Deletion of the membrane complement inhibitor CD59a drives age and gender-dependent alterations to bone phenotype in mice

Abstract

Degenerative joint diseases such as osteoarthritis are characterised by aberrant region-specific bone formation and abnormal bone mineral content. A recent study suggested a role for the complement membrane attack complex in experimental models of osteoarthritis. Since CD59a is the principal regulator of the membrane attack complex in mice, we evaluated the impact of CD59a gene deletion upon maintenance of bone architecture. In vivo bone morphology analysis revealed that male CD59a-deficient mice have increased femur length and cortical bone volume, albeit with reduced bone mineral density. However, this phenomenon was not observed in female mice. Histomorphometric analysis of the trabecular bone showed increased rates of bone homeostasis, with both increased bone resorption and mineral apposition rate in CD59a-deficient male mice. When bone cells were studied in isolation, in vitro osteoclastogenesis was significantly increased in male CD59a-deficient mice, although osteoblast formation was not altered. Our data reveal, for the first time, that CD59a is a regulator of bone growth and homeostasis. CD59a ablation in male mice results in longer and wider bones, but with less density, which is likely a major contributing factor for their susceptibility to osteoarthritis. These findings increase our understanding of the role of complement regulation in degenerative arthritis.

Keywords: Ageing; Bone; CD59a; Micro-CT; Osteoclast.

Fig. 1

Bone growth is increased in male CD59a-deficient mice. Femurs were X-rayed and width measured using digital calliper. (A) Representative images of male 8–50-week-old WT and CD59a^−/− mouse femurs showing differences in bone size. Scale bar (black): 5 mm. (B) Femoral length in male WT and CD59a^−/− mice. (C) Medial–lateral femoral shaft width and (D) anterior–posterior femoral shaft width in male WT and CD59a^−/− mice. (E), (F), (G), and (H), respectively, show representative x-rays, femoral length, medial–lateral femoral shaft width, and anterior–posterior femoral shaft width in age-matched female mice. All values are mean ± SEM for six WT (black bars) and CD59a^−/− (white bars) mice per group. ^⁎P < 0.01, ^⁎⁎P < 0.001 versus WT of the same age and sex.

Fig. 2

Altered cortical bone properties in young-adult CD59a-deficient male mice. Femurs were subjected to micro-CT and cortical bone of the shaft was investigated. A 1 mm section of cortical bone from the metaphysis was analysed; starting 3 mm below a reference point within the growth plate. Trabecular bone architecture (reported in Fig. 3) was measured in a section that started 1 mm below the same reference point. (A) Cortical bone volume (cBV), cross-sectional thickness (Cs.Th) and bone mineral density (BMD) of male and female 8–50-week-old WT (black bars) and CD59a^−/− (white bars) mice are shown. All values are mean ± SEM from a minimum of six mice per group. ^⁎P < 0.05; ^⁎⁎⁎P < 0.001 versus WT of the same sex. (B) Representative images at corresponding relative positions of 8-week-old WT and CD59a^−/− mouse femur cross-sections illustrating variations in bone diameter. Scale bar (white): 1 mm.

Fig. 3

Structural changes in trabecular bone in male CD59a-deficient mice. The secondary spongiosa of femurs were analysed by micro-CT. A 1 mm section of trabecular bone from the metaphysis was analysed; starting 1 mm below a reference point within the growth plate. Cortical bone architecture (reported in Fig. 2) was measured in a section that started 3 mm below the same reference point. (A) Representative images of 8-week-old WT and CD59a^−/− mouse femurs, demonstrating marked increase in trabecular bone architecture compared to controls in male but not female CD59a^−/− mice (white boxes highlight region used for measurements). Scale bar (white), 1 mm. (B) Analysis of relative bone volume (BV/TV), trabecular thickness (Tb.Th), and trabecular number (Tb.N) is shown for 8, 20, and 50-week-old CD59a^−/− (white bars) and WT (black bars) mice. All values are mean ± SEM from a minimum of six mice per group. ^⁎⁎P < 0.01 versus WT of the same sex. (C) Representative three-dimensional reconstructions from male 8-week-old WT and CD59a^−/− mice. Scale bar (black), 1 mm.

