Drugs Flashcards
explain the drug discovery process
(takes 10-15 years)
1. identify POI which is related to a medical need i.e. emergence of certain disease, antibiotic resistance etc.
2. understand POI’s relevant mechanism (deduce from structure, information of active site by X-ray/ NMR/ cryoEM)
3. The information will be used in the selected screening method to identify “hits” or compounds that result in inhibition of the POI activity
4. Select a lead compound from the “hits” (most effective) which can be used in the drug development stage (optimization using medicinal chemistry, analog synthesis, animal testing)
- 2 possible outcomes:
- If the lead compounds end up ineffective/ toxic to animals, it will be reassessed (repeat optimization/ select for another lead compound from “hits”/ repeat screening for new “hits”)
- If it is successful at animal testing stage, drug candidates can then be established for further clinical testing in humans
what are the 3 possible screening methods
- High Throughput Screening (HTS)
- Virtual Screening (VS)
- Combination of HTS and VS
explain High Throughput Screening (HTS)
- Uses a big library of compounds, prepared in an arrayed format, to perform assays w/ POI
- The POI activity w/ the compounds (eg inhibition of POI) can be measured by
fluorescent or luminescent detection, colourimetry, or light scatter - Mostly used by big pharmaceutical companies
what are the disadvantages of HTS
- Expensive (high quality assays & specialized equipment)
- False positives
- Assay variability or errors in data
- Small amount of compound screened (largest = 10^7) relative to possible chemical diversity space (10^60)
explain Virtual Screening/ In silico drug screening
- automatic evaluation of large libraries of compounds using computer programs (selects and scores compounds based on input parameters)
- similar to HTS = relies on assessing a large, diverse library of compounds
-difference= compounds contained in databases rather than physically assembled into arrays. - Used in Computer-Aided Drug Discovery (CADD)
what is CADD
- Entails many different computational methods like virtual screening, virtual library design, lead optimization, de novo design
- classified into ligand-based methods (LBDD) and structure-based (SBDD)
how is the method of CADD chosen
- based on availability of structural data of protein and ligands
what methods are used for both unknown POI & ligand information
- Library design: (when no info about both POI & ligand)
- ligand structure = identify a collection of compounds that can inhibit/ bind the POI by HTS, VS,
- POI structure - use Alphafold to predict 3D structure
- De Novo design: (when have info about POI, not ligand)
-Develop ligands based on small fragments that will occupy active site & inhibit protein
outline LBDD
- 3D structure of POI = not known
- Use information about the molecules that bind to the target of interest (S/Inhibitor)
- Hits are identified, filtered, and optimized to obtain potential drug candidates that will be experimentally tested in vitro.
outline SBDD
- know 3D information of POI
-can classify positives from false positives
describe the prominent steps in SBDD and LBDD approaches
- For SBDD (to obtain known POI structure) = Homology modelling & validation
* The target-template alignment leads to the modelling of 3D structure of POI - model is validated by Ramachandran plot (using PROCHECK) - narrow down database of ligands by virtual screening to obtain set of “candidate ligands” eg. several methods
* SBDD: molecular docking, pharmacophore design
* LBDD: pharmacophore design, QSAR - Docking process by virtual screening: ligands and POI are allowed to interact w/ each other (using docking software)
* SBDD = many ligands are screened against the known POI
* LBDD = the chemical properties of single ligand is used to screen against many POI targets
what are the libraries used in virtual screening
Combinatorial library design = library of ligands constructed based on known structure of POI and compounds that mimic its substrate/ inhibitor = higher success rate because scaffolds can already bind POI
Non-combinatorial library = bought database (files) of compounds
describe LBDD (when its used, why is it important)
- no POI structural information.
- Exploit knowledge of ligand structure which binds POI
- Key to understand pharmacophores of the ligand structure.
- Pharmacophores have to be kept the same, but all else is modifiable = unlikely
to negatively affect binding - LBDD methods can be divided into:
-Similarity searching = based on 2D descriptors/ 3D descriptors
-Pharmacophore mapping
-Machine learning
describe similarity searching (LBDD)
- looks for molecules with similar properties/structures to the reference molecules.
- involves screening the compound of interest against a large database, then ranking the matches in order of high to low similarity to the reference.
