CAMBRIDGE, MASSACHUSETTS—In a recent morning meeting of scientific leaders at Moderna Therapeutics, conversation swerved toward the philosophical. Biochemist Melissa Moore, recently hired to head RNA research at the Boston-area biotech, had something on her mind: hype.
Specifically, she was thinking about Gartner's hype cycle, a glib model cooked up by an IT research firm, in which every new technology ascends a "peak of inflated expectations," sinks into a "trough of disillusionment," then climbs the "slope of enlightenment" to reach a "plateau of productivity." Where on this curve, she wondered to Moderna's president, Stephen Hoge, was their technology?
The question is apt. Moderna was founded on the idea that messenger RNA (mRNA), the molecule that relays genetic instructions from DNA to the cell's proteinmaking machinery, could be re-engineered into a versatile set of drugs and vaccines. These strands of instructions could teach our cells to make whatever was needed to treat or prevent disease—virus-slaying antibodies, wastegobbling enzymes, heart-mending growth factors. The willingness of pharmaceutical giants and investors to bet on that premise to the tune of nearly $2 billion has unleashed waves of both hype and skepticism.
Moderna has shared little detail in published papers about the technology it's developing, though there are clues in its abundant patent filings. Until recently, even the targets of drugs already in clinical trials weren't publicized.
But as more trials get underway, Moderna is gingerly opening up. The company agreed to Science's request for access to some of its researchers and labs over the past few months. And last month, at the annual J.P. Morgan Healthcare Conference in San Francisco, California, CEO Stéphane Bancel unveiled Moderna's first round of drug candidates, which include vaccines for Zika and flu, and a therapy for heart failure.
Expectations are high. Being a startup valued at more than a billion dollars—an anomaly that venture capitalists dub a unicorn—comes with scrutiny, and many wonder whether Moderna's pipeline, consisting mostly of vaccines for now, will expand to match the company's original vision of mRNA as a broad treatment platform. "There were a lot of really big promises made," says Jason Schrum, a biotechnology consultant in San Francisco and a former Moderna employee. "That's what people latched onto; they want the promises to be true, and they want to see the investment really turn it into something meaningful."
In other words, the trough of disillusionment, if it's still ahead, threatens to be deep.
Hacking the kingdom of life
The vision of an mRNA drug has beguiled scientists for decades. "It's a huge idea," says Michael Heartlein, who heads mRNA research at a competing biotech called RaNA Therapeutics just a few blocks away. "Any protein target [where] you can think of a potential therapeutic, you can approach that with mRNA." The single-stranded molecule sets up a temporary protein factory outside a cell's nucleus and attaches to ribosomes. This cellular machinery translates its sequence of four kinds of nucleosides—adenosine, cytidine, uridine, and guanosine—into a protein. Then it degrades within a day.
Assembling these chemical instructions could be a faster and more adaptable way to make drugs than manufacturing the individual proteins themselves in large bioreactors. And it would allow scientists to deliver proteins that act inside cells or span their membranes, which are a challenge to introduce from the outside. An mRNA drug would also be easier to control than traditional gene therapy. Like mRNA, gene therapy can induce cells to make therapeutic proteins, but it typically introduces DNA that can integrate unpredictably into the genome.
If you can hack the rules of mRNA, "essentially the entire kingdom of life is available for you to play with," says Hoge, a physician by training who left a position as a health care analyst to become Moderna's president in 2012. Adjusting mRNA translation to fight disease "isn't actually super high-risk biology," he adds. "It's what your genes would do if they were rational actors."
One key problem, however, is that our bodies would normally destroy incoming mRNA before it could get cranking. It's a relatively large molecule that is prone to degradation, and as far as our cells are concerned, it's supposed to come from the nucleus, where it's transcribed from DNA. RNA invading from outside the cell is the hallmark of a virus, and our immune system has evolved ways to recognize and destroy it.
Biochemist Katalin Karikó heard this argument over and over as she tinkered with mRNA in her University of Pennsylvania (UPenn) biochemistry lab in the early 2000s. But she and her UPenn colleague Drew Weissman found a way to tame cells' typical inflammatory response by modifying one of mRNA's four building blocks, uridine. Assembling mRNA using pseudouridine, a nucleoside variant that occurs naturally in the body, greatly reduced the tendency of immune sentinels called dendritic cells to shoot out inflammatory molecules in response, they reported in 2005.
In mouse studies, this mRNA proved stable enough to stick around in the body and make proteins. Karikó and Weissman founded a company hoping to develop drugs from the discovery, and won nearly a million dollars in small business grants from the U.S. government for animal studies. But shortly after the money came through, Karikó says, the university sold the intellectual property license, and the effort never reached clinical trials. "I could not find any ear," she recalls, "someone that would say, ‘Oh, let's try it.’"
