SEARCH

SEARCH BY CITATION

Abstract

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

Objective

To examine the potential role of high mobility group box chromosomal protein 1 (HMGB-1) in the pathogenesis of arthritis.

Methods

Mice were injected intraarticularly with 1 μg or 5 μg of HMGB-1. Joints were dissected on days 4, 7, and 28 after injection and were evaluated histopathologically and immunohistochemically. To investigate the importance of different white blood cell populations for the development of arthritis, in vivo cell depletion procedures were performed. In addition, spleen cells were cultured in the presence of HMGB-1, and nuclear factor κB (NF-κB) activation was detected by electrophoretic mobility shift assay.

Results

Injection of recombinant HMGB-1 (rHMGB-1) into different mouse strains resulted in an overall frequency of arthritis in 80% of the animals. The inflammation was characterized by mild to moderate synovitis and lasted for at least 28 days. The majority of cells found in the inflamed synovium were Mac-1+ macrophages, whereas only a few CD4+ lymphocytes were detected. Pannus formation was observed in some cases 7 and 28 days after HMGB-1 injection. No significant differences were found with respect to incidence and severity of arthritis between mice depleted of monocytes, granulocytes, or lacking T/B lymphocytes. However, combined removal of monocytes and neutrophils resulted in a 43% lower incidence of arthritis. Mice rendered deficient in the interleukin-1 (IL-1) receptor did not develop inflammation upon challenge with HMGB-1. In vitro data corroborate this finding, showing that rHMGB-1 activated NF-κB, a major pathway leading to IL-1 production.

Conclusion

Our results indicate that HMGB-1 is not a mere expression of inflammatory responses, but on its own, it triggers joint inflammation by activating macrophages and inducing production of IL-1 via NF-κB activation.

High mobility group box chromosomal protein 1 (HMGB-1), named for its rapid mobility on electrophoresis gels, is a ubiquitous, nonhistone, chromatin-associated 215–amino acid protein with highly conserved amino acid sequence identity between rodents and humans (1–3). Nuclear HMGB-1 has been identified and studied for a long time as a DNA binding protein. It participates in maintenance of nucleosomal structure and stability and facilitates the binding of transcription factors to their cognate DNA sequences (4). HMGB-1 also has functions in DNA transcription, recombination (5, 6), repair, cell replication, cell migration, and tumor growth (7, 8).

In contrast to its intranuclear role, extracellular HMGB-1 was recently shown to act as a cytokine mediating delayed endotoxin lethality (9) as well as acute lung injury in mice (10). Moreover, high levels of HMGB-1 have been detected in the blood of patients with sepsis (9) and in the synovial fluid of rheumatoid arthritis (RA) patients (11). Proinflammatory mediators, such as tumor necrosis factor α (TNFα) and interleukin-1 (IL-1), can dose-dependently induce the release of HMGB-1 from monocytes and macrophages (9). Furthermore, once released, HMGB-1 itself can activate an additional downstream cascade by stimulating monocytes to produce proinflammatory cytokines and chemokines (e.g., TNFα, IL-1α, IL-1β, macrophage inflammatory protein 1α, IL-6, and IL-8), whereas the cytokine response to HMGB-1 is totally confined to the monocyte/macrophage population (12). Since proinflammatory cytokine release is important in the mediation of arthritis, we reasoned that HMGB-1 might have a direct role in the development of arthritis.

MATERIALS AND METHODS

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

Mice.

Female NMRI and BALB/c mice, 6–8 weeks old, were purchased from B&K Universal (Stockholm, Sweden). SCID mice and their congenic strain CB17 mice were obtained from M&B (Ry, Denmark). Breeding pairs of type I IL-1R−/− mice and their congenic counterpart C57BL/6 mice were purchased from The Jackson Laboratory (Bar Harbor, ME). The mice were bred and housed in the animal facility of the Department of Rheumatology and Inflammation Research, University of Göteborg. They were kept under standard conditions of temperature and light, and were fed laboratory chow and water ad libitum. The study was approved by the Ethics Committee of the University of Göteborg, and the requirements of the National Board for Laboratory Animals were followed.

Reagents.

