The choice of antibody platform, appropriate in vivo and in vitro standards and the pharmacodynamics and pharmacokinetic models should have been first considered.
It utilizes natural, synthetic and /or semi synthetic sources for potential selection of candidate molecule.
Structure-Activity Relationships are performed to improve potency and selectivity of the candidate molecule.
Improvement of biochemical biophysical, physicochemical and druglike properties like expression, solubility, stability, aggregation and glycosylation.
Optimized lead compound through the stages necessary to allow it to be tested in animal and human trials.
Involvement of pharmacokinetic studies, toxicology testing, and formulation development.
Identifying Bioactive Molecules
Biologically active small molecules have often been discovered by testing a single compound at a time, but such an approach is obviously highly inefficient and cost intensive in terms of both reagents and personnel time. As a result, HTS technologies have been developed to screen large numbers of compounds simultaneously, typically through the miniaturization and automation of assay protocols. What constitutes high throughput will vary depending on technical considerations and the screener’s economic situation, but generally, the number of compounds involved may range from tens of thousands to several million. High-throughput screening assays can be divided into two main classes: “pure protein” and “cell-based”. Pure protein screens generally have optical assay readouts that monitor enzymatic or binding activity. For instance, fluorescence polarization or FRET techniques are commonly used to screen for compounds that affect binding between protein partners. In pure protein assays, every compound screened should have equal access to the target. However, the membrane permeability characteristics of any active molecules that are identified may subsequently pose a major concern if the target is intracellular.
When biologically active small molecules are identified through chemical screens, particularly forward genetic screens, a substantial amount of work is often required to identify the molecular target. Ignorance of the target does not preclude clinical or research use of the molecule; indeed, many clinical agents have been used effectively even when their targets were not known (e.g. aspirin, nitroglycerin, fumagillin, epoxomycin).
Hypothesis-driven Target Identification
For well-characterized cellular pathways, it is sometimes possible to deduce the target of a small molecule by comparing data from across the field. For instance, characteristic phenotypes induced by the small molecule may permit the assignment of the target to a previously identified complementation group. Subsequent hypothesis-driven testing of potential targets can then be undertaken.
With affinity labeling, molecules may be radioactively or chemically labeled; they may be synthetically modified with reactive groups to promote covalent attachment to the target or tagged with specific moieties to facilitate detection. Standard protein fractionation and detection techniques can then be used to identify proteins from crude extracts that are specifically labeled by the molecule. Examples of drugs whose cellular targets were determined by such methods include acetylcholine, the anti-angiogenic agent fumagillin and the antifungal lipopetide echinocandin . Affinity purification of putative targets from cellular extracts is accomplished with molecules immobilized on a solid support used either as a column matrix or as a bead slurry. Small molecules for which cellular targets were identified through affinity purification include the immunosuppressant FK506, the kinase inhibitor purvalanol and the anti-inflammatory compound SB 203580.
Additional studies involving affinity chromatography were instrumental in further characterizing the biochemistry of FK506 and purifying the target of the structurally related compound rapamycin. A natural technological development of affinity-based target identification involves probing proteome chips with small molecules.
Genomic Methods of Target Identification
A. Yeast three-hybrid system
The yeast two-hybrid system has been widely used to study protein–protein interactions and the approach was adapted to create a yeast three-hybrid system for identifying protein–ligand interactions. The method relies on the use of a hybrid “bait” ligand consisting of the query molecule linked to a known ligand. A protein fusion between the known ligand’s receptor and a DNA binding domain serves as the “hook”, while the “fish” is a protein fusion between a transactivation domain and the target protein. If the target protein interacts with the query molecule, the “fish” and “hook” will be brought together by the “bait” and transcription from a reporter gene is activated. Hence it should be possible to identify the target of the query molecule by cloning a library into the “fish” domain and screening or selecting for activity of the reporter.
A three-hybrid approach was recently used to identify targets of various kinase inhibitors. For each inhibitor, a hybrid ligand was synthesized by attachment to methotrexate and the hook protein was a LexA–DHFR fusion. A three-hybrid screen of a human cDNA library with a purvalanol B–methotrexate ligand identified several (but not all) known purvalanol targets and also several new candidate targets, and all identified targets were kinases. Affinity chromatography and enzymatic assays confirmed 12 of 16 novel candidate targets identified in the cDNA screens.
