Advances in computational, biophysical and library technologies have driven the emergence of a new drug discovery paradigm with the power to develop small molecule drugs with unprecedented selectivity and target proteins previously written off as undruggable. In the past, the pharmaceutical industry focused its drug discovery efforts on the lowest-hanging fruit. Enzymes, which have evolved to bind substrates, naturally have a site available for small molecule interactions and are obvious drug targets due to their key role in cell signaling that causes disease. Targeting enzyme activity through the development of orthosteric drugs, which directly bind to the active- or ligand binding-site and compete with the natural ligand, was the major industry focus leaving little room for alternative approaches.
One classic example of a highly successful orthosteric drug is Gleevec, a tyrosine kinase inhibitor approved in 2001 to treat chronic myelogenous and acute lymphocytic leukemias (CML and ALL) that contain the constitutively active kinase and oncoprotein, BCR-ABL1. Gleevec occupies the ATP binding pocket (active-site) of BCR-ABL1, blocking its ability to phosphorylate down-stream substrates and ultimately stops cells from dividing.
The strategy of orthosteric drug development, while yielding drugs to treat a wide range of diseases, has several drawbacks, which are driving the necessity for a more comprehensive drug discovery platform. Most notable is the restriction to only a few families of proteins. Orthosteric drugs require their targets to have an accessible active-site, which is only found in some enzymes, such as kinases. Additionally, within these enzyme families, the enzymatic active-sites are often highly conserved, which makes it very challenging to identify selective inhibitors and increases the possibility of negative side effects (e.g., Gleevec also inhibits off-target kinases c-Kit, PDGFR, ABL2 and DDR1). Since orthosteric drugs bind within a target’s active-site, there is inherent competition from the cellular molecule that naturally binds to that site. Thus, the orthosteric drug will either need to be given at a high dosage, or have a very high binding affinity, which will both result in a higher chance of inhibiting non-target enzymes.
Orthosteric drugs require their targets to have an accessible active-site, which is only found in some enzymes, such as kinases.
Mutational resistance to orthosteric drugs in cancer patients is also a major driver for the development of novel drugs. For example, about 15-20% of patients treated with Gleevec over three years acquire resistance mutations, which often occur at the active-site and interfere with Gleevec binding. Additional orthosteric drugs that target the mutated versions of BCR-ABL1 have been developed to overcome resistance, however they only work on a subset of resistant mutations.
Recently, a new kind of BCR-ABL1 targeting drug has been developed, and is now in clinical tri-als, with the potential to wipe out drug resistance as a limiting treatment factor. ABL001 is a small molecule that binds to the myristoyl pocket within the ABL1 portion of the BCR-ABL1 protein and induces a conformational change that inactivates its kinase activity. In the wild-type ABL1 pro-tein, a myristate group, a saturated fatty acid with 14 carbon atoms, resides at its N-terminus and binds to a hydrophobic pocket within ABL1 causing a conformational change and autoinhibition of its kinase activity. The myristate group is lost in the BCR-ABL1 fusion protein preventing autoinhibition. A combination therapy of ABL001 and drugs like Gleevec that target the active-site would inhibit BCR-ABL1’s kinase activity through two independent biding sites and make mutational drug resistance virtually impossible.
ABL001 belongs to a class of drugs known as allosteric, defined by their modulation of protein function by binding at regulatory sites physically distinct from an active-site. Allostery is a natural regulatory mechanism used within cells to finetune signaling of important pathways. Post-translational modifications, binding of effectors, and internal protein interactions are all ways in which proteins can be allosterically modulated, usually due to a conformational change, to enhance or block their downstream activation.
In addition to providing an additional way to target signaling proteins in cancer, one of the great benefits of allosteric drugs is their high specificity for target proteins. Unlike catalytic sites, allosteric sites don’t have evolutionary pressure to conform to a conserved sequence/structure and instead are more unique. With high specificity comes high safety due to a lack of off-target binding, a feature highly desired in new drugs. While the development of allosteric drugs seems like an obvious road for the industry to travel, there have been considerable limitations that have only recently been overcome with advancements in computational, biophysical and library technologies.
Unlike catalytic sites, allosteric sites don’t have evolutionary pressure to conform to a conserved sequence/structure and instead are more unique. With high specificity comes high safety due to a lack of off-target binding, a feature highly desired in new drugs.
The vast accumulation of protein 3D structures, functional knowledge, and wide spread identification of allosteric regulatory sites, known as the ‘regulome’, has led to significant progress in allosteric drug discovery. Computational modeling has revealed potential allosteric sites among thousands of proteins that can now be verified through advancements in biophysical techniques such as crystallography, cryo-EM, and NMR. Computational modeling has also led to the ability to test billions of molecules in silico to identify highly probable small molecule allosteric inhibitors and allow for more focussed in vitro screens. Dramatic improvements in compound library technologies now allow vast numbers of molecules to be generated and screened in parallel with phenomenal diversity, for example, via DNA encoded libraries, and new compound diversity including macrocycles and cyclic peptides. In the latter case, library sizes can be as diverse as 1016! The combination of these techniques and new insights into the regulome provides a powerful platform for allosteric drug discovery.
Previously undruggable families of proteins are now druggable through the development of allosteric drugs. Phosphatases, for example, are notoriously difficult to treat due to their highly hydrophobic active sites, which results in small molecule inhibitors that are unable to cross the cell membrane. SHP2, an oncogenic tyrosine phosphatase with no approved therapies, is now a drug target for an allosteric drug SHP009, currently in clinical trials, which stabilizes SHP2 into an autoinhibitory conformation. Additionally, transcription factors, which have the potential to be revolutionary drug targets have eluded targeting by the industry. It is well known that transcription factors are highly regulated through co-factors, post-translational modifications, and conformational changes. With the new knowledge and techniques for allosteric drug discovery, transcription factors are next in line as drug targets for the abolishment of a myriad of diseases.
