Maintenance of bone health by osteoblasts in the absence of osteocytes: comparative analysis of gene expression in medaka and zebrafish vertebrae

Osteocytes are regarded as the strain-sensing bone cells; their capacity to sense and convert mechanical loads into biological signals underpins our current understanding of how bone is (re)modelled to normalise local strains. However, most extant fish are anosteocytic and lack osteocytes, yet their bone structure and (re)modelling abilities remain unimpeded, sitting uncomfortably with this paradigm. We propose that alternative, non-osteocytic, mechano-sensing mechanisms are present in anosteocytic fish bones. We examined gene expression in anosteocytic medaka bone and compared it with osteocytic zebrafish bone to identify alternative mechano-sensory modes. In situ hybridization found sost and fgf23 to be expressed by osteocytes in zebrafish, whereas in medaka bone, strong expression of both genes was observed in osteoblasts. snRNA-seq analysis indicated 12 distinct transcriptional clusters in medaka and zebrafish bone, each corresponding to a specific cell type. One such cluster included a concentration of osteoblast- and osteocyte-associated genes, including signature osteocyte genes in both species. Although many were expressed at very low levels by a small proportion of nuclei, expression of irx5a, dkk3b, emilin2a, cfap70 and susd5 was higher in medaka, while fgf23, fbln7, prr15lb, sost, seema3a and ngef expression was higher in zebrafish. We conclude that osteocyte functions are likely evolutionarily transferred to osteoblasts in a model whereby osteoblasts may have developed osteocytic potential at an earlier stage in the osteogenic lineage than has previously been identified.