Drug discovery notes 1 PDF

Preview

Summary

Introduction to drug discovery...

Description

Birth of DD: 1. Concept 2. Target ID + Validation 3. HTS + Hit Confirmation 4. Lead Generation 5. Lead Optimisation 6. Preclinical Development 7. Clinical Development 8. Regulatory Approval 9. Commercialisation 10. Product Pre-industrial DD themes  Plant-based extracts (rather than chemically synthesised)  No link between disease cause + drug mechanism  Efficacy/use via empirical symptoms observations  New drugs discovered by chance over design  Limited ability to regulate efficacy or safety Quinine (anti-malarial) from chinchona (can chew bark to relieve fever), then purified Digoxin (cardiac glycoside) from foxglove (treated swelling and fatigue) from congested heart disease First synthesis of natural product 1868 (chemical synthesis of Alizarin) Carl Grabe + Carl Liebermann Adolf von Baeyer determined structure + nomenclature to name aromatic compounds Paul Ehrlich     

How dyes worked/stained cells Gained access to Baeyer's data (classification system) Biological assay, effect of compounds on organisms (staining) Could relate compound structure to cell function Start of Stucture Activity Relationship (SAR)

SAR - relationship between chemical structure + biological activity, involves cycle between organic synthesis + biological assessment of output   

Suggested chemoreceptors (parts of cells could recognise dyes + structural parts, suggested different affinities) 'Magic bullet' (chemotherapy), suggests CR on microbes + cancer cells are different to host cells so can be targeted Hypothesis proves with discovery of Salvarsan, first treatment of syphillis, based on Atoxyl starting compound that treats sleeping sickness

20th century pharmaceutical discovery Aspirin - Baeyer Insulin - Bantin Penicillin - Fleming Common themes 20thC DD:  Compounds derived from organic chemical synthesis  Systematic 'trial + error' compound screening (no clear reason why they're used)



Drugs often discovered by serendipity rather than design (one found accidentally from use of another)

Gertrude Elion + George Hitchings - rational drug design Understanding biological mechanism that underlies disease + manipulating that Design compounds to interfere with this Purinethol (anti-leukaemia) 3rd person James Black ^ (ligand based drug discovery) Discover drugs from emasculated hormones Propranolol - adrenaline, remove efficacy + produce antagonist, started with isoprenaline (B agonist), 1967 Cimetidine - Histamine efficacy (amine side), adding methyl group onto ring systems to get specificity Approach- start with known endogenous natural ligands/drugs closest 1. Start with clinical problem 2. Identify controlling chemicals/hormones in system 3. Start at most basic molecular level + test similar molecules for relevant in vitro/ex vivo activity Milestones enabling modern day DD+D Problems in 1960/70  Limited starting compounds (Cimetidine only started with histamine, muscarinic antagonists different to acetylcholine)  Pharmacology (assay results turnover very slow, ex vivo bioassays for screening, in vivo bioassays for efficacy/DMPK, used first)  Chemistry (limited synthetic methodologies, requirement for gram quantities, 13 compounds made per week, synthetic choices made by intuition/no clear design strategies) E.g discovery of piroxicam took 18 years 1962-80, team of 2 (pharmacologist + medicinal chemistry, testing in vivo in animal models (dogs), lot of time + money Recombinant DNA technology DNA polymerase 1956 - Arthur Kornberg DNA ligase, reverse transcriptase, exonucleases, DNA phosphorylases/kinases, restriction enzymes, bacterial transformation rDNA - VIRUS b.Berg 1972 + synthetic plasmid vector 1973 DNA sequencing 1975 Fred Sanger/Walter Gilbery Primer-based-site-directed-mutagenesis 1978 Michael Smith PCR 1983 Kary Mullis *No knowledge of techniques needed* Recombinant Protein Production 1. Choose expression system (Bacteria, yeast, human, insect cells) 2. Design/make expression vector (recombinant DNA/rDNA) 3. Move vector into cells (Assay/screen protein for drug activity, SAR from sitedirected mutagenesis, purify protein for structural studies) 4. Culture + maintain cells Screening technologies Combinatorial chemistry, automated workstations do synthesis for you, can create large amounts of compounds rather than linear sequencing Structural biology

