screening for broad spectrum antimicrobials with unknown targets... · 2019-02-13 · screening for...

Screening for Broad Spectrum Antimicrobials with Unknown Targets

A dissertation presented

by

Laura E. Fleck

to

The Department of Biology

In partial fulfillment of the requirements for the degree of

Doctor of Philosophy

in the field of

Biology

Northeastern University

Boston, Massachusetts

October 2, 2013

2

Screening for Broad Spectrum Antimicrobials with Unknown Targets

by

Laura E. Fleck

ABSTRACT OF DISSERTATION

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biology

in the College of Science of Northeastern University

October 2, 2013

3

ABSTRACT

The misuse of antibiotics combined with pathogen tolerance has led to an increase in patients with

multidrug resistant infections, relapse of infections, and fewer effective treatments. We propose

that prodrugs hold potential as the next broad spectrum antimicrobial. Prodrugs are able to diffuse

into the cell where they are converted into a reactive compound by bacterial-specific enzymes. We

developed a unique high-throughput screening assay to identify prodrugs. We hypothesized that

since prodrugs have multiple targets they would rapidly abolish metabolism resulting in cell death.

The viability dye alamar blue was used to test this hypothesis and measured the metabolism of

cells challenged with antiseptics, antibiotics, and the prodrug Nitazol. We employed a high-

throughput screen for broad spectrum antimicrobials using Escherichia coli (E. coli) and

Staphylococcus aureus (S. aureus), and candidate prodrug hits were tested for cytotoxicity and

minimum inhibitory concentrations (MIC) against a panel of pathogens. A validation step

confirmed a strain lacking a converting enzyme was more resistant to a prodrug and a strain over

expressing a converting enzyme was more susceptible. Prodrugs appear to rapidly abolish

metabolism and produce a distinct kinetic curve in the alamar blue reduction screen. Three

compounds were identified as potential broad spectrum prodrugs from this screen. Compound

ADC111, a nitrofuran compound, shows broad spectrum activity and mimimal cytotoxicity against

four mammalian cell lines. In E. coli, ADC111 is converted by nitroreductases NfsA and NfsB.

The second prodrug candidate, ADC112, is an 8-hydroxyquinoline, a class of broad spectrum

antimicrobials where the mechanism of action is unknown. The third prodrug candidate, ADC113,

does not belong to a known class of approved antimicrobials. Different classes of antimicrobials

have different effects on bacterial metabolism, which can be differentiated using the alamar blue

reduction screen. This screen can identify compounds with unknown targets.

4

ACKNOWLEDGEMENTS Over the past five years many people have helped, encouraged, and advised me; I am grateful to

you all. I would like to thank Kim Lewis, my advisor, for all of his imparted knowledge, guidance,

patience, and the amazing opportunity to work with such an intelligent and fun group of scientists

at the Antimicrobial Discovery Center (ADC). Thank you to all of my committee members, Dr’s

Richard Lee, Veronica Godoy-Carter, Eric Stewart, and Gabriele Casadei. Richard- thank you for

always sharing your expert opinion and making the time to travel to Boston. I owe many thanks to

Dr. Jeffrey North at St. Jude Children’s Research Hospital for always synthesizing more

compounds for me. Thank you to Gabriele Casadei for all of your advice, humor, and for teaching

me how to work with animals. Thank you Eric Stewart, for always taking, and making, the time to

read my drafts, abstracts, and lend a helping mind. Thank you to Ken Coleman, Katya Gavrish,

Michael LaFleur and Anna-Barbara Hachmann for sharing your thoughts and ideas which have

helped shape this dissertation. Thank you to Chao Chen, my drug discovery partner in crime, for

everything; especially for your friendship, help with compound dispensing, and starting cultures.

Gabriele, Chao, Mike, Brian Conlon, and Pooja Balani- thank you for letting me be a part of your

experiments, it was a privilege to work and learn alongside all of you. Marin Vulić- thank you for

tirelessly answering my questions regarding genetics (and for making the best macaroon cookies).

Thank you Larry MulCahy for teaching me about whole genome sequencing analysis, your help

with the cell sorter, and all of your constructive comments over the years. Alyssa Theodore- thank

you for all of the Sunday cultures you started for me, always double checking my genetics, great

conversations, and countless fun times. Thank you to everyone in the ADC, past and present, you

have all made the past five years remarkable. I would like to thank Peter Brannen for always

believing in me, and for his feedback and patience while I wrote this dissertation! Lastly, I would

like to thank my parents for their perpetual love and support during my seemingly “never-ending”

education.

5

TABLE OF CONTENTS

ABSTRACT...............................................................................................................................2

ACKNOWLEDGEMENTS ......................................................................................................4

TABLE OF CONTENTS ..........................................................................................................5

LIST OF ABBREVIATIONS ...................................................................................................8

LIST OF FIGURES ..................................................................................................................9

LIST OF TABLES .................................................................................................................. 11

Chapter 1: Introduction .......................................................................................................... 13

1.1 History of Antibiotic Discovery ................................................................................ 14

1.2 The Different Classes of Antibiotics .............................................................................. 16

1.3 Antimicrobial Resistance ............................................................................................... 17

1.4 Chronic infections and Bacterial persister cells ............................................................ 19

1.5 Prodrugs ......................................................................................................................... 21

1.6 Dissertation aims ............................................................................................................ 24

Chapter 2: ................................................................................................................................ 25

Developing a Novel Screen for Prodrug Antimicrobials........................................................ 25

2.1 Introduction ................................................................................................................... 26

2.2 Results ............................................................................................................................ 28

The initial model screen for prodrugs- a targeted prodrug screen based on strains

diminished in activating enzymes. ................................................................................... 28

6

Prodrug screening based on essential protein overexpression ....................................... 36

The screen for compounds with non-specific or unknown targets. ................................ 42

Validation of the alamar blue reduction screen for prodrugs. ....................................... 48

2.3 Discussion ....................................................................................................................... 50

2.4 Materials and Methods .................................................................................................. 52

Chapter 3: ................................................................................................................................ 56

Mechanism of Action and Medicinal Chemistry Optimization of Prodrug Candidates ...... 56

3.1 Introduction ................................................................................................................... 57

3.2 Results ............................................................................................................................ 58

Determining the Mechanism of Action (MOA) of ADC111 ........................................... 58

Selection of resistant mutants to ADC111 ....................................................................... 60

Determining the mechanism of action for ADC112. ....................................................... 61


Determining the mechanism of action for ADC113 ........................................................ 64

Potential targets of ADC113 reactive species .................................................................. 69


3.3 Discussion ....................................................................................................................... 73

3.4 Materials and Methods .................................................................................................. 75

Chapter 4: ................................................................................................................................ 77

A Screen for Prodrug Antimicrobials .................................................................................... 77

7

4.1 Abstract .......................................................................................................................... 78

4.2 Introduction ................................................................................................................... 79

4.3 Results ............................................................................................................................ 81

A screen for prodrugs. ..................................................................................................... 81

Hit validation. ................................................................................................................... 85

Bactericidal activity of hit compounds. ........................................................................... 90

4.4 Discussion ..................................................................................................................... 100

4.5 Materials and Methods ................................................................................................ 103

Chapter 5: .............................................................................................................................. 109

Discussion .............................................................................................................................. 109

5.1 Discussion: .................................................................................................................... 110

Literature Cited .................................................................................................................... 112

8

LIST OF ABBREVIATIONS

CCCP: carbonyl cyanide m-chlorophenyl hydrazine

CHX: Chlorhexidine

CIP: Ciprofloxacin

DNA: Deoxyribonucleic acid

HTS: High Throughput Screening

KAN: Kanamycin

KO: Knock-out

MBC: Minimum Bactericidal Concentration

MDR: Multidrug Resistant Efflux Pump

MIC: Minimum Inhibitory Concentration

MOA: Mechanism of Action

NFT: Nitrofuantoin

NFZ: Nitrofurazone

OE: Over-expressing

PCR: Polymerase Chain Reaction

QIRs: Quiescent Intracellular Reservoirs

RNA: Ribonuclei acid

RNAi: RNA interference

SD: Standard deviation

SPCM: Spectinomycin

TI: Therapeutic Index

UTI: Urinary Tract Infection

9

LIST OF FIGURES

Figure 1-1. Targets of Antibiotics. ............................................................................................. 17

Figure 1-2. Mechanisms of antibiotic resistance. ....................................................................... 18

Figure 1-3. Scheme representing the formation of persister cells. .............................................. 19

Figure 1-4. Model of biofilm resistance to killing based on persister survival. ........................... 20

Figure 1-5. The ideal prodrug model. ........................................................................................ 22

Figure 2-1. Prodrug validation. .................................................................................................. 29

Figure 2-2. Criteria for identifying prodrug candidates in the initial model prodrug screen. ....... 31

Figure 2-3. Prodrug hits from the pilot screen of 3,000 compounds. .......................................... 31

Figure 2-4. Criteria for identifying prodrug candidates in the secondary initial model prodrug

screen. ....................................................................................................................................... 40

Figure 2-5. Hit compounds from initial prodrug pilot screen...................................................... 41

Figure 2-6. Reduction of resazurin to resorufin. ......................................................................... 42

Figure 2-7. Alamar blue reduction in E. coli as a basis for a produg screen. ............................... 44

Figure 2-8. Alamar blue reduction in B. anthracis as a basis for a produg screen. ...................... 45

Figure 2-9. Alamar blue reduction screen workflow. ................................................................. 46

Figure 2-10. Prodrugs and hit compounds. ................................................................................ 47

Figure 2-11. Inhibition of alamar blue reduction by nitrofurans. ................................................ 49

Figure 3-1. Image of the enzyme NfsA’s molecular surface ....................................................... 59

Figure 3-3. Structural analogs of ADC113................................................................................. 65

Figure 3-4. Genes of interest for ADC113 surrounding ybgJ. .................................................... 68

Figure 3-5. Effect of ADC113 on cell membrane potential. ....................................................... 70

Figure 4-1. Prodrug antibiotics. ................................................................................................. 81

10

Figure 4-2. Alamar blue reduction as a basis for a produg screen. .............................................. 83

Figure 4-3. Prodrugs and hit compounds. .................................................................................. 84

Figure 4-4. Inhibition of alamar blue reduction by nitrofurans. .................................................. 90

Figure 4-5. Time and concentration dependent killing of E. coli BW25113 in exponential phase.

................................................................................................................................................. 91

Figure 4-6. Biofilm killing by ADC111, NFT, and ciprofloxacin. .............................................. 92

Figure 4-7. Concentration dependent killing of wild type E. coli in Stationary phase. ................ 93

Figure 4-8. Biofilm killing by ADC112 and Tilbroquinol. ......................................................... 94

Figure 4-9. Killing of stationary phase E. coli with ADC112. .................................................... 95

Figure 4-10. Structural analogs of ADC113. .............................................................................. 96

Figure 4-11. Time and concentration dependent killing of exponentially growing wild type E.

coli. ........................................................................................................................................... 98

Figure 4-12. E. coli biofilm killing with ADC113...................................................................... 99

Figure 4-13. Killing of E. coli in stationary phase. ................................................................... 100

11

LIST OF TABLES

Table 1-1. Prodrugs currently used in the clinic. ........................................................................ 22

Table 2-1. Prodrug Candidates from the initial prodrug screen at ICCB. The MICs were

determined by broth microdilution where the highest concentration tested was 12.5µg/mL. ...... 31

Table 2-2. Activity of PD16 against strains lacking and overexpressing activating enzymes.

MICs were determined by broth microdilution. ......................................................................... 35

Table 2-3. List of genes with conserved essentiality in bacteria. ................................................ 36

Table 2-4. Activity of NCI28002 against strains of E. coli over expressing and lacking DapB. .. 41

Table 2-5. Results of the prodrug screen. ................................................................................... 47

Table 2-6. Activity of ADC111 against strains lacking and overexpressing activating enzymes. 49

Table 2-7. List of strains used in this study. ............................................................................... 53


Table 3-2. Estimated binding affinity of ligands to NfsA. The software PyRx calculates the

estimated binding affinity of ligands to the target NfsA. ADC111 is predicted to have the

strongest binding affinity to NfsA compared to the prodrugs Nitazol and metronidazole which

are also converted into reactive molecules by NfsA, kanamycin is included a control not known

to bind to NfsA. ........................................................................................................................ 60

Table 3-3. Activity of ADC111 against E. coli with compromised DNA repair. ........................ 61

Table 3-4. Potential converting enzymes of ADC112. ............................................................... 62

Table 3-5. MICs of ADC113 and structural analogs against wild type E. coli and efflux mutant

strains lacking TolC and EmrB. The structural analogs of ADC113 found in Figure 3-3 were

tested for activity against wild type E. coli and strains of E. coli lacking TolC and EmrB. ......... 66

Table 3-6. Potential converting enzymes and their MICs against ADC113. ............................... 67

12

Table 3-7. Comparison of MICs between the wild type, knock-out strains, and over expression

strains for ArgC and YbgJ. ........................................................................................................ 68

Table 3-8. Whole Genome Sequencing results of ADC113 resistant mutants. ~10 million 50bp

reads were mapped to the E. coli MG1655 reference sequence (GenBank:U00096.2) using CLC

Genomics Workbench (CLC Bio). Whole genome sequencing of the E. coli BW25113 was

performed at the Tufts University Genomics Core Facility with Illumina HiSeq2000 and at the

Biopolymers Facility at Harvard Medical School with Illumina HiSeq2000. Presence of a

mutation is indicated by ‘Yes’, and ‘NO’indicated no difference in sequence compared to

MG1655. ................................................................................................................................... 71

Table 3-9. Activity of ADC113 against strains lacking potential converting enzymes. ............... 73

Table 4-1. Results of the prodrug screen. ................................................................................... 83

Table 4-2. Cytotoxicity of the hit compounds. ........................................................................... 85

Table 4-3. Spectrum of activity for ADC compounds. ............................................................... 87

Table 4-4. Activity of test compounds against MDR mutant strains. .......................................... 88


Table 4-6. Activity of ADC113 and structural analogs against E. coli........................................ 97

Table 4-7. List of strains used in this study. ............................................................................. 104

13

Chapter 1: Introduction

14

1.1 History of Antibiotic Discovery

Before the time of the scientists Paul Ehrlich and Alexander Flemming, infectious diseases were

the leading cause of human morbidity and mortality in the world, and continue to be a leading

cause of death in developing countries. The work of Ehrlich and Flemming historically mark the

dawn of the antibiotic “Golden Era” (1). Antibiotics, also known as antibacterials, are compounds

or substances that kill or slow down the growth of bacteria. Antimicrobials, derived from the Greek

words ‘anti’ (against), ‘mikros’ (little) and ‘bios’ (life), refer to any substance of natural,

semisynthetic or synthetic origin that kills or inhibits the growth of microorganisms with minimal

or no damage to host cells (2, 3). More specifically, an antibiotic is a substance produced by a

microorganism which kills other microorganisms at low concentrations. People in ancient China,

Greece, Serbia, and Egypt would treat infections by pressing moldy bread and moldy soya beans

against the wounds, preceding the official scientific discovery of penicillin by Alexander

Flemming by over 2,000 years (4, 5).

The first documented microbial by-product with demonstrated antimicrobial properties was

observed by the German scientist E. de Freundenreich 1888. Freundenreich observed that a blue

pigment isolated from Bacillus pyocyaneus (now known as Pseudomonas aeruginosa) stopped the

growth of some bacteria in a test tube (6). In 1889, the pigment was dubbed “pyocyanase” by

Rudolf Emmerich and Oscar Loew. Initial excitement for pyocyanase’s effectiveness against

infectious disease of the time subsided when the compound’s instability and inherent toxicity in

patients became apparent during clinical trials. The world waited twenty more years until Ehrlich,

who had been tirelessly searching for a “magic bullet” that would selectively target a disease-

causing organism while having no negative effect on human tissue, discovered “compound 606”

in 1910. Ehrlich’s compound 606, the chemical dye arsphenamine, was the first chemical

15

compound shown to cure a human disease. Later named Salvarasan, it was used to successfully

treat syphilis, a disease caused by the spirochaete bacterium Treponema pallidum. Interestingly,

the exact structure of Salvarsan was not determined until 2005 (7).

In 1928, Alexander Flemming discovered that a species of contaminating mold of the genus

Penicillium was inhibiting the growth of staphylococci on his agar plates. The active compound

was extracted, named penicillin, and triggered a cascade of discoveries in the field of antibiotics

and antimicrobials. Gerhard J. Domagk, a German scientist working for Bayer Laboratories of the

IG Faben conglomerate, discovered and developed the first sulfonamide in 1935. This compound,

an azo dye, was also notably a prodrug, and was named Prontosil. Prontosil became the first

commercially available antibacterial and resulted in sharp declines in mortality due to meningitis,

child bed fever and pneumonia (8).

Bayer was not the only company searching for new treatments to combat bacterial infections.

Selman A. Waksman worked at Rutgers University, funded by Merk & Co., and is responsible for

setting up the first antibiotic screening platform and discovering aminoglycosides. Waksman first

isolated bacteria from soil samples by pre-treating the environment with pathogens, and then

growing the microbes under varying culture conditions and then testing them against pathogenic

bacteria. He then set up a screen where he looked for growth inhibition zones around a single

colony of an isolated soil microbe. This screen for new antibiotics began in 1943 and about three

months later Waksman isolated Steptomyces griseus, the producer of the antibiotic

Streptomycin(9). Waksman’s efforts resulted in the discovery of 20 natural product antibiotics

(10).

16

1.2 The Different Classes of Antibiotics

Antibiotics are often classified by their spectrum of activity or their mechanism of action. The

main classes of antimicrobial drugs have five main mechanisms of action. Antibiotics such as

penicillin and vancomycin target cell wall biosynthesis. Aminoglycosides and macrolides inhibit

protein biosynthesis, while quinolones like ciprofloxacin and rifamycins such as rifampicin inhibit

DNA and RNA replication. Antibiotics can also inhibit metabolic pathways, like sulfamethoxazole

which inhibits the biosynthesis of folic acid, an important metabolite in DNA biosynthesis.

Lipopeptides like Daptomycin and the peptide antibiotic polymyxin B cause bacterial cell death

by disrupting the structure of the cell membrane (1).

17

Figure 1-1. Targets of Antibiotics.