Mouse recombinant HMGB-1 (rHMGB-1), displaying 100% homology with rat HMGB-1, was expressed in Escherichia coli and purified to homogeneity as previously described (9). Preparations were tested routinely for lipopolysaccharide (LPS) content by the chromogenic Limulus amebocyte cell lysate assay (12). All of them contained <400 pg endotoxin/μg rHMGB-1. The rHMGB1 was dissolved in Tris buffer and kept at −80°C until used. Etoposide was purchased from Bristol-Myers Squibb (Bromma, Sweden) and diluted 1:10 in phosphate buffered saline (PBS) (0.13M NaCl, 10 mM sodium phosphate, pH 7.4) from a stock solution of 20 mg/ml. The hybridoma cells secreting monoclonal antibody RB6-8C5 were a kind gift from Dr. R. Coffmann (DNAX Research Institute, Palo Alto, CA). The antibody was produced and purified as described previously (13). LPS from E coli serotype O55:B5 and murine serum albumin were purchased from Sigma (St. Louis, MO).

Injection protocol.

Following systemic (Hypnorm and Dormicum) or inhalation (Isoflurane) anesthesia, mice were injected intraarticularly in the knee joint with a total volume of 20 μl of solution containing 1 μg or 5 μg HMGB-1 in the first experiment and 5 μg HMGB-1 in the following experiments. Contralateral knee joints of mice injected with HMGB-1 or knee joints of separate control mice received an equivalent volume of PBS or Tris buffer with or without a matching concentration of LPS. Since the results in controls were similar irrespective of whether contralateral knee joints or separate mice were used, we decided to pool these results. Knee joints of control mice (or contralateral knee joints of experimental mice) were injected with 2 ng LPS (n = 3), 0.4 ng LPS (n = 3), 1 ng LPS (n = 5), or 10 μg of murine serum albumin (n = 10). All of these mice were killed after 4 days for histological evaluation of joint appearance. Mice used as controls 7 days following injection consisted of the following subgroups: 2 ng LPS (n = 3), 0.4 ng LPS (n = 3), 1 ng LPS (n = 5), or 10 μg murine serum albumin (n = 10). In addition, knee joints of 6 mice were left unmanipulated.

In vivo cell depletion procedures.

Monocyte depletion.

Etoposide is a drug that has been shown to selectively deplete the monocyte population in mice (14, 15). CB17 mice were injected subcutaneously with 70–150 μl of etoposide (corresponding to 12.5 mg/kg) (14) once a day during the experiment, starting 2 days before the injection of HMGB-1. Mice were monitored and weighed individually once a day, and the dose of etoposide was adjusted to body weight.

Neutrophil depletion.

Monoclonal antibody (mAb) RB6-8C5 is a rat immunoglobulin (IgG2b) that selectively depletes mature mouse neutrophils. CB17 mice were injected intraperitoneally with 1 mg of RB6-8C5 mAb 2 hours before and on days 2 and 6 after the administration of HMGB-1.

Monocyte and neutrophil depletion.

BALB/c mice were injected subcutaneously with 70–150 μl of etoposide once a day during the experiment, starting 3 days before the injection of HMGB-1. Mice were monitored and weighed individually once a day, and the dose of etoposide was adjusted to body weight. Simultaneously, mice were injected intraperitoneally with 200 μg of RB6-8C5 mAb 24 hours and 1 hour before and with 1 mg of RB6-8C5 mAb on days 3 and 5 after the injection of HMGB-1.

Hematologic analyses.

To assess the effect of cell depletion, CB17 mice were bled from a tail vein into heparinized tubes, and total leukocyte counts were determined in a cell counter (Sysmex KX-21; Toa Medical Electronics, Kobe, Japan). To evaluate the percentage of lymphocytes, monocytes, and granulocytes, peripheral blood leukocytes (104 cells/sample) were analyzed on a FACScan flow cytometer (Becton Dickinson, San Jose, CA). The absolute numbers of different leukocyte subsets were then calculated from the total leukocyte counts. Results of the flow cytometric analysis demonstrated that monocytes decreased significantly compared with the levels in untreated animals (mean ± SEM 32.2 ± 4.4 × 104/ml versus 8.6 ± 1.7 × 104/ml; P < 0.04), which has also been demonstrated previously (14). Antineutrophil mAb RB6-8C5 depleted the granulocyte population to a significant degree (13). Before the injection of HMGB-1, the animals pretreated with RB6-8C5 had a significantly lower level of circulating granulocytes than untreated controls (mean ± SEM 40.5 ± 8.5 × 104/ml versus 5.7 ± 0.8 × 104/ml).

