Journal of Bone and Mineral Metabolism

, Volume 28, Issue 2, pp 157–164

Antitumor necrotic factor agent promotes BMP-2-induced ectopic bone formation

Authors

  • Yoshitaka Eguchi
    • Department of Orthopaedic SurgeryOsaka City University Graduate School of Medicine
    • Department of Orthopaedic SurgeryOsaka City University Graduate School of Medicine
  • Yuuki Imai
    • Department of Orthopaedic SurgeryOsaka City University Graduate School of Medicine
  • Yoshifumi Naka
    • Department of Orthopaedic SurgeryOsaka City University Graduate School of Medicine
  • Yuusuke Hashimoto
    • Department of Orthopaedic SurgeryOsaka City University Graduate School of Medicine
  • Hiroaki Nakamura
    • Department of Orthopaedic SurgeryOsaka City University Graduate School of Medicine
  • Kunio Takaoka
    • Department of Orthopaedic SurgeryOsaka City University Graduate School of Medicine
Original Article

DOI: 10.1007/s00774-009-0127-x

Cite this article as:
Eguchi, Y., Wakitani, S., Imai, Y. et al. J Bone Miner Metab (2010) 28: 157. doi:10.1007/s00774-009-0127-x

Abstract

Etanercept (ETN), which is a recombinant human soluble tumor necrosis factor (TNF) receptor that inhibits TNF activity, is effective in the treatment of rheumatoid arthritis. We investigated the effect of ETN on recombinant human bone morphogenetic protein-2 (rhBMP-2)-induced ectopic bone formation in vivo. A block copolymer composed of poly-d,l-lactic acid with random insertion of p-dioxanone and polyethylene glycol (PLA–DX–PEG polymer) was used as the delivery system. Polymer discs (6 mm, 30 mg) containing 5 μg rhBMP-2 were implanted into the left dorsal muscle pouch of mice (n = 50). In the systemic administration groups (n = 5 per group), ETN was subcutaneously injected (25 mg/human = 12.5 μg/mouse) twice per week in a dose-dependent manner (placebo, 12.5 × 10−3, 12.5 × 10−1, 12.5, 125 μg), whereas a single dose of ETN (placebo, 12.5 × 10−3, 12.5 × 10−1, 12.5, 125 μg) was embedded in each rhBMP-2 polymer disc in the local administration groups (n = 5 per group). Three weeks after implantation, the mice were killed and the implants were analyzed. Implants in the optimally dosed groups had increased radiodensity, which was consistent with a significant increase in bone mineral content of the ossicles. Bone histomorphology revealed a significant increase in bone volume/total volume, number of osteoblasts, osteoblast surface/bone surface, and a significant decrease in the number of osteoclasts, osteoclast surface/bone surface in the optimal dosed systemic and locally administered groups. These data suggest that the optimal dose of ETN, administered either systemically or locally, enhanced the bone-inducing capacity of BMP with no apparent adverse systemic effects.

Keywords

Tumor necrosis factor-αBone morphogenetic protein-2EtanerceptOsteoclastOsteoblast

Introduction

Tumor necrosis factor-α (TNF-α) is one of the most potent pro-inflammatory cytokines and is known to be a catabolic factor in the inflammatory reaction of diseases such as rheumatoid arthritis (RA). TNF-α promotes bone destruction through activation of osteoclastogenesis [13] and is also involved in bone formation. Several reports suggested that anti-TNF therapy in the patients with RA improved bone mineral density of lumbar spine and bone metabolic markers [4, 5]. TNF-α reduces bone morphogenetic protein-2 (BMP-2)-induced alkaline phosphatase (ALP) activity in vitro [6], inhibits Smad signaling through the Ras/Rho/MAPK pathway [7], and inhibits Runx2 degradation through upregulation of Smad ubiquitin regulatory factor (Smarf) 1 and 2 [8]. The effect of in vivo TNF-α has not been clarified.

Etanercept (ETN) is a human soluble TNF type II receptor (p75) that binds and inactivates TNFs. Here, we applied ETN systemically and locally in a p-dioxanone and polyethylene glycol (PLA–DX–PEG) polymer containing 5 μg recombinant human BMP-2 (rhBMP-2) and implanted the discs into the dorsal muscle pouch of mice during ectopic bone formation to investigate the effect of TNF-α on rhBMP-2-induced ectopic bone formation in vivo. Our results suggest that the optimal administration of ETN systemically and locally enhanced BMP-2-induced bone formation ectopically. These data extend our understanding of TNF function in the process of bone regeneration and suggest therapeutic strategies that may accelerate skeletal tissue regeneration.

