Detection of lymphangiogenesis by near‐infrared fluorescence imaging and responses to VEGF‐C during healing in a mouse full‐dermis thickness wound model

Noninvasive, longitudinal near‐infrared fluorescence (NIRF) imaging was used to detect and quantify lymphangiogenesis following a full‐dermis thickness incision in the presence and absence of locally administered vascular endothelial growth factor‐C (VEGF‐C), a well‐known regulator of lymphangiogenesis. Peripheral cytokines/chemokines were also measured in treated and sham‐injected animals. Lymphangiogenesis was detected via NIRF imaging by day 7–8 and confirmed by intravital microscopy, while angiogenesis was observed by day 2–3 postincision (PI). All lymph vessel parameters quantified were significantly greater on wounded vs. nonwounded sides of mice. Lymph vessel parameters appeared larger on wounded sides of VEGF‐C– relative to NaCl‐treated mice, although differences were not significant. Interleukin‐1α and interleukin‐22 were significantly elevated at day 7 PI relative to respective preincision levels in VEGF‐C‐treated mice, and decreased by day 21 PI to levels nearing those measured preincision. For the majority of cytokines/chemokines measured, mean responses were significantly greater in VEGF‐C– vs. NaCl‐treated animals. Local VEGF‐C administration may stimulate lymphangiogenesis during tissue repair and regeneration via mediating systemic cytokine/chemokine levels. NIRF imaging can be utilized to detect lymphangiogenesis during wound healing, and offers a promising platform to complement current methods for monitoring wound status and studying the effects of growth factors on healing.     

Reduced COX‐2 Expression in Aged Mice Is Associated With Impaired Fracture Healing

The cellular and molecular events responsible for reduced fracture healing with aging are unknown. Cyclooxygenase 2 (COX‐2), the inducible regulator of prostaglandin E₂ (PGE₂) synthesis, is critical for normal bone repair. A femoral fracture repair model was used in mice at either 7–9 or 52–56 wk of age, and healing was evaluated by imaging, histology, and gene expression studies. Aging was associated with a decreased rate of chondrogenesis, decreased bone formation, reduced callus vascularization, delayed remodeling, and altered expression of genes involved in repair and remodeling. COX‐2 expression in young mice peaked at 5 days, coinciding with the transition of mesenchymal progenitors to cartilage and the onset of expression of early cartilage markers. In situ hybridization and immunohistochemistry showed that COX‐2 is expressed primarily in early cartilage precursors that co‐express col‐2. COX‐2 expression was reduced by 75% and 65% in fractures from aged mice compared with young mice on days 5 and 7, respectively. Local administration of an EP4 agonist to the fracture repair site in aged mice enhanced the rate of chondrogenesis and bone formation to levels observed in young mice, suggesting that the expression of COX‐2 during the early inflammatory phase of repair regulates critical subsequent events including chondrogenesis, bone formation, and remodeling. The findings suggest that COX‐2/EP4 agonists may compensate for deficient molecular signals that result in the reduced fracture healing associated with aging.     

Fluorescent cell tracer dye permits real‐time assessment of re‐epithelialization in a serum‐free ex vivo human skin wound assay

Ex vivo wounded human skin organ culture is an invaluable tool for translationally relevant preclinical wound healing research. However, studies incorporating this system are still underutilized within the field because of the low throughput of histological analysis required for downstream assessment. In this study, we use intravital fluorescent dye to lineage trace epidermal cells, demonstrating that wound re‐epithelialization of human ex vivo wounds occurs consistent with an extending shield mechanism of collective migration. Moreover, we also report a relatively simple method to investigate global epithelial closure of explants in culture using daily fluorescent dye treatment and en face imaging. This study is the first to quantify healing of ex vivo wounds in a longitudinal manner, providing global assessments for re‐epithelialization and tissue contraction. We show that this approach can identify alterations to healing with a known healing promoter. This methodological study highlights the utility of human ex vivo wounds in enhancing our understanding of mechanisms of human skin repair and in evaluating novel therapies to improve healing outcome.     

