Overview of Fracture Healing and Its Assessment

Summary

The study of post natal skeletal repair is of immense clinical interest. Optimal repair of skeletal tissue is necessary in all varieties of elective and reparative orthopedic surgical treatments. However, the repair of fractures is unique in this context in that fractures are one of the most common traumas that humans experience and are the end-point manifestation of osteoporosis, the most common chronic disease of aging. In the first part of this introduction the basic biology of fracture healing is presented. The second part discusses the primary methodological approaches that are used to examine repair of skeletal hard tissue and specific considerations for choosing among and implementing these approaches.

1. INTRODUCTION - OVERVIEW OF FRACTURE HEALING

1.1. Bone Repair Recapitulates Embryological Skeletal Development

Fracture healing and bone repair are unique in that they are post natal processes that mirror many of the ontological events that take place during embryological development of the skeleton [reviewed by (1–5)]. Indeed many of the genes that are preferentially expressed in embryonic stem cells and the morphogenetic pathways that are active during embryonic skeletal development are also expressed in fracture callus and skeletal repair tissues (6,7) . It is generally believed that the recapitulation of these ontological processes during fracture healing facilitates the regeneration of damaged skeletal tissues to their pre-injury structure and biomechanical function. In this regard, the interplay among regenerative processes in a number of different tissues—vascular, hematopoietic, and skeletal—is essential for the unimpeded repair of injured bones. Furthermore, the appropriate temporal differentiation of the various stem cell populations that form the different tissues and make up skeletal organs is dependent on the proper temporal spatial orchestration of specific paracrine, autocrine, and systemic signaling pathways (7, 8).

1.2. Fracture Healing Cascade

The cascade of events that is commonly described for fracture healing involves formation of a blood clot at the site of injury; an inflammatory phase in which specific cells types involved primarily in innate immune response participate; callus generation in which skeletal stem cells are recruited and differentiate in chondrocytes; primary bone formation in which stem cells are recruited to form bone; and secondary bone remodeling involving osteoclasts in which first mineralized cartilage and primary bone are resorbed followed by prolonged coupled remodeling. While these processes take place in a consecutive temporal manner, they overlap significantly and represent a continuum of changing cell populations and signaling processes within the regenerating tissue.

The disruption to the normal bone microenvironments that is caused by the fracture leads to the interactions of cell populations from the medullary space, periosteum, and enveloping muscular tissues. The signaling and cellular contributions from these different tissues and their microenvironments are unique and contribute to the heterogeneous nature of tissue formation at the fracture site (9). Fracture healing and skeletal tissue repair broadly encompasses an initial anabolic phase that is characterized by de novo recruitment and differentiation of skeletal stem cells that form a cartilaginous callus and, subsequently, of those that form the nascent blood vessels that will feed the new bone. This anabolic phase is followed by a very prolonged catabolic period encompassing resorption of the cartilaginous callus with its replacement by primary bone. Finally, the phase of coupled remodeling takes place, during which the marrow space and hematopoietic tissues are reestablished and regeneration of the original structural features of the injured skeletal organ is achieved. A temporal overview of the biological and histological events of fracture healing, the known cell types that are prevalent at each stage of fracture healing and the stages at which specific signaling molecules are produced are presented in Figure 1

Figure 1.

Summary of the multiple stages of fracture healing. Summary of the stages of fracture repair and the timing of the development of these stages as seen in C57B6 strain of mouse are denoted at the top of the figure. Histological sections are presented below for each stage. All histological specimens are from sagittal sections of mouse tibia transverse fractures and were stained with H&E or Safranin-O and fast green, micrographic images are at 200X magnification. Section for the initial injury was taken from the fracture site 24 hours post injury (far left). Sections depicting the initial periosteal response and endochondral formation are from 7 days post injury (left middle). Arrows denote blood vessels (BV) of the vascular in-growth from the peripheral areas of the periosteum. Sections depicting the period of primary bone and cartilage tissue resorption are from 14 days post injury (middle right). Sections depicting the period of secondary bone formation are from 21 days post injury (far right). Insert depicts 400X images of an osteoclast (chondroclast) resorbing an area of calcified cartilage. The major cell types associated with each stage and the relative time frames of the anabolic and catabolic stages of fracture healing are depicted by yellow and orange triangular overlays. The associated molecular processes and regulators next presented in the second boxed area. At the bottom of the figure, the levels of expression of various marker mRNAs for various molecular processes that have been examined in our laboratories and are denoted by three line widths. The levels of expression are by percent over baseline for each and are not comparable between individual mRNAs. Data for expression levels for the pro-inflammatory cytokines and the ECM mRNAs were from Kon et al., 2001 (34) and Gerstenfeld et al., 2003 (4).

