Calcif Tissue Int (2014) 95:349–361 DOI 10.1007/s00223-014-9900-5
ORIGINAL RESEARCH
Nrf2 Deficiency Impairs Fracture Healing in Mice Sebastian Lippross • Rainer Beckmann • Nadine Streubesand • Ferda Ayub • Mersedeh Tohidnezhad • Graeme Campbell • Yuet Wai Kan • Fischer Horst • Tolga Taha So¨nmez • Deike Varoga • Philipp Lichte • Holger Jahr • Thomas Pufe • Christoph Jan Wruck
Received: 19 March 2014 / Accepted: 17 July 2014 / Published online: 6 August 2014 Ó Springer Science+Business Media New York 2014
Abstract Oxidative stress plays an important role in wound healing but data relating oxidative stress to fracture healing are scarce. Nuclear factor erythroid 2-related factor 2 (Nrf2) is the major transcription factor that controls the cellular defence essential to combat oxidative stress by regulating the expression of antioxidative enzymes. This study examined the impact of Nrf2 on fracture healing using a standard closed femoral shaft fracture model in wild-type (WT) and Nrf2-knockout (Nrf2-KO)-mice. Healing was evaluated by histology, real-time RT-PCR, lCT and biomechanical measurements. We showed that Nrf2 expression is activated during fracture healing. Bone healing and remodelling were retarded in the Nrf2-KO compared to the WT-mice. Nrf2-KO-mice developed significantly less callus tissue compared to WT-mice. In addition, biomechanical testing demonstrated lower strength against shear stress in
Sebastian Lippross, Rainer Beckmann, Thomas Pufe, and Christoph Jan Wruck, have contributed equally to this work. All authors state that they have no conflicts of interest. S. Lippross (&) N. Streubesand F. Ayub D. Varoga Department of Trauma Surgery, University Medical Center of Schleswig–Holstein, Campus Kiel, Kiel, Germany e-mail: [email protected] R. Beckmann M. Tohidnezhad T. Pufe C. J. Wruck Department of Anatomy and Cell Biology, Medical Faculty, RWTH Aachen University, Aachen, Germany G. Campbell Section Biomedical Imaging, Department of Diagnostic Radiology, University Hospital Schleswig–Holstein, Campus Kiel, Kiel, Germany Y. W. Kan Department of Laboratory Medicine, University of California, San Francisco, CA, USA
the Nrf2-KO-group compared to WT. The expression of vascular endothelial growth factor (VEGF) and osteocalcin is reduced during fracture healing in Nrf2-KO-mice. Taken together, our results demonstrate that Nrf2 deficiency in mice results in impaired fracture healing suggesting that Nrf2 plays an essential role in bone regeneration. Pharmacological activation of Nrf2 may have therapeutic potential for the enhancement of fracture healing. Keywords mouse
Nrf2 Fracture healing Bone Knock out
Introduction After fracture of long bones a complex cascade of pathophysiological reactions follows the event of discontinuation of the weight-bearing skeleton. The bony injury is accompanied by disruption of blood vessels and soft tissue damage leading to tissue ischemia and acidosis. Damaged F. Horst Department of Dental Materials and Biomaterials Research, RWTH Aachen University Hospital, Aachen, Germany T. T. So¨nmez Department of Cranio-Maxillofacial Surgery, RWTH Aachen University Hospital, Aachen, Germany P. Lichte Department of of Orthopaedic Trauma Surgery, RWTH Aachen University Hospital, Aachen, Germany H. Jahr Department of Orthopaedic Surgery, RWTH Aachen University Hospital, Aachen, Germany
123
350
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
blood vessels cannot supply regeneratory cells with nutrients and oxygen. Within the wound a blood clot is formed, which is covered by platelets and integrated into a fibrin mesh that provides a static base for migrating macrophages, leukocytes and mast cells. Subsequently, regeneratory cells (endothelial cells, smooth muscle cells and pluripotent mesenchymal stem cells) initiate bone repair by forming granulation tissue and differentiation into fibroblasts, chondroblasts and osteoblasts [1]. Toxic metabolites like reactive oxygen species (ROS) are produced immediately after the damaging event causing lipid peroxidation that results in the production of malondialdehyde (MDS) and cell death via DNA breakage [2, 3]. In addition, a number of studies have demonstrated the importance of physiologic antioxidants for the prevention of fractures [4]. Despite these events, fracture healing occurs successfully in most cases. The protective mechanisms that prevent disturbances in fracture healing, especially by toxic ROS, have not been fully elucidated so far. To counteract damage by ROS, animals have developed a complex defence system consisting of detoxification enzymes and antioxidant proteins. In this system the transcription factor Nuclear factor erythroid-derived 2 like 2 (Nrf2) has been identified as a key transcription factor that controls expression of many antioxidant and detoxifying enzymes. Nrf2 binds the Antioxidant Response Element (ARE) in the promoter region of genes coding for antioxidative enzymes [5]. This Nrf2-ARE pathway acts as a master regulator of cellular protection [6, 7]. With respect to fracture repair that involves callus formation as a critical step towards solid bony consolidation, upregulation of Nrf2 was demonstrated in multipotent mesenchymal stem cells (MSC) after adrenaline exposure [8]. Thereby the vulnerability of chondrogenic and osteogenic precursor cells to ROS could be reduced. A recent investigation on the effect of Nrf2 on osteoblasts and osteoclasts further demonstrates that Nrf2 might be a key factor regulating bone acquisition via interfering with both osteoblast and osteoclast differentiation [9]. Because oxidative stress can disturb the complex pathophysiological changes after skeletal injury; we hypothesized that Nrf2 is essential for fracture healing. A standard femur shaft fracture model was used to evaluate callus formation of the Nrf2-KO mouse by micro computed tomography (lCT), histology, immunohistochemistry, molecular biological and biochemical analysis.
specimens of the femoral head from patients suffering from avascular necrosis. The study was approved by the institutional review board. Animals Nrf2-knockout (Nrf2-KO) mice were produced by specifically deleting the Nrf2 gene segment [10]. Wild type (WT) control mice were littermates of the Nrf2-KO mice. C57BL6/J(ARE-luc) mice were purchased from Cgene (Oslo, Norway). All mice used in this study were 10- to 14-weeks old, gender-matched and maintained in our animal facilities under special, pathogen-free conditions. They were housed up to five per cage, with 12:12-h light:dark cycles and access to mouse chow and water ad libitum. Mice were obtained from our breeding colony after approval by our institutional animal studies committee (No. V312-72241.121-9). Femoral Fracture Model Intraperitoneal anaesthesia was achieved by a combination of 0.05 mg/kg Fentanyl, 5 mg/kg Midazolam and 0.5 mg/ kg Medetomidin and antagonized by 1.2 mg/kg Naloxon, 0.5 mg/kg Flumazenil and 2.5 mg/kg Atipamezol. A closed mid-diaphyseal transverse fracture was created in the right femur by three-point bending after insertion of a rod into the medullary canal as described by Bonnarens [11]. Postoperative analgesia of 50–100 mg/kg Metamizol was administered via drinking water. Mice were sacrificed 14 and 28 days after fracture by CO2 asphyxiation. Bones were dissected, fixed in 4 % paraformaldehyde and stored in 70 % Ethanol. Histology Decalcified histology was performed. After lCT scanning, tissue samples were decalcified in 12.5 % EDTA (Schweizerhall Chemie AG, Basel Switzerland) with 1.25 % Sodiumhydroxide (Fluka) as confirmed by radiography and embedded in paraffin. Sections of 4–6 lm were cut longitudinally through the center of the medullary canal using a Leica microtome. They were placed onto Histobond microscope slides (Marienfeld, Germany) and left overnight at 38 °C. Sections were routinely stained with Haematoxylin-Eosin.
Materials and Methods
Immunohistochemistry
Human Bone Tissues
Immunohistochemistry was performed with the following primary antibody: anti-Nrf2 (dilution 1:50, clone 383727; R&D Systems, Minneapolis, MN, USA), 4-hydroxynonoenal (dilution 1:300, abam 46545) and Ki67 clone MIB-5
Bone tissue was derived from the Department of Orthopaedic Surgery. It was obtained from healthy areas of
123
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
351
Fig. 1 Influence of Nrf2 on mechanical stability and connectivity of the corticalis. a The connectivity of the bone lamella of the corticalis is reduced in Nrf2 KO mice (black arrows), haematoxylin eosin staining. b Nrf2 knock out leads to a significant reduction of the force to failure also the vulnerability against bending stress is significantly
reduced in Nrf2 KO mice in comparison to WT mice. Also the resistance against bending stress is significantly reduced in Nrf2 KO mice. Shown are the mean (± SEM), n = 4–9. Statistical analysis was performed using ANOVA * p B 0.05
(dilution 1:30, DAKO M7248). The antibody was omitted in the negative control experiments. We detected primary antibody binding using a streptavidin–biotin sequence with 3-amino-9-ethyl-carbazol (Sigma) as chromogen.
10 rad/s. This resulted in a time to fracture of approximately 0.02 s for a nominal 10’’ deformation. Values were obtained for angular deformation, and torsional strength, stiffness.
Biomechanical Testing
Micro Computed Tomography
Biomechanical testing was performed using a custom-built direct current (DC) motor servoactuated torsion testing device with a deformation rate of approximately 550’’ or
Mineralized tissue formation was assessed by micro computed tomography. The specimens were removed from 70 % ethanol and scanned using a Novotec MicroScope
123
352
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
(Novotec Medical GmbH, Pforzheim) with an isotropic nominal spatial resolution (voxel size) of 15–20 lm. Image analysis was performed using Image-J software (rsb. info.nih.gov/ij). The volume of interest (VOI) consisted of a fixed cylinder encompassing the entire volume of the fracture callus. The greyscale images were thresholded to obtain binary images consisting only of bone and background voxels. The bone voxels were defined as those with greater than 400 Houndsfield units for all measurements. Within the VOI, the bone volume to total volume (BV/TV) ratio was calculated as the number of bone voxels divided by the total number of voxels in the image.