Time course study of delayed wound healing in a biofilm‐challenged diabetic mouse model

Bacterial biofilm has been shown to play a role in delaying wound healing of chronic wounds, a major medical problem that results in significant health care burden. A reproducible animal model could be very valuable for studying the mechanism and management of chronic wounds. Our previous work showed that Pseudomonas aeruginosa (PAO1) biofilm challenge on wounds in diabetic (db/db) mice significantly delayed wound healing. In this wound time course study, we further characterize the bacterial burden, delayed wound healing, and certain aspects of the host inflammatory response in the PAO1 biofilm‐challenged db/db mouse model. PAO1 biofilms were transferred onto 2‐day‐old wounds created on the dorsal surface of db/db mice. Control wounds without biofilm challenge healed by 4 weeks, consistent with previous studies; none of the biofilm‐challenged wounds healed by 4 weeks. Of the biofilm‐challenged wounds, 64% healed by 6 weeks, and all of the biofilm‐challenged wounds healed by 8 weeks. During the wound‐healing process, P. aeruginosa was gradually cleared from the wounds while the presence of Staphylococcus aureus (part of the normal mouse skin flora) increased. Scabs from all unhealed wounds contained 10⁷ P. aeruginosa, which was 100‐fold higher than the counts isolated from wound beds (i.e., 99% of the P. aeruginosa was in the scab). Histology and genetic analysis showed proliferative epidermis, deficient vascularization, and increased inflammatory cytokines. Hypoxia inducible factor expression increased threefold in 4‐week wounds. In summary, our study shows that biofilm‐challenged wounds typically heal in approximately 6 weeks, at least 2 weeks longer than nonbiofilm‐challenged normal wounds. These data suggest that this delayed wound healing model enables the in vivo study of bacterial biofilm responses to host defenses and the effects of biofilms on host wound healing pathways. It may also be used to test antibiofilm strategies for treating chronic wounds.     

The natural behavior of mononuclear phagocytes in HTS formation

Hypertrophic scars (HTS) are caused by trauma or burn injuries to the deep dermis and are considered fibrosis in the skin. Monocytes, M1 and M2 macrophages are mononuclear phagocytes. Studies suggest that M2 macrophages are profibrotic and might contribute to HTS formation. Our lab has established a human HTS‐like nude mouse model, in which the grafted human skin develops red, raised, and firm scarring, resembling HTS seen in humans. In this study, we observed the natural behavior of mononuclear phagocyte system in this nude mouse model of dermal fibrosis at multiple time points. Thirty athymic nude mice received human skin grafts and an equal number of mice received mouse skin grafts as controls. The grafted skin and blood were harvested at 1, 2, 3, 4, and 8 weeks. Wound area, thickness, collagen morphology and level, the cell number of myofibroblasts, M1‐ and M2‐like macrophages in the grafted skin, as well as monocyte fraction in the blood were investigated at each time points. Xenografted mice developed contracted and thickened scars grossly. The xenografted skin resembled human HTS tissue based on enhanced thickness, fibrotic orientation of collagen bundles, increased collagen level, and infiltration of myofibroblasts. In the blood, monocytes dramatically decreased at 1 week postgrafting and gradually returned to normal in the following 8 weeks. In the xenografted skin, M1‐like macrophages were found predominantly at 1–2 weeks postgrafting; whereas, M2‐like macrophages were abundant at later time points, 3–4 weeks postgrafting coincident with the development of fibrosis in the human skin tissues. This understanding of the natural behavior of mononuclear phagocytes in vivo in our mouse model provides evidence for the role of M2‐like macrophages in fibrosis of human skin and suggests that macrophage depletion in the subacute phases of wound healing might reduce or prevent HTS formation.     

Late failure of a split‐thickness skin graft in the setting of homozygous factor V Leiden mutation: a case report and correlative animal model from the Wound Etiology and Healing (WE‐HEAL) study

We present the case of a 53‐year‐old Caucasian male smoker with remote history of left lower extremity deep venous thrombosis (DVT) and a strong family history of thrombosis, who presented to the Center for Wound Healing at MedStar Georgetown University Hospital with spontaneous left leg ulceration. Prothrombotic evaluation showed homozygosity for the factor V Leiden (FVL) mutation. Therapeutic anticoagulation was commenced with warfarin (Coumadin®) and the patient underwent successful debridement and Apligraf® followed by split‐thickness skin graft (STSG) of two wounds. He had an uneventful postoperative course and on the 27th postoperative day the grafts were 95% intact. However, by postoperative day 41 there was 10% graft loss, and over the subsequent 2 weeks both grafts necrosed. On further questioning, it transpired that the patient had discontinued his warfarin on postoperative day 37 because he thought that it was no longer necessary. The patient is enrolled in the Wound Etiology and Healing (WE‐HEAL) study, and at the time of the original graft, residual skin fragments from the STSG were transplanted onto a nude mouse for development of an animal model of wound healing. The mouse graft was successful and was harvested at postoperative day 87 for pathological examination. We review the mechanisms by which prothrombotic states, particularly FVL mutation, can contribute to skin graft failure and delayed wound healing. This case highlights the importance of considering prothrombotic conditions in patients with spontaneous leg ulcerations and the impact of therapeutic anticoagulation on healing. It further allows us to demonstrate the efficacy of the animal model in which residual fragments of STSG tissue are utilised for transplant onto nude mice for manipulation in the laboratory.     

