High-mobility group box chromosomal protein 1 is an abundant and ubiquitous protein found in the nuclei and cytoplasm of nearly all cells. Its functions include maintenance of nucleosome structure and regulation of gene transcription.1 More recently, it has been reported2,3 that HMGB-1 is secreted into extracellular fluid, where it acts as a proinflammatory cytokine. It mediates many inflammatory diseases, including many forms of experimentally induced and naturally developing arthritis.4–6 Extracellular HMGB-1 engages the receptor for advanced glycation end products (ie, RAGE) to activate proinflammatory signaling pathways.7 In 1 study,8 intra-articular injections of recombinant HMGB-1 caused arthritis in mice, and it was concluded that HMGB-1 is not just an expression of inflammatory responses but initiates inflammation by activating macrophages and inducing IL-1 production via activation of nuclear factor κB. In an immunohistochemical studya of HMGB-1 in articular cartilage and synovial membrane from osteoarthritic patients, HMGB-1 was more prevalent in the cytoplasm than in the nuclei of the cells of these tissues, compared with results for normal samples. In that same study, the total amount of HMGB-1 was not significantly different between osteoarthritic and normal tissues; therefore, it was concluded that the location of this protein within cells influences its proinflammatory role.a In other studies,9,10 concentrations of HMGB-1 were significantly higher in SF of patients with rheumatoid arthritis, compared with concentrations in those with osteoarthritis. Samples of SF from normal joints were not included for comparison. To our knowledge, there is no clear evidence that SF HMGB-1 concentrations are increased with osteoarthritis in any species. Although the authors of the aforementioned studies did not address the use of HMGB-1 as a possible biomarker of joint health, the potential for this application should be considered.
Early identification of joint injury and subsequent osteoarthritis could reduce the resulting loss of use and economic burden.11 Because of the relative insensitivity of current diagnostic methods, such as radiography, diagnosis is often made when osteoarthritis is advanced. Difficulties in early diagnosis of joint injury and osteoarthritis have prompted considerable research for biomarkers, including biochemical markers that reflect quantitative and dynamic variations in joint tissue remodeling.12 Synovial fluid biomarkers have been investigated in horses to identify injury-related changes in joints.13–16 Inflammatory biomarkers, such as prostaglandin E2 and nitric oxide, have been used for equine SF,17 but they have the disadvantage that the assays require relatively large volumes of SF. By comparison, HMGB-1 requires a small sample volume (10 μL), which makes it an attractive potential alternative inflammatory biomarker for use in equine SF.
Osteochondral fractures occur commonly in the carpal, MCP, and MTP joints of racehorses.18,19 Most horses with osteochondral injury develop osteoarthritis. Currently, no studies have been published on the association of osteochondral injury with SF HMGB-1 concentrations in horses. Therefore, the purpose of the study reported here was to investigate the association of osteochondral injury with HMGB-1 concentrations in SF obtained from the MCP, MTP, and carpal joints of racing Thoroughbreds. We hypothesized that SF HMGB-1 concentrations would be higher in joints with osteochondral injury than in normal joints.
Materials and Methods
Animals—Two groups of Thoroughbreds were included in the study. Group 1 consisted of 40 horses (age range, 14 to 20 months) purchased from July through September 2004 for resale at 2-year-old in training sales (February through April 2005). All horses in group 1 were unraced and confirmed healthy with no evidence of joint abnormalities on the basis of clinical and radiographic examinations conducted by an attending veterinarian prior to the study. Group 2 consisted of 45 racing Thoroughbreds (age range, 2 to 6 years) undergoing arthroscopy for removal of osteochondral fragments resulting from racing injury. These fragments were removed from the dorsal articular borders of the third, radiocarpal, and intermediate carpal bones; the distal aspect of the radius; and the proximal phalanx. The study protocol was approved by the University of Florida Institutional Animal Care and Use Committee.
