THE ENGLISH PHYSICIAN Edward Jenner achieved his first major scientific recognition after describing the life cycle of the cuckoo bird. But it was the publication of another work in 1798, 10 years later, that would become his legacy and one of medicine’s best ideas. By inoculating humans with a virus found in cows, he was able to trigger a protective immune response against smallpox, resulting in the first demonstration of vaccination. Vaccines for polio, rabies, typhoid and cholera followed, and today the vaccines for measles and five other diseases prevent as many as three million deaths a year, according to the World Health Organization.

Yet as essential as vaccines are, the process of creating new ones has never been quick. It took nine years from the time that the measles virus was first isolated in 1954 until the licensing of the first commercial vaccine. Manufacturing also takes time, and the measles vaccine today is still made much as it was back then—by culturing a large quantity of live virus, and then weakening it by growing it in eggs. That “attenuated” virus, like Jenner’s cowpox virus, doesn’t cause the disease, but it prompts the immune system to produce antibodies and activate T cells, imprinting a lasting “immune memory” that allows the body to respond quickly when it encounters the real thing.

Vaccines made in this time-tested way are highly effective against known threats. But older methods of developing vaccines are no match for a host of emerging—and reemerging—pathogens that call for a tailored and speedy response. The 2015 breakout of an obscure virus called Zika in Brazil “came out of nowhere,” says Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases (NIAID). Last year the Ebola virus roared back with a vengeance, so far taking more than 2,000 lives in the war-torn Democratic Republic of Congo after killing more than 11,000 people earlier this decade. Its comeback prompted the Emergency Committee of the World Health Organization to declare a global health emergency in July, the fifth such declaration since the first emergency committee was formed in 2005.

Fauci points to several factors that have fueled these and other epidemics. Human encroachment on otherwise pristine environments increasingly brings people into contact with animals and their diseases. “About 70% of all new infections that attack people originate in animals,” he says. Floods and other natural disasters, intensified by climate change, displace about 25 million people every year and create conditions that may foster outbreaks. War and civil unrest displace roughly 16 million refugees annually, sending many of them to crowded, unsanitary camps that can become disease incubators. When an outbreak of a new pathogen happens, there’s little time to go through the traditional steps of vaccine development—isolating an infectious agent, growing sufficient amounts to make a batch of vaccine and culturing it long enough to render it safe for humans. By then, says Fauci, “the outbreak will probably be out of control.”

Against this backdrop, infectious disease research has taken on a new urgency, leading to an emerging paradigm for developing vaccines. Rational vaccine design, also known as synthetic vaccinology, uses twenty-first-century technologies to modernize an eighteenth-century idea—replacing the “natural” approach of traditional vaccines with those that are “rationally” engineered to elicit a particular immune response. It begins with the genetic sequence of the pathogen that has been collected in the field. That data can be emailed to a lab, eliminating the risk of shipping and handling infectious material. Aided by a level of computing power unimaginable when the measles vaccine was developed, scientists then are able to identify antigens, the parts of a virus or bacteria that will trigger an immune response, and then create optimized copies of them. These synthetic antigens go into vaccines that can be delivered in new ways—as a nanoparticle, an engineered virus or a snippet of genetic code. The entire process can produce vaccine candidates within months, versus the many years needed by approaches used just a few years ago. In this new age, says Fauci, “forget about growing anything in eggs.”


BEFORE RESEARCHERS CAN DESIGN a new vaccine, they have to know what they’re up against, and that means identifying infectious culprits that may cause a pandemic. Because so many pathogens originate in animals, surveillance efforts have largely focused on areas that are home to many species and where computer modeling suggests a heightened risk. This has given rise to an increasingly sophisticated network of “listening stations” around the world. USAID’s PREDICT project, for example, has monitored hot spots in some 30 countries in Africa, Asia and Latin America since its founding in 2009. By analyzing samples from animals and people who come in contact with them, PREDICT has already discovered hundreds of viruses that could cause disease in humans.

Today, infectious agents are almost always identified by sequencing their genomes. Starting in 1999, when genetic data for the West Nile virus became available during an outbreak, sequencers have become standard diagnostic tools, even at remote clinics on the front lines of new outbreaks. Using computational tools, pathogen hunters scan samples of water, soil or food to detect potentially dangerous pathogens before they infect humans.

While older field tests could detect just a few known pathogens, new tools can help scientists who don’t even know what they’re looking for. VirCapSeq-VERT, for example, is a custom sequencing system developed by W. Ian Lipkin, a professor of epidemiology, pathology and neurology at Columbia University’s Mailman School of Public Health and Vagelos College of Physicians & Surgeons. It works as a universal virus detector, which means that clinicians can take a sample of a person’s blood and detect the genetic signature of virtually any virus known to infect humans and other vertebrates, as well as from novel viruses loosely related to known ones.