Fig. 2 Nrf2- and target gene activity in healthy bone remodelling and c during fracture healing is increased. a Investigation of human bone revealed Nrf2 positive staining in activated lining cells (arrowhead) and active osteoblasts (white arrow), but not in osteocytes (black arrow). b Osteoclasts were also positive for Nrf2. c ARE-activity during fracture healing increased during the first week and declined to basal level on day 16. Shown are the mean (± SD) of luciferase activity normalized to the contralateral healthy limb of the mice. Statistical analysis was performed using ANOVA * p B 0.05; n = 8. d The right limb of C57BL6/J(ARE-luc) mice was fractured and ARE-activity was measured as described in ‘‘Materials and Methods’’. One representative mouse is shown. The colour overlay on the image represents Radiance (photons/s/cm2/sr). e ARE-activity of the fracture callus (white arrow) and the contralateral non-fractured femur 6 days after fracturing The colour overlay on the image represents Radiance (photons/s/cm2/sr). f Nrf2-mRNA is upregulated 2 and 4 weeks after fracturing in soft tissue of the fractured area. n = 3–7. g The expression of the Nrf2 target gene HO-1 in Nrf2 WT mice is significantly induced after 2 weeks and decreased after 4 weeks to basal level. n = 3–4. E and F show mean (±SD), statistical analysis were performed using ANOVA * p B 0.05. h Nrf2 immunohistochemistry revealed no staining of osteocytes. i In chondrocytes of chondral callus from fractured femora from Nrf2WT mice a strong staining was detectable with decreased gradient to hypertrophy chondrocytes.; nuclei were counterstained using hemalum
In Vivo Nrf2-Activity Imaging C57BL6/J(ARE-luc) animals received intraperitoneal injections of luciferin (150 mg/kg) 10 min before imaging. The animals were anesthetized (1–3 % isoflurane) and placed into a light-proof camera box. Images were taken during 10 min in postero-anterior direction with a field of view of 10 cm. The ARE-activity of each animal where measured throughout the experiment. Light emitted from the transgene mice was detected by the CCD camera of the IVIS Imaging System 100 Series (Xenogen Corporation, Alameda, CA). Signal intensity is shown in false-colours. Signals are quantified and archived using the Living Image software. For this study, photons were quantified from the hind leg from the dorsal images using fitted regions of interest (ROI). 2 ROIs were set: one over the fracture area, one at the same area over the contralateral healthy limb. The results are represented as x-fold of the signal of the fractured site compared to the healthy contralateral area. Through imaging and measuring of a ROI on the integrating sphere, counts detected by the CCD camera digitizer can by converted into physical units of radiance in photons/15 min. To calibrate the xenogen imaging device, background signals were evaluated daily and the Living Image software performed background subtraction calculations.
CA, USA). All reagents for qRT-PCR analysis were purchased by ABIsystems, (iScrip cDNA Synthesis Kit, iQ SYBR Green Supermix) and plate assays were analysed by Icycler IQ5 optical system software V 2.0 (Bio-Rad, Paris, France). The methods of DNA amplification were conducted as previously described [12]. In qRT-PCR, all samples were run in duplicates. Each plate contained two negative controls and a positive control. Expression levels were normalized to 18-S ribosomal RNA. Statistical Analysis Differences between different time points were evaluated using ANOVA. The analysis was performed by GraphPad Prism 5 (GraphPad Software Inc., San Diego, CA, USA). Differences were considered statistically significant if p B 0.05.
mRNA Preparation and Expression Analysis Results Collected bone and near-to-bone soft tissue of healthy bone and the fracture site of 5 mice per time point were snapfrozen in liquid nitrogen and stored at -80°. Control samples (day 0 time point) were collected from the middle half of unfractured femora. Tissues were pulverized using an agate mortar under continuous cooling with liquid nitrogen. After homogenisation (Precellys, peqlab), total RNA was isolated from each sample using TRIzol (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s protocol. DNA was eliminated by DNase I digestion (Invitrogen Life Technologies, Carlsbad,
123
Nrf2 Affects Basal Bone Phenotype Histological analysis showed differences of the bone. Corticalis of Nrf2-KO mice were more inhomogenous resulting in more clefts between the bone lamellas after sectioning (Fig. 1a). This observation was reflected in differences of the biomechanical properties of the bone. Nrf2-KO mice show a significant decreased force to failure and resistance again bending stress compared to WT mice (Fig. 1b).
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
353
123
354
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
Nrf2 is Activated in Osteoblasts and Osteoclasts of Human Healthy Bone
antibody against 4–hydroxy-nonenal (HNE). Unfractured bone from Nrf2-WT and -KO mice showed only few positive cells. Cartilaginous callus from Nrf2-WT mice showed lipid peroxidation (Fig. 3a). Genetic disruption of Nrf2 obviously increased the macromolecular oxidative damage in the cartilaginous callus compared to Nrf2-WT mice. To assess the proliferation rate of fibroblasts and chondrocytes of the different callus we used Ki67/Mip5Immunostaining. Nrf2-KO mice showed less proliferating cells after 2 weeks then WT mice. In contrast Nrf2-KO mice had more proliferating cells compared to WT mice after 4 weeks. (Fig. 3b).
Bone samples of healthy areas of human femoral head obtained from patients suffering from avascular necrosis were examined for expression of Nrf2 using anti-Nrf2 immunohistochemistry. Nrf2-expression was found in osteoblasts (Fig. 2a) and osteoclasts (Fig. 2b) but not in osteocytes (Fig. 2a, b). Immunohistochemistry without primary antibody revealed no marked immunoreactivity. Nrf2 is Activated in the Femur of Mice During Fracture Healing First we used ARE-luciferase mice to measure the Nrf2activity in a longitudinal study spanning the entire process of fracture healing. Unfractured legs with inserted rod showed only weak and sparse luciferase activity over the background. Six days after fracture, the luciferase activity detected in the fractured femora rose significantly compared to control conditions (pre OP). The luciferase activity was significantly increased until day 6, and decreased to basal level on day 16 (Fig. 2c, d). The time-dependent progressions of all measurements are graphically displayed in Fig. 2c. ARE-activity could be shown at the fracture callus (Fig. 2e, white arrow) 6 days after fracturing. Next we investigated the amount of Nrf2 on mRNA-level during fracture healing. Figure 2f presents quantitative realtime RT PCR data of Nrf2-mRNA expression of near-tobone soft tissue of unfractured bone compared to near-tobone soft tissue around the fracture callus tissue 2 and 4 weeks after fracture. Two weeks after fracture, the amount of Nrf2-mRNA rose 230-fold compared to unfractured bone. Four weeks after fracture, the Nrf2-mRNA decreased compared to the expression level after 2 weeks (Fig. 2f). The expression of heme oxygenase 1 (HO-1), a target gene of Nrf2, also significantly increased 15 fold after 2 weeks and declined to basal level after 4 weeks (Fig. 2g). Immunohistochemical analysis of unfractured bone and callus tissue of WT mice were examined for Nrf2 protein expression using anti-Nrf2 antibody. We also showed that the Nrf2-protein expression was up-regulated after fracture since Nrf2 could be stained in chondrocytes of the cartilaginous callus after 2 weeks (Fig. 2i), but not in osteocytes of healthy bone (Fig. 2h).
Cartilaginous Callus of Nrf2-KO Mice Suffers From Increased Oxidative Damage One consequence of oxidative stress is membrane lipid peroxidation. Therefore, we stained slices of mouse bone and cartilaginous callus by immunohistochemistry utilizing an
123
Fractured Nrf2-KO Mice Undergo Retarded Bone Formation and Healing lCT analysis demonstrated imperfect fracture healing in Nrf2-KO mice compared to WT mice (Fig. 4). 2 and 4 weeks after fracture, the Nrf2-KO mice had significant decreased callus volume compared to WT mice during the healing prozess (Fig. 4b). Analysis of the lCT data revealed that fracture callus from Nrf2-KO mice had a higher bone volume–total volume ratio (BV/TV) than Nrf2-WT mice after 2 weeks which slightly increases after 4 weeks of healing, while BV/TV of Nrf2 WT mice slightly decreases during 4 weeks of fracture healing (Fig. 4c). There were no differences in trabecular and cortical bone within the callus of Nrf2-WT mice compared to Nrf2-KO mice after 2 weeks. After 4 weeks the trabecular and the cortical bone fraction was slightly lower compared to Nrf2 KO mice (Fig. 4c). Impaired Bone Formation and Callus Remodelling During Fracture Healing in Nrf2-KO Mice Histological characterisation revealed that bone formation and the final remodelling of the bony callus were reduced in the Nrf2-KO compared to the WT mice. HE staining show a decreased fracture callus in Nrf2-KO mice compared to WT (Fig. 5). Particularly, the amount of cartilaginous callus was decreased in Nrf2-KO. The quality of the remodelled bone in these mice was more immature than in WT mice. After 4 weeks, fractures in the WT-mice were healed and cartilage tissue was essentially absent in the callus. In contrast, Nrf2-KO mice healed incompletely during 4 weeks. Bones of WT mice almost returned to normal shape compared to the persisting callus of the bones of Nrf2-KO mice. Nrf2-KO Affects the Biomechanical Properties of the Fracture Callus A crucial aspect of fracture healing is sufficient loadability of the regenerated tissue of the injured limb in order to
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
355
Fig. 3 Nrf2-KO enhances oxidative stress damage in bone during fracture healing and proliferation is decreased in Nrf2-KO. a Bone of Nrf2-WT and -KO mice were fractured as described in ‘‘Materials and Methods’’. Mice were sacrificed and bone and callus specimens prepared for histological evaluation of 4-hydroxynonenal (HNE) immunohistochemistry of representative sections of untreated bone and callus 2 and 4 weeks after fracturing were shown. Unfractured bone from Nrf2-WT and -KO mice showed only few positive cells.
Cartilaginous callus from Nrf2-WT mice showed increased evidence of lipid peroxidation. Genetic disruption of Nrf2 obviously increased the macromolecular oxidative damage in the cartilaginous callus compared to Nrf2-WT mice. b Immunostaining against Ki67 revealed a decreased immunreactivity in Nrf2–KO mice after 2 weeks compared to WT mice. After 4 weeks there are more Ki67 positive cells compared to WT mice. Nuclei were counterstained with hemalum
regain function. To evaluate differences in strength and stability of an injured bone we performed biomechanical tests (Fig. 6). We chose the distraction-to-failure model as a mode of testing because the geometry of a non-stabilized fracture is highly variable, which renders more standard biomechanical tests less accurate. WT- and Nrf2-KO callus
were subjected to a gradual distractive force and the maximum force required to cause failure of the callus was determined. The resistance against shear stress significantly improved during fracture healing in WT-mice while the Nrf2-KO mice rest at the same level during 4 weeks of healing (Fig. 6a). The torsional rigidity was not affected by
123
356
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
Fig. 4 Reduced callus volume in Nrf2-KO. lCt measurements of fractured bone of Nrf2-KO and -WT mice. a Representive lCT slices of a fractured limb of Nrf2-WT and -KO mice and 3D reconstruction of an Nrf2-WT femur 4 weeks after fracturing. b Callus volume,
c bone volume/total volume, d volume of the trabecular bone and e the cortical bone was calculated 2 and 4 weeks after fracturing. Shown are the mean (±SEM), n = 8. Statistical analysis were performed using ANOVA * p B 0.05.; ns not significant
the Nrf2-KO (Fig. 6b). Indicating that Nrf2-KO callus was structurally weaker than WT callus after 4 weeks.
regulated osteocalcin only 3-fold compared to WT control and rested slightly elevated after 4 weeks (Fig. 7b).
