By Megan Bennett
Albuquerque Journal, N.M.
WWR Article Summary (tl;dr) Bette Korber, a theoretical biologist at Los Alamos is truly changing the world with her work to create a vaccine to prevent HIV.
Bette Korber had to fight intense pushback when she first proposed her idea for an HIV vaccine designed on a computer.
The theoretical biologist at Los Alamos National Laboratory was trying to garner support for her “mosaic” vaccine design. Her idea was that with access to hundreds of thousands of different HIV sequences from around the world, a computational code could design synthetic proteins to fight the virus.
Her method “evolves” sequences to solve a particular problem — with a small set of proteins, can science develop the best possible immunological coverage for the global diversity of HIV? In other words, can a vaccine be created that targets HIV generally, despite its massive number of variations?
“That’s the problem I’m trying to solve,” said Korber in during a recent interview in her Los Alamos office.
Skeptics initially didn’t believe that computer-designed, man-made proteins would work. “People just didn’t believe they would fold properly or be stable,” she recalled about those early doubts. “They thought they would just be dead sequences.”
She had the idea for a mosaic as early as 2002. But the first two attempts to fund the project were rejected.
Reviewers thought it was “insane,” she said.
The team finally got the green light the third time around. With funding to further develop the code, a study on the design was published in 2007.
More than a decade later, the mosaic vaccine has reached the rare stage of being tested for efficacy in humans. The work is being supported by Janssen Pharmaceutica, Johnson & Johnson’s pharmaceutical company, The Bill and Melinda Gates Foundation, and the National Institutes of Health.
Within the past year, Korber has received wide recognition for her contributions. In October 2018, she was named R&D Magazine’s Scientist of the Year. She was also awarded LANL’s 2018 Richard P. Feynman Innovation Prize.
Korber has spent her entire career thinking about HIV and the complex diversity of its virus. Different strains of HIV are found around the world, which is why finding a globally effective vaccine proves difficult.
As the virus evades the body’s immune responses, Korber said, it consistently changes within every infected person’s body.
Her desire to start studying HIV was a personal one. Her best friend and former housemate when she was studying immunology at California Institute of Technology, physicist Brian Warr, was diagnosed with AIDS. He died in 1991.
“And there’s nothing you could do at that point,” she said. “There were no drugs, no therapy, there was nothing. You just had to watch the disease take the people you love.”
Korber came to LANL in 1990, after a few years of post-doctoral work at Harvard, for the opportunity to focus more on the theoretical side of biology. Her mentor, the late Gerry Myers, started the lab’s HIV sequence database — the first-ever sequence database for a pathogen — in 1987. She described Myers as a leader in “defining how HIV evolved and how diverse it was.”
Today, more than 850,000 sequences have been inserted and annotated into the database.
“There wasn’t even a web when he started,” Korber remembered. “It was a book. And people had it chained to their desks so no one would steal it, because it was so valuable to be able to look at it.”
Korber “inherited” the sequence database and combined it with her own immunology database that tracks the body’s various immune responses. The site and its interactive tools are accessible to scientists worldwide.
The database is the foundation for all of her team’s work, she said, including the mosaic vaccine. To develop the code more than a decade ago, she partnered with former LANL machine learning expert Simon Perkins.
The code pulls data from the database’s natural sequences to generate artificial sequences. The proteins designed with the mosaic code undergo hours of recombination and evolution to come up with something with optimal coverage for all types of HIV, she explained.
“It evolves the sequences on the computer and selects for the things I want to select for,” Korber said of the code. “So they look and feel like HIV sequences, but they’re not. They don’t exist in nature.”
She also emphasized that mosaic is just one of many vaccine concepts she and her team at LANL are working on.
“It’s just that mosaics are the furthest along,” she explained. “We have a bunch of things going along in parallel that might help, and many other groups have other ideas that may help. We don’t know if any of them will work or if some will work. That remains to be seen.”
Korber described herself as just one small part of the massive government-supported effort in the search for an HIV vaccine. The NIH’s Center for HIV/AIDS Vaccine Immunology & Immunogen Discovery funds long-term research out of Duke University — which she works with — and Scripps University.
And her efforts expand beyond HIV to other “variable viruses.”
A study she co-authored about a vaccine that protected mice from sequences of the Ebola virus and Marburg, a closely related virus, was published at the end of February. Like HIV, those viruses can have different sequences, depending on the outbreak and the species they come from.
“We don’t know why these things move into humans,” she said. “We don’t know what happens. We don’t know what’s lurking, (or) what could happen next year. So we wanted to get something that would cover everything that we know about.”
Along with colleagues, she created a database with all of the Ebola and Marburg sequences over the past 50 years.
And with the help of her husband, fellow LANL scientist James Theiler, she developed a “second-generation” mosaic code. It uses different mathematics and generates results more quickly. In a trial, all but one of the 32 mice that didn’t receive a vaccine — which was developed by a colleague at Oxford — died.
“Every mouse that got the vaccine did fine,” she said of the Ebola and Marburg study. “No symptoms at all. That’s what we were aiming for and it worked — in mice. That doesn’t mean it’s going to work in monkeys or humans. But it’s an important hurdle.”
‘Promising’ so far
Over the past 35 years in the global fight against HIV, the mosaic-based vaccine is just the fifth concept to reach the stage of clinical trials in humans, according to Dan Barouch, a professor at Harvard Medical School and the director of the university’s Center for Virology and Vaccine Research.
Barouch has been working with Korber’s team for the past decade on a variety of projects, including mosaic. He described the technology as an important new concept because of its computational approach for attacking a pathogen with “intrinsic variability.”
After the mosaic design study was published in 2007, the sequences were sent to collaborators at Harvard and Duke. The vaccines worked well in small animals. Then macaque monkeys were given the vaccine in what Korber called a “make or break moment.”
“The immune responses the macaques made to these mosaic vaccines were nicely cross-reactive,” she said, meaning it worked against varying strains of HIV.
The next step was to test the safety of the vaccine in volunteers and see if it caused similar immune responses in humans as seen in monkeys. The studies showed both, she said.
“The effects we were seeing in pre-clinical studies and clinical studies are largely what we had hoped with the promise of the technology,” said Barouch.
In November 2017, the human efficacy stage of its clinical trials began in Africa. Results are expected around 2021.