Background
Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by inflammation in different organ systems and immunologic abnormalities, such as presence of autoreactive T cells and hyperactive B cells producing autoantibodies [
1]. Approximately 25–50 % of SLE patients develop lupus nephritis (LN), which is initiated by the deposition of immune complexes within the renal parenchyma leading to complement activation, mesangial expansion, and induction of inflammatory and fibrotic processes, resulting in glomerulonephritis and progressive renal dysfunction [
2]. LN patients are often treated with strong immunosuppressive regimens, including cytotoxic drugs, antimalarial compounds, immunophilin targeting agents, and glucocorticoids [
3]. In addition, the effect of these therapeutic medications on LN has been determined by the histologic diagnosis of glomerulonephritis [
3]. The treatments can transiently reduce disease activity, but often do not induce remission or prevent end organ damage.
The polymorphisms in the nuclear factor erythroid 2-related factor 2 (Nrf2) genes appear to confer a risk for developing LN [
4]. Moreover, Nrf2 deficiency leads to the development of lupus-like autoimmune nephritis in aged female mice [
5], indicating that Nrf2 plays a crucial role in the development of LN in both human and mice. Nrf2 is a redox-sensitive transcription factor that regulates the expression of antioxidant response element (ARE)-dependent genes, such as those encoding antioxidant and detoxifying enzymes like heme oxygenase-1 (HO-1) and NAD(P)H:quinone oxidoreductase 1 (NQO1) [
6]. Under normal conditions, Nrf2 is retained in the cytoplasm in a “silent” form by its repressor protein, Kelch-like ECH-associated protein 1 (Keap1), which contains a subset of reactive cysteine residues. Oxidative stress or environmental stimuli modifies the cysteine residues of Keap1, thus enabling translocation of Nrf2 to the nucleus where it binds to the ARE [
6]. The activation of Nrf2 leads not only to organ-protective effects but also to anti-inflammatory effects by inhibition of nuclear factor (NF)-kB activation. [
7]. In a previous study, antioxidant response of Nrf2 and NQO1, which functions in the downstream of Nrf2, was demonstrated in the kidney of patients with LN; however, it was not determined whether the response modified the disease activity or not [
8].
Small-molecule compounds such as sulforaphane and dimethyl fumarate (DMF) have been reported as the clinically used Nrf2 activators [
9,
10]. Sulforaphane, an organosulfur compound found in broccoli sprouts and other cruciferous vegetables, has potent anti-inflammatory or anticancer activities, and has been clinically used as a dietary supplement [
11]. It inhibited proinflammatory cytokine production in vitro in lipopolysaccharide (LPS)-stimulated monocytes [
12] and tumor necrosis factor alpha (TNF-α)-stimulated synoviocytes [
13]. Furthermore, sulforaphane improved the development of experimental autoimmune encephalomyelitis (EAE) [
14] and alleviated type II collagen-induced arthritis [
15] and dextran sodium sulfate-induced acute colitis [
16] in mice. In contrast, DMF is the main active ingredient of fumaric acid esters (FAEs) and has been approved in Germany as Fumaderm® for the systemic treatment of psoriasis [
17]. It demonstrated anti-inflammatory activity both in vitro and in vivo. DMF inhibited proinflammatory cytokine production in TNF-α-stimulated human umbilical vein endothelial cells [
18] and LPS/interferon gamma (IFNγ)-stimulated peripheral blood mononuclear cells (PBMCs) [
19] by inhibiting either the activation or translocation of NF-kB in vitro. Moreover, DMF ameliorated the development of mouse EAE [
10]. The efficacy of oral DMF in the treatment of relapsing–remitting multiple sclerosis (RRMS) has been proven in two phase III studies: DEFINE (Determination of the Efficacy and Safety of Oral Fumarate in relapsing–remitting multiple sclerosis) [
20] and CONFIRM (Comparator and an Oral Fumarate in relapsing–remitting multiple sclerosis) [
21]. In 2013, DMF was approved by the US Food and Drug Administration (FDA) as first-line therapy for RRMS patients [
22]. In-vivo studies in psoriasis patients showed that DMF reduced the development of psoriasis [
23], and the clinical efficacy of DMF is being explored in a phase III trial (ClinicalTrials.gov NCT01815723). As already described, the effect of the clinically used Nrf2 activators such as sulforaphane and DMF on the development of various chronic inflammatory diseases has been evaluated. However, there are no reports about the effects of these compounds on the development of human LN.
