COMPOUND PREPARATION AND HIT-TO-LEAD GENERATION

HIT GENERATION

Involves testing the compound library (or a part of it) to find out molecules that act on the desired target (i.e. the hits)

HIT GENERATION

goal

Enforce attrition to immediately discard anything that has no chance of being active

From millions or thousands, hit generation should end with

only a several hundred compounds

Biochemical researchers must select the best screening tests

If one selects “poor” hits with bad properties from the start, no amount of lead optimization can save them (just like spoiled food or 500-year-old, unpreserved books)

HIT GENERATION

Involves testing the compound library (or a part of it) to find out molecules that act on the desired target (i.e. the hits)

• Categorize molecules based on their physicochemical properties

• The selection of these physicochemical properties depends on your target protein.

Hits should also be validated to prevent

accepting “nonsense” or nuisance molecules

Nonsense or nuisance molecules are

false positive results

As early as hit generation, toxic or poor hits are eliminated using

molecular docking and molecular dynamic

BUT SOME HITS ARE NOT REALLY HITS

• Some “hits” don’t actually bind to the desired target

  • They just trick you to thinking they work in weird ways like binding to other proteins, precipitation, aggregation, etc.

BUT SOME HITS ARE NOT REALLY HITS

these are 

nuisance molecules and should be eliminated

Hits must be validated, otherwise they are

merely  called actives

FALSE POSITIVE

• Actives that look like they’re working but are not

Common types of false positives:

• Promiscuous binders

• Pan-assay Interference Compounds (PAINs)

Promiscuous binders

• Actives that actually bind the desired target, but target many other things too

• Usually lead to unwanted toxicity or side effects

Pan-assay Interference Compounds (PAINs)

• Actives that don’t bind to the desired target at ALL

• Just gives positive results to any assay thrown at it

• Gives a positive result but that positive result is not attributed to the activity of the actives or hits

 HIT VALIDATION

• HIT disqualification 

•  use of counterscreens which serve as a sort of trap (a real hit must NOT pass the counterscreen)

HIT disqualification 

• If a test compound gives a positive result for both the screen and counterscreen,

it is “disqualified” from being a hit

Hit disqualification

problem

 amount of possible nuisance mechanisms far exceeds known counterscreens 

If hit disqualification does not give sufficient confidence, the next resort is

use of Biophysical techniques

HIT VALIDATION

goal 

Eliminate false positives as early and efficiently as possible to ensure that only valid, reliable hits advance in the drug discovery process

Counterscreens

are assays that are used to disqualify those molecules that will give false positives.

To be a hit is simple

– it only needs confirmed activity

LEAD GENERATION

• A lead is must have confirmed activity (hit requirement); PLUS

• Show evidence of desired selectivity

• Have activity in cellular systems

• Stability in biologic systems

• Free from toxicity alerts (all these to be expounded within the course)

LEAD GENERATION

• The structure and in vitro activity of hits are confirmed, and their risks are

• characterized sufficiently to warrant investment in optimization

LEAD GENERATION

• Yet another round of filtering using in vitro or in vivo experiments

  • Features sought for in leads (previous slide) are given emphases

  • From hundreds of hits, around a handful to a few dozen leads

Even if a true hit is potent, it may be toxic or have poor ADME =

still not worthy to be a marketed drug

LEAD GENERATION

• goal

• Discharge risk as much and as early as possible

Discharging risk ASAP assures that

millions of dollars won’t be wasted developing a drug that will fail in clinical trials or be withdrawn from the market — a ‘fail fast, fail cheap’ strategy

CHOOSING STARTING POINT

• we will make use of existing molecule and by doing it we will now be ready to do novel molecule 

OPTION A: GET EXISTING DRUG

• Researchers can opt to use existing compounds in the human body

• Many companies use established drugs from their competitors as lead compounds 

Use of existing molecules leads to

me-too drugs 

Although often mocked as ‘me too’ drugs, they can

often offer improvements over the original drug (‘me better’ drugs) as their selling point

OPTION B: NEW DRUG/INDICATION

• Sometimes, indications other than the known can be explored

OPTION B: NEW DRUG/INDICATION

• drug repurposing or repositioning

• selective optimization of side activities (SOSA) 

Drug Repurposing/  Repositioning

• Same drug, different activity

Drug Repurposing/  Repositioning

example 

Sildenafil (from HTN to erectile dysfunction)

Selective Optimization of Side Activities (SOSA)

• Same drug, but converted to an analogue, different activity

• Method where scientists find new uses for existing drugs by improving their side effects or secondary actions