Experimental design—Synovial fluid (1 to 4 mL) was collected from all horses. Horses of group 1 were sedated, and SF samples were collected without lavage from an MCP joint (n = 20 horses), MCJ (10), and RCJ (10). Horses of group 2 were anesthetized, and SF samples were collected without lavage before arthroscopy for removal of osteochondral fragments from an MCP joint (n = 14 horses), MTP joint (4), MCJ (17), or RCJ (10). Synovial fluid samples were aseptically collected via needle arthrocentesis, centrifuged, decanted, and stored at −80°C until assayed.
Procedure for the HMGB-1 immunoassay—Concentrations of HMGB-1 were measured in SF samples by use of a commercially available 2-step sandwich ELISAb in accordance with the manufacturer's instructions. Briefly, a polyclonal antibody specific for human HMGB-1 was precoated onto the wells of the kit, and 10 μL of SF was added. The HMGB-1 in SF bound specifically to the immobilized antibody. After addition of a peroxidase-linked anti–HMGB-1 and -2 monoclonal antibody to form an antigen-antibody complex, color development was initiated by the addition of tetramethyl benzidine. Color developed in proportion to the amount of HMGB-1 in the sample and was measured at 450 nm. Samples of SF were assayed without dilution or digestion.
Arthroscopic and radiographic scores—Available radiographs of osteochondral-injured horses were reviewed, and scores were assigned by use of a scoring system developed by the authors and reported elsewhere.16 Briefly, 10 categories of radiographic changes were graded from 0 to 3 to provide a total radiographic score of 0 to 30. Radiographic scores were determined for joint space, subchondral bone sclerosis and lucency, soft tissue swelling, and size and number of osteophytes, enthesophytes, and fragments. Similarly, available arthroscopic images, videos, and surgical reports for the osteochondral-injured horses were reviewed, and 11 categories were graded to provide a total arthroscopic score of 0 to 37. The total arthroscopic score accounted for 5 categories of inflammation (each graded from 0 to 3), fragment size and number (each graded from 0 to 3), and 4 categories of degenerative cartilage changes related to the fragments (each graded from 0 to 4).16
Statistical analysis—Statistical analysis was performed with personal computer–based statistical software.c Normality plots of the data were assessed. Analysis of box plots identified possible outliers, and extreme Studentized deviate tests were then performed to determine whether the value was > 2 SDs or < 2 SDs from the mean. When the value was > 2 SDs from the mean, it was considered an outlier and eliminated from further analysis. To allow horses to be combined into MCP-MTP joint and carpal joint (MCJ and RCJ) groupings, a 1-way ANOVA was performed to determine whether there were significant differences between the MCP joints, MTP joints, MCJs, and RCJs within horses with normal joints and within osteochondral-injured horses. Differences between normal and osteochondral-injured joints were determined by use of an unpaired t test for the MCP-MTP and carpal joints. Sensitivity, specificity, positive predictive value, negative predictive value, and likelihood ratios of SF HMGB-1 concentrations for identifying osteochondral injury were determined by use of the Fisher exact test. Correlations were assessed via the Spearman correlation.
Discriminant analysis was used to classify the carpal and MCP-MTP joints into the appropriate group (normal or osteochondral-injured) on the basis of SF HMGB-1 concentrations. A quadratic discriminant function was computed from a random sampling of the carpal and MCP-MTP joints by use of separate covariance matrices with the prior probabilities proportional to the population sizes. To determine how well the model could discriminate each joint for its respective group, a subset validation was performed. To eliminate bias, this subset consisted of a random sampling of respective carpal and MCP-MTP joints that were not used in creation of the original model. Values of P < 0.05 were considered significant.
