Science, my boy, is made up of mistakes; but they are mistakes which it is useful to make, because they lead little by little to the truth.
—Jules Verne, Journey to the Centre of the Earth1
BIOLOGISTS HAD BEEN growing cells in culture since 1907, when Ross Harrison, the workaholic embryologist at Johns Hopkins, first coaxed those bits of frog brain to grow in a lab dish. For virtually all of that time, they had operated under this simple piece of received wisdom: cells grown in lab dishes, properly treated, should live indefinitely. If they died, the fault lay not with the cells but with the scientist. His glassware wasn’t clean, or her medium didn’t contain just the right mix of nutrients, or a sloppy technician had sneezed on a plate of cells, launching a fatal infection. This faith in the open-ended life of cells in the lab grew out of the work of a bald, bespectacled, publicity-seeking French scientist named Alexis Carrel. In 1912 Carrel was awarded the Nobel Prize for inventing a much-needed method for surgically joining the two ends of a severed artery. But for Hayflick and other cell culturists, Carrel’s half century of influence resulted from an entirely different experiment.
In the early years of the twentieth century, the charismatic Carrel worked at the Rockefeller Institute in New York City, then the pinnacle of American medical science. One mid-January day in 1912, Carrel took a snippet of tissue from the heart of an eighteen-day-old chick fetus and put it in a lab dish, where he began nourishing it with diluted chicken blood plasma. (“Plasma” is another word for blood serum.) Two months later, having grown so “abundantly” that it had been split and planted in new vessels eighteen times, the heart tissue was still alive and even, he reported, beating in the lab dish. Carrel published a paper reporting on his findings, entitling it “On the Permanent Life of Tissues Outside of the Organism.”2
When, later that same year, he won the Nobel Prize for his artery-joining advance, many scientists and journalists mistakenly thought that Carrel was being honored for the launch of the undying chicken heart. The miraculous beating tissue “was the leading topic of discussion by medical men the world round,” the New York Times reported in an article on Carrel’s Nobel Prize that consumed the entire front page of the broadsheet’s Sunday magazine section.3
Carrel soon handed off the work of maintaining the beating bit of chicken heart to a laboratory colleague, Albert Ebeling, who looked after it for the next thirty-four years, feeding it with medium, dividing the cells when they outgrew their space, and dispensing with most of them while keeping a residual piece of heart always going in a dish. The undying chicken heart became a favorite with the popular press. “Isolated Tissue Holds Life 12 Years in Test … Growth Continues as in Body,” the New York Tribune proclaimed in 1924, as the chicken heart approached its twelfth “birthday.”4 So did Carrel himself, who at one point landed on the cover of Time magazine—although he likely hastened his retirement from the Rockefeller Institute when in 1935 he published a book, Man, the Unknown, propounding the use of gas to euthanize criminals, both sane and insane.5 The heart-in-a-dish outlived him; Carrel died two years before Ebeling finally disposed of the culture in 1946.
The lesson from Carrel’s chicken heart was not lost on the biologists who struggled through the first half of the twentieth century both to repeat his chicken-heart experiment—no one could—and to keep other kinds of cells alive indefinitely in culture. When they failed, as they did repeatedly, it was surely their incompetence that was to blame. That belief was only strengthened when, in 1943, a round-faced, bespectacled cell culturist at the National Cancer Institute coaxed a single cancerous mouse cell into unending lab-dish life.6 Wilton Earle’s mouse cells became the first demonstrably immortal cell line—one that scientists could obtain from Earle and observe for themselves replicating endlessly in the lab.
It was 1951—the year that Hayflick earned his undergraduate degree—before the same feat was repeated with human cells. That year an innovative and determined cancer researcher and cell culturist named George Gey, working at Johns Hopkins University, launched the first human cells to survive indefinitely in the lab, by taking the cells of an extremely aggressive cervical cancer from the womb of a dying woman named Henrietta Lacks. Within a few years the HeLa cell line was being studied and used in biology labs all over the world.
With the perspective of hindsight, it’s easy to say that Earle’s mouse-tumor cells, Gey’s HeLa cells, and other undying cell lines that soon followed in the 1950s were able to live endlessly in the lab only because they were cancerous. After all, it is the very definition of cancer that cells escape the normal constraints on growth and divide uncontrollably and endlessly. But for Hayflick, working nearly sixty years ago, things were not so clear-cut. Indeed, there was no reason to believe that normal, noncancerous fetal cells—just like Carrel’s normal chicken heart cells—shouldn’t also live open-endedly in lab dishes.
So when one day in the winter of 1960 Hayflick noticed something amiss in the incubation room, he started looking for his own mistakes. By now there were several other fetal cell lines in addition to WIHL—the lung cells from the first fetus he had received—multiplying in glass bottles on the wooden shelves.
During late 1959 he had continued to receive fetuses at random intervals from the Hospital of the University of Pennsylvania. With them he had developed cell lines from several fetal organs: skin and muscle; thymus and thyroid glands; kidneys and heart. He found that when the bottles were first planted with cells, it took about ten days for the cells to establish themselves, growing to cover the floor of a bottle in a semiopaque sheen that was visible to a practiced eye. (Hayflick labeled this initial stage of the cells’ growth, when they grew to first cover the bottom of the bottle, phase I.)
Once he had split the contiguous cells, putting half of them into a new bottle, all of the cells began dividing much more rapidly and needed to be split again every three to four days. (He called this period of rapid division, which went on for months, phase II.) He found that no matter how large the variety of cell types in the organ he started with, all of the cultures ended up consisting of only one cell type: the long, tapered fibroblasts that spin the connective tissue that holds organs and cells together. For whatever reasons, other kinds of cells did not thrive in his bottles.