B. Induced haploinsufficiency
One genomic approach to target identification relies on the premise that altering the gene dosage of the target will affect sensitivity to the small molecule. For example, inactivating one copy of the target gene in a diploid organism would in many cases be expected to increase sensitivity. This method of identifying drug targets through induced haploinsufficiency was established by Giaever et al., who constructed and screened a set of 233 heterozygous yeast deletion mutants for those demonstrating increased sensitivity to known drugs. Each mutant was chromosomally tagged with a unique oligonucleotide and the mutants were pooled and grown in the presence of tunicamycin, at a level of drug that is sublethal for wild type. The relative number of each mutant in the pool was monitored at various time points by polymerase chain reaction (PCR) amplification and fluorescence labeling of all tags in the pool, followed by hybridization of the PCR-generated probes to an oligonucleotide microarray. The fluorescence intensity generated at each spot on the array permitted quantitation of the relative abundance of each corresponding heterozygote in the pool. Mutants unaffected by tunicamycin showed no reduction in signal over time, whereas tunicamycin-sensitive heterozygotes showed varying decreases in signal. In this study, one known and two new loci were identified and confirmed as involved in tunicamycin resistance.
C. Expression profiling
Expression profiling was recently employed in identifying the targets of a class of small-molecule antagonists of FK506 in yeast cells subjected to salt stress. These molecules (termed SFKs for suppressors of FK506) were identified in a chemical genetic screen for molecules that rescue yeast growth in the presence of high salt and FK506 . Expression profiling results obtained with SFK2-treated yeast suggested that the Ald6 p pathway was a target of SFK2 and haploinsufficency screening supported this hypothesis. Overexpression of ALD6 was found to suppress the effects of SFK2–SFK4 on growth and the ability of SFKs to inhibit Ald6 p in vitro was subsequently demonstrated. In addition to using DNA microarray technologies in conjunction with haploinsufficiency studies, it is likely that gene expression profiling will be used increasingly as a primary means of identifying the targets of bioactive small molecules. Many research groups have used expression profiling to identify characteristic patterns of gene expression (“fingerprints”) that are associated with certain disease states or biological pathways. If the patterns are sufficiently unique, then it is sometimes possible to assign previously uncharacterized mutants to specific cellular pathways or complementation groups based on their expression profiles. Analogously, by profiling cells grown with and without a small-molecule modulator, it may be possible to identify genes or pathways that are affected by the small molecule.
In analog design, molecular modifications of the lead compound can involve one or more of the following strategies:
- Identification of the active part (pharmacophore)
- Functional group optimization
- Structure activity relationship studies
- Bioisosteric replacement
- Design rigid analogs
- Homologation of alkyl chains or alteration of chain branching, design of aromatic ring position isomers, alteration of ring size, and substitution of an aromatic ring for a saturated one or the converse.
- Alteration of stereochemistry, or design of geometric isomers or stereoisomers.
- Design of fragments of the lead molecule that contain the pharmacophoric group (bond disconnection).
- Alteration of interatomic distances within the pharmacophoric group or in other parts of the molecule.
IDENTIFICATION OF THE ACTIVE PART (PHARMACOPHORE)
Any drug molecule consists of both, essential and nonessential parts. Essential part is important in governing pharmacodynamics (drug receptor interactions) property while non-essential part influences pharmacokinetic features. The relevant groups on a molecule that interact with a receptor are known as bioactive functional groups. They are responsible for the activity. The schematic representation of nature of such bioactive functional groups along with their interatomic distances is known as pharmacophore.
Once such pharmacophore is identified, structural modification can be done to improve pharmacokinetic properties of the drug. For example, the presence of a phenyl ring, asymmetric carbon, ethylene bridge and tertiary nitrogen are found to be minimum structural requirement for a narcotic analgesic to become active. Similarly the presence of two anionic sites and one cationic site must be present in cholinergic agent.
FUNCTIONAL GROUP OPTIMIZATION
The activity of a drug can be correlated to its structure in terms of the contribution of its functional group, to the lipophilicity, electronic and steric features of the drug skeleton. Hence by selecting proper functional group, one can govern the drug distribution pattern and can avoid the occurrence of side-effects.
STRUCTURE ACTIVITY RELATIONSHIPS
The physiological action of a molecule is a function of its chemical constitution. This observation is the basis of interpretation of activity in terms of the structural features of a drug molecule. Generalized conclusion is then can be made after examining a sufficient number of drug analogs.
