Fig 2: RUNX2 increases the accumulation of MDA231 cells as micrometastases in bone marrow. (A) Experimental schedule for the bone colonization of MDA231‐derived cells in SCID mice. An osteogenic premetastatic niche (PMN) was established by injecting MDA231 CDH11high/ITGA5high extracellular vesicles into SCID mice via the tail vein for 3 weeks (2 doses/week). MDA231 RUNX2‐OE cells and control cells were injected into the mice via the left ventricle. (B) H&E staining images demonstrating osteoblasts (indicated by red triangles) in homeostatic bone (HB) and the PMN. (C) X‐ray images showing the bone mass in the HB and the PMN microenvironments. (D) Bar charts quantifying the osteoblast number as the ratio of osteoblast counts to bone perimeter in /mm (N.Ob/B.Pm) and bone mass as the bone volume fraction (BV/TV). (E) Western blot analysis demonstrating increased RUNX2 protein levels in MDA231 RUNX2‐OE cells compared with those in control cells. (F) Representative X‐ray images, pan‐cytokeratin (pan‐CK) immunohistochemical staining images, and TRAP staining images showing osteolytic lesions, tumor cell distribution, and activated osteoclasts, respectively. (G) Pie charts depicting the incidence of DTCs, micrometastases (micromets), and osteolytic lesions formed by control cells and RUNX2‐OE cells within HB and the PMN. The scale bars in the inset images indicate 20 µm. (H) Bar charts quantifying the tumor surface, erosion surface, and the number of TRAP+ osteoclasts normalized to the total bone surface (N.Oc/BS in mm2). (I) Bar charts illustrating the abundance and size of micrometastases in the bone marrow of mice without detectable bone lesions. (J) Representative KI67 fluorescence immunohistochemical staining images. Pan‐CK was used to label the tumor cells, while DAPI was used to stain the nuclei. (K) Bar chart illustrating the reduced numbers of KI67+ tumor cells in micrometastases compared with tumor cells within bone colonization. The data are displayed as the means ± SDs. n.s., not significant; * p <0.05, ** p <0.01, *** p <0.001 compared with the corresponding controls, as determined by Student's t‐test.

Fig 3: RUNX2 promotes the colonization of luminal‐like MCF7 cells within a highly mineralized osteogenic microenvironment induced by E2. (A) Experimental schedule for bone colonization of MCF7‐derived cells in SCID mice. MCF7 RUNX2‐OE cells and control cells were injected into the mice via the left ventricle. E2 (2 mg/kg) was administered via subcutaneous injection weekly to support MCF7 tumor growth, starting 1 week before tumor cell injection. (B) Representative micro‐CT and H&E staining images showing increased bone mass and a reduced marrow cavity in SCID mice administered E2. Bar graph illustrating the bone volume fraction (BV/TV) in mice treated with E2 compared with in control mice. (C) Western blot analysis of RUNX2 protein levels in MCF7 RUNX2‐OE cells and control cells. (D) Representative BLI images showing the systemic distribution of tumor cells following left ventricular inoculation on day 0. X‐ray and H&E staining images displaying bone lesions in mice after sacrifice. TRAP staining images indicating osteoclast activity. (E) Representative micro‐CT images displaying bone lesions in SCID mice on day 38. The osteolytic lesions are marked by asterisks and arrows. (F) Pie charts illustrating the incidence of osteolytic bone colonization in mice injected with MCF7 RUNX2‐OE cells compared with those in mice injected with control cells. (G) Bar charts showing the tumor surface, erosion surface, and TRAP+ surface normalized to the bone surface. The data are displayed as the means ± SDs. ** p <0.01 and *** p <0.001 compared to the control group, as determined by Student's t‐test.