As he planted these cell lines, Hayflick became systematic about naming them. Since he soon developed a second cell line from fetal lungs, he needed to rebrand the first line something other than Wistar Institute Human Lung. He decided to keep it simple and name the lines in numerical order. WIHL he renamed WI-1. The next line he named WI-2, and so on. By September 1960 he would launch WI-25, the last in the series. It came from the lungs of fetus number 19, a female. (There were fewer fetuses than cell lines because Hayflick derived more than one cell line from several of the fetuses, using different organs. For instance, the second line he developed, WI-2, was grown from the skin and muscle of fetus number 1.)7
As was normal in science, there had been mishaps. WI-2 had been lost to bacterial contamination after growing for several months. And cells from some organs, he was learning, grew better in the lab than others. Cell lines from the heart were sluggish: his WI-6 cells had given up and died after not even three months. But the kidney cells seemed to be doing well, and his WI-1 lung cells were star growers. He had begun to suspect that lung fibroblasts were the best suited to life in the lab.
But what was wrong with the WI-3 cells on this day? The WI-3 line had been grown from the lung cells of fetus number 2 and, if his hunch was right, these cells should have been dividing and thriving, just like the lung cells of WI-1. But he had noticed during the past several weeks that the WI-3 cells were taking longer than they had in the past to grow to confluence on the bottom of their bottles. What was more, their culture medium wasn’t turning from pink to yellow as quickly as it once had. The change in the fluid’s color resulted from acid production and meant that the cells were metabolizing actively. A slower progression from pink to yellow meant that the cells were slowing down in their activities of daily living—acting, in oversimplified terms, a lot more like seventy-year-olds than twentysomethings. And now, five months after the launch of the third line, WI-3, there was a new, worrisome sign. The culture medium was normally clear. But today it was cloudy with what he feared was the debris of dying cells.
Hayflick took the bottle of WI-3 cells next door to the lab and peered at the cells through his inverted microscope. He saw what he had expected he might see: there were grainlike bits of debris scattered around. What was more, when he scrutinized the tangled, dense, dark chromosomes that are visible at high magnification in actively dividing cells, they were very few and far between. The cells were slowing, perhaps stopping, their division.
Over the coming months the WI-3 cells would completely degenerate. Healthy fibroblasts tend to line up like soldiers in parallel formations, with each cell immediately next to its neighbors, their finely tapered ends pointing in the same direction. Instead, the WI-3 cells would become spread out in no discernible pattern, pointing in random directions. Black bits of debris, the detritus of dying cells, would litter the white spaces between the cells on Hayflick’s microscope field and cling to the cells’ surfaces like so much washed-up driftwood.
The changes would come on slowly, but at this first sign of the cells’ deterioration, Hayflick began trying to figure out where he was going wrong with WI-3. He adjusted the components of the culture medium—perhaps it was deficient in some essential nutrient—and poured the rejiggered fluid over the lagging cells. But even as he did so, he felt that it wouldn’t make a difference. The medium he used was made by technicians in the basement media room, in large batches that lasted for weeks. If WI-3 was struggling, then the other cells that he had bathed with precisely the same medium should be struggling too. Yet the rest of the cell lines were thriving. His suspicion was confirmed when the readjusted medium failed to resuscitate the WI-3 cells, whose bottles continued to grow cloudier with debris.
Perhaps he was screwing up in some other way. Could dirty glassware be the culprit? But the lab had been using the same glassware, scrubbed and sterilized by the same people working in the same glassware-washing facility in the bowels of the building. Perhaps, occasionally, dirty bottles slipped through due to some oversight, but what were the odds of those random dirty bottles always being used for WI-3 and no other cell lines? Virtually nil.
The only really plausible explanation was that WI-3 had become infected. It was easy enough to rule out bacterial contamination: he planted some sluggish WI-3 cells on plates of agar and left the plates incubating. No bacteria grew. There was no bacterial contamination stunting these cells.
Other microbes were trickier to detect. One key group of suspects was the PPLOs—those nuisance organisms that were smaller than bacteria but larger than viruses and that Hayflick had studied as a graduate student. They were a bane of cell culturists: always popping up like weeds where they weren’t wanted, even after a scientist thought they’d been eliminated with antibiotics. Fortunately for Hayflick, he had made himself into an expert on identifying them microscopically. Colonies of PPLOs had a “fried egg” appearance, but it took a practiced eye to pick this up. He checked the ailing cultures and found no signs of PPLOs.
There was still the possibility that a virus was dooming the WI-3 cells. Hayflick examined them under his microscope for telltale signs of viral invasion. Viruses produce typical microscopic signs when they sicken cells. Cells become strangely shaped. They bloat. They may detach from the glass bottles that they normally adhere to. They can develop “inclusion bodies,” which are clumps of abnormal protein in the nucleus or cytoplasm. Hayflick saw none of these signs in the WI-3 cells. Of course, no matter how hard he looked, he couldn’t prove a negative—that is, he couldn’t prove that an undiscovered, undetectable virus wasn’t lurking in the cells. This was a reality that would eventually come to dog him.
Hayflick did all he could to revive WI-3. He kept splitting the cells into new bottles. There they continued to languish. Then he tried crowding several bottles’ worth of degenerating cells into one bottle. Nothing changed.
Had Hayflick not been so industrious and launched twenty-five cell lines from the organs of nineteen fetuses over the course of months and months—rather than, say, deriving just two or three cell lines—he might at this point have concluded, like so many scientists had before him, that he was simply falling short of the high standard set by the illustrious, Nobel Prize–winning Carrel. He might have concluded that his own ignorance was the culprit and the WI-3 cells were its victim.
