Fig. 4: Spatial analysis of EPHA2- and EFNA1-expressing cells.

a EPHA2 mRNA (pseudocolored red) is shown together with RUNX2 protein (pseudocolored white) and bone matrix visualised by autofluorescence (pseudocolored cyan) in cleared human bone. A maximum intensity projection of the 3D scan is shown. b EPHA2- and RUNX2-double-positive cells are represented as red spheres, with sphere size indicating the distance to the frontal plane. Maximum intensity projection of the segmented 3D scan is shown. c Optical section corresponding to panel (a). YOPRO-stained nuclei (pseudocolored green) were omitted from panel (a) for visual clarity. d Quantification of EPHA2- and RUNX2-positive cells located within 20 μm of the bone surface. Data represent mean ± SE from two femoral heads (two patients), obtained from patients undergoing femoral head replacement. e, f Optical section showing RUNX2-positive cells along the bone surface (e), with additional channels overlaid for EFNA1 mRNA-expressing cells (pseudocolored yellow) and CD31-positive blood vessels (pseudocolored magenta), highlighting their spatial proximity (f). g Distance from EFNA1-positive cells to RUNX2-positive cells on the bone surface. EFNA1-positive cells were stratified into two groups: those associated with blood vessels (within 20 μm, as segmented in Fig. 3f), termed Efna1-EndC (endothelial cells), and those located beyond 20 μm from both blood vessels and the bone surface, termed Efna1-BmC (bone marrow cells). n = 13,699 cells for Efna1-EndC and n = 6945 cells for Efna1-BmC, pooled from two patients within each experimental group. Data are presented as mean ± SD. Statistical analysis was performed using a two-tailed Mann-Whitney test. h Percentage of Efna1-EndC and Efna1-BmC cells located within 20 μm of RUNX2-positive cells detected along the bone surface. Data derived from panel (g). Source data are provided as a Source Data file.