Fig 4: Abnormal activation of the osteogenic microenvironment reactivates quiescent basal‐like cancer cells. (A) Experimental schedule for PTH‐reactivated bone colonization. MDA231 RUNX2‐OE cells were injected into SCID mice via the left ventricle. The mice were then treated with 100 µg/kg PTH for 10 consecutive days. Mice inoculated with PBS served as controls. (B) Representative BLI, X‐ray, H&E staining, immunohistochemical staining for pan‐CK, and TRAP staining images demonstrating the progression of bone lesions and the activity of osteoclasts. (C) Pie charts displaying the incidence of DTCs/micrometastases and osteolytic lesions in mice treated with PTH and in control mice. (D) Bar charts showing the incidence of osteolytic bone colonization, tumor surface, and erosion surface in PTH‐treated mice and control mice. (E) Experimental schedule involving the injection of MDA231 RUNX2‐OE cells through the left ventricle, followed by treatment with either a single dose of low‐dose estradiol cypionate (E2, 0.3 mg/kg) or weekly high‐dose E2 (2 mg/kg) for 3 weeks via subcutaneous injection, starting 7 days post‐inoculation of cancer cells. Mice injected with an equal volume of the solvent corn oil served as controls. (F) Representative BLI, X‐ray, H&E staining, immunohistochemical staining for pan‐CK, and TRAP staining images illustrating the progression of bone lesions and the activity of osteoclasts. (G) Pie charts depicting the incidence of DTCs/micrometastases and osteolytic lesions in mice treated with various doses of E2. (H) Bar charts illustrating the incidence of osteolytic bone colonization, tumor surface, and erosion surface in mice subjected to low‐dose E2, high‐dose E2, and control conditions. The scale bars in the inset images indicate 20 µm. The data are presented as the means ± SDs. * p <0.05 and ** p <0.01 compared with the control group, as determined by Fisher's exact probability method or Student's t test.

Fig 5: The differentiation status of the osteoblastic microenvironment affects the colonization of MDA231 RUNX2‐OE cells. (A) Experimental diagram illustrating the bone colonization of cancer cells in mice following parathyroid hormone (PTH) treatment. SCID mice received daily intraperitoneal injections of 100 µg/kg PTH for 0, 5, 10, 15, or 20 days, followed by the evaluation of the differentiation of osteoblastic microenvironment or the injection of MDA231 RUNX2‐OE cells via the left ventricle. Tetracycline hydrochloride (30 mg/kg) was administered on days 13 and 14, and calcein (6 mg/kg) was injected on days 3 and 4 prior to sacrifice for evaluation of the osteoblastic microenvironment. (B) Representative H&E staining, calcein fluorescence, Masson's staining, calcein/tetracycline fluorescence space, Von Kossa staining, TRAP staining, and micro‐CT images. The red triangles in the H&E‐stained images indicate osteoblasts. (C) Strip plot qualifying the osteoblast number as N.Ob/B.Pm. (D) Plot of the percentage area of calcein fluorescence (Ca.A) normalized to the bone surface area. (E) Strip plot depicting the collagen volume fraction (CVF) quantified by Masson's staining. (F) Strip plot displaying the bone mineralization rate (BMR) calculated from the calcein/tetracycline fluorescence space. (G) Strip plot illustrating the area of Von Kossa staining (VK.A) normalized to the bone surface area. (H) Plot of the percent TRAP+ surface area normalized to the bone surface area (Oc.S/BS). (I) Plot of marrow volume (Ma.V) reflecting cortical bone thickness. (J) The bone volume fraction (BV/TV) and trabecular thickness (Tb.Th) were determined from micro‐CT images. (K) Characterization of the stages of the osteoblastic microenvironment in mice induced by PTH for various durations. (L) Chart showing the colonization of cancer cells in the bone (red bars), lungs (blue bars), lymph nodes (brown bars), and brain (green bars) in mice treated with PTH. (M) Representative bioluminescence (BLI) images demonstrating the systemic distribution of cancer cells in SCID mice on day 0 and the multiorgan colonization observed on day 42. Additional representative X‐ray, H&E, immunohistochemical staining for pan‐CK, and TRAP staining images that indicate bone lesions, tumor cell distribution, and osteoclast activity at the time of sacrifice are shown. (N) Pie charts depicting the incidence of DTCs, micrometastases, and osteolytic lesions in mice treated with PTH for 10 days (PTH10) and in control mice (PTH0). (O) Bar chart showing the tumor surface normalized to the bone surface. (P) Bar charts indicating the erosion surface area and TRAP+ surface normalized to the bone surface. The data are presented as the means ± SDs. * p < 0.05, ** p < 0.01, and *** p < 0.001 compared with the PTH0 group, as determined by Student's t test.