DNA from ancient viral infections helps embryos develop, mouse study reveals

A stretch of DNA in the mouse genome left by ancient viral infections is crucial for early development in the womb, new research shows.

According to the study, published in December in the journal Science Advances, this viral DNA switches on genes that give cells in early-stage mouse embryos the potential to become almost any cell type in the body. The viral DNA — known as MERVL — itself gets activated by a protein called the "Dux transcription factor," which binds to the sequence and essentially kick-starts the embryo's development.

Although it's important in the womb, if Dux stays activated too long, it kills cells. The human version of Dux, called DUX4, causes a progressive muscle-wasting disorder when quirks in its genetic code cause it to be active for too long in muscle cells. That inherited disease, called facioscapulohumeral muscular dystrophy (FSHD), currently has no cure.

The new study not only unravels the roles of MERVL and Dux in the womb but also teases apart these harmful effects that can appear later in life. It's an "important piece of work," said Sherif Khodeer, a postdoctoral research fellow who focuses on stem cell and developmental biology at the university KU Leuven but was not involved in the study.

A powerful gene-editing tool

Researchers at the Medical Research Council Laboratory of Medical Sciences in England used a gene-editing tool called CRISPR activation (CRISPRa) to untangle the close relationship between Dux and MERVL. Unlike traditional CRISPR, which cuts DNA to change its code, CRISPRa boosts the activity of specific genes without changing the underlying DNA sequence.

The team used CRISPRa to switch on either Dux or MERVL in mouse embryonic stem cells. This enabled the researchers to examine how each factor influenced early embryonic development.

When the researchers switched on only MERVL, the stem cells showed "totipotency," or the ability to become any cell type — an important feature of the very earliest embryos. But the cells were missing key traits, the researchers found. This suggests that, while MERVL plays an important role in early mouse embryo development, Dux is also required.

Turning on Dux alone, on the other hand, produced cells that looked much more like natural early embryonic cells. So, the researchers think Dux activates the genes necessary for the embryo's development, independently of MERVL.

Because Dux and MERVL are so closely linked during the earliest stages of embryonic development, scientists previously suspected that MERVL might also contribute to Dux's harmful effects later in life. But the new study suggests this isn't the case.

The researchers tested how Dux causes cell damage by looking at its effects in stem cells with and without a gene called NOXA, which is known to be involved in cell death triggered by various stressors. They found that Dux turns on this NOXA gene, which produces a protein that triggers cell death. When the team removed NOXA, Dux caused much less harm. That showed that NOXA is responsible for the toxicity, not MERVL.

A potential therapeutic target

NOXA was already known to be elevated in FSHD, the human muscle-wasting disease. It's possible that developing a drug to inhibit NOXA could prevent cell death in the condition, thereby helping to improve the survival of muscle cells, the study authors think.

"Facioscapulohumeral muscular dystrophy is a complex disease," senior study author Michelle Percharde, head of the chromatin and development group at the Medical Research Council Laboratory of Medical Sciences , said in a statement.

"Even though all cells of a patient have the genetic changes that cause it, only a subset of cells activate DUX4," she explained. "Understanding what triggers DUX4 activation just in muscle cells, as well as how this compares to activation in early development, are key questions we hope to explore in future research."

It would be "valuable to compare" how mouse Dux and human DUX4 function, Khodeer said, adding that future studies should also explore precisely how MERVL controls nearby genes and when and how MERVL is switched off during mouse embryo development.

Crucially, Khodeer pointed out that MERVL is not present in the human genome. But scientists suspect that certain parts of the human genome could be equivalent to MERVL. As in mice, these stretches of DNA are leftover from ancient viral infections.

Khodeer said the new results raise several questions. For example, do early human embryos develop via the same mechanisms seen in mice? And which bits of ancient viral DNA in humans might play roles similar to MERVL at this early stage of development? "Answering these questions could clarify species-specific differences in early developmental regulation," he told Live Science in an email.