Bacterial virulence factors as targets for new antibacterials: an overview

Dr. Ariel Blocker Cellular and Molecular Medicine & Biochemistry availability timeline

Nature Rev. Micro. 5, 175 (2007) Antibiotic resistance • Mechanisms of antibiotic resistance: -Target modification (direct mutation, upregulation, alternative pathway) -Antibiotic modification -Altered permeability/efflux See Poster presentation • Acquisition of resistance: -Spontaneous mutation (frequency of occurrence ~10-8 to 10-9) -And/or gene acquisition via plasmid/transposons (frequency of transfer 10-3 to 10-4) • The extent of the problem: within a few years resistance is observed to every single antibiotic ever introduced to clinical use Consequence: insidiously rising death rates

Nature 406, 762 (2000) Why we are here I

• The industrial pipeline of novel-mechanism antibacterials is presently empty especially for Gram-negative (Nat. Rev. Microbiol. 6, 29 (2007) and see Plenary Session II): 1) Since the mid 1990s, commercial genome mining has not worked (but see FP7 antiPathoGN in Poster session III) 2) Synthesis/discovery of novel chemical scaffolds that block established targets seems the only way forward now 3) At the rate of present resource investment it could take 5-10 years and 10-15 years to see a novel mechanism drug to treat Gram-positives and Gram negatives, respectively Why we are here II

• It is increasingly apparent that present have significant negative effects in patients because:

1) They non-specifically displace the resident microflora -e.g. Clostridium difficile/ Candida secondary are increasing problems -Possible serious long-term effects on immune system development (see Plenary Session I)

2) They are often initially microbial compounds, can damage cells and/or reduce viable counts and as such they can act as signalling molecules which can enhance antibiotic resistance, virulence, and mutation rates (see Plenary Session III) Besides new antibiotics, what are our options? • Phage or other “natural killer” therapy? • Vaccination! • Host defense antimicrobial peptide derivatisation or development?

Nature Biotech. 24, 1551 (2006) Curr. Opin. Microbiol. 9, 489 (2006)

• Drug development strategies based on our present understanding of bacterial virulence strategies (Plenary Session III) Advantages/disadvantages of antivirulence drugs Pluses: • Targets are intrinsically unique to pathogens and should hence spare the natural flora • There should be little “natural resistance” to any such drugs/reagents already developed by present in the natural/host environment • Antivirulence drugs should not directly kill the , only cripple it, allowing host defense systems to clear it. This might slow evolution of resistance • May act synergistically with conventional antibiotics, because they act through independent mechanisms to block in vivo bacterial replication or survival

Minuses: • Most antivirulence drugs are unlikely to be as “broad-spectrum” as any of the present commonly used antibiotics • As virulence gene expression is a function of time and space in vivo, some antivirulence drugs may not work if administered at the wrong point in the . We need to identify “pathogenesis bottlenecks” and specifically target those Nat. Rev. Microbiol. 6, 17 (2008) What are the possible virulence targets?

• Toxins • Adhesion (when virulence specific) • Virulence-dedicated secretion systems • Virulence-dedicated two-component regulatory systems • Quorum sensing and biofilm generation systems (see Plenary session III talk, Tuesday pm) Different types of toxins

Individual secreted toxins are amongst the most pathogenic virulence factors that bacteria produce “Anti-toxins”

• Antisera raised against inactivated toxin forms (“toxoids”) are one of the oldest known form of antibacterial therapy. • Neutralising IgG fractions were obtained from horses or humans and administered post-exposure to C. tetani, C. diphteriae or C. botulinum toxins while infection with these organisms is cleared. • Humanised monoclonal antibodies are great alternative to horse sera! • In 2003 the FDA approved BabyBIG, a human Botulism Immune Globulin for treatment of infant botulism. • Presently, there are six other such types of antibodies in clinical trials, it making it the most pursued/advanced antivirulence strategy.

• But, how to target toxins directly delivered into cells is non-obvious and for diseases which are rare or where the toxin(s) only in part contributes to the disease symptoms or its rare complications (e.g. dysentery), the market may be too small to justify industrial interest. High diversity of adhesion structures

• Found first in Gram- and now in Gram+ • Afimbrial adhesins: monomeric proteins or protein complexes at cell surface (e.g. Hap autotransporter of H. influenzae or Dr of E. coli) • Fimbrial adhesins: heteropolymeric extracellular fibers called pili or fimbriae (6 different types described in E. coli) • Their expression can vary depending on site/time point of Pili of of UPEC adhering to bladder cells infection • Bind diverse host cell factors (surface/extracellular matrix carbohydrates or protein) Pili function represents a “bottleneck” in UPEC infection

