignificant (n.s.). Error bars represent the standard deviation. Mtb differential gene expression in response to HC102A (panel c) and HC103A (panel d). Genes in red have a p-value <0.05, and indicated gene names are DosR regulated genes. The transcriptional analysis from different assays collectively confirm that the DosR pathway is the target of artemisinin, HC102A and HC103A. Experiments were repeated at least twice with similar results. 77 b Viability (%) 50 Artemisinin Day 15 125 100 HC102A 100 HC103A 75 50 25 0 R os 3A 1 0 ΔD 10 C H C H 0 10 100 Concentration (µM) Ar te m is 10 in 2A in % Inhibition (relative to DMSO control) a d c 0 µM INH 6 1 µM INH 5 µM INH 100 5 µM INH 5 25 µM INH 4 25 µM INH Viability (%) 75 ** 50 * ** * * 25 3 * * artemisinin HC102A HC103A artemisinin HC102A 80 µM 40 µM 80 µM 40 µM 80 µM 40 µM SO 80 µM 40 µM 80 µM 40 µM 80 µM 40 µM DM SO 0 DM Surviving Bacteria (log10[CFU/mL]) 1 µM INH HC103A Supplementary Figure 2.4. Artemisinin, HC102A and HC103A inhibit TAG synthesis, survival and isoniazid tolerance during NRP. (a) Quantification of TAG accumulation for Mtb treated with compounds of interest shows that DosRST regulon inhibitors repress TAG synthesis to the level similar to DdosR mutant. Error bars represent the standard deviation. (b) Dose dependent inhibition of Mtb survival during NRP following 15 d of treatment. Percent viability was calculated relative to the viable bacteria in the DMSO control at day 15. Error bars represent the standard deviation of the mean. (c) Fifteen days of treatment with artemisinin, HC102A or HC103A, with or without INH, significantly reduces bacterial survival (P-value <0.05 based on a t-test) during NRP relative to the respective DMSO controls. (d) Fifteen days of treatment with artemisinin, HC102A and HC103A reduces isoniazid tolerance during NRP. To quantify INH tolerance, percent viability at 1, 5 and 25 µM INH was measured relative to the 0 µM INH control (DMSO control). Significant differences (marked with an asterisk, p<0.05 based on a t-test) were calculated relative to the respective DMSO control sample. For example, cells treated with 5 µM INH and artemisinin, HC102A and HC103A are significantly different from the DMSO control treated with 5 µM INH. Experiments were repeated at least twice with similar results. 78 Absorbance 2.0 1.5 1.0 0.5 0.0 350 b DosS 1.5 Anaerobic DTN 50 µM Art 100 µM Art 200 µM Art 400 µM Art 400 450 500 550 Wavelength (nm) Absorbance a 1.0 0.5 600 0.0 350 DosT G85L Anaerobic DTN 50 µM Art 100 µM Art 200 µM Art 400 450 500 550 Wavelength (nm) 600 Supplementary Figure 2.5. Artemisinin directly modulates DosS and DosT (G85L) heme in a dosedependent manner. UV-visible spectra of DosS (panel a) and DosT (G85L) (panel b) treated with different concentrations of artemisinin. Artemisinin modulated DosT heme at a lower concentration (50 µM, Figure 2.5d) than DosS heme (400 µM), confirming that DosT is more sensitive to artemisinin. DosT (G85L) exhibits a similar profile to WT DosT (Figure 2.5d) 79 a 1.5 DosS + DTN (Fe2+) 1.5 DosS + DTN + HC102A (Fe2+) 1.0 DosS + DTN + HC102A + Art (Fe3+) 0.5 0.0 350 400 450 500 550 600 Wavelength (nm) Absorbance Absorbance b DosS (Fe3+) DosS (Fe3+) DosS + DTN (Fe2+) 1.0 DosS + DTN + HC103A (Fe2+) 0.5 0.0 350 400 450 500 550 Wavelength (nm) 600 Supplementary Figure 2.6. HC102A and HC103A do not modulate DosS redox. DosS treated with HC102A (panel a) or HC103A (panel b) shows a similar overall spectrum as the DMSO control. This result indicates that HC102A and HC103A has no effect in modulating redox status of DosS, and may inhibit DosR regulon by distinct mechanism. 80 dosR hspX 100 DosT DosT G85L DosT G115L 50 0 -50 -100 0.1 1 10 100 % Inhibition % Inhibition 100 0 -50 -100 Concentration (µM) DosT DosT G85L DosT G115L 50 0.1 1 10 Concentration (µM) 100 tgs1 % Inhibition 100 DosT DosT G85L DosT G115L 50 0 -50 -100 0.1 1 10 Concentration (µM) 100 Supplementary Figure 2.7. Substitutions in DosT provide resistance to artemisinin. Mtb was transformed with a replicating plasmid that overexpresses WT dosT, dosT (G85L) or dosT (G115L). Cells were treated with 0.025 µM - 20 µM artemisinin, and total RNA was extracted after 6 d incubation at 37 °C. RT-PCR quantification of DosR-regulated genes (dosR, hspX and tgs1) shows that strains expressing DosT (G115L) exhibited EC50 for artemisinin-mediated inhibition of DosR regulon genes of 1.0-1.6 µM, which is ~5-fold more resistant than WT DosT or DosT (G85L) with EC50 of 0.2-0.3 µM. 81 Supplementary Figure 2.8. Modulation of DosS and DosT kinase activities by HC102A and HC103A. Full western blots from which the cropped images in Figure 2.6 were generated. DosS and DosT protein is marked by the asterisk. 82 Supplementary Notes Chemical synthesis of HC102A (CCG-232500) and HC103A (CCG-257424) Powdered samples of commercially sourced HC102A and HC103A were analyzed by MS and combustion analysis and found to have a molecular formula consistent with the reported structures. 2D-NMR analysis confirmed HC102A to be the racemic (5S,9R)-7,7,9-trimethyl-1,3diazaspiro[4.5]decane-2,4-dione (alpha) isomer. For HC102A synthesis (generating a compound designated CCG-232500), a mixture of 3,3,5-trimethylcyclohexan-1-one (1.1 g, 7.6 mmol) was treated with sodium cyanide (0.92 g, 18.8 mmol), and carbonic acid, diammonia salt (3.6 g, 37.5 mmol). EtOH (10 mL) and water (10 mL) were added and the resulting mixture was heated to 55 °C for 6 h. The mixture was cooled and then diluted with cold water, treated with conc. HCl (2 mL) and filtered. The collected solid was washed with water (2x) and triturated in hot methanol. The resulting solid was collected by filtration and dried under high vacuum overnight at room temp. (5S,9R)-7,7,9-trimethyl-1,3-diazaspiro[4.5]decane-2,4-dione was obtained as a white solid (0.53 g, 2.5 mmol, 32%). HPLC system A (tR = 5.6 min). 1H NMR (400 MHz, DMSO-d6) δ 10.63 (s, 1H), 8.09 (s, 1H), 1.83 (d, J = 9.3 Hz, 1H), 1.44 (dd, J = 13.3, 7.1 Hz, 2H), 1.32 (dd, J = 32.7, 13.7 Hz, 2H), 1.14 (t, J = 12.9 Hz, 1H), 0.96 (s, 3H), 0.92 – 0.62 (m, 7H). ESI-MS m/z 209.1 (MH+). For HC103A synthesis (generating a compound designated CCG-257424), first 3-amino-N(3- hydroxyphenyl)benzamide was synthesized. To a solution of 3-aminobenzoic acid (1 g, 7.3 mmol), 3-aminophenol (0.88 g, 8.0 mmol) and HOBT (1.3 g, 8.7 mmol) in dry DMF cooled to 0 °C was added EDC (1.6 g, 8.75 mmol). The resulting solution was allowed to warm to room temperature and stirred overnight. The reaction was diluted with water and washed with satd. NaHCO3, satd. NaCl solution and dried over MgSO4. After filtration, the organic layer was concentrated in vacuo and purified by flash chromatography (CombiFlash, CH2Cl2/MeOH 83 gradient). 3-amino-N-(3-hydroxyphenyl)benzamide was obtained as an amorphous solid (0.24 g, 7.29 mmol, 14.4 % yield). 1H NMR (400 MHz, DMSO-d6) (Rotomers) δ 10.18 (m, J = 51.4, 10.6 Hz, 1H), 9.90 (s, 1H), 9.35 (m, 2H), 8.25 (s, 1H), 7.95 (m, 1H), 7.83 – 7.34 (m, 2H), 6.82 – 6.32 (m, 2H). 5.2 (s, 2H). To synthesize CGC257424, to a solution of 3-amino-N-(3hydroxyphenyl)benzamide (0.2 g, 0.87 mmol), EDC (0.20 g, 1.1 mmol), and HOBT (0.16 g, 1.0 mmol) in dry DMF cooled to 0 °C was added thiophene-2-carboxylic acid (0.12 g, 0.96 mmol) followed by catalytic 4-dimethylaminopyridine. The resulting solution was allowed to warm to room temperature and stir overnight. The reaction was diluted with water and washed with satd. NaHCO3, satd. NaCl solution and dried, MgSO4. The organic layer was filtered and concentrated in vacuo. Flash chromatography (CH2Cl2/MeOH) was used to obtain N-(3-((3- hydroxyphenyl)carbamoyl)phenyl)thiophene-2-carboxamide as a white solid (0.06 g, 0.17 mmol, 20.2 % yield). HPLC system A (tR = 5.8 min). 1H NMR (400 MHz, DMSO-d6) δ 10.04 (s, 1H), 9.62 (s, 1H), 8.91 (s, 1H), 8.13 (t, J = 1.9 Hz, 1H), 8.02 – 7.90 (m, 2H), 7.59 (d, J = 17.2 Hz, 2H), 7.50 (dd, J = 4.9, 1.2 Hz, 1H), 7.38 – 7.29 (m, 2H), 7.11 – 6.97 (m, 2H), 6.47 (dt, J = 8.3, 1.4 Hz, 1H). ESI-MS m/z 339.0 (M+H+). Starting materials were purchased from Fisher, Sigma-Aldrich Lancaster, Fluka or TCIAmerica and were used without purification. All reaction solvents were purchased from Fisher and used as received. Reactions were monitored by TLC using precoated silica gel 60 F254 plates. Silica gel chromatography was performed by flash chromatography using silica gel (220-240 mesh) obtained from Silicycle or via medium pressure liquid chromatography on a CombiFlash instrument. NMR spectra were recorded on a Varian 400 MHz spectrometer. Chemical shifts are reported in δ (parts per million), by reference to the hydrogenated residues of deuterated solvent as internal standard CDCL3: δ = 7.28 (1H NMR). Mass spectra were recorded on a Micromass 84 LCT time-of-flight instrument utilizing the electrospray ionization mode. The purity of the compounds was assessed via analytical reversed-phase high-performance liquid chromatography with a gradient of 10% acetonitrile/water to 90% acetonitrile/water over 6 min (“System A”, C18 column, 3.5 µm, 4.6x100 mm, 254 nm). 85 REFERENCES 86 REFERENCES 1. Gengenbacher M, Kaufmann SH. 2012. Mycobacterium tuberculosis: success through dormancy. FEMS Microbiol Rev 36:514-532. 2. Baker JJ, Johnson BK, Abramovitch RB. 2014. Slow growth of Mycobacterium tuberculosis at acidic pH is regulated by phoPR and host-associated carbon sources. Mol Microbiol 94:56-69. 3. Wayne LG, Sohaskey CD. 2001. Nonreplicating persistence of Mycobacterium tuberculosis. Annu Rev Microbiol 55:139-163. 4. Betts JC, Lukey PT, Robb LC, McAdam RA, Duncan K. 2002. Evaluation of a nutrient starvation model of Mycobacterium tuberculosis persistence by gene and protein expression profiling. Mol Microbiol 43:717-731. 5. Boon C, Dick T. 2002. Mycobacterium bovis BCG response regulator essential for hypoxic dormancy. J Bacteriol 184:6760-6767. 6. Park HD, Guinn KM, Harrell MI, Liao R, Voskuil MI, Tompa M, Schoolnik GK, Sherman DR. 2003. Rv3133c/dosR is a transcription factor that mediates the hypoxic response of Mycobacterium tuberculosis. Mol Microbiol 48:833-843. 7. Voskuil MI, Schnappinger D, Visconti KC, Harrell MI, Dolganov GM, Sherman DR, Schoolnik GK. 2003. Inhibition of respiration by nitric oxide induces a Mycobacterium tuberculosis dormancy program. J Exp Med 198:705-713. 8. Galagan JE, Minch K, Peterson M, Lyubetskaya A, Azizi E, Sweet L, Gomes A, Rustad T, Dolganov G, Glotova I, Abeel T, Mahwinney C, Kennedy AD, Allard R, Brabant W, Krueger A, Jaini S, Honda B, Yu WH, Hickey MJ, Zucker J, Garay C, Weiner B, Sisk P, Stolte C, Winkler JK, Van de Peer Y, Iazzetti P, Camacho D, Dreyfuss J, Liu Y, Dorhoi A, Mollenkopf HJ, Drogaris P, Lamontagne J, Zhou Y, Piquenot J, Park ST, Raman S, Kaufmann SH, Mohney RP, Chelsky D, Moody DB, Sherman DR, Schoolnik GK. 2013. The Mycobacterium tuberculosis regulatory network and hypoxia. Nature 499:178-183. 9. Ioanoviciu A, Meharenna YT, Poulos TL, Ortiz de Montellano PR. 2009. DevS oxy complex stability identifies this heme protein as a gas sensor in Mycobacterium tuberculosis dormancy. Biochemistry 48:5839-5848. 10. Vos MH, Bouzhir-Sima L, Lambry JC, Luo H, Eaton-Rye JJ, Ioanoviciu A, Ortiz de Montellano PR, Liebl U. 2012. Ultrafast ligand dynamics in the heme-based GAF sensor domains of the histidine kinases DosS and DosT from Mycobacterium tuberculosis. Biochemistry 51:159-166. 87 11. Kumar A, Toledo JC, Patel RP, Lancaster JR, Jr., Steyn AJ. 2007. Mycobacterium tuberculosis DosS is a redox sensor and DosT is a hypoxia sensor. Proc Natl Acad Sci U S A 104:11568-11573. 12. Honaker RW, Leistikow RL, Bartek IL, Voskuil MI. 2009. Unique roles of DosT and DosS in DosR regulon induction and Mycobacterium tuberculosis dormancy. Infect Immun 77:3258-3263. 13. Leistikow RL, Morton RA, Bartek IL, Frimpong I, Wagner K, Voskuil MI. 2010. The Mycobacterium tuberculosis DosR regulon assists in metabolic homeostasis and enables rapid recovery from nonrespiring dormancy. J Bacteriol 192:1662-1670. 14. Converse PJ, Karakousis PC, Klinkenberg LG, Kesavan AK, Ly LH, Allen SS, Grosset JH, Jain SK, Lamichhane G, Manabe YC, McMurray DN, Nuermberger EL, Bishai WR. 2009. Role of the dosR-dosS two-component regulatory system in Mycobacterium tuberculosis virulence in three animal models. Infect Immun 77:12301237. 15. Gautam US, McGillivray A, Mehra S, Didier PJ, Midkiff CC, Kissee RS, Golden NA, Alvarez X, Niu T, Rengarajan J, Sherman DR, Kaushal D. 2015. DosS is required for the complete virulence of Mycobacterium tuberculosis in mice with classical granulomatous lesions. Am J Respir Cell Mol Biol 52:708-716. 16. Mehra S, Foreman TW, Didier PJ, Ahsan MH, Hudock TA, Kissee R, Golden NA, Gautam US, Johnson AM, Alvarez X, Russell-Lodrigue KE, Doyle LA, Roy CJ, Niu T, Blanchard JL, Khader SA, Lackner AA, Sherman DR, Kaushal D. 2015. The DosR Regulon modulates adaptive immunity and is essential for Mycobacterium tuberculosis persistence. Am J Respir Crit Care Med 191:1185-1196. 17. Baek SH, Li AH, Sassetti CM. 2011. Metabolic regulation of mycobacterial growth and antibiotic sensitivity. PLoS Biol 9:e1001065. 18. VanderVen BC, Fahey RJ, Lee W, Liu Y, Abramovitch RB, Memmott C, Crowe AM, Eltis LD, Perola E, Deininger DD, Wang T, Locher CP, Russell DG. 2015. Novel inhibitors of cholesterol degradation in Mycobacterium tuberculosis reveal how the bacterium's metabolism is constrained by the intracellular environment. PLoS Pathog 11:e1004679. 19. Rasko DA, Moreira CG, Li de R, Reading NC, Ritchie JM, Waldor MK, Williams N, Taussig R, Wei S, Roth M, Hughes DT, Huntley JF, Fina MW, Falck JR, Sperandio V. 2008. Targeting QseC signaling and virulence for antibiotic development. Science 321:1078-1080. 20. Rasko DA, Sperandio V. 2010. Anti-virulence strategies to combat bacteria-mediated disease. Nat Rev Drug Discov 9:117-128. 88 21. Anthouard R, DiRita VJ. 2013. Small-molecule inhibitors of toxT expression in Vibrio cholerae. MBio 4. 22. Ng WL, Perez L, Cong J, Semmelhack MF, Bassler BL. 2012. Broad spectrum proquorum-sensing molecules as inhibitors of virulence in vibrios. PLoS Pathog 8:e1002767. 23. Johnson BK, Colvin CJ, Needle DB, Mba Medie F, Champion PA, Abramovitch RB. 2015. The carbonic anhydrase inhibitor ethoxzolamide inhibits the Mycobacterium tuberculosis PhoPR regulon and Esx-1 secretion and attenuates virulence. Antimicrob Agents Chemother 59:4436-4445. 24. Tan S, Sukumar N, Abramovitch RB, Parish T, Russell DG. 2013. Mycobacterium tuberculosis responds to chloride and pH as synergistic cues to the immune status of its host cell. PLoS Pathog 9:e1003282. 25. Johnson BK, Scholz MB, Teal TK, Abramovitch RB. 2016. SPARTA: Simple program for automated reference-based bacterial RNA-seq transcriptome analysis. BMC Bioinformatics 17:66. 26. Taneja NK, Dhingra S, Mittal A, Naresh M, Tyagi JS. 2010. Mycobacterium tuberculosis transcriptional adaptation, growth arrest and dormancy phenotype development is triggered by vitamin C. PLoS One 5:e10860. 27. Mak PA, Rao SP, Ping Tan M, Lin X, Chyba J, Tay J, Ng SH, Tan BH, Cherian J, Duraiswamy J, Bifani P, Lim V, Lee BH, Ling Ma N, Beer D, Thayalan P, Kuhen K, Chatterjee A, Supek F, Glynne R, Zheng J, Boshoff HI, Barry CE, 3rd, Dick T, Pethe K, Camacho LR. 2012. A high-throughput screen to identify inhibitors of ATP homeostasis in non-replicating Mycobacterium tuberculosis. ACS Chem Biol 7:11901197. 28. Deb C, Lee CM, Dubey VS, Daniel J, Abomoelak B, Sirakova TD, Pawar S, Rogers L, Kolattukudy PE. 2009. A novel in vitro multiple-stress dormancy model for Mycobacterium tuberculosis generates a lipid-loaded, drug-tolerant, dormant pathogen. PLoS One 4:e6077. 29. Krishna S, Bustamante L, Haynes RK, Staines HM. 2008. Artemisinins: their growing importance in medicine. Trends Pharmacol Sci 29:520-527. 30. Selmeczi K, Robert A, Claparols C, Meunier B. 2004. Alkylation of human hemoglobin A0 by the antimalarial drug artemisinin. FEBS Lett 556:245-248. 31. Robert A, Dechy-Cabaret O, Cazelles J, Meunier B. 2002. From mechanistic studies on artemisinin derivatives to new modular antimalarial drugs. Acc Chem Res 35:167-174. 89 32. Ioanoviciu A, Yukl ET, Moenne-Loccoz P, de Montellano PR. 2007. DevS, a hemecontaining two-component oxygen sensor of Mycobacterium tuberculosis. Biochemistry 46:4250-4260. 33. Cho HY, Cho HJ, Kim MH, Kang BS. 2011. Blockage of the channel to heme by the E87 side chain in the GAF domain of Mycobacterium tuberculosis DosS confers the unique sensitivity of DosS to oxygen. FEBS Lett 585:1873-1878. 34. Podust LM, Ioanoviciu A, Ortiz de Montellano PR. 2008. 2.3 A X-ray structure of the heme-bound GAF domain of sensory histidine kinase DosT of Mycobacterium tuberculosis. Biochemistry 47:12523-12531. 35. Cho HY, Cho HJ, Kim YM, Oh JI, Kang BS. 2009. Structural insight into the hemebased redox sensing by DosS from Mycobacterium tuberculosis. J Biol Chem 284:1305713067. 36. Sousa EH, Tuckerman JR, Gonzalez G, Gilles-Gonzalez MA. 2007. DosT and DevS are oxygen-switched kinases in Mycobacterium tuberculosis. Protein Sci 16:1708-1719. 37. Messori L, Gabbiani C, Casini A, Siragusa M, Vincieri FF, Bilia AR. 2006. The reaction of artemisinins with hemoglobin: a unified picture. Bioorg Med Chem 14:29722977. 38. Zhang S, Gerhard GS. 2008. Heme activates artemisinin more efficiently than hemin, inorganic iron, or hemoglobin. Bioorg Med Chem 16:7853-7861. 39. Robert A, Coppel Y, Meunier B. 2002. Alkylation of heme by the antimalarial drug artemisinin. Chem Commun (Camb):414-415. 40. Sivaramakrishnan S, de Montellano PR. 2013. The DosS-DosT/DosR mycobacterial sensor system. Biosensors 3:259-282. 41. Roberts DM, Liao RP, Wisedchaisri G, Hol WG, Sherman DR. 2004. Two sensor kinases contribute to the hypoxic response of Mycobacterium tuberculosis. J Biol Chem 279:23082-23087. 42. Koul A, Vranckx L, Dendouga N, Balemans W, Van den Wyngaert I, Vergauwen K, Gohlmann HW, Willebrords R, Poncelet A, Guillemont J, Bald D, Andries K. 2008. Diarylquinolines are bactericidal for dormant mycobacteria as a result of disturbed ATP homeostasis. J Biol Chem 283:25273-25280. 43. Pethe K, Bifani P, Jang J, Kang S, Park S, Ahn S, Jiricek J, Jung J, Jeon HK, Cechetto J, Christophe T, Lee H, Kempf M, Jackson M, Lenaerts AJ, Pham H, Jones V, Seo MJ, Kim YM, Seo M, Seo JJ, Park D, Ko Y, Choi I, Kim R, Kim SY, Lim S, Yim SA, Nam J, Kang H, Kwon H, Oh CT, Cho Y, Jang Y, Kim J, Chua A, Tan BH, Nanjundappa MB, Rao SP, Barnes WS, Wintjens R, Walker JR, Alonso S, Lee S, 90 Kim J, Oh S, Oh T, Nehrbass U, Han SJ, No Z, et al. 2013. Discovery of Q203, a potent clinical candidate for the treatment of tuberculosis. Nat Med 19:1157-1160. 44. Li W, Upadhyay A, Fontes FL, North EJ, Wang Y, Crans DC, Grzegorzewicz AE, Jones V, Franzblau SG, Lee RE, Crick DC, Jackson M. 2014. Novel insights into the mechanism of inhibition of MmpL3, a target of multiple pharmacophores in Mycobacterium tuberculosis. Antimicrob Agents Chemother 58:6413-6423. 45. Gupta RK, Thakur TS, Desiraju GR, Tyagi JS. 2009. Structure-based design of DevR inhibitor active against nonreplicating Mycobacterium tuberculosis. J Med Chem 52:63246334. 46. Rybniker J, Chen JM, Sala C, Hartkoorn RC, Vocat A, Benjak A, Boy-Rottger S, Zhang M, Szekely R, Greff Z, Orfi L, Szabadkai I, Pato J, Keri G, Cole ST. 2014. Anticytolytic screen identifies inhibitors of mycobacterial virulence protein secretion. Cell Host & Microbe 16:538-548. 47. Boshoff HI, Myers TG, Copp BR, McNeil MR, Wilson MA, Barry CE, 3rd. 2004. The transcriptional responses of Mycobacterium tuberculosis to inhibitors of metabolism: novel insights into drug mechanisms of action. J Biol Chem 279:40174-40184. 48. Miller MJ, Walz AJ, Zhu H, Wu C, Moraski G, Mollmann U, Tristani EM, Crumbliss AL, Ferdig MT, Checkley L, Edwards RL, Boshoff HI. 2011. Design, synthesis, and study of a mycobactin-artemisinin conjugate that has selective and potent activity against tuberculosis and malaria. J Am Chem Soc 133:2076-2079. 49. Kim MJ, Park KJ, Ko IJ, Kim YM, Oh JI. 2010. Different roles of DosS and DosT in the hypoxic adaptation of Mycobacteria. J Bacteriol 192:4868-4875. 50. Sander P, B S, Bottger E. 2001. Gene Replacement in Mycobacterium tuberculosis and Mycobacterium bovis BCG Using rpsL as a Dominant Negative Selectable Marker, vol 54. Humana Press, Totowa, NJ. 51. Abramovitch RB, Rohde KH, Hsu FF, Russell DG. 2011. aprABC: a Mycobacterium tuberculosis complex-specific locus that modulates pH-driven adaptation to the macrophage phagosome. Mol Microbiol 80:678-694. 52. Zhang JH, Chung TD, Oldenburg KR. 1999. A Simple Statistical Parameter for Use in Evaluation and Validation of High Throughput Screening Assays. J Biomol Screen 4:6773. 53. Johnson BK, Abramovitch RB. 2015. Macrophage infection models for Mycobacterium tuberculosis. Methods Mol Biol 1285:329-341. 91 54. Rohde KH, Abramovitch RB, Russell DG. 2007. Mycobacterium tuberculosis invasion of macrophages: linking bacterial gene expression to environmental cues. Cell Host Microbe 2:352-364. 55. Schneider CA, Rasband WS, Eliceiri KW. 2012. NIH Image to ImageJ: 25 years of image analysis. Nat Methods 9:671-675. 