Results
Seven SF HMGB-1 concentrations were considered outliers (2 normal MCP-MTP joints, 1 osteochondral-injured MCP-MTP joint, 2 normal carpal joints, and 2 osteochondral-injured carpal joints) and eliminated from further analysis. There were no MTP joints in the normal group. There was no significant (P = 0.141) difference between the mean ± SD HMGB-1 concentrations for MCP joints (11.17 ± 6.10 ng/mL), MCJs (7.24 ± 5.95 ng/mL), and RCJs (10.45 ± 7.00 ng/mL) in the normal group. There was also no significant (P = 0.493) difference between the mean HMGB-1 concentrations for MCP joints (21.54 ± 15.47 ng/mL), MTP joints (23.00 ± 16.46 ng/mL), MCJs (16.61 ± 6.74 ng/mL), and RCJs (16.49 ± 9.07 ng/mL) in the osteochondral-injured group. Therefore, results of the normal and osteochondral-injured groups for the MCP and MTP joints and for the MCJs and RCJs were combined for further analyses.
Mean SF HMGB-1 concentrations in MCP-MTP joints of the osteochondral-injured group were significantly (P = 0.01) higher than in the MCP-MTP joints of the normal group. Similarly, mean SF HMGB-1 concentrations in osteochondral-injured carpal joints were significantly (P < 0.001) higher than in carpal joints of the normal group (Figure 1).
Analysis of the correlation between SF HMGB-1 concentrations in osteochondral-injured joints and radiographic scores or arthroscopic scores was conducted. For the carpal joints, SF HMGB-1 concentrations did not correlate with the radiographic (r = 0.190; P = 0.386) or arthroscopic (r = 0.275; P = 0.240) scores. Similarly for the MCP-MTP joints, SF HMGB-1 concentrations did not correlate with the radiographic (r = −0.061; P = 0.809) or arthroscopic (r = −0.138; P = 0.573) scores.
Synovial fluid HMGB-1 concentrations ≥ 11 ng/mL for MCP-MTP joints and ≥ 9 ng/mL for carpal joints were arbitrarily chosen for determining predictive values for discriminating osteochondral-injured joints from normal joints. This yielded a positive predictive value of 58% and a negative predictive value of 77% for MCP-MTP joints and a positive predictive value of 75% and negative predictive value of 88% for the carpal joints (Table 1). In other words, horses with SF HMGB-1 concentrations ≥ 11 ng/mL for MCP-MTP joints were twice as likely to have an osteochondral injury, and horses with SF HMGB-1 concentrations ≥ 9 ng/mL for carpal joints were 4 times as likely to have an osteochondral injury.
Sensitivity, specificity, positive predictive value, negative predictive value, and likelihood ratio for use of SF HMGB-1 concentrations to discriminate between joints of Thoroughbreds in the normal (n = 40 horses) and osteochondral-injured (45) groups.
Joint | Sensitivity (%) | Specificity (%) | Positive predictive value (%) | Negative predictive value (%) | Likelihood ratio |
---|---|---|---|---|---|
MCP-MTP joints* | 74 | 63 | 58 | 77 | 2 |
Carpal joints† | 86 | 78 | 75 | 88 | 4 |
Cutoff value for the SF HMGB-1 concentration was ≥ 11 ng/mL.
Cutoff value for the SF HMGB-1 concentration was ≥ 9 ng/mL.
For horses with osteochondral-injured MCP-MTP joints, the SF HMGB-1 concentration made a significant contribution to the discriminant analysis model. The discriminant model was developed by use of 30 randomly selected horses (17 in the normal group and 13 in the osteochondral-injured group). This model was significantly able to separate the groups by overall correctly classifying 20 (66.7%) of the horses. In an attempt to verify the predictive nature of this model, an additional subset of 16 horses that were not used to create this model was used, including 10 in the normal group and 6 in the osteochondral-injured group. This discriminant analysis model for the MCP-MTP joints correctly classified 13 of these 16 (81.3%) horses (10/10 in the normal group and 3/6 in the osteochondral-injured group), which suggested that this model can be used to correctly classify a horse into its appropriate group 8 out of 10 times on the basis of SF HMGB-1 concentrations. The model was more accurate for use in identifying horses in the normal group than in the osteochondral-injured group.
