The Power of Three

One day, Michael Greenberg was pondering the many unanswered questions about how the MeCP2 protein malfunctions in the brain to cause Rett Syndrome, an autism spectrum disorder that afflicts 1 in 10,000 girls and much more rarely in boys.

“MeCP2 was proving to be more difficult than anything I had encountered,” says Greenberg, a neurobiologist at Harvard Medical School. “The problem was too complex to solve in one lab alone. We’re not going to figure out how to fix it if we don’t know how it works.”

Across the country and around the world, dozens of research groups were studying MeCP2 and Rett Syndrome, reporting important results at meetings and in scientific journals. But progress was slow to fill essential basic knowledge gaps.

So in 2010, Greenberg started conversations about a new way of doing things. He contacted Monica Coenraads, executive director of the Rett Syndrome Research Trust (RSRT), and proposed a new collaborative framework to support the search for successful therapies by getting a better grasp on the molecular role of MeCP2 in brain function. It made sense to Coenraads, the parent of a then teenaged daughter severely disabled by Rett Syndrome, who brokered the arrangement.

One year later, Greenberg’s group in Boston officially teamed up with the laboratories of Adrian Bird at University of Edinburgh, Scotland, and Gail Mandel, at Oregon Health Science University in Portland, to form the MeCP2 Consortium. The three labs, 3,000 miles to the east and west, were united in the belief that understanding the normal role of MeCP2 is crucial to develop therapies for Rett Syndrome and that working together was more likely to make that happen faster.

Ten years later, collaborative studies from the three labs have reshaped scientific understanding of MeCP2 function. Their joint efforts are spurring new directions in potential treatment strategies. In fact, the Consortium approach has nudged some work in all their labs toward translational experiments necessary to push their basic discoveries toward the clinic.

The Consortium model has made their experiments more rigorous and sparked fruitful new directions, the three senior researchers say. For the postdoctoral fellows, the Consortium has enriched the training, built a stronger web of professional connections, and set a high standard for their independent science careers.

When it began, the Consortium was a leap of faith for each lab. They agreed to meet twice a year in Boston—halfway between Edinburgh and Portland—and have regular conference calls in between. Researchers reveal early findings and discuss their ideas at formative stages, long before scientists usually divulge their data to colleagues outside their labs.

“One has to learn to share those things,” Bird says. “I’ve always been a bit of a blabbermouth regarding what research we are doing, but that’s different from saying what we’re going to do next. In the end, pooling our ideas really did speed up the field.”

Questions remain about MeCP2 and its role in Rett, but Consortium leaders have seen a consensus build around their emerging findings. Their studies show MeCP2 normally acts like a dimmer switch, turning down the activity of hundreds of genes in neurons ever so slightly. MeCP2 proteins blanket the DNA in brain cells, especially the longer neuronal genes. The protein latches onto signposts of gene activity (methylated DNA). Another part of MeCP2 recruits a gene-silencing complex that represses the target gene and others nearby.

In individuals with Rett Syndrome, genetic mutations interfere with these two closely related functions of MeCP2 protein. Based on studies from their labs and others, Consortium researchers and other scientists now believe that the failure of MeCP2 to recognize the methylated DNA or to repress gene activity is a major cause of Rett.

“In terms of the science, the Consortium has helped people in all three groups to receive unbiased critiques of their own experiments before publication and suggestions for new experiments,” Mandel says.

“There are still questions of how the lack of MeCP2 gives rise to Rett Syndrome,” she says. “Is there an MeCP2 function we haven’t found? Or do we have enough information, and it is what it is, that small changes in gene expression destabilize and depress the health of the cell enough so in a circuit it doesn’t function? We still want to know how many cells we need to fix in the brain for a normal phenotype. These are questions for the future that require more thought and experimentation.”

Any uncertainties and trepidation about the risks of collaboration have long vanished. The benefits kicked in at the inaugural three-group meeting, when lab members were cautiously reporting exciting new unpublished data they had gathered before the Consortium formed.

