Genome-wide mouse study yields link to human leukemia

Thanks to a handful of very special mice, scientists have discovered a new tumor suppressor gene and a unique chemical signature implicated in the development of human leukemia, findings that open up a "treasure box" of opportunity and possibility, study authors say.

Researchers in The Ohio State University Comprehensive Cancer Center bred a type of mouse that develops acute lymphoblastic leukemia (ALL). The mouse first goes through a pre-leukemic stage marked by rapidly expanding T cells and natural killer cells, both major components of the immune system.

In comparing the mice in the pre-leukemic stage and those with ALL with normal mice, researchers found that methylation, a chemical process that adds methyl molecules to DNA, silenced a number of genes – but only in the mice with full-blown ALL.

Further tests revealed that the methylation pattern in the mice with leukemia is strikingly similar to the pattern of methylation in human leukemia.

In the process, the researchers also identified a new gene that when methylated, appears to interrupt normal cell death, a process called apoptosis.

"It's given us a whole new way to look at and possibly treat leukemia," says Michael Caligiuri, director of the OSU Comprehensive Cancer Center (OSUCCC) and senior co-author of the study. "It's also validated our mouse model as a good predictor of what happens in the development of human disease," he added.

The findings appear in Nature Genetics online at http://www.nature.com/ng/.

"This is the first time anyone has examined methylation in leukemia on a genome-wide basis in a mouse, and the findings offer important implications for patient care, since we know that methylation, which alters gene function, can be reversed," says Christoph Plass, senior co-author and a member of the OSUCCC's Molecular Biology and Cancer Genetics and Experimental Therapeutics Programs.

While it was Caligiuri's laboratory that designed the mouse model, it was Plass who supervised the methylation studies. He and his colleagues used a system called Restriction Landmark Genome Sequencing (RLGS) to compare methylation patterns among the three groups of mice – a method of using enzymes and gel electrophoresis to map tiny bits of DNA on a grid. The stretches of DNA, referred to as fragments, show up as smudgy blobs on a test film. If a fragment is dark and definite, it is not methylated. If, on the other hand, it loses at least 30 percent of its intensity, it is regarded as methylated.

In the study, the research team tested 2447 fragments in each animal. They found anywhere from 45 to 209 (.8 percent to 8.5 percent) of the fragments methylated in the mice with cancer, but only one or two methylated fragments in the other mice.

"Interestingly, that same range of methylated fragments is exactly what we find in human leukemia, too," says Caligiuri, "so that gives added merit to our mouse model as an investigative tool."

Using data from the methylation studies, Caligiuri and Plass were able to identify a particular stretch of DNA, called Id4, as a tumor suppressor gene.

Tumor suppressor genes help control cancer by identifying and getting rid of defective cells before they have a chance to mature and divide. When tumor suppressor genes lose that ability – as they can if they are silenced through methylation or some other process, it gives cancer a chance to establish a foothold and spread.

Caligiuri says much more work needs to be done, but adds that the identification of Id4 as a likely tumor suppressor gene gives clinicians another possible target for intervention.

"We already have a drug, decitabine, that we know can reverse the effects of methylation," says Plass. "We are just beginning to figure out how it best works in humans, but simply knowing that we have a new target that may be meaningful in treating leukemia is a big step in the right direction."

Grants from the National Cancer Institute and the Leukemia and Lymphoma Society supported the research.

Additional co-authors from Ohio State include Li Yu, Chunhui Liu, Jeff Vandeusen, Brian Becknell, Zunyun Dai, Yue-Zhong Wu, Aparna Raval, Te-Hui Liu, Wei Ding, Charlene Mao, Shujun Liu, Laura Smith, Stephen Lee, Guido Marcucci and John Byrd. Laura Rassenti, from the University of California , San Diego , also contributed to the project.

Source : Ohio State University