Finding a new drug often begins with a simple question: Can we discover a molecule that binds to and changes the behavior of a disease-causing protein? For decades, answering that question has required testing compounds one at a time. In the most advanced pharmaceutical laboratories, scientists may screen one to five million molecules during a research campaign.
With support from a Welch Catalyst grant, Damian Young, Director of the Center for Drug Discovery and Professor of Biochemistry and Molecular Pharmacology and Pathology & Immunology, and his co-investigators Martin M. Matzuk, Zhi Tan, and Mingxing Teng at Baylor College of Medicine are developing DNA-encoded chemical libraries (DELs) that allow researchers to screen billions of compounds together in a single test tube.
Typically, in drug discovery, a protein of interest is identified that researchers think is involved in a disease. Their goal is to find a molecule that may modulate that protein’s function. There are many ways to screen for these molecules, but DELs are unique in scale and efficiency.
In most screening technologies, every individual compound must be “spatially arrayed,” which means each compound is identified and tested separately in its own test tube. With the DELs, in contrast, each compound is tagged with a unique DNA “barcode,” and billions of tagged compounds are put into one mixture in one test tube.
Researchers introduce the protein of interest into their billion-compound mixture, using the protein as bait to fish out the molecules that bind to it. By sequencing the DNA tags attached to the molecules that bound to the protein, scientists can rapidly identify which compounds were successful. Thus, instead of running millions of individual tests, the process becomes a single, powerful selection experiment.
“We screen all those molecules in one test tube, instead of one billion test tubes,” said Dr. Young. “The efficiency of this approach allows us to explore chemical space at a scale that just isn’t possible with traditional screening. And the more chemical possibilities we can sample, the greater our chances of finding the magic bullet in one experiment.”
While DEL technology holds enormous promise, it presents major chemical challenges, especially since, historically, it has been utilized primarily by secretive pharmaceutical companies.
“There was a lot of innovation that needed to occur in terms of the chemistry and how we build these billion-compound mixtures. The technology, at least in the literature, had not progressed very far, since we are one of the first organizations to utilize it in an open, academic environment,” said Dr. Young. “What we’re doing now really gives it a new life that will hopefully be implemented in companies and beyond.”
One major challenge is that all the reactions must be performed in water to keep the DNA tags stable and readable. Yet water can interfere with the progression of many chemical reactions. At the same time, the reactions must not damage the DNA tags or scientists will never be able to identify a successful hit.
To address this, Dr. Young’s team is pioneering what they call “DNA-preserving reactions.” These newly developed chemical strategies are designed to protect the DNA tag while still allowing the creation of diverse, complex molecules.
This seemingly technical advance has broad implications.
If the DNA identifier is damaged during the construction of a library, promising molecules may be missed entirely. By improving DNA integrity during reactions, the team believes they can reveal binding compounds that earlier methods failed to detect.
“We have to think creatively in terms of how we’re going to get a reaction to occur in a solvent that’s really highly reactive and that has all that DNA. It really pushes us to rethink fundamental chemistry in new ways,” said Dr. Young.
Dr. Young’s team is working to use DELs not just to find initial hits, but also to optimize them by building focused follow-up libraries and iteratively selecting for better binders. The project also integrates machine learning to predict promising candidates and guide future library design.
“I am really grateful to The Welch Foundation for recognizing the power of this technology and also valuing that it requires a new type of chemistry. There is a real need for fundamental research that helps researchers succeed in their own respective domains,” he says. “We are innovating these DNA-encoded libraries and DNA-preserving reactions, and then others can access these tools to target cancer, cardiovascular disease, neurological disorders, and whatever else they’re working on. We’re having fun with it, and I’m grateful that it continues to evolve as new ideas emerge.”