“We knew it would take another year or two to have all the data we needed for that publication,” said Matt Lyst, who had joined the Bird lab as a postdoc before the Consortium. “At first we were a little uneasy about revealing everything we knew.”

Then the Bird and Greenberg labs learned they were working on the same small piece of the MeCP2 protein but for different reasons.

“We were studying the fact that this part of the protein interacts with NCoR,” a molecular machine MeCP2 pulls in to slow down gene activity, says Lyst, who was leading the project in the Bird lab. “Mike’s lab was working on the same region of MeCP2 for different reasons. They had found when the neuron fires, this part of MeCP2 becomes phosphorylated, and they were wondering what the phosphorylation would do.”

The postdocs continued to lead their projects, and they pitched in to help each other out behind the scenes. For example, the Bird lab used mice from the Greenberg lab that modeled a specific mutation in that region. Two years later, each lab reported their findings in leading journals, with extra co-authors from both labs on both papers. Certain Rett Syndrome causing mutations in MeCP2 prevent it from interacting with NCoR, they found, and MeCP2 normally regulates this activity after it has been phosphorylated in response to neuronal activity.

Other work in both the Bird and Greenberg labs showed that MeCP2 bound to DNA much more widely than originally documented. “People in the labs initiate projects trying to find something interesting and worthwhile to do and present them at the next Consortium meeting, where they are criticized, suggestions are made, and people take a slightly different path forward.”

Several years earlier, Bird and his colleagues had completely changed the view of Rett Syndrome by reversing Rett symptoms in adult mice models by restoring MeCP2 function. Suddenly, the view of Rett Syndrome was transformed from a developmental disorder to a malfunction of mature neurons, raising the tantalizing question of whether Rett Syndrome can be treated later in the course of disease after children suffer from the symptoms.

The therapeutic targets expanded beyond neurons when the Mandel lab reexamined the brain cells hobbled by MeCP2 malfunctions. Mandel and her postdoctoral fellow, Nurit Ballas, who now runs a Rett Syndrome research group at the State University of New York, Stonybrook. found that MeCP2 was also present in normal glial tissue, including star-shaped astrocytes that support neuronal function in different ways. Then, the Mandel lab, in collaboration with Ballas, repeated the Bird lab’s genetic reversal experiment—but in astrocytes specifically, instead of all cells. In mice, restoring MeCP2 function in astrocytes reversed, stabilized or ameliorated some Rett-like symptoms.

A Consortium project in the Mandel lab addressed lingering questions about the role of astrocytes by showing that some MeCP2 defects in brain circuitry could be traced to the astrocytes. The experiments used another specialized mouse model developed by Bird’s group and pulled in the electrophysiology expertise of the Paul Brehm lab, across the hall from the Mandel lab

Meanwhile, in the earliest days of teaming up, the Bird and Mandel labs learned they had devised two different gene therapy approaches in mice and teamed up to test both. Testing both their approaches in female mice using a common viral vector stabilized or reversed Rett symptoms. The two labs reported these findings together and the postdocs shared co-authorship. Gene therapy to introduce viable copies of MeCP2 is under investigation by other independent labs and at least one biotechnology company.

The Mandel and Bird labs continue to collaborate on another therapeutic idea. Instead of delivering a functional MeCP2 gene, the idea is to repair certain MeCP2 mutations after the gene is transcribed into RNA and before the RNA is translated into the protein (RNA editing). Postdoctoral fellow John Sinnamon in the Mandel lab is leading the project. He reported to the Consortium positive early results from RNA editing in neuronal cells, first in a dish and then in mice. He used an assay developed by Lyst in the Bird lab to insure the repaired MeCP2 protein functioned normally.  Lyst also suggested expanding the approach to mouse models with other mutations sooner rather than later, including mutations studied in the Bird lab, so this was pushed ahead in the priorities. (This editing work funded by RSRT has seeded a new biotech company to work on Rett Syndrome, Vico Therapeutics, co-founded by Mandel and Judith van Deutekom.)

