Erythropoietin modulates bone marrow stromal cell differentiation

Abstract

Erythropoietin is essential for bone marrow erythropoiesis and erythropoietin receptor on non-erythroid cells including bone marrow stromal cells suggests systemic effects of erythropoietin. Tg6 mice with chronic erythropoietin overexpression have a high hematocrit, reduced trabecular and cortical bone and bone marrow adipocytes, and decreased bone morphogenic protein 2 driven ectopic bone and adipocyte formation. Erythropoietin treatment (1 200 IU·kg^-1) for 10 days similarly exhibit increased hematocrit, reduced bone and bone marrow adipocytes without increased osteoclasts, and reduced bone morphogenic protein signaling in the bone marrow. Interestingly, endogenous erythropoietin is required for normal differentiation of bone marrow stromal cells to osteoblasts and bone marrow adipocytes. ΔEpoR_E mice with erythroid restricted erythropoietin receptor exhibit reduced trabecular bone, increased bone marrow adipocytes, and decreased bone morphogenic protein 2 ectopic bone formation. Erythropoietin treated ΔEpoR_E mice achieved hematocrit similar to wild-type mice without reduced bone, suggesting that bone reduction with erythropoietin treatment is associated with non-erythropoietic erythropoietin response. Bone marrow stromal cells from wild-type, Tg6, and ΔEpoR_E-mice were transplanted into immunodeficient mice to assess development into a bone/marrow organ. Like endogenous bone formation, Tg6 bone marrow cells exhibited reduced differentiation to bone and adipocytes indicating that high erythropoietin inhibits osteogenesis and adipogenesis, while ΔEpoR_E bone marrow cells formed ectopic bones with reduced trabecular regions and increased adipocytes, indicating that loss of erythropoietin signaling favors adipogenesis at the expense of osteogenesis. In summary, endogenous erythropoietin signaling regulates bone marrow stromal cell fate and aberrant erythropoietin levels result in their impaired differentiation.

Keywords: Bone; Fat metabolism.

Fig. 1

Bone features of Tg6 mice overexpressing EPO. a Micro-CT generated 3D images of trabecular (top) and cortical bone (bottom) of the femurs of 11-week-old wild-type (wt; left) and Tg6 mice (right). b–f Quantification of trabecular parameters including trabecular bone mineral density (BMD), bone volume/total volume (BV/TV), number (N), thickness (Th), and spacing (Sp) of trabeculae. g–j Quantification of cortical BMD, BV, thickness, and moment of inertia (MOI) (n = 4/group, *P < 0.05, **P < 0.01, ***P < 0.001)

Fig. 2

Bone features of ΔEpoR_E mice with lack of EPOR signaling in non-erythroid cells. a Micro-CT generated images of trabecular bone of 11-week-old male wild-type (wt) and ΔEpoR_Emice. b–c Femurs of 11-week-old male and female ΔEpoR_E mice and age matched wt mice were analyzed using micro-CT. Quantitation of trabecular BMD, BV/TV, number, and spacing of trabecular bone of male (b) and female ΔEpoR_E mice (c) and controls (n = 4/group). d–h EPO induced changes in bones of 8-week-old wild-type (wt) and ΔEpoR_E mice with Epor deletion in non-erythroid cells. d 3D images of trabecular bones of wt and ΔEpoR_E mice receiving 1 200 IU·kg^–1 of EPO for 10 days. Quantification of trabecular parameters including trabecular BMD (e), BV/TV (f), trabecular number (g), and trabecular spacing (h). (n = 5/group, *P < 0.05, **P < 0.01, ***P < 0.001)

