SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

Posttraumatic arthritis is a frequent long-term complication of intraarticular fractures. A model of a closed intraarticular fracture in C57BL/6 mice that progresses to posttraumatic arthritis has been developed. The MRL/MpJ mouse has shown unique regenerative abilities in response to injury. The objective of this study was to determine if the MRL/MpJ mouse is protected from posttraumatic arthritis after intraarticular fractures.

Methods

Intraarticular fractures were created in MRL/MpJ mice and C57BL/6 control mice (n = 16 each). Limbs were analyzed for posttraumatic arthritis 4 and 8 weeks after fracture using microfocal computed tomography bone morphology, subchondral bone thickness evaluation, and histologic evaluation of cartilage degeneration. Serum cytokines and biomarkers were measured after the mice were killed.

Results

Intraarticular fractures were successfully created in all 32 mice. In the experimental fractured limbs, C57BL/6 mice had a decrease in bone density, increased subchondral bone thickness, and increased cartilage degeneration compared with normal contralateral control limbs. In the MRL/MpJ mice, no differences in bone density, subchondral bone thickness, or histologic grading of cartilage degeneration were seen between fractured and contralateral control limbs. Cytokine analysis showed lower systemic levels of the proinflammatory cytokine interleukin-1α (IL-1α) and higher levels of the antiinflammatory cytokines IL-4 and IL-10 in the MRL/MpJ mice.

Conclusion

This study shows that the MRL/MpJ mouse is relatively protected from posttraumatic arthritis after intraarticular fracture. Further investigation into the mechanism involved in this response will hopefully provide new insight into the pathogenesis, prevention, and treatment of posttraumatic arthritis after intraarticular fracture.

Posttraumatic arthritis is defined as the syndrome of osteoarthritic (OA) joint degeneration that develops after joint injuries and is a frequent long-term complication of intraarticular fractures (1). Recent estimates suggest that posttraumatic arthritis is responsible for 12% of the 21 million cases of OA in the US, at a cost of ∼$7 billion annually to the US economy (2). Posttraumatic arthritis also causes disability in young and middle-aged patients, unlike idiopathic OA (3–6).

Despite the impact of posttraumatic arthritis, relatively little is known about the biologic or biomechanical mechanisms underlying the development and progression of this disease (7). A number of factors appear to contribute to the onset of joint degeneration, including the type of injury, inadequate surgical reduction of the articular surface, residual joint instability, and damage to the biomechanical properties or altered viability of the articular cartilage. Malreduction of the articular surface following fracture consistently results in posttraumatic arthritis (8, 9). Additionally, disruption and impaction of the articular surface can lead to chondrocyte necrosis or apoptosis (10–12). In particular, there is growing evidence that proinflammatory cytokines such as interleukin-1 (IL-1) or tumor necrosis factor α (TNFα) are up-regulated in the joint following trauma, and thus may play an important role in the pathogenesis of posttraumatic arthritis (13, 14), similar to their role in primary OA (15, 16).

In order to more precisely study the pathobiology of traumatic cartilage injury, we have developed a novel mouse model of closed articular fractures of the tibial plateau. This model has the advantage of incorporating cartilage impaction and osteochondral fracture without the use of surgical arthrotomy. Following articular fracture, mice showed significant signs of posttraumatic arthritis, including loss of bone density, increase in subchondral bone thickness, and increase in cartilage degeneration (7). An important advantage of this model is its applicability to studying the mechanisms involved in the development and progression of posttraumatic arthritis through genetically modified mice (7). In particular, recent studies have shown that the MRL/MpJ mouse strain exhibits a unique ability to mount a regenerative, as opposed to a reparative, response to injury, providing a novel model for the study of posttraumatic arthritis. It is important to note that a specific genetic cause has not yet been identified to explain the enhanced regenerative abilities of MRL/MpJ mice. However, these mice exhibit increased levels of transforming growth factor β1 (TGFβ1) and reduced inducible levels of IL-1 and TNFα, as measured in isolated splenocytes and peritoneal macrophages (17). These cytokines have also been implicated in the pathogenesis of posttraumatic arthritis.