There are approximately 200 conserved essential proteins in bacteria, but the number of currently

exploited targets is very small. The most successful antibiotics hit only three targets or pathways:

the ribosome (which consists of 50S and 30S subunits), cell wall synthesis and DNA and RNA

synthesis and replication. This figure is adapted from (11).

1.3 Antimicrobial Resistance

Bacteria rapidly develop means to evade antibiotics. For every target that an antibiotic hits, bacteria

have developed a resistance mechanism. These mechanisms are generally well understood and are

depicted in Figure 1-1 (11).

18

Figure 1-2. Mechanisms of antibiotic resistance. The main mechanisms of antibacterial

resistance include the destruction of the antibiotic, modification of the target, titration of the target,

restricted penetration via an efflux pump, and bypass of targeted pathways. The transfer of a

plasmid can confer antibiotic resistance to all of these mechanisms. Adapted from (11).

Bacteria can acquire plasmids carrying genes that encode enzymes which destroy antimicrobial

agents, like β-lactamases. Β-lactamases break open the structural ring of β-lactam antibiotics,

rendering them inactive (12). Efflux pumps like the AcrAB-TolC multidrug pump, extrude

multiple classes of antibiotic out of bacterial cells, never giving the antibiotic a chance to reach its

target (13). Multidrug resistant efflux pumps (MDRs) excrete both natural product antibiotics as

well as synthetic chemotherapeutics out of bacterial cells (14).

19

1.4 Chronic infections and Bacterial persister cells

Bacterial, yeast, and cancer cell populations produce persister cells that are tolerant to

chemotherapeutic treatments (15-17). Persister cells were first observed in 1944 by Joeseph

Bigger. Bigger observed that a culture of Staphyloccus aureus treated with Penicillin had a small

number of surviving cells. When Bigger tested the susceptibility of these surviving cells he found

that they were, in fact, not resistant to Penicillin, but phenotypic variants of the wild type tolerant

to the antibiotic, that arose at a low frequency of 1 in 100 to 1 in 10,000 cells. Pathogens produce

a small subpopulation of dormant persister cells that are tolerant to antibiotics (15). Persister cells

are not resistant mutants, but phenotypic variants that arise stochastically in about 1% of biofilm

and stationary populations (18). When a population of bacterial cells is treated with an antibiotic,

bi-phasic killing results, where the bulk of the population dies and drug-tolerant persister cells

remain (FIG 1-3).

Figure 1-3. Scheme representing the formation of persister cells. A population of bacteria is

treated with bactericidal antibiotics (a) where the bulk of the population dies (blue) leaving only

resistant mutants (dashed line) or persister cells (red) alive. The frequency of isolation of persisters

as a function of the growth phase of the culture is depicted in (b), showing that a stationary phase

culture has the highest number of persister cells (adapted from Lewis, 2007).

20

Persister cells are also present in biofilms, but the biofilm matrix protects them from the immune

system, so once the concentration of the antibiotic drops, persisters resuscitate and repopulate the

biofilm, causing a relapsing infection (FIG 1-4).

Figure 1-4. Model of biofilm resistance to killing based on persister survival. Initial treatment

with antibiotic kills normal cells (colored green) in planktonic and biofilm populations. The

immune system kills planktonic persisters (colored pink), but the biofilm persister cells (colored

pink) are protected from the host defenses by the exopolymer matrix. After the antibiotic

concentration is reduced, persisters resuscitate and repopulate the biofilm and the infection

relapses (adapted from Lewis, 2007).

Biofilm infections are on the rise, largely as a result of medical intervention, and form chronic,

poorly treatable infections. Biofilms form readily on indwelling devices such as catheters and

prostheses, and are responsible for infective endocarditis, recurring urinary tract infections,

infective osteomyelitis, and the incurable infection of lungs of patients with cystic fibrosis (19).

Antibiotics depend on the immune response to clear an infection, and a chronic disease often forms

21

in immune-compromised patients. Importantly, most chronic infections recalcitrant to treatment

are caused by drug-susceptible pathogens (20). Recalcitrance to treatment results from tolerance

rather than resistance. Several mechanisms lead to dormancy in E. coli, and rely mostly on the

action of toxin/antitoxin modules. The toxins responsible for persister formation include mRNA

endonucleases (21, 22), the HipA kinase that inhibits protein synthesis by phosphorylating

elongation factor Ef-Tu (23), and TisB, which decreases the energy level of the cell by creating an

ion channel (24). Bactericidal antibiotics kill by corrupting their targets (21, 25); for example,

fluoroquinolones inhibit the re-ligation step in DNA gyrase and topoisomerase, turning the

enzymes into endonucleases (26). Targets are inactive in dormant persisters, explaining their

tolerance to antibiotics. The high degree of redundancy in the mechanisms of persister formation

precludes development of conventional target-based inhibitors.

1.5 Prodrugs

A prodrug is an inactive compound that penetrates into a cell and is converted into an active form.

The “ideal” antibacterial prodrug is converted by a bacteria-specific enzyme into a reactive

molecule. The reactive form binds covalently to multiple targets creating an irreversible sink,

which leads to accumulation of the drug over time (Fig. 1-5). The accumulation of reactive species

helps solve the permeability problem, making prodrug antibiotics broad spectrum.

22

Figure 1-5. The ideal prodrug model. The ideal prodrug is a benign compound that can diffuse in and out of a bacterial cell. The prodrug is then converted by a bacterial-specific enzyme into a reactive drug. The reactive drug then covalently binds to multiple targets, killing the cell (modified from Lewis, 2007).

All existing classes of prodrugs were discovered in the 1950s, some of which are still used today

(Table 1-1). The newest prodrug, PA-824, is a bicyclic nitroimidazofuran and is activated by

reduction of the aromatic nitro group (27). The exact mechanism of action is unknown but it is

thought that glucose-6-phosphate dehydrogenase or its deazaflavin cofactor are responsible for the

reductive activation of PA-284 (28). PA-824 was deemed safe and efficacious during a phase II

clinical trial conducted in 2011 (29).

Table 1-1. Prodrugs currently used in the clinic.

Prodrug Structure Activating enzyme (encoding gene)

Target Bacteria/Discovery

Isoniazid

Catalase-peroxidase (katG) (30)

M. tuberculosis 1951

Pyrazinamide

Pyrazinamidase/nicotinamidase (pncA) (31)


23

Nitrofurantoin

Nitroreductase (E. coli nfsA/nfsB) (32)

Aerobic & anaerobic Bacteria 1953

Nitrofurazone

Nitroreductase (E. coli nfsA/nfsB) (32)

Aerobic & anaerobic Bacteria 1955

Ethionamide

Monooxygenase (etaA) (33, 34)


Metronidazole

Nitroreductase (H. pylori rdxA/frxA) (35, 36)

Anaerobic & microaerophilic Bacteria 1959

The only prodrug with a relatively broad spectrum is metronidazole, which is converted into an

active form in bacterial cells under anaerobic conditions and acts specifically against anaerobic

species. The nitrofurans have a broad spectrum of activity in vitro, but due to their short half-life

and rapid metabolism by renal tissue, are chiefly used to treat urinary tract infections (UTIs) (3).

Nitrofurans are converted by nitroreductases (32) and nitrofurantoin (NFT) is the best known

compound in the class of nitrofurans. It is a synthetic compound which has been used to treat UTIs

since 1953. The prescribing of NFT decreased in the 1970’s due to reported adverse side effects

(37). Due to the rapid development of fluoroquinolone resistant strains, the use of NFT is being

revived. The unique mode of action of nitrofuran compounds makes co-resistance with other drug

classes unlikely. Several molecular targets in the bacterial cell delay the development of resistance

as multiple mutations in parallel are needed to survive NFT’s action. Nitrofurazone (NFZ) is also

a nitrofuran prodrug but has more toxicity issues than NFT and is less soluble. For this reason,

NFZ is used as a topical ointment and as a preventative for bacterial infection on catheters. The

24

NFZ released by the catheter is not systemically absorbed by the urethra (38). To date, there is not

a single broad spectrum prodrug antibiotic active against both aerobic and anaerobic bacteria.

1.6 Dissertation aims

The goal of this dissertation is to address two main unsolved problems in drug discovery: obtaining

broad-spectrum compounds, and developing antibiotics capable of treating chronic infections by

killing growing and persister cells. The growing and impending need for new broad spectrum

antibiotics, coupled with the need for a sterilizing antibiotic, directed this research to the

development of a novel screening platform for prodrug antimicrobials. The prodrug screen

required validation of bactericidal activity, cytotoxicity, and prodrug activity for the hit

compounds. Once compounds were confirmed for bactericidal and non-cytotoxic activity their

mechanism of action and optimization of chemical structures via medicinal chemistry was studied.

25

Chapter 2:

Developing a Novel Screen for Prodrug Antimicrobials

26

2.1 Introduction:

The initial efforts in antibacterial discovery, which yielded salvarsan, prontosil, sulfa drugs, and

trimethoprim, focused on synthetic compounds (39), but the “golden age” of antibacterial drug

discovery was a time of natural product screening (40). Selman Waksman established the first

successful screening platform in the 1940s (41). Waksman’s method was straightforward: he

isolated strains of bacteria from the soil, mostly Actinomycetes, fermented them in broth medium,

and tested if they inhibited growth of a test pathogen on an overlay plate. By the 1950s the rate of

detection of novel compounds declined, and by the 1970s the discovery of novel compounds

neared zero (11). As reviewed by Baltz (42), work during the 1950s at Merk and Lilly showed that

between 12.5% and 25% of randomly isolated Actinomycetes produced bacterial antibiotics and

that novel products were found in 0.1% of Actinomycetes cultures. Among those novel products,

clinical candidates were identified at a frequency of 2-10%. As more cultures were screened, more

“knowns” accumulated and the frequency of novel products being identified decreased by between

10-6 and 10-7 per culture by 1976.Not only has the discovery of new classes of antimicrobials

decelerated, the last class of broad spectrum antibiotics, the fluroquinolones, was discovered over

50 years ago (43).

The penetration barrier of Gram-negative bacteria is a formidable obstacle for antimicrobial drug

discovery. The envelope of Gram-negative bacteria contains an outer membrane that restricts

penetration of amphipathic compounds, which essentially all drugs are. MDR pumps bind

chemically unrelated amphipathic compounds and extrude them across the outer membrane (44,

45). Efforts to develop discovery platforms for synthetic compounds based on a combination of

new target identification, high-throughput screening (HTS) of compound libraries and rational

design have not been successful- most of the hits did not penetrate into cells of Gram-negative

27

pathogens (46). Blocking efflux is a logical strategy, but so far the only type of MDR inhibitors

with a good spectrum that were identified are polycations, and these compounds have

nephrotoxicity (47, 48). The crystal structures of MDR pumps were obtained (49-52), but did not

prove particularly useful for designing better inhibitors or molecules able to evade the pumps. The

large, poorly structured binding sites of the E. coli AcrAB-TolC pump can accommodate a vast

variety of chemically unrelated compounds (53); without new compounds being introduced,

Gram-negative pathogens have been acquiring resistance largely unchecked, and are now the main

problem in treating infectious diseases. Some pathogens, such as Acinetobacter baumannii (A.

baumannii) or Klebsiella pneumonia (K. pneumonia), are resistant to all available antibiotics (54,

55). The threat of bioweapons based on genetically engineered pathogens such as Yersinia pestis

(Y. pestis) adds to the urgency of the problem.

Not only have no new classes of broad spectrum antimicrobials been discovered in the last 50

years, but no new prodrugs have been discovered since 1959 (11). This is an interesting paradox

directly related to a number of tests implemented during the years following the golden era. These

tests were aimed at discarding toxic and “unattractive” molecules. The tests included cytotoxicity

and specificity tests; the latter follows the effect of test compounds on the rate of label

incorporation into major biopolymers. If a compound inhibits only a certain biosynthetic process

such as protein synthesis, this is a “desirable” antimicrobial hitting a specific target. Compounds

that inhibit all biosynthesis simultaneously were considered antiseptics and discarded. These

criteria eliminate prodrugs as well, since in the specificity test they behave as blindly toxic

antiseptics, leading to the discovery paradox. Since prodrugs were most likely thrown out in earlier

screens, there is an opportunity to discover the missed prodrugs in existing compound libraries.

28

There are two main methods of drug discovery – whole cell, and targeted screening. Rational

structure-based drug design can be viewed as a version of targeted screening with a limited set of

test compounds. Ingenious approaches have been introduced in antimicrobial drug discovery that

combine whole cell and targeted screening (56). It has been reasoned that a cell expressing an

essential protein at a lower level will be highly susceptible to an inhibitor of that target. Thus, the

company, Microcide/Essential Therapeutics, used ts mutants of essential enzymes of

Staphylococcus aureus (S. aureus) to identify hits with a higher activity against these strains (57,

58). The challenge is to design a HTS that will specifically identify prodrugs, will differentiate

them from both generally toxic compounds, and target-specific antibiotics. Our initial screening

approach to identify prodrugs was designed to also report on the nature of the activating enzyme

based on increased susceptibility of a strain over expressing a potential activating enzyme to a

prodrug; or decreased susceptibility of a strain deficient in an activating enzyme. This initial screen

proved too specific for identifying prodrugs with non-specific targets. The rationale for our current

screen for prodrugs is an inverted specificity test: compounds which lack specificity are desired

hits. In order to make the screen practical, a viability dye is used instead of radio label

incorporation. A subsequent cytotoxicity test against mammalian cells then differentiates between

prodrug candidates and generally toxic compounds.

2.2 Results

The initial model screen for prodrugs- a targeted prodrug screen based on strains diminished in activating enzymes.

A pilot screen using the full complement of 4,320 E. coli knockouts of genes that are non-essential

in vitro from the KEIO library was performed. A prodrug should have higher activity against a

strain overexpressing an activating enzyme, and lower activity against a strain attenuated in this

29

enzyme (Fig. 2-1). This will discriminate prodrugs from all other compounds and will serve to

validate the hits.

Figure 2-1. Prodrug validation. Depiction of why over expressing (a) and deleting (b) a converting enzyme and testing the MIC is used as validation for a potential converting enzyme. More converting enzymes equals more reactive molecule, and no converting enzyme yields no reactive molecule.

The rationale of this screening modality was to identify compounds that have a lower activity

against a strain deleted in an activating enzyme as compared to the wild type. A complete, ordered

E. coli K12 knockout library of 4,320 genes and predicted ORFs (the KEIO library) was provided

to the PI by Dr. Hirotada Mori (Baba et al., 2006). All strains of this library were combined in a

mix (BacPool1) for screening. This allowed for screening of the library against all strains

simultaneously, instead of each one at a time, which would have been impractical. Once

compounds that inhibit growth of the wild type, but not the mix were obtained, we then attempted

to identify the resistant strains lacking an activating enzyme for each compound in a secondary

screen. In this secondary screen, a given compound was tested against the individually dispensed

strain library, which would serve to identify the resistant one. Screening was performed by Dr.

Gabriele Casadei at the Harvard NSRB screening facility, which hosts a collection of 150,000

compounds. We first tested the general procedure to establish the Z´-factor, which measures the

overall quality of the screen (see materials and methods). The Z´-factor for our assay was 0.75,

E E

E

a) b)

30

suggesting that we could proceed to pilot screening. It is important to note that the deviation from

perfect results in this screen was due to normal variations in OD reading among wells, rather than

to false positives or false negatives. We did not have any case of substantial growth in a well with

ciprofloxacin or lack of growth in a well without an antibiotic (data not shown).

The pilot screen of 3,000 compounds was performed in duplicate to reduce variability. The

controls were E. coli W3100 cells, which were compared to a pool of 4,320 knockout strains from

the Keio library (BacPool). The pool was prepared by growing each mutant overnight in microtiter

plates in LBB at 37⁰C and then mixing all of them in equal amounts. The compounds were

dispensed at a final concentration of 46 μg/ml in 275 nl volume. For each tested molecule there

could have been three possible scenarios when scoring for growth/no growth:

1) Growth in both the BacPool and K12 wells, indicating lack of antimicrobial activity.

2) Growth inhibition in the both the BacPool and K12 wells: a possible antibiotic (direct

activity) or generally toxic compound is present, but is not a prodrug.

3) Growth of a BacPool well and no growth of the K12 well: a prodrug hit (Figure 2).

31

Figure 2-2. Criteria for identifying prodrug candidates in the initial model prodrug screen.

Compounds that did not inhibit growth of the pooled knock-out strains, but did inhibit growth of

the wild type E. coli were considered prodrug candidates.

Of the 3,000 molecules tested, we obtained 3 prodrug hits (Fig. 2-3), resulting in a hit rate of

0.1%. The screen was performed in duplicate, so these hits are unlikely to be false positives.

(a) (b) (c)

Figure 2-3. Prodrug hits from the pilot screen of 3,000 compounds. The pilot screen at ICCB

produced 3 hits, (a) is carbonylcyanide-p-trifluoromethoxyphenylhydrazone (FCCP), which

came from ICCB’s known bioactive plate. (b) compound ICCB-00566120, and (c) compound

ICCB-00566457, yielding a hit rate of 0.1%.

There were 59 prodrug hits identified in the screen against 152,796 compounds at ICCB, giving a

0.04% hit rate. 35 of the 59 prodrug hits validated as prodrugs in a secondary screen against

BacPool (Table 2-1).

Table 2-1. Prodrug Candidates from the initial prodrug screen at ICCB. The MICs were

determined by broth microdilution where the highest concentration tested was 12.5µg/mL.

32

Prodrug Candidates from ICCB Screen Prodrug

Candidate Structure MIC (µg/mL)

PD01

>12.5

PD02

>12.5

PD03

>12.5

PD04

>12.5

PD05

>12.5

PD06

12.5

PD07

>12.5

PD08

>12.5

PD09

>12.5

PD10

>12.5

PD11

>12.5

PD12

>12.5

33

PD13

>12.5

PD14

>12.5

PD15

>12.5

PD16

12.5

PD17

>12.5

PD18

>12.5

PD19

>12.5

PD20

>12.5

PD21

12.5

PD22 *unavailable >12.5

PD23

>12.5

PD24

>12.5

PD25

>12.5

PD26

>12.5

PD27

>12.5

34

PD28

>12.5

PD29

6.25

PD30

12.5

PD31

>12.5

PD32

0.39

PD33

12.5

PD34

0.78

PD35

>12.5

*The MIC of PD22 (Commercial ID: Maybridge BTB 136666CB) was tested using a small amount received as a cherry pick from ICCB. ICCB and Maybridge no longer carry PD22 or have the structure information available.