Histopathologic examination.

Four, 7, and 28 days after the injection of HMGB-1 and the control buffer, the mice were killed and knee joints were removed, fixed in 4% paraformaldehyde, decalcified, and embedded in paraffin. Tissue sections were stained with hematoxylin and eosin. All the slides were coded and assessed by 3 observers (RP, I-MJ, AT) in a blinded manner with regard to synovial hypertrophy (membrane thickness >2 layers), pannus formation (synovial tissue overlaying joint cartilage), and cartilage and bone destruction (16). The extent of synovitis was judged on a scale from grade 0 to grade 3 (0 = no signs of inflammation, 1 = mild synovial hypertrophy consisting of up to 5 cell layers, 2 = moderate inflammation characterized by hyperplasia of synovial membrane up to 10 cell layers and influx of inflammatory cells throughout the synovial tissue, and 3 = marked synovial hypertrophy consisting of >10 cell layers, and the synovial tissue infiltrated by inflammatory cells).

Immunohistochemical evaluation.

Seven days after the injection of HMGB-1 and the control buffer, the mice were killed and knee joints were removed and demineralized by a procedure detailed previously (17). The demineralized specimens were then embedded (Tissue-Tek; Miles, Elkhart, IN), frozen in isopentane prechilled with liquid nitrogen, and kept at −80°C until cryosectioned. Sections (5 μm thick) were cut sagitally, fixed in cold acetone for 5 minutes, washed in PBS, and thereafter incubated overnight in a humid chamber at 4°C with unlabeled primary rat monoclonal anti-CD11b (Mac-1, clone M1/70) (18), anti-CD4 (H129.19) (19), anti-CD8 (Lyt 2; PharMingen, San Diego, CA), or bovine serum albumin (BSA) as a negative control, at appropriate dilutions. Biotin-labeled rabbit anti-rat IgG (Vector, Burlingame, CA) was used as secondary antibody followed by depletion of endogenous peroxidase with H2O2 and incubation with StreptABComplex (Dako, Glostrup, Denmark) and 3-amino-9-ethyl-carbazole buffer containing H2O2. All sections were counterstained with Mayer's hematoxylin.

Nuclear extracts.

Spleens were obtained from healthy NMRI mice, and splenocyte cultures were prepared as described previously (14). Cells (2 × 106/ml) were incubated for 2 hours at 37°C, 5 ml/well in a 6-well microtiter plate with different concentrations of rHMGB-1. The stimulation was stopped with ice-cooled PBS, and the cell suspension was washed twice with ice-cooled PBS, resuspended in 2 ml of hypoton buffer (10 mM HEPES [pH 7.9] at 4°C, 0.1 mM EDTA, 0.1 mM EGTA, 10 mM KCl, 0.75 mM spermidin, 0.15 mM spermin, 1M dithiothreitol [DTT], and proteinase inhibitors [Complete Mini Tablets; 1 tablet/10 ml] [Roche Diagnostics, Mannheim, Germany]), and homogenized. Following centrifugation at 14,000g for 10 minutes at 4°C, the supernatant was carefully removed and the pellet was resuspended in ice-cold extraction buffer (20 mM HEPES [pH 7.9] at 4°C, 25% glycerol, 0.42M NaCl, 1 mM EDTA, 1 mM EGTA, 1M DTT, and proteinase inhibitors [Complete Mini Tablets; 1 tablet/10 ml]). After extraction at 4°C overnight on a rotator, the tubes were centrifuged at 14,000g for 1 hour at 4°C. The supernatant was collected and stored at −80°C in aliquots until the analysis of nuclear factor κB (NF-κB) could be performed. Before freezing, the protein concentration was determined using Bradford reagent (Sigma).

Electrophoretic mobility shift assay (EMSA).