However, as WI-3 languished in front of him, Hayflick was watching over his other cell lines. They were growing well until one day a few weeks later, when he again visited his incubation room and noticed that some other Blake bottles were becoming cloudy with debris. They were the bottles bearing WI-4, a cell line launched from the kidneys of fetus number 3 that had been doing very well—until recently, when the cells had begun multiplying more slowly. Now, like the WI-3 cells before them, they were grinding to a halt, succumbing to something. What was more, his bottles of WI-5, derived from the muscle of fetus number 3, were beginning to take longer to grow to confluence. By the time these muscle cells stopped dividing, the next group of cells, WI-7, from the thymus and thyroid glands of fetus number 4, had begun to grow sluggishly.
As Hayflick watched his newer cell lines flourish in their bottles while the cells from the first fetuses languished and finally died around them, he took his perplexity to several Wistar colleagues. One of them was Lionel Manson, a portly, avuncular immunologist with a dry, self-deprecating sense of humor and a razor-sharp intellect.
“I was telling Lionel what I found and I said, ‘I’m weighing several different explanations,’” Hayflick recalled in a 2014 interview. “And he just said cavalierly: ‘Have you thought about it having to do with aging?’ And I said: ‘No. But,’ I said, ‘aging is a wastebasket’—at that time it was; it still is to some extent—‘a trash basket into which you put everything you can’t explain.’”8
Still, Hayflick took Manson’s flippant suggestion back to his lab and pondered it. And the more he thought about it, the more convinced he became of its merit. It was the only theory that was supported by his now-voluminous data and his months of observations. And it led him inescapably to one conclusion: there was nothing wrong in his methods. What was wrong was the scientific faith in the immortality of cells in lab dishes that dated back nearly fifty years, to Carrel and his never-dying chicken heart. True, cancer cells like Wilton Earle’s mouse cells and Gey’s HeLa cells had been living in labs for years now, and it seemed they would go on doing so. But normal, noncancerous cells in a lab bottle were not immortal. They aged and died, just like human beings. The truth of it was staring him in the face.
Hayflick rehearsed the objections that he could already hear a chorus of critics raising. True, he couldn’t explain how it was that Carrel’s chicken heart tissue had continued to live for decades. But wasn’t the standard of scientific credibility the repeatability of an experiment? No one had been able to repeat Carrel’s experiment. The reason, he was now increasingly certain, was because it wasn’t repeatable.
There was also the matter of WI-1, the first cell line he had launched, back in September of 1959. It was still going strong, dividing energetically, six months later. But in this waning winter of 1960, he was now almost certain that this could be explained: WI-1 was a particularly hardy, long-lived line, but not an immortal one. He would bet money that sooner or later it would die too.
There are two subtle but important points worth making at this juncture about nomenclature; both will bear on this story as it moves forward. The first arises from what is actually happening in a bottle of fetal fibroblasts lying on its side incubating. If one takes a bottle, the flat side of which is covered with cells that have grown to confluence, and splits the cells in half, planting half of them in a new bottle of the same size, it might seem to make sense to conclude that, once the bottoms of the two bottles are covered with cells, every cell in that initial bottle (the original bottle is commonly called the “mother” and the new one the “daughter”) has divided once. One would be wrong.
Cells, like people, vary in their vigorousness. Some cells divide more sluggishly, while some are eager, rapid replicators. So over a given period of time, some cells will replicate fewer times than others, some perhaps not at all. Which means that the only conclusion that can be drawn when the floors of the two bottles are eventually covered with cells is that the initial population in the mother bottle has doubled in size. For it now covers twice the area that it did.
This is why biologists don’t speak about individual cells doubling or say that “these cells have now doubled in number five times.” Instead they refer to “population doubling levels,” which they describe with the acronym “PDL.” Given the inherent variability of individual cells, it’s the only accurate term to use.
There is a second issue with terminology that can be confusing. It is this: Whenever a biologist splits a confluent culture of cells and puts some of them in one or more new bottles, this is called a “passage,” because cells have “passed” from one bottle into another. However, scientists will often use the term “passage” as a synonym for a cell population’s doubling. This is accurate if, and only if, as the cells move through a sequence of passages, half of them are placed into just one new daughter bottle of the same size as the mother bottle. This is called a 2:1 split.
However, cells can be placed into any number of new bottles. For instance, if three quarters of the cells are removed from a confluent mother bottle and placed in equal portions into three additional daughter bottles of the same size, and then the cells in all four bottles are allowed to grow to confluence again, the original cell population will clearly have quadrupled in size. Yet it will have been through only one “passage,” in that the cells will have been “passed” into new bottles just once. The terms “passage” and “PDL” will be used interchangeably in this book—only because, in the experiments involved, the cells were routinely put through 2:1 splits.
As the winter of 1960 turned into spring, the U.S. Food and Drug Administration approved the world’s first officially sanctioned oral contraceptive pill. The University of Pennsylvania prepared to open its new Women’s Residence Hall. And in a speech to newspaper editors in Washington, DC, the Roman Catholic Democratic presidential candidate John F. Kennedy announced that “my religion is hardly, in this critical year of 1960, the dominant issue of our time.”9
The Hayflick family had recently moved into a modest, three-bedroom brick house in the leafy suburb of Penn Wynne, just northwest of the city on the edge of Philadelphia’s fashionable Main Line. Across the street was an expansive, hilly park where the Hayflick children spent hours playing hide-and-seek, hunting for frogs in a stream, and swinging on a rope swing with a wooden seat. The couple would soon add a fourth bedroom to accommodate their growing family: daughters Rachel and Annie were born in 1963 and 1965 respectively.
