Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2008 Feb;118(2):491-504.
doi: 10.1172/JCI33102.

Pharmacologic Targeting of a Stem/Progenitor Population in Vivo Is Associated With Enhanced Bone Regeneration in Mice

Free PMC article

Pharmacologic Targeting of a Stem/Progenitor Population in Vivo Is Associated With Enhanced Bone Regeneration in Mice

Siddhartha Mukherjee et al. J Clin Invest. .
Free PMC article


Drug targeting of adult stem cells has been proposed as a strategy for regenerative medicine, but very few drugs are known to target stem cell populations in vivo. Mesenchymal stem/progenitor cells (MSCs) are a multipotent population of cells that can differentiate into muscle, bone, fat, and other cell types in context-specific manners. Bortezomib (Bzb) is a clinically available proteasome inhibitor used in the treatment of multiple myeloma. Here, we show that Bzb induces MSCs to preferentially undergo osteoblastic differentiation, in part by modulation of the bone-specifying transcription factor runt-related transcription factor 2 (Runx-2) in mice. Mice implanted with MSCs showed increased ectopic ossicle and bone formation when recipients received low doses of Bzb. Furthermore, this treatment increased bone formation and rescued bone loss in a mouse model of osteoporosis. Thus, we show that a tissue-resident adult stem cell population in vivo can be pharmacologically modified to promote a regenerative function in adult animals.