92 CHAPTER 3 - Chemical inhibition of the Mycobacterium tuberculosis DosRST twocomponent regulatory system by targeting sensor kinase heme and response regulator DNA binding This work is in preparation for journal submission. Contributions and authors for the work described below: Huiqing Zheng, Bilal Aleiwi, Edmund Ellsworth, and Robert B. Abramovitch H.Z., E.E. and R.B.A conceived the experiments. H.Z. performed the Mtb physiology and biochemistry experiments. B.A. synthesized the analogs. E.E. directed the medicinal chemistry optimizations. 93 Summary Mycobacterium tuberculosis (Mtb) possesses a two-component regulatory system, DosRST, that enables Mtb to sense host immune cues and establish a state of non-replicating persistence (NRP). NRP bacteria are tolerant to several anti-mycobacterial drugs and are thought to play a role in the long course of tuberculosis (TB) therapy. Therefore, small molecules that inhibit Mtb from establishing or maintaining NRP could reduce the reservoir of drug tolerant bacteria and function as an adjunct therapy to reduce treatment time. Previously, we reported the discovery of six novel chemical inhibitors of DosRST, named HC101A-106A, from a whole cell, reporter-based phenotypic high-throughput screen. Here, we report functional and mechanism of action studies of HC104A and HC106A. RNAseq transcriptional profiling shows that the compounds downregulate genes of the DosRST regulon. Both compounds reduce hypoxia-induced triacylglycerol synthesis by ~50%. HC106A inhibits Mtb survival during hypoxia-induced NRP, however, HC104A did not inhibit survival during NRP. An electrophoretic mobility assay shows that HC104A inhibits DosR DNA binding in a dosedependent manner, indicating that HC104A may function by directly targeting DosR. In contrast, UV-visible spectroscopy studies suggest HC106A directly targets the histidine kinase heme, via a mechanism that is distinct from the oxidation and alkylation of heme previously observed with artemisinin (HC101A). Synergistic interactions were observed when DosRST inhibitors were examined in pair-wise combinations with the strongest potentiation observed between artemisinin paired with HC102A, HC103A, or HC106A. Our data collectively show that the DosRST pathway can be inhibited by multiple distinct mechanisms. 94 Introduction Mycobacterium tuberculosis (Mtb) is the causative agent of tuberculosis (TB). Mtb is an intracellular pathogen that can latently infect the host without causing disease symptoms (1). During chronic infection, it can establish a dormant state known as non-replicating persistence (NRP) where Mtb modulates its metabolism in response to environmental and host immune cues, such as hypoxia, acidic pH, and nutrient starvation (2, 3). DosRST is a two-component regulatory system that regulates Mtb persistence (4-6). It consists of two sensor histidine kinases, DosS and DosT, and the cognate response regulator DosR, which regulates expression of about 50 genes in the DosRST regulon (6-8). The pathway can be induced by host intracellular stimuli, such as nitric oxide (NO), carbon monoxide (CO) and hypoxia, through DosS and DosT (9-11). DosS is an oxygen and redox sensor, whereas DosT acts an oxygen sensor (12-14). Both kinases sense ligands via the heme group, and are inactive when the heme exists as either the Met (Fe3+) form (DosS) or the oxy (Fe2+-O2) form (DosT) in the presence of O2 (13). However, hypoxic conditions activate the kinases by inducing the conversion of DosS to the ferrous form and DosT to the deoxy form. Therefore, DosS/T play overlapping and distinct roles in sensing the redox status and oxygen level of the environment to turn on the DosR pathway (11, 15). Non-replicating bacilli are naturally tolerant to many anti-mycobacterial drugs, such as isoniazid (INH) that only kills replicating Mtb (16-18). During TB infection, the disease presents as a spectrum of replicating and non-replicating bacteria; the NRP population of bacteria are thought to be responsible, in part, for the 6-month long course of TB treatment. This long antibiotic regimen makes controlling the TB epidemic challenging and has likely contributed to the evolution of drug-resistant Mtb strains. Therefore, there exists an urgent need to identify new strategies to treat the disease, with a particular focus on discovering new ways to shorten the course of TB 95 therapy. Targeting the DosRST pathway is a promising strategy, because dosRST mutants are attenuated in in vitro models of hypoxia-driven NRP (19) and in animal models that generate hypoxic granulomas, including non-human primates, guinea pigs, and C3HeB/FeJ mouse models of TB infection (4, 20-22). Furthermore, deletion of DosR-regulated gene tgs1, which is involved in triacylglycerol (TAG) synthesis, causes reduced antibiotic tolerance (23, 24). Therefore, inhibiting the DosRST pathway and killing the reservoir of NRP bacteria may function to narrow the spectrum of TB disease and shorten the course of TB therapy. In an effort to discover the new chemical probes that inhibit Mtb persistence, we previously performed a whole cell phenotypic high-throughput screening (HTS) of a >540,000 compound library using the DosRST regulon reporter strain CDC1551 (hspX’::GFP) (25). We discovered six compounds that inhibit the DosR-dependent, hypoxia-induced GFP fluorescence. In the previous report, we showed that the HC101, HC102 and HC103 series functioned to inhibit NRP associated physiologies, including TAG accumulation, survival during hypoxia and isoniazid tolerance. Mechanism of action studies showed that the HC101 series, composed of artemisinin and related analogs, functioned by oxidizing and alkylating the DosS and DosT heme. HC102 and HC103 did not modulate the DosS/T heme, and were instead found to inhibit sensor kinase autophosphorylation. Here, we report the characterization of two additional compounds, HC104A and HC106A. Transcriptional and biochemical analysis demonstrate that both compounds function to downregulate the DosRST pathway and inhibit persistence associated physiologies. Biochemical studies show HC104A and HC106A function by distinct mechanisms of action, with HC104A inhibiting DosR DNA binding and HC106A interacting with DosS and DosT heme to block environmental sensing. Studies examining the pair-wise interactions between the five DosRST inhibitors revealed synergistic interactions, including strong potentiating interactions of 96 artermisinin and HC106A. Structure activity relationship studies of HC106 identified functional groups of HC106A that are required for activity and enabled optimization of HC106 potency to nanomolar effective concentrations against whole cell Mtb. Results Inhibition of the DosR regulon by HC104A and HC106A Characterization studies were undertaken with two putative DosRST regulon inhibitors, HC104A and HC106A (Fig. 3.1a) (25). Half-maximal effective concentration (EC50) studies using the CDC1551 (hspX’::GFP) DosRST-dependent fluorescent reporter strain, show that HC104A and HC106A inhibit DosRST-dependent GFP florescence with EC50 values of 9.8 µM and 2.48 µM, respectively (Fig. 3.1b and 3.1c). The compounds have minimal impact on Mtb growth, suggesting they are also potential DosRST inhibitors, as DosRST is not required for survival under the conditions of mild hypoxia used in the reporter-based assay. RNA-seq-based transcriptional profiling was undertaken to determine if the DosRST regulon was inhibited by the compounds. Mtb was treated with 40 µM HC104A, HC106A or dimethyl sulfoxide (DMSO) control for 6 d in a standing flask, and following incubation RNA was extracted, sequenced and analyzed for differential gene expression relative to the DMSO control. As a control for the DosR regulon, transcriptional profiling was also previously conducted on a DMSO treated CDC1551(DdosR) mutant strain (25). The transcriptional profiles showed that the genes strongly repressed by HC104A and HC106A (>2-fold; q<0.05) are from the dosR regulon (Fig. 3.2a-c, Supplementary Dataset 3.1, 3.3). HC106A exhibited a remarkably strong reduction of gene expression, with transcripts for tgs1 and hspX being almost undetectable by RNA-seq following HC106A treatment. Interestingly, while HC106A broadly inhibited genes of the DosRST regulon, HC104A only 97 strongly inhibited part of the DosR regulon, with the strongest inhibition reserved for hspX, the promoter used to drive reporter fluorescence in the screen. These RNA-seq results were validated by semi-quantitative RT-PCR, with HC104A causing downregulation of dosR, hspX, and tgs1 in vitro by 6-, 570- and 13-fold, respectively; whereas HC106A downregulated these three genes by 49-, 1360-, and 1424-fold, respectively (Supplementary Fig. 3.1), with hspX and tgs1 transcripts being below the level of detection by qRT-PCR. Comparisons of transcriptional profiles from the inhibitor treated wild type (WT) Mtb strain to a CDC1551(DdosR) mutant strain showed that there are a total of 26 genes and 53 genes from dosR regulon inhibited by HC104A and HC106A, respectively. Notably, HC104A and HC106A have an additional 119 genes and 35 genes repressed that were not repressed in the CDC1551(DdosR) mutant strain (Fig. 3.2c). This observation suggested that these two compounds exhibit some DosR-independent activities. To confirm the specificity of the compounds, RNA-seq was also performed on CDC1551(DdosR) mutant background (Supplementary Dataset 3.2-3) treated with HC104A or HC106A. This analysis identified 171 genes and 51 genes that are downregulated (>2-fold; q<0.05) by HC104A and HC106A, respectively (Fig. 3.2d). This finding indicates that HC104A and HC106A impact other targets beside DosR regulon, with HC106A showing greater on-target specificity than HC104A. Based on these findings, we conclude that: 1) HC106A strongly and specifically inhibits the DosRST regulon; and 2) HC104A strongly inhibits a portion of the DosRST regulon, with several notable off-target activities. To assess the impact of the inhibitors on the DosRST pathway on intracellular Mtb, murine bone marrow-derived mouse macrophages were infected with Mtb and treated with 40 µM HC104A and HC106A for 48 h. Total bacterial RNA was isolated and analyzed by RT-PCR for hspX and tgs1 gene differential expression. The results demonstrate that the induction of hspX and 98 tgs1 were inhibited 185- and 10-fold by HC104A and 6- and 4-fold by HC106A, respectively (Fig. 3.3a). These finding confirm that HC104A and HC106A can access Mtb inside the macrophage, however, the reduced repression of the pathway by HC106A as compared to broth culture, suggests that the molecule may not be able to efficiently target intracellular Mtb. The DosRST pathway is also induced by redox signals such as vitamin C and NO (25). To examine whether HC104A and HC106A can repress the induction of DosRST pathway by vitamin C or NO, Mtb cells were pretreated with HC104A or HC106A for 24 h followed by vitamin C or NO induction for 2 h. The expression of DosR-regulated genes (hspX and tgs1) was examined by real time-PCR. Vitamin C and DETA-NONOate (NO donor) strongly induced hspX and tgs1 as previously reported ((25), Fig. 3.3b). For instance, vitamin C induced hspX and tgs1 by 2162- and 58-fold, respectively; whereas DETA-NONOate upregulated hspX and tgs1 3024- and 113-fold, respectively (Fig. 3.3b). Mtb cells pretreated with HC106A showed strong inhibition of hspX and tgs1 induction by vitamin C and DETA-NONOate. For example, HC106A inhibited the hspX and tgs1 transcripts by 78- and 14-fold following vitamin C treatment, respectively, and 362- and 151fold following DETA-NONOate treatment. Following vitamin C treatment, HC104 showed inhibition of hspX by 3.4-fold and no effect on tgs1, or 302- and 6.6-fold inhibition of hspX and tgs1 following DETA-NONOate treatment. These findings show that HC104A and HC106A act as inhibitors of the DosRST pathway in response to hypoxia and redox signals. 99 a b O N O 100 % Inhibition N HC104A GFP HC104A OD 50 0 1 HC104A Cl O N O Cl N H N H HC106A GFP HC106A OD 50 0 HC106A 1000 100 % Inhibition c Br 10 100 Concentration (µM) 0.01 0.1 1 10 100 1000 Concentration (µM) Figure 3.1. HC104A and HC106A inhibit DosRST reporter fluorescence. (a) Chemical structures of HC104A and HC106A. HC104A (b) and HC106A (c) inhibited DosR-driven GFP fluorescence signal in a dose-dependent manner, while having minimal impact of Mtb growth. The EC50 values of fluorescence inhibition for HC104A and HC106A are 9.8 µM and 2.5 µM, respectively. 100 Fold change (log2) -2 0 2 Fold change (log2) -4 -2 0 4 2 6 b a Rv1738 pfkB -4 Rv2028c Rv2028c -6 -6 Rv2030c Rv3131 narX fdxA tgs1 hspX -8 -8 hspX 2 c 4 6 8 10 12 Average counts (logCPM) D dosR mutant HC104A 14 2 4 6 8 10 12 Average counts (logCPM) d HC104A 68 4 HC106A 107 22 157 12 31 14 14 27 23 HC106A Figure 3.2. Transcriptional profiling shows HC104A and HC106A inhibited DosR regulon during hypoxia. Magnitude-amplitude plots showing differential gene expression of Mtb cells treated with 40 µM HC104A (a) or HC106A (b). The labeled genes represent selected genes that belong to the DosRST regulon. The red dots represent genes with significant differential expression, q<0.05. (c) A Venn diagram for the downregulated genes (>2-fold; q< 0.05) of WT CDC1551 treated with HC104A or HC106A compared to that of CDC1551 (DdosR). (d) Venn diagram for downregulated genes (>2-fold; q< 0.05) of CDC1551 (DdosR) treated with HC104A or HC106A. 101 Vitamin C R os 6A D C om p 0 R HC106A HC104A DMSO HC106A HC104A DMSO 0.1 HC106A 1 50 os 10 D 100 100 Δ % Survival (relative to DMSO control) n.s 1000 HC104A Fold change (relative to DMSO control) hspX tgs1 C d 10000 H b 0 10 0.001 C hspX tgs1 H 0.01 50 H C 10 6A 0.1 100 4A HC106A 10 1 HC104A H C 10 4A c % inhibition of TAG accumulation (relative to DMSO control) Fold change (relative to DMSO control) a NO Figure 3.3. Inhibition of DosR regulon and persistence-associated physiologies by HC104A and HC106A. (a) Inhibition of DosR regulon in murine macrophages infected with Mtb and treated with HC104A and HC106A for 48 h. Bacterial RNA was isolated after incubation, and the differential gene expression of hspX and tgs1 bone-marrow derived were quantified by qRT-PCR. The error bars represent the standard derivation of three biological replicates. (b) HC104A and HC106A inhibit DosR regulon induction by vitamin C and NO. Cells were pretreated with the compounds or DMSO for 24 h, and induced with vitamin C or NO for 2 h. Total bacterial RNA was isolated, and the transcripts of DosR-regulated genes, hspX and tgs1, were quantified by qRTPCR. The difference in the drug treated samples compared to DMSO treated samples in response to vitamin C or NO are significant with a P-value <0.001 based on a t-test, except the one marked as non-significant (n.s.). The error bars represent the standard deviation of three replicates. The experiment was repeated twice with similar results. (c) Inhibition of TAG accumulation of Mtb treated with HC104A or HC106A. Mtb cells were treated with 40 µM of the compounds and labeled with [1,2-14C] sodium acetate for 6 d. Total lipid was isolated and analyzed by TLC. TAG accumulation was quantified from the TLC. The error bars represent the standard derivation of two biological replicates. (d) Mtb cell survival during NRP when treated with HC104A or HC106A during NRP. Mtb cells were pretreated with 40 µM of compounds for 48 h in an anaerobic chamber, and continued incubation for 10 d. Surviving bacteria were enumerated on 7H10 agar. The error bars represent the standard derivation of three biological replicates. The experiment was repeated twice with similar results. 102 Inhibition of Mtb persistence-associated physiologies Mtb accumulates TAG during hypoxia (25-27). DosR directly regulates tgs1, which encodes a TAG synthase that is involved in the last step of TAG biosynthesis and is required for TAG accumulation during hypoxia (28). Transcriptional profiling showed that HC104A and HC106A repress expression of tgs1. Based on this transcriptional profiling, we hypothesized these two compounds may inhibit TAG biosynthesis during NRP. To test our hypothesis, Mtb cells were radiolabeled with 14C-acetate and treated with HC104A or HC106A for 6 d. Lipids were isolated and analyzed by thin layer chromatography (TLC). As previously observed, DMSO treated CDC1551(DdosR) mutant displayed a strong (87%) reduction of TAG accumulation as compared to DMSO treated WT (Fig. 3.3c and Supplementary Fig. 3.2). Mtb cells treated with HC104A or HC106A showed a ~50% reduction of TAG accumulation, supporting our hypothesis that the compounds can inhibit TAG biosynthesis. DosRST has been previously reported to be required for survival during NRP, where deletion of DosR causes greatly reduced survival during prolonged hypoxic stress (19). The impact of HC104A and HC106A on Mtb survival during NRP was examined using the hypoxic shiftdown model (25). Mtb survival was examined following 10 d of treatment with the compounds at 40 µM. The DdosR mutant control had 15% survival relative to DMSO, and was partially complemented, supporting the proposal that survival during hypoxia is DosR dependent. Mtb cells treated with HC106A displayed 50% survival relative to DMSO control (Fig. 3.3d), whereas, HC104A had no impact on Mtb survival during NRP, an observation that suggests the portion of the DosR regulon inhibited by HC104A is not essential for survival during NRP. 103 HC104A inhibits DosR DNA binding There are several potential targets of HC104A and HC106A to directly inhibit DosRST signaling, including: 1) direct inhibition of DosS/T sensor kinase activity, 2) modulation of the heme in the sensor, or 3) inhibition of DosR binding of DNA. To investigate the biochemical mechanisms of action of HC104A and HC106A, inhibition of DosS autophosphorylation was initially evaluated. The DosS protein was treated with different concentrations of HC104A and HC106A from 10 µM to 40 µM, or with 40 µM HC103A as a positive control that was previously discovered to be a DosS/T autophosphorylation inhibitor (25). As previously observed, HC103A strongly inhibited DosS autophosphorylation, but HC104A and HC106A had no inhibitory activity (Supplementary Fig. 3.3). This suggests HC104A and HC106A are not directly inhibiting DosS/T autophosphorylation activity. Next, a UV-visible spectroscopy assay was employed to investigate if HC104A targets to the heme of the sensor kinase DosS. Treatment of DosS protein with the reducing agent dithionite (DTN) caused a shift of the Soret peak from 403 nm to 430 nm as previously described (13, 25). Addition of HC104A to reduced DosS did not shift the peak to the oxidized position, suggesting that HC104A does no modulate DosS heme redox (Supplementary Fig. 3.4). Together, these data indicate that HC104A does not inhibit DosRST signaling by targeting sensor kinase autophosphoryation or heme and supported examining DosR as a potential target. Inspection of the HC104A structure revealed it had significant similarity to the compound virstatin (29). Virstatin is an anti-virulence compound that inhibits Vibrio cholera cholera toxin production by targeting the transcription regulator ToxT (29). Virstatin inhibits ToxT protein dimerization and subsequently interferes with DNA binding, thereby inhibiting the transcription of downstream genes involved in toxin production. Based on the similarity of chemical structure 104 between HC104A and virstatin, we hypothesized that HC104A may be targeting DosR, and interfering with DNA binding. To test this hypothesis, an electrophoretic mobility shift assay (EMSA) was employed to investigate the impact of HC104A on DosR DNA binding. Recombinant DosR protein, ranging from 0.5 µM to 4 µM, was treated with 40 µM HC104A or a DMSO control and tested for binding to fluorescently labeled hspX promoter DNA. In the DMSO treated control, DosR bound promoter DNA beginning at a concentration 2 µM DosR protein (Fig. 3.4a). Treating the reaction containing 2 µM DosR protein with HC104A significantly inhibited DNA binding by ~22-fold compared to DMSO control (Fig. 3.4b). To further characterize the impact of HC104A on DosR binding of DNA, a dose-response study was performed. Reactions containing 2 µM recombinant DosR proteins were treated with different concentrations of HC104A or virstatin ranging from 1 – 80 µM. HC104A inhibited DosR binding of DNA beginning at 10 µM HC104A. The fraction of free DNA increased as HC104A concentration increased (Fig. 3.4c). For example, the fraction of free DNA was 72%, 89%, 92%, 96% and 100% for 20 µM, 30 µM, 40 µM, 60 µM and 80 µM HC104A, respectively, whereas DMSO control had 15% free DNA (Fig. 3.4d). Thus, HC104A significantly inhibits DosR-DNA binding in a dose-dependent manner. Reactions treated with virstatin had no impact on DosR binding of DNA (Supplementary Fig. 3.5a). Consistent with these observations, virstatin did not have any impact on DosRST signaling in the whole cell Mtb fluorescence reporter assay (Supplementary Fig. 3.5b). These findings support the hypothesis that HC104A may function by inhibiting DosR DNA binding activity and has an activity that is distinct from the related molecule virstatin. 