Animal models are frequently employed to mimic serious bone injury in biomedical research. Due to their small size, establishment of stabilized bone lesions in mice are beyond the capabilities of most research groups. Herein, we describe a simple method for establishing and analyzing experimental femoral defects in mice.
While bone has a remarkable capacity for regeneration, serious bone trauma often results in damage that does not properly heal. In fact, one tenth of all limb bone fractures fail to heal completely due to the extent of the trauma, disease, or age of the patient. Our ability to improve bone regenerative strategies is critically dependent on the ability to mimic serious bone trauma in test animals, but the generation and stabilization of large bone lesions is technically challenging. In most cases, serious long bone trauma is mimicked experimentally by establishing a defect that will not naturally heal. This is achieved by complete removal of a bone segment that is larger than 1.5 times the diameter of the bone cross-section. The bone is then stabilized with a metal implant to maintain proper orientation of the fracture edges and allow for mobility.
Due to their small size and the fragility of their long bones, establishment of such lesions in mice are beyond the capabilities of most research groups. As such, long bone defect models are confined to rats and larger animals. Nevertheless, mice afford significant research advantages in that they can be genetically modified and bred as immune-compromised strains that do not reject human cells and tissue.
Herein, we demonstrate a technique that facilitates the generation of a segmental defect in mouse femora using standard laboratory and veterinary equipment. With practice, fabrication of the fixation device and surgical implantation is feasible for the majority of trained veterinarians and animal research personnel. Using example data, we also provide methodologies for the quantitative analysis of bone healing for the model.
It is estimated that half of the US population experience a fracture by the age of 651. For those patients with fractures treated surgically, 500,000 procedures involve the use of a bone graft2 and this number is expected to rise with an increasingly aging population3. Although bone is one of the few organs that has the capacity to completely heal without scarring, there are instances where the process fails3,4. Depending on the circumstances and quality of treatment, 2-30% of long bone fractures fail, resulting in non-union3,5. While there remains some debate on the definition, pseudoarthrosis, critical-sized or non-union bone injuries generally refers to an injury that does not heal over the natural lifetime of the subject6. For experimental purposes, this duration is shortened to the average time required for complete healing of an average-sized bone injury. Non-union bone lesions occur for numerous reasons, but major factors include extreme trauma resulting in a critically-sized gap, infection, poor angiogenesis, tobacco use, or inhibited osteoregenerative capacity due to disease or age7. Even if non-unions are successfully treated, it can cost in excess of $60,000 per procedure, depending on the type of injury and the approaches employed8.
In moderate cases, autologous bone grafting is employed. This strategy involves recovery of bone from a donor site and implantation at the site of injury. While this approach is extremely effective, the volume of available donor-derived bone is limited and the procedure involves an additional surgery, which results in persistent pain in many patients9,10. In addition, the efficacy of the autologous bone graft is dependent of the health of the patient. Bone substitutes made from synthetic materials or processed cadaveric bone are abundantly available11-13, but they have significant limitations, including poor host-cell adhesion properties, reduced osteoconductivity, and the potential for immune rejection14. There is therefore an urgent need for bone regeneration technologies that are safe, effective and widely available.
Our ability to improve bone regenerative strategies is critically dependent on the ability to mimic serious bone trauma in test animals, but the generation and stabilization of large bone lesions is technically challenging. In most cases, serious long bone trauma is mimicked experimentally by establishing a defect that will not naturally heal. Although it can vary with species15, this is achieved by complete removal of a bone segment that is larger than 1.5 times the diameter of the bone cross-section16. The bone is then stabilized with a metal implant to maintain proper orientation of the fracture edges and allow for mobility. Due to their small size and the fragility of their long bones, establishment of such lesions in mice are beyond the capabilities of most research groups. As such, long bone defect models are confined to rats and larger animals. Nevertheless, mice afford significant research advantages in that they can be genetically modified and bred as immune-compromised strains that do not reject human cells and tissue.