Of these 35 prodrug candidates, only 8 (PD06, PD16, PD21, PD29, PD30, PD32, PD33, and PD34)

had an MIC of 12.5µg/mL or lower. Following the logic of the prodrug model, 1,536 strains from

a library over expressing single genes in E. coli, the ASKA library (59), were screened at

concentrations 8X less than the MICs (with 0.1mM IPTG) of the respective 7 prodrug candidates

with good activity against the wild type E. coli (PD32 is no longer commercially available). None

35

of the 1,536 strains over-expressing enzymes showed an increased sensitivity to PD06, PD16,

PD21, PD29, PD30, PD33, or PD34. PD16 is a known antimicrobial, nitazole, a nitroaromatic

compound and a structural analog of the prodrug metronidazole (Table 1-1). Metronidazole is

converted by nitroreductases (60), and in order to determine if PD16 is also converted by

nitroreductases activity against strains over expressing and lacking the nitroreductases in E. coli

were tested (Table 2-2). The concentration of PD16 in the screen against the 1,536 over-expression

strains was too low to detect NfsA and NfsB as the converting enzymes.

Table 2-2. Activity of PD16 against strains lacking and overexpressing activating enzymes.

MICs were determined by broth microdilution.

Strain MIC (µg/mL)

Wild Type BW25113 12.5

BW25113pZS*24 12.5

BW25113pZS*24nfsA 3

BW25113pZS*24nfsB 3

BW25113ΔnfsAΔnfsB >100

BW25113ΔnfsAΔnfsBpZS*24nfsA 3

BW25113ΔnfsAΔnfsBpZS*24nfsB 3

When both nitroreductases are deleted from E. coli, there is no inhibition of growth even at

100µg/mL of PD16. Over-expressing either of the nitroreductases yields a four-fold decrease in

the MIC.

A complete screen of the ASKA and KEIO libraries with the remaining 6 prodrug candidates was

not feasible due to a lack of compound. The obvious disadvantage of the screen described above

36

is that resistance to compounds identified by this screen will develop easily due to null mutations

in the non-essential activating enzymes. For example, metronidazole resistance in Helicobacter

pylori (H. pylori) is due to a null mutation in the gene encoding the single non-essential activating

enzyme nitroreductase, RdxA (61). However, our aim is to identify compounds whose action

depends on essential enzymes. This led to another approach for identifying prodrugs.

Prodrug screening based on essential protein overexpression.

There are ~300 essential genes in E. coli (62). A set of 50 E. coli strains from the ASKA library

overexpressing conserved essential genes coding for potential prodrug-converting enzymes were

used for screen development. The work on identifying activating enzyme candidates was

performed in collaboration with Dr. Michael Galperin (NCBI), a specialist in bioinformatics. The

50 selected enzymes share homology to their counterparts in Mycobacterium tuberculosis (M.

tuberculosis), and do not have close homologs in humans (Table 2-3).

Table 2-3. List of genes with conserved essentiality in bacteria.

Gene

Essentiality (Yes-‘Y’, No-‘N’)

Gene Annotation E. coli M. tuberculosis

Present

in

Humans

ackA Y Y N Acetate kinase (EC 2.7.2.1)

argC Y Y N N-acetyl-gamma-glutamyl-phosphate

reductase (EC 1.2.1.38)

asd Y Y N Aspartate-semialdehyde dehydrogenase

(EC 1.2.1.11)

btuR Y Y N COB(I)alamin adenosyltransferase (EC

2.5.1.17)

37

coaD Y Y N Phosphopantetheine adenylyltransferase

(EC 2.7.7.3)

cysE Y Y N Serine acetyltransferase (EC 2.3.1.30)

dapA Y Y Y/N Dihydrodipicolinate synthase (EC 4.2.1.52)

dapB Y Y N Dihydrodipicolinate reductase (EC

1.3.1.26)

dapD Y Y N Tetrahydrodipicolinate N-

succinyltransferase (EC 2.3.1.117)

dapF Y Y N Diaminopimelate epimerase (EC 5.1.1.7)

ddlB Y Y N D-alanine--D-alanine ligase B (EC 6.3.2.4)

dxr Y Y N 1-deoxy-D-xylulose 5-phosphate

reductoisomerase (EC 1.1.1.267)

elaA Y Y N GTP-binding protein ElaA

fabA Y Y N Dehydratase/isomerase that plays a

specific and essential role in the synthesis

of unsaturated fatty acids. (EC 2.3.1.85)

frlD Y Y N Fructoselysine kinase

ftsI Y Y N Peptidoglycan synthetase

hemD Y Y N Uroporphyrinogen-III synthase (EC

4.2.1.75)

ispE Y Y N 4-diphosphocytidyl-2-C-methyl-D-

erythritol kinase (EC 2.7.1.148)

ispF Y Y N 2-C-methyl-D-erythritol 2,4-

cyclodiphosphate synthase (EC 4.6.1.12)

ispG Y Y N 1-hydroxy-2-methyl-2-(E)-butenyl 4-

diphosphate synthase (gcpE)

ispH Y Y N 4-hydroxy-3-methylbut-2-enyl diphosphate

reductase

lgt Y Y N Prolipoprotein diacylglyceryl transferase

(EC 2.4.99.-)

38

menE Y Y Y/N O-succinylbenzoic acid--CoA ligase (EC

6.2.1.26)

mesJ Y Y N tRNA(Ile)-lysidine synthetase

mrdA Y Y N Penicillin-binding protein 2,

transglycosylase/transpeptidase

mtn Y Y N MTA/SAH nucleosidase

murC Y Y N UDP-N-acetylmuramate--alanine ligase

(EC 6.3.2.8)

murD Y Y N UDP-N-acetylmuramoylalanine--D-

glutamate ligase (EC 6.3.2.9)

murE Y Y N UDP-N-acetylmuramoylalanyl-D-

glutamate--2,6-diaminopimelate ligase (EC

6.3.2.13)

murG Y Y N UDP-N-acetylglucosamine--N-

acetylmuramyl-(Pentapeptide)

pyrophosphoryl-undecaprenol N-

acetylglucosamine transferase (EC 2.4.1.-)

murI Y Y N Glutamate racemase (EC 5.1.1.3)

paaY Y Y N Phenylacetic acid degradation protein,

predicted acyltansferase

pspE Y Y N Rhodanese-related sulfurtransferase

pyrH Y Y N Uridylate kinase (EC 2.7.4.-)

ribB Y Y N 3,4-dihydroxy-2-butanone 4-phosphate

synthase

ribD Y Y N Riboflavin-specific deaminase/HTP

reductase (EC 3.5.4.26/EC 1.1.1.193)

tdcG Y Y N L-serine dehydratase (EC 4.2.1.13)

thiL Y Y N Thiamine-monophosphate kinase (EC

2.7.4.16)

39

yagS Y Y N Putative xanthine dehydrogenase yagS,

FAD binding subunit (EC 1.1.1.204)

yahF Y Y N Predicted acyl-CoA synthetase subunit

ybeY Y Y N Predicted metal-dependent hydrolase

ybhA Y Y N Predicted HAD family hydrolase

ycdX Y Y N Predicted PHP family hydrolase

yciL Y Y N Predicted RluB-like pseudouridylate

synthase

ydjQ Y Y N Predicted nuclease

yeaZ Y Y N Predicted protease

yfcH Y Y N Predicted nucleoside-diphosphate sugar

epimerase

yfgE Y Y N DnaA paralog, predicted DNA replication

initiation ATPase

yjeE Y Y N Predicted ATP/GTPase

From this list of 50 conserved essential enzymes, we chose two enzymes (bolded in Table 2-3)

with enzymatic reactions likely to convert prodrugs (personal communication, R. Lee). These

enzymes were DapB and FabA. DapB is Figure 2-2. The criteria for identifying prodrug candidates

in the initial model prodrug screen are depicted in (Figure 2-4), where compounds that did not

inhibit growth of the wild type E. coli, but did inhibit growth of the E. coli strains over-expressing

an essential enzyme, were considered prodrug candidates.

40

Figure 2-4. Criteria for identifying prodrug candidates in the secondary initial model

prodrug screen. Compounds that did not inhibit growth of the wild type E. coli, but did inhibit

growth of the E. coli strains over expressing an essential enzyme, were considered prodrug

candidates.

3,000 compounds from the Chembridge library were screened at 10µg/mL against strains from the

ASKA library (59) over-expressing (OE) dapB and fabA, induced with 0.1mM IPTG and the E.

coli wild type.

This screen resulted in a 0.6% hit rate with no compounds specific to DapB or FabA. 8,000

compounds from the NCI library were also screened at 10µg/mL against strains from the ASKA

library (59) over-expressing dapB and fabA, induced with 0.1mM IPTG and the E. coli wild type.

This screen had a 0.025% hit rate for compounds specific to DapB, 0.047% hit rate for compounds

with activity against both DapB OE and FabA OE but not the wild type, and a 0.18% hit rate for

compounds with activity against all three strains.

3,000 compounds from the Chembridge library were screened at 10µg/mL against strains from the

ASKA library (59) over-expressing (OE) dapB and fabA, induced with 0.1mM IPTG and the E.

coli wild type. This screen resulted in a 0.6% hit rate with no compounds specific to DapB or

FabA. 8,000 compounds from the NCI library were also screened at 10µg/mL against strains from

41

the ASKA library (59) overexpressing dapB and fabA, induced with 0.1mM IPTG and the E. coli

wild type. This screen had a 0.025% hit rate for the single compound specific to DapB, a 0.04%

hit rate for compounds with activity against both DapB OE and FabA OE but not the wild type (3

compounds), and a 0.18% hit rate for compounds with activity against all three strains (14

compounds). An auxotroph of DapB was obtained and tested against the one confirmed compound

(Fig. 2-5) specific to DapB, NCI 28002 (Table 2-3). NCI28002 is structurally similar to berberine

and ethidium bromide.

(a) (b) (c)

Figure 2-5. Hit compounds from initial prodrug pilot screen. NCI 28002 (a) is structurally

similar to the antimicrobial berberine (b) and ethidium bromide (c).

Table 2-4. Activity of NCI28002 against strains of E. coli over expressing and lacking DapB.

Strain

MIC (µg/mL)

Berberine Ethidium

Bromide NCI28002

WT BW25113 1,500 250 >100

dapA OE >500 50 50

dapB OE >500 50 12.5

fabA OE >500 50 25

WT MG1655 >500 >100 >100

ATM783 >500 >100 >100

42

The ASKA strains were more sensitive to both ethidium bromide and NCI28002. It likely that

NCI28002 is converted by DapB, but this was not investigated further due to high cytotoxicity

against mammalian cell lines IMR90 and FaDu (data not shown).

This very specific screen yielded no non-toxic prodrug candidates and alternative approaches

were taken.

The screen for compounds with non-specific or unknown targets.

The rationale for this screen is an inverted specificity test; compounds which lack specificity are

desired hits. In order to make this screen practical, a viability dye is used instead of label

incorporation. A subsequent cytotoxicity test against mammalian cells then differentiates between

prodrug candidates and generally toxic compounds.

Once activated, prodrugs are expected to hit unrelated targets, and we used this property to develop

a whole-cell screen. The logic is that prodrugs with a non-specific mode of action will inhibit

general metabolism, unlike traditional single-target antibiotics, and could be identified with a

viability dye such as alamar blue. The active ingredient in alamar blue is the compound resazurin.

Actively metabolizing cells reduce the blue-colored resazurin in the pink colored resorufin (Fig.

2-6). This change in color can be detected by eye and by reading fluorescence.

Figure 2-6. Reduction of resazurin to resorufin. Resazurin is oxidatively reduced by

metabolically active cells into resorufin.

43

Different classes of antimicrobials have different mechanisms of action and consequently have

diverse effects on the cells’ metabolism. These effects can be recorded over time and differentiate

between classes of drugs (Fig. 2-7). Known nitrofuran prodrugs nitrofurazone (NFZ) and

nitrofurantoin (NFT) rapidly inhibited alamar blue reduction by E. coli (Fig. 2-7). The

protonophore CCCP had a similar effect. By contrast, ciprofloxacin, a specific inhibitor of DNA

gyrase/topoisomerase, and kanamycin, a protein synthesis inhibitor, had no effect on the initial

rate of alamar blue reduction. Similar results were obtained with Bacillus anthracis (Fig. 2-8). This

differential action of non-specific versus specific compounds enabled the development of a

prodrug screen (Fig. 2-9). We determined that with a starting inoculum of ~2x106 CFU/mL, the

optimal time point to read fluorescence and differentiate between the rapid shutdown of

metabolism by compounds with non-specific targets and target-specific antibiotics was 4 hours.

The Z’ for the screen was >0.9, showing high fidelity of the approach.

44

Figure 2-7. Alamar blue reduction in E. coli as a basis for a produg screen. The prodrugs

Nitrofurazone (NFZ), and Nitrofurantoin (NFT), protonophore CCCP, ciprofloxacin (CIP) and

kanamycin (KAN) were added at 50 µg/mL to wells of a microtiter plate containing alamar blue

and E. coli cells. Fluorescence was detected using excitation at 544nm and emission at 590nm.

Data are the mean of three independent trials ± SD.

45

Figure 2-8. Alamar blue reduction in B. anthracis as a basis for a produg screen. The prodrug

Nitrofurantoin (NFT), the antiseptic chlorhexidine (CHX), Spectinomycin (SPCM), and

kanamycin (KAN) were added at 50 µg/mL to wells of a microtiter plate containing alamar blue

and B. anthracis cells. Fluorescence was detected using excitation at 544nm and emission at

590nm. Data are the mean of three independent trials ± SD.

46

Figure 2-9. Alamar blue reduction screen workflow.

A screen was then performed against E. coli and Bacillus anthracis (B. anthracis), measuring

alamar blue reduction at the 4 hour time point after delivering compounds at 35µg/mL. We started

with a pilot screen of 11,280 compounds from the Chembridge library. The hit rate for non-specific

compounds inhibiting alamar blue reduction was 0.05% for E. coli and 14% for B. anthracis (Table

2-4). This stark contrast in susceptibility reflects the known differences in the permeability barriers

between the Gram-negative E. coli and the Gram-positive B. anthracis. 14% is not a useful hit rate

because it is too high, and a direct screen against B. anthracis would have to be performed at a

considerably lower concentration of compounds. Screening a total of 54,480 compounds against

E. coli produced a cumulative hit rate of approximately 0.1%. On the basis of potential medicinal

chemistry properties, 20 compounds were selected for further analysis (data not shown) and the

three least cytotoxic compounds (Fig. 2-6) will be discussed in chapters 3 and 4.

47

Table 2-5. Results of the prodrug screen.

The Prodrug Screen- 54,480 Compounds

Chembridge Library: 11,280 compounds

E. coli - 4 PD hits, 0.05% HR

B. anthracis – 1,577 PD hits, 14% HR

Chemdiv, MPEX Library- 23,040 compounds

E. coli – 41 PD hits, 0.18% HR B. anthracis – 2,350 PD hits, 10% HR

Enamine Library- 20,160 compounds

E. coli – 5 PD hits, 0.025% HR B. anthracis – 1,766 PD hits, 8.8% HR

Overall E. coli Hit rate = 0.09% Overall B. anthracis Hit rate = 10.5%

Compounds were tested against E. coli and B. anthracis at 35µg/mL in wells of a 96 well microtiter plate with Mueller-Hinto broth containing alamar blue. Fluorescence changes were detected after incubation for 4 hours.

Figure 2-10. Prodrugs and hit compounds. ADC111 (a) and nitrofurantoin (b) are nitrofurans.

ADC112 (c) is an analog of tiliquinol (d) and tilbroquinol (e). ADC113 (f) is a β-diketone.

ADC111 is a nitrofuran resembling nitrofurantoin (Fig. 2-10), a known prodrug which is used to

treat urinary tract infection (UTI) caused by E. coli and other pathogens (37). ADC112 is an

analog of tiliquinol and tilbroquinol (Fig. 2-8), the two 8-hydroxyquinoline compounds of Intetrix,

a mix of tiliquinol and tilbroquinol. ADC112 differs from tilbroquinol only in the R7 group, having

48

a bromine instead of a methyl group. The mechanism of action for 8-hydroxyquinolines remains

unknown, although it has been proposed that these compounds chelate metals necessary for

multiple enzymatic catalysis reactions including DNA synthesis (63, 64). ADC113 is a β-diketone

and has been previously reported to be a putative dehydratase inhibitor in mycobacteria. It should

be noted that ADC113 is the same compound as PD30 (Fig. 2-8), which was identified in our

initial screen for prodrugs against pooled knock-out strains.

Validation of the alamar blue reduction screen for prodrugs.

Given that ADC111 is a nitrofuran, we examined the activity of this compound against strains

lacking the activating enzymes for this class of prodrugs, the nitroreductases NfsA and NfnB. In

the alamar blue test, ADC111 and nitrofurantoin rapidly inhibited reduction of the dye (Fig. 7A).

This inhibition was largely relieved in an ΔnfsAΔnfnB double mutant (Fig. 2-11). Similarly, the

MIC of ADC111 increased, 8 fold, in the ΔnfsAΔnfnB strain (Table 2-5). The nitrofurantoin MIC

increased 4 fold in the ΔnfsAΔnfnB strain. There are likely additional, not yet characterized

nitroreductases in E. coli (65) accounting for residual activity of the prodrugs in the ΔnfsAΔnfnB

mutant. Overexpression of the activating enzymes produced an opposite effect, an increase in

sensitivity (Table 2-5). This behavior is the opposite of what is expected from conventional

inhibitors of targets, where downregulation causes increased susceptibility and overexpression

leads to reduced activity but is consistent with prodrugs. This contrasting behavior of prodrugs

may serve as a good validation tool for this type of compounds. For ADC111, MIC dropped to 15

ng/ml in a strain overexpressing NfsA.

49

Table 2-6. Activity of ADC111 against strains lacking and overexpressing activating

enzymes.

Strain MIC (µg/mL) ADC111 Nitrofurantoin (NFT)

BW25113 2 12.5 BW25113ΔnfsAΔnfsB 16 50 BW25113ΔnfsA 4 12.5 BW25113ΔnfsB 2 12.5 nfsA++ 0.0156 0.025 nfsB++ 1 12.5

MIC was determined by broth microdilution.