We hypothesized that MRL/MpJ mice would exhibit improved bone and cartilage healing after intraarticular fracture, which would lead to reduced severity of posttraumatic OA compared with C57BL/6 mice, a common inbred strain of mice generally used as a control for this strain (17–26). The severity of OA changes was assessed by measurements of bone density, subchondral bone thickness, histologic cartilage degeneration, and levels of serum cytokines and biomarkers reflective of cartilage degeneration.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Animals and fracture procedure.

All animal procedures were performed in accordance with a protocol approved by the Duke University Institutional Animal Care and Use Committee. Male MRL/MpJ mice (n = 22) were obtained from The Jackson Laboratory (Bar Harbor, ME) and male C57BL/6 mice (n = 22) were obtained from Charles River (Wilmington, MA). Mice were housed until 22 weeks of age in order to ensure skeletal maturity (27), at which time 16 mice from each strain were subjected to intraarticular fracture (7). The mice were anesthetized using an intraperitoneal injection of pentobarbital, and positioned with the left lower limb in a custom cradle that held the limb at 90° of flexion and secured at the ankle. The knee was then bluntly probed to determine the location of the tibial plateau. A stainless steel, wedge-shaped indenter mounted to a materials testing system (ElectroForce 3200; Bose, Eden Prairie, MN) was used to apply a 10N compressive preload to the left tibial plateau and to create the fracture by loading the tibial plateau in compression at a rate of 20N/second to a maximum load of 55N. Displacement of the indenter is strongly correlated with the energy of the fracture, as determined by load displacement curves for each joint (7). Therefore, a displacement limit was used during loading in order to limit the severity of the fractures. The displacement limits were scaled to the relative size of the tibial plateau in the mice strains; a limit of 3.2 mm was used for the MRL/MpJ mice and a limit of 2.5 mm was used for the C57BL/6 mice. The C57BL/6 mice tibial plateau is approximately three-fourths the area of the MRL/MpJ tibial plateau as calculated by microfocal computed tomography (micro-CT) of unfractured limbs.

Using anteroposterior, lateral, and oblique high-resolution digital radiography (MX-20; Faxitron, Wheeling, IL), fractures were categorized according to the Orthopaedic Trauma Association (OTA) classification system (28, 29). The goal was to create tibial plateau intraarticular fractures of lower fracture energies and “moderate” severity (OTA classes B1 and B2). After fracture, no restrictions or surgical interventions were applied to the mice in order to evaluate the natural healing response to fractures. An additional 3 mice from each strain were subjected to a sham procedure, in which the mice were anesthetized, the knee was probed to determine the location of the tibial plateau, but the application of the fracture load to the knee was omitted. Eight mice from each strain were killed at both 4 weeks and 8 weeks after fracture. Mice subjected to the sham procedure were killed at 8 weeks. Three additional mice from each strain, with no fracture, were used as controls to measure systemic levels of inflammatory cytokines and biomarkers.

When the mice were killed, ∼150 μl of serum was collected via a retroorbital bleed and cardiac stick for analyses of circulating systemic levels of inflammatory cytokines, TGFβ1, as well as hyaluronic acid (HA) and cartilage oligomeric matrix protein (COMP), which are biomarkers of cartilage metabolism associated with progressive arthritic joint pathology (30–33). Collected blood was centrifuged at 3,500 revolutions per minute for 15 minutes, and the serum was stored at −80°C until analyzed.

Inflammatory cytokines were quantified using the Bio-Plex protein array system (Bio-Rad, Hercules, CA) with the Bio-Plex mouse cytokine 23-plex panel (171-F11241; Bio-Rad), which included IL-1α, IL-1β, IL-4, IL-10, and TNFα. All samples were analyzed according to the manufacturer's instructions using a standard range of 0–3,200 pg/ml and a sample dilution of 1:2, using 30 μl of serum. Commercial enzyme-linked immunosorbent assay kits (MB100B; R&D Systems, Minneapolis, MN) were used to measure HA (029-001; Corgenix, Westminster, CO) and COMP (14-2004-86; AnaMar Medical, Lund, Sweden). Samples were measured in duplicate for each of the above assays according to the manufacturer's instructions, using the following volumes of sera: 20 μl TGFβ1, 20 μl HA, and 12 μl COMP.