105 a 1 2 3 4 0.5 1 2 3 150 4 μM DosR % Free DNA 0.5 b HC104A DMSO DMSO HC104A ** 100 50 0 0 4. 0 3. 0 2. 5 0. 1. 0 0 4. 0 3. 0 2. 1. 0. 5 0 DosR (µM) d c 100 HC104A (μM) 1 2.5 5 10 20 30 40 60 80 % Free DNA 0 * ** ** ** 50 0 0 20 40 60 80 HC104A (µM) Figure 3.4. Inhibition of DosR DNA-binding by HC104A. (a) DosR protein ranging from 0.5 µM to 4 µM was treated with DMSO or 40 µM HC104A and binding to the hspX promoter was examined by EMSA. HC104A inhibits DosR DNA binding at 2 µM concentration. (b) The free DNA of each reaction was quantified in ImageJ and the percentage of free DNA was normalized using reactions containing 0.5 µM DosR as 100% free DNA. Differences between reactions containing 2 µM DosR treated with DMSO or HC104A are significant (**P value <0.005 based on a t-test). The error bars represent the standard derivation of two biological replicates. (c) Dose-dependent impact of HC104A on DosR DNA binding. DosR protein at 2 µM was treated with HC104A at concentrations from 1 µM to 80 µM. (c) The free DNA of each reaction was also quantified in ImageJ with the percentage of free DNA is normalized using the reaction containing 80 µM HC104A as 100% free DNA. The differences between treated reactions as compared to DMSO control are significant. (*P value <0.05 and **P value <0.005 based on a t-test). The error bars represent the standard derivation of two biological reps. 106 b 1.5 Absorbance Absorbance a DosS DTN 100 µM 1.0 0.5 400 1.0 0.5 450 500 550 0.0 350 600 0.5 HC106A 0.0 350 400 450 500 550 450 500 550 600 600 1 pVV16 DosS 0.1 100 µM 0.001 0.01 DosS_G117L DTN Fold change (relative to DMSO control) Absorbance d 1.0 400 Wavelength (nm) Wavelength (nm) c DosS DTN 100 µM CO HC106A 0.0 350 1.5 *** hspX tgs1 Wavelength (nm) Figure 3.5. Interactions between HC106A and DosS heme. WT DosS protein was treated with dithionite (DTN) and then 100 µM HC106A (a) or 100 µM pf CORM-2 (a CO donor) (b). The UV-visible spectra of the two treatments exhibited a shift of the Soret peak to a common position of 422 nm. (c) DosS with a G117L amino acid substitution, that is predicted to block the heme exposing channel, provides resistance to HC106A. The spectrum of the mutant protein did not change, after HC106A treatment, indicating resistance to HC106A. (d) Overexpression of DosS protein promotes resistance to HC106A treatment in Mtb. Mtb cells with the pVV16 empty vector or the DosS overexpression plasmid were treated with 20 µM HC106A for 6 d. Bacterial RNA was isolated for analysis of the differential gene expression of hspX and tgs1 and analyzed by qRTPCR. Overexpression of DosS caused 23- and 16.5-fold increase of hspX and tgs1 transcripts, respectively, compared to the empty vector control (***P value <0.0001 based on a t-test). The error bar represents the standard derivation of the mean for three technical replicates. The experiments were repeated twice with a similar result. 107 HC106A modulates DosS heme DosS and DosT have a channel that exposes the heme to the environment and enables interactions with gases (30, 31). This channel is an Archilles heel that can be targeted by small molecules. Previously, it was shown that the artemisinin modulates DosS/T by oxidizing and alkylating heme carried by the kinases (25). UV-visible spectroscopy studies were conducted to examine if HC106A modulated DosS heme. Recombinant DosS was purified from E. coli, degassed and the change of DosS heme spectrum was monitored under anaerobic conditions by UV-visible spectroscopy. Treating DosS with the reducing agent dithionite (DTN) caused the Soret peak to shift to 430 nm as shown previously (13, 25). HC106A was added to the reaction following DTN treatment to observe the impact on the DosS heme UV-visible spectrum. HC106A caused the DosS Soret peak to immediately shift to 422 nm, where the peak was stably maintained for 2 h (Fig. 3.5a). This spectrum shift is different from artemisinin, where under identical conditions, artemisinin causes the DosS Soret peak to gradually shift back to the oxidized state at 403 nm (25). These findings show that HC106A may also interact with sensor kinase heme, but via a mechanism that is distinct from artemisinin-heme interactions. The Soret peak at 422 nm is consistent with previously described spectra that are observed when DosS heme interacts with NO or CO (13). To confirm this observation, DosS was treated with 100 µM CORM-2 (a CO donor) which caused a shift of the Soret peak to 422 nm, similar to what was observed for HC106A (Fig. 3.5b). This finding supports a hypothesis that HC106A may also be directly binding to the heme. Notably, CO activates DosS kinase function, whereas HC106A functions to inactivate the regulon, demonstrating that the impact of heme binding by CO or HC106A has differing impacts on the sensor kinase switch. 108 Amino acid substitutions in the channel exposing the DosS heme to the environment, such as DosS E87L or G117L, can limit access of artemisinin to modulate heme (25). To support the hypothesis that HC106A directly targets DosS heme, we tested the impact of these amino acid substitutions on HC106A/DosS heme interactions. Treating DosS(E87L) with HC106A exhibited a profile similar to wild type DosS with the Soret peak shifting to 422 nm (Supplementary Fig. 3.6). However, DosS(G117L) had no change to the overall spectrum after HC106A treatment (Fig. 3.5c). This finding indicates that DosS(G117L) is resistant to HC106A, and confirms that HC106A accesses the heme via a similar mechanism as artemisinin. To confirm DosS is a target of HC106A in Mtb, we examined the impact of overexpressing DosS protein in Mtb. If DosS is the target of HC106A, overexpression may reduce the effectiveness of HC106A. WT DosS protein was constitutively expressed from the hsp60 promoter in Mtb. The vector control showed that both hspX and tgs1 genes were downregulated by HC106A by 2331- and 3470-fold, respectively (Fig. 3.5d). Overexpressing DosS provided significant resistance to HC106A, with hspX and tgs1 showing 23- and 16.5-fold less inhibition, respectively, relative to the empty vector control This observation of resistance in Mtb is consistent with the biochemical data supporting the view that DosS is a direct target of HC106A. Synergistic interactions between DosRST inhibitors With multiple distinct inhibitory activities of the HC101A, HC102A, HC103A, HC104A and HC106A, we sought to examine if potentiating or antagonistic interactions existed between the molecules, when targeting the DosRST pathway. To examine these interactions, checkboard assays were performed with pairwise comparisons of artemisinin (HC101A), HC102A, HC103A, HC104A and HC106A. CDC1551 (hspX’::GFP) was treated with combinations of two compounds 109 ranging from 50 µM to 0.08 µM in 96-well plates. DosR-driven GFP fluorescence and optical density were measured following 6 d of treatment. The Combination Index (CI) was calculated for each drug pair based on the Chou-Talalay method in the CompuSyn software package (32, 33), where CI values of < 1, = 1 and >1 indicate synergistic, additive or antagonistic interactions, respectively. Among all 64 compound pairs, artemisinin combined with HC102A, HC103A, HC104A and HC106, showed 46, 49, 41, and 50 combinations that have CI <1, respectively (Fig. 3.6). Notably, some CI values are below 0.1 when artemisinin was paired with HC102A, HC103A or HC106A combinations. Example dose response curves illustrate these synergistic interactions (Fig. 3.6). Several other pairwise comparisons also demonstrated synergy (Supplementary Fig. 3.7), however, in general, these interactions had CI between 1 and 0.1, supporting weaker synergistic interactions, as compared to combinations with artemisinin. Overall, these studies provide the evidence that the inhibitors function by distinct mechanisms and may be combined to improve potency. 110 b HC102A (µM) 50 Artemisinin (µM) 50 50 20 20 88 3.2 3.2 1.28 1.28 0.512 0.51 0.205 0.21 0.082 0.08 0.1 0.04 0.02 0.01 0.01 0.02 0.06 0.2 20 0.26 0.1 0.05 0.02 0.02 0.02 0.14 0.5 8 0.75 0.2 0.09 0.05 0.03 0.06 0.51 1.03 3.2 3.2 1.17 0.42 0.2 0.1 0.06 0.19 1.29 1.91 2.2 0.81 0.35 0.18 0.11 0.23 1.47 3.2 2.48 0.86 0.44 0.23 0.16 0.37 2 4.57 3.35 1.23 0.54 0.29 0.2 0.92 3.47 8.52 4.4 1.46 0.58 0.3 0.2 0.73 4.38 8.89 5050 2020 88 3.2 3.2 1.28 1.28 0.51 0.512 0.21 0.205 0.08 0.082 5050 0.12 0.1 0.08 0.09 0.08 0.11 0.41 0.45 2020 0.09 0.05 0.05 0.04 0.04 0.06 0.66 0.81 88 0.18 0.05 0.04 0.04 0.04 0.13 0.55 1.07 3.2 3.2 0.24 0.11 0.06 0.04 0.05 0.14 1.09 1.37 1.28 1.28 0.6 0.25 0.12 0.07 0.07 0.16 1.4 1.87 % inhibition 50 Art HC102A Art + 8 µM HC102A 50 0 0.01 100 20 20 0.52 0.36 0.3 0.27 0.35 0.44 0.99 1.33 88 1.39 0.44 0.34 0.26 0.35 0.73 1.51 1.76 3.2 3.2 1.3 0.52 0.4 0.21 0.29 0.56 1.33 1.73 1.28 1.28 1.89 0.75 0.45 0.24 0.28 0.54 1.64 1.92 0.51 0.51 2.88 1.02 0.5 0.27 0.3 0.74 1.7 2.12 0.21 0.21 4.07 0.99 0.53 0.3 0.24 0.66 2.65 3.83 0.08 0.08 4.06 1.36 0.61 0.29 0.22 0.58 3.22 10.4 % inhibition 100 100 5050 2020 88 3.2 3.2 1.28 1.28 0.512 0.51 0.205 0.21 0.082 0.08 5050 1.16 1.11 1.03 1.09 1.06 0.97 1.04 0.95 2020 0.5 0.5 0.45 0.45 0.45 0.46 0.43 0.45 88 0.33 0.23 0.2 0.19 0.19 0.18 0.2 0.35 3.2 3.2 0.2 0.12 0.09 0.09 0.08 0.09 0.31 1.44 1.28 1.28 0.24 0.09 0.06 0.05 0.04 0.1 0.9 0.85 0.51 0.51 0.07 0.11 0.07 0.37 0.03 0.04 0.47 0.44 0.21 0.21 0.89 0.26 0.14 0.08 0.05 0.07 1.58 1.72 0.08 0.08 2.03 0.66 0.3 0.16 0.15 0.16 6.67 13.5 100 50 Art HC104A Art + 3.2 µM HC104A Art + 1.28 µM HC104A 0 0.01 0.08 0.08 2 0.89 0.49 0.26 0.22 1.12 9.62 18.5 HC106A (µM) 0.1 1 10 100 % inhibition Artemisinin (µM) 50 50 0.89 0.67 0.66 0.72 0.64 0.63 0.62 0.6 0.1 1 10 Concentration [µM] d HC104A (µM) Artemisinin (µM) 0.1 1 10 Concentration (µM) c 50 50 20 20 88 3.2 3.2 1.28 1.28 0.512 0.51 0.205 0.21 0.082 0.08 0.21 0.21 2.44 0.65 0.42 0.21 0.19 0.73 7.04 8.91 Art HC103A Art + 3.2 µM HC103A Art + 1.28 µM HC103A Art + 3.2 µM HC102A 0 0.01 0.51 0.51 1.5 0.42 0.24 0.13 0.25 0.7 3.66 3.34 100 100 % inhibition HC103A (µM) 1.28 0.51 0.51 0.21 0.21 0.08 0.08 1.28 Artemisinin (µM) a Art HC106A 50 Art + 0.21 µM HC106A Art + 0.08 µM HC106A 0 0.01 Concentration (µM) 0.1 1 10 Concentration (µM) 100 0 1 CI 10 Figure 3.6. Synergistic interactions between DosRST inhibitors. CDC1551 (hspX’::GFP) was treated with pairwise combinations of two compounds at concentrations of 50 µM to 0.08 µM. GFP fluorescence was measured and used to calculate percentage inhibition. The data were analyzed in the CompuSyn software to determine the combination index (CI) for the panel of each drug combination, including (a) artemisinin and HC102A; (b) artemisinin and HC103A; (c) artemisinin and HC104A; and (d) artemisinin and HC106A. Example EC50 curves are presented with individual compounds or a selected synergistic combination to illustrate the potentiating interactions. 111 Structure-activity relationship study for HC104A and HC106 We conducted a catalog search for HC104 and HC106 analogs and obtained 10 commercial analogs for each series to define initial structure activity relationships (SAR). For HC104A we observed that a bromine in the 5-position is required for activity and that the R2 dimethylamine group is not required (Supplementary Table 3.1). For example, HC104B is identical to HC104A except for the removal of the bromine (R1), which results in a complete loss of activity in the whole cell assay. Whereas, replacement of the R2 group with a methyl (HC104G) results in an active compound, although ~5-fold less active than HC104A. Although not highly potent, its ligand efficiency, cLogD and druglikeness are in the range of what would be considered acceptable to good as a starting point for further manipulation. For HC106A (Supplementary Table 3.2), catalog SAR work led to new understandings of the nature of the series. We first found that the simple removal of an ortho chloro on the “A” ring of HC106A leads to ~ 2-fold enhanced activity, with an EC50 in the whole cell Mtb assay for DosRST inhibition of 1.33 µM (HC106F). It was also found that the use of an alternative isomer of the isoxazole had no activity (HC106C). To further understand the SAR of the HC106 series, additional analogs were synthesized to examine the need of the central urea functionality and whether modifications can be tolerated (Supplementary Table 3.3). A pyridyl analog (MSU-41425), designed to replace the isoxazole also demonstrated no activity as was the symmetrical 4-chloroaniline derived urea (MSU41324). However, the bis-isoxazole urea (MSU-39444) provides an EC50 of 1.7 µM, indicating that the isoxazole is important for binding. Isoxazoles are unique among heterocycles in that they exist in multiple tautomeric forms as supported by initial NMR studies (34). We next explored the need of one of the -NHs of the urea, capping it with a methyl (MSU-39451), integrating it into a ring for conformational restriction (MSU-39453), and replacing with a methylene unit 112 (MSU-39449). In all cases, reduced activity (0.5 - 1 log) was observed but not all activity was lost. Thus, HC106A is a potent whole cell inhibitor of the DosRST pathway, with flexibility to be improved via SAR. To further test the SAR, we conducted a Topliss Tree evaluation of the “A-ring” aniline (35). To reliably prepare the derivatives, we explored and established a general preparation (Supplementary Fig. 3.8). This route is preferred relative to alternative approaches for its cleanliness, yields and ease of purification, usually by trituration. It is also anticipated that it will allow access to future derivatives. Using HC106F and HC106A as starting points, we prepared the 3,4-diclorochloro and 3-chloro derivatives (MSU-39452 and MSU-39445, respectively). Both the 3- and 4-chloro derivatives demonstrated greater activity than 3,4dichloro (MSU-39452). We found that replacing the 4-chlorophenyl ring with pyridyl analogs (MSU-39448 and MSU-39450) lead to increased activity. Focusing on 4-position derivatives, we found that fluoro (MSU-39446), bromo (MSU-41464) and methoxy (MSU-39447), as electron p-orbital donating substituents, also lead towards increased activity. Para-t-butyl phenyl (MSU-41442), also did the same. Electron withdrawing substituents, such as 4-CO2Me (MSU4165), 4-trifluoromethyl (MSU-41463) and biphenyl (MSU-41443) saw similar activity. Overall, several analogs were discovered with significantly ~4-fold enhanced potency, with several inhibitors having whole cell DosRST inhibitory EC50 below 1 µM, including EC50 of 0.63 µM, 0.61 µM, 0.54 µM and 0.75 µM for MSU-33189, -39447, -39446, and -39455. respectively. The parent analog HC106A had an EC50 of 2.48 µM. Kinetic solubility assays were conducted for selected analogs and all exhibited excellent aqueous solubility greater than >100 µM, except for MSU-41443 (Supplementary Table 3.3). This finding shows that the urea group present in the HC106A does not have a detrimental impact on HC106 aqueous solubility. 113 Overall, the nanomolar whole cell potency, flexible SAR and excellent solubility, confirm that the HC106 series is a suitable series for continued optimizations to identify a drug-like lead compound. Discussion DosRST is a two-component regulatory system required for Mtb environmental sensing, adaptation, and persistence. By using a fluorescent reporter strain CDC1551 (hspX’::GFP), we have previously discovered small molecule inhibitors of DosRST, named HC101A-HC106A, from the whole-cell phenotypic HTS. Here, we report the characterizations of HC104A and HC106A as DosRST inhibitors. Both compounds downregulated genes in the DosR regulon, and inhibited TAG biosynthesis. HC106A also reduced Mtb survival during NRP. In a UV-visible spectroscopy assay, we observed that HC106A interacts with the heme of sensor kinase DosS. The UV-visible spectrum of HC106A-treated DosS is similar to those of COor NO-treated DosS. The overlap between the CO and HC106A spectra supports that HC106A may also directly bind to the heme of DosS. Interestingly, CO activates sensor kinases, whereas HC106A inhibits them. This could be due to the difference in conformational changes induced by CO and HC106A, or binding of HC106A may lock the sensor kinases into an inactive state. Furthermore, the DosS G117L substitution, recombinant DosS blocks the heme exposing channel, provides resistance to HC106A. This means that, similar to artemisinin, this channel is also important for the activity of HC106A. These findings provide additional evidence that the hemeexposing channel in DosS/T can be exploited by small molecules to prevent or inhibit the heme from sensing signals and to disrupt signal transduction of a two-component regulatory system. To 114 our knowledge, these two distinct mechanisms of actions are novel mechanisms to inhibit the DosRST pathway. Mechanistic studies via EMSA indicates that HC104A may function by targeting DosR and inhibiting DosR DNA binding. Virstatin had no effect on DosR DNA binding and no impact on the DosRST signaling in whole cells, showing that although both compounds share a similar structure, they function by distinct mechanisms. From the catalog SAR study, we found that the bromine group is required for the compound. Virstatin does not have the bromine, and the R-group is butyric acid instead of dimethylamine. These two differences are enough to differentiate the activity of the compounds. Moreover, transcriptional profiling shows that the most repressed genes by HC104A are from the DosR regulon, providing additional evidence that HC104A is somewhat selective for DosR regulated genes. Interestingly, the genes most downregulated by HC104A, including hspX, Rv2030c, pfkB, and Rv2028c, are from the same operon under control of hspX promoter, which is strongly induced by DosR in hypoxia. This result suggests that HC104A is more specific to target hspX operon genes as compared to other DosR regulated genes. This finding leads to the speculation that HC104A may be more efficient to prevent DosR binding to the hspX promoter than the other DosR promoters. HC104A may fit better in the pocket in the interface of DosR-hspX’ complex. This postulation also supports the idea that HC104A may not have an impact on DosR protein dimerization, which would lead to universal downregulation of DosRregulated genes. Detailed characterizations of HC104A on inhibiting DosR dimerization and promoter binding specificity is the subject of ongoing characterizations. This study together with the previous report reveals multiple distinct inhibitory mechanisms for the five compounds, HC101A-104A, HC106A. Four out five compounds, HC101A-HC103A and HC106A, are proposed to function by targeting the sensor kinases 115 DosRST. Furthermore, these four compounds are effective at decreasing Mtb survival during NRP in hypoxic-shift down assay. In contrast, the compound targeting DosR, HC104A, had no impact on Mtb survival during NRP. This result is consistent with a previous report that showed only the DdosS mutant exhibits a survival defect in the C3HeB/FeJ mouse model (22). Notably, the dosR mutant strain used in our study is also lacking dosS expression due to polar impacts of the deletion, supporting the hypothesis that the reduced survival during NRP may be dependent on DosS. Alternatively, we noted that HC104A only inhibits a portion of the DosR regulon with only 26 genes being downregulated. This finding suggests that the DosR regulated genes that are inhibited by HC104A (Fig. 3.2, and Supplementary Dataset 3.1, 3.3) are not required for DosRST-dependent persistence during NRP or not sufficiently inhibited to have an impact on survival. The transcriptional profiling data showed off-target impacts in the treated dosR mutant strain. These effects may be due to inhibition of DosS/T signaling that functions independent of the response regulator. Several recent studies have demonstrated the occurrence of crossinteractions between histidine kinases and response regulators in Mtb. For instance, Lee et al. shows that DosT can interact with the other non-cognate response regulators, including NarL and PrrA (36). Our transcriptional profiling of the compound treated dosR mutant strain suggests that some genes downregulated by HC101A, HC103A and HC106A may be DosS-dependent but DosR-independent (Supplementary Fig 3.9). In the prior study, six genes are similarly regulated between artemisinin and HC103A in the treated dosR mutant strain, including Rv0260c that encodes a putative response regulator (25). HC106A and HC103A share four differentially regulated genes in the dosR mutant, including argC, argJ, argB, and argF, which are genes involved in arginine biosynthesis. Our data presented here and in the literature point to the possibility that DosST modulate gene expression independently of DosR. 116 The synergy studies show significant synergistic interactions between artemisinin, HC102A, HC103A, HC104, or HC106A. Moreover, artemisinin exhibited the greatest synergistic activities with HC102A, HC103A or HC106A, indicating that inhibition of histidine kinases by a second inhibitor can lead to synergistic