Figure 2-11. Inhibition of alamar blue reduction by nitrofurans. (A), wild type E. coli; (B),

ΔnfsAΔnfsB double mutant with resistance to kanamycin. The experiment was performed in a

microtiter plate and fluorescence was measured every ten minutes at excitation 544nm, and

emission 590nm.

50

2.3 Discussion

The major reason for the lack of new compounds is well understood – most antibiotics were

produced by soil actinomycetes, and over-mining of this limited resource lead to the end of the

golden era of discovery (43). However, most classes of synthetic compounds were also discovered

during the golden era of the 50s and 60s, and this is harder to understand, given enormous advances

in chemistry and biology (11). A particularly puzzling case is that of antibiotics with a prodrug

mode of action. As mentioned in the Introduction, all of these compounds were discovered in the

1950s. Prodrugs have the features of a theoretically “ideal” antibiotic – a compound that is broad-

spectrum, non-toxic, and has the ability to kill both growing and dormant persister cells (66); (Fig.

1-5). Poor penetration across the complex envelope of Gram-negative bacteria is the main reason

synthetic approaches have not been successful in obtaining broad-spectrum compounds (53), with

the sole exception of fluoroquinolones. The irreversible binding of activated prodrugs to their

targets creates a sink, ensuring good accumulation over time. In this study, we tested these

assumptions and also developed a screen for prodrug compounds. Prodrugs such as metronidazole

do not have a specific target (67, 68), and historically, once validation steps such as specificity

tests were introduced, it excluded this type of antimicrobials from the discovery process. We

decided to revisit prodrugs, and develop a screen to identify prodrugs. Our initial platforms to

identify prodrugs based on identifying converting enzymes proved too specific with very low hit

rates (0.04% and 0.06%) and did not produce enough hit compounds. We determined that a screen

based on lack of specificity, essentially looking at compounds that are discarded in conventional

HTS campaigns would prove more successful. Using the viability dye alamar blue, we identified

hits acting against E. coli that inhibited reduction of this reporter of general metabolism, and then

tested the compounds for cytotoxicity. The screen successfully differentiates target-specific

51

compounds from known prodrugs such as nitrofurantoin. A pilot HTS with 55,000 compounds

produced hits, and 3 of these with low cytotoxicity were examined. Importantly, one of the

compounds, ADC111, is a nitrofuran analog of nitrofurantoin which is used to treat UTI (3, 37).

The activity of ADC111 depends on the presence of nitroreductases in E. coli. The finding of a

prodrug validates the screen. ADC112 is an analog of another antimicrobial that has been used in

the clinic, Intetrix. The mode of action of intetrix is unknown. We also identified a compound

ADC113 which has been previously reported to be a putative dehydratase inhibitor in

mycobacteria (69), however, in our hands we find it to act as a broad spectrum prodrug, suggesting

an alternative mechanism of action. Taken together, the results of the alamar blue reduction pilot

screen suggest that a larger HTS is likely to produce a number of novel prodrug leads.

52

2.4 Materials and Methods

Growth of bacterial strains

Strains of E. coli and B. anthracis were grown in cation adjusted Mueller Hinton II broth (CA-

MHB; BD cat. # 212322).

Construction of Bacterial strains

Mutations were introduced into the parental strain, E. coli K12 BW25113, by P1 transduction (70).

The kanamycin resistance cassette from the mutant alleles originated from the KEIO collection

(71) and was cured when needed by expressing the FLP recombinase from the helper plasmid

pCP20 (72). The MDR pump deletion mutant acrB and tolC are derived from the KEIO collection.

Strain BW25113 pZS*24nfsA was constructed by amplifying the nfsA ORF using primers

nfsAfwKpn1 (5’- gtagtagtaGGTACC CCGTCCACCGCAATATTCACGTT-3’) and

nfsArevCla1 (5’- gtagtagtaATCGAT GGTTGGGCGACGCGCTAA- 3’), and cloning into the

Kpn1/Cla1 digested sites of pZS*24 (73). Strain BW25113 pZS*24nfsB was constructed by

amplifying the nfsB ORF using primers nfsBfwCla1 (5’-

gtagtagtaATCGATGCTGGCACGCAAAATTACTTTCAC- 3’) and nfsBrevMlu1 (5’-

gtagtagtaACGCGTCCGGCAAGAGAGAATTACACTTCGG- 3’) and cloning into the

Cla1/Mlu1 digested sites of pZS*24 (73). All DNA amplification used for mutant construction and

screening was performed with Phusion high fidelity polymerase (New England Biolabs). Cloning

and PCR techniques were performed in accordance to standard protocols (74-76). All restriction

enzymes were purchased from New England Biolabs.

53

Table 2-7. List of strains used in this study.

Name Genotype Parent/ Source

Reference

BW25113 K12 rrnB3ΔlacZ4787hsdR514 Δ(araBAD)567Δ(rhaBAD)568rph-1

(71)

nfsB- FRT ΔnfsB::FRT JW0567 (71)

MV1970 ΔnfsB::FRTΔnfsA::kan JW0835 into nfsB-FRT

(71)

tolC ΔtolC::kan JW5503 (71) acrB ΔacrB::kan JW0451 (71) nfsA+ BW25113pZS*24nfsA (73) nfsA- ΔnfsA::kan JW0835 (71) nfsB+ BW25113pZS*24nfsB (73) nfsB- ΔnfsB::kan JW0567 (71) MG1655 E. coli K-12 prototroph (77) ATM783 MG1655ΔdapB::kan (78) dapB+ MG1655pCA24N-dapB JW0029 (59) dapA+ MG1655pCA24N-dapA JW2463 (59) fabA+ MG1655pCA24N-fabA JW0937 (59)

Initial Prodrug Screen validation

We first tested the general procedure to establish the Z´-factor. Since the output of our screening

is a typical growth/no growth assay, this was performed by comparing growth in control wells to

those containing a model antimicrobial, ciprofloxacin at 30 μg/mL. E. coli K12 were cultured in

LBB medium, and exponentially growing cells were dispensed at 106CFU/mL in 384-well

microtiter plates, 30 μL/well. Six plates were used in this experiment. Ciprofloxacin was added

to half of the wells (3 μL in LBB, bringing the final volume to 33μL). After an overnight

incubation at 37C, the OD600 of the plates were read, and the values of each well were used to

calculate Z´-factor:

푍 = 1 −3(휎 + 휎 )휇 + 휇

54

Where: 휎 is equal to the standard deviation of the positive control, 휎 is equal to the standard

deviation of the negative control, 휇 is equal to the average of the positive control, and 휇 is

equal to the average of the negative control. Following advice of S. Rudnicki, the ICCB

screening facility manager, we used the following table for Z´-factor interpretation:

High-throughput Screening Assay Fitness:

1 > Z´ > 0.9 An excellent assay

0.9 > Z´ > 0.7 A good assay

0.7 > Z´ > 0.5 Hit selection will benefit significantly from any improvement

0.5 = Z´ The absolute minimum recommend for high throughput screening

Alamar blue Reduction Screen

In order to test the fidelity of the screen, Z prime (Z’) scores were determined as described by (79),

by using NFT at 4X MIC for the positive control in six columns of a 96 well plate, and 1% DMSO

as the negative control in the other 6 columns of the 96 well plate. The “Prodrug Hit” Z’ was

determined from fluorescence readings after four hours of incubation and the direct activity Z’ was

determined by analyzing OD600 data after 24 hours of incubation. Each screening plate also

contained a negative and positive control column at each end of the 96 well plate.

Antimicrobials of interest were added to separate wells of a 96 well plate with black sides and a

clear bottom (Costar no. 3094). A 10% solution of alamar blue in MHB along with antibiotics and

test compounds were added to the screening plate. Broth cultures of E. coli were grown in MHB

to exponential phase, ~107 CFU/mL, diluted 1:10 in MHB, and added to the screening plate

containing MHB with 10% alamar blue and compound, resulting in a final antibiotic concentration

55

of 50 µg/mL. Control wells contained cells with alamar blue, but no antimicrobial compound or

alamar blue and antimicrobial compound without cells. A fluorometer (Spectra MAX GeminiXS)

was used to take readings at an excitation wavelength of 544 nm and emission wavelength of 590

nm. Kinetic readings were taken at 37˚C every 10 minutes for 240 minutes.

MIC Determination

For killing experiments and MIC determination, bacterial cells were grown in CA-MHB. MICs

were determined according to CLSI recommendations (80, 81).

56

Chapter 3:

Mechanism of Action and Medicinal Chemistry Optimization of Prodrug Candidates

57

3.1 Introduction:

Determining the mechanism of action (MOA) of new compounds poses an interesting challenge.

The MOA of a drug typically refers to the biochemical reaction between the drug and its target.

Since we hypothesize that our screen will identify prodrug antimicrobials, we need to determine if

our hit compounds are benign, and then converted into reactive species via a converting enzyme(s).

The prodrug model Fig. 2-1 depicts a useful validation test. The E. coli knock-out KEIO library

(71) is convenient and easy to test against compounds and can reveal the activating enzyme. The

obvious short-coming of using this library is that essential enzymes are excluded from the

collection and our ideal prodrug will be converted by an essential enzyme. We also utilize the E.

coli ASKA collection, where genes are over expressed on a plasmid (59). This can reveal the

activating enzyme even if it is essential.

Selecting for resistant mutants against the prodrug candidates is the preferred method for

determining the MOA. Resistant mutants can provide insightful information in regards to how a

drug acts. Bacterial mechanisms of resistance are well understood and were discussed in chapter

1. Prodrugs hit multiple targets, and spontaneous mutations occurring in parallel to all targets is

not expected. We therefore reason that prodrug resistant mutants will have mutations in their

converting enzymes. Resistance to nitrofurantoin (NFT) in E. coli results from mutations in the

converting enzymess nfsA and nfsB (82).

Hit compounds are rarely put directly into clinical trials; rather, they provide a starting point in the

drug discovery process. Compounds often need chemical modifications in order to be more

“drugable”, i.e. having good solubility, potency, and less toxicity. It is important to test analogs of

hit compounds to determine the essential core structure needed for activity, and to check for

58

increased and decreased potency. This testing process is used to determine the structure activity

relationship (SAR).

3.2 Results:

Determining the Mechanism of Action (MOA) of ADC111.

As previously discussed in chapters 2 and 3, ADC111 is a nitrofuratoin derivative, which lead to

the hypothesis that nitroreductases in E. coli will convert this prodrug candidate. In the alamar blue

test, ADC111 and nitrofurantoin rapidly inhibited reduction of the dye (Fig. 2-11A). This

inhibition was largely relieved in an ΔnfsAΔnfnB double mutant (Fig. 2-11B). Similarly, the MIC

of ADC111 increased, 8 fold, in the ΔnfsAΔnfnB strain (Table 3-1). Nitrofurantoin MIC increased

4 fold in the ΔnfsAΔnfnB strain. There are likely additional not yet characterized nitroreductases

in E. coli (65) accounting for residual activity of the prodrugs in the ΔnfsAΔnfnB mutant.

Overexpression of the activating enzymes produced an opposite effect, an increase in potency

(Table 3-1). This behavior is the opposite of what is expected from conventional inhibitors of

targets, where downregulation causes increased susceptibility and overexpression leads to reduced

activity. This contrasting behavior of prodrugs serves as a good validation tool for these types of

compounds. For ADC111, MIC dropped to 15 ng/ml in a strain overexpressing NfsA.

59


enzymes.

Strain MIC (µg/mL)

ADC111 Nitrofurantoin (NFT)

BW25113 2 12.5

BW25113ΔnfsAΔnfsB 16 50

BW25113ΔnfsA 4 12.5

BW25113ΔnfsB 2 12.5

nfsA++ 0.0156 0.025

nfsB++ 1 12.5


We used the virtual screening software PyRx (83)to construct the molecular surface image of the

enzyme NfsA (PDB 1F5V) (Figure 3-1a), dock ADC111 in the active site of NfsA (Figure 3-1b

and c), and predict the binding affinity of ADC111 to NfsA (Table 3-2).

Figure 3-1. Image of the enzyme NfsA’s molecular surface. The virtual Screening Software

PyRx was used to construct the molecular surface image of the enzyme NfsA, PDB 1F5V (a).

PyRx was also used to virtually dock FL1 into the active site of NfsA (b and c).

(a) (b) (c)

60

Table 3-2. Estimated binding affinity of ligands to NfsA. The software PyRx calculates the

estimated binding affinity of ligands to the target NfsA. ADC111 is predicted to have the strongest

binding affinity to NfsA compared to the prodrugs Nitazol and metronidazole which are also

converted into reactive molecules by NfsA, kanamycin is included a control not known to bind to

NfsA.

Ligand Target Binding Affinity (kcal/mol)

ADC111 NfsA -8.8

Nitazol NfsA -4.9

Metronidazole NfsA -4.9

Kanamycin NfsA 18

Selection of resistant mutants to ADC111

Selecting for resistant mutants and then sequencing the genome of the mutant strain and can

provide crucial information about a drug’s target, or in our case, the converting enzyme. We were

not able to obtain stable resistant mutants for ADC111. When ADC111 is removed from the

growth medium the cells become susceptible to ADC111. This suggests a plausible frequency of

mutation <10-9. We suspect this low frequency is due to multiple enzymes ability to reduce

ADC111. This speculation is also supported by the existence of an MIC when both nitroreductases

NfnB and NfsA are deleted (Table 3-1).

We hypothesized that DNA is one of the multiple targets that the activated, reactive species of

ADC111 hits. We tested ADC111 against four strains with mutations or deletions causing

compromised DNA repair (Table 3-3). The lexA3 mutant has a mutation in the lexA gene, leading

to no SOS induction. The lexA3 mutant is still proficient in homologous recombination, but not in

nucleotide excision repair (NER). RecA is needed for SOS induction and recombination; in the

absence of RecA no induction of the SOS response takes place. This effect is clearly demonstrated

61

by the ΔrecA mutant’s increased sensitivity to ciprofloxacin. recF is a recombination gene and

repairs single strand breaks in DNA. The uvrB gene plays a role in single strand break repair and

NER. The ΔrecA mutant and the ΔuvrB mutant have decreased MICs to ADC111, implying that

the reactive species of ADC111 do in fact damage DNA in bacterial cells.

Table 3-3. Activity of ADC111 against E. coli with compromised DNA repair.

MIC (µg/mL)

Strain (E. coli) Ciprofloxacin Chloramphenicol ADC111

Wild Type BW25113 0.01 6.25 1.0

Wild Type LVM 0.02 6.25 1.5

lexA3 mutant 0.01 3.125 0.39

ΔrecA 0.0025 3.125 ≤0.02

ΔrecF 0.02 6.25 0.39

ΔuvrB 0.01 3.125 0.19

MICs were determined by broth microdilution.

Determining the mechanism of action for ADC112.

ADC112 belongs to a known class of molecules, the 8-hydroxyquinolines. Drugs for urinary use

(nitroxoline) and intestinal antisepsis (Intetrix) are part of this class and act as antibacterial agents

as well as antiprotozoals. Their spectrum includes activity against E.coli, Salmonella, Shigella,

Proteus, Vibrio cholera, staphylococci, streptococci, Lamblia, Trichomonas, Entamoeba

histolytica, and Candida albicans (63). ADC112 is structurally very similar to tilbroquinol,

differing only in the R7 group where ADC112 has a bromine and tilbroquinol a methyl group. The

exact mechanism of action for this class of molecules is unknown, although it has been proposed

that these compounds chelate metals necessary for multiple enzymatic catalysis reactions including

62

DNA synthesis (63, 64). Intetrix is not approved by the FDA and not available in the United States.

The literature did not provide us with any clues for a mechanism of action of ADC112. Following

our previous logic for how prodrugs work, (without an activating enzyme a prodrug should have

less or no activity compared to the wild type) we chose to try to identify the converting enzyme(s)

by utilizing a collection of E. coli “long-chromosomal deletion” mutants supplied by the Japanese

National Biological Resource Project (84). These E. coli mutant strains contain long chromosomal

deletions of 10-100kb between essential genes. Initially, at concentrations 10X the MIC, there

were no wells with growth. At 4X the MIC there were three wells with growth: OCR48, 49-8,

OCR05, and OCR08-7, for a total of 42 genes knocked out (Table 3-4). The individual knock-out

strains from the KEIO library were tested for MICs against ADC112 dissolved in DMSO. ADC112

has increased solubility in PEG400 and was re-tested against the KEIO, but when the MICs for

these strains were re-tested, there was still no single deletion that conferred resistance to ADC112.

Table 3-4. Potential converting enzymes of ADC112.