For horses with osteochondral-injured carpal joints, the SF HMGB-1 concentration made a significant (P < 0.001) contribution to the discriminant analysis model. The discriminant model was developed by use of 45 randomly selected horses (29 in the normal group and 16 in the osteochondral-injured group). This model was significantly (P < 0.001) able to separate the groups by overall correctly classifying 33 (73.3%) of the horses. In an attempt to verify the predictive nature of this model, an additional subset of 20 horses that were not used to create this model was used, including 8 in the normal group and 12 in the osteochondral-injured group. This discriminant analysis model for the carpal joints correctly classified 14 of these 20 (70%) horses (7/8 in the normal group and 7/12 in the osteochondral-injured group), which suggested that this model can be used to correctly classify a horse into its appropriate group 7 out of 10 times on the basis of SF HMGB-1 concentrations. The model was more accurate for use in identifying horses in the normal group than in the osteochondral-injured group.
Discussion
Osteochondral injury was associated with significantly higher SF HMGB-1 concentrations in MCP-MTP and carpal joints, compared with concentrations in normal joints. It has been suggested that HMGB-1 plays a role in both the inflammatory and destructive processes of rheumatoid arthritis and adjuvant-induced arthritis.4 To date, no association has been detected between osteoarthritis and increases in SF HMGB-1 concentrations. In addition, HMGB-1 has been identified in chondrocytes and synoviocytes.4 In 1 study,a investigators determined that articular cartilage and synovial membrane from humans with osteoarthritis differed from normal tissues in the cellular location of HMGB-1 but not the total content. The increase in SF HMGB-1 concentrations in the horses reported here indicated that osteochondral injury was associated with release of HMGB-1 from these cells into the SF. This would require translocation of HMGB-1 from the cells into the extracellular fluid to cause an increase in SF concentrations of HMGB-1.
The release mechanism for HMGB-1 is not clearly understood.20 Macrophages and monocytes in synovial membranes translocate HMGB-1 from the nucleus to the cytoplasm before it is released from cells.4 In a recent study,10 it was found that hypoxia causes substantial secretion or release of HMGB-1 from many kinds of cells. The investigators of that study suggested that the appearance of extracellular HMGB-1 might be associated with tissue hypoxia rather than with nuclear factor κB–mediated inflammatory pathways. They reported that HMGB-1 concentrations were significantly correlated with those of lactic acid, a marker of tissue hypoxia. Furthermore, those investigators proposed that the role of hypoxia in HMGB-1 release suggests a new concept for treatment. Treatments (such as hyperbaric oxygen) that reduce hypoxia may decrease extracellular HMGB-1 concentrations and thereby reduce inflammation and cartilage degradation during arthritis.10
A reciprocal relationship between the early (TNF and IL-1) and late (HMGB-1) cytokines has been proposed.21 Early proinflammatory cytokines have been defined as those released within the first few hours after onset of endotoxemia, whereas late cytokines are released later during endotoxemia.21 Although osteoarthritis does not involve endotoxemia, these terms serve a useful purpose in the description of the inflammatory response. Interferon-γ, TNF, and IL-1 cause HMGB-1 release.2,22 On the other hand, HMGB-1 stimulates monocytes to release TNF, IL-1α, IL-1β, IL-6, IL-8, macrophage inflammatory protein-1α, and macrophage inflammatory protein-1β.21 It has been suggested23 that TNF and IL-1 play key roles in the pathogenesis of osteoarthritis. Both of these cytokines increase the production of matrix metalloproteinases (1, 3, 9, and 13) and a disintegrin and metalloprotease with thrombospondin motifs 4 (ie, ADAMTS4).23 This strongly suggests that the interaction of HMGB-1 with other cytokines is of considerable importance in osteoarthritis. It would have been interesting to compare HMGB-1 concentrations with concentrations of other cytokines or markers of tissue hypoxia in the SF of these horses, but unfortunately, adequate volumes of SF were not available. Additional studies need to be conducted to determine this interaction.
