Prostaglandins (PGs) are key regulatory factors that affect bone metabolism. Prostaglandin E2 (PGE2) regulates bone resorption and bone formation. Prostacyclin (PGI2) is one of the major products derived from arachidonic acid by the action of cyclooxygenase and PGI2 synthase (PGIS). Unlike PGE2, there are few reports about the role of PGI2 in bone regulation. Therefore, we investigated the potential effect of PGI2 on bone metabolism. We used PGIS knockout (PGIS−/−), PGIS heterozygous (PGIS+/−), and wild-type mice to investigate the role of PGI2. Notably, PGIS−/− mice gradually displayed an increase in trabecular bone mass in adolescence. Adult PGIS−/− mice showed an increase in trabecular bone volume/tissue volume. Histomorphometric analysis showed that PGIS−/− mice displayed increases in both bone formation and bone resorption parameters. Levels of serum osteocalcin and C-telopeptides were increased in adult PGIS−/− mice. Furthermore, the increased bone mass patterns were rescued in PGIS−/tg mice. In conclusion, adult PGIS−/− mice displayed an overall increase in the levels of both bone formation and bone resorption parameters, which suggests that PGI2 deficiency accelerates high bone turnover activity with a greater increase in bone mass in aging. These results indicated that PGI2 may contribute to the maintenance of normal bone mass and micro-architecture in mice in age-dependent manner. Our findings demonstrate for the first time that PGI2 is involved in bone metabolism in vivo.
Prostaglandins (PGs) are one of the regulatory factors that affect bone metabolism (Raisz 1995, 1999, Harada & Rodan 2003). They are produced by the sequential oxygenation of arachidonic acid by cyclooxygenases (COXs) and terminal PG synthases (Simmons et al. 2004). The metabolites of COXs are involved in skeletal development. PGE2 regulates bone remodeling by mediating osteoblast differentiation (Yoshida et al. 2002, Arikawa et al. 2004) and regulates bone resorption by increasing cAMP levels through EP2 and EP4 receptor activation (Li et al. 2000, Miyaura et al. 2000, Yoshida et al. 2002, Sugimoto & Narumiya 2007). Prostacyclin (PGI2), a product converted by COXs and PGI2 synthase (PGIS), also up-regulates cAMP levels through the activation of a PGI2 receptor (IP; Namba et al. 1994). Resident bone cells; osteoblasts and osteocytes, can produce both PGE2 and PGI2 under mechanical loading (Rawlinson et al. 1991, 2000). Osteoblasts also produce PGI2 in response to growth factors such as vascular endothelial growth factor (Clarkin et al. 2008).
Genetic manipulation in mice has been a powerful tool for a better understanding of mammalian physiology. COX-2 gene deletion in mice induced a negative effect on bone formation by reducing osteoblastogenesis and delaying bone healing (Zhang et al. 2002, Naik et al. 2009), while skeletal abnormalities have not yet been reported in membrane-associated PGE synthase 1 null mice (Uematsu et al. 2002, Kamei et al. 2004).
Bone regulation during development processes has been widely studied (Harada & Rodan 2003). In contrast, information about controlling bone remodeling and maintaining bone mass in adults, which could contribute to therapeutic applications, remains limited. Although the critical role of PGE2 in bone has been widely demonstrated, the role of PGI2 is poorly understood. In a recent paper, it has been demonstrated that the injected iloprost, a stable PGI2 analog, stimulated calvarial bone formation in postnatal rats (Tuncbilek et al. 2008). Hence, the focus of the present study was to investigate the potential roles of PGI2 in bone metabolism.
In the present study, we investigated skeletal abnormalities in both young and adult PGI2-deficient mice. Bone abnormalities were assessed by radiography, histomorphometric analysis, and the measurement of biological markers. We found an increase in the bone mass of adult PGI2-deficient mice. Based on our results, the high bone mass phenotype in aged PGIS-null mice was a result from net favor of bone formation and mineralization. This phenotype was rescued in the offspring of PGIS-knockout mice crossbred with PGIS-transgenic mice. This evidence suggests that PGI2 is involved in bone metabolism and contributes to the lifelong maintenance of normal bone mass and micro-architecture in mice.
