The pathogenesis of bisphosphonate-related osteonecrosis of the jaw (BRONJ) is highly controversial. We have previously reported the development of osteonecrosis by periodontal pathogenic stimulation in the jaw and femur of rats treated with bisphosphonate. Since the major toxicity factor of Gram-negative bacteria is lipopolysaccharide (LPS), the present study aimed to evaluate the relationship between osteonecrosis and LPS in a rat model of BRON-like lesions. Seventeen male rats were injected subcutaneously with zoledronic acid weekly for 4 weeks and divided into three groups: LPS (LPS administered into the bone marrow of the mandible and femur) and LPS plus polymyxin B (PMB) and saline groups (given neutralized LPS with PMB or saline, respectively, using the same protocol). At 4 weeks after the procedure, harvested specimens were analyzed using histomorphology ( n = 5 from each group) and histochemistry ( n = 1 each from LPS and LPS plus PMB groups). There was a significantly wider area of osteonecrosis in the LPS group as compared to the saline and LPS plus PMB groups in both the mandible ( P = 0.030 and P = 0.009, respectively) and femur ( P = 0.002 and P = 0.020, respectively). Our results indicate that LPS stimulation is deeply involved in the development and promotion of BRON.
Since 2003, the number of reports on bisphosphonate-related osteonecrosis of the jaw (BRONJ), a known adverse effect of bisphosphonate (BP) use, has been increasing. Clinically BRONJ only occurs in the jaw in spite of BP being administered to the whole body. However, a previous study by our group showed that limited osteonecrosis in Wistar rats administrated BP occurred not only in the jaw but also in the femur following mechanical stimulation (drill-hole fenestration of cortical bone). Interestingly, our findings also confirmed that additional inflammatory stimulation with a periodontal pathogen, freeze-dried Aggregatibacter actinomycetemcomitans , promoted that osteonecrosis significantly. From those results, we concluded that the mechanism of BRONJ, developing only in the jaw in the clinical setting, must involve an environmental factor related to the fact that the jaw is readily exposed to oral bacteria.
Lipopolysaccharide (LPS) is one of the components of the outer membrane of Gram-negative bacterial cells, and shows strong mitogen activity as an endotoxin. Since it is both thermally and chemically stable, the major antigenic factor of freeze-dried A. actinomycetemcomitans used in our previous study must have been LPS.
In the present study, we examined whether purified LPS stimulation similarly promotes osteonecrosis and if neutralization of LPS toxicity with polymyxin B (PMB) causes a reduction in osteonecrosis in rats administered BP. PMB is known to reduce the mitogenic effect of LPS, an endotoxin with a negative charge. PMB is a positively charged cationic antibiotic that reduces the bioactivity of LPS for binding to the lipid A portion of LPS.
The aim of this study was to evaluate the effects of LPS on the development of BRONJ-like lesions.
Materials and methods
The protocol of the present study was approved by the institutional animal care and use committee.
Seventeen male Wistar rats (8 weeks old; body weight 265–295 g) were used in the experiments. All animals had access to laboratory food and water ad libitum. They were individually housed in plastic cages in a monitored environment (temperature 22 ± 1 °C, humidity 50 ± 5%, 12-h light/12-h dark cycle). All rats were treated with a subcutaneous injection of zoledronic acid (ZA; Novartis Pharma, Basel, Switzerland) at 0.1 mg/kg of body weight weekly for 4 weeks. The dosage of ZA was determined in reference to the dose for adult cancer patients.
Fifteen rats were divided into three groups, with five in each group, according to the stimulation method employed, as follows. One week after the final injection (13 weeks old), the rats underwent a surgical procedure during which a drill hole was made in the bilateral mandibular borders and in the diaphysis of the bilateral femurs. The LPS group then received an injection of 50 μg of A. actinomycetemcomitans LPS into the bone marrow of the bilateral mandibles and femurs. The LPS plus PMB group received an injection of 50 μg of A. actinomycetemcomitans LPS mixed with 500 U of PMB (Nacalai Tesque, Inc., Kyoto, Japan) using the same protocol. The LPS was extracted from lyophilized cells of A. actinomycetemcomitans Y4 using a hot phenol–water method. The saline group received an injection of saline using the same protocol.