Here, we investigated the effects of clinically used Nrf2 activators on renal functions both in vitro and in vivo. To assess the in-vitro functions, human renal mesangial cells (HRMCs) were used. Mesangial cells typically have physiological functions including phagocytosis, synthesis of the mesangial extracellular matrix, and control of glomerular hemodynamics via mesangial cell contraction [
24]. In response to immunological, toxic, ischemic, or mechanical injury, mesangial cells produce chemokines or cytokines that selectively induce inflammatory cell migration and activation at the site of injury [
25]. Pristane-injected mice were treated with DMF, an Nrf2 activator with ensured clinical safety [
22], to evaluate its in-vivo effects on the development of LN. Pristane is a natural hydrocarbon oil, also called TMPD (2,6,10,14-tetramethylpentadecane). Pristane-induced LN is associated with hypergammaglobulinemia, lupus-related autoantibody production, and immune complex-mediated glomerulonephritis, and resembles human LN [
26]. In this model, Nrf2-deficient mice developed a greater degree of renal damage [
8]. We examined possible therapy with Nrf2 activators both in vitro and in vivo, and glucocorticoid, the standard therapy for human LN, was used as the reference compound.
Methods
Test compounds
DMF (BG-12), sulforaphane, and prednisolone were purchased from Sigma-Aldrich (St. Louis, MO, USA). For the cellular functional assay, these compounds were dissolved in dimethylsulfoxide (DMSO; Nacalai Tesque, Kyoto, Japan) and stored in aliquots at –20 °C. For the animal model, DMF and prednisolone were suspended in 0.5 % methylcellulose as vehicle.
Antibodies and flow cytometric analysis
Monoclonal antibodies (mAbs) used were PE-conjugated anti-human CD120b (hTNFR-M1), biotin-conjugated anti-human CD120a (MABTNFR1-B1), and isotype-matching control mAb. Allophycocyanin-conjugated streptavidin was used for staining biotin antibodies. All of these were purchased from BD Biosciences (San Jose, CA, USA). For Fc receptor blocking, we used FcR Blocking Reagent (Miltenyi Biotech, Auburn, CA, USA). Cell surface staining was performed according to standard techniques, and flow cytometry was performed with a FACSCalibur flow cytometer and analyzed using CellQuest™ Pro software (BD Biosciences).
Cell culture
HRMCs were purchased from ScienCell Research Laboratories (Carlsbad, CA, USA), and maintained in the Mesangial cell medium (ScienCell Research Laboratories) in a humidified atmosphere containing 5 % CO2 at 37 °C. HRMCs were used between passages 3 and 6.
Cellular treatments
HRMCs were plated in 96-well poly-l-lysine-coated plates (BD Biosciences) at a density of 1.5 × 104 cells per well. Nrf2 activators, DMF and sulforaphane, or prednisolone were added to the culture media, and DMSO at a final concentration of 0.1 % in culture medium was used as the vehicle control. To evaluate the cytotoxicity of the compounds, the CellTiter-Glo luminescent cell viability assay was performed according to the manufacturer’s instructions (Promega, Madison, WI, USA). In experiments examining the anti-inflammatory effects of Nrf2 activators or prednisolone on TNF-α-induced monocyte chemoattractant protein-1 (MCP-1) or interleukin-6 (IL-6) production, these compounds were added 1 hour before the addition of TNF-α (R&D Systems, Minneapolis, MN, USA). Cells were cultured for 18 hours, and then the supernatants were collected and stored at –80 °C until they were assayed.
Measurement of NF-kB or Nrf2 transcriptional activity
The activity of nuclear NF-kB or Nrf2 was determined using TransAM transcriptional factor assaying kits for NF-kB p65 or Nrf2 according to the manufacturer’s instructions (Active Motif, Carlsbad, CA, USA). The nuclear extract kit was from Active Motif. NF-kB p65 protein (Active Motif) or Nrf2 protein (Abnova, Taipei, Taiwan) was used as the standard in each kit.