EMSA was performed as described previously (20), with minor modifications. The sequences of synthetic oligonucleotides used for the NF-κB probes were as follows: sense 5′-GGCTCAAACAGGGGGCTTTCCCTCCTCAATAT-3′, antisense 5′-GGATATTGAGGAGGGAAAGCCCCCTGTTTGAG-3′. Oligonucleotides were annealed at 56°C. The double-stranded product was purified by elution from the electrophoretic gel. Double-stranded oligonucleotides were labeled with α-32P-labeled deoxynucleotide (Amersham Pharmacia Biotech, Uppsala, Sweden) using Klenow polymerase (Roche Diagnostics) and then used in the binding reaction with the nuclear protein extracts.

Binding reaction mixtures in a volume of 20 μl containing the same amount of nuclear protein (4.7 μg), 5 μg of poly(dI-dC)poly(dI-dC) (Amersham Pharmacia Biotech), 1 μl 32P-labeled probe containing NF-κB binding sites, 1 mM DTT, 0.2 mg/ml BSA, and binding buffer (0.02M Tris HCl [pH 7.9], 30 mM NaCl, 5 μM EGTA, 5% glycerol) were mixed and incubated for 20 minutes at room temperature. The final NaCl concentration was adjusted to 80 mM equally in all samples. The complexes were separated by electrophoresis through a native 5% polyacrylamide gel containing 0.25× Tris–borate–EDTA, 3% glycerol, and 0.1% ammonium persulfate. The gel was vacuum-dried and exposed to x-ray film for 48 hours at −80°C. For the competition assay, the nuclear extract was preincubated with an excess of unlabeled NF-κB oligonucleotide for 10 minutes on ice prior to the addition of the radiolabeled probe. For the supershift assay, nuclear extract was incubated with 1 μl of p50 antibody (Santa Cruz Biotechnology, Heidelberg, Germany) for 20 minutes on ice prior to EMSA.

Statistical analysis.

Statistical comparisons were made using the Mann-Whitney U test, Kruskal-Wallis nonparametric test, and Fisher's exact test. All values are expressed as the mean ± SEM. P values less than 0.05 were considered significant.

RESULTS

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

Induction of arthritis by HMGB-1 in the healthy host.

To assess the potential inflammatory role of HMGB-1, mice were given a single intraarticular injection of 1 μg or 5 μg of this protein. Macroscopic signs of arthritis were not clinically observed in any of the joints. NMRI mice injected with 5 μg of HMGB-1 developed arthritis in 92% and 79% of cases, in contrast to 9% and 6% of mice injected with control buffer after 4 and 7 days, respectively (Figure 1A). Moreover, a 5-fold lower dose (1 μg) triggered arthritis in 67% of mice after 4 days (Figure 1A). The severity of arthritis was most evident on days 4 and 7 after administration of 5 μg HMGB-1 (Figure 1B). Even 28 days after injection of HMGB-1, 50% of mice displayed mild synovitis, indicating that this molecule triggers long-lasting inflammation in the singularly exposed joint.

thumbnail image

Figure 1. Incidence of arthritis (A) and its severity (B) in NMRI mice 4 (n = 13), 7 (n = 19), and 28 (n = 6) days after intraarticular injection with 5 μg of high mobility group box chromosomal protein 1 (HMGB-1). Incidence and arthritis severity scores after administration of 1 μg HMGB-1 from another experiment (n = 6) are also included. The results of control (phosphate buffered saline [PBS] + lipopolysaccharide [LPS]) injections are pooled since they were similar irrespective of whether contralateral knee joints or joints from separate mice were used (see Materials and Methods). The second control group of mice received 10 μg of murine serum albumin and joints were histologically evaluated 4 days (n = 10) and 7 days (n = 10) after the injections. The statistical significance of arthritis frequency was calculated using Fisher's exact test, and differences between groups with respect to arthritis severity were tested for significance using the Mann-Whitney U test. Values are the mean and SEM.