• Both Type 1 and P pili key to the initial attachment of UPEC to bladder cells •Both are assembled by the “chaperone-usher pathway”, for which there are 15 other operons in E. coli and which are wide- spread in other Gram- •Thus, inhibiting this pathway has “broad spectrum potential” Molecular mechanism of chaperone- usher pathway

Donor strand exchange determined through crystalisation of most pilin and chaperone subunits in this pathway (Scott Hultgren and Gabriel Waksman) Rational design of “Pilicides” Fred Almqvist, Waksman and Hultgren PNAS 103, 17897 (2006) The chaperone R8-K112 donor strand exchange cleft region was originally targeted, followed by compounds 1a and 1b with two substituents (R1 and R2) that bind P and type 1 pili chaperones, respectively. However, poor water solubility limited their utility. Substitution of the open position in the 2-pyridone scaffold (R3) resolved this issue, resulting in 2- pyridones, 2a–2d, where 2a–2c bound chaperones in the low millimolar range and 2d is a negative control.

However, crystal structure of the PapD-2c complex revealed that 2c binds to the surface of the chaperone interacting Growth in pilicides inhibits hemagglutination, with the usher so it can only prevent host cell adhesion and biofilm formation in vitro. pili assembly, not diassemble them Two-component response systems (TCRSs)

• Dominant system by which bacteria respond to their environment • 4000 different ones identified in 400 sequenced genomes, only a few are virulence specific • Consist of 1) receiver histidine kinase that is activated by extracellular signals and 2) response regulator that transmits the signal to an intracellular target that modulates gene expression • Broad-spectrum inhibitors could non-selectively target TCRSs common to many bacteria • More selective inhibitors are required to inhibit expression of specific virulence genes Regulation of V. cholerae virulence gene expression Environmental signals: temperature, bile salts, pH

Low cell density sensing cascade

John Mekalanos, Victor DiRita Infec. Immun. 75, 5542 (2007) Identification of Virstatin, a ToxT inhibitor • HTS of a 50,000-compound small molecule library • Screening used V. cholerae strain carrying a chromosomally integrated tetA resistance gene controlled by the ctx promoter • 101 hits, 15 minimally generally toxic to bacteria • 4-[N-(1,8- naphthalimide)]-n-butyric acid (virstatin) was selected for further study • Inhibition of ctx and tcp (but not toxRS, tcpPH or toxT) expression confirmed in WT strains at 5-50 µM drug • Global transcriptional profiling used to verify that only ToxT-regulated genes where affected • ToxT drug-resistant mutant isolated, in dimerisation domain • Tested in in vivo model: orogastric administration of virstatin protects infant mice from intestinal colonization by V. cholerae. Even delayed administration reduced bacterial recovery by 3 logs Mekalanos Science 310, 670 (2005) Secretion systems involved in virulence Gram- and Gram+

• General secretion pathway (GSP) or •Basic knowledge still fairly rudimentary sec system (into and across IM to (molecular genetics was harder). periplasm) • Type I secretion (ABC-transporters) •Several systems similar to Gram- predicted from • Type II secretion (main terminal branch genome sequences (sec, Tat, ABC, T4SS, flagella and type IV pili). Probably many undiscovered of GSP and type IV pili) new ones. • Type III secretion (flagella and virulence) •ESAT-6/ESX-1/Snm secretion system in • Type IV secretion (conjugation and Mycobacterium, Staphyloccus (and probably virulence) etc…). Called Type VII. • Type V secretion (autotransporters) • Type VI secretion (very new!) • Tat-pathway (twin arginine transport), Caveat: Those systems mostly equivalent to ΔpH pathway of involved in virulence may not be chloroplasts. virulence-dedicated. Genetic, structural and functional similarities by flagella and virulence T3SS The most highly conserved part is the export machinery

F0-like?

F1-ATPase like Where does our work fit in?

Recent papers on small molecule T3SS inhibitors (Plenary and poster sessions III)

Initial screening of 9,400 chemicals for Yersinia yopE gene expression (Kauppi et al., 2003)

Inhibition of Yersinia motility (Kauppi et al., 2003)

Inhibition of Yersinia Yop secretion (Nordfelth et al., 2005)

Inhibition of Chlamydia virulence (Muschiol et al., 2006; Wolf et al., 2006; Bailey et al., 2007)

Inhibition of Salmonella T3SS-1 (Hudson et al., 2007; Negrea et al., 2007)

Results suggest conserved drug target within virulence and flagellar T3SS acylated hydrazones of different salicylaldehydes The drugs inhibit all forms of T3S in Shigella…

“Leakage” (slow secretion of effector proteins) Induced secretion (burst of effector protein secretion upon Congo red addition )

Constitutive secretion (unregulated secretion by mxiH mutants)

Fast constitutive secretion (unregulated secretion by …but only if added ipaB/D knockouts) during bacterial growth and kept in throughout, not if added at assay start. The inhibitors affect epithelial cell invasion and induction of macrophage apoptosis by Shigella

1 Effect of compounds on HeLa cell invasiveness Compound Relative invasiona) INP0400 40 ± 13b) INP0402 17 ± 11 INP0406 81 ± 13 2 Data derived are from experiments in duplicate. 3 a) Percentages given are relative to the result for the 4 DMSO control (set to 100%). 5 b) Errors given are standard deviations.