Materials and methods

Reagents

ETN (Embrel®), which is a recombinant human soluble TNF receptor fusion protein with an immunoglobulin G1 dimer, was purchased from Takeda Pharmaceutical (Tokyo, Japan) and dissolved in distilled water at a concentration of 25 mg/ml before use. ETN has similar binding affinities to both human and mouse TNF-α [9, 10]. Recombinant human bone morphogenetic protein-2 derived from Escherichia coli (E-BMP-2) was provided by Osteophama (Osaka, Japan). Biodegradable copolymer composed of poly-d,l-lactic acid with random insertion of p-dioxanone and polyethylene glycol (PLA–DX–PEG polymer, MW10800, PLA/DX/PEG molar ratio) was used as the drug delivery system [11] and donated by Taki Chemical (Kakogawa, Japan).

Preparation of implants

The implants were prepared based on previous reports using biodegradable polymer and 5 μg rhBMP-2 with or without several doses of ETN [1216]. Our previous data suggested that the minimal optimal content of rhBMP-2 required to induce osteogenesis in mice is approximately 5 μg in 30 mg polymer (0.005%) [12].

Implantation into mouse muscle

Male ICR mice (4 weeks old) were purchased from Nippon SLC (Hamamatsu, Japan). The mice were housed and acclimated in cages with free access to food and water. After 1 week, the mice were operated on under intramuscular anesthesia (ketamine, 2.6 μg/g, and xylasin, 0.8 μg/g). The test implants were aseptically placed into the left dorsal muscle pouch (one pellet per animal). The whole-body weight of each mouse was monitored weekly. Animals were killed by high-dose inhalation of diethyl ether gas at 3 weeks after implantation. The experimental protocol was performed under the Regulations for Animal Experiments of Osaka City University.

Systemic and local administration of ETN

Fifty mice were divided into a local and a systemic administration group (25 mice per group). The doses of ETN were determined based on the concentration of ETN in clinical use for humans (25 mg/kg). Mice in each group were further divided into five subgroups based on the doses of ETN (5 mice per subgroup). For the systemic administration group, E-BMP-2 containing the polymer composite was implanted, and several concentrations of ETN (0, 12.5 × 10−3, 12.5 × 10−1, 12.5, or 125 μg) were injected subcutaneously at the base of the tail twice per week for 3 weeks (Table 1). For the local administration group, the implants contained several concentrations of ETN (0, 12.5 × 10−3, 12.5 × 10−1, 12.5, or 125 μg/implant) transplant, respectively, and no additional ETN was administered during the course of the experiment. Newly formed ectopic bone was harvested at 3 weeks after implantation in both groups and analyzed by radiologic and histological examination.
Table 1

Adverse systemic effects in the etanercept (ETN) systemic and local administration group

 

1

2

3

4

5

P*

Systemic group

 Body weight (g)

  0 w

30.0 (1.1)

31.0 (1.6)

29.5 (1.7)

29.3 (5.4)

29.9 (2.3)

n.s.

  1 w

35.8 (2.1)

36.5 (2.4)

35.7 (2.2)

35.4 (2.4)

36.2 (3.3)

n.s.

  2 w

40.3 (2.5)

40.9 (2.9)

39.3 (2.7)

38.9 (2.7)

39.5 (3.9)

n.s.

  3 w

42.8 (3.1)

42.5 (3.2)

41.1 (3.0)

40.9 (2.6)

41.3 (4.0)

n.s.

 BMC (tibia; mg)

  3 w

140.0 (0.6)

14.6 (1.3)

14.3 (1.3)

13.7 (1.3)

13.9 (1.0)

n.s.

 ALP (IU/dl)

  3 w

314.0 (46.4)

297.0 (66.6)

374.3 (130.1)

273.6 (88.7)

301.8 (81.0)

n.s.

 Ca (mg/dl)

  3 w

8.9 (0.1)

8.6 (0.1)

8.6 (0.1)

8.8 (0.2)

8.9 (0.1)

n.s.

 P (mg/dl)

  3 w

9.2 (1.3)

8.4 (1.0)

8.4 (0.5)

8.8 (1.7)

8.2 (0.7)

n.s.

Local group

 Body weight (g)

  0 w

26.9 (1.7)

26.9 (1.9)

26.7 (1.6)

26.8 (2.0)

26.8 (1.1)

n.s.

  1 w

32.9 (2.0)

34.2 (2.4)

35.0 (0.9)

33.9 (4.0)

35.0 (1.3)

n.s.

