One morning in spring 2000, Michael Levin flopped in his chair and clicked on his desktop computer. A newly minted assistant professor at Harvard, Levin, then 30, was looking to solve a riddle that had baffled science for centuries: How do our dividing embryonic cells know on which side of our bodies to grow our hearts, our livers, our gall bladders? Countless people throughout history have been born with some, even all, of their organs transposed, and yet functioning. Levin suspected DNA alone was not to blame; there had to be some other trigger. Days earlier, he had ordered an imaging test on a half dozen chick embryos at the verge of organized development. As he pulled up the results, he stared, amazed. Electrical charges, rendered in yellows and reds, lay across the cells in patches, left to right, as clearly as a neon “This Way” arrow. Levin sat back and blinked. He was witnessing, for the first time in history, embryo cells telling each other left from right via electricity.

For decades, genetics taught us a simple truth: Each cell in our body (and there are billions) contains the blueprint that tells us how to grow. That might not be the whole story. Levin and a few others now say that tiny bioelectric signals surging through and among our cells act as an instruction to kick-start gene expression. These signals point cells in the right direction as they start to grow into things like hearts, and influence the shape and function of the body. For two decades, Levin has set about proving it.

In doing so, he has created a startling Island of Dr. Moreau zoo of freaks. He forced tadpoles to grow an eye on their gut; induced frogs to sprout six legs; and caused worms to grow two heads, which, when severed, will grow back just like a salamander’s severed tail—all by manipulating the faintest of bioelectric signals.

He now thinks—no, he knows—he will one day do the same for humans. So if a solider loses an arm to a bomb on the battlefield, he will simply grow a new one. “I don’t know if it will be faster than the normal process of human fetal growth,” says Levin, sitting in his laboratory office at Tufts University where he now works, tending his creations as well as a jungle of houseplants. “Worst-case scenario: If you get your arm blown off at 25, by 35 you will have a teenager’s hand, which is very functional.”

To do this, Levin traverses the most infinitesimal of passages. Across the surface of each cell sit hollow proteins known as ion channels. Charged molecules (or ions) surge through these pathways, entering and exiting cells and changing cell polarities and voltage gradients (the difference in voltage across the body). Tiny gates inside the channels control flow, swinging open and closed based on certain signals; when enough gates open, ions flood the cell and change its charge. The cell passes information to its neighbors through another group of gated proteins called “gap junctions.” By deploying such microscopic tools as neurotoxins, Levin easily opens or blocks channels, flooding them with ions or strangling them. In the process, he creates creatures that nature never designed and that few of us have ever imagined.

“The endgame of this field,” Levin says, “is complete specification of shape. You’d be able to sit down on a computer, like in Photoshop, and draw what you want, and out it comes. If you said, ‘I want a triangular frog with seven legs, and the eyes should be over here,’ I don’t see any reason you couldn’t do that.”

It’s a goal that sounds megalomaniacal, preposterous, Frankensteinian. Even supporters—including his former Harvard mentor, developmental biologist Cliff Tabin—question his controversial claims. While the genetics field now believes that ion channels play a role in creating and differentiating organ placement in the body, many doubt Levin can drive the mechanism. “How do you control it?” asks Tabin. “If you are designing the logic of the system, how do you decide where to make a head as opposed to where you make a tail? You might need channel proteins to make these decisions, but that might not be the linchpin of the decision itself.”

Levin disagrees. And his life’s work, his motivating animation and purpose, is to prove he can use bioelectricity to fix anything.

At 47 years old, on the short side, with pale blue eyes, and an often untidy beard, Levin looks a lot like Pavel Chekov, the navigator in the original Star Trek TV show. Mostly it’s the lax brown hair pasted across his forehead. But he’s also Russian, as was Chekov. Despite immigrating to Swampscott, Massachusetts, a seaside town north of Boston, when he was child, his slight ­Boston accent is sometimes colored by a subtle Slavic inflection.

“That’s a six-legged frog we made, showing you can trigger ectopic limb formation by appropriate voltage gradients,” Levin says, in his usual clinical deadpan. He rarely betrays amazement, humor, or even a glint of smug at what he has wrought. He is standing in the corridor outside his office, a stretch of hall dominated by an unsettling gallery of his creations. Like a music producer’s gold-record collection, he has hung at least a dozen poster-size blowups of his scientific journal covers, such as a 2007 Development, depicting a frog with two legs, one left arm and three ­crab-arm-like protrusions blooming from the right side of its body.

When Levin was a kid, about 10 years old, his father would bring home computers from his programming job at Digital Equipment Corporation, which made those boxy computers you see in Eddie Murphy cop movies. Levin used them to log into the company’s mainframe to learn coding. By 15, he wrote a version of Pac-Man, created a software-graphics editor, and published a journal paper, showing how to use trigonometry to draw 3D-like shapes on a 2D screen.

