Today’s topic in these GEITP pages is an excellent example of controversial conclusions (over more than two decades) able to be overturned by additional experiments; this is an example of how rigorous science is supposed to correct itself. Osteocalcin (OCN) is a 46-amino-acid protein, synthesized and secreted almost exclusively by osteoblasts (terminally-differentiated cells responsible for synthesis and mineralization of bone matrix during development of the skeleton — and its periodic regeneration throughout life). Osteoblasts originate from mesenchymal progenitors and are short-lived cells that are constantly replaced — depending on demand for bone formation. OCN secreted by osteoblasts contains three γ-carboxyglutamic acid residues that confer high affinity to the bone hydroxyapatite matrix. However, when bone is resorbed by osteoclasts (a macrophage-derived cell-type), the carboxyl groups on OCN are removed, and decarboxylated OCN is released into the circulation. Circulating levels of decarboxylated OCN are therefore dependent on the rate of bone turnover, also known as bone-remodeling.
Originally thought to function exclusively in bone, a more expansive view of decarboxylated OCN as an endocrine hormone has evolved over the past two decades; this all began with characterization of “an OCN-knockout mouse.” OCN had been proposed to act on multiple end organs and tissues — including pancreas, liver, fat cells, muscle, male gonads, and brain — to regulate functions ranging from bone mass accumulation to body weight, adiposity, glucose and energy metabolism, male fertility, brain development, and cognition. Because of such a wide range or purported target cell-types, OCN has therefore been considered a hormone (any signaling-molecule, produced by glands in multicellular organisms, that is transported by the circulatory system to target distant organs to regulate physiology and behavior).
However, many puzzling differences between mouse and human OCN function(s) have been reported. One possible explanation (for differences between results in mice and in humans) is that OCN genetics and function differ between humans and mice; humans have a single OCN gene named BGLAP (bone gamma-carboxyglutamate protein), whereas mice have two adjacent genes, Bglap and Bglap2. That OCN is an endocrine hormone with pleiotropic effects — is widely cited in textbooks and reviews and has provided the rationale for numerous clinical studies on the relationship between OCN and diabetes or obesity. This view has now been seriously challenged by two studies of independent OCN-knockout mouse models: PLoS Genet Jun 2020; 16: e1008586 and PLoS Genet Jun 2020; 16: e1008361.
In the first paper, authors replaced DNA encoding BGLAP and BGLAP2 proteins with a neo cassette in embryonic stem cells, and then investigated the role of OCN on bone formation and mineralization, as well as glucose metabolism,
testosterone production, and muscle mass. In contrast to results reported more than two decades ago, OCN was found NOT to participate in bone formation (or resorption) and bone mass in either estrogen-sufficient or estrogen-deficient
states. Instead, OCN was shown to be indispensable for alignment of biological apatite crystallites parallel to collagen fibers (see Figure in attached review). Loss of OCN function had NO EFFECT on collagen orientation. However, bone strength was decreased in the OCN-deficient mice — indicating that alignment of crystallites with collagen fibers is one of the elusive determinants of bone quality that (together with bone mass) determines the ability of bone to resist fractures. In addition, OCN was demonstrated TO PLAY NO ROLE in exercise-induced bone formation, glucose metabolism, improvement of glucose metabolism caused by exercise, testosterone synthesis, spermatogenesis, or muscle mass. ☹
In the second paper, authors used CRISPR/Cas9-mediated gene editing to delete most of the Bglap and Bglap2 protein-coding regions. Those gene-edited mice have no circulating OCN — but revealed normal bone mass as well as normal blood glucose and normal male fertility. Also, RNA-seq transcriptomics of cortical bone samples from the OCN-deficient mice showed minimal differences from non-gene-edited control mice. The gene-edited mice DO exhibit increased bone crystal size and maturation of hydroxyapatite, consistent with that found in the first paper, as well as earlier evidence by many other groups, and the general consensus that OCN plays a role in mineralization.
How can we explain the apparent discrepancies between these two recent articles and other conclusions made for more than two decades? Genetic background of the mouse strain, modifier genes, and differences in the molecular genetics of the knockout alleles — remain possible explanations. However, both groups explicitly state that these latest mouse lines will be publicly available so that their findings can be confirmed and extended by other interested investigators; indeed, the importance of resource-sharing is one of the most valuable take-home messages of this story. 😊
PLoS Genet Jun 2020; 16: e1008714
Well, the take-home message of that OCN story was that there can be discrepancies among knockout mouse models — due to issues such as genetic background of the mouse line, modifier genes, and differences in how the knockout alleles had been created. From the time of the first OCN-knockout mouse study, from which numerous textbooks and review articles had written what they believed was 100% correct, it took 24 years before two new knockout mouse lines proved that original study to be inaccurate.
A second take-home message of that OCN story is simply to demonstrate that “science is never settled.” New approaches and new experiments can disprove a previously-established “fact” based on what had been done at that time.
From: W T. P
What about vitamin K2 variability — genetic differences in response, varying levels of dietary intake? Vitamin K2 is understood to be required for OCN function as well, correct? It would be most interesting to know/examine whether Vitamin K2-deficient mice have normal bone mass, assuming lethality under conditions of Vitamin K2 deficiency does not preclude such experiments (while maintaining adequate levels of Vitamin K1, of course).
Because of its practical significance to human health, vitamin K2 is more significant genetic variability among OCN-knockout mouse lines. Thank you for an interesting topic.