Having introduced the world to the hydra’s wondrous ability to regenerate itself, Abraham Trembley was brutally honest about what he had accomplished. That finding, he acknowledged, was not “the fruit of long patience and great wisdom, but a gift of chance.”

Two hundred years later, the embryologist Leroy Stevens—Roy, to his family and friends—would similarly insist that his discovery of a tumor-laden mouse made shortly after he joined the staff of the Roscoe B. Jackson Memorial Laboratory in Bar Harbor, Maine, was due to luck, not genius. And what luck. The odds of bumping into a mouse strain bearing such a singular burden were infinitesimal.

He was fortunate as well, he confided to a colleague some years later, to have stumbled on this unusual sighting while working at an establishment that gave its scientists the freedom to pursue whatever they bumped into. “I felt perfectly free to do anything I wanted, and didn’t have to account to anybody,” remarked Stevens, a compact fellow with a long nose, dark wavy hair, a ready sense of humor, and an independent streak a mile long. “I very slowly got this thing off the ground. I mean, what do you do when you find something as rare as that!”

Today, it is widely held that Stevens’s lucky break and all his science thereafter went a long way toward launching the field of stem cell biology. For a century logic had been whispering that cells inside an animal’s early embryo must serve as the basis for all the differentiated, specialized cells that form organs and systems, and build the adult. Zoologists suspected that similar embryonic cells must lurk in the tissues of adult starfish, crabs, salamanders, leeches, and other regenerative creatures. What other explanation was there for the ability of these animals to replace a missing part out of thin air? Once his attention was given over to the strange sight he beheld in Mouse Strain 129, Stevens was able to peer into a microscope at a rodent’s version of these immature, unspecialized cells. Although he never studied their potential to differentiate in depth, his research made the existence of these unique cells so noteworthy and real that scientists the world over would become mesmerized by them and hopeful that their special powers might one day revolutionize medicine.

It all began in the winter of 1953. Stevens, then a junior researcher at the Jackson Laboratory, had been assigned to look for outward differences in inbred mice that might signal inherited defects in certain genes. He was going about this task when he noticed that a chinchilla-colored male mouse in the strain he was investigating had a distinct abnormality—an enlarged scrotum. The mouse was immediately dissected and slides of its tissues examined. As a colleague confirmed, the testicular growth was a tumor that also arises in humans. It was the notorious teratoma—Greek for “swollen monster”—and true to its name, it could be as monstrous in appearance as it could be deadly. (For our purposes here, the term teratoma will be used for both varieties of the tumor, benign and malignant.)

Of the great range of tumors that strike humans, teratomas rep-

resent but a small fraction. Most are found in the gonads—usually the ovaries, where they are almost always benign. By contrast, they rarely develop in the testes, but when present in that location, they are almost always malignant and a death sentence if left untreated. Very occasional reports place them in other parts of the body: head and neck, heart, liver, stomach, chest cavity, uterus, spinal cord, and even under the eyebrow. By the early ’50s, they had been spotted in only a few nonhuman animals—cow, sheep, and stallion, with some also seen in the ovaries of occasional female mice. The one beheld by Stevens was the first ever described in the testis of a male mouse.

The grisly contents of teratomas had fascinated pathologists for centuries. Unlike so many other tumors, which usually contain one type of cell, a teratoma could wind up a jambalaya of different cell types and tissues. Those that had been growing for a while and getting larger might contain muscle fiber, nerve, nodules of cartilage and bone, skin, intestinal epithelium, pigment, fat, or other cells and tissues, intermixed with fluids. Wads of hair often were the tumor’s most prominent feature, while it wasn’t uncommon to find a few baby teeth sitting in the midst of this strange mélange. Especially in younger tumors, undifferentiated cells—small, round, and plain—were fairly easy to spot under a microscope’s lens, sometimes surrounded by mature tissue. These immature cells were stem cells—referred to in those days as “embryonic” cells. Generally, the older the tumor became, the more its stem cells proliferated and differentiated, just as happens in an organized way to immature cells inside an embryo. But in a teratoma it was unorganized, a jumble.

What was bizarre, even for a pathologist, was finding a teratoma’s confusion of tissues senselessly growing where they didn’t belong. What could be more disconcerting, for instance, than to find a teratoma with teeth inside it growing in a woman’s ovary?—the only example of tooth development outside the mouth. A physician in the seventeenth century had reported the even stranger sight of a bone with indentations shaped like eye sockets inside the testicular

teratoma of a young Frenchman. Stevens, himself, when studying teratomas from freshly killed mice would come across “twitching and pulsating movements that correspond with the histological findings of abundant striated skeletal and cardiac muscle,” as he noted in one paper. Cardiac muscle cells inside a mouse’s testicular teratoma pulsated in unison, just as they did in the heart.

