Normal inflammation and regeneration of muscle following injury require osteopontin from both muscle and non-muscle cells

BACKGROUND: Osteopontin is secreted by skeletal muscle myoblasts and macrophages, and its expression is upregulated in muscle following injury. Osteopontin is present in many different structural forms, which vary in their expression patterns and effects on cells. Using a whole muscle autograft model of muscle injury in mice, we have previously shown that inflammation and regeneration of muscle following injury are delayed by the absence of osteopontin. The current study was undertaken to determine whether muscle or non-muscle cells provide the source of osteopontin required for its role in muscle regeneration. METHODS: The extensor digitorum longus muscle of wild-type and osteopontin-null mice was removed and returned to its bed in the same animal (autograft) or placed in the corresponding location in an animal of the opposite genotype (allograft). Grafts were harvested at various times after surgery and analysed by histology, flow cytometry and quantitative polymerase chain reaction. Data were analysed using one- or two-way ANOVA or Kruskal-Wallis test. RESULTS: Immunohistochemistry confirmed that osteopontin was expressed by macrophages in osteopontin-null muscle allografts in wild-type hosts and by myoblasts in wild-type allografts in osteopontin-null hosts. The decrease in muscle fibre number observed in wild-type autografts following grafting and the subsequent appearance of regenerating fibres were delayed in both types of allografts to a similar extent as in osteopontin-null autografts. Infiltration of neutrophils, macrophages and M1 and M2 macrophage subtypes were also delayed by lack of osteopontin in the muscle and/or host. While the proportion of macrophages showing the M1 phenotype was not affected, the proportion showing the M2 phenotype was decreased by osteopontin deficiency. Expression of tumour necrosis factor-α and interleukin-4 was lower in osteopontin-null than in wild-type autografts, and there was no difference between the osteopontin-null graft types. CONCLUSIONS: Osteopontins from muscle and non-muscle sources are equally important in the acute response of muscle to injury, and cannot substitute for each other, suggesting that they play distinct roles in regulation of cell behaviour. Future studies of mechanisms of osteopontin’s roles in acute muscle inflammation and regeneration will need to investigate responses to osteopontins derived from both myoblasts and macrophages.