As awesome as our immune system is, though, it doesn’t win every time. Right now it’s facing a stern challenge from a foe it doesn’t always know how to fight. While some countries have managed to control Covid-19 through social measures such as mask-wearing, others—notably the U.S.—have manifestly failed. According to the World Health Organization, Covid-19 has killed almost 700,000 people and sickened some 18 million since its identification in Wuhan, China, last December.

That’s where a vaccine comes in. It’s the injectable answer, the deus ex machina that sets right the botch we’ve made of things. The scientists working in vaccine laboratories from Beijing to Ahmedabad, India, to Plymouth Meeting, Pa., are doing the most important work on the planet. If they succeed, some orator somewhere will surely say of them what British Prime Minister Winston Churchill said of Royal Air Force pilots during World War II: “Never in the field of human conflict was so much owed by so many to so few.”

Failure, or at least partial failure, is very much a possibility. Experts warn that the new vaccines may provide only some protection, and only temporarily—more like the annual influenza jab than, say, the knockout polio and measles vaccines. Old people whose immune systems can’t handle Covid-19 may get the smallest boost. Some vaccines that look good in the lab could fall short in clinical trials, and some anti-vaxxers will refuse to get shots. There could be side effects; a bad vaccine could harm people if it’s authorized for emergency use before human trials are complete. (On Aug. 11, Russian President Vladimir Putin announced the world’s first registration of a Covid-19 vaccine for public use even though Phase 3 trials have just begun.) Vaccines that require refrigeration or multiple doses will be hard to administer in poor countries. Then there’s vaccine nationalism: Countries are racing to find a vaccine and deliver it to their people first rather than pooling resources with others for a faster, better solution.

To appreciate how big the challenge is, consider that there’s still no vaccine against HIV, the retrovirus that causes AIDS, even though it has been identified since 1983 and has killed more than 33 million people. The fallback for AIDS has been antiretroviral drugs, which keep HIV from spreading to other people and arrest its spread in the body. That could be the primary avenue of attack against Covid-19 if vaccines prove to be short-lived or only partly effective.

On the plus side, today’s vaccine developers have better tools and a deeper understanding of viruses and the immune system than their predecessors who developed vaccines against smallpox, diphtheria, tetanus, whooping cough, polio, measles, mumps, and rubella. “It’s a running start,” says Stanley Plotkin, 88, who played key roles in developing rubella and rotavirus vaccines and continues to consult for more than 40 companies from his perch as a professor emeritus of pediatrics at the University of Pennsylvania. “The end is not yet in sight, but I remain cautiously optimistic,” Plotkin says. “Underline cautiously.”

Distinguishing self from nonself, or friend from foe, is the immune system’s first job. That’s more easily said than done, considering that most cells found in our bodies, though seen as self, aren’t strictly human. The body is a “scaffold” for microbes, including bacteria and viruses, that outnumber human cells by a factor of 10. The innate immune system identifies foes approximately, by recognizing broad classes of pathogens. It existed before those jawed fish came along in the Ordovician Period, and it remains the first line of defense. Its weapons include macrophages, neutrophils, and dendritic cells that engulf viruses, and natural killer cells that destroy them. The gene for the proteins on those soldiers that detect the enemy—known as Toll-like receptors—was identified first in fruit flies, in 1985.

The adaptive immune system, the smart one, zeroes in on specific enemies using a strange, Darwinian process whose discovery won a Nobel Prize for Japanese-born Susumu Tonegawa in 1987. There aren’t enough genes in our DNA to code for all the proteins specific to every possible foe, but the body finds a way using a trick called gene rearrangement. The genes in immature white blood cells mix and match segments of DNA to randomly spew out an immense variety of cells with different pathogen detectors. The vast majority are useless. It’s as if each soldier in an army is given a unique weapon that works against only one potential foe. The soldier who happens to have the right weapon for the occasion is then cloned on a massive scale, as in the Star Wars prequels.

Because of this complexity, it takes a few days for the adaptive immune system to kick in after an infection, starting in the lymph nodes, which swell when you’re sick. The defensive play consists of B cells, which emanate from bone marrow, and T cells, which start in the marrow but migrate to the thymus, a small organ between the breastplate and the heart. Some B cells become plasma cells that manufacture antibodies, the Y-shaped molecules that glom onto viruses and other invaders. Some T cells kill pathogens; others help other immune cells. A different set of B and T cells become memory cells that go into action rapidly if a vanquished enemy attempts to return.

This snapshot only scratches the surface of the immune system’s power and complexity. Antibodies alone are a whole textbook. Some neutralize pathogens by clogging their binding sites, while others coat the pathogen (or “opsonize” it) to be recognized by other immune-system attackers. Scientists are still trying to figure out why some people fight off Covid-19 better than others. Age, obesity, and chronic health conditions are clearly factors. Vulnerable people produce fewer T cells, and the ones they do make don’t act as rapidly. They also don’t clone B cells as quickly, so they have fewer antibodies. What they do produce a lot of is cytokines—small proteins that cause inflammation. While cytokines can aid healing, an oversupply can damage healthy tissue. Inflammation from a “cytokine storm” is the cause of death for many Covid-19 victims.