Selective Optimization of Side Activities (SOSA)

example 

Iproniazid (antidepressant) to isoniazid (anti-TB)

Many of the hits obtained from synthesis don’t have a

‘drug-like’ structure and it may require far more effort to optimize them

• Thus, much time and money is saved when existing structures are used

COMPOUND PREPATION

• natural products

• synthetic molecules

OPTION A: NATURAL PRODUCTS

• Often follows forward approach to discovery

• Skips the need for getting many reagents and reacting them by organic synthesis

• very complex compounds, and thus, original for most of the time

OPTION A: NATURAL PRODUCTS

• Often follows forward approach to discovery

• Screens are done to find actives, then followed by determination of mechanism of action and drug targets 

OPTION A: NATURAL PRODUCTS

Much more challenging than synthetic route in terms of 

compound isolation

OPTION B: SYNTHETHIC MOLECULES

• Employed in

• both forward and reverse approaches of discovery

Some synthetic compound libraries can be

tested immediately (forward method),

libraries can be designed to obey _

predictive data that may come from previous CADD data (reverse method)

OPTION B: SYNTHETHIC MOLECULES

• Easier isolation, and skips the horrors of endless extraction and chromatography 

• Can deliver up to thousands or millions of compounds

OPTION B: SYNTHETHIC MOLECULES

provides 

•  limited access to chemical space and requires more reaction steps (more quantity at the expense of small diversity)

  • Would require you to strictly follow guidelines or processes

“Tuklas Lunas” is a program aiming to discover drugs

• Once they discover the drug compound from out of the national products. They are now checking if this product is possible for drug development.”

Combinatorial synthesis

• Produce mixtures of different compounds within each reaction vessel

• More random, and is more applicable for hit-to-lead generation

Parallel synthesis

• Produce a single product in each vessel

• More focused and organized, and is more applicable for lead optimization

Parallel and combinatorial methods are used not just for small molecule synthesis, but also

peptide synthesis

Parallel and combinatorial methods generally involve

• the use of solid phase techniques – that is, carrying out syntheses on a solid beads rather than in solution

• The essential requirements for solid phase synthesis are:

• Polymeric Support

• Anchor/Linker

• Bond Between Substrate and Linker

• Cleavage Method

Polymeric Support

A cross-linked insoluble polymeric support (e.g., a resin bead) that is inert to the synthetic conditions.

Anchor/Linker

A molecule covalently linked to the resin; it contains a reactive functional group that allows attachment of the substrate.

Bond Between Substrate and Linker

A covalent bond that connects the substrate to the linker and remains stable under the reaction conditions used in the synthesis

Cleavage Method

A chemical or physical means to cleave (release) the final product or intermediates from the linker.

The techniques of solid phase synthesis have been used to

produce large quantities of compounds from a particular reaction sequence

Swelling

- its purpose is to double check if a certain starting material or reagents will attach to the cavity of the resin bead 

Magnification of the resin bead leads to  

the string-like structure

Cavities

= ‘to dock tadpole structure’

Best to synthesize __ than tadpole-like molecules

spider-like molecules

Chances of success are greater if the “arms” are 

evenly spread to allow more thorough exploration of chemical space

Tadpole-like molecules

only explore chemical space on a limited area, and wastes the chance of seeing what can happen if other regions in space are filled

COMBINATORIAL SYNTHESIS

• Mixtures of compounds are produced in each vessel, allowing production of up to millions of novel structures in a span of time that only a few compounds are usually made with conventional labwork

COMBINATORIAL SYNTHESIS

• Each reaction vessel is n

ot purified, and is tested for biological activity as a whole

COMBINATORIAL SYNTHESIS

• There is an

• economy of effort, as a negative result for a mixture of compounds saves the effort of synthesizing, purifying, and identifying each component there

PARALLEL SYNTHESIS

• Carried out in a series of wells such that each well contains a single product

• Reagents are added one by one without jumbling everything

• Lesser products made, but purer in quality

• Often used for lead optimization rather than generation

PARALLEL SYNTHESIS

• Starting material (SM); sort of like a reagent but in the form of a solid phase. It is linked to the bead. SM can attach multiple functional groups and those functional group attachments can have their own functional group attachments

• The means of cleaving in the product from the intermediate from the SM which will create a new compound—complex compound. 

• The three compounds will be subjected to the binding site to check

• Then the recipe is checked, yung bilog—where SM attaches; they try to check it and in order for the recipe to come up with more compounds so we utilize SM/reagents/solvents → swelling and formation of a new compound.