Drugs added at start of bacterial growth only, infection was for 5 min Drugs added at start of assay which then lasted 5 hr The secretion inhibitors lead to an increase in ‘needleless’ T3S secretons

Intrabacterial T3S protein levels upon growth in presence of the drugs

‘ghost’ cell preparation (osmotic shock) Intact secretons Only bases The secretion inhibitors lead to an increase in NCs with shorter needles

Using a G-test, the distribution of needle lengths in the INP400 sample was found to differ from that of all other samples at p=7.6x10-9, whereas the distribution of needle lengths in all other samples was similar to that 1 Effect of inhibitors on secreton needle length of the expected size (nm) 400 400C 406 DMSO n.a. frequencies in the control a) <35 18 ± 4 11 ± 7 9 ± 4 11 ± 1 13 ± 1 samples (p values all 35-63 66 ± 10 73 ± 4 79 ± 5 81 ± 9 81 ± 6 63-91 15 ± 13 15 ± 8 10 ± 8 8 ± 10 5 ± 4 above 0.02). 91-112 1 ± 1 1 ± 1 1 ± 0 0 ± 0 1 ± 1 2 Data derived are from experiments in duplicate or triplicate. 3 The values represent percentages of needles that fall into a class of needle 4 length. N.a. means no addition. a) 5 Errors given are standard deviations . Summary

Our Shigella data: - INP0400 is the strongest and most consistent Shigella T3SS inhibitor - Inhibitors affect assembly of the needle within the secreton, probably by generating a partial, kinetic block in the secretion of the first T3S substrates. This produces a phenocopy of an export apparatus mutant, albeit with incomplete penetrance. Due to the existence of checkpoint(s) for needle assembly, this may lead to a more complete (but quickly reversible) block in all forms of effector secretion.

Our early Salmonella work and that of others: - INP0404/405 are strongest Salmonella motility inhibitors - 2X lower number of flagella for cells grown on motility plates with drugs, probable inhibition at the level of hook synthesis (see also Negrea et al. 2007). NB: All unpublished data has been removed for online posting Next basic questions for antivirulence drugs

• Will they actually work in vivo? Only tested in very few cases so far (Poster session II) • For those that may be relatively “broad spectrum”, what will they do to commensals? • What will be their effectiveness in immunocompromised individuals? • Can they be used as adjuncts to bacteriostatic/bacteriocidal therapies? • And, what will be the in vivo resistance development/transmission frequencies? Requirements for implementation of antivirulence drug therapies • Use detailed mechanistic and/or structural understanding of virulence mechanisms of all major bacterial pathogens to develop new drugs. The adacemics are about here now • Accept that the “one-drug-fits-all” strategy for new broad spectrum antibacterials has failed. Industrial acceptance of additional development costs required. • Health systems to accept need and cost of rapid diagnosis in the clinic of organism causing infection (and possibly ability to profile its virulence genes). • Consider each specific disease, pathogen and virulence mechanism and combine the strength of synergistic therapies to minimize evolution of resistance. Health systems to accept added treatment costs. Authors and Acknowledgements

•Shigella authors: Andreas Veenendaal, Charlotta Sundin and Ariel Blocker •Salmonella authors: Andreas Veenendaal, Tohru Minamino, Isabel Martinez-Argudo, Giulia Franzoni, Simon Foster, David Studholme, Keiichi Namba and Ariel Blocker. •Acknowledgement of: Sally Makady and John Oskoff for help with unsuccessful Salmonella genetic mapping experiments and Mark Pallen for suggesting the genomic sequencing. •Enormous thanks to: Pia Keyser, Mikael Elofsson and Hans Wolf-Watz (Creative Antibiotics Ltd, Umea) for their support and for providing bucket loads of the drugs. •Funds from: EEC Marie-Curie Actions/EMBO (AKJV), MRC (IMA, ABJ), Wellcome Trust (IMA, AJB), Guy Newton Research Trust (AJB), ICORP/JST (TM, KN).