Materials and Methods
Female PGIS-null mice in a C57BL/6 strain (Yokoyama et al. 2002) were used in our study. PGIS+/− animals have no abnormal phenotype. PGIS-transgenic mice were generated in a C57BL/6 strain by the microinjection method. Mouse PGIS cDNA was inserted into the multicloning site of the modified pCI-neo mammalian expression vector (Promega), in which the CMV immediate early enhancer/promoter was replaced with the CMV early enhancer/chicken β-actin (CAG) promoter. The linearized plasmid was microinjected in fertilized eggs from C57BL/6J mice (CLEA Japan Inc., Osaka, Japan). The PGIStg/tg mice were crossbred with PGIS−/− mice to generate PGIS−/−/PGIStg/tg (PGIS−/tg) littermates. PGIS gene expression levels in the PGIS−/tg mice were confirmed by quantitative real-time PCR by using the DyNAmo probe qPCR kit (Finnzymes, Espoo, Finland). Mice were allowed food (CE-2; CLEA Japan Inc.) and distilled water ad libitum. Mice were maintained in a 12 h light:12 h darkness cycle. The experimental procedures were reviewed and approved by the Animal Care and Use Committee of Tokyo Medical and Dental University.
Radiographic assessment of bone
Peripheral quantitative computed tomography analysis
For the longitudinal assessment of bone, 5- to 34-week-old mice were kept alive during the bone density measurement by peripheral quantitative computed tomography (pQCT) (XCT Research SA+; Stratec Medizintechnik GmbH, Pforzheim, Germany) (Aoki et al. 2006, Alles et al. 2009). To conduct high-resolution scanning mode pQCT, mice were anesthetized with injections of medetomidine hydrochloride (0.5 mg/kg; Meijiseika, Tokyo, Japan) and ketamine hydrochloride (50 mg/kg; Sankyo, Tokyo, Japan; Saito et al. 2007).
Soft X-ray, dual X-ray absorptiometry and microfocal-computed tomography analysis
The mice were killed by using ether anesthesia. Their hind limbs (tibias and femurs) were harvested. Muscles and soft tissue were roughly removed from the bones, and then the bones were fixed in phosphate-buffered glutaraldehyde (2.5%)–formalin (4%) fixative (pH 7.4) for 2 days. Then, the samples were rinsed and kept in PBS for 2 days and water for 1 day. Afterward, the prepared bone samples were used for the soft X-ray radiographic analysis (SRO-M50; Sofron, Tokyo, Japan).
Trabecular and cortical regions of long bones were assessed by conducting a longitudinal study with dual X-ray absorptiometry (DXA; DCS-600R; Aloka, Tokyo, Japan; Saito et al. 2007, Hussain Mian et al. 2008, Alles et al. 2009) and determining the trabecular bone volume/tissue volume (%) (BV/TV) by microfocal CT (micro-CT; SMX-90CT, Shimadzu, Kyoto, Japan) respectively. Trabecular bone in secondary spongiosa was measured at a 450–500 μm distance from the growth plate of the tibia to avoid primary spongiosa.
Histological assessment of bone
Mice received s.c. injections of demeclocycline hydrochloride (8 mg/kg of body weight; Sigma) and calcein (10 mg/kg of body weight; Sigma) at 10 days and 3 days prior to killing respectively. After micro-CT analysis, the bone specimens were embedded as undecalcified sections in mixtures of methyl methacrylate and 2-hydroxyethylmethacrylate resins, as described previously (Saito et al. 2007). Polymerization was performed at 4 °C. Standard sections (3 μm thick) were prepared with a Reichert-Jung microtome 2050 Supercut (Cambridge Instrument, Heidelberger, Switzerland). Un-decalcified sections were stained with tartrate-resistant acid phosphatase (TRAP) and counterstained with toluidine blue. Some sections were stained with Von Kossa for visualizing mineralization. Standard histomorphometric analysis of the tibial metaphysis was performed by using an image analyzing system (KS400, Carl Zeiss, Jena, Germany), as described elsewhere (Parfitt et al. 1987, Aoki et al. 2006).
Measurement of bone biological markers
After blood samples were collected, serum osteocalcin levels were measured by using a Mouse osteocalcin EIA kit (Biomedical Technologies Inc., Stoughton, MA, USA; Tomomatsu et al. 2009). The quantity of C-terminal telopeptides (CTX), which are generated by osteoclasts as a degradation product of type I collagen, was measured by RatLaps EIA (Immunodiagnostic system, Boldon, UK; Garnero et al. 2003).
Statistical analysis was performed by using the Student's t-test and one-way ANOVA. When a significant F ratio was identified, groups were compared by using Tukey's honestly significant differences (HSD) protected least-significant difference post hoc test.
Skeletal phenotypic change was detected in PGIS−/− mice
PGIS−/− mice were born at the expected Mendelian ratio with grossly normal bone shape and size. Body weights of PGIS−/− mice were similar to those of PGIS+/− and wild-type (WT) littermates. PGIS+/− mice showed no obvious phenotype change. We examined the change in bone content in 5-week-old anesthetized PGIS-null mice by pQCT measurement, and the PGIS−/− mice showed a significant decrease in trabecular bone density. The pQCT measurement was repeated periodically. Over time, PGIS−/− mice developed an increase in trabecular bone density that became significantly different as compared to PGIS+/− littermates (by week 27; Fig. 1A). Cortical bone density in PGIS−/− mice at the later period of life was not significantly different (Fig. 1B).