2. ASSESSING SKELETAL REPAIR

2.1. Using an Integrated Approach to Assessing Tissue Repair

All skeletal healing can be defined both functionally (i.e., by the injured skeletal tissue’s regain of its original structure and biomechanical properties) and in terms of the biological processes that facilitate the regain in function. Because of this complexity, an integrated approach should be taken to asses skeletal tissue repair at the level of the whole organ (via biomechanical and micro-computed tomography (μCT) assessments), at tissue and cellular levels (via histological assessments and some methods of contrast-enhanced μCT), and molecular levels (via immuno-histological, in situ hybridizaton, and other assessments of mRNA and protein expression) to define most clearly the mechanisms that promote healing. A description of the use of these techniques presented below provides examples of how biological and functional characteristics are relatable to each other during the normal processes of fracture healing. Many of the methods described here in the context of fracture healing are also detailed in the various chapters of this book. Within this introduction a specific review is provided to illustrate how whole-organ and molecular assessments can be used in conjunction with the many other, tissue- and cell-based assays that are outlined in greater technical detail within the other chapters.

When considering use of multiple assessments, it is important to recognize that performing several assessments on the same sample can be advantageous from the standpoint of cost savings and statistical power. For example, following μCT imaging with mechanical testing allows direct examination of relationships between mechanical behavior and both structure and mineralization. With this particular sequential approach, one can also use the measured mechanical properties to provide an important calibration or validation of finite element models created from the μCT data (10). Alternatively if the tissue is first fixed, such as with formalin, μCT imaging can be followed by histological assessments. This sequence allows one to relate cellular composition to structure and mineral content. Use of multiple approaches may also be used as means of validating biological and structural findings. As an example, qRT-PCR analyses of mRNA expression of markers of chondrogenesis or osteogenesis, or osteoclastogenesis may be used to confirm histological and histomorphometric data of cartilage, bone compositions as well as to define the remodeling and developmental progression of these tissues. By assessing bone healing at multiple levels (organ, tissue, cellular, and molecular) relationships and interactions between the various mechanisms that work at each levels during fracture healing can then be developed. A summary of the most common measurements that are made in the assessment of skeletal repair tissues is presented in Tables 1.

Table 1.

