Forensic scientists analyzing bones found in the Gobi desert discovered that the DNA within them could be surprisingly easily extracted. In an experiment designed to mimic the conditions that affected those bones, baking a particularly difficult sample made the DNA much more easily extracted, probably because it makes it easier to break open more cells and expose more of the DNA molecules.

Mummies have always held secrets, but now one of them has led to a new DNA technique.

Our fascination with mummies has sold millions at the box office. Now these preserved people -- mummies more than 800 years old -- are helping scientists reveal the mysteries of the past.

University of New Haven forensic scientist Dr. Heather Coyle is experimenting with a new technique by going back in time.

These are skeletal remains recently gathered from a Gobi desert cave. Surprisingly, Dr. Coyle was able to extract DNA from these mummies, but when she tried the same method on a body found in the USA, she was not as successful. "We realized that the bone we were trying to process was not yielding DNA from the case we were working on," Dr. Coyle said.

Standard DNA procedure for bones is to freeze them. When Coyle and her team re-examined the mummy remains they realized the Gobi desert created a natural bone baking process.

"It makes the bone more brittle so it makes it easier to grind and break open more cells, so we think we are accessing more DNA to begin with," Dr. Coyle said. Dr. Coyle decided to mimic nature by baking the cold case bones for 72 hours. Liquid nitrogen was then poured into a pulverizer. The bone was placed inside, ready to be crushed. After a short cycle the bone was turned to powder and ready for DNA extraction.

Coyle hopes her new technique will someday help close the book on several cold case files.

What is DNA? DNA is the blueprint that encodes all the data for building a human body, along with instructions on how it should operate. Every cell in a person's body contains a copy of this DNA.

DNA typing is based on an unusual feature found in the human genome. There are multiple copies of certain short sequences, 3 to 30 base pairs long, that are repeated one after another as many as 100 times. These groups of repeat sequences are widely scattered through the genome. Everyone has these repeat units, but the number varies from person to person.

Only identical twins will have the same numbers and patterns of these sequences. These genetic data aren't instructions to make anything; scientists think they might exist to get mixed up in the regular genes and provide some variety for evolution.

Scientists had thought DNA-binding proteins primarily used full-body hugs for accurate readings of the information coded in the DNA's sequence.

Even a protein known to use the hug method, called direct readout, can effectively pinpoint sites on DNA using indirect readout, found researcher Nancy C. Horton and her colleagues.

"It was a total surprise," said Horton, a UA associate professor of biochemistry and molecular biophysics. "No one had ever seen it before."

Doing the quick squeezes that scientists call indirect readout probably works faster than requiring full-body contact with all the DNA, the researchers suggest. Quick and accurate identification of key sites on DNA is important for the health of all kinds of cells, from bacteria to humans.

To detect the protein-DNA connection in such detail, Horton and her co-authors Elizabeth J. Little and Andrea C. Babic studied a DNA-binding protein that bacteria use to protect themselves from viral infections.

The finding has implications for the development of designer drugs.

"People have and are developing DNA-binding proteins to turn genes on and off," Horton said. Such designer proteins can be used to cut out the bad copy of a gene and help replace it with good copy.

"We found that indirect readout is important for finding the right sequence, and we now think indirect readout is also important for finding it quickly," she said.

The team published their paper, "Early Interrogation and Recognition of DNA Sequence by Indirect Readout," in the December issue of the journal Structure. First author Little and co-author Babic were postdoctoral research associates in Horton's laboratory when they did the research. The two are now senior scientists at Ventana Medical Systems, Inc. in Tucson, Ariz.

The National Institutes of Health funded the research.

Horton studies proteins that bind to DNA.

Seven years ago, she figured out the structure of a protein called HincII that snips up DNA. The protein is a type of enzyme called a restriction endonuclease and comes from Haemophilus influenzae bacteria.

Since that time, Horton has been trying to learn how HincII interrogates the DNA to find the right place to cut.

The protein protects bacteria by cutting up DNA from invading viruses. Without the protective protein, viral DNA would commandeer the bacterium's cellular machinery to produce viruses and ultimately kill the bacterial cell.

The HincII protein distinguishes between bacterial DNA and viral DNA by recognizing certain sequences on DNA. Such a defense requires speed to prevent the marauding virus from killing the cell and also accuracy so the protein doesn't accidentally hack up the bacterium's own DNA.

Horton knew from her previous work that the HincII protein used the direct readout method to find the particular sequence of DNA that corresponded to enemy DNA. The protein seemed to distort the DNA to read it.

Removing the direct readout contact between the protein and the DNA might show whether the DNA distortion or the contact itself was important, Horton said.

Therefore Little and Babic created a mutant protein that couldn't hug DNA closely and therefore couldn't use the direct readout method. Little described the mutant protein as missing the fingers the normal protein used to probe the DNA.

"If the finger was doing all the recognition, then the mutant should cut any DNA sequence," Horton said.

To see how the mutant interacted with DNA, the researchers crystallized the mutant protein-DNA complex in action.

Initially, Horton thought the assay had gone wrong and almost threw the results in the trash, she wrote in an e-mail.

The mutant protein had chosen the proper site on the DNA with 100 percent specificity, which was opposite from what she expected. In addition, the DNA was distorted, even though the mutant couldn't make the strong contact a normal protein would.

"I did a double-take. I was just taking a picture to have a record that it was non-specific," she said.

Understanding how endonucleases and other DNA binding-proteins recognize a particular DNA sequence provides insight into key cellular processes including the replication, transcription and repair of DNA.

Little said, "In every single one of your cells are proteins looking for the proper sequences in DNA in order to make the proteins you need to stay alive."

Horton added, "Understanding how these processes work helps in the understanding of diseases so that we could potentially cure the disease."