Combining Metabolite-Based Pharmacophores with Bayesian Machine Learning Models for Mycobacterium...

Combining Metabolite-Based Pharmacophores

with Bayesian Machine Learning Models for

Mycobacterium tuberculosis Drug Discovery

Sean Ekins1,2*, Peter B. Madrid3*, Malabika Sarker3, Shao-Gang Li4,

Nisha Mittal4, Pradeep Kumar5, Xin Wang4, Thomas P. Stratton4,

Matthew Zimmerman,6 Carolyn Talcott3, Pauline Bourbon3, Mike

Travers1, Maneesh Yadav3 and Joel S. Freundlich4*

1Collaborative Drug Discovery Inc., 1633 Bayshore Highway, Suite 342, Burlingame, CA 94010, USA.2Collaborations in Chemistry, 5616 Hilltop Needmore Road, Fuquay-Varina, NC 27526, USA.

3SRI International, 333 Ravenswood Avenue, Menlo Park, CA 94025, USA.4Departments of Pharmacology & Physiology and Medicine, Center for Emerging and Reemerging Pathogens,

Rutgers University – New Jersey Medical School, 185 South Orange Avenue, Newark, NJ 07103, USA.5Department of Medicine, Center for Emerging and Reemerging Pathogens, Rutgers University – New Jersey Medical

School, 185 South Orange Avenue, Newark, NJ 07103, USA.

.

streptomycin (1943)

para-aminosalicyclic acid (1949)

isoniazid (1952)

pyrazinamide (1954)

cycloserine (1955)

ethambutol (1962)

rifampicin (1967)

Globally ~$500M in R&D /yr

Multi drug resistance in 4.3%

of cases

Extensively drug resistant

increasing incidence

one new drug (bedaquiline) in

40 yrs

TB key points

Tested >350,000 molecules Tested ~2M 2M >300,000

>1500 active and non toxic Published 177 100s 800

Big Data: Screening for New Tuberculosis Treatments

How many will become a new drug?How do we learn from this big data?

TBDA screened over 1 million, 1 million more to go

TB Alliance + Japanese pharma screens

~ 20 public datasets

for TB

Including Novartis

data on TB hits

>300,000 cpds

Patents, Papers

Annotated by CDD

Open to browse by

anyone

http://www.collaborativedrug.

com/register

Molecules with activity

against

Over 8000 molecules with dose

response data for Mtb in CDD

Public from NIAID/SRI

Phase I - Mimic strategy

1. The enzymes around these metabolites are

"in vivo essential".

2. These enzymes have no human homolog.

3. These enzyme targets are not yet explored

though some enzymes from the same

pathways are drug targets (experimental or

predicted).

Multi-step process

1. Identification of essential in vivo enzymes of Mtb involved intensive

literature mining and manual curation, to extract all the genes essential

for Mtb growth in vivo across species.

2. Homolog information was collated from other studies.

3. Collection of metabolic pathway information involved using TBDB.

4. Identifying molecules and drugs with known or predicted targets

involved searching the CDD databases for manually curated data. The

structures and data were exported for combination with the other data.

5. All data were combined with URL links to literature and TBDB and

deposited in the CDD database.

Initially over 700 molecules in dataset

Dataset Curation: TB molecules and target information

database connects molecule, gene, pathway and literature

Sarker et al., Pharm Res 2012, 29, 2115-2127.

TB molecules and target information database connects

molecule, gene, pathway and literature


Pharmacophore developed (using AccelrysDiscovery Studio) from 3D conformations of the substrate

van der Waals surface for the metabolite mapped onto it

pharmacophore plus shape searched in 3D compound databases from vendors

In silico hits collated

Filtered for TB whole cell activity and reactivity

Compounds filtered based on Bayesian score using models derived from NIAID / Southern Research

Inst data to retrieve ideal molecular properties for in vitro TB activity


Two Proposed Mimics of D-fructose 1,6 bisphosphate

Computationally searched >80,000 molecules – and used bayesian

models for filter - narrowed to 842 hits -tested 23 compounds in vitro (3

picked as inactives), lead to 2 proposed as mimics of D-fructose 1,6

bisphosphate

Sarker et al., Pharm Res 2012, 29: 2115-2127

a.

b.

1R41AI088893-01

Phase II – Mimic approach expanded

Specific Aim 1: Develop molecular mimics of at least 20 additional

substrates of in vivo essential enzymes.

SPECIFIC AIM 2: Progress molecules discovered in phase I and

identify the putative target/s.

