Chemical biologists break barriers of druggability

Most researchers don’t associate weight loss surgery with chemical biology, but A. Sloan Devlin is not most researchers.

A chemical biologist at Harvard Medical School, in Boston, Massachusetts, Devlin knew that bariatric surgery was frequently associated with an anti-diabetic effect. A set of metabolites, called bile acids, were thought to play a role, though no-one knew how. She set out to determine which particular molecule was therapeutically enhancing sugar metabolism in the guts of bariatric surgery patients.

Together with her surgical colleagues Eric Sheu and David Harris, along with postdoc Snehal Chaudhari, Devlin profiled all the bile acids found in the upper large intestines of obese mice subject to sleeve gastrectomy, a common weight-loss surgical procedure. Among more than 50 bile acids analyzed, the presence of just one increased significantly as a result of the surgery. What’s more, the same molecule, called cholic-acid-7-sulfate, or CA7S, was enriched in fecal samples from human patients after a sleeve gastrectomy.

“That was striking to us,” Devlin says. “When you see the same thing in mice and humans, it’s a really good sign that you’re on to something biologically important.”

Devlin and her colleagues went on to describe precisely how the gut-restricted bile acid improves glucose regulation. It activates a protein called Takeda G protein-coupled receptor 5 (TGR5), which induces the production of a hormone needed for insulin secretion. The researchers discovered a novel signaling axis between the intestines and liver that seems to implicate surgery-induced shifts in gut microbial communities with the synthesis of CA7S. They also showed how, in mice at least, oral administration of the molecule seems to mimic the anti-diabetic benefits of bariatric surgery, no scalpel required, an exciting observation with broad implications.

Devlin is among the growing cadre of researchers who are pushing the boundaries of chemical biology, and with it translational sciences. She presented her results in October at the Translational Chemical Biology meeting, a Nature Conference held in partnership with the Dana-Farber Cancer Institute of Boston, Nature Chemical Biology and Nature Reviews Drug Discovery.

The goal of this event was to bring together a community of scientists who, like Devlin, are using the tools of chemical biology to reshape the drug discovery landscape. Other chemical biologists are similarly broadening the possibilities of drug development with new technologies for disrupting protein-protein interactions, scaffolding proteins, RNA molecules and other targets that have eluded scientists in the past.

“Only a tiny fraction of the available drug targets is currently accessed given the state of today’s pharmacopeia,” notes Dana-Farber chemical biologist, Nathanael Gray, one of the co-organizers of the Translational Chemical Biology meeting. But with the advances discussed at the conference, he says, “we can hugely expand the druggable universe.”

Frontiers in post-transcriptional and post-translational modifications

In drug discovery, one of the frontiers lies in targeting RNA. Unlike proteins, stringy RNAs weren’t believed to have well-defined folds and pockets that small molecule drugs could latch on to. In recent years, researchers have proven that view incorrect.

Portions of some RNAs are indeed rich in potentially druggable structures, notes Amanda Hargrove, a biochemist at Duke University. “It’s become clear that it is possible to target RNA functionally in a biological system,” she says.

In August, US regulators approved the first small molecule drug directed at an RNA target outside the ribosome, a therapy known as risdiplam, for spinal muscular atrophy. Many companies are now racing to bring others like it to market.

One of those companies is Arrakis Therapeutics, based in Waltham, Massachusetts. At the conference, chief scientific officer Jennifer Petter described how risdiplam and other clinical-stage candidates like it were initially discovered in phenotypic screens, and only later shown to work by acting upon RNA targets. Now, she says, “we’d like to purposefully go after these kinds of targets.”

To do so, Petter’s team developed a new platform for analyzing small molecule–RNA interactions. The platform, called PEARL-seq, provides information both on binding sites within RNA targets and on drug selectivity in a single experiment. In a proof-of-concept report, the Arrakis researchers outlined how their method helped identify small molecules with high affinity and specificity against a synthetic RNA. They are taking the same approach to find drug leads for RNA associated with chylomicronaemia syndrome, a heritable disease of lipid metabolism, and two RNAs implicated in cancer.

For researchers targeting RNAs, Petter says, “there’s nothing remarkable or magical about going after it.” You just need the right tool, which is typically a chemical probe with demonstrated activity against its target in a biological system.

To help researchers work faster, Hargrove and her colleagues compiled a list of such RNA probes — around 100 traditional small molecules and 40 multivalent ligands in total. They incorporated detailed experimental information about each one to create an online resource called the RNA-Targeted BIoactive ligaNd Database, or R-BIND.

