DARPA thinks it has a solution to Ebola (and all other infectious diseases)

Despite decades of work on Ebola and about six months of lead time while the latest outbreak was slowly gathering strength, the medical establishment did not have good answers about what might happen if the disease jumped the epidemiological containments that were put in place in the U.S. and elsewhere.

The Ebola outbreak looked a lot like the doomsday scenario that military planners at the Defense Advance Research Projects Agency had feared. They worried both about weaponized diseases and the national security crises that might ensue if a really bad disease swept over unstable areas of the world (not to mention the United States). To prepare for such a possibility, four years ago, they gave Colonel Wattendorf the funding to develop new types of vaccines that could be deployed much more rapidly than the current best technology. The requirements of the military meant he had to develop a radically different approach to infection protection.

A traditional vaccine is a lot like a practice run. Doctors inject an attenuated infectious agent—called an adjuvant— that allows the body to become familiarized with a virus or bacteria. In response, the body rummages around in the genome, which has blueprints for all kinds of antibodies, and starts to crank out defenders that can bind to and neutralize the disease. The process is so simple that people had figured out how to inoculate (or variolate, as it was known) themselves long before we understood what a virus was (which, for the record, didn’t definitively shake out until the 20^th-century).

In the last couple decades, a few teams have experimented with what have been called DNA vaccines. But these experiments have encoded instructions for attaching an adjuvant—a substance that enhances the effect of a given vaccine—onto a strand of DNA, which has been pushed into a body’s cells through highly technical processes. Once inside the body, the modified cells crank out the proteins that the DNA encodes, and the body sees the signs of the intruders and generates an immune response.

The DARPA program takes the normal kind of genetic vaccine and cuts out a few steps. Rather than injecting DNA with code for an adjuvant and hoping the body makes the right kind of antibodies on its own, Wattendorf wants to inject the code for the antibody itself, which would allow the body to do less immunological guesswork before starting to fight the infection. Essentially, instead of giving the immune system practice for the real test of an infection, this method would provide the right answer ahead of time.

Like many hot Silicon Valley start-ups, DARPA’s new vaccination method isn’t just a product: it’s a scalable platform. If it works, the method could work with any infectious disease, even the worst of the worst. And they had, in the Ebola outbreak, what Wattendorf described as a “live-fire test” of their emerging capabilities.

All they needed to test their theory was the blood of the Ebola survivors.

---

In a secluded and secret wing of Emory University Hospital, there are two rooms meant for only the sickest and most dangerous patients. The Serious Communicable Diseases Unit is one of four centers in the U.S. that were created to deal with the worst diseases that can be transmitted from human to human. Air there doesn’t circulate into the rest of the hospital. Biological specimens from the sick don’t leave either: laboratory tests are done right there. Doctors and nurses don special gear any time they go in. This is where four of America’s Ebola patients were treated.

It’s also where Colonel Wattendorf’s program could get the antibodies they needed to test their platform. He reached out to the doctors taking care of the Ebola patients — Rafi Ahmed, director of the Emory Vaccine Center, and Aneesh Mehta, MD, assistant professor of medicine at Emory University School of Medicine — and asked them to participate in the research. DARPA has since funded them and a coalition of other top researchers with more than $10 million to characterize the body’s immune response to Ebola.

The doctors drew a vial of blood from each of the four survivors, eight milliliters apiece, not even enough to fill up a shot glass. And they began looking for the antibody that could prevent an Ebola pandemic.

Imagine an antibody. It would look like this Y, but in three dimensions.

The stick of the Y attaches to the outside of cells that make antibodies and the V part of the Y is what attaches to things attacking the body, attempting to neutralize them.

Antibodies are proteins, long strings of amino acids that act as the body’s tiny machines. They’re produced by B-cells. Each B-cell only makes one kind of antibody. When a new B-cell comes into being, it is called “naïve,” which means that it doesn’t yet make any particular kind of antibody. Over the course of a few hours, the B-cell will imprint on a particular infectious agent based on what’s floating around the body. Then, using the different codes for antibodies that reside in our genomes, they start cranking out antibodies of a particular shape. To do so, the body transcribes the code from DNA into single-stranded RNA. That messenger RNA travels out to the ribosomes—the cell’s factories—which then build the antibody proteins, these intricately-folded machines with highly specific shapes for binding onto invaders.

Each B-cell might make a different antibody, but what the DARPA program is after is the best antibody. Just a few years ago, finding one B-cell and sequencing the antibody it was making would have been impossible. But the agency has developed what are known as “microfluidics techniques” that make the sorting possible. Once the B-cells are isolated, they are dropped into one of the indents on a 96-well plate, where the experiments can begin. They can read the sequence for the antibody and bring it into contact with Ebola virus to see how effective it is at fighting the disease.