inhibition of the DosRST pathway. This interaction could be due to both sensor kinases being required for full induction of the DosR regulon, where DosT responds early during hypoxia and DosS further induces the regulon at later times (37), and it is possible that the inhibitors have different affinities for DosS or DosT. Thus, multiple inhibitors could inhibit both DosS and DosT better than an inhibitor alone. Interestingly, artemisinin shows the greatest synergism with HC106A. Both compounds are proposed to target the heme of DosS/T, but through different mechanisms. This finding shows that both inhibitors can enter the channel of DosST to interact with the heme and do so without antagonizing interactions. The identification of new antibacterial agents and tuberculosis drugs has been associated with the realization that these compounds can occupy a different region of chemical space relative to drugs in most other therapeutic areas. Series HC106 is easily in the range of the characteristics for compounds currently in use, in development and anti-TB compounds. Series HC104 needs further exploration to fully assess its suitability. The physicochemical properties of a drug, such as solubility and permeability, impact its oral bioavailability as these factors influence absorption, distribution, metabolism, and excretion. Series HC106 demonstrates excellent aqueous solubility (Table 3.3, with compounds generally having solubility >100 µM). The melting points of most of the compounds in the series and the corresponding cLogP are also supportive appropriate solubility. The ability of an anti-TB drug to reach its target site is greatly hampered by the highly impermeable Mtb cell envelope. Both series (HC104 and HC106) demonstrate low micromolar and nanomolar whole cell EC50 values, suggesting adequate cell wall permeability. SAR studies 117 of HC106 show that there remain significant opportunities for optimization of HC106 for potency and drug properties and future studies will be focused on such optimizations. Materials and methods Bacterial strains and growth conditions. Mtb CDC1551, CDC1551 (DdosR) strains were used in this study. All cultures were grown at 37°C and 5% CO2 in 7H9 Middlebrook medium supplemented with 10% OADC (oleic acid albumin dextrose catalase) and 0.05% Tween-80 in standing, vented tissue culture flasks, unless stated otherwise. EC50 assays. The assay was performed as previously described (25). Briefly, the (hspX’::GFP) reporter strain culture was diluted to an OD600 of 0.05 in fresh 7H9 media, pH 7.0, and 200 µL of diluted culture was aliquoted in clear-bottom, black, 96-well plates (Corning). Cells were treated with an 8-point (2.5-fold) dilution series ranging from 200 µM – 0.32 µM. For the structure relationship studies for the HC106 series, a 12-point (2.5-fold) dilution series of HC106 analogs ranging from 200 µM – 8.4 nM were used. GFP fluorescence and optical density were measured following 6 d incubation. Percentage fluorescence and growth inhibitions were normalized to a rifampin-positive control (100% inhibition) and DMSO-negative control (0% inhibition). EC50 values were calculated for each compound using GraphPad Prism software package (version 6). Each experiment was performed with two technical replicates per plate and two biological replicates, and the error bar represents the s.d. of the biological replicates. Experiments were performed twice with similar results. 118 Transcriptional profiling and data analysis. Transcriptional profiling studies were conducted as previously described in Zheng et al. (25). Briefly, CDC1551 or CDC1551 (DdosR) cultures were treated with 40 µM HC104A, HC106A or DMSO control for 6 d. The starting OD600 was 0.1 in 8 mL of 7H9 medium in standing T25 vented tissue culture flasks. Bacterial growth consumes oxygen and stimulates the DosRST pathway. The total bacterial RNA from two biological replicates was isolated and prepared for sequencing as previously described (38). The RNA-seq data were processed and analyzed using the SPARTA software package (39). Real-time PCR assays. The vitamin C and NO assays were performed as previously described (25). Briefly, cultures at an OD600 of 0.6 were pretreated with 80 µM HC104A, HC106A or a DMSO control for 24 h, and induced with 50 µM DETA-NONOate or 20 mM vitamin C for 2 h. For the HC106A resistance assays, CDC1551 was transformed with the empty replicating plasmid pVV16 or the plasmid expressing dosS from the strong hsp60 promoter (pVV16-DosS), and treated with 20 µM HC106A for 6 d. Total bacterial RNA was isolated and differential gene expression of DosR-regulated genes, including hspX and tgs1, was quantified. The experiment was performed in three technical replicates and error bars represent the s.d from the mean. The experiment was repeated twice with similar results. To examine Mtb gene expression in macrophages, murine bone-marrow derived macrophages were isolated as previously described (40) and seeded in T75 vented, tissue culture flasks. Macrophages were infected with CDC1551 with multiplicity of infection ratio of 1:20 as previously described (40). After infection, the flasks were treated with 40 µM HC104A or HC106A or DMSO for 48 h, with three individual flasks for each treatment. Total bacterial RNA was isolated after treatment, and the transcripts of DosR- 119 controlled genes (hspX and tgs1) were quantified in RT-PCR. The experiment was conducted with three biological replicates. The error bar represents the s.d. of the biological replicates. TAG biosynthesis. The lipid labelling and TAG TLCs were performed as previously described (25). Briefly, CDC1551 was cultured at an initial OD600 of 0.1 and radiolabeled with 8 µCi of [1,214 C] sodium acetate in T25 vented tissue culture flasks. The cultures were treated with 40 µM HC104A, HC106A or DMSO for 6 d at 37°C. CDC1551 (DdosR) and dosRS complement strains were also examined. Total lipids were extracted and 14 C incorporation was determined by scintillation counting. 20,000 c.p.m. of total lipids were analyzed by TLC using silica gel 60 aluminum sheets (EMD Millipore). To determine TAG accumulation, the lipids were developed in hexane-diethyl ether-acetic acid (80:20:1; vol/vol/vol) solvent system. The TLC was exposed to a phosphor screen for 3 d, and imaged on a Typhoon imager and TAG was quantified using ImageJ software (41). The experiment was repeated twice with similar results, and the error bar represents the s.d. of two biological replicates. NRP survival assays. Survival during NRP was examined using the hypoxic shift down assays as previously described (25, 42). Briefly, CDC1551 cells were treated with 40 µM HC104A, HC106A or DMSO control in a 24-well plate (1 mL/well). CDC1551 (DdosR) and dosRS complement strains were also examined. Plates were incubated in an anaerobic chamber (BD GasPak) for 12 d. It took 48 h for cultures to become anaerobic, as monitored by a methylene blue control. Bacterial CFUs were numerated on 7H10 agar plates following incubation. The experiment was repeated twice with similar results. 120 DosR protein purification. DosR full length protein was purified as previously described (43). Briefly, the dosR gene (Rv3133c) was cloned into pET15b (Novagen Darmstadt, Germany) using the primer set: forward primer 5’-TTTCATATGGTGGTAAAGGTCTTCTTGGTCGATGAC-3’; reverse primer 5’-TTTGGATCCTCATGGTCCATCACCGGGTGG-3’. The His6-DosR protein was expressed in E. coli BL21(DE3) strain. The culture was grown to OD600 0.5-06, and induced with 1 mM IPTG for 6.5 h at 29°C. The cell pellet was suspended in lysis buffer (20 mM TrisHCl, pH 8.0, 10% glycerol, 500 mM NaCl, 0.5 mg/ml lysozyme and 0.1 mg/ml PMSF), and incubated at 37°C for 30 min. The soluble fraction of lysate was collected after centrifugation and applied to a TALON metal affinity Co2+ column (Clontech). The column was washed twice with washing buffer (20 mM Tris-HCl, pH 8.0, 10% glycerol, 500 mM NaCl) without imidazole, then with 20 mM imidazole. The protein was eluted with the same buffer containing 300 mM imidazole. The fractions with the most DosR protein (as determined by SDS-PAGE) were pooled together for dialysis in 25 mM Tris-HCl, pH 8.0. The final protein concentration was determined using a Qubit kit (Invitrogen). Electrophoretic mobility assay. The assay is fluorescence-based using 6-carboxyfluorescein (6FAM) labeled 385 bp probe from the hspX promoter. In designing the primer set, 6FAM was added to the 5’ ends of forward and reverse primers. The hspX probe was synthesized via PCR using the primer set: forward primer 5’-6FAM-CAACTGCACCGCGCTCTTGATG-3’; reverse primer 5’-6FAM-CATCTCGTCTTCCAGCCGCATCAAC-3’. The probe was purified by Qiagen PCR purification kit. The DosR protein was pre-phosphorylated in 10 µL of phosphorylation buffer (40 mM Tris-HCl, pH 8.0, 5 mM MgCl2, 50 mM lithium potassium acetyl phosphate), and incubated at room temperature for 30 min. The protein was then transferred to binding buffer in a 121 final volume of 20 µL (final concentration, 25 mM Tris-HCl, pH 8.0, 0.5 mM EDTA, 20 mM KCl, 6 mM MgCl2, 10 nM probe, 1 µg poly-dI-dC (Sigma Aldrich)), and treated with HC104A or an equal volume of DMSO or virstatin (Santa Cruz Biotech). Two different assays were performed. Firstly, different DosR protein concentrations from 0.5 µM to 4 µM were treated with 40 µM. Second, dose response assays were performed with 2 µM DosR treated with different concentrations of HC104A or virstatin from 1 µM to 80 µM. After incubating on ice for 30 min, the reactions were terminated by adding 1 µL 80% glycerol, and loaded on a native 5% Tris/Borate/EDTA (TBE) polyacrylamide gel. The gel was run at 50 V, for 5-6 h at 4°C in 1X TBE buffer, and was imaged using a Typhoon scanner with appropriate filters that can detect florescence at excitation = 495 nm, emission = 520 nm. Binding of the unbound probe was quantified using ImageJ (41). The assay was repeated at least twice with similar results. The error bar represents the s.d. of two biological replicates. UV-visible spectroscopy assay. DosS and mutant proteins were purified and analyzed as previously described (25). Briefly, 7.5 µM of recombinant DosS protein was deoxygenated with argon gas in a sealed cuvette. The protein was reduced with 400 µM DTN for 20 min. The reaction was then treated with 100 µM HC106A, 400 µM HC104A, 100 µM CORM-2 (tricarbonyldichlororuthenium (II) dimer) or equal volume of DMSO. The UV-visible spectra were recorded for kinetic changes over 2 h. The experiment was repeated at least twice with similar results. Checkboard synergy studies. The reporter strain CDC1551 (hspX::GFP) was treated with pairs of DosRST inhibitors from 50 µM – 0.08 µM in 96-well plates, including HC101A-HC104A and 122 HC106A. GFP fluorescence and OD600 were measured after 6 d incubation. The percentage of fluorescence inhibition (FI) and growth inhibition were calculated for each drug pair, with limited growth inhibition observed. The FI data was utilized for further analysis of interactions using CompuSyn software (32). The Combination Index (CI) value was calculated for each drug pair according to the Chou-Talalay method, which is based on the Median-Effect equation derived from the Mass-Action Law principle (33, 44). The resulting CI values provide quantitative determination of drug interactions, including synergism (CI < 1), additive effect (C = 1), and antagonism (C > 1). Autophosphorylation assay. The DosS autophosphorylation assay was performed as previously described (25). Recombinant DosS protein was treated with 10 µM, 20 µM or 40 µM of HC104A, or HC106A. DMSO and 40 µM HC103A were also included as positive and negative controls, respectively. Kinetic solubility assay. The assay was performed with 7-point (2-fold) dilutions from 200 µM 3.125 µM for HC106 analogs. Mebendazole, benxarotene and aspirin were also included as controls. The drug dilutions were added to PBS, pH 7.4, with the final DMSO concentration of 1%, and incubated at 37 °C for 2 h. The absorbance at 620 nm was measured for each drug dilution to estimate of the compound solubility. Three replicates were examined for each dilution. Acknowledgements The MSU RTSF provided technical support for the RNA-seq library preparation and sequencing. This project was supported by start-up funding and a Molecular Discovery Grant from Michigan 123 State University, AgBioResearch, a grant from the NIH-NIAID (R21AI105687) and Grand Challenges Explorations grants from the Bill and Melinda Gates Foundation (OPP1059227 and OPP1119065). 124 APPENDIX 125 APPENDIX Supplementary information Supplementary Dataset 3.1. Differential gene expression data of WT Mtb treated with inhibitors Supplementary Dataset 3.2. Differential gene expression data of the DosR mutant treated with the inhibitors. Supplementary Dataset 3.3. Complete gene expression tables for transcriptional profiling experiments. 126 N N N O Compound N O O O O N O Br ID# MW (g/mole) EC50 (μM) Ligand Efficiency CLogP (cLogD at pH 7.4) Druglikeness Br HC104A 361.2 9.8 HC104G 290.1 43.8 HC104B 282.3 >200 0.34 0.36 0.26 3.0 (0.55) 2.3 (2.4) 2.9 (-0.24) 4.4 0.55 6.2 Supplementary Table 3.1. Structure and properties of HC104 series. Catalog structure activity relationship study performed for HC104 analogs with different R-groups. The reporter strain CDC1551 (hspX’::GFP) was treated across doses of each analog from 200 µM to 0.328 µM. The EC50 values of fluorescence inhibition calculated for each analog to determine their potency. The other chemical properties of the analogs are also included. 127 ID# Compound Cl N H Cl N H Cl Cl N H N H Cl MSU41425 O N N O O HC106C Cl MSU39444 N H O HC106A MSU39450 O N O HC106F N H O N O N N H N H N O O O N N H Cl N H ID# 1.33 MSU41324 2.48 MSU39449 >200 MSU39451 1.95 MSU39453 1.7 MSU41422 EC50 (μM) Compound Cl Cl O N H >200 N H Cl O N O N H Cl O N O N Me N H 16.62 N H Cl O N H 4.14 O N O N 5.2 O N >200 Cl O N H EC50 (μM) N H N >200 Supplementary Table 3.2. Initial SAR studies of the HC106 series. The HC106 analogs with different R-groups were synthesized or purchased. The reporter strain CDC1551 (hspX’::GFP) was treated with across doses of each analog from 200 µM to 0.328 µM. The EC50 values of fluorescence inhibition calculated for each analog to determine their potency. 128 O N O R EC50 (μM) MW (g/mol) TPSA (ang2) Ligand Efficiency (LE) m.p. (°C) cLogP Solubility (μM) * 1.33 237.6 63 0.54 183 2.2 >200 * 2.48 272.1 63 0.63 203.1 63 0.40 183184 1.6 0.61 233.2 72 0.36 166167 1.5 * 1.42 238.6 75 0.38 158159 1.3 * 1.95 238.6 75 0.38 122123 1.3 * 0.54 221.2 63 0.38 176177 1.7 * 1.2 282.1 67 183184 2.3 0.75 237.6 63 0.37 159 (dec) 2.1 2.47 272.1 63 0.36 164 2.8 2.08 259.3 63 163 3.1 >100 11.2 279.3 63 193 3.2 14 1.67 271.2 67 178 2.4 1.16 261.2 94 194 1.5 1.12 217.2 67 128130 1.3 Compound R= ID# N H Cl HC106F N H Cl HC106A N H Cl MSU33189 MSU39447 MeO * N H MSU39448 Cl MSU39450 Cl N N H N N H F MSU39446 MSU41464 * N H 164 N H Br N H 110 >100 >100 Cl MSU39445 * N H >100 Cl MSU39452 MSU41442 Cl t-Bu * N H MSU41443 MSU41463 * N H * N H F 3C * N H O MSU41465 MSU41462 Me O N H N H * * Supplementary Table 3.3. Initial Topliss Tree evaluation of “A-ring” aniline of series HC106. The HC106 analogs with different R-groups were synthesized. The reporter strain CDC1551 (hspX’::GFP) was treated with across doses of each analog from 200 µM to 0.328 µM. The EC50 values of fluorescence inhibition were calculated for each analog to determine their potency. The other chemical properties of the analogs are also included. Kinetic solubility was experimentally determined. 129 Fold change (relative to DMSO control) 1 HC104A HC106A dosR hspX tgs1 0.1 0.01 0.001 0.0001 Supplementary Figure 3.1. Inhibition of the DosR regulon by HC104 and HC106A during hypoxia. Mtb cells were treated with 40 µM compounds for 6 d, and total bacterial RNA was isolated. The DosR regulated genes, dosR, hspX, and tgs1 were quantified in qRT-PCR. The error bars represent the standard derivation of three replicates. The experiment was repeated at least twice with similar results. 130 6A ΔD os R 10 5A HC 10 HC 4A 10 HC SO DM * Supplementary Figure 3.2. TLC of TAG reduction in Mtb treated with HC104A and HC106A. Mtb cells were treated with 40 µM of the compounds and labeled with [1,2-14C] sodium acetate in T25 vented tissue culture flasks for 6 d. Total lipid was isolated and analyzed BY TLC. The experiment was repeated twice with similar results. 131 HC105A HC104A 0 10 20 40 10 20 HC106A HC103A 40 20 40 40 μM Supplementary Figure 3.3. Autoradiograph examining the impact of HC104A and HC106A on DosS autophosphorylation. DosS protein was treated with 10 µM, 20 µM or 40 µM of the compounds, with DMSO and HC103A as positive and negative controls, respectively. The results show that HC104A and HC106A have no effect on DosS autophosphorylation. 132 0.2 DosS DTN HC104A 1.0 0.5 0.0 350 400 450 500 550 600 Absorbance Absorbance 1.5 DosS DTN HC104A 0.1 0.0 500 Wavelength (nm) 550 600 Wavelength (nm) Supplementary Figure 3.4. Investigation of interaction between HC104A and DosS. WT DosS treated with 400 µM HC104A shows no impact on shifting of the Soret peak in the UV-visible spectroscopy assay. The experiment was repeated at least twice with similar results. 133 b a 100 Virstatin FI Virstatin GI OH N O O % inhibition O 50 0 0.1 1 10 100 1000 Concentration (µM) c Virstatin (μM) 0 1 2.5 5 10 20 30 40 60 80 Supplementary Figure 3.5. The impact of virstatin on DosR DNA-binding and DosRST signaling in Mtb. (a) Chemical structure of virstatin. (b) Dose-response curve of virstatin shows no effect on inhibition of Mtb DosR-driven GFP fluorescence. (c) DosR protein at 2 µM was treated with 9 point dose response of virstatin from 1 µM to 80 µM. The reactions were analyzed on native PAGE gel. The experiment was repeated at least twice with similar results. 134 Absorbance 0.6 DosS_E87L DTN 0.4 100 µM 0.2 0.0 350 400 450 500 550 600 Wavelength (nm) Supplementary Figure 3.6. Investigating the interaction between HC106A and DosS heme. DosS E87L protein was treated with 100 µM HC106A after being reduced with DTN. The UV-visible spectra were recorded after each treatment, and showed no change on the overall spectrum compared to WT protein. The experiment was repeated at least twice with similar results. 135 HC103A (µM) HC102A (µM) 5050 2020 88 3.2 3.2 1.28 1.28 0.512 0.51 0.205 0.21 0.082 0.08 5050 0.14 0.19 0.21 0.3 0.44 0.43 0.43 0.43 2020 0.28 0.39 0.45 0.55 0.72 0.85 0.94 0.94 88 0.43 0.32 0.38 0.61 0.9 1.11 1.19 1.2 3.2 3.2 0.34 0.33 0.47 0.71 1.14 1.24 1.37 1.32 HC104A (µM) 1.28 1.28 0.38 0.41 0.62 0.99 1.22 1.28 1.4 1.27 0.51 0.51 0.44 0.58 0.91 1.18 1.55 1.35 1.59 1.18 0.21 0.21 0.5 0.75 1.13 1.49 1.87 1.86 3.61 1.51 0.08 0.08 0.49 0.82 1.33 1.89 2.99 4.2 13.8 5.84 100 5050 0.33 0.33 0.31 0.28 0.28 0.27 0.29 0.27 2020 0.53 0.73 0.89 1.06 1.11 1.11 1.12 1.09 88 0.85 1.06 1.36 1.58 1.65 1.66 1.68 1.61 3.2 3.2 0.77 1.12 1.64 1.91 2.05 1.95 2.06 1.9 1.28 1.28 0.76 1.19 1.76 2.13 2.3 2.27 2.26 2.17 50 HC102A HC103A 50 0 0.01 HC102A + 3.2 µM HC103A HC102A + 8 µM HC104A 0.1 1 10 Concentration (µM) HC102A + 3.2 µM HC104A 0 100 0.01 0.1 1 10 50 50 0.82 0.85 0.64 0.71 0.72 0.72 0.7 0.71 20 20 0.37 0.4 0.42 0.42 0.42 0.44 0.41 0.43 88 0.33 0.43 0.8 1.36 1.64 1.74 1.78 1.66 3.2 3.2 0.31 0.46 0.9 1.29 1.69 1.59 1.87 1.67 100 Concentration (µM) HC104A (µM) 1.28 1.28 0.33 0.49 0.81 1.18 1.4 1.47 1.52 1.37 0.51 0.51 0.39 0.58 0.82 1.62 1.3 1.21 1.35 1.25 0.21 0.44 0.63 0.99 1.12 1.3 1.24 1.49 1.14 0.08 0.08 0.5 0.7 1.05 1.27 1.47 1.32 1.72 0.91 HC103A (µM) 50 50 20 20 88 3.2 3.2 1.28 1.28 0.512 0.51 0.205 0.21 0.082 0.08 HC102A (µM) 0.08 0.08 0.74 1.04 1.57 2.32 7.27 46.9 5.7 9.43 HC104A HC106A (µM) 100 50 50 20 20 88 3.2 3.2 1.28 1.28 0.51 0.512 0.21 0.205 0.08 0.082 5050 0.67 0.52 0.45 0.42 0.39 0.4 0.39 0.16 2020 0.29 0.5 0.76 1.01 1.07 1.13 1.16 1.18 88 0.35 0.79 1.03 1.23 1.34 1.43 1.43 1.44 3.2 3.2 0.44 0.96 1.22 1.35 1.42 1.47 0.94 1.47 1.28 1.28 0.48 1.11 1.38 1.49 1.52 1.5 1.41 1.44 0.51 0.51 0.64 1.3 1.56 1.87 1.76 1.53 1.41 1.24 0.21 0.21 0.56 1.17 1.54 1.92 1.87 0.95 1.78 1.26 0.08 0.08 0.55 1.22 1.47 1.72 1.91 2.02 3.2 3.37 100 50 % inhibition % inhibition 0.21 0.21 0.75 1.13 1.69 2.18 3.05 3.81 5.22 3.1 HC102A HC102A + 8 µM HC103A HC102A HC106A HC102A + 1.28 µM HC104A 50 HC103A HC104A HC103A + 8 µM HC104A HC102A + 0.51 µM HC104A 0 0.01 0.1 1 10 HC103A + 3.2 µM HC104A 0 0.01 100 0.1 Concentration (µM) 1 2020 0.54 0.44 0.43 0.41 0.44 0.48 0.48 0.46 88 0.3 0.2 0.3 0.6 1.1 1.5 1.9 2.2 3.2 3.2 0.22 0.25 0.72 0.89 1.56 1.95 2.16 2.08 1.28 1.28 0.2 0.39 0.65 0.96 1.38 1.64 1.93 2.29 0.51 0.51 0.27 0.74 0.79 1 1.37 1.61 1.73 1.79 0.21 0.21 0.35 0.81 1 1.13 1.29 1.59 1.99 2.1 0.08 0.08 0.51 1.03 1.19 1.24 1.3 1.44 2.5 2.74 % inhibition 100 5050 2020 88 3.2 3.2 1.28 1.28 0.51 0.512 0.21 0.205 0.08 0.082 50 1.79 1.35 1.05 0.84 0.77 0.88 0.8 0.79 20 1.05 0.64 0.53 0.48 0.51 0.52 0.4 0.44 8 0.65 0.52 0.78 0.96 1.29 1.44 1.72 1.69 3.2 3.2 0.51 0.65 0.97 1.11 1.68 1.45 1.62 1.59 1.28 1.28 0.45 0.83 1.11 1.22 1.42 1.47 0.94 1.43 0.51 0.51 0.44 0.99 1.38 1.45 1.73 1.64 1.3 1.5 0.21 0.21 0.86 1.11 1.52 1.58 1.67 1.68 1.76 1.46 0.08 0.08 0.55 1.16 1.55 1.62 1.64 1.46 1.51 1.35 100 50 HC103A HC106A HC103A + 1.28 µM HC106A HC104A 50 HC106A HC104A + 1.28 µM HC106A HC103A + 0.51 µM HC106A 0 0.01 HC104A (µM) 5050 1.23 1.04 0.96 0.78 0.87 0.82 0.89 0.82 100 HC106A (µM) % inhibition 50 50 20 20 88 3.2 3.2 1.28 1.28 0.512 0.51 0.205 0.21 0.082 0.08 10 Concentration (µM) HC106A (µM) HC103A (µM) 0.51 0.51 0.77 1.19 1.75 2.06 2.67 2.6 3.21 2.28 100 % inhibition % inhibition 50 50 20 20 88 3.2 3.2 1.28 1.28 0.512 0.51 0.205 0.21 0.082 0.08 HC102A (µM) A 0.1 1 Concentration (µM) 10 100 HC104A + 0.51 µM HC106A 0 0.01 0.1 1 Concentration (µM) 10 100 0 1 CI 10 Supplementary Figure 3.7. Checkerboard assays examining paired interactions of DosRST inhibitors. CDC1551 (hspX’::GFP) was treated with different combination of two compounds from 50 µM to 0.08 µM. GFP fluorescence was measured and used to calculate percentage inhibition. The data were analyzed in the CompuSyn software to determine the combination index (CI) for the panel of each drug combination, including (a) artemisinin and HC104A; (b) HC102A and HC103A; (c) HC102A and HC104A; (d) HC102A and HC106A; (e) HC103A and HC104A; (f) HC103A and 136 Supplementary Figure 3.7. (cont’d) HC106A; (g) HC104A and HC106A. Selected dose response curves are presented to illustrate synergistic interactions. The experiment was repeated twice with similar results. 