The concept of bioisosterism is derived from Langmuir’s observation that certain physical properties of chemically different substances are strikingly similar. These similarities were rationalized on the basis that carbon monoxide and nitrogen both have 14 orbital electrons and similarly, diazomethane and ketone both have 22 orbital electrons.
Bioisosteres are substituents or groups that have chemical or physical similarities and which produce broadly similar biological properties. Bioisosteres have been classified as either classical or nonclassical. Classical bioisosteres are those which have similar steric and electronic features and have the same number of atoms as the substituent moiety for which they are used as replacement. Non classical bioisosteres do not obey the strict steric and electronic definition of classical isosteres and they don not have the same number of atoms as the substituent moiety for which they are used as a replacement. These isosteres are capable of maintaining similar biological activity be mimicking the spatial arrangement, electronic properties, or some other physicochemical property of the molecule of functional group that is critical for the retention of the biological activity.
DESIGN OF RIGID ANALOGS
Imposition of some degree of molecular rigidity on a flexible organic molecule (e.g., by incorporation of elements of the flexible molecule into a rigid ring system or by introduction of a carbon-carbon double or triple bond) may result in potent, biologically active agents that show a higher degree of specificity of pharmacological effect. The three-dimensional geometry of the pharmacophore can be determined and the key functional groups are held in one steric disposition, or in the case of a semirigid structure, the key functional groups are constrained to a limited range of steric dispositions and interatomic distances. By the rigid analog strategy, it is possible to approximate “frozen” conformations of a flexible lead molecule that, if an enhanced pharmacological effect results, may assists in defining and understanding structure activity parameters.
HOMOLOGATTION OF ALKYL CHAIN OR ALTERATION OF CHAIN BRANCHING, CHANGES IN RING SIZE, AND RING POSITION ISOMERS.
Change in size or branching of an alkyl chain on a bioactive molecule may have profound and sometimes unpredictable effects on physical and pharmacological properties. Alteration of the size and shape of an alkyl substituent can affect the conformational preference of a flexible molecule and may alter the spatial relationships of the components of the pharmacophore, which may be reflected in the ability of the molecule to achieve complementarily with its receptor or with the catalytic surface of a metabolizing enzyme. The alkyl group itself may represent a binding site with the receptor through hydrophobic interactions, and alteration of the chain may alter its binding capacity. Position isomers of substituents even alkyl groups on an aromatic ring may possess different pharmacological properties. In addition to their ability to affect electron distribution over an aromatic ring system, position isomers may differ in their complementarily to receptors, and the position of a substituent on a ring may influence the spatial occupancy of the ring system.
ALTERATION OF STEREOCHEMISTRY AND DESIGN OF STERIOISOMERS AND GEOMETRIC ISOMERS.
In the case of chiral molecules, one enantiomer would be expected to demonstrate pharmacological activity and the other enantiomer should be expected to be pharmacologically inert, is not valid.
At high doses (R)-enantiomer selectively stimulated presynaptic dopaminergic receptor sites, while at lower doses, it selectively stimulated postsynaptic receptor sites. In contrast, the (S)-enantiomer stimulated presynaptic dopamine receptors and at the same dose level, it blocked postsynaptic dopamine receptors. Thus, this enantiomer exhibits a bifunctional mode of dopaminergic attenuation: that of presynaptic agonism and postsynaptic antagonism.
FRAGMENT OF THE LEAD MOLECULE
Design of fragments of a lead molecule is based on the premise that some lead molecules, especially polycyclic natural products, may be much more structurally complex than is necessary for optimal pharmacologic effect. A bond disconnection strategy may be employed, in which bonds in the polycyclic structure are broken or removed or destroy one or more of the rings.
VARIATION IN INTERATOMIC DISTANCES.
Alteration of distances between portions of the pharmacophore of a molecule or even between other portions of the molecule may [produce profound qualitative and quantitative changes in pharmacological actions.
The screening cascade is designed to decrease the number of compounds examined at each level in order to ensure that compounds with flaws are removed as early as possible. The cascade, also referred to as a screening tree, begins with in vitro profiling and then transitions into in vivo studies designed to determine a compound’s pharmacokinetic profile and demonstrate efficacy in an appropriate animal model.
At the top of the cascade, compounds are screened for activity against the biological target and a threshold of interest is generally set to determine if compounds are active enough to warrant further investigation. Potency is, of course, an important issue, as dosing requirements are lower for compounds that are more potent. All other things being equal, compounds with higher potency can be dosed at lower levels, decreasing the likelihood of side effects. A compound with target potency of 5nM in theory could be provided to a patient at a significantly lower dose than a compound serving the same function but with a potency of 5μM.
