It was previously reported that the expression of cyclo-oxigenase-2 (COX-2) is induced by prostaglandin E(2) (PGE(2)) in vitro in an osteogenic cell line and organ culture, suggesting an autoamplification mechanism. In this study, we first tested whether this phenomenon also occurs in bone tissue in vivo and found that a single anabolic dose of PGE(2) (5 mg/kg) induced (between 30 and 120 min) in rat tibiae, an increase in the mRNA level of COX-2 (2.5- to 9-fold) but not that of COX-1. Secondly, to test whether COX-2 activity in generating endogenous prostaglandins (PGs) is required for the in vivo anabolic properties of PGE(2), young male rats were injected daily with either vehicle (8% ethanol) or 5 mg/kg PGE(2) for 21 days. PGE(2)-injected rats received, 45 min prior to PGE(2), either dimethyl sulphoxide (as vehicle) or one of two doses of NS-398, a selective COX-2 inhibitor: a low dose (3 mg/kg) or a high dose (10 mg/kg). PGE(2) increased bone formation (measured as cancellous mineralizing surface, mineral apposition rate and bone formation rate) and bone mass (measured as cancellous bone area and surface and cortical width). None of these increases was suppressed by pre-administration of NS-398. In contrast, the high dose of NS-398 effectively suppressed an increase in rat hind-paw volume induced by a local carrageenan injection. Furthermore, since COX-2 inactivation may affect PG receptor expression, we found that pre-administration of NS-398 did not abolish the induction in EP(4) receptor mRNA levels, caused by PGE(2) in rat bone tissue. For in vitro testing, rat femoral bone marrow stromal cell cultures were initiated and were incubated in the absence or presence of PGE(2) at 100 nM (as an inducer) and with increasing concentrations of NS-398 (10(-8) M to 10(-5) M) for 21 days, after which time mineralized (Von-Kossa positive) nodules were counted. PGE(2) increased nodule formation as previously reported; however, NS-398 reduced nodule formation in both control and PGE(2)-treated cultures to the same extent. We conclude that while the level of COX-2 mRNA is increased in vivo by administration of PGE(2), inhibition of its activity (i.e. generation of endogenous PGs) does not abolish the anabolic effect of PGE(2).

Prostaglandin E(2) (PGE(2)) has been shown to exert a bone anabolic effect in young and adult rats. In this study we tested whether it possesses a similar effect on bone formation and bone mass in aging rats. Fifteen-month-old rats were injected daily with either PGE(2) at 5 mg/kg or vehicle for 14 days. PGE(2) treatment stimulated the rate of cancellous bone formation (a approximately 5.5-fold increase in bone formation rate), measured by the incorporation of calcein into bone-forming surfaces at the tibial proximal metaphysis. This effect resulted in increased cancellous bone area (+54%) at the same site. Since PGE(2) treatment resulted in a much higher proportion of bone surface undergoing bone formation and thus lined with osteoblasts, we tested the hypothesis that PGE(2) stimulates osteoblast differentiation from bone marrow precursor cells both in vivo and in vitro. We found that ex vivo cultures of bone marrow stromal cells from rats injected for 2 weeks with PGE(2) at 5 mg/kg per day yielded more ( approximately 4-fold) mineralized nodules and exhibited a greater (by 30-40%) alkaline phosphatase activity compared with cultures from vehicle-injected rats, attesting to a stimulation of osteoblastic differentiation by PGE(2). We also compared the osteogenic capacity of bone marrow from aging (15-month-old) versus young (5-week-old) rats and its regulation by PGE(2) in vitro. Bone marrow stromal cell cultures from aging rats exhibited a greatly diminished osteogenic capacity, reflected in reduced nodule formation ( approximately 6% of young animals) and lower alkaline phosphatase activity ( approximately 60% of young animals). However, these parameters could be stimulated in both groups of animals by incubation with 10-100 nM PGE(2). The magnitude of this stimulation was greater in cultures from aging rats (+550% vs +70% in nodule formation of aging compared with young rats). In conclusion, we demonstrate here that PGE(2) exerts a bone anabolic effect in aging rats, similar to the effect we and others have reported in young, growing rats. The PGE(2)-stimulated bone formation, which augments bone mass, most likely results from recruitment of osteoblasts from their bone marrow stromal precursors.