Figure 1
Figure 1. Bzb treatment increases osteoblasts in vivo.
(A) Mice treated with Bzb (0.3 mg/kg i.p.; 3 times/week) showed an increase in the levels of serum osteocalcin over 3 weeks compared with control-treated (saline-treated) mice. In contrast, serum osteocalcin decreased over time in control mice. P = 0.01 by Student’s t test; n = 5 mice each. (B) Micro-CT analysis revealed an increase in trabecular bone volume in drug-treated mice. A representative reconstruction of trabecular bone is shown as 3D reconstruction (left panel, mock treated; right panel, Bzb treated) and as cross sections (left panel, control; right panel, Bzb treated). Scale bar: 1.0 mm. (C) Colony formation from Bzb-treated versus saline-treated (mock-treated) animals showed increased CFU-F, increased Ops (CFU-Alk), and decreased adipocytic colonies (CFU-Adipo). *P = 0.007, n = 8 wells for CFU-F; P < 0.01, n = 9 wells for CFU-Alk; P = 0.01, n = 12 wells for CFU-Adipo. (D) Histomorphometric analysis of Bzb-treated animals showed increased trabecular connectivity, trabecular volume occupied by bone, and trabecular number. ΧP = 0.05, trabecular connectivity; ζP = 0.02, trabecular bone volume; #P = 0.03, trabecular number/mm. n = 4 femurs. (E) Histological sections of treated animals showed increased bone with normal architecture in trabeculae (H&E-stained samples) but with increased bone volume. Original magnification, ×40. (F) Increased osteoblast number per BSpm in distal femur was observed in Bzb-treated animals. **P = 0.02; n = 3. (G) Increase in mineralization rate in animals treated with Bzb. ††P < 0.01; n = 6. (H) No significant change was observed in TRAP-stained osteoclasts, quantified by osteoclasts/μm of BSpm. P = 0.53 by Student’s t test.
Figure 2
Figure 2. Bzb increases osteoblastogenesis in MSCs in vitro.
(A) Murine MSCs are multilineage-potent fibroblastoid cells that can differentiate into osteoblasts and adipocytes; they are hematopoietic lineage negative, CD45, and CD105+. Ops are Osx positive. Upon differentiation, MSCs lose 105 staining, form von Kossa staining ECM nodules, and express high levels of BSP. (B) Isolated murine MSCs (CD105-selected cells) showed increased alkaline phosphatase activity, blue staining (left panel, control; right panel, 1 nM). Alkaline phosphatase–positive cells per well increased with 1 and 2 nM of Bzb. *P = 0.01; P = 0.01; n = 6 wells each. Original magnification, ×100. (C) CD105-selected cells exposed to osteogenic medium with or without 1.5 nM Bzb were analyzed by qPCR. BSP transcript levels and Runx-2 transcript levels (plotted against the baseline value of untreated MSCs assigned as 1.00) increased significantly in Bzb-treated samples at 30 hours (P = 0.04; ΧP = 0.03; n = 2), while preexposure to osteogenic differentiation (MSC→Og) for 8 days followed by Bzb treatment abrogated the response (white bars for 0 and 1.5 nM; P = NS). (D) Von Kossa–positive area increased with 1 and 2 nM Bzb. ζP = 0.05; #P = 0.01; n = 3 wells. Original magnification, ×100.
Figure 3
Figure 3. Osx-positive Op cells do not respond to Bzb, while Osx-negative MSCs respond to Bzb in vitro.
Ops were recovered from the compact bone of Osx-GFP mice by isolation of CD45GFP+ cells. CD45GFP+ Ops plated in vitro showed no change in collagen I clusters, while CD45GFP MSCs isolated from the same mice showed an increase (arrows). *P < 0.05; n = 4. Original magnification, ×100.
Figure 4
Figure 4. Bzb acts on hMSCs.
(A) hMSCs (CD105+CD73+CD45) were subjected to Bzb in vitro, and API was assessed after 6 days in culture. API measurements revealed an increase in activity with 0.5, 1, and 2 nM Bzb. *P = 0.001; n = 10. (B) Alizarin red staining was increased upon Bzb treatment. **P = 0.002; n = 5 fields. (C) Adipocytes (oil red O–positive cells) were decreased. Original magnification, ×100. P = 0.03; n = 4 fields.
Figure 5
Figure 5. Runx-2 modulates the responsiveness to Bzb.
(A) Steady state levels of Runx-2 protein in MSCs from WT (lanes 1, 3, and 5) or Shn3–/– mice (lanes 2, 4, and 6) were detected by Western blotting of MSCs cultured either with DMSO vehicle, Bzb, or lactacystin. In WT MSCs, p62 Runx-2 (red circle) was upregulated by Bzb treatment (lane 1 versus lane 3) and by lactacystin (lane 1 versus lane 5) to levels comparable to Runx-2 levels in Shn3–/– MSCs (lane 2). Bzb and lactacystin also increased Runx-2 levels in Shn3–/– MSCs, suggesting that Shn3-independent proteasomal activity targeting Runx-2 still remains functional in these MSCs. Bottom panel shows GAPDH used as a loading control. (B) C3HT1/2 cells were transfected with a luciferase construct containing 6 Runx-2–binding elements from the osteocalcin gene. Cotransfection of a Runx-2 expression construct (Rx-2) increased luciferase expression above baseline. Addition of either Bzb or lactacystin (third and fourth bars) further increased luciferase expression (*P = 0.0009; **P = 0.0003) compared with Runx-2–cotransfected cells without drug treatment. (C) MSCs from WT or Shn3–/– animals were plated in osteogenic medium. Von Kossa staining (brown) revealed a dose-dependent increase in osteogenic activity with Bzb treatment in WT mice, while in Shn3–/– mice, there was constitutively elevated von Kossa granule formation that was not increased by Bzb treatment. P = 0.04; n = 4 wells. Original magnification, ×100. (D) In WT embryonic mesodermal fibroblasts (MEFs), treatment with Bzb showed an increase in expression of Alp and BSP, while in Runx-2–/– cells, there was no response (upper 2 panels); however, expression of collagen I (bottom panel) was increased in both WT (ΧP = 0.02) and Runx-2–/– cells (#P = 0.03). Stimulation of WT cells to promote Op formation with exogenous BMP-2 (two bars on far right of each panel) abolished the Bzb responsiveness of all 3 genes.
Figure 6
Figure 6. Bzb increases bone formation from MSCs in vivo.
(A) Sponges embedded with cells differentiated in osteogenic medium (upper 2 panels) or loaded with undifferentiated MSCs (lower panels) were transplanted into immunocompromised mice, and recipient mice were treated with Bzb at 0.3 mg/kg i.p. for 10 doses. Bzb treatment had no observable impact on osteoblast-loaded sponges (upper 2 panels; stained with H&E). In MSC-loaded sponges (lower panels), osteoid and bone were increased upon Bzb treatment (yellow arrows), shown in low power (original magnification, ×40) or in high power (HP; original magnification, ×100). Bzb treatment also increased matrix-depositing cells, seen as cells (white arrows) adjacent to xylenol orange (XO) fluorescence (red) with DAPI nuclear counterstain. Increased bone upon Bzb treatment was also observed with trichrome stain and with alizarin red stain on control- and Bzb-treated sponges. Graph shows quantification of alizarin red–stained fraction of tissue in control- and Bzb-treated sponges. *P = 0.007; n = 3. (B) Mice were implanted intrafemorally with GFP+ MSCs and then treated with control or Bzb i.p. Upper panels show femurs containing implanted GFP+ MSCs after treatment. Femurs were stained with alkaline phosphatase (purple). Increased ectopic bone is seen in Bzb-treated bones (arrows). Lower 4 panels show bright-field and epifluorescence images of the same section at higher magnification. Original magnification, ×20 (upper 2 panels); ×100 (lower 4 panels). In saline-treated animals, most GFP+ cells remained Alkaline phosphatase negative (long arrow). Occasional GFP+Alkaline phosphatase–positive cells were seen lining the endosteal surface (short arrows). In contrast, in Bzb-treated animals, increased ectopic bone was observed. Ectopic alkaline phosphatase–positive cells derived from GFP+ MSCs (green) were frequently observed (lower 2 right panels, white arrows) not only next to the bone, but also in the marrow space. (C) In saline-treated animals, GFP+ cells retained fibroblastoid appearance and CD105 positivity (GFP, green; CD105, red; double labeled, yellow; double-labeled cells are shown with white arrows in panel). In Bzb-treated animals, most cells lost their CD105 positivity (green). Images are from a single optical slice, using a confocal microscope. Original magnification, ×100.
Figure 7
Figure 7. Bzb rescues osteoporosis in ovariectomized mice.
FVB/N female mice were ovariectomized at week 9 and treated with saline or 0.3 mg/kg Bzb for 6 weeks (18 doses) starting at week 11.5. (A) Representative micro-CT analysis of cross sections of trabecular bone are shown for mock-treated, ovariectomized, and ovariectomized/Bzb-treated mice. Scale bar: 1.0 mm. (B) Baseline distal femur BMD increased upon Bzb treatment in both mock-treated and ovariectomized mice. *P = 0.006; **P = 0.03; n = 10 femurs/group. Quantitative analysis by micro-CT scanning revealed that ovariectomized animals had decreases in trabecular bone volume fraction and that Bzb treatment partly rescued this decrease. P = 0.001; P = 0.01; n = 10 femurs/group. Trabecular number/mm decreased with ovariectomy but was rescued by Bzb treatment. ζP = 0.01; ΧP = 0.002; n = 10 femurs/group. (C) BFR per BSpm (BFR/B pm) was increased in ovary-intact animals by Bzb treatment (#P = 0.007; n = 7 femurs) and also increased in ovariectomized animals (††P = 0.05; n = 7 femurs). (D) Mineral apposition rate was increased in ovary-intact animals (‡‡P = 0.01; n = 7 femurs) with Bzb treatment. In ovariectomized animals, there was increased mineral apposition, but the effect was not significant. P = 0.1.

Comment in

  • Bone building with bortezomib.
    Roodman GD. Roodman GD. Version 2. J Clin Invest. 2008 Feb;118(2):462-4. doi: 10.1172/JCI34734. J Clin Invest. 2008. PMID: 18219395 Free PMC article.

Similar articles

See all similar articles

Cited by 73 articles

See all "Cited by" articles

Publication types