“We are hoping this will be useful for lead development, library design or even just identifying controls that people might want to use in their assays,” Hargrove says.

One workaround to drugging RNA directly is to modulate the enzymes that write, erase or recognize chemical modifications that are added to RNA molecules to alter their activity. That’s the approach taken by Chuan He, a chemist at the University of Chicago in Illinois. At the meeting, He explained how disruption of RNA epigenetic mechanisms can lead to huge changes in cell function. Many companies, including one he cofounded, have formed around the idea of targeting RNA-modifying proteins for therapeutic gain.

Several conference speakers also discussed ways of modulating the chemical changes that occur to proteins, called post-translational modifications, as a means of drugging targets that have proven intractable in the past.

Chemists Kate Carroll, of the Scripps Research Institute in Jupiter, Florida, and Eranthie Weerapana, of Boston College in Massachusetts, presented strategies for perturbing the oxidized cysteine residues found on proteins involved in human disease. Cheryl Arrowsmith, a cancer epigeneticist of the University of Toronto and Structural Genomics Consortium in Toronto, Canada, described how pharmacologically inhibiting enzymes that add methyl tags to proteins helps counter the genetic dysregulation involved in tumour growth. And Christina Woo, a chemical biologist at Harvard University in Cambridge, Massachusetts, talked about controlling the dynamics of a sugar adornment called N-acetyl glucosamine (O-GlcNAc) to maintain cellular health.

In all these cases, the goal is the same: as Woo puts it, “To enable targeting of the undruggable proteome.”

A focus on targeted protein degradation

Sometimes targets are simply too difficult to modulate. In those cases, many in the pharmaceutical industry are now exploring how to get rid of them altogether.

Much of the focus is on how to hijack the so-called ubiquitin-proteasome system, which tags certain proteins for degradation. Known as proteolysis-targeting chimeras, or PROTACs, the two-headed small molecules bind to a target protein on one end and recruit a ubiquitin ligase enzyme on the other, leading to target degradation.

“We and others have shown that, indeed, we can degrade a wide variety of different targets,” says Alessio Ciulli, a biochemist at the University of Dundee, UK. “We therefore now have proof-of-concept validation for a novel modality [of drug development] where we can really achieve significant differentiation from more conventional [inhibitor-based] approaches.”

At the meeting, Ciulli discussed his team’s efforts to convert small molecule binders directed against bromodomain-containing proteins into more selective and potent PROTAC degraders. Bromodomain-containing proteins play a critical role in the regulation of gene expression and have been implicated in cancer.

Giulio Superti-Furga, a systems biologist at the CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences in Vienna, described a new push to unlock the therapeutic potential of targeting solute carrier protein transporters, including through the development of PROTACs.

In the virtual poster session, several students and postdocs described innovations at the forefront of the targeted-protein degradation field. These include novel PROTACs directed against KRAS, an oncoprotein implicated in about 30% of all cancers, and PARP1, a protein linked to aberrant cellular stress responses in many diseases. They also include new tools for expanding the potential of next-generation degraders with novel ubiquitin ligase binders, or with systems for trafficking target proteins to the lysosomal pathway.

That clean-up system can do away with proteins found outside of cells, overcoming a drawback of proteasomal targeting, which is limited to proteins inside the cell interior. Speaker Anna Greka, director of the Kidney Disease Initiative at the Broad Institute in Cambridge, Massachusetts, also spoke to lysosomal degradation. She described a newly discovered conventional small molecule that helps spare the kidney from the toxicity of a mutant protein called MUC1 by rerouting MUC1-filled vesicles inside renal tubular epithelial cells to the lysosome.

New horizons in protein degradation and inhibition

All these modular degrader technologies “open up the world of the druggable,” says William Kaelin, Jr., a molecular oncologist at Dana-Farber. “But, this is only one way a chemical might degrade a protein.”

As Kaelin has shown, some drugs, including thalidomide-like drugs such as lenalidomide, a common therapy for people with multiple myeloma, can act like a sort of molecular glue to stabilize the interaction between the ubiquitin ligase enzyme and disease-associated proteins, leading to proteolysis of the target. He and others are now engaged in identifying novel anti-cancer agents that function in similar ways.

“We think that other small molecules may also function as these types of molecular glues able to cause the degradation of their target proteins,” says Nicolas Thomä, a structural biologist at the Friedrich Miescher Institute for Biomedical Research in Basel. In collaboration with Dana-Farber leukemia biologist Benjamin Ebert, Thomä recently conducted a systematic search for compounds that were highly toxic to cancer cells that expressed high levels of a specific ubiquitin ligase component. This led the researchers to show how CR8, a known inhibitor of cyclin-dependent kinase (CDK) enzymes, actually works as a molecular glue degrader of cyclin K, an important binding partner of a CDK.