Fig. 4

Elevated osteoclast activity and increased number of OPN-positive cells in trabecular bone in male CD59a-deficient mice. Femurs were stained with TRAP to identify osteoclasts. (A) Osteoclast surface per unit bone surface (OcS/BS) and osteoclast numbers per unit tissue area (OC.N/TA) were significantly increased in male CD59a^−/− (white bars) compared with WT (black bars) mice. All values are mean ± SEM from a minimum of six mice per group at 8 and 20 weeks of age. ^⁎P < 0.05; ^⁎⁎P < 0.01 versus WT. (B) Representative images of TRAP-positive osteoclasts (stained red and highlighted by arrow). M, marrow; Tb, trabecular bone. Scale bar (black), 100 μm. Femurs were stained for OPN by immunohistochemistry. Summary of analysis in (C) epiphysis, (D) metaphysis, and (E) cortical bone of male WT (black bars) and CD59a^−/− (white bars) mice at 8 and 20 weeks. All values are mean ± SEM from seven mice per group. ^⁎⁎P < 0.01, ^⁎⁎⁎P < 0.001 versus WT. (F) Representative images of isotype control and OPN staining (brown appearance and highlighted by yellow arrows) at 20 weeks. Megakaryocytes are shown by green arrows. Insets show higher magnification of staining. M, marrow; Tb, trabecular bone. Scale bar (black), 50 μm.

Fig. 5

Reduced osteoid and increased bone formation in male CD59a-deficient mice at 8 weeks of age. Femurs were stained with von Kossa/van Gieson to reveal osteoids. (A) Histomorphometric investigation of femoral osteoid surface over total bone surface (OS/BS) of WT (black bars) and CD59a^−/− (white bars) male mice at 8 and 20 weeks. All values are mean ± SEM from a minimum of six mice per group at 8 and 20 weeks of age. ^⁎⁎⁎P < 0.001 versus WT. (B) Representative images of osteoid staining with van Gieson (pink appearance and highlighted by arrows) at 8 weeks. M, marrow; Tb, trabecular bone. Scale bar (black), 100 μm. (C) Mineral apposition rate (MAR) and (D) bone formation rate (BFR/BS) calculated from calcein double labelling are shown in WT (black bars) and CD59a^−/− (white bars) mice. Bone formation rate is the amount of mineralised bone formed per unit of time per unit of bone surface. All values are mean ± SEM from four mice per group at 8 weeks of age. ^⁎P < 0.05 versus WT. (E) Representative images of calcein double labelling in WT and CD59a^−/− mice. Scale bar (white), 100 μm.

Fig. 6

Increased osteoclastogenesis in CD59a-deficient mice in vitro. Osteoclast differentiation assays were conducted with RANKL-induced primary BMC from male and female WT and CD59a^−/− mice at 8–10 weeks of age. (A) Representative images of TRAP-negative control samples stimulated with M-CSF only and TRAP-positive cultures stimulated with M-CSF and RANKL (multinucleated cells [MNCs], arrow). Scale bar (black), 0.25 mm. (B) Quantification of TRAP-positive MNCs produced from RANKL-induced cultures. Culture supernatants of cells grown in M-CSF only and M-CSF and RANKL from male mice were harvested and mKc concentrations were measured by ELISA in WT (black bars) and CD59a^−/− (white bars) samples. (C) Osteoblasts were generated from enriched primary osteoprogenitor cells from male mice and, after 14 days in mineralisation medium, stained for ALP and Alizarin red. Representative images are shown. (D) ALP and Alizarin red coverage of wells was quantified. All values are mean ± SEM of triplicate cultures from at least five separate mouse bone marrows for each group at 8–10 weeks of age. ^⁎P < 0.05; ^⁎⁎P < 0.01; ^⁎⁎⁎P < 0.001 versus WT cultures.