Strain Annotation OCR48,49-8 bfP

major pilin structural unit bundlin ybfP b0689 lipoprotein ybfG b0691 ybfG, ECK0678, JW5094, ybfH, pseudogene wbfG L-fucosamine transferase ybfI b4636 pseudogene potE b0692 putrescine/proton symporter: putrescine/ornithine antiporter

putrescine transport protein ECP-0710 speF b0693 ornithine decarboxylase ybfK b4590 hypothetical protein b4590 kdpE b0694 DNA-binding response regulator in two-component regulatory

system with KdpD kdpD b0695 fused sensory histidine kinase in two-component regulatory system

with KdpE: signal sensing protein kdpC b0696 potassium translocating ATPase, subunit C kdpB b0697 potassium translocating ATPase, subunit B kdpA b0698 potassium translocating ATPase, subunit A kdpF b4513 potassium ion accessory transporter subunit vbfA Unknown function

63

ybfA b0699 predicted protein OCR05 udp b3813 uridine phosphorylase rmuC b3832 predicted recombination limiting protein ubiE b3833 bifunctional 2-octaprenyl-6-methoxy-1,4-benzoquinone methylase/S-

adenosylmethionine:2-DMK methyltransferase yigP b3834 conserved protein, SCP2 family ubiB b3835 2-octaprenylphenol hydroxylase tatA b3836 TatABCE protein translocation system subunit, sec-independent

protein translocase protein TatA tatB b3838 TatABCE protein translocation system subunit tatC b3839 TatABCE protein translocation system subunit, sec-independent

protein translocase protein TatC tatD b4483 quality control of Tat-exported FeS proteins; Mg-dependent

cytoplasmic DNase (EC:3.1.21.-) rfaH b3842 DNA-binding transcriptional antiterminator ubiD b3843 3-octaprenyl-4-hydroxybenzoate decarboxylase fre b3844 NAD(P)H-flavin reductase (EC:1.5.1.29) fadA b3845 3-ketoacyl-CoA thiolase (thiolase I) fadB b3846 fused 3-hydroxybutyryl-CoA epimerase/delta(3)-cis-delta(2)-trans-

enoyl-CoA isomerase/enoyl-CoA hydratase/3-hydroxyacyl-CoA dehydrogenase

pepQ b3847 proline dipeptidase yigZ b3848 conserved protein, UPF0029 family trkH b3849 potassium transporter hemG b3850 protoporphyrin oxidase, flavoprotein (EC:1.3.3.4) rrlA b3854 23S ribosomal RNA of RrnA operon rrfA b3855 5S ribosomal RNA of RrnA operon mobB b3856 molybdopterin-guanine dinucleotide biosynthesis protein B mobA b3857 molybdopterin-guanine dinucleotide synthase yihD b3858 conserved protein rdoA b3859 Thr/Ser kinase implicated in Cpx stress response dsbA b3860 periplasmic protein disulfide isomerase I (EC:5.3.4.1) yihF b3861 conserved protein, DUF945 family OCR08-7 glpK b3926 glycerol kinase (EC:2.7.1.30) glpF b3927 glycerol facilitator yiiU b3928 zapB, ECK3920, JW3899, yiiU,septal ring assembly factor,

stimulates cell division

64


Attempts to isolate resistant mutants via spontaneous mutagenesis on agar plates containing

ADC112 are unsuccessful to date. The most difficult challenge is that ADC112 falls out of solution

in the agar, leaving a lawn of non-resistant mutants. We hoped that Intetrix would have better

solubility and that if we could obtain a resistant mutant to Intetrix, it would help solve the

mechanism of action for the 8-hydroxyquinolines. Unfortunately, Intetrix also crashes out of

solution, and we have not yet been able to obtain resistant mutants.

Determining the mechanism of action for ADC113

ADC113 does not belong to a known class of approved antimicrobials, and we started by

performing a limited SAR with the aim of obtaining a more potent compound. Three analogs were

synthesized (Fig. 3-3, 1690, 1689, & 1650), and the rest were obtained from Enamine.

65

Figure 3-3. Structural analogs of ADC113. Lee 1690, Lee 1689, and Lee 1650were synthesized, and

EN300-11952, STK08845, B020549, EN300-14313, EN300-34629, and EN300-13710 were obtained

from Enamine.

None of the compounds were more potent against E. coli than ADC113 (Table 3-5). The two compounds

with the best MICs had electronegative halogens on the right hand aryl ring and a di-ketone functionality

(ADC113 and B020549). Replacement of the electronegative halogen with an electron donating group,

or reduction of the left-hand ketone generally had a negative impact on the MIC.

66

Table 3-5. MICs of ADC113 and structural analogs against wild type E. coli and efflux

mutant strains lacking TolC and EmrB. The structural analogs of ADC113 found in Figure 3-3

were tested for activity against wild type E. coli and strains of E. coli lacking TolC and EmrB.

Compound MIC (µg/mL)

WT ΔtolC ΔemrB ADC113 6.25/12 1.5 3.12 Lee 1690 25 1.25 1.25 EN300-13710 50 5 5 STK088045 50 1.25 2.5 B020549 50 12.5 25 EN300-14313 100 12.5 25 Lee 1689 >100 50 >100 Lee 1650 >100 50 50 EN300-34629 >100 >100 >100 EN300-11952 >100 >100 >100

The decrease in MIC in the efflux mutant strains indicated that ADC113 is subject to some level

of efflux. In some cases, like EJNC-013, the change in MIC is at least two fold, indicating that the

lack of fluorine molecules are crucial for activity.

We also utilized the same collection of E. coli “long-chromosomal deletion” mutants supplied by

the Japanese National Biological Resource Project (84) that we used in our efforts to determine

ADC112’s converting enzyme. We tested the collection at 4X the MIC of ADC113, 25µg/mL, and

saw growth in a single well with a strain containing 30 gene deletions. We tested individual knock-

out strains from the KEIO collection (71), and found two potential converting enzymes. Both

single knock-out strains had MICs at lease 8X higher than that of the wild type (Table 3-6). ArgC,

acetylglutamylphosphate, which is a reductase that carries out the third step in arginine

biosynthesis, and YbgJ, a putative carboxylase.

67

Table 3-6. Potential converting enzymes and their MICs against ADC113.

KEIO Strain MIC (µg/mL) KEIO Strain MIC (µg/mL)

ΔyecT 6.25 ΔotsA 25

ΔflhE 6.25 ΔaraH 12.5

ΔflhB 6.25 ΔaraG 6.25

ΔcheZ 6.25 ΔaraF 6.25

ΔcheY 6.25 ΔclpX 6.25

ΔcheB 6.25 Δlon 6.25

Δtap 6.25 ΔdegQ 6.25

Δtar 6.25 ΔargC 50

ΔmotA 6.25 ΔallA 6.25

ΔcheW 6.25 ΔypfI 6.25

ΔcheA 6.25 ΔycbC 12.5

ΔmotB 6.25 ΔybgJ 100

ΔflhC 6.25 ΔybdH 6.25

ΔflhD 6.25 ΔwcaK 3.13

We also tested strains over expressing ArgC and YbjJ from the ASKA library (59) (Table 13).

These strains had the same MIC as the wild type. Our prodrug model suggests that when a

converting enzyme is over expressed, the MIC should decrease. This is not the case for ArgC and

YbgJ and a possible explanation is that more enzymes than the wild type level do not increase the

overall rate of drug conversion. A strain lacking both ArgC and YbgJ was constructed to see if the

double deletion would have a cumulative effect on the MIC; surprisingly, there was no difference

in MIC compared to the wild type E. coli. We decided to test single gene knock outs from the

KEIO collection (71) for all of the genes surrounding ybgJ (Fig. 3-4).

68

Figure 3-4. Genes of interest for ADC113 surrounding ybgJ.

Table 3-7. Comparison of MICs between the wild type, knock-out strains, and over

expression strains for ArgC and YbgJ.

Strain MIC (µg/mL) of ADC113

BW25113 12.5

ΔemrR::kan 100

ΔemrB::kan 3.125

ΔtolC::kan 0.78-1.5

ΔybgJ ::FRT ΔargC 12.5

ΔybgJ::kan 100

ΔybgJ::FRT 50

ΔybgJ::Cam 25

ΔybgL::kan 25

ΔybgI::kan 25

ΔybgK::kan 12.5

Δnei::kan 25

ΔabrB::kan 25

ybgJ++ 12.5

ybgI++ 12.5

69

ybgK++ 12.5

ybgL++ 12.5

nei++ 12.5

abrB++ 12.5

The MIC data above implies that the insertion of the kanamycin cassette confers a higher MIC to

ADC113 in the ΔybgJ KEIO strain. This phenomenon was not investigated further.

Potential targets of ADC113 reactive species

We tested the effects of ADC113 on the cell membrane potential using the BacLight™ Bacterial

Membrane Potential Kit (B34950) from Invitrogen. The kit uses the carbocyanine dye DiOC2(3)

and the proton ionophore CCCP (carbonyl cyanide 3-chlorophenylhydrazone) as a control to

measure destruction of the membrane potential using a flow cytometer. ADC113 does appear to

affect the proton gradient, collapsing the membrane potential (Figure 3-5). This result alone cannot

draw insightful conclusions about ADC113’s MOA. If ADC113 is acting as our prodrug model

suggests, binding to multiple targets and corrupting proteins, it is no surprise that the membrane

potential is affected.

70

(a) (b)

(c) (d)

Figure 3-5. Effect of ADC113 on cell membrane potential. A flow cytometer was used to measure how ADC113 affects the membrane potential of E. coli. The x-axis represents Green/Yellow fluorescence, and the the y-axis represents red fluorescence. (a) Cells treated with 30mM DiOC2 (Has Red fluorescence)(3) (b) Cells treated with 30mM DiOC2(3) + 5mM CCCP (has Yellow fluorescence) as a positive control for cell membrane depolarization (c) Cells treated with 30mM DiOC2(3) + 50µg/mL ADC113 (d) Cells with 30mM DiOC2(3) + 15µg/mL Chloramphenicol as a negative control for cell membrane depolarization.


Another approach we took identify the converting enzyme was to select for resistant mutants. After

2.5 years, we were able to obtain resistant mutants in the wild type E.coli. The mutation rate for

the wild type BW25113 strain (background to the KEIO collection) was 10-7 when selecting at

100µg/mL of PD30. This resistant mutant was sent out for whole genome sequencing and the

mutations were mapped to identify potential target genes.

71

Whole genome sequencing of the ADC113 resistant mutants was performed at the Tufts University

Genomics Core Facility with Illumina HiSeq2000. We mapped ~10 million 50bp reads to the E.

coli MG1655 reference sequence (GenBank:U00096.2) using CLC Genomics Workbench (CLC

Bio). Whole genome sequencing of the E. coli BW25113 was performed at the Tufts University

Genomics Core Facility with Illumina HiSeq2000 and at the Biopolymers Facility at Harvard

Medical School with Illumina HiSeq2000. We identified a single point mutation in each of the

resistant mutant strains (Table 3-8) compared to the wild type strain BW25113. BW25113 is the

background strain for the KEIO collection.

Table 3-8. Whole Genome Sequencing results of ADC113 resistant mutants. ~10 million 50bp

reads were mapped to the E. coli MG1655 reference sequence (GenBank:U00096.2) using CLC

Genomics Workbench (CLC Bio). Whole genome sequencing of the E. coli BW25113 was

performed at the Tufts University Genomics Core Facility with Illumina HiSeq2000 and at the

Biopolymers Facility at Harvard Medical School with Illumina HiSeq2000. Presence of a mutation

is indicated by ‘Yes’, and ‘NO’indicated no difference in sequence compared to MG1655.

# Chromo-somal position in reference genome MG1655

Gene Base Change

Amino acid change

Inter-genic position

BW25113 Wild Type

LF0020 mutant

LF0030 mutant

1 66528 araD T→C Yes - Yes Yes Yes

2 70289 - G→T No araB-araC

Yes Yes Yes

3 502653 - C→T No ybaL-fsr Yes Yes Yes

4 547694 ylbE A→G No - Yes Yes Yes

72

5 704236 nagE T→C Yes - Yes Yes Yes

6 1335418 acnA A→G Yes - Yes Yes Yes

7 1650355 intQ T→C Yes - Yes yes Yes

8 2809048 mprA C→T Yes - No NO (YES*)

Yes

9 2842032 hycG G→T No - Yes Yes Yes

10 3927295 viaA C→T Yes - No Yes NO

11 3957957 - C→T No ppiC-yfiN

Yes Yes Yes

12 4091793 rhaD Yes

13 4159271 fabR G→T Yes - Yes Yes Yes

14 4162146 btuB C→G Yes - Yes Yes Yes

15 4472857 - A→G No yjgJ-tabA Yes Yes Yes

16 4600532 yjjP C→T Yes - Yes Yes Yes

17 4614692 yjjI G→T yes -

Yes

Yes Yes

The mutation in the viaA gene was a null mutation. viaA was amplified from the resistant mutant

and put into a clean MG1655 (77) background, viaA*. This strain had no change in MIC compared

to that of the wild type (Table 3-9). Upon closer and manual inspection of the whole genome

sequencing data, a mutation in the mprA gene was found in the other resistant mutant, LF0020,

73

explaining the resistance to ADC113. MprA, otherwise known as EmrR, is the repressor for the

emrAB efflux pump (85). When a deletion mutant of mprA was tested there was an increase in

MIC, indicating that the mutation in the resistant mutant resulted in a loss of function.

Table 3-9. Activity of ADC113 against strains lacking potential converting enzymes.

Strain MIC (µg/mL)

Wild Type E. coli BW25113 6.25

ΔmprA 100

ΔviaA 6.25

viaA* 6.25

This discovery was surprising because most mutations in efflux pumps appear at a high frequency

when selecting for resistant mutants. We have since been trying to obtain a resistant mutant in a

ΔtolC mutant strain to avoid mutations in efflux pumps.

3.3 Discussion:

ADC111 is a nitrofuran prodrug and is converted by nitroreductases NfsA and NfsB. In the absence

of NfsA and NfsB, ADC111 does not rapidly shut down metabolism in the cell, as shown in Fig.

2-11. The reactive species of ADC111 bind DNA, and when DNA repair is compromised there is

increased sensitivity to ADC111 (Table 3-3.)

The literature reports that 8-hydroxyquinolines chelate metal ions that may be necessary for

enzymatic catalysis reactions to occur (63, 64). ADC112 sterilizes non-growing cells and it would

be interesting to measure NADH and NAD+ levels over time in a treated culture to see if metal

chelation inhibits energy consumption. Testing MICs in defined medium where metal ion

74

concentrations vary would also be potentially informative; if ADC112 does chelate metal, then

increased concentrations of metal ions would result in an increased MIC.

We were able to determine that ADC113 disturbs membrane potential, and the active structure

requires fluorine atoms for activity against E. coli. Obtaining resistant mutants in a TolC mutant

strain to help avoid mutations related to efflux are currently ongoing. It would be useful to know

if ADC113 is hitting multiple targets. Label incorporation would provide valuable insight to this

question.

Our alamar blue reduction screen is the inverse of a target specific screen, where we are looking

for compounds that have multiple targets. Labeling compounds would serve as useful validation

to determine if our prodrug candidates bind multiple targets. Compounds can be labeled either

fluorescently or radioactively. The macromolecular synthesis (MMS) assay monitors inhibition of

key pathways including RNA synthesis (transcription), protein translation (translation), DNA

replication, cell wall (peptidoglycan) synthesis, and fatty acid (lipid) biosynthesis. This assay

utilizes radioactively labeled precursors that are incorporated into specific macromolecules (86).

Free precursors are soluble in trichloracetic acid (TCA) and the macromolecules produced by the

cells are not, so the radiolabeled macromolecules can be selectively precipitated and separated

from the un-incorporated free precursors by filtration, and then quantified. A target of a drug can

be identified by the lack of incorporation of the radioactively labeled precursor into the

macromolecule. Antibiotics known to disrupt specific pathways can be used as positive controls.

75


Growth of Bacterial Strains

Strains of E. coli were grown in cation adjusted Mueller Hinton II broth (CA-MHB; BD cat. #

212322). All strains were grown at 37⁰C with aeration at 220 rpm.

















76



Reference


(71)



(71)

tolC ΔtolC::kan JW5503 (71) acrB ΔacrB::kan JW0451 (71) nfsA+ BW25113pZS*24nfsA (73) nfsA- ΔnfsA::kan JW0835 (71) nfsB+ BW25113pZS*24nfsB (73) nfsB- ΔnfsB::kan JW0567 (71) MG1655 E. coli K-12 prototroph (77) ATM783 MG1655ΔdapB::kan (78) viaA* LVMviaA A374V This study viaA- ΔviaA::kan JW5610 (71) mprA- ΔmprA::kan JW2659 (71) lexA3 lexA3 (87) dapB+ MG1655pCA24N-dapB JW0029 (59) dapA+ MG1655pCA24N-dapA JW2463 (59) fabA+ MG1655pCA24N-fabA JW0937 (59)

MIC Determination


were determined according to CLSI recommendations (80, 81).

77

Chapter 4:

A Screen and Validation of Prodrug Antimicrobials

Accepted to the Journal of Antimicrobial Agents and Chemotherapy on 12/09/2013

Laura E Fleck1, E Jeffrey North2, Richard E Lee2, Lawrence R Mulcahy1, Gabriele Casadei3, and Kim Lewis1

1Antimicrobial Discovery Center, Department of Biology, Northeastern University, Boston, Massachusetts, USA 2 Department of Chemical Biology & Therapeutics, St. Jude Children’s Research Hospital, Memphis, Tennessee, USA 3Istituto Zooprofilattico Sperimentale della Lombardia e dell’ Emilia Romagna, Via dei Mercati, 13/4 - 43126 Parma, Italy

78

4.1 Abstract

The rise of resistant pathogens and chronic infections tolerant to antibiotics presents an unmet need

for novel antimicrobial compounds. Identifying broad-spectrum leads is challenging due to the

effective penetration barrier of Gram-negative bacteria, formed by an outer membrane restricting

amphipathic compounds, and MDR pumps. In chronic infections, pathogens are shielded from the

immune system by biofilms or host cells, and dormant persisters tolerant to antibiotics are

responsible for recalcitrance to chemotherapy with conventional antibiotics. We reasoned that the

dual need for broad-spectrum and sterilizing compounds could be met by developing prodrugs that

are activated by bacteria-specific enzymes, and the generally reactive compounds could kill

persisters and accumulate over time due to irreversible binding to targets. We report the

development of a screen for prodrugs, based on identifying compounds that non-specifically

inhibit reduction of a viability dye alamar blue, and then eliminate generally-toxic compound by

testing for cytotoxicity. A large pilot of 55,000 compounds against E. coli produced 20 hits, 3 of

which were further examined. One compound, ADC111, is an analog of a known nitrofuran

prodrug nitrofurantoin, and its activity depends on the presence of activating nitroreductase

enzymes. ADC112 is an analog of another known antimicrobial tilbroquinol with unknown

mechanism of action, and ADC113 does not belong to a known class of FDA approved antibiotic.

All three compounds had a good spectrum and showed good to excellent activity against persister

cells in biofilm and stationary cultures. These results suggest that screening for overlooked

prodrugs may present a viable platform for antimicrobial discovery.