Level of VEGF and Osteocalcin in the Fracture Callus Discussion Expression of VEGF-protein in soft tissue after fracture was measured by ELISA technique. 2 and 4 weeks after fracture VEGF protein expression is upregulated in Nrf2WT while Nrf2-KO mice totally failed to upregulate VEGF after fracture (Fig. 7a). Expression of osteocalcin in mineralized bone tissue was analysed by real-time RT-PCR and normalized to the expression of untreated Nrf2-WT t0 samples. Osteocalcin was significantly upregulated (12-fold) in tissue of Nrf2WT mice 2 weeks after fracture and relapsed back to basal level after 4 weeks. In contrast, Nrf2-KO mice up
123
Fracture healing is a complex process that is poorly understood. Every surgeon remembers cases of patients with impaired bone healing. Many patients ultimately lose a limb after a long period of surgical and conservative attempts to treat the disease. The clinical scenario calls for specific treatment options i.e. the application of osteoinductive and osteoconductive substances. With respect to such therapeutic intentions, the need for more detailed knowledge of the molecular regulating mechanisms becomes most evident. Many studies have dealt with
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
357
Fig. 5 Reduced callus area in Nrf2-KO. Representative bone sections of healthy and fractured WT and Nrf2-KO mice stained with haematoxylin and eosin. The increased fracture callus observed by lCt analysis mainly consist of cartilage. Especially the amount of
cartilaginous callus was more increased in KO mice compared to WT. 4 weeks after fracture the quality of the remodelled bone was more immature in Nrf2-KO mice compared to WT mice (black arrow)
predisposing factors for impaired bone development revealing several gene defects for example osteogenesis imperfecta [13]. In contrast, there are few reports on genetic mechanisms of impaired bone healing. Previous work in our laboratory provided first evidence of a role of Nrf2 in bone remodelling. We could show an increased number of spontaneously microfractured bones
in a mouse model of rheumatoid arthritis in Nrf2-KO mice [14]. To investigate the phenotypic differences in bone formation in Nrf2-KO mice biomechanical and histological experiments were performed that demonstrated significant differences. Histologic analysis revealed brittle cortical bone in Nrf2-KO Mice (Fig. 1a). The force to failure as well as the resistance against bending stress are reduced in
123
358
Fig. 6 No detectable increase of biomechanical properties during fracture healing in Nrf2-KO mice. Biomechanical testing of murine Nrf2-KO and -WT femora 2 and 4 weeks after fracturing. The tests were performed as described in ‘‘Materials and methods’’. Values were obtained for shear stress (a), and torsional rigidity (b). Shown are the mean (±SD); n = 4–8. Statistical analysis was performed using ANOVA *: p B 0.05
Nrf2-KO (Fig. 1b). Nrf2 negatively regulates osteoblast maturation and activity while Nrf2 deficiency leads to increased differentiation of osteoclasts as well as to enhanced activity in vitro, but coculture of osteoblasts from Nrf2-KO mice with bone marrow derived macrophages lead to decreased osteoclastic differentiation [9]. Park et al. [9] could demonstrate in vivo an increased mineral apposition rate in Nrf2-KO mice with an increase of osteoblasts per bone perimeter and osteoblast surface per bone while the number of osteoclasts per bone perimeter did not differ between KO and WT mice. Bone remodelling is a well orchestrated permanently ongoing process in order to maintain the bone. We here provide a rationale for Nrf2 influencing this process. The enhanced bone apposition with reduced bone resorption may lead to inferior bone quality as supported by reduced biomechanical properties in KO mice (Fig. 1). Based on the differences in the basal bone phenotype of Nrf2-KO mice we address the role of Nrf2 in fracture healing. Immunohistochemical staining of Nrf2 in fractured bone and callus showed Nrf2-expression in chondrocytes of the cartilaginous callus (Fig. 2i), but not in osteocytes of healthy bone (Fig. 2h). Since unstressed cells
123
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
Fig. 7 VEGF expression in soft tissue and osteocalcin in mineralized bone tissue is not increased in Nrf2-KO. a Expression of VEGFprotein in soft tissue of fractured femora measured by ELISA. b Expression of osteocalcin in mineralized bone tissue. Expression of osteocalcin was shown as fold induction of untreated Nrf2-WT. Statistical analysis was performed with an ANOVA * p B 0.05; n = 3–7
have a very low steady state protein-level of Nrf2, these results indicate that an induction of Nrf2 occurs during bone remodelling. This observation is in accordance with our finding that Nrf2 is expressed in cells actively involved in bone remodelling: osteoblasts and osteoclasts (Fig. 2a, b). The Nrf2 activity was indirectly measured by AREdriven luciferase reportergen mice, by this we could demonstrate an increasing Nrf2-activity in the fractured limb during the first 6 days (Fig. 2c, d). To localize the signal to the fracture callus we measured the ARE-activity at day 6 of a fleshed femur (Fig. 2e). On the other hand we show that Nrf2 expression is up-regulated in soft tissue around the fractured bone. We added a mechanistic confirmation by the detection the target gene HO-1 (Fig. 1f, g). Noticeably the expression of Nrf2 and HO-1 peaks around day 14 post fracture whereas the Nrf2 activity seems greatest after 1 week. A possible explanation is that AREdriven luciferase can be a detectable effect in the course of the initial inflammatory response to fracture whereas the expression of Nrf2 and HO-1 in the surrounding tissue may well be part of a later but important regenerative reaction. In conclusion, we claim that Nrf2 is activated in the callus
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
359
of the regenerating bone and in the surrounding soft tissue during fracture healing. Since Nrf2 is the major transcription factor that regulates the detoxification of ROS we estimated the extent of oxidative stress level during fracture healing. Because of the highly reactive character of ROS, it is difficult to directly demonstrate their presence in vivo. It is more appropriate to measure the ‘footprints’ of ROS, e.g. their effects on various lipids, proteins, and nucleic acids. Therefore, we used immunohistochemistry against 4-hydroxy-2-nonenal (HNE) a specific product of lipid peroxidation, which we detected during fracture healing in the callus. The oxidative stress level was higher in Nrf2KO mice compared to WT-mice. (Fig. 3a). Arai et al. [15] demonstrated a decreased expression of osteoanbolic markers like the osteoanabolic key transcription factor Runx2, alkaline phosphatase and bone sialoprotein after H2O2 exposure of osteoblast like cells MC3T3-E1 in the presence of ROS. On the other hand a moderate level of ROS is required for chondrocyte survival and differentiation in early stages of chondrogenesis [16]. However, high levels of ROS lead to inhibition of proliferation and to hypertrophy of chondrocytes [17]. Nrf2 is also a negative regulator of chondrogenic proliferation and maturation. Increased oxidative stress leads to reduced differentiation of chondrogenic progenitor cells and in accordance with the reduced callus volume observed in the Nrf2-KO mice after fracturing. In addition, the delayed healing process could be referred to the reduced differentiation caused by the lack of Nrf2. Hinoi et al. [19] could demonstrate a decreased expression of chondrogenic differentiation markers like collagen type II, type X and osteopontin in Nrf2 overexpressing ATDC5 cells. Nrf2-KO mouse shows higher levels of oxidative stress, which inhibits the proliferation and leads in combination with the Nrf2-KO earlier to a hypertrophic phenotype. The effect of Nrf2 of bone is controversial discussed in literature. Hinoi and colleagues reported a negative effect of Nrf2 on osteoblast and chondrocyte differentiation in vitro. They showed an Nrf2-dependent inhibition of a RUNX2-dependent osteocalcin promoter activity [18, 19]. In contrast, Arai et al. [15] showed an increased osteocalcin and Nrf2 expression after ROS exposure in the same cell line. Related to these findings, we showed here that osteocalcin-mRNA is significantly decreased in injured tissue of Nrf2-KO animals after 2 weeks compared to WT animals, indicating that Nrf2 binds to the osteocalcin promoter and up-regulates the osteocalcin expression in vivo (Fig. 7b). Our data are supported by recently published data demonstrating that osteocalcin is Nrf2-regulated under arthritic conditions [20]. Osteocalcin is involved in bone mineralization and calcium ion homeostasis. It is used as a marker for the bone formation process and higher serum-
osteocalcin levels are correlated with an increase in bone mineral density. Osteocalcin is also discussed as a crucial factor to stimulate bone mineral maturation, which could be a hint for the difference in the cortical structure in healthy Nrf2-KO mice (Fig. 5b) [21]. Another aspect is that Nrf2 is also involved in osteoblast differentiation by inorganic phosphate [22]. Park et al. reported an inhibitory effect of Nrf2 on osteoblast activity by siRNA-experiments and an additional inhibitory effect on osteoclast maturation in coculture experiments with osteoblasts from Nrf2-KO mice in vivo. In vitro they could demonstrate an increased mineral apposition rate and bone formation rate in KO-mice [9]. The negative effect on osteoclasts may lead to decreased resorption of the cartilage callus but it could also influence the bone remodelling so that the depletion of Nrf2 may lead to an immature bone of lower quality after fracture healing. Consistent with our recent finding that Nrf2 is a critical factor for VEGF-expression [23], we showed here that Nrf2-KO mice were not able to up-regulate VEGF-protein expression in soft tissue after fracturing (Fig. 7b). VEGF has an important osteoanabolic capacity by triggering osteoblast proliferation and migration as well as osteoblast differentiation. By doing this, VEGF could influence the quality of the bone remodelling [24, 25]. The classical function of VEGF to induce neovascularisation of the fracture callus is an important parameter influencing the healing process. VEGF is the most important growth factor initiating the regeneration of the vascular system at the fracture site [26–28]. The increased oxidative stress, as well as the decreased osteocalcin expression and the failure of VEGF induction in Nrf2-KO mice negatively influence remodelling of bone. lCT analysis (Fig. 4) and histological observations (Fig. 5) of fracture callus volume demonstrate a decrease of callus volume in the Nrf2-KO mice in comparison to the WT littermates. During the healing process the ratio between bone fraction and total callus volume decreases. Although the callus volume in Nrf2-KO mice remained constant between 2 and 4 weeks, the BV/TV slightly increased. The general increased BV/TV might be caused by the general bone phenotype. Park et al. [9] could demonstrate a increased BV/TV ratio in Nrf2-KO mice in comparison to WT but the missing reduction during fracture healing indicates a delayed bone formation. The fraction of trabecular and cortical bone of the callus is similar in WT and Nrf2-KO mice after 2 weeks and further decreased slightly after 4 weeks (Figs. 3e and 4d). In the murine fracture model the remodelling process is nearly finished after 4 weeks and the original structure of trabecular and cortical bone is restored. In Nrf2-KO mice this seem to be retarded. The trabecular and cortical bone fraction reduces during time but are slightly increased after 4 weeks compared to
123
360
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
WT-mice. This is also visible in the more immature bon in Nrf2-KO after 4 weeks in HE-stained slices (Fig. 5d) and the persisting callus of the bones of Nrf2-KO mices (Fig. 5c). The differences in callus structure between Nrf2-KO and -WT are also reflected in the biomechanical properties. While torsional rigidity seems unaffected by the Nrf2-KO, Nrf2 influence the resistance against shear stress. The Nrf2KO callus does not further improve until week 4 (Fig. 6a, b). Since shear stress is inversely related to the cross-sectional area of the callus the reduced callus volume after 4 weeks in Nrf2-KO mice may be one reason for the lower resistance against shear stress [29]. The exposure to high oxidative stress in Nrf2-KO also influences the callus volume and quality by reduced proliferation and maturation of chondroblasts. Also further healing processes are affected by the Nrf2 depletion. The reduced callus size in Nrf2-KO mice and bone quality may be caused by the retarded bone remodelling process by the inhibitory effect of the Nrf2 depletion on osteoclast formation. Also the lack of VEGF induction in Nrf2-KO mice has an negative impact on the healing process. VEGF regulates the vascularisation and therefore the conversion of the cartilage callus to bony callus but it have also an direct impact of the bone quality. VEGF positively influences the osteoblast maturation and activity. Also osteocalcin a marker of bone remodelling implicated in bone mineralization and calcium ion homeostasis is decreased in Nrf2 KO mice and contributes to the retarded bone remodelling in Nrf2-KO mice after fracture healing.