For human cell-based applications, immune-compromised mice are attractive to work with because they are physiologically well-characterized, easy to house, cost effective, and easily analyzed radiologically and histologically. Of paramount importance is that immune-compromised mice do not reject cells from different species including humans. Their small size also permits the testing of very small numbers of cells or volumes of experimental scaffolds in orthopedic applications. Several murine orthopedic models have been reported that afford various degrees of bone stability17,18. Those systems that result in very high levels of stability, such as external fixators and locking plates predominantly heal by intramembranous ossification although endochondral healing has been reported19. In contrast, those that permit some micro- and/or macro-motion, such as those employing unfixed or partially-fixed medullary pins, generally heal with a predominance of endochondral ossification20,21. Delayed union or non-union defects of long bone are particularly difficult to achieve in mice due to the extra level of stabilization required. However, a number of approaches have been reported, including medullary pins with interlocking nails, locking plates and external fixators22. These systems generally work well, but given their complicated design they can be technically challenging to install. For example, Garcia et al.23 devised an elegant interlocking pin system for use in mice, but the procedure involves incisions at two separate sites and extensive modification of the femur to accommodate the pins. These procedures were performed under a dissecting microscope.
Herein, we describe a simple femoral medullary pin with a central collar designed to prevent closing of a 3 mm bone deficit and also delineate the original edges of the defect. While the pin was not fixed to the bone itself, precise sizing of the pin diameter and reaming of the medullary cavity results in sufficient interference to minimize torsional motion (Figure 1). With careful selection of inbred age, gender and strain-matched mice, the result is a highly reproducible hypertrophic non-uniondefect22 which can be easily evaluated radiologically. Moreover regions of interest can be reproducibly defined after micro-computed tomography (µCT) for measurement of de novo bone formation and histomorphological parameters. The pins were prototyped in our laboratory using readily available tools.
Figure 1: Experimental principle. Diagrammatic summary of the segmental defect model. The central 3 mm segment of a 9-10 mm murine femur is excised surgically (left). A 3 mm long, 19 gauge surgical steel tube is passed over a 9 mm long, 22 G stainless steel tube and fixed with adhesive at the exact center (right). The resultant pin is fitted into the medullary canals of the remaining proximal and distal portions of the femur with the 19 G collar replacing the 3 mm segment of bone (below, center).
NOTE: This protocol is designed for female nude (Nu/J) mice (18-25 g, 6 weeks) acquired from Jackson Laboratories. Since strains of mice vary slightly in terms of anatomy and growth rate, we advise that the fabrication of pins is optimized to the strain, gender and age of the recipients prior to implantation into live subjects. If the strains are carefully matched, the interference fit between the pin and marrow cavity is highly reproducible. Procedures for housing, diet and general animal husbandry are beyond the scope of this protocol, but all mice were housed in accordance with the Guide for the Care and Use of Laboratory Animals (8th Edition) and institutional policies set by the Institutional Animal Care and Use Committee (IACUC) and Department of Comparative Medicine (DCM) at Scott and White Hospital.
1. Fabrication of Pins
Figure 2: Pin assembly. (A) Photographs of the surgical steel tubing at various stages of assembly and a 4 mm diameter cylinder cut from 5 mm thick Gelfoam sheet. (B) After polishing the assembled pin (above), the Gelfoam cylinder is trimmed, positioned over the steel collar (middle), then autoclaved, resulting in a sterile pin coated with dried Gelfoam. This improves cell attachment at the site of injury and maintains the direction of healing (below). (C) An excised femur fitted with the medullary pin. (D) Examples of pin assemblies at various thicknesses and lengths demonstrating the flexibility of this approach. (E) µCT reconstruction of a pin-stabilized femoral defect illustrating the interaction of the medullary pin with the trabecular bone in the proximal and distal ends of the femur.
2. Surgical Technique
NOTE: The following procedure is written observing all guidelines set by the Guide for the Care and Use of Laboratory Animals (8th Edition). Please observe all additional policies set by the local IACUC and ensure an IACUC-approved animal use protocol (or equivalent) is in place before proceeding with the following sections.