Protein quantification
Cells were rinsed with phosphate-buffered saline (PBS) and solubilized in lysis buffer, which was PBS containing 1 mM ethylenediaminetetraacetic acid, 0.5 % Triton X-100, 10 μg/ml leupeptin, 10 μg/ml pepstatin, 100 μM phenylmethanesulfonyl fluoride, and 3 μg/ml aprotinin (all from Sigma-Aldrich). Intracellular HO-1 (R&D Systems) and NQO1 (Cusabio, Wuhan, China) expressions were detected using enzyme-linked immunosorbent assay (ELISA) kits according to the manufacturers’ instructions. The total protein level in the cell lysate was detected using the Pierce™ BCA Protein Assay kit according to the manufacturer’s instructions (Thermo Fisher Scientific, Waltham, MA, USA). The amount of MCP-1 and IL-6 released into the supernatants of the cultured cells was determined using ELISA kits (R&D Systems).
RNA isolation and real-time PCR
RNAs were isolated from HRMCs using the GenElute mammalian total RNA kit (Sigma-Aldrich). Quantitative reverse transcription PCR (RT-PCR) was performed using the RNA-to-CT 1-step kit (Thermo Fisher Scientific) and the ABI PRISM 7900HT system (Thermo Fisher Scientific). The primers (HO-1: Hs01110250_m1, NQO1: Hs00168547_m1, glucocorticoid receptor alpha (GRα): Hs00230818_m1, GRβ: Hs00354508_m1, glyceraldehyde 3-phosphate dehydrogenase (GAPDH): Hs02758991_g1) were from Thermo Fisher Scientific. The comparative cycle time (CT) was used to normalize transcripts to GAPDH.
Mice
Female BALB/c mice were obtained from SLC (Shizuoka, Japan) and maintained under specific pathogen-free conditions in the Animal Research Institute of the Tohoku University Graduate School of Medicine, Sendai, Japan. All experiments performed in this study conformed to the ethical guidelines of Tohoku University regarding animal experimentation.
Determination of the effect of medications on pristane-induced nephritis
Female BALB/c mice were injected intraperitoneally with either 0.5 ml pristane (Sigma-Aldrich) or PBS at 7 weeks of age. The medications were administered daily by oral gavage from the starting day until the experimental end point (for 20 weeks), as described previously [
27]. Each treatment group consisted of five to seven mice: vehicle alone, 75 mg/kg body weight (bw) DMF, or 2 mg/kg bw prednisolone. The DMF dose was decided based on the recommended human dose of 480 mg/day for multiple sclerosis (MS) [
22]. Conversely, the dose of prednisolone was equal to that in the previous reports [
28,
29]. At 20 weeks after injection, urine was collected from these mice for 24 hours using the metabolic cages. After killing the mice under anesthesia, the kidneys were removed and frozen for protein and RNA extraction, or fixed for histology analysis. The concentration of urine albumin was measured by ELISA (Shibayagi, Gunma, Japan) and urine creatinine was measured using Liquitech creatinine PAP II (Roche Diagnostics, Basel, Switzerland) and an automatic analyzer (Hitachi 7180; Hitachi High-Technologies, Tokyo, Japan). The urine albumin-to-creatinine ratio (uACR) was expressed as milligrams of albumin per gram of creatinine.
Serum samples were stored at –80 °C until they were assayed. ELISA kits were used to measure the autoantibody titers, which are anti-double strand (ds) DNA (Shibayagi) and anti-nuclear ribonucleoprotein (nRNP) antibody (Alpha Diagnostic, San Antonio, TX, USA).
The kidneys were homogenized in the lysis buffer as already described and both MCP-1 (R&D Systems) and HO-1 (Takara Bio, Shiga, Japan) levels in the kidney were measured by ELISA kit. The isolation of RNAs from the kidney and the quantitative RT-PCR were performed as already described. Oligonucleotide primers were purchased from Thermo Fisher Scientific: Mm00516005_m1 for HO-1, Mm01253561_m1 for NQO1, Mm00441242_m1 for MCP-1, Mm00446190_m1 for IL-6, Mm00443258_m1 for TNF-α, Mm01256744_m1 for fibronectin (Fn1), Mm01178820_m1 for tumor growth factor beta 1 (TGF-β1), and Mm99999915_g1 for GAPDH.