137 N HO O Oxalyl Chloride CH2Cl2 DMF (1 drop) O 1 N Cl O 1) TMS-Azide THF, RT 2) 110 , 3 hrs Toluene O O C N R1 O N 3 2 Supplementary Figure 3.8. Synthetic scheme for the HC106 analogs. 138 O NH 2 R1 N H N N H 4 O a Artemisinin 62 HC103A 6 b Artemisinin HC103A 3 0 HC103A 36 1.0 HC106A PE Rv0260c Rv0885 fprB PE Rv3049c Rv1652 Rv1653 Rv1654 Rv1656 argC argJ argB argF HC106A 4 1 Rv0160c Rv0260c Rv0885 Rv0886 Rv1386 Rv3049c 5.0 Fold change 25.0 Supplementary Figure 3.9. Comparison between artemisinin, HC103 and HC106 for interactions. (a) Venn diagram for the downregulated genes (>2-fold; q< 0.05) of CDC1551 (DdosR) treated with artemisinin, HC103A, or HC106A. (b) Overlap exists between artemisinin and HC103A or HC103A and HC106A differentially expressed genes (downregulated >2X, q<0.05). The heatmaps represent the commonly downregulated genes between the two compounds. 139 Experimental procedures for urea formation Formation of acyl chloride 2. To a stirred solution of isoxazole acid 1 (1 eq.) in dry tetrahydrofuran (THF, 0.4 M) under N2 atmosphere was added oxalyl chloride (1.5 eq.) dropwise over 5-10 min followed by dimethylformamide (DMF) (cat.) and the reaction mixture was continued to stir at room temperature. Upon completion, the reaction mixture was concentrated into a residue in vacuo and the residue was dissolved in THF and concentrated again to ensure the removal of excess oxalyl chloride. The crude acyl chloride 2 was used directly in the next step without further purification. Formation of acyl azide and rearrangement into isocyanate 3. The crude acyl chloride 2 was dissolved in THF (0.4 M) and stirred at room temperature under N2 atmosphere. Trimethylsilyl (TMS) azide (2 eq.) was added dropwise over 5 min and stirring was continued. Upon completion of the reaction, the mixture was diluted with ethyl acetate (0.4 M) and quenched with H2O (0.4 M). The two layers were separated, and the organic layer was dried over anhydrous Na2SO4 and filtered. The ethyl acetate solvent was swapped into toluene (0.1 M) by the addition of toluene followed by removal of the ethyl acetate in vacuo. Care was taken not to concentrate the toluene. The toluene acyl azide solution was heated at reflux conditions under N2 atmosphere for 4 h to give the desired isocyanate 3 which was used as a solution in toluene in the next step. Formation of urea 4. The crude isocyanate solution in toluene was mixed with different amines (1.5 eq.) and stirred at room temperature overnight. Isolation of the ureas was done by diluting the reaction mixture with hexanes, stirring for few hours and filtration of the formed precipitate. The solid material was washed with hexanes and dried under high vacuum. The urea products usually do not require further purifications. All products were analyzed by 1H NMR and high-resolution mass spectrometry. 140 Synthesis of MSU-41422 (amide) from acyl chloride 2. Acyl chloride 2 (1 eq.) was dissolved in dichloromethane (DCM, 0.2 M) and 4-chloroaniline (1.2 eq.) was added. The reaction mixture was stirred at room temperature overnight. The reaction mixture was concentrated in vacuo and the residue was purified by flash column chromatography. Synthesis of MSU-39449. A slurry of aaa (0.10 g, 0.60 mmoles) in 2.0 mL of DCM was treated with oxalyl chloride (0.12 mL, 0.9 mmoles) and 1 drop of DMF. The mixture (with gas evolution) gradually became homogeneous and was stirred for 30 min. The mixture was concentrated in vacuo, diluted with 5 mL of DCM and again concentrated in vacuo, the process repeated three times. The resulting residue was again dissolved in DCM (2 mL) and treated with 5-aminoisoxazole (0.030 g., 0.36 mmoles), followed by pyridine (0.48 mL, 0.6 mmoles). The mixture was then allowed to stir overnight. The mixture was then quenched with 1.0 N HCl and extracted with DCM. The organic layers were combined, washed with saturated KHCO3, dried with Na2SO4 and concentrated in vacuo. Medium pressure liquid chromatography (SiO2, 100% DCM to 3% methanol / DCM) to provide a solid (0.023 g). MSU-39444. 1H NMR (500 MHz, DMSO-d6) δ 10.54 (s, 2H), 8.44 (d, J = 1.9 Hz, 2H), 6.11 (d, J = 2.0 Hz, 2H). MSU-39449. 1H NMR (500 MHz, DMSO-d6) δ 11.91 (s, 1H), 8.41 (s, 1H), 7.39 (d, J = 2.0 Hz, 2H), 7.35 (d, J = 2.0 Hz, 2H), 6.19 (s, 1H), 3.73 (s, 2H). MSU-39445. 1H NMR (500 MHz, DMSO-d6) δ 10.41 (s, 1H), 9.14 (s, 1H), 8.40 (d, J = 2.0 Hz, 1H), 7.82 – 7.58 (m, 1H), 7.31 (dd, J = 4.9, 1.8 Hz, 2H), 7.07 (dt, J = 6.4, 2.3 Hz, 1H), 6.06 (d, J = 1.9 Hz, 1H). MSU-39452. 1H NMR (500 MHz, DMSO-d6) δ 10.50 (s, 1H), 9.23 (s, 1H), 8.51 – 8.28 (m, 1H), 7.86 (d, J = 2.5 Hz, 1H), 7.64 – 7.43 (m, 1H), 7.37 (dd, J = 8.9, 2.5 Hz, 1H), 6.26 – 5.80 (m, 1H). 141 MSU-39447. 1H NMR (500 MHz, DMSO-d6) δ 8.83 (s, 1H), 8.43 – 8.29 (m, 1H), 7.36 (d, J = 8.4 Hz, 2H), 6.86 (d, J = 8.6 Hz, 2H), 6.00 (d, J = 1.9 Hz, 1H), 3.70 (s, 3H). MSU-39451. 1H NMR (500 MHz, DMSO-d6) δ 10.13 (s, 1H), 8.34 (d, J = 1.9 Hz, 1H), 7.56 – 7.41 (m, 2H), 7.41 – 7.22 (m, 2H), 6.04 (d, J = 2.0 Hz, 1H), 3.25 (s, 3H). MSU-39453. 1H NMR (500 MHz, DMSO-d6) δ 10.49 (s, 1H), 8.41 (d, J = 1.9 Hz, 1H), 7.84 (d, J = 8.7 Hz, 1H), 7.27 (dt, J = 2.3, 1.1 Hz, 1H), 7.18 (dt, J = 8.8, 1.6 Hz, 1H), 6.14 (d, J = 1.9 Hz, 1H), 4.13 (dd, J = 9.1, 8.2 Hz, 2H), 3.17 (t, J = 8.6 Hz, 2H). MSU-39448. 1H NMR (500 MHz, DMSO-d6) δ 10.55 (s, 1H), 9.25 (s, 1H), 8.64 – 8.24 (m, 2H), 7.98 (d, J = 8.6 Hz, 1H), 7.45 (d, J = 8.6 Hz, 1H), 6.08 (d, J = 18.1 Hz, 1H). MSU-39450. 1H NMR (500 MHz, DMSO-d6) δ 11.00 (s, 1H), 9.66 (s, 1H), 8.42 (d, J = 1.9 Hz, 1H), 8.34 (d, J = 2.6 Hz, 1H), 7.96 – 7.79 (m, 1H), 7.73 (d, J = 8.9 Hz, 1H), 6.10 (q, J = 2.7, 2.2 Hz, 2H). MSU-39446. 1H NMR (500 MHz, DMSO-d6) δ 10.24 (s, 1H), 8.90 (s, 1H), 8.38 (d, J = 1.9 Hz, 1H), 7.54 – 7.37 (m, 2H), 7.26 – 7.02 (m, 2H), 6.03 (d, J = 2.0 Hz, 1H). MSU-41422. 1H NMR (500 MHz, Chloroform-d) δ 8.41 (d, J = 1.8 Hz, 1H), 8.26 (s, 1H), 7.71 – 7.52 (m, 2H), 7.44 – 7.33 (m, 2H), 7.06 (d, J = 1.8 Hz, 1H). MSU-41324. 1H NMR (500 MHz, DMSO-d6) δ 9.91 (s, 1H), 9.50 (s, 1H), 8.31 (d, J = 2.6 Hz, 1H), 7.85 (dd, J = 9.0, 2.7 Hz, 1H), 7.67 (d, J = 8.9 Hz, 1H), 7.39 – 7.30 (m, 2H). MSU-41425. 1H NMR (500 MHz, DMSO-d6) δ 8.83 (s, 1H), 7.53 – 7.39 (m, 2H), 7.39 – 7.16 (m, 2H). MSU-41443. 1H NMR (500 MHz, DMSO-d6) δ 10.36 (d, J = 28.4 Hz, 1H), 9.13 (d, J = 45.7 Hz, 1H), 8.40 (d, J = 1.9 Hz, 1H), 7.70 – 7.59 (m, 5H), 7.59 – 7.49 (m, 3H), 7.43 (t, J = 7.7 Hz, 3H), 7.32 (t, J = 7.4 Hz, 1H), 6.06 (d, J = 1.9 Hz, 1H). 142 MSU-41442. 1H NMR (500 MHz, DMSO-d6) δ 10.15 (s, 1H), 8.77 (s, 1H), 8.38 (d, J = 1.9 Hz, 1H), 7.44 – 7.27 (m, 4H), 6.03 (d, J = 2.0 Hz, 1H), 1.25 (s, 9H). MSU-33189. 1H NMR (500 MHz, DMSO-d6) δ 10.20 (s, 1H), 8.85 (s, 1H), 8.39 (d, J = 1.9 Hz, 1H), 7.54 – 7.37 (m, 2H), 7.35 – 7.21 (m, 2H), 7.02 (t, J = 7.4 Hz, 1H), 6.04 (d, J = 1.9 Hz, 1H). MSU-33231. 1H NMR (500 MHz, DMSO-d6) δ 10.29 (s, 1H), 9.01 (s, 1H), 8.39 (d, J = 1.9 Hz, 1H), 7.56 – 7.41 (m, 2H), 7.41 – 7.27 (m, 2H), 6.05 (d, J = 1.9 Hz, 1H). MSU-41462. 1H NMR (500 MHz, DMSO-d6) δ 10.18 (s, 1H), 8.31 (d, J = 2.0 Hz, 1H), 7.43 – 7.07 (m, 3H), 6.92 (s, 1H), 5.94 (d, J = 1.9 Hz, 1H), 4.30 (d, J = 6.0 Hz, 2H). MSU-41463. 1H NMR (500 MHz, DMSO-d6) δ 10.39 (s, 1H), 9.29 (s, 1H), 8.41 (d, J = 1.9 Hz, 1H), 7.67 (d, J = 1.0 Hz, 4H), 6.08 (d, J = 1.9 Hz, 1H). MSU-41464. 1H NMR (500 MHz, DMSO-d6) δ 10.29 (s, 1H), 9.01 (s, 1H), 8.39 (d, J = 1.9 Hz, 2H), 7.50 – 7.45 (m, 1H), 7.45 – 7.40 (m, 1H), 6.05 (s, 1H). MSU-41465. 1H NMR (500 MHz, DMSO-d6) δ 10.38 (s, 1H), 9.27 (s, 1H), 8.41 (d, J = 2.0 Hz, 1H), 8.00 – 7.79 (m, 2H), 7.66 – 7.48 (m, 2H), 6.08 (d, J = 1.9 Hz, 1H), 3.81 (s, 3H). 143 REFERENCES 144 REFERENCES 1. Boon C, Dick T. 2012. How Mycobacterium tuberculosis goes to sleep: the dormancy survival regulator DosR a decade later. Future Microbiol 7:513-518. 2. Rustad TR, Sherrid AM, Minch KJ, Sherman DR. 2009. Hypoxia: a window into Mycobacterium tuberculosis latency. Cell Microbiol 11:1151-1159. 3. Wayne LG, Sohaskey CD. 2001. Nonreplicating persistence of Mycobacterium tuberculosis. Annu Rev Microbiol 55:139-163. 4. Mehra S, Foreman TW, Didier PJ, Ahsan MH, Hudock TA, Kissee R, Golden NA, Gautam US, Johnson AM, Alvarez X, Russell-Lodrigue KE, Doyle LA, Roy CJ, Niu T, Blanchard JL, Khader SA, Lackner AA, Sherman DR, Kaushal D. 2015. The DosR Regulon Modulates Adaptive Immunity and Is Essential for Mycobacterium tuberculosis Persistence. Am J Respir Crit Care Med 191:1185-1196. 5. Hudock TA, Foreman TW, Bandyopadhyay N, Gautam US, Veatch AV, LoBato DN, Gentry KM, Golden NA, Cavigli A, Mueller M, Hwang SA, Hunter RL, Alvarez X, Lackner AA, Bader JS, Mehra S, Kaushal D. 2017. Hypoxia sensing and persistence genes are expressed during the intragranulomatous survival of Mycobacterium tuberculosis. Am J Respir Cell Mol Biol 56:637-647. 6. Dasgupta N, Kapur V, Singh KK, Das TK, Sachdeva S, Jyothisri K, Tyagi JS. 2000. Characterization of a two-component system, devR-devS, of Mycobacterium tuberculosis. Tuber Lung Dis 80:141-159. 7. Park HD, Guinn KM, Harrell MI, Liao R, Voskuil MI, Tompa M, Schoolnik GK, Sherman DR. 2003. Rv3133c/dosR is a transcription factor that mediates the hypoxic response of Mycobacterium tuberculosis. Mol Microbiol 48:833-843. 8. Saini DK, Malhotra V, Dey D, Pant N, Das TK, Tyagi JS. 2004. DevR-DevS is a bona fide two-component system of Mycobacterium tuberculosis that is hypoxia-responsive in the absence of the DNA-binding domain of DevR. Microbiology 150:865-875. 9. Voskuil MI, Schnappinger D, Visconti KC, Harrell MI, Dolganov GM, Sherman DR, Schoolnik GK. 2003. Inhibition of respiration by nitric oxide induces a Mycobacterium tuberculosis dormancy program. J Exp Med 198:705-713. 10. Kumar A, Deshane JS, Crossman DK, Bolisetty S, Yan BS, Kramnik I, Agarwal A, Steyn AJ. 2008. Heme oxygenase-1-derived carbon monoxide induces the Mycobacterium tuberculosis dormancy regulon. J Biol Chem 283:18032-18039. 145 11. Lee JM, Cho HY, Cho HJ, Ko IJ, Park SW, Baik HS, Oh JH, Eom CY, Kim YM, Kang BS, Oh JI. 2008. O2- and NO-sensing mechanism through the DevSR twocomponent system in Mycobacterium smegmatis. J Bacteriol 190:6795-6804. 12. Sousa EH, Tuckerman JR, Gonzalez G, Gilles-Gonzalez MA. 2007. DosT and DevS are oxygen-switched kinases in Mycobacterium tuberculosis. Protein Sci 16:1708-1719. 13. Kumar A, Toledo JC, Patel RP, Lancaster JR, Jr., Steyn AJ. 2007. Mycobacterium tuberculosis DosS is a redox sensor and DosT is a hypoxia sensor. Proc Natl Acad Sci U S A 104:11568-11573. 14. Ioanoviciu A, Yukl ET, Moenne-Loccoz P, de Montellano PR. 2007. DevS, a hemecontaining two-component oxygen sensor of Mycobacterium tuberculosis. Biochemistry 46:4250-4260. 15. Kim MJ, Park KJ, Ko IJ, Kim YM, Oh JI. 2010. Different roles of DosS and DosT in the hypoxic adaptation of Mycobacteria. J Bacteriol 192:4868-4875. 16. Rao SP, Alonso S, Rand L, Dick T, Pethe K. 2008. The protonmotive force is required for maintaining ATP homeostasis and viability of hypoxic, nonreplicating Mycobacterium tuberculosis. Proc Natl Acad Sci U S A 105:11945-11950. 17. Adams KN, Takaki K, Connolly LE, Wiedenhoft H, Winglee K, Humbert O, Edelstein PH, Cosma CL, Ramakrishnan L. 2011. Drug tolerance in replicating mycobacteria mediated by a macrophage-induced efflux mechanism. Cell 145:39-53. 18. Grant SS, Kaufmann BB, Chand NS, Haseley N, Hung DT. 2012. Eradication of bacterial persisters with antibiotic-generated hydroxyl radicals. Proc Natl Acad Sci U S A 109:12147-12152. 19. Leistikow RL, Morton RA, Bartek IL, Frimpong I, Wagner K, Voskuil MI. 2010. The Mycobacterium tuberculosis DosR regulon assists in metabolic homeostasis and enables rapid recovery from nonrespiring dormancy. J Bacteriol 192:1662-1670. 20. Converse PJ, Karakousis PC, Klinkenberg LG, Kesavan AK, Ly LH, Allen SS, Grosset JH, Jain SK, Lamichhane G, Manabe YC, McMurray DN, Nuermberger EL, Bishai WR. 2009. Role of the dosR-dosS two-component regulatory system in Mycobacterium tuberculosis virulence in three animal models. Infect Immun 77:12301237. 21. Gautam US, Mehra S, Kaushal D. 2015. in-vivo gene signatures of Mycobacterium tuberculosis in C3HeB/FeJ Mice. PLoS One 10:e0135208. 22. Gautam US, McGillivray A, Mehra S, Didier PJ, Midkiff CC, Kissee RS, Golden NA, Alvarez X, Niu T, Rengarajan J, Sherman DR, Kaushal D. 2015. DosS is required for 146 the complete virulence of Mycobacterium tuberculosis in mice with classical granulomatous lesions. Am J Respir Cell Mol Biol 52:708-716. 23. Deb C, Lee CM, Dubey VS, Daniel J, Abomoelak B, Sirakova TD, Pawar S, Rogers L, Kolattukudy PE. 2009. A novel in vitro multiple-stress dormancy model for Mycobacterium tuberculosis generates a lipid-loaded, drug-tolerant, dormant pathogen. PLoS One 4:e6077. 24. Basak A, Abouelhassan Y, Zuo R, Yousaf H, Ding Y, Huigens RW. 2017. Antimicrobial peptide-inspired NH125 analogues: bacterial and fungal biofilm-eradicating agents and rapid killers of MRSA persisters. Org Biomol Chem 15:5503-5512. 25. Zheng H, Colvin CJ, Johnson BK, Kirchhoff PD, Wilson M, Jorgensen-Muga K, Larsen SD, Abramovitch RB. 2017. Inhibitors of Mycobacterium tuberculosis DosRST signaling and persistence. Nat Chem Biol 13:218-225. 26. Daniel J, Deb C, Dubey VS, Sirakova TD, Abomoelak B, Morbidoni HR, Kolattukudy PE. 2004. Induction of a novel class of diacylglycerol acyltransferases and triacylglycerol accumulation in Mycobacterium tuberculosis as it goes into a dormancy-like state in culture. J Bacteriol 186:5017-5030. 27. Sirakova TD, Dubey VS, Deb C, Daniel J, Korotkova TA, Abomoelak B, Kolattukudy PE. 2006. Identification of a diacylglycerol acyltransferase gene involved in accumulation of triacylglycerol in Mycobacterium tuberculosis under stress. Microbiology 152:27172725. 28. Baek SH, Li AH, Sassetti CM. 2011. Metabolic regulation of mycobacterial growth and antibiotic sensitivity. PLoS Biol 9:e1001065. 29. Hung DT, Shakhnovich EA, Pierson E, Mekalanos JJ. 2005. Small-molecule inhibitor of Vibrio cholerae virulence and intestinal colonization. Science 310:670-674. 30. Podust LM, Ioanoviciu A, Ortiz de Montellano PR. 2008. 2.3 A X-ray structure of the heme-bound GAF domain of sensory histidine kinase DosT of Mycobacterium tuberculosis. Biochemistry 47:12523-12531. 31. Cho HY, Cho HJ, Kim YM, Oh JI, Kang BS. 2009. Structural insight into the hemebased redox sensing by DosS from Mycobacterium tuberculosis. J Biol Chem 284:1305713067. 32. Chou T, Martin N. 2005. CompuSyn for drug combinations: PC software and user's guide: a computer program for quantitation of synergism and antagonism in drug combinations, and the determination of IC50 and ED50 and LD50 values. 33. Chou TC, Talalay P. 1984. Quantitative analysis of dose-effect relationships: the combined effects of multiple drugs or enzyme inhibitors. Adv Enzyme Regul 22:27-55. 147 34. Boulton AJ, Katritzky AR. 1961. The tautomerism of heteroaromatic compounds with five-membered rings—I: 5-hydroxyisoxazoles-isoxazol-5-ones. Tetrahedron 12:41-50. 35. Topliss JG. 1972. Utilization of operational schemes for analog synthesis in drug design. Journal of Medicinal Chemistry 15:1006-1011. 36. Lee HN, Jung KE, Ko IJ, Baik HS, Oh JI. 2012. Protein-protein interactions between histidine kinases and response regulators of Mycobacterium tuberculosis H37Rv. J Microbiol 50:270-277. 37. Honaker RW, Leistikow RL, Bartek IL, Voskuil MI. 2009. Unique roles of DosT and DosS in DosR regulon induction and Mycobacterium tuberculosis dormancy. Infect Immun 77:3258-3263. 38. Baker JJ, Johnson BK, Abramovitch RB. 2014. Slow growth of Mycobacterium tuberculosis at acidic pH is regulated by phoPR and host-associated carbon sources. Mol Microbiol 94:56-69. 39. Johnson BK, Scholz MB, Teal TK, Abramovitch RB. 2016. SPARTA: Simple Program for Automated reference-based bacterial RNA-seq Transcriptome Analysis. BMC Bioinformatics 17:66. 40. Johnson BK, Abramovitch RB. 2015. Macrophage infection models for Mycobacterium tuberculosis. Methods Mol Biol 1285:329-341. 41. Schneider CA, Rasband WS, Eliceiri KW. 2012. NIH Image to ImageJ: 25 years of image analysis. Nat Methods 9:671-675. 42. Mak PA, Rao SP, Ping Tan M, Lin X, Chyba J, Tay J, Ng SH, Tan BH, Cherian J, Duraiswamy J, Bifani P, Lim V, Lee BH, Ling Ma N, Beer D, Thayalan P, Kuhen K, Chatterjee A, Supek F, Glynne R, Zheng J, Boshoff HI, Barry CE, 3rd, Dick T, Pethe K, Camacho LR. 2012. A high-throughput screen to identify inhibitors of ATP homeostasis in non-replicating Mycobacterium tuberculosis. ACS Chem Biol 7:11901197. 43. Vashist A, Prithvi Raj D, Gupta UD, Bhat R, Tyagi JS. 2016. The alpha10 helix of DevR, the Mycobacterium tuberculosis dormancy response regulator, regulates its DNA binding and activity. Febs j 283:1286-1299. 44. Chou TC. 2010. Drug combination studies and their synergy quantification using the Chou-Talalay method. Cancer Res 70:440-446. 148 CHAPTER - 4 HC2091 kills Mycobacterium tuberculosis by targeting the MmpL3 mycolic acid transporter This work has been submitted to Antimicrobial Agents and Chemotherapy and is currently under peer-review. Contributions and authors for the work described below: Huiqing Zheng, Garry B. Coulson, Elizabeth R. Haiderer and Robert B. Abramovitch 
 H.Z., G.B.C and R.B.A conceived the experiments. H.Z. performed the Mtb physiology and biochemistry experiments. G.B.C. performed initial characterization and secondary assays prioritizing HC2091 for study. E.R.H conducted the spectrum of activity studies cherry-pick of the compounds. 149 Summary Tuberculosis, caused by the intracellular pathogen Mycobacterium tuberculosis (Mtb), is a deadly disease that requires a long course of treatment. The emergence of drug resistant strains has been driven efforts to discover new small molecules that can kill the bacterium. Here we report characterizations of the compound HC2091 that kills Mtb in a time- and dose-dependent manner in vitro, as well as killing Mtb in macrophages. Whole genome sequencing of mutants that are spontaneously resistant to HC2091 identified single nucleotide variants in the mmpL3 mycolic acid transporter gene. HC2091 resistant mutants do not exhibit cross-resistance with the wellcharacterized MmpL3 inhibitor SQ109, suggesting a distinct mechanism of interaction with MmpL3. Additionally, HC2091 does not modulate bacterial membrane potential or kill nonreplicating Mtb, thus acting differently from other known MmpL3 inhibitors. RNA-seq transcriptional profiling and lipid profiling of Mtb treated with HC2091 or SQ109 show that both compounds target a similar pathway. HC2091 has a dissimilar chemical structure from previously described MmpL3 inhibitors, supporting that HC2091 is a new class of MmpL3 inhibitor. 
