Impaired bone resorption by lipopolysaccharide in vivo in mice deficient in the prostaglandin E receptor EP4 subtype

Abstract

In a previous study we showed that the involvement of EP4 subtype of the prostaglandin E (PGE) receptor is crucial for lipopolysaccharide (LPS)-induced osteoclast formation in vitro. The present study was undertaken to test whether EP4 is actually associated with LPS-induced bone resorption in vivo. In wild-type (WT) mice, osteoclast formation in vertebrae and tibiae increased 5 days after systemic LPS injection, and urinary excretion of deoxypyridinoline, a sensitive marker for bone resorption, statistically increased 10 days after injection. In EP4 knockout (KO) mice, however, LPS injection caused no significant changes in these parameters throughout the experiment. LPS exposure for 4 h strongly induced osteoclast differentiation factor (ODF) mRNA expression in primary osteoblastic cells (POB) both from WT and EP4 KO mice, and this expression was not inhibited by indomethacin, suggesting prostaglandin (PG) independence. LPS exposure for 24 h further induced ODF expression in WT POB, but not in EP4 KO POB. Indomethacin partially inhibited ODF expression in WT POB, but not in EP4 KO POB. These data suggest that ODF is induced both PG dependently and PG independently. LPS exposure for 24 h induced slightly greater osteoclastgenesis inhibitory factor (OCIF) mRNA expression in EP4 KO than in WT POB. These findings suggest that the reduced ODF expression and apparently increased OCIF expression also are responsible for the markedly reduced LPS-induced osteoclast formation in EP4 KO mice. Our results show that the EP4 subtype of the PGE receptor is involved in LPS-induced bone resorption in vivo also. Since LPS is considered to be largely involved in bacterially induced bone loss, such as in periodontitis and osteomyelitis, our study is expected to help broaden our understanding of the pathophysiology of these conditions.

FIG. 1

Effect of systemic LPS injection on urinary excretion of D-Pyr. LPS (20 mg/kg of body weight) was injected subcutaneously into 11-week-old WT mice or EP4 KO mice. The ratios of D-Pyr (nanomolar) creatinine (Cr) (millimolar) before injection (day 0) and on days 5, 7, 10, and 14 after injection are shown. Values for the basal level (before injection) were assigned 100% in each case. Data are expressed as the mean + standard deviation (error bar). (n = 7). ∗, P < 0.05.

FIG. 2

(A) Histological analysis of vertebrae from WT mice (top) and EP4 KO mice (bottom). Figures show representative photographs of TRAP staining of lumbar vertebrae before (0d) and on days 5 and 10 (5d and 10d, respectively) after systemic LPS injection into WT or EP4 KO mice. Cells stained with red (showed by arrows) represent osteoclasts. (B) Osteoclast formation in sections of the vertebrae and the tibiae from WT mice and EP4 KO mice after systemic LPS injection. Mice were sacrificed before injection (day 0) and on days 5 and 10, and their vertebrae and tibiae were processed for histomorphometry. Values for the basal level (before injection) were assigned 100% in each case. Data are expressed as the mean + standard deviation (error bar). ∗, P < 0.05.

FIG. 3

Serum osteocalcin levels in WT or EP4 KO mice before (day 0) and on days 3, 5, 7, 10, and 14 after injection. Levels are expressed as the percentage of the value before injection for either group. Data are expressed as the mean + standard deviation (error bar). (n = 7). ∗, P < 0.05.

FIG. 4

Effect of LPS (100 ng/ml) on ODF and OCIF mRNA expression in POB from WT and EP4 KO mice. At confluency, POB were cultured in a starvation medium for 24 h and exposed to LPS with or without indomethacin (IND) (10⁻⁷ M) for 4 or 24 h. Fifteen micrograms of total RNA from each of the POB cultures was hybridized to ODF, OCIF, and β-actin probes. (A) Representative results of ODF and OCIF expression among similar results from four independent experiments. (B) The levels of ODF and OCIF mRNAs were normalized to that of β-actin at each time point by means of northern blot hybridization. Values for the basal level (after 24 h of starvation) were assigned 1.0 in each case.