Download figure to PowerPoint

We examined different healthy mouse strains (NMRI, C57BL/6, and CB17) to assess their susceptibility to arthritis. HMGB-1 was arthritogenic in all of these strains. Joint inflammation occurred in 80% and 79% of CB17 and NMRI mice, respectively, 7 days after injection of 5 μg HMGB-1. However, in the case of C57BL/6 mice, only 40% of healthy recipients developed arthritis (data not shown). Also, the severity of arthritis in C57BL/6 mice was significantly lower compared with that in NMRI and CB17 mice (Figure 2).

thumbnail image

Figure 2. Severity of arthritis in different healthy mouse strains: NMRI (n = 19), CB17 (n = 10), and C57BL/6 (n = 10) 7 days after a single administration of 5 μg of high mobility group box chromosomal protein 1. Differences between groups regarding arthritis severity were tested for significance using the Mann-Whitney U test. Values are the mean and SEM.

Download figure to PowerPoint

To exclude the possible role of LPS in the induction of arthritis, control mice were injected with an LPS dose corresponding to that present in the HMGB-1 preparation used for the injection. Signs of arthritis were not seen in mice receiving 2 ng LPS, which is the amount found in the highest dose of HMGB-1 (i.e., 5 μg) injected. These findings indicate that the induction of arthritis was due to HMGB-1 rather than to LPS contamination. To further ascertain the specificity of arthritis triggered by HMGB-1, we injected highly purified murine serum albumin (10 μg/joint) as an irrelevant protein control. Joints were histologically evaluated 4 days (n = 10) and 7 days (n = 10) after the injections. Minor inflammatory infiltrates were seen in 20% of mice 4 days after the albumin injection, whereas no signs of arthritis were found in any of the 10 joints examined on day 7 (Figures 1A and B).

Histopathologic characterization of HMGB-1–triggered inflammation in the joints.

Morphologic evaluation of the injected joints with respect to influx of inflammatory cells into synovial tissue, hypertrophy of synovial membrane, and pannus formation was performed on days 4, 7, and 28 after administration of HMGB-1. Arthritis was characterized by mild to moderate synovitis. There was an influx of inflammatory cells throughout the synovial tissue as well as around blood vessels. Also, the synovial membrane lining layer was clearly thickened (Figures 3A and B). The highest arthritis severity scores were found on days 4 and 7, whereas they significantly decreased by day 28. In addition, pannus formation (Figure 3C) and cartilage erosions (Figure 3D) were observed in some cases 7 days after injection of 5 μg of HMGB-1. Pannus formation was occasionally observed until the termination of the experiment (day 28) (Figure 3E).

thumbnail image

Figure 3. A–E, Impact of injection of either control buffer (A) or high mobility group box chromosomal protein 1 (HMGB-1) (B–E) on mouse knee joints. Photomicrographs showing a normal NMRI mouse knee joint from a control animal (A) and the histopathologic appearance of arthritis in a knee joint injected with HMGB-1 (B) obtained on day 7 after injection of control buffer or 5 μg of HMGB-1, respectively. Influx of inflammatory cells into synovial tissue (ST) (arrows) is evident; arrowheads indicate thickened synovial membrane lining layer. Pannus (C and D) (arrowheads) and cartilage erosions (D) (arrows) were occasionally observed as early as 7 days after injection with HMGB-1. Twenty-eight days after HMGB-1 injection, fibrotic changes (E) (arrow) in the ST in parallel with pannus (E) (arrowheads) were observed. Histologic staining was performed using hematoxylin and eosin. F and G, Immunohistochemical staining of NMRI mouse knee joints obtained 7 days after injection with 5 μg of HMGB-1 for Mac-1 (brown) in most inflammatory cells in synovial tissue is observed (F) (arrowheads). In control studies (G), the primary, Mac-1 antibody was omitted. JC = joint cavity; C = cartilage; V = blood vessel; BM = bone marrow. (Original magnification × 200 in A–E; × 800 in F and G.)

Download figure to PowerPoint

Immunohistochemical features of arthritic joints.

In sections of arthritic joints obtained 7 days after administration of HMGB-1, a large fraction of inflammatory cells was stained with Mac-1 antibody, which targets granulocytes and monocyte/macrophages. These cells were found both within the thickened synovial membrane lining layer as well as in the deeper synovial tissue constituting the major population of inflammatory cells (Figures 3F and G). A majority of these cells displayed macrophage morphology, i.e., mononuclear rather than polymorphonuclear cells. In contrast, only a few CD4+ lymphocytes and no CD8+ cells were detected.