Like others before him, Roy Stevens was instantly captivated by the teratoma. Having earned his Ph.D. in experimental embryology from the University of Rochester the previous spring, where he had been whipped into shape by the famous embryologist Hans Holtfreter, Stevens appreciated a teratoma’s resemblance to an embryo, albeit a disorganized and nonsensical embryo. It, too, was a growth that most likely began as a single embryonic cell and grew larger as cells divided, proliferated, and differentiated into specialized cells of the ectoderm, mesoderm, and endoderm—the body’s outer, middle, and inner leaves. Strange as it seems, some scientists saw a teratoma and its differentiating cells as an excellent chance to learn more about a mammal’s embryo and its cells, and how they underwent specialization. The tumor’s malignant state also provided an instructive window into abnormal development, and why some cells miss various cues and fall out of control.

Any watch post into the earliest days of a mammal’s development would be a boost to embryologists. Sequestered inside the body, a mammal’s developing embryo wasn’t as retrievable as those of a multitude of other classes of vertebrates, whether bird or reptile or fish, and certainly not nearly as accessible as the fertilized eggs of the European newts, American frogs, and other amphibians that Stevens had studied at Rochester. Amphibians had long been embryologists’ favorite friends: Their fertilized eggs, which develop outside their bodies, are easy to collect in large numbers, and the size of the eggs—four to five times larger than fertilized mammalian eggs—makes them relatively easy to examine. Since the mammalian embryo was less accessible, even by the middle of the twentieth century there weren’t many mammalian embryologists. There just wasn’t enough

raw material to solve embryology’s foremost mystery as it applied to mammals: How did a single cell—the zygote, or fertilized egg—manage the colossal feat of expanding into a highly complex individual?

The chief stages of early development in all vertebrates include cleavage, wherein the zygote undergoes a series of divisions that create a ball of many smaller cells; gastrulation, wherein the enlarging embryo develops the three germ layers (ectoderm, endoderm, and mesoderm); organogenesis, wherein the germ layers interact to form specific tissues and organs; and finally birth. In the human case, a little being comprised of trillions of superbly calibrated cells enters the world already knowing how to wail for his or her supper.

Autopsying hundreds more mice from Strain 129, Stevens came across several other males that had the same testicular growth. There weren’t many of them, but he knew he was on to something. For the tumor to keep popping up, however infrequently, a defective gene or genes must be getting passed down, generation after generation.

While an inherited teratoma had never been glimpsed in a mouse, or in any other mammal for that matter, this was precisely the sort of genetic phenomenon that had compelled Clarence Cook Little, a distinguished biologist and former college president, to start the Jackson Laboratory in 1929. “Prexy” Little was a mouseologist and had been since his boyhood, when he had raised squeaking mice along with pedigreed pigeons, rabbits, and guinea pigs in his family’s Brookline, Massachusetts, home. It was while he was a student at Harvard, where he had been swept up in the new science of genetics and taught by the influential William Ernest Castle, that Little began the mouse matings that would lead to the first-ever inbred mouse strain. Later, at Cold Spring Harbor Laboratory, he helped to produce others of the earliest, most enduring inbred colonies. By mating a litter’s brothers and sisters over and over—for at least twenty generations—you could create a line of mice that were almost as alike genetically as identical twins, and any observable difference between two such strains of mice might reveal a genetic difference.

Little believed that inbred strains would validate his hunch that

some cancers were the vicious outcome of defective genes. When Roy Stevens came on board in the summer of ’52, Little was still at the helm, still directing his scientists to hunt for evidence that could demonstrate that at least some varieties of cancers arose because of the inheritance of aberrant genes, and still remarkably ahead of most other scientists in this pursuit. The idea that inherited mutations could account for cancers in families was still some years away from gaining wide acceptance.

Only one percent of Mouse Strain 129, it turned out, was afflicted by teratomas. Still, the value of this unusual inbred mouse line wasn’t lost on Stevens. By examining these mice, perhaps he could figure out where in the developing embryo a teratoma started. It seemed a good bet that the pathology began before birth, because the tumor could be spotted a week after birth. Maybe he could even discover what caused the tumor, and establish some biological mechanism common to all cancers, about which so little was known.

With Dr. Little behind him, Stevens plotted out how he would locate a mouse’s teratoma beginning cells. Essentially, he would inspect tissue from the testis region of younger and younger fetuses, working backward through development until he arrived at the tumor’s earliest appearance. To get up to speed, he read everything he could find in the Jax Lab’s library on teratomas, not without sneaking a Lucky Strike or two, which put him on thin ice with the librarian, Joan Staats. (Stevens, it seems, was a bit of a devil. Back during his stint in Officer Training School, he had graduated with the most demerits, an accomplishment he takes some pride in and wants inscribed on his tombstone, according to his daughter Anne Wheeler.)