That summer and fall Hayflick’s fetal cell lines continued to age and die. He dubbed the stage in which their dividing slowed and stopped and they degenerated and finally died “phase III.” And he wrestled with the question of how to prove that it was something intrinsic to the cells—some inherent property and not anything in their environment—that was the cause of their mortality. Hayflick’s former colleague and friend from Galveston, the chromosome expert Paul Moorhead, had since moved to the Wistar. In a 2012 interview Moorhead recalled that it was he who proposed the simple, elegant experiment that did the trick.10
Moorhead’s lab at the Wistar reflected his lowly designation as a postdoctoral fellow: it was a tiny cupboard of a place up on the third floor. But the chromosome aficionado from Arkansas had there what counted most: a Zeiss Jena made by the Reichert Company—in his opinion, the best microscope that money could buy. Leaning over the microscope’s sturdy black base, he could paste both eyes to the eyepieces and peer at chromosomes at eight hundred magnifications.
About the time that U.S. voters went to the polls and elected the forty-three-year-old Kennedy to replace the seventy-year-old Dwight D. Eisenhower, Hayflick took his oldest and his youngest fetal cell lines and mixed them. The first were the now-elderly WI-1 cells, which, as he had expected, had stopped dividing in late summer, after eleven months. They were now in phase III—still alive, still metabolizing, but ever so slowly. They had been split in their bottles forty-nine times and might perhaps divide once or twice more, if they could screw up the energy. But basically, they were reaching the end.
To these WI-1 cells Hayflick added youthful WI-25 cells that he had launched just weeks earlier and that were vigorously replicating. They would do so, he expected, for months to come, for the WI-25 bottles had been split a mere thirteen times. Hayflick then left the mix of young and old cells, both in the same bottle, being nourished by the same medium, to incubate.11
Moorhead’s stratagem relied on the fact that, apart from their ages, there was a singular difference between the two groups of cells. The WI-1 cells had come from a male fetus and thus bore one X chromosome and one Y chromosome. The WI-25 cells were from a female fetus, so they bore two X chromosomes. Both kinds of chromosome would be visible under the microscope to Moorhead’s expert eye.
After the cells had been incubating for about two months and had been split into new bottles seventeen times, Hayflick handed the mixture to Moorhead. The Arkansan expert carefully prepared the cells, using techniques that stained and spread the chromosomes so that they were individually visible, rather than messed together in a tangled pile. Then he studied them at hundreds of magnifications. He saw virtually only X chromosomes. The younger, female cells, now having undergone about thirty total divisions, were thriving. As for Y chromosomes, Moorhead spotted vanishingly few. The male cells, which had already been elderly at the start of the mixing experiment, were gone. Dead.
The mixing experiment had clinched the case. If the cell-killing factor had been in the glassware or in the medium, or if it had been some other technical slip, all of the cells would have been exposed to the problem and all of them would have been dead. What was more, as Hayflick and Moorhead wrote archly in a paper that would be named a “Citation Classic” because other papers referred to it so frequently, “If a latent virus had been responsible … it seems unlikely that it would have been able to discriminate between male and female cells.”12
Hayflick could now confidently assert that something intrinsic to the cells was behind what he was seeing over and over again with his own eyes. Something inside them was causing them to die. Admittedly, WI-25, the last cell line he had launched, was still dividing, as were several other more recently derived lines. But he was now sure that these too would eventually slow their replicating and then stop. Then they would degenerate and die. Not one of the twenty-five cell lines he had launched was immortal.
Hayflick was thirty-two years old and a virtual unknown in the rarefied universe of top biologists that he would have loved to inhabit. He was faced with the prospect of making an audacious claim. A claim that would challenge fifty years of received wisdom, along with the reputation of the Nobel Prize–winning Carrel. A claim that normal cells aged and, finally, died in their lab dishes. He was nervous. And his confidence wasn’t helped by a warning from one of the most respected cell biologists of the time. Gey, the talented developer of the HeLa cells, was visiting the Wistar Institute one day when Hayflick confided in him his new findings. In a 2012 interview Hayflick recalled Gey’s response: “Be careful, Lenny. You’re going to ruin your career.”13
Gey couldn’t have been more mistaken. Hayflick’s finding would one day make his name and distinguish him as the man who opened the door to a whole new realm—the study of cellular aging. It is an area of huge relevance to two of our top health preoccupations today: aging and cancer. But Hayflick had a long road to travel before recognition came.
And how to explain Alexis Carrel’s normal chicken-heart cells, replicating faithfully in the lab from their launch in 1912 until Carrel’s colleague Ebeling finally dispensed with them in 1946? Years later, in the 1960s, a woman who had worked as a technician for Carrel in the 1930s approached Hayflick after he gave a talk at a scientific meeting. In his lecture Hayflick had speculated that Carrel’s method had a fatal flaw related to the fluid that he and his technicians extracted from chick embryos and used to “glue” the chicken-heart cells to the floor of a new culture vessel whenever they overgrew their current dish and needed to be divided. This fluid extract from chick embryos was also used daily to feed the cells. In preparation for this, it was spun in an antiquated centrifuge, a process that was supposed to remove cells, leaving only nutritious fluid. Hayflick proposed that, with or without Carrel’s knowledge, the fluid extract actually contained errant fresh cells from the chick embryos; that the culture stayed alive for decades because it was frequently replenished with these new young cells. The woman told Hayflick that he was on the mark; that in the 1930s she had raised questions with Carrel’s chief technician indicating that she thought this might be happening. She had been told to forget what she was seeing or risk losing her job. In the midst of the Depression, that was not something she was eager to do.14
It is a measure of Hayflick’s productivity, energy, and ambition that in the autumn of 1960, as he and Moorhead conducted the male-and-female-cell-mixing experiment, he was also using his new human fetal cells to develop a first-of-its-kind polio vaccine—more on this will follow—as well as running several studies that he knew would be crucial to the successful reception of the landmark paper that he and Moorhead were now putting together.