A year later, in 1986, Levin’s father took the family to Vancouver for Expo 86 (aka the World’s Fair). The experience would change Levin’s life. A monorail hummed overhead. Eurythmics played. General Motors showed off a new holographic technology. But Levin’s big lightbulb moment didn’t happen in a pavilion or aboard a magnetic levitating train. It happened in a tiny bookstore, in downtown Vancouver, far from the crowds.

One day, rummaging the store’s shelves, Levin found a copy of the 1985 The Body Electric: Electromagnetism and the Foundation of Life by Robert Becker and Gary Selden. An orthopedic surgeon for the U.S. Veterans Administration, Becker had become fixated on bioelectricity: the way our bodies interact with magnetic fields (like power lines) as well as the impulses that animate our muscles and brains. As far back as the 1780s, Italian physicist Luigi Galvani discovered the presence of animal electricity by attaching electrodes to a dead frog’s legs and making them twitch. Other scientists later figured out that ions carried this energy through the body. It wasn’t until the 1930s and ’40s that new tools helped researchers learn that ion flows could control a cell’s polarity.

Becker cited these studies and added details from his own experiments. He amputated the limbs of frogs and salamanders, and applied a voltmeter to the wound sites. He found that within 24 hours of an amputation, the wound-site voltage in both species spiked from -10 mV to +20. But the voltage in the salamander later plummeted to -30 mV, a pattern that preceded limb regeneration.

Becker wondered if you could change the voltage in the frog, would it do the same: regrow the lopped-off limb? He thought so. But at the time, he did not have tools precise enough to try it.

Levin, just 16 years old, was electrified by this question. When he returned home, he tracked down each study Becker cited, read it, looked up the references, read those, and followed the trail back to Galvani, Xeroxing hundreds of papers along the way.

His obsession remained a hobby, a side project to his real work: coding. Still, it kept nudging its way into his work. Later, as a ­computer science major at Tufts, he wanted to create an artificial intelligence, which he felt would need the ability to self-repair. But to figure out how to make a machine do it, he first needed to figure out how nature does it. So he borrowed magnetic coils from the physics lab, wrapped them around sea urchin embryos, and measured the way electromagnetic waves changed the rate of cell division, findings that led to his first two scientific papers. By senior year, he had started a software company. But what he really wanted was to join a research lab to make science-altering discoveries. So he left the software company and soon found ­himself in Tabin’s lab at Harvard Medical School.

At the time, Tabin’s team had identified a signaling gene that seemed to express on the left side of the body early in development. They knew a bit about what it did in later stages, but they hadn’t looked into why it was located where it was. Not one of his postdocs would tackle the deeper questions of why and how. “I have a number of really smart, talented, and ambitious students; none of them would touch it with a 10-foot pole,” says Tabin. They did not want to take the risk of throwing years of their lives into a black hole. As soon as he entered the lab, Levin—despite doubts from his thesis adviser—jumped on it. Levin realized, “correctly,” says Tabin, that it was an richly unexplored bit of science. “Mike, when he sees something he thinks is a really cool idea, he doesn’t worry about what other people think,” says Tabin.

Levin then found other genes that controlled the left-right symmetry of body parts, and eventually ­illuminated the genetic pathway that helped direct the action. But he still believed something else drove the signals forward. By 2000, he knew it was bioelectricity, but he needed to know more about how it worked. A colleague had access to a tool that made the cells fluoresce red, green, yellow, and blue based on voltage. Levin asked him to try it on some chick embryos. And then, on that spring day in 2000, there it was: proof electricity played a key role in gene expression, in influencing where and when organs grow.

Sparking life into severed stumps to make them regrow is not all that new. In the 1970s, pioneers such as biologists Lionel Jaffe and Richard Borgens showed they could ignite the beginnings of limb regeneration in frogs by applying electrical currents. But they had conducted their experiments with simple batteries. Levin is the first to precisely tweak bioelectric signals at the cellular level, and to try to crack the code for what it means to those cells. At Tufts, he built a complex toolbox to do this. Among those tools: neurotoxins and drugs that block ion channels that would otherwise stay open, or open those that would stay shut; RNA that also codes for new channels, which Levin injects into cells via glass micropipette; molecules that can transport ions through cell membranes; and genes that code for ion channels (discovered by brain, kidney, and gut specialists). He tracks the impact of voltage changes using fluorescent proteins and dyes, which grow brighter as the voltage gradient rises.

Each cell surface hosts hundreds of ion channels. But only one or two dominate these voltage gradients, so Levin can easily manipulate them. For instance, just four channels act as the master control knobs that determine if organs grow on the correct side of the body. Tweaking any one of them randomizes the organ placement. Levin grew an eye on a tadpole’s gut, just by adding one extra channel. “If you ask the question, ‘Where does the eye come from in the first place?’ you look in the embryo, and you can see that there’s a particular bioelectric pattern that sets up the endogenous eye field,” Levin explains. “Now if I set up that same pattern somewhere else, will I get an eye? The answer, as we know, is yes.”