Historically, teratomas and their weird jumble of tissues had been blamed on demons. More recently, those that grew in the ovary were thought to represent an ectopic pregnancy, which occurs when an embryo develops outside the uterus. But the theory that made the most sense to Stevens, as it did to other researchers, proposed that a teratoma formed when a primordial germ cell in the embryo—a pre-

cursor of egg and sperm cells—fell out of line. Another Jackson Lab scientist, Elizabeth Russell, had helped trace the pathway that early germs cells took when they migrate from the yolk sac into the embryo’s gonadal region, where they mature into egg or sperm. With his mouse pups in mind, Stevens imagined that all it took to start a teratoma was for one migrating germ cell to go astray in the developing embryo. Although the resulting tumor usually manifested itself shortly after a mouse, or child, was born, sometimes a rogue cell sat dormant for decades, which is why teratomas could suddenly make an appearance later in life.

If Stevens was to prove that a misbehaving germ cell in an embryo initiated a teratoma, he would need plenty of mouse mothers supplying him with plenty of tumor-laden embryos. Within months, however, he ran into a serious obstacle. The laborious work of dissecting 3,557 mice and inspecting their testicular tissue had yielded only thirty teratomas. “I could have abandoned the project then and there,” he told an interviewer years later. “But I was young and stubborn, and it seemed to me my work had a very real potential if I could find a way to increase the incidence of tumors dramatically and thus avoid having to raise, kill, dissect, and examine hundreds of thousands of mice in the hope of tracing the development of the tumors.”

Stevens had plunged into his new project with the same dauntlessness he had shown as an Army lieutenant in World War II, when he participated in the Sicilian Campaign under General George S. Patton and served as a forward observer in reconnaissance aircraft. “Roy was a war hero,” recounts Barry Pierce, a pathologist and friend of Stevens. “He was decorated several times for bravery. When he received his final medal from General Patton, Patton looked at him and said, ‘You’re racking up quite a collection of these, aren’t you!’”

Now, more than bravery, Stevens would search for resiliency and determination to get him through a long dry spell when his science seemed to be going nowhere. After two years’ time and 17,000 inspected mice, he finally hit upon a substrain in which one of every

fifty males had a teratoma. In a blink, he had twice as many tumorous mouse pups to work with. But he needed still more, or the research might drag on forever. He tried feeding his mice known carcinogens, even blitzing them with radiation. But these attempts went nowhere.

His work was creeping along so slowly, he worried that the American Cancer Society would give up on him and cease its funding. To keep his ACS contacts happy, he wrote up detailed papers about Mouse Strain 129’s bizarre tumors and their chaotic land-scapes of undifferentiated and differentiated cells. It was “generally accepted,” he frequently stated, although not yet proven, that a teratoma’s growth in the testis or ovary of a developing mouse was due to “undifferentiated pluripotent cells” going astray. In his own estimation, the culprit was a germ cell in the gonads of a late embryo. Yet for all he knew, the culprit might instead be a stem cell in the early embryo.

Stevens’s papers were often accompanied by photographs that depicted fields of small, round embryonic cells inside a teratoma. As the tumor grew bigger, these stem cells made more and more of themselves, along with differentiated fare. Its versatility made a stem cell a most unusual cell. Other cells upon dividing usually resulted in two equal daughter cells that resembled the mother cell. But a stem cell had choices. It, too, upon dividing could give rise to two daughter cells—stem cells, in its case. But it could also give rise to one stem cell plus a differentiated cell; or two differentiated cells.

Stevens spent more and more hours in the lab, and fewer hours at home with his wife and, by the late ’50s, two young daughters and younger son. He tried to set Sundays aside, however, as a day in which he and his family could explore one scenic corner of Mount Desert or another. Having grown up in the Buffalo, New York, suburb of Kenmore, Stevens had felt cheated as a boy not to live closer to the ocean. (Coincidentally, his family had resided at #129 Lasalle Avenue, the very same number of his mouse strain.) When the Jackson Laboratory hired him, and he and his wife, Jean, came to Mount

Desert, it felt as though they had definitely arrived in Eden, the name originally given to Bar Harbor when it was settled in 1796.