Metrics of Bone Repair

A. Radiographic Data	Units
Faxitron Radiography
Areal BMD	gray level
Projected Bone Area	μm2
Computed Tomography
Total Callus Volume	mm3
Mineralized Callus Volume	mm3
Mineralized Callus Volume Fraction	-
Callus Mineral Content	mg HA/cm3 ‡
Average Tissue Mineral Density	mg HA/cm3 ‡
Standard Deviation of Tissue Mineral Density	mg HA/cm3 ‡
B. Biomechanical Data (Torsion)	Units
Ultimate Torque	N-mm
Torsional Stiffness	N-mm/deg
Torsional Rigidity	N-mm2/deg
Twist to Failure	deg
Toughness (Work to Failure)	N-deg
C. Histomorphometric Data†	Units
Callus Diameter (CDm)*
Mean value for measurements made, in two orthogonal planes, of the diameter of the midpoint of the fracture callus.	Mm
Total Callus Area (CAr)
Mean value for measurements of the total callus areas inclusive of all tissues both within and outside the original bone cortices	mm2
Area of Cartilage (Cg/Ar)
Mean value for measurements of the total cartilage in the callus. May alternatively be expressed as the percent of total callus volume that is cartilage. %Cg	mm2
Area of Total Osseous Tissues TOT/Ar
Mean value for measurements of total callus area that is osseous tissue. (Includes pre-exiting cortical bone, new woven bone and surfaces lined by osteoblasts). May alternatively be expressed as the percent of total callus volume that is osseous tissues. %TOT	mm2
Area of Void (V/Ar)
Mean value for measurements of total callus area that includes the marrow cavity, hematopoietic elements and empty unstained space. May alternatively be expressed as the percent of total callus volume that is the void. %V	mm2
Area of Fibrous Tissue FT/Ar
Mean value for measurements of traced areas of fibrous tissue within the callus. May alternatively be expressed as the percent of total callus volume that is fibrous tissue (%FT)	mm2
Osteoclast Volume Density (Oc/Ar)
Mean value for measurements of tartrate resistant acid phosphatase stained cells calculated as the number of osteoclasts per unit area of callus.	(#/mm2 )
Osteoblast Bone Surface Density (Ob/Ar)
Number of osteoblasts lining a bone or mineralized cartilage surface as calculated per unit surface area of new trabecular bone	#/mm2
Number of Vessels per Callus Area	(#/mm2 )

^*Values in parenthesis ( ) denotes nomenclature as is the agreed standard²

^‡HA = hydroxyapatite

While area, diameter, and cartilage values are not listed in Table 1 and are not commonly used measurements in metabolic bone studies they are defined and the their nomenclature as is in the original agreed standard.

^†Relational Determinations to Biomechanical Testing

CDm and CAr measurements provide comparative values to X-ray, qCT and biomechanical determinations. Diameters can be used in calculations of moments of inertia.

C:TOT measurements can be used in material assessments in relationship to stiffness and strength determinations

Vd Since the external callus tissue initially is devoid of a hematopoietic containing marrow space the assessment and the progression of Vd area gives total measurements of rates of resorption and subsequent remodeling of the external callus and may be related to OcD.

2.2. Selection of a Model to Assess Skeletal Repair

The model that one chooses to use in assessing skeletal tissue repair should be carefully considered in relationship to the research questions that are being asked. In this context four considerations come into play:

• 1) If one is using a surgical model as a means of extrapolating developmental and regenerative characteristics about skeletal tissues, the nature of the type of skeletal formation process that one wishes to examine (intramembraneous bone formation versus endochondral bone formation) is important to consider.

• 2) The nature of both the bone (cortical versus intermedullary) and the surrounding soft tissue compartments that are effected and contribute to the repair process should be considered.

• 3) The skeletal organ and its developmental background should be considered. At its simplest level this would divide surgical models into those that assess bone repair in appendicular, axial and cranial tissues or in different soft tissue elements.

• 4) If one is assessing a therapeutic modality the surgical model that is chosen should most closely approximate the orthopedic application and therapeutic modality that is being assessed.

It is also important to note that, due to the effects of systemic interactions and the heterogeneity in cellular composition, in vivo models cannot be used to fully dissect the molecular mechanisms of the various biological processes that effect repair. It is therefore ideal that in vivo studies should be complimented with in vitro methods of cell or organ culture that are presented in this textbook.

A fracture or any surgical repair model may be tracked temporally and isolated spatially. In the case a fracture of a long bone, the injury induces one round of endochondral bone formation in which callus cells differentiate in a synchronous manner that temporally phenocopies the spatial/temporal variation of the cell zones from the top to the bottom of the growth plate. This round is followed by a prolonged period of coupled remodeling. As such fractures, represent an ideal biological process to examine in a post natal context many cellular and molecular mechanisms that underlie both the endochondral bone formation that takes place during skeletal tissue development and the coupled remodeling that takes place during skeletal tissue homeostasis.

3. Whole Organ Assessments

The two central, functional attributes of skeletal tissues are their ability to regulate apatite mineral deposition and resorption and to assemble and model the microstructure of the mineralized tissue to meet the biomechanical needs of the animal. These unique functional attributes make radiographic approaches particularly useful in examining skeletal tissue repair since these approaches focus on the mineralized tissues within the callus.