In the DARPA program, Wattendorf told me that they selected the top 20 antibodies to do further research on. Your body might end up producing one of these exact same antibodies some day in the future, after a doctor injects you with the genetic instructions for making them.

---

There is a beautiful simplicity to the idea of using good antibodies to heal people or prevent disease. But almost all previous efforts have focused on finding a good antibody and then manufacturing lots of them at a factory. After many years of research, pharmaceutical companies can now produce large quantities of antibodies in 20,000-liter vats of Chinese hamster ovary cells and also, more recently, in genetically modified tobacco plants.

This technique became known by the shorthand of monoclonal antibodies (because all the antibodies are clones of the good antibody) and it was once one of the hottest topics in biotechnology.

But human bodies are complex. For all the elegance of the theory, in practice, very few of these kinds of medications have come onto the market. And even for those that have made it through FDA trials, producing antibodies in hamster ovary cells is expensive, takes a year to get running, and requires a perfectly functioning logistics chain. None of these things make the process a fit for DARPA’s plan to slow or stop infectious disease outbreaks.

If we don’t use our bio-industrial facilities to produce the antibodies, however, then researchers have to find a way to harness our own body’s cellular machinery. And that means finding a way to get outside genetic material into cells, where they can work with the cellular machinery there to pump out antibodies.

That’s where the work on DNA vaccines comes in. For 25 years, researchers have known that if they inject DNA containing the genetic code of a virus into mammals, they can induce an immune response. And Wattendorf knew a lot of this work could be ported into his new vaccine program.

He tapped DNA vaccine specialists David Weiner at the University of Pennsylvania and David Ho at the Aaron Diamond AIDS Research Center, an affiliate of Rockefeller University, for help with the DARPA program. Weiner has been working on the problem since the very beginning, and has seen the hype around DNA vaccines peak and then collapse when there were disappointing clinical trials. He’s written most of the scientific reviews of progress in the field and guest-edited a special edition of the journal Vaccine on DNA vaccines.

Weiner says that the techniques for getting genetic information into cells has improved a lot since the failures of a decade ago. They’ve figured out how to get much better uptake of DNA into cells, which leads to a much better immune response. He credited the company Inovio, for whom he is a paid consultant, for coming up with a new technique for driving DNA into cells. “They have really dramatically improved those delivery systems for humans,” Weiner said. “It has really simplified the process and increases our ability to express a gene a thousand-fold.”

The new technique, called “electroporation,” works by delivering millisecond-long electrical impulses to a cell, which opens the cell membranes up to make DNA injection easier. “The brief electrical pulses are computer designed, so they not only open the pores of the cell membrane,” Weiner continued, “but they also change direction to move the DNA through these very brief and very small openings DNA can go through.” In essence, the electrical pulses help the DNA shimmy through the complex chemistry of the cell membrane and into the nucleus where it can be used.

Last week, the NIH announced that Weiner’s lab will be the lead research group on a $16 million program to work on HIV with their techniques. “DARPA is not the only group that’s been funding this research now that things have turned around,” he told me.

Other groups are working on injecting RNA that’s been modified to last longer in the body than it normally does. The RNA is then packaged up in lipids, which can wiggle into cells and be used to directly produce the proteins we call antibodies.

These new techniques seem to be working, at least in animals. “We have data with replicating RNA and DNA that we can have four months of [antibody] expression in large animal models,” Wattendorf said. T

---

DARPA is a place for high-risk, high-reward research. And, as anyone in Silicon Valley will tell you, its importance to the modern-day technology industry can’t be understated. DARPA catalyzed research on self-driving cars, robots, brain implants, laser weapons, and of course, the Internet itself.

Wattendorf believes his program can succeed because of the very structure of how DARPA works. The key hurdles, Wattendorf argues, aren’t scientific, but institutional. And DARPA can jump over and around them. For the overall DNA vaccine program to work, it must combine many disciplines and subfields in new ways. As Wattendorf described his work to me, he talked about how the people who work on monoclonal antibodies don’t get together with the people who work on anti-infectives who don’t get together with the vaccine people who don’t get together with the Ebola specialists. Lots of interesting technologies don’t get applied outside their local domains. And that’s why DARPA could play a fascinating role, by serving as a central magnet that draws all of these entities together on single projects. The key skill for preventing diseases, according to Wattendorf’s theory, isn’t brilliant flashes of House-like biological insight, but practical institutional political knowledge. (Well, that, and access to the government’s deep pockets.)