Exogenous calreticulin improves diabetic wound healing

A serious consequence of diabetes mellitus is impaired wound healing, which largely resists treatment. We previously reported that topical application of calreticulin (CRT), an endoplasmic reticulum chaperone protein, markedly enhanced the rate and quality of wound healing in an experimental porcine model of cutaneous repair. Consistent with these in vivo effects, in vitro CRT induced the migration and proliferation of normal human cells critical to the wound healing process. These functions are particularly deficient in poor healing diabetic wounds. Using a genetically engineered diabetic mouse (db/db) in a full‐thickness excisional wound healing model, we now show that topical application of CRT induces a statistically significant decrease in the time to complete wound closure compared with untreated wounds by 5.6 days (17.6 vs. 23.2). Quantitative analysis of the wounds shows that CRT increases the rate of reepithelialization at days 7 and 10 and increases the amount of granulation tissue at day 7 persisting to day 14. Furthermore, CRT treatment induces the regrowth of pigmented hair follicles observed on day 28. In vitro, fibroblasts isolated from diabetic compared with wild‐type mouse skin and human fibroblasts cultured under hyperglycemic compared with normal glucose conditions proliferate and strongly migrate in response to CRT compared with untreated controls. The in vitro effects of CRT on these functions are consistent with CRT's potent effects on wound healing in the diabetic mouse. These studies implicate CRT as a potential powerful topical therapeutic agent for the treatment of diabetic and other chronic wounds.     

Novel differences in the expression of inflammation‐associated genes between mid‐ and late‐gestational dermal fibroblasts

While cutaneous wounds of late‐gestational fetuses and on through adulthood result in scar formation, wounds incurred early in gestation have been shown to heal scarlessly. Unique properties of fetal fibroblasts are believed to mediate this scarless healing process. In this study, microarray analysis was used to identify differences in the gene expression profiles of cultured fibroblasts from embryonic day 15 (E15; midgestation) and embryonic day 18 (E18; late‐gestation) skin. Sixty‐two genes were differentially expressed and 12 of those genes are associated with inflammation, a process that correlates with scar formation in fetal wounds. One of the differentially expressed inflammatory genes was cyclooxygenase‐1 (COX‐1). COX‐1 was more highly expressed in E18 fibroblasts than in E15 fibroblasts, and these differences were confirmed at the gene and protein level. Differences in COX‐1 protein expression were also observed in fetal skin by immunohistochemical and immunofluorescence staining. The baseline differences in gene expression found in mid‐ and late‐gestational fetal fibroblasts suggest that developmental alterations in fibroblasts could be involved in the transition from scarless to fibrotic fetal wound healing. Furthermore, baseline differences in the expression of inflammatory genes by fibroblasts in E15 and E18 skin may contribute to inflammation and scar formation late in gestation.     

Noncontact,low‐frequency ultrasound as an effective therapy against Pseudomonas aeruginosa–infected biofilm wounds

Bacterial biofilms, a critical chronic wound mediator, remain difficult to treat. Energy‐based devices may potentially improve healing, but with no evidence of efficacy against biofilms. This study evaluates noncontact, low‐frequency ultrasound (NLFU) in the treatment of biofilm‐infected wounds. Six‐millimeter dermal punch wounds in rabbit ears were inoculated with 10⁷ colony‐forming units of Pseudomonas aeruginosa or left as sterile controls. A biofilm was established in vivo using our published model. NLFU treatment was carried out every other day or every day, with contralateral ear wounds acting as internal, untreated controls. Wounds were harvested for several quantitative endpoints and scanning electron microscopy to evaluate the biofilm structure. The P. aeruginosa biofilm consistently impaired wound epithelialization and granulation. NLFU, both every other day and every day, improved healing and reduced bacterial counts relative to untreated controls (p < 0.05). Scanning electron microscopy confirmed a qualitative decrease in bacteria after both treatments. NLFU also reduced inflammatory cytokine expression (p < 0.05). Our study suggests that NLFU is an effective therapy against P. aeruginosa wound biofilm. This represents the first in vivo evidence of energy‐based modalities' impact on wound biofilm, setting the foundation for future mechanistic studies. Continued wound care technology research is essential to improving our understanding, and treatment, of biofilm‐infected chronic wounds.     