At 5 weeks of age, lower bone mineral density (BMD) was observed in PGIS−/− mice by soft X-ray photographs (Fig. 2A), micro-CT (Fig. 2B), and DXA analysis (Fig. 2C). To determine the skeletal change in more detail, we performed bone analysis by using micro-CT analysis. Fractional trabecular bone volume BV/TV (%) in PGIS−/− mice was decreased as compared to that in the WT mice (Fig. 2D).
Adult PGIS−/− mice have increased trabecular bone mass
The increasing bone mass phenotype was observed in PGIS−/− mice (Fig. 3A and B). Corresponding to the time-dependent study by pQCT measurement, 34-week-old PGIS−/− mice showed an increase in the trabecular BV/TV as compared to PGIS+/− mice (Fig. 3C) (each group, n=8). This increase can be visualized in two-dimensional micro-CT images at the dense trabeculation in metaphyseal regions along the midshaft area. Although female mice were used in our study, these bone phenotypes were observed in both females and males (Supplementary Figure 1A and B, see section on supplementary data given at the end of this article).
Based on histochemical sections of tibia examined by Von Kossa staining, the PGIS−/− mice showed a greater mass of trabecular bone along the marrow space of the tibia distal metaphyseal area, as compared with the PGIS+/− mice. The results of Von Kossa staining indicated that the trabecular bones filling in the marrow space were well mineralized in adult PGIS−/− mice (Fig. 3D). Toluidine blue and TRAP staining showed an increase in the number of osteoclast-like cells along the newly formed bone surface (Fig. 3E and F). In a histomorphometric analysis of adult PGIS−/− mice, both bone formation and bone resorption parameters were increased (Fig. 4A–G). To determine whether the increase in trabecular bone mass was due to an acceleration in bone formation, a dynamic histomorphometric analysis was performed by using double-fluorochrome labeling at a 7-day interval with demeclocycline and calcein, which are markers of the amount of newly formed bone (Fig. 5A). The results demonstrated increases in the bone formation parameters of mineral surface per bone surface, bone formation rate, and mineral appositional rate (Fig. 5B–D). The serum osteocalcin level, which is a biochemical parameter for bone formation, was significantly increased in PGIS−/− mice (Fig. 5E). Serum CTX, which represents bone resorption activity, was also increased in PGIS−/− mice (Fig. 5F).
Bone phenotype was rescued in crossbred PGIS−/tg mice
The PGIS gene was abundantly expressed in PGIStg/tg mice. The mRNA expression in PGIStg/tg mice was increased 2.5- to 3-fold compared to that in WT mice (data not shown). PGIStg/tg mice displayed no skeletal phenotypic change. We next examined whether changes in skeletal phenotypes occurred in the offspring of overexpressed PGIS mice crossbred with PGIS-null mice. The PGIS−/tg mice showed the rescuing of the bone phenotype, as confirmed by micro-CT analysis (Fig. 6A). The trabecular BV/TV value was recovered by ∼50% as compared to PGIS−/− mice (Fig. 6B). To confirm that the morphological changes corresponded to the biochemical aspect, we measured serum osteocalcin and CTX. Similar to the radiographic results, the serum osteocalcin level in PGIS−/tg mice tended to be recovered to the level in PGIS+/− mice (Fig. 6C); the serum CTX level in PGIS−/tg mice was significantly lower than that in PGIS−/− mice and was similar to the level in PGIS+/− mice (Fig. 6D). Notably, PGIS+/−/tg mice were not different from the PGIS+/− group by micro-CT and histological analysis (Supplementary Figure 2A–C, see section on supplementary data given at the end of this article).
PGs usually have a biphasic effect on bone formation, which is mediated by cAMP regulation (Raisz & Fall 1990, Raisz 1999). The roles of endogenous PGs in skeletal growth are suggested by studies of fracture repair and heterotropic ossification, which probably involve the production of both PGE2 and PGI2 (Rawlinson et al. 1991, 2000, Raisz 1999). PGs have important effects on the skeleton, and it is strongly suggested that these effects are moderated by different receptors and specific pathways, which still remain unidentified. A large number of specific PG receptor gene-deleted animals have been cloned (Sugimoto & Narumiya 2007). PGE2 stimulates bone formation and bone resorption via the EP4 receptor. EP4 agonist is an effective bone anabolic agent in rats (Hagino et al. 2005, Ito et al. 2006); however, the efficacy of the anabolic effect in mice is limited (Kato et al. 2007). In regards to the bone phenotype in EP4-knockout animals, mice with a global deletion of the EP4 receptor developed osteopenia and impaired fracture healing with age (Li et al. 2005), while a recent study showed that targeted deletion of the EP4 receptor showed no skeletal change in aged mice (Gao et al. 2009). Thus, the roles of PGI2 in bone metabolism may be associated with different underlying mechanisms. It has been demonstrated that the injected iloprost, a stable PGI2 analog, stimulated calvarial bone formation in postnatal rats (Tuncbilek et al. 2008). Their data may support our present observation in which at 5 weeks of age, the BMD was lower in PGIS−/− mice.