3.1. Plain-Film X-Ray

This assessment is the most common clinical tool to assess hard tissue repair although this assessment is limited by its relatively qualitative nature (11-13). Figure 2 presents series of plain-film X-ray assessments made across a time course of fracture healing. This type of study provides a first approximation of the progression of tissue repair (Figure 2). For these studies, an X-ray device with a high-energy beam and capability of high resolution, such as a Faxitron® cabinet X-ray system, is ideal. The best resolution is obtained if the bone is removed from animal and cleaned of a large amount of surrounding soft tissue. Fixation devices should be left in place to maintain the integrity of the construct up until the tissue construct is stable. At least two separate anatomical orientations should be used to make such measurements of callus dimensions and bone bridging since the tissues that form during repair can be irregular in shape and cortices of most bones are not true cylindrical structures.

Figure 2.

Radiographic Series for assessment of fracture healing. A) Time series of evolution of mouse femur fracture callus structure and mineral content. B) Anterior and posterior views of mopsue femur fracture callus at 10 days post fracture. All images were produced with a Faxitron® device set at 40s and 30kV.

3.2. μ CT

Given that the size, shape, and composition of the bone repair tissues change over the time-course of healing, μCT can provide important, qualitative and quantitative assessment of these changes so as to provide non-destructive, and even non-invasive, evaluation of repair progress. We and others have developed “standard” methods of μCT evaluation of mineralized tissues in fracture healing as well as contrast-enhanced μCT methods for examining contributions of non-mineralized tissues and vascular elements.

3.21. μCT Assessment of Mineralized Tissues in the Callus

Many of the technical considerations for μCT studies of intact bones (14) also apply to studies of fracture healing. These considerations include scanning parameters (voltage, current, integration time, and resolution) and methods of image processing (noise filtering, thresholding and defining volumes of interest for analysis). Typically, image resolutions of 16 μm/voxel and 12 μm/voxel are sufficient for rat and mouse calluses, respectively. A Gaussian filter is also commonly used. These resolutions and filter are standard for commercially available desktop μCT systems.

The boundaries of the callus must be defined in μCT assessments, if one is interested in quantifying callus size and the fraction or percentage of the callus occupied by mineralized tissue, just as described below for histomorphometric assessments of callus tissues. Identification, or segmentation, of the callus can be achieved by defining the outer and inner boundaries of the callus on transverse cross-sectional images distributed along the length of the callus (Figure 3). Although we often use an inner boundary that excludes the cortex and medullary space, this boundary can be omitted for complete analysis of the extent of bridging (since some bridging of the cortex might be present) and if healing is so advanced that the boundary between the cortex and mineralized callus tissue cannot be reliably identified.

Figure 3.

(A) Definition of the outer (green) and inner (red) callus boundaries on a transverse cross-section of micro-computed tomography scan of a fracture callus at 14 days post fracture. (B) The 3-D volume of interest that is defined by the area enclosed by these two boundaries on each transverse cross-section along the length of the callus is shown rendered in entirety on the left and in a longitudinal cut-away view on the right. Reprinted from Morgan et al., 2009 (15).

Upon defining the callus or some portion of the callus as the volume of interest, one needs to choose one or more thresholds for specific quantification and visualization of the mineralized portions of the callus. A threshold is a gray value above which a voxel in the μCT scan will be considered to contain mineralized tissue. If a simple categorization of mineralized vs. unmineralized tissue is desired, then one threshold is sufficient. If three or more categories of the relative extent of mineralization are desired, or if a scaffold or implant material is present, then more than one threshold may be used. At present, there is no standard method for choosing a threshold. We recommend that regardless of how many thresholds are used, a threshold should be defined as a percentage of the average grayvalue of the pre-existing cortex or implant material (15). Although commercial μCT systems have software algorithms for quantifying trabecular structure, this type of analysis is not appropriate for repair tissues, because of the wide range of mineralization that is present in these tissues renders the measures of “trabecular” thickness, “trabecular” number, connectivity, etc. extremely sensitive to the choice of threshold.

