The strain that Alexis Carrel’s laboratory kept from 1912 to 1946 was fed on two things: the plasma of young hens, and a clear fluid the laboratory called embryonic juice, made by mincing chick embryos a week old and spinning the pulp. The plasma was the clot the cells lived in. The juice was what made them grow. This piece is about the juice: where it came from, what the people who used it believed it was, and why, since 1965, it has been the likeliest explanation of the thirty-four years.

Where it came from

Carrel came to the juice by way of wounds. In 1907 and 1908, at the Rockefeller Institute, he had laid the pulp of thyroid and other organs on small wounds in dogs and watched the tissue around them grow faster than it should: thyroid pulp on skin raised what he called exuberant granulations, and on bone it thickened the periosteum. He set out to find what in the pulp did it, and it was for this, as much as to answer his critics, that he took up the culture of tissues: a wound in a dog could not be measured, a fragment of tissue in a drop of plasma could.

The first extract went into the strain on February 1, 1912, at its fourth passage: two drops of diluted plasma and one of juice “prepared from seven day old chick embryo plus 3 drops of Ringer solution.” The next day’s note reads “active growth.” From the middle of March fresh juice went in at almost every passage. In June 1912 Carrel sent the Journal of Experimental Medicine the paper that made it standard. Fragments of heart, skin and periosteum grown in plasma with extract always outgrew their controls in plasma alone, by anything from three to forty times. The juices of embryo, of adult spleen and of the Rous chicken sarcoma were the strongest; kidney and heart were weaker; blood corpuscles barely worked. Juice from an animal of another species did little. Diluted juice did less. Juice heated to 56 degrees Celsius for ten minutes lost a third of its power, heated for thirty-five minutes two thirds, and at 70 degrees it lost all of it. Filter paper did not weaken it, a Berkefeld filter weakened it, and a Chamberland porcelain filter, the kind that holds back bacteria, removed its power entirely. Whatever the active thing was, it was large, it was destroyed by moderate heat, and it would not pass through porcelain. Carrel had opened the paper with wounds in mind: if the rate of repair could be raised tenfold, a broken leg might knit in days.

By 1919 the recipe had settled. Ebeling described it exactly. The plasma came from healthy hens not over two years old, starved for twenty-four hours so that the blood would not be cloudy with fat, drawn from the carotid artery through an oiled cannula into chilled paraffined tubes, spun, and kept corked in the cold. The extract came from embryos seven to eight days old: “The embryos are first washed in Ringer’s solution to remove the amniotic fluid and traces of blood which adhere to them after removal from the shell. Then the solution is drawn off in order to obtain an undiluted tissue juice. The embryos are minced in a watch-glass with sharp, curved scissors. The pulp is centrifuged for 10 minutes and the supernatant fluid obtained is drawn off.” The medium was equal parts plasma and juice, where in 1914 it had been two parts to one. “As soon as the tissues were washed in Ringer’s solution, or in salt solution every 2 days, and embryonic juice was used in the plasma of the medium, the rate of growth increased very much and became almost constant.” Raymond Parker’s manual, in 1950, told the story the way the laboratory remembered it: in the strain’s first year “the subcultures were becoming smaller and smaller with each subsequent transfer,” because plasma alone “did not stimulate the cells to divide fast enough to compensate for the amount of tissue lost in transplanting the fragments to new media. With the addition of tissue extract, however, this situation was completely remedied.”

What they thought it was

For twenty years the laboratory tried to find the ingredient. Carrel’s first guess was a specific growth hormone of the embryo. In the 1920s Lillian Baker and Carrel fractionated the juice and found the activity concentrated in the protein fraction, in what they called nucleoprotein and glycoprotein, and then found that digests of pure proteins, the proteoses, also drove growth; by 1928 they had given up the idea of a special hormone and were treating the juice as rich food. In Copenhagen Albert Fischer precipitated it with ammonium sulphate, carbon dioxide and alcohol and concluded that no single protein was responsible. The Cambridge culturist E. N. Willmer, in his textbook of 1935, wrote that how the extract worked was “still wrapped in the mists of obscurity, which, although they frequently appear to be lifting, come down again as thick as ever.” The question was never answered in the strain’s lifetime, and when chemically defined media arrived after the war the juice fell out of general use. It was still being used in 1980 for one thing, muscle cells, which grow badly without it.