(Hayflick also, at about this time, codiscovered the cause of walking pneumonia. The culprit was a species of Mycoplasma, the tiny microbes that he had studied as a graduate student. Mycoplasma pneumoniae was the first of these microbes discovered to cause disease in humans, and the New York Times splashed the discovery at the top of its front page.)15
That paper would boldly hypothesize that the cells’ deaths in their bottles were the outcome of “[aging] at the cellular level.”16 And it would define what later came to be known as “the Hayflick limit”—the number of divisions that a normal cell in culture can undergo before it ceases to divide. Based on the data from his twenty-five fetal cell lines, Hayflick estimated this number at fifty divisions, plus or minus ten. Importantly, freezing the cells didn’t affect the Hayflick limit. For instance, when Hayflick froze some WI-1 cells that had divided just nine times, and then thawed them months later and put them in the incubation room, they began dividing again and, seeming to “remember” their age, went through forty-one more divisions over five months before dying in their bottles. What was more, this fact—that the cells would commence dividing again after being frozen and thawed—meant that freezing appropriate numbers of them at young ages could ensure an all-but-endless supply of cells into the future.
But for what practical purpose would such an endless supply of cells ever be needed? It’s in the answer to this question that the enormous and lasting public-health impact of Hayflick’s work rests. For in his groundbreaking paper with Moorhead, Hayflick went beyond an iconoclastic assault on the immortality of normal cells. He also suggested that the new cells could make a big contribution to vaccine making.17 He had conducted the experiments to prove it.
In 1960 the making of new viral vaccines was a top priority for virologists and a goal that was eminently within reach, thanks to the technical breakthroughs of the previous two decades. The fight against polio, fresh in everyone’s minds, had shown as much. It had been a terrifying bane—the disease stirred something like the fear that Ebola does today. But in the space of five years, since the launch of Jonas Salk’s killed vaccine in 1955, it had been reduced to a preventable disease. What was more, it was becoming clear that a stronger, longer-lasting live polio vaccine was within a year or two of approval by U.S. regulators.
Now the prospect of developing vaccines against other viral diseases beckoned to scientists. Measles, mumps, and German measles, also called rubella, were regular childhood afflictions. Hepatitis was rarer, and gravely serious. Chicken pox was a particular bane for children with weakened immune systems. For some diseases, like rubella and hepatitis, vaccines couldn’t yet be made, because the viruses hadn’t yet been captured in lab dishes. Virologists set out to hunt them down. For others, like measles and mumps, the viruses had been isolated, and scientists were hurrying to develop vaccines. The U.S. government soon provided the money and muscle to make sure that the new vaccines were used. In 1962 President John F. Kennedy signed the Vaccination Assistance Act, allowing the Communicable Disease Center (CDC) to support mass immunization campaigns and ongoing maintenance programs and to funnel vaccine money and resources directly to state and local health departments.
As well as the challenge of creating new vaccines, there were vexing problems with existing ones. One of two existing rabies vaccines, made in dried animal brains, could produce fatal allergic reactions; the other, made in duck embryos, was not as effective as the animal brain–produced version. And a silent monkey virus had been discovered in monkey kidney cells used to make the Salk polio vaccine. As he prepared his paper in the autumn of 1960, Hayflick, unlike the public, was keenly aware of the monkey-virus problem.
It wasn’t lost on Hayflick that his human fetal cells might provide clean, safe alternative microfactories in which to produce viral vaccines, if—and it was a big “if”—they could be infected with disease-causing viruses. There was precisely one way to find out if that was the case. Hayflick began infecting bottles of the WI cells with different viruses. They turned out to have a huge range of virus susceptibility: thirty-one viruses invaded and damaged the cells. These viruses included measles, rabies, herpes simplex, adenovirus, influenza, and polio. The cells even succumbed to invasion by varicella, the virus that causes both chicken pox and shingles but is extremely choosy about which cells it will grow in.18
The implications were exciting, at least in Hayflick’s eyes. Public and scientific interest in antiviral vaccine making was surging. At this juncture it could be a big advance to introduce a new, safe, plentiful supply of cells for vaccine making. But Hayflick also knew that it was going to be an uphill battle to convince vaccine regulators that his cells were safe. After all, the hot area in cancer research was the purported potential of as-yet-unidentified hidden viruses to cause cancer. Vaccine-approving agencies were going to want assurances that no such unidentified viruses lurked in his human fetal cells. Nothing would put them off the cells more quickly than the possibility of such a virus getting into a vaccine and later causing cancer in vaccinated people.
Hayflick examined his two dozen fetal cell lines under the microscope repeatedly. He found no telltale signs of viral infection. He took the fluid bathing the cells and injected it into cultures of other kinds of cells, and into animals. Neither the cells nor the animals showed any signs of infection.19
Then he turned again to Moorhead to scrutinize the cells’ chromosomes. Virtually all cancers had abnormal-looking chromosomes and abnormal numbers of chromosomes. If any of the WI cell lines showed such deviations, that too would mean a no-go from regulators. Their thinking would run like this: If the fetal cells harbored abnormal chromosomes, they were either cancerous or would soon become so. If this was the case, then there was every chance that a hidden virus in the cells had caused the malignant changes. And if a hidden, cancer-causing virus was in the cells and they were used to make a vaccine—well, there was an epidemic of cancer just waiting to happen.
Hayflick waited while Moorhead stared at sample after sample of the fetal cells, counting chromosomes painstakingly, hour after hour, week after week. He looked at young cells, which had divided just nine times. And he was careful to look at old ones too: the oldest cells he examined had been through forty replications. Since chromosomes were particularly vulnerable to developing cancer-associated abnormalities during cell division, the older cells, having divided more times, were most at risk of showing anomalies.
