High-Mobility Group Box 1 Inhibition Alleviates Lupus-Like Disease in BXSB Mice



High-mobility group box 1 protein (HMGB1), a ubiquitous nuclear DNA-binding protein, functions as a potent proinflammatory factor. In this study, we evaluated the effects of HMGB1 inhibition on murine lupus using the lupus-prone model. We treated male BXSB mice with neutralizing anti-HMGB1 monoclonal antibody (HMGB1 mAb) from age 16 weeks to 26 weeks. The control group received the same amount of control IgG. Lupus-prone male BXSB mice treated with HMGB1mAb showed attenuated proteinuria, glomerulonephritis, circulating anti-dsDNA and immune complex deposition. Levels of serum IL-1β, IL-6, IL-17 and IL-18 were also significantly decreased by administration of HMGB1mAb in lupus-prone BXSB mice. HMGB1mAb treatment also decreased the caspase-1 activity in the kidneys of BXSB mice and reduced the mouse mortality. Our study supports that HMGB1 inhibition alleviates lupus-like disease in BXSB mice and might be a potential treatment option for human SLE.


High-mobility group box 1 (HMGB1), a highly conserved protein previously known as a DNA-binding protein involved in maintenance of nucleosome structure and regulation of gene transcription, was first identified as a chromosomal protein and later on also as a membrane-bound form on neuronal cells and was termed ‘amphoterin’ [1]. Recently, HMGB1 has been found to act as a potent proinflammatory cytokine [2]. Extracellular HMGB1 has been identified as a crucial cytokine involved in many inflammatory conditions such as intestinal inflammatory disorder [3], stroke [4] and rheumatoid arthritis [5, 6].

Systemic lupus erythematosus (SLE) is a systemic autoimmune disease characterized by the involvement of multiple organ systems. Recently, studies suggested that HMGB1 is associated with nucleosomes released from apoptotic cells and that HMGB1 contributes to the immunostimulatory effect of nucleosomes [7]. In addition, HMGB1 has been found to be significantly elevated in sera from patients with lupus, and HMGB1 has been regarded as one of the components in DNA-containing immune complexes [8, 9]. However, the role of HMGB1 in SLE is not clearly understood. In this study, we explored the potential of HMGB1 inhibition as a treatment option for lupus-like disease using the lupus-prone model BXSB mice and the underlying mechanisms. This study reveals that HMGB1 represents a potential therapeutic target in SLE and that HMGB1 inhibition delays lupus progression.

Materials and methods

BXSB mice

The lupus-prone male BXSB mice were purchased from Jackson Laboratory (Bar Harbor, ME, USA). The animal experimentation was conducted according to the Principles of Laboratory Animal Care and approved by Ethics Committee of Shandong University.

Treatment with HMGB1mAb

Neutralizing anti-HMGB1 monoclonal antibodies (HMGB1mAb) were generated by Coowinbiotech Company, China, as described previously [10, 11]. Groups of 16-week-old male BXSB mice (n = 12/group) received HMGB1mAb as previously described [11]. Anti-HMGB1 antibody (10 μg/mouse) was intraperitoneally administered three times a week from week 16 until week 26. As the control, BXSB mice were given the same amount of control IgG by intraperitoneal injection. All mice were monitored regularly for different parameters over 10 weeks. For the survival analysis, two additional groups of 16-week-old male BXSB mice (n = 20/group) were employed and observed.

Urinary albumin/creatinine concentration ratios

Mice were placed in metabolic cages for 24-h urine collection. Urinary albumin was determined using the mouse albumin ELISA Kit (Newberg, OR, USA) according to the instructions. Urinary creatinine concentrations were detected with a Beckman autoanalyzer (Beckman Coulter, Shanghai, China). Urinary albumin was normalized to creatinine excretion.


The spleens of the mice were harvested and weighed after 10 weeks of treatment. Splenomegaly was evaluated by a spleen index [SI = spleen weight/body weight) * 1000].

