Lecture 2- Drug Discovery: From target Identification to Lead Optimization

Project Initiation

• Drug companies target markets, not disease:

  • Each new drug now costs more than $1 billion to develop

  • Companies must sell enough drug to recover R&D costs and achieve profitability

• Market analysis

• Competitive assessment

• Research Analysis

Market Analysis

• Number of potential customers 

• Type of potential customers

  • Can they pay for the drug

• Nature of disease

  • Life-threatening, quality of life, lifestyle

  • Acute(short-term) vs chronic (Long term)

• Current treatment

  • Cost

  • Effectives

  • Safety 

  • Convenience

Competitive assessment

• What are competitors doing?

• TIming of market entry and profitability:

  • 1st drug to market makes the most revenue

  • 2nd drug makes less overall revenue but often still profitable due to lower development costs

  • 3rd drug may only achieve limited profitability

  • 4th and beyond have difficult recovering costs (R&D)

• What are the advantages/disadvantages?

  • Safety and efficacy for existing therapies

  • Cost and accessibility

  • Convenience and patient adherence

Feasibility Study (Biochemistry)

• Focus on solvable problems

• Is the disease well understood?

  • Clear diagnosis

  • Can it realistically be treated with drugs?

• Is the mechanism of disease understood?

  • Black box diseases are difficult to target

• Are biochemical tools available?

  • Reliable assays to test drug activity

  • Suitable animal models

• Are there existing drugs?

  • Potential starting point for research

  • Proof of principle that the approach can work

  • Competition or patent barriers

Biochemical studies

• Proof of principle

  • Does the idea have a chance of working?

  • Is there experimental evidence supporting the concept?

• Drug testing methods

  • Reliable ways to test candidate drugs

  • Often multiple methods are developed (fast, accurate, biologically relevant)

• Development of animal models

  • Create or exaggerate disease state to study drug effect

Medicinal chemistry starts with a lead

• Show activity in the test being used

• Some specificity

  • Not active in unrelated tests

• Pattern of “drug-like” properties

  • Must be able to enter and function in the body

• Patent opportunity

• Chemistry must be feasible

  • Structure can be modified

  • Practical and efficient synthetic routes

Lead Identification

• Medical chemistry requires a starting point

  • Lead

• Methods used to identify leads

  • High throughput screening (HTS)

  • Natural products

  • Rational drug design

  • Combinatorial chemistry

  • De-novo design

High-Throughput Screening (HTS)

• Thousands of compounds are tested for activity

  • Typical libraries contain >500,000 compounds

• Companies maintain large compound collections, including:

  • Every compound that was ever made in the company

  • Compounds purchased from academic labs

  • Natural products

• Requires a biological assay that can be automated

• Modern robotic systems can screen >200,00 compounds per week

  • Compounds showing activity (hits) are tested more carefully’

Importance of testing a Variety of Compounds

• No way to predict which compounds will work in advance

• Drugs and biological molecules interact in many unpredictable ways

• Testing only small variations of the same compound limits discovery

• A large variety of chemical structures and types increases the chance of finding hits

HTS Testing

• All compounds are tested at the same dose

  • Typical concentration: ~30 nanometers

• Compounds tested one at a time

  • Clear individual results

• Biological assay provides a simple yes/no answer

• Counter-screens help remove false positives

  • Assay with related biological functions

  • Desired compounds are specific (active only in the main assay)

• Good HTS generates ~500 hits

• Hit = compounds that tests positive in HTS

Hit- to Lead Removes False positives

• Most initial hits are false (>99%)

  • Impurities

  • Decomposition products

  • Compound reactivity

    • Detergent-like behaviour

    • Redox reactions

    • Strong electrophiles or nucleophiles

  • Interference with assay

    • Hydrophobic compounds(stick to protein or plastic)

    • Assay artifacts (colour, fluorescence)

    • Poor solubility

• Goal: eliminate false hits quickly

Pan-Assay Interference Compounds (PAINS)

• Promiscuous bioactive compounds

  • Show activity in almost any biological test

• Certain chemical structures react nonspecifically with biological targets ( some interfere with test methods)

  • Redox activity

  • Detergent-like properties

  • Strong acid/base behaviour

  • Strong nucleophile/electrophole interactions

  • Photoreactivity

  • Chelating properties

  • High lipophilicity

Re-testing hits

• Re-test compounds using purified samples

• Test at multiple concentrations

  • Compounds act in concentration-dependent manner

  • Dose response curve

Test in other Assays

• Test in available assays

  • Previously collected data for the compounds is re-examined

• Reject promiscuous compounds (PAINS)

Structure confirmation

• Spectroscopy and Independent Synthesis

• Several compounds in collections are very old

  • Decomposition (during storage)

