One company’s antibody discovery technology is breaking open challenging drug targets
G-protein-coupled receptors (GPCRs) are the most diverse group of cell surface receptors in eukaryotes. There are 750 GPCRs in the human genome, transmitting a multitude of messages to cells about their environment and affecting everything from cardiovascular disease to obesity, autoimmunity to cancer, asthma to reproductive drive, infertility to psychiatry. As such, they are a popular target for drug developers, but taking perfect aim at GPCRs has been be tricky.
Small molecule drugs struggle to specifically discriminate a single member of the 750 member family, creating off-target side effects. Antibodies, renowned for their exquisite specificity as therapeutics, would be a perfect solution to target GPCRs in many cases. However, because of the unique biology of GPCRs, targeting them with antibodies has been traditionally extremely challenging. Distributed Bio, a biotech company that entered an exclusive partnership with Charles River last year, has developed two technologies to routinely produce antibodies against GPCRs.
“What we have is the ability to easily make antibodies against GPCRs,” said Dr. Jacob Glanville, Chief Science Officer and founder of Distributed Bio. “That already is a remarkable feat, because that is extremely difficult, but on top of that we have the ability to make antibodies that are totally human, that look like normal human antibodies that you could have produced in your own blood.”
Although about 30% of all small molecule therapeutics target GPCRs, they are notoriously difficult to work with since they do not hold together when they are removed from the cell surface membrane. “GPCRs are beautiful but complex. They are woven seven times, back and forth through the membrane of the cell, like stitching a button.” Said Sawsan Youssef, Senior Director of Immunology and Immune Oncology at Distributed Bio. “Like a stitched button, when the thread is pulled out of the cloth, everything falls apart.” This can make them hard to study and to raise antibodies against, since they are impossible to isolate without destroying them. “Antibody engineers like to raise antibodies against their drug target in isolation,” says Valerie Chiou, Scientist at Distributed Bio. “This helps the antibody discovery technologies focus. However, with GPCRs, you can’t remove them from the cell or they fall apart, so you have this challenge of trying to produce antibodies against the GPCR while there are thousands of other distracting proteins on that same cell surface.” It also makes it difficult to create drugs to target certain GPCRs while avoiding others, since there are hundreds of GPCRs to navigate.
Distributed Bio attacked this problem on two fronts. First, they created the most comprehensive library of antibodies possible. Second, they optimized a high-throughput genomic sequencing technique to distinguish between antibodies that were talking with GPCRs and those that were not.
“The combination of both of those things has allowed us to peer through the haze of billions of antibodies that are irrelevant and focus on the ones that are hitting the GPCRs,” Dr. Glanville said.
Antibody libraries made the news last year when Sir Gregory Winter and Professor George Smith shared half of the Nobel Prize for their work on phage display of peptides and antibodies. Since 1989 researchers have used this technique to test synthetic antibodies against various cell surface proteins like GPCRs, but the volume of work began to get out of hand.
“For years they were operating in the dark,” said Dr. Glanville. “Their technology let them build libraries much larger than they could actually read, so vast that they had no way to really check to make sure they were as diverse as they thought they were. It was like the internet – billions and billions of pages, but if you invented an AI to actually go out and read every page, it would come back and tell you that most of the internet is garbage. That’s essentially what I did back in 2009 for antibody library technology. And we’ve been trying to fix the problems that we found ever since.”
Using this deep library analysis method, Jacob and his teams iterated through 7 generations of library designs since 2009. Now, Distributed Bio’s SuperHuman 2.0 library holds about 76 billion antibodies, and every one of those antibodies is well-folded and unique, making it one of the most diverse libraries available. “We’ve improved library sizes by about 2000x since 2008,” says Jacob. “This definitely helps finding hits against hard targets like GPCRs, but even with this larger library, we would still have the problem of lots of distracting other proteins on the outside of cells that contain our GPCRs.”
Refining the antibody search
To address that, Distributed Bio couples their library with a genomic sequencing technique to help identify isolate, and synthesize only the antibodies that are specific to their GPCR of interest. This is crucial, as it helps them identify antibodies that are extremely specific to just their GPCR of interest, and allows them to ignore the oceans of useless antibodies against other cell surface proteins. “Whole departments at multiple institutions worked for years without finding hits against GPCRs like CXCR5,” said Jacob. “With our library and discovery method, Valerie Chiou generated multiple functionally active, fully human anti-GPCR antibodies against CXCR5 while working with us as a graduate student intern.” Valerie has since been hired by Distributed Bio.
Last year the FDA approved the first GPCR-targeted monoclonal antibody drug with erenumab, which prevents migraines by targeting calcitonin gene-related peptide receptors (CGRPR). This was one of several drugs approved in 2018 with novel targets, demonstrating the potential that previously untargeted molecules can bring to treating patients. Given the interest in new molecular targets, GPCRs and Distributed Bio are poised to attract a great deal of research attention in the coming years.