Now Lipkin is working on tools that can also allow researchers to assess qualities about a virus that might make it harder to treat and vaccinate against, and has built sequencing systems that detect bacteria and genes for virulence and antibiotic resistance. Lipkin says the new tools will be able to provide a report within hours of obtaining a sample, helping researchers “recognize a threat and appreciate it in its full complexity.” On the receiving end of this information from the field might be someone like Mark Poznansky, director of the Vaccine and Immunotherapy Center at Massachusetts General Hospital. In 2014 Poznansky led a multi-institution project called VaxCelerate II that looked into Lassa, a virus that causes hemorrhagic fever in humans. His team received Lassa’s genomic sequence by email and developed a vaccine candidate within just 90 days, a feat that offers a kind of step-by-step playbook for rapid rational vaccine design.

The first task for the MGH researchers was to “decode” the digital version of the Lassa genomic sequence, which meant determining the full set of proteins produced by the virus. Then, using bioinformatics tools that search databases on virus biology and protein structures, they started to home in on specific antigens that could be used in a vaccine to trigger the immune system.

Traditional vaccines contain lots of antigens because they include whole pathogens in killed or attenuated form. Many of these structures serve no useful purpose in the vaccine, however, and some can even trigger a potentially dangerous phenomenon called antibody-dependent enhancement, in which antibodies responding to a virus actually help it enter a cell or replicate. Rationally designed vaccines aim to be more selective, presenting the immune system with just a few select proteins or fragments of proteins that stand in for the whole virus. In this case, after determining which antigens to focus on, Poznansky’s team identified specific features on the protein’s surface—called epitopes—that immune system cells were most likely to recognize and attack. Then the researchers chemically synthesized antigens that mimicked those features.

Compared to vaccines made of whole viruses, these rationally designed vaccines are safer to manufacture because there’s no need to work with infectious agents. The process is also faster, sidestepping the long process of weakening the virus in an egg medium. The main question is how best to deliver the synthetic antigens into the body. In the case of the Lassa vaccine, the researchers chose to inject them in the blood as a mixture of proteins that would assemble themselves into free-floating chains that would, in turn, trigger the production of Lassa antibodies.

The researchers in Poznansky’s lab then tested the vaccine, first in human cell cultures and later in “humanized” mice engineered to generate human-like immune responses. Testing an immune response can be complex, involving many types of cells, and until recently, this was done mostly through flow cytometry, a method in which cell samples are suspended in fluid, labeled with fluorescent markers and injected in a device that illuminates them with a laser beam, one cell at a time. Analyzing the intensity of each fluorescent marker allows scientists to measure up to 20 aspects of a cell simultaneously, at a rate of thousands of cells per second.

In an effort to maximize the information gained from samples, researchers in the Vaccine and Immunotherapy Center used a newer method, mass cytometry. For this method, the same antibodies that are usually fluorescently tagged are instead labeled with heavy metal ions, which can provide an extremely detailed analysis in a single experiment. Using a tiny blood sample from a humanized mouse, for example, mass cytometry can measure more than 40 parameters simultaneously, showing how a test subject’s entire inflammatory and immune system is responding to a vaccine.

Mass cytometry produces “hundreds of thousands of data points,” says Poznansky, and making sense of that massive amount of data can be challenging. To interpret these datasets, Patrick Reeves, senior scientist and project leader in the Vaccine and Immunotherapy Center, and his team are using new visualization tools that show mass cytometry data, and other forms of big data, as “architectural layouts” of the immune and inflammatory response. Some visualizations resemble abstract art, with clusters of colors indicating surges of immune cell activity and expression of immune and inflammatory molecules. Others resemble mandalas, with the spokes and concentric rings of a wheel and squares of color showing activation of particular immune system components. “Patrick’s images let us see multiple aspects of a mouse or human immune response to a vaccine or infection over time,” says Poznansky.

The VaxCelerate II Lassa vaccine platform has been licensed to a biotechnology company, which means that it could soon be ready for use in high-risk areas. And Lassa is not the team’s only success. Q fever is an animal-borne disease that can infect humans; it garnered attention when U.S. troops in Iraq and Afghanistan became infected with it when they inhaled bacteria from local livestock. In 2018, using a process similar to that used for Lassa, VaxCelerate produced a vaccine candidate for Q fever that is now being tested.