But when stem cell biologist Derrick Rossi's team at Boston Children's Hospital used pseudouridine-containing mRNA to encode proteins that transformed mature cells into stem cells, he found quite a few ears. Serial entrepreneur Robert Langer of the Massachusetts Institute of Technology (MIT) and Noubar Afeyan, CEO of the venture capital firm Flagship Pioneering, both in Cambridge, saw the makings of a whole new class of drugs—and the idea of Moderna was born.
The company, which launched operations in 2011 with Flagship funding, quickly set its sights on new (and patentable) nucleoside modifications that would provoke an even smaller immune response than pseudouridine. "This stuff was working a little bit," says Hoge, "so why not make it work a lot?"
Initially operating in "stealth mode"—without announcement of its existence—Moderna's team screened mRNA assembled from various modified nucleosides and hit on one called 1-methylpseudouridine. It bore a chemical "bump" that the team suspected kept it from locking into key receptors on the surface of immune cells.
As the data flowed in during 2011 and 2012, Bancel, who had come to Moderna from the French diagnostics company bioMérieux, began to work up a pitch. He was catching potential investors at an inauspicious time: Many were smarting from disappointing trials of RNA interference therapies, which use short, double-stranded RNA to disrupt the production of disease-causing proteins. "No one had cracked how to make RNA stable enough to be a therapeutic," says Mene Pangalos, who heads the Innovative Medicines and Early Development Biotech Unit at AstraZeneca in Cambridge, U.K.
Bancel showed Pangalos and his team two studies in which an injection of modified mRNA containing pseudouridine prompted nonhuman primates to express two human proteins. Among dozens of mouse studies, he presented work led by Moderna Co-Founder Kenneth Chien, then at Harvard Medical School in Boston, showing that mice recovering from induced heart attacks survived longer and had stronger hearts when injected with mRNA encoding a protein that drives blood vessel formation—vascular endothelial growth factor (VEGF).
"That got us excited," says Pangalos, who was eager to build up AstraZeneca's pipeline of cardiovascular drugs. "It was incredibly high risk. It was untried and untested." But if it could work for one disease, it would likely work for many. Changing the disease target didn't require developing or identifying a whole new drug, just altering the mRNA sequence. And although many of the initial animal studies used mRNAs with pseudouridine, Moderna's new chemistry was already starting to outperform that first generation in rodent studies. "I don't think it was such a stretch to imagine the technology would continue to improve, given what they were doing," Pangalos says. In March 2013, a few months after Moderna announced itself to the world, AstraZeneca put an up-front $240 million into a partnership to pursue up to 40 drug candidates using Moderna's technology.
Schrum, who led early chemistry research at Moderna and made some of the discoveries behind its initial patents, had left the company by the time the AstraZeneca deal was sealed. To him, the sum was astonishing, given the preliminary findings he had seen. "There was a lot of excitement that this [technology] can be applied to anything, and that this is a panacea," he says. Before meetings with potential investors and partners, he remembers the Moderna team being "frantic to get some sort of data, just general data, without a whole lot of specifics attached." Winning those early investments, by his estimate, "comes down to salesmanship."
Moderna's bold premise inspired headlines comparing it to a young Genentech, the most famously successful of all biotechs. Bancel, meanwhile, insists that he never hyped the company. "We never said, ‘Oh look at mRNA; we're going to cure 2 million diseases.’ No, we said, ‘What if? What if this could work?’" But as more cash poured in—$100 million from Alexion Pharmaceuticals to pursue rare diseases, $100 million from Merck for a set of antiviral drugs—the image of Bancel as a brash newcomer with a crisp suit and an audacious pitch became part of the company's mystique.
Afeyan at Flagship, who recruited Bancel, calls such a portrayal irrelevant "social science" that gives Moderna's technology short shrift. "There is real science here," he says. "There's real data, there's real molecules."
Moderna now has more money to throw at those molecules than most biotechs can dream of, though it's far from the only group chasing mRNA drugs. The German biotech CureVac, for example, has brought mRNA-based vaccines for rabies and cancer to clinical trials, and Karikó now heads a research team at BioNTech in Mainz, Germany, that focuses on mRNA-based drugs.
But few companies have delved into nucleoside engineering the way Moderna has, or pursued such a broad range of diseases from the start. Beyond its $100-millionper-year platform research, Moderna runs four wholly owned ventures focused on drugs for infectious diseases, rare diseases, immuno-oncology, and personalized cancer vaccines. It has about 430 full-time employees, spilling across three buildings around biotech-dense Kendall Square. Higher-ups are identified by black-and-white headshots hanging at their office doors.