After the mice were killed, both hind limbs (experimental and contralateral control) were harvested and fixed in 10% buffered formalin in neutral limb alignment. After fixation, limbs were imaged by micro-CT (μCT 40; Scanco Medical, Brüttisellen, Switzerland) for quantitative 3-dimensional evaluation of the fracture and subsequent healing, as previously described (7), in 3 regions: cancellous bone of the distal femoral condyles, the tibial plateau immediately distal to the subchondral plate, and the metaphyseal region of the tibial plateau. Bone volume (mm3) and bone density (mg/cm3) of each region were reported, as well as cancellous bone fraction (bone volume/total volume) in the femoral condyles. Bone density was determined using a phantom to calibrate linear attenuation to hydroxyapatite concentration. In addition, the volume (mm3) and density (mg/cm3) of incompletely calcified fracture repair callus were determined for both the tibial plateau and the tibial metaphyseal regions.

Two-dimensional coronal cross-sectional images of the weight-bearing region of the joint were generated by micro-CT and used to quantify displacement of the articular surface and subchondral bone thickness for the lateral and medial femoral condyles and the lateral and medial aspects of the tibia. Within each location, 10 thickness measurements were obtained from the articular surface to the beginning of cancellous bone using image analysis software, and the average was reported.

Both limbs from each mouse then underwent standard histologic preparation, including decalcification (Cal-Ex; Fisher Scientific, Fair Lawn, NJ). The entire joint was sectioned in the coronal plane at a thickness of 6 μm. Sections were stained with Safranin O and fast green. Cartilage degeneration was graded in the medial and lateral aspects of both the tibia and the femoral condyles by 3 graders under blinded conditions, and scores were averaged using a previously described modified Mankin scoring scale (maximum score 30) (7, 33, 34). Scores from each location were summed for each joint for a total score within a range of 0– 120.

Statistical analysis.

Statistical analysis was performed using multifactorial analysis of variance, with significance reported at the 95% confidence level. Regression analysis was used to analyze the relationship between measured outcomes.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Intraarticular fracture creation.

Intraarticular fractures were successfully created in all mice. All fractures involved the articular cartilage and underlying subchondral bone. The energy of fracture in the C57BL/6 mice ranged from 34.5 mJ to 134.2 mJ (mean ± SD 79.9 ± 29.6 mJ), and the energy of fracture in the MRL/MpJ mice ranged from 41.1 mJ to 247.9 mJ (mean ± SD 116.3 ± 59.2 mJ). The ratio of fracture energies in the C57BL/6 and MRL/MpJ mice was ∼0.7, which is consistent with the difference in their relative tibial plateau size. The fracture energy was not significantly different between the MRL/MpJ mice at 4 and 8 weeks; however, in the C57BL/6 group, the average fracture energy was 100.4J in the 4-week group and 59.4J in the 8-week group (P = 0.01).

From high-resolution digital radiography and micro-CT scans, fractures in the C57BL/6 mice were categorized using the OTA system as class B1 (n = 8), class B2 (n = 7), or class C1 (n = 1). In the MRL/MpJ mice, fractures were categorized as class B1 (n = 9), class B2 (n = 6), or class B3 (n = 1). Most fractures were characterized as stable fractures with localized displacement of the articular cartilage and underlying subchondral bone (Figure 1). However, 2 fractures from each strain were unstable fractures, resulting in significant displacement of the lateral articular surface of the tibial plateau: 0.635 mm and 0.280 mm (C57BL/6), and 0.632 mm and 0.386 mm (MRL/MpJ). The goal was to create fractures of lower fracture energy and “moderate” severity, which are classified as B1 or B2. The more severe B3 and C1 fractures were associated with higher fracture energies: mean ± SD 92 ± 43 mJ for B1 and B2 versus 185 ± 9 mJ for B3 and C1. One mouse from each strain received a more severe fracture than intended, resulting in a fracture error rate of <7%.

thumbnail image

Figure 1. Classification of intraarticular tibial plateau fractures. Microfocal computed tomography scans of the contralateral control limb and the experimental (fractured) limb in A, a C57BL/6 mouse and B, an MRL/MpJ mouse show similar intraarticular fractures in the experimental limb in each mouse strain, with disruption of subchondral and periarticular bone in the lateral region of the tibial plateau after fracture (arrows).

Download figure to PowerPoint

Sham-operated animals.

The sham procedure resulted in no degenerative changes in the limb compared with the contralateral control limb in either strain in all measured outcomes (P > 0.05). Additionally, no differences were found in systemic levels of cytokines and biomarkers when compared with age-matched animals with no fractures (Table 1).