A number of outcome measures can be quantified using μCT (Table 1). Some of these measures describe callus size and quantity of mineralized tissue (total volume, mineralized volume, mineralized volume fraction, mineral content), while others describe the mineralization (average and standard deviation of the tissue mineral density) or overall structure (moment of inertia).

3.22. Contrast-Enhanced μCT Imaging of Cartilaginous Tissues in the Callus

Formation of the cartilage tissues is a key phase of many skeletal tissue studies. The during skeletal repair cartilage tissues provide initial stability at the surgery or fracture site and serves as a template for subsequent formation of mineralized tissue. In order to provide non-destructive assessment of the soft callus with μCT, a contrast agent is required to increase the x-ray attenuation of these tissues. We have used a cationic, iodinated contrast agent (16) for this purpose. On account of the large, fixed negative charge in cartilage, the contrast agent, via electrostatic attraction, preferentially accumulates in regions of cartilage within the callus. These regions incur the largest increase in attenuation from pre- to post-incubation images. The attenuation of the non-cartilaginous soft tissues is moderately increased, allowing clear delineation of the callus boundaries, while the attenuation of bone tissue is unchanged. The basic experimental approach in this contrast-enhanced μCT (CECT) method is to perform μCT scans both before and after incubation of the callus in the contrast agent (Figure 4A). Analysis of the pre-incubation images, post-incubation images, and images formed by subtracting the former from the latter, enables discrimination among cartilage, non-cartilaginous soft tissue, and mineralized tissue in the callus. The respective locations of these different tissues within the callus can be non-destructively visualized and quantified in both 2-D and 3-D (Figure 4). Measurements of callus area and cartilage area made with CECT compare well to those made using histomorphometry (Figure 4 C,E).

Figure 4.

(A) CECT method: Labeling of tissues in the far right image is as follows: callus boundary = green outline, cartilage = red/pink; cortex = purple; other mineralized tissue = grey. (B) Comparison of (left) a histological section (bright orange-red = cartilage) and (right) CECT cross-section (blue = cartilage; red = mineralized cartilage; grey = bone) of a murine fracture callus (post-operative day 10). (C) 3-D rendering of a callus imaged with CECT (red = cortex; blue = cartilage; yellow = non-cartilaginous soft callus). Comparison of histomorphometric and CECT measurements of (D) cartilage area and (E) total callus area. Each symbol represents a different callus (n=4 measurements per callus, each corresponding to one quadrant of the cross-section).

3.22. Contrast-Enhanced μCT Imaging of Vessel Structure in the Callus

A different contrast agent that is perfused at the time of euthanasia allows μCT assessment of the vasculature during bone repair. In this method, a contrast agent such as a mixture of lead chromate and silicone rubber (Microfil MV-122; Carver, MA), is injected through the left ventricle of the heart and allowed to perfuse with drainage into the body cavity by cutting the vena cave. After perfusion the contrast agent is allowed to polymerize (17,18). The choice of contrast agent and imaging procedure depends on the type of CT scanner available and the type of vasculature to be quantified (19). For imaging of only intermediate- to large-sized vessels or when using synchrotron μCT, a very highly attenuating contrast material, such as bismuth (20), could be used and discrimination between the contrast-enhanced vascular casts and the surrounding mineralized tissue can be achieved purely with a threshold (21-23). However, for analyses of small vascular elements using a desktop μCT system, only a μCT scan performed after decalcification of the host bone is likely to allow clear discrimination between vessel and mineralized tissue.

The disadvantages of performing the μCT scan after decalcification are two-fold. First, decalcification results in large changes in shape and size of the callus, and the original anatomic positions of the vessels are thus lost. Second, the spatial relationship between vascular elements and mineralized tissue cannot be determined. An extension of the aforementioned method is to perform μCT scans both before and after decalcification. The post-decalcification images are registered to and then subtracted from the pre-decalcification images to yield data on the vessels (in their original anatomic position), mineralized tissue, and the respective locations of these two tissues (Figure 5).

