Amjad Javed, Ph.D., of the University of Alabama at Birmingham, has taken a major step forward in understanding the bone development function of a gene called runx2, which could lead to future ways to speed bone healing, aid bone bioengineering, stem osteoporosis and reduce arthritis.
Javed, a professor in the UAB School of Dentistry's Department of Oral and Maxillofacial Surgery, says the results will contribute to future personalized medicine. This month, Javed presented this work to a standing-room-only audience at the International Association for Dental Research Annual Meeting in Boston. The work was published recently in two articles in the Journal of Bone and Mineral Research.
It was well-known that the deletion of both copies of the runx2 gene is lethal and the organism cannot form bone, teeth or cartilage.
To learn about the function of runx2 in specific cells types, Javed and his colleagues developed mice in which both copies of the runx2 gene were removed in only one of two key cells for bone tissue — either chondrocytes or osteoblasts.
"Our objective was to dissect and tease out which cell is really contributing what in bone development," Javed said. "Runx2 is vital. But when we talk up personalized medicine, we need to identify which specialized cells to target within bone tissue."
Study of these mice (technically known as the next-generation conditional knockout runx2 model) shows that chondrocytes and osteoblasts have surprisingly different functions in bone formation during gestation or after birth:
•Chondrocytes are involved in bone mineralization during embryonic development.
•Osteoblasts are involved in bone growth during postnatal development.
This is a major step forward in understanding the biology of bones — the dynamic, complex organs that are actively remodeled throughout life. Bones have cartilage-producing cells (chondrocytes), bone-creating cells (osteoblasts), bone-eating cells (osteoclasts), neuronal cells and blood-forming (hematopoietic) cells. Connective tissue and muscle surround the bones.
Javed's model began with the cartilage-producing cells. "We first removed the runx2 gene in chondrocytes, cells that are fundamental for every cartilage tissue in the body," Javed said. "Our first surprise was lethality at birth."
The skull of the mouse neonates was normal (skull bones are formed through a different bone-creation process); but the cartilage of all the other bones in the body failed to mature and get replaced by mineralized bone, a process known as endochondral ossification. So runx2, which had previously been thought to function only in the developing terminally mature chondrocyte, now appears to act earlier.
Without the runx2 gene, chondrocytes are unable to proliferate and differentiate into the column of cells needed for bone formation and lengthening. The deletion mutant showed that runx2 directly regulates a unique set of four cell-cycle genes to control the proliferative capacity of chondrocytes. The runx2 mutant mice also suffered dwarfism due to a near absence of a proliferative zone in the growth plates of bones.
When Javed's group removed runx2 in osteoblast cells, the results were again surprising. "Here we expected lethality," Javed said. "To our surprise, they were born alive. When we saw that, we thought it was a mistake. We started questioning, 'What happened?'"
It had been thought that the runx2 mutation would prevent the osteoblasts from differentiating into their final developmental stage of mature osteoblasts, and thus leave the mice boneless. But the results showed that the runx2 gene is not essential after the cells have already committed to becoming mature osteoblasts. Thus, most of the bones of the mice developed normally, though they had poor calcification.
The exception was the skull, which is formed through a different bone-creation process (intramembranous ossification). The skulls of the runx2 mutant mice had open fontanelles and sutures — the soft spots of the head between the six plates of the skull, spots that are supposed to fuse into bone after birth. Open fontanelles are one of the hallmarks seen in the human hereditary congenital disorder cleidocranial dysplasia (CCD), caused by a heterozygous runx2 mutation; but the mice did not show the other hallmark of human CCD, a missing collarbone. Lacking a collarbone, CCD patients are able to touch their shoulders together in front of their chest.
"Then, we had an additional surprise," Javed said of the mice with runx2 deletion in osteoblasts. "It is the postnatal skeletal growth that is affected. The mouse starts normally, but by three months of age — which is equivalent to 18-21 years in humans — the mice had 30 percent less weight than wild-type mice."
The runx2 mutation in osteoblasts caused poor alignment of the collagen scaffold that provides structural strength to mineralized bone. This left the bones brittle, less stiff and prone to fractures. As an additional surprise, the runx2 mutation in osteoblasts caused a significant reduction in osteoclasts (the bone-eating cells that work together with osteoblasts in bone remodeling).
The runx family
Runx2 — as well as two related genes called runx1 and runx3 — is a master transcription factor that controls at least a thousand other genes. The developmental impact of these runx genes (pronounced either "runks" or "run-ex") has long been known by making deletion mutants of each gene. These deletion-mutants are lethal during gestational development, but each master transcription factor controls a very different tissue.
While runx2 controls bone, teeth and cartilage, organisms with loss of the runx1 gene suffer problems in hematopoiesis. A deletion of runx3 gene leads to hyperplasia of the gastrointestinal tract.
Javed's colleagues in the runx2 work, which took three years to fully develop and test, are Haiyan Chen, M.D., Ph.D., Farah Ghori-Javed, M.D., Harunur Rashid, Ph.D., Mitra Adhami Ph.D., and John Clarke, all of the Department of Oral and Maxillofacial Surgery, Institute of Oral Health Research in the UAB School of Dentistry; Rosa Serra, Ph.D., of the UAB Department of Cell, Developmental and Integrative Biology, UAB School of Medicine; Yang Yang, M.D., Ph.D., of the UAB Division of Molecular and Cellular Pathology, UAB Department of Pathology in the UAB School of Medicine; and Soraya Gutierrez, Ph.D., of the Universidad de Concepción, Chile, Departamento do Bioquímica y Biología Molecular.
University of Alabama at Birmingham