We found it useful to use different cutoff values for SF HMGB-1 (≥ 11 ng/mL for MCP-MTP joints and ≥ 9 ng/mL for carpal joints) to calculate predictive values and likelihood ratios. The assay yielded good positive and negative predictive values, with higher values for the carpal joints (Table 1). Use of different cutoff values for these 2 joints for calculation of predictive values is consistent with the methods used in other studies16,24,d conducted by our laboratory group. Thus, it may not be prudent to extrapolate biomarker values among joints, even though there were no significant differences in HMGB-1 concentrations among joints in this study.
The discriminant analysis model may prove to be useful in the diagnostic application for biomarkers of joint injury and disease. When separating normal and injured carpal or MCP-MTP joints, discriminant analysis yielded results similar to positive and negative predictive values. The model performed well at identifying carpal and MCP-MTP joints in the normal group.
In other studies conducted by our laboratory group, we have detected a correlation between radiographic and arthroscopic scores of joint injury and biomarker concentrations in SF or serum. These correlations were detected for serum concentrations and SF-to-serum concentration ratios of bone alkaline phosphatase,16 an enzyme believed to contribute to calcification of bone matrix25 and thus assumed to be a biomarker of bone synthesis. We have also detected a correlation between collagen degradation biomarkers in serum and SF with radiographic and arthroscopic scores.24,d Therefore, we expected to detect a correlation of SF HMGB-1 concentrations, presumably a biomarker of joint inflammation, with radiographic and arthroscopic scores of joint injury. We have no definitive explanation for this lack of correlation. However, complete medical records were not available for some of the horses with osteochondral injury. Previous treatment and the duration of the osteochondral injury before collection of SF samples typically were unknown, which is often a limitation of clinical case material. Racehorses with musculoskeletal injury are commonly treated with anti-inflammatory drugs, such as intra-articularly administered corticosteroids, as well as phenylbutazone, an NSAID. It is possible that administration of these drugs may have indirectly affected SF HMGB-1 concentrations and thus may have accounted for the lack of correlation with radiographic and arthroscopic scores. In rodents with experimentally induced sepsis, extremely high doses of corticosteroids (such as dexamethasone and cortisone) and NSAIDs (such as aspirin, ibuprofen, and indomethacin) failed to reduce high serum concentrations of HMGB-1 caused by endotoxemia.26 However, in a study27 of 31 humans with chronic arthridites, evaluation of synovial membranes revealed that intra-articular administration of corticosteroids virtually eliminated extracellular HMGB-1 staining and markedly reduced cytoplasmic HMGB-1 staining in synoviocytes and macrophage-like cells. The potential effect of these drugs on SF HMGB-1 concentrations in the study reported here is unknown, but it may have been a factor. Thus, although there was no correlation between radiographic and arthroscopic scores and SF HMGB-1 concentrations, the assay was of use in discriminating between osteochondral-injured and normal joints. This lack of correlation would not preclude the use of HMGB-1 as a biomarker of joint injury. Additional studies in animals with experimentally induced osteoarthritis may allow further characterization of the effect of anti-inflammatory drugs on the HMGB-1 response.
The study reported here had other limitations. It was determined that 7 HMGB-1 concentrations were outliers; thus, they were eliminated from further analysis. Even though this was performed in accordance with accepted statistical analyses, it increased the possibility of a type I error. In other words, there was a slightly greater chance of stating that a difference existed between groups when there actually was no difference. We do not believe that elimination of these 7 concentrations dramatically affected our findings because there was no bias toward any one group (2 MCP-MTP joints in the normal group, 1 MCP-MTP joint in the osteochondral-injured group, 2 carpal joints in the normal group, and 2 carpal joints in the osteochondral-injured group were eliminated). Ages of horses in the normal and osteochondral-injured groups differed (horses in the normal group were younger). Although synthesis of HMGB-1 decreases in the liver with age,28 the effect of age on SF concentrations is unknown. We evaluated only SF concentrations of HMGB-1. It is possible that comparison of these values with those in serum or urine may have provided useful information regarding the relative concentrations of HMGB-1 between these compartments.
Another limitation of the study was that a full validation of the assay for application in equine SF was not performed. However, the equine HMGB-1 protein sequence has 99% homology with that of human HMGB-1e; therefore, there is a high likelihood that the assay is valid in equine body fluids.