Figure 5.

(A) 3-D renderings of the mineralized tissue (yellow) without (top row) and with (bottom row) the vasculature (red) in calluses from a murine model of distraction osteogenesis (DO). “FA” denotes femoral artery. (B) Vessel volume in the gap periphery vs. gap center for the DO time-course and un-operated controls (* p<0.05) Reprinted from Rundle, 2006 (41).

3.3. Mechanical Approaches

In the laboratory setting, the mechanical properties of a healing bone are also commonly assessed by mechanical tests that load the bone in torsion or in three-point bending. The choice of the type of test is dictated by technical as well as physiological considerations. Tension and compression tests are not commonly used, because variability in the alignment of the fracture and in the asymmetry of the callus will lead the applied tensile or compressive displacements to induce variable amounts of bending and shear within the callus. Bending and torsion are logical choices when studying fracture healing in long bones, because these bones experience bending and torsional moments in vivo. However, whereas torsion tests subject every cross-section of the callus to the same torque, three-point bending creates a non-uniform bending moment throughout the callus. As a result, failure of the callus during a three-point bend test does not necessarily occur at the weakest cross-section of the callus.

Regardless of the type of mechanical test, the outcome measures that can be obtained are the strength, stiffness, rigidity, and toughness of the healing bone (Figure 6). For torsion tests, an additional parameter, twist to failure, can be used as a measure of the ductility of the callus. Although strength, a measure of the force or moment that causes failure, can only be measured once for a given callus, it is possible to obtain more than one measure of stiffness and rigidity. Multi-stage testing protocols have been reported that apply non-destructive loads to the callus in planes or in loading modes that are different from those used for the stage of the test in which the callus is loaded to failure. With these protocols, it is possible to quantify the bending stiffness in multiple planes (24) or the torsional as well as compressive stiffness (25).

Figure 6.

Representative torque-twist curve for a mouse tibia 21 days post fracture. The curve is annotated to show definitions of basic biomechanical parameters. Torsional rigidity is computed by multiplying the torsional stiffness by the gage length. Analogous definitions hold for bending tests.

The mechanical properties illustrated in Figure 7 are structural, rather than material, properties. Material properties describe the intrinsic mechanical behavior of a particular type of material (tissue), such as woven bone, fibrocartilage, or granulation tissue. The structural properties of a fracture callus depend on the material properties of the individual callus tissues as well as the spatial arrangement of the tissues and the overall geometry of the callus. While it is possible to use measurements of callus geometry together with those of structural properties to gain some insight into callus tissue material properties (26) true measurement of these material properties requires direct testing of individual callus tissues (27,28).

Figure 7.

Prediction of torsional strength by BV, TMD, and σ_TMD. The beta weight (β) and partial correlation coefficient (r_partial) are given for each μCT measure to indicate its relative contribution. Data from Kakar et al., 2007 (37).

Although there is no substitute for mechanical testing to assess the extent of healing at intermediate-to-late time-points, it is of biological and translational interest to identify relationships between the mechanical properties of the callus and non-destructively obtained measurements of callus composition and structure. In a large, composite dataset of murine calluses at multiple time-points post fracture (15), torsional strength was best predicted (as determined by stepwise regression) by the combination of average tissue mineral density, mineralized callus volume, and the standard deviation of mineral density (r²=0.62, p<0.001) (Figure 7, Table 1). Torsional rigidity was best predicted by the combination of average tissue mineral density, callus mineral content, mineralized volume fraction, and the standard deviation of mineral density (r²=0.70, p<0.001). Changes in the calluses over time were characterized primarily by increase in average tissue mineral density, while variability among calluses at a given time-point was seen primarily in the measures that quantify the absolute and relative amounts of mineralized tissue in the callus, i.e. mineralized callus volume, callus mineral content, and mineralized volume fraction. Overall, these results illustrate how the mechanical properties of the callus depend on measures of both the quantity and mineral density of the hard tissue.

