Code breakers: scientists tease out the secrets of proteins that DNA wraps around.

Jamming a week's worth of clothing into a carry-on suitcase is tough, but consider the challenge a human cell faces with its DNA. More than 6 feet of this double-stranded molecule, making up a cell's 23 pairs of chromosomes, must get stuffed into the cell's microscopic nucleus. Just as people might roll or fold their clothes in special ways to stuff a piece of luggage, cells have devised tricks of their own to cram in all their DNA. One trick is to tightly wind the DNA around complexes of proteins called histones, much as thread is coiled around a spool. The histone-DNA combos, in turn, are folded and refolded to make up individual chromosomes.

When scientists originally discovered this packing system, they were befuddled. To make new proteins, certain cellular enzymes must read the sequence of nucleotides that make up a cell's DNA. But the enzymes can't do their job, the scientists reasoned, if the genetic sequences are locked in a tight embrace with histones.

Scientists have learned more recently that cells use various chemical modifications of histones to sometimes expose and sometimes sequester, thus turning genes off or on. As biologists start to understand these alterations, appreciation for the importance of histones is growing.

"They're not just spools on which DNA is organized and packed into the nucleus. They're intimately involved in regulating access to genes," says Shelley L. Berger of the Wistar Institute in Philadelphia.

In fact, 4 years ago, Brian D. Stahl and C. David Allis, both then at the University of Virginia Health Science Center in Charlottesville, coined the term histone code to represent the idea that specific histone modifications can be paired with specific genetic activity within a cell. For example, one pattern of histone chemistry turns up when a cell is dividing, while another pattern forecasts the death of a cell. More-permanent histone modifications may maintain a cell's specific identity, such as brain cell or liver cell.

Further deciphering this histone code and developing ways to manipulate it could have major medical payoffs, say both Berger and Allis. Compounds that interfere with how cells modify histones have already shown promise in treating tumors and Huntington's disease.

"The implications for human health are quite strong," says Allis, who now leads a histone-biology research team at Rockefeller University in New York.

TALE OF THE TAILS When it comes to biology, the genetic code has earned fame. Check any life science textbook and it will describe how different triplets of DNA nucleotides represent the 20 amino acids that make up natural proteins. The three-nucleotide sequences generally signal for one or another amino acid, but some simply tell a cell when to stop building a protein.

A histone code may be much more complex. Peer inside the nucleus of a human cell and zoom in on a chromosome. Scientists compare its structure to that of a string of beads, with each bead consisting of a 146-nucleotide-long DNA strand wrapped almost twice around a complex of eight histones.

Each bead contains two copies of four histones: H2A, H2B, H3, and H4. Several decades ago, as scientists began to piece together the structure of the beads, they observed that small groups of atoms known as acetyl or methyl groups frequently adorn the histones. "It began to emerge that the histone proteins were phenomenally decorated by these chemical flags," says Allis.

Scientists also noticed general correlations between certain patterns of histone decoration and gene activity. In particular, parts of chromosomes in which histones are covered with acetyl groups tend to have active genes, whereas deacetylated histones tend to harbor inactive genes. DNA near methylated histones is generally shut down.

Slowly, as researchers learned more about the structure of histones, it became evident that patterns of acetylation and methylation could be quite precise. It turns out that each histone has a tail, a flexible string of amino acids jutting out from the DNA-wrapped spool. Acetyl and methyl groups tend to plant themselves on particular amino acids in the tails, scientists found.

Histone tails have a lot to teach biologists, according to Allis. From species to species, he notes, these tails are nearly identical, implying that they are important to the cell. "Nature has held these things constant for a reason," says Allis.

In 1992, Bryan Turner of the University of Birmingham Medical School in England and his colleagues discovered that the male fruit fly's single X chromosome, but not its other chromosomes or the female's two X chromosomes, has a specific acetylated amino acid on the tail of histone H4. Since genes on the male fly's X chromosome are extra-active to compensate for their absence on the smaller Y chromosome, the investigators suggested that the acetylation accounts for the increased male-gene activity.

Another major breakthrough in histone biology occurred in 1996. That year, Allis' group identified the first histone acetylase, an enzyme that places acetyl groups on histone tails. A month after that work was published, a research team led by Stuart Schreiber of Harvard University reported the discovery of a histone deacetylase, an enzyme that strips the tails of such groups. Moreover, the acetylase and deacetylase had already been implicated in turning genes on and off, respectively. Together, the two reports made it clear that the enzymes regulate genes via histone tails. "All of a sudden, a beautiful mechanism emerged," recalls Berger.

By 2000, when Stahl and Allis proposed that there's an elaborate code of histone modifications, scientists had tallied several histone-tail decorations beyond methylation and acetylation and identified additional enzymes involved in this chemical accessorizing. In some cases, sugars or whole proteins, albeit small ones, mark histones. Biologists found, for example, that the protein ubiquitin, which was originally thought only to mark proteins for destruction, attaches to histone tails that remain intact.