Industry perspectives at the meeting came from Gregory Michaud, a chemical biologist at Novartis Institutes for BioMedical Research in Cambridge, Massachusetts, and Ingrid Wertz, a molecular biologist at Genentech in South San Francisco, California. Michaud described unpublished internal data on a molecular glue degrader that could reprogram the VHL E3 ligase to degrade a metabolic enzyme that has an important role in maintaining the hepatic concentration of intracellular free cysteine.

Wertz showed how the cytotoxic payload of a CD79b-targeted antibody-drug conjugate called polatuzumab vedotin works, in part, by promoting proteasome-mediated degradation of MCL-1, an anti-apoptotic protein whose activity often underpins resistance to the BCL-2 inhibitor venetoclax. That observation led Genentech to begin testing the combination of polatuzumab vedotin, which was approved in 2019 for the treatment of diffuse large B-cell lymphoma, together with venetoclax in patients with relapsed and refractory lymphoma.

At Rockefeller University in New York City, meanwhile, chemical biologist Tarun Kapoor has devised a strategy for minimizing the chances of drug resistance arising in the first place. Termed resistance analysis during design, or RADD, the approach involves integrating analyses of resistance early in the drug development process by engineering biochemically silent mutations into the target protein and then evaluating candidate inhibitors against them. Mutations that alter inhibitor–protein interactions are good indicators that the drug may be prone to resistance mechanisms arising.

“The approach not only identifies a chemical compound that is potent and selective, but also identifies potential resistance-conferring mutations,” says Kapoor. “At the time of drug design, when we are completing optimization of the chemical moiety, we already know the mutation in the target that will bump out the compound.”

Questioning what drugs can be

Another chemical strategy to overcome resistance involves the design of covalent inhibitors. These potent, selective and long-acting agents block protein function generally by forming irreversible bonds with their targets — although as Jack Taunton, a chemist at the University of California, San Francisco (UCSF), pointed out, covalent drugs can be made into reversible inhibitors as well.

One such reversible covalent drug that Taunton and his collaborators designed with fine-tuned target occupancy times and minimal off-target effects — a selective inhibitor of Bruton’s tyrosine kinase called rilzabrutinib — is now in late-stage clinical trials for the treatment of a blistering skin disease called pemphigus vulgaris, and a blood disorder known as immune thrombocytopenia.

A decade ago, covalent inhibitors had fallen out of favour among drugmakers because of concerns about potential toxicity. Now, as computational chemical biologist Nir London from the Weizmann Institute of Science in Rehovot, Israel, puts it: “Covalent is the new black”. Efforts to find promising covalent drug candidates, including for previously undruggable targets, have been helped by high-throughput assays of the kind London and others have developed.

Matthew Bogyo, for example, recently created a phage display–based approach to identifying covalently binding agents. A chemical biologist at Stanford University in California, Bogyo is now putting the technique to use in the hunt for novel anti-infective agents. London, meanwhile, devised a crystallographic screening protocol that led to the discovery of the first known inhibitors against a pair of cancer-associated enzymes. Using the same approach, his team has since found compounds with covalent activity against the main protease enzyme of the novel coronavirus responsible for COVID-19.

Ellen Berg, chief scientific officer for translational biology at DiscoverX in South San Francisco, discussed another screening option. Her BioMAP system, which has undergone more than 10 years of refinement, offers a platform for researchers to interrogate the function of candidate drugs in an exhaustive panel of human primary cell-based models of tissue and disease biology.

Rather than targeting proteins for therapeutic gain, UCSF computational biologist, Tanja Kortemme is focused on engineering proteins to take on new therapeutic functions. She and her colleagues have developed design strategies for creating proteins with distinct shapes and folding patterns. With an eye to protein biosensors that can detect and respond to molecular signals in living cells, Kortemme’s team has written the playbook for fashioning proteins with new deliberate binding site conformations as well. “We can build new functional type geometries into the existing proteins by computational design,” she says.

Approaches like those, and the many others under the spotlight at the October meeting collectively, show how chemical biologists are chipping away at the barriers of druggability.

“It opens everyone’s eyes to things we previously thought impossible actually being possible,” says Kaelin. And whether with conventional small molecules, protein degraders, covalent inhibitors, molecular glues — or even gut metabolites — the community that gathered virtually for the Translational Chemical Biology forum is challenging the conventional wisdom on what a drug should be.