Table 1. Circulating serum levels of biomarkers, cytokines, and growth factors*
 Nonfracture controlsFracture (4 weeks)Fracture (8 weeks)Sham-operated (8 weeks)
  • *

    Values are the mean ± SD. Three mice from each strain were used as controls, 8 mice from each strain were killed at 4 weeks after intraarticular fracture, 8 mice from each strain were killed 8 weeks postfracture, and 3 mice from each strain underwent a sham procedure and were killed 8 weeks later. HA = hyaluronic acid; COMP = cartilage oligomeric matrix protein; IL-1α = interleukin-1α; TNFα = tumor necrosis factor α; TGFβ1 = transforming growth factor β1; NA = not available.

  • P < 0.03 versus sham-operated mice.

  • P = 0.06 versus all other categories.

  • §

    P < 0.045 versus C57BL/6 mice for all categories.

  • Sample from only 1 animal was available for analysis.

  • #

    P = 0.001 versus C57BL/6 mice for all categories; P = 0.004 versus IL-1β for all categories.

  • **

    P < 0.05 versus controls and sham-operated mice.

  • ††

    P = 0.07 versus 4 weeks postfracture.

Biomarkers, ng/ml    
 C57BL/6 mice    
  HA357.5 ± 120.4546.8 ± 150.3457.8 ± 176.9318.8 ± 54.0
  COMP20.6 ± 11.917.5 ± 6.515.5 ± 4.625.2 ± 13.7
 MRL/MpJ mice    
  HA352.8 ± 41.6528.9 ± 273.7326.2 ± 67.3311.3 ± 42.9
  COMP§29.0 ± 4.843.7 ± 5.633.2 ± 9.440.6 ± 28.4
Proinflammatory cytokines, μg/ml    
 C57BL/6 mice    
  IL-1α148.3 ± 20.779.1 ± 34.3108.9 ± 52.3137.8 ± 57.2
  IL-1β54.3 ± 28.770.9 ± 24.667.0 ± 18.773.0 ± 11.6
  TNFα661.4 ± 104.2725.4 ± 519.4468.2 ± 217.2473.9 ± 527.3
  IL-622.7 ± 4.912.4 ± 8.111.1 ± 5.524.2
 MRL/MpJ mice    
  IL-1α#39.7 ± 13.847.2 ± 19.946.5 ± 26.434.7 ± 10.4
  IL-1β92.9 ± 11.395.8 ± 4.994.8 ± 30.487.1 ± 33.0
  TNFα580.6 ± 221.11,044.5 ± 778.1819.8 ± 714.4803.0 ± 860.9
  IL-612.9 ± 4.819.5 ± 12.529.0 ± 12.815.6 ± 13.9
Antiinflammatory cytokines, μg/ml    
 C57BL/6 mice    
  IL-41.0 ± 0.20.7 ± 0.3**0.4 ± 0.3**1.4 ± 0.7
  IL-1054.7 ± 7.420.7 ± 10.0**29.5 ± 12.4**63.0 ± 40.9
 MRL/MpJ mice    
  IL-4§0.5 ± 0.11.2 ± 0.81.2 ± 1.91.0 ± 1.2
  IL-10§28.0 ± 4.647.9 ± 15.947.3 ± 51.132.1 ± 40.5
Growth factors, ng/ml    
 C57BL/6 mice    
  TGFβ191.997.5 ± 29.0113.6 ± 16.8††NA
 MRL/MpJ mice    
  TGFβ194.2 ± 5.392.6 ± 12.6105.4 ± 13.2††88.5 ± 4.6

Micro-CT evaluation.

The tibial plateau in the fractured limb of mice of both strains showed a slight increase in bone volume from 4 to 8 weeks postfracture (Figure 2A), although the difference was not significant compared with the contralateral control limb (P = 0.058). Both strains also showed lower bone density in the fractured limb compared with the control limb at all measured time points since fracture (P = 0.04). However, within the tibial metaphysis (Figure 2B), the C57BL/6 strain showed a significant increase in bone volume (P < 0.001) accompanied by a significant decrease in bone density (P < 0.001) in the fractured limb that was not observed in the MRL/MpJ strain. In the femoral condyles, there was a decrease in cancellous bone fraction in the fractured limb (mean ± SD 0.71 ± 0.08) compared with the control limb (0.73 ± 0.08) in both strains at all measured time points since fracture (P = 0.04).