 150 Introduction Mycobacterium tuberculosis (Mtb) is the causative agent of tuberculosis (TB). Standard TB therapy requires months long treatment with four drugs, isoniazid, rifampin, pyrazinamide, and ethambutol. However, with the evolution and spread of drug resistant Mtb, currently available therapies have become inadequate to control multi-drug resistance and extensively drug resistant TB. Therefore, discovery efforts focused on identifying new antibiotics is critical for the long-term control of TB disease. Several new classes of compounds that kill Mtb, by inhibiting new targets, have been discovered in recent years (1), including inhibitors of the MmpL3 mycolic acid transporter. MmpL3 (Mycobacterium membrane protein Large 3) has proven to be a common target of small molecule inhibitors of mycobacterial growth identified by high-throughput screens (2, 3). MmpL3 is a membrane transporter belonging to the resistance, nodulation, and division superfamily. It is a mycolic acid flippase (4) that moves trehalose monomycolate (TMM) to the pseudoperiplasmic space, from where TMM is further modified to trehalose dimycolate (TDM) and incorporated into the mycomembrane (5, 6). Initial genetic studies suggested that mmpL3 may be essential, including the absence of mutants in high density transposon mutant library (7, 8) and an inability to generate deletion mutants by homologous recombination (6, 9). Mtb and M. smegmatis mmpL3 knockdown strains provide genetic evidence that mmpL3 is required for survival both in vitro and in vivo in a mouse TB infection model (3, 10-12). Depletion of MmpL3 arrests bacterial cell division and causes rapid cell death. This phenotype makes MmpL3 an attractive therapeutic target and supports efforts to characterize small molecules targeting MmpL3. Reported MmpL3 inhibitors include a highly diverse collection of chemical classes: 1,2ethylenediamine, SQ109 (13); diphenylether-modified adamantyl 1,2-diamine, TBL-140 (14); 151 indole-2-carboxamide, NITD-304 and NITD-349 (15-17); adamantyl ureas, AU1235 (6, 18); 1,5diarylpyrrole, BM212 (19, 20); benzimidazole, C215 (21); tetrahydropyrazo[1,5-a]pyrimidine-3carboxamide, THPP-2 (22) and N-benzyl-6’,7’-dihydro-spiro[piperidine-4,4’-thieno[3,2-c]pyrans (23) and N-[(4-methoxyphenyl)methyl]-5-(4-methylphenyl)-7-(trifluoromethyl)-1,5,6,7- 
tetrahydropyrazolo[1,5-a]pyrimidine-3-carboxamide, Compound 2 (24). MmpL3 inhibitors decrease TDM synthesis due to ineffective transport of precursor TMM. Additionally, several MmpL3 inhibitors decrease intracellular ATP concentration and reduce the protein motive force (PMF) by disrupting membrane potential (14, 25, 26), however, this activity is not universally shared by compounds targeted MmpL3. We previously performed a high-throughput phenotypic screen of a >200,000 compound library for inhibitors of Mtb PhoPR and DosRST two-component regulatory systems (27, 28), and as part of these screens we also identified compounds that inhibited Mtb growth. Through a series of high-throughput secondary assays, the compound HC2091 was identified in both screens as an inhibitor of Mtb growth with no detectable macrophage cytotoxicity, supporting its continued study. Here we report that HC2091 is an MmpL3 inhibitor that kills Mtb in vitro and in macrophages. Results HC2091 kills Mtb in vitro and in macrophages HC2091 (N-[2-(4-chlorophenyl)ethyl]-4-thiophen-2-yloxane-4-carboxamide; Fig. 4.1a) inhibits Mtb growth with a half maximal effective concentration (EC50) of 6.4 µM (Fig. 4.1b). The structure is distinct from other previously identified MmpL3 inhibitors, however it does share a carboxamide group that is present in indolecarboxamides and THPP, suggesting this functional group may be associated with activity. The spectrum of activity of HC2091 was 152 examined in M. smegmatis and other non-mycobacteria. HC2091 was active against M. smegmatis with a growth inhibition EC50 of 20 µM, but had no impact on other tested Gramnegative or -positive bacteria (Supplementary Table 4.1), supporting the notion that HC2091 inhibits a mycobacterium specific target. Dose- and time- dependent killing of Mtb by HC2901 was examined by treating Mtb for 7 d across a 0.25 µM to 80 µM dose response curve, and enumerating CFUs (Fig. 4.1c). At concentrations below 1 µM HC2091 had no effect on growth, at 5 µM HC2091 was bacteriostatic and above 10 µM HC2091 was bactericidal with an ~3 log reduction of CFU at day 7, relative to dimethyl sulfoxide (DMSO) treated controls. Killing of Mtb was observed beginning at 2 d of treatment. Collectively these data indicate that HC2091 bactericidal activity is dose and time dependent. Furthermore, HC2091 did not show any cytotoxicity against primary murine bonemarrow derived macrophages up to concentrations of 200 µM (Supplementary Table 4.2) and was fully soluble in a kinetic solubility assay (Supplementary Table 4.3). The observed high solubility and low cytotoxicity properties of HC2091 suggested that the compound could kill Mtb in macrophages. To examine the activity of HC2091 against intracellular Mtb, murine bone marrow derived macrophages were infected with Mtb and treated with HC2091 across a 6-point dose response curve from 1 µM to 40 µM. Bacterial CFUs were enumerated at various timepoints over 6 d of treatment. Consistent with the potency in broth culture (Figure 4.1c), bacterial growth inhibition was observed at concentrations of 5 µM and greater, with a 1 log reduction of CFUs compared to the DMSO control (Fig. 4.2a). These data show that HC2091 can kill intracellular Mtb. During infection, Mtb can present as a spectrum of disease, with some bacteria actively replicating and others existing in a state of non-replicating persistence (NRP). Non-replicating 153 bacteria are often tolerant to antibiotics and we sought to determine if HC2091 could kill them. Using the hypoxic-shift down model of NRP to examine HC2091 activity (29), Mtb cells were pre-incubated in an anaerobic chamber for 48 h to promote NRP, and then treated with HC2091 from 0.5 µM to 80 µM for 5 and 10 d. As a positive control, cells were treated with 10 µM and 20 µM SQ109, which has been shown to kill Mtb during NRP. To show that the bacteria were drug tolerant, cells were treated with 5 µM and 25 µM isoniazid (INH), which is less effective against NRP bacteria. It was observed that HC2091, like INH, does not kill NRP Mtb. In contrast, SQ109 reduced Mtb CFUs by 1 log at day 5 and 2 logs at day 10. This supports the hypothesis that HC2091 is inactive against non-replicating Mtb cells and likely acts in a manner unique from that of SQ109. 154 a Cl O S N H O b % Growth Inhibition 100 50 0 0.1 1 10 100 Concentration (µM) Bacterial growth ([log10 CFU/mL]) c 109 DMSO 0.25 µM 1 µM 5 µM 10 µM 20 µM 40 µM 80 µM 108 107 106 105 104 0 2 4 6 8 Time (Days) Figure 4.1. HC2091 is bactericidal to Mtb. (a) Chemical structure of HC2091; (b) HC2091 inhibits Mtb growth with EC50 value of 6.4 µM; (c) HC2091 kills Mtb in a dose- and time- dependent manner. The error bars represent the standard deviation of three technical replicates. The experiment was repeated twice with similar results. 155 Bacterial growth ([log10 CFU/mL]) a 107 DMSO 1 µM 5 µM 10 µM 20 µM 40 µM 106 105 104 0 2 4 6 Time (Days) Bacterial growth ([log10 CFU/mL]) b 108 Day 5 Day 10 107 106 105 D M SO 0. 5 µM 2 µM 5 µM 10 µM 20 µM 40 µM 80 µM 25 µM 10 µM 104 HC2091 INH SQ109 Figure 4.2. HC2091 is active against intracellular Mtb and inactive against non-replicating Mtb. (a) Mtb-infected murine bone marrow derived macrophages were treated with doses of HC2091 ranging from 1 µM to 40 µM. CFUs were enumerated on 7H10 agar plates at the indicated days post infection. (b) The impact of HC2091 on non-replicating Mtb was characterized using the hypoxic shift-down assay (29). SQ109 and INH were used as controls. Both experiments were repeated twice with similar results and the error bars represent the standard deviation of three technical replicates. 156 HC2091 resistant strains have mutations in mmpL3 To identify the potential target of HC2091, 1010 Mtb cells were plated on 7H10 plates containing 10 µM HC2091. Putative spontaneous resistant mutants were isolated and purified and EC50 values were determined for 23 mutants to confirm their resistance to HC2091. All mutants exhibited 8- to 30- fold increased resistance to HC2091 compared to WT (Supplemental Fig. 4.1). Whole genome sequences for eleven mutants were obtained to discover the potential target of HC2091. Using a GATK-based analysis pipeline, single nucleotide variants in each mutant were identified. All of the sequenced mutants harbored mutations in the mmpL3 gene (Table 4.1). Notably, four amino acid substitutions were concentrated in the region of amino acid 643-670 (Supplementary Fig. 4.2), which is a distinct region from where many other mmpL3 resistance mutations have been identified. This observation suggests that cross-resistance may not be observed with other MmpL3 inhibitors. To test this hypothesis, the EC50 of SQ109 was examined in several of the HC2091 mutants. HC2091 resistant mutants exhibited a small 1- to 3- fold enhanced resistance to SQ109 (Table 4.1), as compared to a greater than 10-fold resistance to HC2091. This observation suggests HC2091 may target a different region of MmpL3 than SQ109 and other MmpL3 inhibitors where the HC2091 resistant alleles have not been isolated. 157 Mutant clones SNV location Nucleotide change Quality Amino acid changes HC2091 EC50 (μM) SQ109 EC50 (μM) WT N/A N/A N/A N/A 6.4 0.95 #01/17 245733 CTG -> CCG 3030/3503 L567P 40-45 2.2 #02 245506 GTG -> ATG 2491 V643M >100 3.77 #03 245488 ATG -> CTG 3441 M649L >100 1.03 #04/14/16 245503 TTC -> CTC 2899/3179/ 3662 F644L >100 3.75 #05/06/12 245424 ACC -> AAC 2685/2474/ 2191 T670N >100 1.02 #23 245335 GCC -> ACC 2615 A700T >100 2.5 Table 4.1. Identification of HC2091 resistance mutations. Whole genome sequencing of selected HC2091 resistant strains identified single nucleotide variants (SNVs) in mmpL3 in all of the sequenced mutants. The mutants did not show cross resistance with SQ109. 158 Inhibition of mycolic acid biosynthesis by HC2091 A primary function of MmpL3 is to export TMM from the cytoplasm for mycomembrane formation (4-6). To assess the impact of HC2091 on biosynthesis of mycolic acids, Mtb cells were grown in the presence of 14 C-acetate and treated with HC2091. Radiolabelled total lipids were isolated and analyzed by thin layer chromatography (TLC, Fig. 4.3a). TDM accumulation in HC2091-treated Mtb was decreased in a dose dependent manner, with a 66% reduction of TDM relative to a DMSO control when treated with 40 µM HC2091 (Fig. 4.3b). Consistent with the mechanism of an MmpL3 inhibitor, the TDM-precursor TMM accumulated in all of the treatments. The results are consistent with previous reports that TDM biosynthesis is inhibited due to blockage of the exporting precursor TMM (5, 6). Modulation of mycolic acid accumulation provides further evidence supporting that MmpL3 is the target of HC2091. Transcriptional analysis of Mtb treated with HC2091 and SQ109 Transcriptional profiling was undertaken to define pathways that are regulated by HC2091. Mtb cells were treated with 10 µM HC2091 or SQ109 for 24 h, and total bacterial RNA was isolated. HC2091 at this concentration and time point was observed to have a defined impact on Mtb transcriptional profiling with 11 genes downregulated (>2-fold, q<0.05, Fig. 4.4ab and Supplementary Database 4.1-2). These genes completely overlapped with genes downregulated by treatment with SQ109, suggesting that both compounds are targeting the same pathway (Fig. 4.4c). The most highly repressed genes include Rv1886c (fbpB) and Rv2244 (acpM) both of which are involved in cell wall mycolic acid biosynthesis. The RNA-seq data show that HC2091 and SQ109 likely inhibit the same target. Given the limited number of genes regulated by HC2091, it is likely that HC2091 does not have other strong activities independent of targeting MmpL3. 159 HC2091 has no effect on Mtb membrane potential. Several MmpL3 inhibitors inhibit the mycobacterial membrane potential, including SQ109, BM212, TBL-140, indolecarboxamides, AUs, and THPPs (14, 22). Additionally, specific mmpL3 mutations have been shown to modulate membrane potential (30). We next sought to determine if HC2091 had any impact on Mtb membrane potential. Using the ratiometric membrane potential reporter dye DiOC2, kinetic changes of membrane potential were monitored for 1 h after treatment with doses of HC2091, at 5 µM, 20 µM and 80 µM, or treatment with SQ109 and carbonyl cyanide m-chlorophenyl hydrazone (CCCP) as positive controls. The results showed that the fluorescence ratio was decreased over time with CCCP and SQ109 treatments (Supplementary Figure 4.3), consistent with the previous reports that these two compounds inhibit Mtb membrane potential. However, for HC2091 treatments, the fluorescence ratio was stable and similar to the DMSO control. These data support the idea that HC2091 does not impact on Mtb membrane potential. 160 a HC2091 SQ109 b TDM TMM TDM TMM * Percentage relative to DMSO control 200 150 100 50 HC2091 µM 10 µM 40 µM 10 5 µM 0 SQ109 Figure 4.3. HC2091 inhibits TDM accumulation. (a) Mtb cells were treated with HC2091 or SQ109 and lipids were labeled with 14C-acetate for 6 d. Lipids were isolated and analyzed by TLC. (b) Quantification of TDM and TMM accumulation for Mtb cells treated with HC2091 or SQ109 as controls. The error bars represent the standard deviation of two biological replicates. 161 b 2 Fold change (log2) 0 2 Fold change (log2) -0.5 0.0 0.5 1.0 4 1.5 a -1.0 Rv0950 Rv2450c Rv2450c Rv1885c 2 c Rv1886c Rv1885c -4 -1.5 Rv1886c Rv0950 4 6 8 10 Average counts (logCPM) 14 2 4 HC2091 HC2091 0 12 6 8 10 Average counts (logCPM) Rv1885c Rv1886c Rv0950c Rv2450c MT1627 Rv1013 Rv1592c Rv2451 Rv2190c Rv2189c Rv2244 27 1.0 14 SQ109 SQ109 11 12 5.0 Fold change CHP fbpB CHP rpfE HP pks16 CHP HP CHP CHP acpM 25.0 Figure 4.4. Shared transcriptional profiles between HC2091 and SQ109. Transcriptional profiling of Mtb cells treated with HC2091 (a) or SQ109 (b). Cells were treated with 10 µM of the compounds for 24 h. The genes in red are significantly differentially regulated (q<0.05). (c) Overlap exists between the HC2091 and SQ109 differentially expressed genes (downregulated >2X, q<0.05). The heatmap represents the 11 commonly downregulated genes. CHP, conserved hypothetical protein; HP, hypothetical protein. 162 Structure-activity relationship studies A small-scale structure-activity relationship (SAR) study of HC2091 was conducted using 24 commercially available compounds to identify the functional group/s of the compound responsible for its activity. Mtb was treated with a dose response profile from 0.32 µM to 200 µM, growth was measured after 6 d incubation and EC50 values were determined for each analog (Fig. 4.6). We observed that HC2091M, N, W had a similar activity to HC2091A, with EC50 values are 9.7 µM, 10 µM, and 8.5 µM, respectively. HC2091D, E, F have EC50 values of 33 µM, 22 µM, and 35 µM, respectively. HC2091G, J, L have EC50 values of 81.3 µM, 66.8 µM, and 63.8 µM, respectively. HC2091I and K were inactive (>200 µM). Comparing the activity of HC2091A to HC2091E shows that the extra carbon linker to the benzene ring improves the activity. Additional observations include that a chloride (HC2091E) at the para-position had higher activity as compared to a fluoride (HC2091J), with EC50 of 21.9 µM and 66.8 µM, respectively. However, other functional groups, such as trifluoromethyl, ethyl, and isobutyl, provided activity similar to the primary hit HC2091A. Eukaryotic cytotoxicity was determined for HC2091A, M, N, W (Supplemental Table 4.2) and toxicity was >200 µM for HC2091A, M, and N and 80 µM for HC2091W, suggesting that the analogs have limited macrophage cytotoxicity. Analogs HC2091A, M and N were also soluble in water at concentrations of >200 µM (Supplementary Fig. 4.3). Overall, these initial SAR studies indicate that there exists opportunities to improve the HC2091 for potency or pharmacokinetic properties, as several distinct analogs of HC2091 exhibited activity, cytotoxicity and solubility properties comparable to the parent compound. 
 163 O S N H R O Analogs EC50 (μM) R Analogs * * 6.4 A Cl F M F * 200 N H * 10 N * * O C 9.7 F O B EC50 (μM) R 200 Cl O 200 N H Cl * * D N+ O- 32.7 P 200 Br O * E 21.9 * Q 200 Cl * * F 34.4 R O O 200 O * * G 81.3 S 121 O O O * H 200 T 200 U 66.8 V * 200 S * I * 200 F * J * Cl 200 F * * F K 200 W 8.5 * L * F 63.8 X 200 O NH 2 Table 4.2. HC2091 Structure activity relationship study. Analogs of HC2091 were purchased from commercial sources and EC50 values were calculated for each analog. The primary hit HC2091A has EC50 value of 6 µM. Three analogs with a similar EC50 value as HC2091A were identified, including HC2091M, N, W. 164 Discussion This study describes the identification of a new class of MmpL3 inhibitor. HC2091 has several features that distinguish it from other MmpL3 inhibitors, including a distinct chemical structure, limited impact on membrane potential, resistance amino acid substitutions clustered in the MmpL3 C-terminus, and high solubility. In contrast to SQ109, HC2091 has no effect on nonreplicating Mtb, consistent with its lack of impact on membrane potential, which is essential for Mtb survival during NRP (31). HC2019 resistant mutants do not show cross-resistance with SQ109, and HC2091 does not impact the membrane potential, suggesting that HC2091 is a distinct anti-mycobacterial agent from SQ109. Whole genome sequencing of HC2091 resistant mutants strongly suggests that HC2091 targets MmpL3. This conclusion is further supported by the phenotype of decreasing TDM accumulation observed by TLC. Analogs of HC2091A were identified from an SAR study, showing that optimization of this molecule is possible with additional chemistry efforts. Such optimizations could include increased potency and improvements to pharmacokinetic properties, such as replacing the thiophene group which may be prone to in vivo metabolism (32). MmpL3 has the essential function of exporting mycolic acids to the mycomembrane. mmpL3 knockdown studies show that MmpL3 is essential to Mtb and required for virulence in a mouse model of TB infection (10, 11). Therefore, MmpL3 is an attractive drug target for new TB therapies. Several MmpL3 inhibitors have been discovered recently, with a variety of distinct chemical structures. None of the known MmpL3 inhibitors known to date share the structure of HC2091. It is apparent that MmpL3 is particularly susceptible to chemical inhibition, however, it is unclear why it is such a common target in whole cell phenotypic screens. One possibility is that the mutations in mmpL3 are enough to offset the activity of the compound on other targets. For 165 instance, THPP was originally discovered and characterized as an MmpL3 inhibitor based on spontaneous resistance-conferring mutations and phenotypes. Later research, however, showed that it actually targets EchA6, which is predicted to be an enoyl-coenzyme A (CoA) hydrolase (33). This result raises the question of whether other putative MmpL3 inhibitors are actually targeting to MmpL3 or have other targets. Further biochemical studies are needed to characterize these compounds to define their direct interaction and modulation of the MmpL3 protein, such as a recent demonstration that BM212 directly binds and inhibits MmpL3 (4). In membrane potential studies, MmpL3 inhibitors, including SQ109, BM212, indolecarboxamides, AUs, and THPPs, were all reported to disrupt membrane potential (22), which abolishes PMF-driven translocation processes, including TMM translocation via MmpL3. To further support the requirement of PMF for TMM translocation, the impacts of known PMF inhibitors, CCCP and valinomycin, on TDM synthesis were examined (22). The results showed CCCP caused 3.7-fold decrease in TDM biosynthesis, while there was 10-fold increase in TMM accumulation (22). Similar effects were also observed when Mtb was treated with valinomycin. This may explain why these MmpL3 inhibitors have broader targets, as PMF is required for many cellular functions. HC2091 exhibited a very narrow transcriptional response when treated with HC2091, however, the genes that were regulated overlapped with SQ109. These observations support the hypothesis that MmpL3 may be the primary target of HC2091. Additionally, HC2091 has no impact on non-replicating Mtb, while other MmpL3 inhibitors, including SQ109, DA5, BM212, THPP-2 have significant impact on killing non-replicating Mtb. It is not known whether MmpL3 is required for Mtb survival during NRP, whereas agents that target electron transport and membrane potential are known to kill NRP bacteria (29, 31, 34). However, it is known that cell 166 wall synthesis, including mycolic acid biosynthesis, is reduced during NRP, therefore it is reasonable to postulate that the translocation of TMM is minimal, and inhibition of MmpL3, without disrupting PMF, is not lethal to non-replicating Mtb. Therefore, the impact on nonreplicating Mtb by other MmpL3 inhibitors may not be due to inhibition of TDM synthesis, but rather due to disruption of PMF. Overall, HC2091 represents a new class of MmpL3 inhibitor with distinct chemical and physiological properties that support its continued study as a potential new anti-mycobacterial agent. Materials and methods Bacterial strains and growth conditions. Mtb strain CDC1551 was used for this study. It was cultured in 7H9 Middlebrook medium (supplemented with 10% OADC (oleic acid dextrose catalase) and 0.05% Tween-80) at 37°C and 5% CO2 in standing vented tissue culture flasks. M. smegmatis was grown in Luria-Bertani (LB) broth with standing at 37°C. Non-mycobacterial strains include S. aureus strains Wichita (29213) or Seattle (25923), E. coli (Migula) Castellani and Chalmers, P. aeruginosa (Schroeter) Migula, P. vulgaris Hauser emend Judicial Commission, and E. faecalis (Andrewes and Horder) Schleifer and Kilpper-Balz. All of non-mycobacterial strains were grown in LB broth with shaking at 37°C with the exception of E. faecalis, which was grown in brain heart infusion medium in standing flasks at 37°C. EC50 determination and spectrum of activity in other mycobacteria and non-mycobacteria. Mtb cultures were grown to mid-exponential phase, diluted into fresh 7H9 medium to an OD600 of 0.05, and 200 µL of diluted culture was aliquoted to wells of a 96-well plate and treated with an 8-point (2.5-fold) dilution series of HC2091 ranging from 200 µM-0.32 µM. Optical density (OD600) was 167 measured after 6 d of incubation, and growth of cultures was normalized based on OD600 relative to the 0.3 µM rifampin (100% growth inhibition) and DMSO (0% growth inhibition) controls. The EC50 values were calculated using the GraphPad Prism software package (version 6). To examine the spectrum of activity of HC2091, the EC50 of HC2091 was also determined against M. smegmatis and other non-mycobacteria, including S. aureus, E. coli, P. aeruginosa, P. vulgaris, and E. faecalis. Tests were performed in 96-well plates in LB broth shaking at 37°C, with the exception of E. faecalis, which was grown in brain heart infusion medium in standing flasks at 37°C, and M. smegmatis, which was also grown standing at 37°C. Culture was diluted to a starting OD600 of 0.05. Bacteria were incubated in the presence of an 8-point (2-fold) dilution series of HC2091 ranging from 200 µM - 1.5 µM for 6-8 h, except for M. smegmatis, which was incubated for 60 h. Growth was monitored using by measuring OD600 and normalized based on kanamycin (100% growth inhibition) and DMSO (0% growth inhibition) controls with the exception of P. aeruginosa, for which tobramycin was used as the control for 100% growth inhibition. The experiments were performed with three technical replicates per plate and two biological replicates. EC50 values were calculated based on a variable slope, four-parameter non-linear least squares regression model in the GraphPad Prism software package (ver. 6). 