When Moorhead finally looked up at the end of weeks of work, in the autumn of 1960, he had good news for Hayflick: these cells, young and old alike, were normal cells, not aberrant ones. They were diploid cells, with twenty-three pairs of chromosomes, for a reassuring, normal total of forty-six. And so Hayflick dubbed the fetal cells “human diploid cell strains.” The term “cell strain” was a very deliberate substitution for “cell line” on Hayflick’s part. In Hayflick’s nomenclature a “line” is a group of cells that will go on dividing endlessly, whereas a “strain” denotes a group of cells that is mortal—that will reach the Hayflick limit and then expire.
Hayflick was now armed with reassuring chromosome data, which Moorhead documented with striking photos showing the chromosomes of several of the WI lines neatly laid out in twenty-three numbered pairs. But Hayflick wasn’t finished amassing evidence for the safety of the cells. He enlisted a scientist friend of his, Anthony Girardi, who worked at Merck in nearby West Point, to conduct another key experiment, this time on hamsters—an additional study aimed at persuading regulators that these cells were not cancers-in-waiting.
Hamsters’ big cheek pouches don’t have the normal immunological defenses that attack foreign invaders, including cancer cells. That makes them useful to biologists, who, as Hayflick worked, had already shown that cancer cells formed ever-enlarging tumors when they were injected into hamster cheek pouches. Hayflick’s friend Girardi injected the animals’ cheek pouches with vigorously growing WI-25 cells. If the cells had any propensity to turn into cancers, several weeks growing in hamsters’ cheeks would give them plenty of opportunity to do so. For comparison, Girardi injected five additional, control hamsters with aggressively cancerous HeLa cells.20
The HeLa-injected animals sprouted tumors in their cheek pouches, and those tumors were still growing after three weeks. None of the WI-25–injected animals developed cancer.21 These findings were reassuring, but for Hayflick they still weren’t enough to counter the fears he anticipated among vaccine regulators. Hamsters were not human beings. The cells, he determined, had to be put into people.
Hayflick believed that the risks of such an experiment would be minimal. He was aware that a few years earlier, a high-profile cancer researcher named Chester Southam had taken microscopically normal-looking cells—fibroblasts, the same cell type as Hayflick’s fetal cells—from a human embryo and injected them under the skin of three dying cancer patients whom he described as “volunteers.” Southam, who was chief of virology at the prestigious Sloan-Kettering Institute for Cancer Research in New York City, reported that the embryonic cells had not grown in the dying patients, unlike the cancer cells that he had simultaneously injected in the same patients.22
Southam’s name has come to be notorious among medical ethicists, and his studies, which also included injecting aggressive cancer cells into healthy prisoners at the Ohio State Penitentiary, would ultimately lead to a public outcry in 1963. That year three doctors at the Jewish Chronic Disease Hospital in Brooklyn, New York, would resign in protest rather than inject unsuspecting patients with cancer cells for Southam.23 Their action launched a lawsuit, a state attorney general’s investigation, and a media storm that both accelerated and reflected changing public perceptions. Southam ended up being put on probation by New York’s medical licensing authority for one year. Many in the medical establishment were not convinced that he had done anything wrong, and soon after his probation ended he was elected president of the American Association for Cancer Research.24 But by the mid-1960s, ordinary people were becoming less willing to give scientists carte blanche to tinker with human beings on a “Trust me, I know what’s best for you” basis.
That was not the case in 1957, which is the year that the Southam study injecting the normal-looking embryonic cells, along with cancer cells, in dying cancer patients was published in Science, the leading American science journal. The notion that dying patients, who had nothing to gain by participating and who were enlisted in a study by their doctors, who had everything to gain by their participation, could be called “volunteers” was accepted then. So was the practice of putting such patients at risk. Southam implied as much when he wrote in Science that they “had advanced incurable cancer and a very short life expectancy.”25 And so, as they planned for the injection of their own human fetal cells into dying cancer patients, Hayflick and Moorhead felt not only that they were doing nothing wrong, but also that they were following in the footsteps of an eminent man of science. Indeed, their paper would state that Southam’s experiment had laid the groundwork for their own.26
It made sense for them to turn to Robert Ravdin, the son of the powerful HUP surgeon in chief I. S. Ravdin, for help injecting the diploid cells into dying cancer patients; Ravdin specialized in cancer surgery. A comment that he made in this era suggests that he likely had no compunction about the injections. In 1964, when Chester Southam was on the public hot seat, Robert Ravdin, defending Southam, would tell a reporter that if every subject in a human trial had to be fully informed, everyone would need a PhD.27
Being a hard-charging surgeon—a prince of the hospital, if not the king—the junior Ravdin did what surgeons do. He delegated the task downward, to a second-year surgical resident named William Elkins. The twenty-eight-year-old Elkins had an impeccable pedigree: he had attended Princeton as an undergraduate, then gone on to Harvard Medical School. But as a surgical resident at HUP, he was becoming convinced that he was not cut out for the punishing, testosterone-driven world of the operating room. What he wanted to do was science. The next year he would move to the Wistar and begin a long research career in transplantation immunology.28 As Thanksgiving 1960 approached, he was still slogging it out on the surgical wards when his boss, the junior Ravdin, asked Elkins to do a menial chore: inject some fetal cells belonging to Wistar scientists into a few dying patients.
It’s not clear if, how, or when these patients were informed about the experiment or what they understood of the process and its purposes. Both Hayflick and Elkins, recalling these events more than fifty years later, conjectured that Ravdin approached the patients and explained the experiment. However, it would emerge in the mid-1960s that in dozens of studies in this era patients were not informed that they were experimental subjects. It seems at least possible that that was the case here.