Figure 3: Surgical instruments. (A) Surgical kit. Micro-drill (1), cutting wheel (2), absorbable suture (3), outer suture (4), scalpel blades (5), scalpel handle (6), sterile medullary pins wrapped in foil (7), medullary depth gauge made form hypodermic tubing (8), fine nosed forceps (9), rat-teeth forceps (10), blunt reaming needles (11), needle drivers (12), small hemostats (13), fine scissors (14), periosteal elevator (15), modified Kern forceps (16). (B): Modifications required to produce Kern-style bone-grips for mice.
Figure 4: Surgical procedure. (A-C) Approach and exposure of the femur. (D-E) Elevation and cuts. (F-K) Reaming and sizing of medullary canals.
Figure 5: Surgical procedure. (A-B) Installation of pin. (C-D) Closure. (E) Typical image of properly positioned pin 24 hr after surgery.
3. Postoperative Procedures
4. Analysis of Specimens
NOTE: Bone healing can be assessed by a wide variety of methodologies that are beyond the scope of this protocol. The following is a method that we have successfully employed using a specimen microCT (µCT) imager. It is recommended that the following parameters are tested initially, then optimized for the specific needs of the project.
Mice typically recover consciousness and hind-limb mobility 5-10 minutes after withdrawal of anesthesia. During the first 5 days, it is advisable to house mice individually and introduce environmental enrichment to prevent excess use of the limb. For this purpose, igloo-type nests reduce the need for nest building and encourage resting. We have also observed that provision of food and hydrogel on the floor of the cage reduces the probability of pin displacement. During the 5-day period, analgesia should be administered as required in accordance with institutionally-approved policies. Weight loss up to approximately 15% of the original pre-operative weight is possible during the 5-day postoperative period. Twenty-four hours after surgery, X-rays of the affected limb are recommended to assess pin-placement. The pins should be completely inserted into the proximal and distal medullary canals with the edges of the defect flush against the collar (Figure 2C, 2E, Figure 5E, and Figure 6A). In rare instances (<5% of cases), pins can become dislocated at early stages of healing and animal use protocols should provide for revision surgery if this occurs. While it is technically challenging to quantify torsional motion on murine femurs, manual palpation of specimens confirmed that torsional and longitudinal motion is marginal after 7 days as connective tissue accumulates around the pin. After 5 days of post-operative monitoring, the mice can be returned to standard communal housing as per institutionally-approved policies.
Without therapeutic intervention, the edges of the defect typically extend 0.5 mm during 21 days but in rare cases, new bone growth can be up to 1 mm (Figure 6A). Bone growth arrests after this period as the inflammatory and anabolic stages of regeneration cease resulting in a non-union defect. After 14-21 days of healing, the volume of de novo bone can be readily determined from axial images generated by µCT scanning. Using the scanning, axial reconstruction and ROI selection procedures described in the protocol, the volume of new bone generated increases with time, but does not generally exceed a total of 1 mm3 in the absence of therapeutic intervention (Figure 6C) compared to 6-7 mm3 in an anatomically equivalent region of uninjured femur. While conventional biomechanical testing is technically challenging due to the specimen size and nature of the fixation method, the polar moment of inertia (PMI), an estimation of the ability of a material to resist torsion based on cross-sectional area and density, has been shown to represent a suitable estimation of strength in long bones26,27. Using this method, axial cross-sections at various distances from the lesion edges can be selected for analysis. After 21 days, the PMI of de novo bone 0.25 mm from the lesion edges ranges from between 0.05-0.35 mm4 compared to 0.02-0.08 mm4 at the center of the lesion further confirming the presence of non-union in untreated cases (Figure 6D). The PMI of uninjured femur at an anatomically equivalent location typically ranges between 0.5-0.7 mm4 using the conditions described here.
