The mRNA bet: how early DARPA money helped seed the pla...

When SARS-CoV-2's genome was published in January 2020, Moderna was able to lock in a vaccine candidate within days. That speed did not come from nowhere. It rested on roughly a decade of prior investment in an unproven idea: that you could inject the genetic instructions for a protein and let a person's own cells manufacture the antigen. A piece of that early money came from the U.S. Defense Advanced Research Projects Agency (DARPA), which started betting on nucleic-acid vaccines years before there was a pandemic or a market.

This is a success story, and the honest version of it is more interesting than the myth. DARPA did not invent the mRNA vaccine. Neither did any single agency or company. What the record shows is a layered chain of public and private investment, each link doing a job the others could not, that together produced one of the highest-return outcomes in the history of federal research spending. DARPA's role was real and early, and it was one bet among several. Getting the credit right is the whole point.

What DARPA actually funded, and when

DARPA's involvement runs through a program called ADEPT (Autonomous Diagnostics to Enable Prevention and Therapeutics). According to DARPA's own COVID-19 retrospective, the agency began investing in nucleic-acid vaccines under ADEPT in 2011, working from the hypothesis that delivering the genes that encode an antigen, rather than the antigen itself, could trigger protective immunity (DARPA). That is a platform bet: fund the underlying capability, not a specific product.

Two early awards anchor the story.

On October 2, 2013, DARPA awarded Moderna Therapeutics a grant of up to $25 million under the ADEPT: PROTECT program to develop messenger-RNA therapeutics that produce antibodies against infectious diseases and engineered biological threats. The structure was a $24.6 million award plus a $0.7 million "seedling" grant received in March 2013 (Moderna). The phrase "up to $25 million" matters: it is a ceiling over roughly five years, not a lump sum handed over on day one.
In 2011, DARPA co-funded a four-year project with a total value of $33.1 million awarded to CureVac (a German mRNA firm), Sanofi Pasteur, and In-Cell-Art to advance CureVac's RNActive mRNA platform (Fierce Biotech). The $33.1 million was the total project value co-funded by DARPA, not necessarily DARPA's full share, and some outlets reported the figure in euros rather than dollars.

Two details keep this honest. First, the 2013 Moderna money predates COVID by about seven years, and per Knowledge Ecology International's later analysis, Moderna applied that DARPA funding to its Chikungunya and Zika vaccine programs. It paid for platform maturation, not for a COVID vaccine that did not yet have a target. Second, DARPA backed more than one horse, and not every horse finished. CureVac's own mRNA COVID vaccine went on to fail its Phase 3 trial in 2021. That is not a footnote to bury. It is what a real research portfolio looks like: several modest early bets, some of which do not pay off, underwriting the occasional outsized win.

The invention DARPA did not produce

Here is the load-bearing correction to any single-agency origin story. The scientific breakthrough that made mRNA vaccines viable did not come from DARPA. It came from Katalin Karikó and Drew Weissman at the University of Pennsylvania.

Their work solved a fundamental problem: unmodified mRNA provokes a strong inflammatory response and gets destroyed before it can do useful work. Karikó and Weissman showed that modifying the nucleosides in the mRNA, using pseudouridine, sharply reduced that immunogenicity and boosted protein output. They first published the finding in the journal Immunity in 2005, and in 2023 they received the Nobel Prize in Physiology or Medicine "for their discoveries concerning nucleoside base modifications that enabled the development of effective mRNA vaccines against COVID-19" (Nobel Assembly at Karolinska Institutet).

This was NIH-supported academic science, separate from and predating DARPA's platform awards (NIH/NIAID). And it underlies both of the COVID mRNA vaccines. Per Penn Medicine, both Pfizer/BioNTech's and Moderna's vaccines rely on the Karikó and Weissman modified-nucleoside technology licensed from the University of Pennsylvania (Penn Medicine).

