Yongheng and collaborators present novel data demonstrating that systemic administration of magnesium ascorbyl phosphate (MAP, a sable form of ascorbic acid) by oral gavage promotes bone formation in rodents. They also show that MAP stimulates osteogenic differentiation of human bone marrow-derived mesenchymal stem cells cultured in vitro, and that this effect is mediated by calcium/calmodulin-dependent serine/threonine kinase II (CaMKII).
Although the manuscript is well written and presents some interesting aspects, the following points limit the impact and the overall quality of this study:
1- The effects of ascorbic acid on bone homeostasis in vivo and osteoblast differentiation in vitro has been widely documented. The signaling pathways activated and transcription factors upregulated downstream have been also characterized, at least in part. The effect of MAP itself on human MG63 cells in vitro has been previously reported. Although the description of the interaction of MAP and CaMKII is novel and interesting, most of the effects of MAP presented here were expected. It would be important to show the effect of MAP treatment on non-ovariectomized mice. Moreover, the in vivo phenotype in figure 2 could be completed by showing osteoprogenitor and osteoblast numbers. The RANKL/OPG ratio could be also evaluated.
2- There is no demonstration that CaMKII is mediating MAP effects in vivo. Hence, the title appears misleading, since it is not specified that this conclusion is drawn from in vitro data only. Mice floxed for CaMKIIa exist and could be used to conditionally knock CaMKIIa out of osteoblast lineage cells. Of note, it is not totally clear which CaMKII isoform is studied in this manuscript. We assume that it is the alpha isoform based on what is written page 5 lines 13 and 14, but it is not clear if this isoform is specifically targeted in the experiments performed by the authors.
3- The results obtained in vitro using siRNA against CaMKII and the inhibitor NG-93 are surprising, since experiments are conducted for several days (3, 7 and 14 days) while we do not expect long lasting effects of siRNAs, and even more so for the inhibitor. Cells have been treated for 30 minutes with NG-93 prior to MAP treatment over several days. How could CaMKII be inhibited for up to 14 days after removing the inhibitor?
4- It would be important to compare the in vivo effects of MAP treatment, at least with AA, and perhaps other bone anabolic agents to get a better sense of the relative efficacy of MAP. This aspect is addressed only in vitro in supplementary figure 2 (which could be presented in the results section, not just in the discussion).
5- It would be interesting to assess the effect of MAP on chondrogenesis and adipogenesis.
6- AA is known to affect HIF signaling. We understand that all experiments have been performed in non-physiological (hyperoxic) conditions (21% O2). It would be important to evaluate the effect of MAP in conditions where HIF signaling is activated.
7- Cell proliferation rates (Ki67+ %) and OSX+ osteoprogenitor numbers should be provided separately.
8- The experimental design appears sometimes confusing or inappropriate. MPA treatments are performed in vitro using osteogenic media that already contains ascorbic acid. This complicates the interpretation of the results. The duration of MAP treatment and the nature of the media used are not always specified. This information should be always mentioned in the figure legend, and materials and methods.

Yongheng and collaborators have adequately addressed all concerns and present novel data in mice indicating that systemic administration of magnesium ascorbyl phosphate (MAP, a stable form of ascorbic acid) may be a relevant alternative as bone anabolic agent to treat osteoporotic patients.