In the study reported here, we determined that osteochondral injury was associated with a significant increase in SF HMGB-1 concentration in the MCP-MTP and carpal joints of horses. Synovial fluid HMGB-1 concentrations were not correlated with radiographic or arthroscopic scores. Selection of arbitrary SF HMGB-1 concentrations for each joint yielded high negative predictive values and positive predictive values for the discrimination between normal and osteochondral-injured joints. This provides a rationale for using different SF HMGB-1 concentrations to establish criteria for the use of this biomarker as a diagnostic tool for various joints. On the basis of these findings, SF HMGB-1 analysis may be useful for evaluation of joint injury in horses.
ABBREVIATIONS
HMGB-1 | High-mobility group box chromosomal protein 1 |
IL | Interleukin |
MCJ | Middle carpal joint |
MCP | Metacarpophalangeal |
MTP | Metatarsophalangeal |
RCJ | Radiocarpal joint |
SF | Synovial fluid |
TNF | Tumor necrosis factor |
Zheng Y, Dong X, Lu J, et al. A study on high mobility group box chromosomal protein 1 in articular cartilage and synovial membranes in osteoarthritis (abstr). APLAR J Rheumatol 2006;9(suppl 1):A40.
HMGB-1, IBL-Transatlantic Corp, Toronto, ON, Canada.
SPSS, version 15.0 for Windows, SPSS Inc, Chicago, Ill.
Cleary OB, Trumble TN, Brown MP, et al. Effects of exercise and osteochondral fragmentation on CTX-II concentrations in equine synovial fluid and serum (abstr). Trans Orthop Res Soc 2008;33:746.
Equine HMGB-1 protein sequence (accession No. BAF33339); human HMGB-1 protein sequence (accession No. CAG33144); protein BLAST program. Available at: blast.ncbi.nlm.nih.gov/Blast.cgi. Accessed Oct 15, 2008.
References
- 1.↑
Goodwin GH, Sanders C, Johns EW. A new group of chromatin-associated proteins with a high content of acidic and basic amino acids. Eur J Biochem 1973;38:14–19.
- 2.
Wang H, Bloom O, Zhang M, et al. HMG-1 as a late mediator of endotoxin lethality in mice. Science 1999;285:248–251.
- 3.
Yang H, Wang H, Czura CJ, et al. The cytokine activity of HMGB1. J Leukoc Biol 2005;78:1–8.
- 4.↑
Kokkola R, Sundberg E, Ulfgren AK, et al. High mobility group box chromosomal protein 1: a novel proinflammatory mediator in synovitis. Arthritis Rheum 2002;46:2598–2603.
- 5.
Wittemann B, Neuer G, Michels H, et al. Autoantibodies to nonhistone chromosomal proteins HMG-1 and HMG-2 in sera of patients with juvenile rheumatoid arthritis. Arthritis Rheum 1990;33:1378–1383.
- 6.
Palmblad K, Sundberg E, Diez M, et al. Morphological characterization of intra-articular HMGB1 expression during the course of collagen-induced arthritis. Arthritis Res Ther 2007;9:R35–R45.
- 7.↑
Kokkola R, Andersson A, Mullins G, et al. RAGE is the major receptor for the proinflammatory activity of HMGB1 in rodent macrophages. Scand J Immunol 2005;61:1–9.
- 8.↑
Pullerits R, Jonsson IM, Verdrengh M, et al. High mobility group box chromosomal protein 1, a DNA binding cytokine, induces arthritis. Arthritis Rheum 2003;48:1693–1700.
- 9.
Taniguchi N, Kawahara K, Yone K, et al. High mobility group box chromosomal protein 1 plays a role in the pathogenesis of rheumatoid arthritis as a novel cytokine. Arthritis Rheum 2003;48:971–981.
- 10.↑
Hamada T, Torikai M, Kuwazuru A, et al. Extracellular high mobility group box chromosomal protein 1 is a coupling factor for hypoxia and inflammation in arthritis. Arthritis Rheum 2008;58:2675–2685.
