Prostaglandins (PGs) are produced by osteoblastic cells and are abundant in bone matrix . These compounds are complex regulators of bone remodelling that can stimulate bone formation and resorption. The main enzyme regulating the conversion of arachidonic acid released from membrane phospholipids to PGs is prostaglandin G/H synthase (PGHS), commonly called cyclooxygenase (COX). There are two COX isoenzymes encoded by separate genes, COX-1 and COX-2 . COX-1 is usually constitutively expressed, while COX-2 is rapidly and transiently induced. Both COX-1 and COX-2 are expressed in osteoblasts, and COX-2 is the main enzyme regulating the production of PG in response to various hormones and cytokines .

Non-steroidal anti-inflammatory drugs (NSAIDs) and selective COX-2 inhibitors are some of the most commonly prescribed medications worldwide . They are used to treat painful inflammatory conditions such as arthritis, traumatic injuries, back pain, and dysmenorrhea, and are becoming part of comprehensive pain management. Recent studies have reported that COX-2 function is essential for skeletal fracture repair . Some researchers using standardized models of fracture repair in the femur or tibia have shown histological and mechanical evidence that COX-2 inhibitors delay healing ; however, the effects of COX-2 activity on fracture repair of the membranous viscerocranium are poorly understood. For example, a mouse calvarial defect model in which calvarial bone is removed has been used , but this is not suitable as a fracture model, because there is a large bone defect.

Previous studies have shown that fracture repair in the long bones develops from endochondral ossification (osteogenesis cartilaginea) . While craniofacial bones such as the calvaria, maxilla, and mandible develop from intramembranous ossification (osteogenesis membranacea). To the best of the authors’ knowledge, craniofacial fracture repair in COX-2 ^−/− mice has not been investigated. Understanding the mechanism of fracture repair that involves intramembranous ossification is important in oral and maxillofacial surgery.

In this study, the authors initially established an in vivo experimental protocol for craniofacial fracture repair in mice. The effect of COX-2 on bone response in the calvaria of COX-2 wild-type ( COX-2 ^+/+ ) and knockout ( COX-2 ^−/− ) mice was then examined.

Fracture repair model

The animals were anaesthetized by intraperitoneal injection of a mixture of 0.1 ml/kg of ketamine hydrochloride (Ketaral; Sankyo Corp.) and xylazine (Cerectal; Bayer, Leverkusen, Germany) in a ratio of 1:1. The skin covering the surgical region was disinfected with 70% ethanol. A linear incision was made from the centre of the occipital bone to the frontal bones over the parietal bones to reveal soft tissue, and the periosteum was stripped. A 4-mm circular fracture was created on the right side of the parietal bone with a sterile disposable trephine (Biopsy Punch: Maruho, Osaka, Japan), by hand. To determine ossification in the fracture site, it was considered important in this fracture model that the calvarial disk was not removed. Care was taken to leave the calvarial disks in place to avoid dural perforation, and the incisions were sutured closed ( Fig. 1 ).

Surgical procedure. (A) An incision was made in the calvaria, and the bone surface was exposed. (B) A 4-mm fracture was created on the right side with a trephine, by hand. (C) The calvarial disks were left in place with the underlying dura mater.

RNA analysis

Four mice in each group were killed under general anaesthesia 0, 1, 2, and 4 h after fracture, because COX-2 mRNA is rapidly and transiently induced. The calvarial bone was removed. It was difficult to extract only the fracture site, so the bone bordering the fracture anteriorly and posteriorly was dissected free of overlying soft tissue and snap frozen in liquid nitrogen. The bone was homogenized in ISOGEN (Wako Pure Chemicals Co., Osaka, Japan), using a Polytron tissue homogenizer (Kinematica Ag, Switzerland) ( Fig. 2 A) . Total cellular RNA was extracted according to the manufacturer’s instructions, and 2 μg of total RNA was reverse-transcribed with a Takara RNA polymerase chain reaction (PCR) kit (AMV), ver. 2.1 (Takara, Tokyo, Japan), to generate single-stranded cDNA. PCR was performed with an ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster, CA, USA). The following oligonucleotide primers were used: COX-2 forward, 5′-TGGTGCCTGGTCTGATGATG-3′; COX-2 reverse, 5′-GTGGTAACCGCTCAGGTGTTG-3′; actin forward, 5′-AGATGTGGATCAGCAAGCAG-3′; and actin reverse, 5′-GCGCAAGTTAGGTTTTGTCA-3′. The PCR assays used 1× QuantiTect SYBR Green PCR Master Mix (Qiagen Inc., Valencia, CA, USA), 0.3 μM specific primers, and 500 ng of cDNA. The mRNA copy number of a specific gene in each sample of total RNA was calculated with a standard curve generated with serially diluted plasmids containing PCR amplicon sequences and was normalized to rodent total RNA (Applied Biosystems), with actin as an internal control. Standard plasmids were synthesized with a TOPO TA Cloning Kit (Invitrogen Co., Carlsbad, CA, USA) according to the manufacturer’s instructions. All reactions were run in triplicate.

(A) Diagram showing the region from which RNA was extracted. The box indicates the part of the calvarial bone from which RNA was extracted. The calvarial bone was harvested, and soft tissue was removed. (B) Messenger RNA level of COX-2 in the fracture region. Total RNA was extracted at 0, 1, 2, and 4 h, and real-time PCR was performed as described in the text. The relative expression of COX-2 mRNA, standardized to actin, was determined as described in the text. Data are means ± SEM from three experiments, each with four replicates per treatment group. *Significant difference from 0 h group, P < 0.01.