Experiments in mice and observations in humans have suggested the bone protein osteocalcin acts as a hormone regulating, among other things, metabolism, fertility, exercise capacity and acute stress. That interpretation is now partially in doubt. Two independent papers published yesterday (May 28) in PLOS Genetics, each of which presents a new osteocalcin knockout mouse strain, report that glucose metabolism and fertility were unaffected in the animals. While some researchers praise the studies, others highlight weaknesses.
“I thought they were very good papers. I think the authors should be congratulated for very comprehensive studies of both skeletal and extraskeletal functions of osteocalcin,” says emeritus bone researcher Caren Gundberg of Yale School of Medicine who was not involved in the research.
Skeletal biologist Gerard Karsenty of Columbia University disagrees. “There have been 25 laboratories in the world . . . that have shown osteocalcin is a hormone,” says Karsenty. These two papers “do not affect the work of [those] groups,” he adds, “because they are . . . technically flawed.”
This tiny protein, one of the most abundant in the body, is produced and secreted by bone-forming osteoblast cells. In the 40 or so years since osteocalcin’s discovery, its precise function, or functions—whether in the bone or endocrine system—have not been fully pinned down.
Studies from Karsenty’s lab more than 10 years ago were the first to indicate that osteocalcin could act as a hormone, regulating glucose metabolism. But the suggested hormonal function has been questioned for its relevance to humans. For example, while studies in people have shown that levels of osteocalcin in the blood are correlated with diabetes, whether this is a cause or effect is unclear.
“If the findings were relevant to human disease, one would have anticipated more activity in clinical trials, but this has not been evident,” says Gundberg. “Where are the pharmaceutical companies?”
She’s not alone in her doubts. “There are some very outrageous claims in that [hormone] research that, if you have a little bit of understanding of physiology, pathophysiology, and clinical medicine, doesn’t make sense,” says bone metabolism expert Stavros Manolagas of the University of Arkansas for Medical Sciences who was not involved in the studies but who has written a review of the two recent papers.
Much of the work on osteocalcin has been based on mice created by Karsenty’s group that have a knockout of the genes encoding the protein. In the latest work, Bart Williams of the Van Andel Institute and Toshihisa Komori of Nagasaki University have independently created their own osteocalcin knockout mice to examine the protein’s functions.
Williams’s team used CRISPR-Cas9 gene editing to delete the two genes—Bglap1 and Bglap2—that encode two identical osteocalcins in mice. Komori’s used homologous recombination similar to the original method of Karsenty and colleagues to delete the genes. In all three knockout mice—the two new ones and Karsenty’s—the nearby gene Bglap3 has been left intact. This gene encodes a protein that differs by four amino acids to osteocalcin and is expressed in kidney not bone.
Williams’s and Komori’s teams observed some differences in bone structures between their knockout mice and controls. For example, Komori and colleagues found osteocalcin to be necessary for the alignment of minerals in the bone matrix, while Williams and his group observed differences in collagen maturity and phosphate-carbonate ratios (Komori and colleagues did not find this). The teams found no significant differences in glucose levels or fertility as had been reported previously.
“We expected to recapitulate some of the phenotypes that were reported,” says Williams. “I was surprised that we didn’t.”
“We analyzed the mice thoroughly, but we could not find any hormonal functions,” writes Komori in an email to The Scientist.
These metabolic studies were not thorough enough, nor were they performed optimally, to conclude that osteocalcin doesn’t function as a hormone, according to endocrinologist Clifford Rosen of the Maine Medical Center Research Institute who was not involved in the research. Glucose levels are “notoriously unreliable” in mice, for example, he says. A better approach would have been to use a glucose clamp technique, which controls glucose levels in an animal to assess how the body responds.
More importantly, Rosen adds, neither team injected osteocalcin into wildtype mice. Deleting a gene and not seeing an effect could be down to a number of reasons, he says, such as genetic background, off-target effects of gene-editing, or compensation by other genes, including possibly that of Bglap3, which Williams and colleagues noted increased expression by eight-fold in the knockout mouse bones. But, “the proof of principal for an endocrine hormone [is] showing that adding it has an impact on target tissues . . . [and] that is what’s missing from these papers.”
As of yet there’s no clear explanation as to why the new mice do produce the same results as those used in other studies, but says Williams, “a lot of times when there is confusion it’s because there’s something important underlying it and maybe working out why there are these differences might be very illuminating in ways we can’t even anticipate right now.”
C.R. Diegel et al., “An osteocalcin-deficient mouse strain without endocrine abnormalities,” PLOS Genet, 16:e1008361, 2020.
T. Moriishi et al., “Osteocalcin is necessary for the alignment of apatite crystallites, but not glucose metabolism, testosterone synthesis, or muscle mass,” PLOS Genet, 16:e1008586, 2020.