Figure 6: Typical results. (A) X-ray scans of femurs at day 7, 14 and 21 post-surgery demonstrating limited ingrowth of bone. Panel B: Diagrammatic representation of the ROI parameters for volumetric bone measurement. Representative axial cross-section with an exclusion zone (EX) over the pin-collar including a 100 µm margin to exclude measurement of artifacts at the metal-tissue interface (left). Full bone scan (right, top) demonstrating the longitudinal ROI defined by the sections of bone between the edges of the collar with original bone tissue excluded (EX). The original lesion edges are located in the axial sections using the difference between the diameter of the pin and the collar (right, below). (C) Typical volumetric measurements at day 7, 14 and 21 post surgery with no therapeutic intervention (means with standard deviations, n = 3). (D) PMI measurements at axial sections 0.25 mm from the proximal and distal lesion edges and mid-point (means with standard deviations, n = 3). (E) Masson’s trichrome-stained paraffin-embedded section (left) cut in the longitudinal direction demonstrating bone outgrowth consisting of cartilage (ca) and cancellous bone (b) (scale bar = 0.5 mm). The position of the field is indicated on the X-ray scan (right). (F) Non-decalcified, methyl methacrylate embedded coronal section of 1 day-old defect stained with hematoxylin and eosin (scale bar = 1.0 mm). Please click here to view a larger version of this figure.
Decalcification of the tissues typically takes 10-14 days using standard conditions, and is readily monitored by X-ray scanning. It is advisable to retain some musculature on the specimens during decalcification to improve stability during handling. After decalcification, pins can be carefully removed with a sharp scalpel permitting paraffin embedding and histology. Masson’s trichrome staining of longitudinal sections permits visualization of endochondral bone outgrowth with a leading edge of cartilage (ca) followed by cancellous bone (cb) (Figure 6E). While some damage will inevitably occur during dissection, the histological structure of bone and connective tissue remains clear if the pin is removed carefully. Alternatively, methyl methacrylate embedding and sectioning of non-demineralized sections may be performed with the pin in place (Figure 6F).
Herein, we describe a simple method to generate a critical-sized pin-stabilized defect of the murine femur using standard laboratory and veterinary equipment. While the assembly of the pins and the surgical procedure itself requires practice, it is well within the capabilities of a well-trained biomedical research scientist or veterinarian.
The pin is positioned into the medullary canal without additional fixation, making the procedure technically more feasible than more complicated approaches that employ external fixators or interlocking screws. While some torsional motion may occur during the early stages of healing, this is minimized by careful attention to the pin diameter and adequate reaming of the medullary canal so as to attain a firm interference fit between the implant and endosteum. With careful selection of inbred strain and matching of age and gender, the fit becomes reproducibly robust within a few days. Nevertheless, with the advent of 3D printing techniques, it is expected that torsional motion can be further reduced by more sophisticated versions of the pin that incorporate roughened surfaces and/or barbed attachment sites. The ease of pin fabrication and the availability of a wide variety of hypodermic tubing sizes also permit the optimization of the technique for virtually any adult inbred mouse, irrespective of natural or experimental bone phenotype.
The unique pin-collar design serves two purposes: (i) to prevent aberrant narrowing of the defect and damage of the bone extremities through longitudinal slippage, and (ii) to provide landmarks that define the original edges of the defect. As such, volumetric and PMI measurements can be made easily using a specimen µCT scanner such as a Skyscan 1174. Indeed, this approach permits a level of quantitation that is not easily obtained with standard non-critical sized fracture techniques that often exhibit variable or poorly defined injuries. While a µCT unit is preferable for quantitation of healing, evaluation by objective assessment of orthogonal X-ray images or 2D image analysis techniques may represent feasible alternatives. Due to their small size and relatively low mineral content, murine limbs can be readily prepared for histology and mounted as whole specimens for conventional histomorphometry. This dismisses sampling issues frequently faced by researchers performing histomorphometric analyses of large animal fractures.
In the experiments described here, the healing time was relatively short at 3 weeks, which corresponds to the rapid, anabolic phase of bone healing. Thereafter, bone remodeling is a very slow process28. Generally, if bridging is not observed after 4 weeks, healing is unlikely to occur and in agreement, we observe very little additional bone growth after 4 weeks in this system. Furthermore, a 3 mm gap meets the criteria of Key et al.16 for a critical sized defect and Garcia et al. demonstrated that a gap as narrow as 1.8 mm does not sufficiently heal after 10 weeks and this could be delayed to 15 weeks with stripped perichondrium23.