  • Incorrectly identified

• Synthesize, purify, confirm structure and re-evaluate

• Test a series of compounds with related structure

  • Compounds in the series should all be active, but show difference that can be related to structural patterns

Hit-to-Lead Required 3 to 6 Months

Drug developed by HTS

• Nevirapine

• Gleevec

• Quinolones

  • All contain Aromatic rings

Natural products as Leads

• Chemicals isolated from plants, animals, microbes

Secondary metabolites

• Chemical not directly required for life

• Produced by the organism for a secondary purpose

  • Poisons for defense or against predation

  • Colors, fragrances

  • Hormones, pheromones

  • Modify the organism's environment

Searching for natural products

• ollect a large amount of a source organism

  • Plant, animal, microorganisms

• Isolate and purify substance showing biological activity

• Determine the chemical structure

• Confirm biological activity

• May use information about the source organism to narrow search

  • Poisonous organism

  • Local knowledge

Natural product leads

• Isolate from living things

  • May require a large amount of raw material (tons)

• Often difficult to perform SAR

  • Complex structures

  • Limited to chemical modification of the natural product

• Academic labs only

  • Companies closed NP departments in 1980s

Natural product structures

• Penicillin

• Taxol

• Digoxin

Drugs produced in Industrial quantities

• More than 20,000 tons of penicillin per year

• Must be farmable or produced by industrially compatible method

Spongistain

• 400kg of marine sponge provided only 13.8g of spongistain

• Found to be one of the most potent tumor-growth inhibitors ever discovered

• Attempt to re-isolate the compound unsuccessful

  • 13 tons of wet sponge gave just 35 mg

• The sponge was extinct by the 3rd attempt

Rational drug design

• Design a lead using a known chemical structure

  • Enzyme substrate

  • Natural inhibitor

  • Ligand for a biological receptor

  • Existing drug

• Knowledge of the mechanism of action

• Design for improved properties

Captopril Resembles snake venom

Acycolvir resembles guanine

Rationally Designed drugs

• Captopril

• Acyclovir

• Indinavir

Combinatorial Chemistry

• Semi-random synthesis of a large library of related compounds

  • Robotic synthesis

• Text mixtures (academic)

• Test impure single product (industry)

• Based on the idea of Darwinian selection

Combinatorial search

• Make a large mixture (library of molecules)

• Structures have a common “core) and random “appendages)

  • Made using similar chemical reactions

• Test the compounds as a mixture

• “Deconvoulte” the result from the mixture to identify molecules that show desired properties

One drug discovered this way

• Sofrafenib

  • Kidney and liver cancer (rare)

  • Combinatorial chemistry used to identify an alternative core “lead” after hit-to-lead

De-novo design

• Use a computer to design a lead from scratch

• Required the 3D structure of biological target

  • X-ray crystal structure

  • NMR structure

De-novo Drug Design

• Start with a protein structure

• Build a computer model of the binding site

• Design a molecule that fits into this pocket

Lead optimization

• nce a lead is available, optimization begins

  • Focus on medical chemistry

• Structure Activity Relationships (SAR)

  • Structure properties Relationships (SPR)

• Make small changes to a drug's structure and observe changes in activity

  • Identify parts of the molecule that interact with biological target

  • Optimize as many properties as possible

Molecules made one at a time (SAR)

• Chemist designs new molecules based on lead structure

• Chemistry synthesizes the designed molecule (structure)

• Biochemist tests the synthesized molecule (activity)

• Chemistry analyzes the results to determine the relationship between structural modifications and changes in activity

• Chemist uses this information to design new compounds

SAR as a process

Structure Property Relationships (SPR)

• Optimize several properties simultaneously

  • Potency

    • Dose that produces biological activity

    • Aim for the lowest dose possible (low dose = high potency)

  • Selectivity

    • Ratio of dose that gives desired biological activity vs undesired activity

    • Aim for the highest possible dose for undesired biological activity (high dose = high selectivity)

  • Solubility (in water)

  • Lipophilicity (solubility in non-polar solvents)

  • Chemical stability (time and conditions for decomposition)

  • Acid-base behaviour

  • Susceptibility to metabolism (rate and type)

  • Toxicity (nature and dose)

  • Ease of synthesis (speed and efficiency)

SPR as a process

Focuses on molecules that can enter the body

• One of the hardest problem in medical chemistry is getting drugs into the body

• Avoid wasting effort on compounds that can never become drugs

  • Focus on “drug-like” molecules

• Certain properties correlate well with drug performance, optimize these simultaneously with potency

• Requires extensive computational support

Drug Candidate

• End result of the discovery phase

  • Chemical properties are known

  • Basic biological profile is known

• Process required 1 to 3 years

  • Approx 2000 to 5000 chemical compounds are made and tested to find each candidate

• Companies often identify 3 or 4 backup compounds

  • If the primary compounds, minimize recovery time