Some incoming scientists discovered they weren’t meant for the Maine coast, which could be gray and forlorn in winter, and quickly moved on. But not Stevens. A consummate naturalist, he would never tire of the island’s rugged beauty. Anne Wheeler recounts that when she, her sister, and brother were growing up, their father’s excitement over natural wonders was unfailingly contagious. “There was the time in the middle of the night,” she remembers, “when the cat brought in a bat, and he woke all of us up and dissected it on the spot.” Also high on her list of childhood memories are the many Sundays in summer when the Stevens and other Jackson Lab families converged on Compass Harbor with their picnics. There, after throwing their sandwich crusts to the resident gulls George and Martha, the embryologist and his children would spend blissful hours wading in the cove’s immense tidal pool, hunting for crawly, slippery creatures hidden under rocks and seaweed, Stevens as content as a clam in mud.

While savoring the island’s spectacles, this observant scientist would feel the tug of a connection between the plant and animal species around him and the featureless embryonic cells—stem cells—that he witnessed scattered in the midst of the teratomas he was extracting from mice. It had been quietly appreciated for some time that Nature’s fecundity must have lots to do with the embryonic cells that underlie the cellular construction of plants and animals. These young cells were the road to cellular diversity, the fountain through which every kind of specialized cell flowed. From stem cells, as well as progenitor cells—cells that were a touch more differentiated but still had stemness—evolved myriad lineages of maturing cells that diverged down further branching avenues of form, function, nuance, and color, culminating in the creative detail and specialization that allow creatures to thrive on land, in water, or in air. An osprey’s talon or a wolf spider’s eight eyes, wild indigo’s bright flower or the light bones of a ruby-throated hummingbird or the

nectar the hummer sips on—every bit of biological variety was descended from stem cells. Everything had to come from something, and that something was the stem cell.

In the mouse teratomas that Stevens examined in the lab, the more advanced specimens could contain as many as fourteen different mature cell types—bone, intestine, blood, hair, muscle, cartilage, nerve, and more—the whole lot likely springing from a single cell. The proficiency of a stem cell, whether from a tumor of the gonads or an early embryo, was as mystifying as a wizard’s ability to change willy-nilly into anything he desires to be. It had something to do, Stevens knew, with the genes inside a stem cell’s nucleus and their getting switched on and off. But knowing this didn’t make a stem cell seem any less magical.

Many of the marine creatures that Roy Stevens and his children spied in the tidal pool at Compass Harbor—starfish, crabs, sea anemones, jellyfish, sea urchins, and others—were among the very critters that had been the objects of unending experiments in the late nineteenth century and had gone a long way toward enlightening naturalists about embryonic cells and their distinctive capabilities. By the year 1900, cells were a fairly new addition to science’s Book of Knowledge. Only in 1838 and 1839 had they been identified as the vital “pores” of living tissue. Even less time had passed since, in 1855, Rudolf Virchow had observed that “every cell comes from a cell,” which had led to the realization that the microscopic world was teeming with dividing cells, multiplying cells, and differentiating cells.

In the 1890s, experiments with sea urchins brought the great surprise that cells in the early embryo were far more plastic—that is, flexible—than previously imagined. The German biologist Hans Driesch had expected to find that when he separated an urchin’s

beginning two-cell embryo, each cell would develop into only the part of the animal it was fated to become. So the work of other scientists had indicated. “But things turned out as they were bound to do, and not as I had expected,” Driesch later confessed. He was dumbfounded to discover that each of the two cells had the capacity to develop into a complete larva, a complete organism. It happened again when he broke apart a four-cell embryo: an entire larva ensued from each of the four cells. The larvae were smaller than usual but normal nonetheless.

The fertilized egg obviously possessed a power that enabled it to give rise to a complete individual, but now Driesch was showing that even after it had divided once, then once again, its daughter cells also had this prospektiv Potenz—“prospective potency,” as he called it. Their fate was far from sealed, apparently, although the older the embryo, the more the potency of its cells seemed to diminish. The early embryo, Driesch declared, was endowed with “harmonious equipotential,” and perhaps as a direct result, the terms totipotent cell and pluripotent cell gained regular use by the early 1900s. A totipotent cell can generate every cell connected to an organism; a pluripotent cell, nearly every cell. A stem cell bearing such potency was described by other adjectives as well, including indifferent or undetermined, both words referring to a cell’s undifferentiated state. As for the term stem cell, stammzelle appeared in the German literature of the nineteenth century, although its English-language equivalent did not cross many lips until Roy Stevens’s era. Even up through the 1970s, researchers were more apt to refer to stem cells as “embryonic” or “embryonal” cells, according to scientists of that generation.

Driesch had presciently stated, “Every cell, during development, carries the totality of all primordia.” But he and his contemporaries fell short of knowing why a cell had such potential. Although Gregor Mendel had theorized in 1865 that minuscule entities inside pea plants were the messengers of inherited characteristics, it would be well into the next century before the genes that lay tucked away in a cell’s nucleus would be fully valued as the blueprint for inherited

traits, as well as the means by which a cell knew to specialize down a certain lineage pathway, be it blood or bone.