Small footprint work stations often enable one to

perform up to 24 reactions followed by 24 simultaneous workups on a heater stirrer unit

Some stations have miniaturized everything,

from the reaction portion to the separatory procedures

BEFORE THE ACTUAL SCREENING

• Effective management of compound libraries is a key element and is usually handled by a dedicated logistics group

• Most advanced systems use fully automated robotics for handling compounds

• Effective management of compound libraries is a key element and is usually handled by a dedicated logistics group

• Must ensure compound accessibility, integrity, error-free handling, efficient use, and rapid response to requests

• Samples need to be organized so that the compound directory can be easily updated once more compounds arrive

THE ACTUAL SCREENING PROCESS

• OPTION A: IN VITRO/ HIGH-THROUGHPUT SCREENING (HTS)

• OPTION B: IN SILICO/ VIRTUAL SCREENING

OPTION A: IN VITRO/ HIGH-THROUGHPUT SCREENING (HTS)

• An assay has to be developed, allowing quantification of the interaction of molecules with the chosen target

Most in vitro screens in H2L are

colorimetric in nature, and so requires:

• well plates (containing purified protein targets or whole cells)

• UV-vis plate readers

Well plates may hold specific reactions equivalent to

what is done in many test tubes, then the results interpreted by the reader all at once

Since this churns out data on so many molecules at once, such screening is called

high-throughput screening (HTS)

Throughput is essential if speed is a critical factor in the project, especially for a large compound library:

Before 2000's

- 96 well plate, “high throughput” this time means testing thousands of compounds per WEEK

Throughput is essential if speed is a critical factor in the project, especially for a large compound library:

• Last 2 decades

- 384 or 1536-well plates can handle thousands of compounds per DAY, and is today’s definition of high throughput

Typically, large combinatorial libraries are screened (primary screen) and numerous bioactive compounds are

classified as primary hits

Counterscreens validate hits →

• decision to become leads

In recent years, biophysical methods like _have been greatly used

NMR, XRC, and surface plasmon resonance (SPR)

Mistakes of HTS can be checked by

NMR to ensure that the compounds concerned are binding in the correct binding site

OPTION B: IN SILICO/ VIRTUAL SCREENING

• Molecular recognition events are simulated and input as very large virtual compound libraries to be screened in silico

• Biological screening accounts for about 15% of the total expenditures of an industry (that’s a lot of money!)

  • Thus, VS offers economy, speed and flexibility to drive drug discovery projects

Examples of the Virtual Screening are:

Pharmacophore Activity, Molecular Docking, Similarity Searching, and these compounds are evaluated which will be subjected to in vivo test until such time the lead compounds are identified.

Molecular Docking

Used to test thousands of compounds against a known target protein, especially when the protein’s X-ray structure is available.

Reliability of Docking

When the X-ray structure of the protein is accurate and well-characterized, the docking results are more reliable and predictive of potential hits.

Molecular Dynamics (MD)

A computationally intensive method that cannot be applied to thousands of compounds; has to wait when leads are narrowed down during optimization

LIGAND-BASED DRUG DESIGN (LBDD) VIRTUAL DESIGN

• Pharmacophore models are useful especially when full elucidation of the drug target structure lags behind the results of screening experiments (which is a common case)

• Once the pharmacophore model is constructed, virtual screening can be done

• Pharmacophore modelling still finds use even when target structure is available, if faster or less demanding computation is needed (vs. molecular docking)

Pharmacophore models are useful especially when

full elucidation of the drug target structure lags behind the results of screening experiments (which is a common case)

Once the pharmacophore model is constructed,

virtual screening can be done

Pharmacophore modelling still finds use even when target structure is available,

if faster or less demanding computation is needed (vs. molecular docking)

Prospective

• If there is no information on activity yet (ex. there are no existing drugs for your target), CADD may be used to predict novel compounds that may work 

• They could predict and establish information on the mechanism of action of that particular drug, how the novel compound may work based on their testing, assessment or validation under the CADD.

Retrospective

• If there are already known drugs (ex. you want to make me-better drugs), their structure may be analyzed with CADD to trace reasons why they worked

OPTION A&B: THE COMMON PRACTICE

• HTS (High-Throughput Screening well-plate method): no or little information is provided about the mechanism of action, BUT it does give direct results (i.e. did it work on our cell samples or not?)

• VS (Virtual Screening): not as convincing as HTS, but provides useful information about the mechanism of action (i.e. why and how did it work at the molecular level?)

Mechanism: less detailed

Translability more confidence 

HTS (in vitro)

Mechanism: more detailed

Translability less confidence 

VS  (in silico)