-The top-ranking compounds are selected for biological testing based on several quantitative measuring methods
what is the measure of similarity based on (LBDD)
- Physicochemical properties: eg. MW, MR, logP (partition coefficient: shows solubility of substrate in an aq environment or membrane)
- 2D properties: eg. fingerprints, topological indices, maximum common substructures
* need quantitative basis to measure and rank compounds according to similar structures to known ligand
- Similarity coefficient: A quantitative measure of similarity between 2 molecules (parent compound and a similar compound in the library) - 3D properties: eg fingerprints, molecular fields
explain 2D fingerprinting (2D properties) (LBDD)
- use a binary vector model (bit-string)
- Each bit in the bit string represents 1 molecular fragment
– can design the bit string according to preference by selecting certain functional groups as a bit - The bit string for a molecule records the presence 1 or absence 0 of each fragment in the molecule
- each molecule in library will be assigned a unique bit string
- Similarity is based on determining the number of bits that are common between structures of query and compounds in the library
- More (1) bits in common = greater similarity
- Will only get info. on what functional groups are present
explain hashed fingerprinting (2D properties) (LBDD)
- uses bit string, but also contains information on the structure (topology) – how the atoms are connected to each other in 2D space, but still no info. about stereochemistry (3D structure)
- Each fragment is processed using many different hashing functions, each of which sets a single bit in the fingerprint
- When assigning the bit string, each bit also considers the order of functional groups – only when the functional groups are in same order, will consider as present (1)
what is the problem with hashed fingerprinting
Bit collision = the same bit can equate to multiple patterns (generate false positive)
* A few bit collisions in the fingerprint are ok, but too many may result in losing information in the fingerprint
what happens once the 2D/hashed fingerprints are defined (LBDD)
generate: Similarity (Tanimoto/ Jaccard) coefficient
* define % similarity according to the bit strings for each compound in the library
T(a, b) = Nc / (Na + Nb - Nc)
a = bits in reference structure
b= bits in database structure
c = bits in common in ref and database
- can run search against the library to filter for certain compounds that meet a certain % similarity criteria: If 2D = similar, mostly just assume that will have similar 3D structure
explain scaffold hopping (3D properties) (LBDD)
- purpose is to find a molecule with the desired pharmacophore (same activity/ function) as the query compound but w/ different core structure
- Considers if the molecules occupy same 3D volume, space and binding site as the substrate
what is the principle of 3D fingerprinting (3D properties) (LBDD)
- to break down different functional groups in a compound of interest to reconstruct a model for potential drugs
*like in 2D, once 3D library search is done, must score & rank obtained hits to narrow down the compounds selected for further experiments in the wet lab
explain the 3D fingerprinting process (LBDD)
- Define point of functionalities in the molecule (defined by same function does not have to be the exact same functional group)
- Connect them into triplets and obtain associated distance - can now deduce how they connect in 3D space (more point of functionalities = more triplets detected = more accuracy)
- Develop bit string for diff. triplets
- Use the defined bit string to search for compound in database that satisfy the
following properties:
- pairs of atoms at given distance range
- triplets of atoms and associated distance
- pharmacophore pairs and triplets (donors, acceptors, aromatic centers etc.)
- valence angles
- torsion angles
explain the flexibility of parameters in 3D fingerprinting
- can be strict or flexible w/ parameters
- hits can have completely diff. sequence/ structure but must satisfy functional properties
explain 3D shape search (LBDD)
- Molecules are aligned in 3D
- Similarity score is based on common volume occupied
- Method = generate spherical volume of known ligand, then find molecules with volumes that are sufficiently similar to the spherical volume
-Molecules can be dissimilar in 2D but similar in 3D or vice-versa
-Due to diff. rotations around bonds & etc., must decide how the screening is conducted
what is the principle of pharmacophore mapping (LBDD)
Pharmacophore generation & searching
- to determine pharmacophores of a molecule & search for them in library
(molecular features that are vital for the biological activity of a compound) - Protein structure is not required
- Assumes that all (majority) of the known actives bind to the same location
explain pharmacophore generation (LBDD)
- Identify pharmacophoric features and how they are organized in 3D space
explain pharmacophore searching (LBDD)
- Given a pharmacophore, find all molecules in a database that can match it in a low- energy conformation
- Scaffold-hopping possible
- Doesn’t require structural similarity
- Just needs to match the pharmacophore
- Can also do 4-point model = better feeling for dimensionality of compound but too computationally demanding
in a database, you must: (LBDD)
- Ensure diversity of ligands (tautomers/protonations/ conformational space
should be accounted for) - Must define molecules in the database using same protocol and parameters used to generate pharmacophores
- Generate cutoffs/ allowed tolerance (what Angstrom range to give such that a match is
indicated?) - more or less strict - have actives and decoys
-but not ideal for scaffold hopping because the system looks for dissimilarity in core structure
what are actives
a compound known to bind the POI
- A good similarity measure will cluster the known actives at the top of the ranking
- if you pick up actives = ensure accuracy of searching criteria
- if you don’t pick up actives = defined parameters/ criteria= not optimal
what are decoys
compounds known to not bind or inhibit POI
- if decoys are shown as similar to the reference molecule = design algorithm is erroneous, not optimal
explain quantitative-structure activity relationship (QSAR) (LBDD)
machine learning
* tries to establish quantitative relationships between descriptors and the target property – so can predict activities of compounds from 2D properties
* eg. inhibition property, toxicity – can eliminate certain compounds known that will not pass animal testing, etc. or eliminate false positives
* QSAR models require descriptors that accurately convey chemically-relevant information to the machine learning models
what are the 3 main methods of SBDD
- Structure and known inhibitor design
* Known inhibitor or co-factor is modified to improve binding affinity or selectivity - Virtual High Throughput Screening (vHTS)
* Docking of small molecules into the crystal structure which are scored and ranked - De novo design
* A molecule is designed from scratch to bind in the active site by docking and connecting fragments to create full molecules. These molecules are then scored and ranked.