References
Conclusions Taken together, our results show that the transcription factor Nrf2 is a positive regulator of fracture healing, possibly by influencing chondrocytic differentiation within the fracture callus whereas the formation of cortical bone appears unaffected. Acknowledgments We thank S. Echterhagen, C. Jaeschke, M. Nicolau, A. Ru¨ben and L. Shen for their excellent technical assistance. This work has been supported in part by the ChristianAlbrecht-University of Kiel and by a Grant from the AO Germany to S.L. This research project was also in part supported by the Excellence Initiative of the German federal and state governments, the START-Program and by the Interdisciplinary Centre for Clinical Research within the faculty of Medicine at the RWTH Aachen University (T9-3; T9-5; T11-3). This work was in part supported by the DFG (PU 214/5-2, 4-2, 3-2). Human and Animal Rights and Informed Consent All experiments were in keeping with the local ethics committees guidelines and permission was obtained.
123
1. Einhorn TA (1998) The cell and molecular biology of fracture healing. Clin Orthop Relat Res:S7–21 2. Turgut A, Gokturk E, Kose N, Kacmaz M, Ozturk HS, Seber S, Acar S (1999) Oxidant status increased during fracture healing in rats. Acta Orthop Scand 70:487–490 3. Yeler H, Tahtabas F, Candan F (2005) Investigation of oxidative stress during fracture healing in the rats. Cell Biochem Funct 23:137–139 4. Kelpke SS, Reiff D, Prince CW, Thompson JA (2001) Acidic fibroblast growth factor signaling inhibits peroxynitrite-induced death of osteoblasts and osteoblast precursors. J Bone Miner Res 16:1917–1925 5. Lee JM, Li J, Johnson DA, Stein TD, Kraft AD, Calkins MJ, Jakel RJ, Johnson JA (2005) Nrf2, a multi-organ protector? Faseb J 19:1061–1066 6. Wruck CJ, Gotz ME, Herdegen T, Varoga D, Brandenburg LO, Pufe T (2008) Kavalactones protect neural cells against amyloid beta peptide-induced neurotoxicity via extracellular signal-regulated kinase 1/2-dependent nuclear factor erythroid 2-related factor 2 activation. Mol Pharmacol 73:1785–1795 7. Wruck CJ, Claussen M, Fuhrmann G, Romer L, Schulz A, Pufe T, Waetzig V, Peipp M, Herdegen T, Gotz ME (2007) Luteolin protects rat PC12 and C6 cells against MPP ? induced toxicity via an ERK dependent Keap1-Nrf2-ARE pathway. J Neural Transm Suppl: 57–67 8. Takahata Y, Takarada T, Iemata M, Yamamoto T, Nakamura Y, Kodama A, Yoneda Y (2009) Functional expression of beta2 adrenergic receptors responsible for protection against oxidative stress through promotion of glutathione synthesis after Nrf2 upregulation in undifferentiated mesenchymal C3H10T1/2 stem cells. J Cell Physiol 218:268–275 9. Park CK, Lee Y, Kim KH, Lee ZH, Joo M, Kim HH (2014) Nrf2 is a novel regulator of bone acquisition. Bone 10. Chan K, Lu R, Chang JC, Kan YW (1996) NRF2, a member of the NFE2 family of transcription factors, is not essential for murine erythropoiesis, growth, and development. Proc Natl Acad Sci USA 93:13943–13948 11. Bonnarens F, Einhorn TA (1984) Production of a standard closed fracture in laboratory animal bone. J Orthop Res 2:97–101 12. Brandenburg LO, Kipp M, Lucius R, Pufe T, Wruck CJ (2010) Sulforaphane suppresses LPS-induced inflammation in primary rat microglia. Inflamm Res 59:443–450 13. Ben Amor IM, Glorieux FH, Rauch F (2011) Genotype-phenotype correlations in autosomal dominant osteogenesis imperfecta. J Osteoporos 2011:540178 14. Wruck CJ, Fragoulis A, Gurzynski A, Brandenburg LO, Kan YW, Chan K, Hassenpflug J, Freitag-Wolf S, Varoga D, Lippross S, Pufe T (2011) Role of oxidative stress in rheumatoid arthritis: insights from the Nrf2-knockout mice. Ann Rheum Dis 70:844–850 15. Arai M, Shibata Y, Pugdee K, Abiko Y, Ogata Y (2007) Effects of reactive oxygen species (ROS) on antioxidant system and osteoblastic differentiation in MC3T3-E1 cells. IUBMB Life 59:27–33 16. Kim KS, Choi HW, Yoon HE, Kim IY (2010) Reactive oxygen species generated by NADPH Oxidase 2 and 4 are required for chondrogenic differentiation. J Biol Chem 285:40294–40302 17. Morita K, Miyamoto T, Fujita N, Kubota Y, Ito K, Takubo K, Miyamoto K, Ninomiya K, Suzuki T, Iwasaki R, Yagi M, Takaishi H, Toyama Y, Suda T (2007) Reactive oxygen species induce chondrocyte hypertrophy in endochondral ossification. J Exp Med 204:1613–1623
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
361
18. Hinoi E, Fujimori S, Wang L, Hojo H, Uno K, Yoneda Y (2006) Nrf2 negatively regulates osteoblast differentiation via interfering with Runx2-dependent transcriptional activation. J Biol Chem 281:18015–18024 19. Hinoi E, Takarada T, Fujimori S, Wang L, Iemata M, Uno K, Yoneda Y (2007) Nuclear factor E2 p45-related factor 2 negatively regulates chondrogenesis. Bone 40:337–344 20. Maicas N, Ferrandiz ML, Brines R, Ibanez L, Cuadrado A, Koenders MI, Van den Berg WB, Alcaraz MJ (2011) Deficiency of Nrf2 accelerates the effector phase of arthritis and aggravates joint disease. Antioxid Redox Signal 15:889–901 21. Boskey AL, Gadaleta S, Gundberg C, Doty SB, Ducy P, Karsenty G (1998) Fourier transform infrared microspectroscopic analysis of bones of osteocalcin-deficient mice provides insight into the function of osteocalcin. Bone 23:187–196 22. Beck GR, Moran E, Knecht N (2003) Inorganic phosphate regulates multiple genes during osteoblast differentiation, including Nrf2. Exp Cell Res 288:288–300 23. Kweider N, Fragoulis A, Rosen C, Pecks U, Rath W, Pufe T, Wruck CJ (2011) Interplay between vascular endothelial growth factor (VEGF) and nuclear factor erythroid 2-related factor-2 (Nrf2): implications for preeclampsia. J Biol Chem 286: 42863–42872 24. Mayr-Wohlfart U, Waltenberger J, Hausser H, Kessler S, Gunther KP, Dehio C, Puhl W, Brenner RE (2002) Vascular endothelial
growth factor stimulates chemotactic migration of primary human osteoblasts. Bone 30:472–477 Street J, Bao M, deGuzman L, Bunting S, Peale FV, Ferrara N, Steinmetz H, Hoeffel J, Cleland JL, Daugherty A, van Bruggen N, Redmond HP, Carano RAD, Filvaroff EH (2002) Vascular endothelial growth factor stimulates bone repair by promoting angiogenesis and bone turnover. P Natl Acad Sci USA 99: 9656–9661 Keramaris NC, Calori GM, Nikolaou VS, Schemitsch EH, Giannoudis PV (2008) Fracture vascularity and bone healing: a systematic review of the role of VEGF. Injury 39(Suppl 2):S45– S57 Pufe T, Wildemann B, Petersen W, Mentlein R, Raschke M, Schmidmaier G (2002) Quantitative measurement of the splice variants 120 and 164 of the angiogenic peptide vascular endothelial growth factor in the time flow of fracture healing: a study in the rat. Cell Tissue Res 309:387–392 Weiss S, Zimmermann G, Pufe T, Varoga D, Henle P (2009) The systemic angiogenic response during bone healing. Arch Orthop Trauma Surg 129:989–997 Nazarian A, Entezari V, Vartanians V, Muller R, Snyder BD (2009) An improved method to assess torsional properties of rodent long bones. J Biomech 42:1720–1725
25.
26.
27.
28.