Lavish funding has allowed Moderna to set up production facilities that can manufacture more than 1000 new, made-to-order mRNA a month. ("Moderna has probably made more RNA by in vitro transcription than all of humankind ever," quips Edward Miracco, a senior scientist on its process innovation team.) And it has allowed for many parallel animal experiments to characterize different mRNA and select the most promising. "If you need to run a 25-arm experiment, just do it," Bancel recalls telling his team. "We have the money, we have the infrastructure. Just do the right science."
We've had failures. We've gone down blind alleys. But because we've been quiet about it, nobody's seen that.
It has taken a lot of science to make mRNA act like a drug. Some of Moderna's most promising early candidates, although they could tiptoe past the immune system, produced underwhelming amounts of protein in animal studies. The same nucleoside modifications that made mRNA more stealthy also made it less recognizable to the ribosome. "If you're trying to sneak in there and make a thing, you have to look pretty darn natural," Hoge says. Moderna needed to figure out what features of naturally occurring mRNA were most important for translation, and how to restore them.
By the summer of 2013, word of the company's ambitions was wafting through academic labs, including Melissa Moore's at the University of Massachusetts Medical School in Worcester. Moore had spent her career studying the intricacies of how nascent mRNA gets spliced in the nucleus and loaded with proteins to become a complex known as a messenger ribonucleoprotein (mRNP). Over those years, she had also grown frustrated by how many more male than female scientists held consulting roles at biotech companies. When a colleague told her about Moderna, she decided to go out on a limb.
"Although we have many common connections, I don't believe you and I have ever met," she wrote in an email to Tony de Fougerolles, who was then Moderna's chief scientific officer. "I am arguably the world's expert on how the synthetic history and protein complements of mRNPs contribute to gene expression." Maybe, Moore suggested, her knowledge could improve Moderna's product. "I remember going home and being emotionally depleted, because I had completely just put myself out there," she says. "I had never done anything like that before, but I knew I had to do it."
De Fougerolles invited Moore to give a seminar, which led to a sponsored research agreement, and, eventually, a position on the scientific advisory board. Last year, Moore left her tenured position to become chief scientific officer of Moderna's research platform. "I could have spent the next 15 years turning the crank, putting out more papers, training more students," she says, "but when I'm 80 or 90 and I look back at my life, I would regret that decision."
Moore's academic work has advanced a counterintuitive theory about mRNA. It might seem that secondary structure—the folds and loops caused by bonding between nucleosides in the strand—should hinder protein production. Too much structure could force the ribosome to do extra work untangling the strand or even stall translation altogether. But findings in Moore's lab supported the view that mRNA strands with more of the nucleosides that tend to form tight bonds are, in fact, easier for ribosomes to translate.
The bioinformatics team at Moderna was making parallel discoveries. Even between mRNAs with the same sequence, they were finding that different modified nucleosides produced different amounts of protein. And nucleosides with a tendency to form tighter structures were more productive. The team knew that the frequency and placement of the modified nucleosides in the strand changed how it folded, and hence how it interacted with the ribosome. And because trillions upon trillions of different nucleoside sequences can code for the same protein, there were plenty of ways to engineer more efficient ones—providing they could be predicted.
Doing so took the Moderna team deep into the structure of mRNA. To model how single-atom changes affected bonding between nucleosides, they enlisted a quantum chemistry expert, Michelle Hall. "When I started looking for industry jobs, people were like, ‘Oh that's adorable. Nobody does that in industry,’" Hall remembers. "Turns out, not true."
Her calculations informed an algorithm that predicts, for a given protein, what mRNA sequence would produce the structure most appealing to a ribosome. Across many drug candidates, the team saw a several-fold increase in protein production using the new designs. Bancel recalls the meeting when they described this breakthrough: "They blew my brain on the walls."
Avoiding the hype curve
Outside researchers can't yet weigh in on how mind-blowing Moderna's fundamental research might be. "It would be stupendous to see the data out of Moderna," says Paul Agris, an analytical biochemist at the State University of New York in Albany's RNA Institute who has spent decades studying the consequences of modifying RNA nucleosides.
But for now, the company's only published paper is the one from Chien's group on producing VEGF in mice. It hasn't revealed which modified nucleoside is in its newest generation of drug candidates. And it launched its first two phase I trials without announcing what diseases they targeted—a decision Bancel attributes to fears that financial markets would prematurely pigeonhole the company into a particular field. (Investigators are not required to register phase I trials with ClinicalTrials.gov.)
Moderna's leaders argue that they've disclosed research the way most private companies do—by detailing it in patent filings. "It wasn't a deliberate effort to be secretive," Hoge says. "The act of publication was not, in and of itself, a focus for us. In fact, it wasn't even clear that it was anywhere on our priority list."
For many researchers who have worked with companies, that isn't surprising. "It's a highly competitive field, and they've made the decision that they don't want to publish a bunch of papers. That makes sense," says Daniel Anderson, a molecular geneticist who develops drug delivery systems at MIT. "Publishing papers can generate excitement. … But if you have a whole lot of people and a whole lot of money, it may be smart just to stay quiet and develop your technology and patent the heck out of it."