Pathological evaluation of LN
The kidneys were fixed in 10 % formalin in 0.01 M phosphate buffer (pH 7.2) and embedded in paraffin wax. Tissue sections were stained with periodic acid–Schiffs (PAS) and used for evaluations. The kidneys were scored by histological criteria as follows: 0 = positive staining in <25 % of glomerular cross-section, 1 = positive staining in 25–50 % of glomerular cross-section, 2 = positive staining in 50–75 % of glomerular cross-section, and 3 = positive staining in >75 % of glomerular cross-section. At least 40 glomeruli per animal were assessed to determine the average score. The scoring systems of histology were determined by a pathologist (MO) and a cytologist (HT) blinded to the treatment.
Statistical analyses
The significance of average differences among multiple groups was evaluated by the Dunnett’s parametric or nonparametric test. p < 0.05 was considered significant.
Discussion
Several reports have shown that glucocorticoid can effectively inhibit proinflammatory cytokine or chemokine production in human immune cells, including PBMCs or monocytes [
33,
34]. However, glucocorticoid-insensitive NF-kB activation and proinflammatory cytokine production have been demonstrated in some nonimmune human cells, for example brain cells [
35] and airway epithelial cells [
36]. Consistent with the described reports, glucocorticoid (prednisolone) showed only a small effect on IL-6 (Fig.
3a) and MCP-1 (Fig.
3b) production in TNF-α-stimulated HRMCs in the present study. The association of glucocorticoid residence and expression of GRβ, which is a dominant negative inhibitor of GRα, have been clinically studied in immune-mediated diseases such as rheumatoid arthritis [
37] and ulcerative colitis [
38]. The present study revealed that the exposure of mesangial cells to glucocorticoid under inflammatory stimulation increased GRβ expression (Fig.
2e). Taken together, the reaction of human mesangial cells in inflammatory response may be associated with glucocorticoid-insensitive renal diseases.
Among the many Nrf2 activators, we selected DMF that has ensured clinical safety and evaluated its effect in an in-vivo model. Another Nrf2 activator, bardoxolone methyl, was developed for the treatment of chronic kidney diseases and type II diabetes mellitus. However, the phase III BEACON (Bardoxolone Methyl Evaluation in Patients with Chronic Kidney Disease and Type 2 Diabetes Mellitus: the Occurrence of Renal Events) trial was terminated owing to severe renal dysfunction and cardiovascular failure [
39]. The safety data observed with DMF suggest that bardoxolone methyl’s side effects may be compound specific rather than target related. Bardoxolone methyl as an Nrf2 activator was expected to be a new therapeutic approach for renal diseases. In fact, bardoxolone methyl has ameliorated kidney disease in models of ischemic acute kidney injury [
40], and furthermore in genetic lupus mice [
41]. The other compounds which have been shown to activate Nrf2, such as epigallocatechin-3-gallate [
42] or hemin [
43], have also suppressed renal disease in spontaneous lupus mice, such as NZB/W F1 or MRL/
lpr mice.
In the present study, we used pristane-induced LN, which resembles the pathology of human LN [
26], as the animal model. Pristane treatment of the peritoneal cavity induced apoptosis of peritoneal cells [
44] leading to the production of autoantibodies, resulting in the development of LN. Extranodal lymphoid tissue (also called tertiary lymphoid neogenesis) formed in the peritoneal cavity [
45] and was indicative of the inflammatory lesion in this pristane-treated model as well as in human LN with tubulointerstitial inflammation [
46]. It has been indicated that chronic inflammatory condition leading to lymphoid neogenesis is associated with autoantibody production [
47].
There are no histopathological studies on fibrosis of the kidney in pristane-induced LN. In the present study, the upregulation of TGF-β1 expression in the kidney of pristane-injected mice was detected at the experimental end point (Fig.