explain structure and known inhibitor design (SBDD)
- Need high-resolution X-ray structure of target protein.
- Solved structure should be in the presence of the substrate -residue orientations in active site are known
- Use this structure as a template for screening of small molecules
- Can redesign the molecules:
-if modification leads to increased activity = groups modified allow molecules to fit and/or bind better
-if modification leads to decreased activity = groups modified are biologically important for binding.
-if modification leads to no change = group modified are not important, & can thus be variable
explain the principle of Virtual High Throughput Screening (vHTS) (SBDD)
- dock small molecules with the crystal structure to identify leads- then scored and ranked
- lead = toxic = can modify to avoid toxic side effects
- Must test lead activity using assay (enzymatic and pharmacokinetics)
- Ideally should also solve crystal structure of ligand to verify binding mode
what are the challenges of vHTS
The energetics of protein-ligand interactions are complicated:
* Must also consider involvement of solvation of binding site (water molecules mediating H bond between ligand and substrate)
* Both proteins and ligands can be quite flexible
* Many target-binding ligands are not good drug candidates
* The structures of many important drug targets are difficult to determine = helped by Alpha fold
what is the method of vHTS
- Perform docking to identify leads – which are scored and ranked
* need high resolution Xray of POI to know residues involved in binding -so when docking, can fit the ligand by selecting proper stereochemistry - Once identified leads, can perform Pose prediction: to identify which chemical groups of the ligands are important for binding and specificity
Once identify database of leads (Substrate/inhibitor), describe by molecular descriptors:
* 1D-: chemical composition & physicochemical properties (eg. MW,hydrophobicity)
* 2D-: chemical topology: how functional groups are associated with each other (eg. Connectivity indices, degree of branching, degree of flexibility)
* 3D-: the spatial arrangement of chemical groups (eg. shape, volume, functionality, surface area)
- Can also identify activities associated with certain structures of leads using Structure Activity Relationships (SAR)
* Correlations that are constructed between the features of chemical structure in a set of candidate compounds and parameters of biological activity, such as potency, selectivity and toxicity
* Used to: Identify groups of the lead compound that are important to biological activity
* X-ray crystallography can also be used to identify important interactions between drug and protein
what is the Lipinski’s rule of 5s for good drug candidiates (SBDD)
- MW < 500
- Fewer than 5 H-bond donating functions
- Fewer than 10 H-bond accepting functions
- Calculated logP between –1 and +5
* Calculated from partition of compound in artificial membrane (made by mixing n-
octanol = mimicking lipid bilayer; with water = mimic aq environment)
* Describes solubility
* High log P = hydrophobic; negative = very water soluble
* P = [D] lipid / [D] water = log P
Explain rigid docking (SBDD)
- The ligand is treated as a rigid structure and only the translational and rotational degrees of freedom (rotation of whole molecules in 3D) are considered – no rotation between bonds of functional groups etc.