thumbnail image

Figure 2. Morphologic bone changes in the tibia following intraarticular fracture (Fx), as measured by microfocal computed tomography. A, Bone volume and bone density of the tibial plateau immediately distal to the subchondral plate in C57BL/6 and MRL/MpJ mice at 4 weeks postfracture (n = 8 per strain) and 8 weeks postfracture (n = 8 per strain) and 8 weeks after the sham procedure (n = 3 per strain). Mice of both strains had lower bone density in the fractured limb compared with the control limb at all measured time points since fracture (∗ = P = 0.04). B, Bone volume and bone density of the metaphyseal region of the tibia in C57BL/6 and MRL/MpJ mice at 4 weeks postfracture (n = 8 per strain) and 8 weeks postfracture (n = 8 per strain) and 8 weeks after the sham procedure (n = 3 per strain). C57BL/6 mice had a significant increase in bone volume (∗∗ = P < 0.001) and a significant decrease in bone density at all measured time points since fracture. Values are the mean ± SD.

Download figure to PowerPoint

There was a significant contribution of fracture callus in the tibial plateau and the metaphyseal region of the tibia in both strains. In the tibial plateau, there was a significantly higher callus density in the fractured limbs at 4 weeks postfracture compared with 8 weeks (P < 0.01): C57BL/6 mice, mean ± SD 365 ± 25 mg/cm3 at 4 weeks and 338 ± 10 mg/cm3 at 8 weeks, and MRL/MpJ mice, 347 ± 17 mg/cm3 at 4 weeks and 335 ± 8 mg/cm3 at 8 weeks. The metaphyseal region of the tibia showed more callus volume than the tibial plateau in both strains (P < 0.001). Within the metaphyseal region, there was significantly more callus volume at 4 weeks postfracture compared with 8 weeks in both strains (P > 0.01): C57BL/6 mice, 0.77 ± 0.18 mm3 at 4 weeks and 0.57 ± 0.06 mm3 at 8 weeks; MRL/MpJ mice, 0.81 ± 0.12 mm3 at 4 weeks and 0.77 0.10 mm3 at 8 weeks. Both of these changes in callus volume and density were associated with the transitioning of fracture callus to fully calcified bone.

Subchondral bone thickness.

In the C57BL/6 mice, subchondral bone thickness was significantly higher in the experimental limb (P < 0.05) when compared with the control limb at both 4 and 8 weeks postfracture. The MRL/MpJ mice did not show any subchondral thickening in the experimental limb after fracture (P = 0.99) compared with the control limb at either 4 or 8 weeks postfracture (Figure 3B).

thumbnail image

Figure 3. Subchondral bone thickness. A, Coronal cross-sectional image showing the thickness of subchondral bone. Lines indicate the location of subchondral thickness measurements in the lateral and medial femoral condyles and in the lateral and medial aspects of the tibia. B, Mean ± SD subchondral thickness at 4 weeks and 8 weeks postfracture and at 8 weeks after the sham procedure in the C57BL/6 and MRL/MpJ mice. In C57BL/6 mice, subchondral bone thickness was significantly greater in the fractured limb (Fx) compared with the control limb at both 4 and 8 weeks postfracture (∗ = P < 0.05).

Download figure to PowerPoint

Histologic evaluation of cartilage degeneration.

The C57BL/6 mice showed more cartilage degeneration postfracture than the MRL/MpJ mice. The C57BL/6 mice showed loss of cartilage, loss of proteoglycan staining, and development of thick fibrocartilage. The MRL/MpJ mice did show some loss of cartilage at the fracture site, but did not show a loss of proteoglycan staining or the development of fibrocartilage (Figure 4). The total joint Mankin score of cartilage degeneration (sum of all locations) was significantly higher (P = 0.03) for the fractured limb in C57BL/6 mice than for the fractured limb in MRL/MpJ mice (C57BL/6 22.8 ± 10.3 versus MRL/MpJ 15.0 ± 8.5).

thumbnail image

Figure 4. Histologic evaluation of articular cartilage. A, Lateral compartment of the contralateral control joint and the experimental (fractured) joint (in the coronal plane) from the same C57BL/6 mouse shown in Figure 1, demonstrating cartilage disruption and degeneration, fibrocartilage, and loss of proteoglycan staining (red). B, Lateral compartment of the contralateral control joint and the fractured joint (in the coronal plane) from the same MRL/MpJ mouse shown in Figure 1, demonstrating articular cartilage disruption with no fibrocartilage and minimal loss of staining. Boxed areas show sites of histologic damage. F = femur; T = tibia; M = meniscus. Bars = 100 μm. (Safranin O stained; original magnification × 100.)