4j), suggesting that the fibrosis in the kidney of this model may be caused after LN onset. Pristane caused arthritis in some strains of mice with characteristics such as synovial hyperplasia, periostitis, and marginal erosions [
48]. The pristane-induced arthritis was strongly inhibited by treatment with prednisolone (2 mg/kg bw) [
49]. However, the present study revealed that treatment with the same dose of prednisolone did not inhibit the development of pristane-induced LN (Figs.
3 and
4), suggesting that the difference in the effects may be dependent on organ specificity, including renal cells such as mesangial cells or other nonimmune cells. Interestingly, different from the strongly glucocorticoid-sensitive LN in NZB/W F1 [
28] or MRL/
lpr [
29] mice, pristane-induced LN was observed to be glucocorticoid-insensitive LN (Fig.
3). Our present findings suggest upregulation of GRβ expression, which remains to be identified in mice LN models, in nonimmune cells such as mesangial cells.
Although prednisolone inhibited autoantibody production, kidney disease was not improved in the mice (Figs.
3 and
4). The clinical trial in human LN has also demonstrated that the elimination of autoantibodies by plasmapheresis was unsuccessful [
50]. Furthermore, the results of another phase III clinical trial—the LUNAR (Lupus Nephritis Assessment with Rituximab) study—revealed that the treatment of rituximab to deplete B cells in LN (class III or class IV) has not resulted in more efficient clinical improvement compared with conventional therapy, despite the reduction of serum anti-dsDNA antibody levels [
51]. Taken together, it is evident that it is very important for the therapy against LN to directly prevent the kidney disease and not inhibit the autoantibody production including the glomerular deposition of immune complexes.
In conclusion, Nrf2 activators had not only anti-inflammatory effects but also organ-protective effects on LN pathology. In other words, targeted Nrf2 drugs may be more effective than the currently used immunosuppressants. The results of two phase III studies about DMF revealed that there were no cases of renal adverse events [
20,
21] and hence DMF is expected to be a new therapeutic approach for LN. Thus, Nrf2 activators such as DMF, which do not cause serious renal adverse events, are promising alternatives to glucocorticoids for treating excessive immunological activation in damaged or inflamed kidney as well as other inflammatory diseases, such as MS and psoriasis. Our findings provide useful information for the possible application of Nrf2 activators for the treatment of glucocorticoid-resistant LN patients.
Abbreviations
ARE, antioxidant response element; BEACON, Bardoxolone Methyl Evaluation in Patients with Chronic Kidney Disease and Type 2 Diabetes Mellitus, the Occurrence of Renal Events; bw, body weight; CONFIRM, Comparator and an Oral Fumarate in Relapsing–Remitting Multiple Sclerosis; DEFINE, Determination of the Efficacy and Safety of Oral Fumarate in Relapsing–Remitting Multiple Sclerosis; DMF, dimethyl fumarate; DMSO, dimethylsulfoxide; ds, double strand; EAE, experimental autoimmune encephalomyelitis; ELISA, enzyme-linked immunosorbent assay; FAE, fumaric acid ester; FDA, Food and Drug Administration; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GR, glucocorticoid receptor; HO-1, heme oxygenase-1; HRMC, human renal mesangial cell; IFNγ, interferon gamma; IL-6, interleukin-6; Keap1, Kelch-like ECH-associated protein 1; LN, lupus nephritis; LPS, lipopolysaccharide; LUNAR, Lupus Nephritis Assessment with Rituximab; mAb, monoclonal antibody; MCP-1, monocyte chemoattractant protein-1; MMF, monomethyl fumarate; MS, multiple sclerosis; NF, nuclear factor; NQO1, NAD(P)H:quinone oxidoreductase 1; Nrf2, nuclear factor erythroid 2-related factor 2; nRNP, nuclear ribonucleoprotein; PAS, periodic acid–Schiffs; PBMC, peripheral blood mononuclear cell; PBS, phosphate-buffered saline; RRMS, relapsing–remitting multiple sclerosis; RT-PCR, reverse transcription PCR; SLE, systemic lupus erythematosus; TGF-β1, tumor growth factor beta 1; TMPD, 2,6,10,14-tetramethylpentadecane; TNF-α, tumor necrosis factor alpha; uACR, urine albumin-to-creatinine ratio