The secret appears to be a secreted signaling protein called Wnt10b. Known to inhibit the development of adipose tissue in mice, Wnt10b also stimulates the growth of bone cells, according to a new study that will be published February 21 in the Online Early Edition of the Proceedings of the National Academy of Sciences.
"High levels of Wnt10b expression in bone marrow directly increased bone mass and density in our experimental mice," says Ormond A. MacDougald, Ph.D., associate professor of molecular and integrative physiology in the U-M Medical School. "This is the first identification of a specific signaling protein in the Wnt family that regulates bone formation."
Wnt10b is one of a family of 19 related proteins. Wnts (pronounced "wints") regulate the complex changes that take place as an embryo develops. One step in this process determines the fate of primitive cells called mesenchymal stem cells.
"In bone marrow, mesenchymal stem cells have the potential to become either fat cells called adipocytes or bone-forming cells called osteoblasts," MacDougald says. "In adult animals, including humans, there's a reciprocal relationship between bone and marrow fat. Our research indicates that Wnt10b's signal blocks the fat cell pathway and stimulates the osteoblast pathway, which means less fat and more bone."
To study the effect of Wnt10b gene expression on tissue development, MacDougald's research team created an artificial sequence of DNA called a transgene linking Wnt10b to the FABP4 promoter, which is expressed in fatty tissue and in bone marrow. U-M scientists injected the transgene DNA into fertilized mouse eggs, and then bred mice that inherited the new gene to create the transgenic animals used in their research.
Kurt D. Hankenson, D.V.M., Ph.D., a U-M assistant professor of orthopedic surgery and laboratory animal medicine, and Christina N. Bennett, a U-M graduate student and first author of the PNAS paper, used a technology called micro-computerized tomography to scan femur (leg) bones from mice that inherited the FABP4-Wnt10b gene combination and compare them to scans from normal mice.
Bennett and Hankenson discovered that femurs from the transgenic mice had almost four times as much bone, and were mechanically stronger than femurs from control mice. (Note to editors: An image showing the femur scan comparison is available.)
"It was a very exciting moment the first time we saw scans showing increased bone mass in transgenic mice," Bennett says. "Visually, we don't see any abnormal side-effects in bone from the transgenic mice. Its development and morphology appear to be completely normal."
Loss of bone often develops with aging, but Wnt10b transgenic mice maintained their high levels of bone mass up to the ripe old age of 23 months, when the study was concluded.
Estrogen deficiency in females is another common cause of bone loss. When U-M scientists removed ovaries from normal mice in the study, they developed reduced bone mineral density and bone volume. But the Wnt10b females showed no bone loss after their ovaries were removed. "Because the transgenic mice have more trabecular bone, or bone within the marrow cavity, to begin with, they are doubly protected from the usual loss of bone density due to estrogen deficiency," MacDougald adds.
To confirm that Wnt10b was the key to increased bone formation, Bennett and Hankenson scanned bones from a strain of laboratory mice that didn't have a gene for Wnt10b. Lacking the ability to produce Wnt10b protein in bone marrow cells, these mice had 30 percent lower bone volume and bone mineral density than normal mice.
Using PCR analysis of Wnt10b-expressing cells in bone marrow, MacDougald found high levels of collagen and alkaline phosphatase, and expression of transcription factors that turn on genes involved in bone formation.
Bennett discovered another important clue when she found that Wnt10b expression shuts down activity of a gene called PPAR-gamma, which is required for the development of adipocytes or fat cells. "It suggests that Wnt10b's role may be to block PPAR-gamma, shifting development from the adipocyte pathway to the osteoblast pathway," she says.
In future research, MacDougald hopes to unravel the molecular mechanism for Wnt10b's bone-building effect. "It's not only an important scientific question, it's important to the understanding and potential treatment of osteoporosis and other human diseases," he says. "Right now, there is a need for drugs on the market to stimulate new bone formation. Being able to activate Wnt signaling in bone marrow and osteoblasts might help prevent the loss of bone associated with aging or menopause."
The research was funded by the National Institutes of Health, the U-M Diabetes Research and Training Center, the U-M Core Center for Musculoskeletal Disorders, and the Nathan Shock Mutant and Transgenic Rodent Core. Fellowships to Christina Bennett were from the Tissue Engineering and Regeneration Training Grant and the American Physiological Society Porter Fellowship. Kenneth Longo was supported by a mentor-based postdoctoral fellowship from the American Diabetes Association.
The experimental mice used in the study were produced in the U-M's Transgenic Animal Model Core facility. The University of Michigan has filed for patent protection on the Wnt10b transgenic mouse.
Additional collaborators on the study include Kenneth A. Longo, Ph.D., a former research fellow in MacDougald's lab who is now a postdoctoral fellow in the U-M School of Dentistry; Wendy S. Wright, research associate; Larry J. Suva, Ph.D., Center for Orthopaedic Research, University of Arkansas for Medical Sciences; and Timothy F. Lane, Ph.D., Jonsson Comprehensive Cancer Center, University of California, Los Angeles, who developed the Wnt10b knock-out mouse.
MacDougald and his research team published a paper in the August 2004 issue of the Journal of Biological Chemistry, which showed that Wnt10b over-expression in adipocytes produced mice with 50 percent less body fat and fewer fat cells.
Reference: Regulation of osteoblastogenesis and bone mass by Wnt10b, PNAS Early Online Edition, www.pnas.org_cgi_doi_10.1073_pnas.0408742102. To appear in print March 1, 2005, vol. 102, no. 9