Download figure to PowerPoint

In examining site-specific changes within the joint, the lateral tibial plateau in the C57BL/6 mice showed the most degeneration postfracture (Figure 5A). In the MRL/MpJ mice, there was no significant difference in Mankin scores between the contralateral control and experimental limbs at either 4 or 8 weeks postfracture. Analysis by anatomic location within the joint showed no significant differences at any location postfracture (Figure 5B).

thumbnail image

Figure 5. Histologic grading of articular cartilage degeneration, by location in the joint, at 4 and 8 weeks postfracture and 8 weeks after the sham procedure. A, In the C57BL/6 mice, the modified Mankin score was significantly higher in the experimental (fractured) limb (Fx) compared with the control limb at 4 weeks postfracture (∗ = P < 0.001). B, In the MRL/MpJ mice, there was no difference in modified Mankin scores. LF = lateral femoral condyle; MF = medial femoral condyle; LT = lateral tibial plateau; MT = medial tibial plateau. Values are the mean ± SD.

Download figure to PowerPoint

There was a significant correlation between the type of fracture (OTA) and the resultant articular cartilage degeneration in both strains. The 2 severe fracture types, B3 and C1, had significantly higher Mankin scores than the moderate B1 and B2 fractures (P = 0.04): B1 18.5 ± 9.0, B2 15.7 ± 6.6, B3 37.7, and C1 47.7. Both severe fractures randomly occurred in the 4-week group. In the C57BL/6 strain, this resulted in significantly higher Mankin scores at 4 weeks postfracture compared with 8 weeks.

Biomarkers of cartilage degeneration.

Circulating levels of HA and COMP were measured from serum (Table 1). There was no overall strain effect on systemic levels of HA. However, in the C57BL/6 strain, the mice with fractures had significantly higher systemic HA levels than the sham-operated mice (P = 0.03). In the MRL/MpJ strain, there was a trend toward higher systemic levels of HA at 4 weeks postfracture compared with 8 weeks postfracture and compared with sham-operated and control mice (P = 0.06). Additionally, a positive correlation was found between systemic levels of HA and histologic Mankin scores of cartilage degeneration in all animals (r2 = 0.50, P = 0.003).

For COMP, the MRL/MpJ strain showed significantly higher systemic levels than the C57BL/6 strain (P < 0.001). Although there were contrasting patterns of change in serum COMP following fracture in the 2 strains (increase in MRL/MpJ, decrease in C57BL/6), there was no significant effect of time since fracture on the correlation (Table 1).

Systemic levels of cytokines.

For the proinflammatory cytokines measured, there was no significant effect of fracture (Table 1). However, the MRL/MpJ strain had significantly lower levels of IL-1α compared with the C57BL/6 strain (P = 0.004) at all measured time points since fracture. Additionally, the MRL/MpJ strain had significantly lower levels of IL-1α compared with IL-1β (P = 0.001). There were no differences in levels of IL-1β, TNFα, or IL-6 between the strains. For the antiinflammatory cytokines measured, there was a significant effect of fracture in the C57BL/6 strain (Table 1). The C57BL/6 mice with fractures showed significantly lower levels of IL-4 and IL-10 than did the sham-operated mice and the controls (P < 0.05). Additionally, the MRL/MpJ strain showed significantly higher levels of both IL-4 and IL-10 than the C57BL/6 strain (P < 0.04). There was a significant effect of fracture type by strain on IL-10 levels. All fracture types in the C57BL/6 mice (B1, 20.9 ± 8.8 μg/ml; B2, 32.2 ± 11.5 μg/ml; C1, 9.2 μg/ml) had significantly lower levels of IL-10 than did the B1 fractures in the MRL/MpJ mice (67.8 ± 57.7 μg/ml; P < 0.014), and within the C57BL/6 strain, the most severe C1 fracture had significantly lower levels of IL-10 than did the more moderate B2 fractures (P = 0.026). For the growth factor TGFβ1, there was no strain effect, but there was a trend toward increasing systemic levels of TGFβ1 from 4 weeks to 8 weeks postfracture in both strains (P = 0.07).

