When computer science meets the alpaca immune system

After earning bachelor's and Ph.D. degrees in bioengineering with a focus on molecular biology, Will White, a postdoctoral scholar in the Department of Computer Science at Tufts, was pretty certain his future research would involve things like DNA and proteins. What he didn't know was that alpacas would also be part of the equation.

For the past two years, White has worked with the furry mammals known not only for their teddy bear–like fleece, but also for the special antibodies they produce called nanobodies. The work culminated with a new paper in the journal Communications Biology called "Functional mapping of immune nanobody repertoires with NanoMAP."

Small antibodies, a slow discovery process

The paper stems from a research project started in 2022 by Charles Shoemaker, a professor in the Department of Infectious Disease and Global Health at Cummings School of Veterinary Medicine at Tufts. For decades, Shoemaker had been working with antibodies, the proteins produced to fight off viruses and disease, and specifically with nanobodies, the mini version of antibodies that sharks and members of the camelid family—including alpacas—uniquely produce.

Like other researchers, Shoemaker was interested in the way that nanobodies, which are 10x smaller than standard antibodies, easily slip into tight spaces and reach invaders. They are also easier to work with; standard antibodies are made of four protein chains (two heavy, two light) and matching them can be difficult. Nanobodies are made only of heavy chain proteins, eliminating the need for pairing.

What Shoemaker wanted with this new project, White says, was to find a way to modernize his strategy for discovering new nanobodies.

"Chuck and a lot of other people working with nanobodies tended to use more traditional methods for discovery," White says.

Researchers would pick random samples from a large group, test their properties in labs, and hope they had good candidates to use in experiments—a process that is time-consuming, may miss stronger candidates to study, and often needs repeating.

From alpaca blood to candidate treatments

Shoemaker and his team at Cummings School decided to partner with Lenore Cowen, a professor in the Department of Computer Science at the School of Engineering, and research assistant Edward Moseley, EG24, to create Nanobody Meta-clustering Analysis Platform (NanoMAP), a custom computational tool that makes it easier and more efficient for researchers to identify nanobodies that strongly bind to antigen targets. White joined the project two years ago.

The process starts with alpacas rented from a local farm, White says. Following National Institutes of Health guidelines, the alpacas are injected with a small amount of protein antigens from a pathogen—in this case from COVID-19, botulinum neurotoxin, or a parasitic blood fluke. This triggers their immune systems to produce nanobodies and fight back.

Researchers collect blood samples, separate B cells (the immune cells that produce nanobodies), extract the nanobody DNA, and then reformat it into an easier-to-handle format to test its ability to bind to the antigen target and to all known variants of the antigen.

After non-binders are removed, NanoMAP analyzes these sequences, groups them into "clonal families" based on the similarity of their amino acid sequences, then helps the researchers decide which nanobodies are the best candidates to develop into diagnostics or treatments.

"This allows us to capture nanobodies with the same or similar sequence that should have very similar properties in terms of how well they're going to do as a drug or as a diagnostic," White says. "If we group nanobodies with similar sequence, we can aggregate the information and get a more accurate picture of the behavior of that group."

Diverse targets and cross-disciplinary teamwork

White says what Tufts is doing is unique.

"There are other people trying to do somewhat similar things," he says, "but what we have that I haven't seen anywhere else is the size of our data set in terms of the diversity of the different target proteins that we can go after, like the botulinum toxin and the parasitic proteins and the COVID spike."

Another distinction—aside from working with alpacas—has been having two different departments work side by side.

"A lot of times, people in different fields, in this case biology and computer science, talk past each other without realizing that neither one of them really actually understood what the other was saying," White says. "But this is a good example of fairly disparate groups collaborating.

"Both sides are getting a lot out of it. Chuck has a much more comprehensive tool for analyzing his data, and we get access to a really cool set of data that we can work with for projects that we're progressing on now that this initial paper is published."