Inducing a limb to regrow requires a little extra TLC. To make a tadpole regrow a tail, Levin soaks the wound in a solution so charged ions flood its cells. Soak time: one hour. Eight days later: new tail. To regrow a limb takes a 24-hour soak. A functional leg takes about six months. The purpose of the soak, Levin says, is that it “kick-starts all these other cascades of gene expression, of cell behavior.”

Levin faces challenges to do the same for humans, or any warm-blooded mammal. First, warm-blooded animals have much higher blood pressure than reptiles. So there’s a huge risk of bleeding out if the wound is not papered over with a scab. Second, warm-blooded limbs tend to grow more slowly, allowing a greater risk that infection will take hold. And just as with any animal, the body attacks infection with inflammation, which could inhibit cellular growth. Also, to conduct electrical current around a wound, it must stay moist and be protected from air.

Levin, along with David Kaplan, chairman of Tufts department of biomedical engineering, developed a watertight BioDome that they place over an animal’s wound site. Levin’s hope is that a human amputee would have to wear it only a few hours, just long enough to give the cells the initial signal to start growing. Made of silicone, rubber, and silk, it would contain an aquatic habitat similar to what you find surrounding an embryo, but filled with the sort of ion-manipulating tools that would trigger limb regrowth. The pair have put BioDomes over frogs’ severed limbs, which helped the frogs regrow functioning legs. “The tools are there,” says Kaplan. “It’s just getting everything to work together, so it’s only a matter of time.”

Levin’s work could soon change cancer treatment. This past March, he and colleagues made global ­headlines when they reversed cancerous ­tumors in frogs using light to manipulate bioelectric signals. Many cancerous tumors, Levin says, have abnormal bioelectric signaling, in the form of massive cell depolarization. It’s this wonky signal, he believes, that causes them to grow and spread.

Instead of nuking the body with chemo, it might be possible to one day coax deviant cells back into normal tissue. He has also shown he can reverse embryonic birth defects, such as a malformed forebrain in a frog, a defect that bears similarities to those caused in human embryos by a parent’s alcohol abuse.

Already doctors use ion-channel drugs to treat certain cardiac and neurological illnesses. Levin says those same drugs might be used to treat cancers and correct birth defects, if detected in the embryo, by restoring the necessary signals. “I’d put money on the fact this is going to happen within the next 25 years,” Levin says. “I’m being conservative, but I think in my lifetime, we’ll see that.”

Not everyone is sure. Most of the work in regenerative ­medicine takes place around the genome and stem cells. And while some scientists think that sort of singular focus neglects other potential factors—such as bioelectricity—science as a whole is not quite ready to accept Levin’s assertions that bioelectricity is a ­primary ­trigger.

“For us to fully buy into a lot of the things he is talking about, I think there should be a bit more mechanistic insight,” says ­Andre Levchenko, a biomedical engineer who directs the Yale Systems Biology Institute. “We should understand at the same level and clarity as we understand genetic information that controls cell function. We don’t have such an understanding for electrical ­potential. If his aim is to gain the same level of understanding, it’s laudable. He should be supported. It’s clearly part of the story.”

Despite lingering doubts, Levin has landed powerful backing for his experiments, which have been funded by the National ­Institutes of Health. Last April, the Paul G. Allen Frontiers Group, launched by the billionaire Microsoft co-founder, bestowed a $10 million grant, which could balloon to $30 million. Thomas C. Skalak, the group’s executive director, recalled the reaction after Levin gave a lecture the previous winter with a slide show of his creations. “It was earth-shattering,” Skalak says. “People were saying it changed their whole view of biology, that they had never seen data that showed one could have permanent changes in morphology of an organism above the level of a genetic change. It really was an eye-opener.” His hope is that Levin will crack open a new field in bioscience. “We expect it to mushroom,” he says.

So does Levin. His most ambitious goal is to grow any shape he wants, in the lab or in utero. That level of understanding would mean he could fix any malady. And he is using his computer skills to do it. He is designing computational models and artificial ­intelligence programs that will analyze and predict how changing gradients affect an organism’s shape and function—in essence, cracking life’s bioelectric code in order to fully control it.

“We know only a little bit about it right now,” Levin says. “We need to do much more to really have good control.” He likens it to brain science. We know memories are embedded in the brain, but neuroscientists don’t know how to tweak specific neuron states to edit them. “Same with us,” says Levin. “We know electrical properties encode a sort of pattern memory in tissues that cause morphological change. But we are only beginning to understand the formula that connects those patterns.” He adds: “I’m optimistic we will see the long-term stuff. It’s very hard. This is frontier stuff. But you and I will see it in our lifetime.”

Adam Piore is the author of Body Builders: Inside the Science of the Engineered Human.

This article was originally published in the January/February 2017 issue of Popular Science, under the title “The Body Electrician.”