Effect of monocytes and granulocytes on HMGB-1–induced inflammation.

To evaluate the role of different leukocyte populations in HMGB-1–triggered arthritis, CB17 mice were rendered monocytopenic or neutropenic by using multiple injections of etoposide and antineutrophil mAb RB6-8C5, respectively. CB17 SCID mice were used as a strain that lacks B and T lymphocytes. No significant differences in arthritis incidence and severity compared with untreated littermates were observed in these experiments after administration of HMGB-1. In contrast, simultaneous depletion of monocytes and neutrophils resulted in a lower frequency of arthritis, since only 37% of mice displayed mild synovitis (Figure 4A). Also, arthritis severity was significantly decreased in the monocyte- and neutrophil-depleted group compared with untreated mice (Figure 4B).

thumbnail image

Figure 4. Incidence of arthritis (A) and its severity (B) after administration of high mobility group box chromosomal protein 1 (HMGB-1), in mice depleted of neutrophils, monocytes, or lacking T/B lymphocytes (SCID mice). Also, mice depleted of both neutrophils and monocytes were assayed with respect to their propensity to develop chromosomal protein 1 HMGB-1–triggered arthritis. The histopathologic evaluation was done 6 or 7 days after administration of 5 μg of HMGB-1. Values are the mean and SEM. NS = not significant; ∗ = Fisher's exact test; ∗∗ = Mann-Whitney U test.

Download figure to PowerPoint

IL-1 as a key mediator in HMGB-1–triggered arthritis.

To determine whether rHMGB-1–triggered arthritis is mediated by IL-1, we administered 5 μg of HMGB-1 to IL-1R−/− mice and to backcrossed controls. Histologic examination of joint sections revealed a total absence of arthritis in IL-1R−/− mice, whereas 40% of the backcrossed controls displayed synovitis.

Activation of NF-κB by HMGB-1 in mouse spleen cells.

Since synthesis of IL-1 is mediated by the activation of NF-κB, we investigated whether the exposure of mouse spleen cells to rHMGB-1 leads to the activation of NF-κB. Spleen cells were stimulated for 2 hours with different concentrations (0.01 μg/ml, 1 μg/ml, and 10 μg/ml) of rHMGB-1, and EMSA was used to determine the activation of NF-κB. Incubation of spleen cells with rHMGB-1 resulted in dose-dependent activation of NF-κB (Figure 5).

thumbnail image

Figure 5. A, Image of an electrophoretic mobility shift assay (EMSA) gel showing nuclear factor κB (NF-κB) activation in unstimulated mouse spleen cells (lane 1), and cells incubated for 2 hours with 0.01 μg/ml (lane 2), 1 μg/ml (lane 3), or 10 μg/ml (lane 4) of high mobility group box chromosomal protein 1 (HMGB-1). B, Image of an EMSA gel showing supershift and cold competition experiments. Nuclear extracts from unstimulated mouse spleen cells (lane 2) and from cells stimulated for 2 hours with 10 μg/ml of HMGB-1 (lanes 3–5) were incubated in the absence (lane 3) or in the presence (lane 4) of an excess of homologous oligonucleotide or subjected to supershift with p50 antibody (lane 5). Arrowhead indicates the supershift band; arrows indicate the NF-κB band. A free probe with no nuclear extract is also shown (lane 1).

Download figure to PowerPoint

DISCUSSION

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

Our results indicate that the HMGB-1 molecule by itself triggers arthritis in healthy recipients. Also, it is found in joints of patients with RA as well as in animals with adjuvant arthritis (11, 21), and therefore it might play an important pathogenic role in inflammatory joint disease. This proinflammatory effect was not caused by LPS contamination because mice receiving LPS at a dose equivalent to that which could be found in the highest dose of HMGB-1 did not display any signs of arthritis.