During the tail end of the nineteenth century, meanwhile, interest in the phenomenal regenerative skill that abided in so many grown creatures was booming. A burgeoning fleet of American and European naturalists, among them Driesch and other revered experimental biologists of the day, were sinking their teeth into the subject, hoping to solve the countless questions that had stymied Trembley, Bonnet, Réaumur, and other earlier regeneration pioneers. Why had Nature given this remarkable ability to smaller creatures but not to more complex animals? You only had to get up to birds on the evolutionary ladder, and the talent was practically gone. Some smaller creatures, however, were as regeneratively lackluster as humans. Regeneration appeared to be a quirky talent that was distributed unevenly across the animal kingdom.

Could it be, these scientists asked themselves, that embryonic cells endowed with Potenz, the likes of which Driesch had caught sight of in the early embryo, resided in the tissues of grown animals as well? What better explanation could there possibly be for the miraculous regrowth of, for instance, a lobster’s claw? Yet when a naturalist took to his microscope to find these powerful cells, he didn’t necessarily see them. In the late 1800s, the biologist Francis Herrick, fascinated by how a grown lobster could drop a claw and handily replace it with a new one, methodically scrutinized a lobster’s anatomy for a “store of embryonic cells” that might explain this trick. His search proved fruitless, and as he noted in his monograph The American Lobster (1895), “The examination of serial sections through this part of the limb reveals nothing but normal tissue cells. Embryonic cells may be present but are not discernible.”

The emerging consensus was that the animal kingdom exhibited two varieties of regeneration—the ordinary kind and the extraordinary kind. The former consisted of the routine repair and replacement of tissue that every organism underwent. The continuous

replenishment of skin, for example, was as important to a person’s everyday survival as a tree’s ability to grow back its leaves.

But it was the second and far less frequent type of regeneration that was more interesting to regeneration scientists, particularly since humans lacked it—an animal’s expedient restoration of an entire missing part. This capability is termed epimorphic regeneration, referring to the remodeling of multiple tissues. By the early twentieth century, biologists perceived that a “bud,” or group of cells, forms at the boundary of the lost part, and that these cells rapidly proliferate and differentiate into the range of tissues necessary for remaking the structure. Salamanders and other tailed amphibians were observed to be especially able regenerators. Lose a leg, a tail, a jaw, or an eye lens? A salamander could grow a replacement in a matter of weeks or months. Impressive, too, were the American lobster and numerous crab species that could so deftly regenerate a claw lost to battle or caught under a rock. How many more pickpockets there might be, biologists joked, if they too could drop an arm and scamper off, knowing they’d soon have another?

(There’s the somewhat parallel situation of a kelp crab’s clever means of escape, as pointed out to me by Robin Cooper, a biologist at the University of Kentucky. If a seal tries to eat a kelp crab, the crab will give the seal a good pinch, dropping its claw while doing so. Which is why every once in a great while a seal will be spotted with a crab’s claw dangling from its nose.)

At field stations from Wood’s Hole to Pacific Grove, regeneration scientists struggled to divine information about the restorative prowess of earthworms, silkworms, flatworms, fiddler and gulf-weed crabs, hydra and various fish, leeches, frog larvae, lizards, salamanders, and Mediterranean medusae (jellyfish). If you left a hydra in darkness as opposed to sunlight for twenty-five days, did regeneration proceed more slowly? If you starved a salamander, did that affect its limb restoration? Was there a relationship between a crustacean’s molting period and regeneration? How many arms could

you remove from a Brittle-Star and still see recovery? Darkness didn’t appear to markedly slow regenerative growth; nor did starvation, although the restored legs of starved salamanders were quite a bit more spindly than those of well-fed salamanders, as the biologist Thomas Hunt Morgan assiduously observed in a 1906 study. Molting and regeneration did seem to share anatomical features. And, as might be expected, the Brittle-Star didn’t survive the removal of all five of its legs. Poor starfish.

As in Trembley’s day, most early-twentieth-century reports fell short of deep explanation. But while regeneration would have to wait for the Molecular Age for further unraveling, some tangible observations were pocketed. Zoologists agreed that for regeneration to take place in either grown plants or animals, sufficient embryonic cells must be present. But where did they come from? One possibility cited by Eugen Korschelt in his tome Regeneration and Transplantation (1927) was that cells “left over from embryonic development” still lingered in the adult body, “very insignificant complexes of indifferent cells which we cannot perceive.” A second possibility noted by Korschelt was that mature cells at the site of a missing appendage first dedifferentiated back into an unspecialized embryonic state, and then, upon regaining totipotency, they and their cell offspring could proliferate and differentiate anew into the sundry cell types that were needed to reform an entire appendage from scratch.