  2 w

36.1 (1.9)

36.8 (2.1)

38.0 (2.0)

36.7 (3.6)

36.9 (1.2)

n.s.

  3 w

39.6 (2.9)

39.2 (2.9)

40.8 (1.7)

39.0 (3.7)

39.7 (1.8)

n.s.

 BMC (tibia; mg)

  3 w

15.4 (1.3)

15.7 (1.2)

15.5 (0.7)

15.1 (1.3)

15.0 (1.3)

n.s.

 ALP (IU/dl)

      

  3 w

287.8 (40.3)

335.0 (9.7)

320.3 (59.0)

296.5 (49.9)

285.0 (85.4)

n.s.

 Ca (mg/dl)

  3 w

8.6 (0.2)

8.7 (0.2)

8.5 (0.3)

8.3 (0.3)

8.4 (0.4)

n.s.

 P (mg/dl)

  3 w

12.6 (0.9)

10.0 (1.0)

9.0 (0.8)

9.0 (0.9)

8.7 (0.8)

n.s.

Values indicate the mean (SE in parentheses). Body weight measured at preimplantation (0 w) and at 1 week (1 w), 2 weeks (2 w), and 3 weeks (3 w) after implantation. BMC indicates the bone mineral content of the left tibia; serum alkaline phosphatase (ALP), serum calcium (Ca), and serum phosphate (P) were evaluated at 3 weeks postimplantation. Each subgroup was divided according to quantity of ETN administered (μg): 1, 0; 2, 12.5 × 10−3; 3, 12.5 × 10−1; 4, 12.5 μg; 5, 125 μg

n.s. not significant

* P < 0.05 versus control

Radiologic examination

All ossicles were harvested and fixed in 70% ethanol and then radiographed with a soft X-ray apparatus (Sofron©, Tokyo, Japan). Bone mineral content (BMC; mg per ossicle) of each ossicle was measured by dual-energy X-ray absorptiometry (DXA) using a bone mineral analyzer (DCS-600EX; Aloka©, Tokyo, Japan).

Histological examination

All ossicles were treated with a graded series of ethanol and embedded in methyl methacrylate (MMA; Wako, Japan). Samples were sliced into 7-μm-thick sections and prepared for von Kossa (vK), tartrate-resistant acid phosphatase (TRAP), and toluidine blue (TB) staining as previously described [17]. The sections were visualized and measured with an optical microscope connected to a computer-controlled display (Axio Imager® A1; ©Zeiss, Germany). Each part of the resulting digital images were photographed at the same magnification (2.5 × objective) and merged with Adobe® Photoshop® CS2 software to visualize the whole section as a “Macro” view (see Fig. 3). We chose and marked a fixed, rectangular region of interest (ROI) in each section with the software and photographed each at the same magnification (40 × objective).

Each section was visualized with an optical microscope connected to a computer-controlled display (Olympus©, Tokyo, Japan) for bone histomorphometry. Eight areas of trabecular bone in each ossicle at the same magnification (20 × objective) were randomly selected, and the bone was analyzed histomorphologically according to a standard protocol [18] using computer software (OsteoMeasure®; ©OsteoMetrics, Decatur, GA, USA). Bone volume/tissue volume (BV/TV, %), osteoid surface/bone surface (OS/BS, %), number of osteoblasts (N.Ob, mm2), osteoblast surface/bone surface (Ob. S/BS, %), number of osteoclasts (N.Oc, mm2), and osteoclast surface/bone surface (Oc. S/BS, %) were measured in this experiment.

Adverse systemic effects

The body weights of all mice were measured every week. Whole blood was collected from the right ventricle at the same time that implants were harvested. Serum levels of alkaline phosphatase (ALP), calcium (Ca), and phosphate (P) were analyzed by Nippon SRL (Tokyo, Japan) to evaluate the effect of ETN administration on general bone metabolism. The left tibia of each mouse was harvested at 3 weeks after surgery, and BMC was measured in the left tibia from the proximal end to the tibiofibular joint as described previously [19].

Statistical analysis

All sample data were analyzed with statistical software (GraphPad Prism® version 5.0; ©GraphPad Software) and presented as mean ± SE. Statistical significance was assessed by Dunnett’s test. A probability (P value) < 0.05 was considered statistically significant.