Holding its data close doesn't seem to have hurt Moderna's ability to raise money and advance its drugs. But now that treatments are being injected into people, "there's a certain obligation to patients to start to tell that story," Hoge says. The company has submitted several manuscripts to journals, and last month described the collection of drugs in its pipeline.
Human safety trials have already begun for vaccines against two flu strains and the Zika virus, and for a fourth undisclosed viral vaccine developed in collaboration with Merck. In each case, the mRNA encodes viral proteins that infected cells would normally present to activate the immune system and beat back an infection. Last month, Moderna also began trials of its VEGF drug, developed with AstraZeneca. Intended to treat cardiovascular diseases as well as slow wound healing in diabetes, the growth factor-encoding mRNA is first being injected under the skin of trial participants to evaluate safety.
Moderna is also doing animal safety tests of a personalized cancer vaccine that would code for immune-activating proteins unique to a person's cancer cells, based on genetic sequencing of their tumor. Another possible cancer drug, awaiting regulatory approval for a clinical trial, consists of mRNA for a surface protein called OX40L that would, when injected into a tumor, prompt T cells to proliferate and attack.
Last month's presentation also got attention for what it didn't describe—trials of drugs that replace missing or deficient proteins to treat chronic diseases. Most of Moderna's advanced candidates are vaccines, which require just a low dose of mRNA that makes enough protein to kick the immune system into gear. And all of them are administered locally, under the skin or into a muscle or tumor. To tackle lifelong diseases where patients are missing a key protein, such as an enzyme that removes toxic compounds from the body, mRNA drugs will likely have to be delivered intravenously for decades. That makes even mild toxicity or subtle immune reactions a potential deal-breaker.
Much of the risk comes down to formulation—the molecular packaging that ferries mRNA into cells and protects it from being hacked apart by enzymes along the way. "That's where the breakthroughs are really needed," says RaNA's Heartlein. Many RNA drugs to date have encapsulated the nucleic acid in nanoparticles made of lipids. But because mRNA is so large—roughly 100 times the length of the RNA used for interference therapies—it's harder to stabilize and to encapsulate. And many lipid nanoparticles are not easily degraded in the body, so they can cause toxic buildup in the liver. "We're going to find applications [for mRNA drugs]," Heartlein says, but "it may not be as broadly applicable at the end of the day as people are thinking."
Hoge acknowledges that some conditions may be off limits to mRNA drugs simply because they require higher levels of protein than the mRNA can make at a safe dose. Muscular dystrophies or skin disorders where patients lack a key structural protein, for example, are a long shot. "A lot of people think that gene therapy might be the only solution for some of these diseases. And certainly for some of them, it might be," he says.
Moderna is developing delivery systems that may limit toxicity. Among its proprietary nanoparticles is a family of engineered lipids that its scientists have found to be more biodegradable—and thus more tolerable at higher doses—than existing formulations. A separate "delivery innovation" team is developing nonlipid formulations, such as polymers that form solid, porous structures interspersed with mRNA.
AstraZeneca's Pangalos says his group has its sights set firmly on mRNA drugs for chronic use, and expects a drug intended for repeated dosing to enter trials in the next 18 months. But Moderna has had to retreat from optimistic predictions about a partnership with Alexion to treat a rare disease called Crigler-Najjar syndrome. The mRNA treatment would code for an enzyme that breaks down bilirubin, a toxic substance that builds up in patients' blood. Before it can enter human testing, the companies must be sure the dose needed to impact the disease is many-fold lower than the dose that causes toxicity.
In 2015, Moderna and Alexion predicted that the drug would advance to clinical trials in 2016, but late last year they informed investors that the trials would be delayed, so that the formulation could be optimized. "Lavishly funded Moderna hits safety problems," announced an article published by STAT after Bancel left the drug out of last month's presentation.
A missed milestone, particularly in preclinical studies, hardly signals a catastrophe, says Eric Schmidt, a biotech analyst at Cowen Group in New York City. "I'm just surprised at the drama around the situation," he says. "Why, just because this company has been successful at raising money, is it being treated differently in the popular press?" That may be the price of Moderna's unicorn status: The higher the hopes are for a new treatment approach, the more consequential its warts and blunders become.
But wealth and secrecy may also be protective. Maybe, as Moore and Hoge concluded from their morning meeting, you don't have to ride up and down Gartner's hype curve if you can work through the biggest setbacks before the public ever sees them.
Most small biotechs have to publicize every step of their early research in a scramble to raise money, Moore notes. "Then people get to see all the failures. We've had failures. We've gone down blind alleys. But because we've been quiet about it, nobody's seen that," she says. "That's why I think we're going to end up on the slope of enlightenment without passing the trough of disillusionment."