Researchers at the National Institutes of Health and other institutions have found a second genetic defect that accounts for previously unexplained forms of osteogenesis imperfecta (OI), a disorder that weakens bones, sometimes results in frequent fractures and is sometimes fatal.

The affected gene contains the information for a protein designated P3H1 (prolyl 3-hydroxylase 1), also known as leprecan. P3H1 is part of a complex of proteins that is crucial for refining collagen to its final form. Collagen is an important building block for bone. When the gene does not function, little or no P3H1 is produced, resulting in defective collagen, and a form of OI.

"The findings provide new insight into the causes of osteogenesis imperfecta and information on how bone is formed," said Duane Alexander, M.D., Director of NIH's National Institute of Child Health and Human Development. "This information may lead to advances in understanding osteogenesis imperfecta as well as understanding a variety of other bone disorders."

The study appears in the online version of "Nature Genetics".

In addition to NIH researchers, other authors of the study are from the University of Washington in Seattle, Shriner's Hospital for Children, in Chicago, and from Children's National Medical Center and Georgetown University Hospital, both in Washington, D.C.

Although there is no treatment for the disorder, the finding does allow OI experts to test families who have lost a child to OI for the presence of the defective gene. Couples with a child affected by this form of OI could be apprised of their risk for conceiving another child with the disorder. Similarly, siblings of children affected by the defective gene can also be counseled about their likelihood of carrying the gene.

OI is an uncommon disorder that occurs in 1 out of 15,000 to 20,000 births. About 85 percent of all OI cases are caused by mutations in the genes that contain the information needed to make collagen. Some cases of OI, however, could not be explained by mutations in the collagen gene. The authors had earlier predicted that these unexplained OI cases might be caused by absence of proteins that interact with, and chemically modify, type I collagen.

In a previous study, NIH scientists first discovered a defect in the gene that codes for cartilage-associated protein (CRTAP), one of the proteins that work with P3H1 during collagen synthesis. That finding appeared in the December 28, 2006 "New England Journal of Medicine".

Patients who have a loss of function of either CRTAP or P3H1 will develop severe OI, explained the senior author of both studies, Joan Marini, M.D., Ph.D., Chief of NICHD's Bone and Extracellular Matrix Branch. Of the two genes, mutations in P3H1 are not always fatal, as is the case when individuals do not have a functioning gene for CRTAP.

Dr. Marini said that since 1979, when OI was first classified, scientists have suspected that another form of the disease existed because about 15 percent of people affected with the disorder did not have mutations in the collagen genes known to account for OI.

The classical form of the disorder results from a defect in the genes for type I collagen, which serves as a kind of molecular scaffolding that holds together bone, tendons, skin, and other tissues. These collagen defects are dominant mutations, requiring only one copy of a mutant gene to cause bone disease.

The two defective genes found by the NIH researchers explain cases of the disease that did not have a mutation in a collagen gene, as occurs with the dominant form of OI. As a result of the current and previous study, these formerly unexplained cases are now understood to be a recessive form of OI-in which two defective copies of a gene are needed to cause a particular trait.

There are several known forms of OI, which vary in severity. In the most severe forms, infants may die shortly after birth. About one-fourth of the known collagen mutations are lethal. In other forms, individuals may lead relatively normal lives, but have bones that fracture easily.

The current paper in "Nature Genetics" described five children with severe or lethal OI who either did not produce P3H1 or had greatly reduced levels of it. Of the five, three children had a lethal form of OI; the other two had nonfatal OI with severely abnormal bone development. Dr. Marini and her team found that the children had a common genetic defect -- a lack of P3H1 -- and thus discovered a second form of recessive OI.

Dr. Marini estimated that the recessive form of OI caused by a mutation in the gene for P3H1 might occur in 4 percent to 6 percent of lethal cases -- about twice as frequently as the CRTAP defect. "This discovery has completely changed the ability to diagnose babies with lethal and severe OI," she said, adding that her lab had already performed a genetic test confirming that a child's unknown disease was recessive OI.

Dr. Marini pointed out that the cases referred to in the study were not selected by race, but by medical and biochemical findings. She said that, interestingly, four out of the five were either West Africans or African Americans of West African descent.

An identical genetic mutation was found among these four patients. Two infants who each had two copies of the West African mutation died within a month of birth. Two other children who had one copy of the West African mutation in combination with a different P3H1 defect have lived longer than 1 year but have severe abnormalities of bone development.

Defective P3H1 was also found in a teenager from Pakistan. His P3H1 mutation was different from the African mutation. His parents are cousins and he inherited the same P3H1 defect from each of them. His gene defect is not lethal but his bone development has been abnormal. His bones are extremely fragile and his growth has been deficient. At age 16, his height is comparable to that of a 3 year old.

Dr. Marini added that these findings suggest that it is possible West Africans and African Americans share a common P3H1 mutation for defective bone formation. Further research will investigate if this mutation does in fact have a high frequency in these groups.