Chapter I
The Blood's Reserve
After the atomic bombs of 1945, doctors wanted to know why radiation kills. At moderate doses the main cause is the loss of the bone marrow, which makes blood. In Toronto in 1961, James Till, a physicist, and Ernest McCulloch, a physician, injected marrow into irradiated mice and saw lumps on their spleens. Each lump grew from a single cell and held several kinds of blood cells, as well as cells that could make new lumps in another mouse. They had found the blood stem cell, a cell that renews itself and makes specialised descendants. Doctors had already begun trying bone marrow transplants, without knowing which cells did the work.
Chapter II
Cells from Embryos
Stem cells in adults are rare and usually limited to one tissue. The early embryo has cells that can become anything, but only for a day or two. In 1981 Martin Evans with Matthew Kaufman, and Gail Martin, found how to keep them in that state in a dish. Mouse embryonic stem cells divide indefinitely and, put back into an embryo, take part in building every tissue. Combined with methods for altering a chosen gene, they let researchers make mice lacking any gene they wished, which became a basic tool of developmental genetics.
In 1998 James Thomson grew human embryonic stem cells. Hopes rose for replacement cells for diabetes, Parkinson's disease and spinal injury, and so did objections to using human embryos. One way round seemed to be cloning, which would make stem cells matched to a patient. In 2004 and 2005 Hwang Woo-suk in Seoul claimed to have done it. The data were fabricated. A later analysis suggested that his one genuine line had arisen from an unfertilised egg that began dividing on its own, not from cloning.
Chapter III
Four Genes
John Gurdon's frogs, in experimental embryology, had shown that an egg can reset a specialised nucleus. In 1987 Harold Weintraub's group had shown that a single gene, MyoD, can turn a fibroblast, a connective-tissue cell, into a muscle cell. Shinya Yamanaka wondered whether the genes that keep embryonic stem cells pluripotent could reset a skin cell without an egg. With Kazutoshi Takahashi he picked 24 candidates, put them all into mouse skin cells together, and got colonies that looked like embryonic stem cells. They then worked out which genes mattered, and in 2006 reported that four were enough. Their induced pluripotent stem cells, or iPS cells, avoided both the embryo and the egg.
Chapter IV
A Closer Look: Finding Four Genes Among Twenty-Four
Yamanaka's team suspected that some combination of their 24 candidate genes could reprogram a cell. How many combinations are there? Each gene can be in or out of the mixture, so
possible mixtures. Even restricted to mixtures of exactly four genes, there are
Each test took weeks, so testing combinations one by one was hopeless. Instead they used subtraction. First, all 24 together: colonies formed. Then 24 more experiments, each leaving out one gene. Where leaving a gene out made the colonies disappear or become much rarer, that gene mattered. This narrowed the list to 10. The same trick on the 10, with 10 more experiments, gave 4:
| Step | Experiments |
|---|---|
| All 24 genes together | 1 |
| Leave out each of 24 in turn | 24 |
| The 10 remaining together | 1 |
| Leave out each of 10 in turn | 10 |
| Total | 36 |
In outline, about thirty-six experiments instead of ten thousand. The published work ran more controls than this, but the logic was the same. The method works because the effect needs every essential gene at once: leaving out any one of them breaks it.
Reprogramming was rare. In the early experiments only a small fraction of cells, commonly quoted as between 0.01 and 0.1 per cent, became iPS colonies. From a dish of one million cells that means 100 to 1,000 colonies. It was enough to find, pick and grow them, and later studies showed that the rest were not incapable, only slower: given time, nearly every cell can make the switch.
Chapter V
Cells to Organs
iPS cells spread quickly because anyone could make them from a skin sample, with no embryo involved. Patients' own cells could be turned into neurons or heart cells to study their disease in a dish. In 2009 Toshiro Sato and Hans Clevers showed that single stem cells from the gut, given the right signals, build miniature gut linings, a founding result for the modern study of organoids. The field also suffered a second scandal in 2014, when claims that an acid bath could make stem cells collapsed within months. In 2026 Japan gave conditional, time-limited approval to the first two treatments made from iPS cells: nerve cells for Parkinson's disease and sheets of heart muscle for heart failure. Replacing a whole organ remains out of reach. It will need what developmental genetics has learned about how organs are laid out, and what cell biology knows about how cells hold their shape and identity.