“It’s a wonderful environment to speak openly with others in the field and to speak freely without worrying about competition,” Sinnamon says. “We share reagents and ideas. There’s a spirit of: when you benefit, I benefit. People have no hesitation about saying, let me send you what I can. There’s the opportunity to bounce ideas off each other—‘we have this crazy idea, what do you think?’—and inevitably the discussion goes on for an hour.”

In fact, the Consortium experience seems to have made MeCP2 projects more attractive to junior researchers.

Consider the difference in the Greenberg lab. When Harrison Gabel started as a postdoctoral fellow a year before the Consortium formed, for example, MeCP2 had a reputation “as a place where postdoc projects could go to die,” he says. Gabel accepted the challenge anyway and soon thrived on the new collaborative problem-solving among the three labs. He ended up leading a Consortium project that found MeCP2 preferentially turned down preferred long neuronal genes. Now heading his own lab at Washington University, St. Louis, he calls the Consortium “an ideal way to do science—not only to produce new insights into MeCP2 biology, but also to serve as an example for us young investigators on how to embrace collaboration to attack these challenging problems.”

Six years later, when Lisa Boxer began her postdoctoral fellowship, she factored in what she had heard about the early Consortium experience, along with a personal desire to work on a neurological disorder, and chose MeCP2. “At the Consortium meetings people are willing to present early stage projects, which can be very useful to change what you’re doing or change direction,” she says.

That came in handy with Boxer’s project. MeCP2 dampens gene activity. She wanted to know at what step does that happen. After Gabel’s finding, the prevailing theory was the “speed bump” model—MeCP2 was putting the brakes on transcription of long genes, maybe at the many places where MeCP2 had latched on.

“We all thought this was how it worked, but everyone’s experiments were turning out inconsistent with this model,” said Boxer, referring to complementary studies in the Greenberg, Gabel and Bird labs. “One meeting, Adrian pointed out a paper he had published 20 years ago that showed MeCP2 could act on a gene from a distance.”

They took another look at their data. It affirmed that MeCP2 somehow was acting on the transcription start site, the one place it could not be found on the gene. Boxer published her findings in January 2020. The next question—how MeCP2 does that—will be in the hands of another postdoctoral fellow.

Boxer predicts MeCP2 will prove to be fundamentally different from other proteins that control gene activity. “MeCP2 is such a mystery,” Boxer says. “No one can agree on exactly what it does, yet clearly it is so important because mutations in it cause Rett. I don’t think it functions like other proteins.”

The Consortium officially involves three labs, but the Rett Syndrome Research Trust (RSRT) is a fourth presence. RSRT executive director Monica Coenraads not only helped broker the Consortium, she connected the idea with a large gift by RSRT Chairman Tony Schoener and his wife Kathy.

Coenraads and RSRT staff attend the Consortium meetings and interact with the scientists. The Consortium researchers value the opportunity to meet other parents and the girls whose conditions they are trying to understand at the molecular level.

Those interactions can bear scientific fruit. At one meeting, Lyst was presenting structural data to learn which mutations interfered with the normal interaction between MeCP2 protein and the NCoR complex. He remembers Coenraads pointing out that a person with a specific mutation in the NCoR complex—not MeCP2—also had a Rett diagnosis.

“Then we were able to ask whether that mutation had an effect on NCoR binding to MeCP2 and found there was a compelling reason to believe that binding was absolutely central to Rett Syndrome,” Lyst says.

After a decade, expertise in the Consortium labs has evolved from complementary to overlapping. Researchers in each lab still rely on their colleagues for fuller explanations of their specialty areas, but they will nurture and refine promising ideas. Mandel credits both the Consortium intellectual and financial support for her ability to pursue transformative work that showed certain MeCP2 RNA mutations could be edited.

For all their different perspectives and lively critiques, Consortium researchers can explain each other’s scientific contributions to outsiders, and they share a similar impatience to unlock therapeutic possibilities.

“All this frantic activity aimed at trying to make a therapy for Rett depends on evidence that, unlike some others, this neurological disorder is reversible,” Bird says. “None of us will be happy until we find something of practical use. I’m sure it’s curable, so it’s deeply frustrating that we haven’t been able to make that happen it yet.”