That last point closes off a common confusion. Pfizer/BioNTech's vaccine was built on the Penn-licensed science plus BioNTech's own work. BioNTech did not receive the DARPA ADEPT mRNA awards; those went to Moderna and the CureVac consortium. So the public-science lineage of the Pfizer/BioNTech shot runs through NIH-funded Penn research, not through DARPA. Lumping BioNTech into "DARPA-funded" is simply wrong.

The money that carried a vaccine to market

DARPA seeded the platform. Something much larger scaled it into a deployed product.

Moderna, a former DARPA ADEPT performer, received FDA Emergency Use Authorization for its mRNA COVID-19 vaccine in December 2020, and both mRNA vaccines were deployed at scale across 2020 and 2021 (DARPA). But the vaccine itself was co-designed rapidly in 2020 with the NIH, specifically the NIAID Vaccine Research Center, and it was carried to market on pandemic-era funding that dwarfs the 2013 grant.

During the pandemic, the U.S. committed large sums to Moderna alone:

Up to roughly $955 million toward development through BARDA (the Biomedical Advanced Research and Development Authority): $483 million in April 2020, then a further $472 million in July 2020 (PolitiFact).
Up to a further $1.525 billion, in a deal announced in August 2020, to manufacture and deliver 100 million doses.

Taken together, that is a potential total of about $2.48 billion committed to one company's vaccine through BARDA and Operation Warp Speed (GEN). These are commitment ceilings rather than cash already spent, but even as ceilings they make the scale clear. The 2013 DARPA award, capped at $25 million, was seed capital. The BARDA and Warp Speed money was the scale-up.

A peer-reviewed accounting puts the whole ledger in perspective. A BMJ retrospective cohort study by Lalani and colleagues, published in March 2023, estimated total identifiable U.S. public investment related to mRNA COVID-19 vaccines at $31.9 billion from 1985 through March 2022 (Lalani et al., BMJ). Of that, roughly $337 million was direct pre-pandemic research and development, which the authors split as about $116 million from NIH (35 percent), about $148 million from BARDA (44 percent), and about $72 million from the Department of Defense (21 percent). DARPA's mRNA grants sit inside that $72 million DoD bucket.

Two cautions about those numbers. The percentage split describes the roughly $337 million of pre-pandemic R&D, not the $31.9 billion headline; the large majority of the $31.9 billion was vaccine purchases, not research. And the study's authors call the pre-pandemic figure conservative and likely much higher. What the figures establish is proportion: DARPA was a modest slice of a modest pre-pandemic R&D total, which was itself a tiny fraction of total pandemic spending.

The disclosure dispute

The one genuine controversy attached to this story is about credit, not conduct. In August 2020, the advocacy group Knowledge Ecology International (KEI) examined a set of Moderna patents and reported that none of 126 granted patents (and 154 applications) disclosed the early DARPA support, flagging 11 that in its view should have (KEI). Under the Bayh-Dole Act, federally funded inventions generally carry disclosure obligations, and KEI argued Moderna had not met them for the DARPA money.

DARPA publicly acknowledged that it opened an investigation into the question. That is worth stating precisely, because it is easy to overstate in either direction. An investigation being opened is confirmed; it is not the same as wrongdoing being proven. KEI is an advocacy organisation, and its analysis should be read as advocacy. The honest characterisation is a contested disclosure dispute over whether early public support was properly acknowledged, not evidence that DARPA "owns" the vaccine or that Moderna was adjudicated to have hidden funding.

The ROI case

On pure return, this is one of the strongest bets in the federal research portfolio, and the reason is leverage rather than attribution.