79

4.2 Introduction

The need for novel antibiotics to combat drug-resistant pathogens is well understood (55). Less

recognized, but no less important, is the unmet need for compounds capable of effectively killing

dormant forms of pathogens (15). Biofilm infections are on the rise, largely a result of medical

intervention, and form chronic, poorly treatable infections. Biofilms form readily on indwelling

devices – catheters and prostheses. Biofilms are also responsible for infective endocarditis,

recurring urinary tract infections, infective osteomyelitis, and the incurable infection of lungs of

patients with cystic fibrosis (19). Antibiotics depend on the immune response to clear an infection,

and a chronic disease often forms in immune-compromised patients. Importantly, most chronic

infections recalcitrant to treatment are caused by drug-susceptible pathogens. Recalcitrance to

treatment results from tolerance rather than resistance. Pathogens produce a small subpopulation

of dormant persister cells that are tolerant to antibiotics (15), and the biofilm matrix protects them

from the immune system. Once the concentration of the antibiotic drops, persisters resuscitate and

repopulate the biofilm, causing a relapsing infection. Several mechanisms lead to dormancy in E.

coli, and rely mostly on the action of toxin/antitoxin modules. The toxins responsible for persister

formation include mRNA endonucleases (21, 22); the HipA kinase that inhibits protein synthesis

by phosphorylating elongation factor Ef-Tu (23); and TisB, which decreases the energy level of

the cell by opening an ion channel (24). Bactericidal antibiotics kill by corrupting their targets (21,

25); for example, fluoroquinolones inhibit the re-ligation step in DNA gyrase and topoisomerase,

turning the enzymes into endonucleases (26). Targets are inactive in dormant persisters, explaining

their tolerance to antibiotics. The high degree of redundancy in the mechanisms of persister

formation precludes development of conventional target-based inhibitors.

80

We considered prodrugs as a type of compounds that could kill persister cells and eradicate a

chronic infection. Nitroaromatic prodrugs such as metronidazole or nitrofurantoin are benign

compounds that enter into the cell and are converted into a reactive drug by nitroreductases specific

to microorganisms (Fig. 4-1). Since the highly activated species produced hit multiple targets, this

could in principle kill both growing and dormant cells. Given that redox activated prodrugs bind

covalently to their targets, this creates an irreversible sink, ensuring accumulation over time. The

sink is likely to counter efflux by MDR pumps. The dual barrier of the outer membrane and MDRs

prevents most compounds from entering cells of Gram-negative bacteria (44, 45), and is largely

responsible for the paucity of broad-spectrum antibiotics. The last class of broad-spectrum

compounds, the fluoroquinolones, was discovered over 50 years ago (11).

It is interesting to note that all prodrug antibiotics were discovered in the 50s (11). It seems that

subsequently developed validation tests based on determining specificity of action of hits

precluded prodrug discovery. In an effort to eliminate generally toxic compounds, the specificity

test was introduced, where the ability of a test compound to inhibit label incorporation into major

biopolymers is measured (88). Compounds that inhibit all biosynthesis are non-specific and are

eliminated. By this test, metronidazole is a nuisance compound. Interestingly, metronidazole is a

broad-spectrum antibiotic, but its use is limited, since nitroreductases are expressed primarily

under anaerobic/micro-aerophilic conditions. Given the potential of prodrugs for both broad-

spectrum and sterilizing activity, we considered developing a screen for these compounds. The

rationale for the screen is an inverted specificity test: compounds which lack specificity are desired

hits. In order to make the screen practical, a vital die is used instead of label incorporation. A

subsequent cytotoxicity test against mammalian cells then differentiates between prodrug

81

candidates and generally toxic compounds. In this study, we report development of a prodrug

screen and validation of hit compounds, including their ability to kill persister cells.

Figure 4-1. Prodrug antibiotics. An ideal prodrug is an inactive compound that penetrates into

the cell and is converted by a bacteria-specific enzyme into a reactive molecule. The reactive form

binds covalently to unrelated targets, killing both regular and dormant cells. Importantly, covalent

binding creates an irreversible sink, which leads to accumulation of the drug over time.

4.3 Results

A screen for prodrugs. Once activated, prodrugs are expected to hit multiple targets, and we used

this property to develop a whole-cell screen. We reasoned that prodrugs with a non-specific mode

of action will inhibit general metabolism, and could be identified with a vital die such as alamar

blue (resazurin). Actively metabolizing cells reduce blue resazurin into red-colored resofurin,

which is accompanied by a strong change in fluorescence as well. A cytotoxicity test would then

discriminate prodrug candidates from generally-toxic compounds.

Known nitrofuran prodrugs nitrofurazone (NFZ) and nitrofurantoin (NFT) rapidly inhibit alamar

blue reduction by E. coli (Fig. 3-2). The protonophore carbonyl cyanide m-chlorophenyl hydrazine

82

(CCCP) had a similar effect. In contrast, ciprofloxacin, a specific inhibitor of DNA

gyrase/topoisomerase, and kanamycin, a protein synthesis inhibitor, had no effect on the initial

rate of resazurin reduction. Similar results were obtained with Bacillus anthracis (not shown). This

differential action of non-specific vs. specific compounds enabled development of a prodrug

screen. We determined that with a starting inoculum of ~2x106 CFU/mL, the optimal time point

to read fluorescence and differentiate between the rapid shutdown of metabolism by compounds

with non-specific targets and target-specific antibiotics was 4 hours. The Z’ for the screen was

>0.9 (data not shown), showing high fidelity of the approach (see Materials and Methods).

A screen was then performed against E. coli, measuring alamar blue reduction at the 4 hour time

point after delivering compounds at 35 µg/mL. We started with a pilot screen of 11,000 structurally

diverse compounds from ChemBridge. The hit rate for non-specific compounds inhibiting alamar

blue reduction was 0.05% for E. coli (Table 4-1) and 14% for B. anthracis. This stark contrast in

susceptibility reflects the known differences in the permeability barriers between the Gram-

negative E. coli and the Gram-positive B. anthracis. 14% is not a useful hit rate, and a direct screen

against B. anthracis would have to be performed at a considerably lower concentration of

compounds. Screening a total of 55,000 compounds against E. coli produced a cumulative hit rate

of approximately 0.1%. Based on potential medicinal chemistry properties including low

molecular weight (<500), low lipophilicity (cLogP <5) and after removal of reactive electrophilic

and known promiscuous species, 20 synthetically tractable compounds were selected for further

analysis. Of these 20 compounds, three compounds are described in this paper (Fig. 4-3).

83

Figure 4-2. Alamar blue reduction as a basis for a produg screen. The prodrugs Nitrofurazone

(NFZ), and Nitrofurantoin (NFT), protonophore CCCP, ciprofloxacin (CIP) and kanamycin

(KAN) were added at 50 µg/mL to wells of a microtiter plate containing alamar blue and E. coli

cells. Fluorescence was detected using excitation at 544nm and emission at 590nm. Data are the

mean of three independent trials ± SD.

Table 4-1. Results of the prodrug screen.

Chemical Library No. of Prodrug (PD) hits and Hit Rate

Chembridge Library: 11,280 compounds

E. coli - 4 PD hits, 0.05% Hit rate

Chemdiv, MPEX Library- 23,040 compounds

E. coli – 41 PD hits, 0.18% Hit rate

Enamine Library- 20,160 compounds E. coli – 5 PD hits, 0.025% Hit rate

Overall E. coli Hit rate = 0.09% Total compounds screened: 54,480

Total hits: 50

84

Compounds were tested against E. coli at 35 µg/mL in wells of a microtiter plate with Mueller-

Hinton broth containing alamar blue. Fluorescence changes were detected after incubation for 4

hours.

Figure 4-3. Prodrugs and hit compounds. ADC111 (a) and nitrofurantoin (b) are nitrofurans.

ADC112 (c) is an analog of tiliquinol (d) and tilbroquinol (e). ADC113 (f) is a β-diketone.

ADC111 is a nitrofuran resembling nitrofurantoin, a known prodrug which is used to treat urinary

tract infection (UTI) caused by E. coli and other pathogens (37). ADC112 is an analog of tiliquinol

and tilbroquinol, the two 8-hydroxyquinoline compounds of Intetrix. ADC112 differs from

tilbroquinol only in the R7 group, having a bromine instead of a methyl group. The mechanism of

action for 8-hydroxyquinolines remains unknown, although it has been proposed that these

compounds chelate metals necessary for multiple enzymatic catalysis reactions including DNA

synthesis (63, 64). ADC113 is a β-diketone and has been previously reported to be a putative

dehydratase inhibitor in mycobacteria, however, in our hands we find it to be acting as a prodrug

85

in a panel of Gram-positive and Gram-negative bacteria. This suggests an alternative mechanism

of action.

Hit validation. The aim of the prodrug screen was to find bactericidal compounds that are not

generally toxic without a specific target in bacteria. We therefore examined the cytotoxicity of the

hits against mammalian cells. Compounds with a therapeutic index (TI) of ≥10 were considered

desirable hits. ADC111 was less toxic to mammalian cells, and considerably more active against

E. coli as compared to its approved analog nitrofurantoin, which translated into an excellent

therapeutic index (Table 2). For example, the TI of ADC111 with FaDu cells was 320, as compared

to nitrofurantoin TI of only 10. ADC112 was also considerably less toxic then compared to its

analog tilbroquinol, with TI 21/42 vs. 4, correspondingly. ADC113 had a TI ranging from 5 to 16

depending on the cell type, a reasonable number for a hit compound. The cytotoxicity data

suggested that further evaluation of these compounds was warranted.

Table 4-2. Cytotoxicity of the hit compounds.

Compound

MIC (µg/mL) with 10% FBS

Cytotoxicity

TI FaDu Caco2 HepG2

ADC111 0.78 250 31.25 31.25 320/40

Nitrofurantoin 12.5 125 125 125 10

ADC112 1.5 31.25 62.5 62.5 21/42

Tilbroquinol 25 100 100 100 4

ADC113 6.25-12.5 100 100 62.5 8-16 &

5-10

86

MIC values listed are for wild type E. coli. Compounds were considered cytotoxic at

concentrations where there was less than 50% survival compared to that of the untreated control.

Next, we examined the spectrum of activity of the hits. All 3 compounds showed a reasonably

broad spectrum with good activity against Gram-positive and Gram-negative species (Table 3-3).

ADC111 was considerably more active than its analog nitrofurantoin. For example, the E. coli

MIC of ADC111 is 0.78 µg/mL, which favorably compares to the nitrofurantoin MIC of 12.5

µg/mL. The compounds showed excellent activity against important drug-resistant pathogens,

including methicillin-resistant S. aureus and vancomycin-resistant E. faecalis. The compounds

also showed good activity against pathogens important for biodefense: Bacillus anthracis, Yersinia

pestis and Francisella tularensis.

87

Table 4-3. Spectrum of activity for ADC compounds.

MIC (µg/mL) Species ADC111 ADC112 ADC113 CIP STR DOX RIF VAN TIG Francisella tularensis (SchuS4)*

nd 0.097 0.195-0.34

nd 8 nd nd nd nd

Bacillus anthracis (Ames)*

0.35 1.5 0.78 0.015 nd nd nd nd nd

Yersinia pestis (KIM)*

≤0.78 1.5 0.78-1.56 nd nd 9 nd nd nd

Escherichia coli (ATCC 25922)

0.78 1.5 6.25-12.5 0.015 nd nd nd nd nd

Staphylococcus aureus (NCTC 8325)

1.56 ≤1.5 3.12 0.15 nd nd ≤0.06 1 nd

Enterococcus faecalis (OG1RF, ATCC 47077)

3.12 ≤1.5 6.25 0.031 nd nd nd nd nd

Salmonella Typhimurium (LT2)

5 nd nd nd 2 nd nd nd nd

Staphylococcus aureus (MRSA, NRS54)

6.25 ≤1.5 6.25 nd nd nd ≤0.06 4 nd

Clostrium difficile (CD196)

6.25 ≤1.5 6.25 nd nd nd nd 1.5 nd

Enterococcus faecium (VRE BM4147)

12.5 6.25 6.25 nd nd nd nd nd 0.125

Acinetobacter baumannii (AB17978)

12.5 6.25 12.5 0.06 nd nd nd nd nd

Pseudomonas aeruginosa (PAO1)

25 >50 >50 0.031 nd nd nd nd nd

Compounds where an MIC was not determined are labeled as, “nd”. BSL3 agents are denoted by an “*”, and were tested at the NERCE/BEID facility in Boston, MA. MICs were determined accorinding to CLSI guidelines and by broth microdilution.

One of the expected features of a prodrug mode of action is good permeability. Since activated

prodrugs bind covalently to their targets, this will create an accumulation sink, countering

multidrug efflux pumps. The activity of hit compounds was therefore tested in strains lacking

88

either the transenvelope pump AcrB (89), or lacking TolC, a common porin with a gated channel

that services several MDRs (45, 90).

In a control experiment, the activity of erythromycin against E. coli increases 128 fold in a tolC

mutant (Table 3-4), reflecting the prominent role of efflux in protecting the cells from antibiotics.

By contrast, there is only a 2 to 4 fold increase in potency of the known prodrugs nitrofurazone

and nitrofurantoin in a tolC strain, and a similar 4 fold increase in the potency of the nitrofuran

ADC111. Other candidate prodrugs tested, ADC112 and ADC113, also showed moderate

increases in potency against the MDR mutant strains.

Table 4-4. Activity of test compounds against MDR mutant strains.

Strain MIC (µg/mL)

ADC111 ADC112 ADC113 NFT NFZ ERY

BW25113 1.25 2.5 12.5 12.5 6.25 200

ΔtolC 0.4 0.67 1.25 3.125 3.12 1.56

ΔacrB 1.25 2.5 3.12 6.25 3.12 1.56


Given that ADC111 is a nitrofuran, we examined the activity of this compound against strains

lacking the activating enzymes for this class of prodrugs, the nitroreductases NfsA and NfnB. In

the alamar blue test, ADC111 and nitrofurantoin rapidly inhibited reduction of the dye (Fig. 3-

4A). This inhibition was largely relieved in an ΔnfsAΔnfnB double mutant (Fig. 4-4B). Similarly,

the MIC of ADC111 increased, 8 fold, in the ΔnfsAΔnfnB strain (Table 4-5). The nitrofurantoin

MIC increased 4 fold in the ΔnfsAΔnfnB strain. There are likely additional, not-yet characterized

nitroreductases in E. coli (65) accounting for residual activity of the prodrugs in the ΔnfsAΔnfnB

89

mutant. Overexpression of the activating enzymes produced an opposite effect, an increase in

potency (Table 4-5). This behavior is the opposite of what is expected from conventional inhibitors

of targets, where downregulation causes increased susceptibility and overexpression leads to

reduced activity. This contrasting behavior of prodrugs may serve as a good validation tool for this

type of compounds. For ADC111, MIC dropped to 15 ng/ml in a strain overexpressing NfsA,

showing the potential of a prodrug as a potent therapeutic. We were unable to obtain mutants

resistant to ADC112 and ADC113 so far. This may indicate the presence of more than one

activating enzymes/targets.


enzymes.

Strain MIC (µg/mL)

ADC111 Nitrofurantoin (NFT)

BW25113 2 12.5

BW25113ΔnfsAΔnfsB 16 50

BW25113ΔnfsA 4 12.5

BW25113ΔnfsB 2 12.5

nfsA++ 0.0156 0.025

nfsB++ 1 12.5


90

Figure 4-4. Inhibition of alamar blue reduction by nitrofurans. (A), wild type E. coli; (B),

ΔnfsAΔnfsB double mutant (resistant to kanamycin). The experiment was performed in a microtiter

plate and fluorescence was measured every ten minutes at excitation 544nm, and emission 590nm.

Bactericidal activity of hit compounds. An ability to hit unrelated targets suggests that prodrugs

may be able to effectively kill both growing and dormant cells. Both ADC111 and nitrofurantoin

were highly bactericidal against an exponentially growing culture of E. coli. The killing was

biphasic, with persisters surviving considerably better than regular cells (Fig. 4-5).

91

Figure 4-5. Time and concentration dependent killing of E. coli BW25113 in exponential

phase. Cells were challenged with varying concentrations of ADC111 (MIC = 2 µg/mL), NFT

(MIC = 12.5 µg/mL), and ciprofloxacin (MIC = 0.01 µg/mL). Cell count was determined by

plating on nutrient agar. The limit of detection is the x-axis. Data are the mean of three independent

trials ± SD.

Next, we tested the ability of ADC111 to kill cells growing in a biofilm. E. coli produces biofilms

on indwelling catheters and (19) forms intracellular biofilms in bladder epithelial cells (91).

ADC111 was considerably more effective than nitrofurantoin used to treat UTI in killing a biofilm

within a 24 hour period (Fig. 4-6).

92

Figure 4-6. Biofilm killing by ADC111, NFT, and ciprofloxacin. 24 hour old biofilms of E. coli

were grown on pegs, then moved into fresh medium containing antimicrobials for 24 hours, cells

were dislodged and plated for colony count. Data are the mean of three independent trials ± SD.

Unlike slow growing biofilms, a stationary culture produces more persisters and is harder to

eradicate (92, 93). This is clear from observing the effect of ciprofloxacin on a stationary culture.

Even at high concentrations, the killing is less than 3 log (Fig. 4-7). Ciprofloxacin reaches a very

high concentration in the bladder, up to 400 µg/mL (94). This however does not help eradicate the

pathogen due to the presence of persister cells, and because of the paradoxical relationship between

killing efficiency and concentration, which inverts at higher levels of this drug (95). Nitrofurantoin

was considerably more effective than ciprofloxacin in killing a stationary culture, and ADC111

93

had an even greater effect than nitrofurantoin, decreasing the cell count by 7 orders of magnitude

(Fig. 4-7).

Figure 4-7. Concentration dependent killing of wild type E. coli in Stationary phase. Cells

were challenged for 24 hours. The limit of detection is the x-axis. Data are the mean of three

independent experiments ± SD.

Next, we tested the ability of ADC112 to kill biofilm and stationary cultures of E. coli.

Tilbroquinol, an analog of ADC112, killed comparably well to ciprofloxacin, but showed a distinct

paradoxical effect at higher concentrations (Fig. 4-8). ADC112 was similar in effectiveness to

intetrix, but lacked the paradoxical effect, showing a gradual increase in killing at higher

concentrations.

94

Figure 4-8. Biofilm killing by ADC112 and Tilbroquinol. 24 hour old biofilms were treated

with compounds for 24 hours. The number of surviving cells per peg was determined by colony

count. The limit of detection is Log1.6 (CFU/mL). Data are representative of three independent

trials ± SD.

ADC112 was effective in killing stationary cells of E. coli, achieving complete sterilization (Fig.

4-9).

95

Figure 4-9. Killing of stationary phase E. coli with ADC112. Cells were treated for 24 hours

and then plated for colony count. The limit of detection is the x-axis. Data are representative of

three independent trials ± SD.

ADC113 does not belong to a known class of approved antimicrobials, and we started by

performing a limited SAR with the aim of obtaining a more potent compound. Three analogs were

synthesized (Fig. 10, Lee 1690, Lee 1689, & Lee1650), and the rest were obtained from Enamine.