Results

Systemic and local administration of ETN enhanced BMP-induced ectopic bone formation

At 3 weeks after surgery, the implants with or without ETN had been completely replaced by new bone (Figs. 1a, 2a). As expected, the implants treated with ETN alone (without BMP-2) showed no new bone formation regardless of group (data not shown). Soft radiographs revealed differences in size and radiodensity of the ossicles among the subgroups in both the systemic (Fig. 1a) and local (Fig. 2a) administration groups when compared with the control group. The DXA analysis revealed that BMC was significantly higher in the systemic group treated with 12.5 × 10−1 μg ETN and in the local group treated with 12.5 μg ETN compared to the control group (P < 0.05) (Figs. 1b, 2b).
https://static-content.springer.com/image/art%3A10.1007%2Fs00774-009-0127-x/MediaObjects/774_2009_127_Fig1_HTML.gif
Fig. 1

a Soft X-ray image of bone morphogenetic protein-2 (BMP-2, 5 μg) ectopically induced ossicles in the systemic administration group (n = 5, respectively), with or without various doses of etanercept (ETN). Bar 1 mm. Etanercept was injected subcutaneously six times in mice during the 3 weeks after implantation. b Bone mineral content (BMC) of ossicles in the systemic administration group. Asterisk indicates a statistically significant difference (P < 0.05) from the control group measured by Dunnett’s test. Data are represented by the mean ± SE

https://static-content.springer.com/image/art%3A10.1007%2Fs00774-009-0127-x/MediaObjects/774_2009_127_Fig2_HTML.gif
Fig. 2

a Soft X-ray image of bone morphogenetic protein-2 (BMP-2, 5 μg) ectopically induced ossicles in the local administration group with or without various dose of etanercept (ETN). Bar 1 mm. No additional etanercept was administered during the course of the experiment. b Bone mineral content (BMC) of ossicles in the local administration group. Asterisks indicate a statistically significant difference (P < 0.05) from the control group, measured by Dunnett’s test. Data are represented by the mean ± SE

ETN involved in osteoclast and osteoblast parameters in BMP-2-induced bone formation

Larger ossicles were confirmed in the low magnification of sections from the groups treated with ETN (Fig. 3b) when compared to the controls (Fig. 3a), and the group treated with ETN systemically and locally exhibited increased trabecular bone volume (Fig. 3d), increased numbers of osteoblasts (arrows in Fig. 3f), and decreased numbers of osteoclasts (arrowheads in Fig. 3h) when compared with the control group, regardless of ETN administrative method. Moreover, the bone histomorphometric analysis indicated that ETN treatment resulted in a significant increase in BV/TV (Figs. 4a, 5a), N.Ob, and Ob.S/BS (Figs. 4c, d, 5c, d) and a significant decrease in N.Oc and OcS/BS (Figs. 4e, f, 5e, f) compared with the control group regardless of administration method. There were no significant differences in OS/BS among the groups treated with or without ETN (Figs. 4b, 5b).
https://static-content.springer.com/image/art%3A10.1007%2Fs00774-009-0127-x/MediaObjects/774_2009_127_Fig3_HTML.gif
Fig. 3

Histological findings with or without etanercept (ETN) in the systemic administration group. Whole sections (Macro) of bone morphogenetic protein-2 (BMP-2)-induced newly formed ossicles with (b) or without (a) ETN (vK, von Kossa staining; 2.5 × objective). Bara 500 μm. Boxed areas show the sections with higher magnification (×40). Lower panels: vK staining (c, d), toluidine blue (TB) staining (e, f), and tartrate acid-resistant phosphatase (TRAP) staining (g, h). Arrows and arrowheads indicate osteoblasts and osteoclasts, respectively. Barg 50 μm

https://static-content.springer.com/image/art%3A10.1007%2Fs00774-009-0127-x/MediaObjects/774_2009_127_Fig4_HTML.gif
Fig. 4

Bone histomorphometric analysis with or without etanercept (ETN) in the systemic group. a BV/TV (bone volume/trabecular volume, %). b OS/BS (osteoid surface/bone surface, %). c N.Ob (number of osteoblast/mm2). d ObS/BS (osteoblast surface/bone surface, %). e N.Oc (number of osteoclast/mm2). f OcS/BS (osteoclast surface/bone surface, %). Asterisks indicate P < 0.05, representing a statistically significant difference from control group as measured by Dunnett’s test. Data are represented by the mean ± SE

https://static-content.springer.com/image/art%3A10.1007%2Fs00774-009-0127-x/MediaObjects/774_2009_127_Fig5_HTML.gif
Fig. 5

Bone histomorphometric analysis with or without etanercept (ETN) in the local group. a BV/TV (bone volume/trabecular volume). b OS/BS (osteoid surface/bone surface). c N.Ob (number of osteoblasts). d ObS/BS (osteoblast surface/bone surface). e N.Oc (number of osteoclasts). f OcS/BS (osteoclast surface/bone surface). Asterisks indicate P < 0.05, representing a statistically significant difference from control group as measured by Dunnett’s test. Data are represented by the mean ± SE

No significant differences in body weight gain during the experiment, BMC in the left proximal tibia, serum ALP, Ca, or P were noted among the groups treated with or without various concentrations of ETN systemically or locally at 3 weeks postoperatively (see Table 1).