Previous studies have shown that inflammatory cytokines, such as TNFα and IL-1, induce the release of HMGB-1 from monocytes and macrophages (12). Furthermore, in vitro studies have revealed that extracellular HMGB-1 itself can activate an additional downstream cascade by stimulating monocyte/macrophages to secrete a subset of proinflammatory cytokines and chemokines. This proinflammatory response to HMGB-1 is highly restricted to the monocyte/macrophage subset since no cytokine production by lymphocytes was seen (12). Indeed, it seems that in the case of HMGB-1–triggered arthritis, the innate immune system plays a major role. The finding strongly supporting this conclusion is that inflammation in the joints occurs at the same frequency and severity in SCID mice lacking T and B cells as in congenic controls. Finally, the scarcity of joint-infiltrating T cells supports this suggestion. In contrast, an abundance of mononuclear Mac-1+ macrophages was found in the inflamed synovial tissue. Moreover, simultaneous depletion of both these cell types diminished the severity as well as the incidence of synovitis, indicating the concomitant action of these 2 major components of innate immunity.

IL-1 is one of the principal mediators of the host response to inflammatory stimuli. The major source of IL-1 is activated mononuclear phagocytes (22). This cytokine, together with TNFα, is considered to play a pivotal role in the development and progression of joint inflammation in rheumatic diseases, such as RA. TNFα is a master cytokine that causes joint swelling, whereas IL-1, the production of which may occur independent of TNFα, has been shown to be more responsible for cartilage and bone destruction (23). In the present study, IL-1R−/− mice did not develop arthritis in response to administration of HMGB-1, suggesting a major role for this macrophage-derived cytokine in HMGB-1–triggered joint inflammation. The reason arthritis developed in 40% of mice of the congenic, IL-1R+/+ strain (C57BL/6) rather than in ∼80% as seen in NMRI and CB17 mice is unknown, but probably reflects differences in background genome and the propensity to develop inflammation following exposure to HMGB-1.

The signaling mechanisms by which HMGB-1 activates cells are not fully understood. HMGB-1 has at least one high-affinity receptor, the receptor for advanced glycation end products (RAGE) (24), which is a multiligand receptor of the immunoglobulin superfamily and is expressed on many cell types, including endothelial cells, mononuclear phagocytes, and smooth muscle cells (25). In this respect, all these cell types are readily found in healthy synovial tissue. In addition, the fact that HMGB-1 increases expression of adhesion molecules and production of chemokines (26) may cause it to serve as an efficient stimulator of the influx of inflammatory cells to the synovial tissue and might contribute to the development of inflammation in the joints.

NF-κB is a critical transcription factor involved in the production of many cytokines and adhesion molecules. In this study, we show that stimulation of spleen cells with HMGB-1 leads to the activation of NF-κB. One needs to keep in mind that the ligation of RAGE by HMGB-1 leads to activation of NF-κB, at least in neural cells (27). Therefore, it remains possible that activation of NF-κB in lymphoid cells found in our experiments might also be mediated via the RAGE receptor pathway. Nevertheless, our results suggest that the proinflammatory effect of HMGB-1 is at least partly mediated by activation of NF-κB.

Previous reports have indicated that HMGB-1 is a late mediator of endotoxin lethality: its release from cultured macrophages was delayed compared with the release of early proinflammatory cytokines (9). The release of HMGB-1 is not only a late response to proinflammatory stimuli, but it also itself provokes a delayed response (12), thereby prolonging and maintaining inflammation. Indeed, our findings indicate that 50% of mice injected with HMGB-1 displayed synovitis up to 28 days following a single injection. The fact that HMGB-1 is not stably associated with the chromosomes (28) and can be released from damaged as well as from necrotic cells (21, 29, 30) might also contribute to the prolongation of the arthritis process because the cells undergoing necrosis at the site of inflammation could serve as an additional source of HMGB-1.

In conclusion, we propose that HMGB-1 is of importance in the pathogenesis of arthritis because it can be found in the inflamed joint and it can induce joint inflammation on its own by activating monocyte/macrophages and inducing the release of the proinflammatory cytokine IL-1 via NF-κB activation. This finding implicates HMGB-1 as a potential target for future therapies.

Acknowledgements

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

We thank Berit Ericsson for excellent technical assistance and Jan Bjersing for help with the EMSA.

REFERENCES

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