Histopathology and immunofluorescence

Kidney samples were collected at the time of harvest and fixed in 10% buffered formalin, embedded in paraffin. Haematoxylin and eosin (H&E) or periodic acid–Schiff–methenamine (PASM) staining was performed. For immunofluorescence staining, the kidney tissues were snap-frozen in liquid nitrogen and placed in optimum cutting temperature (OCT) compound (Sakura Japan, Osaka, Japan). Frozen sections (5 mm thick) were stained with fluorescein-conjugated anti-mouse IgG (1:50) or anti-mouse C3 (1:30, Cedarlane, Burlington, Canada) after being blocked with 10% foetal bovine serum. Bright-field and fluorescent photomicrographs were captured on an Olympus microscope and NP70 camera using a 20× objective lens. Image analysis was performed with ImageJ (ImageJ, NIH). We evaluated glomerular pathology by examining changes in each glomerulus identified in 20 glomerular cross sections (GCS) per kidney. Changes were scored on a semiquantitative scale of 0–3 as previously reported [12] by two independent experienced renal pathologists in a blinded manner, where 0 = normal (35–40 cells/GCS), 1 = mild (few lesions, with slight proliferative changes and mild hypercellularity [41–50 cells/GCS] and/or minor exudation), 2 = moderate (moderate hypercellularity [50–60 cells/GCS], including segmental and/or diffuse proliferative changes, hyalinosis and/or moderate exudates) and 3 = severe (segmental or global sclerosis and/or severe hypercellularity [>60 cells/GCS], necrosis, crescent formation and/or heavy exudation). Slides were scored 0–3 in a blinded fashion by two different researchers for the intensity and coverage of immunofluorescence in the glomerulus (0, no fluorescence; or just detectable above background; 1, fluorescence scattered and light; 2, bright but not diffuse; and 3, bright and diffuse).

Serum anti-dsDNA antibodies and circulating immune complex

Serum anti-dsDNA antibody levels and circulating IgG were determined by ELISA as previously described [13]. Means of the triplicate OD490 values were recorded for the serum. Levels of anti-dsDNA and immune complex were expressed as U/ml, using a positive reference standard of pooled serum from diseased BXSB mice.

Cytokines detection

Levels of serum IL-1β, IL-6, IL-17, IL-18 and IFNα were detected using a mouse inflammation cytometric bead array kit (CBA; BD Biosciences, California, USA) and were analysed on a FACSCalibur flow cytometer. Standard curves were determined for each cytokine according to the manufacturer.

Caspase-1 activity assay

The enzymatic activity of caspase-1 was assayed using a Caspase-1 Activity Kit according to the manufacturer's protocol (Beyotime Institute of Biotechnology, Beijing, China), which is based on the ability of caspase-1 to change acetyl-Tyr-Val-Ala-Asp p-nitroanilide (Ac-YVAD-pNA) into the yellow formazan product p-nitroaniline (pNA). Renal lysates were centrifuged at 12,000 g for 10 min, and the protein concentrations were determined by the Bradford protein assay. Cellular extracts (30 μg of protein) were incubated in a 96-well microtitre plate with 20 ng of Ac-DEVD-pNA overnight at 37 °C. The absorbance values of pNA at 405 nm, OD405, were measured using a 96-well plate reader (BioTek, Santa Barbara, CA, USA). An increase in the OD405 indicated activation of caspase-1.

Statistical analysis

Statistics were performed using SPSS 17.0 software. Data are presented as means ± standard deviation (SD). Group comparisons were analysed by unpaired two-tailed Student's t-test or Kruskal–Wallis test. Survival was analysed by Kaplan–Meier plot and log-rank test. The P < 0.01 was considered significant.


HMGB1mAb reduced proteinuria and splenomegaly

To determine the effect of HMGB1 inhibition on urine protein excretion, we measured 24-h urinary protein excretion every 2 weeks. Both groups of mice developed proteinuria from 16 weeks of age (Fig. 1A). In contrast, mice treated with HMGB1mAb developed statistically attenuated proteinuria compared with controls starting from 8 weeks after treatment (P < 0.01). In addition, splenomegaly was observed in adult BXSB mice, and treatment with HMGB1mAb significantly decreased the splenomegaly, as indicated by the SI (Fig. 1B).

