Proustian DNA

For centuries, scientists have debated how cellular complexity develops. Today, the “unopened question” of how cells acquire different identities centers on epigenetics — the study of how genes are turned on and off, rather than the changing of the DNA sequence itself.

Humans develop from a single cell into an organism with 37 trillion cells and thousands of cell types. The original sperm and egg that meet to create a zygote contain all of the genetic material that a person will have for their lifetime. So, how does rich cellular complexity develop within human embryos? Ramesh A. Shivdasani, a professor at the Harvard Medical School and deputy director of the Dana-Farber Cancer Institute, calls this “the big, big, unopened question in biology.”

For centuries, scientists have debated how cellular complexity develops. During the Enlightenment, some scientists believed humans grew from a perfectly-formed, miniature person in the sperm or egg. By the 1950s, scientists were beginning to understand the importance of DNA in providing a blueprint for human development from embryo to organism. Today, the “unopened question” of how cells acquire different identities centers on epigenetics — the study of how genes are turned on and off, rather than the changing of the DNA sequence itself.

Which genes are on or off determines a cell’s identity, Shivdasani explains. As cells differentiate, different organs are created by different parts of the DNA. Fully-formed organs use different parts of the DNA than developing organs — once a kidney is formed, it turns off the DNA used in development and turns on other sequences. Prior to his lab’s most recent study, published on March 21 and led by post-doctoral fellow Unmesh Jadhav, the prevailing scientific assumption was that adult cells contain no record of which genes were active during their development.


Shivdasani has been curious about cellular identity since he began his medical training about 25 years ago. Jadhav joined Shivdasani’s lab six years ago, in part, to study the identity of intestinal stem cells. “I came and joined the lab and we started looking at all of this from an epigenetic angle,” Jadhav says.

One form of epigenetic control is the regulation of the dozens of regions of DNA, called enhancers, associated with each gene. Enhancers are like switches: they turn on and off in various arrangements. This allows different cell types to activate the same genes in different ways — if a gene has twenty switches, a brain cell might use twelve, while a kidney cell might use four of those and seven others. One way the body regulates these switches is through a protein complex — called Polycomb Repressive Complex 2, or PRC2 — that turns the switches them off.

About three years ago, Jadhav says, he began researching a potential relationship between PRC2 and another DNA repression mechanism, called methylation. They were probing a highly specific biological phenomena which Shivdasani described over email as “arguably esoteric.”


The scientific process is often taught as simple and straightforward: hypothesize, test your hypothesis, gain crucial insights, and so on.

“In truth, the path to scientific discovery is a torturous and serpentine path with a lot of diversions and false leads, wrong hypotheses, and a lot of serendipity,” Shivdasani says. Some of the most revolutionary scientific discoveries were accidental — the people who discovered penicillin and X-rays did not set out to do so. The same serendipity also led Shivdasani’s lab to cell memory.

“We had no reason to think memory would exist. We were frankly not even that interested in the problem,” Shivdasani says.

While researching PRC2, the protein complex that turns genes off, the lab was surprised that gene expression didn’t change substantially. So, Jadhav hypothesized that the other gene repression mechanism, methylation, was turning off the genes instead.

This hypothesis proved false. “It still remains a mystery to us of how these two balance each other out in an invivo system,” Jadhav says. But in the process of being proven wrong, they identified a new research angle.

“We always had our eye on how cells behave in development and what kind of transitions occur in particular stem cell behavior,” Jadhav says.

To test Jadhav’s original hypothesis, they removed PRC2-based gene repression and sequenced the genome of intestinal cells in mice to see which DNA switches were on or off. “Having studied the switches that are on or off in the intestine for years, we know every switch that is on in the adult intestinal cell,” Shivdasani says. In the absence of PRC2, they found far more active DNA switches than expected — three times as many.

The next step was to make sense of the data. Jadhav and his fellow researchers had previously studied developing intestinal cells, and used this experience when analyzing the data. They observed that the additional switches were near regions of DNA used in developing embryos.

“We looked back in the developing intestine, we looked in the embryo, we looked in the fetus, and we saw — lo and behold — that over the course of development those are the same set of switches that are being turned on,” Shivdasani says. The adult cell contained memories of the developing cell’s active switches.

Once they developed this new hypothesis about cell memory, they had to perform further experiments to test it. “Switching directions of research is a fraught endeavor, but when the payout is potentially profound, then one does it with a great deal of excitement and energy and enthusiasm, and that’s what we did,” Shivdasani says. “We took that ball and ran with it.”

The final study looks at intestinal and red blood cells in mice and analyzes publicly available data on brain and skin cells. This confirmed the existence of a fossil record of cellular development. “It’s almost as if you arrive at the station of a sedimentary rock and you cut through and you can see the Precambrian, the Cambrian, the Pleistocene, and you find all the fossils very perfectly preserved,” Shivdasani says.

Shivdasani finds the paper’s second discovery potentially more exciting. When they removed PRC2 gene regulation in specific circumstances, many of the genes used during development turned on. The genes that initially made the intestine into an intestine started working again.

“We did not think just by having modulation of one repressor like this would allow this entire history of development to be resurrected,” Jadhav explains.

Significantly, cells do not recover a general archive of development but instead have specific memories — intestinal cells can only resurrect the developmental history of intestinal cells. “It only recalls the memory that it has, which is only the memory of its tissue, and that’s why it’s sort of a Proustian kind of recovery of memory,” Shivdasani says.

According to Jadhav, though potential medical applications are speculative, they could be important to regenerative and cancer medicine. Cancer spreads throughout the body by turning on genes that should be off — perhaps cancer cells are tapping into the developmental archive. Knowing which genes are used during development could also contribute to significant advances in organ regeneration, and even provide insight into how tissues regenerate on a daily basis.

“This opens a lot of avenues about thinking on the lines of how cells maintain memory and how that can be linked to the epigenome,” Jadhav says. “That is an exciting possibility for me when I think about myself as an independent investigator in the future.”

Yet what, exactly, future investigations will yield is difficult to know. “Serendipity,” Shivdasani says, “is what all scientists live for.”

—Magazine writer Matteo N. Wong can be reached at