 Bactericidal activity of HC2091. Mtb cells were initially cultured to an OD600 0.1 in 7H9 media in standing T25 tissue culture flasks, and incubated at 37°C. The cultures were treated with different doses of HC2091 from 0.25 µM to 80 µM including a DMSO control. At each indicated time point, 100 µL of culture was removed from the flasks, and 10-fold serial dilutions were performed and bacterial CFUs were enumerated by dilution plating on 7H10 agar plates. Bactericidal activity of the compound was determined by comparing bacterial CFUs after 168 treatment to the initial inoculum. The experiments were performed with three technical replicates per treatment and two biological replicates. Activity of HC2091 against intracellular Mtb. Murine bone-marrow derived macrophages were seeded in 24-well plates before infection as previously described (35). Macrophages were infected with Mtb at a multiplicity of infection ratio of 1:2 for 1 h. Mtb-infected macrophages were then treated with indicated doses of HC2091 with a DMSO control. Culture medium was changed every 2 d with fresh antibiotics added at each medium change. Macrophages were lysed by adding 0.1% Tween-80 in water. Viable Mtb cells were dilution plated on 7H10 agar medium and enumerating CFUs. The experiment was performed with three technical replicates per treatment and two biological replicates. Whole genome sequencing to identify the target of HC2091. Mtb cells were grown to lateexponential phase with an OD600 of ~1.0, 1010 cells were plated on 7H10 agar medium supplemented with 10 µM HC2091. The plates were incubated at 37°C until isolated resistant colonies were observed. Spontaneous resistant mutants were isolated, cultured and dilution plated to purify single colonies. EC50 values were determined as described above. Genomic DNA from eleven resistant mutants with different EC50 values were submitted for whole genome sequencing, and results were analyzed using a GATK based workflow as previously described (36, 37). Mycolic acid accumulation analysis. Mtb cells were diluted to OD600 of 0.1 in 8 mL of 7H9 medium in standing T-25 tissue culture flasks. Cells were radiolabeled by adding 8 µCi of [1,214 C] sodium acetate to the culture. Cultures were treated with the indicated doses of HC2091, 169 SQ109 or a DMSO control for 24 h at 37°C. Total lipid was extracted and analyzed in thin-layer chromatography (TLC), as previously described (38). Total extractable lipid 14C incorporation was determined in scintillation counting, and 20,000 c.p.m. (counts per minute) were spotted on a 100cm2 high-performance TLC silica gel 60 aluminum sheet (EMD Millipore). A chloroformmethanol-water (20:4:0.5, vol/vol/vol) solvent system was used to resolve lipids to analyze mycolic acid accumulation. The TLC was exposed to a phosphor screen for 3 d, imaged on a Typhon imager and lipids were quantified by ImageJ (39). The experiment was repeated with two biological replicates with similar results. Transcriptional profiling and data analysis. Mtb cells were inoculated to an initial OD600 of 0.1 in 8 mL 7H9 medium. Cell were treated with 10 µM HC2091, SQ109 or DMSO for 24 h and grown at 37°C and 5% CO2 without shaking in T-25 tissue culture flasks. Total bacterial RNA from two biological replicates was extracted and sequenced as previously described (38). RNAseq data were analyzed in the SPARTA software package (40). Transcriptional profiling data is available at the GEO Database (GEO accession number GSE107485). Membrane potential. The membrane potential assay was performed as previously described (14). Briefly, CDC1551 cells were labeled with 30 µM DiOC2 (diethyloxacarbocyanine iodide, ThermoScientific) in 1 mL PBS, pH 7.4, supplemented with 50 mM KCl, and incubated at 37°C for 15 min. Cells were washed twice and suspended in PBS at final concentration of 0.2 OD600, 200 µL of labeled cells were aliquoted to 96-well plates, and treated with 20 µM SQ109, or 5 µM, 20 µM, and 80 µM HC2091. A DMSO negative control and a 25 µM CCCP (Sigma Aldrich) positive controls were also included in the assay. Each treatment included 4 replicates per plate. 170 The kinetics of fluorescence (excitation = 485 nm, emission = 610 nm / 515 nm) was measured every 2 min for 60 min. The red / green (610 nm / 515 nm) fluorescence intensity ratio was calculated and used to quantify membrane potential. The experiment was repeated at least twice with similar result. The error bars represent the standard deviation of four replicates. Structure activity relationship (SAR) and eukaryotic cytotoxicity of HC2091. HC2091 analogs with different chemical properties were purchased from ChemDiv for catalog SAR studies. The EC50 values of compounds B-X were calculated by using 8-point (2.5-fold) dilution series ranging from 200-0.32 µM as described above. Eukaryotic cytotoxicity of HC2091 and its selected analogs were examined as previously described (28). Briefly, mouse bone marrow derived macrophages were seeded in 96-well plates and grown at 37°C with 5% CO2 in medium without antibiotics. After overnight incubation, cells were treated with a dilution series of the compounds from 200 0.32 µM, and continued incubation for 4 d at 37°C with 5% CO2. Cell viability was assessed using the CellTiter Glo luciferase reagent (Promega). The data were normalized with the 1% Triton X100 (100% toxicity) and DMSO (0% toxicity) controls. EC50 and half-maximal cell toxicity concentration (CC50) values were calculated in the GraphPad Prism software package (version 6). Kinetic solubility assay. The solubility assay was performed with 7-point (2-fold) dilutions from 200 µM - 3.125 µM for HC2091 analogs. Mebendazole, benxarotene and aspirin were also included as controls. The drug dilutions were added to PBS, pH 7.4, with the final DMSO concentration of 1%, and incubated at 37 °C for 2 h. The absorbance at 620 nm was measured for each drug dilution to estimate the compound solubility. Three replicates were done for each dilution. 171 Acknowledgements We thank Edmund Ellsworth and the MSU Chemical Biology and MedicinalChemistry Core for conducting the kinetic solubility assays. The MSU RTSF provided technical support for the RNAseq library preparation and sequencing. This project was supported by start-up funding from Michigan State University, AgBioResearch, grants from the NIH-NIAID (R21AI105687 and R01AI116605) and Grand Challenges Explorations grants from the Bill and Melinda Gates foundation (OPP1059227 and OPP1119065) and the Jean P. Schultz Endowed Biomedical Research Fund at the MSU College of Human Medicine. 172 APPENDIX 173 APPENDIX Supplementary information Supplementary Dataset 4.1. Differential gene expression data of WT Mtb treated with inhibitors Supplementary Dataset 4.2. Complete gene expression tables for transcriptional profiling experiments. 174 Bacterial strain EC50 (μM) E. coli > 200 E. faecalis > 200 M. smegmatis 20 P. aeruginosa > 200 P. vulgaris > 200 S. aureus 25923 > 200 S. aureus 29213 > 200 Supplementary Table 4.1. Spectrum of activity of HC2091. The EC50 studies were performed for the mycobacteria and non-mycobacteria. Cells were treated with a dose from 200 µM to 0.32 µM, and the optical density was measured after time that allowed for bacterial growth. The EC50 values were calculated for each strain. The experiment was repeated twice with similar results. 175 HC2091 analogs EC50 (μM) A > 200 M > 200 N >200 W > 80 Supplementary Table 4.2. Cytotoxicity of HC2091 analogs. Murine bone-marrow derived macrophages were treated with a series of (2.5-fold) dilutions of HC2091 analogs ranging from 200 µM to 0.32 µM for 4 d. Cell survival was measured with CellTiter-Glo (Promega). 176 Compound Name Solubility (µM) Mebendazole 37.0 Benxarotene 23.8 Aspirin >100 HC2091A > 200 HC2091M > 200 HC2091N >200 HC2091W 161.8 Supplementary Table 4.3. Solubility of HC2091 analogs. The assay was performed with 7-point (2-fold) dilutions from 200 µM - 3.125 µM for HC2091 analogs. Mebendazole, benxarotene and aspirin were also included as controls. The absorbance at 620 nm was measured for each drug dilution to estimate of the compound solubility. The experiment was repeated with three replicates. 177 % Inhibition 50 0 1 10 100 50 0 1 1000 10 100 1000 Concentration (µM) Concentration (µM) -50 WT 13 14 15 16 17 18 19 20 21 22 23 100 % Inhibition WT 1 2 3 4 5 6 7 8 10 11 12 100 -50 Supplementary Figure 4.1. Dose-response curves for HC2091 resistant mutants. Each mutant was treated with a series of (2.5-fold) dilutions of HC2091 ranging from 200 µM to 0.32 µM for 6 d, and the OD600 was measured after allowing 6 d for growth. The experiment was repeated twice with similar results. 178 Extracellular 31 184 228 237 307 301 415 561 611 1 Cytoplasmic 201 14 205 257 283 336 397 579 624 2 583 690 696 6 3 646 672 4 5 726 1 HC2091 (L567P) 2 HC2091 (V643M) 3 HC2091 (F644L) 4 HC2091 (M649L) 5 HC2091 (T670N) 6 HC2091 (AN700T) Supplementary Figure 4.2. The map of HC2091 resistance substitutions on the MmpL3 protein. Six different single mutations on mmpL3 were identified from whole genome sequencing of HC2091 spontaneous resistant mutants. The majority of the amino acid substitutions are located in the Cterminus of the protein. The model of MmpL3 protein was derived from the PHYRE2 protein fold recognition server (41). 179 DMSO 25 µM CCCP 20 µM SQ109 5 µM HC2091 20 µM HC2091 80 µM HC2091 Red/Green Fluorescence Ratio 0.15 0.10 0.05 0.00 0 20 40 60 Time (minutes) Supplementary Figure 4.3. Impact of HC2091 on Mtb membrane potential. Mtb cells were labeled with DiOC2 and treated with 20 µM SQ109, or 5 µM, 20µM, and 80µM HC2091. DMSO-negative and CCCP-positive controls were also included. The red / green (610 nm / 515 nm) fluorescence intensity ratio was calculated for each treatment over 60 min. SQ109 and CCCP showed reduced fluorescence intensity ratio, while HC2091 maintained the curve similar to DMSO control, suggesting HC2091 has no impact on Mtb membrane potential. The experiment was repeated twice with similar results. The error bars represent the standard deviation of four technical replicates. 180 REFERENCES 181 REFERENCES 1. Hoagland DT, Liu J, Lee RB, Lee RE. 2016. New agents for the treatment of drugresistant Mycobacterium tuberculosis. Adv Drug Deliv Rev 102:55-72. 2. Poce G, Consalvi S, Biava M. 2016. MmpL3 inhibitors: diverse chemical scaffolds inhibit the same target. Mini Rev Med Chem 16:1274-1283. 3. Belardinelli JM, Yazidi A, Yang L, Fabre L, Li W, Jacques B, Angala SK, Rouiller I, Zgurskaya HI, Sygusch J, Jackson M. 2016. Structure-function profile of MmpL3, the essential mycolic acid transporter from Mycobacterium tuberculosis. ACS Infect Dis 2:702-713. 4. Xu Z, Meshcheryakov VA, Poce G, Chng SS. 2017. MmpL3 is the flippase for mycolic acids in mycobacteria. Proc Natl Acad Sci U S A 114:7993-7998. 5. Tahlan K, Wilson R, Kastrinsky DB, Arora K, Nair V, Fischer E, Barnes SW, Walker JR, Alland D, Barry CE, 3rd, Boshoff HI. 2012. SQ109 targets MmpL3, a membrane transporter of trehalose monomycolate involved in mycolic acid donation to the cell wall core of Mycobacterium tuberculosis. Antimicrob Agents Chemother 56:1797-1809. 6. Grzegorzewicz AE, Pham H, Gundi VA, Scherman MS, North EJ, Hess T, Jones V, Gruppo V, Born SE, Kordulakova J, Chavadi SS, Morisseau C, Lenaerts AJ, Lee RE, McNeil MR, Jackson M. 2012. Inhibition of mycolic acid transport across the Mycobacterium tuberculosis plasma membrane. Nat Chem Biol 8:334-341. 7. Lamichhane G, Tyagi S, Bishai WR. 2005. Designer arrays for defined mutant analysis to detect genes essential for survival of Mycobacterium tuberculosis in mouse lungs. Infect Immun 73:2533-2540. 8. Griffin JE, Gawronski JD, Dejesus MA, Ioerger TR, Akerley BJ, Sassetti CM. 2011. High-resolution phenotypic profiling defines genes essential for mycobacterial growth and cholesterol catabolism. PLoS Pathog 7:e1002251. 9. Domenech P, Reed MB, Barry CE, 3rd. 2005. Contribution of the Mycobacterium tuberculosis MmpL protein family to virulence and drug resistance. Infect Immun 73:3492-3501. 10. Li W, Obregon-Henao A, Wallach JB, North EJ, Lee RE, Gonzalez-Juarrero M, Schnappinger D, Jackson M. 2016. Therapeutic potential of the Mycobacterium tuberculosis mycolic acid transporter, MmpL3. Antimicrob Agents Chemother 60:51985207. 182 11. Degiacomi G, Benjak A, Madacki J, Boldrin F, Provvedi R, Palu G, Kordulakova J, Cole ST, Manganelli R. 2017. Essentiality of mmpL3 and impact of its silencing on Mycobacterium tuberculosis gene expression. Sci Rep 7:43495. 12. Rock JM, Hopkins FF, Chavez A, Diallo M, Chase MR, Gerrick ER, Pritchard JR, Church GM, Rubin EJ, Sassetti CM, Schnappinger D, Fortune SM. 2017. Programmable transcriptional repression in mycobacteria using an orthogonal CRISPR interference platform. Nat Microbiol 2:16274. 13. Sacksteder KA, Protopopova M, Barry CE, 3rd, Andries K, Nacy CA. 2012. Discovery and development of SQ109: a new antitubercular drug with a novel mechanism of action. Future Microbiol 7:823-837. 14. Foss MH, Pou S, Davidson PM, Dunaj JL, Winter RW, Pou S, Licon MH, Doh JK, Li Y, Kelly JX, Dodean RA, Koop DR, Riscoe MK, Purdy GE. 2016. DiphenyletherModified 1,2-Diamines with Improved Drug Properties for Development against Mycobacterium tuberculosis. ACS Infect Dis 2:500-508. 15. Rao SP, Lakshminarayana SB, Kondreddi RR, Herve M, Camacho LR, Bifani P, Kalapala SK, Jiricek J, Ma NL, Tan BH, Ng SH, Nanjundappa M, Ravindran S, Seah PG, Thayalan P, Lim SH, Lee BH, Goh A, Barnes WS, Chen Z, Gagaring K, Chatterjee AK, Pethe K, Kuhen K, Walker J, Feng G, Babu S, Zhang L, Blasco F, Beer D, Weaver M, Dartois V, Glynne R, Dick T, Smith PW, Diagana TT, Manjunatha UH. 2013. Indolcarboxamide is a preclinical candidate for treating multidrug-resistant tuberculosis. Sci Transl Med 5:214ra168. 16. Stec J, Onajole OK, Lun S, Guo H, Merenbloom B, Vistoli G, Bishai WR, Kozikowski AP. 2016. Indole-2-carboxamide-based MmpL3 inhibitors show exceptional antitubercular activity in an animal model of tuberculosis infection. J Med Chem 59:62326247. 17. Lun S, Guo H, Onajole OK, Pieroni M, Gunosewoyo H, Chen G, Tipparaju SK, Ammerman NC, Kozikowski AP, Bishai WR. 2013. Indoleamides are active against drug-resistant Mycobacterium tuberculosis. Nat Commun 4:2907. 18. Scherman MS, North EJ, Jones V, Hess TN, Grzegorzewicz AE, Kasagami T, Kim IH, Merzlikin O, Lenaerts AJ, Lee RE, Jackson M, Morisseau C, McNeil MR. 2012. Screening a library of 1600 adamantyl ureas for anti-Mycobacterium tuberculosis activity in vitro and for better physical chemical properties for bioavailability. Bioorg Med Chem 20:3255-3262. 19. La Rosa V, Poce G, Canseco JO, Buroni S, Pasca MR, Biava M, Raju RM, Porretta GC, Alfonso S, Battilocchio C, Javid B, Sorrentino F, Ioerger TR, Sacchettini JC, Manetti F, Botta M, De Logu A, Rubin EJ, De Rossi E. 2012. MmpL3 is the cellular target of the antitubercular pyrrole derivative BM212. Antimicrob Agents Chemother 56:324-331. 183 20. Deidda D, Lampis G, Fioravanti R, Biava M, Porretta GC, Zanetti S, Pompei R. 1998. Bactericidal activities of the pyrrole derivative BM212 against multidrug-resistant and intramacrophagic Mycobacterium tuberculosis strains. Antimicrob Agents Chemother 42:3035-3037. 21. Stanley SA, Grant SS, Kawate T, Iwase N, Shimizu M, Wivagg C, Silvis M, Kazyanskaya E, Aquadro J, Golas A, Fitzgerald M, Dai H, Zhang L, Hung DT. 2012. Identification of novel inhibitors of M. tuberculosis growth using whole cell based highthroughput screening. ACS Chem Biol 7:1377-1384. 22. Li W, Upadhyay A, Fontes FL, North EJ, Wang Y, Crans DC, Grzegorzewicz AE, Jones V, Franzblau SG, Lee RE, Crick DC, Jackson M. 2014. Novel insights into the mechanism of inhibition of MmpL3, a target of multiple pharmacophores in Mycobacterium tuberculosis. Antimicrob Agents Chemother 58:6413-6423. 23. Remuinan MJ, Perez-Herran E, Rullas J, Alemparte C, Martinez-Hoyos M, Dow DJ, Afari J, Mehta N, Esquivias J, Jimenez E, Ortega-Muro F, Fraile-Gabaldon MT, Spivey VL, Loman NJ, Pallen MJ, Constantinidou C, Minick DJ, Cacho M, RebolloLopez MJ, Gonzalez C, Sousa V, Angulo-Barturen I, Mendoza-Losana A, Barros D, Besra GS, Ballell L, Cammack N. 2013. Tetrahydropyrazolo[1,5-a]pyrimidine-3carboxamide and N-benzyl-6',7'-dihydrospiro[piperidine-4,4'-thieno[3,2-c]pyran] analogues with bactericidal efficacy against Mycobacterium tuberculosis targeting MmpL3. PLoS One 8:e60933. 24. Ioerger TR, O'Malley T, Liao R, Guinn KM, Hickey MJ, Mohaideen N, Murphy KC, Boshoff HI, Mizrahi V, Rubin EJ, Sassetti CM, Barry CE, 3rd, Sherman DR, Parish T, Sacchettini JC. 2013. Identification of new drug targets and resistance mechanisms in Mycobacterium tuberculosis. PLoS One 8:e75245. 25. Feng X, Zhu W, Schurig-Briccio LA, Lindert S, Shoen C, Hitchings R, Li J, Wang Y, Baig N, Zhou T, Kim BK, Crick DC, Cynamon M, McCammon JA, Gennis RB, Oldfield E. 2015. Antiinfectives targeting enzymes and the proton motive force. Proc Natl Acad Sci U S A 112:E7073-7082. 26. Li K, Schurig-Briccio LA, Feng X, Upadhyay A, Pujari V, Lechartier B, Fontes FL, Yang H, Rao G, Zhu W, Gulati A, No JH, Cintra G, Bogue S, Liu YL, Molohon K, Orlean P, Mitchell DA, Freitas-Junior L, Ren F, Sun H, Jiang T, Li Y, Guo RT, Cole ST, Gennis RB, Crick DC, Oldfield E. 2014. Multitarget drug discovery for tuberculosis and other infectious diseases. J Med Chem 57:3126-3139. 27. Johnson BK, Colvin CJ, Needle DB, Mba Medie F, Champion PA, Abramovitch RB. 2015. The carbonic anhydrase inhibitor ethoxzolamide inhibits the Mycobacterium tuberculosis PhoPR regulon and Esx-1 secretion and attenuates virulence. Antimicrob Agents Chemother 59:4436-4445. 184 28. Zheng H, Colvin CJ, Johnson BK, Kirchhoff PD, Wilson M, Jorgensen-Muga K, Larsen SD, Abramovitch RB. 2017. Inhibitors of Mycobacterium tuberculosis DosRST signaling and persistence. Nat Chem Biol 13:218-225. 29. Mak PA, Rao SP, Ping Tan M, Lin X, Chyba J, Tay J, Ng SH, Tan BH, Cherian J, Duraiswamy J, Bifani P, Lim V, Lee BH, Ling Ma N, Beer D, Thayalan P, Kuhen K, Chatterjee A, Supek F, Glynne R, Zheng J, Boshoff HI, Barry CE, 3rd, Dick T, Pethe K, Camacho LR. 2012. A high-throughput screen to identify inhibitors of ATP homeostasis in non-replicating Mycobacterium tuberculosis. ACS Chem Biol 7:11901197. 30. McNeil MB, Dennison D, Parish T. 2017. Mutations in MmpL3 alter membrane potential, hydrophobicity and antibiotic susceptibility in Mycobacterium smegmatis. Microbiology 163:1065-1070. 31. Rao SP, Alonso S, Rand L, Dick T, Pethe K. 2008. The protonmotive force is required for maintaining ATP homeostasis and viability of hypoxic, nonreplicating Mycobacterium tuberculosis. Proc Natl Acad Sci U S A 105:11945-11950. 32. Jaladanki CK, Taxak N, Varikoti RA, Bharatam PV. 2015. Toxicity originating from thiophene containing drugs: exploring the mechanism using quantum chemical methods. Chem Res Toxicol 28:2364-2376. 33. Cox JA, Abrahams KA, Alemparte C, Ghidelli-Disse S, Rullas J, Angulo-Barturen I, Singh A, Gurcha SS, Nataraj V, Bethell S, Remuinan MJ, Encinas L, Jervis PJ, Cammack NC, Bhatt A, Kruse U, Bantscheff M, Futterer K, Barros D, Ballell L, Drewes G, Besra GS. 2016. THPP target assignment reveals EchA6 as an essential fatty acid shuttle in mycobacteria. Nat Microbiol 1:15006. 