The operative procedure used was the same as in our previous study. Briefly, the animals were anesthetized with an intraperitoneal administration of chloral hydrate (3.5 mg/kg). A skin incision was then made on the ventral surface over the mandibular borders and femurs with the rat in a supine position. After blunt dissection, the periosteum of the mandible and femur was cut and reflected to expose the bone surface. A drill hole was then made in the cortical bone using a small round bur (1.8 mm in diameter) with sufficient irrigation in the posterior portion of the bilateral mandibular borders and middle portion of the ventral surface of the bilateral femurs. A microfibrillar collagen haemostat (Zeria Pharmaceutical Co. Ltd., Tokyo, Japan) soaked with A. actinomycetemcomitans LPS (50 μg/site, dissolved in 5 μl saline), a mixture of LPS and PMB (LPS 50 μg/site, PMB 500 U, dissolved in 5 μl saline), or saline (5 μl/site) was packed into each hole in the LPS, LPS plus PMB, and saline groups, respectively. Finally, the holes were covered with yellow bone wax (Alfresa Pharma Corporation, Osaka, Japan). After thorough irrigation of the wounds, the incisions were closed in layers ( Fig. 1 ).
Four weeks after the operation, all rats were euthanized with an overdose of chloral hydrate and fixed with a perfusion of 2% paraformaldehyde. The mandibles and femurs were then removed and 10 samples from each group (both sides of each rat) were subjected to analysis. Soft radiographic images of all specimens were obtained before the histological examinations (25 kVp, 2.5 mA, 10 min; SRO-M50; Sofron, Tokyo, Japan).
The harvested mandibles and femurs were fixed in 10% formalin, then decalcified in 10% ethylenediaminetetraacetic acid (EDTA; pH 7.2) at room temperature for 28 days and embedded in paraffin. The mandibular specimens were sliced in a sagittal direction into 6-μm-thick sections that included the middle region of the drilled site at the mandibular border. The femur specimens were sliced along the long axis, including the middle region of the drilled site, at the same thickness. Histological staining was performed using haematoxylin and eosin.
We defined osteonecrosis as a region of empty lacuna >500 μm 2 , in accordance with the definition of Allen and Burr. The area of necrotic bone with continuity to the drilled site was measured using a DP2-BSW (Olympus, Tokyo, Japan). Differences in the area of osteonecrosis among the three groups were assessed using a Steel–Dwass post hoc test after a Kruskal–Wallis test with the Statcel 3 add-in form on Excel (version 3; Saitama, Japan). Statistical significance was set at P < 0.05.
The remaining two rats were used for observations of alkaline phosphatase (ALP) and tartrate-resistant acid phosphatase (TRAP) activity following LPS or LPS plus PMB stimulation. The surgical protocol for the LPS and LPS plus PMB groups was performed as described previously. After harvesting and formalin fixation of the specimens, non-decalcified frozen sections were cut into 7-μm-thick sections with a Leica CM3050S cryostat. Double staining with ALP and TRAP, which indicate osteoblasts and osteoclasts, respectively, was performed for histochemistry analysis.
The ZA injection did not cause any adverse effects. No abnormal activity was observed during the experimental period and animal body weights increased during the study. At 4 weeks after the surgical procedure, there were no significant differences in body weight among the groups. None of the animals developed exposed bone lesions at the surgical site during the study period.
At 4 weeks after the surgical procedure, soft X-ray imaging revealed callus formation in all of the operated mandibles and femurs as a response to the surgical procedure. The femurs had greater callus formation as compared to the mandibles. However, no apparent osteonecrosis findings, such as sequester formation, were observed in either the mandibles or femurs in any of the radiographic images.
Histological and histochemical examination
External callus formation was seen in all of the operated mandibles and femurs as a response to the surgical procedure. Furthermore, osteonecrosis was observed to some degree in the cortical bone around the drill hole in all of the specimens. However, in both mandibles and femurs, the area of osteonecrosis in the LPS group was much wider than that in the other two groups ( Figs 2 and 3 ). In addition, the area of osteonecrosis observed in the LPS plus PMB group was the same as that in the saline group.
With regard to the double staining with ALP and TRAP, there were fewer ALP-positive cells in the LPS specimens than in the LPS plus PMB specimens ( Fig. 4 ). There were no TRAP-positive cells around the drill hole in any of the specimens from the LPS and LPS plus PMB groups.
In the mandibles, the median areas of osteonecrosis in the LPS, LPS plus PMB, and saline groups were 0.61 (interquartile range (IQR) 0.44–0.85), 0.32 (IQR 0.28–0.41), and 0.30 (IQR 0.26–0.44) mm 2 , respectively. Osteonecrosis observed in the LPS group was significantly wider than that in the saline group ( P = 0.030) and LPS plus PMB group ( P = 0.009), while the osteonecrosis observed in the LPS plus PMB group was comparable to that in the saline group ( Fig. 5 ). In the femurs, the median areas of osteonecrosis were 1.97 (IQR 1.73–2.15), 1.19 (IQR 0.95–1.62), and 1.20 (IQR 0.92–1.32) mm 2 , respectively. The area of osteonecrosis in the LPS group was significantly wider than that in the saline group ( P = 0.002) and LPS plus PMB group ( P = 0.020), while there was no significant difference between the LPS plus PMB and saline groups ( Fig. 6 ).