- If have diff ligand conformations (known to have rotatable bonds), each conformation is docked separately (pre-rotate bonds before docking)
what are the pros/cons of rigid docking
- Advantage: fast & not computationally demanding
- Dis adv: can’t rotate any bonds although rotations are known to prevent steric clash –
lose many possible hits
what is the process of rigid docking
- Define enzyme active site in geometrical shapes (triangles/ spheres)
- Match & rank ligand that can satisfy the defined shape
* Can also define functionalities within the active site – if active site = known to have H bond donor, set one of the ligand ranking criteria to be having a H bond acceptor at a certain position
explain flexible docking
- Allows some degree of ligand flexibility along torsion angles during docking process + small conf. changes at binding site can be accommodated
what are the pros/cons of flexible docking
- Dis Adv: more comp demanding
- Adv: allow for molecules to adopt optimal binding pose
Why is scoring necessary (for ranking) (SBDD)
- Many different poses of the same ligand need to be ranked based on their affinity with receptor to identify positive hits from other poses
explain first principle scoring (SBDD)
-Rank ligand based on force fields generated with molecular mechanic information, can be:
* Intra molecular = bond lengths, angles, dihedrals,
* Inter molecular VDW contacts (non-polar), Electrostatic interactions (polar)
E bind = E intra + E nonpolar + E polar
explain empirical scoring (SBDD)
- Based on Gibbs free energy of binding - better since also account for entropy/enthalpy in binding
- Also include a penalty function to account for unfavorable interactions
- Values are empirically determined from experiments
ΔGbind = ΔG0 + (ΔG polar x Σ f-complex + (ΔG nonpolar x Σ f-complex) + ΔG rot x non-rotatable bonds
what are some considerations before docking (SBDD)
- Water in structure
- Tautomeric forms (eg. keto-enol forms can appear to have the same electron densities
but are diff. in terms of being H bond acceptor/donors, so may interact differently) - Weak electron density of sidechain of aa = can result in wrong assignment
- pKa and Protonation state (if protons are not resolved in the structure, will affect H
bond interpretation required in docking programs) – a problem with charged residue since easily lose H, but if know crystallisation buffer & pH of environment, can deduce protonation state - Complications from rotatable bonds due to flexible torsion angles
- Ring conformations may not be distinguishable (chair or boat)
explain de novo design (SBDD)
- When No info. known about ligand
- Active site is treated as an empty pocket = molecules are designed from scratch by searching small fragments that form favourable interactions with the active site
what is the process in de novo
- define binding pocket & interaction sites (aa residues involved in binding)
- search for fragments that can satisfy the active site 3D space and volume OR can use lattice strategy, in which binding pocket is defined as a 2D lattice & search fragments that satisfy defined lattice
what are some parameters involved in de novo screening
- Search of possible poses & conformations (search algorithm) = Orientation of molecule in binding site
- Predicting energetics of protein-ligand binding (scoring function) = Binding affinity, and thermodynamics
what are the 3 main methods to form a complete lead molecule in De Novo design
Docking:
* Dock larger molecular entities that you already know from experiments, have favourable
interactions in the active site
Building (lead identification by fragment evolution)
* start with 1 fragment known to make favorable contacts with an interaction site
* then build from that fragment to form a complete molecule
Linking:
-lead identification by fragment linking
-lead identification by fragment self-assembly (using enzymes)
explain Linking (lead identification by fragment linking)
- separately place fragments (small functional groups) identified to make favorable
contacts with each interaction site and join them to form a complete molecule - fragments are joined by a linking group/ core template
explain linking (Lead identification by fragment self-assembly)
- Only works with enzymes – the protein target is an enzyme, which is used to perform the linking resulting in its own inhibition
- Advantageous as both components may be too large to accommodate active site, but are able to bind effectively individually
optimization (SBDD)
- Once complete molecule has been formed:
- Optimize or modify properties of the lead compound
- Re-engineered to address optimization of a particular property (eg. selectivity, cell- based activity, oral activity or efficacy)
what must you ensure throughout the whole method (SBDD)
For each method & after every step, must assess the compounds’ properties after growing/ linking the fragment if more/ less desirable
Once hits have been identified from the screening, they are validated by re-testing them and checking the purity and structure of the compounds.
what criteria must the leads fulfill (SBDD)
- Potency = the amount of drug required for its specific effect
- Efficacy= the maximum strength of the effect itself
- Pharmacokinetics = rate of adsorption, distribution, metabolism, and excretion (ADME).
- Pharmacodynamics= determining the biochemical and physiological effects of drugs, the mechanism of drug action, and the relationship between drug concentration and effect.
