C3a-C3aR signaling is a novel modulator of skeletal homeostasis

Abstract

Osteoimmune studies have identified complement signaling as an important regulator of the skeleton. Specifically, complement anaphylatoxin receptors (i.e., C3aR, C5aR) are expressed on osteoblasts and osteoclasts, implying that C3a and/or C5a may be candidate mediators of skeletal homeostasis. The study aimed to determine how complement signaling influences bone modeling/remodeling in the young skeleton. Female C57BL/6J C3aR^-/-C5aR^-/- vs. wildtype and C3aR^-/- vs. wildtype mice were examined at age 10 weeks. Trabecular and cortical bone parameters were analyzed by micro-CT. In situ osteoblast and osteoclast outcomes were determined by histomorphometry. Osteoblast and osteoclast precursors were assessed in vitro. C3aR^-/-C5aR^-/- mice displayed an increased trabecular bone phenotype at age 10 weeks. In vitro studies revealed C3aR^-/-C5aR^-/- vs. wildtype cultures had less bone-resorbing osteoclasts and increased bone-forming osteoblasts, which were validated in vivo. To determine whether C3aR alone was critical for the enhanced skeletal outcomes, wildtype vs. C3aR^-/- mice were evaluated for osseous tissue outcomes. Paralleling skeletal findings in C3aR^-/-C5aR^-/- mice, C3aR^-/- vs. wildtype mice had an enhanced trabecular bone volume fraction, which was attributed to increased trabecular number. There was elevated osteoblast activity and suppressed osteoclastic cells in C3aR^-/- vs. wildtype mice. Furthermore, primary osteoblasts derived from wildtype mice were stimulated with exogenous C3a, which more profoundly upregulated C3ar1 and the pro-osteoclastic chemokine Cxcl1. This study introduces the C3a/C3aR signaling axis as a novel regulator of the young skeleton.

Keywords: Complement; Osteoblast; Osteoclast; Osteoimmunology.

Fig. 1

Knockdown of complement C3aR and C5aR enhances bone morphology. Wildtype (WT) and C3aR^−/-C5aR^−/− mice were euthanized, and femurs were analyzed. (A-G) WT vs. C3aR^−/-C5aR^−/− micro-CT trabecular analysis, (n = 4–5 mice/group). A. Representative micro-CT images of trabecular bone at the distal femur. B. Trabecular (Tb) bone mineral density (BMD). C. Trabecular bone volume fraction (BV/TV). D. Connectivity density (Conn.D). E. Trabecular number (Tb.N). F. Trabecular thickness (Tb.Th). G. Trabecular Separation (Tb.Sp). (H-K) WT vs. C3aR^−/-C5aR^−/− micro-CT cortical analysis, (n = 4–5 mice/group). H. Representative micro-CT images of femur mid-diaphysis. I. Cortical BMD. J. Cortical bone area fraction (Ct.Ar/T.Ar). K. Cortical thickness (Ct.Th). Data are expressed as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001.

Fig. 2

Elevated osteoblast differentiation and function in C3aR^−/-C5aR^−/− vs. wildtype cultures. (A–D) Wildtype (WT) and C3aR^−/-C5aR^−/− mice were euthanized, bone marrow harvested, and BMSCs were isolated for in vitro assays. A. BMSC expansion over time in culture. B. BMSC differentiation potential assay; BMSCs were cultured for 4 days and harvested pre-confluent for qRT-PCR analysis. mRNA markers for osteoblastic (Runx2, Sp7), chondrogenic (Col2a1), and adipogenic (Pparg) potential were evaluated, and relative quantification of mRNA was performed via the comparative CT method (2^-ΔΔCT). C. BMSC osteogenic potential assay: BMSCs stimulated with osteogenic media for 5 days were isolated for qRT-PCR analysis of Bglap (Ocn) mRNA. D. Representative images of day 11 mineralization cultures stained by the von Kossa method. Cultures (n = 3–4 mice/group) were carried out in duplicate. (E-F) WT and C3aR^−/-C5aR^−/− mice were euthanized, and tibiae were processed for paraffin-embedded in situ immunofluorescent staining for osterix (OSX+) bone-lining osteoblasts (n = 4 mice/group). E. Representative images of OSX-stained proximal tibia immunofluorescence. F. Number of osteoblasts per bone perimeter (N.Ob/B.Pm). G. Serum was isolated from whole blood (n = 4 mice/group); ELISA analysis of P1NP levels. Data are expressed as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001.

Fig. 3

Suppressed osteoclastogenesis in C3aR^−/-C5aR^−/− vs. wildtype mice. (A-E) Wildtype (WT) and C3aR^−/-C5aR^−/− mice were euthanized, and whole marrow cells were isolated and sorted for CD11b⁻ cells. CD11b⁻ cells were treated with CSF1 alone or CSF1 + RANKL for 5 days (n = 5–7 mice/group). A. Representative images of TRAP-stained cultures. B. Number of osteoclasts (N.Oc/Well) in day 5 cultures. C. Osteoclast size (Oc.Ar/Oc). D. Number of nuclei per osteoclast (N.Nc/Oc). E. qRT-PCR gene expression studies were conducted in CD11b^neg OCP cultures on day 4 to detect Dcstamp mRNA transcription level alterations in RANKL-stimulated osteoclast differentiation. Relative quantification of mRNA was performed via 2^-ΔΔCT; data expressed as treatment (CSF1 and RANKL) fold change relative to control (CSF1). (F-I) WT and C3aR^−/-C5aR^−/− mice were euthanized, and tibiae were processed for paraffin-embedded TRAP+ staining for osteoclasts in vivo (n = 4 mice/group). F. Representative images of TRAP-stained proximal tibia histology. G. Number of osteoclasts per bone perimeter (N.Oc/B.Pm). H. Osteoclast size (Oc.Ar/Oc). I. Osteoclast perimeter per bone perimeter (Oc.Pm/B.Pm). Data are expressed as mean ± SEM, *p < 0.05, **p < 0.01.