29.
123
ORIGINAL RESEARCH
Nrf2 Deficiency Impairs Fracture Healing in Mice Sebastian Lippross • Rainer Beckmann • Nadine Streubesand • Ferda Ayub • Mersedeh Tohidnezhad • Graeme Campbell • Yuet Wai Kan • Fischer Horst • Tolga Taha So¨nmez • Deike Varoga • Philipp Lichte • Holger Jahr • Thomas Pufe • Christoph Jan Wruck
Received: 19 March 2014 / Accepted: 17 July 2014 / Published online: 6 August 2014 Ó Springer Science+Business Media New York 2014
Abstract Oxidative stress plays an important role in wound healing but data relating oxidative stress to fracture healing are scarce. Nuclear factor erythroid 2-related factor 2 (Nrf2) is the major transcription factor that controls the cellular defence essential to combat oxidative stress by regulating the expression of antioxidative enzymes. This study examined the impact of Nrf2 on fracture healing using a standard closed femoral shaft fracture model in wild-type (WT) and Nrf2-knockout (Nrf2-KO)-mice. Healing was evaluated by histology, real-time RT-PCR, lCT and biomechanical measurements. We showed that Nrf2 expression is activated during fracture healing. Bone healing and remodelling were retarded in the Nrf2-KO compared to the WT-mice. Nrf2-KO-mice developed significantly less callus tissue compared to WT-mice. In addition, biomechanical testing demonstrated lower strength against shear stress in
Sebastian Lippross, Rainer Beckmann, Thomas Pufe, and Christoph Jan Wruck, have contributed equally to this work. All authors state that they have no conflicts of interest. S. Lippross (&) N. Streubesand F. Ayub D. Varoga Department of Trauma Surgery, University Medical Center of Schleswig–Holstein, Campus Kiel, Kiel, Germany e-mail: [email protected] R. Beckmann M. Tohidnezhad T. Pufe C. J. Wruck Department of Anatomy and Cell Biology, Medical Faculty, RWTH Aachen University, Aachen, Germany G. Campbell Section Biomedical Imaging, Department of Diagnostic Radiology, University Hospital Schleswig–Holstein, Campus Kiel, Kiel, Germany Y. W. Kan Department of Laboratory Medicine, University of California, San Francisco, CA, USA
the Nrf2-KO-group compared to WT. The expression of vascular endothelial growth factor (VEGF) and osteocalcin is reduced during fracture healing in Nrf2-KO-mice. Taken together, our results demonstrate that Nrf2 deficiency in mice results in impaired fracture healing suggesting that Nrf2 plays an essential role in bone regeneration. Pharmacological activation of Nrf2 may have therapeutic potential for the enhancement of fracture healing. Keywords mouse
Nrf2 Fracture healing Bone Knock out
Introduction After fracture of long bones a complex cascade of pathophysiological reactions follows the event of discontinuation of the weight-bearing skeleton. The bony injury is accompanied by disruption of blood vessels and soft tissue damage leading to tissue ischemia and acidosis. Damaged F. Horst Department of Dental Materials and Biomaterials Research, RWTH Aachen University Hospital, Aachen, Germany T. T. So¨nmez Department of Cranio-Maxillofacial Surgery, RWTH Aachen University Hospital, Aachen, Germany P. Lichte Department of of Orthopaedic Trauma Surgery, RWTH Aachen University Hospital, Aachen, Germany H. Jahr Department of Orthopaedic Surgery, RWTH Aachen University Hospital, Aachen, Germany
123
350
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
blood vessels cannot supply regeneratory cells with nutrients and oxygen. Within the wound a blood clot is formed, which is covered by platelets and integrated into a fibrin mesh that provides a static base for migrating macrophages, leukocytes and mast cells. Subsequently, regeneratory cells (endothelial cells, smooth muscle cells and pluripotent mesenchymal stem cells) initiate bone repair by forming granulation tissue and differentiation into fibroblasts, chondroblasts and osteoblasts [1]. Toxic metabolites like reactive oxygen species (ROS) are produced immediately after the damaging event causing lipid peroxidation that results in the production of malondialdehyde (MDS) and cell death via DNA breakage [2, 3]. In addition, a number of studies have demonstrated the importance of physiologic antioxidants for the prevention of fractures [4]. Despite these events, fracture healing occurs successfully in most cases. The protective mechanisms that prevent disturbances in fracture healing, especially by toxic ROS, have not been fully elucidated so far. To counteract damage by ROS, animals have developed a complex defence system consisting of detoxification enzymes and antioxidant proteins. In this system the transcription factor Nuclear factor erythroid-derived 2 like 2 (Nrf2) has been identified as a key transcription factor that controls expression of many antioxidant and detoxifying enzymes. Nrf2 binds the Antioxidant Response Element (ARE) in the promoter region of genes coding for antioxidative enzymes [5]. This Nrf2-ARE pathway acts as a master regulator of cellular protection [6, 7]. With respect to fracture repair that involves callus formation as a critical step towards solid bony consolidation, upregulation of Nrf2 was demonstrated in multipotent mesenchymal stem cells (MSC) after adrenaline exposure [8]. Thereby the vulnerability of chondrogenic and osteogenic precursor cells to ROS could be reduced. A recent investigation on the effect of Nrf2 on osteoblasts and osteoclasts further demonstrates that Nrf2 might be a key factor regulating bone acquisition via interfering with both osteoblast and osteoclast differentiation [9]. Because oxidative stress can disturb the complex pathophysiological changes after skeletal injury; we hypothesized that Nrf2 is essential for fracture healing. A standard femur shaft fracture model was used to evaluate callus formation of the Nrf2-KO mouse by micro computed tomography (lCT), histology, immunohistochemistry, molecular biological and biochemical analysis.
specimens of the femoral head from patients suffering from avascular necrosis. The study was approved by the institutional review board. Animals Nrf2-knockout (Nrf2-KO) mice were produced by specifically deleting the Nrf2 gene segment [10]. Wild type (WT) control mice were littermates of the Nrf2-KO mice. C57BL6/J(ARE-luc) mice were purchased from Cgene (Oslo, Norway). All mice used in this study were 10- to 14-weeks old, gender-matched and maintained in our animal facilities under special, pathogen-free conditions. They were housed up to five per cage, with 12:12-h light:dark cycles and access to mouse chow and water ad libitum. Mice were obtained from our breeding colony after approval by our institutional animal studies committee (No. V312-72241.121-9). Femoral Fracture Model Intraperitoneal anaesthesia was achieved by a combination of 0.05 mg/kg Fentanyl, 5 mg/kg Midazolam and 0.5 mg/ kg Medetomidin and antagonized by 1.2 mg/kg Naloxon, 0.5 mg/kg Flumazenil and 2.5 mg/kg Atipamezol. A closed mid-diaphyseal transverse fracture was created in the right femur by three-point bending after insertion of a rod into the medullary canal as described by Bonnarens [11]. Postoperative analgesia of 50–100 mg/kg Metamizol was administered via drinking water. Mice were sacrificed 14 and 28 days after fracture by CO2 asphyxiation. Bones were dissected, fixed in 4 % paraformaldehyde and stored in 70 % Ethanol. Histology Decalcified histology was performed. After lCT scanning, tissue samples were decalcified in 12.5 % EDTA (Schweizerhall Chemie AG, Basel Switzerland) with 1.25 % Sodiumhydroxide (Fluka) as confirmed by radiography and embedded in paraffin. Sections of 4–6 lm were cut longitudinally through the center of the medullary canal using a Leica microtome. They were placed onto Histobond microscope slides (Marienfeld, Germany) and left overnight at 38 °C. Sections were routinely stained with Haematoxylin-Eosin.
Materials and Methods
Immunohistochemistry
Human Bone Tissues
Immunohistochemistry was performed with the following primary antibody: anti-Nrf2 (dilution 1:50, clone 383727; R&D Systems, Minneapolis, MN, USA), 4-hydroxynonoenal (dilution 1:300, abam 46545) and Ki67 clone MIB-5
Bone tissue was derived from the Department of Orthopaedic Surgery. It was obtained from healthy areas of
123
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
351
Fig. 1 Influence of Nrf2 on mechanical stability and connectivity of the corticalis. a The connectivity of the bone lamella of the corticalis is reduced in Nrf2 KO mice (black arrows), haematoxylin eosin staining. b Nrf2 knock out leads to a significant reduction of the force to failure also the vulnerability against bending stress is significantly
reduced in Nrf2 KO mice in comparison to WT mice. Also the resistance against bending stress is significantly reduced in Nrf2 KO mice. Shown are the mean (± SEM), n = 4–9. Statistical analysis was performed using ANOVA * p B 0.05
(dilution 1:30, DAKO M7248). The antibody was omitted in the negative control experiments. We detected primary antibody binding using a streptavidin–biotin sequence with 3-amino-9-ethyl-carbazol (Sigma) as chromogen.
10 rad/s. This resulted in a time to fracture of approximately 0.02 s for a nominal 10’’ deformation. Values were obtained for angular deformation, and torsional strength, stiffness.
Biomechanical Testing
Micro Computed Tomography
Biomechanical testing was performed using a custom-built direct current (DC) motor servoactuated torsion testing device with a deformation rate of approximately 550’’ or
Mineralized tissue formation was assessed by micro computed tomography. The specimens were removed from 70 % ethanol and scanned using a Novotec MicroScope
123
352
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
(Novotec Medical GmbH, Pforzheim) with an isotropic nominal spatial resolution (voxel size) of 15–20 lm. Image analysis was performed using Image-J software (rsb. info.nih.gov/ij). The volume of interest (VOI) consisted of a fixed cylinder encompassing the entire volume of the fracture callus. The greyscale images were thresholded to obtain binary images consisting only of bone and background voxels. The bone voxels were defined as those with greater than 400 Houndsfield units for all measurements. Within the VOI, the bone volume to total volume (BV/TV) ratio was calculated as the number of bone voxels divided by the total number of voxels in the image.
Fig. 2 Nrf2- and target gene activity in healthy bone remodelling and c during fracture healing is increased. a Investigation of human bone revealed Nrf2 positive staining in activated lining cells (arrowhead) and active osteoblasts (white arrow), but not in osteocytes (black arrow). b Osteoclasts were also positive for Nrf2. c ARE-activity during fracture healing increased during the first week and declined to basal level on day 16. Shown are the mean (± SD) of luciferase activity normalized to the contralateral healthy limb of the mice. Statistical analysis was performed using ANOVA * p B 0.05; n = 8. d The right limb of C57BL6/J(ARE-luc) mice was fractured and ARE-activity was measured as described in ‘‘Materials and Methods’’. One representative mouse is shown. The colour overlay on the image represents Radiance (photons/s/cm2/sr). e ARE-activity of the fracture callus (white arrow) and the contralateral non-fractured femur 6 days after fracturing The colour overlay on the image represents Radiance (photons/s/cm2/sr). f Nrf2-mRNA is upregulated 2 and 4 weeks after fracturing in soft tissue of the fractured area. n = 3–7. g The expression of the Nrf2 target gene HO-1 in Nrf2 WT mice is significantly induced after 2 weeks and decreased after 4 weeks to basal level. n = 3–4. E and F show mean (±SD), statistical analysis were performed using ANOVA * p B 0.05. h Nrf2 immunohistochemistry revealed no staining of osteocytes. i In chondrocytes of chondral callus from fractured femora from Nrf2WT mice a strong staining was detectable with decreased gradient to hypertrophy chondrocytes.; nuclei were counterstained using hemalum
In Vivo Nrf2-Activity Imaging C57BL6/J(ARE-luc) animals received intraperitoneal injections of luciferin (150 mg/kg) 10 min before imaging. The animals were anesthetized (1–3 % isoflurane) and placed into a light-proof camera box. Images were taken during 10 min in postero-anterior direction with a field of view of 10 cm. The ARE-activity of each animal where measured throughout the experiment. Light emitted from the transgene mice was detected by the CCD camera of the IVIS Imaging System 100 Series (Xenogen Corporation, Alameda, CA). Signal intensity is shown in false-colours. Signals are quantified and archived using the Living Image software. For this study, photons were quantified from the hind leg from the dorsal images using fitted regions of interest (ROI). 2 ROIs were set: one over the fracture area, one at the same area over the contralateral healthy limb. The results are represented as x-fold of the signal of the fractured site compared to the healthy contralateral area. Through imaging and measuring of a ROI on the integrating sphere, counts detected by the CCD camera digitizer can by converted into physical units of radiance in photons/15 min. To calibrate the xenogen imaging device, background signals were evaluated daily and the Living Image software performed background subtraction calculations.
CA, USA). All reagents for qRT-PCR analysis were purchased by ABIsystems, (iScrip cDNA Synthesis Kit, iQ SYBR Green Supermix) and plate assays were analysed by Icycler IQ5 optical system software V 2.0 (Bio-Rad, Paris, France). The methods of DNA amplification were conducted as previously described [12]. In qRT-PCR, all samples were run in duplicates. Each plate contained two negative controls and a positive control. Expression levels were normalized to 18-S ribosomal RNA. Statistical Analysis Differences between different time points were evaluated using ANOVA. The analysis was performed by GraphPad Prism 5 (GraphPad Software Inc., San Diego, CA, USA). Differences were considered statistically significant if p B 0.05.
mRNA Preparation and Expression Analysis Results Collected bone and near-to-bone soft tissue of healthy bone and the fracture site of 5 mice per time point were snapfrozen in liquid nitrogen and stored at -80°. Control samples (day 0 time point) were collected from the middle half of unfractured femora. Tissues were pulverized using an agate mortar under continuous cooling with liquid nitrogen. After homogenisation (Precellys, peqlab), total RNA was isolated from each sample using TRIzol (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s protocol. DNA was eliminated by DNase I digestion (Invitrogen Life Technologies, Carlsbad,
123
Nrf2 Affects Basal Bone Phenotype Histological analysis showed differences of the bone. Corticalis of Nrf2-KO mice were more inhomogenous resulting in more clefts between the bone lamellas after sectioning (Fig. 1a). This observation was reflected in differences of the biomechanical properties of the bone. Nrf2-KO mice show a significant decreased force to failure and resistance again bending stress compared to WT mice (Fig. 1b).
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
353
123
354
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
Nrf2 is Activated in Osteoblasts and Osteoclasts of Human Healthy Bone
antibody against 4–hydroxy-nonenal (HNE). Unfractured bone from Nrf2-WT and -KO mice showed only few positive cells. Cartilaginous callus from Nrf2-WT mice showed lipid peroxidation (Fig. 3a). Genetic disruption of Nrf2 obviously increased the macromolecular oxidative damage in the cartilaginous callus compared to Nrf2-WT mice. To assess the proliferation rate of fibroblasts and chondrocytes of the different callus we used Ki67/Mip5Immunostaining. Nrf2-KO mice showed less proliferating cells after 2 weeks then WT mice. In contrast Nrf2-KO mice had more proliferating cells compared to WT mice after 4 weeks. (Fig. 3b).
Bone samples of healthy areas of human femoral head obtained from patients suffering from avascular necrosis were examined for expression of Nrf2 using anti-Nrf2 immunohistochemistry. Nrf2-expression was found in osteoblasts (Fig. 2a) and osteoclasts (Fig. 2b) but not in osteocytes (Fig. 2a, b). Immunohistochemistry without primary antibody revealed no marked immunoreactivity. Nrf2 is Activated in the Femur of Mice During Fracture Healing First we used ARE-luciferase mice to measure the Nrf2activity in a longitudinal study spanning the entire process of fracture healing. Unfractured legs with inserted rod showed only weak and sparse luciferase activity over the background. Six days after fracture, the luciferase activity detected in the fractured femora rose significantly compared to control conditions (pre OP). The luciferase activity was significantly increased until day 6, and decreased to basal level on day 16 (Fig. 2c, d). The time-dependent progressions of all measurements are graphically displayed in Fig. 2c. ARE-activity could be shown at the fracture callus (Fig. 2e, white arrow) 6 days after fracturing. Next we investigated the amount of Nrf2 on mRNA-level during fracture healing. Figure 2f presents quantitative realtime RT PCR data of Nrf2-mRNA expression of near-tobone soft tissue of unfractured bone compared to near-tobone soft tissue around the fracture callus tissue 2 and 4 weeks after fracture. Two weeks after fracture, the amount of Nrf2-mRNA rose 230-fold compared to unfractured bone. Four weeks after fracture, the Nrf2-mRNA decreased compared to the expression level after 2 weeks (Fig. 2f). The expression of heme oxygenase 1 (HO-1), a target gene of Nrf2, also significantly increased 15 fold after 2 weeks and declined to basal level after 4 weeks (Fig. 2g). Immunohistochemical analysis of unfractured bone and callus tissue of WT mice were examined for Nrf2 protein expression using anti-Nrf2 antibody. We also showed that the Nrf2-protein expression was up-regulated after fracture since Nrf2 could be stained in chondrocytes of the cartilaginous callus after 2 weeks (Fig. 2i), but not in osteocytes of healthy bone (Fig. 2h).
Cartilaginous Callus of Nrf2-KO Mice Suffers From Increased Oxidative Damage One consequence of oxidative stress is membrane lipid peroxidation. Therefore, we stained slices of mouse bone and cartilaginous callus by immunohistochemistry utilizing an
123
Fractured Nrf2-KO Mice Undergo Retarded Bone Formation and Healing lCT analysis demonstrated imperfect fracture healing in Nrf2-KO mice compared to WT mice (Fig. 4). 2 and 4 weeks after fracture, the Nrf2-KO mice had significant decreased callus volume compared to WT mice during the healing prozess (Fig. 4b). Analysis of the lCT data revealed that fracture callus from Nrf2-KO mice had a higher bone volume–total volume ratio (BV/TV) than Nrf2-WT mice after 2 weeks which slightly increases after 4 weeks of healing, while BV/TV of Nrf2 WT mice slightly decreases during 4 weeks of fracture healing (Fig. 4c). There were no differences in trabecular and cortical bone within the callus of Nrf2-WT mice compared to Nrf2-KO mice after 2 weeks. After 4 weeks the trabecular and the cortical bone fraction was slightly lower compared to Nrf2 KO mice (Fig. 4c). Impaired Bone Formation and Callus Remodelling During Fracture Healing in Nrf2-KO Mice Histological characterisation revealed that bone formation and the final remodelling of the bony callus were reduced in the Nrf2-KO compared to the WT mice. HE staining show a decreased fracture callus in Nrf2-KO mice compared to WT (Fig. 5). Particularly, the amount of cartilaginous callus was decreased in Nrf2-KO. The quality of the remodelled bone in these mice was more immature than in WT mice. After 4 weeks, fractures in the WT-mice were healed and cartilage tissue was essentially absent in the callus. In contrast, Nrf2-KO mice healed incompletely during 4 weeks. Bones of WT mice almost returned to normal shape compared to the persisting callus of the bones of Nrf2-KO mice. Nrf2-KO Affects the Biomechanical Properties of the Fracture Callus A crucial aspect of fracture healing is sufficient loadability of the regenerated tissue of the injured limb in order to
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
355
Fig. 3 Nrf2-KO enhances oxidative stress damage in bone during fracture healing and proliferation is decreased in Nrf2-KO. a Bone of Nrf2-WT and -KO mice were fractured as described in ‘‘Materials and Methods’’. Mice were sacrificed and bone and callus specimens prepared for histological evaluation of 4-hydroxynonenal (HNE) immunohistochemistry of representative sections of untreated bone and callus 2 and 4 weeks after fracturing were shown. Unfractured bone from Nrf2-WT and -KO mice showed only few positive cells.
Cartilaginous callus from Nrf2-WT mice showed increased evidence of lipid peroxidation. Genetic disruption of Nrf2 obviously increased the macromolecular oxidative damage in the cartilaginous callus compared to Nrf2-WT mice. b Immunostaining against Ki67 revealed a decreased immunreactivity in Nrf2–KO mice after 2 weeks compared to WT mice. After 4 weeks there are more Ki67 positive cells compared to WT mice. Nuclei were counterstained with hemalum
regain function. To evaluate differences in strength and stability of an injured bone we performed biomechanical tests (Fig. 6). We chose the distraction-to-failure model as a mode of testing because the geometry of a non-stabilized fracture is highly variable, which renders more standard biomechanical tests less accurate. WT- and Nrf2-KO callus
were subjected to a gradual distractive force and the maximum force required to cause failure of the callus was determined. The resistance against shear stress significantly improved during fracture healing in WT-mice while the Nrf2-KO mice rest at the same level during 4 weeks of healing (Fig. 6a). The torsional rigidity was not affected by
123
356
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
Fig. 4 Reduced callus volume in Nrf2-KO. lCt measurements of fractured bone of Nrf2-KO and -WT mice. a Representive lCT slices of a fractured limb of Nrf2-WT and -KO mice and 3D reconstruction of an Nrf2-WT femur 4 weeks after fracturing. b Callus volume,
c bone volume/total volume, d volume of the trabecular bone and e the cortical bone was calculated 2 and 4 weeks after fracturing. Shown are the mean (±SEM), n = 8. Statistical analysis were performed using ANOVA * p B 0.05.; ns not significant
the Nrf2-KO (Fig. 6b). Indicating that Nrf2-KO callus was structurally weaker than WT callus after 4 weeks.
regulated osteocalcin only 3-fold compared to WT control and rested slightly elevated after 4 weeks (Fig. 7b).