In the Nov. 11, 2003 Proceedings of the National Academy of Sciences, Yuzuru Shiio and Robert N. Eisenman of the Fred Hutchinson Cancer Research Center in Seattle report that genes are turned off when members of a family of ubiquitin-like proteins are added to the tails of histone H4.

"There are all kinds of sites [on histone tails] that can be modified," says Berger. She adds, "The possibilities for a code are really quite enormous. It's not going to be a simple code."

How does acetylation or any other histone-tail modification influence gene activity? The original theory was that chemically modifying histones would make their electrical charge less positive. As a result, they wouldn't hold on as tightly to their DNA, which is negatively charged.

Today, biologists are more inclined to argue that modified histone tails act as landing pads for other proteins that influence the accessibility of DNA for gene activity. Turner put forth this idea a decade ago, but it proved a challenge to identify proteins that recognize specific histone-tail configurations.

In 2001, however, two research teams reported that heterchromatin protein 1, a molecule known to mediate the silencing of genes, binds to the amino acid lysine on the tail of histone H3 only if methyl groups adorn the lysine. Other proteins that bind specifically to modified histones have subsequently turned up.

"Now, we're finding these docking molecules," says Allis.

LIFE AND DEATH Allis notes that some investigators may quibble with the notion of a histone code. "Code suffers a bit from being a buzzword," he says.

Turner agrees. "I think we have to agree on what we mean by the histone code and what we expect from it," he cautions. "I think if the code is going to be worth anything, it has to have predictive value. It has to be passed on from one cell generation to the next"

Histone methylation, for example, appears stable, says Turner. Scientists haven't yet found an enzyme that strips methyl groups off a histone, and the patterns of this chemical tag appear to be transferred into both sister cells when a cell divides.

As a cell specializes, it may use histone methylation to permanently turn off unneeded genes and activate those that are essential to the cell's function. Histone methylation "looks like it's less transient and like it's more involved in the long-term setting of the genome," says Berger.

In contrast, a cell's pattern of histone acetylation may not qualify as a code, says Turner. Acetyl groups frequently hop on and off of histone tails, making it difficult to argue that they provide a cell with a discernible identity.

Arguing that there is a histone code, Allis cites other additional instances in which he can make predictions about a cell by reading its histones. He and his colleagues have found that if a cell has a phosphorus-containing chemical group tacked onto several amino acids on the tail of H3, the cell is dividing. On the other hand, if a particular serine on the tail of H2B has a similar phosphate group, the cell is about to commit suicide, the researchers reported in the May 16, 2003 Cell.

Allis refers to these two distinct histone markings as life codes and death codes. "It's cool to think that there may be a unique property of the H2B tail that encodes death" for a cell, he says.

If so, perhaps researchers can use the death code to kill tumor cells. Or, they could thwart the death code and thereby stop cells from dying in a variety of human illnesses, such as degenerative brain disorders. "This opens up therapeutic opportunities," says Allis.

Drugs that disrupt the putative histone code are already being wielded against lymphomas, leukemia, and other cancers. After cell and animal studies suggested that inhibiting histone-deacetylase activity could kill cancer cells, physicians cautiously began to test inhibitors of the enzyme on people. The initial concern was that such drugs would have dangerous side effects because they would also affect deacetylation in normal cells. That hasn't been a major problem so far.

"These drugs don't seem overly toxic," says Allis. "People are tolerating them reasonably well. More importantly, their tumors are disappearing."

Animal studies have also indicated that histone-deacetylase inhibitors might thwart the brain-cell loss characteristic of Huntington's disease (SN: 2/15/03,p.102). Mutant proteins generated in the illness seem to gum up the workings of histone acetyltransferases, so blocking deacetylase function may create a more normal histone state within cells.

DESIGNER HISTONES Even as researchers attempt to exploit their new understanding of histones for medical purposes, many questions remain about these DNA-wrapped entities. Consider the mystery of how a dividing cell produces two cells with seemingly identical methylation of the histones' tails. "How [cells] inherit the histone code is something we have to scratch our heads about," says Allis.

Scientists are coming up with new tools to probe histone biology. In the Oct. 14, 2003 Proceedings of the National Academy of Sciences, a research team described a strategy to synthesize large quantities of a histone with a chosen modification, such as acetylation of a particular amino acid in the histone's tail.

In essence, the investigators manufacture just the tail, with the chosen modification, and chemically glue it to a tailless histone created separately. The researchers can then wrap these designer histones with DNA and examine how combinations of histone-tail modifications influence a gene's activity.

"We're looking at creating these different types of molecules en masse. We're now making entire libraries of all possible modifications" says study coauthor Dewey McCafferty of University of Pennsylvania School of Medicine in Philadelphia.

With such designer histones, it seems that researchers are on their way to having in their hands all the words of the histone code. But, it may still be a stiff challenge to figure out what those words mean.