- 13.
Brown MP, Trumble TN, Plaas AH, et al. Exercise and injury increase chondroitin sulfate chain length and decrease hyaluronan chain length in synovial fluid. Osteoarthritis Cartilage 2007;15:1318–1325.
- 14.
Frisbie DD, Ray CS, Ionescu M, et al. Measurement of synovial fluid and serum concentrations of the 846 epitope of chondroitin sulfate and of carboxy propeptides of type II procollagen for diagnosis of osteochondral fragmentation in horses. Am J Vet Res 1999;60:306–309.
- 15.
Fuller CJ, Barr AR, Sharif M, et al. Cross-sectional comparison of synovial fluid biochemical markers in equine osteoarthritis and the correlation of these markers with articular cartilage damage. Osteoarthritis Cartilage 2001;9:49–55.
- 16.↑
Trumble TN, Brown MP, Merritt KA, et al. Joint dependent concentrations of bone alkaline phosphatase in serum and synovial fluids of horses with osteochondral injury: an analytical and clinical validation. Osteoarthritis Cartilage 2008;16:779–786.
- 17.↑
van den Boom R, van de Lest CH, Bull S, et al. Influence of repeated arthrocentesis and exercise on synovial fluid concentrations of nitric oxide, prostaglandin E2 and glycosaminoglycans in healthy equine joints. Equine Vet J 2005;37:250–256.
- 18.
McIlwraith CW, Yovich JV, Martin GS. Arthroscopic surgery for the treatment of osteochondral chip fractures in the equine carpus. J Am Vet Med Assoc 1987;191:531–540.
- 19.
Yovich JV, McIlwraith CW. Arthroscopic surgery for osteochondral fractures of the proximal phalanx of the metacarpophalangeal and metatarsophalangeal (fetlock) joints in horses. J Am Vet Med Assoc 1986;188:273–279.
- 20.↑
Chen G, Ward MF, Sama AE, et al. Extracellular HMGB1 as a proinflammatory cytokine. J Interferon Cytokine Res 2004;24:329–333.
- 21.↑
Andersson U, Wang H, Palmblad K, et al. High mobility group 1 protein (HMG-1) stimulates proinflammatory cytokine synthesis in human monocytes. J Exp Med 2000;192:565–570.
- 22.
Rendon-Mitchell B, Ochani M, Li J, et al. IFN-gamma induces high mobility group box 1 protein release partly through a TNFdependent mechanism. J Immunol 2003;170:3890–3897.
- 23.↑
Bondeson J, Wainwright SD, Lauder S, et al. The role of synovial macrophages and macrophage-produced cytokines in driving aggrecanases, matrix metalloproteinases, and other destructive and inflammatory responses in osteoarthritis. Arthritis Res Ther 2006;8:R187–R199.
- 24.↑
Trumble TN, Scarbrough AB, Brown MP. Osteochondral injury increases type II collagen degradation products (C2C) in synovial fluid of Thoroughbred racehorses. Osteoarthritis Cartilage 2009;17:371–374.
- 25.↑
Beertsen W, van den Bos T. Alkaline phosphatase induces the mineralization of sheets of collagen implanted subcutaneously in the rat. J Clin Invest 1992;89:1974–1980.
- 26.↑
Li W, Li J, Ashok M, et al. A cardiovascular drug rescues mice from lethal sepsis by selectively attenuating a late-acting proinflammatory mediator, high mobility group box 1. J Immunol 2007;178:3856–3864.
- 27.↑
af Klint E, Grundtman C, Engstrom M, et al. Intraarticular glucocorticoid treatment reduces inflammation in synovial cell infiltrations more efficiently than in synovial blood vessels. Arthritis Rheum 2005;52:3880–3889.
- 28.↑
Thakur MK, Prasad S. Analysis of age-associated alteration in the synthesis of HMG nonhistone proteins of the rat liver. Mol Biol Rep 1991;15:19–24.