Fig. 4

C3aR^−/− vs. wildtype mice have a superior trabecular bone phenotype. Wildtype (WT) and C3aR^−/− mice were euthanized, and femurs were analyzed. (A-G) WT vs. C3aR^−/− micro-CT trabecular analysis, (n = 4–5 mice/group). A. Representative micro-CT images of trabecular bone at the distal femur. B. Trabecular (Tb) bone mineral density (BMD). C. Trabecular bone volume fraction (BV/TV). D. Connectivity density (Conn.D). E. Trabecular number (Tb.N). F. Trabecular thickness (Tb.Th). G. Trabecular Separation (Tb.Sp). (H-K) WT vs. C3aR^−/− micro-CT cortical analysis, (n = 4–5 mice/group). H. Representative micro-CT images of femur mid-diaphysis. I. Cortical BMD. J. Cortical bone area fraction (Ct.Ar/T.Ar). K. Cortical thickness (Ct.Th). Data are expressed as mean ± SEM, *p < 0.05, **p < 0.01.

Fig. 5

Enhanced osteoblastogenesis in C3aR^−/− vs. wildtype animals. (A–D) Wildtype (WT) and C3aR^−/− mice were euthanized, bone marrow harvested, and BMSCs were isolated for in vitro assays. A. BMSC expansion over time in culture. B. BMSC differentiation potential assay; BMSCs were cultured for 4 days and harvested pre-confluent for qRT-PCR analysis. mRNA markers for osteoblastic (Runx2, Sp7), chondrogenic (Col2a1), and adipogenic (Pparg) potential were evaluated, and relative quantification of mRNA was performed via the comparative CT method (2^-ΔΔCT). C. BMSC osteogenic potential assay: BMSCs stimulated with osteogenic media for 5 days were isolated for qRT-PCR analysis of Bglap (Ocn) mRNA. D. Representative images of day 11 mineralization cultures stained by the von Kossa method. Cultures (n = 5 mice/group) were carried out in duplicate. (E-F) WT and C3aR^−/− mice were euthanized, and tibiae were processed for paraffin-embedded in situ immunofluorescent staining for osterix (OSX+) bone-lining osteoblasts (n = 5 mice/group). E. Representative images of OSX-stained proximal tibia immunofluorescence. F. Number of osteoblasts per bone perimeter (N.Ob/B.Pm). G. Serum was isolated from whole blood (n = 5 mice/group); ELISA analysis of P1NP levels. Data are expressed as mean ± SEM, *p < 0.05, **p < 0.01, ***p < 0.001.

Fig. 6

C3aR^−/− vs. wildtype mice show a decreased osteoclast phenotype. (A-E) Wildtype (WT) and C3aR^−/− mice were euthanized, and whole marrow cells were isolated and sorted for CD11b⁻ cells. CD11b⁻ cells were treated with CSF1 alone or CSF1 + RANKL for 5 days (n = 5 mice/group). A. Representative images of TRAP-stained cultures. B. Number of osteoclasts per well (N.Oc/Well) in day 5 cultures. C. Osteoclast size in day 5 cultures (Oc.Ar/Oc). D. Number of nuclei per osteoclast (N.Nc/Oc) in day 5 cultures. E. qRT-PCR gene expression studies were conducted in CD11b^neg OCP cultures on day 4 to detect Dcstamp mRNA transcription level alterations in RANKL-stimulated osteoclast differentiation. Relative quantification of mRNA was performed via 2^-ΔΔCT; data expressed as treatment (CSF1 and RANKL) fold change relative to control (CSF1). (F-I) WT and C3aR^−/− mice were euthanized, and tibiae were processed for paraffin-embedded TRAP+ staining for osteoclasts in vivo (n = 4 mice/group). F. Representative images of TRAP-stained proximal tibia histology. G. Number of osteoclasts per bone perimeter (N.Oc/B.Pm). H. Osteoclast size (Oc.Ar/Oc). I. Osteoclast perimeter per bone perimeter (Oc.Pm/B.Pm). Data are expressed as mean ± SEM. *p < 0.05, ***p < 0.001.

Fig. 7

Exogenous C3a induces C3ar1 and osteoclastic genes Cxcl1 and Ccl2 in osteoblasts in vitro. Wildtype bone marrow stromal cells treated with osteogenic media were stimulated with or without C3a for 2 h; cells & supernatants were collected for qRT-PCR and ELISA analyses. A. qRT-PCR analysis of C3ar1 mRNA in osteoblasts stimulated with varying concentrations (0-20 ng) of C3a; n = 4 mice/group. B. qRT-PCR analysis of Cxcl1 and Ccl2 mRNA in osteoblasts stimulated with Ctrl or 20 ng C3a; n = 4 mice/group. mRNA quantification of data via the 2^-ΔΔCT method. C. ELISA analysis of C3a in cell supernatants derived from osteoblasts stimulated with Ctrl or 20 ng C3a; n = 4 mice/group. Data are expressed as mean ± SEM. *p < 0.05.