Fig. 3

Histology analysis by H&E staining of femurs sections of Tg6- and ΔEpoR_E-mice. a Femurs of 11-week-old male and female Tg6 mice and controls. Red arrows indicate marrow adipocytes, which appear as white circles in the marrow, pink colored regions are bones and blue regions are the marrow (BM). White arrows in Tg6 histology images show sinusoids (n = 4/group). b Femurs of 11-week-old male and female ΔEpoR_E mice and controls (n = 4/group). c Femur sections of 8-week-old wt and ΔEpoR_E mice receiving 1 200 IU·kg^–1 of EPO for ten days. d Number of marrow adipocytes in femurs of male and female Tg6 mice (n = 5/group). e Number of marrow adipocytes in the femurs of wild-type and ΔEpoR_E mice. f Number of adipocytes in the femurs of wild-type and ΔEpoR_Emice receiving 1 200 IU·kg^–1 of EPO or saline treatment for ten days. *P < 0.05, **P < 0.01, ***P < 0.001

Fig. 4

Osteogenic cultures of Tg6- and ΔEpoR_E-mice. a Calvarial osteoprogenitors isolated from wt and Tg6-mice were cultured in osteogenic medium and stained for ALP expression on day 7 and mineralization deposits by alizarin red staining on day 21. b Relative expression of Alp, Runx2 and Osterix in osteogenic cultures on day 7 determined by real-time PCR. c Cell proliferation of calvarial osteoblasts was determined using MTT assay at 24, 48, and 72 h of culture in osteogenic medium and data expressed as absorbance values. d ALP activity was measured using a colorimetric assay and expressed as absorbance normalized to the cell number determined by MTT assay. e Calvarial osteoprogenitors isolated from wt and ΔEpoR_E-mice cultured in osteogenic medium and stained for ALP (day 7) and alizarin red (day 21). f Relative expression of Alp, Runx2 and Osterix mRNA levels from primary calvarial osteogenic cells (day 7) determined by real-time PCR. g–h Calvarial osteogenic cells from wt and ΔEpoR_E-mice treated with 5 U and 10 U of EPO and measured for their cell proliferation rate by MTT assay (g) and ALP activity of the cells were measured using colorimetric assays, absorbance was normalized to the cell number (h). (n = 4/group, **P < 0.01, ***P < 0.001)

Fig. 5

Osteoclasts in Tg6- and ΔEpoR_E mice. a Representative images of TRAP stained osteoclasts in the femur sections of 11-week-old wt and Tg6-mice. Arrows indicate purple stained osteoclasts along the surface of trabecular bones. b Images of in vitro osteoclast differentiation of bone marrow cells of wild-type (wt) and Tg6-mice using M-CSF and RANKL. Osteoclasts appear after 4–5 days and were stained for TRAP. c TRAP-stained sections of femurs of 11-week-old wild-type (wt) and ΔEpoR_E-mice showing osteoclasts. Arrows indicate purple stained osteoclasts along the surface of trabecular bone. d In vitro osteoclast differentiation from the bone marrow of wt and ΔEpoR_E-mice identified by TRAP staining. e–h Number of osteoclasts on the bone surface of wt and Tg6-mice (e), number of osteoclasts in the bone marrow cultures of wt and Tg6-mice (f), number of osteoclasts on the bone surface of wt and ΔEpoR_E-mice (g), and number of osteoclasts in the bone marrow cultures of wt and ΔEpoR_E-mice (h) analyzed using ImageJ. i TRAP stained femur sections of 8-week-old wt and ΔEpoR_E-mice receiving 1200 U·kg^–1 of EPO or saline for 10 days. Arrows indicate purple stained osteoclasts along the trabecular bone surface. j Enumeration of osteoclasts in the wt and ΔEpoR_E-mice receiving EPO or saline. k Relative expression of Cathepsin K levels in the whole bone marrow determined by real-time PCR. (n = 5/group, *P < 0.05, **P < 0.01, ***P < 0.001)

