Extended Data Fig. 1: Analysis of tissue sources of regeneration in distraction.

a, Experimental timeline of distraction model. b, Illustration of the tissue response (at POD15) in the mouse model of mandibular distraction (left), fracture (middle) and acute lengthening (right). c, Putative cellular sources of bone regeneration in mandibular distraction, including periosteum, endosteum and circulating progenitors. d, Experimental scheme for detecting circulating progenitor cells in mandibular distraction. GFP mice are surgically fused to their wild-type (WT) littermates through parabiosis. Peripheral blood chimaerism is confirmed via flow cytometry. After GFP-positive cells are confirmed in the WT through FACS, mandibular distraction is performed on the WT parabiont. FSC-A, forward scatter area; FSC-H, forward scatter height. e, Upon detection of 1/1 blood chimaerism, WT parabionts undergo mandibular distraction according to the timeline outlined. f, Quantification of the haematopoietic fraction of GFP-positive cells obtained from distraction calluses (n = 4 biological replicates per time point). g, Representation of the Rainbow reporter construct at the R26 locus and the colours produced by random recombination. h, One-year tracing of mandibles under normal homeostasis (uninjured), with confocal micrographs of whole-mount periosteum one year after recombination. Clones are shown with coloured dotted outlines. The white dotted line demarcates skeletal muscle (upper left quadrant) from the periosteum. The view is a buccal-to-lingual view of the posterior periosteum overlying the body of the mandible. n = 50 clones, with 16–151 cells per clone. i, Confocal micrograph of the transverse mandible section from a Rainbow mouse at POD15 after targeted labelling of the periosteum for subsequent mandibular distraction (n = 8). Coloured outlines indicate single clones; the white dotted outline indicates mandibular bone (mb) at the distraction site. n refers to the number of animals in each independent experiment.