THE VAXCELERATE VACCINE APPROACH delivers antigen epitopes, tiny fragments of protein, directly into the patient’s body. But other new vaccines, including some of the first developed for Zika, take a different approach, using DNA or RNA with genetic instructions that allow the body’s own cells to manufacture the vaccinating proteins. To make these vaccines, scientists insert short DNA sequences into small circular molecules called plasmids—a sort of key ring to hang genes on. When the plasmids are injected into a muscle cell, that genetic code is translated, causing the cell to produce key proteins that trigger an immune response. One DNA vaccine for Zika is now being tested in an international clinical trial.

As efficient and direct as this method is, however, it does require one interim step. DNA relies on a molecule called messenger RNA, or mRNA, to forward genetic coding instructions to the cellular machinery that actually makes proteins. RNA-based vaccines skip that step by delivering mRNA directly into the cells. In several clinical studies, RNA vaccines have performed better than DNA vaccines. That may be because, unlike DNA vaccines, RNA vaccines don’t have to find their way into the cell nucleus, where DNA is transcribed. Instead, they need only penetrate the outer cell wall. Also, delivery via RNA greatly reduces the risk of its code becoming integrated in the patient’s genome, which could have unintended effects.

RNA vaccines have their own drawbacks, however. They tend to become unstable in the body and quickly lose effectiveness. To address that problem, researchers are trying new approaches, such as manipulating the RNA sequence to make it easier to store in the cell, or binding RNA to other molecules that protect it. These stabilized RNA vaccines also can be held for long periods at room temperature, making them attractive for use in countries in which refrigeration may not always be available. An RNA vaccine against rabies is currently in clinical trials, and other vaccines against influenza, HIV, tuberculosis and Zika are at earlier stages of development.

“Gene-based vaccines are potentially very fast and flexible,” says John Mascola, director of the Vaccine Research Center at NIAID. Both DNA and RNA vaccines spur the body to make antigens that closely resemble natural viral epitopes, adding to the vaccines’ effectiveness. Producing gene-based vaccines is also less expensive than making a traditional vaccine, and the new vaccines can be manufactured in quantity in a matter of weeks.

Gene-based vaccines can be delivered in other ways as well. Viral-vector vaccines use common human or animal viruses that have been engineered to be noninfectious and incapable of replicating. Because they have a real virus as a “chassis,” viral-vector vaccines are highly effective at infiltrating host cells and transferring their cargo of genes, which makes the cell produce antigenic proteins for as long as a few days. These vaccines also have inherent “adjuvant” properties—they trigger the production of cytokines, inflammatory proteins that spur a stronger response from antibodies and T cells, which contain and clear a pathogen—and they are relatively easy to produce in large amounts in industrial bioreactors. This is the approach used to make several current vaccines, including the first Ebola vaccine to be widely deployed, which uses a common pig virus as a vector. It was found to be safe and protective in a large 2015 study and has been used to vaccinate more than 200,000 people in the current outbreak in the Eastern Democratic Republic of Congo.

NEW WAYS TO DESIGN and manufacture vaccines also promise to transform the fight against a more familiar but no less deadly foe—the seasonal influenza virus, which infects tens of millions in the United States alone. While the flu is not a new disease, there are many strains of influenza, and they mutate quickly, making it hard to predict exactly what will show up in a particular flu season. If a vaccine is a poor match, there is seldom time to prepare a new formulation, which normally must be slowly cultured in eggs. Researchers have long hoped to circumvent this guessing game by creating a universal flu vaccine that would work for all or most flu strains.

Several universal flu vaccines are now in clinical trials, including Biondvax’s M-001, a peptide vaccine (similar to Poznansky’s Lassa vaccine) that contains nine viral epitopes common to 40,000 influenza viruses listed in the NIH database. In spring 2019, the NIH’s Vaccine Research Center Clinical Trials Programs started enrolling volunteers for the first in-human trial of its experimental universal flu vaccine, H1ssF_3928. Another—Medicago’s MT-2271—delivers key proteins from multiple flu strains in the form of virus-like particles (VLPs). A VLP is a type of nanoparticle that has a dense, repetitive arrangement of proteins on its surface. Its structure closely resembles that of real viral particles, so it often elicits a more robust immune response than other kinds of vaccines that isolate a particular antigen or epitope. There are already VLP vaccines approved for hepatitis B and for human papilloma virus, and VLPs can be produced rapidly through a variety of methods, including growing them in the leaves of tobacco plants.

If rational design—through nanoparticles, DNA vaccines or other novel means—can master the common flu vaccine, which is given 150 million times each year, the lives saved in the United States alone could be in the hundreds of thousands. “We’re certainly not there yet,” says Anthony Fauci, “but we’re way ahead of where we were a couple of years ago, and we are clearly making very positive progress toward a universal flu vaccine.” The discovery of that vaccine and others will be a major leap forward for one of medicine’s best ideas.