Hayflick chose WI-1 cells for the injections. These were not young cells, and he wanted it that way. If the WI cells were going to morph into cancer cells, these older cells were the likeliest to do so. Conversely, if these aged cells weren’t cancerous, his case that they were normal would be all the stronger for the fact that he had used older cells.
Hayflick gave Elkins two lots of WI-1 cells. He had grown both of them, then frozen them for months, and then thawed them and grown them some more. One group had divided thirty-seven times. The other was still older, having replicated forty-five times.29 If any one of the cells was cancerous, he would expect it to form a tumor at the injection site. A biopsy of the tumor would reveal abundantly growing cells with the microscopic hallmarks of cancer.
One day late in 1960, Elkins put a small syringe fitted with a fine needle on a tray alongside a test tube containing a sterile solution of salts and water. In each ounce of that saline solution there were some 177 million living WI-1 cells. Then he headed to the surgical wards.
Visiting a dying cancer patient, Elkins turned one of the patient’s forearms over, revealing its softer underside. He swabbed the skin with disinfectant and then turned to his tray. He sucked half a milliliter—less than two hundredths of an ounce but containing about three million WI-1 cells—into the syringe, eased the needle under the patient’s skin, and pushed the plunger. Then, using a marker, he drew a circle around the injection site. He would be back to check the result, he told the patient.30
This faceless patient and five others are identified only by their initials in Hayflick and Moorhead’s landmark paper. In it there is a table describing the injections and their results.31
The first patient had cancer that had spread throughout his or her abdomen. The doctors were at the point of simply treating symptoms, not trying to stop the disease. After the patient was injected, a nodule, or bump, developed at the injection site but disappeared on the sixth day. A biopsy of the nodule, to look at its cells under the microscope, was not done.
Three of the patients had breast cancer that had spread. All three were on drugs that suppressed their immune systems, making them less able to fight off WI-1 if it was in fact cancerous. One of them developed a slight fibrous hardening at the injection site on the seventh day; she died on the eighth day. A biopsy of her injection site was negative for WI-1 cells or anything else abnormal. A second developed a nodule on the sixth day that disappeared by the tenth, four days before she died. A biopsy wasn’t done. The third breast cancer patient didn’t develop a nodule. Nonetheless, her injection site was biopsied on the seventh day, with negative results.
The other two patients had lung cancer and colon cancer respectively. Their diseases had spread through their bodies and they were on chemotherapy. One developed a small nodule on the sixth day. A biopsy showed no WI-1 cells, normal or cancerous. As for the other patient, the paper simply says that nothing had happened at the injection site after nine days, and that no biopsy was taken.
The sum of the results was reassuring: of the nodules developed by four of the patients, three of them melted away, and a biopsy of the fourth showed no cells of any kind. And no nodules had appeared at the injection sites of the other two patients.
By the summer of 1962, another eleven dying cancer patients would be injected with Hayflick’s cells, bringing the total to seventeen. A report by Hayflick and others to the World Health Organization in July of that year states that the cells had been “implanted in 17 patients dying of cancer.” Again the nodules that developed in some patients disappeared within ten days, and biopsies didn’t reveal any living cells.32 The results from these seventeen anonymous patients would be a key basis on which Hayflick would testify innumerable times over decades that his human diploid cells were normal and didn’t cause cancer.
For Hayflick the injections by Elkins were the final experiment in what was now eighteen months of painstaking trial-and-error work—work that had tested not only his scientific thinking and resourcefulness but his, and Moorhead’s, capacity for repetitive, monotonous tasks and meticulous observation. He had established what were, to his knowledge, the first cell lines that had been proven to be, and to remain, certifiably normal when grown in the lab. They could be grown for months and months and yet they still reliably retained the diploid number of forty-six chromosomes. And those chromosomes, scrutinized by Moorhead under the microscope, looked just as they ought to: his diploid cells did not harbor the broken, disjointed, frayed, and otherwise strangely constructed chromosomes that signified cancer. Nor did they appear to cause cancer in hamsters or human beings.
And just as reliably as they harbored forty-six chromosomes, the diploid cells aged and finally died in culture; they could replicate only through fifty or so divisions. They were as mortal as all flesh. The two sets of observations—the cells’ normalcy and their inevitable deaths—he now knew, went hand in hand.
His findings could be summarized like this:
Normal (diploid) chromosome number and appearance = finite growth in the lab
A few years later he would take this thinking further:
Normal (diploid) chromosome number and appearance = finite growth in the lab ↔ corresponds to normal cells in human beings
Abnormal (heteroploid) chromosome number and appearance = indefinite growth in the lab ↔ corresponds to cancer in human beings33
Normal cells could escape from aging only by acquiring the properties of cancer cells—whether in the lab or in a living human being, he would write.fn1 34
Hayflick had also created, he believed, a promising tool to deploy against infectious diseases. For he had demonstrated that his diploid cells could be infected with dozens of disease-causing viruses, making them near-perfect factories for making viral vaccines. They appeared to be clean—free of hidden, lurking, unknown viruses that would scare off regulators—and they could be produced and then frozen, thawed, and expanded into near-infinite quantities of cells for just such a purpose.
For one junior scientist and his chromosome-expert colleague, it seemed that there was a tremendous amount to be proud of in this paper, which, as Hayflick wrote it up, stretched to thirty-five pages. He knew precisely where he wanted to place it.
In late 1960 the Journal of Experimental Medicine was an enviable place for a young, ambitious scientist to be published. Already sixty-five years old, which was a considerable age in the young world of U.S. research, the journal had been founded by William Welch, a giant of American medical research. It had published many of the great biologists. It had also published Carrel’s 1912 paper establishing, via the beating chicken heart, that cells could live indefinitely in a culture bottle if they were just treated properly. The journal was put out by what was arguably the pinnacle U.S. research institution: the Rockefeller Institute in New York City. There, none other than Peyton Rous presided over it as editor.
