We may gauge how we’re aging based on visible changes, such as wrinkles. For years, scientists have been trying to gauge aging based on changes inside our cells.
Many alterations occur to our DNA as we age. Some of these changes are epigenetic—they modify DNA without altering the genetic sequence itself. These changes affect how cells in different parts of the body use the same genetic code. By controlling when specific genes are turned on and off, or “expressed,” they tell cells what to do, where to do it, and when to do it.
One such type of modification occurs when chemical tags known as methyl groups attach to DNA in specific places. This process, known as methylation, affects interactions between DNA and protein-making machinery. Changes in DNA methylation—both increases and decreases—occur with aging.
Dr. Steve Horvath from the University of California, Los Angeles, examined the relationship between DNA methylation and aging. He took advantage of publicly available methylation datasets, including ones from The Cancer Genome Atlas, a joint effort of NIH’s National Cancer Institute (NCI) and National Human Genome Research Institute (NHGRI). The datasets were developed by hundreds of researchers and comprised almost 8,000 samples of 51 healthy tissues and cell types. Samples came from people ranging in age from newborns to 101 years. They included tissues from throughout the body, including the brain, breast, skin, colon, kidney, liver, lung, and heart.
Horvath first developed an age predictor using 39 datasets. The tool was based on 353 specific DNA sites where methyl groups increased or decreased with age. He then tested the predictor in 32 additional datasets. Results appeared in the October 21, 2013, issue of Genome Biology.
Horvath found that the computed biological age based on DNA methylation closely predicted the chronological age of numerous tissues and cells to within just a few years. There were some tissues, however, where the biological age did not match the chronological age. These included skeletal muscle, heart tissue, and breast tissue. The clock also worked well in chimpanzees.
In both embryonic and induced pluripotent stem cells—genetically altered adult cells with characteristics of embryonic stem cells—the DNA methylation age proved to be near zero.
Horvath also analyzed nearly 6,000 samples from 20 different cancers and found that cancer greatly affected DNA methylation age. However, in most cancers the age acceleration didn’t reflect the tumor grade and stage.
The rate of ticking of the biological clock, as measured by the rates of change in DNA methylation, wasn’t constant. It was faster from birth to adulthood, and then slowed to a constant rate around the age of 20.
Horvath didn’t find evidence of a relationship with DNA methylation age in B cells (a type of white blood cell) from people with a premature aging disease (progeria).
“Pinpointing a set of biomarkers that keeps time throughout the body has been a 4-year challenge,” Horvath says. “My goal in inventing this age-predictive tool is to help scientists improve their understanding of what speeds up and slows down the human aging process.”
UCLA has filed a provisional patent on the age-predictive tool, which is freely available to scientists online. Horvath plans to examine whether DNA methylation is only a marker of aging or itself affects aging.