There were no significant correlations found between circulating cytokine levels and histologic severity of cartilage degeneration, except for IL-1α. A positive correlation was found between systemic levels of IL-1α and Mankin scores of cartilage degeneration in all animals (r2 = 0.41, P = 0.016). There was no effect of strain or time since fracture on the correlation.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

The findings of this study support the hypothesis that MRL/MpJ mice exhibit improved bone and cartilage healing after intraarticular fracture that results in a reduced severity of posttraumatic OA compared with C57BL/6 mice. After 4 and 8 weeks of healing, the MRL/MpJ mice did not show the signs of posttraumatic arthritis exhibited by the C57BL/6 mice. The MRL/MpJ mice therefore appear to be protected from posttraumatic arthritis after intraarticular fractures.

A number of analyses were used to detect the development of posttraumatic arthritis. Micro-CT analysis was used to evaluate morphologic bone changes that are consistent with OA. A decrease in trabecular bone thickness and a loss of bone density has been shown in OA (35). The C57BL/6 mice in this study had a decrease in bone density in the fractured limb compared with the control limb at 4 and 8 weeks postfracture, which is consistent with OA changes previously reported clinically (35, 36) and in animal models (7, 37, 38). However, the MRL/MpJ mice did not show a decrease in bone density in the fractured limb as compared with the control limb in the metaphyseal region of the tibia.

In addition to density changes, the C57BL/6 mice showed thickening of the subchondral bone in the fractured limb, which increased from 4 to 8 weeks postfracture. The MRL/MpJ mice showed no subchondral bone thickening. Increased subchondral bone thickness is seen clinically and has also been observed in various animal models of arthritis (7, 33).

The C57BL/6 mice showed a significantly higher Mankin score of cartilage degeneration in the experimental limb compared with the control limb 4 weeks postfracture. Eight weeks postfracture, the experimental limb had a higher Mankin score than the control limb, although this difference was not significant. The MRL/MpJ mice did not show a significant increase in degenerative cartilage in the fractured limb compared with the contralateral control limb at any time.

In the C57BL/6 mice, the mean Mankin score of the experimental limbs was higher at 4 weeks than at 8 weeks postfracture due to the random assignment of the more severe C1 type fracture and a B1 fracture that extended from the anterior to the posterior aspect of the tibial plateau in the 4-week group. All the fractures in the 8-week group were the more moderate B1 and B2 fractures. Additionally, the average energy of fracture was significantly higher in the 4-week group than in the 8-week group (100.4 mJ versus 59.4 mJ; P = 0.01).

Systemic levels of HA and COMP were evaluated from serum collected when the mice were killed. These 2 markers have been used in human studies and are shown to be indicative of various aspects of OA. HA is a glycosaminoglycan found in synovium and cartilage that predicts disease outcome in knee (39) and hip OA (40) and correlates with OA progression (39–45). To date, there are few data regarding serum or synovial fluid biomarker levels following articular fracture and the subsequent development of arthritis. In a study of patients with acute knee injury, there was no association between serum or synovial fluid HA levels and the extent of cartilage damage (46). In this study, HA was significantly correlated with the histologic severity of cartilage degeneration in both strains, demonstrating that HA may be useful for monitoring the development of disease in this model.

COMP is a noncollagenous protein primarily isolated from the extracellular matrix of cartilage, although it is not unique to this tissue. COMP has been used successfully to monitor tissue involvement in a rat model of experimental arthritis (47) and has been found to be predictive of cartilage degeneration in a canine meniscectomy model of OA (33). COMP levels were not associated with degenerative changes in the articular cartilage. The MRL/MpJ strain had higher systemic levels of COMP than did the C57BL/6 strain. Although COMP levels were not significantly affected by fracture in either strain, they tended to decrease in C57BL/6 mice and increase in MRL/MpJ mice postfracture. These findings are consistent with a trend toward a more robust bone repair response in the MRL/MpJ mice.