34. Pethe K, Bifani P, Jang J, Kang S, Park S, Ahn S, Jiricek J, Jung J, Jeon HK, Cechetto J, Christophe T, Lee H, Kempf M, Jackson M, Lenaerts AJ, Pham H, Jones V, Seo MJ, Kim YM, Seo M, Seo JJ, Park D, Ko Y, Choi I, Kim R, Kim SY, Lim S, Yim SA, Nam J, Kang H, Kwon H, Oh CT, Cho Y, Jang Y, Kim J, Chua A, Tan BH, Nanjundappa MB, Rao SP, Barnes WS, Wintjens R, Walker JR, Alonso S, Lee S, Kim J, Oh S, Oh T, Nehrbass U, Han SJ, No Z, et al. 2013. Discovery of Q203, a potent clinical candidate for the treatment of tuberculosis. Nat Med 19:1157-1160. 35. Johnson BK, Abramovitch RB. 2015. Macrophage infection models for Mycobacterium tuberculosis. Methods Mol Biol 1285:329-341. 36. Williams EA, Mba Medie F, Bosserman RE, Johnson BK, Reyna C, Ferrell MJ, Champion MM, Abramovitch RB, Champion PA. 2017. A nonsense mutation in Mycobacterium marinum that is suppressible by a novel mechanism. Infect Immun 85:e00653-16 185 37. McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, Garimella K, Altshuler D, Gabriel S, Daly M, DePristo MA. 2010. The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res 20:1297-1303. 38. Baker JJ, Johnson BK, Abramovitch RB. 2014. Slow growth of Mycobacterium tuberculosis at acidic pH is regulated by phoPR and host-associated carbon sources. Mol Microbiol 94:56-69. 39. Schneider CA, Rasband WS, Eliceiri KW. 2012. NIH Image to ImageJ: 25 years of image analysis. Nat Methods 9:671-675. 40. Johnson BK, Scholz MB, Teal TK, Abramovitch RB. 2016. SPARTA: Simple program for automated reference-based bacterial RNA-seq transcriptome analysis. BMC Bioinformatics 17:66. 41. Kelley LA, Mezulis S, Yates CM, Wass MN, Sternberg MJE. 2015. The Phyre2 web portal for protein modeling, prediction and analysis. Nature Protocols 10:845. 186 CHAPTER 5 – Concluding Remarks Bacterial persistence is increasingly recognized as a threat to human health, as it may contribute to infectious disease treatment failure. Although currently available antibiotics are effective at killing replicating bacteria, the remaining persisters or non-replicating bacteria that survive antibiotic treatment represent a reservoir that can promote disease relapse. Additionally, survival of persisters can select for the evolution of drug-resistant bacterial strains, which further exacerbates the control of infectious diseases. Therefore, bacterial persistence is a fundamental problem in infectious disease research. Approaches for controlling bacterial persistence or eradicating the survivors of drug treatment represent important new avenues for drug development. This dissertation demonstrates new discoveries that may be applied to control bacterial persistence. Mycobacterium tuberculosis (Mtb) is an excellent pathogen to study bacterial persistence because Mtb is well-known for shifting its physiology to a non-replicating persistence (NRP) state to survive in stressful environmental conditions. Although Mtb is a strict aerobe, it can remain viable for years in anaerobic conditions during dormancy (1). Non-replicating Mtb naturally tolerates many anti-mycobacterial drugs (2), and is a driving force of the long course of TB therapy. Therefore, one strategy to shorten and improve TB treatment is to inhibit Mtb from establishing and maintaining NRP. Mtb uses the DosRST two-component regulatory system to promote NRP and enable it to establish and maintain a chronic infection. DosRST is not required in vitro under conditions of mild hypoxia, but is required for full virulence in vivo (1). Animal model studies show that deletion of dosR significantly reduces Mtb survival in different animal species that form hypoxic granulomas, including C3Heb/FeJ mice, non-human primates, and guinea pigs, suggesting the 187 DosRST pathway is a valid drug target (3-5). This dissertation presents a new approach to target the DosRST pathway using small molecule inhibitors discovered from whole cell phenotypic highthroughput screening (HTS). The reporter strain CDC1551 (hspX’::GFP) was used to screen for compounds that inhibit DosR-driven GFP fluorescence under hypoxic conditions, while having minimal impact on Mtb growth (Chapter 2). Six novel molecules HC101A-HC106A with distinct chemical structures were identified and characterized as DosRST inhibitors. To our knowledge, these are the first small molecules that inhibit the DosRST pathway by the presented mechanisms. The success of this HTS validates the use of a reporter strain to screen for virulence inhibitors. A series of physiology studies show that DosRST inhibitors can inhibit persistence-associated physiologies, including triacylglerol biosynthesis, survival and antibiotic tolerance, providing proof-of-concept evidence that inhibiting Mtb from establishing or maintaining persistence is an effective way to kill Mtb (Chapter 2-3). Biochemical studies revealed novel mechanisms of action for the compounds (Chapter 23). HC101A is the first-line antimalarial drug artemisinin that targets DosS/T heme, by modulating the heme iron redox state and alkylating the heme carried by the sensor histidine kinases. Interestingly, HC106A also interacts with the heme of DosS/T, likely by directly binding to the ferrous ion in a manner similar to natural ligands, such as carbon monoxide (CO) and nitric oxide (NO). Although both HC101A and HC106A target the heme of DosST, their inhibitory mechanisms are distinct. This result demonstrates that the heme is particularly vulnerable to chemical modulation to repress the DosRST pathway. This dissertation further shows that HC102A and HC103A may target the kinase domains of DosST by inhibiting their autophosphorylation activity, although further genetic and biochemical studies are required to define the precise mechanism of action. HC104A disrupts DosR DNA binding activity, without 188 modulating sensor kinase heme or autophosphorylation activity. Overall, these mechanistic studies demonstrate there are multiple mechanisms by which chemical probes can disrupt DosRST signaling. This dissertation also shows that characterization of small molecules can be applied as chemical biology tools to study protein structure and function. For instance, modeling of the protein DosT with artemisinin revealed a channel on the protein that artemisinin can access to interact with the heme. To support the modeling results, amino acid substitutions E85L and G115L in DosS or G87L and G117L in DosT were tested in biochemical assays and shown to block the channel and provide resistance to artemisinin (Chapter 2). This study provides evidence that the channel is important for the protein function and can be targeted by small molecules and thus opens the door to screen for or design small molecules that target this channel. For instance, any molecules that can enter the channel and block it or modulate the heme may efficiently inactivate the protein DosS/T from sensing host signals. Moreover, HC106A binds to the ferrous ion of the heme like CO or NO (Chapter 3). Interestingly, CO or NO binding of heme activates the protein (6), whereas HC106A presumably inactivates the protein. Future structural biology comparisons examining conformational changes of DosS/T induced by the binding of HC106A, CO or NO can be performed to elucidate the mechanism of kinase activation. This information is important because the structural profiles can be used to design drugs that lock the protein in the inactive form via simulation. Furthermore, incorporating the protein channel information and ligand binding, more novel drugs can be designed to target the sensor domains of heme-based histidine kinases. This study also supports the hypothesis that other heme-based sensor kinases may also be targeted by similar mechanisms. For instance, the newly discovered NO sensor, NosP, is a hemecontaining sensor protein that regulates biofilm formation and quorum sensing in many bacteria, 189 including Pseudomonas aeruginosa, Vibrio. cholera, Legionella pneumophilia and Shewanella oneidensis (7, 8). Binding of NO to NosP inhibits the activity of the associated histidine kinase, and thus blocks phosphate flux to activate the signaling pathways. NtrY/NtrX is the twocomponent system that regulates gene expression of a nitrogen respiration enzyme in Brucella abortus (9). This system functions more like DosRST, where NtrY is a redox sensor that employs the heme group in the PAS domain to sense the environmental cues, including NO, CO and oxygen level. It can be quickly oxidized in the presence of oxygen, whereas it is activated under low oxygen level to induce the NtrX regulon. RcoM is a single-component transcriptional regulator that induces the CO oxidation operon in responding to CO in Burkholderia xenovorans (10). It has a PAS sensor domain in the N-terminus, and a LytTR DNA-binding domain in the C-terminus. The heme group of RcoM can bind to CO to activate the protein for binding to the targeted genes. In the future, we hope to examine the impact of heme-targeting molecules on these diverse environmental sensors. Two-component regulatory systems play essential roles in bacterial signaling, and are often required for bacterial pathogenesis. Many inhibitors have been discovered and proven as an effective way to control bacterial pathogenesis (11-13). An early study found the chemical inhibitors of the AlgR2/AlgR1 two-component regulatory system that is involved in the regulation of alginate production in P. aeruginosa (14). These inhibitors repressed autophosphorylation activity of AlgR2, analogous to how we propose HC102A and HC103A may inhibit DosST, and another inhibitor targeted DNA-binding activity of AlgR1, analogous to how we think HC104A may target DosR. They have therapeutic potential to treat P. aeruginosa infections in cystic fibrosis patients (14). Another notable inhibitor, LED209, was found to have a broad spectrum of activity against multiple Gram-negative bacteria by inhibiting the conserved histidine kinase, 190 QseC (15). QseC can activate multiple response regulators that control the induction of the locus enterocyte effacement, as well as Shiga toxin production in different bacterial species. Animal studies show that LED209 significantly improves survival of mice infected with Salmonella typhimurim and Francisella tularensis (16). Ethoxzolamide was discovered as an inhibitor of the PhoPR system involved in sensing several environmental cues, such as acidic pH and chloride (17). Mtb treated with ethoxzolamide have reduced virulence lipid production, as well as inhibition of the Esx-1 secretion system. Mice treated with ethoxzolamide also have reduced bacterial burdens following TB infection. Together, data presented in this dissertation, along with these other examples support the conclusion that two-component regulatory systems are good drug targets to treat pathogenic bacteria to control infectious diseases. Additionally, this dissertation discovered an impact of artemisinin on Mtb. The recent advances in antimalarials have identified several novel synthetic endoperoxides that are effective as antimalarials (18-22). It would be interesting to see if those synthetic endoperoxides also have an impact on modulating the DosRST signaling pathway. Furthermore, modification of the chemical structure to become more TB specific would be ideal, so there is reduced overlap between malaria and TB treatments. Co-infection of malaria and TB is not commonly reported, but cases have been described in some developing countries (23). Given the high prevalence of latent TB in the developing world, it is highly likely many individuals with malaria are chronically infected with TB. Findings from this dissertation raise potential concerns about the impact of artemisinin treatment on TB in co-infected patients. It is unknown whether treating co-infected patients with artemisinin to cure malaria has any impact on latent TB. There might exist a risk that artemisinin treatment can reactivate Mtb from dormancy by inhibiting the DosRST pathway. 191 Lastly, this dissertation describes the discovery and characterization of another compound, HC2091, that kills Mtb by targeting MmpL3 (Chapter 4). Many MmpL3 inhibitors have been reported in the literature, however many of the previously discovered inhibitors have liabilities including poor solubility (24, 25). HC2091 exhibits no cross-resistance with the well characterized MmpL3 inhibitor SQ109, and the HC2091 resistance mutations are observed in distinct areas of the gene. Together, these data suggest that the compounds might use different mechanisms to inhibit MmpL3. HC2091 has high solubility (>200 µM) and low cytotoxicity, characteristics that support continued studies of HC2091 as a candidate for treating TB disease. Further efforts to optimize potency and ascertain drug metabolism and pharmacokinetic properties of HC2091 are required to determine if it could be pursued as a potential TB therapeutic. There remain significant future studies that will need to be undertaken to follow up the research presented in this dissertation. This dissertation presents the initial discovery and characterization of the DosRST inhibitors HC101A-HC106A, and shows that these inhibitors reduce Mtb survival and antibiotic tolerance in vitro. However, in order to further develop these inhibitors, several additional experiments are required. For instance, testing DosRST inhibitors in TB infection animal models, such as C3Heb/FeJ mice, and specifically examining the ability of these compounds to reduce Mtb survival and antibiotic tolerance. Large scale structure-function relationship studies should be performed for the compounds discovered in this dissertation, including DosRST inhibitors and HC2091, in order to optimize their potency and pharmacokinetic properties and to prioritize compounds for animal efficacy studies. This dissertation presents chemical biology based approaches to define novel mechanisms of inhibiting the Mtb two-component regulatory system DosRST. The studies provide proof-ofconcept evidence supporting our original hypothesis that inhibiting Mtb from establishing NRP 192 via chemical inhibition of the DosRST pathway will reduce Mtb survival and antibiotic tolerance. In addition, I have shown multiple, distinct mechanisms of action for the inhibitors to repress DosRST signaling. This work has furthered our understanding of how to target two-component regulatory systems that are important for virulence. Moreover, I discovered an MmpL3 inhibitor that is effective against Mtb. Together, these studies advance our knowledge of Mtb physiology and biochemistry. The findings support new avenues to combat this deadly disease and, thus, this research has the potential to improve TB treatment and global health. 193 REFERENCES 194 REFERENCES 1. Leistikow RL, Morton RA, Bartek IL, Frimpong I, Wagner K, Voskuil MI. 2010. The Mycobacterium tuberculosis DosR regulon assists in metabolic homeostasis and enables rapid recovery from nonrespiring dormancy. J Bacteriol 192:1662-1670. 2. Rao SP, Alonso S, Rand L, Dick T, Pethe K. 2008. The protonmotive force is required for maintaining ATP homeostasis and viability of hypoxic, nonreplicating Mycobacterium tuberculosis. Proc Natl Acad Sci U S A 105:11945-11950. 3. Converse PJ, Karakousis PC, Klinkenberg LG, Kesavan AK, Ly LH, Allen SS, Grosset JH, Jain SK, Lamichhane G, Manabe YC, McMurray DN, Nuermberger EL, Bishai WR. 2009. Role of the dosR-dosS two-component regulatory system in Mycobacterium tuberculosis virulence in three animal models. Infect Immun 77:12301237. 4. Gautam US, McGillivray A, Mehra S, Didier PJ, Midkiff CC, Kissee RS, Golden NA, Alvarez X, Niu T, Rengarajan J, Sherman DR, Kaushal D. 2015. DosS is required for the complete virulence of Mycobacterium tuberculosis in mice with classical granulomatous lesions. Am J Respir Cell Mol Biol 52:708-716. 5. Mehra S, Foreman TW, Didier PJ, Ahsan MH, Hudock TA, Kissee R, Golden NA, Gautam US, Johnson AM, Alvarez X, Russell-Lodrigue KE, Doyle LA, Roy CJ, Niu T, Blanchard JL, Khader SA, Lackner AA, Sherman DR, Kaushal D. 2015. The DosR regulon modulates adaptive immunity and is essential for Mycobacterium tuberculosis persistence. Am J Respir Crit Care Med 191:1185-1196. 6. Kumar A, Toledo JC, Patel RP, Lancaster JR, Jr., Steyn AJ. 2007. Mycobacterium tuberculosis DosS is a redox sensor and DosT is a hypoxia sensor. Proc Natl Acad Sci U S A 104:11568-11573. 7. Hossain S, Boon EM. 2017. Discovery of a novel nitric oxide binding protein and nitricoxide-responsive signaling pathway in Pseudomonas aeruginosa. ACS Infect Dis 3:454461. 8. Williams DE, Nisbett LM, Bacon B, Boon E. 2017. Bacterial heme-based sensors of nitric oxide. Antioxid Redox Signal 00:1-16. 9. Carrica Mdel C, Fernandez I, Marti MA, Paris G, Goldbaum FA. 2012. The NtrY/X two-component system of Brucella spp. acts as a redox sensor and regulates the expression of nitrogen respiration enzymes. Mol Microbiol 85:39-50. 10. Kerby RL, Youn H, Roberts GP. 2008. RcoM: a new single-component transcriptional regulator of CO metabolism in bacteria. J Bacteriol 190:3336-3343. 195 11. Gotoh Y, Eguchi Y, Watanabe T, Okamoto S, Doi A, Utsumi R. 2010. Two-component signal transduction as potential drug targets in pathogenic bacteria. Curr Opin Microbiol 13:232-239. 12. Matsushita M, Janda KD. 2002. Histidine kinases as targets for new antimicrobial agents. Bioorg Med Chem 10:855-867. 13. Johnson BK, Abramovitch RB. 2017. Small molecules that sabotage bacterial virulence. Trends Pharmacol Sci 38:339-362. 14. Roychoudhury S, Zielinski NA, Ninfa AJ, Allen NE, Jungheim LN, Nicas TI, Chakrabarty AM. 1993. Inhibitors of two-component signal transduction systems: inhibition of alginate gene activation in Pseudomonas aeruginosa. Proc Natl Acad Sci U S A 90:965-969. 15. Rasko DA, Moreira CG, Li de R, Reading NC, Ritchie JM, Waldor MK, Williams N, Taussig R, Wei S, Roth M, Hughes DT, Huntley JF, Fina MW, Falck JR, Sperandio V. 2008. Targeting QseC signaling and virulence for antibiotic development. Science 321:1078-1080. 16. Curtis MM, Russell R, Moreira CG, Adebesin AM, Wang C, Williams NS, Taussig R, Stewart D, Zimmern P, Lu B, Prasad RN, Zhu C, Rasko DA, Huntley JF, Falck JR, Sperandio V. 2014. QseC inhibitors as an antivirulence approach for Gram-negative pathogens. MBio 5:e02165. 17. Johnson BK, Colvin CJ, Needle DB, Mba Medie F, Champion PA, Abramovitch RB. 2015. The carbonic anhydrase inhibitor ethoxzolamide inhibits the Mycobacterium tuberculosis PhoPR regulon and Esx-1 secretion and attenuates virulence. Antimicrob Agents Chemother 59:4436-4445. 18. Charman SA, Arbe-Barnes S, Bathurst IC, Brun R, Campbell M, Charman WN, Chiu FC, Chollet J, Craft JC, Creek DJ, Dong Y, Matile H, Maurer M, Morizzi J, Nguyen T, Papastogiannidis P, Scheurer C, Shackleford DM, Sriraghavan K, Stingelin L, Tang Y, Urwyler H, Wang X, White KL, Wittlin S, Zhou L, Vennerstrom JL. 2011. Synthetic ozonide drug candidate OZ439 offers new hope for a single-dose cure of uncomplicated malaria. Proc Natl Acad Sci U S A 108:4400-4405. 19. Vennerstrom JL, Arbe-Barnes S, Brun R, Charman SA, Chiu FC, Chollet J, Dong Y, Dorn A, Hunziker D, Matile H, McIntosh K, Padmanilayam M, Santo Tomas J, Scheurer C, Scorneaux B, Tang Y, Urwyler H, Wittlin S, Charman WN. 2004. Identification of an antimalarial synthetic trioxolane drug development candidate. Nature 430:900-904. 20. O'Neill PM, Amewu RK, Nixon GL, Bousejra ElGarah F, Mungthin M, Chadwick J, Shone AE, Vivas L, Lander H, Barton V, Muangnoicharoen S, Bray PG, Davies J, Park BK, Wittlin S, Brun R, Preschel M, Zhang K, Ward SA. 2010. Identification of a 196 1,2,4,5-tetraoxane antimalarial drug-development candidate (RKA 182) with superior properties to the semisynthetic artemisinins. Angew Chem Int Ed Engl 49:5693-5697. 21. Rudrapal M, Chetia D. 2016. Endoperoxide antimalarials: development, structural diversity and pharmacodynamic aspects with reference to 1,2,4-trioxane-based structural scaffold. Drug Des Devel Ther 10:3575-3590. 22. O'Neill PM, Amewu RK, Charman SA, Sabbani S, Gnadig NF, Straimer J, Fidock DA, Shore ER, Roberts NL, Wong MH, Hong WD, Pidathala C, Riley C, Murphy B, Aljayyoussi G, Gamo FJ, Sanz L, Rodrigues J, Cortes CG, Herreros E, AnguloBarturen I, Jimenez-Diaz MB, Bazaga SF, Martinez-Martinez MS, Campo B, Sharma R, Ryan E, Shackleford DM, Campbell S, Smith DA, Wirjanata G, Noviyanti R, Price RN, Marfurt J, Palmer MJ, Copple IM, Mercer AE, Ruecker A, Delves MJ, Sinden RE, Siegl P, Davies J, Rochford R, Kocken CHM, Zeeman AM, Nixon GL, Biagini GA, Ward SA. 2017. A tetraoxane-based antimalarial drug candidate that overcomes PfK13-C580Y dependent artemisinin resistance. Nat Commun 8:15159. 23. Li XX, Zhou XN. 2013. Co-infection of tuberculosis and parasitic diseases in humans: a systematic review. Parasit Vectors 6:1-12. 24. Jia L, Tomaszewski JE, Hanrahan C, Coward L, Noker P, Gorman G, Nikonenko B, Protopopova M. 2005. Pharmacodynamics and pharmacokinetics of SQ109, a new diamine-based antitubercular drug. Br J Pharmacol 144:80-87. 25. Jia L, Noker PE, Coward L, Gorman GS, Protopopova M, Tomaszewski JE. 2006. Interspecies pharmacokinetics and in vitro metabolism of SQ109. Br J Pharmacol 147:476485. 197