Limitations of the db/db mouse in translational wound healing research: Is the NONcNZO10 polygenic mouse model superior?

Murine models have provided valuable insights into the pathogenesis of both diabetes and chronic wounds. However, only a few published reports to date have investigated wound healing differences among the differing diabetic mouse models. The goal of the present study was to further define the wound healing deficiency phenotypes of streptozotocin‐induced (STZ‐induced), Akita, and db/db diabetic mice in comparison with a promising new polygenic strain of Type 2 diabetes (NONcNZO10) by using three specific wound models that targeted different critical processes in the pathogenesis of chronic wounds. Incisional, excisional, and ischemia/reperfusion wound models were established on mice of each strain. Wound healing parameters including tensile strength, epithelial gap, and wound necrosis were evaluated. In contrast to the other diabetic mice, the NONcNZO10 strain was found to have significant wound healing impairments in all wound healing models. Not only do the NONcNZO10 mice appear to better model human Type 2 diabetes, these provocative findings suggest that the mice may show more clinically relevant wound healing deficiencies than previous diabetic mouse models.     

Protective role of IL‐1β against post‐arthroplasty Staphylococcus aureus infection

MyD88 is an adapter molecule that is used by both IL‐1R and TLR family members to initiate downstream signaling and promote immune responses. Given that IL‐1β is induced after Staphylococcus aureus infections and TLR2 is activated by S. aureus lipopeptides, we hypothesized that IL‐1β and TLR2 contribute to MyD88‐dependent protective immune responses against post‐arthroplasty S. aureus infections. To test this hypothesis, we used a mouse model of a post‐arthroplasty S. aureus infection to compare the bacterial burden, biofilm formation and neutrophil recruitment in IL‐1β‐deficient, TLR2‐deficient and wild‐type (wt) mice. By using in vivo bioluminescence imaging, we found that the bacterial burden in IL‐1β‐deficient mice was 26‐fold higher at 1 day after infection and remained 3‐ to 10‐fold greater than wt mice through day 42. In contrast, the bacterial burden in TLR2‐deficient mice did not differ from wt mice. In addition, implants harvested from IL‐1β‐deficient mice had more biofilm formation and 14‐fold higher adherent bacteria compared with those from wt mice. Finally, IL‐1β‐deficient mice had ～50% decreased neutrophil recruitment to the infected postoperative joints than wt mice. Taken together, these findings suggest a mechanism by which IL‐1β induces neutrophil recruitment to help control the bacterial burden and the ensuing biofilm formation in a post‐surgical joint. © 2011 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 29: 1621–1626, 2011     

17β‐Estradiol administration promotes delayed cutaneous wound healing in 40‐week ovariectomised female mice

This study investigated the effect of 17β‐estradiol on wound healing in 40‐week ovariectomised female mice. Thirty‐six‐week‐old female mice were divided into three groups: medication with 17β‐estradiol after ovariectomy (OVX + 17β‐estradiol), ovariectomy (OVX) and sham (SHAM). The mice received two full‐thickness wounds, and the OVX + 17β‐estradiol group was administered 17β‐estradiol at 0·01 g/day until healing. In the OVX + 17β‐estradiol group, the ratio of wound area was significantly smaller than those of the OVX and SHAM groups on days 1–3, 5, 6, 8–12 and 9–12, respectively, the numbers of neutrophils and macrophages were significantly smaller than those on days 3 and 7, the ratio of re‐epithelialisation was significantly higher than those on days 3 and 11, the ratio of myofibroblasts was significantly higher than those on day 11 and smaller on day 14, and the ratio of collagen fibres was significantly larger than that of the OVX group on days 7–14. We found that 17β‐estradiol administration promotes cutaneous wound healing in 40‐week female mice by reducing wound area, shortening inflammatory response, and promoting re‐epithelialisation, collagen deposition and wound contraction. Our results suggest that cutaneous wound healing that is delayed because of ageing is promoted by exogenous and continuous 17β‐estradiol administration.     

Full-thickness wounding of the mouse tail as a model for delayed wound healing: accelerated wound closure in Smad3 knock-out mice

Experimentally induced wounds in animal models are useful in gaining a better understanding of the cellular and molecular processes of wound healing, and in the initial evaluation of the safety and effectiveness of potential therapeutic agents. However, studying delayed healing has proved difficult in animals, whose wounds heal within a few days. In this report, we describe a novel method for establishing mouse wounds that require up to 3 weeks or more for complete closure, and we show the validity of this model in Smad3 null mice, which are known to display accelerated healing. Full-thickness wounds, measuring 0.3 by 1.0 cm, were made down to fascia on the dorsal aspect of the mouse tail in Smad3 knock-out mice and control littermates, approximately 1 cm distal to the body of the animal. The wounds were left to heal by secondary intention and were assessed histologically by computerized planimetry for wound closure at various times after wounding. The wounds in wild-type mice displayed delayed healing, with full closure occurring between 14 and 25 days after wounding. Complete closure of similar wounds in Smad3 null mice healed 30 percent faster (p < 0.01). By immunostaining for ki67, a marker for proliferation, Smad3 null animals also showed increased proliferation of dermal wound cells by day 4 after wounding. Cultured dermal fibroblasts from Smad3 null mice had increased baseline DNA synthesis and, interestingly, an enhanced response to transforming growth factor-beta1. By Western blot analysis, Smad3 null mice fibroblasts showed a compensatory increase in mitogen-activated protein kinase phosphorylation in response to transforming growth factor-beta1, suggesting that mitogen-activated protein kinase overcompensation together with loss of Smad3 may be involved in the modulation of faster healing. We conclude that this novel tail-wounding model may be useful for studying delayed wound closure.     

Survival of human cadaver skin on severe combined immune deficiency pigs: Proof of concept

Transplantation of human xenografts onto immunocompromised mice is a powerful research tool for studying wound healing. However, differences in healing between humans and mice and their small size limit this model. We determined whether human cadaver skin xenografts transplanted onto pigs with severe combined immune deficiency (SCID) would survive and not be rejected. Meshed (1:1.5), cryopreserved human cadaver skin was transplanted onto 10 partial thickness dermatome wounds in each of two normal domestic pigs and two SCID pigs. Autografts (n = 2/animal) from the four animals were used as controls. In normal pigs, all autografts were engrafted and healed with a minimal, if any, inflammation and scarring. All human xenografts were rejected by the normal pigs within 5–11 days and associated with an intense T‐cell inflammatory response. In contrast, both autografts and xenografts were engrafted and survived the 28‐day study in the SCID pigs with a minimal inflammation and no gross scarring.     

Disruption of thrombospondin‐2 accelerates ischemic fracture healing

Thrombospondin‐2 (TSP2) is a matricellular protein that is highly up‐regulated during fracture healing. TSP2 negatively regulates vascularity, vascular reperfusion following ischemia, and cutaneous wound healing. As well, TSP2‐null mice show increased endocortical bone formation due to an enhanced number of mesenchymal progenitor cells and show increased cortical thickness. Mice deficient in TSP2 (TSP2‐null) show an alteration in fracture healing, that is unrelated to their cortical bone phenotype, which is characterized by enhanced vascularization with a shift towards an intramembranous healing phenotype; thus, we hypothesized that there would be enhanced ischemic fracture healing in the absence of TSP2. We investigated whether an absence of TSP2 would enhance ischemic fracture healing utilizing Laser doppler, µCT and histological analysis. Ischemic tibial fractures were created in wildtype (WT) and TSP2‐null mice and harvested 10, 20, or 40 days post‐fracture. TSP2‐null mice show enhanced vascular perfusion following ischemic fracture. At day 10 post‐fracture, TSP2‐null mice have 115% greater bone volume than WT mice. This is associated with a 122% increase in vessel density, 20% increase in cell proliferation, and 15% decrease in apoptosis compared to WT. At day 20, TSP2‐null mice have 34% more bone volume, 51% greater bone volume fraction, and 37% more bone tissue mineral density than WT. By 40 days after fracture the TSP2‐null mice have a 24% increase in bone volume fraction, but other parameters show no significant differences. These findings indicate TSP2 is a negative regulator of ischemic fracture healing and that in the absence of TSP2 bone regeneration is enhanced. © 2012 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 31: 935–943, 2013     

The efficacy of basic fibroblast growth factor‐loaded poly(lactic‐co‐glycolic acid) nanosheet for mouse wound healing