SPECIFIC AIM 3: Develop the approach into a commercial product

66 Pharmacophores of substrates and metabolites

Developed for Mtb Enzymes

Green = Hydrogen bond acceptor, Purple = hydrogen bond donor, cyan = hydrophobe

Grey – van der Waals surface

Filter hits with Bayesian Models

Top scoring molecules

assayed for

Mtb growth inhibition

Mtb screening

molecule

database/s

High-throughput

phenotypic

Mtb screening

Descriptors + Bioactivity (+Cytotoxicity)

Bayesian Machine Learning classification Mtb Model

Molecule Database

(e.g. GSK malaria

actives)

virtually scored

using Bayesian

Models

New bioactivity data

may enhance models

Identify in vitro hits and test models3 x published prospective tests ~750

molecules were tested in vitro

198 actives were identified

>20 % hit rate

Multiple retrospective tests 3-10 fold

enrichment

NH

S

N

Ekins et al., Pharm Res 31: 414-435, 2014

Ekins, et al., Tuberculosis 94; 162-169, 2014

Ekins, et al., PLOSONE 8; e63240, 2013

Ekins, et al., Chem Biol 20: 370-378, 2013

Ekins, et al., JCIM, 53: 3054−3063, 2013

Ekins and Freundlich, Pharm Res, 28, 1859-1869, 2011

Ekins et al., Mol BioSyst, 6: 840-851, 2010

Ekins, et al., Mol. Biosyst. 6, 2316-2324, 2010,

Hits found in Phase II

• Screened > 200,000 compounds, 14,733 retrieved, tested 110

• 3 actives based on standard Alamar Blue assay

• Most promising – quinoxaline di-N-oxides

Menadione Menadione Indole-3-acetamide Lipoamide

Ekins et al.,PLOS ONE, 10: e0141076, 2015

Synthetic routes to the A) arylamide and B) quinoxaline di-N-oxide families

Arylamides MIC ≥50µM – not pursued further

• quinoxaline di-N-oxide initial hit previously shown to have MIC

3.13ug/ml, no cytotox (Villar et al., J Antimicrobial Chemotherapy, 62, 547-54,

2008)


Structures of quinoxaline di-N-oxides with the most

promising antitubercular activities and selectivities


Physiochemical and ADME data

Mouse liver microsomal

stability

Kinetic

Solubility

Caco-2 Cell Permeability

Molecule Compound

remaining

after 1h in

the presence

of NADPH

(%)

Compound

remaining

after 1h in the

absence of

NADPH (%)

Solubility

Limit at 2 h

(µM)

Mean A-

>B Papp

(10-6 cm

s-1)

Mean B->A

Papp (10-6

cm s-1)

Efflux

ratioPapp

(B->A)/Papp

(A->B)

SRI50 0.06 77.5 125 0.0 0.0 N/A

SRI54 0 79.1 15.6 0.66 0.10 0.15

SRI58 63.6 110 125 2.3 0.57 0.25

SRI58 did not exhibit quantifiable

blood levels in mice

Activity of SRI50 against wild type and clinical MDR-TB

strains

Strain Drug Resistancea Strain TypeSRI50

(µg/mL)

H37Rv None Laboratory 0.16

210 None Clinical 0.31

692 pan-susceptible Clinical 0.16

91 RIF, EMB Clinical 0.16

36 INH, RIF, EMB Clinical 0.16

116 INH, EMB, PAS Clinical 0.16

31 INH, RIF, EMB, KAN, SM, CAP Clinical 0.31

a RIF = rifampicin; EMB = ethambutol; INH = isoniazid; PAS = p-aminosalicyclic acid;

KAN = kanamycin; SM = streptomycin; CAP = capreomycin


Mtb transcriptional response to SRI54 as compared to other

small molecule antituberculars and environmental stresses

131 genes up-regulated, 184 down-regulated


TB Mobile Vers.2

Ekins et al., J Cheminform 5:13, 2013Clark et al., J Cheminform 6:38 2014

Predict targetsCluster molecules

http://goo.gl/vPOKS

http://goo.gl/iDJFR

Target prediction with TB Mobile

FtsZ, CysH, DprE1 and Rv1885c

CysS


Summary

• Combining bioinformatics and cheminformatics leads to synergies

• In this study we computationally searched >206,000 Asinex Gold

molecules with over 60 pharmacophores of Mtb essential substrates or

metabolites and ultimately tested ~110 compounds in vitro

• We identified 3 compounds possessing whole cell activity against Mtb

(MIC 2.5 – 40 microg/mL)

• More stringent hit criteria than for phase I

• Our strategy identified a series that previously led to in vivo actives

• Current work confounded by poor in vivo PK

All at CDD, SRI, and many others …Funding: 2R42AI088893-02 NIAID, CDD TB has been

developed thanks to funding from the Bill and Melinda Gates Foundation (Grant#49852)

Combining Metabolite-Based Pharmacophores with Bayesian Machine Learning Models for Mycobacterium...

Science

Transcript of Combining Metabolite-Based Pharmacophores with Bayesian Machine Learning Models for Mycobacterium...