In the present study, we primarily observed the skeletal changes in PGIS−/− mice by pQCT and micro-CT analysis. Lower bone mass was observed in 5-week-old PGIS−/− mice and was corrected at puberty period at 6–8 weeks old, which may have occurred from adapting process or changes in endogenous hormone levels. Unexpectedly, PGIS−/− mice showed a gradual increase in trabecular density and BV/TV values in age-dependent manner, while changes in cortical component in adult period of life were limited.
By histology and histomorphometry, we found that adult PGIS−/− mice displayed increases in bone formation and bone resorption parameters. This phenomenon was confirmed by measurements of serum bone markers, osteocalcin and CTX, which were significantly increased. The total net result in increased bone mass may be caused by the favoring of bone formation over bone resorption. Therefore, we performed a mineralized nodule formation assay, osteoclast formation, and pit assay with bone marrow cells derived from these mice. However, these in vitro assays did not show any differences between PGIS−/− and PGIS+/− mice (Supplementary Figure 3A–D, see section on supplementary data given at the end of this article). Thus, our in vitro data suggest that the lack of PGI2 does not affect the ability of bone precursor cells to differentiate.
While PGIStg/tg mice showed no distinct skeletal changes, bone mass and serum bone markers of the crossbred PGIS−/tg mice were rescued back to normal, as compared with those of knockout animals. These data confirmed that the lack of PGI2 in mice caused the accelerated bone turnover and that PGI2 may play a role as a negative regulator of bone metabolism, although it is not clear how the mechanism of PGI2 is involved.
Bone is a highly vascular structure. Blood supply in bone is important for bone modulation and remodeling. PGI2 is mainly released by vascular endothelial cells. The signaling of PGI2 via its cell surface receptor, the IP receptor, is well documented in vascular biology and inflammatory responses. IP receptor serves the critical roles with the vasodilatation, mediating platelet aggregation and controlling vascular smooth muscle cell proliferation (Namba et al. 1994, Murata et al. 1997, Stitham et al. 2007). It is demonstrated that there is a communication between endothelial and bone cells, which is important for controlling vascular supply during bone growth, remodeling, and repair (Deckers et al. 2000, Clarkin et al. 2008). However, there is no report of any changes in the bone phenotype of IP receptor knockout mice. Meanwhile, PGI2 action via nuclear receptors in other physiological responses is gradually being more appreciated. PGI2 exerts its effects by interacting with a nuclear hormone receptor, the peroxisome proliferator-activated receptor (PPAR). Therefore, the response to PGI2 should be considered by both IP and PPAR in certain cell types (Namba et al. 1994, Hertz et al. 1996, Hatae et al. 2001). One of the three subtypes of PPARs, the PPARδ uses PGI2 as an endogenous ligand to modulate specific cellular and metabolic functions (Lim et al. 1999, Gupta et al. 2000, Hatae et al. 2001). Further studies are needed to determine whether PGI2 is involved with PPARδ signaling in bone tissues.
In conclusion, we demonstrated that the bone mass in adult PGI2-deficient mice is increased. This increase in bone mass was associated with an increase in both bone formation and bone resorption, which suggests that PGI2 deficiency accelerates high bone turnover activity with a greater increase in bone mass associated with aging. Furthermore, this imbalance was reversed in knockout animals by the insertion of PGIS overexpression. Our finding suggests that PGI2 may play a role as a regulator in maintaining normal bone mass and micro-architecture.
This is linked to the online version of the paper at http://dx.doi.org/10.1677/JOE-09-0376.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
This study was supported by grants from the Ministry of Education, Culture, Sports, Science, and Technology of Japan to IM (20659306) and CY (18592060, 20592196).
We thank Drs Wantida Sriarj and Niroshani Soysa (TMD, University of Tokyo, Japan) for their technical support; and we also thank Drs Susumu Ohkawara, Yuji Arai, and Takayuki Morisaki (National Cardiovascular Center Research Institute, Osaka, Japan) for their technical support of the generation of PGIS-transgenic mice.
SaitoHKojimaTTakahashiMHorneWCBaronRAmagasaTOhyaKAokiK2007A tumor necrosis factor receptor loop peptide mimic inhibits bone destruction to the same extent as anti-tumor necrosis factor monoclonal antibody in murine collagen-induced arthritis. Arthritis and Rheumatism561164–1174.