It may be that Wattendorf’s dreams never pan out. DARPA has never tested DNA vaccines on humans, which means they are a long way off from the kinds of clinical trials that would allow people to actually acquire these kinds of medicines. The work I saw at the DARPA conference and that Wattendorf described to me is not peer-reviewed, and few outsiders have seen the work.

Dimiter Dimitrov, a senior investigator at the National Cancer Institute who focuses on antibody engineering and vaccines, said that he was familiar with Wattendorf’s program and found it “highly promising.” Two other post-doc researchers who work in adjacent fields also reviewed what Wattendorf told me and found it elegant and interesting.

But University of Maryland virologist Alan Schmaljohn, who has worked on Ebola and other emerging viruses for decades, cautioned against getting too excited about the novelty of Wattendorf’s idea. “To someone ‘skilled in the arts,’ most of the ‘new’ ideas occurred to many of us long ago,” he told me in an e-mail exchange. “There is often novelty in the newest technical approaches, and (fueling an aging scientist’s skepticism) even more often there are obligatory claims of novelty (to secure funding and notoriety).”

Schmaljohn cited a basket full of other research groups who had tried to do similar things as the DARPA program, mostly working on HIV, and beginning in 1996. They generally called it “vectored immunoprophylaxis,” and it is certainly an area of active research among a small number of labs.

But Schmaljohn noted that there were several reasons for why “most of us didn’t pursue the idea when its obviousness first occurred to us.” First, it’s difficult to achieve “a sufficient dose of antibody.” It’s a math game: X number of cells have to begin producing Y number of antibodies. And it could be difficult to reach good numbers. Second, viruses are mutate quickly, and develop variants that can give antibodies the slip, so “expressed antibodies can rapidly become useless.” And finally, he wondered about the safety of the antibody expression, especially if there was no built-in “off switch” for when the body should stop producing those antibodies.

Wattendorf admits the technical challenges are formidable, but his presentation tried to pre-empt each critique. To the first objection, he says that the electroporation that Penn’s David Weiner describes above can increase the uptake of genetic material in order to produce a sufficient dose of antibody quickly. The second problem might be accounted for in the design of the program: Wattendorf isn’t trying to protect people from disease forever, as you might with HIV, but rather for a few months. And that’s one reason he’s interested in using RNA—which can be engineered to degrade within a few months—instead of DNA.

“We do not want permanent gene therapy producing an antibody against Ebola,” Wattendorf told me. “That would not be a possible safe solution. We want an ephemeral, transient system.”

DARPA alone can’t build these technologies from the ground up—most biotech is way too expensive to be a solo project—but they can evaluate the potential of recombining extant ideas into new programs and then fund that exploration. It’s a risky investment, but if it works, DARPA gets the ability to protect its troops from naturally occurring or weaponized viruses and the rest of us get a bulwark to stave off the apocalypse.

There’s little doubt that the world will face a pandemic in our lifetime, something at least as bad as the 1918 influenza outbreak, which killed somewhere between 20 and 40 million people in a single year. (For comparison, HIV/AIDS has killed about 40 million people in the last 35 years.) We already share our diseases globally. And Wattendorf’s system would extend our defenses to the global scale, too. We could have shared cellular machinery working across all of humanity.

That wouldn’t just be an academic achievement—it would be a profound humanitarian good, one that that could save the lives of people working on the front lines of communicable disease. Last week, after Colonel Wattendorf had finished spinning his vision of an Ebola-free future, I heard about another health care worker who had contracted the disease and was flown back to Maryland for treatment. Hearing about that case, my first thought was my son. As a father, the most horrible psychological component to Ebola isn’t just the disease. It’s the thought of not being able to touch your child if they contracted the disease, for fear of contracting the disease yourself through bodily fluids. Or maybeyou would touch them and risk your own life. No one should face such a horrible choice, and someday maybe no one will have to.

Wattendorf knows it’s not going to be easy. He finished scribbling a series of diagrams to illustrate how his program works, and I told him how I worried that computer metaphors made it seem as if biotechnology was easier than it actually is. He agreed.

“There are a lot of things here. Discover an antibody—that’s hard. Deliver an antibody—that’s hard. And do it in a way that’s gene therapy, but not gene therapy, for long enough and high enough—that’s hard,” he said. “It’s easy to say all this, but actually linking all these steps together and making all those handoffs, and doing it in 90-120 days, is what makes it a DARPA problem.”