This alleged retro ability was reported as far back as the mid-1800s, shortly after the cell theory put cells on the map as the common ingredient of all living matter. “Some facts concerning the regressive differentiation of cells have long been familiar,” noted zoologist Henry Wilson in 1911. He refers to an 1856 journal report that states that a sponge’s choanocytes, mature flagellated cells, could “de-specialize” and possibly make “a complete return to the indifferent (totipotent) state.” Wilson himself imparts that the activity of straining small pieces of a mature sponge through “fine bolting cloth such as is used for tow nets” produced immature cells that proliferated and then redifferentiated into mature sponge tissue. There might

be “other animals besides sponges and hydroids in which the somatic cells when forcibly disjoined, will fuse and give rise to totipotent regenerative material,” Wilson elsewhere comments.

Not everyone believed that mature cells could dedifferentiate. Korschelt mentions that there were those who stuck to the opinion “that once cells have developed along specific lines, they are in no case able later to change tack.” Yet Korschelt felt that due to the abundance of evidence, “there can scarcely be any denying the possibility that development processes are reversible.” Immature cells could mature, and so apparently could mature cells regress, he believed.

Day in and day out Roy Stevens was tipping the lid on stem cells. They lay just below his nose, bathed in his microscope’s circle of light. Granted, the stem cells he examined were from mouse tumors. Most probably had malignant properties and were a shade different from normal stem cells in embryos. Nevertheless they were stem cells, and for anyone interested in investigating the far reaches of their capacity to differentiate into a range of mature cells, Stevens’s Mouse Strain 129 and its teratomas made them available. Normal stem cells had not yet been isolated from embryos, not even a mouse’s; so the tumor kind were most unique and alluring. Another scientist might have put everything aside to concentrate on these special cells and their talents, but Stevens’s sole bent was to understand how a teratoma began and grew, in hopes of “gaining control of the disease in humans,” he would explain. So focused was he on digging down to the roots of this cancer, that when evaluating him for a possible fellowship, one Jackson Lab colleague observed that if Stevens had any shortcomings, they arose “from his single-mindedness” at the bench. Continued this observer, “I think he has chosen an independent course and will continue so.”

To make progress, Stevens needed still more tumor-beset animals. In 1958 he gambled that among the Jackson Laboratory’s considerable stocks of mice, he would find a mutant mouse that, when mated with his 129 strain, would increase the incidence of tumors. It was a shot in the dark. The hoped-for mouse had to have a specific genetic susceptibility for teratomas. A long three years and eight generations of mice later, Stevens met with success. Thirty percent of his mouse offspring were now heritably doomed to teratomas, and he could finally move on to a full-blown investigation.

If the tumor was caused by a deviant germ cell, he guessed he would catch the evidence in a fetus’s genital ridge, the tiny region that develops into the testis or ovary. He and his technician Donald Varnum painstakingly worked backward through mouse fetal time: gestation day 19 (birth), day 18, day 17, and so on. After endless dissections, in 1964 they at last came upon the first sign of tumor growth—on day 12. There, in the genital ridge, a germ cell that should have become a sperm cell started erroneously growing like an embryo, its cells dividing and multiplying. “By tracing the testicular teratoma back to its precise origin, we had focused on the very beginning of a cancerous process for the first time in history!” Stevens later exulted to a Jackson Lab chronicler. The journey, which had taken twelve long years, helped prove Prexy Little’s premise: Inherited mutations in genes could indeed bring about a cancer.

Now that he had the tumor on its genetic knees, Stevens wanted to design drugs to combat it. By the late ’60s, he could perpetuate a near limitless supply of tumors by transplanting a fetal mouse’s tiny genital region into the testis of an adult mouse. Teratomas resulted because the genital ridge’s germ cells, its stem cells, raced into chaos when buried in unfamiliar territory. Then Stevens did an experiment that established once and for all that an errant germ cell commenced a teratoma. He transplanted fetal genital ridges that contained almost no germ cells into the testes of adult mice—and no teratomas bloomed. What stronger proof was needed that errant germ cells caused mouse teratomas? Case closed!

An experimental embryologist to the hilt, he next grafted mouse embryos into odd places in adult mice—the kidney, for instance—and embryos frequently turned into disorganized teratomas. This was further confirmation that stem cells, in this case those in an embryo, didn’t like being plunged into alien surroundings. As witnessed here, put an embryo and its stem cells where they didn’t belong and a teratoma ensued. “This showed how easily an embryonic cell and a cancer cell could move one to another. There was a certain similarity between these two cells,” notes Davor Solter, who in 1970 was in a laboratory in Yugoslavia that made the same finding.