96

Figure 4-10. Structural analogs of ADC113. Lee 1690, Lee 1689, and Lee 1650 were

synthesized, and EN300-11952, STK08845, B020549, EN300-14313, EN300-34629, and EN300-

13710 were obtained from Enamine.

None of the compounds were more potent against E. coli than ADC113 (Table 3-6). The two

compounds with the best MICs had electronegative halogens on the aryl ring and a di-ketone

functionality (ADC113 and B020549). Replacement of the halogen with an electron donating

group, or removal of fluorogroups from the trifluoromethyl ketone had a negative impact on the

MIC.

97

Table 4-6. Activity of ADC113 and structural analogs against E. coli.

Compound

MIC

(µg/mL)

E. coli

ADC113 6.25/12

Lee 1690 25

EN300-13710 50

STK088045 50

B020549 50

EN300-14313 100

Lee 1689 >100

Lee 1650 >100

EN300-34629 >100

EN300-11952 >100

ADC113 had excellent activity against exponentially growing E. coli cells (Fig. 4-11).

98

Figure 4-11. Time and concentration dependent killing of exponentially growing wild type

E. coli. Cells were incubated with ADC113, and the cell count was determined by plating. The

limit of detection is the x-axis. Data are representative of three independent trials ± SD.

ADC113 had good killing against biofilms (Fig. 4-12), and was comparable to ciprofloxacin in

killing stationary cells (Fig. 4-13).

99

Figure 4-12. E. coli biofilm killing with ADC113. 24 hour old biofilms were treated with 10X

MIC of ADC 113, and 10X and 20X MIC of CIP for 24 hours. The number of surviving cells per

peg was determined by colony count. The limit of detection is 101.6 CFU/peg.

1

100

Figure 4-13. Killing of E. coli in stationary phase. Cells were challenged for 24 hours with

32X MIC of ADC113 and CIP.

4.4 Discussion

The antibiotic crisis we are facing is due to the lack of good starting compounds (46), rapid rise of

resistance that makes existing antibiotics ineffective, and otherwise successful medical

interventions which prolong life of patients with deficient immune responses, leading to the rise

of recalcitrant chronic infections. The major reason for the lack of new compounds is well

understood – most antibiotics are produced by soil actinomycetes, and over-mining of this limited

resource lead to the end of the golden era of discovery (43). However, most classes of synthetic

compounds were also discovered during the golden era of the 50s and 60s, and this is harder to

understand, given enormous advances in chemistry and biology (11). A particularly puzzling case

101

is that of antibiotics with a prodrug mode of action. As mentioned above, all of these compounds

were discovered in the 50s. Prodrugs have the features of a theoretically “ideal” antibiotic – a

compound that is broad-spectrum, non-toxic, and able to kill both growing cells and dormant

persister cells (66); (Fig. 4-1).

Poor penetration across the complex envelope of Gram-negative bacteria is the main reason





do not have a specific target (67, 68), and, historically, once validation steps such as specificity

testing were introduced, these types of antimicrobials were extruded from the discovery process.

We decided to revisit prodrugs, and developed a screen based on the lack of specificity of

mechanism of action, essentially looking at compounds that are discarded in conventional HTS

campaigns. Using the viability dye alamar blue, we identified hits acting against E. coli that

inhibited reduction of this reporter of general metabolism, and then tested the compounds for

cytotoxicity. Testing for cytotoxicity in mammalian cells eliminates general target-specific energy

poisons, but not bacteria-specific energy poisons which are desirable compounds. The screen

successfully differentiates target-specific compounds from known prodrugs such as nitrofurantoin.

A pilot HTS with 55,000 compounds produced 50 hits, and 3 of these with low cytotoxicity were

examined. Importantly, one of the compounds, ADC111, is a nitrofurazone analog of

nitrofurantoin which is used to treat UTI (3, 37). The activity of ADC111 depended on the presence

of nitroreductases in E. coli. The finding of a prodrug validates the screen. ADC112 is an analog

of another antimicrobial that has been used in the clinic, intetrix. The mode of action of intetrix is

102

unknown. We also identified a compound ADC113 that does not belong to known classes of

approved antimicrobials, but has been previously reported to be a putative dehydratase inhibitor

in mycobacteria (69). The rapid reduction of alamar blue combined with ADC113’s broad

spectrum of activity suggests an alternative mechanism of action. Taken together, the results of

the pilot suggest that a larger HTS is likely to produce a number of novel prodrug leads.

We also examined the main predictions of the prodrug model – the ability of compounds to have

a broad spectrum of action, and kill dormant cells. There was little effect of MDRs on the MIC of

prodrugs, in contrast to erythromycin, a model substrate of MDRs. This suggests that irreversible

binding to targets may indeed impart prodrugs with good penetration properties, enabling a broad

spectrum of action. Importantly, prodrugs were able to effectively kill biofilms of E. coli and

stationary cells that produce large amounts of dormant persisters. Biofilms of E. coli are often

associated with recurring urinary tract infection (UTI). E. coli forms intracellular biofilms in the

bladder epithelial cells (91), and these are likely to contain persister cells. In some patients with

recurring UTI, E. coli form “quiescent intracellular reservoirs” (QIRs) residing in Lamp1+

endosomes of bladder epithelial cells (96). QIRs are protected from the immune system and are

not killed by antibiotics (97). When bladder epithelial turnover occurs, and antibiotic treatment

has ceased, viable E. coli reemerge and cause a recurrence of the UTI (98). QIRs may be equivalent

to persisters. Pathways of persister formation are highly redundant, precluding development of

specific anti-persister compounds. Activation of the ClpP protease in S. aureus by

acyldepsipeptide leads to the degradation of proteins and death of persisters, resulting in pathogen

eradication in vitro and in vivo (99).Target-specific antibiotics such as ciprofloxacin kill the bulk

of the population, leaving approximately 1% of surviving persisters. Increasing the concentration

103

of ciprofloxacin does not lead to more killing. Nitrofurantoin killed stationary cells much better

than ciprofloxacin, and ADC111 almost completely eradicated a stationary population. An

irreversibly binding compound will actually be more effective against non-growing dormant cells;

in rapidly propagating bacteria, a prodrug will be diluted, diminishing the sink effect.

Taken together, results of this study suggest that prodrugs are a promising type of antimicrobials

capable of sterilizing and broad-spectrum activity. The screen we developed provides a platform

for the discovery of prodrugs that have been overlooked in conventional screening campaigns.


Growth of Bacterial Strains

Strains of E. coli, S. aureus, F. tularensis, B. anthracis, Y. pestis, S. Typhimurium, A. baumannii, and P.

aeruginosa were grown in cation adjusted Mueller Hinton II broth (CA-MHB; BD cat. # 212322). Strains

of E. faecalis, E. faecium, and C. difficile were grown in Brain Heart Infusion (BHI; BD cat. #211059)

broth supplemented with yeast extract (5 g/L), cysteine (1 g/L), and hemin (15 mg/L). All strains were

grown at 37˚C with aeration at 220 rpm except for E. faecium, which was grown statically.








104












Reference/ Strain designationd


(71)



(71)

tolC ΔtolC::kan JW5503 (71) acrB ΔacrB::kan JW0451 (71) nfsA+ BW25113pZS*24nfsA (73) nfsA- ΔnfsA::kan JW0835 (71) nfsB+ BW25113pZS*24nfsB (73) nfsB- ΔnfsB::kan JW0567 (71) Francisella tularensis

SchuS4a

Bacillus anthracis Amesa Yersinia pestis KIMa Escherichia coli ATCC 25922

105

Staphylococcus aureus

NCTC 8325

Enterococcus faecalis

ATCC 47077

Salmonella Typhimurium (LT2)

ATCC 700720

Staphylococcus aureus (MRSA)

NRS54, Novobiotic

Clostrium difficile (CD196)

Clinical isolate, CD196c

Enterococcus faecium (VRE BM4147)

BM4147, Novobiotic

Acinetobacter baumannii

ATCC 17978

Pseudomonas aeruginosa (PAO1)

ATCC BAA-47

aStrain designation from the NERCE/BEID facility. b Strain designation from the Network on Antimicrobial Resistance in Staphylococcus aureus, NARSA. cThis clinical isolate was a gift from Dr. Linc Sonenshein at Tufts University. d Abbreviations: ATCC, American Type Culture Collection; NCTC, National Collection of Type Cultures.

Alamar Blue Reduction Screen

In order to test the fidelity of the screen, Z prime (Z’) scores were determined as described by (79),

by using NFT at 4X MIC for the positive control in six columns of a 96 well plate, and 1% DMSO

as the negative control in the other 6 columns of the 96 well plate. The “Prodrug Hit” Z’ was

determined from fluorescence readings after four hours of incubation and the direct activity Z’ was

determined by analyzing OD600 data after 24 hours of incubation. Each screening plate also

contained a negative and positive control column at each end of the 96 well plate.

Antimicrobials of interest were added to separate wells of a 96 well plate with black sides and a

clear bottom (Costar no. 3094). A 10% solution of Alamar Blue (Thermo Scientific Cat.

No.:PI88952) in MHB along with antibiotics and test compounds were added to the screening

plate. Broth cultures of E. coli were grown in MHB to exponential phase, ~107 CFU/mL, diluted

1:10 in MHB, and added to the screening plate containing MHB with 10% Alamar Blue and

106

antimicrobial, resulting in a final antibiotic concentration of 50 µg/mL. Control wells contained

cells with Alamar Blue, but no antimicrobial compound or Alamar Blue and antimicrobial

compound without cells. A fluorometer (Spectra MAX GeminiXS ) was used to take readings at

an excitation wavelength of 544 nm and emission wavelength of 590 nm. Kinetic readings were

taken at 37˚C every 10 minutes for 240 minutes.

MIC Determination


were determined according to CLSI recommendations (80, 81). MIC determination for all BSL3

agents were performed at the New England Regional Center of Excellence/Biodefense and

Emerging Infectious Diseases (NERCE/BEID) facility at Harvard Medical School in Boston, MA.

Time and Concentration Dependent Killing Assays

Prior to the addition of antibiotics and test compounds, overnight cultures were diluted 100-fold

into 3 mL of fresh medium in 17- by 100-mm polypropylene tubes and incubated for 1.5 hours

with aeration at 220 rpm to a cell concentration of ~2x108 CFU/mL. For determination of CFU

counts, cells were washed in 1% NaCl solution, serially diluted by ten-fold and plated on LB

(Luria-Bertani medium) agar plates supplemented with 20 mM MgSO4.

Biofilm Killing

E. coli biofilms were grown by the hanging-peg model as previously described (100). Briefly, a

device containing 96 polystyrene pegs was used (Nunc no. 445497), with a single peg hanging

into each well of a microtiter plate (Nunc no. 269787). For biofilm formation, the pegs were placed

107

in a sterile 96 well plate filled with MHB and cells (105/mL) and incubated for 24 hours at 37°C.

Once the biofilms formed on the pegs they were washed in MHB and placed into a sterile microtiter

plate with fresh MHB for drug susceptibility testing. Following 24 hour incubation in the presence

of an antimicrobial agent, the pegs were washed twice in MHB. The pegs were then moved into a

fresh sterile microtiter plate with MHB and incubated for 15 minutes in a water bath sonicator

(Branson Ultrasonic Cleaner). For each antimicrobial concentration tested, cells were collected

from four parallel pegs, serially diluted ten-fold and plated for colony counting.

Cytotoxicity

All mammalian cell lines were grown in vented 75 cm2 tissue culture treated flasks (BD Falcon

no. 353136) with 5% CO2 at 37°C in Eagle’s Minimal Essential Medium (EMEM) with L-

Glutamine (ATCC 30-2003), supplemented with either 10% or 20% of Fetal Bovine Serum (FBS)

(ATCC 30-2020). Attached cells were removed from flasks using Trypsin-EDTA (0.25% Trypsin

0.53mM EDTA) and counted using a hemocytometer. 2 x 104 cells were added to each well of a

black sided, clear flat bottom, tissue culture treated 96 well microtiter plate (Costar 3904)

containing appropriately diluted test compounds. Amphotericin B was used as a control. Cells

were incubated in the presence of compounds for 24 hours. The challenged cells were then washed

three times with fresh media and left to recover in fresh EMEM with FBS for 24 hours. The media

was then aspirated, and fresh medium containing 10% alamar blue was added to each well. A

fluorometer (Spectra MAX GeminiXS) was used to take readings at an excitation wavelength of

544nm and emission wavelength of 590nm. The cytotoxic concentration is determined as greater

than or equal to 50% transmittance of the control.

108

Acknowledgments

This study was supported by NIH grant T-RO1 AI085585 and the American Lebanese Syrian

Associated Charities (ALSAC), St. Jude Children’s Research Hospital.

We thank Marin Vulić for assistance with E. coli strain construction.

We would like to thank the NERCE/BEID facility at Harvard Medical School in Boston, MA for MIC

and MBC testing against BSL3 agents.

We would like to thank Novobiotic Pharmaceuticals for sharing bacterial strains.

We would like to thank Dr. Linc Sonenshein at Tufts University for sharing bacterial strains.

109

Chapter 5:

Discussion

110

5.1 Discussion:

Prodrugs have the features of a theoretically “ideal” antibiotic – a compound that is broad-

spectrum, non-toxic, and has the ability to kill both growing and dormant persister cells (66); (Fig.

1-5). Poor penetration across the complex envelope of Gram-negative bacteria is the main reason





do not have a specific target (67, 68), and once validation steps such as specificity tests were

introduced, it excluded this type of antimicrobial from the discovery process. We decided to

reexamine prodrugs, and developed a screen based on lack of specificity, essentially looking at

compounds that are discarded in conventional HTS campaigns. Using the viability dye alamar

blue, we identified hits acting against E. coli that inhibited reduction of this reporter of general

metabolism, and then tested the compounds for cytotoxicity. The screen successfully differentiates

target-specific compounds from known prodrugs such as nitrofurantoin. A pilot HTS with 55,000

compounds produced hits, and 3 of these with low cytotoxicity were examined. Importantly, one

of the compounds, ADC111, is a nitrofurazone analog of nitrofurantoin which is used to treat UTI

(3, 37). The activity of ADC111 depended on the presence of nitroreductases in E. coli. The finding

of a prodrug validates the screen. ADC112 is an analog of another antimicrobial that has been used

in the clinic, Intetrix. The mode of action of Intetrix is unknown. We also identified a compound

ADC113 that does not belong to known classes of antimicrobials. Taken together, the results of

this project suggest that a larger HTS is likely to produce a number of novel broad spectrum

prodrug leads.

111

We also examined the main predictions of the prodrug model – the ability of compounds

to have a broad spectrum of action, and kill dormant cells. There was little effect of MDRs on the

MIC of prodrugs, in contrast to erythromycin, a model substrate of MDRs. This suggests that

irreversible binding to targets may indeed impart prodrugs with good penetration properties,

enabling a broad spectrum of action. Importantly, prodrugs were able to effectively kill biofilms

of E. coli and stationary cells that produce large amounts of dormant persisters.

112

Literature Cited

1. Gootz TD. 1990. Discovery and development of new antimicrobial agents. Clinical

microbiology reviews 3:13-31.

2. Waksman SA, Woodruff HB. 1940. The Soil as a Source of Microorganisms Antagonistic

to Disease-Producing Bacteria. Journal of Bacteriology 40:581-600.

3. Shah RR, Wade G. 1989. Reappraisal of the risk/benefit of nitrofurantoin: review of

toxicity and efficacy. Adverse drug reactions and acute poisoning reviews 8:183-201.

4. Haas LF. 1999. Papyrus of Ebers and Smith. Journal of Neurology and Neurosurgical

Psychiatry 67:578.

5. Bickel L. 1972. Rise Up to Life. Charles Scribner's Sons, New York.

6. Hays EE, Wells IC, Katzman PA, Cain CK, Jacobs FA, Thayer SA, Doisy EA, Gaby

WL, Roberts EC, Muir RD, Carroll CJ, Jones LR, Wade NJ. 1945. ANTIBIOTIC

SUBSTANCES PRODUCED BY PSEUDOMONAS AERUGINOSA. Journal of

Biological Chemistry 159:725-750.

7. Lloyd NC, Morgan HW, Nicholson BK, Ronimus RS. 2005. The composition of

Ehrlich's salvarsan: resolution of a century-old debate. Angew Chem Int Ed Engl 44:941-

944.

8. Finkelstein R, Birkeland JM. 1938. The Mode of Action of Sulfanilamide and Prontosil.

Science 87:441-442.

9. Waksman SA, Reilly HC, Schatz A. 1945. Strain Specificity and Production of Antibiotic

Substances: V. Strain Resistance of Bacteria to Antibiotic Substances, Especially to

Streptomycin. Proceedings of the National Academy of Sciences of the United States of

America 31:157-164.

113

10. Kresge N, Simoni RD, Hill RL. 2004. Selman Waksman: the Father of Antibiotics.

Journal of Biological Chemistry 279:e7.

11. Lewis K. 2013. Platforms for antibiotic discovery. Nature reviews. Drug discovery 12:371-

387.

12. Waxman D, Strominger, Jl. 1983. Penicillin-Binding Proteins and the Mechanism of

Action of beta lactam Antibiotics. Annual Review of Biochemistry 52:825-869.

13. Blais J, Cho D, Tangen K, Ford C, Lee A, Lomovskaya O. 1999. Efflux pump inhibitors

enhance the activity of antimicrobial agents against a broad selection of bacteria. 39th

Interscience Conference for Antimicrobial Agents and Chemotherapy, San Francisco.

14. Lewis K. 1994. Multidrug resistance pumps in bacteria: variations on a theme. Trends

Biochem Sci 19:119-123.

15. Lewis K. 2010. Persister cells. Annu Rev Microbiol 64:357-372.

16. Lafleur MD, Kumamoto CA, Lewis K. 2006. Candida albicans biofilms produce

antifungal-tolerant persister cells. Antimicrobial agents and chemotherapy 50:3839-3846.

17. Sharma SV, Lee DY, Li B, Quinlan MP, Takahashi F, Maheswaran S, Mcdermott U,

Azizian N, Zou L, Fischbach MA, Wong KK, Brandstetter K, Wittner B, Ramaswamy

S, Classon M, Settleman J. 2010. A chromatin-mediated reversible drug-tolerant state in

cancer cell subpopulations. Cell 141:69-80.