4. TISSUE LEVEL HISTOLOGICAL APPROACHES

Of the recommendations and conventions defined by Parfitt et al., 1987 (29) for general histological assessments of skeletal tissue, we have put forth our perspective on the aspects of these assessments that would be most appropriate for bone repair (30). Extensive details of histological methods including in situ hybridization and immunohistochemistry are discussed in other chapters in this book and will not be discussed in detail here. Rather, the present focus is on general recommendations as to how to approach the use of histological and histomorphometric approaches in the analysis of fracture callus tissues. It is important to note that repair tissues are very heterogeneous, being composed of cartilage, bone, fibrous, and hematopoietic tissues, which are changing throughout the repair process. Therefore, the greatest challenge confronting any histological assessment of a bone reparative process is how to sample this heterogeneous tissue so that any quantitative measurements are representative of the repair tissue that is formed.

In this regard two sets of issues should be considered in any histological assessment of skeletal repair tissues. The first is related to the anatomical plane and sampling of the repair tissue. While fracture calluses may be examined in a longitudinal plane such approaches provide primarily a qualitative overview of the tissue heterogeneity. This in part is due to difficultly of reproducibly positioning a bone during embedding such that a uniform longitudinal plane is always sectioned between individual specimens. This problem largely is related to the fact that long bones are not perfectly cylindrical as illustrated by comparisons of the anterior and posterior X-ray images in Figure 2B and the MicroCT reconstructions presented in Figure 3B. Sampling of tissue compositions is further complicated by the fact that endochondral bone formation, and cartilage resorption and bone remodeling that occurs during tissue repair arises also in a non- uniform manner. This is underscored by Figure 8 which the presents the immense structural heterogeneity in representative longitudinal and transverse sections of fracture calluses at 14 days post injury. To address this challenge, transverse sections can be collected at fixed increments along the long axis of the callus. This sectioning approach provides an optimal means for both observing the variability of the tissue formed in all three planes of the callus and obtaining accurate cross-sectional diameters and area measurements at precisely defined anatomical positions within the callus and in relationship to the fracture site. We have used serial sections to reconstruct in three dimensions the tissue compositions of whole calluses (31) and to obtain measurements of the mechanical properties of callus tissues in conjunction with histological assessment (27).

Figure 8.

Histological approach to assess fracture. A) Demonstration of the tissue heterogeneity and morphological irregularity of both the in longitudinal and transverse dimensions. Three transverse slices taken from center at three increments from the center to edge of the callus showing how the cartilage content varies. B) Demonstration of pseudo coloring of Safranin-O/Fast green stained sections segmenting cartilage from voids and bone areas.

The second practical aspect to be considered in the application of histomorphometric techniques to bone healing studies is to identify appropriate histological stains to assess the callus. It is important to recognize that a tissue stain should not be used to determine tissue phenotype, but rather to improve and enhance visualization of that tissue by distinguishing it, by color, from a different histologically contiguous tissue. Multiple stains can be used to discriminate cartilage from bone and have been used in studies assessing fracture healing. In our studies we have used Safranin-O/ fast green staining, which has been widely used and shown to be effective in measurements of cartilage thickness in studies of osteoarthritis (32) and to quantify the amount of cartilage tissue repair of joint surfaces (33). Other studies also have used combinations of alcian blue with hematoxylin and eosin in order to obtain differential staining of cartilage and bone (34,35). In general, a staining protocol should provide consistent staining outcome that will enhance histological tissue discrimination of cartilage from bone and non-osseous tissue and can be used with semi-automated measurement techniques.

5. MOLECULAR APPROACHES

The assessment of the expression of mRNAs in conjunction with immunohistochemical approaches provides a means of identifying the underlying morphogenetic signaling and regulatory processes that regulate bone repair. In this final section three separate approaches to assay mRNAs (in situ hybridization, individual candidate mRNA assays and large scale microarray will be discussed.