As Virginia Papaioannou, a developmental biologist at Columbia University, describes, “Both an embryo and a teratoma are made up of stem cells that divide rapidly. The main difference is that stem cells in the embryo have controls put on them at certain points—they know when to stop dividing.” But if they are placed outside the uterus, “those growth controls fail, leaving a tumor to develop.”

Stevens was convinced that, in a reverse situation, if one put a teratoma’s stem cells into the uterus where they belonged, they would switch from malignant to normal. Don Varnum recalls Stevens’s conviction: “‘Don,’ Roy would say to me, ‘if we could get some of those cells in the right environment, I think they would participate in normal development.’” Adds Varnum, “Well we tried, but we couldn’t get it to work.”

Despite his many trials and setbacks, Stevens nevertheless felt that luck had constantly visited him. And in the early ’70s, it did so again when a Jackson Lab colleague, Seldon Bernstein, encountered and brought to Stevens’s attention an inbred mouse strain in which roughly fifty percent of newborn females had ovarian teratomas by the age of three months. The strain and its aberration were “a great shot in the arm for me,” Stevens later recalled, for he would discover that the ovarian tumor’s cell of origin, just like that of its counterpart in the testis, was a deviating germ cell that began growing in the fetus as if it were fertilized. This false embryo “developed beautifully for quite a while,” Stevens described. “It looked just like a normal

embryo, but then it would all get mixed up. It became disorganized and you had a tumor instead of a baby mouse.”

In the 1970s, Roy Stevens’s mouse-derived teratomas and their stem cells would begin attracting worldwide attention. But before they did, there was only one other scientist on the planet who was as fixated on teratomas as Stevens, and our story at this point would be incomplete without mention of Gordon Barry Pierce and the way he so productively straddled pathology and embryology. By “the late 1950s and early 1960s, Roy or I formed a rather small, but intensely interested audience whenever the other gave a paper,” Pierce, a luminous and loquacious pathologist, later related. “To most oncologists, we were the fellows with ‘the funny little tumors.’” If people laughed at them, it wasn’t for long. Together Stevens and Pierce would become the famed pioneers of teratoma biology.

Pierce’s immersion in teratomas started in 1953, the very year that Stevens happened onto his first teratoma-burdened mouse. During his pathology residency at the University of Alberta Hospital, the Canadian had been in the impossible position of trying to stave off a testicular tumor in a three-year-old boy. Chemotherapy was not yet an effective option, and the drugs that would one day save children from this deadly tumor were still years away. When the boy died, Pierce was enormously sad and also frustrated by how little was known about testicular tumors. Determined to render them less mystifying and less lethal, he began requesting teratoma tissue from Roy Stevens in the late ’50s, and due to their mutual interest in teratomas, the two men quickly became close friends.

By ’64, Pierce had wrung some crucial findings from teratomas. He and Lewis Kleinsmith, a medical student of Pierce’s at the University of Michigan, Ann Arbor, proved what had only been postulated before, that a teratoma’s assortment of specialized cells definitely

arose from one pluripotent stem cell. (Stevens, at the time, was pinning down the actual type of pluripotent cell involved: a germ cell.) Kleinsmith and Pierce had removed a single cell from a teratoma, a cell they imagined was pluripotent, and implanted it in the abdominal cavity of a mouse, and the implanted cell had duly produced a teratoma with a teratoma’s typical bedlam of cells and tissues.

Working with Frank Dixon at the University of Pittsburgh, Pierce had also shown the extent of a teratoma’s malignancy, and their revelations proved stunning. The widespread belief was that every cell in a malignant teratoma must be malignant, both its stem cells and its mature cells. Yet Pierce and Dixon uncovered that the tumor’s malignant state was due entirely to its stem cells, for as long as they remained undifferentiated. When the stem cells differentiated, they actually stopped being cancerous. Pathologists already had a name for a teratoma’s undifferentiated stem cells: embryonal carcinoma cells, or EC cells.

The great surprise that a teratoma’s cancer appeared to lie in its stem cells, and not its mature cells, produced an uproar among oncologists. The notion that malignant cells could differentiate into benign cells “took dogma right by the throat,” says Pierce, because it ran counter to the conviction that once a cancer cell, always a cancer cell. A cancer cell was only supposed to “divide and divide; that’s part of the ritual of being malignant.” It wasn’t supposed to be able to change to a benign state, or become anything other than what it was.