Systemic levels of the proinflammatory cytokine IL-1α were positively correlated with cartilage degeneration, and IL-1α levels were significantly lower in the MRL/MpJ strain. Conversely, this strain had higher levels of the antiinflammatory cytokines IL-4 and IL-10 postfracture compared with the C57BL/6 strain. Under certain circumstances, IL-10 may exhibit either antiinflammatory or proinflammatory activity, but in this model system, IL-10 is likely to have antiinflammatory effects. Previous studies have shown lower inducible levels of the proinflammatory cytokines TNFα and IL-1β levels in MRL/MpJ mice compared with those in C57BL/6 mice (17). This combination of lower proinflammatory and higher antiinflammatory cytokines may aid in the unique healing characteristics of the MRL/MpJ mice.

There were no significant differences in systemic levels of TGFβ1 between the MRL/MpJ and C57BL/6 strains. However, there was a trend toward increasing levels of TGFβ1 with time in both strains (P = 0.07). Systemic levels of TGFβ1 in patients with long bone fractures have been reported at levels ranging from 54.57 ng/ml at 5 days postfracture to a peak level of 112.71 ng/ml at 42 days postfracture (48). This trend and the levels measured are similar to those described in our study. Future studies may investigate the role of TGFβ1 immediately following fracture.

The mechanisms behind the regenerative abilities of the MRL/MpJ mouse are not well understood, nor is it known whether the protection against posttraumatic arthritis is related to the regenerative capabilities. Several researchers have investigated genetic differences in the MRL/MpJ mice. Multiple candidate genes have been found within genomic quantitative trait loci that explained 70% of the variance in wound healing and included multiple growth factors and cytokines (49–51). Chadwick et al identified 400 genes that were differentially expressed in the MRL/MpJ mice compared with C57BL/6 and DBA/2 mice with amputated digit tip regrowth (52). Li et al, using gene expression in wound repair, showed that MRL/MpJ mice had less inflammatory response and an earlier transition into tissue repair than C57BL/6 mice (23, 24). Further mechanistic work showed that MRL/MpJ mice have different inducible levels of cytokines. Studies of stimulated macrophages and splenocytes from MRL/MpJ mice showed an increased inducible level of TGFβ1 and decreased inducible levels of TNFα and IL-1β (17).

In future studies, we hope to investigate these genetic variations that play a role in articular fracture healing and which mechanisms may play a protective role in preventing the development of posttraumatic arthritis. In a recent study, it has also been shown that these mice exhibit accelerated repair of full-thickness cartilaginous defects (53). Interestingly, that study showed sexual dimorphism in the repair response, with male mice showing a significant increase in the repair response, while female mice did not (53). In the present study, we examined only male mice, which consistently showed improved cartilage and bone repair.

No effect of the sham procedure was seen in either strain. Although the contralateral limb may be affected by alterations in gait and systemic effects of the articular fracture and subsequent arthritic changes, there were no significant differences in any measured outcome between the contralateral control limbs of the experimental mice and the mice that underwent the sham operation (P > 0.40), which suggests that the changes that occurred were due to the articular fracture.

The fractures created in this model were not surgically reduced and weight bearing, although the injured leg was not restricted so that the natural history of healing following articular fracture could be followed. Additionally, this study was limited to “moderate” fractures, so the ability of MRL/MpJ mice to heal highly displaced or comminuted fractures is unknown.

This study identifies a unique mouse strain that is protected from posttraumatic arthritis after closed intraarticular fracture. MRL/MpJ mice do not develop decreased bone density, increased subchondral bone thickness, or increased cartilage degeneration after intraarticular fracture, while these changes related to posttraumatic arthritis are consistently seen in the C57BL/6 strain. Identification of the regenerative mechanism that prevents the development of posttraumatic arthritis in the MRL/MpJ mouse will provide new insights into the pathogenesis, prevention, and treatment of this debilitating disease.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Dr. Olson had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Ward, Furman, Guilak, Olson.

Acquisition of data. Ward, Furman, Huebner, Kraus.

Analysis and interpretation of data. Ward, Furman, Huebner, Kraus, Guilak, Olson.

Manuscript preparation. Ward, Furman, Huebner, Kraus, Guilak, Olson.

Statistical analysis. Ward, Furman, Huebner.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

We would like to thank Steve Johnson for excellent technical assistance, and Dr. Timothy Griffin and Dr. Holly Leddy for assistance with the histologic grading.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES