Abstract
Bone marrow (BM) adipose tissue (BMAT) and BM adipocytes are developmentally distinct from other fat depots and differ from white and brown adipose tissue adipocytes in their lipid composition. BMAT also negatively regulates hematopoiesis. Obesity results in BMAT expansion, which increases bone resorbing osteoclasts and is associated with reduced osteoblast bone forming potential leading to increased risk of fractures. Factors that could limit the expansion of BMAT and decrease the negative effects on bones and hematopoiesis could be useful in restoring bone and BM homeostasis.
Erythropoietin (EPO) is essential for erythroid differentiation. In animal models of diet induced obesity, EPO lowers inflammation, reduces white fat mass and promotes glucose tolerance. Here, we investigated the EPO effect on BMAT, bones and distinct BM cell populations under obesogenic conditions. Young C57BL6/J female mice (6 weeks) were placed on normal chow diet (ND) or on 60% kcal high fat diet (HFD) for 4 weeks and were treated with EPO 1200 IU/kg or saline daily for 10 more days on HFD. This course of EPO treatment increased hematocrit without affecting whole body fat mass. For bone analysis, femurs were harvested and analyzed using micro-computed tomography (micro-Ct), and hematoxylin and eosin staining to identify bone and marrow features. HFD decreased trabecular number and increased trabecular spacing, but increased the cortical bone volume, suggesting the adaptation of the bones to withstand the increased body mass in HFD fed mice. EPO treatment reduced trabecular mineral density, trabecular number and spacing, and bone volume/tissue volume under both ND and HFD conditions. However, with EPO treatment, cortical bone volume decreased only in ND fed mice but not in HFD fed mice and cortical bone mineral density increased in HFD fed mice. Consistent with these selective EPO associated changes in bone loss were decreased trabecular osteocytes and decreased trabecular lining osteoblasts in ND fed mice and analogous trends in HFD fed mice, while EPO treatment reduced osteocytes in the cortical bones and osteoblasts in the periosteal region of cortical bones in ND fed mice but not in HFD fed mice. Histology sections showed that EPO treatment markedly reduced BMAT in ND and HFD fed mice, and inhibited the increase in BMAT with HFD feeding. External fat depots did not exhibit any decrease in fat mass following the ten day EPO treatment, suggesting that BMAT is more sensitive and has a quicker response to EPO treatment in both ND and HFD fed mice.
Contrary to published reports that link EPO stimulated bone loss to increased osteoclast number and activity, we did not find that EPO increased osteoclasts detected by tartarate resistant acid phosphatase (TRAP) staining. Although femur sections showed an increase in osteoclasts with HFD feeding, EPO treatment reduced osteoclasts in HFD fed mice and had no significant effect in ND fed mice. These data suggest that bone loss, especially trabecular bone, accompanying EPO stimulated increase in hematocrit is not due to increase osteoclast activity but rather is due in part to decreased osteoblasts and osteocytes. Flow cytometry analysis of BM cell populations demonstrated that EPO treatment in ND fed mice decreased hematopoietic cells by 40% and increased erythroid precursor cells by 70%. Although EPO stimulated increase in hematocrit was comparable in ND and HFD fed mice, corresponding changes in % hematopoietic cells and % erythroid precursor cells in BM cell populations were not evident in HFD fed mice. This suggests that in obese mice, EPO treatment abrogates the suppression of hematopoiesis exerted by the expanding BMAT and results in the increased hematopoietic cell proliferation concomitant with EPO stimulated erythropoiesis.
In summary, EPO treatment decreases trabecular bone in ND and HFD fed mice, decreases cortical bone volume in ND fed mice, but increases cortical bone mineral density in HFD fed mice. Concomitant with EPO stimulated increased hematocrit is a decrease in BMAT and abrogation of BMAT expansion in diet-induced obese mice. Changes in BMAT are prior to detectable changes in fat mass in external fat depots, suggesting increased sensitivity of BMAT to EPO treatment compared with white adipose tissue. These changes exemplify EPO response in non-hematopoietic tissue of bone and BMAT that accompany bone marrow stimulated erythropoiesis.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.