Today, Pierce speaks glowingly of the experiments Armin Braun performed at the Rockefeller Institute (now Rockefeller University) in the early ’60s that also left convincing evidence that tumor cells could become normal cells. What might seem curious, yet shouldn’t given all that plants and animals share at the molecular level, is that plants can also be afflicted with a teratoma. Braun used the cells of teratomas from tobacco plants affected by crown gall disease—these plant tumors being as much of a jumble of immature and mature cells and tissues as an animal teratoma—and he showed that tumorous buds, if grafted to normal plant tissue, could regain normalcy.

Braun recognized that genes within the cells were behind this changeover. “The cells were made tumorous simply by activation of some of the ordinarily inactive genes; the cells returned to the normal state when those genes were again repressed and rendered non-functional,” he observed in a Scientific American article.

Pierce concluded in the early ’70s that malignant stem cells were but “caricatures” of normal stem cells and originated when normal stem cells slipped out of their regular routine. The observation was not a new one. “But Barry is the one who really provided the experimental basis” for what he referred to as a stem cell theory of cancer, says Ralph Parchment, a pharmacologist at Wayne State University and former student of Pierce’s. At the core of the theory: If too many undifferentiated cells form and not enough of them mature, their nonstop self-renewal could lead to cancer’s destructive growth. Pierce’s observations prompted him to recommend a treatment measure. Rather than attempt to kill cancer cells through such toxic means as chemotherapy, he entreated, why not try to control their rampant division by getting them to differentiate into benign tissue?

To this day, Pierce stands adamantly behind this idea. However, the evidence for “differentiation therapy”—the practice of attempting to cure cancer by getting its immature cells to mature—is far from conclusive. One of several questions posed by skeptics is, if a cancer stem cell can be changed into a benign state, say by treating it with a certain chemical, would its genes really and truly be normal?

A remarkable experiment in the mid-’70s nonetheless provided confirmation, in the opinion of many scientists, that cancerous stem cells indeed could be nudged into a normal state. The question posed by three separate groups was, if you place malignant EC cells from mouse teratomas in the early embryo, will this normal environment impose the right controls and induce EC cells to differentiate into benign, normal cells and tissues? Roy Stevens, as we saw, often said as much to his assistant Don Varnum: “Don, if we could put these cancerous stem cells into a normal embryo, I swear they’d go back to being normal.” Upon injecting a teratoma’s stem cells into early

mouse embryos, two American teams and one British team basically got the same results. Most of the time these implanted cells contributed to the development of normal adults; their malignancy swung toward normalcy. The groups that independently published this evidence were led by Ralph Brinster at the University of Pennsylvania, Beatrice Mintz at the Institute for Cancer Research, and Richard Gardner at Oxford University.

With the blossoming of the above experiment, a vision was coming alive. What if you could alter the genes of stem cells in ways that diseases alter genes, and then return the modified stem cells to the embryo? As long as the modification ended up in the developing embryo’s egg and sperm cells, you might be able to create entire lines of changed mice that mimic a specific human disease. These transgenic animals—animals that have been permanently changed in this fashion—might be invaluable for studying all sorts of diseases. Since normal embryonic stem cells were not yet obtainable from mice, a teratoma’s malignant stem cells were at the center of this vision.

It all hinged, however, on getting the gene changes into an animal’s germ line—the lineage of egg and sperm cells that carry the genetic code of one generation over to another with ceaseless continuation. If successful, scientists would then know how to modify Life.

These days, as long as a child’s or adult’s teratoma is discovered in time, this cancerous tumor is virtually curable through the administration of cisplatin, a chemotherapy drug. Cisplatin is one of a trio of drugs that helped restore bicyclist Lance Armstrong to health when he was diagnosed in 1996 with a testicular tumor that contained a small element of teratoma.

Cisplatin and other chemotherapy drugs arrived irrespective of the ground covered by Roy Stevens and Barry Pierce. And yet these famed teratoma pioneers produced breakthrough observations that would impact the far corners of biology, and therefore the study of disease and medicine as well. Barry Pierce, who is now retired and spends much of his time attending to his bonsai collection and working at his bench—his woodworking bench—alerted many scientists to the possibility that some cancers might be a stem cell disease. Today, more than ever, cancer researchers are investigating this claim.

Roy Stevens, who is also retired, still lives within an arm’s length of the ocean. Because of his lucky sighting back in ’53 of a most unusual mouse, biology’s horizons would expand in every direction. Researchers throughout the world would request his famous mouse, eager to study its tumors, their precocious little cells, and the rainbow of mature cells derived from them. Even a tumor’s stem cells had lots to say, and embryologists, cell biologists, geneticists, pathologists, oncologists, and other investigators would lean in to listen.