Induced pluripotent stem cells explained: how adult cells are wound back to the start

In 2006, Kazutoshi Takahashi and Shinya Yamanaka at Kyoto University reported in the journal Cell that four genes, added to ordinary skin cells from mice, could turn them into cells that looked and behaved like embryonic stem cells. In 2007 the Kyoto lab did the same with human cells, as did James Thomson's group at the University of Wisconsin using a slightly different set of genes. In 2012 Yamanaka shared the Nobel Prize in Physiology or Medicine with John Gurdon, whose frog experiments in the early 1960s had shown that a specialized cell's nucleus still holds every instruction needed to build a whole animal.

The cells are called induced pluripotent stem cells, usually shortened to iPSCs. "Induced" because they are made in the lab rather than found in the body. "Pluripotent" because, like embryonic stem cells, they can in principle give rise to nearly any cell type: neurons, heart muscle, liver cells, the pigmented cells at the back of the eye. If those terms are new, our piece on what makes a stem cell a stem cell covers potency and self-renewal from the ground up.

The four factors

Yamanaka's team started with 24 genes known to be active in embryonic stem cells, put them all into mouse fibroblasts, and then took genes away one at a time to find the ones the effect could not do without. Four survived the cut.

All four are transcription factors, proteins that bind DNA and turn sets of genes on or off. Together they are often called the Yamanaka factors, or OSKM. Thomson's human recipe used OCT4, SOX2, NANOG and LIN28, an early sign that there is more than one route to the same place.

What reprogramming actually changes

A skin cell and a neuron from the same person carry essentially the same DNA. What differs is which genes are in use, and that pattern is held in place by chemical marks on the DNA and on the proteins it is wrapped around, known as epigenetic marks. Reprogramming does not rewrite the genetic code. It wipes most of those marks and lets the cell settle into the pattern of an early embryonic cell.

The process is slow and wasteful. It takes a few weeks, and in the early experiments well under 1 percent of treated cells became true iPSCs. The first experiments also used retroviruses that stitched the factor genes permanently into the cell's DNA, a problem if the cells are ever going into a person. Lines made for clinical use now mostly rely on methods that leave no permanent insertion, such as Sendai virus, episomal plasmids that are diluted out as cells divide, or synthetic RNA. Hongkui Deng's group at Peking University has gone further and reprogrammed human cells with small molecules alone. The starting material has widened too: a routine blood draw now works as well as a skin biopsy.

Three reasons labs care

1. Disease in a dish

Take cells from a person with an inherited disease, reprogram them, then coax the iPSCs into the cell type the disease damages. Researchers gain a renewable supply of that person's neurons or heart cells, tissue that could never be biopsied from a living brain. Early examples included motor neurons from people with ALS and spinal muscular atrophy, and heart muscle cells from people with inherited long QT syndrome that showed the same electrical fault in the dish.

Gene editing sharpened the approach. With CRISPR, a lab can correct the suspected mutation in a copy of a patient's line; since the edit is the only difference, any change in behavior points at that mutation.

2. Drug screening and safety testing

Cells made from iPSCs can be spread across plates with hundreds of wells and exposed to thousands of compounds. In Japan, screens on iPSC-derived motor neurons pointed researchers toward existing drugs, including bosutinib and ropinirole, which then went into small early-stage ALS trials. Neither is approved for ALS. Drug developers also test candidates on iPSC-derived heart cells to catch rhythm problems, an approach an FDA-backed effort called CiPA has studied.

3. Replacement cells

The most ambitious use is to make the cells a disease has destroyed and put them back. Targets under study include retinal cells lost in macular degeneration, dopamine neurons lost in Parkinson's disease, insulin-producing islet cells for type 1 diabetes, heart muscle damaged by a heart attack, and corneal cells. For background, see our explainers on stem cells and type 1 diabetes and cell therapy after a heart attack.

From dish to patient: a short timeline

The eye

In September 2014, a team led by ophthalmologist Masayo Takahashi of RIKEN treated a woman in her seventies who had wet age-related macular degeneration. Her skin cells were reprogrammed, turned into retinal pigment epithelium (the support layer under the light-sensing cells) and grown into a sheet about 1.3 by 3 millimeters that surgeons placed under her retina. The aim was safety, not a cure. In the one-year report in the New England Journal of Medicine in 2017, the sheet was still intact and her vision had neither improved nor worsened.

The second planned patient never received cells. Sequencing found small genetic changes in that person's iPSC line, and the team chose not to go ahead. The program moved to donor cells, and in 2017 patients began receiving retinal cells made from a donor line.

The brain

In April 2025, Nature published two Parkinson's papers side by side. The Kyoto trial, led by Jun Takahashi, transplanted dopamine neuron precursors made from donor iPSCs into both sides of the brain in seven people aged 50 to 69. There were no serious adverse events. Of the six people assessed for effect, four improved on a standard motor score measured off medication and five improved on medication, and brain scans suggested more dopamine was being made. The companion paper, from Memorial Sloan Kettering Cancer Center and BlueRock Therapeutics, reported a 12-person safety trial of similar cells made from embryonic stem cells.

Both were small, early trials without a control group, so some of the improvement could reflect expectation and normal ups and downs. Sumitomo Pharma, which is developing the Kyoto approach, has said it intends to seek approval in Japan.

What can go wrong

• Tumors. A pluripotent cell that never finished differentiating can, in principle, grow into a teratoma, a tumor of mixed tissues. Clinical products have to show that leftover pluripotent cells have been removed or reduced to a level judged safe. Because c-Myc is a cancer gene, many labs now use a milder relative, L-Myc, or leave Myc out.
• Genetic changes. Cells pick up mutations during reprogramming and months of culture. Some, such as extra copies of a small region of chromosome 20 or mutations in the tumor suppressor gene TP53, give cells a growth advantage and can take over a culture. Lines meant for patients are now routinely sequenced.
• Epigenetic memory. iPSCs can keep traces of the cell type they came from, which can tilt how readily they turn into other types.
• Immaturity. Heart cells made from iPSCs beat, and neurons fire, but many resemble fetal cells more than adult ones.

Your own cells or a bank's?

HLA markers are what the immune system uses to tell self from non-self, as in bone marrow transplants. Some people inherited identical HLA types from both parents, and their cells match a larger share of recipients. That is why Kyoto University's Center for iPS Cell Research and Application (CiRA) built an iPSC stock from such donors, aiming to cover much of Japan's population with a fairly small number of lines.

A third route edits cells so the immune system overlooks them, producing so-called hypoimmune lines meant to suit almost anyone. These are at an earlier stage, and they raise a question of their own: a cell the immune system cannot see is also harder to clear if it misbehaves.

Reading the next iPSC headline

Three details in any report tell you most of what you need:

1. Whose cells? The patient's own or a donor line. That decides cost, timing and whether immune suppression is part of the picture.
2. Did patients take immunosuppressants? If so, some of the risk comes from those drugs rather than the cells.
3. Was there a comparison group? Early cell therapy trials are usually open-label, with a handful of people who all know what they received. They can show that cells survive and cause no obvious harm. Showing how well a treatment works is the job of the larger, controlled trials that come after.