Enzymes Display Intricacy in Repackaging DNA

Demethylase Family Undoes Histone Triple Methylation

In 1884, a German physiologist Albrecht Kossel spotted a peculiar protein in the nucleus of bird blood cells. What stood out was the way the protein associated with the main nuclear substance, the nucleic acids. In some cells the protein bound the acids loosely, in others so tightly that chemical agents could barely pry it away. Kossel, who would win the Nobel Prize in 1910, dubbed the quirky protein “histone”—German for “of unknown origin.” It would be an apt designation. For nearly a century, histones were shrouded in mystery and misconception. For example, in the 1960s, researchers discovered that histones—thought to provide an inert scaffolding for the newly discovered DNA—could be altered at certain spots by the addition of up to three methyl groups. The changes were thought to be permanent.

Photo by Graham Ramsay

“Histone regulation is highly dynamic,” said Yang Shi (center). “We are still at the discovery stage, cataloguing changes and finding players that mediate these changes. Figuring out what they mean in the context of a biological output is something that will take decades.” Shi appears with Monica Colaiacovo and Johnathan Whetstine.

“That became the dominant view of the field for almost a half century—that histone methylation is likely irreversible,” said Yang Shi, HMS professor of pathology. Two years ago, Shi and colleagues discovered an enzyme that could remove a methyl group from histones bearing two such groups, shattering the longstanding myth and sending researchers racing to find more demethylases. Now, Johnathan Whetstine, a research fellow working with Shi and colleagues, reports that they have uncovered a whole family of enzymes, capable in this case of stripping methyl groups not from dimethylated, but from the more heavily decorated, or trimethylated, histones.

What’s more, they have identified the JMJD2 family of demethylases in cultured human cells and also in a living organism, the nematode C. elegans. The research appears in the May 5 Cell. In addition, they have completed a crystal structure of the active core of one of the enzymes, JMJD2A, which was published on May 4 in Cell online.

Methyl-group Modifications
The findings from the Shi lab are bound once again to shake up the fast-moving field of histone biology. While their 2004 study demonstrated that demethylases exist, this paper reveals their activity to be unexpectedly intricate. Histones consist of two main parts—a cylindrical core, around which DNA is wrapped, and a protruding tail, which can receive methyl groups at lysine and arginine residues. The dimethylation-reversing enzyme, LSD1, acts at a single lysine, while some members of the newly described family appear to act at two. In addition, certain JMJD2 members may act in a slightly different fashion. Most remove only a single methyl group, but one member, JMJD2D, can remove up to two. “There is some way that the system is fine-tuning the demethylation,” said Whetstine.

The existence of a family of enzymes that can do slightly different things further suggests that demethylation is an unexpectedly versatile process. “The balance between methylation and demethylation at specific sites can generate a whole array of methylated states, each of which can serve as a docking site for other proteins that, in turn, carry out biological functions,” said Shi.

For years, the clamping of methyl groups onto histones was thought to play a largely negative biological role, preventing transcription factors and other proteins from gaining access to genes. Over the past few years, however, methylation has been shown, in some cases, to activate transcription. In their new paper, Whetstine, Shi, and colleagues show that demethylation plays an unexpected role in another critical function, regulating the cell’s response to DNA damage. They came to this conclusion by blocking the activity of JMJD2A in worms. The worms exhibited an unusually high number of apoptotic cells in their germline. On closer inspection, these cells displayed a higher level of RAD51, a marker of DNA damage. “This is really a new area. We are now in a position to better understand the relationship between histone methylation and DNA damage,” said Whetstine.

The findings break new ground in other ways—this is the first time anyone has traced the function of a demethylase in a living organism. The new crystal structure of JMJD2A is also the first one to be completed of a demethylase. The structure, which reveals intriguing features, appears at an opportune time. Histone methylation, along with other modifications such as acetylation, is increasingly being recognized as playing a role in diseases, including cancer. “If demethylases are playing a role, they will be very good targets for chemicals because of their high level of specificity,” said Shi. “You might be able to design drugs by rational design.”

Enzyme as Quarry
Given the breadth of the discovery, it is remarkable how quickly it came together. After finding LSD1 in 2004, Shi began looking for a second demethylase—as did researchers all over the country. “Yang’s paper showing that demethylases actually exist ignited something ferocious. Chromatin labs ran right to it,” said Whetstine. Shi persuaded Whetstine to share the lead in the effort. Whetstine compiled a list of candidate enzymes and was on the verge of pinpointing a particularly promising dimethylation-reversing enzyme when Shi learned that another lab was about to publish it.

“We figured, OK—we lost on the identification of the second demethylase, we should focus on an enzyme that will reverse trimethylation,” said Shi. In fact, they had been looking for such a protein while carrying out their original candidate screen and, during the fall, Whetstine had spotted a very promising protein. Homing in on the enzyme, JMJD2A, he and his colleagues incubated it with histone tail segments, each trimethylated at a different site. The trimethylated proteins have a precise molecular weight and charge, which changes slightly upon demethylation. Using mass spectrometry to pick up these subtle changes, they found that JMJD2A removes methyl groups not at one but at two lysines, designated K9 and K36.

To test its demethylating powers, the researchers overexpressed JMJD2A in cultured human cells. Levels of trimethylated K9 and K36 plunged. Next, they used RNAi to knock down JMJD2A in the cells. The reverse happened—trimethylated K9 and K36 histone levels shot up. They repeated the RNAi experiments in C. elegans, paying special attention to the effects on the worm’s germ cells, which exhibit a dynamic pattern of histone methylation. Collaborating with co-author Monica Colaiacovo, HMS assistant professor of genetics, Whetstine and colleagues noticed large numbers of condensed, dark-staining nuclei. They turned out to belong to apoptotic cells. The cells also exhibited high levels of RAD51, the marker for DNA damage.

Convinced that JMJD2A was a functioning demethylase, they explored the possibility that its siblings—JMJD2B, JMJD2C, and JMJD2D—might have trimethylation-reversing abilities too. Mass spectrometry showed that they did, to varying degrees. One of them, JMJD2D, was capable of removing two methyl groups at K9.

They rushed to finish and submitted their paper in March. “Basically, Johnathan was not sleeping much,” Shi said.

“It was just a matter of physically working insane. From March 6th to the 8th, I did not leave the lab. I watched the sun come up and go down,” Whetstine said.

Dawn and dusk are reflective times, but it was Shi who offered two philosophical lessons about the experience of working in a highly competitive field of science. “In Chinese, the character for crisis also means opportunity. There are always these two sides that play into every aspect of biology,” he said. And the second? “The fear factor plays a huge role in scientific discovery.”

—Misia Landau

top