Why it is the suspect

The suspicion was, in a sense, in Carrel’s first paper. In February 1912, to prove that the cells in his cultures divided rather than merely crawled out of the fragment, he minced a nine-day embryo in an equal volume of Ringer’s solution, spun it for five minutes, and put a drop of the fluid from the top of the tube into a culture medium with no tissue in it. “On February 2 many isolated cells were seen in the peripheral part of the medium. On February 5 the number of these cells had increased very much.” After five passages they “formed a real tissue, and had the appearance of an ordinary culture.” The supernatant of a spun embryo pulp, in other words, could carry living cells, and the cells could found a culture. Carrel used the fact to win an argument and did not return to it.

Others noticed the risk. In 1924 A. H. Drew, lecturing in London on the cultivation of tissues, recommended freezing and thawing the extract “to disintegrate the cells” in it. In 1938 Parker’s manual, written in Carrel’s own department, warned of the danger of carrying cells over in the juice and recommended the same freezing and thawing, and gave a centrifuge speed, 2,400 revolutions a minute, that ought to have been enough to throw the cells down. Nobody wrote down when Carrel’s laboratory began to freeze its extract, or whether it did. Keith Porter, who worked elsewhere at the Institute in the 1930s, remembered the juice as highly viscous and thought some cells might never have settled out of it. Charity Waymouth remembered Fischer in Copenhagen still not freezing his extract in 1946. And when Helen Morton, who had learned Parker’s method from Parker himself in Toronto, tested it in the early 1950s, she found living cells in two preparations out of five. They did not grow when she tried them; she thought it probable all the same that the strain had been topped up with living cells at various times in its life.

Leonard Hayflick made the suggestion in print in 1965, four years after he and Paul Moorhead had shown that normal cells stop dividing. He had read Ebeling’s table of 1912 and seen that the strain’s spurts of good growth fell in the weeks when extract went into the medium. His later wording was blunt: the culture “was fed in those years with an extract of chick embryo tissue prepared daily and extracted under conditions that permitted the addition of fresh living cells to the alleged immortal culture at each feeding.”

Jan Witkowski, reassessing the case in 1980, accepted that it probably happened and doubted that it explained thirty-four years. Cells carried in with the juice would have been spread evenly through the clot, and what the laboratory cut out and carried forward at each passage was the growing edge around the fragment, not scattered colonies; and above all, every other laboratory in the world made its extract the same way, from the same embryos, and none of them grew an immortal strain. “I believe that although some cell contamination probably occurred, it is unlikely that it could have done so consistently over a thirty-four-year period, and particularly in the last ten years of the immortal strain’s life when Carrel’s laboratory was well aware of the problem and presumably took great care to prevent cell contamination of the extract.” His own candidate was the deliberate version of the same thing, told to a visitor by a technician in 1930: “we just add a few embryo cells now and then.” The main entry weighs the two. What they have in common is the juice.

What it comes to

For thirty-four years the one ingredient added at every passage was the one ingredient never run blank against the strain. A culture on plasma alone dwindled passage by passage, and its dwindling was read as proof that the juice was food. The juice may well have been food. It was also, by the laboratory’s own demonstration of February 1912, a possible source of cells, and the two readings of the same growth were never put to a test that could tell them apart. When the thing you add to make an experiment work could also manufacture the result, leave it out once, and believe what happens.

Sources

• Alexis Carrel, “Artificial Activation of the Growth in Vitro of Connective Tissue,” Journal of Experimental Medicine 17 (1913), 14–19; “On the Permanent Life of Tissues outside of the Organism,” Journal of Experimental Medicine 15 (1912), 516–528; Albert H. Ebeling, “The Permanent Life of Connective Tissue outside of the Organism,” Journal of Experimental Medicine 17 (1913), 273–285, and “A Strain of Connective Tissue Seven Years Old,” Journal of Experimental Medicine 30 (1919), 531–537. All public domain, on PubMed Central and in the issue scans at archive.org. pmc.ncbi.nlm.nih.gov, pmc.ncbi.nlm.nih.gov, pmc.ncbi.nlm.nih.gov, pmc.ncbi.nlm.nih.gov
• Raymond C. Parker, Methods of Tissue Culture, second edition, Hoeber, 1950, chapter VII, “Preparation of Tissue Extracts,” for the laboratory’s own account of the discovery and the chemistry of Baker and Carrel. archive.org
• Jan A. Witkowski, “Dr. Carrel’s Immortal Cells,” Medical History 24 (1980), 129–142, section 4(b), for Drew, Parker’s 1938 warning, Porter, Morton, Waymouth, Willmer and Fischer, and his own judgement. pmc.ncbi.nlm.nih.gov
• Leonard Hayflick, “A Brief History of the Mortality and Immortality of Cultured Cells,” Keio Journal of Medicine 47 (1998), 174–182, for his suggestion in his own words. jstage.jst.go.jp