This is the land-mine-strewn biological battlefield that BioNTech, Moderna, Oxford, and other vaccine developers are trying to navigate. One realization every vaccinologist comes to early is that human beings have no hope of creating a defensive system better than the one our bodies already have. Their goal is more modest: to put the immune system on alert, the way a coach gets a team ready to face a new opponent.

Preparing the body to fight by giving it a taste of the enemy is an old trick. Historians now believe that people in China and possibly Africa and India were inoculating one another against smallpox with bits of fresh matter from ripe pustules for a century or more before the practice was introduced to Europe in the early 18th century. In 1796 the English physician Edward Jenner went a step beyond by proving, through an experiment on an 8-year-old boy (this was before institutional review boards existed), that exposure to cowpox generated immunity to smallpox, a more serious disease.

What’s new is that scientists can both view and simulate down to the molecular level how viruses infect cells, how the immune system reacts, and how vaccines bolster that reaction. “I was trained in the ’60s, when a lot of the things that we’re talking about in immunology were imaginary. Now, in my lifetime, they have become pure proteins that we can isolate and identify and work with,” says Caltech’s Huang.

The new tools have led to a profusion of creativity and potential pathways to a Covid-19 vaccine. For example, a partnership between France’s Sanofi SA and Britain’s GlaxoSmithKline Plc uses armyworm moth cells as a factory for churning out copies of the virus’s spike protein—the tip of the spear—that are then injected into the body, triggering an immune response. Sanofi first used that approach for a flu vaccine, Flublok. Tobacco companies, which could use some good publicity, are using tobacco plants as vaccine factories.

Several of the leading companies in the race treat the coronavirus like a piece of software. They isolate the genetic code for the spike protein and put it into a segment of messenger ribonucleic acid, which then goes into the body. The cells that receive the mRNA carry out its instructions to manufacture the spike protein and then show it on their surfaces, which puts the immune system on high alert for the actual virus. Some companies embed the instructions in a ring of DNA called a plasmid. In either case, the body essentially mass-produces its own vaccine.

The software development approach to vaccines moves incredibly fast. On Jan. 11, Chinese authorities shared the genetic sequence of what was then called the novel coronavirus. Just two days later, Massachusetts-based Moderna Inc. and researchers at the U.S. National Institutes of Health isolated the part of the sequence that codes for the virus’s spike protein, which gave them the makings of mRNA-1273, their entry in the vaccine race. Moderna shipped its first batch of mRNA-1273 for animal testing on Feb. 24 and dosed the first human volunteer on March 16. Having been shown to be effective and safe in early tests, it began Phase 3 clinical trials on July 27.

While Moderna delivers its mRNA inside microscopic blobs of fat, some competitors’ nucleic acid vaccines are being delivered via other viruses—such as measles or human or chimpanzee adenovirus—that have been weakened so they can infect cells but not spread. Once the weakened virus enters a cell, the mRNA spliced inside it pumps out spike proteins. There’s poetic justice in pitting one virus against another.

The wave of investment has made this a golden age for vaccine development. Under Operation Warp Speed, the U.S. government has committed $2 billion to the venture of Sanofi and GlaxoSmithKline and $1.2 billion to the British-Swedish company AstraZeneca, as well as splashing out billions on U.S. companies: $1.95 billion to Pfizer, $1.6 billion to Novavax, $483 million to Moderna, and $456 million to Johnson & Johnson, among others. Almost any idea is worth a shot. Some scientists are even trying old methods such as the bacillus Calmette−Guérin (BCG) vaccine against tuberculosis in hopes it will stimulate the innate immune system. Many of the 200 or so vaccine projects worldwide will fail, but that’s to be expected. The International Monetary Fund estimated this spring that Covid-19 will decrease world economic output by $9 trillion in 2020 and 2021. Given that, each month a vaccine can be accelerated is worth $375 billion, says Harvard economist Michael Kremer.

The creativity extends to testing and distribution. Companies are trying out vaccines in countries such as the U.S. where the chance of catching the disease is high. (New Zealand would be a rotten place for a vaccine trial.) In countries with poor recordkeeping such as Bangladesh, organizations are experimenting with using biometrics such as iris scans to make sure each person gets inoculated once, and only once. If a vaccine whose development was partly funded by the U.S. government gets snagged by patent fights, the government has the right to seize control and let others use the patent.

It’s easy to forget that not long ago, vaccines weren’t even dreamed of for most diseases. Poliomyelitis, for example, was a vicious killer and crippler of children until the 1950s, when Jonas Salk created a vaccine. With World War II still raging in 1944, President Franklin Roosevelt, who used a wheelchair himself, devoted a radio address to raising money for the National Foundation for Infantile Paralysis, saying, “The dread disease that we battle at home, like the enemy we oppose abroad, shows no concern, no pity for the young.”

With Covid-19 it’s the old rather than the young who are most at risk, but the sense of being at war with a silent killer is just as strong. The best hope for humanity is a successful alliance between the immune system within us and the vaccine researchers among us.