Right kind of risk. In 2011 to 2013, mRNA as a vaccine platform was speculative. The companies were pre-revenue, the immunogenicity problem had only recently been solved in an academic lab, and no private funder was rushing to underwrite a rapid-response biothreat platform with no market. Funding the unproven thing that could matter enormously in a crisis is precisely DARPA's mandate.
The payoff dwarfs the seed. The seeded platform helped Moderna design a candidate within days of the genome's release and reach FDA authorisation by December 2020. Vaccines deployed at scale across 2020 and 2021, billions of doses, lives and economic activity preserved: the return is many multiples of the tens of millions in early seed money, no matter how you slice the accounting.
Portfolio logic. A CureVac thread failed in Phase 3, and the biggest single scientific enabler came from someone else's lab. That is not a knock on the model. It is the model. A few modest early bets, some of which fail, are how portfolio R&D earns occasional outsized wins.

The honest caveats

Shared credit is the point, not a caveat to explain away. The load-bearing invention, Karikó and Weissman's modified nucleosides, was NIH-funded Penn science, licensed to both vaccines. DARPA's dollars bought platform maturation, not the core discovery.
Keep the magnitudes straight. DARPA seeded; BARDA and Operation Warp Speed scaled. The 2013 grant was capped at $25 million and funded Chikungunya and Zika work, not a COVID vaccine. The money that carried a deployed vaccine to market ran to the billions and came later.
Success was not guaranteed at funding time. CureVac's DARPA-seeded COVID vaccine failed Phase 3 in 2021. Outcomes were genuinely uncertain when the bets were placed.
Credit was contested. The KEI and DoD patent-disclosure inquiry is a real dispute about whether the public support was properly acknowledged. It is unresolved, and it does not make DARPA the owner of the technology.

The most persuasive version of this story does not need to inflate DARPA's role. A functioning public-research ecosystem, running from Nobel-winning bench science at Penn, through DARPA's early platform bets on Moderna and CureVac, to BARDA and Warp Speed scale-up, delivered a return no single actor could have produced alone. That is the return on a high-risk public research bet, told straight.

Fact-check notes and sources

DARPA began investing in nucleic-acid vaccines under ADEPT in 2011, on the hypothesis that gene-encoded antigens could trigger immunity: DARPA COVID-19 retrospective.
The October 2, 2013 Moderna award of up to $25 million ($24.6 million plus a $0.7 million March 2013 seedling) under ADEPT: PROTECT, for mRNA therapeutics: Moderna press release. The "up to" figure is a ceiling over roughly five years.
The 2011 $33.1 million total-value project co-funded by DARPA with CureVac, Sanofi Pasteur, and In-Cell-Art: Fierce Biotech. Currency reporting varies between dollars and euros; the figure is total project value, not DARPA's full share. CureVac's mRNA COVID vaccine later failed Phase 3 in 2021.
Karikó and Weissman's nucleoside base modification, published in Immunity in 2005 and honoured with the 2023 Nobel Prize in Physiology or Medicine: Nobel Assembly at Karolinska Institutet. NIH framing of the NIH-funded basic science: NIH/NIAID.
Both COVID mRNA vaccines rely on the Karikó and Weissman modified-nucleoside technology licensed from Penn; Pfizer/BioNTech built on that plus its own work and did not receive the DARPA mRNA awards: Penn Medicine.
Moderna, a former DARPA ADEPT performer, received FDA Emergency Use Authorization in December 2020: DARPA.
Pandemic-era commitments to Moderna: up to about $955 million in development ($483 million April 2020 plus $472 million July 2020) via BARDA, per PolitiFact; plus up to $1.525 billion (announced August 2020) for 100 million doses, for a potential total near $2.48 billion, per GEN. These are commitment ceilings.
Total U.S. public investment of $31.9 billion (1985 to March 2022), including about $337 million in pre-pandemic R&D split NIH $116 million / BARDA $148 million / DoD $72 million: Lalani et al., BMJ. The percentages describe the pre-pandemic total, not the $31.9 billion; most of the headline figure was vaccine purchases, and the authors call the pre-pandemic figure conservative.
The Bayh-Dole patent-disclosure dispute, and DARPA's confirmed investigation: Knowledge Ecology International (advocacy source). Report as a contested inquiry, not adjudicated wrongdoing.

The mRNA bet: how early DARPA money helped seed the platform behind the COVID vaccines