Fig. 6

Elevated EPO and lack of EPOR signaling inhibit BMP2-induced ectopic bone formation. a Micro-CT images of ectopic bone formed by BMP2 in collagen scaffolds transplanted in wild-type (wt) and Tg6-mice 4 weeks post transplantation. Shown are ossicles formed in wt mice (top) and in Tg6-mice (bottom), and inner trabecular bone (right) and cortical bone (left). b Micro-CT images of ectopic bone formed by BMP2 in collagen scaffolds transplanted in wild-type (wt) and ΔEpoR_E-mice 4 weeks post transplantation. Shown are ossicles formed in wt mice (top) and in ΔEpoR_E-mice (bottom), and inner trabecular bone (right) and cortical bone (left). c H&E staining of decalcified ossicles formed in the wt and Tg6-mice, showing bone as pink colored regions, adipocytes as white circular regions and marrow as blue regions. d H&E stained images of ossicles formed in wt and ΔEpoR_E-mice bone, marrow and adipocytes. e, f Micro-CT quantification of trabecular (e) and cortical (f) volume of the ectopic bone formed in wt and Tg6-mice. g Quantification of marrow adipocytes in the wt and Tg6-ossicle sections. h Space occupied by marrow (excluding adipocytes and bones). i Number of TRAP stained osteoclasts in the wt and Tg6-ossicles. j, k Micro-CT quantification of trabecular (j) and cortical (k) volume of the BMP2-induced ectopic bone formed in wt and ΔEpoR_E-mice. l Quantification of marrow adipocytes in the wt and ΔEpoR_E-ossicle sections. m Space occupied by marrow (excluding adipocytes and bone). n Number of TRAP stained osteoclasts in the wt and ΔEpoR_E-ossicles. Quantification of adipocytes, marrow space and osteoclast numbers were performed using ImageJ analysis. (n = 4–5/group, *P < 0.05, **P < 0.01)

Fig. 7

In vivo bone formation potential of BMSCs from Tg6- and ΔEpoR_E-mice transplanted in immunocompromised mice. a Micro-CT images of ectopic bone formed by the differentiation of BMSCs from wild-type (wt) and Tg6-mice 4 weeks post transplantation. Shown are wt BMSC ossicles (top) and Tg6-BMSC ossicles (bottom), and inner trabecular bone (right) and cortical bone (left). b Micro-CT images of ectopic bone formed by BMSCs from wild-type (wt) and ΔEpoR_E-mice. Shown are wt BMSC ossicles (top) and ΔEpoR_E-BMSC ossicles (bottom), and inner trabecular bone (right) and cortical bone (left). c, d H&E staining of wt, Tg6 (c), and ΔEpoR_E (d) BMSC ossicles showing cortical bone (CB), trabecular bone (TB), adipocytes (A), gelfoam (G), and bone marrow (BM). e, f Quantification of trabecular (e) and cortical bone volume (f) of wt and Tg6-BMSC ossicles. g, h Area of adipocytes (g) and bone marrow (h) in wt and Tg6–BMSC ossicles quantified using ImageJ software analysis. i Quantification of the number of osteoclasts present in the ossicles developed from the wt and Tg6-BMSCs. j, k Micro-CT quantification of trabecular (j) and cortical bone volume (k) of ossicles developed from the differentiation of wt and ΔEpoR_E-BMSCs. l, m Area of adipocytes (l) and bone marrow (m) inside the wt and ΔEpoR_E-ossicles. n Quantification of osteoclasts present on the wt and ΔEpoR_E-BMSC ossicles. (n = 4/group, **P < 0.01, ***P < 0.001)

Fig. 8

Schematic for erythropoietin regulation of BMSC differentiation. a EPO-EPOR signaling in BMSCs is required for maintaining the balance between osteogenic and adipogenic differentiation in the marrow. b Elevated EPO signaling in Tg6-mice inhibits both osteogenesis and adipogenesis. c Absence of endogenous EPO signaling in BMSCs in ΔEpoR_E-mice reduces osteogenesis and favors adipogenesis resulting in fatty marrow. d EPO treatment in mice inhibits both osteogenesis and adipogenesis