Although human recombinant basic fibroblast growth factor (bFGF) is widely used for wound healing, daily treatment with bFGF is required because of its short half‐life. An effective controlled‐release system of bFGF is, therefore, desired in clinical settings. To investigate the efficacy of a bFGF‐loaded nanosheet for wound healing, focusing on the controlled‐release of bFGF, bFGF‐loaded poly(lactic‐co‐glycolic acid) (PGLA) nanosheets were developed, and their in vitro release profile of bFGF and their in vivo efficacy for wound healing were examined. A polyion complex of positively charged human recombinant bFGF and negatively charged alginate was sandwiched between PLGA nanosheets (70 nm thick for each layer). The resulting bFGF‐loaded nanosheet robustly adhered to silicon skin by observation using a microscratch test. bFGF was gradually and continuously released over three days in an in vitro incubation study. Treatment with the bFGF‐loaded nanosheets (every 3 day for 15 days) as well as with a conventional bFGF spray effectively promoted wound healing of mouse dorsal skin defects with accelerated tissue granulation and angiogenesis, although the dose of bFGF used in the treatment with the bFGF nanosheets was approximately 1/20 of the sprayed bFGF. In conclusion, we developed a bFGF‐loaded nanosheet that sustained a continuous release of bFGF over three days and effectively promoted wound healing in mice.     

Selective and non‐selective cyclooxygenase inhibitors delay stress fracture healing in the rat ulna

Anti‐inflammatory drugs are widely used to manage pain associated with stress fractures (SFxs), but little is known about their effects on healing of those injuries. We hypothesized that selective and non‐selective anti‐inflammatory treatments would retard the healing of SFx in the rat ulna. SFxs were created by cyclic loading of the ulna in Wistar rats. Ulnae were harvested 2, 4 or 6 weeks following loading. Rats were treated with non‐selective NSAID, ibuprofen (30 mg/kg/day); selective COX‐2 inhibition, [5,5‐dimethyl‐3‐3 (3 fluorophenyl)‐4‐(4 methylsulfonal) phenyl‐2 (5H)‐furanone] (DFU) (2.0 mg/kg/day); or the novel c5a anatagonist PMX53 (10 mg/kg/day, 4 and 6 weeks only); with appropriate vehicle as control. Quantitative histomorphometric measurements of SFx healing were undertaken. Treatment with the selective COX‐2 inhibitor, DFU, reduced the area of resorption along the fracture line at 2 weeks, without affecting bone formation at later stages. Treatment with the non‐selective, NSAID, ibuprofen decreased both bone resorption and bone formation so that there was significantly reduced length and area of remodeling and lamellar bone formation within the remodeling unit at 6 weeks after fracture. The C5a receptor antagonist PMX53 had no effect on SFx healing at 4 or 6 weeks after loading, suggesting that PMX53 would not delay SFx healing. Both selective COX‐2 inhibitors and non‐selective NSAIDs have the potential to compromise SFx healing, and should be used with caution when SFx is diagnosed or suspected. © 2012 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 31: 235–242, 2013     

BMP‐7–induced ectopic bone formation and fracture healing is impaired by systemic NSAID application in C57BL/6‐mice

Nonsteroidal antiinflammatory drugs (NSAIDs) are known to potentially impair the fracture healing process. The aim of the present study was to determine if the impairment of bone healing by systemic NSAID application is, at least in part, due to an interaction of NSAIDs with the bone anabolic BMP‐7 pathway. Therefore, we first analyzed fracture healing in control and diclofenac‐treated mice, where we not only found a significant impairment of fracture healing due to diclofenac treatment as assessed by biomechanical testing and µCT imaging, but also found high coexpression of bone morphogenetic protein‐7 (BMP‐7) and cyclooxygenase‐2 (COX‐2) within the fracture callus of both groups. To experimentally address the possible interaction between BMP‐7 and COX‐2, we then induced ectopic bone formation in control (n = 10) and diclofenac‐treated mice (n = 10) by application of BMP‐7 (recombinant human OP‐1, rhOP‐1) into the hamstring muscles. After 20 days of treatment, each ectopic bone nodule was analyzed by contact‐radiography, µCT, histology, and histomorphometry. Diclofenac application decreased the trabecular number and bone mass in the ectopic bone nodules significantly due to reduced osteoblast number and activity. These data demonstrate that the bone anabolic effect of BMP‐7 and fracture healing is impaired by diclofenac application, and suggest that the potential negative impact of NSAIDs on fracture healing is, at least in part, due to interference with BMP‐7 signaling. © 2010 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 28:785–791, 2010