- Chemical optimization= binding affinity and favorable accommodation in active site
- Patentability
describe a case study of SBDD
Thymidylate synthase
- generates dTMP from dUMP using 5,10 Methylene tetrahydrofolate
- dTMP = critical for DNA replication and repair, so the enzyme has been of interest as a
target for cancer chemotherapeutic agents. - A lead was identified by docking
- favorable binding is verified by Xray crystallography
- Original lead identified = hydrophobic, so performed in silico screening to obtain compound with more solubility & in silico synthesis by adding more H bonding group
-but results in diff. interaction than expected, thus decreased binding affinity - further analysis and redesign: add amide group & result in favorable binding + increased binding affinity
- must verify in silico screening by structural analysis eg. Xray crystallography, NMR, CryoEM or functional assay
describe HIV-1 protease inhibitor example
- Squanivir (Roche) & Indinavir (Merck) designed by modeling chemical structures on the computer to fit inside of the active site of HIV-1 protease using the X-ray crystal structure
describe the neuramidase drugs
- Relenza and Tamiflu designed from known structure of neuramidase pocket and its substrate – sialic acid
what is drug metabolism defined as
metabolic breakdown of drugs by living organisms, usually through specialized enzymatic systems
what happens to a drug after ingestion
- drug compounds can enter the body’s biosynthetic pathways of endogenous substrates (eg. hormones, cholesterol, bile acids) to be metabolized. This is possible because drugs resemble the natural compounds.
- These chemical alterations are known as “biotransformation”, which occur primarily in the liver and sometimes, in other tissues (depending on the type of drug compound)
what are the results of drug metabolism
-Most metabolic products are inactive because they have been broken down into non-toxic compounds which are then excreted from the body. However, there are some exceptions:
* Inactive drug compound is metabolized and become active eg. prodrugs
* Drug compound is metabolized and become more toxic/carcinogenic
describe drug conversion to lipophilic compounds
Drug metabolism is required to convert lipophilic compounds into more hydrophilic compounds to be excreted
* If the lipid soluble non-polar compounds are not metabolized, they will remain in the blood and tissues and maintain their pharmacological effects for much longer
describe the metabolic pathway of hydrophilic drugs
-enters the stomach
- enters the circulation system
-enters liver where it is metabolized.
-metabolites secreted out of the body by kidney in urine
(problem: can be secreted too quickly – can have techniques to prolong life of hydrophilic drugs)
describe the metabolic pathway of lipophilics drugs
If no metabolism: stomach - circulation system- not metabolized by liver -remain hydrophobic - can’t be secreted by kidney – remain in blood & tissue - can result in side effects from its prolonged activity
If some are metabolized: stomach - circulation system - some metabolized to hydrophilic compounds in the liver -hydrophilic metabolites secreted by kidney in the urine - some that are not metabolized will still be hydrophobic and remain in the bloodstream
- This partial metabolism results in attenuation of the prolonged effect of the drug
If all are metabolized: stomach - circulation system & travel to target tissues -completely metabolized by the liver - all are converted to hydrophilic molecules - secreted by the kidney via urine.
what is pharmacokinetics (PK)
the study of how an organism affects a drug or the study of the absorption, distribution, metabolism, and excretion (ADME) processes of a drug
explain ADME at different organs
- Absorption: orally administered drug dissolves in GI tract and is absorbed by the
stomach and small intestine - Metabolism: liver
- Distribution: circulatory system (bloodstream)
- Elimination: kidney (urine), colon (feces)
why is pharmacokinetics important
- provide understanding about the physical and chemical properties of a drug which is important in drug development because they will determine the drug’s success to reach its target.
- affect’:
-Chemical stability eg. is it stable in the stomach?
-Metabolic stability ex. half-life: how well/ how fast a drug is metabolized?
-Successful Absorption eg. ability to cross membranes - helps establish optimal dosage
explain optimal dosage
Optimal dosage = drug concentration in plasma in therapeutic window to elicit desired effects of the drug
- below = no effect on target – because metabolized too quickly/ doesn’t absorb efficiently
- exceed = toxic overdose – because metabolic system can’t cope with conc. of the ingested drug, resulting in prolonged effect & will already generate damage to the body by the time it is detoxified
explain how drug metabolism occurs in 3 phases
Phase 0: where absorption of the drug compound occurs eg. via passive diffusion through the membrane or through a transporter.
- The drug will then enter Phase I, where the drug gets metabolized/detoxified. Usually phase I = sufficient to make the compounds inactive, so it can be secreted out via urine.
- if phase I could not completely inactivate the drug, it will enter Phase II, where inactive compounds will be secreted as faeces.
- Some compounds would enter Phase III when there is a resistance against the drug. This is when different transporters pump the drug out of the target cell without undergoing metabolism.
explain phase I transformations of drug metabolism
involve introduction or unmasking of a functional group ex. (OH, -SH, -NH2, -COOH, etc.)
* These metabolites are often inactive & can be excreted readily via urine
- oxidation
- reduction
- hydrolysis