While the size and fragility of their bones present serious technical challenges for orthopedic research, the use of mice is advantageous in numerous ways. For example, there is a variety of immune-compromised strains that permit testing of human cells and proteins without fear of immunological rejection, and their small size reduces the need for excessive amounts of valuable experimental materials, cells or compounds. This is exemplified by our recent study demonstrating the efficacy of adult human stem cells and their extracellular proteins for osteoregeneration29. The relatively short lifespan of mice also present the opportunity for research into aging30 and the wide variety of inbred strains permit the study of global genotype on healing31. There are also a number of disease models that are easily established in mice such as diabetes and osteoporosis32,33. Of significant note is the availability of many transgenic mice that may be used with this technique to further our understanding of regenerative bone physiology under conditions of extreme trauma.
The authors have nothing to disclose.
We thank the staff and veterinarians at the Scott & White Hospital Department of Comparative Medicine, Temple, Texas, for their invaluable advice and assistance during the development of this technique. This work was funded in part by The Institute for Regenerative Medicine Program Funds, Scott & White RGP grant #90172, NIH 2P40RR017447-07 and NIH R01AR066033-01 (NIAMS). We thank Dr Suzanne Zeitouni for proofing the manuscript.
Name of Equipment/Material* | Company | Catalog or model | Notes |
Pin Assembly | |||
Dremel rotary tool | Dremel | 8220 | or equivalent |
Heavy duty cut off wheel | Dremel | 420 | |
Surgical tubing 19G | Small Parts (Amazon) | B000FMZ8LY | OD 1.07mm, ID 0.889mm |
Surgical tubing 21G | Small Parts (Amazon) | B000FMZ8YQ | OD 0.82mm, ID 0.635mm |
Surgical tubing 22G | Small Parts (Amazon) | B000FMYLZS | OD 0.719mm, ID 0.502mm |
Surgical tubing 23G | Small Parts (Amazon) | B000FN0SY0 | OD 0.643mm, ID 0.444mm |
Cyanoacrylate adhesive | Loctite | 1365882 | |
Emery disc | Dremel | 413 | |
Rubber polishing point | Dremel | 462 | |
Felt polishing disc | Dremel | 414 | |
Gelatin sponge | Surgifoam/Ethicon | 1974 | |
Punch biopsy cutter | Miltex | 33-34 | |
Surgery/post-operative | |||
Warm pad and circulator pump | Stryker/Thermocare | TP700, TP700C, TPP722 | |
Coverage quaternary spray | Steris | 1429-77 | |
Bead sterilizer | Germinator/CellPoint Scentific | Germinator 500 | |
Anesthesia system | VetEquip Inc | 901806 or 901807/901809 | |
Isofluorane anesthetic | VETone/MWI | 501017, 502017 | |
Surgical disinfectant | Chloraprep/CareFusion | 260449 | |
Surgical tools | Fine Science Tools | various | recommend German made |
Face protection | Splash Shield | 4505 | |
Rechargable high speed drill | Fine Science Tools | 18000-17 | |
Diamond cutting wheel | Strauss Diaiond | 361.514.080HP | |
Absorbable sutures | Covidien | UM-213 | |
Outer sutures | Ethicon | 668G | or equivalent |
Vetbond | 3M | 1469SB | or equivalent |
Hydration gel | Clear H2O | 70-01-1082 | |
Diet gel | Clear H2O | 72-01-1062 | |
Buprenorphine | Reckitt and Benckser | 12496-0757-01 | controlled substance |
Mouse igloos | Bio Serv | K3328, 3570,3327 | |
Euthanasia cocktail | Euthasol/Virbac | 710101 | controlled substance |
Analysis | |||
Live animal imager | Orthoscan | FD Pulse | or equivalent |
Micro-CT unit and software | Bruker | Skyscan1174 | or equivalent |
Sealing film/Parafilm M | VWR or Fisher | 100501-338, S37441 | |
*Generic sources are suitable for all other items such as gause, drapes, protective clothing, animal care equipment. |