Once a compound has satisfied the potency criteria, selectivity and physicochemical properties criteria are typically examined. Nature has developed exquisite systems to accomplish very specific tasks with highly selective systems, but many of these systems overlap structurally, and this can have a significant impact on the biological properties of a given test compound. Thus, the next biological screening step in a typical screening cascade is often an assessment of a compound’s potency at biological systems that are closely related to the target of interest. The Kv1.5 channel, for example, is a voltage-gated potassium channel that has been the target of research programs atrial arrhythmia, and many compounds have been identified that can block this channel with a high level of potency.82 There are, however, over 70 other voltage-gated potassium channels with varying degrees of similarity to the Kv1.5 channel, and undesired activity at any of these related channels could create unwanted side effects in human or animal studies. For example, the Kv1.5 channel is closely related to the hERG channel. Blockade of the hERG channel has been linked to torsade de pointes and sudden cardiac death,83 so any compound moving forward in this area would need to be counterscreened for hERG activity in order to ensure that advancing compounds do not present a risk of sudden cardiac death in a clinical setting. This is a rather extreme example of the importance of proper selectivity, but it should be clear that failure to achieve proper target selectivity in this area represents a significant barrier to moving a program forward.
Similarly, there are over 500 known kinase enzymes,84 and any drug discovery program designed to target a single kinase, or even a family of kinases, has an associated risk of identifying compounds that are active at multiple members of this large family of related enzymes. In order to mitigate this risk, kinase programs routinely screen test compounds against panels of related kinases in order to understand the risks associated with off-target activity.
In general, compounds that are potent at the target of interest, but are also potent at a variety of other targets (“promiscuous compounds”), do not move forward in a drug discovery program, as the risk of undesired (or unpredicted) side effects is too high. The level of selectivity required, however, is dependent on the program, the nature of the potential side effect presented by off-target activity (off-target activity leading to excessive hair growth might be tolerable, whereas sudden cardiac death through poor hERG selectivity is not), the target patient population, duration of treatment (some side effects only appear upon extended exposure to a drug), and a variety of other factors. Overall, target selectivity is a major factor to consider.
An active and suitably selective compound, however, is not necessarily a good drug candidate. Physicochemical properties also play a major role in determining whether or not a compound is suitable for further investigation. In vitro screens designed to predict absorption, distribution, metabolism, and excretion (in vitro ADME) are generally performed early in a program in order to ensure that candidates reaching the drug development pipeline are “druglike” in nature. Compounds that have poor aqueous solubility, for instance, are often difficult to develop as drugs. In order for a drug to exert an influence on a biological target, it must be soluble in biological fluid at a level consistent with its potency. Thus, the level of solubility required for a given compound is directly linked to its potency. As potency increases, the requisite solubility decreases, as less drug is required to provide the intended effect. Solubility also has a direct impact on absorption, as a compound must be soluble in biological fluids in order for it to successfully pass through a biological membrane and reach its intended target.
The ability of a compound to penetrate cellular membrane (its permeability) can also be a determining factor in the success or failure of a given candidate compound. If a compound is potent, selective, and soluble, but unable to pass through a biological membrane, it may not be able to reach the target of interest and fail to demonstrate the desired efficacy. Orally active drugs must be absorbed in the gastrointestinal tract, and drugs that target intracellular system must also pass through the cell membrane in order to reach their intended targets. Extracellular targets, of course, do not face this added issue, but CNS drug candidates face the added complexity of required permeability through the blood–brain barrier (BBB). Additionally, there are efflux pumps (e.g., P-glycoproteins (Pgps))85 designed to remove xenobiotic material that can limit permeability, preventing efficacy. The inability of a compound to penetrate a cell membrane represents a significant issue that could prevent further investigation of the candidate compound.
Metabolic and chemical stability are also important considerations. If all of the previously mentioned criteria are met, and a compound is able to enter the body, but is immediately metabolized, efficacy studies will fail to show the desired results. However, the relative rate of metabolism that can be allowed for a successful compound depends on the goals of the project. If, for example, the goal is to develop a new antibacterial agent, then high-metabolic stability will likely be desired so that the potential drug candidate will be available in the circulation long enough to kill the invading organism. If, however, the goal is to develop a new surgical anesthetic, metabolic stability may be less of an issue, as it may be desirable for drug efficacy to fade rapidly upon termination of dosing regimens.