Figure 1.

Effects of HMGB1mAb treatment on the proteinuria and splenomegaly in lupus-prone BXSB mice. (A) Urine protein/creatinine was assessed every 2 weeks in mice treated with HMGB1mAb or control. (B) Spleen index was determined at the end of the study. Results shown are the mean ± SD of 12 mice per group. #< 0.01 compared with control-treated group.

HMGB1mAb reduced the serum levels of anti-dsDNA autoantibody and circulating IgG

There was a remarkable rise in serum anti-dsDNA antibody levels in the control group at week 26, which was obviously inhibited by HMGB1mAb treatment (Fig. 2A). Similarly, HMGB1mAb-treated group showed significantly lower circulating IgG compared with the control group at week 26 (Fig. 2B).

Figure 2.

HMGB1mAb reduced the serum levels of anti-dsDNA autoantibody (A) and circulating immune complex (B) in BXSB mice. At week 26, sera were collected from mice and subjected to ELISA. Values are the means ± SD (= 12 mice per group). #< 0.01 compared with control-treated group.

HMGB1mAb alleviated renal pathology and renal immune complex deposition

H&E-stained kidney sections were assessed by histological scoring for overall glomerular inflammation. BXSB mice in the control group exhibited lupus-like diffuse glomerulonephritis. HMGB1mAb-treated mice, however, exhibited a significant improvement in glomerular histology in the kidneys (Fig. 3A, B).

Figure 3.

HMGB1mAb reduced renal pathology and immune complex deposition. (A) Renal pathology (H&E), (B) quantitative analysis of histology, (C) renal deposition of IgG and C3, (D) quantitative analysis of fluorescence intensity. Data were mean ± SD from 12 mice age 26 weeks per group, #< 0.01 compared with control-treated group.

We also determined the renal IgG and C3 deposition. Glomerular IgG and C3 deposition was significantly reduced in HMGB1mAb-treated mice compared with controls (Fig. 3C, D).

HMGB1mAb reduced proinflammatory cytokines

To ascertain whether HMGB1 blockade inhibits the proinflammatory cytokines, HMGB1mAb-treated and control-treated BXSB mice were bled at the end of the study. As measured by ELISA. The HMGB1mAb-treated mice showed lower serum levels of IL-1β, IL-6, IL-17 and IL-18 than the control-treated mice (Fig. 4). In addition, the IFNα levels were also reduced.

Figure 4.

HMGB1mAb reduced the production of proinflammatory cytokines. We found serum IL-1β, IL-6, IL-17 and IL-18 were downregulated by HMGB1mAb treatment. Data are the means ± SD (= 12 mice per group); #< 0.01 compared with control-treated group.

HMGB1mAb reduced the caspase-1 activity and mouse mortality

Considering that HMGB1 is involved in activation of caspase-1 [14], importantly, an essential role for caspase-1 in the induction of lupus has recently been reported [15]. We also evaluated the caspase-1 activity in the kidneys of lupus-prone BXSB mice. We found that HMGB1mAb significantly reduced the caspase-1 activity as compared with the control group (Fig. 5A).

Figure 5.

HMGB1mAb treatment reduced the caspase-1 activity and improved mouse survival. (A) Activity of caspase-1 in the kidneys of the mice age 26 weeks (= 12 mice per group). (B) Kaplan–Meier cumulative survival plot shows the survival was significantly prolonged in mice treated with HMGB1mAb compared with control (= 20 mice per group). In the histograms, the data are the mean ± SD. #< 0.01 compared with control-treated group.

In addition, we assessed the benefit of HMGB1mAb treatment on the mouse survival in BXSB mice (n = 20 per group). The proportion of surviving animals treated with HMGB1mAb was significantly higher than that in control-treated mice (Fig. 5B, P < 0.001). This suggests that HMGB1mAb treatment significantly prolonged the survival of BXSB mice.