5.1. In situ Hybridization

At the single mRNA level in situ hybridization is used to identify the nature of specific cell populations and their anatomical localization within the callus surrounding muscle and marrow spaces that are expressing a given mRNA. Such approaches can be used both to identity and to validate simple histological assessments of the cell type (chondrocytes, osteoblasts, etc.) and the differentiated states of these cells (31). This approach can also be used in conjunction with immunohistological methods to place in anatomical context the cells expressing a given regulatory factor or morphogenetic protein relative to those expressing receptors that make them responsive to these signals.

5.2. Individual mRNA Assessments

Isolation of total RNAs from the entire callus allows one to look at expression individual mRNAs within the total cell population of the callus tissue. This approach can be used to provide an extremely sensitive temporal road map of the differentiation and development of the tissues and cell types of the callus as well as the expression or any regulatory factors that one is interested, which may be functional in fracture healing. It is critical that calluses be isolated quickly and in as reproducible manner devoid of surrounding muscle as possible with manual dissection methods. Tissues should be dissected as exactly as possible from the point the callus initially rises from cortical surface and since it is physically impossible to separate the original bone and marrow elements from the external callus we isolate the mRNAs from the intact tissue specimen. At very early time points before the callus has condensed (day 5 for mouse and day 7 for rat) a margin of muscle may be needed to ensure that the early regenerative cells are isolated, but some caution should be taken to interpret the anatomical localization of the cells that are expressing a given mRNA specie. Confirmatory analysis using in situ hybridization or immunohistochemistry is recommended.

It is important to note that there are some differences of opinion on how to assess replicates. For our studies in mouse, we generally use three biological replicates, each representing a pool of total RNAs from calluses of three animals. Reference mRNAs are made from unfractured bones (three replicates pooled from groups of three animals) isolated from the mid-diaphyseal region of these bones. This approach incorporates aspects of obtaining reproducibility from assaying replicates and takes into account biological variability since the replicates represent experimental repeats of multiple pooled animal samples. We have taken this approach since mRNA yields from single mouse calluses are very small and insufficient to carry out assays for large numbers of genes or for use in microarray studies. In general this approach has provided very sensitive means of seeing reproducible differences in temporal profiles of expressed genes under differing experimental conditions (34,36,37). For studies performed in rats, individual calluses suffice to provide sufficient yields of total mRNA for running multiple mRNA assays (38).

5.3. MicroArray Approaches

Sequencing of the entire genomes of multiple species has provided the means by which the transcriptome of fracture healing can be assessed by microarray analysis. Given the expense of such studies these approaches should only be carried out in conjunction with a core facility that has technical staff, instrumentation, and a track record of executing this type of study. The setup of a microarray study involves a number of crucial steps the first and most important being having sufficient replicates per experimental group (treatment condition or animal genotype, and time point). In planning a study, sufficient RNA is needed to carry out the array twice in the event of a technical problem with the execution of the array and to carry out mRNA candidate validation by qRT-RT-PCR. A minimum of three biological replicates is recommended as described above. Analysis of microarray data should follow four main steps: 1) quality control to identify microarray artifacts from hybridization and correct for chip batch effects; 2) data pre-processing to eliminate outliers; 3) identification of differentially expressed genes (DEG); and 4) extraction of biological knowledge from DEG. It is recommended that, given the expense of microarray experiments, appropriate expertise in statistical and computational analysis also be available either through collaborative or fee-for-service arrangements. The most basic approach in step 3 uses basic statistical analysis to identify genes that showed the greatest quantitative changes in their expression (39-41). In this context this type of approach can be used to identify both known and novel genes that show the greatest changes in expression over the time course of healing. Other more sophisticated statistical approaches have been carried out that cluster genes based on common temporal profiles of expression and hat examine gene functions in each cluster across the time course of fracture healing (42-44).

6. CONCLUSION

In this brief introductory chapter we have reviewed the biological events of fracture repair. We have also laid out a number of the general methodological approaches including a summary of various radiographic and molecular technique that are used to assess bone repair, using fracture healing as the example.. General guidance is presented on integrating multiple technical approaches to best assess skeletal repair that may be also be applicable for other areas of skeletal biology research.