Level of VEGF and Osteocalcin in the Fracture Callus Discussion Expression of VEGF-protein in soft tissue after fracture was measured by ELISA technique. 2 and 4 weeks after fracture VEGF protein expression is upregulated in Nrf2WT while Nrf2-KO mice totally failed to upregulate VEGF after fracture (Fig. 7a). Expression of osteocalcin in mineralized bone tissue was analysed by real-time RT-PCR and normalized to the expression of untreated Nrf2-WT t0 samples. Osteocalcin was significantly upregulated (12-fold) in tissue of Nrf2WT mice 2 weeks after fracture and relapsed back to basal level after 4 weeks. In contrast, Nrf2-KO mice up
123
Fracture healing is a complex process that is poorly understood. Every surgeon remembers cases of patients with impaired bone healing. Many patients ultimately lose a limb after a long period of surgical and conservative attempts to treat the disease. The clinical scenario calls for specific treatment options i.e. the application of osteoinductive and osteoconductive substances. With respect to such therapeutic intentions, the need for more detailed knowledge of the molecular regulating mechanisms becomes most evident. Many studies have dealt with
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
357
Fig. 5 Reduced callus area in Nrf2-KO. Representative bone sections of healthy and fractured WT and Nrf2-KO mice stained with haematoxylin and eosin. The increased fracture callus observed by lCt analysis mainly consist of cartilage. Especially the amount of
cartilaginous callus was more increased in KO mice compared to WT. 4 weeks after fracture the quality of the remodelled bone was more immature in Nrf2-KO mice compared to WT mice (black arrow)
predisposing factors for impaired bone development revealing several gene defects for example osteogenesis imperfecta [13]. In contrast, there are few reports on genetic mechanisms of impaired bone healing. Previous work in our laboratory provided first evidence of a role of Nrf2 in bone remodelling. We could show an increased number of spontaneously microfractured bones
in a mouse model of rheumatoid arthritis in Nrf2-KO mice [14]. To investigate the phenotypic differences in bone formation in Nrf2-KO mice biomechanical and histological experiments were performed that demonstrated significant differences. Histologic analysis revealed brittle cortical bone in Nrf2-KO Mice (Fig. 1a). The force to failure as well as the resistance against bending stress are reduced in
123
358
Fig. 6 No detectable increase of biomechanical properties during fracture healing in Nrf2-KO mice. Biomechanical testing of murine Nrf2-KO and -WT femora 2 and 4 weeks after fracturing. The tests were performed as described in ‘‘Materials and methods’’. Values were obtained for shear stress (a), and torsional rigidity (b). Shown are the mean (±SD); n = 4–8. Statistical analysis was performed using ANOVA *: p B 0.05
Nrf2-KO (Fig. 1b). Nrf2 negatively regulates osteoblast maturation and activity while Nrf2 deficiency leads to increased differentiation of osteoclasts as well as to enhanced activity in vitro, but coculture of osteoblasts from Nrf2-KO mice with bone marrow derived macrophages lead to decreased osteoclastic differentiation [9]. Park et al. [9] could demonstrate in vivo an increased mineral apposition rate in Nrf2-KO mice with an increase of osteoblasts per bone perimeter and osteoblast surface per bone while the number of osteoclasts per bone perimeter did not differ between KO and WT mice. Bone remodelling is a well orchestrated permanently ongoing process in order to maintain the bone. We here provide a rationale for Nrf2 influencing this process. The enhanced bone apposition with reduced bone resorption may lead to inferior bone quality as supported by reduced biomechanical properties in KO mice (Fig. 1). Based on the differences in the basal bone phenotype of Nrf2-KO mice we address the role of Nrf2 in fracture healing. Immunohistochemical staining of Nrf2 in fractured bone and callus showed Nrf2-expression in chondrocytes of the cartilaginous callus (Fig. 2i), but not in osteocytes of healthy bone (Fig. 2h). Since unstressed cells
123
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
Fig. 7 VEGF expression in soft tissue and osteocalcin in mineralized bone tissue is not increased in Nrf2-KO. a Expression of VEGFprotein in soft tissue of fractured femora measured by ELISA. b Expression of osteocalcin in mineralized bone tissue. Expression of osteocalcin was shown as fold induction of untreated Nrf2-WT. Statistical analysis was performed with an ANOVA * p B 0.05; n = 3–7
have a very low steady state protein-level of Nrf2, these results indicate that an induction of Nrf2 occurs during bone remodelling. This observation is in accordance with our finding that Nrf2 is expressed in cells actively involved in bone remodelling: osteoblasts and osteoclasts (Fig. 2a, b). The Nrf2 activity was indirectly measured by AREdriven luciferase reportergen mice, by this we could demonstrate an increasing Nrf2-activity in the fractured limb during the first 6 days (Fig. 2c, d). To localize the signal to the fracture callus we measured the ARE-activity at day 6 of a fleshed femur (Fig. 2e). On the other hand we show that Nrf2 expression is up-regulated in soft tissue around the fractured bone. We added a mechanistic confirmation by the detection the target gene HO-1 (Fig. 1f, g). Noticeably the expression of Nrf2 and HO-1 peaks around day 14 post fracture whereas the Nrf2 activity seems greatest after 1 week. A possible explanation is that AREdriven luciferase can be a detectable effect in the course of the initial inflammatory response to fracture whereas the expression of Nrf2 and HO-1 in the surrounding tissue may well be part of a later but important regenerative reaction. In conclusion, we claim that Nrf2 is activated in the callus
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
359
of the regenerating bone and in the surrounding soft tissue during fracture healing. Since Nrf2 is the major transcription factor that regulates the detoxification of ROS we estimated the extent of oxidative stress level during fracture healing. Because of the highly reactive character of ROS, it is difficult to directly demonstrate their presence in vivo. It is more appropriate to measure the ‘footprints’ of ROS, e.g. their effects on various lipids, proteins, and nucleic acids. Therefore, we used immunohistochemistry against 4-hydroxy-2-nonenal (HNE) a specific product of lipid peroxidation, which we detected during fracture healing in the callus. The oxidative stress level was higher in Nrf2KO mice compared to WT-mice. (Fig. 3a). Arai et al. [15] demonstrated a decreased expression of osteoanbolic markers like the osteoanabolic key transcription factor Runx2, alkaline phosphatase and bone sialoprotein after H2O2 exposure of osteoblast like cells MC3T3-E1 in the presence of ROS. On the other hand a moderate level of ROS is required for chondrocyte survival and differentiation in early stages of chondrogenesis [16]. However, high levels of ROS lead to inhibition of proliferation and to hypertrophy of chondrocytes [17]. Nrf2 is also a negative regulator of chondrogenic proliferation and maturation. Increased oxidative stress leads to reduced differentiation of chondrogenic progenitor cells and in accordance with the reduced callus volume observed in the Nrf2-KO mice after fracturing. In addition, the delayed healing process could be referred to the reduced differentiation caused by the lack of Nrf2. Hinoi et al. [19] could demonstrate a decreased expression of chondrogenic differentiation markers like collagen type II, type X and osteopontin in Nrf2 overexpressing ATDC5 cells. Nrf2-KO mouse shows higher levels of oxidative stress, which inhibits the proliferation and leads in combination with the Nrf2-KO earlier to a hypertrophic phenotype. The effect of Nrf2 of bone is controversial discussed in literature. Hinoi and colleagues reported a negative effect of Nrf2 on osteoblast and chondrocyte differentiation in vitro. They showed an Nrf2-dependent inhibition of a RUNX2-dependent osteocalcin promoter activity [18, 19]. In contrast, Arai et al. [15] showed an increased osteocalcin and Nrf2 expression after ROS exposure in the same cell line. Related to these findings, we showed here that osteocalcin-mRNA is significantly decreased in injured tissue of Nrf2-KO animals after 2 weeks compared to WT animals, indicating that Nrf2 binds to the osteocalcin promoter and up-regulates the osteocalcin expression in vivo (Fig. 7b). Our data are supported by recently published data demonstrating that osteocalcin is Nrf2-regulated under arthritic conditions [20]. Osteocalcin is involved in bone mineralization and calcium ion homeostasis. It is used as a marker for the bone formation process and higher serum-
osteocalcin levels are correlated with an increase in bone mineral density. Osteocalcin is also discussed as a crucial factor to stimulate bone mineral maturation, which could be a hint for the difference in the cortical structure in healthy Nrf2-KO mice (Fig. 5b) [21]. Another aspect is that Nrf2 is also involved in osteoblast differentiation by inorganic phosphate [22]. Park et al. reported an inhibitory effect of Nrf2 on osteoblast activity by siRNA-experiments and an additional inhibitory effect on osteoclast maturation in coculture experiments with osteoblasts from Nrf2-KO mice in vivo. In vitro they could demonstrate an increased mineral apposition rate and bone formation rate in KO-mice [9]. The negative effect on osteoclasts may lead to decreased resorption of the cartilage callus but it could also influence the bone remodelling so that the depletion of Nrf2 may lead to an immature bone of lower quality after fracture healing. Consistent with our recent finding that Nrf2 is a critical factor for VEGF-expression [23], we showed here that Nrf2-KO mice were not able to up-regulate VEGF-protein expression in soft tissue after fracturing (Fig. 7b). VEGF has an important osteoanabolic capacity by triggering osteoblast proliferation and migration as well as osteoblast differentiation. By doing this, VEGF could influence the quality of the bone remodelling [24, 25]. The classical function of VEGF to induce neovascularisation of the fracture callus is an important parameter influencing the healing process. VEGF is the most important growth factor initiating the regeneration of the vascular system at the fracture site [26–28]. The increased oxidative stress, as well as the decreased osteocalcin expression and the failure of VEGF induction in Nrf2-KO mice negatively influence remodelling of bone. lCT analysis (Fig. 4) and histological observations (Fig. 5) of fracture callus volume demonstrate a decrease of callus volume in the Nrf2-KO mice in comparison to the WT littermates. During the healing process the ratio between bone fraction and total callus volume decreases. Although the callus volume in Nrf2-KO mice remained constant between 2 and 4 weeks, the BV/TV slightly increased. The general increased BV/TV might be caused by the general bone phenotype. Park et al. [9] could demonstrate a increased BV/TV ratio in Nrf2-KO mice in comparison to WT but the missing reduction during fracture healing indicates a delayed bone formation. The fraction of trabecular and cortical bone of the callus is similar in WT and Nrf2-KO mice after 2 weeks and further decreased slightly after 4 weeks (Figs. 3e and 4d). In the murine fracture model the remodelling process is nearly finished after 4 weeks and the original structure of trabecular and cortical bone is restored. In Nrf2-KO mice this seem to be retarded. The trabecular and cortical bone fraction reduces during time but are slightly increased after 4 weeks compared to
123
360
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
WT-mice. This is also visible in the more immature bon in Nrf2-KO after 4 weeks in HE-stained slices (Fig. 5d) and the persisting callus of the bones of Nrf2-KO mices (Fig. 5c). The differences in callus structure between Nrf2-KO and -WT are also reflected in the biomechanical properties. While torsional rigidity seems unaffected by the Nrf2-KO, Nrf2 influence the resistance against shear stress. The Nrf2KO callus does not further improve until week 4 (Fig. 6a, b). Since shear stress is inversely related to the cross-sectional area of the callus the reduced callus volume after 4 weeks in Nrf2-KO mice may be one reason for the lower resistance against shear stress [29]. The exposure to high oxidative stress in Nrf2-KO also influences the callus volume and quality by reduced proliferation and maturation of chondroblasts. Also further healing processes are affected by the Nrf2 depletion. The reduced callus size in Nrf2-KO mice and bone quality may be caused by the retarded bone remodelling process by the inhibitory effect of the Nrf2 depletion on osteoclast formation. Also the lack of VEGF induction in Nrf2-KO mice has an negative impact on the healing process. VEGF regulates the vascularisation and therefore the conversion of the cartilage callus to bony callus but it have also an direct impact of the bone quality. VEGF positively influences the osteoblast maturation and activity. Also osteocalcin a marker of bone remodelling implicated in bone mineralization and calcium ion homeostasis is decreased in Nrf2 KO mice and contributes to the retarded bone remodelling in Nrf2-KO mice after fracture healing.