Lupus nephritis (LN) is a common complication of SLE, associated with significant morbidity and mortality. Although therapeutic effects have significantly improved in the last decades, the high rates of progression to end-stage renal disease coupled with the adverse effects of drugs have led to an intensive search for more effective and less toxic therapies for LN [16]. In this study, we showed that administration of HMGB1mAb significantly ameliorated the severity of murine lupus in BXSB mice, as reflected by reduced proteinuria and glomerulonephritis. Further studies suggest that this therapeutic effect was closely associated with reducing caspase-1 activity and multiple proinflammatory cytokines. These data support the HMGB1 played a role in the pathogenesis of murine LN.

Production of numerous autoantibodies against a variety of self-components including antibodies to nucleosomes and DNA is a classical feature of SLE. It has been proposed that apoptotic cells are the source of autoantigens. Due to deficient clearance mechanisms in SLE, apoptotic cells accumulate to become late apoptotic cells or even secondary necrotic cells [17]. Studies have suggested that HMGB1 is associated with nucleosomes released from apoptotic cells and that HMGB1 contributes to the immunostimulatory effect of nucleosomes [7]. In addition, nucleosomes as well as HMGB1 have been found to be significantly elevated in sera from patients with SLE, and HMGB1 has been regarded as one of the components in DNA-containing immune complexes [8, 18]. However, the mode of action of HMGB1 in the pathophysiology of SLE is still unclear. Increased concentrations of DNA-containing immune complexes in the serum are associated with human SLE, and data demonstrate a mechanism by which HMGB1 and RAGE activate plasmacytoid dendritic cells and B cells in response to DNA and contribute to autoimmune pathogenesis [19]. Moreover, HMGB1 is able to activate B cells [20]. Elevated levels of HMGB1 and antibodies targeted against HMGB1 have been detected in human lupus sera, so HMGB1 has therefore been identified as a new autoantigen [21]. In the present study, we found that HMGB1mAb treatment attenuated the circulating anti-dsDNA and immune complex deposition. These results suggest that HMGB1 plays a functional role in the presence of in immune complexes.

Another important finding in this study is that the therapeutic effects of HMGB1 blockade might also be related to a systemic blunting of autoimmunity and proinflammatory responses, as reflected by reduced serum levels of proinflammatory cytokines in serum. Notably, caspase-1 is regarded as a key mediator of IL-17-related inflammatory processes [22]. Caspase-1 is an essential component of the nucleotide binding and oligomerization domain-like receptor family pyrin domain containing 3 (NLRP3) inflammasome, which is responsible for the production of IL-1 and IL-18. Recent studies have demonstrated the crucial role for NLRP3 inflammasome in human lupus [23, 24]. Activation of the NLRP3 inflammasome is enhanced in lupus-prone mice and human lupus [25, 26]. Importantly, Kahlenberg et al. have shown that caspase-1 is an essential component in the development of lupus and may play an important role in the crosstalk between environmental exposures and autoimmunity development, thus identifying a novel pathway for therapeutic targeting [14]. Our results show that the levels of serum IL-1β, IL-6, IL-17 and IL-18 were significantly decreased by HMGB1mAb treatment in lupus-prone BXSB mice. HMGB1mAb treatment also decreased the caspase-1 activity in the kidneys of BXSB mice and the mouse mortality.

In summary, our study demonstrates that HMGB1 blockade confers benefit on murine lupus-like disease. Our analysis demonstrates the therapeutic efficacy of HMGB1 blockade in the lupus-prone BXSB mice. The effects involve inhibition of autoimmunity, caspase activity and inflammatory responses in murine lupus-like disease. Taken with the current data, HMGB1 blockade might be a potential target for treating human SLE.


The study was supported by the National Nature Science Foundation of China (No. 81172861), Science and Technology Development Project of Shandong Province (No. 2011GGH21829).

Conflict of interests

There are no competing interests.

Author contributions

Drs. Yuanchao Zhang, Qingrui Yang, Chunling Zhang and Chun Li contributed in designing of study and data analysis; Drs. Shuli Jia and Pingbo Yao participated in doing experiments and preparing the article.