X-ray crystallography Nuclear magnetic resonance (receptor structure, dynamics, drug activity in binding site) Cryo-electron microscopy Computational Chemistry + Biology Chemoinformatics (chemical compound info stored in computers) Bioinformatics (Genetic info stored + accessed) QSAR/quantitative structure activity relationships (defines relationship between physiochemical drug parameters + biological activities, finds ideal drugs/binding sites) Virtual screening (chemoinformatics/structural bio, screening drugs in silico, see new compounds or find compounds that have been made that can be potential ligands/hits) In vitro ADME/DMPK + Toxicology Pharmacokinetics In vitro pitch + predictive toxicology Predicting PK properties before getting to development stage of drug Safety + regulation FDA EMA E.g thalidomide 1962 Stages in modern day DD+D The process of taking a therapeutic concept + converting it into a physical entity that is marketed for medicinal benefit + usually financial profit Discovery: Concept Target ID + Validation HTS + Hit confirmation Lead generation Lead optimisation Drug - single chemical entity that is conventionally marketed for use in medicine  Includes both conventional 'small molecule' drugs + biopolymers/vaccines  Doesn't include herbal remedies, dietary supplements (complex mixtures), recreational drugs (not conventionally marketed or medicinal or hits/leads/candidates that are created in drug discovery process (although pharmacologically classed as drugs SAR - relationship between drug's chemical structure + bio activity Pharmacophore - 3D ensemble of steric + electronic molecule features necessary to ensure optimal molecular interactions with a specific bio target structure in order to bind with sufficient activity to activate/block the bio response (defines SAR) Analytical pharmacology: 1. Recombinant DNA technology - generates recombinant cell lines that express drug target using target cDNA 2. Screening technologies - robotics + assays Work out affinities, intrinsic activities, potencies, pharmacological outputs More compounds needed from medicinal chemists Large compound collections with screening technologies allowed high throughput screening to create more ligands Pharmacophore leads to lead compound

Lead optimization needs to be done before development into animals (making sure PK are robust), concentrate on ADME profile Pre-clinical development candidate goes in + needs to be re-synthesised, development chemists start process to find most efficient way to make large quantities (more regulated) Pre-clinical stages look at toxicology + in vivo PK Clinical trials in humans (after ethics + safety regulations) then develops therapeutic drug Structure-based drug discovery Recombinant DNA technology + structural biology shows target 3d structure Computational approaches (molecular dynamics + modelling) help understand structure + binding site Virtual compounds + HTS in silico can generate new ligands + compound fragments that can bind to binding site Concept - from translational chemistry (observation in interesting substance exhibiting bio activity of interest to find link between chemical + activity, or translational bio/medicine more recently (observation of bio property or clinical presentation + development of a concept that can be altered for therapeutic benefit) Target ID - analyse pathophysiological pathways underlying disease + identify 'amendable' proteins in pathway that can be targets, knock-outs or known drug targets can help Validation - efficacy in clinic is needed, evidence from other approved drugs that act on same target is clear evidence, more difficult if it is a novel target (in vivo disease models or knock-out/transgenic) Hit finding - full scale HTS may involve entire compound collection, limited conc of each is used + data looks at 'yes' or 'no' to hits, HTS screen can be more focused subsets of compound collection Hit confirmation - separating true hits from artefacts, hit becomes confirmed which are further checked (e.g whole groups) + become qualifies hits Lead generation - most promising qualified hits are starting points for further modification, may be several classes of compound (keep options open), use combinatorial chem to generate lots of focus libraries around each hit/family, look at SAR + predicted PK + toxicology issues to learn more about pharmacophore, best candidate is classed as lead compound and taken forward to optimisation Lead optimisation - balancing efficacy with PK, all about balance, good in vitro + ex vivo activity at target but needs good PK + low toxicity aswell, in vitro ADME/DMPK important as new compounds being generated, most promising are testing in vivo animal studies to ensure PK predictions translate Challenges + progress in recent DD+D HTS started in 90's, gradually drop in new molecular entities coming into market from 99 while research + development costs rising Drug failures (2000): 39% poor PK/ADME (wrong library/not drug-like) 30% poor efficacy (wrong target/not validated) 11% toxicity in animal trials 10% adverse effects in humans 5% commercial reasons 2008:

28% poor efficacy 26% commercial reasons (competitive nature, risk averse, will bail early) 19% toxicity in animal trials 12% adverse effects in humans 9% poor PK/ADME (companied won't pursue non drug-like products) Target-based screening Pro - chemical knowledge can be applied via molecular hypothesis (targets + interactions known) Con - hypothesis may not be relevant to pathogenesis/therapy (target validation, target can be affected but won't have efficacy for disease) Required hypothesis: 1. Selected target is important in human disease 2. Molecular MOA of drug can achieve bio response 3. Assay activity translates effectively into therapeutic activity Return to phenotypic screening? E.g James Black Pro - 'direct read-out', no prior understanding of molecular MOA, if drug works then is accepted Con - low through-put assays + difficult to optimise molecule's properties *Swinney, D & Anthony, J 2011 comparing phenotypic vs target based*  More drugs came through for phenotypic screening  Phenotypic better for new molecular entity 1st in-class  Target best for 'follower drug'  Followers often best in class  Target better if target already validated, if molecular MOA is defined, improving in 1st in-class  Final conclusion recommends using both approaches 90% failure in development 2006-2015 Average 51 NTD's a year since 2013 now Predicted 50/50 split in biotechnology + conventional/unclassified drugs in 2024 Blockbuster drugs - >US $1 billion sales per year High risk but high reward Small molecules make up 64% of current blockbusters Antibody-therapies grew from 13% to 26% Exceptional patient benefit in trials typically recognised by regulatory agencies which can accelerate development path (blockbusters 5x better)...