References
Conclusions Taken together, our results show that the transcription factor Nrf2 is a positive regulator of fracture healing, possibly by influencing chondrocytic differentiation within the fracture callus whereas the formation of cortical bone appears unaffected. Acknowledgments We thank S. Echterhagen, C. Jaeschke, M. Nicolau, A. Ru¨ben and L. Shen for their excellent technical assistance. This work has been supported in part by the ChristianAlbrecht-University of Kiel and by a Grant from the AO Germany to S.L. This research project was also in part supported by the Excellence Initiative of the German federal and state governments, the START-Program and by the Interdisciplinary Centre for Clinical Research within the faculty of Medicine at the RWTH Aachen University (T9-3; T9-5; T11-3). This work was in part supported by the DFG (PU 214/5-2, 4-2, 3-2). Human and Animal Rights and Informed Consent All experiments were in keeping with the local ethics committees guidelines and permission was obtained.
123
1. Einhorn TA (1998) The cell and molecular biology of fracture healing. Clin Orthop Relat Res:S7–21 2. Turgut A, Gokturk E, Kose N, Kacmaz M, Ozturk HS, Seber S, Acar S (1999) Oxidant status increased during fracture healing in rats. Acta Orthop Scand 70:487–490 3. Yeler H, Tahtabas F, Candan F (2005) Investigation of oxidative stress during fracture healing in the rats. Cell Biochem Funct 23:137–139 4. Kelpke SS, Reiff D, Prince CW, Thompson JA (2001) Acidic fibroblast growth factor signaling inhibits peroxynitrite-induced death of osteoblasts and osteoblast precursors. J Bone Miner Res 16:1917–1925 5. Lee JM, Li J, Johnson DA, Stein TD, Kraft AD, Calkins MJ, Jakel RJ, Johnson JA (2005) Nrf2, a multi-organ protector? Faseb J 19:1061–1066 6. Wruck CJ, Gotz ME, Herdegen T, Varoga D, Brandenburg LO, Pufe T (2008) Kavalactones protect neural cells against amyloid beta peptide-induced neurotoxicity via extracellular signal-regulated kinase 1/2-dependent nuclear factor erythroid 2-related factor 2 activation. Mol Pharmacol 73:1785–1795 7. Wruck CJ, Claussen M, Fuhrmann G, Romer L, Schulz A, Pufe T, Waetzig V, Peipp M, Herdegen T, Gotz ME (2007) Luteolin protects rat PC12 and C6 cells against MPP ? induced toxicity via an ERK dependent Keap1-Nrf2-ARE pathway. J Neural Transm Suppl: 57–67 8. Takahata Y, Takarada T, Iemata M, Yamamoto T, Nakamura Y, Kodama A, Yoneda Y (2009) Functional expression of beta2 adrenergic receptors responsible for protection against oxidative stress through promotion of glutathione synthesis after Nrf2 upregulation in undifferentiated mesenchymal C3H10T1/2 stem cells. J Cell Physiol 218:268–275 9. Park CK, Lee Y, Kim KH, Lee ZH, Joo M, Kim HH (2014) Nrf2 is a novel regulator of bone acquisition. Bone 10. Chan K, Lu R, Chang JC, Kan YW (1996) NRF2, a member of the NFE2 family of transcription factors, is not essential for murine erythropoiesis, growth, and development. Proc Natl Acad Sci USA 93:13943–13948 11. Bonnarens F, Einhorn TA (1984) Production of a standard closed fracture in laboratory animal bone. J Orthop Res 2:97–101 12. Brandenburg LO, Kipp M, Lucius R, Pufe T, Wruck CJ (2010) Sulforaphane suppresses LPS-induced inflammation in primary rat microglia. Inflamm Res 59:443–450 13. Ben Amor IM, Glorieux FH, Rauch F (2011) Genotype-phenotype correlations in autosomal dominant osteogenesis imperfecta. J Osteoporos 2011:540178 14. Wruck CJ, Fragoulis A, Gurzynski A, Brandenburg LO, Kan YW, Chan K, Hassenpflug J, Freitag-Wolf S, Varoga D, Lippross S, Pufe T (2011) Role of oxidative stress in rheumatoid arthritis: insights from the Nrf2-knockout mice. Ann Rheum Dis 70:844–850 15. Arai M, Shibata Y, Pugdee K, Abiko Y, Ogata Y (2007) Effects of reactive oxygen species (ROS) on antioxidant system and osteoblastic differentiation in MC3T3-E1 cells. IUBMB Life 59:27–33 16. Kim KS, Choi HW, Yoon HE, Kim IY (2010) Reactive oxygen species generated by NADPH Oxidase 2 and 4 are required for chondrogenic differentiation. J Biol Chem 285:40294–40302 17. Morita K, Miyamoto T, Fujita N, Kubota Y, Ito K, Takubo K, Miyamoto K, Ninomiya K, Suzuki T, Iwasaki R, Yagi M, Takaishi H, Toyama Y, Suda T (2007) Reactive oxygen species induce chondrocyte hypertrophy in endochondral ossification. J Exp Med 204:1613–1623
S. Lippross et al.: Impaired Fracture Healing in Nrf2 Deficient Mice
361
18. Hinoi E, Fujimori S, Wang L, Hojo H, Uno K, Yoneda Y (2006) Nrf2 negatively regulates osteoblast differentiation via interfering with Runx2-dependent transcriptional activation. J Biol Chem 281:18015–18024 19. Hinoi E, Takarada T, Fujimori S, Wang L, Iemata M, Uno K, Yoneda Y (2007) Nuclear factor E2 p45-related factor 2 negatively regulates chondrogenesis. Bone 40:337–344 20. Maicas N, Ferrandiz ML, Brines R, Ibanez L, Cuadrado A, Koenders MI, Van den Berg WB, Alcaraz MJ (2011) Deficiency of Nrf2 accelerates the effector phase of arthritis and aggravates joint disease. Antioxid Redox Signal 15:889–901 21. Boskey AL, Gadaleta S, Gundberg C, Doty SB, Ducy P, Karsenty G (1998) Fourier transform infrared microspectroscopic analysis of bones of osteocalcin-deficient mice provides insight into the function of osteocalcin. Bone 23:187–196 22. Beck GR, Moran E, Knecht N (2003) Inorganic phosphate regulates multiple genes during osteoblast differentiation, including Nrf2. Exp Cell Res 288:288–300 23. Kweider N, Fragoulis A, Rosen C, Pecks U, Rath W, Pufe T, Wruck CJ (2011) Interplay between vascular endothelial growth factor (VEGF) and nuclear factor erythroid 2-related factor-2 (Nrf2): implications for preeclampsia. J Biol Chem 286: 42863–42872 24. Mayr-Wohlfart U, Waltenberger J, Hausser H, Kessler S, Gunther KP, Dehio C, Puhl W, Brenner RE (2002) Vascular endothelial
growth factor stimulates chemotactic migration of primary human osteoblasts. Bone 30:472–477 Street J, Bao M, deGuzman L, Bunting S, Peale FV, Ferrara N, Steinmetz H, Hoeffel J, Cleland JL, Daugherty A, van Bruggen N, Redmond HP, Carano RAD, Filvaroff EH (2002) Vascular endothelial growth factor stimulates bone repair by promoting angiogenesis and bone turnover. P Natl Acad Sci USA 99: 9656–9661 Keramaris NC, Calori GM, Nikolaou VS, Schemitsch EH, Giannoudis PV (2008) Fracture vascularity and bone healing: a systematic review of the role of VEGF. Injury 39(Suppl 2):S45– S57 Pufe T, Wildemann B, Petersen W, Mentlein R, Raschke M, Schmidmaier G (2002) Quantitative measurement of the splice variants 120 and 164 of the angiogenic peptide vascular endothelial growth factor in the time flow of fracture healing: a study in the rat. Cell Tissue Res 309:387–392 Weiss S, Zimmermann G, Pufe T, Varoga D, Henle P (2009) The systemic angiogenic response during bone healing. Arch Orthop Trauma Surg 129:989–997 Nazarian A, Entezari V, Vartanians V, Muller R, Snyder BD (2009) An improved method to assess torsional properties of rodent long bones. J Biomech 42:1720–1725
25.
26.
27.
28.
29.
123
