Scientists develop molecules that may treat Crohn’s disease

An estimated 3 million Americans have an inflammatory bowel disease (IBD) such as Crohn’s disease or ulcerative colitis. But a lucky few individuals are far less likely to develop IBD because they have a rare variant of a gene called CARD9. This protective gene variant prevents the long-term digestive tract inflammation that can cause tissue damage and lead to disease.

Now, researchers at the Broad Institute, Mass General Brigham, Harvard Medical School, and Johnson & Johnson Innovative Medicine have developed small-molecule drug candidates that mimic the effects of this rare gene variant and could potentially treat Crohn’s and other inflammatory bowel diseases.

The protective CARD9 variant was first identified by Broad researchers in 2011, who then discovered in 2015 how it worked to reduce IBD risk. The new study, published in Cell, reveals a set of molecules that have the same effect as the protective variant and decrease sustained inflammation in mice. The work demonstrates a complete genetics-to-therapeutics pipeline—from the initial discovery of a protective gene variant to the design of molecules that reproduce its beneficial effects.

“The deep investments in human genetics are starting to pay off,” said study senior author Ramnik Xavier. “This study shows we can translate genetic insights about disease all the way to new drug candidates.”

Xavier is a Broad core institute member, the Kurt J. Isselbacher Professor of Medicine at Harvard Medical School, and director of the Center for Computational and Integrative Biology and core member in the Department of Molecular Biology at Massachusetts General Hospital.

The new study offers a roadmap for how to systematically translate genetic discoveries into new medicines. While this work took more than a decade, the team says similar efforts in the future can be accelerated thanks to new technologies such as gene editing in cells and disease models, novel chemistry approaches, and AI in drug discovery.

“The genetics-to-therapeutics approach demonstrated in this CARD9 work is the full realization of the promise, and ultimate purpose, of our work in human genetics. This paradigm can and should be applied to other diseases where genetics has now identified causal genes and mechanisms,” said Mark Daly, co-director of the Program in Medical and Population Genetics at the Broad and an institute member who led the 2011 research that identified IBD genes including CARD9.

Daly is also the founding chief of the Analytic and Translational Genetics Unit at Massachusetts General Hospital and a professor at the Harvard Medical School.

The study’s first authors are Jason Rush of the Center for the Development of Therapeutics at Broad, and Joshua Wertheimer and Steven Goldberg of Johnson & Johnson Innovative Medicine.

Genetics first

In 2011, Daly, Xavier, and colleagues sequenced the genomes of tens of thousands of people with or without Crohn’s disease or ulcerative colitis. The team identified dozens of genes linked to the conditions, including CARD9, which is known to be involved in inflammatory responses. People who had one common variant of CARD9 had an increased risk of developing IBD, while those with a rarer variant that truncated the protein were protected from the disease.

But translating this genetic discovery into a drug was far from straightforward. Completely shutting off the function of the CARD9 protein would prevent the immune system from fighting infections in the gut. The protective variant somehow allowed this initial immune response while blocking longer-term inflammation. Moreover, CARD9 is what scientists call an “undruggable” target—a scaffolding protein with no obvious binding pockets for small molecules.

Before attempting drug discovery, Xavier’s team first needed to understand exactly how the protective variant worked and which sections of the CARD9 protein were most important. In 2015, they published data showing how the protective, shortened version of CARD9 acted like a brake on inflammation. They now knew which section of the protein—a so-called coiled-coil domain—must be targeted with a drug. 

“This work underscores how important it is to understand the nuanced biology of a protein before we target it with new drugs,” said Daniel Graham, a co-author on the latest study, and senior director of functional genomics in the Infectious Disease and Microbiome Program at the Broad, where he is also an institute scientist. “The old perspective of immediately trying to develop an activator or inhibitor isn’t going to serve us well with targets emerging from genetics.”

Screen time

In the new study, Xavier and his colleagues worked closely with Johnson & Johnson's Janssen research division to tackle the drug discovery challenge. They began with a massive screen of 20 billion molecules, searching for any compound that could bind to CARD9. The initial molecules they found didn’t have an impact on inflammation.

Rather than abandoning the project, the team developed a tool to forge ahead. They obtained the first-ever crystal structure of CARD9 that showed how a small molecule could bind to a coiled-coil domain, confirming that they were targeting the right place on CARD9. Then, they converted one of the binding molecules into a fluorescent probe and tested additional compounds from Janssen's library—this time looking for molecules that pushed the fluorescent probe out of the way and bind to CARD9 in its place.

Xavier said that using this “binder-first” approach was key to the team’s success. They pinpointed a new class of molecules that successfully blocked CARD9’s inflammatory signaling.

“The binder-first strategy gave us two critical pieces of information,” said Xavier. “It proved CARD9 was druggable and gave us a clearer picture of where the precise binding site was on the coiled-coil domain. That allowed us to narrow our search and find molecules that actually work.”

In human immune cells, the new molecules selectively reduced inflammatory signaling without affecting other immune pathways. In mice with the human CARD9 gene, treatment with the drug candidates reduced inflammation. 

Roadmap for the future

While the new compounds require additional optimization before they can be tested in humans, the work demonstrates a complete pipeline from genetic discovery to therapeutic candidates. 

Xavier’s group is continuing to develop the new set of compounds starting with mechanistic validation of human genetics and explore their potential as therapies for patients with immune-mediated disease. Because the protective CARD9 variant exists in humans without causing problems, the researchers are optimistic about the safety profile of drugs that mimic its effects.

“This is similar to the PCSK9 story for cholesterol drugs,” said Xavier, referring to a now-commonly prescribed class of drug that mimics the effects of a variant of the PCSK9 gene, which is linked to low cholesterol levels and reduced heart disease risk. “This hard-to-drug protective variant gave us the blueprint and the confidence that targeting this pathway can be done safely.”