18. Lewis K. 2007. Persister cells, dormancy and infectious disease. Nat Rev Microbiol 5:48-

56.

19. Costerton JW, Stewart PS, Greenberg EP. 1999. Bacterial biofilms: a common cause of

persistent infections, p. 1318-1322, Science, vol. 284.

114

20. Mulcahy L, Burns J, 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.

21. Keren I, Shah D, Spoering A, Kaldalu N, Lewis K. 2004. Specialized persister cells and

the mechanism of multidrug tolerance in Escherichia coli. J Bacteriol 186:8172-8180.

22. Maisonneuve E, Shakespeare LJ, Jorgensen MG, Gerdes K. 2011. Bacterial persistence

by RNA endonucleases. Proceedings of the National Academy of Sciences of the United

States of America 108:13206-13211.

23. Ma S, Km P, W X, S H, K L, Rg B. 2009. Molecular mechanisms of HipA-mediated

multidrug tolerance and its neutralization by HipB. Science 323:396.

24. Dorr T, Vulić M, Lewis K. 2010. Ciprofloxacin causes persister formation by inducing

the TisB toxin in Escherichia coli. Plos Biol 8:e1000317.

25. Keren I, Wu Y, Innocencio J, Mulcahy L, Lewis K. 2013. Killing by antibiotics does

not depend on reactive oxygen species. Science.

26. Hooper DC. 2001. Mechanisms of action of antimicrobials: focus on fluoroquinolones.

Clinical infectious diseases : an official publication of the Infectious Diseases Society of

America 15:S9-S15.

27. Singh R, Manjunatha U, Boshoff HI, Ha YH, Niyomrattanakit P, Ledwidge R, Dowd

CS, Lee IY, Kim P, Zhang L, Kang S, Keller TH, Jiricek J, Barry CE, 3rd. 2008. PA-

824 kills nonreplicating Mycobacterium tuberculosis by intracellular NO release. Science

322:1392-1395.

115

28. Ellner JJ. 2009. The emergence of extensively drug-resistant tuberculosis: a global health

crisis requiring new interventions: Part II: scientific advances that may provide solutions.

Clinical and translational science 2:80-84.

29. Diacon AH, Dawson R, Du Bois J, Narunsky K, Venter A, Donald PR, Van Niekerk

C, Erondu N, Ginsberg AM, Becker P, Spigelman MK. 2012. Phase II dose-ranging

trial of the early bactericidal activity of PA-824. Antimicrobial agents and chemotherapy

56:3027-3031.

30. Zhang Y, Heym B, Allen B, Young D, Cole S. 1992. The catalase-peroxidase gene and

isoniazid resistance of Mycobacterium tuberculosis. Nature 358:591-593.

31. Hirano K, Takahashi M, Kazumi Y, Fukasawa Y, Abe C. 1997. Mutation in pncA is a

major mechanism of pyrazinamide resistance in Mycobacterium tuberculosis. Tuber Lung

Dis 78:117-122.

32. Mcosker CC, Fitzpatrick PM. 1994. Nitrofurantoin: mechanism of action and

implications for resistance development in common uropathogens. J Antimicrob

Chemother 33 Suppl A:23-30.

33. Debarber AE, Mdluli K, Bosman M, Bekker LG, Barry CE, 3rd. 2000. Ethionamide

activation and sensitivity in multidrug-resistant Mycobacterium tuberculosis. Proceedings

of the National Academy of Sciences of the United States of America 97:9677-9682.

34. Vannelli TA, Dykman A, Ortiz De Montellano PR. 2002. The antituberculosis drug

ethionamide is activated by a flavoprotein monooxygenase. The Journal of biological

chemistry 277:12824-12829.

35. Goodwin A, Kersulyte D, Sisson G, Veldhuyzen Van Zanten SJ, Berg DE, Hoffman

PS. 1998. Metronidazole resistance in Helicobacter pylori is due to null mutations in a gene

116

(rdxA) that encodes an oxygen-insensitive NADPH nitroreductase. Mol Microbiol 28:383-

393.

36. Sisson G, Goodwin A, Raudonikiene A, Hughes NJ, Mukhopadhyay AK, Berg DE,

Hoffman PS. 2002. Enzymes associated with reductive activation and action of

nitazoxanide, nitrofurans, and metronidazole in Helicobacter pylori. Antimicrobial agents

and chemotherapy 46:2116-2123.

37. Guay DR. 2001. An update on the role of nitrofurans in the management of urinary tract

infections. Drugs 61:353-364.

38. Johnson JR, Berggren T, Conway AJ. 1993. Activity of a nitrofurazone matrix urinary

catheter against catheter-associated uropathogens. Antimicrobial agents and chemotherapy

37:2033-2036.

39. Aminov RI. 2010. A brief history of the antibiotic era: lessons learned and challenges for

the future. Frontiers in microbiology 1:134.

40. Baltz RH. 2008. Renaissance in antibacterial discovery from actinomycetes. Curr Opin

Pharmacol 8:557-563.

41. Schatz A, Bugie E, Waksman SA. 1944. Streptomycin, a substance exhibiting antibiotic

activity against gram-positive and gram-negative bacteria. Proc. Soc. Exp. Biol. Med.

55:66-69.

42. Baltz RA. 2005. Antibiotic discovery from actinomycetes: Will a renaissance follow the

decline and fall? . SIM News 55:188-196.

43. Lewis K. 2012. Antibiotics: Recover the lost art of drug discovery. Nature 485:439-440.

117

44. Lomovskaya O, Lewis K. 1992. Emr, an Escherichia coli locus for multidrug resistance.

Proceedings of the National Academy of Sciences of the United States of America

89:8938-8942.

45. Li XZ, Nikaido H. 2004. Efflux-mediated drug resistance in bacteria. Drugs 64:159-204.

46. Payne DJ, Gwynn MN, Holmes DJ, Pompliano DL. 2007. Drugs for bad bugs:

confronting the challenges of antibacterial discovery. Nature reviews. Drug discovery

6:29-40.

47. Renau T, Leger R, Flamme E, Sangalang J, She, Yen, Ford, Mathias, Lee A,

Lomovskaya O, Chamberland S, Lee VL, Hecher S, Otha T, Nakayama K. 1999a.

Inhibitors of efflux pumps in Pseudomonas aerruginosa potentiate the activity of the

fluoroquinolone antibacterial levofloxacin. 39th Interscience Conference for Antimicrobial

Agents and Chemotherapy, San Francisco.

48. Lomovskaya O, Warren MS, Lee A, Galazzo J, Fronko R, Lee M, Blais J, Cho D,

Chamberland S, Renau T, Leger R, Hecker S, Watkins W, Hoshino K, Ishida H, Lee

VJ. 2001. Identification and characterization of inhibitors of multidrug resistance efflux

pumps in Pseudomonas aeruginosa: novel agents for combination therapy. Antimicrobial

agents and chemotherapy 45:105-116.

49. Symmons MF, Bokma E, Koronakis E, Hughes C, Koronakis V. 2009. The assembled

structure of a complete tripartite bacterial multidrug efflux pump. Proceedings of the

National Academy of Sciences of the United States of America 106:7173-7178.

50. Murakami S, Nakashima R, Yamashita E, Yamaguchi A. 2002. Crystal structure of

bacterial multidrug efflux transporter AcrB. Nature 419:587-593.

118

51. Yu EW, Mcdermott G, Zgurskaya HI, Nikaido H, Koshland DE, Jr. 2003. Structural

basis of multiple drug-binding capacity of the AcrB multidrug efflux pump. Science

300:976-980.

52. Nakashima R, Sakurai K, Yamasaki S, Nishino K, Yamaguchi A. 2011. Structures of

the multidrug exporter AcrB reveal a proximal multisite drug-binding pocket. Nature

480:565-569.

53. Alekshun MN, Levy SB. 2007. Molecular mechanisms of antibacterial multidrug

resistance. Cell 128:1037-1050.

54. Boucher HW, Talbot GH, Bradley JS, Edwards JE, Gilbert D, Rice LB, Scheld M,

Spellberg B, Bartlett J. 2009. Bad bugs, no drugs: no ESKAPE! An update from the

Infectious Diseases Society of America. Clinical infectious diseases : an official

publication of the Infectious Diseases Society of America 48:1-12.

55. Bush K, Courvalin P, Dantas G, Davies J, Eisenstein B, Huovinen P, Jacoby GA,

Kishony R, Kreiswirth BN, Kutter E, Lerner SA, Levy S, Lewis K, Lomovskaya O,

Miller JH, Mobashery S, Piddock LJ, Projan S, Thomas CM, Tomasz A, Tulkens PM,

Walsh TR, Watson JD, Witkowski J, Witte W, Wright G, Yeh P, Zgurskaya HI. 2011.

Tackling antibiotic resistance. Nat Rev Microbiol 9:894-896.

56. Miesel L, Greene J, Black TA. 2003. Genetic strategies for antibacterial drug discovery.

Nat Rev Genet 4:442-456.

57. Sun DEA. 2001, p 77. 41st Interscience Conference on Antimicrobial Agents and

Chemotherapy, Chicago.

119

58. Schmid MB. 2001. New targets and strategies for identification of novel classes of

antibiotics, p. 197-208. In Hughes D, Andersson DI (ed.), Antibiotic Development and

Resistance. Taylor and Francis, New York.

59. Kitagawa M, Ara T, Arifuzzaman M, Ioka-Nakamichi T, Inamoto E, Toyonaga H,

Mori H. 2005. Complete set of ORF clones of Escherichia coli ASKA library (a complete

set of E. coli K-12 ORF archive): unique resources for biological research. DNA Res

12:291-299.

60. Muller M. 1986. Reductive activation of nitroimidazoles in anaerobic microorganisms.

Biochem Pharmacol 35:37-41.

61. Van Der Wouden EJ, Thijs JC, Kusters JG, Van Zwet AA, Kleibeuker JH. 2001.

Mechanism and clinical significance of metronidazole resistance in Helicobacter pylori.

Scand J Gastroenterol Suppl:10-14.

62. Gerdes SY, Scholle MD, Campbell JW, Balazsi G, Ravasz E, Daugherty MD, Somera

AL, Kyrpides NC, Anderson I, Gelfand MS, Bhattacharya A, Kapatral V, D'souza

M, Baev MV, Grechkin Y, Mseeh F, Fonstein MY, Overbeek R, Barabasi AL, Oltvai

ZN, Osterman AL. 2003. Experimental determination and system level analysis of

essential genes in Escherichia coli MG1655. J Bacteriol 185:5673-5684.

63. Bryskier A, Ed. 2005. Antimicrobial Agents: Antibacterials and Antifungals. ASM Press,

Washington, DC.

64. Gershon H, Parmegiani R. 1963. Antimicrobial activity of 8-quinolinol, its salts with

salicylic acid and 3-hydroxy-2-naphthoic acid, and the respective copper (II) chelates in

liquid culture. Appl Microbiol 11:62-65.

120

65. Whiteway J, Koziarz P, Veall J, Sandhu N, Kumar P, Hoecher B, Lambert IB. 1998.

Oxygen-insensitive nitroreductases: analysis of the roles of nfsA and nfsB in development

of resistance to 5-nitrofuran derivatives in Escherichia coli. J Bacteriol 180:5529-5539.

66. K L. 2007. Persister cells, dormancy and infectious disease. Nat. Rev. Microbiol. 5:48.

67. Edwards DI. 1993. Nitroimidazole drugs-action and resistance mechanisms I. Mechanism

of action. Journal of Antimicrobial Chemotherapy 31:9-20.

68. Müller M. 1986. Reductive activation of nitroimidazoles in anaerobic microorganisms.

Biochemical Pharmacology 35:37-41.

69. Bhowruth V, Brown AK, Besra GS. 2008. Synthesis and biological evaluation of NAS-

21 and NAS-91 analogues as potential inhibitors of the mycobacterial FAS-II dehydratase

enzyme Rv0636. Microbiology 154:1866-1875.

70. Miller JJH. 1992. Short Course in Bacterial Genetics CSH Press.

71. Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, Baba M, Datsenko KA, Tomita

M, Wanner BL, Mori H. 2006. Construction of Escherichia coli K-12 in-frame, single-

gene knockout mutants: the Keio collection. Mol Syst Biol 2.

72. Datsenko KA, Wanner BL. 2000. One-step inactivation of chromosomal genes in

Escherichia coli K-12 using PCR products. Proceedings of the National Academy of

Sciences of the United States of America 97:6640-6645.

73. Lutz R, Bujard H. 1997. Independent and tight regulation of transcriptional units in

Escherichia coli via the LacR/O, the TetR/O and AraC/I1-I2 regulatory elements. Nucleic

acids research 25:1203-1210.

74. Ausubel FM, Brent, R., Kingston, R.E., Moore, D.D.,, Seidman JG, Smith, J.A., and

Struhl, K. . 1987. Current Protocols in Microbiology. John Wiley & Sons, New York.

121

75. Ausubel FM, Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A. 1992.

Short Protocols in Microbiology. Green Publishing Associates and John Wiley & Sons,

New York, NY.

76. Sambrook J, Fritsch EF, Maniatis T. 1989. Molecular Cloning: A Laboratory Manual,

Second Edition ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor.

77. Blattner FR, Plunkett G, 3rd, Bloch CA, Perna NT, Burland V, Riley M, Collado-

Vides J, Glasner JD, Rode CK, Mayhew GF, Gregor J, Davis NW, Kirkpatrick HA,

Goeden MA, Rose DJ, Mau B, Shao Y. 1997. The complete genome sequence of

Escherichia coli K-12. Science 277:1453-1462.

78. Mccoy AJ, Adams NE, Hudson AO, Gilvarg C, Leustek T, Maurelli AT. 2006. L,L-

diaminopimelate aminotransferase, a trans-kingdom enzyme shared by Chlamydia and

plants for synthesis of diaminopimelate/lysine. Proceedings of the National Academy of

Sciences of the United States of America 103:17909-17914.

79. 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 4:67-

73.

80. Clsi. 2012. Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria That Grow

Aerobically; Approved Standard-Ninth Edition, CLSI Document M07-A9. Clinical

Laboratory Standards Institute, Wayne, PA.

81. Clsi. 2012. Methods for Antimicrobial Susceptibility Testing of Anaerobic Bacteria;

Approved Standard-Eighth Edition, CLSI Document M11-A8. CLSI, Wayne, PA.

82. Sandegren L, Lindqvist A, Kahlmeter G, Andersson DI. 2008. Nitrofurantoin resistance

mechanism and fitness cost in Escherichia coli. J Antimicrob Chemother 62:495-503.

122

83. Wolf LK. 2009. Digital Briefs, Chemical & Engineering News, vol. 87.

84. Kato J-I, Hashimoto M. 2007. Construction of consecutive deletions of the Escherichia

coli chromosome. Mol Syst Biol 3.

85. Lomovskaya O, Lewis K, Matin A. 1995. EmrR is a negative regulator of the Escherichia

coli multidrug resistance pump EmrAB. J Bacteriol 177:2328-2334.

86. Cotsonas King A, Wu L. 2009. Macromolecular synthesis and membrane perturbation

assays for mechanisms of action studies of antimicrobial agents. Current protocols in

pharmacology / editorial board, S.J. Enna Chapter 13:Unit 13A 17.

87. Mount DW, Low KB, Edmiston SJ. 1972. Dominant mutations (lex) in Escherichia coli

K-12 which affect radiation sensitivity and frequency of ultraviolet lght-induced mutations.

J Bacteriol 112:886-893.

88. Hilliard JJ, Goldschmidt RM, Licata L, Baum EZ, Bush K. 1999. Multiple mechanisms

of action for inhibitors of histidine protein kinases from bacterial two-component systems.

Antimicrobial agents and chemotherapy 43:1693-1699.

89. Okusu H, Ma D, Nikaido H. 1996. AcrAB efflux pump plays a major role in the antibiotic

resistance phenotype of Escherichia coli multiple-antibiotic-resistance (Mar) mutants. J.

Bacteriol. 178:306-308.

90. Tegos G, Stermitz FR, Lomovskaya O, Lewis K. 2002. Multidrug pump inhibitors

uncover remarkable activity of plant antimicrobials. Antimicrobial agents and

chemotherapy 46:3133-3141.

91. Anderson GG, Palermo JJ, Schilling JD, Roth R, Heuser J, Hultgren SJ. 2003.

Intracellular bacterial biofilm-like pods in urinary tract infections. Science 301:105-107.

123

92. Keren I, Kaldalu N, Spoering A, Wang Y, Lewis K. 2004. Persister cells and tolerance

to antimicrobials. FEMS Microbiol Lett 230:13-18.

93. Spoering AL, Lewis K. 2001. Biofilms and planktonic cells of Pseudomonas aeruginosa

have similar resistance to killing by antimicrobials. J Bacteriol 183:6746-6751.

94. Gilbert DN. 2006. Urinary Tract Infections in Patients with Chronic Renal Insufficiency.

Clinical Journal of the American Society of Nephrology 1:327-331.

95. Lewin CS, Morrissey I, Smith JT. 1991. The mode of action of quinolones: the paradox

in activity of low and high concentrations and activity in the anaerobic environment. Eur J

Clin Microbiol Infect Dis 10:240-248.

96. Mysorekar IU, Hultgren SJ. 2006. Mechanisms of uropathogenic Escherichia coli

persistence and eradication from the urinary tract. Proceedings of the National Academy

of Sciences of the United States of America 103:14170-14175.

97. Schilling JD, Hultgren SJ. 2002. Recent advances into the pathogenesis of recurrent

urinary tract infections: the bladder as a reservoir for uropathogenic Escherichia coli. Int J

Antimicrob Agents 19:457-460.

98. Silverman JA, Schreiber HLT, Hooton TM, Hultgren SJ. 2013. From Physiology to

Pharmacy: Developments in the Pathogenesis and Treatment of Recurrent Urinary Tract

Infections. Current urology reports.

99. 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.

124

100. Ceri H, Olson ME, Stremick C, Read RR, Morck D, Buret A. 1999. The Calgary

Biofilm Device: new technology for rapid determination of antibiotic susceptibilities of

bacterial biofilms. Journal of clinical microbiology 37:1771-1776.

screening for broad spectrum antimicrobials with unknown targets... · 2019-02-13 · screening for...

Documents