What happens when you clone mice for 20 years straight?

This story begins in 1997, the same year I was born. An up-and-coming postdoc at the University of Hawaii, Dr. Teruhiko Wakayama, had just succeeded in creating the first cloned mouse, Cumulina. This was just a year after the birth of Dolly the Sheep, and cloning was still cutting-edge technology. By taking an adult mouse cell and injecting it into an enucleated egg, then stimulating the egg to divide, Dr. Wakayama was able to make a mouse that was the genetic copy of the original cell.

By 2005, cloning was more routine, and efficiency had increased due to various optimizations, like histone deacetylase inhibitors to help address epigenetic issues. Dr. Wakayama, who had recently started his own lab at RIKEN, and his wife, Dr. Sayaka Wakayama, had an interesting idea: what if we cloned a mouse? And then took a cell from the cloned mouse and made another clone? And then did it again?

And again?

57 times?

Those … are doing a LOT of work here. (CC-BY-NC-ND)

Can’t stop, won’t stop cloning

Long-term experiments aren’t a new thing — in fact, they’ve been around for a while.

Some, like the E. coli long-term evolution experiment, have tracked genetic changes over time in a lineage of cells. But doing this in cloned mice is next-level difficult. With E. coli, all you need to do to make a new generation of cells is pipette a small drop into new growth medium. With cloning mice, the process involves biopsies, cell culture, egg retrieval, microinjection, and embryo transfer. There are only a few people in the world who can do this, and very few as good as those in the Wakayama lab.

And because DNA sequencing technology was still in its infancy when the project started, many of the tools that the project ended up using to analyze the mice were only developed partway through. This project, which started in 2005, persisted through lab moves, an earthquake in 2011, and the COVID-19 pandemic. Over 20 years of diligent effort, these researchers repeatedly cloned mice, creating a total of 58 generations. Due to the low efficiency of cloning, this took 30,947 individual attempts. On the 58th generation, the project ended. Not because the researchers got tired of it, but because the mouse cells just wouldn’t make clones.

So what happened?

In an interim report in 2013, the research group reported successful cloning for up to 25 generations with no decrease in the health of the mice or in cloning efficiency rates. Although the cloned mice had epigenetic abnormalities typical of clones (including enlarged placentas), these abnormalities did not worsen with time. Telomeres likewise did not shorten with increasing generations. However, in 2013 whole genome sequencing was much more expensive than it is today, and the researchers did not investigate the accumulation of genetic mutations over time.

It turns out that mutations were in fact accumulating. Not at a terribly high rate, but enough to matter. This brings us to the current paper.

In 2026, the Wakayama group reported their final results. Things appeared to go reasonably well up until around generation 40, but after that point, the efficiency started to drop, and by generation 58, the researchers simply couldn’t continue.

Fig 1B: cloning success rate by generation and year. (CC-BY-NC-ND)

Although the mice that were born were healthy and lived to a normal mouse lifespan (Fig. 1D), the cloning efficiency by generation 58 was nearly zero (Fig. 1B).

With the benefit of advanced sequencing technology, the researchers found a steady accumulation of mutations over the generations. By generation 57, the cloned mice had acquired over 3400 single-base changes relative to the starting mouse, whereas 62 generations of natural reproduction (in an inbred mouse strain) had accumulated 752. In other words, cloning caused 3.1 times more single-nucleotide mutations per generation than natural reproduction.

Note: F60 and F62 refer to 60 and 62 generations of natural breeding, respectively. (CC-BY-NC-ND)

On its own this increased mutation rate is not terrible. But unlike with cloning, harmful mutations that arise in sexually reproducing organisms can be lost through recombination. Cloning lacks this mechanism, so harmful mutations are more likely to accumulate.

On top of the single-nucleotide changes, there were also much larger issues with the genomes of the cloned mice. In particular, one entire X chromosome was lost at some point between generations 25 and 45, and never regained. Mice can tolerate having a single X chromosome reasonably well (much better than humans), but this is still not very healthy. There were also various other chromosomal deletions, inversions, and translocations. Between the structural abnormalities and the point mutations, it’s clear why the cloning process failed after 58 generations.

Conclusions

Although the media reports that this study showed “a big unexpected problem with cloning”, I don’t think this characterization is accurate. The accumulation of mutations was definitely a known issue (as far back as 1932) with asexual reproduction. Also, keep in mind that their first 25 generations were quite healthy, and even though the efficiency decreased in later generations, the mice that survived had normal lifespans.

Fig. 1D: mouse lifespan by generation (CC-BY-NC-ND)

A 3.1-fold increased mutation rate isn’t great, but I think if the researchers had been able to do whole genome sequencing for quality control at each generation (instead of only retrospectively), and only clone mice without severe genetic abnormalities, they likely could have avoided most of the negative effects they saw here. And the lack of epigenetic abnormalities accumulating over generations is also encouraging, both for cloning as well as for other technologies (like IVG) that create offspring from adult somatic cells.

Overall, I’d be worried if someone proposed to do human cloning for more than 25 generations. But that long term experiment would take a lot more than 20 years!