Discussion

Our data indicated that systemic and local ETN administration promoted BMP-2-induced bone formation at 3 weeks postimplantation. The BMC of each implant significantly increased in response to an increase in the ETN concentration (see Figs. 1, 2). There were no significant differences in body weight, BMC of the tibia, serum ALP, Ca, or P among groups treated with or without ETN (Table 1). Thus, we concluded that ETN promoted cartilage formation [20] at an early stage and subsequent osteogenesis during BMP-2-induced ectopic bone formation. Our data also revealed a significant increase in BV/TV, N.Ob, and OBs/BS, and a decrease in N.Oc and OcS/BS, in the optimally dosed groups, but no significant differences in OS/BS were found among the groups (see Figs. 3, 4, 5). BMP-2-induced bone formation may be accelerated by local promotion of osteoblastogenesis and inhibition of osteoclastogenesis in the presence of ETN. Our objective was to gain insight into the role of the anti-TNF-α effect during BMP-2-induced ectopic bone formation.

Rheumatoid arthritis (RA) is a common chronic inflammatory disease, which results in inflammatory synovitis and subsequent joint dysfunction. Joint damage occurs early in the course of RA; 30% of patients have radiographic findings of bony erosion at the time of diagnosis and this proportion increases to 60% after 2 years [21]. The early control of cartilage and bone destruction is crucial for successful RA treatment [22]. The cause of RA is unknown; however, TNF-α is a key cytokine in RA and is mainly involved in osteoclastogenesis [13]. ETN is a recombinant human soluble TNF type II receptor (p75) linked to an IgG1-Fc moiety that binds to and inactivates the TNF interaction [23]. The efficacy of TNF inhibitors for preventing bone destruction is well known [24], and the administration of ETN prevents the worsening of joint erosion in patients with early RA [25], while TNF-α may play a key role in bone formation. Anti-TNF therapy in patients with RA demonstrated an improved bone formation/resorption marker ratio [4, 5]. It was previously suggested that TNF-α interacts with BMP in bone formation and regulates the physiological bone formation process [26, 27]. Our data revealed that ETN regulated osteoblastogenesis and osteoclastogenesis and that the effect reached a peak at a submaximal concentration in a BMP-2-induced ectopic bone formation model. The marked blockade of TNF action might delay the physiological process.

Local expression of TNF-α in our experiment is unclear. The procedure for its implantation into the muscle may cause inflammation. TNF is, however, widely expressed in the body, involved in normal bone formation, and induced by BMP [2628]. The 3-week period in our experiment may not have been sufficient to observe any changes in systemic bone mass because we did not intend to explore the influence of TNF on systemic bone mass but instead local bone formation ectopically at an early stage. Further study might be needed to evaluate the adverse systemic effect in a longer observation period.

BMP-2, which strongly induces ectopic bone formation, promotes the differentiation of early mesenchymal stem cells to chondroblasts and osteoblasts [2932]. RhBMP-2 is widely used as a fracture treatment [33] and for spinal fusion [34]. We previously described the safety and efficacy of a biodegradable co-polymer as a drug delivery system (DDS) for rhBMP-2 [11]. Osteogenesis using the DDS with rhBMP-2 has been confirmed in a critical size bone defect [35] and spinal fusion model [36] in rabbits. We also reported that rhBMP-2 in combination with the prostaglandin E EP4 receptor selective agonist, EP4A, enhances osteogenesis [14]. As shown in this study, a TNF inhibitor might enhance rhBMP-2 osteogenesis, although these agents are expensive.

In summary, the systemic or local administration of ETN enhanced the bone-inducing capacity of BMP with no apparent adverse systemic effects in vivo.

Acknowledgments

This work was supported in part by Grants-in-Aid from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (Project Grants 16109009 and 1679085 to KT, and 19791018 to YI). We extend our appreciation to Ms K. Kamei, A. Inagaki, K. Hata, and Y. Hanamoto for their technical assistance.

Copyright information

© The Japanese Society for Bone and Mineral Research and Springer 2009