Rous was the biologist who, fifty years earlier, had discovered that he could cause a sarcoma, a malignant connective-tissue tumor, in chickens by infecting them with finely filtered fluid from chickens that already had such tumors.35 He had posited that an “agent” separate from the tumor cells caused the sarcoma; since then this “agent” had been named the Rous sarcoma virus. With the resurgence in the 1950s of the idea that viruses might cause cancer, Rous’s fame and eminence had grown. In 1966 he would be recognized with a Nobel Prize, fifty-five years after publishing his groundbreaking paper.
Before Hayflick put the bulky manuscript and its glossy accompanying photos in the mail, he did two things. According to Moorhead, he suggested to Moorhead that the two men flip a coin to determine whose name should go first on the manuscript. Moorhead remembers declining the offer, saying that Hayflick had done the lion’s share of the work and deserved to be in the prestigious place of first author.36
Hayflick also took his findings to Koprowski. He briefed his boss on the many disease-causing viruses that infected the diploid cells, and their consequent potential for vaccine making. He told him about their normal chromosomes and their lack of cancerous transformation. And he laid out the audacious claim that the paper would make—that cells aged and eventually died in lab dishes—throwing into question Carrel’s “immortal” chicken heart and upending decades of received wisdom. “The next thing you’re going to tell me is that these lung cells of yours are breathing,” Hayflick remembers Koprowski saying at one point, Hayflick recalled in a 2014 interview.37
Nonetheless, the junior scientist invited his boss to join him and Moorhead as an author on the paper. It was a common practice, then as now, for top bosses to share authorship of papers, no matter how little of the actual work they had done. Koprowski demurred. His message was clear: Hayflick and Moorhead could own the paper, and the consequences of their temerity, all by themselves. He did agree, though, to do them the favor of drafting and signing a cover letter to Rous.
Hayflick sent off the paper with its cover letter and waited for what seemed like an eternity for a reply. While he was waiting, he fought back doubts. He was an unknown, a greenhorn, and no matter how good his data, some were going to dismiss him, to say that he had gone off the rails. Gey’s cautionary comment came back to haunt him. Was he, in fact, about to ruin his career?
Then an opportunity to buy some peace of mind presented itself. The huge Federation of American Societies for Experimental Biology was holding its 1961 annual meeting in nearby Atlantic City, New Jersey. In mid-April twelve thousand biologists would take over Gilded Age hotels like the Shelburne and the Marlborough-Blenheim and attend events like “The Thyroid Smoker” on the Skyline Terrace at the Traymore. Among the nearly three thousand talks to be given at the seaside resort, one of the most listened-to would be by Theodore “Ted” Puck.
Puck was a powerful, influential figure in cell culture. He was the Colorado-based scientist who had reported growing normal human cells from the skin samples of adult volunteers and from leftover surgical tissues. In the 1958 paper where he had reported this, however, Puck and his coauthor, Tjio, had not noted anything about the cells grinding to a halt. They had reported that the cells were still dividing heartily months after being launched, after dozens of replications.38 The implication was that they kept dividing indefinitely.
Puck, a Nobel-caliber biologist, was a man whom people turned to with their questions about how to best nurture lab-dish cells. He was known to be meticulous with his culture procedures and to independently test the ingredients of his medium to rule out contamination. (Hayflick had bought his ingredients off the shelf, and used them as is.) What was more, Puck was established as a brilliant man of science and a preeminent cell culturist. How could he have missed something as basic as cells dying, repeatedly, under his nose?
Hayflick wanted badly to put a single question to Puck, and the Atlantic City conference would give him his chance to do that. He just needed to screw up his courage to ask that question. In a 2012 interview Hayflick recalled Puck giving his talk in a big, packed hall. He thought that Puck was pompous, and at the same time, he felt intimidated. He forced himself to raise his hand. Puck called on him.39
Hayflick asked Puck if his normal human cells had ever ceased dividing and died, in spite of his assiduous attention to getting every component of the medium just right. Oh, sure, Puck replied. It happened from time to time. But it wasn’t a big problem. He would just go back to the freezer and thaw another ampule. That was all Hayflick needed to hear. He knew in that moment that Puck had looked at cells aging and dying in culture. But he hadn’t seen them.
Hayflick returned to Philadelphia full of confidence. It was a good thing, because ten days later, on April 24, 1961, Peyton Rous at last responded to the Hayflick-Moorhead paper, in a letter to Koprowski. His words were devastating. After apologizing for keeping the young authors waiting because of his busy spring meeting schedule, Rous reported that he and three other editors had read the manuscript and discussed it. Not only was it “too specialized” for the Journal of Experimental Medicine, but it was poorly structured and all over the map, bogged down with extraneous observations like the fact that the fetal diploid cells were susceptible to infection by many different viruses.
As for the authors’ proposal that cells aged and died in lab bottles, it “seems notably rash,” Rous wrote. “The largest fact to have come out from tissue culture in the last fifty years is that cells inherently capable of multiplying will do so indefinitely” if they’re properly cared for.
“What a broadside!” Rous observed of his own skewering of Hayflick and Moorhead. “Yet I write it with complete good will.”40
The man who in 1911 had been ignored and sometimes disparaged when he dared to suggest that viruses might cause cancer was, fifty years later, quite ready to poke holes in Hayflick’s audacious challenge to conventional wisdom. Yet Hayflick, with every stubborn, determined bone in his body, wasn’t for a moment going to be deterred. Three weeks after receiving Rous’s rejection letter, he sent the paper off to the less-venerable but still-respected journal Experimental Cell Research. That journal immediately accepted it for publication.